diff --git a/.basedpyright/baseline.json b/.basedpyright/baseline.json new file mode 100644 index 000000000..14bf1024e --- /dev/null +++ b/.basedpyright/baseline.json @@ -0,0 +1,966 @@ +{ + "files": { + "./git/cmd.py": [ + { + "code": "reportGeneralTypeIssues", + "range": { + "startColumn": 13, + "endColumn": 32, + "lineCount": 1 + } + }, + { + "code": "reportOptionalOperand", + "range": { + "startColumn": 27, + "endColumn": 35, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 28, + "endColumn": 31, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 28, + "endColumn": 40, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 19, + "endColumn": 68, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 19, + "endColumn": 31, + "lineCount": 1 + } + } + ], + "./git/config.py": [ + { + "code": "reportGeneralTypeIssues", + "range": { + "startColumn": 11, + "endColumn": 26, + "lineCount": 1 + } + }, + { + "code": "reportInvalidTypeVarUse", + "range": { + "startColumn": 43, + "endColumn": 45, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 41, + "endColumn": 48, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 32, + "endColumn": 39, + "lineCount": 1 + } + }, + { + "code": "reportCallIssue", + "range": { + "startColumn": 25, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 30, + "endColumn": 39, + "lineCount": 1 + } + } + ], + "./git/db.py": [ + { + "code": "reportIncompatibleMethodOverride", + "range": { + "startColumn": 8, + "endColumn": 12, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 61, + "endColumn": 79, + "lineCount": 1 + } + }, + { + "code": "reportIncompatibleMethodOverride", + "range": { + "startColumn": 8, + "endColumn": 14, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 70, + "endColumn": 88, + "lineCount": 1 + } + } + ], + "./git/index/base.py": [ + { + "code": "reportArgumentType", + "range": { + "startColumn": 30, + "endColumn": 36, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 28, + "endColumn": 34, + "lineCount": 1 + } + }, + { + "code": "reportAssignmentType", + "range": { + "startColumn": 38, + "endColumn": 76, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 45, + "endColumn": 53, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 60, + "endColumn": 63, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 56, + "endColumn": 60, + "lineCount": 1 + } + }, + { + "code": "reportSelfClsParameterName", + "range": { + "startColumn": 30, + "endColumn": 33, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 51, + "endColumn": 57, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 55, + "endColumn": 61, + "lineCount": 1 + } + }, + { + "code": "reportIncompatibleMethodOverride", + "range": { + "startColumn": 8, + "endColumn": 12, + "lineCount": 1 + } + }, + { + "code": "reportAssignmentType", + "range": { + "startColumn": 20, + "endColumn": 46, + "lineCount": 1 + } + } + ], + "./git/index/fun.py": [ + { + "code": "reportArgumentType", + "range": { + "startColumn": 8, + "endColumn": 36, + "lineCount": 1 + } + } + ], + "./git/index/typ.py": [ + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 37, + "endColumn": 46, + "lineCount": 1 + } + } + ], + "./git/objects/base.py": [ + { + "code": "reportArgumentType", + "range": { + "startColumn": 20, + "endColumn": 27, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 29, + "endColumn": 36, + "lineCount": 1 + } + } + ], + "./git/objects/blob.py": [ + { + "code": "reportIncompatibleVariableOverride", + "range": { + "startColumn": 4, + "endColumn": 8, + "lineCount": 1 + } + } + ], + "./git/objects/commit.py": [ + { + "code": "reportIncompatibleVariableOverride", + "range": { + "startColumn": 4, + "endColumn": 8, + "lineCount": 1 + } + }, + { + "code": "reportIncompatibleMethodOverride", + "range": { + "startColumn": 8, + "endColumn": 31, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 20, + "endColumn": 33, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 20, + "endColumn": 30, + "lineCount": 1 + } + } + ], + "./git/objects/submodule/base.py": [ + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 19, + "endColumn": 24, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 19, + "endColumn": 24, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 23, + "endColumn": 25, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 23, + "endColumn": 25, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 41, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 52, + "endColumn": 84, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 20, + "endColumn": 40, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 35, + "endColumn": 42, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 15, + "endColumn": 20, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 15, + "endColumn": 20, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 19, + "endColumn": 33, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 19, + "endColumn": 33, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 15, + "endColumn": 27, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 15, + "endColumn": 27, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 15, + "endColumn": 19, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 15, + "endColumn": 19, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 18, + "endColumn": 20, + "lineCount": 1 + } + } + ], + "./git/objects/tag.py": [ + { + "code": "reportIncompatibleVariableOverride", + "range": { + "startColumn": 4, + "endColumn": 8, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 72, + "endColumn": 78, + "lineCount": 1 + } + } + ], + "./git/objects/tree.py": [ + { + "code": "reportIncompatibleVariableOverride", + "range": { + "startColumn": 4, + "endColumn": 8, + "lineCount": 1 + } + }, + { + "code": "reportIncompatibleMethodOverride", + "range": { + "startColumn": 8, + "endColumn": 31, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 19, + "endColumn": 83, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 15, + "endColumn": 54, + "lineCount": 1 + } + } + ], + "./git/objects/util.py": [ + { + "code": "reportAssignmentType", + "range": { + "startColumn": 23, + "endColumn": 34, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 22, + "endColumn": 26, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 30, + "endColumn": 34, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 15, + "endColumn": 54, + "lineCount": 1 + } + } + ], + "./git/refs/log.py": [ + { + "code": "reportArgumentType", + "range": { + "startColumn": 30, + "endColumn": 34, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 17, + "endColumn": 22, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 28, + "endColumn": 30, + "lineCount": 1 + } + } + ], + "./git/refs/reference.py": [ + { + "code": "reportInvalidTypeVarUse", + "range": { + "startColumn": 22, + "endColumn": 34, + "lineCount": 1 + } + }, + { + "code": "reportIncompatibleVariableOverride", + "range": { + "startColumn": 13, + "endColumn": 17, + "lineCount": 1 + } + } + ], + "./git/refs/symbolic.py": [ + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 15, + "endColumn": 20, + "lineCount": 1 + } + } + ], + "./git/refs/tag.py": [ + { + "code": "reportIncompatibleMethodOverride", + "range": { + "startColumn": 8, + "endColumn": 14, + "lineCount": 1 + } + }, + { + "code": "reportIncompatibleMethodOverride", + "range": { + "startColumn": 8, + "endColumn": 14, + "lineCount": 1 + } + } + ], + "./git/remote.py": [ + { + "code": "reportArgumentType", + "range": { + "startColumn": 25, + "endColumn": 36, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 29, + "endColumn": 40, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 26, + "endColumn": 38, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 26, + "endColumn": 38, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 26, + "endColumn": 38, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 26, + "endColumn": 38, + "lineCount": 1 + } + } + ], + "./git/repo/base.py": [ + { + "code": "reportRedeclaration", + "range": { + "startColumn": 4, + "endColumn": 15, + "lineCount": 1 + } + }, + { + "code": "reportAttributeAccessIssue", + "range": { + "startColumn": 18, + "endColumn": 22, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 35, + "endColumn": 41, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 15, + "endColumn": 46, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 15, + "endColumn": 51, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 15, + "endColumn": 28, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 16, + "endColumn": 31, + "lineCount": 1 + } + }, + { + "code": "reportTypedDictNotRequiredAccess", + "range": { + "startColumn": 21, + "endColumn": 31, + "lineCount": 1 + } + }, + { + "code": "reportTypedDictNotRequiredAccess", + "range": { + "startColumn": 34, + "endColumn": 44, + "lineCount": 1 + } + }, + { + "code": "reportTypedDictNotRequiredAccess", + "range": { + "startColumn": 65, + "endColumn": 79, + "lineCount": 1 + } + }, + { + "code": "reportTypedDictNotRequiredAccess", + "range": { + "startColumn": 82, + "endColumn": 102, + "lineCount": 1 + } + }, + { + "code": "reportTypedDictNotRequiredAccess", + "range": { + "startColumn": 50, + "endColumn": 69, + "lineCount": 1 + } + }, + { + "code": "reportTypedDictNotRequiredAccess", + "range": { + "startColumn": 68, + "endColumn": 85, + "lineCount": 1 + } + }, + { + "code": "reportTypedDictNotRequiredAccess", + "range": { + "startColumn": 88, + "endColumn": 111, + "lineCount": 1 + } + }, + { + "code": "reportTypedDictNotRequiredAccess", + "range": { + "startColumn": 51, + "endColumn": 73, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 12, + "endColumn": 26, + "lineCount": 1 + } + }, + { + "code": "reportAssignmentType", + "range": { + "startColumn": 22, + "endColumn": 38, + "lineCount": 1 + } + } + ], + "./git/repo/fun.py": [ + { + "code": "reportArgumentType", + "range": { + "startColumn": 25, + "endColumn": 33, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 35, + "endColumn": 43, + "lineCount": 1 + } + }, + { + "code": "reportAssignmentType", + "range": { + "startColumn": 18, + "endColumn": 28, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 11, + "endColumn": 14, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 24, + "endColumn": 27, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 11, + "endColumn": 14, + "lineCount": 1 + } + }, + { + "code": "reportReturnType", + "range": { + "startColumn": 11, + "endColumn": 20, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 25, + "endColumn": 28, + "lineCount": 1 + } + }, + { + "code": "reportArgumentType", + "range": { + "startColumn": 24, + "endColumn": 27, + "lineCount": 1 + } + } + ], + "./test/deprecation/test_basic.py": [ + { + "code": "reportUnusedExpression", + "range": { + "startColumn": 12, + "endColumn": 62, + "lineCount": 1 + } + } + ] + } +} diff --git a/.github/dependabot.yml b/.github/dependabot.yml index 2fe73ca77..7f803ee43 100644 --- a/.github/dependabot.yml +++ b/.github/dependabot.yml @@ -1,11 +1,15 @@ version: 2 +enable-beta-ecosystems: true updates: -- package-ecosystem: "github-actions" +- package-ecosystem: github-actions directory: "/" schedule: - interval: "weekly" + interval: weekly -- package-ecosystem: "gitsubmodule" +- package-ecosystem: pre-commit directory: "/" schedule: - interval: "weekly" + interval: monthly + groups: + pre-commit: + patterns: ["*"] diff --git a/.github/workflows/alpine-test.yml b/.github/workflows/alpine-test.yml index bd09a939b..b10336a15 100644 --- a/.github/workflows/alpine-test.yml +++ b/.github/workflows/alpine-test.yml @@ -1,6 +1,13 @@ name: test-alpine -on: [push, pull_request, workflow_dispatch] +on: + push: + branches: [main] + pull_request: + workflow_dispatch: + +permissions: + contents: read jobs: test: @@ -23,7 +30,7 @@ jobs: adduser runner docker shell: sh -exo pipefail {0} # Run this as root, not the "runner" user. - - uses: actions/checkout@v4 + - uses: actions/checkout@v7 with: fetch-depth: 0 @@ -44,20 +51,36 @@ jobs: # and cause subsequent tests to fail cat test/fixtures/.gitconfig >> ~/.gitconfig - - name: Set up virtualenv + - name: Set up virtual environment run: | python -m venv .venv - name: Update PyPA packages run: | - # Get the latest pip, wheel, and prior to Python 3.12, setuptools. . .venv/bin/activate - python -m pip install -U pip $(pip freeze --all | grep -ow ^setuptools) wheel + python -m pip install -U pip 'setuptools; python_version<"3.12"' wheel - name: Install project and test dependencies run: | . .venv/bin/activate - pip install ".[test]" + pip install ./smmap ./gitdb '.[test]' + + - name: Show POSIX file ownership + run: | + ls -ld -- \ + "$(pwd)" \ + "$(pwd)/.git" \ + "${HOME:?HOME is not set}/.gitconfig" \ + 2>&1 || true + + - name: Show safe.directory entries + # `actions/checkout`'s safe.directory add is only durable for the + # checkout itself (it writes under a throwaway HOME override and + # then discards it), so by the time this step runs the runner + # user's `~/.gitconfig` has no entries -- and the Alpine container + # chowns the workspace to runner:docker to match the test user, so + # git accepts the ownership without one. Expected: `(none)`. + run: git config --global --get-all safe.directory || echo "(none)" - name: Show version and platform information run: | diff --git a/.github/workflows/codeql.yml b/.github/workflows/codeql.yml index ae5241898..eeb68cd0c 100644 --- a/.github/workflows/codeql.yml +++ b/.github/workflows/codeql.yml @@ -13,44 +13,59 @@ name: "CodeQL" on: push: + branches: [main] pull_request: schedule: - cron: '27 10 * * 3' jobs: analyze: - name: Analyze + name: Analyze (${{ matrix.language }}) # Runner size impacts CodeQL analysis time. To learn more, please see: # - https://gh.io/recommended-hardware-resources-for-running-codeql # - https://gh.io/supported-runners-and-hardware-resources - # - https://gh.io/using-larger-runners - # Consider using larger runners for possible analysis time improvements. + # - https://gh.io/using-larger-runners (GitHub.com only) + # Consider using larger runners or machines with greater resources for possible analysis time improvements. runs-on: ${{ (matrix.language == 'swift' && 'macos-latest') || 'ubuntu-latest' }} - timeout-minutes: ${{ (matrix.language == 'swift' && 120) || 360 }} permissions: - actions: read - contents: read security-events: write strategy: fail-fast: false matrix: - language: [ 'python' ] - # CodeQL supports [ 'c-cpp', 'csharp', 'go', 'java-kotlin', 'javascript-typescript', 'python', 'ruby', 'swift' ] - # Use only 'java-kotlin' to analyze code written in Java, Kotlin or both - # Use only 'javascript-typescript' to analyze code written in JavaScript, TypeScript or both - # Learn more about CodeQL language support at https://aka.ms/codeql-docs/language-support - + include: + - language: actions + build-mode: none + - language: python + build-mode: none + # CodeQL supports the following values keywords for 'language': 'actions', 'c-cpp', 'csharp', 'go', 'java-kotlin', 'javascript-typescript', 'python', 'ruby', 'swift' + # Use `c-cpp` to analyze code written in C, C++ or both + # Use 'java-kotlin' to analyze code written in Java, Kotlin or both + # Use 'javascript-typescript' to analyze code written in JavaScript, TypeScript or both + # To learn more about changing the languages that are analyzed or customizing the build mode for your analysis, + # see https://docs.github.com/en/code-security/code-scanning/creating-an-advanced-setup-for-code-scanning/customizing-your-advanced-setup-for-code-scanning. + # If you are analyzing a compiled language, you can modify the 'build-mode' for that language to customize how + # your codebase is analyzed, see https://docs.github.com/en/code-security/code-scanning/creating-an-advanced-setup-for-code-scanning/codeql-code-scanning-for-compiled-languages steps: - name: Checkout repository - uses: actions/checkout@v4 + uses: actions/checkout@v7 + + # Add any setup steps before running the `github/codeql-action/init` action. + # This includes steps like installing compilers or runtimes (`actions/setup-node` + # or others). This is typically only required for manual builds. + # - name: Setup runtime (example) + # uses: actions/setup-example@v1 # Initializes the CodeQL tools for scanning. - name: Initialize CodeQL - uses: github/codeql-action/init@v3 + uses: github/codeql-action/init@v4 with: languages: ${{ matrix.language }} - setup-python-dependencies: false + build-mode: ${{ matrix.build-mode }} + config: | + paths-ignore: + - gitdb/gitdb/test/** + - smmap/smmap/test/** # If you wish to specify custom queries, you can do so here or in a config file. # By default, queries listed here will override any specified in a config file. # Prefix the list here with "+" to use these queries and those in the config file. @@ -58,23 +73,23 @@ jobs: # For more details on CodeQL's query packs, refer to: https://docs.github.com/en/code-security/code-scanning/automatically-scanning-your-code-for-vulnerabilities-and-errors/configuring-code-scanning#using-queries-in-ql-packs # queries: security-extended,security-and-quality - - # Autobuild attempts to build any compiled languages (C/C++, C#, Go, Java, or Swift). - # If this step fails, then you should remove it and run the build manually (see below) - - name: Autobuild - uses: github/codeql-action/autobuild@v3 - + # If the analyze step fails for one of the languages you are analyzing with + # "We were unable to automatically build your code", modify the matrix above + # to set the build mode to "manual" for that language. Then modify this step + # to build your code. # ℹī¸ Command-line programs to run using the OS shell. # 📚 See https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions#jobsjob_idstepsrun - - # If the Autobuild fails above, remove it and uncomment the following three lines. - # modify them (or add more) to build your code if your project, please refer to the EXAMPLE below for guidance. - - # - run: | - # echo "Run, Build Application using script" - # ./location_of_script_within_repo/buildscript.sh + - if: matrix.build-mode == 'manual' + shell: bash + run: | + echo 'If you are using a "manual" build mode for one or more of the' \ + 'languages you are analyzing, replace this with the commands to build' \ + 'your code, for example:' + echo ' make bootstrap' + echo ' make release' + exit 1 - name: Perform CodeQL Analysis - uses: github/codeql-action/analyze@v3 + uses: github/codeql-action/analyze@v4 with: category: "/language:${{matrix.language}}" diff --git a/.github/workflows/cygwin-test.yml b/.github/workflows/cygwin-test.yml index d42eb0587..b8150e93d 100644 --- a/.github/workflows/cygwin-test.yml +++ b/.github/workflows/cygwin-test.yml @@ -1,36 +1,52 @@ name: test-cygwin -on: [push, pull_request, workflow_dispatch] +on: + push: + branches: [main] + pull_request: + workflow_dispatch: + +permissions: + contents: read jobs: test: - runs-on: windows-latest - strategy: + matrix: + selection: [fast, perf] + include: + - selection: fast + additional-pytest-args: --ignore=test/performance + - selection: perf + additional-pytest-args: test/performance + fail-fast: false + runs-on: windows-latest + env: CHERE_INVOKING: "1" CYGWIN_NOWINPATH: "1" defaults: run: - shell: C:\tools\cygwin\bin\bash.exe --login --norc -eo pipefail -o igncr "{0}" + shell: C:\cygwin\bin\bash.exe --login --norc -eo pipefail -o igncr "{0}" steps: - name: Force LF line endings run: | git config --global core.autocrlf false # Affects the non-Cygwin git. - shell: bash # Use Git Bash instead of Cygwin Bash for this step. + shell: pwsh # Do this outside Cygwin, to affect actions/checkout. - - uses: actions/checkout@v4 + - uses: actions/checkout@v7 with: fetch-depth: 0 - - name: Set up Cygwin - uses: egor-tensin/setup-cygwin@v4 + - name: Install Cygwin + uses: cygwin/cygwin-install-action@v6 with: - packages: python39 python39-pip python39-virtualenv git + packages: git python39 python-pip-wheel python-setuptools-wheel python-wheel-wheel + add-to-path: false # No need to change $PATH outside the Cygwin environment. - name: Arrange for verbose output run: | @@ -55,19 +71,58 @@ jobs: # and cause subsequent tests to fail cat test/fixtures/.gitconfig >> ~/.gitconfig - - name: Set up virtualenv + - name: Set up virtual environment run: | - python -m venv .venv + python3.9 -m venv .venv echo 'BASH_ENV=.venv/bin/activate' >>"$GITHUB_ENV" - name: Update PyPA packages run: | - # Get the latest pip, wheel, and prior to Python 3.12, setuptools. - python -m pip install -U pip $(pip freeze --all | grep -ow ^setuptools) wheel + python -m pip install -U pip 'setuptools; python_version<"3.12"' wheel - name: Install project and test dependencies run: | - pip install ".[test]" + pip install ./smmap ./gitdb '.[test]' + + - name: Show POSIX file ownership + # Cygwin's `ls -ld` reports the NTFS Owner SID via Cygwin's SID-to-uid + # mapping (well-known SIDs by their RID, machine-local accounts by + # 0x30000+RID). That mapping is what Cygwin git's + # `is_path_owned_by_current_user` reduces to, so this is the view that + # determines whether `safe.directory` is consulted. + run: | + ls -ld -- \ + "$(pwd)" \ + "$(pwd)/.git" \ + "${HOME:?HOME is not set}/.gitconfig" \ + 2>&1 || true + + - name: Show NTFS file ownership + # Authoritative NTFS Owner via Get-Acl, with no Cygwin SID-to-uid layer + # in between -- useful for confirming what the Cygwin view reports as + # "Administrators" is the BUILTIN\Administrators SID (S-1-5-32-544). + shell: pwsh + run: | + $paths = @( + "$pwd", + "$pwd\.git", + "$env:USERPROFILE\.gitconfig" + ) + foreach ($p in $paths) { + if (Test-Path -LiteralPath $p) { + try { + $owner = (Get-Acl -LiteralPath $p).Owner + } catch { + $owner = "ERROR: $($_.Exception.Message)" + } + "{0,-44} {1}" -f $owner, $p + } else { + "(missing: $p)" + } + } + + - name: Show safe.directory entries + run: git config --global --get-all safe.directory - name: Show version and platform information run: | @@ -77,6 +132,6 @@ jobs: python --version python -c 'import os, sys; print(f"sys.platform={sys.platform!r}, os.name={os.name!r}")' - - name: Test with pytest + - name: Test with pytest (${{ matrix.additional-pytest-args }}) run: | - pytest --color=yes -p no:sugar --instafail -vv + pytest --color=yes -p no:sugar --instafail -vv ${{ matrix.additional-pytest-args }} diff --git a/.github/workflows/dependencies.yml b/.github/workflows/dependencies.yml new file mode 100644 index 000000000..4803e733b --- /dev/null +++ b/.github/workflows/dependencies.yml @@ -0,0 +1,42 @@ +name: Dependency packages + +on: + push: + branches: [main] + pull_request: + workflow_dispatch: + +permissions: + contents: read + +jobs: + test: + runs-on: ${{ matrix.python-version == '3.7' && 'ubuntu-22.04' || 'ubuntu-latest' }} + strategy: + fail-fast: false + matrix: + project: [smmap, gitdb] + python-version: ["3.7", "3.8", "3.9", "3.10", "3.11", "3.12", "3.13", "3.13t"] + + steps: + - uses: actions/checkout@v7 + with: + fetch-depth: 0 + - uses: actions/setup-python@v7 + with: + python-version: ${{ matrix.python-version }} + allow-prereleases: true + - name: Install test dependencies + run: | + python -m pip install --upgrade pip + python -m pip install pytest flake8 ./smmap + if test "${{ matrix.project }}" = gitdb; then python -m pip install ./gitdb; fi + - name: Lint + run: flake8 "${{ matrix.project }}/${{ matrix.project }}" --count --select=E9,F63,F7,F82 --show-source --statistics + - name: Test + run: | + if test "${{ matrix.project }}" = gitdb; then + GITDB_TEST_GIT_REPO_BASE="$(git rev-parse --git-common-dir)" pytest -o addopts= -v gitdb/gitdb/test + else + pytest -o addopts= -v smmap/smmap/test + fi diff --git a/.github/workflows/lint.yml b/.github/workflows/lint.yml index a0e81a993..d22640951 100644 --- a/.github/workflows/lint.yml +++ b/.github/workflows/lint.yml @@ -1,15 +1,22 @@ name: Lint -on: [push, pull_request, workflow_dispatch] +on: + push: + branches: [main] + pull_request: + workflow_dispatch: + +permissions: + contents: read jobs: lint: - runs-on: ubuntu-latest + runs-on: ubuntu-slim steps: - - uses: actions/checkout@v4 + - uses: actions/checkout@v7 - - uses: actions/setup-python@v5 + - uses: actions/setup-python@v7 with: python-version: "3.x" diff --git a/.github/workflows/pythonpackage.yml b/.github/workflows/pythonpackage.yml index 68b4fd8f9..7dfa2e8f2 100644 --- a/.github/workflows/pythonpackage.yml +++ b/.github/workflows/pythonpackage.yml @@ -3,7 +3,11 @@ name: Python package -on: [push, pull_request, workflow_dispatch] +on: + push: + branches: [main] + pull_request: + workflow_dispatch: permissions: contents: read @@ -13,18 +17,29 @@ jobs: strategy: matrix: os-type: [ubuntu, macos, windows] - python-version: ["3.7", "3.8", "3.9", "3.10", "3.11", "3.12", "3.13"] + python-version: ["3.7", "3.8", "3.9", "3.10", "3.11", "3.12", "3.13", "3.14", "3.14t", "3.15", "3.15t"] exclude: - os-type: macos python-version: "3.7" # Not available for the ARM-based macOS runners. - - os-type: windows - python-version: "3.13" # FIXME: Fix and enable Python 3.13 on Windows (#1955). + - os-type: macos + python-version: "3.14t" + - os-type: macos + python-version: "3.15t" include: - os-ver: latest - os-type: ubuntu python-version: "3.7" os-ver: "22.04" + - build-docs: true # We ensure documentation builds, except on very old interpreters. + - python-version: "3.7" + build-docs: false + - python-version: "3.8" + build-docs: false - experimental: false + - python-version: "3.15" + experimental: true + - python-version: "3.15t" + experimental: true fail-fast: false @@ -35,22 +50,21 @@ jobs: shell: bash --noprofile --norc -exo pipefail {0} steps: - - uses: actions/checkout@v4 + - uses: actions/checkout@v7 with: fetch-depth: 0 - name: Set up Python ${{ matrix.python-version }} - uses: actions/setup-python@v5 + uses: actions/setup-python@v7 with: python-version: ${{ matrix.python-version }} allow-prereleases: ${{ matrix.experimental }} - name: Set up WSL (Windows) if: matrix.os-type == 'windows' - uses: Vampire/setup-wsl@v4.1.0 + uses: Vampire/setup-wsl@v7.0.0 with: - distribution: Alpine - additional-packages: bash + wsl-version: 1 - name: Prepare this repo for tests run: | @@ -66,12 +80,59 @@ jobs: - name: Update PyPA packages run: | - # Get the latest pip, wheel, and prior to Python 3.12, setuptools. - python -m pip install -U pip $(pip freeze --all | grep -ow ^setuptools) wheel + python -m pip install -U pip 'setuptools; python_version<"3.12"' wheel - name: Install project and test dependencies run: | - pip install ".[test]" + pip install ./smmap ./gitdb '.[test]' + + - name: Show POSIX file ownership + # Linux and macOS only. On Windows, Git Bash's `ls -ld` reports a + # uniform uid+gid for every path regardless of NTFS Owner (MSYS2's + # SID-to-uid mapping doesn't have Cygwin's fidelity), so it would + # not be informative here. The NTFS Owner check below covers Windows. + if: matrix.os-type != 'windows' + run: | + ls -ld -- \ + "$(pwd)" \ + "$(pwd)/.git" \ + "${HOME:?HOME is not set}/.gitconfig" \ + 2>&1 || true + + - name: Show NTFS file ownership + # Windows only. Reads NTFS Owner directly via Get-Acl, which is the + # authoritative view for Windows-side ownership questions; the POSIX + # view via Git Bash's MSYS2 layer is not a reliable proxy here. + if: matrix.os-type == 'windows' + shell: pwsh + run: | + $paths = @( + "$pwd", + "$pwd\.git", + "$env:USERPROFILE\.gitconfig" + ) + foreach ($p in $paths) { + if (Test-Path -LiteralPath $p) { + try { + $owner = (Get-Acl -LiteralPath $p).Owner + } catch { + $owner = "ERROR: $($_.Exception.Message)" + } + "{0,-44} {1}" -f $owner, $p + } else { + "(missing: $p)" + } + } + + - name: Show safe.directory entries + # `actions/checkout`'s safe.directory add is only durable for the + # checkout itself (it writes under a throwaway HOME override and + # then discards it), so by the time this step runs the runner + # user's `~/.gitconfig` has no entries -- and git accepts the + # workspace's ownership anyway: Git for Windows via its + # Admins-group exemption on the windows matrix; on Linux/macOS + # the workspace is owned by the test user. Expected: `(none)`. + run: git config --global --get-all safe.directory || echo "(none)" - name: Show version and platform information run: | @@ -91,14 +152,18 @@ jobs: continue-on-error: true - name: Check types with mypy + if: matrix.python-version != '3.7' && matrix.python-version != '3.8' run: | - mypy --python-version=${{ matrix.python-version }} + mypy --python-version="${PYTHON_VERSION%t}" # Version only, with no "t" for free-threaded. env: MYPY_FORCE_COLOR: "1" TERM: "xterm-256color" # For color: https://github.com/python/mypy/issues/13817 - # With new versions of mypy new issues might arise. This is a problem if there is - # nobody able to fix them, so we have to ignore errors until that changes. - continue-on-error: true + PYTHON_VERSION: ${{ matrix.python-version }} + + - name: Check types with basedpyright + if: matrix.os-type == 'ubuntu' && matrix.python-version == '3.12' + run: | + basedpyright --warnings - name: Test with pytest run: | @@ -106,7 +171,7 @@ jobs: continue-on-error: false - name: Documentation - if: matrix.python-version != '3.7' + if: matrix.build-docs run: | - pip install ".[doc]" + pip install '.[doc]' make -C doc html diff --git a/.gitignore b/.gitignore index d85569405..eab294a65 100644 --- a/.gitignore +++ b/.gitignore @@ -10,6 +10,8 @@ __pycache__/ # Transient editor files *.swp *~ +\#*# +.#*# # Editor configuration nbproject diff --git a/.gitmodules b/.gitmodules deleted file mode 100644 index 251eeeec4..000000000 --- a/.gitmodules +++ /dev/null @@ -1,3 +0,0 @@ -[submodule "gitdb"] - url = https://github.com/gitpython-developers/gitdb.git - path = git/ext/gitdb diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index 424cc5f37..75a29d6db 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -1,29 +1,31 @@ +exclude: ^(?:gitdb|smmap)/ + repos: - repo: https://github.com/codespell-project/codespell - rev: v2.3.0 + rev: v2.4.3 hooks: - id: codespell additional_dependencies: [tomli] exclude: ^test/fixtures/ - repo: https://github.com/astral-sh/ruff-pre-commit - rev: v0.6.0 + rev: v0.16.0 hooks: - - id: ruff + - id: ruff-check args: ["--fix"] exclude: ^git/ext/ - id: ruff-format exclude: ^git/ext/ - repo: https://github.com/shellcheck-py/shellcheck-py - rev: v0.10.0.1 + rev: v0.11.0.1 hooks: - id: shellcheck args: [--color] exclude: ^test/fixtures/polyglot$|^git/ext/ - repo: https://github.com/pre-commit/pre-commit-hooks - rev: v4.6.0 + rev: v6.0.0 hooks: - id: end-of-file-fixer exclude: ^test/fixtures/|COPYING|LICENSE @@ -33,6 +35,6 @@ repos: - id: check-merge-conflict - repo: https://github.com/abravalheri/validate-pyproject - rev: v0.19 + rev: v0.25 hooks: - id: validate-pyproject diff --git a/AUTHORS b/AUTHORS index 45b14c961..95205f439 100644 --- a/AUTHORS +++ b/AUTHORS @@ -55,5 +55,8 @@ Contributors are: -Eliah Kagan -Ethan Lin -Jonas Scharpf +-Gordon Marx +-Enji Cooper +-Harshita Yadav Portions derived from other open source works and are clearly marked. diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 8536d7f73..76f276323 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -9,6 +9,28 @@ The following is a short step-by-step rundown of what one typically would do to - Feel free to add yourself to AUTHORS file. - Create a pull request. +## Quality expectations + +Contributions must be made with care and meet the quality bar of the surrounding code. +That means a change should not leave GitPython worse than it was before: it should be +readable, maintainable, tested where practical, documented and consistent with the +existing style and behavior. + +A contribution that works only narrowly but lowers the quality of the +codebase may be declined. The maintainers may not always be able to provide +detailed feedback. + +## Prevent agent impersonation + +AI agents communicating through a person's account must identify themselves, for +example in issue or PR descriptions and comments. AI assistance that does not replace +the person as the speaker, such as proofreading or wording polish, does not require +identification. + +Attributing AI assistance in commit metadata, for example with a `Co-authored-by` +trailer, is welcome but not required. Code is reviewed the same way regardless of its +origin. + ## Fuzzing Test Specific Documentation For details related to contributing to the fuzzing test suite and OSS-Fuzz integration, please diff --git a/README.md b/README.md index 59c6f995b..c5a069bfe 100644 --- a/README.md +++ b/README.md @@ -46,7 +46,7 @@ by setting the `GIT_PYTHON_GIT_EXECUTABLE=` environment variable. - Git (1.7.x or newer) - Python >= 3.7 -The list of dependencies are listed in `./requirements.txt` and `./test-requirements.txt`. +The list of dependencies are listed in [`./requirements.txt`](https://github.com/gitpython-developers/GitPython/blob/main/requirements.txt) and [`./test-requirements.txt`](https://github.com/gitpython-developers/GitPython/blob/main/test-requirements.txt). The installer takes care of installing them for you. ### INSTALL @@ -101,15 +101,13 @@ In the less common case that you do not want to install test dependencies, `pip #### With editable *dependencies* (not preferred, and rarely needed) -In rare cases, you may want to work on GitPython and one or both of its [gitdb](https://github.com/gitpython-developers/gitdb) and [smmap](https://github.com/gitpython-developers/smmap) dependencies at the same time, with changes in your local working copy of gitdb or smmap immediately reflected in the behavior of your local working copy of GitPython. This can be done by making editable installations of those dependencies in the same virtual environment where you install GitPython. - -If you want to do that *and* you want the versions in GitPython's git submodules to be used, then pass `-e git/ext/gitdb` and/or `-e git/ext/gitdb/gitdb/ext/smmap` to `pip install`. This can be done in any order, and in separate `pip install` commands or the same one, so long as `-e` appears before *each* path. For example, you can install GitPython, gitdb, and smmap editably in the currently active virtual environment this way: +GitPython, [gitdb](gitdb), and [smmap](smmap) live in this repository as independently packaged projects. To work on all three at once, install each one editably in the same virtual environment: ```sh -pip install -e ".[test]" -e git/ext/gitdb -e git/ext/gitdb/gitdb/ext/smmap +pip install -e smmap -e gitdb -e ".[test]" ``` -The submodules must have been cloned for that to work, but that will already be the case if you have run `./init-tests-after-clone.sh`. You can use `pip list` to check which packages are installed editably and which are installed normally. +You can use `pip list` to check which packages are installed editably and which are installed normally. To reiterate, this approach should only rarely be used. For most development it is preferable to allow the gitdb and smmap dependencices to be retrieved automatically from PyPI in their latest stable packaged versions. @@ -180,7 +178,7 @@ Style and formatting checks, and running tests on all the different supported Py #### Configuration files -Specific tools are all configured in the `./pyproject.toml` file: +Specific tools are all configured in the [`./pyproject.toml`](https://github.com/gitpython-developers/GitPython/blob/main/pyproject.toml) file: - `pytest` (test runner) - `coverage.py` (code coverage) @@ -189,9 +187,9 @@ Specific tools are all configured in the `./pyproject.toml` file: Orchestration tools: -- Configuration for `pre-commit` is in the `./.pre-commit-config.yaml` file. -- Configuration for `tox` is in `./tox.ini`. -- Configuration for GitHub Actions (CI) is in files inside `./.github/workflows/`. +- Configuration for `pre-commit` is in the [`./.pre-commit-config.yaml`](https://github.com/gitpython-developers/GitPython/blob/main/.pre-commit-config.yaml) file. +- Configuration for `tox` is in [`./tox.ini`](https://github.com/gitpython-developers/GitPython/blob/main/tox.ini). +- Configuration for GitHub Actions (CI) is in files inside [`./.github/workflows/`](https://github.com/gitpython-developers/GitPython/tree/main/.github/workflows). ### Contributions @@ -212,8 +210,8 @@ Please have a look at the [contributions file][contributing]. ### How to make a new release -1. Update/verify the **version** in the `VERSION` file. -2. Update/verify that the `doc/source/changes.rst` changelog file was updated. It should include a link to the forthcoming release page: `https://github.com/gitpython-developers/GitPython/releases/tag/` +1. Update/verify the **version** in the [`VERSION`](https://github.com/gitpython-developers/GitPython/blob/main/VERSION) file. +2. Update/verify that the [`doc/source/changes.rst`](https://github.com/gitpython-developers/GitPython/blob/main/doc/source/changes.rst) changelog file was updated. It should include a link to the forthcoming release page: `https://github.com/gitpython-developers/GitPython/releases/tag/` 3. Commit everything. 4. Run `git tag -s ` to tag the version in Git. 5. _Optionally_ create and activate a [virtual environment](https://packaging.python.org/en/latest/guides/installing-using-pip-and-virtual-environments/#creating-a-virtual-environment). (Then the next step can install `build` and `twine`.) @@ -240,7 +238,7 @@ Please have a look at the [contributions file][contributing]. [3-Clause BSD License](https://opensource.org/license/bsd-3-clause/), also known as the New BSD License. See the [LICENSE file][license]. -One file exclusively used for fuzz testing is subject to [a separate license, detailed here](./fuzzing/README.md#license). +One file exclusively used for fuzz testing is subject to [a separate license, detailed here](https://github.com/gitpython-developers/GitPython/blob/main/fuzzing/README.md#license). This file is not included in the wheel or sdist packages published by the maintainers of GitPython. [contributing]: https://github.com/gitpython-developers/GitPython/blob/main/CONTRIBUTING.md diff --git a/VERSION b/VERSION index e6af1c454..dfabc766a 100644 --- a/VERSION +++ b/VERSION @@ -1 +1 @@ -3.1.44 +3.1.58 diff --git a/doc/Makefile b/doc/Makefile index ddeadbd7e..7e0d325fe 100644 --- a/doc/Makefile +++ b/doc/Makefile @@ -3,7 +3,7 @@ # You can set these variables from the command line. BUILDDIR = build -SPHINXOPTS = -W +SPHINXOPTS = SPHINXBUILD = sphinx-build PAPER = diff --git a/doc/requirements.txt b/doc/requirements.txt index 81140d898..24472ba39 100644 --- a/doc/requirements.txt +++ b/doc/requirements.txt @@ -1,3 +1,3 @@ -sphinx >= 7.1.2, < 7.2 +sphinx >= 7.4.7, < 8 sphinx_rtd_theme sphinx-autodoc-typehints diff --git a/doc/source/changes.rst b/doc/source/changes.rst index 00a3c660e..714cc7ffc 100644 --- a/doc/source/changes.rst +++ b/doc/source/changes.rst @@ -2,6 +2,180 @@ Changelog ========= +3.1.59 +====== + +Security fixes for + +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-5xxx-qhh7-9287 +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-3wxw-xv34-2frg + +If you can, also try and provide feedback on the upcoming v4 branch +https://github.com/gitpython-developers/GitPython/pull/2177 - patches welcome. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.59 + +3.1.58 +====== + +Security fixes for + +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-hh9p-6wh2-4mfc +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-9rj7-rf2p-w77r +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-4gmw-gg2m-w46p +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-wvpp-8hx9-p66j +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-jm78-9fvv-mhgr +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-hmq2-w58f-27jc + +If you can, also try and provide feedback on the upcoming v4 branch +https://github.com/gitpython-developers/GitPython/pull/2177 - patches welcome. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.58 + +3.1.57 +====== + +A security fix for + +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-3f7w-8rr8-f37f +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-539m-9xh6-q6rr + +If you can, also try and provide feedback on the upcoming v4 branch +https://github.com/gitpython-developers/GitPython/pull/2177 - patches welcome. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.57 + +3.1.56 +====== + +A security fix for + +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-p538-c434-8v24 + +If you can, also try and provide feedback on the upcoming v4 branch +https://github.com/gitpython-developers/GitPython/pull/2177 - patches welcome. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.56 + +3.1.55 +====== + +A security fix for + +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-94p4-4cq8-9g67 + +If you can, also try and provide feedback on the upcoming v4 branch +https://github.com/gitpython-developers/GitPython/pull/2177 - patches welcome. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.55 + +3.1.54 +====== + +A security fix for + +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-r9mr-m37c-5fr3 +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-6p8h-3wgx-97gf +* https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-fjr4-x663-mwxc + +If you can, also try and provide feedback on the upcoming v4 branch +https://github.com/gitpython-developers/GitPython/pull/2177 - patches welcome. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.54 + +3.1.53 +====== + +A security fix for +https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-3rp5-jjmw-4wv2 + +If you can, also try and provide feedback on the upcoming v4 branch +https://github.com/gitpython-developers/GitPython/pull/2177 - patches welcome. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.53 + +3.1.52 +====== + +A security fix for +https://github.com/gitpython-developers/GitPython/security/advisories/GHSA-rwj8-pgh3-r573. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.52 + +3.1.51 +====== + +This is primarily a security release. It prevents additional argument-injection +paths that could allow execution of arbitrary commands or writing to arbitrary +files through unsafe Git options (GHSA-956x-8gvw-wg5v), and closes bypasses of +the existing protections using abbreviated long options or joined short options +(GHSA-2f96-g7mh-g2hx). + +The release also improves support for relative worktree paths and diffs against +the empty tree, preserves stderr from failed diff processes, supports relative +configuration include paths, and includes assorted documentation, typing, test, +and CI improvements. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.51 + +3.1.50 +====== + +Save setting of configuration values, this time sections as well, as follow-up to 3.1.49. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.50 + +3.1.49 +====== + +Save setting of configuration values, +which could be used to inject other more configuration. + +Also more conforming `rev-parse` implementation. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.49 + +3.1.48 +====== + +Safe reference creation in the face of untrusted input. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.48 + + +3.1.47 +====== + +Address various security issues related to bypassing injection-protection +of unsafe Git flags. + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.47 + +3.1.46 +====== + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.46 + +3.1.45 +====== + +See the following for all changes. +https://github.com/gitpython-developers/GitPython/releases/tag/3.1.45 + 3.1.44 ====== @@ -105,7 +279,7 @@ https://github.com/gitpython-developers/gitpython/milestone/61?closed=1 but a necessary fix for https://github.com/gitpython-developers/GitPython/issues/1515. Please take a look at the PR for more information and how to bypass these protections in case they cause breakage: https://github.com/gitpython-developers/GitPython/pull/1521. - + See the following for all changes. https://github.com/gitpython-developers/gitpython/milestone/60?closed=1 @@ -170,38 +344,38 @@ https://github.com/gitpython-developers/gitpython/milestone/53?closed=1 * General: - Remove python 3.6 support - + - Remove distutils ahead of deprecation in standard library. - + - Update sphinx to 4.1.12 and use autodoc-typehints. - + - Include README as long_description on PyPI - + - Test against earliest and latest minor version available on Github Actions (e.g. 3.9.0 and 3.9.7) - + * Typing: - Add types to ALL functions. - + - Ensure py.typed is collected. - + - Increase mypy strictness with disallow_untyped_defs, warn_redundant_casts, warn_unreachable. - + - Use typing.NamedTuple and typing.OrderedDict now 3.6 dropped. - + - Make Protocol classes ABCs at runtime due to new behaviour/bug in 3.9.7 & 3.10.0-rc1 - + - Remove use of typing.TypeGuard until later release, to allow dependent libs time to update. - + - Tracking issue: https://github.com/gitpython-developers/GitPython/issues/1095 * Runtime improvements: - Add clone_multi_options support to submodule.add() - + - Delay calling get_user_id() unless essential, to support sand-boxed environments. - + - Add timeout to handle_process_output(), in case thread.join() hangs. See the following for details: diff --git a/doc/source/intro.rst b/doc/source/intro.rst index d053bd117..e1075b1c9 100644 --- a/doc/source/intro.rst +++ b/doc/source/intro.rst @@ -97,11 +97,6 @@ and cloned using:: $ git clone https://github.com/gitpython-developers/GitPython git-python -Initialize all submodules to obtain the required dependencies with:: - - $ cd git-python - $ git submodule update --init --recursive - Finally verify the installation by running unit tests:: $ python -m unittest diff --git a/doc/source/reference.rst b/doc/source/reference.rst index 13dd38d02..49d732ca2 100644 --- a/doc/source/reference.rst +++ b/doc/source/reference.rst @@ -91,6 +91,7 @@ Objects.Util .. automodule:: git.objects.util :members: :undoc-members: + :private-members: _get_intermediate_items, _list_traverse, _traverse :special-members: Index.Base diff --git a/doc/source/tutorial.rst b/doc/source/tutorial.rst index fd3b14c57..d095d3be3 100644 --- a/doc/source/tutorial.rst +++ b/doc/source/tutorial.rst @@ -513,6 +513,12 @@ The GitDB is a pure-python implementation of the git object database. It is the repo = Repo("path/to/repo", odbt=GitDB) +.. warning:: + ``GitDB`` may fail or become extremely slow when traversing trees in + repositories with very large commits (thousands of changed files in a + single commit). If you encounter ``RecursionError`` or excessive + slowness during tree traversal, switch to ``GitCmdObjectDB`` instead. + GitCmdObjectDB ============== diff --git a/git/cmd.py b/git/cmd.py index 2048a43fa..193dfd4f6 100644 --- a/git/cmd.py +++ b/git/cmd.py @@ -60,6 +60,11 @@ overload, ) +if sys.version_info >= (3, 10): + from typing import TypeAlias +else: + from typing_extensions import TypeAlias + from git.types import Literal, PathLike, TBD if TYPE_CHECKING: @@ -207,7 +212,7 @@ def pump_stream( ) if stderr_handler: error_str: Union[str, bytes] = ( - "error: process killed because it timed out." f" kill_after_timeout={kill_after_timeout} seconds" + f"error: process killed because it timed out. kill_after_timeout={kill_after_timeout} seconds" ) if not decode_streams and isinstance(p_stderr, BinaryIO): # Assume stderr_handler needs binary input. @@ -268,12 +273,12 @@ def _safer_popen_windows( if shell: # The original may be immutable, or the caller may reuse it. Mutate a copy. env = {} if env is None else dict(env) - env["NoDefaultCurrentDirectoryInExePath"] = "1" # The "1" can be an value. + env["NoDefaultCurrentDirectoryInExePath"] = "1" # The "1" can be any value. # When not using a shell, the current process does the search in a # CreateProcessW API call, so the variable must be set in our environment. With # a shell, that's unnecessary if https://github.com/python/cpython/issues/101283 - # is patched. In Python versions where it is unpatched, and in the rare case the + # is patched. In Python versions where it is unpatched, in the rare case the # ComSpec environment variable is unset, the search for the shell itself is # unsafe. Setting NoDefaultCurrentDirectoryInExePath in all cases, as done here, # is simpler and protects against that. (As above, the "1" can be any value.) @@ -308,6 +313,234 @@ def dict_to_slots_and__excluded_are_none(self: object, d: Mapping[str, Any], exc ## -- End Utilities -- @} + +class _AutoInterrupt: + """Process wrapper that terminates the wrapped process on finalization. + + This kills/interrupts the stored process instance once this instance goes out of + scope. It is used to prevent processes piling up in case iterators stop reading. + + All attributes are wired through to the contained process object. + + The wait method is overridden to perform automatic status code checking and possibly + raise. + """ + + __slots__ = ("proc", "args", "status") + + # If this is non-zero it will override any status code during _terminate, used + # to prevent race conditions in testing. + _status_code_if_terminate: int = 0 + + def __init__(self, proc: Union[None, subprocess.Popen], args: Any) -> None: + self.proc = proc + self.args = args + self.status: Union[int, None] = None + + def _terminate(self) -> None: + """Terminate the underlying process.""" + if self.proc is None: + return + + proc = self.proc + self.proc = None + if proc.stdin: + proc.stdin.close() + if proc.stdout: + proc.stdout.close() + if proc.stderr: + proc.stderr.close() + # Did the process finish already so we have a return code? + try: + if proc.poll() is not None: + self.status = self._status_code_if_terminate or proc.poll() + return + except OSError as ex: + _logger.info("Ignored error after process had died: %r", ex) + + # It can be that nothing really exists anymore... + if os is None or getattr(os, "kill", None) is None: + return + + # Try to kill it. + try: + proc.terminate() + status = proc.wait() # Ensure the process goes away. + + self.status = self._status_code_if_terminate or status + except (OSError, AttributeError) as ex: + # On interpreter shutdown (notably on Windows), parts of the stdlib used by + # subprocess can already be torn down (e.g. `subprocess._winapi` becomes None), + # which can cause AttributeError during terminate(). In that case, we prefer + # to silently ignore to avoid noisy "Exception ignored in: __del__" messages. + _logger.info("Ignored error while terminating process: %r", ex) + # END exception handling + + def __del__(self) -> None: + self._terminate() + + def __getattr__(self, attr: str) -> Any: + return getattr(self.proc, attr) + + # TODO: Bad choice to mimic `proc.wait()` but with different args. + def wait(self, stderr: Union[None, str, bytes] = b"") -> int: + """Wait for the process and return its status code. + + :param stderr: + Previously read value of stderr, in case stderr is already closed. + + :warn: + May deadlock if output or error pipes are used and not handled separately. + + :raise git.exc.GitCommandError: + If the return status is not 0. + """ + if stderr is None: + stderr_b = b"" + stderr_b = force_bytes(data=stderr, encoding="utf-8") + status: Union[int, None] + if self.proc is not None: + status = self.proc.wait() + p_stderr = self.proc.stderr + else: # Assume the underlying proc was killed earlier or never existed. + status = self.status + p_stderr = None + + def read_all_from_possibly_closed_stream(stream: Union[IO[bytes], None]) -> bytes: + if stream: + try: + return stderr_b + force_bytes(stream.read()) + except (OSError, ValueError): + return stderr_b or b"" + else: + return stderr_b or b"" + + # END status handling + + if status != 0: + errstr = read_all_from_possibly_closed_stream(p_stderr) + _logger.debug("AutoInterrupt wait stderr: %r" % (errstr,)) + raise GitCommandError(remove_password_if_present(self.args), status, errstr) + return status + + +_AutoInterrupt.__name__ = "AutoInterrupt" +_AutoInterrupt.__qualname__ = "Git.AutoInterrupt" + + +class _CatFileContentStream: + """Object representing a sized read-only stream returning the contents of + an object. + + This behaves like a stream, but counts the data read and simulates an empty stream + once our sized content region is empty. + + If not all data are read to the end of the object's lifetime, we read the rest to + ensure the underlying stream continues to work. + """ + + __slots__ = ("_stream", "_nbr", "_size") + + def __init__(self, size: int, stream: IO[bytes]) -> None: + self._stream = stream + self._size = size + self._nbr = 0 # Number of bytes read. + + # Special case: If the object is empty, has null bytes, get the final + # newline right away. + if size == 0: + stream.read(1) + # END handle empty streams + + def read(self, size: int = -1) -> bytes: + bytes_left = self._size - self._nbr + if bytes_left == 0: + return b"" + if size > -1: + # Ensure we don't try to read past our limit. + size = min(bytes_left, size) + else: + # They try to read all, make sure it's not more than what remains. + size = bytes_left + # END check early depletion + data = self._stream.read(size) + self._nbr += len(data) + + # Check for depletion, read our final byte to make the stream usable by + # others. + if self._size - self._nbr == 0: + self._stream.read(1) # final newline + # END finish reading + return data + + def readline(self, size: int = -1) -> bytes: + if self._nbr == self._size: + return b"" + + # Clamp size to lowest allowed value. + bytes_left = self._size - self._nbr + if size > -1: + size = min(bytes_left, size) + else: + size = bytes_left + # END handle size + + data = self._stream.readline(size) + self._nbr += len(data) + + # Handle final byte. + if self._size - self._nbr == 0: + self._stream.read(1) + # END finish reading + + return data + + def readlines(self, size: int = -1) -> List[bytes]: + if self._nbr == self._size: + return [] + + # Leave all additional logic to our readline method, we just check the size. + out = [] + nbr = 0 + while True: + line = self.readline() + if not line: + break + out.append(line) + if size > -1: + nbr += len(line) + if nbr > size: + break + # END handle size constraint + # END readline loop + return out + + # skipcq: PYL-E0301 + def __iter__(self) -> "Git.CatFileContentStream": + return self + + def __next__(self) -> bytes: + line = self.readline() + if not line: + raise StopIteration + + return line + + next = __next__ + + def __del__(self) -> None: + bytes_left = self._size - self._nbr + if bytes_left: + # Read and discard - seeking is impossible within a stream. + # This includes any terminating newline. + self._stream.read(bytes_left + 1) + # END handle incomplete read + + +_CatFileContentStream.__name__ = "CatFileContentStream" +_CatFileContentStream.__qualname__ = "Git.CatFileContentStream" + + _USE_SHELL_DEFAULT_MESSAGE = ( "Git.USE_SHELL is deprecated, because only its default value of False is safe. " "It will be removed in a future release." @@ -321,7 +554,7 @@ def dict_to_slots_and__excluded_are_none(self: object, d: Mapping[str, Any], exc ) -def _warn_use_shell(extra_danger: bool) -> None: +def _warn_use_shell(*, extra_danger: bool) -> None: warnings.warn( _USE_SHELL_DANGER_MESSAGE if extra_danger else _USE_SHELL_DEFAULT_MESSAGE, DeprecationWarning, @@ -337,12 +570,12 @@ class _GitMeta(type): def __getattribute(cls, name: str) -> Any: if name == "USE_SHELL": - _warn_use_shell(False) + _warn_use_shell(extra_danger=False) return super().__getattribute__(name) def __setattr(cls, name: str, value: Any) -> Any: if name == "USE_SHELL": - _warn_use_shell(value) + _warn_use_shell(extra_danger=value) super().__setattr__(name, value) if not TYPE_CHECKING: @@ -416,6 +649,18 @@ class Git(metaclass=_GitMeta): re_unsafe_protocol = re.compile(r"(.+)::.+") + unsafe_git_ls_remote_options = [ + # This option allows arbitrary command execution in git-ls-remote. + "--upload-pack", + "--exec", + ] + + unsafe_git_pathspec_from_file_options = [ + # Reads pathspecs from a caller-controlled file. Some commands include an + # unmatched pathspec in their error output, which can disclose the file. + "--pathspec-from-file", + ] + def __getstate__(self) -> Dict[str, Any]: return slots_to_dict(self, exclude=self._excluded_) @@ -665,29 +910,34 @@ def is_cygwin(cls) -> bool: @overload @classmethod - def polish_url(cls, url: str, is_cygwin: Literal[False] = ...) -> str: ... + def polish_url(cls, url: str, is_cygwin: Literal[False] = ..., expand_vars: bool = ...) -> str: ... @overload @classmethod - def polish_url(cls, url: str, is_cygwin: Union[None, bool] = None) -> str: ... + def polish_url(cls, url: str, is_cygwin: Union[None, bool] = None, expand_vars: bool = True) -> str: ... @classmethod - def polish_url(cls, url: str, is_cygwin: Union[None, bool] = None) -> PathLike: + def polish_url(cls, url: str, is_cygwin: Union[None, bool] = None, expand_vars: bool = True) -> PathLike: """Remove any backslashes from URLs to be written in config files. Windows might create config files containing paths with backslashes, but git stops liking them as it will escape the backslashes. Hence we undo the escaping just to be sure. + + :param expand_vars: + Expand environment variables and an initial ``~``. Disable this for values + obtained from an untrusted source, such as remote URLs. """ if is_cygwin is None: is_cygwin = cls.is_cygwin() if is_cygwin: - url = cygpath(url) + url = cygpath(url, expand_vars=expand_vars) else: - url = os.path.expandvars(url) - if url.startswith("~"): - url = os.path.expanduser(url) + if expand_vars: + url = os.path.expandvars(url) + if url.startswith("~"): + url = os.path.expanduser(url) url = url.replace("\\\\", "\\").replace("\\", "/") return url @@ -712,237 +962,117 @@ def check_unsafe_protocols(cls, url: str) -> None: ) @classmethod - def check_unsafe_options(cls, options: List[str], unsafe_options: List[str]) -> None: - """Check for unsafe options. + def _canonicalize_option_name(cls, option: str) -> str: + """Return the option name used for unsafe-option checks. - Some options that are passed to ``git `` can be used to execute - arbitrary commands. These are blocked by default. + Examples: + ``"--upload-pack=/tmp/helper"`` -> ``"upload-pack"`` + ``"upload_pack"`` -> ``"upload-pack"`` + ``"--config core.filemode=false"`` -> ``"config"`` """ - # Options can be of the form `foo`, `--foo bar`, or `--foo=bar`, so we need to - # check if they start with "--foo" or if they are equal to "foo". - bare_unsafe_options = [option.lstrip("-") for option in unsafe_options] + option_name = option.lstrip("-").split("=", 1)[0] + option_tokens = option_name.split(None, 1) + if not option_tokens: + return "" + return dashify(option_tokens[0]) + + @classmethod + def check_unsafe_options( + cls, options: List[str], unsafe_options: List[str], clusterable_short_options: str = "46flnqsv" + ) -> None: + """Raise :class:`~git.exc.UnsafeOptionError` for blocked option spellings. + + In addition to exact matches, this rejects abbreviated long options accepted + by Git (for example, ``--upl`` for ``--upload-pack``) and unsafe short options + whose values are joined to the same token, including after clusterable flags + (for example, ``-uVALUE`` and ``-fuVALUE``). + + A list containing only bare names is treated as normalized keyword arguments, + so multi-character names such as ``upload_p`` are checked as long-option + abbreviations. If any item starts with ``-``, the list is treated as tokenized + command-line input: bare items can be option values and are not checked as + abbreviations. Thus ``["--origin", "upload"]`` is allowed. Single-dash options + use short-option parsing rather than broad prefix matching, preserving safe + attached values such as ``-oupstream`` and ``-bcurrent``. + + Some options passed to ``git `` can execute arbitrary commands and + are therefore blocked by default unless the caller explicitly allows them. + """ + # Options can be of the form `foo`, `--foo`, `--foo bar`, or `--foo=bar`. + # Git accepts any unambiguous prefix of a long option, so an abbreviated + # spelling such as `--upl` for `--upload-pack` must be rejected too. An + # option is unsafe if its canonical name is a prefix of any blocked + # option's canonical name. Only long options and multi-character kwargs + # can be abbreviations; single-character short options remain exact-match + # only. + canonical_unsafe_options = {cls._canonicalize_option_name(option): option for option in unsafe_options} + unsafe_short_options = { + canonical: option + for canonical, option in canonical_unsafe_options.items() + if option.startswith("-") and not option.startswith("--") and len(canonical) == 1 + } + # These value-less Git flags can be clustered before another short option + # (for example, ``-fuVALUE``). Stop at any other character because it may + # begin an attached value, as ``o`` does in the safe option ``-oupstream``. + clusterable_short_options_set = frozenset(clusterable_short_options) + options_are_kwargs = all(not option.startswith("-") for option in options) for option in options: - for unsafe_option, bare_option in zip(unsafe_options, bare_unsafe_options): - if option.startswith(unsafe_option) or option == bare_option: + candidate = cls._canonicalize_option_name(option) + if not candidate: + continue + unsafe_option = canonical_unsafe_options.get(candidate) + if unsafe_option is not None: + raise UnsafeOptionError(f"{unsafe_option} is not allowed, use `allow_unsafe_options=True` to allow it.") + option_token = option.split("=", 1)[0].split(None, 1)[0] + if option_token.startswith("-") and not option_token.startswith("--"): + for option_char in option_token[1:]: + unsafe_option = unsafe_short_options.get(option_char) + if unsafe_option is not None: + raise UnsafeOptionError( + f"{unsafe_option} is not allowed, use `allow_unsafe_options=True` to allow it." + ) + if option_char not in clusterable_short_options_set: + break + if not (option.startswith("--") or (options_are_kwargs and len(candidate) > 1)): + continue + for canonical, unsafe_option in canonical_unsafe_options.items(): + if canonical.startswith(candidate): raise UnsafeOptionError( f"{unsafe_option} is not allowed, use `allow_unsafe_options=True` to allow it." ) - class AutoInterrupt: - """Process wrapper that terminates the wrapped process on finalization. - - This kills/interrupts the stored process instance once this instance goes out of - scope. It is used to prevent processes piling up in case iterators stop reading. - - All attributes are wired through to the contained process object. - - The wait method is overridden to perform automatic status code checking and - possibly raise. - """ - - __slots__ = ("proc", "args", "status") - - # If this is non-zero it will override any status code during _terminate, used - # to prevent race conditions in testing. - _status_code_if_terminate: int = 0 - - def __init__(self, proc: Union[None, subprocess.Popen], args: Any) -> None: - self.proc = proc - self.args = args - self.status: Union[int, None] = None - - def _terminate(self) -> None: - """Terminate the underlying process.""" - if self.proc is None: - return - - proc = self.proc - self.proc = None - if proc.stdin: - proc.stdin.close() - if proc.stdout: - proc.stdout.close() - if proc.stderr: - proc.stderr.close() - # Did the process finish already so we have a return code? - try: - if proc.poll() is not None: - self.status = self._status_code_if_terminate or proc.poll() - return - except OSError as ex: - _logger.info("Ignored error after process had died: %r", ex) - - # It can be that nothing really exists anymore... - if os is None or getattr(os, "kill", None) is None: - return - - # Try to kill it. - try: - proc.terminate() - status = proc.wait() # Ensure the process goes away. - - self.status = self._status_code_if_terminate or status - except OSError as ex: - _logger.info("Ignored error after process had died: %r", ex) - # END exception handling - - def __del__(self) -> None: - self._terminate() - - def __getattr__(self, attr: str) -> Any: - return getattr(self.proc, attr) - - # TODO: Bad choice to mimic `proc.wait()` but with different args. - def wait(self, stderr: Union[None, str, bytes] = b"") -> int: - """Wait for the process and return its status code. - - :param stderr: - Previously read value of stderr, in case stderr is already closed. - - :warn: - May deadlock if output or error pipes are used and not handled - separately. - - :raise git.exc.GitCommandError: - If the return status is not 0. - """ - if stderr is None: - stderr_b = b"" - stderr_b = force_bytes(data=stderr, encoding="utf-8") - status: Union[int, None] - if self.proc is not None: - status = self.proc.wait() - p_stderr = self.proc.stderr - else: # Assume the underlying proc was killed earlier or never existed. - status = self.status - p_stderr = None - - def read_all_from_possibly_closed_stream(stream: Union[IO[bytes], None]) -> bytes: - if stream: - try: - return stderr_b + force_bytes(stream.read()) - except (OSError, ValueError): - return stderr_b or b"" - else: - return stderr_b or b"" - - # END status handling - - if status != 0: - errstr = read_all_from_possibly_closed_stream(p_stderr) - _logger.debug("AutoInterrupt wait stderr: %r" % (errstr,)) - raise GitCommandError(remove_password_if_present(self.args), status, errstr) - return status - - # END auto interrupt - - class CatFileContentStream: - """Object representing a sized read-only stream returning the contents of - an object. - - This behaves like a stream, but counts the data read and simulates an empty - stream once our sized content region is empty. - - If not all data are read to the end of the object's lifetime, we read the - rest to ensure the underlying stream continues to work. - """ - - __slots__ = ("_stream", "_nbr", "_size") - - def __init__(self, size: int, stream: IO[bytes]) -> None: - self._stream = stream - self._size = size - self._nbr = 0 # Number of bytes read. - - # Special case: If the object is empty, has null bytes, get the final - # newline right away. - if size == 0: - stream.read(1) - # END handle empty streams - - def read(self, size: int = -1) -> bytes: - bytes_left = self._size - self._nbr - if bytes_left == 0: - return b"" - if size > -1: - # Ensure we don't try to read past our limit. - size = min(bytes_left, size) - else: - # They try to read all, make sure it's not more than what remains. - size = bytes_left - # END check early depletion - data = self._stream.read(size) - self._nbr += len(data) - - # Check for depletion, read our final byte to make the stream usable by - # others. - if self._size - self._nbr == 0: - self._stream.read(1) # final newline - # END finish reading - return data - - def readline(self, size: int = -1) -> bytes: - if self._nbr == self._size: - return b"" - - # Clamp size to lowest allowed value. - bytes_left = self._size - self._nbr - if size > -1: - size = min(bytes_left, size) - else: - size = bytes_left - # END handle size - - data = self._stream.readline(size) - self._nbr += len(data) - - # Handle final byte. - if self._size - self._nbr == 0: - self._stream.read(1) - # END finish reading - - return data - - def readlines(self, size: int = -1) -> List[bytes]: - if self._nbr == self._size: - return [] - - # Leave all additional logic to our readline method, we just check the size. - out = [] - nbr = 0 - while True: - line = self.readline() - if not line: - break - out.append(line) - if size > -1: - nbr += len(line) - if nbr > size: - break - # END handle size constraint - # END readline loop - return out - - # skipcq: PYL-E0301 - def __iter__(self) -> "Git.CatFileContentStream": - return self - - def __next__(self) -> bytes: - line = self.readline() - if not line: - raise StopIteration - - return line + @classmethod + def _option_candidates(cls, args: Sequence[Any] = (), kwargs: Optional[Mapping[str, Any]] = None) -> List[str]: + """Collect possible option spellings before command-line transformation.""" + options = [ + option for option in cls._unpack_args([arg for arg in args if arg is not None]) if option.startswith("-") + ] + if kwargs: + split_single_char_options = kwargs.get("split_single_char_options", True) + for key, value in kwargs.items(): + values = value if isinstance(value, (list, tuple)) else (value,) + if any(value is True or (value is not False and value is not None) for value in values): + key = str(key) + if len(key) != 1: + options.append(f"--{dashify(key)}") + elif split_single_char_options: + options.append(f"-{key}") + options.extend( + str(value) + for value in values + if value is not True and value not in (False, None) and str(value).startswith("-") + ) + else: + options.extend( + f"-{key}" if value is True else f"-{key}{value}" + for value in values + if value is True or (value is not False and value is not None) + ) + return options - next = __next__ + AutoInterrupt: TypeAlias = _AutoInterrupt - def __del__(self) -> None: - bytes_left = self._size - self._nbr - if bytes_left: - # Read and discard - seeking is impossible within a stream. - # This includes any terminating newline. - self._stream.read(bytes_left + 1) - # END handle incomplete read + CatFileContentStream: TypeAlias = _CatFileContentStream def __init__(self, working_dir: Union[None, PathLike] = None) -> None: """Initialize this instance with: @@ -971,7 +1101,7 @@ def __init__(self, working_dir: Union[None, PathLike] = None) -> None: def __getattribute__(self, name: str) -> Any: if name == "USE_SHELL": - _warn_use_shell(False) + _warn_use_shell(extra_danger=False) return super().__getattribute__(name) def __getattr__(self, name: str) -> Any: @@ -997,6 +1127,22 @@ def set_persistent_git_options(self, **kwargs: Any) -> None: self._persistent_git_options = self.transform_kwargs(split_single_char_options=True, **kwargs) + def ls_remote( + self, + *args: Any, + allow_unsafe_options: bool = False, + **kwargs: Any, + ) -> Union[str, bytes, Tuple[int, Union[str, bytes], str], "Git.AutoInterrupt"]: + """List references in a remote repository. + + :param allow_unsafe_options: + Allow unsafe options, like ``--upload-pack`` or ``--exec``. + """ + if not allow_unsafe_options: + candidate_options = self._option_candidates(args, kwargs) + Git.check_unsafe_options(options=candidate_options, unsafe_options=self.unsafe_git_ls_remote_options) + return self._call_process("ls_remote", *args, **kwargs) + @property def working_dir(self) -> Union[None, PathLike]: """:return: Git directory we are working on""" @@ -1098,9 +1244,28 @@ def execute( information (stdout). :param command: - The command argument list to execute. - It should be a sequence of program arguments, or a string. The - program to execute is the first item in the args sequence or string. + The command to execute. A sequence of program arguments is recommended. + A string is also accepted, but its meaning is strongly platform-dependent. + + By default, a shell is not used. On Unix-like systems, a string is the whole + program name (so ``"git log -n 1"`` raises :class:`GitCommandNotFound`). On + Windows, the program parses the arguments itself, so multi-word strings can + work but are not portable. + + Avoid ``shell=True`` (and :attr:`Git.USE_SHELL`): this runs the command in + a shell, which is generally unsafe. The shell interprets metacharacters + such as ``;``, ``|``, ``&``, ``$(...)``, ``$VAR``, ``%VAR%``, and ``^`` + (depending on the platform) as syntax. Any untrusted text in the command + can then execute arbitrary OS commands. See :attr:`Git.USE_SHELL`. + + Producing a sequence automatically by :func:`shlex.split` and passing it + as the command is far safer than ``shell=True``. But :func:`shlex.split` + parses POSIX shell syntax on all systems, and the result is still unsafe + for anything but *fixed, fully trusted* strings. Do not use it on strings + built by interpolating values: whitespace or quoting in an untrusted value + can still inject arguments. For input derived in any way from untrusted + data, build the argument sequence yourself, while ensuring each argument + is fully sanitized. :param istream: Standard input filehandle passed to :class:`subprocess.Popen`. @@ -1168,6 +1333,11 @@ def execute( needed (nor useful) to work around any known operating system specific issues. + On Unix-like systems, when migrating away from passing string commands with + ``shell=True``, :func:`shlex.split` may serve as a transitional step in rare + cases, with extreme care. (Drop ``shell=True`` and pass the resulting + sequence as the command.) See the `command` parameter above on the risks. + :param env: A dictionary of environment variables to be passed to :class:`subprocess.Popen`. @@ -1212,7 +1382,7 @@ def execute( # Allow the user to have the command executed in their working dir. try: - cwd = self._working_dir or os.getcwd() # type: Union[None, str] + cwd = self._working_dir or os.getcwd() # type: Optional[PathLike] if not os.access(str(cwd), os.X_OK): cwd = None except FileNotFoundError: @@ -1282,6 +1452,10 @@ def execute( if as_process: return self.AutoInterrupt(proc, command) + watchdog: Optional[threading.Timer] = None + kill_check: Optional[threading.Event] = None + timeout_error: Optional[Callable[[], Union[str, bytes]]] = None + if sys.platform != "win32" and kill_after_timeout is not None: # Help mypy figure out this is not None even when used inside communicate(). timeout = kill_after_timeout @@ -1307,6 +1481,7 @@ def kill_process(pid: int) -> None: except OSError: pass # Tell the main routine that the process was killed. + assert kill_check is not None kill_check.set() except OSError: # It is possible that the process gets completed in the duration @@ -1314,23 +1489,25 @@ def kill_process(pid: int) -> None: pass return + def make_timeout_error() -> Union[str, bytes]: + err = f'Timeout: the command "{" ".join(redacted_command)}" did not complete in {timeout:g} secs.' + return err if universal_newlines else err.encode(defenc) + def communicate() -> Tuple[AnyStr, AnyStr]: + assert watchdog is not None + assert kill_check is not None watchdog.start() out, err = proc.communicate() watchdog.cancel() if kill_check.is_set(): - err = 'Timeout: the command "%s" did not complete in %d ' "secs." % ( - " ".join(redacted_command), - timeout, - ) - if not universal_newlines: - err = err.encode(defenc) + err = make_timeout_error() return out, err # END helpers kill_check = threading.Event() watchdog = threading.Timer(timeout, kill_process, args=(proc.pid,)) + timeout_error = make_timeout_error else: communicate = proc.communicate @@ -1343,25 +1520,38 @@ def communicate() -> Tuple[AnyStr, AnyStr]: if output_stream is None: stdout_value, stderr_value = communicate() # Strip trailing "\n". - if stdout_value.endswith(newline) and strip_newline_in_stdout: # type: ignore[arg-type] + if stdout_value is not None and stdout_value.endswith(newline) and strip_newline_in_stdout: # type: ignore[arg-type] stdout_value = stdout_value[:-1] - if stderr_value.endswith(newline): # type: ignore[arg-type] + if stderr_value is not None and stderr_value.endswith(newline): # type: ignore[arg-type] stderr_value = stderr_value[:-1] status = proc.returncode else: max_chunk_size = max_chunk_size if max_chunk_size and max_chunk_size > 0 else io.DEFAULT_BUFFER_SIZE - stream_copy(proc.stdout, output_stream, max_chunk_size) - stdout_value = proc.stdout.read() - stderr_value = proc.stderr.read() + if watchdog is not None: + watchdog.start() + try: + if proc.stdout is not None: + stream_copy(proc.stdout, output_stream, max_chunk_size) + stdout_value = proc.stdout.read() + if proc.stderr is not None: + stderr_value = proc.stderr.read() + status = proc.wait() + finally: + if watchdog is not None: + watchdog.cancel() # Strip trailing "\n". - if stderr_value.endswith(newline): # type: ignore[arg-type] + if stderr_value is not None and stderr_value.endswith(newline): # type: ignore[arg-type] stderr_value = stderr_value[:-1] - status = proc.wait() + if kill_check is not None and kill_check.is_set(): + assert timeout_error is not None + stderr_value = timeout_error() # END stdout handling finally: - proc.stdout.close() - proc.stderr.close() + if proc.stdout is not None: + proc.stdout.close() + if proc.stderr is not None: + proc.stderr.close() if self.GIT_PYTHON_TRACE == "full": cmdstr = " ".join(redacted_command) @@ -1475,7 +1665,7 @@ def transform_kwargs(self, split_single_char_options: bool = True, **kwargs: Any return args @classmethod - def _unpack_args(cls, arg_list: Sequence[str]) -> List[str]: + def _unpack_args(cls, arg_list: Sequence[Any]) -> List[str]: outlist = [] if isinstance(arg_list, (list, tuple)): for arg in arg_list: @@ -1551,7 +1741,7 @@ def _call_process( turns into:: - git rev-list max-count 10 --header master + git rev-list --max-count=10 --header master :return: Same as :meth:`execute`. If no args are given, used :meth:`execute`'s diff --git a/git/config.py b/git/config.py index de3508360..c300de499 100644 --- a/git/config.py +++ b/git/config.py @@ -66,12 +66,18 @@ CONFIG_LEVELS: ConfigLevels_Tup = ("system", "user", "global", "repository") """The configuration level of a configuration file.""" -CONDITIONAL_INCLUDE_REGEXP = re.compile(r"(?<=includeIf )\"(gitdir|gitdir/i|onbranch):(.+)\"") +CONDITIONAL_INCLUDE_REGEXP = re.compile(r"(?<=includeIf )\"(gitdir|gitdir/i|onbranch|hasconfig:remote\.\*\.url):(.+)\"") """Section pattern to detect conditional includes. See: https://git-scm.com/docs/git-config#_conditional_includes """ +UNSAFE_CONFIG_CHARS_RE = re.compile(r"[\r\n\x00]") +"""Characters that cannot be safely written in config names or values.""" + +VALID_CONFIG_OPTION_NAME_RE = re.compile(r"^[A-Za-z0-9_.-]+$") +"""Pattern for option names that can be written without changing config syntax.""" + class MetaParserBuilder(abc.ABCMeta): # noqa: B024 """Utility class wrapping base-class methods into decorators that assure read-only @@ -87,15 +93,15 @@ def __new__(cls, name: str, bases: Tuple, clsdict: Dict[str, Any]) -> "MetaParse mutating_methods = clsdict[kmm] for base in bases: methods = (t for t in inspect.getmembers(base, inspect.isroutine) if not t[0].startswith("_")) - for name, method in methods: - if name in clsdict: + for method_name, method in methods: + if method_name in clsdict: continue method_with_values = needs_values(method) - if name in mutating_methods: + if method_name in mutating_methods: method_with_values = set_dirty_and_flush_changes(method_with_values) # END mutating methods handling - clsdict[name] = method_with_values + clsdict[method_name] = method_with_values # END for each name/method pair # END for each base # END if mutating methods configuration is set @@ -496,19 +502,26 @@ def string_decode(v: str) -> str: if mo: # We might just have handled the last line, which could contain a quotation we want to remove. optname, vi, optval = mo.group("option", "vi", "value") + optname = self.optionxform(optname.rstrip()) + if vi in ("=", ":") and ";" in optval and not optval.strip().startswith('"'): pos = optval.find(";") if pos != -1 and optval[pos - 1].isspace(): optval = optval[:pos] optval = optval.strip() - if optval == '""': - optval = "" - # END handle empty string - optname = self.optionxform(optname.rstrip()) - if len(optval) > 1 and optval[0] == '"' and optval[-1] != '"': + + if len(optval) < 2 or optval[0] != '"': + # Does not open quoting. + pass + elif optval[-1] != '"': + # Opens quoting and does not close: appears to start multi-line quoting. is_multi_line = True optval = string_decode(optval[1:]) - # END handle multi-line + elif optval.find("\\", 1, -1) == -1 and optval.find('"', 1, -1) == -1: + # Opens and closes quoting. Single line, and all we need is quote removal. + optval = optval[1:-1] + # TODO: Handle other quoted content, especially well-formed backslash escapes. + # Preserves multiple values for duplicate optnames. cursect.add(optname, optval) else: @@ -542,11 +555,21 @@ def _included_paths(self) -> List[Tuple[str, str]]: :return: The list of paths, where each path is a tuple of (option, value). """ + + def _all_items(section: str) -> List[Tuple[str, str]]: + """Return all (key, value) pairs for a section, including duplicate keys.""" + return [ + (key, value) + for key, values in self._sections[section].items_all() + if key != "__name__" + for value in values + ] + paths = [] for section in self.sections(): if section == "include": - paths += self.items(section) + paths += _all_items(section) match = CONDITIONAL_INCLUDE_REGEXP.search(section) if match is None or self._repo is None: @@ -557,8 +580,12 @@ def _included_paths(self) -> List[Tuple[str, str]]: if keyword in ["gitdir", "gitdir/i"]: value = osp.expanduser(value) + git_dir = os.fspath(self._repo.git_dir) if self._repo.git_dir else None + if sys.platform == "win32": + git_dir = git_dir.replace("\\", "/") if git_dir else None - if not any(value.startswith(s) for s in ["./", "/"]): + drive, _tail = osp.splitdrive(value) + if not drive and not any(value.startswith(s) for s in ["./", "/"]): value = "**/" + value if value.endswith("/"): value += "**" @@ -567,12 +594,11 @@ def _included_paths(self) -> List[Tuple[str, str]]: if keyword.endswith("/i"): value = re.sub( r"[a-zA-Z]", - lambda m: "[{}{}]".format(m.group().lower(), m.group().upper()), + lambda m: f"[{m.group().lower()!r}{m.group().upper()!r}]", value, ) - if self._repo.git_dir: - if fnmatch.fnmatchcase(str(self._repo.git_dir), value): - paths += self.items(section) + if git_dir and fnmatch.fnmatchcase(git_dir, value): + paths += _all_items(section) elif keyword == "onbranch": try: @@ -582,8 +608,12 @@ def _included_paths(self) -> List[Tuple[str, str]]: continue if fnmatch.fnmatchcase(branch_name, value): - paths += self.items(section) - + paths += _all_items(section) + elif keyword == "hasconfig:remote.*.url": + for remote in self._repo.remotes: + if fnmatch.fnmatchcase(remote.url, value): + paths += _all_items(section) + break return paths def read(self) -> None: # type: ignore[override] @@ -610,6 +640,8 @@ def read(self) -> None: # type: ignore[override] files_to_read = list(self._file_or_files) # END ensure we have a copy of the paths to handle + files_to_read = [osp.abspath(path) if isinstance(path, (str, os.PathLike)) else path for path in files_to_read] + seen = set(files_to_read) num_read_include_files = 0 while files_to_read: @@ -622,8 +654,6 @@ def read(self) -> None: # type: ignore[override] file_path = cast(IO[bytes], file_path) self._read(file_path, file_path.name) else: - # Assume a path if it is not a file-object. - file_path = cast(PathLike, file_path) try: with open(file_path, "rb") as fp: file_ok = True @@ -757,8 +787,9 @@ def _assure_writable(self, method_name: str) -> None: if self.read_only: raise IOError("Cannot execute non-constant method %s.%s" % (self, method_name)) - def add_section(self, section: str) -> None: + def add_section(self, section: "cp._SectionName") -> None: """Assures added options will stay in order.""" + self._assure_config_name_safe(section, "section") return super().add_section(section) @property @@ -863,6 +894,48 @@ def _value_to_string(self, value: Union[str, bytes, int, float, bool]) -> str: return str(value) return force_text(value) + def _value_to_string_safe(self, value: Union[str, bytes, int, float, bool]) -> str: + value_str = self._value_to_string(value) + if UNSAFE_CONFIG_CHARS_RE.search(value_str): + raise ValueError("Git config values must not contain CR, LF, or NUL") + return value_str + + def _assure_config_name_safe(self, name: "cp._SectionName", label: str) -> None: + if isinstance(name, str) and UNSAFE_CONFIG_CHARS_RE.search(name): + raise ValueError("Git config %s names must not contain CR, LF, or NUL" % label) + if label == "option" and isinstance(name, str) and not VALID_CONFIG_OPTION_NAME_RE.fullmatch(name): + raise ValueError("Git config option names may contain only letters, digits, '-', '_', or '.'") + if label == "section" and isinstance(name, str): + in_quotes = False + escaped = False + for index, char in enumerate(name): + if escaped: + escaped = False + elif in_quotes and char == "\\": + escaped = True + elif char == '"': + if not in_quotes and (index == 0 or name[index - 1] not in " \t"): + raise ValueError("Git config quoted subsection names must begin after whitespace") + in_quotes = not in_quotes + elif char == "]" and not in_quotes: + raise ValueError("Git config section names must not contain an unquoted closing bracket") + if in_quotes: + raise ValueError("Git config section names must not contain an unterminated quote") + + @needs_values + @set_dirty_and_flush_changes + def set( + self, + section: str, + option: str, + value: Union[str, bytes, int, float, bool, None] = None, + ) -> None: + self._assure_config_name_safe(section, "section") + self._assure_config_name_safe(option, "option") + if value is not None: + value = self._value_to_string_safe(value) + return super().set(section, option, value) + @needs_values @set_dirty_and_flush_changes def set_value(self, section: str, option: str, value: Union[str, bytes, int, float, bool]) -> "GitConfigParser": @@ -883,9 +956,12 @@ def set_value(self, section: str, option: str, value: Union[str, bytes, int, flo :return: This instance """ + self._assure_config_name_safe(section, "section") + self._assure_config_name_safe(option, "option") + value_str = self._value_to_string_safe(value) if not self.has_section(section): self.add_section(section) - self.set(section, option, self._value_to_string(value)) + super().set(section, option, value_str) return self @needs_values @@ -910,9 +986,12 @@ def add_value(self, section: str, option: str, value: Union[str, bytes, int, flo :return: This instance """ + self._assure_config_name_safe(section, "section") + self._assure_config_name_safe(option, "option") + value_str = self._value_to_string_safe(value) if not self.has_section(section): self.add_section(section) - self._sections[section].add(option, self._value_to_string(value)) + self._sections[section].add(option, value_str) return self def rename_section(self, section: str, new_name: str) -> "GitConfigParser": @@ -929,6 +1008,7 @@ def rename_section(self, section: str, new_name: str) -> "GitConfigParser": """ if not self.has_section(section): raise ValueError("Source section '%s' doesn't exist" % section) + self._assure_config_name_safe(new_name, "section") if self.has_section(new_name): raise ValueError("Destination section '%s' already exists" % new_name) diff --git a/git/diff.py b/git/diff.py index 9c6ae59e0..d1963b84f 100644 --- a/git/diff.py +++ b/git/diff.py @@ -3,13 +3,13 @@ # This module is part of GitPython and is released under the # 3-Clause BSD License: https://opensource.org/license/bsd-3-clause/ -__all__ = ["DiffConstants", "NULL_TREE", "INDEX", "Diffable", "DiffIndex", "Diff"] +__all__ = ["DiffConstants", "NULL_TREE", "NULL_TREE_SHA", "INDEX", "Diffable", "DiffIndex", "Diff"] import enum import re import warnings -from git.cmd import handle_process_output +from git.cmd import Git, handle_process_output from git.compat import defenc from git.objects.blob import Blob from git.objects.util import mode_str_to_int @@ -23,18 +23,18 @@ List, Match, Optional, + Sequence, Tuple, TYPE_CHECKING, TypeVar, Union, cast, ) -from git.types import Literal, PathLike +from git.types import PathLike, Literal if TYPE_CHECKING: from subprocess import Popen - from git.cmd import Git from git.objects.base import IndexObject from git.objects.commit import Commit from git.objects.tree import Tree @@ -83,6 +83,9 @@ class DiffConstants(enum.Enum): :const:`git.NULL_TREE` and :const:`Diffable.NULL_TREE`. """ +NULL_TREE_SHA = "4b825dc642cb6eb9a060e54bf8d69288fbee4904" +"""SHA of Git's canonical empty tree object.""" + INDEX: Literal[DiffConstants.INDEX] = DiffConstants.INDEX """Stand-in indicating you want to diff against the index. @@ -186,6 +189,7 @@ def diff( other: Union[DiffConstants, "Tree", "Commit", str, None] = INDEX, paths: Union[PathLike, List[PathLike], Tuple[PathLike, ...], None] = None, create_patch: bool = False, + allow_unsafe_options: bool = False, **kwargs: Any, ) -> "DiffIndex[Diff]": """Create diffs between two items being trees, trees and index or an index and @@ -215,6 +219,10 @@ def diff( applied makes the self to other. Patches are somewhat costly as blobs have to be read and diffed. + :param allow_unsafe_options: + If ``True``, allow options such as ``--output`` and ``-O`` that can write to + or read from arbitrary filesystem paths. + :param kwargs: Additional arguments passed to :manpage:`git-diff(1)`, such as ``R=True`` to swap both sides of the diff. @@ -227,6 +235,13 @@ def diff( an instance of :class:`~git.objects.tree.Tree` or :class:`~git.objects.commit.Commit`, or a git command error will occur. """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([other], kwargs), + unsafe_options=self.repo.unsafe_git_diff_options, + clusterable_short_options="46abceflmnpqrstuvwzBCDMNRW", + ) + args: List[Union[PathLike, Diffable]] = [] args.append("--abbrev=40") # We need full shas. args.append("--full-index") # Get full index paths, not only filenames. @@ -270,7 +285,10 @@ def diff( # END paths handling kwargs["as_process"] = True - proc = diff_cmd(*self._process_diff_args(args), **kwargs) + args = self._process_diff_args(args) + if create_patch: + self.repo.git(c="diff.mnemonicPrefix=false") + proc = diff_cmd(*args, **kwargs) diff_method = Diff._index_from_patch_format if create_patch else Diff._index_from_raw_format index = diff_method(self.repo, proc) @@ -289,7 +307,7 @@ class DiffIndex(List[T_Diff]): The class improves the diff handling convenience. """ - change_type = ("A", "C", "D", "R", "M", "T") + change_type: Sequence[Literal["A", "C", "D", "R", "M", "T"]] = ("A", "C", "D", "R", "M", "T") # noqa: F821 """Change type invariant identifying possible ways a blob can have changed: * ``A`` = Added @@ -598,7 +616,14 @@ def _index_from_patch_format(cls, repo: "Repo", proc: Union["Popen", "Git.AutoIn # FIXME: Here SLURPING raw, need to re-phrase header-regexes linewise. text_list: List[bytes] = [] - handle_process_output(proc, text_list.append, None, finalize_process, decode_streams=False) + stderr_list: List[bytes] = [] + + def finalize_process_with_stderr(proc: Union["Popen", "Git.AutoInterrupt"]) -> None: + finalize_process(proc, stderr=b"".join(stderr_list)) + + handle_process_output( + proc, text_list.append, stderr_list.append, finalize_process_with_stderr, decode_streams=False + ) # For now, we have to bake the stream. text = b"".join(text_list) @@ -764,11 +789,16 @@ def _index_from_raw_format(cls, repo: "Repo", proc: "Popen") -> "DiffIndex[Diff] # :100644 100644 687099101... 37c5e30c8... M .gitignore index: "DiffIndex" = DiffIndex() + stderr_list: List[bytes] = [] + + def finalize_process_with_stderr(proc: Union["Popen", "Git.AutoInterrupt"]) -> None: + finalize_process(proc, stderr=b"".join(stderr_list)) + handle_process_output( proc, lambda byt: cls._handle_diff_line(byt, repo, index), - None, - finalize_process, + stderr_list.append, + finalize_process_with_stderr, decode_streams=False, ) diff --git a/git/ext/gitdb b/git/ext/gitdb deleted file mode 160000 index f36c0cc42..000000000 --- a/git/ext/gitdb +++ /dev/null @@ -1 +0,0 @@ -Subproject commit f36c0cc42ea2f529291e441073f74e920988d4d2 diff --git a/git/index/base.py b/git/index/base.py index 39cc9143c..a3c915242 100644 --- a/git/index/base.py +++ b/git/index/base.py @@ -23,6 +23,7 @@ from gitdb.db import MemoryDB from git.compat import defenc, force_bytes +from git.cmd import Git import git.diff as git_diff from git.exc import CheckoutError, GitCommandError, GitError, InvalidGitRepositoryError from git.objects import Blob, Commit, Object, Submodule, Tree @@ -33,6 +34,8 @@ LockedFD, join_path_native, file_contents_ro, + _is_path_rooted, + _to_relative_path, to_native_path_linux, unbare_repo, to_bin_sha, @@ -57,6 +60,7 @@ Any, BinaryIO, Callable, + cast, Dict, Generator, IO, @@ -129,6 +133,9 @@ class IndexFile(LazyMixin, git_diff.Diffable, Serializable): index directly before operating on it using the git command. """ + unsafe_git_checkout_index_options = ["--prefix"] + unsafe_git_read_tree_options = ["--index-output"] + __slots__ = ("repo", "version", "entries", "_extension_data", "_file_path") _VERSION = 2 @@ -253,7 +260,12 @@ def write( @post_clear_cache @default_index - def merge_tree(self, rhs: Treeish, base: Union[None, Treeish] = None) -> "IndexFile": + def merge_tree( + self, + rhs: Treeish, + base: Union[None, Treeish] = None, + allow_unsafe_options: bool = False, + ) -> "IndexFile": """Merge the given `rhs` treeish into the current index, possibly taking a common base treeish into account. @@ -267,6 +279,9 @@ def merge_tree(self, rhs: Treeish, base: Union[None, Treeish] = None) -> "IndexF Optional treeish reference pointing to the common base of `rhs` and this index which equals lhs. + :param allow_unsafe_options: + Allow options that may write to arbitrary paths. + :return: self (containing the merge and possibly unmerged entries in case of conflicts) @@ -277,6 +292,12 @@ def merge_tree(self, rhs: Treeish, base: Union[None, Treeish] = None) -> "IndexF yourself, you have to commit the changed index (or make a valid tree from it) and retry with a three-way :meth:`index.from_tree ` call. """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([base, rhs]), + unsafe_options=self.unsafe_git_read_tree_options, + ) + # -i : ignore working tree status # --aggressive : handle more merge cases # -m : do an actual merge @@ -321,7 +342,13 @@ def new(cls, repo: "Repo", *tree_sha: Union[str, Tree]) -> "IndexFile": return inst @classmethod - def from_tree(cls, repo: "Repo", *treeish: Treeish, **kwargs: Any) -> "IndexFile": + def from_tree( + cls, + repo: "Repo", + *treeish: Treeish, + allow_unsafe_options: bool = False, + **kwargs: Any, + ) -> "IndexFile": R"""Merge the given treeish revisions into a new index which is returned. The original index will remain unaltered. @@ -345,6 +372,9 @@ def from_tree(cls, repo: "Repo", *treeish: Treeish, **kwargs: Any) -> "IndexFile :param kwargs: Additional arguments passed to :manpage:`git-read-tree(1)`. + :param allow_unsafe_options: + Allow options that may write to arbitrary paths. + :return: New :class:`IndexFile` instance. It will point to a temporary index location which does not exist anymore. If you intend to write such a merged Index, @@ -362,6 +392,12 @@ def from_tree(cls, repo: "Repo", *treeish: Treeish, **kwargs: Any) -> "IndexFile if len(treeish) == 0 or len(treeish) > 3: raise ValueError("Please specify between 1 and 3 treeish, got %i" % len(treeish)) + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates(treeish, kwargs), + unsafe_options=cls.unsafe_git_read_tree_options, + ) + arg_list: List[Union[Treeish, str]] = [] # Ignore that the working tree and index possibly are out of date. if len(treeish) > 1: @@ -407,7 +443,7 @@ def raise_exc(e: Exception) -> NoReturn: r = str(self.repo.working_tree_dir) rs = r + os.sep for path in paths: - abs_path = str(path) + abs_path = os.fspath(path) if not osp.isabs(abs_path): abs_path = osp.join(r, path) # END make absolute path @@ -508,7 +544,7 @@ def iter_blobs( :param predicate: Function(t) returning ``True`` if tuple(stage, Blob) should be yielded by - the iterator. A default filter, the `~git.index.typ.BlobFilter`, allows you + the iterator. A default filter, the :class:`~git.index.typ.BlobFilter`, allows you to yield blobs only if they match a given list of paths. """ for entry in self.entries.values(): @@ -530,7 +566,10 @@ def unmerged_blobs(self) -> Dict[PathLike, List[Tuple[StageType, Blob]]]: stage. That is, a file removed on the 'other' branch whose entries are at stage 3 will not have a stage 3 entry. """ - is_unmerged_blob = lambda t: t[0] != 0 + + def is_unmerged_blob(t: Tuple[StageType, Blob]) -> bool: + return t[0] != 0 + path_map: Dict[PathLike, List[Tuple[StageType, Blob]]] = {} for stage, blob in self.iter_blobs(is_unmerged_blob): path_map.setdefault(blob.path, []).append((stage, blob)) @@ -649,13 +688,12 @@ def _to_relative_path(self, path: PathLike) -> PathLike: :raise ValueError: """ - if not osp.isabs(path): - return path if self.repo.bare: - raise InvalidGitRepositoryError("require non-bare repository") - if not osp.normpath(str(path)).startswith(str(self.repo.working_tree_dir)): - raise ValueError("Absolute path %r is not in git repository at %r" % (path, self.repo.working_tree_dir)) - return os.path.relpath(path, self.repo.working_tree_dir) + drive, _tail = osp.splitdrive(os.fspath(path)) + if drive or _is_path_rooted(path): + raise InvalidGitRepositoryError("paths with a drive or root require a non-bare repository") + return path + return _to_relative_path(cast(PathLike, self.repo.working_tree_dir), path) def _preprocess_add_items( self, items: Union[PathLike, Sequence[Union[PathLike, Blob, BaseIndexEntry, "Submodule"]]] @@ -687,12 +725,17 @@ def _store_path(self, filepath: PathLike, fprogress: Callable) -> BaseIndexEntry This must be ensured in the calling code. """ st = os.lstat(filepath) # Handles non-symlinks as well. + if S_ISLNK(st.st_mode): # In PY3, readlink is a string, but we need bytes. # In PY2, it was just OS encoded bytes, we assumed UTF-8. - open_stream: Callable[[], BinaryIO] = lambda: BytesIO(force_bytes(os.readlink(filepath), encoding=defenc)) + def open_stream() -> BinaryIO: + return BytesIO(force_bytes(os.readlink(filepath), encoding=defenc)) else: - open_stream = lambda: open(filepath, "rb") + + def open_stream() -> BinaryIO: + return open(filepath, "rb") + with open_stream() as stream: fprogress(filepath, False, filepath) istream = self.repo.odb.store(IStream(Blob.type, st.st_size, stream)) @@ -717,14 +760,17 @@ def _entries_for_paths( ) -> List[BaseIndexEntry]: entries_added: List[BaseIndexEntry] = [] if path_rewriter: + working_tree_dir = self.repo.working_tree_dir + if working_tree_dir is None: + raise InvalidGitRepositoryError("Cannot rewrite paths without a working tree") + working_tree_dir = str(working_tree_dir) for path in paths: if osp.isabs(path): abspath = path - gitrelative_path = path[len(str(self.repo.working_tree_dir)) + 1 :] + gitrelative_path = path[len(working_tree_dir) + 1 :] else: gitrelative_path = path - if self.repo.working_tree_dir: - abspath = osp.join(self.repo.working_tree_dir, gitrelative_path) + abspath = osp.join(working_tree_dir, gitrelative_path) # END obtain relative and absolute paths blob = Blob( @@ -767,7 +813,7 @@ def add( - path string Strings denote a relative or absolute path into the repository pointing - to an existing file, e.g., ``CHANGES``, `lib/myfile.ext``, + to an existing file, e.g., ``CHANGES``, ``lib/myfile.ext``, ``/home/gitrepo/lib/myfile.ext``. Absolute paths must start with working tree directory of this index's @@ -786,7 +832,7 @@ def add( They are added at stage 0. - - :class:~`git.objects.blob.Blob` or + - :class:`~git.objects.blob.Blob` or :class:`~git.objects.submodule.base.Submodule` object Blobs are added as they are assuming a valid mode is set. @@ -812,7 +858,7 @@ def add( - :class:`~git.index.typ.BaseIndexEntry` or type - Handling equals the one of :class:~`git.objects.blob.Blob` objects, but + Handling equals the one of :class:`~git.objects.blob.Blob` objects, but the stage may be explicitly set. Please note that Index Entries require binary sha's. @@ -978,6 +1024,7 @@ def remove( self, items: Union[PathLike, Sequence[Union[PathLike, Blob, BaseIndexEntry, "Submodule"]]], working_tree: bool = False, + allow_unsafe_options: bool = False, **kwargs: Any, ) -> List[str]: R"""Remove the given items from the index and optionally from the working tree @@ -995,7 +1042,7 @@ def remove( The path string may include globs, such as ``*.c``. - - :class:~`git.objects.blob.Blob` object + - :class:`~git.objects.blob.Blob` object Only the path portion is used in this case. @@ -1008,6 +1055,10 @@ def remove( physically removing the respective file. This may fail if there are uncommitted changes in it. + :param allow_unsafe_options: + Allow unsafe options such as ``--pathspec-from-file`` to be passed to + :manpage:`git-rm(1)`. + :param kwargs: Additional keyword arguments to be passed to :manpage:`git-rm(1)`, such as ``r`` to allow recursive removal. @@ -1019,13 +1070,18 @@ def remove( This is interesting to know in case you have provided a directory or globs. Paths are relative to the repository. """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), + unsafe_options=Git.unsafe_git_pathspec_from_file_options, + ) args = [] if not working_tree: args.append("--cached") args.append("--") # Preprocess paths. - paths = self._items_to_rela_paths(items) + paths = list(map(os.fspath, self._items_to_rela_paths(items))) # type: ignore[arg-type] removed_paths = self.repo.git.rm(args, paths, **kwargs).splitlines() # Process output to gain proper paths. @@ -1122,6 +1178,7 @@ def commit( author_date: Union[datetime.datetime, str, None] = None, commit_date: Union[datetime.datetime, str, None] = None, skip_hooks: bool = False, + trailers: Union[None, "Dict[str, str]", "List[Tuple[str, str]]"] = None, ) -> Commit: """Commit the current default index file, creating a :class:`~git.objects.commit.Commit` object. @@ -1158,6 +1215,7 @@ def commit( committer=committer, author_date=author_date, commit_date=commit_date, + trailers=trailers, ) if not skip_hooks: run_commit_hook("post-commit", self) @@ -1198,6 +1256,7 @@ def checkout( paths: Union[None, Iterable[PathLike]] = None, force: bool = False, fprogress: Callable = lambda *args: None, + allow_unsafe_options: bool = False, **kwargs: Any, ) -> Union[None, Iterator[PathLike], Sequence[PathLike]]: """Check out the given paths or all files from the version known to the index @@ -1224,6 +1283,9 @@ def checkout( no explicit paths are given. Otherwise progress information will be send prior and after a file has been checked out. + :param allow_unsafe_options: + Allow unsafe options, such as ``--prefix``. + :param kwargs: Additional arguments to be passed to :manpage:`git-checkout-index(1)`. @@ -1247,6 +1309,12 @@ def checkout( i.e. if you want :manpage:`git-checkout(1)`-like behaviour, use ``head.checkout`` instead of ``index.checkout``. """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), + unsafe_options=self.unsafe_git_checkout_index_options, + ) + args = ["--index"] if force: args.append("--force") @@ -1333,8 +1401,11 @@ def handle_stderr(proc: "Popen[bytes]", iter_checked_out_files: Iterable[PathLik kwargs["as_process"] = True kwargs["istream"] = subprocess.PIPE proc = self.repo.git.checkout_index(args, **kwargs) + # FIXME: Reading from GIL! - make_exc = lambda: GitCommandError(("git-checkout-index",) + tuple(args), 128, proc.stderr.read()) + def make_exc() -> GitCommandError: + return GitCommandError(("git-checkout-index", *args), 128, proc.stderr.read()) + checked_out_files: List[PathLike] = [] for path in paths: @@ -1345,11 +1416,11 @@ def handle_stderr(proc: "Popen[bytes]", iter_checked_out_files: Iterable[PathLik try: self.entries[(co_path, 0)] except KeyError: - folder = str(co_path) + folder = co_path if not folder.endswith("/"): folder += "/" for entry in self.entries.values(): - if str(entry.path).startswith(folder): + if os.fspath(entry.path).startswith(folder): p = entry.path self._write_path_to_stdin(proc, p, p, make_exc, fprogress, read_from_stdout=False) checked_out_files.append(p) @@ -1385,6 +1456,7 @@ def reset( working_tree: bool = False, paths: Union[None, Iterable[PathLike]] = None, head: bool = False, + allow_unsafe_options: bool = False, **kwargs: Any, ) -> "IndexFile": """Reset the index to reflect the tree at the given commit. This will not adjust @@ -1416,6 +1488,9 @@ def reset( The paths need to exist at the commit, otherwise an exception will be raised. + :param allow_unsafe_options: + Allow options that may write to arbitrary paths. + :param kwargs: Additional keyword arguments passed to :manpage:`git-reset(1)`. @@ -1432,15 +1507,15 @@ def reset( """ # What we actually want to do is to merge the tree into our existing index, # which is what git-read-tree does. - new_inst = type(self).from_tree(self.repo, commit) + new_inst = type(self).from_tree(self.repo, commit, allow_unsafe_options=allow_unsafe_options) if not paths: self.entries = new_inst.entries else: nie = new_inst.entries for path in paths: path = self._to_relative_path(path) + key = entry_key(path, 0) try: - key = entry_key(path, 0) self.entries[key] = nie[key] except KeyError: # If key is not in theirs, it mustn't be in ours. @@ -1464,12 +1539,11 @@ def reset( return self - # FIXME: This is documented to accept the same parameters as Diffable.diff, but this - # does not handle NULL_TREE for `other`. (The suppressed mypy error is about this.) def diff( self, - other: Union[ # type: ignore[override] + other: Union[ Literal[git_diff.DiffConstants.INDEX], + Literal[git_diff.DiffConstants.NULL_TREE], "Tree", "Commit", str, @@ -1477,6 +1551,7 @@ def diff( ] = git_diff.INDEX, paths: Union[PathLike, List[PathLike], Tuple[PathLike, ...], None] = None, create_patch: bool = False, + allow_unsafe_options: bool = False, **kwargs: Any, ) -> git_diff.DiffIndex[git_diff.Diff]: """Diff this index against the working copy or a :class:`~git.objects.tree.Tree` @@ -1489,6 +1564,13 @@ def diff( Will only work with indices that represent the default git index as they have not been initialized with a stream. """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([other], kwargs), + unsafe_options=self.repo.unsafe_git_diff_options, + clusterable_short_options="46abceflmnpqrstuvwzBCDMNRW", + ) + # Only run if we are the default repository index. if self._file_path != self._index_path(): raise AssertionError("Cannot call %r on indices that do not represent the default git index" % self.diff()) @@ -1496,6 +1578,46 @@ def diff( if other is self.INDEX: return git_diff.DiffIndex() + if other == git_diff.NULL_TREE or other == git_diff.NULL_TREE_SHA: + args: List[Union[PathLike, str]] = [ + "--cached", + git_diff.NULL_TREE_SHA, + "--abbrev=40", + "--full-index", + ] + + if not any(x in kwargs for x in ("find_renames", "no_renames", "M")): + args.append("-M") + + if create_patch: + args.append("-p") + args.append("--no-ext-diff") + else: + args.append("--raw") + args.append("-z") + + args.append("--no-color") + + if paths is not None and not isinstance(paths, (tuple, list)): + paths = [paths] + + if paths: + args.append("--") + args.extend(paths) + + kwargs["as_process"] = True + if create_patch: + self.repo.git(c="diff.mnemonicPrefix=false") + proc = self.repo.git.diff(*args, **kwargs) + + diff_method = ( + git_diff.Diff._index_from_patch_format if create_patch else git_diff.Diff._index_from_raw_format + ) + index = diff_method(self.repo, proc) + + proc.wait() + return index + # Index against anything but None is a reverse diff with the respective item. # Handle existing -R flags properly. # Transform strings to the object so that we can call diff on it. @@ -1507,7 +1629,13 @@ def diff( # Invert the existing R flag. cur_val = kwargs.get("R", False) kwargs["R"] = not cur_val - return other.diff(self.INDEX, paths, create_patch, **kwargs) + return other.diff( + self.INDEX, + paths, + create_patch, + allow_unsafe_options=allow_unsafe_options, + **kwargs, + ) # END diff against other item handling # If other is not None here, something is wrong. @@ -1515,4 +1643,4 @@ def diff( raise ValueError("other must be None, Diffable.INDEX, a Tree or Commit, was %r" % other) # Diff against working copy - can be handled by superclass natively. - return super().diff(other, paths, create_patch, **kwargs) + return super().diff(other, paths, create_patch, allow_unsafe_options=allow_unsafe_options, **kwargs) diff --git a/git/index/fun.py b/git/index/fun.py index 59cce6ae6..886e10de9 100644 --- a/git/index/fun.py +++ b/git/index/fun.py @@ -26,7 +26,7 @@ from gitdb.base import IStream from gitdb.typ import str_tree_type -from git.cmd import handle_process_output, safer_popen +from git.cmd import Git, handle_process_output, safer_popen from git.compat import defenc, force_bytes, force_text, safe_decode from git.exc import HookExecutionError, UnmergedEntriesError from git.objects.fun import ( @@ -36,7 +36,7 @@ ) from git.util import IndexFileSHA1Writer, finalize_process -from .typ import BaseIndexEntry, IndexEntry, CE_NAMEMASK, CE_STAGESHIFT +from .typ import CE_EXTENDED, BaseIndexEntry, IndexEntry, CE_NAMEMASK, CE_STAGESHIFT from .util import pack, unpack # typing ----------------------------------------------------------------------------- @@ -64,10 +64,114 @@ def hook_path(name: str, git_dir: PathLike) -> str: return osp.join(git_dir, "hooks", name) +def _commit_hook_path(name: str, index: "IndexFile") -> str: + """:return: path to the named commit hook, respecting Git's core.hooksPath.""" + with index.repo.config_reader() as config: + hooks_dir = config.get("core", "hooksPath", fallback="") + + if not hooks_dir: + return hook_path(name, index.repo.git_dir) + + return osp.abspath(osp.join(index.repo.working_dir, osp.expanduser(hooks_dir), name)) + + def _has_file_extension(path: str) -> str: return osp.splitext(path)[1] +def _is_in_windows_system_root(path: str) -> bool: + """Return whether ``path`` is inside the Windows installation directory.""" + system_root = os.environ.get("SystemRoot") + if not system_root: + return False + + system_root = osp.normcase(osp.realpath(system_root)) + path = osp.normcase(osp.realpath(path)) + try: + return osp.commonpath((system_root, path)) == system_root + except ValueError: + # Paths on different drives have no common path on Windows. + return False + + +def _which_from_path(command: str) -> Union[str, None]: + """Resolve ``command`` from PATH, excluding the Windows installation.""" + for directory in os.get_exec_path(): + # Unlike POSIX, Windows does not define an empty PATH entry as the current + # directory. Skip it rather than letting abspath() turn it into one. + if not directory: + continue + directory = osp.abspath(directory) + candidate = osp.join(directory, command) + # SystemRoot contains the WSL launcher stubs. They are valid executables but + # not suitable for running a Windows Git hook: the hook path and environment + # were prepared for Git for Windows, and WSL may have no distribution at all. + if _is_in_windows_system_root(candidate): + continue + if osp.isfile(candidate) and os.access(candidate, os.X_OK): + return candidate + return None + + +_GIT_FOR_WINDOWS_PREFIXES = ("mingw64", "mingw32", "clangarm64", "clang64", "clang32", "ucrt64") + + +def _git_for_windows_root() -> Union[str, None]: + """Infer a standard Git for Windows root from GitPython's selected executable.""" + git_executable = os.fspath(Git.GIT_PYTHON_GIT_EXECUTABLE or Git.git_exec_name) + if osp.dirname(git_executable): + # CreateProcess resolves a relative executable path containing a directory + # from the parent process cwd, even when Popen supplies a different child cwd. + git_executable = osp.abspath(git_executable) + else: + # GitPython deliberately retains a bare executable name so later PATH changes + # affect Git commands. Resolve it with the same PATH snapshot used for Bash. + names = (git_executable,) if _has_file_extension(git_executable) else (git_executable, f"{git_executable}.exe") + for name in names: + resolved = _which_from_path(name) + if resolved is not None: + git_executable = resolved + break + else: + git_executable = "" + if not git_executable: + return None + if osp.basename(git_executable).lower() not in ("git", "git.exe"): + return None + + executable_dir = osp.dirname(git_executable) + directory_name = osp.basename(executable_dir).lower() + if directory_name == "cmd": + # The normal system-wide PATH entry is /cmd. + return osp.dirname(executable_dir) + if directory_name == "bin": + prefix = osp.dirname(executable_dir) + if osp.basename(prefix).lower() in _GIT_FOR_WINDOWS_PREFIXES: + # Git Bash commonly exposes //bin/git.exe. + return osp.dirname(prefix) + if osp.basename(prefix).lower() != "usr": + # An explicitly configured Git may be the root-level bin/git.exe. Do + # not make the same inference from usr/bin: unlike the recognized + # platform prefixes, "usr" has no reliably bounded parent layout. + return prefix + return None + + +def _git_for_windows_bash() -> Union[str, None]: + """Return Bash from the Git for Windows installation selected by GitPython.""" + git_root = _git_for_windows_root() + if git_root is None: + return None + + # Match gix-path's precedence: prefer the lightweight bin shim, then the + # underlying usr/bin executable. Both belong to the same installation as Git. + for relative_path in ("bin/bash.exe", "usr/bin/bash.exe"): + candidate = osp.join(git_root, *relative_path.split("/")) + if osp.isfile(candidate) and os.access(candidate, os.X_OK): + return candidate + return None + + def run_commit_hook(name: str, index: "IndexFile", *args: str) -> None: """Run the commit hook of the given name. Silently ignore hooks that do not exist. @@ -82,20 +186,31 @@ def run_commit_hook(name: str, index: "IndexFile", *args: str) -> None: :raise git.exc.HookExecutionError: """ - hp = hook_path(name, index.repo.git_dir) + hp = _commit_hook_path(name, index) if not os.access(hp, os.X_OK): return env = os.environ.copy() - env["GIT_INDEX_FILE"] = safe_decode(str(index.path)) + env["GIT_INDEX_FILE"] = safe_decode(os.fspath(index.path)) env["GIT_EDITOR"] = ":" cmd = [hp] try: if sys.platform == "win32" and not _has_file_extension(hp): # Windows only uses extensions to determine how to open files # (doesn't understand shebangs). Try using bash to run the hook. - relative_hp = Path(hp).relative_to(index.repo.working_dir).as_posix() - cmd = ["bash.exe", relative_hp] + try: + bash_hp = osp.relpath(hp, index.repo.working_dir) + except ValueError: + # Different drives have no relative path on Windows. Git Bash accepts + # an absolute path in this form, although a relative path is preferable + # because it also works with the Windows Subsystem for Linux wrapper. + bash_hp = hp + # Prefer Bash associated with GitPython's selected Git installation. If + # that layout is not recognized, use an explicitly configured non-system + # PATH entry. Preserve the bare fallback for installations that previously + # relied on WSL or another CreateProcess-resolved Bash. + bash_executable = _git_for_windows_bash() or _which_from_path("bash.exe") or "bash.exe" + cmd = [bash_executable, Path(bash_hp).as_posix()] process = safer_popen( cmd + list(args), @@ -158,7 +273,7 @@ def write_cache( write = stream_sha.write # Header - version = 2 + version = 3 if any(entry.extended_flags for entry in entries) else 2 write(b"DIRC") write(pack(">LL", version, len(entries))) @@ -172,6 +287,8 @@ def write_cache( plen = len(path) & CE_NAMEMASK # Path length assert plen == len(path), "Path %s too long to fit into index" % entry.path flags = plen | (entry.flags & CE_NAMEMASK_INV) # Clear possible previous values. + if entry.extended_flags: + flags |= CE_EXTENDED write( pack( ">LLLLLL20sH", @@ -185,6 +302,8 @@ def write_cache( flags, ) ) + if entry.extended_flags: + write(pack(">H", entry.extended_flags)) write(path) real_size = (tell() - beginoffset + 8) & ~7 write(b"\0" * ((beginoffset + real_size) - tell())) @@ -206,8 +325,7 @@ def read_header(stream: IO[bytes]) -> Tuple[int, int]: unpacked = cast(Tuple[int, int], unpack(">LL", stream.read(4 * 2))) version, num_entries = unpacked - # TODO: Handle version 3: extended data, see read-cache.c. - assert version in (1, 2) + assert version in (1, 2, 3), "Unsupported git index version %i, only 1, 2, and 3 are supported" % version return version, num_entries @@ -260,12 +378,15 @@ def read_cache( ctime = unpack(">8s", read(8))[0] mtime = unpack(">8s", read(8))[0] (dev, ino, mode, uid, gid, size, sha, flags) = unpack(">LLLLLL20sH", read(20 + 4 * 6 + 2)) + extended_flags = 0 + if flags & CE_EXTENDED: + extended_flags = unpack(">H", read(2))[0] path_size = flags & CE_NAMEMASK path = read(path_size).decode(defenc) real_size = (tell() - beginoffset + 8) & ~7 read((beginoffset + real_size) - tell()) - entry = IndexEntry((mode, sha, flags, path, ctime, mtime, dev, ino, uid, gid, size)) + entry = IndexEntry((mode, sha, flags, path, ctime, mtime, dev, ino, uid, gid, size, extended_flags)) # entry_key would be the method to use, but we save the effort. entries[(path, entry.stage)] = entry count += 1 diff --git a/git/index/typ.py b/git/index/typ.py index 974252528..927633a9f 100644 --- a/git/index/typ.py +++ b/git/index/typ.py @@ -32,6 +32,9 @@ CE_VALID = 0x8000 CE_STAGESHIFT = 12 +CE_EXT_SKIP_WORKTREE = 0x4000 +CE_EXT_INTENT_TO_ADD = 0x2000 + # } END invariants @@ -87,6 +90,8 @@ class BaseIndexEntryHelper(NamedTuple): uid: int = 0 gid: int = 0 size: int = 0 + # version 3 extended flags, only when (flags & CE_EXTENDED) is set + extended_flags: int = 0 class BaseIndexEntry(BaseIndexEntryHelper): @@ -102,7 +107,7 @@ def __new__( cls, inp_tuple: Union[ Tuple[int, bytes, int, PathLike], - Tuple[int, bytes, int, PathLike, bytes, bytes, int, int, int, int, int], + Tuple[int, bytes, int, PathLike, bytes, bytes, int, int, int, int, int, int], ], ) -> "BaseIndexEntry": """Override ``__new__`` to allow construction from a tuple for backwards @@ -134,6 +139,14 @@ def stage(self) -> int: """ return (self.flags & CE_STAGEMASK) >> CE_STAGESHIFT + @property + def skip_worktree(self) -> bool: + return (self.extended_flags & CE_EXT_SKIP_WORKTREE) > 0 + + @property + def intent_to_add(self) -> bool: + return (self.extended_flags & CE_EXT_INTENT_TO_ADD) > 0 + @classmethod def from_blob(cls, blob: Blob, stage: int = 0) -> "BaseIndexEntry": """:return: Fully equipped BaseIndexEntry at the given stage""" @@ -179,7 +192,7 @@ def from_base(cls, base: "BaseIndexEntry") -> "IndexEntry": Instance of type :class:`BaseIndexEntry`. """ time = pack(">LL", 0, 0) - return IndexEntry((base.mode, base.binsha, base.flags, base.path, time, time, 0, 0, 0, 0, 0)) + return IndexEntry((base.mode, base.binsha, base.flags, base.path, time, time, 0, 0, 0, 0, 0)) # type: ignore[arg-type] @classmethod def from_blob(cls, blob: Blob, stage: int = 0) -> "IndexEntry": @@ -198,5 +211,5 @@ def from_blob(cls, blob: Blob, stage: int = 0) -> "IndexEntry": 0, 0, blob.size, - ) + ) # type: ignore[arg-type] ) diff --git a/git/index/util.py b/git/index/util.py index e59cb609f..982a5afb7 100644 --- a/git/index/util.py +++ b/git/index/util.py @@ -15,7 +15,7 @@ # typing ---------------------------------------------------------------------- -from typing import Any, Callable, TYPE_CHECKING, Optional, Type +from typing import Any, Callable, TYPE_CHECKING, Optional, Type, cast from git.types import Literal, PathLike, _T @@ -106,7 +106,7 @@ def git_working_dir(func: Callable[..., _T]) -> Callable[..., _T]: @wraps(func) def set_git_working_dir(self: "IndexFile", *args: Any, **kwargs: Any) -> _T: cur_wd = os.getcwd() - os.chdir(str(self.repo.working_tree_dir)) + os.chdir(cast(PathLike, self.repo.working_tree_dir)) try: return func(self, *args, **kwargs) finally: diff --git a/git/objects/base.py b/git/objects/base.py index eeaebc09b..faf600c6b 100644 --- a/git/objects/base.py +++ b/git/objects/base.py @@ -122,7 +122,7 @@ def new(cls, repo: "Repo", id: Union[str, "Reference"]) -> AnyGitObject: :return: New :class:`Object` instance of a type appropriate to the object type behind `id`. The id of the newly created object will be a binsha even though the - input id may have been a `~git.refs.reference.Reference` or rev-spec. + input id may have been a :class:`~git.refs.reference.Reference` or rev-spec. :param id: :class:`~git.refs.reference.Reference`, rev-spec, or hexsha. @@ -218,7 +218,7 @@ class IndexObject(Object): """Base for all objects that can be part of the index file. The classes representing git object types that can be part of the index file are - :class:`~git.objects.tree.Tree and :class:`~git.objects.blob.Blob`. In addition, + :class:`~git.objects.tree.Tree` and :class:`~git.objects.blob.Blob`. In addition, :class:`~git.objects.submodule.base.Submodule`, which is not really a git object type but can be part of an index file, is also a subclass. """ diff --git a/git/objects/blob.py b/git/objects/blob.py index 58de59642..f7d49c9cc 100644 --- a/git/objects/blob.py +++ b/git/objects/blob.py @@ -6,6 +6,7 @@ __all__ = ["Blob"] from mimetypes import guess_type +import os import sys if sys.version_info >= (3, 8): @@ -44,5 +45,5 @@ def mime_type(self) -> str: """ guesses = None if self.path: - guesses = guess_type(str(self.path)) + guesses = guess_type(os.fspath(self.path)) return guesses and guesses[0] or self.DEFAULT_MIME_TYPE diff --git a/git/objects/commit.py b/git/objects/commit.py index 0ceb46609..45843eac2 100644 --- a/git/objects/commit.py +++ b/git/objects/commit.py @@ -86,6 +86,12 @@ class Commit(base.Object, TraversableIterableObj, Diffable, Serializable): # INVARIANTS default_encoding = "UTF-8" + # Options to :manpage:`git-rev-list(1)` that can overwrite files. + unsafe_git_rev_options = [ + "--output", + "-o", + ] + type: Literal["commit"] = "commit" __slots__ = ( @@ -263,13 +269,21 @@ def summary(self) -> Union[str, bytes]: else: return self.message.split(b"\n", 1)[0] - def count(self, paths: Union[PathLike, Sequence[PathLike]] = "", **kwargs: Any) -> int: + def count( + self, + paths: Union[PathLike, Sequence[PathLike]] = "", + allow_unsafe_options: bool = False, + **kwargs: Any, + ) -> int: """Count the number of commits reachable from this commit. :param paths: An optional path or a list of paths restricting the return value to commits actually containing the paths. + :param allow_unsafe_options: + Allow unsafe options, like ``--output``. + :param kwargs: Additional options to be passed to :manpage:`git-rev-list(1)`. They must not alter the output style of the command, or parsing will yield incorrect @@ -278,6 +292,11 @@ def count(self, paths: Union[PathLike, Sequence[PathLike]] = "", **kwargs: Any) :return: An int defining the number of reachable commits """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), unsafe_options=self.unsafe_git_rev_options + ) + # Yes, it makes a difference whether empty paths are given or not in our case as # the empty paths version will ignore merge commits for some reason. if paths: @@ -289,7 +308,7 @@ def name_rev(self) -> str: """ :return: String describing the commits hex sha based on the closest - `~git.refs.reference.Reference`. + :class:`~git.refs.reference.Reference`. :note: Mostly useful for UI purposes. @@ -302,6 +321,7 @@ def iter_items( repo: "Repo", rev: Union[str, "Commit", "SymbolicReference"], paths: Union[PathLike, Sequence[PathLike]] = "", + allow_unsafe_options: bool = False, **kwargs: Any, ) -> Iterator["Commit"]: R"""Find all commits matching the given criteria. @@ -330,6 +350,11 @@ def iter_items( raise ValueError("--pretty cannot be used as parsing expects single sha's only") # END handle pretty + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([rev], kwargs), unsafe_options=cls.unsafe_git_rev_options + ) + # Use -- in all cases, to prevent possibility of ambiguous arguments. # See https://github.com/gitpython-developers/GitPython/issues/264. @@ -349,7 +374,7 @@ def iter_items( return cls._iter_from_process_or_stream(repo, proc) def iter_parents(self, paths: Union[PathLike, Sequence[PathLike]] = "", **kwargs: Any) -> Iterator["Commit"]: - R"""Iterate _all_ parents of this commit. + R"""Iterate *all* parents of this commit. :param paths: Optional path or list of paths limiting the :class:`Commit`\s to those that @@ -374,6 +399,11 @@ def stats(self) -> Stats: """Create a git stat from changes between this commit and its first parent or from all changes done if this is the very first commit. + :note: + If this commit is at the boundary of a shallow clone, this will + raise :exc:`~git.exc.GitCommandError`, since the parent object + was never fetched and only exists as a reference on this commit. + :return: :class:`Stats` """ @@ -450,14 +480,7 @@ def trailers_list(self) -> List[Tuple[str, str]]: :return: List containing key-value tuples of whitespace stripped trailer information. """ - cmd = ["git", "interpret-trailers", "--parse"] - proc: Git.AutoInterrupt = self.repo.git.execute( # type: ignore[call-overload] - cmd, - as_process=True, - istream=PIPE, - ) - trailer: str = proc.communicate(str(self.message).encode())[0].decode("utf8") - trailer = trailer.strip() + trailer = self._interpret_trailers(self.repo, self.message, ["--parse"], encoding=self.encoding).strip() if not trailer: return [] @@ -469,6 +492,27 @@ def trailers_list(self) -> List[Tuple[str, str]]: return trailer_list + @classmethod + def _interpret_trailers( + cls, + repo: "Repo", + message: Union[str, bytes], + trailer_args: Sequence[str], + encoding: str = default_encoding, + ) -> str: + message_bytes = message if isinstance(message, bytes) else message.encode(encoding, errors="strict") + cmd = [repo.git.GIT_PYTHON_GIT_EXECUTABLE, "interpret-trailers", *trailer_args] + proc: Git.AutoInterrupt = repo.git.execute( # type: ignore[call-overload] + cmd, + as_process=True, + istream=PIPE, + ) + try: + stdout_bytes, _ = proc.communicate(message_bytes) + return stdout_bytes.decode(encoding, errors="strict") + finally: + finalize_process(proc) + @property def trailers_dict(self) -> Dict[str, List[str]]: """Get the trailers of the message as a dictionary. @@ -531,11 +575,14 @@ def _iter_from_process_or_stream(cls, repo: "Repo", proc_or_stream: Union[Popen, if hasattr(proc_or_stream, "wait"): proc_or_stream = cast(Popen, proc_or_stream) - if proc_or_stream.stdout is not None: - stream = proc_or_stream.stdout + stream = proc_or_stream.stdout + if stream is None: + raise ValueError("Process has no stdout stream") elif hasattr(proc_or_stream, "readline"): proc_or_stream = cast(IO, proc_or_stream) # type: ignore[redundant-cast] stream = proc_or_stream + else: + raise TypeError("Expected a process or stream") readline = stream.readline while True: @@ -570,6 +617,7 @@ def create_from_tree( committer: Union[None, Actor] = None, author_date: Union[None, str, datetime.datetime] = None, commit_date: Union[None, str, datetime.datetime] = None, + trailers: Union[None, Dict[str, str], List[Tuple[str, str]]] = None, ) -> "Commit": """Commit the given tree, creating a :class:`Commit` object. @@ -609,6 +657,14 @@ def create_from_tree( :param commit_date: The timestamp for the committer field. + :param trailers: + Optional trailer key-value pairs to append to the commit message. + Can be a dictionary mapping trailer keys to values, or a list of + ``(key, value)`` tuples (useful when the same key appears multiple + times, e.g. multiple ``Signed-off-by`` trailers). Trailers are + appended using ``git interpret-trailers``. + See :manpage:`git-interpret-trailers(1)`. + :return: :class:`Commit` object representing the new commit. @@ -678,6 +734,21 @@ def create_from_tree( tree = repo.tree(tree) # END tree conversion + # APPLY TRAILERS + if trailers: + trailer_args: List[str] = [] + if isinstance(trailers, dict): + for key, val in trailers.items(): + trailer_args.append("--trailer") + trailer_args.append(f"{key}: {val}") + else: + for key, val in trailers: + trailer_args.append("--trailer") + trailer_args.append(f"{key}: {val}") + + message = cls._interpret_trailers(repo, str(message), trailer_args) + # END apply trailers + # CREATE NEW COMMIT new_commit = cls( repo, @@ -900,7 +971,7 @@ def co_authors(self) -> List[Actor]: if self.message: results = re.findall( r"^Co-authored-by: (.*) <(.*?)>$", - self.message, + str(self.message), re.MULTILINE, ) for author in results: diff --git a/git/objects/fun.py b/git/objects/fun.py index fe57da13a..ad5fbd59b 100644 --- a/git/objects/fun.py +++ b/git/objects/fun.py @@ -115,11 +115,11 @@ def tree_entries_from_data(data: bytes) -> List[EntryTup]: # Default encoding for strings in git is UTF-8. # Only use the respective unicode object if the byte stream was encoded. name_bytes = data[ns:i] - name = safe_decode(name_bytes) + name = safe_decode(bytes(name_bytes)) # Byte is NULL, get next 20. i += 1 - sha = data[i : i + 20] + sha = bytes(data[i : i + 20]) i = i + 20 out.append((sha, mode, name)) # END for each byte in data stream diff --git a/git/objects/submodule/base.py b/git/objects/submodule/base.py index fa60bcdaf..da0e09af4 100644 --- a/git/objects/submodule/base.py +++ b/git/objects/submodule/base.py @@ -6,11 +6,13 @@ import gc from io import BytesIO import logging +import ntpath import os import os.path as osp import stat import sys import uuid +import urllib import git from git.cmd import Git @@ -27,6 +29,7 @@ from git.util import ( IterableList, RemoteProgress, + _to_relative_path, join_path_native, rmtree, to_native_path_linux, @@ -48,6 +51,7 @@ Callable, Dict, Iterator, + List, Mapping, Sequence, TYPE_CHECKING, @@ -65,7 +69,7 @@ if TYPE_CHECKING: from git.index import IndexFile from git.objects.commit import Commit - from git.refs import Head + from git.refs import Head, RemoteReference from git.repo import Repo # ----------------------------------------------------------------------------- @@ -300,8 +304,20 @@ def _config_parser_constrained(self, read_only: bool) -> SectionConstraint: parser.set_submodule(self) return SectionConstraint(parser, sm_section(self.name)) + @classmethod + def _validated_name(cls, name: str) -> str: + if ( + not name + or name.startswith(("/", "\\")) + or ntpath.splitdrive(name)[0] + or ".." in name.replace("\\", "/").split("/") + ): + raise ValueError("Invalid submodule name %r" % name) + return name + @classmethod def _module_abspath(cls, parent_repo: "Repo", path: PathLike, name: str) -> PathLike: + name = cls._validated_name(name) if cls._need_gitfile_submodules(parent_repo.git): return osp.join(parent_repo.git_dir, "modules", name) if parent_repo.working_tree_dir: @@ -351,7 +367,12 @@ def _clone_repo( module_abspath_dir = osp.dirname(module_abspath) if not osp.isdir(module_abspath_dir): os.makedirs(module_abspath_dir) - module_checkout_path = osp.join(str(repo.working_tree_dir), path) + module_checkout_path = osp.join(repo.working_tree_dir, path) # type: ignore[arg-type] + + if url.startswith("../"): + remote_name = cast("RemoteReference", repo.active_branch.tracking_branch()).remote_name + repo_remote_url = repo.remote(remote_name).url + url = os.path.join(repo_remote_url, url) clone = git.Repo.clone_from( url, @@ -372,23 +393,14 @@ def _to_relative_path(cls, parent_repo: "Repo", path: PathLike) -> PathLike: :raise ValueError: If path is not contained in the parent repository's working tree. """ - path = to_native_path_linux(path) + if parent_repo.working_tree_dir: + path = _to_relative_path(parent_repo.working_tree_dir, path) + else: + path = to_native_path_linux(path) if path.endswith("/"): path = path[:-1] - # END handle trailing slash - - if osp.isabs(path) and parent_repo.working_tree_dir: - working_tree_linux = to_native_path_linux(parent_repo.working_tree_dir) - if not path.startswith(working_tree_linux): - raise ValueError( - "Submodule checkout path '%s' needs to be within the parents repository at '%s'" - % (working_tree_linux, path) - ) - path = path[len(working_tree_linux.rstrip("/")) + 1 :] - if not path: - raise ValueError("Absolute submodule path '%s' didn't yield a valid relative path" % path) - # END verify converted relative path makes sense - # END convert to a relative path + if not path or path == ".": + raise ValueError("Submodule checkout path must not be the repository root") return path @@ -516,6 +528,7 @@ def add( raise InvalidGitRepositoryError("Cannot add submodules to bare repositories") # END handle bare repos + cls._validated_name(name) path = cls._to_relative_path(repo, path) # Ensure we never put backslashes into the URL, as might happen on Windows. @@ -535,7 +548,7 @@ def add( if sm.exists(): # Reretrieve submodule from tree. try: - sm = repo.head.commit.tree[str(path)] + sm = repo.head.commit.tree[os.fspath(path)] sm._name = name return sm except KeyError: @@ -601,7 +614,7 @@ def add( # END verify url ## See #525 for ensuring git URLs in config-files are valid under Windows. - url = Git.polish_url(url) + url = Git.polish_url(url, expand_vars=False) # It's important to add the URL to the parent config, to let `git submodule` know. # Otherwise there is a '-' character in front of the submodule listing: @@ -727,129 +740,144 @@ def update( prefix = "DRY-RUN: " # END handle prefix - # To keep things plausible in dry-run mode. - if dry_run: - mrepo = None + mrepo = None # END init mrepo + def fetch_remotes(module_repo: "Repo") -> None: + rmts = module_repo.remotes + len_rmts = len(rmts) + for i, remote in enumerate(rmts): + op = FETCH + if i == 0: + op |= BEGIN + # END handle start + + progress.update( + op, + i, + len_rmts, + prefix + "Fetching remote %s of submodule %r" % (remote, self.name), + ) + # =============================== + if not dry_run: + remote.fetch(progress=progress) + # END handle dry-run + # =============================== + if i == len_rmts - 1: + op |= END + # END handle end + progress.update( + op, + i, + len_rmts, + prefix + "Done fetching remote of submodule %r" % self.name, + ) + # END fetch new data + try: + self._validated_name(self.name) + # ENSURE REPO IS PRESENT AND UP-TO-DATE ####################################### try: mrepo = self.module() - rmts = mrepo.remotes - len_rmts = len(rmts) - for i, remote in enumerate(rmts): - op = FETCH - if i == 0: - op |= BEGIN - # END handle start - - progress.update( - op, - i, - len_rmts, - prefix + "Fetching remote %s of submodule %r" % (remote, self.name), - ) - # =============================== - if not dry_run: - remote.fetch(progress=progress) - # END handle dry-run - # =============================== - if i == len_rmts - 1: - op |= END - # END handle end - progress.update( - op, - i, - len_rmts, - prefix + "Done fetching remote of submodule %r" % self.name, - ) - # END fetch new data + fetch_remotes(mrepo) except InvalidGitRepositoryError: mrepo = None if not init: return self # END early abort if init is not allowed - # There is no git-repository yet - but delete empty paths. checkout_module_abspath = self.abspath - if not dry_run and osp.isdir(checkout_module_abspath): - try: - os.rmdir(checkout_module_abspath) - except OSError as e: - raise OSError( - "Module directory at %r does already exist and is non-empty" % checkout_module_abspath - ) from e - # END handle OSError - # END handle directory removal - - # Don't check it out at first - nonetheless it will create a local - # branch according to the remote-HEAD if possible. - progress.update( - BEGIN | CLONE, - 0, - 1, - prefix - + "Cloning url '%s' to '%s' in submodule %r" % (self.url, checkout_module_abspath, self.name), - ) - if not dry_run: - mrepo = self._clone_repo( - self.repo, - self.url, - self.path, - self.name, - n=True, - env=env, - multi_options=clone_multi_options, - allow_unsafe_options=allow_unsafe_options, - allow_unsafe_protocols=allow_unsafe_protocols, - ) - # END handle dry-run - progress.update( - END | CLONE, - 0, - 1, - prefix + "Done cloning to %s" % checkout_module_abspath, - ) + module_abspath = self._module_abspath(self.repo, self.path, self.name) - if not dry_run: - # See whether we have a valid branch to check out. + # ``git submodule deinit`` leaves the repository in + # ``.git/modules`` and empties the checkout. Reconnect that retained + # repository instead of trying to clone over it. + if not dry_run and osp.isdir(module_abspath): try: - mrepo = cast("Repo", mrepo) - # Find a remote which has our branch - we try to be flexible. - remote_branch = find_first_remote_branch(mrepo.remotes, self.branch_name) - local_branch = mkhead(mrepo, self.branch_path) - - # Have a valid branch, but no checkout - make sure we can figure - # that out by marking the commit with a null_sha. - local_branch.set_object(Object(mrepo, self.NULL_BIN_SHA)) - # END initial checkout + branch creation - - # Make sure HEAD is not detached. - mrepo.head.set_reference( - local_branch, - logmsg="submodule: attaching head to %s" % local_branch, + git.Repo(module_abspath) + except InvalidGitRepositoryError: + pass + else: + if osp.lexists(checkout_module_abspath) and ( + osp.islink(checkout_module_abspath) + or not osp.isdir(checkout_module_abspath) + or os.listdir(checkout_module_abspath) + ): + raise OSError( + "Module directory at %r does already exist and is non-empty" % checkout_module_abspath + ) + os.makedirs(checkout_module_abspath, exist_ok=True) + self._write_git_file_and_module_config(checkout_module_abspath, module_abspath) + mrepo = git.Repo(checkout_module_abspath) + mrepo.head.reset(mrepo.head.commit, index=True, working_tree=True) + fetch_remotes(mrepo) + with self.repo.config_writer() as writer: + writer.set_value(sm_section(self.name), "url", self.url) + + if mrepo is None: + # There is no git-repository yet - but delete empty paths. + if not dry_run and osp.isdir(checkout_module_abspath): + try: + os.rmdir(checkout_module_abspath) + except OSError as e: + raise OSError( + "Module directory at %r does already exist and is non-empty" % checkout_module_abspath + ) from e + # END handle directory removal + + # Don't check it out at first - nonetheless it will create a local + # branch according to the remote-HEAD if possible. + progress.update( + BEGIN | CLONE, + 0, + 1, + prefix + + "Cloning url '%s' to '%s' in submodule %r" % (self.url, checkout_module_abspath, self.name), + ) + if not dry_run: + if self.url.startswith("."): + url = urllib.parse.urljoin(self.repo.remotes.origin.url + "/", self.url) + else: + url = self.url + mrepo = self._clone_repo( + self.repo, + url, + self.path, + self.name, + n=True, + env=env, + multi_options=clone_multi_options, + allow_unsafe_options=allow_unsafe_options, + allow_unsafe_protocols=allow_unsafe_protocols, ) - mrepo.head.reference.set_tracking_branch(remote_branch) - except (IndexError, InvalidGitRepositoryError): - _logger.warning("Failed to checkout tracking branch %s", self.branch_path) - # END handle tracking branch - - # NOTE: Have to write the repo config file as well, otherwise the - # default implementation will be offended and not update the - # repository. Maybe this is a good way to ensure it doesn't get into - # our way, but we want to stay backwards compatible too... It's so - # redundant! - with self.repo.config_writer() as writer: - writer.set_value(sm_section(self.name), "url", self.url) - # END handle dry_run + progress.update(END | CLONE, 0, 1, prefix + "Done cloning to %s" % checkout_module_abspath) + + if not dry_run: + # See whether we have a valid branch to check out. + try: + mrepo = cast("Repo", mrepo) + remote_branch = find_first_remote_branch(mrepo.remotes, self.branch_name) + local_branch = mkhead(mrepo, self.branch_path) + local_branch.set_object(Object(mrepo, self.NULL_BIN_SHA)) + mrepo.head.set_reference( + local_branch, + logmsg="submodule: attaching head to %s" % local_branch, + ) + mrepo.head.reference.set_tracking_branch(remote_branch) + except (IndexError, InvalidGitRepositoryError): + _logger.warning("Failed to checkout tracking branch %s", self.branch_path) + + with self.repo.config_writer() as writer: + writer.set_value(sm_section(self.name), "url", self.url) # END handle initialization # DETERMINE SHAS TO CHECK OUT ############################# binsha = self.binsha hexsha = self.hexsha + is_detached = False if mrepo is not None: # mrepo is only set if we are not in dry-run mode or if the module # existed. @@ -999,6 +1027,7 @@ def move(self, module_path: PathLike, configuration: bool = True, module: bool = raise ValueError("You must specify to move at least the module or the configuration of the submodule") # END handle input + self._validated_name(self.name) module_checkout_path = self._to_relative_path(self.repo, module_path) # VERIFY DESTINATION @@ -1139,6 +1168,7 @@ def remove( raise ValueError("Need to specify to delete at least the module, or the configuration") # END handle parameters + self._validated_name(self.name) # Recursively remove children of this submodule. nc = 0 for csm in self.children(): @@ -1192,6 +1222,7 @@ def remove( for remote in mod.remotes: num_branches_with_new_commits = 0 rrefs = remote.refs + rref = None for rref in rrefs: num_branches_with_new_commits += len(mod.git.cherry(rref)) != 0 # END for each remote ref @@ -1257,6 +1288,36 @@ def remove( return self + @unbare_repo + def deinit(self, force: bool = False) -> "Submodule": + """Run ``git submodule deinit`` on this submodule. + + This is a thin wrapper around ``git submodule deinit ``, + which unregisters the submodule (removes its entry from + ``.git/config`` and empties the working-tree directory) + without deleting the submodule from ``.gitmodules`` + or its checked-out repository under ``.git/modules/``. + A subsequent :meth:`update` will re-initialize the + submodule from the retained contents. + + :param force: + If ``True``, pass ``--force`` to ``git submodule deinit``. This + allows deinitialization even when the submodule's working tree has + local modifications that would otherwise block the command. + + :return: + self + + :note: + Doesn't work in bare repositories. + """ + args: List[str] = [] + if force: + args.append("--force") + args.extend(["--", str(self.path)]) + self.repo.git.submodule("deinit", *args) + return self + def set_parent_commit(self, commit: Union[Commit_ish, str, None], check: bool = True) -> "Submodule": """Set this instance to use the given commit whose tree is supposed to contain the ``.gitmodules`` blob. @@ -1365,6 +1426,9 @@ def rename(self, new_name: str) -> "Submodule": if self.name == new_name: return self + self._validated_name(self.name) + self._validated_name(new_name) + # .git/config with self.repo.config_writer() as pw: # As we ourselves didn't write anything about submodules into the parent @@ -1412,6 +1476,7 @@ def module(self) -> "Repo": If a repository was not available. This could also mean that it was not yet initialized. """ + self._validated_name(self.name) module_checkout_abspath = self.abspath try: repo = git.Repo(module_checkout_abspath) diff --git a/git/objects/submodule/root.py b/git/objects/submodule/root.py index d93193fa3..d068049c1 100644 --- a/git/objects/submodule/root.py +++ b/git/objects/submodule/root.py @@ -7,6 +7,7 @@ import git from git.exc import InvalidGitRepositoryError +from git.util import IterableList from .base import Submodule, UpdateProgress from .util import find_first_remote_branch @@ -19,7 +20,6 @@ if TYPE_CHECKING: from git.repo import Repo - from git.util import IterableList # ---------------------------------------------------------------------------- @@ -162,6 +162,7 @@ def update( # type: ignore[override] prefix = "DRY-RUN: " repo = self.repo + sms: "IterableList[Submodule]" = IterableList("name") try: # SETUP BASE COMMIT @@ -182,7 +183,7 @@ def update( # type: ignore[override] # END handle previous commit psms: "IterableList[Submodule]" = self.list_items(repo, parent_commit=previous_commit) - sms: "IterableList[Submodule]" = self.list_items(repo) + sms = self.list_items(repo) spsms = set(psms) ssms = set(sms) diff --git a/git/objects/tree.py b/git/objects/tree.py index 09184a781..a3d611c80 100644 --- a/git/objects/tree.py +++ b/git/objects/tree.py @@ -5,6 +5,7 @@ __all__ = ["TreeModifier", "Tree"] +import os import sys import git.diff as git_diff @@ -50,7 +51,9 @@ # -------------------------------------------------------- -cmp: Callable[[str, str], int] = lambda a, b: (a > b) - (a < b) + +def cmp(a: str, b: str) -> int: + return (a > b) - (a < b) class TreeModifier: @@ -228,7 +231,7 @@ def _iter_convert_to_object(self, iterable: Iterable[TreeCacheTup]) -> Iterator[ raise TypeError("Unknown mode %o found in tree data for path '%s'" % (mode, path)) from e # END for each item - def join(self, file: str) -> IndexObjUnion: + def join(self, file: PathLike) -> IndexObjUnion: """Find the named object in this tree's contents. :return: @@ -239,6 +242,7 @@ def join(self, file: str) -> IndexObjUnion: If the given file or tree does not exist in this tree. """ msg = "Blob or Tree named %r not found" + file = os.fspath(file) if "/" in file: tree = self item = self @@ -267,7 +271,7 @@ def join(self, file: str) -> IndexObjUnion: raise KeyError(msg % file) # END handle long paths - def __truediv__(self, file: str) -> IndexObjUnion: + def __truediv__(self, file: PathLike) -> IndexObjUnion: """The ``/`` operator is another syntax for joining. See :meth:`join` for details. diff --git a/git/refs/head.py b/git/refs/head.py index 683634451..7f563374e 100644 --- a/git/refs/head.py +++ b/git/refs/head.py @@ -19,10 +19,10 @@ from typing import Any, Sequence, TYPE_CHECKING, Union +from git.cmd import Git from git.types import Commit_ish, PathLike if TYPE_CHECKING: - from git.objects import Commit from git.refs import RemoteReference from git.repo import Repo @@ -44,9 +44,6 @@ class HEAD(SymbolicReference): __slots__ = () - # TODO: This can be removed once SymbolicReference.commit has static type hints. - commit: "Commit" - def __init__(self, repo: "Repo", path: PathLike = _HEAD_NAME) -> None: if path != self._HEAD_NAME: raise ValueError("HEAD instance must point to %r, got %r" % (self._HEAD_NAME, path)) @@ -66,6 +63,7 @@ def reset( index: bool = True, working_tree: bool = False, paths: Union[PathLike, Sequence[PathLike], None] = None, + allow_unsafe_options: bool = False, **kwargs: Any, ) -> "HEAD": """Reset our HEAD to the given commit optionally synchronizing the index and @@ -88,12 +86,21 @@ def reset( Single path or list of paths relative to the git root directory that are to be reset. This allows to partially reset individual files. + :param allow_unsafe_options: + Allow unsafe options such as ``--pathspec-from-file`` to be passed to + :manpage:`git-reset(1)`. + :param kwargs: Additional arguments passed to :manpage:`git-reset(1)`. :return: self """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([commit], kwargs), + unsafe_options=Git.unsafe_git_pathspec_from_file_options, + ) mode: Union[str, None] mode = "--soft" if index: @@ -149,7 +156,7 @@ class Head(Reference): k_config_remote_ref = "merge" # Branch to merge from remote. @classmethod - def delete(cls, repo: "Repo", *heads: "Union[Head, str]", force: bool = False, **kwargs: Any) -> None: + def delete(cls, repo: "Repo", *heads: "Union[Head, str]", force: bool = False, **kwargs: Any) -> None: # type: ignore[override] """Delete the given heads. :param force: @@ -238,7 +245,12 @@ def rename(self, new_path: PathLike, force: bool = False) -> "Head": self.path = "%s/%s" % (self._common_path_default, new_path) return self - def checkout(self, force: bool = False, **kwargs: Any) -> Union["HEAD", "Head"]: + def checkout( + self, + force: bool = False, + allow_unsafe_options: bool = False, + **kwargs: Any, + ) -> Union["HEAD", "Head"]: """Check out this head by setting the HEAD to this reference, by updating the index to reflect the tree we point to and by updating the working tree to reflect the latest index. @@ -250,6 +262,10 @@ def checkout(self, force: bool = False, **kwargs: Any) -> Union["HEAD", "Head"]: If ``False``, :exc:`~git.exc.GitCommandError` will be raised in that situation. + :param allow_unsafe_options: + Allow unsafe options such as ``--pathspec-from-file`` to be passed to + :manpage:`git-checkout(1)`. + :param kwargs: Additional keyword arguments to be passed to git checkout, e.g. ``b="new_branch"`` to create a new branch at the given spot. @@ -265,6 +281,11 @@ def checkout(self, force: bool = False, **kwargs: Any) -> Union["HEAD", "Head"]: the HEAD detached which is allowed and possible, but remains a special state that some tools might not be able to handle. """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), + unsafe_options=Git.unsafe_git_pathspec_from_file_options, + ) kwargs["f"] = force if kwargs["f"] is False: kwargs.pop("f") diff --git a/git/refs/log.py b/git/refs/log.py index 17e3a94b3..0681943bf 100644 --- a/git/refs/log.py +++ b/git/refs/log.py @@ -4,7 +4,6 @@ __all__ = ["RefLog", "RefLogEntry"] from mmap import mmap -import os.path as osp import re import time as _time @@ -126,7 +125,7 @@ def from_line(cls, line: bytes) -> "RefLogEntry": elif len(fields) == 2: info, msg = fields else: - raise ValueError("Line must have up to two TAB-separated fields." " Got %s" % repr(line_str)) + raise ValueError("Line must have up to two TAB-separated fields. Got %s" % repr(line_str)) # END handle first split oldhexsha = info[:40] @@ -145,7 +144,7 @@ def from_line(cls, line: bytes) -> "RefLogEntry": actor = Actor._from_string(info[82 : email_end + 1]) time, tz_offset = parse_date(info[email_end + 2 :]) # skipcq: PYL-W0621 - return RefLogEntry((oldhexsha, newhexsha, actor, (time, tz_offset), msg)) + return RefLogEntry((oldhexsha, newhexsha, actor, (time, tz_offset), msg)) # type: ignore [arg-type] class RefLog(List[RefLogEntry], Serializable): @@ -212,8 +211,11 @@ def path(cls, ref: "SymbolicReference") -> str: :param ref: :class:`~git.refs.symbolic.SymbolicReference` instance + + :raise ValueError: + If `ref.path` is invalid or escapes the repository's reflog directory. """ - return osp.join(ref.repo.git_dir, "logs", to_native_path(ref.path)) + return to_native_path(ref._get_validated_reflog_path(ref.repo, ref.path)) @classmethod def iter_entries(cls, stream: Union[str, "BytesIO", mmap]) -> Iterator[RefLogEntry]: @@ -267,6 +269,7 @@ def entry_at(cls, filepath: PathLike, index: int) -> "RefLogEntry": return RefLogEntry.from_line(fp.readlines()[index].strip()) # Read until index is reached. + line = b"" for i in range(index + 1): line = fp.readline() if not line: diff --git a/git/refs/reference.py b/git/refs/reference.py index e5d473779..0c4327225 100644 --- a/git/refs/reference.py +++ b/git/refs/reference.py @@ -3,6 +3,7 @@ __all__ = ["Reference"] +import os from git.util import IterableObj, LazyMixin from .symbolic import SymbolicReference, T_References @@ -65,7 +66,7 @@ def __init__(self, repo: "Repo", path: PathLike, check_path: bool = True) -> Non If ``False``, you can provide any path. Otherwise the path must start with the default path prefix of this type. """ - if check_path and not str(path).startswith(self._common_path_default + "/"): + if check_path and not os.fspath(path).startswith(self._common_path_default + "/"): raise ValueError(f"Cannot instantiate {self.__class__.__name__!r} from path {path}") self.path: str # SymbolicReference converts to string at the moment. super().__init__(repo, path) diff --git a/git/refs/remote.py b/git/refs/remote.py index b4f4f7b36..e16ae70f8 100644 --- a/git/refs/remote.py +++ b/git/refs/remote.py @@ -58,17 +58,20 @@ def delete(cls, repo: "Repo", *refs: "RemoteReference", **kwargs: Any) -> None: `kwargs` are given for comparability with the base class method as we should not narrow the signature. """ + for ref in refs: + cls._check_ref_name_valid(ref.path) + repo.git.branch("-d", "-r", *refs) # The official deletion method will ignore remote symbolic refs - these are # generally ignored in the refs/ folder. We don't though and delete remainders # manually. for ref in refs: try: - os.remove(os.path.join(repo.common_dir, ref.path)) + os.remove(cls._get_validated_path(repo.common_dir, ref.path)) except OSError: pass try: - os.remove(os.path.join(repo.git_dir, ref.path)) + os.remove(cls._get_validated_path(repo.git_dir, ref.path)) except OSError: pass # END for each ref diff --git a/git/refs/symbolic.py b/git/refs/symbolic.py index 510850b2e..824d0c46c 100644 --- a/git/refs/symbolic.py +++ b/git/refs/symbolic.py @@ -4,6 +4,7 @@ __all__ = ["SymbolicReference"] import os +from pathlib import Path from gitdb.exc import BadName, BadObject @@ -39,8 +40,8 @@ if TYPE_CHECKING: from git.config import GitConfigParser from git.objects.commit import Actor - from git.refs import Head, TagReference, RemoteReference, Reference from git.refs.log import RefLogEntry + from git.refs.reference import Reference from git.repo import Repo @@ -76,10 +77,10 @@ class SymbolicReference: def __init__(self, repo: "Repo", path: PathLike, check_path: bool = False) -> None: self.repo = repo - self.path = path + self.path: PathLike = path def __str__(self) -> str: - return str(self.path) + return os.fspath(self.path) def __repr__(self) -> str: return '' % (self.__class__.__name__, self.path) @@ -103,12 +104,38 @@ def name(self) -> str: In case of symbolic references, the shortest assumable name is the path itself. """ - return str(self.path) + return os.fspath(self.path) @property def abspath(self) -> PathLike: return join_path_native(_git_dir(self.repo, self.path), self.path) + @staticmethod + def _get_validated_path(base: PathLike, path: PathLike) -> str: + path = os.fspath(path) + base_path = os.path.realpath(os.fspath(base)) + abs_path = os.path.realpath(os.path.join(base_path, path)) + try: + common_path = os.path.commonpath([base_path, abs_path]) + except ValueError as e: + raise ValueError("Reference path %r escapes the repository" % path) from e + if common_path != base_path: + raise ValueError("Reference path %r escapes the repository" % path) + return abs_path + + @classmethod + def _get_validated_ref_path(cls, repo: "Repo", path: PathLike) -> str: + """Return the absolute filesystem path for a ref after validating it.""" + cls._check_ref_name_valid(path) + ref_path = os.fspath(path) + return cls._get_validated_path(_git_dir(repo, ref_path), ref_path) + + @classmethod + def _get_validated_reflog_path(cls, repo: "Repo", path: PathLike) -> str: + """Return the absolute filesystem path for a reflog after validating it.""" + cls._check_ref_name_valid(path) + return cls._get_validated_path(os.path.join(repo.git_dir, "logs"), path) + @classmethod def _get_packed_refs_path(cls, repo: "Repo") -> str: return os.path.join(repo.common_dir, "packed-refs") @@ -178,7 +205,7 @@ def _check_ref_name_valid(ref_path: PathLike) -> None: """ previous: Union[str, None] = None one_before_previous: Union[str, None] = None - for c in str(ref_path): + for c in os.fspath(ref_path): if c in " ~^:?*[\\": raise ValueError( f"Invalid reference '{ref_path}': references cannot contain spaces, tildes (~), carets (^)," @@ -212,7 +239,7 @@ def _check_ref_name_valid(ref_path: PathLike) -> None: raise ValueError(f"Invalid reference '{ref_path}': references cannot end with a forward slash (/)") elif previous == "@" and one_before_previous is None: raise ValueError(f"Invalid reference '{ref_path}': references cannot be '@'") - elif any(component.endswith(".lock") for component in str(ref_path).split("/")): + elif any(component.endswith(".lock") for component in Path(ref_path).parts): raise ValueError( f"Invalid reference '{ref_path}': references cannot have slash-separated components that end with" " '.lock'" @@ -235,7 +262,7 @@ def _get_ref_info_helper( tokens: Union[None, List[str], Tuple[str, str]] = None repodir = _git_dir(repo, ref_path) try: - with open(os.path.join(repodir, str(ref_path)), "rt", encoding="UTF-8") as fp: + with open(os.path.join(repodir, ref_path), "rt", encoding="UTF-8") as fp: # type: ignore[arg-type] value = fp.read().rstrip() # Don't only split on spaces, but on whitespace, which allows to parse lines like: # 60b64ef992065e2600bfef6187a97f92398a9144 branch 'master' of git-server:/path/to/repo @@ -387,19 +414,25 @@ def set_object( # set the commit on our reference return self._get_reference().set_object(object, logmsg) - commit = property( - _get_commit, - set_commit, # type: ignore[arg-type] - doc="Query or set commits directly", - ) + @property + def commit(self) -> "Commit": + """Query or set commits directly""" + return self._get_commit() + + @commit.setter + def commit(self, commit: Union[Commit, "SymbolicReference", str]) -> "SymbolicReference": + return self.set_commit(commit) + + @property + def object(self) -> AnyGitObject: + """Return the object our ref currently refers to""" + return self._get_object() - object = property( - _get_object, - set_object, # type: ignore[arg-type] - doc="Return the object our ref currently refers to", - ) + @object.setter + def object(self, object: Union[AnyGitObject, "SymbolicReference", str]) -> "SymbolicReference": + return self.set_object(object) - def _get_reference(self) -> "SymbolicReference": + def _get_reference(self) -> "Reference": """ :return: :class:`~git.refs.reference.Reference` object we point to @@ -411,7 +444,7 @@ def _get_reference(self) -> "SymbolicReference": sha, target_ref_path = self._get_ref_info(self.repo, self.path) if target_ref_path is None: raise TypeError("%s is a detached symbolic reference as it points to %r" % (self, sha)) - return self.from_path(self.repo, target_ref_path) + return cast("Reference", self.from_path(self.repo, target_ref_path)) def set_reference( self, @@ -478,7 +511,7 @@ def set_reference( # END handle non-existing # END retrieve old hexsha - fpath = self.abspath + fpath = self._get_validated_ref_path(self.repo, self.path) assure_directory_exists(fpath, is_file=True) lfd = LockedFD(fpath) @@ -496,12 +529,14 @@ def set_reference( return self # Aliased reference - reference: Union["Head", "TagReference", "RemoteReference", "Reference"] - reference = property( # type: ignore[assignment] - _get_reference, - set_reference, # type: ignore[arg-type] - doc="Returns the Reference we point to", - ) + @property + def reference(self) -> "Reference": + return self._get_reference() + + @reference.setter + def reference(self, ref: Union[AnyGitObject, "SymbolicReference", str]) -> "SymbolicReference": + return self.set_reference(ref) + ref = reference def is_valid(self) -> bool: @@ -606,7 +641,7 @@ def to_full_path(cls, path: Union[PathLike, "SymbolicReference"]) -> PathLike: full_ref_path = path if not cls._common_path_default: return full_ref_path - if not str(path).startswith(cls._common_path_default + "/"): + if not os.fspath(path).startswith(cls._common_path_default + "/"): full_ref_path = "%s/%s" % (cls._common_path_default, path) return full_ref_path @@ -623,7 +658,7 @@ def delete(cls, repo: "Repo", path: PathLike) -> None: Alternatively the symbolic reference to be deleted. """ full_ref_path = cls.to_full_path(path) - abs_path = os.path.join(repo.common_dir, full_ref_path) + abs_path = cls._get_validated_ref_path(repo, full_ref_path) if os.path.exists(abs_path): os.remove(abs_path) else: @@ -686,9 +721,8 @@ def _create( symbolic reference. Otherwise it will be resolved to the corresponding object and a detached symbolic reference will be created instead. """ - git_dir = _git_dir(repo, path) full_ref_path = cls.to_full_path(path) - abs_ref_path = os.path.join(git_dir, full_ref_path) + abs_ref_path = cls._get_validated_ref_path(repo, full_ref_path) # Figure out target data. target = reference @@ -698,7 +732,7 @@ def _create( if not force and os.path.isfile(abs_ref_path): target_data = str(target) if isinstance(target, SymbolicReference): - target_data = str(target.path) + target_data = os.fspath(target.path) if not resolve: target_data = "ref: " + target_data with open(abs_ref_path, "rb") as fd: @@ -780,8 +814,8 @@ def rename(self, new_path: PathLike, force: bool = False) -> "SymbolicReference" if self.path == new_path: return self - new_abs_path = os.path.join(_git_dir(self.repo, new_path), new_path) - cur_abs_path = os.path.join(_git_dir(self.repo, self.path), self.path) + new_abs_path = self._get_validated_ref_path(self.repo, new_path) + cur_abs_path = self._get_validated_ref_path(self.repo, self.path) if os.path.isfile(new_abs_path): if not force: # If they point to the same file, it's not an error. @@ -834,7 +868,7 @@ def _iter_items( # Read packed refs. for _sha, rela_path in cls._iter_packed_refs(repo): - if rela_path.startswith(str(common_path)): + if rela_path.startswith(os.fspath(common_path)): rela_paths.add(rela_path) # END relative path matches common path # END packed refs reading @@ -908,8 +942,7 @@ def from_path(cls: Type[T_References], repo: "Repo", path: PathLike) -> T_Refere SymbolicReference, ): try: - instance: T_References - instance = ref_type(repo, path) + instance = cast(T_References, ref_type(repo, path)) if instance.__class__ is SymbolicReference and instance.is_detached: raise ValueError("SymbolicRef was detached, we drop it") else: @@ -923,4 +956,4 @@ def from_path(cls: Type[T_References], repo: "Repo", path: PathLike) -> T_Refere def is_remote(self) -> bool: """:return: True if this symbolic reference points to a remote branch""" - return str(self.path).startswith(self._remote_common_path_default + "/") + return os.fspath(self.path).startswith(self._remote_common_path_default + "/") diff --git a/git/refs/tag.py b/git/refs/tag.py index 1e38663ae..3a7d946c1 100644 --- a/git/refs/tag.py +++ b/git/refs/tag.py @@ -16,6 +16,7 @@ from typing import Any, TYPE_CHECKING, Type, Union +from git.cmd import Git from git.types import AnyGitObject, PathLike if TYPE_CHECKING: @@ -42,11 +43,13 @@ class TagReference(Reference): __slots__ = () + unsafe_git_tag_options = ["--file", "-F"] + _common_default = "tags" _common_path_default = Reference._common_path_default + "/" + _common_default - @property - def commit(self) -> "Commit": # type: ignore[override] # LazyMixin has unrelated commit method + @property # type: ignore[misc] + def commit(self) -> "Commit": # LazyMixin has unrelated commit method """:return: Commit object the tag ref points to :raise ValueError: @@ -80,8 +83,8 @@ def tag(self) -> Union["TagObject", None]: return None # Make object read-only. It should be reasonably hard to adjust an existing tag. - @property - def object(self) -> AnyGitObject: # type: ignore[override] + @property # type: ignore[misc] + def object(self) -> AnyGitObject: return Reference._get_object(self) @classmethod @@ -92,6 +95,7 @@ def create( reference: Union[str, "SymbolicReference"] = "HEAD", logmsg: Union[str, None] = None, force: bool = False, + allow_unsafe_options: bool = False, **kwargs: Any, ) -> "TagReference": """Create a new tag reference. @@ -121,14 +125,25 @@ def create( :param force: If ``True``, force creation of a tag even though that tag already exists. + :param allow_unsafe_options: + Allow unsafe options, such as ``--file``. + :param kwargs: Additional keyword arguments to be passed to :manpage:`git-tag(1)`. :return: A new :class:`TagReference`. """ - if "ref" in kwargs and kwargs["ref"]: - reference = kwargs["ref"] + legacy_ref = kwargs.pop("ref", None) + if legacy_ref: + reference = legacy_ref + + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([path, reference], kwargs), + unsafe_options=cls.unsafe_git_tag_options, + clusterable_short_options="46adefilnqsv", + ) if "message" in kwargs and kwargs["message"]: kwargs["m"] = kwargs["message"] diff --git a/git/remote.py b/git/remote.py index 20e42b412..e2d5cbc1d 100644 --- a/git/remote.py +++ b/git/remote.py @@ -517,6 +517,9 @@ def iter_items(cls, repo: "Repo", *args: Any, **kwargs: Any) -> NoReturn: # -> raise NotImplementedError +Progress = Union[RemoteProgress, "UpdateProgress", Callable[..., RemoteProgress], None] + + class Remote(LazyMixin, IterableObj): """Provides easy read and write access to a git remote. @@ -805,7 +808,7 @@ def create(cls, repo: "Repo", name: str, url: str, allow_unsafe_protocols: bool """ scmd = "add" kwargs["insert_kwargs_after"] = scmd - url = Git.polish_url(url) + url = Git.polish_url(url, expand_vars=False) if not allow_unsafe_protocols: Git.check_unsafe_protocols(url) repo.git.remote(scmd, "--", name, url, **kwargs) @@ -872,7 +875,7 @@ def update(self, **kwargs: Any) -> "Remote": def _get_fetch_info_from_stderr( self, proc: "Git.AutoInterrupt", - progress: Union[Callable[..., Any], RemoteProgress, None], + progress: Progress, kill_after_timeout: Union[None, float] = None, ) -> IterableList["FetchInfo"]: progress = to_progress_instance(progress) @@ -1000,7 +1003,7 @@ def _assert_refspec(self) -> None: def fetch( self, refspec: Union[str, List[str], None] = None, - progress: Union[RemoteProgress, None, "UpdateProgress"] = None, + progress: Progress = None, verbose: bool = True, kill_after_timeout: Union[None, float] = None, allow_unsafe_protocols: bool = False, @@ -1068,7 +1071,10 @@ def fetch( Git.check_unsafe_protocols(ref) if not allow_unsafe_options: - Git.check_unsafe_options(options=list(kwargs.keys()), unsafe_options=self.unsafe_git_fetch_options) + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), + unsafe_options=self.unsafe_git_fetch_options, + ) proc = self.repo.git.fetch( "--", self, *args, as_process=True, with_stdout=False, universal_newlines=True, v=verbose, **kwargs @@ -1081,7 +1087,7 @@ def fetch( def pull( self, refspec: Union[str, List[str], None] = None, - progress: Union[RemoteProgress, "UpdateProgress", None] = None, + progress: Progress = None, kill_after_timeout: Union[None, float] = None, allow_unsafe_protocols: bool = False, allow_unsafe_options: bool = False, @@ -1122,7 +1128,10 @@ def pull( Git.check_unsafe_protocols(ref) if not allow_unsafe_options: - Git.check_unsafe_options(options=list(kwargs.keys()), unsafe_options=self.unsafe_git_pull_options) + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), + unsafe_options=self.unsafe_git_pull_options, + ) proc = self.repo.git.pull( "--", self, refspec, with_stdout=False, as_process=True, universal_newlines=True, v=True, **kwargs @@ -1135,7 +1144,7 @@ def pull( def push( self, refspec: Union[str, List[str], None] = None, - progress: Union[RemoteProgress, "UpdateProgress", Callable[..., RemoteProgress], None] = None, + progress: Progress = None, kill_after_timeout: Union[None, float] = None, allow_unsafe_protocols: bool = False, allow_unsafe_options: bool = False, @@ -1195,7 +1204,10 @@ def push( Git.check_unsafe_protocols(ref) if not allow_unsafe_options: - Git.check_unsafe_options(options=list(kwargs.keys()), unsafe_options=self.unsafe_git_push_options) + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), + unsafe_options=self.unsafe_git_push_options, + ) proc = self.repo.git.push( "--", diff --git a/git/repo/base.py b/git/repo/base.py index db89cdf41..583e96ca4 100644 --- a/git/repo/base.py +++ b/git/repo/base.py @@ -126,6 +126,7 @@ class Repo: working_dir: PathLike """The working directory of the git command.""" + # stored as string for easier processing, but annotated as path for clearer intention _working_tree_dir: Optional[PathLike] = None git_dir: PathLike @@ -141,6 +142,14 @@ class Repo: re_author_committer_start = re.compile(r"^(author|committer)") re_tab_full_line = re.compile(r"^\t(.*)$") + unsafe_git_init_options = [ + # Can install hooks that execute during later Git commands: + "--template", + # Redirects the repository metadata to a caller-controlled path: + "--separate-git-dir", + ] + """Options to :manpage:`git-init(1)` that permit unsafe code execution or I/O.""" + unsafe_git_clone_options = [ # Executes arbitrary commands: "--upload-pack", @@ -148,8 +157,12 @@ class Repo: # Can override configuration variables that execute arbitrary commands: "--config", "-c", + # Can install hooks that execute during clone: + "--template", + # Fetches from an additional caller-controlled URI: + "--bundle-uri", ] - """Options to :manpage:`git-clone(1)` that allow arbitrary commands to be executed. + """Options to :manpage:`git-clone(1)` that permit unsafe command execution or I/O. The ``--upload-pack``/``-u`` option allows users to execute arbitrary commands directly: @@ -158,8 +171,47 @@ class Repo: The ``--config``/``-c`` option allows users to override configuration variables like ``protocol.allow`` and ``core.gitProxy`` to execute arbitrary commands: https://git-scm.com/docs/git-clone#Documentation/git-clone.txt---configltkeygtltvaluegt + + The ``--template`` option can install hooks that execute during clone: + https://git-scm.com/docs/git-clone#Documentation/git-clone.txt---templatetemplate-directory + + The ``--bundle-uri`` option fetches from an additional URI before fetching from the + clone URL. An untrusted value can therefore make Git access local files or + unintended network resources: + https://git-scm.com/docs/git-clone#Documentation/git-clone.txt---bundle-uriuri """ + unsafe_git_archive_options = [ + # Allows arbitrary command execution through the remote git-upload-archive command. + "--exec", + # Writes output to a caller-controlled filesystem path. + "--output", + "-o", + # Reads from a caller-controlled filesystem path: + "--add-file", + # Injects a caller-controlled path and contents: + "--add-virtual-file", + ] + + unsafe_git_revision_options = [ + # This option allows output to be written to arbitrary files before revision parsing. + "--output", + "-o", + ] + + unsafe_git_blame_options = unsafe_git_revision_options + [ + # These options read from arbitrary files and expose their contents through blame output. + "--contents", + "-S", + "--ignore-revs-file", + ] + + unsafe_git_diff_options = unsafe_git_revision_options + [ + # Reads caller-controlled order patterns from an arbitrary file. + "-O", + "--orderfile", + ] + # Invariants config_level: ConfigLevels_Tup = ("system", "user", "global", "repository") """Represents the configuration level of a configuration file.""" @@ -215,15 +267,13 @@ def __init__( epath = path or os.getenv("GIT_DIR") if not epath: epath = os.getcwd() + epath = os.fspath(epath) if Git.is_cygwin(): # Given how the tests are written, this seems more likely to catch Cygwin # git used from Windows than Windows git used from Cygwin. Therefore # changing to Cygwin-style paths is the relevant operation. - epath = cygpath(str(epath)) + epath = cygpath(epath) - epath = epath or path or os.getcwd() - if not isinstance(epath, str): - epath = str(epath) if expand_vars and re.search(self.re_envvars, epath): warnings.warn( "The use of environment variables in paths is deprecated" @@ -243,6 +293,28 @@ def __init__( # It's important to normalize the paths, as submodules will otherwise # initialize their repo instances with paths that depend on path-portions # that will not exist after being removed. It's just cleaner. + if ( + osp.isfile(osp.join(curpath, "gitdir")) + and osp.isfile(osp.join(curpath, "commondir")) + and osp.isfile(osp.join(curpath, "HEAD")) + ): + git_dir = curpath + + if "GIT_WORK_TREE" in os.environ: + self._working_tree_dir = os.getenv("GIT_WORK_TREE") + else: + # Linked worktree administrative directories store the path to the + # worktree's .git file in their gitdir file (without "gitdir: " prefix). + with open(osp.join(git_dir, "gitdir")) as fp: + worktree_gitfile = fp.read().strip() + + if not osp.isabs(worktree_gitfile): + worktree_gitfile = osp.normpath(osp.join(git_dir, worktree_gitfile)) + + self._working_tree_dir = osp.dirname(worktree_gitfile) + + break + if is_git_dir(curpath): git_dir = curpath # from man git-config : core.worktree @@ -274,7 +346,8 @@ def __init__( sm_gitpath = find_worktree_git_dir(dotgit) if sm_gitpath is not None: - git_dir = expand_path(sm_gitpath, expand_vars) + # worktrees can use relative paths as of Git 2.48, so we join to curpath + git_dir = osp.normpath(osp.join(curpath, sm_gitpath)) self._working_tree_dir = curpath break @@ -354,21 +427,19 @@ def __ne__(self, rhs: object) -> bool: def __hash__(self) -> int: return hash(self.git_dir) - # Description property - def _get_description(self) -> str: + @property + def description(self) -> str: + """The project's description""" filename = osp.join(self.git_dir, "description") with open(filename, "rb") as fp: return fp.read().rstrip().decode(defenc) - def _set_description(self, descr: str) -> None: + @description.setter + def description(self, descr: str) -> None: filename = osp.join(self.git_dir, "description") with open(filename, "wb") as fp: fp.write((descr + "\n").encode(defenc)) - description = property(_get_description, _set_description, doc="the project's description") - del _get_description - del _set_description - @property def working_tree_dir(self) -> Optional[PathLike]: """ @@ -514,7 +585,7 @@ def create_submodule(self, *args: Any, **kwargs: Any) -> Submodule: def iter_submodules(self, *args: Any, **kwargs: Any) -> Iterator[Submodule]: """An iterator yielding Submodule instances. - See the `~git.objects.util.Traversable` interface for a description of `args` + See the :class:`~git.objects.util.Traversable` interface for a description of `args` and `kwargs`. :return: @@ -522,7 +593,7 @@ def iter_submodules(self, *args: Any, **kwargs: Any) -> Iterator[Submodule]: """ return RootModule(self).traverse(*args, **kwargs) - def submodule_update(self, *args: Any, **kwargs: Any) -> Iterator[Submodule]: + def submodule_update(self, *args: Any, **kwargs: Any) -> RootModule: """Update the submodules, keeping the repository consistent as it will take the previous state into consideration. @@ -686,11 +757,7 @@ def _config_reader( git_dir: Optional[PathLike] = None, ) -> GitConfigParser: if config_level is None: - files = [ - self._get_config_path(cast(Lit_config_levels, f), git_dir) - for f in self.config_level - if cast(Lit_config_levels, f) - ] + files = [self._get_config_path(f, git_dir) for f in self.config_level if f] else: files = [self._get_config_path(config_level, git_dir)] return GitConfigParser(files, read_only=True, repo=self) @@ -759,6 +826,7 @@ def iter_commits( self, rev: Union[str, Commit, "SymbolicReference", None] = None, paths: Union[PathLike, Sequence[PathLike]] = "", + allow_unsafe_options: bool = False, **kwargs: Any, ) -> Iterator[Commit]: """An iterator of :class:`~git.objects.commit.Commit` objects representing the @@ -776,6 +844,9 @@ def iter_commits( Arguments to be passed to :manpage:`git-rev-list(1)`. Common ones are ``max_count`` and ``skip``. + :param allow_unsafe_options: + Allow unsafe options in the revision argument, like ``--output``. + :note: To receive only commits between two named revisions, use the ``"revA...revB"`` revision specifier. @@ -786,7 +857,18 @@ def iter_commits( if rev is None: rev = self.head.commit - return Commit.iter_items(self, rev, paths, **kwargs) + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([rev], kwargs), unsafe_options=self.unsafe_git_revision_options + ) + + return Commit.iter_items( + self, + rev, + paths, + allow_unsafe_options=allow_unsafe_options, + **kwargs, + ) def merge_base(self, *rev: TBD, **kwargs: Any) -> List[Commit]: R"""Find the closest common ancestor for the given revision @@ -872,26 +954,31 @@ def is_valid_object(self, sha: str, object_type: Union[str, None] = None) -> boo return False def _get_daemon_export(self) -> bool: - if self.git_dir: - filename = osp.join(self.git_dir, self.DAEMON_EXPORT_FILE) + git_dir = getattr(self, "git_dir", None) + if git_dir is None: + return False + filename = osp.join(git_dir, self.DAEMON_EXPORT_FILE) return osp.exists(filename) def _set_daemon_export(self, value: object) -> None: - if self.git_dir: - filename = osp.join(self.git_dir, self.DAEMON_EXPORT_FILE) + git_dir = getattr(self, "git_dir", None) + if git_dir is None: + return + filename = osp.join(git_dir, self.DAEMON_EXPORT_FILE) fileexists = osp.exists(filename) if value and not fileexists: touch(filename) elif not value and fileexists: os.unlink(filename) - daemon_export = property( - _get_daemon_export, - _set_daemon_export, - doc="If True, git-daemon may export this repository", - ) - del _get_daemon_export - del _set_daemon_export + @property + def daemon_export(self) -> bool: + """If True, git-daemon may export this repository""" + return self._get_daemon_export() + + @daemon_export.setter + def daemon_export(self, value: object) -> None: + self._set_daemon_export(value) def _get_alternates(self) -> List[str]: """The list of alternates for this repo from which objects can be retrieved. @@ -899,8 +986,7 @@ def _get_alternates(self) -> List[str]: :return: List of strings being pathnames of alternates """ - if self.git_dir: - alternates_path = osp.join(self.git_dir, "objects", "info", "alternates") + alternates_path = osp.join(self.common_dir, "objects", "info", "alternates") if osp.exists(alternates_path): with open(alternates_path, "rb") as f: @@ -929,11 +1015,14 @@ def _set_alternates(self, alts: List[str]) -> None: with open(alternates_path, "wb") as f: f.write("\n".join(alts).encode(defenc)) - alternates = property( - _get_alternates, - _set_alternates, - doc="Retrieve a list of alternates paths or set a list paths to be used as alternates", - ) + @property + def alternates(self) -> List[str]: + """Retrieve a list of alternates paths or set a list paths to be used as alternates""" + return self._get_alternates() + + @alternates.setter + def alternates(self, alts: List[str]) -> None: + self._set_alternates(alts) def is_dirty( self, @@ -959,7 +1048,7 @@ def is_dirty( if not submodules: default_args.append("--ignore-submodules") if path: - default_args.extend(["--", str(path)]) + default_args.extend(["--", os.fspath(path)]) if index: # diff index against HEAD. if osp.isfile(self.index.path) and len(self.git.diff("--cached", *default_args)): @@ -1045,13 +1134,23 @@ def active_branch(self) -> Head: :raise TypeError: If HEAD is detached. + :raise ValueError: + If HEAD points to the ``.invalid`` ref Git uses to mark refs as + incompatible with older clients. + :return: :class:`~git.refs.head.Head` to the active branch """ - # reveal_type(self.head.reference) # => Reference - return self.head.reference + active_branch = self.head.reference + if active_branch.name == ".invalid": + raise ValueError( + "HEAD points to 'refs/heads/.invalid', which Git uses to mark refs as incompatible with older clients" + ) + return active_branch - def blame_incremental(self, rev: str | HEAD | None, file: str, **kwargs: Any) -> Iterator["BlameEntry"]: + def blame_incremental( + self, rev: str | HEAD | None, file: str, allow_unsafe_options: bool = False, **kwargs: Any + ) -> Iterator["BlameEntry"]: """Iterator for blame information for the given file at the given revision. Unlike :meth:`blame`, this does not return the actual file's contents, only a @@ -1062,6 +1161,9 @@ def blame_incremental(self, rev: str | HEAD | None, file: str, **kwargs: Any) -> uncommitted changes. Otherwise, anything successfully parsed by :manpage:`git-rev-parse(1)` is a valid option. + :param allow_unsafe_options: + Allow unsafe options in revision argument, like ``--output`` or ``--contents``. + :return: Lazy iterator of :class:`BlameEntry` tuples, where the commit indicates the commit to blame for the line, and range indicates a span of line numbers in @@ -1070,6 +1172,12 @@ def blame_incremental(self, rev: str | HEAD | None, file: str, **kwargs: Any) -> If you combine all line number ranges outputted by this command, you should get a continuous range spanning all line numbers in the file. """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([rev], kwargs), + unsafe_options=self.unsafe_git_blame_options, + clusterable_short_options="46bceflnpqstvw", + ) data: bytes = self.git.blame(rev, "--", file, p=True, incremental=True, stdout_as_string=False, **kwargs) commits: Dict[bytes, Commit] = {} @@ -1148,7 +1256,8 @@ def blame( rev: Union[str, HEAD, None], file: str, incremental: bool = False, - rev_opts: Optional[List[str]] = None, + rev_opts: Optional[Sequence[str]] = None, + allow_unsafe_options: bool = False, **kwargs: Any, ) -> List[List[Commit | List[str | bytes] | None]] | Iterator[BlameEntry] | None: """The blame information for the given file at the given revision. @@ -1158,6 +1267,9 @@ def blame( uncommitted changes. Otherwise, anything successfully parsed by :manpage:`git-rev-parse(1)` is a valid option. + :param allow_unsafe_options: + Allow unsafe options in revision argument, like ``--output`` or ``--contents``. + :return: list: [git.Commit, list: []] @@ -1167,9 +1279,15 @@ def blame( appearance. """ if incremental: - return self.blame_incremental(rev, file, **kwargs) - rev_opts = rev_opts or [] - data: bytes = self.git.blame(rev, *rev_opts, "--", file, p=True, stdout_as_string=False, **kwargs) + return self.blame_incremental(rev, file, allow_unsafe_options=allow_unsafe_options, **kwargs) + rev_opts_list = list(rev_opts or []) + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([rev, rev_opts_list], kwargs), + unsafe_options=self.unsafe_git_blame_options, + clusterable_short_options="46bceflnpqstvw", + ) + data: bytes = self.git.blame(rev, *rev_opts_list, "--", file, p=True, stdout_as_string=False, **kwargs) commits: Dict[str, Commit] = {} blames: List[List[Commit | List[str | bytes] | None]] = [] @@ -1189,6 +1307,7 @@ class InfoTD(TypedDict, total=False): keepends = True for line_bytes in data.splitlines(keepends): + line_str = "" try: line_str = line_bytes.rstrip().decode(defenc) except UnicodeDecodeError: @@ -1303,6 +1422,7 @@ def init( mkdir: bool = True, odbt: Type[GitCmdObjectDB] = GitCmdObjectDB, expand_vars: bool = True, + allow_unsafe_options: bool = False, **kwargs: Any, ) -> "Repo": """Initialize a git repository at the given path if specified. @@ -1327,6 +1447,10 @@ def init( information disclosure, allowing attackers to access the contents of environment variables. + :param allow_unsafe_options: + Allow unsafe options to be used, such as ``--template`` and + ``--separate-git-dir``. + :param kwargs: Keyword arguments serving as additional options to the :manpage:`git-init(1)` command. @@ -1334,6 +1458,11 @@ def init( :return: :class:`Repo` (the newly created repo) """ + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), + unsafe_options=cls.unsafe_git_init_options, + ) if path: path = expand_path(path, expand_vars) if mkdir and path and not osp.exists(path): @@ -1359,9 +1488,9 @@ def _clone( ) -> "Repo": odbt = kwargs.pop("odbt", odb_default_type) - # When pathlib.Path or other class-based path is passed - if not isinstance(path, str): - path = str(path) + # url may be a path and this has no effect if it is a string + url = os.fspath(url) + path = os.fspath(path) ## A bug win cygwin's Git, when `--bare` or `--separate-git-dir` # it prepends the cwd or(?) the `url` into the `path, so:: @@ -1377,17 +1506,21 @@ def _clone( if multi_options: multi = shlex.split(" ".join(multi_options)) + clone_url = Git.polish_url(url, expand_vars=False) if not allow_unsafe_protocols: - Git.check_unsafe_protocols(str(url)) + Git.check_unsafe_protocols(clone_url) if not allow_unsafe_options: - Git.check_unsafe_options(options=list(kwargs.keys()), unsafe_options=cls.unsafe_git_clone_options) - if not allow_unsafe_options and multi_options: - Git.check_unsafe_options(options=multi_options, unsafe_options=cls.unsafe_git_clone_options) + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), + unsafe_options=cls.unsafe_git_clone_options, + ) + if not allow_unsafe_options and multi: + Git.check_unsafe_options(options=multi, unsafe_options=cls.unsafe_git_clone_options) proc = git.clone( multi, "--", - Git.polish_url(str(url)), + clone_url, clone_path, with_extended_output=True, as_process=True, @@ -1427,7 +1560,7 @@ def _clone( # escape the backslashes. Hence we undo the escaping just to be sure. if repo.remotes: with repo.remotes[0].config_writer as writer: - writer.set_value("url", Git.polish_url(repo.remotes[0].url)) + writer.set_value("url", Git.polish_url(repo.remotes[0].url, expand_vars=False)) # END handle remote repo return repo @@ -1482,7 +1615,7 @@ def clone( self.common_dir, path, type(self.odb), - progress, + progress, # type: ignore[arg-type] multi_options, allow_unsafe_protocols=allow_unsafe_protocols, allow_unsafe_options=allow_unsafe_options, @@ -1543,7 +1676,7 @@ def clone_from( url, to_path, GitCmdObjectDB, - progress, + progress, # type: ignore[arg-type] multi_options, allow_unsafe_protocols=allow_unsafe_protocols, allow_unsafe_options=allow_unsafe_options, @@ -1555,6 +1688,8 @@ def archive( ostream: Union[TextIO, BinaryIO], treeish: Optional[str] = None, prefix: Optional[str] = None, + allow_unsafe_options: bool = False, + allow_unsafe_protocols: bool = False, **kwargs: Any, ) -> Repo: """Archive the tree at the given revision. @@ -1577,6 +1712,12 @@ def archive( repository-relative path to a directory or file to place into the archive, or a list or tuple of multiple paths. + :param allow_unsafe_options: + Allow unsafe options, like ``--exec`` or ``--output``. + + :param allow_unsafe_protocols: + Allow unsafe protocols to be used in ``remote``, like ``ext``. + :raise git.exc.GitCommandError: If something went wrong. @@ -1587,6 +1728,14 @@ def archive( treeish = self.head.commit if prefix and "prefix" not in kwargs: kwargs["prefix"] = prefix + remote = kwargs.get("remote") + if not allow_unsafe_protocols and remote is not None: + Git.check_unsafe_protocols(str(remote)) + if not allow_unsafe_options: + Git.check_unsafe_options( + options=Git._option_candidates([], kwargs), + unsafe_options=self.unsafe_git_archive_options, + ) kwargs["output_stream"] = ostream path = kwargs.pop("path", []) path = cast(Union[PathLike, List[PathLike], Tuple[PathLike, ...]], path) @@ -1627,8 +1776,9 @@ def currently_rebasing_on(self) -> Commit | None: ``None`` if we are not currently rebasing. """ - if self.git_dir: - rebase_head_file = osp.join(self.git_dir, "REBASE_HEAD") + if not self.git_dir: + return None + rebase_head_file = osp.join(self.git_dir, "REBASE_HEAD") if not osp.isfile(rebase_head_file): return None with open(rebase_head_file, "rt") as f: diff --git a/git/repo/fun.py b/git/repo/fun.py index 125ba5936..66e7eba69 100644 --- a/git/repo/fun.py +++ b/git/repo/fun.py @@ -20,6 +20,7 @@ import os import os.path as osp from pathlib import Path +import re import stat from string import digits @@ -28,19 +29,21 @@ from git.cmd import Git from git.exc import WorkTreeRepositoryUnsupported from git.objects import Object +from git.objects.util import parse_date from git.refs import SymbolicReference from git.util import cygpath, bin_to_hex, hex_to_bin # Typing ---------------------------------------------------------------------- -from typing import Optional, TYPE_CHECKING, Union, cast, overload +from typing import Iterator, Optional, TYPE_CHECKING, Tuple, Union, cast, overload from git.types import AnyGitObject, Literal, PathLike if TYPE_CHECKING: from git.db import GitCmdObjectDB - from git.objects import Commit, TagObject + from git.objects import Commit from git.refs.reference import Reference + from git.refs.log import RefLog, RefLogEntry from git.refs.tag import Tag from .base import Repo @@ -139,6 +142,23 @@ def short_to_long(odb: "GitCmdObjectDB", hexsha: str) -> Optional[bytes]: # END exception handling +def _describe_to_long(repo: "Repo", name: str) -> Optional[bytes]: + """Resolve git-describe style names to the abbreviated object they contain.""" + match = re.match(r"^.+-\d+-g([0-9A-Fa-f]{4,40})(?:-dirty)?$", name) + if match is None: + match = re.match(r"^.+-g([0-9A-Fa-f]{4,40})(?:-dirty)?$", name) + if match is None: + match = re.match(r"^([0-9A-Fa-f]{4,40})-dirty$", name) + if match is None: + return None + # END handle match + + hexsha = match.group(1) + if len(hexsha) == 40: + return hexsha.encode("ascii") + return short_to_long(repo.odb, hexsha) + + @overload def name_to_object(repo: "Repo", name: str, return_ref: Literal[False] = ...) -> AnyGitObject: ... @@ -192,6 +212,10 @@ def name_to_object(repo: "Repo", name: str, return_ref: bool = False) -> Union[A # END for each base # END handle hexsha + if hexsha is None: + hexsha = _describe_to_long(repo, name) + # END handle describe output + # Didn't find any ref, this is an error. if return_ref: raise BadObject("Couldn't find reference named %r" % name) @@ -227,6 +251,361 @@ def to_commit(obj: Object) -> "Commit": return obj +def _object_from_hexsha(repo: "Repo", hexsha: str) -> AnyGitObject: + return Object.new_from_sha(repo, hex_to_bin(hexsha)) + + +def _current_reflog_ref(repo: "Repo") -> SymbolicReference: + try: + return repo.head.ref + except TypeError: + return repo.head + # END handle detached head + + +def _common_reflog_path(repo: "Repo", ref: SymbolicReference) -> Optional[str]: + if repo.common_dir == repo.git_dir: + return None + # END handle normal repository + return SymbolicReference._get_validated_path(osp.join(repo.common_dir, "logs"), ref.path) + + +def _ref_log(repo: "Repo", ref: SymbolicReference) -> "RefLog": + try: + return ref.log() + except FileNotFoundError: + common_path = _common_reflog_path(repo, ref) + if common_path and osp.isfile(common_path): + from git.refs.log import RefLog + + return RefLog.from_file(common_path) + # END handle linked-worktree branch logs + try: + if ref.path == repo.head.ref.path: + return repo.head.log() + # END handle linked-worktree current branch logs + except TypeError: + pass + # END handle detached head + raise + # END handle missing branch log + + +def _ref_log_entry(repo: "Repo", ref: SymbolicReference, index: int) -> "RefLogEntry": + try: + return ref.log_entry(index) + except FileNotFoundError: + common_path = _common_reflog_path(repo, ref) + if common_path and osp.isfile(common_path): + from git.refs.log import RefLog + + return RefLog.entry_at(common_path, index) + # END handle linked-worktree branch logs + try: + if ref.path == repo.head.ref.path: + return repo.head.log_entry(index) + # END handle linked-worktree current branch logs + except TypeError: + pass + # END handle detached head + raise + # END handle missing branch log + + +def _find_reflog_entry_by_date(repo: "Repo", ref: SymbolicReference, spec: str) -> str: + try: + timestamp, _offset = parse_date(spec) + except ValueError as e: + raise NotImplementedError("Support for additional @{...} modes not implemented") from e + # END handle unsupported dates + log = _ref_log(repo, ref) + if not log: + raise IndexError("Invalid revlog date: %s" % spec) + # END handle empty log + + for entry in reversed(log): + if entry.time[0] <= timestamp: + return entry.newhexsha + # END found candidate + # END for each entry + return log[0].newhexsha + + +def _previous_checked_out_branch(repo: "Repo", nth: int) -> AnyGitObject: + if nth <= 0: + raise ValueError("Invalid previous checkout selector: -%i" % nth) + # END handle invalid input + + seen = 0 + for entry in reversed(_ref_log(repo, repo.head)): + message = entry.message or "" + prefix = "checkout: moving from " + if not message.startswith(prefix): + continue + # END skip non-checkouts + + previous_branch = message[len(prefix) :].split(" to ", 1)[0] + seen += 1 + if seen == nth: + return name_to_object(repo, previous_branch) + # END found selector + # END for each entry + raise IndexError("Invalid previous checkout selector: -%i" % nth) + + +def _tracking_branch_object(repo: "Repo", ref: Optional[SymbolicReference]) -> AnyGitObject: + from git.refs.head import Head + + if ref is None: + try: + head = repo.active_branch + except TypeError as e: + raise BadName("@{upstream}") from e + elif isinstance(ref, Head): + head = ref + elif os.fspath(ref.path).startswith("refs/heads/"): + head = Head(repo, ref.path) + else: + raise BadName("%s@{upstream}" % ref.name) + # END handle head + + tracking_branch = head.tracking_branch() + if tracking_branch is None: + raise BadName("%s@{upstream}" % head.name) + # END handle missing upstream + return tracking_branch.commit + + +def _apply_reflog(repo: "Repo", ref: Optional[SymbolicReference], content: str) -> AnyGitObject: + if content.startswith("+"): + content = content[1:] + # END handle explicit positive sign + + if content.startswith("-"): + if ref is not None: + raise ValueError("Previous checkout selectors do not take an explicit ref") + if content == "-0": + raise ValueError("Negative zero is invalid in reflog selector") + # END handle invalid negative zero + try: + return _previous_checked_out_branch(repo, int(content[1:])) + except ValueError as e: + raise ValueError("Invalid previous checkout selector: %s" % content) from e + # END handle previous checkout branch + + content_lower = content.lower() + if content_lower in ("u", "upstream", "push"): + return _tracking_branch_object(repo, ref) + # END handle sibling branches + + ref = ref or _current_reflog_ref(repo) + try: + entry_no = int(content) + except ValueError: + hexsha = _find_reflog_entry_by_date(repo, ref, content) + else: + if entry_no >= 100000000: + hexsha = _find_reflog_entry_by_date(repo, ref, "%s +0000" % entry_no) + elif entry_no == 0: + return ref.commit + else: + try: + entry = _ref_log_entry(repo, ref, -(entry_no + 1)) + except IndexError as e: + raise IndexError("Invalid revlog index: %i" % entry_no) from e + # END handle index out of bound + hexsha = entry.newhexsha + # END handle offset or date-like timestamp + # END handle content + return _object_from_hexsha(repo, hexsha) + + +def _find_closing_brace(rev: str, start: int) -> int: + depth = 1 + escaped = False + for idx in range(start + 1, len(rev)): + char = rev[idx] + if escaped: + escaped = False + elif char == "\\": + escaped = True + elif char == "{": + depth += 1 + elif char == "}": + depth -= 1 + if depth == 0: + return idx + # END found end + # END handle char + # END for each char + raise ValueError("Missing closing brace to define type in %s" % rev) + + +def _parse_search(pattern: str) -> Tuple[str, bool]: + if not pattern: + raise ValueError("Revision search requires a pattern") + # END handle empty pattern + + if pattern.startswith("!-"): + return pattern[2:], True + if pattern.startswith("!!"): + return pattern[1:], False + if pattern.startswith("!"): + raise ValueError("Need one character after /!, typically -") + return pattern, False + + +def _unescape_braced_regex(pattern: str) -> str: + out = [] + idx = 0 + while idx < len(pattern): + char = pattern[idx] + if char == "\\" and idx + 1 < len(pattern): + next_char = pattern[idx + 1] + if next_char in "{}\\": + out.append(next_char) + else: + out.append(char) + out.append(next_char) + # END handle escaped char + idx += 2 + continue + # END handle backslash + out.append(char) + idx += 1 + # END for each char + return "".join(out) + + +def _find_commit_by_message( + repo: "Repo", rev: Optional[AnyGitObject], pattern: str, braced: bool = False +) -> AnyGitObject: + pattern, negated = _parse_search(_unescape_braced_regex(pattern) if braced else pattern) + try: + regex = re.compile(pattern) + except re.error as e: + raise ValueError("Invalid commit message regex %r" % pattern) from e + # END handle invalid regex + if rev is None: + commits = _all_ref_commits(repo) + else: + commits = _reachable_commits([to_commit(cast(Object, rev))]) + # END handle starting point + + for commit in commits: + message = commit.message + if isinstance(message, bytes): + message = message.decode(commit.encoding, "replace") + # END handle bytes message + matches = regex.search(message or "") is not None + if matches != negated: + return commit + # END found commit + # END for each commit + raise BadName("No commit found matching message pattern %r" % pattern) + + +def _all_ref_commits(repo: "Repo") -> Iterator["Commit"]: + starts = [] + for ref in repo.references: + try: + starts.append(to_commit(cast(Object, ref.object))) + except (BadName, ValueError): + pass + # END skip refs that do not point to commits + # END for each ref + try: + starts.append(repo.head.commit) + except ValueError: + pass + # END handle unborn head + return _reachable_commits(starts) + + +def _reachable_commits(starts: list["Commit"]) -> Iterator["Commit"]: + seen = set() + pending = starts[:] + while pending: + pending.sort(key=lambda commit: commit.committed_date, reverse=True) + commit = pending.pop(0) + if commit.binsha in seen: + continue + # END skip seen commit + seen.add(commit.binsha) + yield commit + pending.extend(commit.parents) + # END while commits remain + + +def _index_lookup(repo: "Repo", spec: str) -> AnyGitObject: + if not spec: + raise ValueError("':' must be followed by a path") + # END handle empty lookup + + stage = 0 + path = spec + if len(spec) >= 2 and spec[1] == ":" and spec[0] in "0123": + stage = int(spec[0]) + path = spec[2:] + # END handle stage + + try: + return repo.index.entries[(path, stage)].to_blob(repo) + except KeyError as e: + raise BadName("Path %r did not exist in the index at stage %i" % (path, stage)) from e + + +def _tree_lookup(obj: AnyGitObject, path: str) -> AnyGitObject: + if obj.type != "tree": + obj = to_commit(cast(Object, obj)).tree + # END get tree + if not path: + return obj + return obj[path] + + +def _peel(obj: AnyGitObject, output_type: str, repo: "Repo", rev: str) -> AnyGitObject: + if output_type.startswith("/"): + return _find_commit_by_message(repo, obj, output_type[1:], braced=True) + if output_type == "": + return deref_tag(obj) if obj.type == "tag" else obj + if output_type == "object": + return obj + if output_type == "commit": + return to_commit(cast(Object, obj)) + if output_type == "tree": + return to_commit(cast(Object, obj)).tree if obj.type != "tree" else obj + if output_type == "blob": + obj = deref_tag(obj) if obj.type == "tag" else obj + if obj.type == output_type: + return obj + # END handle matching type + raise ValueError("Could not accommodate requested object type %r, got %s" % (output_type, obj.type)) + if output_type == "tag": + if obj.type == output_type: + return obj + # END handle matching type + raise ValueError("Could not accommodate requested object type %r, got %s" % (output_type, obj.type)) + # END handle known types + raise ValueError("Invalid output type: %s ( in %s )" % (output_type, rev)) + + +def _first_rev_token(rev: str) -> Optional[int]: + for idx, char in enumerate(rev): + if char in "^~:": + return idx + if char == "@": + next_char = rev[idx + 1] if idx + 1 < len(rev) else None + if idx == 0 and next_char in (None, "^", "~", ":", "{"): + return idx + if next_char == "{": + return idx + # END handle reflog selector + # END handle at symbol + # END for each char + return None + + def rev_parse(repo: "Repo", rev: str) -> AnyGitObject: """Parse a revision string. Like :manpage:`git-rev-parse(1)`. @@ -253,137 +632,82 @@ def rev_parse(repo: "Repo", rev: str) -> AnyGitObject: :raise IndexError: If an invalid reflog index is specified. """ - # Are we in colon search mode? if rev.startswith(":/"): - # Colon search mode - raise NotImplementedError("commit by message search (regex)") - # END handle search + return _find_commit_by_message(repo, None, rev[2:]) + if rev.startswith(":"): + return _index_lookup(repo, rev[1:]) + # END handle top-level colon modes obj: Optional[AnyGitObject] = None ref = None - output_type = "commit" - start = 0 - parsed_to = 0 lr = len(rev) - while start < lr: - if rev[start] not in "^~:@": - start += 1 - continue - # END handle start + first_token = _first_rev_token(rev) + if first_token is None: + return name_to_object(repo, rev) + # END handle plain name + + if first_token == 0: + if rev[0] != "@": + raise ValueError("Revision specifier must start with an object name: %s" % rev) + # END handle invalid leading token + ref = _current_reflog_ref(repo) + obj = ref.commit + start = 0 if rev.startswith("@{") else 1 + else: + if rev[first_token] == "@": + ref = cast("Reference", name_to_object(repo, rev[:first_token], return_ref=True)) + obj = ref.commit + else: + obj = name_to_object(repo, rev[:first_token]) + # END handle anchor + start = first_token + # END initialize anchor + while start < lr: token = rev[start] - if obj is None: - # token is a rev name. - if start == 0: - ref = repo.head.ref - else: - if token == "@": - ref = cast("Reference", name_to_object(repo, rev[:start], return_ref=True)) - else: - obj = name_to_object(repo, rev[:start]) - # END handle token - # END handle refname - else: - if ref is not None: - obj = cast("Commit", ref.commit) - # END handle ref - # END initialize obj on first token + if token == "@": + if start + 1 >= lr or rev[start + 1] != "{": + raise ValueError("Invalid @ token in revision specifier: %s" % rev) + # END handle invalid @ + end = _find_closing_brace(rev, start + 1) + obj = _apply_reflog(repo, ref if first_token != 0 and start == first_token else None, rev[start + 2 : end]) + ref = None + start = end + 1 + continue + # END handle reflog - start += 1 + if token == ":": + return _tree_lookup(obj, rev[start + 1 :]) + # END handle path - # Try to parse {type}. - if start < lr and rev[start] == "{": - end = rev.find("}", start) - if end == -1: - raise ValueError("Missing closing brace to define type in %s" % rev) - output_type = rev[start + 1 : end] # Exclude brace. - - # Handle type. - if output_type == "commit": - obj = cast("TagObject", obj) - if obj and obj.type == "tag": - obj = deref_tag(obj) - else: - # Cannot do anything for non-tags. - pass - # END handle tag - elif output_type == "tree": - try: - obj = cast(AnyGitObject, obj) - obj = to_commit(obj).tree - except (AttributeError, ValueError): - pass # Error raised later. - # END exception handling - elif output_type in ("", "blob"): - obj = cast("TagObject", obj) - if obj and obj.type == "tag": - obj = deref_tag(obj) - else: - # Cannot do anything for non-tags. - pass - # END handle tag - elif token == "@": - # try single int - assert ref is not None, "Require Reference to access reflog" - revlog_index = None - try: - # Transform reversed index into the format of our revlog. - revlog_index = -(int(output_type) + 1) - except ValueError as e: - # TODO: Try to parse the other date options, using parse_date maybe. - raise NotImplementedError("Support for additional @{...} modes not implemented") from e - # END handle revlog index - - try: - entry = ref.log_entry(revlog_index) - except IndexError as e: - raise IndexError("Invalid revlog index: %i" % revlog_index) from e - # END handle index out of bound - - obj = Object.new_from_sha(repo, hex_to_bin(entry.newhexsha)) - - # Make it pass the following checks. - output_type = "" - else: - raise ValueError("Invalid output type: %s ( in %s )" % (output_type, rev)) - # END handle output type - - # Empty output types don't require any specific type, its just about - # dereferencing tags. - if output_type and obj and obj.type != output_type: - raise ValueError("Could not accommodate requested object type %r, got %s" % (output_type, obj.type)) - # END verify output type + start += 1 - start = end + 1 # Skip brace. - parsed_to = start + if token == "^" and start < lr and rev[start] == "{": + end = _find_closing_brace(rev, start) + obj = _peel(obj, rev[start + 1 : end], repo, rev) + ref = None + start = end + 1 continue # END parse type - # Try to parse a number. num = 0 - if token != ":": - found_digit = False - while start < lr: - if rev[start] in digits: - num = num * 10 + int(rev[start]) - start += 1 - found_digit = True - else: - break - # END handle number - # END number parse loop - - # No explicit number given, 1 is the default. It could be 0 though. - if not found_digit: - num = 1 - # END set default num - # END number parsing only if non-blob mode - - parsed_to = start - # Handle hierarchy walk. + found_digit = False + while start < lr: + if rev[start] in digits: + num = num * 10 + int(rev[start]) + start += 1 + found_digit = True + else: + break + # END handle number + # END number parse loop + + if not found_digit: + num = 1 + # END set default num + try: - obj = cast(AnyGitObject, obj) if token == "~": obj = to_commit(obj) for _ in range(num): @@ -391,35 +715,22 @@ def rev_parse(repo: "Repo", rev: str) -> AnyGitObject: # END for each history item to walk elif token == "^": obj = to_commit(obj) - # Must be n'th parent. - if num: + if num == 0: + pass + else: obj = obj.parents[num - 1] - elif token == ":": - if obj.type != "tree": - obj = obj.tree - # END get tree type - obj = obj[rev[start:]] - parsed_to = lr + # END handle parent else: raise ValueError("Invalid token: %r" % token) # END end handle tag except (IndexError, AttributeError) as e: raise BadName( - f"Invalid revision spec '{rev}' - not enough " f"parent commits to reach '{token}{int(num)}'" + f"Invalid revision spec '{rev}' - not enough parent commits to reach '{token}{int(num)}'" ) from e # END exception handling # END parse loop - # Still no obj? It's probably a simple name. - if obj is None: - obj = name_to_object(repo, rev) - parsed_to = lr - # END handle simple name - if obj is None: raise ValueError("Revision specifier could not be parsed: %s" % rev) - if parsed_to != lr: - raise ValueError("Didn't consume complete rev spec %s, consumed part: %s" % (rev, rev[:parsed_to])) - return obj diff --git a/git/types.py b/git/types.py index cce184530..100fff43f 100644 --- a/git/types.py +++ b/git/types.py @@ -13,7 +13,6 @@ Sequence as Sequence, Tuple, TYPE_CHECKING, - Type, TypeVar, Union, ) @@ -130,7 +129,8 @@ https://git-scm.com/docs/gitglossary#def_object_type """ -Lit_commit_ish: Type[Literal["commit", "tag"]] +if TYPE_CHECKING: + Lit_commit_ish = Literal["commit", "tag"] """Deprecated. Type of literal strings identifying typically-commitish git object types. Prior to a bugfix, this type had been defined more broadly. Any usage is in practice diff --git a/git/util.py b/git/util.py index 9e8ac821d..02f57c132 100644 --- a/git/util.py +++ b/git/util.py @@ -36,7 +36,7 @@ import logging import os import os.path as osp -import pathlib +from pathlib import Path import platform import re import shutil @@ -217,7 +217,7 @@ def rmtree(path: PathLike) -> None: couldn't be deleted are read-only. Windows will not remove them in that case. """ - def handler(function: Callable, path: PathLike, _excinfo: Any) -> None: + def handler(function: Callable[[str], Any], path: str, _excinfo: Any) -> None: """Callback for :func:`shutil.rmtree`. This works as either a ``onexc`` or ``onerror`` style callback. @@ -272,9 +272,9 @@ def stream_copy(source: BinaryIO, destination: BinaryIO, chunk_size: int = 512 * def join_path(a: PathLike, *p: PathLike) -> PathLike: R"""Join path tokens together similar to osp.join, but always use ``/`` instead of possibly ``\`` on Windows.""" - path = str(a) + path = os.fspath(a) for b in p: - b = str(b) + b = os.fspath(b) if not b: continue if b.startswith("/"): @@ -289,19 +289,19 @@ def join_path(a: PathLike, *p: PathLike) -> PathLike: if sys.platform == "win32": - def to_native_path_windows(path: PathLike) -> PathLike: - path = str(path) + def to_native_path_windows(path: PathLike) -> str: + path = os.fspath(path) return path.replace("/", "\\") def to_native_path_linux(path: PathLike) -> str: - path = str(path) + path = os.fspath(path) return path.replace("\\", "/") to_native_path = to_native_path_windows else: # No need for any work on Linux. def to_native_path_linux(path: PathLike) -> str: - return str(path) + return os.fspath(path) to_native_path = to_native_path_linux @@ -315,6 +315,68 @@ def join_path_native(a: PathLike, *p: PathLike) -> PathLike: return to_native_path(join_path(a, *p)) +def _is_path_rooted(path: PathLike) -> bool: + r"""Whether ``path`` has a root, including one encoded in a UNC drive. + + On Windows, ``\directory`` is rooted on the current drive without being + absolute, while ``C:\directory`` has both a drive and a root. In contrast, + ``directory`` and the drive-relative ``C:directory`` have no root. + UNC paths are rooted: ``\\server\share`` stores the share in the drive + returned by :func:`os.path.splitdrive`, while ``\\server\share\directory`` + additionally has a rooted tail. + On POSIX, which has no drive concept, this simply distinguishes absolute + paths such as ``/directory`` from relative paths such as ``directory``. + """ + drive, tail = osp.splitdrive(os.fspath(path)) + separators = (os.sep,) if os.altsep is None else (os.sep, os.altsep) + return tail.startswith(separators) or drive.startswith(separators) + + +def _to_relative_path(root: PathLike, path: PathLike) -> str: + r"""Return a normalized Git-style path confined to ``root``. + + A Windows path such as ``\directory`` is rooted but not absolute. Resolve it + against the drive of ``root`` rather than treating it as relative to ``root``. + Drive-relative paths such as ``C:directory`` are rejected because their meaning + depends on process-global per-drive state. + + For example, with ``root`` set to ``C:\repo`` on Windows: + + * ``directory\file`` -> ``directory/file`` + * ``directory\`` -> ``directory/`` + * ``C:\repo\directory\file`` -> ``directory/file`` + * ``\repo\directory\file`` -> ``directory/file`` + * ``C:directory\file`` -> :exc:`ValueError` + * ``C:\other\file`` -> :exc:`ValueError` + + On POSIX, ``/repo/directory/file`` under ``/repo`` similarly becomes + ``directory/file``. A trailing separator is preserved as a Git-style ``/``. + """ + path_str = os.fspath(path) + if not path_str: + return path_str + + drive, _tail = osp.splitdrive(path_str) + rooted = _is_path_rooted(path_str) + if drive and not rooted: + raise ValueError("Drive-relative path %r is not supported" % path_str) + + root_abs = osp.abspath(os.fspath(root)) + path_abs = osp.abspath(osp.join(root_abs, path_str)) + try: + common_path = osp.commonpath([root_abs, path_abs]) + except ValueError as e: + raise ValueError("Path %r is not in repository at %r" % (path_str, root_abs)) from e + if common_path != root_abs: + raise ValueError("Path %r is not in repository at %r" % (path_str, root_abs)) + + relative_path = to_native_path_linux(osp.relpath(path_abs, root_abs)) + separators = (os.sep,) if os.altsep is None else (os.sep, os.altsep) + if path_str.endswith(separators) and relative_path != "." and not relative_path.endswith("/"): + relative_path += "/" + return relative_path + + def assure_directory_exists(path: PathLike, is_file: bool = False) -> bool: """Make sure that the directory pointed to by path exists. @@ -372,7 +434,7 @@ def is_exec(fpath: str) -> bool: progs = [] if not path: path = os.environ["PATH"] - for folder in str(path).split(os.pathsep): + for folder in os.fspath(path).split(os.pathsep): folder = folder.strip('"') if folder: exe_path = osp.join(folder, program) @@ -382,13 +444,13 @@ def is_exec(fpath: str) -> bool: return progs -def _cygexpath(drive: Optional[str], path: str) -> str: +def _cygexpath(drive: Optional[str], path: str, expand_vars: bool = True) -> str: if osp.isabs(path) and not drive: # Invoked from `cygpath()` directly with `D:Apps\123`? # It's an error, leave it alone just slashes) p = path # convert to str if AnyPath given else: - p = path and osp.normpath(osp.expandvars(osp.expanduser(path))) + p = path and osp.normpath(osp.expandvars(osp.expanduser(path)) if expand_vars else path) if osp.isabs(p): if drive: # Confusing, maybe a remote system should expand vars. @@ -397,11 +459,11 @@ def _cygexpath(drive: Optional[str], path: str) -> str: p = cygpath(p) elif drive: p = "/proc/cygdrive/%s/%s" % (drive.lower(), p) - p_str = str(p) # ensure it is a str and not AnyPath + p_str = os.fspath(p) # ensure it is a str and not AnyPath return p_str.replace("\\", "/") -_cygpath_parsers: Tuple[Tuple[Pattern[str], Callable, bool], ...] = ( +_cygpath_parsers: Tuple[Tuple[Pattern[str], Callable[..., str], bool], ...] = ( # See: https://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx # and: https://www.cygwin.com/cygwin-ug-net/using.html#unc-paths ( @@ -416,20 +478,23 @@ def _cygexpath(drive: Optional[str], path: str) -> str: ) -def cygpath(path: str) -> str: +def cygpath(path: str, expand_vars: bool = True) -> str: """Use :meth:`git.cmd.Git.polish_url` instead, that works on any environment.""" - path = str(path) # Ensure is str and not AnyPath. + path = os.fspath(path) # Ensure is str and not AnyPath. # Fix to use Paths when 3.5 dropped. Or to be just str if only for URLs? if not path.startswith(("/cygdrive", "//", "/proc/cygdrive")): for regex, parser, recurse in _cygpath_parsers: match = regex.match(path) if match: - path = parser(*match.groups()) + if parser is _cygexpath: + path = parser(*match.groups(), expand_vars=expand_vars) + else: + path = parser(*match.groups()) if recurse: - path = cygpath(path) + path = cygpath(path, expand_vars=expand_vars) break else: - path = _cygexpath(None, path) + path = _cygexpath(None, path, expand_vars=expand_vars) return path @@ -438,7 +503,7 @@ def cygpath(path: str) -> str: def decygpath(path: PathLike) -> str: - path = str(path) + path = os.fspath(path) m = _decygpath_regex.match(path) if m: drive, rest_path = m.groups() @@ -464,6 +529,12 @@ def _is_cygwin_git(git_executable: str) -> bool: # Just a name given, not a real path. uname_cmd = osp.join(git_dir, "uname") + + if not (Path(uname_cmd).is_file() and os.access(uname_cmd, os.X_OK)): + _logger.debug(f"Failed checking if running in CYGWIN: {uname_cmd} is not an executable") + _is_cygwin_cache[git_executable] = is_cygwin + return is_cygwin + process = subprocess.Popen([uname_cmd], stdout=subprocess.PIPE, universal_newlines=True) uname_out, _ = process.communicate() # retcode = process.poll() @@ -484,7 +555,9 @@ def is_cygwin_git(git_executable: PathLike) -> bool: ... def is_cygwin_git(git_executable: Union[None, PathLike]) -> bool: - if sys.platform == "win32": # TODO: See if we can use `sys.platform != "cygwin"`. + # TODO: when py3.7 support is dropped, use the new interpolation f"{variable=}" + _logger.debug(f"sys.platform={sys.platform!r}, git_executable={git_executable!r}") + if sys.platform != "cygwin": return False elif git_executable is None: return False @@ -497,7 +570,7 @@ def get_user_id() -> str: return "%s@%s" % (getpass.getuser(), platform.node()) -def finalize_process(proc: Union[subprocess.Popen, "Git.AutoInterrupt"], **kwargs: Any) -> None: +def finalize_process(proc: Union["subprocess.Popen[Any]", "Git.AutoInterrupt"], **kwargs: Any) -> None: """Wait for the process (clone, fetch, pull or push) and handle its errors accordingly.""" # TODO: No close proc-streams?? @@ -509,26 +582,27 @@ def expand_path(p: None, expand_vars: bool = ...) -> None: ... @overload -def expand_path(p: PathLike, expand_vars: bool = ...) -> str: +def expand_path(p: PathLike, expand_vars: bool = ...) -> Optional[PathLike]: # TODO: Support for Python 3.5 has been dropped, so these overloads can be improved. ... def expand_path(p: Union[None, PathLike], expand_vars: bool = True) -> Optional[PathLike]: - if isinstance(p, pathlib.Path): - return p.resolve() + if p is None: + return None try: - p = osp.expanduser(p) # type: ignore[arg-type] + if isinstance(p, Path): + return p.resolve() + expanded_path = osp.expanduser(os.fspath(p)) if expand_vars: - p = osp.expandvars(p) - return osp.normpath(osp.abspath(p)) + expanded_path = osp.expandvars(expanded_path) + return osp.normpath(osp.abspath(expanded_path)) except Exception: return None def remove_password_if_present(cmdline: Sequence[str]) -> List[str]: - """Parse any command line argument and if one of the elements is an URL with a - username and/or password, replace them by stars (in-place). + """Redact credentials in URLs and HTTP Authorization extra headers in a command line. If nothing is found, this just returns the command line as-is. @@ -538,6 +612,16 @@ def remove_password_if_present(cmdline: Sequence[str]) -> List[str]: new_cmdline = [] for index, to_parse in enumerate(cmdline): new_cmdline.append(to_parse) + config_key, separator, header = to_parse.partition("=") + header_name, colon, _ = header.partition(":") + if ( + separator + and colon + and config_key.lower().endswith(".extraheader") + and header_name.strip().lower() == "authorization" + ): + new_cmdline[index] = "%s%s%s%s *****" % (config_key, separator, header_name, colon) + continue try: url = urlsplit(to_parse) # Remove password from the URL if present. @@ -598,13 +682,17 @@ def __init__(self) -> None: self.error_lines: List[str] = [] self.other_lines: List[str] = [] - def _parse_progress_line(self, line: AnyStr) -> None: + def _parse_progress_line(self, line: AnyStr) -> Optional[object]: """Parse progress information from the given line as retrieved by :manpage:`git-push(1)` or :manpage:`git-fetch(1)`. - Lines that do not contain progress info are stored in :attr:`other_lines`. - Lines that seem to contain an error (i.e. start with ``error:`` or ``fatal:``) are stored in :attr:`error_lines`. + + The base implementation returns ``None``. Subclasses may return another + value for compatibility with existing overrides, but callers should treat + the return value as unspecified. """ # handle # Counting objects: 4, done. @@ -619,7 +707,7 @@ def _parse_progress_line(self, line: AnyStr) -> None: if self._cur_line.startswith(("error:", "fatal:")): self.error_lines.append(self._cur_line) - return + return None cur_count, max_count = None, None match = self.re_op_relative.match(line_str) @@ -629,7 +717,7 @@ def _parse_progress_line(self, line: AnyStr) -> None: if not match: self.line_dropped(line_str) self.other_lines.append(line_str) - return + return None # END could not get match op_code = 0 @@ -660,7 +748,7 @@ def _parse_progress_line(self, line: AnyStr) -> None: self.line_dropped(line_str) # Note: Don't add this line to the other lines, as we have to silently # drop it. - return + return None # END handle op code # Figure out stage. @@ -686,6 +774,7 @@ def _parse_progress_line(self, line: AnyStr) -> None: max_count and float(max_count), message, ) + return None def new_message_handler(self) -> Callable[[str], None]: """ @@ -695,7 +784,7 @@ def new_message_handler(self) -> Callable[[str], None]: """ def handler(line: AnyStr) -> None: - return self._parse_progress_line(line.rstrip()) + self._parse_progress_line(line.rstrip()) # END handler @@ -756,7 +845,7 @@ class CallableRemoteProgress(RemoteProgress): __slots__ = ("_callable",) - def __init__(self, fn: Callable) -> None: + def __init__(self, fn: Callable[..., Any]) -> None: self._callable = fn super().__init__() @@ -835,7 +924,7 @@ def _main_actor( cls, env_name: str, env_email: str, - config_reader: Union[None, "GitConfigParser", "SectionConstraint"] = None, + config_reader: Union[None, "GitConfigParser", "SectionConstraint[GitConfigParser]"] = None, ) -> "Actor": actor = Actor("", "") user_id = None # We use this to avoid multiple calls to getpass.getuser(). @@ -871,7 +960,9 @@ def default_name() -> str: return actor @classmethod - def committer(cls, config_reader: Union[None, "GitConfigParser", "SectionConstraint"] = None) -> "Actor": + def committer( + cls, config_reader: Union[None, "GitConfigParser", "SectionConstraint[GitConfigParser]"] = None + ) -> "Actor": """ :return: :class:`Actor` instance corresponding to the configured committer. It @@ -886,7 +977,9 @@ def committer(cls, config_reader: Union[None, "GitConfigParser", "SectionConstra return cls._main_actor(cls.env_committer_name, cls.env_committer_email, config_reader) @classmethod - def author(cls, config_reader: Union[None, "GitConfigParser", "SectionConstraint"] = None) -> "Actor": + def author( + cls, config_reader: Union[None, "GitConfigParser", "SectionConstraint[GitConfigParser]"] = None + ) -> "Actor": """Same as :meth:`committer`, but defines the main author. It may be specified in the environment, but defaults to the committer.""" return cls._main_actor(cls.env_author_name, cls.env_author_email, config_reader) @@ -969,11 +1062,11 @@ class IndexFileSHA1Writer: __slots__ = ("f", "sha1") - def __init__(self, f: IO) -> None: + def __init__(self, f: IO[bytes]) -> None: self.f = f self.sha1 = make_sha(b"") - def write(self, data: AnyStr) -> int: + def write(self, data: bytes) -> int: self.sha1.update(data) return self.f.write(data) @@ -1135,7 +1228,7 @@ def _obtain_lock(self) -> None: # END endless loop -class IterableList(List[T_IterableObj]): +class IterableList(List[T_IterableObj]): # type: ignore[type-var] """List of iterable objects allowing to query an object by id or by named index:: heads = repo.heads @@ -1170,6 +1263,7 @@ def __new__(cls, id_attr: str, prefix: str = "") -> "IterableList[T_IterableObj] return super().__new__(cls) def __init__(self, id_attr: str, prefix: str = "") -> None: + super().__init__() self._id_attr = id_attr self._prefix = prefix @@ -1199,25 +1293,23 @@ def __getattr__(self, attr: str) -> T_IterableObj: # END for each item return list.__getattribute__(self, attr) - def __getitem__(self, index: Union[SupportsIndex, int, slice, str]) -> T_IterableObj: # type: ignore[override] - assert isinstance(index, (int, str, slice)), "Index of IterableList should be an int or str" - + def __getitem__( # type: ignore[override] # pyright: ignore[reportIncompatibleMethodOverride] + self, index: Union[SupportsIndex, int, slice, str] + ) -> T_IterableObj: if isinstance(index, int): return list.__getitem__(self, index) elif isinstance(index, slice): raise ValueError("Index should be an int or str") else: try: - return getattr(self, index) + return getattr(self, cast(str, index)) except AttributeError as e: - raise IndexError("No item found with id %r" % (self._prefix + index)) from e + raise IndexError(f"No item found with id {self._prefix}{index}") from e # END handle getattr def __delitem__(self, index: Union[SupportsIndex, int, slice, str]) -> None: - assert isinstance(index, (int, str)), "Index of IterableList should be an int or str" - delindex = cast(int, index) - if not isinstance(index, int): + if isinstance(index, str): delindex = -1 name = self._prefix + index for i, item in enumerate(self): @@ -1281,7 +1373,7 @@ def list_items(cls, repo: "Repo", *args: Any, **kwargs: Any) -> IterableList[T_I :return: list(Item,...) list of item instances """ - out_list: IterableList = IterableList(cls._id_attribute_) + out_list: IterableList[T_IterableObj] = IterableList(cls._id_attribute_) out_list.extend(cls.iter_items(repo, *args, **kwargs)) return out_list @@ -1290,7 +1382,8 @@ class IterableClassWatcher(type): """Metaclass that issues :exc:`DeprecationWarning` when :class:`git.util.Iterable` is subclassed.""" - def __init__(cls, name: str, bases: Tuple, clsdict: Dict) -> None: + def __init__(cls, name: str, bases: Tuple[type, ...], clsdict: Dict[str, Any]) -> None: + super().__init__(name, bases, clsdict) for base in bases: if type(base) is IterableClassWatcher: warnings.warn( diff --git a/gitdb/.coveragerc b/gitdb/.coveragerc new file mode 100644 index 000000000..71b6ef701 --- /dev/null +++ b/gitdb/.coveragerc @@ -0,0 +1,7 @@ +[run] +source = gitdb + +; to make nosetests happy +[report] +include = */gitdb/* +omit = */gitdb/ext/* diff --git a/gitdb/.gitignore b/gitdb/.gitignore new file mode 100644 index 000000000..8b7da9277 --- /dev/null +++ b/gitdb/.gitignore @@ -0,0 +1,10 @@ +MANIFEST +.coverage +build/ +dist/ +*.pyc +*.o +*.so +.noseids +*.sublime-workspace +*.egg-info diff --git a/gitdb/AUTHORS b/gitdb/AUTHORS new file mode 100644 index 000000000..6c7e9b997 --- /dev/null +++ b/gitdb/AUTHORS @@ -0,0 +1,4 @@ +Creator: Sebastian Thiel + +Contributors: + - Ram Rachum (@cool-RR) diff --git a/gitdb/LICENSE b/gitdb/LICENSE new file mode 100644 index 000000000..0d6fe8bdb --- /dev/null +++ b/gitdb/LICENSE @@ -0,0 +1,42 @@ +Copyright (C) 2010, 2011 Sebastian Thiel and contributors +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions +are met: + +* Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + +* Redistributions in binary form must reproduce the above copyright +notice, this list of conditions and the following disclaimer in the +documentation and/or other materials provided with the distribution. + +* Neither the name of the GitDB project nor the names of +its contributors may be used to endorse or promote products derived +from this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED +TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR +PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF +LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING +NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + + +Additional Licenses +------------------- +The files at +gitdb/test/fixtures/packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.idx +and +gitdb/test/fixtures/packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.pack +are licensed under GNU GPL as part of the git source repository, +see http://en.wikipedia.org/wiki/Git_%28software%29 for more information. + +They are not required for the actual operation, which is why they are not found +in the distribution package. diff --git a/gitdb/MANIFEST.in b/gitdb/MANIFEST.in new file mode 100644 index 000000000..b939b5ded --- /dev/null +++ b/gitdb/MANIFEST.in @@ -0,0 +1,18 @@ +include VERSION +include LICENSE +include CHANGES +include AUTHORS +include README +include MANIFEST.in + +include gitdb/_fun.c +include gitdb/_delta_apply.c +include gitdb/_delta_apply.h + +graft gitdb/test + +global-exclude .git* +global-exclude *.pyc +global-exclude *.so +global-exclude *.dll +global-exclude *.o diff --git a/gitdb/Makefile b/gitdb/Makefile new file mode 100644 index 000000000..ad0912ffe --- /dev/null +++ b/gitdb/Makefile @@ -0,0 +1,7 @@ +.PHONY: all clean + +all: + @grep -Ee '^[a-z].*:' Makefile | cut -d: -f1 | grep -vF all + +clean: + rm -rf build/ dist/ .eggs/ .tox/ diff --git a/gitdb/README.rst b/gitdb/README.rst new file mode 100644 index 000000000..45344a078 --- /dev/null +++ b/gitdb/README.rst @@ -0,0 +1,69 @@ +GitDB +===== + +GitDB allows you to access bare git repositories for reading and writing. It aims at allowing full access to loose objects as well as packs with performance and scalability in mind. It operates exclusively on streams, allowing to handle large objects with a small memory footprint. + +Installation +============ + +.. image:: https://img.shields.io/pypi/v/gitdb.svg + :target: https://pypi.python.org/pypi/gitdb/ + :alt: Latest Version +.. image:: https://img.shields.io/pypi/pyversions/gitdb.svg + :target: https://pypi.python.org/pypi/gitdb/ + :alt: Supported Python versions +.. image:: https://readthedocs.org/projects/gitdb/badge/?version=latest + :target: https://readthedocs.org/projects/gitdb/?badge=latest + :alt: Documentation Status + +From `PyPI `_:: + + pip install gitdb + +SPEEDUPS +======== + +If you want to go up to 20% faster, you can install gitdb-speedups with:: + + pip install gitdb-speedups + +However, please note that gitdb-speedups is not currently maintained. + +REQUIREMENTS +============ + +* smmap - declared as a dependency, automatically installed +* pytest - for running the tests + +SOURCE +====== + +The source is available in the GitPython repository on GitHub: + +https://github.com/gitpython-developers/GitPython/tree/main/gitdb + +Run the tests with:: + + pytest + +DEVELOPMENT +=========== + +.. image:: https://github.com/gitpython-developers/GitPython/actions/workflows/dependencies.yml/badge.svg + :target: https://github.com/gitpython-developers/GitPython/actions/workflows/dependencies.yml + +The library is considered mature, and not under active development. Its primary (known) use is in GitPython. + +INFRASTRUCTURE +============== + +* Discussions + * https://github.com/gitpython-developers/GitPython/discussions + +* Issue Tracker + * https://github.com/gitpython-developers/GitPython/issues + +LICENSE +======= + +New BSD License diff --git a/gitdb/SECURITY.md b/gitdb/SECURITY.md new file mode 100644 index 000000000..95389ff6e --- /dev/null +++ b/gitdb/SECURITY.md @@ -0,0 +1,3 @@ +# Security Policy + +See [GitPython](https://github.com/gitpython-developers/GitPython/blob/main/SECURITY.md). Vulnerabilities found in `gitdb` can be reported there. diff --git a/gitdb/doc/.gitignore b/gitdb/doc/.gitignore new file mode 100644 index 000000000..567609b12 --- /dev/null +++ b/gitdb/doc/.gitignore @@ -0,0 +1 @@ +build/ diff --git a/gitdb/doc/Makefile b/gitdb/doc/Makefile new file mode 100644 index 000000000..b10926ae2 --- /dev/null +++ b/gitdb/doc/Makefile @@ -0,0 +1,89 @@ +# Makefile for Sphinx documentation +# + +# You can set these variables from the command line. +SPHINXOPTS = +SPHINXBUILD = sphinx-build +PAPER = +BUILDDIR = build + +# Internal variables. +PAPEROPT_a4 = -D latex_paper_size=a4 +PAPEROPT_letter = -D latex_paper_size=letter +ALLSPHINXOPTS = -d $(BUILDDIR)/doctrees $(PAPEROPT_$(PAPER)) $(SPHINXOPTS) source + +.PHONY: help clean html dirhtml pickle json htmlhelp qthelp latex changes linkcheck doctest + +help: + @echo "Please use \`make ' where is one of" + @echo " html to make standalone HTML files" + @echo " dirhtml to make HTML files named index.html in directories" + @echo " pickle to make pickle files" + @echo " json to make JSON files" + @echo " htmlhelp to make HTML files and a HTML help project" + @echo " qthelp to make HTML files and a qthelp project" + @echo " latex to make LaTeX files, you can set PAPER=a4 or PAPER=letter" + @echo " changes to make an overview of all changed/added/deprecated items" + @echo " linkcheck to check all external links for integrity" + @echo " doctest to run all doctests embedded in the documentation (if enabled)" + +clean: + -rm -rf $(BUILDDIR)/* + +html: + $(SPHINXBUILD) -b html $(ALLSPHINXOPTS) $(BUILDDIR)/html + @echo + @echo "Build finished. The HTML pages are in $(BUILDDIR)/html." + +dirhtml: + $(SPHINXBUILD) -b dirhtml $(ALLSPHINXOPTS) $(BUILDDIR)/dirhtml + @echo + @echo "Build finished. The HTML pages are in $(BUILDDIR)/dirhtml." + +pickle: + $(SPHINXBUILD) -b pickle $(ALLSPHINXOPTS) $(BUILDDIR)/pickle + @echo + @echo "Build finished; now you can process the pickle files." + +json: + $(SPHINXBUILD) -b json $(ALLSPHINXOPTS) $(BUILDDIR)/json + @echo + @echo "Build finished; now you can process the JSON files." + +htmlhelp: + $(SPHINXBUILD) -b htmlhelp $(ALLSPHINXOPTS) $(BUILDDIR)/htmlhelp + @echo + @echo "Build finished; now you can run HTML Help Workshop with the" \ + ".hhp project file in $(BUILDDIR)/htmlhelp." + +qthelp: + $(SPHINXBUILD) -b qthelp $(ALLSPHINXOPTS) $(BUILDDIR)/qthelp + @echo + @echo "Build finished; now you can run "qcollectiongenerator" with the" \ + ".qhcp project file in $(BUILDDIR)/qthelp, like this:" + @echo "# qcollectiongenerator $(BUILDDIR)/qthelp/GitDB.qhcp" + @echo "To view the help file:" + @echo "# assistant -collectionFile $(BUILDDIR)/qthelp/GitDB.qhc" + +latex: + $(SPHINXBUILD) -b latex $(ALLSPHINXOPTS) $(BUILDDIR)/latex + @echo + @echo "Build finished; the LaTeX files are in $(BUILDDIR)/latex." + @echo "Run \`make all-pdf' or \`make all-ps' in that directory to" \ + "run these through (pdf)latex." + +changes: + $(SPHINXBUILD) -b changes $(ALLSPHINXOPTS) $(BUILDDIR)/changes + @echo + @echo "The overview file is in $(BUILDDIR)/changes." + +linkcheck: + $(SPHINXBUILD) -b linkcheck $(ALLSPHINXOPTS) $(BUILDDIR)/linkcheck + @echo + @echo "Link check complete; look for any errors in the above output " \ + "or in $(BUILDDIR)/linkcheck/output.txt." + +doctest: + $(SPHINXBUILD) -b doctest $(ALLSPHINXOPTS) $(BUILDDIR)/doctest + @echo "Testing of doctests in the sources finished, look at the " \ + "results in $(BUILDDIR)/doctest/output.txt." diff --git a/gitdb/doc/source/algorithm.rst b/gitdb/doc/source/algorithm.rst new file mode 100644 index 000000000..2e01b3fbe --- /dev/null +++ b/gitdb/doc/source/algorithm.rst @@ -0,0 +1,97 @@ +######################## +Discussion of Algorithms +######################## + +************ +Introduction +************ +As you know, the pure-python object database support for GitPython is provided by the GitDB project. It is meant to be my backup plan to ensure that the DataVault (http://www.youtube.com/user/ByronBates99?feature=mhum#p/c/2A5C6EF5BDA8DB5C ) can handle reading huge files, especially those which were consolidated into packs. A nearly fully packed state is anticipated for the data-vaults repository, and reading these packs efficiently is an essential task. + +This document informs you about my findings in the struggle to improve the way packs are read to reduce memory usage required to handle huge files. It will try to conclude where future development could go to assure big delta-packed files can be read without the need of 8GB+ RAM. + +GitDB's main feature is the use of streams, hence the amount of memory used to read a database object is minimized, at the cost of some additional processing overhead to keep track of the stream state. This works great for legacy objects, which are essentially a zip-compressed byte-stream. + +Streaming data from delta-packed objects is far more difficult though, and only technically possible within certain limits, and at relatively high processing costs. My first observation was that there doesn't appear to be 'the one and only' algorithm which is generally superior. They all have their pros and cons, but fortunately this allows the implementation to choose the one most suited based on the amount of delta streams, as well as the size of the base, which allows an early and cheap estimate of the target size of the final data. + +********************************** +Traditional Delta-Apply-Algorithms +********************************** + +The brute-force CGit delta-apply algorithm +========================================== +CGit employs a simple and relatively brute-force algorithm, which resolves all delta streams recursively. When the recursion reaches the base level of the deltas, it will be decompressed into a buffer, then the first delta gets decompressed into a second buffer. From that, the target size of the delta can be extracted, to allocated a third buffer to hold the result of the operation, which consists of reading the delta stream byte-wise, to apply the operations in order, as described by single-byte opcodes. During recursion, each target buffer of the preceding delta-apply operation is used as base buffer for the next delta-apply operation, until the last delta was applied, leaving the final target buffer as result. + +About Delta-Opcodes +------------------- +There are only two kinds of opcodes, 'add-bytes' and 'copy-from-base'. One 'add-bytes' opcode can encode up to 7 bit of additional bytes to be copied from the delta stream into the target buffer. +A 'copy-from-base' opcode encodes a 32 bit offset into the base buffer, as well as the amount of bytes to be copied, which are up to 2^24 bytes. We may conclude that delta-bases may not be larger than 2^32+2^24 bytes in the current, extensible, implementation. +When generating the delta, git prefers copy operations over add operations, as they are much more efficient. Usually, the most recent, or biggest version of a file is used as base, whereas older and smaller versions of the file are expressed by copying only portions of the newest file. As it is not efficiently possible to represent all changes that way, add-bytes operations fill the gap where needed. All this explains why git can only add 128 bytes with one opcode, as it tries to minimize their use. +This implies that recent file history can usually be extracted faster than old history, which may involve many more deltas. + +Performance considerations +-------------------------- +The performance bottleneck of this algorithm appear to be the throughput of your RAM, as both opcodes will just trigger memcpy operations from one memory location to another, times the amount of deltas to apply. This in fact is very fast, even for big files above 100 MB. +Memory allocation could become an issue as you need the base buffer, the target buffer as well as the decompressed delta stream in memory at the same time. The continuous allocation and deallocation of possibly big buffers may support memory fragmentation. Whether it really kicks in, especially on 64 bit machines, is unproven though. +Nonetheless, the cgit implementation is currently the fastest one. + +The brute-force GitDB algorithm +=============================== +Despite of working essentially the same way as the CGit brute-force algorithm, GitDB minimizes the amount of allocations to 2 + num of deltas. The amount memory allocated peaks whiles the deltas are applied, as the base and target buffer, as well as the decompressed stream, are held in memory. +To achieve this, GitDB retrieves all delta-streams in advance, and peaks into their header information to determine the maximum size of the target buffer, just by reading 512 bytes of the compressed stream. If there is more than one delta to apply, the base buffer is set large enough to hold the biggest target buffer required by the delta streams. +Now it is possible to iterate all deltas, oldest first, newest last, and apply them using the buffers. At the end of each iteration, the buffers are swapped. + +Performance Results +------------------- +The performance test is performed on an aggressively packed repository with the history of cgit. 5000 sha's are extracted and read one after another. The delta-chains have a length of 0 to about 35. The pure-python implementation can stream the data of all objects (totaling 62,2 MiB) with an average rate of 8.1 MiB/s, which equals about 654 streams/s. +There are two bottlenecks: The major is the collection of the delta streams, which involves plenty of pack-lookup. This lookup is expensive in python, and is overly expensive. Its not overly critical though, as it only limits the amount of streams per second, not the actual data rate when applying the deltas. +Applying the deltas happens to be the second bottleneck, if the files to be processed get bigger. The more opcodes have to be processed, the more python slow function calls will dominate the result. As an example, it takes nearly 8 seconds to unpack a 125 MB file, where cgit only takes 2.4 s. + +To eliminate a few performance concerns, some key routines were rewritten in C. This changes the numbers of this particular test significantly, but not drastically, as the major bottleneck (delta collection) is still in place. Another performance reduction is due to the fact that plenty of other code related to the deltas is still pure-python. +Now all 5000 objects can be read at a rate of 11.1 MiB /s, or 892 streams/s. Fortunately, unpacking a big object is now done in 2.5s, which is just a tad slower than cgit, but with possibly less memory fragmentation. + + +************************************** +Paving the way towards delta streaming +************************************** + +GitDB's reverse delta aggregation algorithm +=========================================== +The idea of this algorithm is to merge all delta streams into one, which can then be applied in just one go. + +In the current implementation, delta streams are parsed into DeltaChunks (->**DC**). Each DC represents one copy-from-base operation, or one or multiple consecutive add-bytes operations. DeltaChunks know about their target offset in the target buffer, and their size. Their target offsets are consecutive, i.e. one chunk ends where the next one begins, regarding their logical extend in the target buffer. +Add-bytes DCs additional store their data to apply, copy-from-base DCs store the offset into the base buffer from which to copy bytes. + +During processing, one starts with the latest (i.e. topmost) delta stream (->**TDS**), and iterates through its ancestor delta streams (->ADS) to merge them into the growing toplevel delta stream.. + +The merging works by following a set of rules: + * Merge into the top-level delta from the youngest ancestor delta to the oldest one + * When merging one ADS, iterate from the first to the last chunk in TDS, then: + + * skip all add-bytes DCs. If bytes are added, these will always overwrite any operation coming from any ADS at the same offset. + * copy-from-base DCs will copy a slice of the respective portion of the ADS ( as defined by their base offset ) and use it to replace the original chunk. This acts as a 'virtual' copy-from-base operation. + + * Finish the merge once all ADS have been handled, or once the TDS only consists of add-byte DCs. The remaining copy-from-base DCs will copy from the original base buffer accordingly. + +Applying the TDS is as straightforward as applying any other DS. The base buffer is required to be kept in memory. In the current implementation, a full-size target buffer is allocated to hold the result of applying the chunk information. Here it is already possible to stream the result, which is feasible only if the memory of the base buffer + the memory of the TDS are smaller than a full size target buffer. Streaming will always make sense if the peak resulting from having the base, target and TDS buffers in memory together is unaffordable. + +The memory consumption during the TDS processing is only the condensed delta-bytes, for each ADS an additional index is required which costs 8 byte per DC. When applying the TDS, one requires an allocated base buffer too.The target buffer can be allocated, but may be a writer as well. + +Performance Results +------------------- +The benchmarking context was the same as for the brute-force GitDB algorithm. This implementation is far more complex than the said brute-force implementation, which clearly reflects in the numbers. It's pure-python throughput is at only 1.1 MiB/s, which equals 89 streams/s. +The biggest performance bottleneck is the slicing of the parsed delta streams, where the program spends most of its time due to hundred thousands of calls. + +To get a more usable version of the algorithm, it was implemented in C, such that python must do no more than two calls to get all the work done. The first prepares the TDS, the second applies it, writing it into a target buffer. +The throughput reaches 15.2 MiB/s, which equals 1221 streams/s, which makes it nearly 14 times faster than the pure python version, and amazingly even 1.35 times faster than the brute-force C implementation. As a comparison, cgit is able to stream about 20 MiB when controlling it through a pipe. GitDBs performance may still improve once pack access is reimplemented in C as well. + +A 125 MB file took 2.5 seconds to unpack for instance, which is only 20% slower than the c implementation of the brute-force algorithm. + + +Future work +=========== + +Another very promising option is that streaming of delta data is indeed possible. Depending on the configuration of the copy-from-base operations, different optimizations could be applied to reduce the amount of memory required for the final processed delta stream. Some configurations may even allow it to stream data from the base buffer, instead of pre-loading it for random access. + +The ability to stream files at reduced memory costs would only be feasible for big files, and would have to be paid with extra pre-processing time. + +A very first and simple implementation could avoid memory peaks by streaming the TDS in conjunction with a base buffer, instead of writing everything into a fully allocated target buffer. diff --git a/gitdb/doc/source/api.rst b/gitdb/doc/source/api.rst new file mode 100644 index 000000000..9dea95764 --- /dev/null +++ b/gitdb/doc/source/api.rst @@ -0,0 +1,113 @@ +.. _api-label: + +############# +API Reference +############# + +**************** +Database.Base +**************** + +.. automodule:: gitdb.db.base + :members: + :undoc-members: + +**************** +Database.Git +**************** + +.. automodule:: gitdb.db.git + :members: + :undoc-members: + +**************** +Database.Loose +**************** + +.. automodule:: gitdb.db.loose + :members: + :undoc-members: + +**************** +Database.Memory +**************** + +.. automodule:: gitdb.db.mem + :members: + :undoc-members: + +**************** +Database.Pack +**************** + +.. automodule:: gitdb.db.pack + :members: + :undoc-members: + +****************** +Database.Reference +****************** + +.. automodule:: gitdb.db.ref + :members: + :undoc-members: + +************ +Base +************ + +.. automodule:: gitdb.base + :members: + :undoc-members: + +************ +Functions +************ + +.. automodule:: gitdb.fun + :members: + :undoc-members: + +************ +Pack +************ + +.. automodule:: gitdb.pack + :members: + :undoc-members: + +************ +Streams +************ + +.. automodule:: gitdb.stream + :members: + :undoc-members: + +************ +Types +************ + +.. automodule:: gitdb.typ + :members: + :undoc-members: + + +************ +Utilities +************ + +.. automodule:: gitdb.util + :members: + :undoc-members: + + + + + + + + + + + diff --git a/gitdb/doc/source/changes.rst b/gitdb/doc/source/changes.rst new file mode 100644 index 000000000..b4340e4dc --- /dev/null +++ b/gitdb/doc/source/changes.rst @@ -0,0 +1,141 @@ +######### +Changelog +######### + +****** +4.0.12 +****** + +- various improvements - please see the release on GitHub for details. + +****** +4.0.11 +****** + +- various improvements - please see the release on GitHub for details. + +****** +4.0.10 +****** + +- improvements to the way external packages are imported. + +***** +4.0.9 +***** + +- re-release of 4.0.8 to get a valid signature. + +***** +4.0.8 +***** + +* drop support for python 3.4 and 3.5 due to EOL +* Updated upper bound for smmap requirement in setup.py + (`#69 `_) + +***** +4.0.7 +***** + +* Updated upper bound for smmap requirement in setup.py + (`#69 `_) + +***** +4.0.6 +***** + +* Bumped upper bound for smmap requirement + (`#67 `_, + `#68 `_) + +***** +4.0.5 +***** + +* Re-release of 4.0.4, with known signature + +***** +4.0.4 +***** + +* Support for PyOxidizer + +***** +4.0.2 +***** + +* Updated to release as Pure Python Wheel rather than Universal Wheel + (`#62 `_) + +***** +4.0.1 +***** + +* Switched back to the gitdb package name on PyPI and fixed the gitdb2 mirror package + (`#59 `_) +* Switched back to require smmap package and fixed version requirement to >= 3.0.1, < 4 + (`#59 `_) +* Updated smmap submodule + +*********** +3.0.3.post1 +*********** + +* Fixed changelogs for v3.0.2 and v3.0.3 + +***** +3.0.3 +***** + +* Changed ``force_bytes`` to use UTF-8 encoding by default + (`#49 `_) +* Restricted smmap2 version requirement to < 3 +* Updated requirements.txt + +***** +3.0.2 +***** + +* Removed Python 2 compatibility shims + (`#56 `_) + +***** +0.6.1 +***** + +* Fixed possibly critical error, see https://github.com/gitpython-developers/GitPython/issues/220 + + - However, it only seems to occur on high-entropy data and didn't reoccour after the fix + +***** +0.6.0 +***** + +* Added support got python 3.X +* Removed all `async` dependencies and all `*_async` versions of methods with it. + +***** +0.5.4 +***** +* Adjusted implementation to use the SlidingMemoryManager by default in python 2.6 for efficiency reasons. In Python 2.4, the StaticMemoryManager will be used instead. + +***** +0.5.3 +***** +* Added support for smmap. SmartMMap allows resources to be managed and controlled. This brings the implementation closer to the way git handles memory maps, such that unused cached memory maps will automatically be freed once a resource limit is hit. The memory limit on 32 bit systems remains though as a sliding mmap implementation is not used for performance reasons. + +***** +0.5.2 +***** +* Improved performance of the c implementation, which now uses reverse-delta-aggregation to make a memory bound operation CPU bound. + +***** +0.5.1 +***** +* Restored most basic python 2.4 compatibility, such that gitdb can be imported within python 2.4, pack access cannot work though. This at least allows Super-Projects to provide their own workarounds, or use everything but pack support. + +***** +0.5.0 +***** +Initial Release diff --git a/gitdb/doc/source/conf.py b/gitdb/doc/source/conf.py new file mode 100644 index 000000000..b387f6094 --- /dev/null +++ b/gitdb/doc/source/conf.py @@ -0,0 +1,193 @@ +# +# GitDB documentation build configuration file, created by +# sphinx-quickstart on Wed Jun 30 00:01:32 2010. +# +# This file is execfile()d with the current directory set to its containing dir. +# +# Note that not all possible configuration values are present in this +# autogenerated file. +# +# All configuration values have a default; values that are commented out +# serve to show the default. + +import sys +import os + +# If extensions (or modules to document with autodoc) are in another directory, +# add these directories to sys.path here. If the directory is relative to the +# documentation root, use os.path.abspath to make it absolute, like shown here. +sys.path.append(os.path.abspath('../../')) + +# -- General configuration ----------------------------------------------------- + +# Add any Sphinx extension module names here, as strings. They can be extensions +# coming with Sphinx (named 'sphinx.ext.*') or your custom ones. +extensions = ['sphinx.ext.autodoc'] + +# Add any paths that contain templates here, relative to this directory. +templates_path = ['.templates'] + +# The suffix of source filenames. +source_suffix = '.rst' + +# The encoding of source files. +#source_encoding = 'utf-8' + +# The master toctree document. +master_doc = 'index' + +# General information about the project. +project = 'GitDB' +copyright = '2011, Sebastian Thiel' + +# The version info for the project you're documenting, acts as replacement for +# |version| and |release|, also used in various other places throughout the +# built documents. +# +# The short X.Y version. +version = '0.5' +# The full version, including alpha/beta/rc tags. +release = '0.5.3' + +# The language for content autogenerated by Sphinx. Refer to documentation +# for a list of supported languages. +#language = None + +# There are two options for replacing |today|: either, you set today to some +# non-false value, then it is used: +#today = '' +# Else, today_fmt is used as the format for a strftime call. +#today_fmt = '%B %d, %Y' + +# List of documents that shouldn't be included in the build. +#unused_docs = [] + +# List of directories, relative to source directory, that shouldn't be searched +# for source files. +exclude_trees = [] + +# The reST default role (used for this markup: `text`) to use for all documents. +#default_role = None + +# If true, '()' will be appended to :func: etc. cross-reference text. +#add_function_parentheses = True + +# If true, the current module name will be prepended to all description +# unit titles (such as .. function::). +#add_module_names = True + +# If true, sectionauthor and moduleauthor directives will be shown in the +# output. They are ignored by default. +#show_authors = False + +# The name of the Pygments (syntax highlighting) style to use. +pygments_style = 'sphinx' + +# A list of ignored prefixes for module index sorting. +#modindex_common_prefix = [] + + +# -- Options for HTML output --------------------------------------------------- + +# The theme to use for HTML and HTML Help pages. Major themes that come with +# Sphinx are currently 'default' and 'sphinxdoc'. +html_theme = 'default' + +html_theme_options = { + "stickysidebar": "true" +} + +# Add any paths that contain custom themes here, relative to this directory. +#html_theme_path = [] + +# The name for this set of Sphinx documents. If None, it defaults to +# " v documentation". +#html_title = None + +# A shorter title for the navigation bar. Default is the same as html_title. +#html_short_title = None + +# The name of an image file (relative to this directory) to place at the top +# of the sidebar. +#html_logo = None + +# The name of an image file (within the static path) to use as favicon of the +# docs. This file should be a Windows icon file (.ico) being 16x16 or 32x32 +# pixels large. +#html_favicon = None + +# Add any paths that contain custom static files (such as style sheets) here, +# relative to this directory. They are copied after the builtin static files, +# so a file named "default.css" will overwrite the builtin "default.css". +#html_static_path = ['.static'] + +# If not '', a 'Last updated on:' timestamp is inserted at every page bottom, +# using the given strftime format. +#html_last_updated_fmt = '%b %d, %Y' + +# If true, SmartyPants will be used to convert quotes and dashes to +# typographically correct entities. +#html_use_smartypants = True + +# Custom sidebar templates, maps document names to template names. +#html_sidebars = {} + +# Additional templates that should be rendered to pages, maps page names to +# template names. +#html_additional_pages = {} + +# If false, no module index is generated. +#html_use_modindex = True + +# If false, no index is generated. +#html_use_index = True + +# If true, the index is split into individual pages for each letter. +#html_split_index = False + +# If true, links to the reST sources are added to the pages. +#html_show_sourcelink = True + +# If true, an OpenSearch description file will be output, and all pages will +# contain a tag referring to it. The value of this option must be the +# base URL from which the finished HTML is served. +#html_use_opensearch = '' + +# If nonempty, this is the file name suffix for HTML files (e.g. ".xhtml"). +#html_file_suffix = '' + +# Output file base name for HTML help builder. +htmlhelp_basename = 'GitDBdoc' + + +# -- Options for LaTeX output -------------------------------------------------- + +# The paper size ('letter' or 'a4'). +#latex_paper_size = 'letter' + +# The font size ('10pt', '11pt' or '12pt'). +#latex_font_size = '10pt' + +# Grouping the document tree into LaTeX files. List of tuples +# (source start file, target name, title, author, documentclass [howto/manual]). +latex_documents = [ + ('index', 'GitDB.tex', 'GitDB Documentation', + 'Sebastian Thiel', 'manual'), +] + +# The name of an image file (relative to this directory) to place at the top of +# the title page. +#latex_logo = None + +# For "manual" documents, if this is true, then toplevel headings are parts, +# not chapters. +#latex_use_parts = False + +# Additional stuff for the LaTeX preamble. +#latex_preamble = '' + +# Documents to append as an appendix to all manuals. +#latex_appendices = [] + +# If false, no module index is generated. +#latex_use_modindex = True diff --git a/gitdb/doc/source/index.rst b/gitdb/doc/source/index.rst new file mode 100644 index 000000000..5223e6bd9 --- /dev/null +++ b/gitdb/doc/source/index.rst @@ -0,0 +1,26 @@ +.. GitDB documentation master file, created by + sphinx-quickstart on Wed Jun 30 00:01:32 2010. + You can adapt this file completely to your liking, but it should at least + contain the root `toctree` directive. + +Welcome to GitDB's documentation! +================================= + +Contents: + +.. toctree:: + :maxdepth: 2 + + intro + tutorial + api + algorithm + changes + +Indices and tables +================== + +* :ref:`genindex` +* :ref:`modindex` +* :ref:`search` + diff --git a/gitdb/doc/source/intro.rst b/gitdb/doc/source/intro.rst new file mode 100644 index 000000000..434138616 --- /dev/null +++ b/gitdb/doc/source/intro.rst @@ -0,0 +1,38 @@ +######## +Overview +######## + +The *GitDB* project implements interfaces to allow read and write access to git repositories. In its core lies the *db* package, which contains all database types necessary to read a complete git repository. These are the ``LooseObjectDB``, the ``PackedDB`` and the ``ReferenceDB`` which are combined into the ``GitDB`` to combine every aspect of the git database. + +For this to work, GitDB implements pack reading, as well as loose object reading and writing. Data is always encapsulated in streams, which allows huge files to be handled as well as small ones, usually only chunks of the stream are kept in memory for processing, never the whole stream at once. + +Interfaces are used to describe the API, making it easy to provide alternate implementations. + +================ +Installing GitDB +================ +Its easiest to install gitdb using the *pip* program:: + + $ pip install gitdb + +As the command will install gitdb in your respective python distribution, you will most likely need root permissions to authorize the required changes. + +If you have downloaded the source archive, the package can be installed by running the ``setup.py`` script:: + + $ python setup.py install + +=============== +Getting Started +=============== +It is advised to have a look at the :ref:`Usage Guide ` for a brief introduction on the different database implementations. + +================= +Source Repository +================= +The latest source can be cloned using git from github: + + * https://github.com/gitpython-developers/gitdb + +License Information +=================== +*GitDB* is licensed under the New BSD License. diff --git a/gitdb/doc/source/tutorial.rst b/gitdb/doc/source/tutorial.rst new file mode 100644 index 000000000..55a737f6b --- /dev/null +++ b/gitdb/doc/source/tutorial.rst @@ -0,0 +1,114 @@ +.. _tutorial-label: + +########### +Usage Guide +########### +This text briefly introduces you to the basic design decisions and accompanying types. + +****** +Design +****** +The *GitDB* project models a standard git object database and implements it in pure python. This means that data, being classified by one of four types, can can be stored in the database and will in future be referred to by the generated SHA1 key, which is a 20 byte string within python. + +*GitDB* implements *RW* access to loose objects, as well as *RO* access to packed objects. Compound Databases allow to combine multiple object databases into one. + +All data is read and written using streams, which effectively prevents more than a chunk of the data being kept in memory at once mostly [#]_. + +******* +Streams +******* +In order to assure the object database can handle objects of any size, a stream interface is used for data retrieval as well as to fill data into the database. + +Basic Stream Types +================== +There are two fundamentally different types of streams, **IStream**\ s and **OStream**\ s. IStreams are mutable and are used to provide data streams to the database to create new objects. + +OStreams are immutable and are used to read data from the database. The base of this type, **OInfo**, contains only type and size information of the queried object, but no stream, which is slightly faster to retrieve depending on the database. + +OStreams are tuples, IStreams are lists. Both, OInfo and OStream, have the same member ordering which allows quick conversion from one type to another. + +**************************** +Data Query and Data Addition +**************************** +Databases support query and/or addition of objects using simple interfaces. They are called **ObjectDBR** for read-only access, and **ObjectDBW** for write access to create new objects. + +Both have two sets of methods, one of which allows interacting with single objects, the other one allowing to handle a stream of objects simultaneously and asynchronously. + +Acquiring information about an object from a database is easy if you have a SHA1 to refer to the object:: + + + ldb = LooseObjectDB(fixture_path("../../../.git/objects")) + + for sha1 in ldb.sha_iter(): + oinfo = ldb.info(sha1) + ostream = ldb.stream(sha1) + assert oinfo[:3] == ostream[:3] + + assert len(ostream.read()) == ostream.size + # END for each sha in database + +To store information, you prepare an *IStream* object with the required information. The provided stream will be read and converted into an object, and the respective 20 byte SHA1 identifier is stored in the IStream object:: + + data = "my data" + istream = IStream("blob", len(data), StringIO(data)) + + # the object does not yet have a sha + assert istream.binsha is None + ldb.store(istream) + # now the sha is set + assert len(istream.binsha) == 20 + assert ldb.has_object(istream.binsha) + +********************** +Asynchronous Operation +********************** +For each read or write method that allows a single-object to be handled, an *_async* version exists which reads items to be processed from a channel, and writes the operation's result into an output channel that is read by the caller or by other async methods, to support chaining. + +Using asynchronous operations is easy, but chaining multiple operations together to form a complex one would require you to read the docs of the *async* package. At the current time, due to the *GIL*, the *GitDB* can only achieve true concurrency during zlib compression and decompression if big objects, if the respective c modules where compiled in *async*. + +Asynchronous operations are scheduled by a *ThreadPool* which resides in the *gitdb.util* module:: + + from gitdb.util import pool + + # set the pool to use two threads + pool.set_size(2) + + # synchronize the mode of operation + pool.set_size(0) + + +Use async methods with readers, which supply items to be processed. The result is given through readers as well:: + + from async import IteratorReader + + # Create a reader from an iterator + reader = IteratorReader(ldb.sha_iter()) + + # get reader for object streams + info_reader = ldb.stream_async(reader) + + # read one + info = info_reader.read(1)[0] + + # read all the rest until depletion + ostreams = info_reader.read() + + + +********* +Databases +********* +A database implements different interfaces, one if which will always be the *ObjectDBR* interface to support reading of object information and streams. + +The *Loose Object Database* as well as the *Packed Object Database* are *File Databases*, hence they operate on a directory which contains files they can read. + +File databases implementing the *ObjectDBW* interface can also be forced to write their output into the specified stream, using the ``set_ostream`` method. This effectively allows you to redirect its output to anywhere you like. + +*Compound Databases* are not implementing their own access type, but instead combine multiple database implementations into one. Examples for this database type are the *Reference Database*, which reads object locations from a file, and the *GitDB* which combines loose, packed and referenced objects into one database interface. + +For more information about the individual database types, please see the :ref:`API Reference `, and the unittests for the respective types. + + +---- + +.. [#] When reading streams from packs, all deltas are currently applied and the result written into a memory map before the first byte is returned. Future versions of the delta-apply algorithm might improve on this. diff --git a/gitdb/gitdb.pro b/gitdb/gitdb.pro new file mode 100644 index 000000000..a682615c4 --- /dev/null +++ b/gitdb/gitdb.pro @@ -0,0 +1,48 @@ + +OTHER_FILES += \ + setup.py \ + README.rst \ + MANIFEST \ + Makefile \ + LICENSE \ + AUTHORS \ + gitdb/util.py \ + gitdb/typ.py \ + gitdb/stream.py \ + gitdb/pack.py \ + gitdb/__init__.py \ + gitdb/fun.py \ + gitdb/exc.py \ + gitdb/base.py \ + doc/source/tutorial.rst \ + doc/source/intro.rst \ + doc/source/index.rst \ + doc/source/conf.py \ + doc/source/changes.rst \ + doc/source/api.rst \ + doc/source/algorithm.rst \ + gitdb/db/ref.py \ + gitdb/db/pack.py \ + gitdb/db/mem.py \ + gitdb/db/loose.py \ + gitdb/db/__init__.py \ + gitdb/db/git.py \ + gitdb/db/base.py \ + gitdb/test/test_util.py \ + gitdb/test/test_stream.py \ + gitdb/test/test_pack.py \ + gitdb/test/test_example.py \ + gitdb/test/test_base.py \ + gitdb/test/lib.py \ + gitdb/test/__init__.py \ + gitdb/test/performance/test_stream.py \ + gitdb/test/performance/test_pack_streaming.py \ + gitdb/test/performance/test_pack.py \ + gitdb/test/performance/lib.py + +HEADERS += \ + gitdb/_delta_apply.h + +SOURCES += \ + gitdb/_fun.c \ + gitdb/_delta_apply.c diff --git a/gitdb/gitdb.pro.user b/gitdb/gitdb.pro.user new file mode 100644 index 000000000..3ca1e2182 --- /dev/null +++ b/gitdb/gitdb.pro.user @@ -0,0 +1,266 @@ + + + + ProjectExplorer.Project.ActiveTarget + 0 + + + ProjectExplorer.Project.EditorSettings + + Default + + + + ProjectExplorer.Project.Target.0 + + Desktop + + Qt4ProjectManager.Target.DesktopTarget + 0 + 0 + 1 + + + 0 + Build + + ProjectExplorer.BuildSteps.Build + + + + Make + + Qt4ProjectManager.MakeStep + true + + clean + + + + 1 + Clean + + ProjectExplorer.BuildSteps.Clean + + 2 + false + + Qt in PATH Release + + Qt4ProjectManager.Qt4BuildConfiguration + 0 + /home/byron/projects/git-python/git/ext/gitdb/gitdbpro-build-desktop + 3 + 0 + false + + + + + qmake + + QtProjectManager.QMakeBuildStep + + + + Make + + Qt4ProjectManager.MakeStep + false + + + + 2 + Build + + ProjectExplorer.BuildSteps.Build + + + + Make + + Qt4ProjectManager.MakeStep + true + + clean + + + + 1 + Clean + + ProjectExplorer.BuildSteps.Clean + + 2 + false + + Qt in PATH Debug + + Qt4ProjectManager.Qt4BuildConfiguration + 2 + /home/byron/projects/git-python/git/ext/gitdb/gitdbpro-build-desktop + 3 + 0 + true + + + + + qmake + + QtProjectManager.QMakeBuildStep + + + + Make + + Qt4ProjectManager.MakeStep + false + + + + 2 + Build + + ProjectExplorer.BuildSteps.Build + + + + Make + + Qt4ProjectManager.MakeStep + true + + clean + + + + 1 + Clean + + ProjectExplorer.BuildSteps.Clean + + 2 + false + + Qt 4.6.2 OpenSource Release + + Qt4ProjectManager.Qt4BuildConfiguration + 0 + /home/byron/projects/git-python/git/ext/gitdb/gitdbpro-build-desktop + 2 + 0 + true + + + + + qmake + + QtProjectManager.QMakeBuildStep + + + + Make + + Qt4ProjectManager.MakeStep + false + + + + 2 + Build + + ProjectExplorer.BuildSteps.Build + + + + Make + + Qt4ProjectManager.MakeStep + true + + clean + + + + 1 + Clean + + ProjectExplorer.BuildSteps.Clean + + 2 + false + + Qt 4.6.2 OpenSource Debug + + Qt4ProjectManager.Qt4BuildConfiguration + 2 + /home/byron/projects/git-python/git/ext/gitdb/gitdbpro-build-desktop + 2 + 0 + true + + 4 + + + 0 + Deploy + + ProjectExplorer.BuildSteps.Deploy + + 1 + No deployment + + ProjectExplorer.DefaultDeployConfiguration + + 1 + + gitdb + + Qt4ProjectManager.Qt4RunConfiguration + 2 + + gitdb.pro + false + false + + false + + 3768 + true + false + + + + /usr/bin/pytest + gitdb/test/test_pack.py + + 2 + python + false + + $BUILDDIR + Run python + test-pack + ProjectExplorer.CustomExecutableRunConfiguration + 3768 + true + false + + 2 + + + + ProjectExplorer.Project.TargetCount + 1 + + + ProjectExplorer.Project.Updater.EnvironmentId + {d38778e3-6b24-4419-8fc3-c8cc320f55e0} + + + ProjectExplorer.Project.Updater.FileVersion + 8 + + diff --git a/gitdb/gitdb/__init__.py b/gitdb/gitdb/__init__.py new file mode 100644 index 000000000..04920428d --- /dev/null +++ b/gitdb/gitdb/__init__.py @@ -0,0 +1,16 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Initialize the object database module""" + +__author__ = "Sebastian Thiel" +__contact__ = "byronimo@gmail.com" +__homepage__ = "https://github.com/gitpython-developers/GitPython/tree/main/gitdb" +version_info = (4, 0, 12) +__version__ = '.'.join(str(i) for i in version_info) + +# default imports +from gitdb.base import * +from gitdb.db import * +from gitdb.stream import * diff --git a/gitdb/gitdb/base.py b/gitdb/gitdb/base.py new file mode 100644 index 000000000..9a23a4f71 --- /dev/null +++ b/gitdb/gitdb/base.py @@ -0,0 +1,315 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Module with basic data structures - they are designed to be lightweight and fast""" +from gitdb.util import bin_to_hex + +from gitdb.fun import ( + type_id_to_type_map, + type_to_type_id_map +) + +__all__ = ('OInfo', 'OPackInfo', 'ODeltaPackInfo', + 'OStream', 'OPackStream', 'ODeltaPackStream', + 'IStream', 'InvalidOInfo', 'InvalidOStream') + +#{ ODB Bases + + +class OInfo(tuple): + + """Carries information about an object in an ODB, providing information + about the binary sha of the object, the type_string as well as the uncompressed size + in bytes. + + It can be accessed using tuple notation and using attribute access notation:: + + assert dbi[0] == dbi.binsha + assert dbi[1] == dbi.type + assert dbi[2] == dbi.size + + The type is designed to be as lightweight as possible.""" + __slots__ = tuple() + + def __new__(cls, sha, type, size): + return tuple.__new__(cls, (sha, type, size)) + + def __init__(self, *args): + tuple.__init__(self) + + #{ Interface + @property + def binsha(self): + """:return: our sha as binary, 20 bytes""" + return self[0] + + @property + def hexsha(self): + """:return: our sha, hex encoded, 40 bytes""" + return bin_to_hex(self[0]) + + @property + def type(self): + return self[1] + + @property + def type_id(self): + return type_to_type_id_map[self[1]] + + @property + def size(self): + return self[2] + #} END interface + + +class OPackInfo(tuple): + + """As OInfo, but provides a type_id property to retrieve the numerical type id, and + does not include a sha. + + Additionally, the pack_offset is the absolute offset into the packfile at which + all object information is located. The data_offset property points to the absolute + location in the pack at which that actual data stream can be found.""" + __slots__ = tuple() + + def __new__(cls, packoffset, type, size): + return tuple.__new__(cls, (packoffset, type, size)) + + def __init__(self, *args): + tuple.__init__(self) + + #{ Interface + + @property + def pack_offset(self): + return self[0] + + @property + def type(self): + return type_id_to_type_map[self[1]] + + @property + def type_id(self): + return self[1] + + @property + def size(self): + return self[2] + + #} END interface + + +class ODeltaPackInfo(OPackInfo): + + """Adds delta specific information, + Either the 20 byte sha which points to some object in the database, + or the negative offset from the pack_offset, so that pack_offset - delta_info yields + the pack offset of the base object""" + __slots__ = tuple() + + def __new__(cls, packoffset, type, size, delta_info): + return tuple.__new__(cls, (packoffset, type, size, delta_info)) + + #{ Interface + @property + def delta_info(self): + return self[3] + #} END interface + + +class OStream(OInfo): + + """Base for object streams retrieved from the database, providing additional + information about the stream. + Generally, ODB streams are read-only as objects are immutable""" + __slots__ = tuple() + + def __new__(cls, sha, type, size, stream, *args, **kwargs): + """Helps with the initialization of subclasses""" + return tuple.__new__(cls, (sha, type, size, stream)) + + def __init__(self, *args, **kwargs): + tuple.__init__(self) + + #{ Stream Reader Interface + + def read(self, size=-1): + return self[3].read(size) + + @property + def stream(self): + return self[3] + + #} END stream reader interface + + +class ODeltaStream(OStream): + + """Uses size info of its stream, delaying reads""" + + def __new__(cls, sha, type, size, stream, *args, **kwargs): + """Helps with the initialization of subclasses""" + return tuple.__new__(cls, (sha, type, size, stream)) + + #{ Stream Reader Interface + + @property + def size(self): + return self[3].size + + #} END stream reader interface + + +class OPackStream(OPackInfo): + + """Next to pack object information, a stream outputting an undeltified base object + is provided""" + __slots__ = tuple() + + def __new__(cls, packoffset, type, size, stream, *args): + """Helps with the initialization of subclasses""" + return tuple.__new__(cls, (packoffset, type, size, stream)) + + #{ Stream Reader Interface + def read(self, size=-1): + return self[3].read(size) + + @property + def stream(self): + return self[3] + #} END stream reader interface + + +class ODeltaPackStream(ODeltaPackInfo): + + """Provides a stream outputting the uncompressed offset delta information""" + __slots__ = tuple() + + def __new__(cls, packoffset, type, size, delta_info, stream): + return tuple.__new__(cls, (packoffset, type, size, delta_info, stream)) + + #{ Stream Reader Interface + def read(self, size=-1): + return self[4].read(size) + + @property + def stream(self): + return self[4] + #} END stream reader interface + + +class IStream(list): + + """Represents an input content stream to be fed into the ODB. It is mutable to allow + the ODB to record information about the operations outcome right in this instance. + + It provides interfaces for the OStream and a StreamReader to allow the instance + to blend in without prior conversion. + + The only method your content stream must support is 'read'""" + __slots__ = tuple() + + def __new__(cls, type, size, stream, sha=None): + return list.__new__(cls, (sha, type, size, stream, None)) + + def __init__(self, type, size, stream, sha=None): + list.__init__(self, (sha, type, size, stream, None)) + + #{ Interface + @property + def hexsha(self): + """:return: our sha, hex encoded, 40 bytes""" + return bin_to_hex(self[0]) + + def _error(self): + """:return: the error that occurred when processing the stream, or None""" + return self[4] + + def _set_error(self, exc): + """Set this input stream to the given exc, may be None to reset the error""" + self[4] = exc + + error = property(_error, _set_error) + + #} END interface + + #{ Stream Reader Interface + + def read(self, size=-1): + """Implements a simple stream reader interface, passing the read call on + to our internal stream""" + return self[3].read(size) + + #} END stream reader interface + + #{ interface + + def _set_binsha(self, binsha): + self[0] = binsha + + def _binsha(self): + return self[0] + + binsha = property(_binsha, _set_binsha) + + def _type(self): + return self[1] + + def _set_type(self, type): + self[1] = type + + type = property(_type, _set_type) + + def _size(self): + return self[2] + + def _set_size(self, size): + self[2] = size + + size = property(_size, _set_size) + + def _stream(self): + return self[3] + + def _set_stream(self, stream): + self[3] = stream + + stream = property(_stream, _set_stream) + + #} END odb info interface + + +class InvalidOInfo(tuple): + + """Carries information about a sha identifying an object which is invalid in + the queried database. The exception attribute provides more information about + the cause of the issue""" + __slots__ = tuple() + + def __new__(cls, sha, exc): + return tuple.__new__(cls, (sha, exc)) + + def __init__(self, sha, exc): + tuple.__init__(self, (sha, exc)) + + @property + def binsha(self): + return self[0] + + @property + def hexsha(self): + return bin_to_hex(self[0]) + + @property + def error(self): + """:return: exception instance explaining the failure""" + return self[1] + + +class InvalidOStream(InvalidOInfo): + + """Carries information about an invalid ODB stream""" + __slots__ = tuple() + +#} END ODB Bases diff --git a/gitdb/gitdb/const.py b/gitdb/gitdb/const.py new file mode 100644 index 000000000..6391d796f --- /dev/null +++ b/gitdb/gitdb/const.py @@ -0,0 +1,4 @@ +BYTE_SPACE = b' ' +NULL_BYTE = b'\0' +NULL_HEX_SHA = "0" * 40 +NULL_BIN_SHA = NULL_BYTE * 20 diff --git a/gitdb/gitdb/db/__init__.py b/gitdb/gitdb/db/__init__.py new file mode 100644 index 000000000..20fd2280d --- /dev/null +++ b/gitdb/gitdb/db/__init__.py @@ -0,0 +1,11 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ + +from gitdb.db.base import * +from gitdb.db.loose import * +from gitdb.db.mem import * +from gitdb.db.pack import * +from gitdb.db.git import * +from gitdb.db.ref import * diff --git a/gitdb/gitdb/db/base.py b/gitdb/gitdb/db/base.py new file mode 100644 index 000000000..7312fe00f --- /dev/null +++ b/gitdb/gitdb/db/base.py @@ -0,0 +1,278 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Contains implementations of database retrieveing objects""" +from gitdb.util import ( + join, + LazyMixin, + hex_to_bin +) + +from gitdb.utils.encoding import force_text +from gitdb.exc import ( + BadObject, + AmbiguousObjectName +) + +from itertools import chain +from functools import reduce + + +__all__ = ('ObjectDBR', 'ObjectDBW', 'FileDBBase', 'CompoundDB', 'CachingDB') + + +class ObjectDBR: + + """Defines an interface for object database lookup. + Objects are identified either by their 20 byte bin sha""" + + def __contains__(self, sha): + return self.has_obj + + #{ Query Interface + def has_object(self, sha): + """ + Whether the object identified by the given 20 bytes + binary sha is contained in the database + + :return: True if the object identified by the given 20 bytes + binary sha is contained in the database""" + raise NotImplementedError("To be implemented in subclass") + + def info(self, sha): + """ :return: OInfo instance + :param sha: bytes binary sha + :raise BadObject:""" + raise NotImplementedError("To be implemented in subclass") + + def stream(self, sha): + """:return: OStream instance + :param sha: 20 bytes binary sha + :raise BadObject:""" + raise NotImplementedError("To be implemented in subclass") + + def size(self): + """:return: amount of objects in this database""" + raise NotImplementedError() + + def sha_iter(self): + """Return iterator yielding 20 byte shas for all objects in this data base""" + raise NotImplementedError() + + #} END query interface + + +class ObjectDBW: + + """Defines an interface to create objects in the database""" + + def __init__(self, *args, **kwargs): + self._ostream = None + + #{ Edit Interface + def set_ostream(self, stream): + """ + Adjusts the stream to which all data should be sent when storing new objects + + :param stream: if not None, the stream to use, if None the default stream + will be used. + :return: previously installed stream, or None if there was no override + :raise TypeError: if the stream doesn't have the supported functionality""" + cstream = self._ostream + self._ostream = stream + return cstream + + def ostream(self): + """ + Return the output stream + + :return: overridden output stream this instance will write to, or None + if it will write to the default stream""" + return self._ostream + + def store(self, istream): + """ + Create a new object in the database + :return: the input istream object with its sha set to its corresponding value + + :param istream: IStream compatible instance. If its sha is already set + to a value, the object will just be stored in the our database format, + in which case the input stream is expected to be in object format ( header + contents ). + :raise IOError: if data could not be written""" + raise NotImplementedError("To be implemented in subclass") + + #} END edit interface + + +class FileDBBase: + + """Provides basic facilities to retrieve files of interest, including + caching facilities to help mapping hexsha's to objects""" + + def __init__(self, root_path): + """Initialize this instance to look for its files at the given root path + All subsequent operations will be relative to this path + :raise InvalidDBRoot: + **Note:** The base will not perform any accessablity checking as the base + might not yet be accessible, but become accessible before the first + access.""" + super().__init__() + self._root_path = root_path + + #{ Interface + def root_path(self): + """:return: path at which this db operates""" + return self._root_path + + def db_path(self, rela_path): + """ + :return: the given relative path relative to our database root, allowing + to pontentially access datafiles""" + return join(self._root_path, force_text(rela_path)) + #} END interface + + +class CachingDB: + + """A database which uses caches to speed-up access""" + + #{ Interface + def update_cache(self, force=False): + """ + Call this method if the underlying data changed to trigger an update + of the internal caching structures. + + :param force: if True, the update must be performed. Otherwise the implementation + may decide not to perform an update if it thinks nothing has changed. + :return: True if an update was performed as something change indeed""" + + # END interface + + +def _databases_recursive(database, output): + """Fill output list with database from db, in order. Deals with Loose, Packed + and compound databases.""" + if isinstance(database, CompoundDB): + dbs = database.databases() + output.extend(db for db in dbs if not isinstance(db, CompoundDB)) + for cdb in (db for db in dbs if isinstance(db, CompoundDB)): + _databases_recursive(cdb, output) + else: + output.append(database) + # END handle database type + + +class CompoundDB(ObjectDBR, LazyMixin, CachingDB): + + """A database which delegates calls to sub-databases. + + Databases are stored in the lazy-loaded _dbs attribute. + Define _set_cache_ to update it with your databases""" + + def _set_cache_(self, attr): + if attr == '_dbs': + self._dbs = list() + elif attr == '_db_cache': + self._db_cache = dict() + else: + super()._set_cache_(attr) + + def _db_query(self, sha): + """:return: database containing the given 20 byte sha + :raise BadObject:""" + # most databases use binary representations, prevent converting + # it every time a database is being queried + try: + return self._db_cache[sha] + except KeyError: + pass + # END first level cache + + for db in self._dbs: + if db.has_object(sha): + self._db_cache[sha] = db + return db + # END for each database + raise BadObject(sha) + + #{ ObjectDBR interface + + def has_object(self, sha): + try: + self._db_query(sha) + return True + except BadObject: + return False + # END handle exceptions + + def info(self, sha): + return self._db_query(sha).info(sha) + + def stream(self, sha): + return self._db_query(sha).stream(sha) + + def size(self): + """:return: total size of all contained databases""" + return reduce(lambda x, y: x + y, (db.size() for db in self._dbs), 0) + + def sha_iter(self): + return chain(*(db.sha_iter() for db in self._dbs)) + + #} END object DBR Interface + + #{ Interface + + def databases(self): + """:return: tuple of database instances we use for lookups""" + return tuple(self._dbs) + + def update_cache(self, force=False): + # something might have changed, clear everything + self._db_cache.clear() + stat = False + for db in self._dbs: + if isinstance(db, CachingDB): + stat |= db.update_cache(force) + # END if is caching db + # END for each database to update + return stat + + def partial_to_complete_sha_hex(self, partial_hexsha): + """ + :return: 20 byte binary sha1 from the given less-than-40 byte hexsha (bytes or str) + :param partial_hexsha: hexsha with less than 40 byte + :raise AmbiguousObjectName: """ + databases = list() + _databases_recursive(self, databases) + partial_hexsha = force_text(partial_hexsha) + len_partial_hexsha = len(partial_hexsha) + if len_partial_hexsha % 2 != 0: + partial_binsha = hex_to_bin(partial_hexsha + "0") + else: + partial_binsha = hex_to_bin(partial_hexsha) + # END assure successful binary conversion + + candidate = None + for db in databases: + full_bin_sha = None + try: + if hasattr(db, 'partial_to_complete_sha_hex'): + full_bin_sha = db.partial_to_complete_sha_hex(partial_hexsha) + else: + full_bin_sha = db.partial_to_complete_sha(partial_binsha, len_partial_hexsha) + # END handle database type + except BadObject: + continue + # END ignore bad objects + if full_bin_sha: + if candidate and candidate != full_bin_sha: + raise AmbiguousObjectName(partial_hexsha) + candidate = full_bin_sha + # END handle candidate + # END for each db + if not candidate: + raise BadObject(partial_binsha) + return candidate + + #} END interface diff --git a/gitdb/gitdb/db/git.py b/gitdb/gitdb/db/git.py new file mode 100644 index 000000000..a1ed14280 --- /dev/null +++ b/gitdb/gitdb/db/git.py @@ -0,0 +1,85 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +from gitdb.db.base import ( + CompoundDB, + ObjectDBW, + FileDBBase +) + +from gitdb.db.loose import LooseObjectDB +from gitdb.db.pack import PackedDB +from gitdb.db.ref import ReferenceDB + +from gitdb.exc import InvalidDBRoot + +import os + +__all__ = ('GitDB', ) + + +class GitDB(FileDBBase, ObjectDBW, CompoundDB): + + """A git-style object database, which contains all objects in the 'objects' + subdirectory + + ``IMPORTANT``: The usage of this implementation is highly discouraged as it fails to release file-handles. + This can be a problem with long-running processes and/or big repositories. + """ + # Configuration + PackDBCls = PackedDB + LooseDBCls = LooseObjectDB + ReferenceDBCls = ReferenceDB + + # Directories + packs_dir = 'pack' + loose_dir = '' + alternates_dir = os.path.join('info', 'alternates') + + def __init__(self, root_path): + """Initialize ourselves on a git objects directory""" + super().__init__(root_path) + + def _set_cache_(self, attr): + if attr == '_dbs' or attr == '_loose_db': + self._dbs = list() + loose_db = None + for subpath, dbcls in ((self.packs_dir, self.PackDBCls), + (self.loose_dir, self.LooseDBCls), + (self.alternates_dir, self.ReferenceDBCls)): + path = self.db_path(subpath) + if os.path.exists(path): + self._dbs.append(dbcls(path)) + if dbcls is self.LooseDBCls: + loose_db = self._dbs[-1] + # END remember loose db + # END check path exists + # END for each db type + + # should have at least one subdb + if not self._dbs: + raise InvalidDBRoot(self.root_path()) + # END handle error + + # we the first one should have the store method + assert loose_db is not None and hasattr(loose_db, 'store'), "First database needs store functionality" + + # finally set the value + self._loose_db = loose_db + else: + super()._set_cache_(attr) + # END handle attrs + + #{ ObjectDBW interface + + def store(self, istream): + return self._loose_db.store(istream) + + def ostream(self): + return self._loose_db.ostream() + + def set_ostream(self, ostream): + return self._loose_db.set_ostream(ostream) + + #} END objectdbw interface diff --git a/gitdb/gitdb/db/loose.py b/gitdb/gitdb/db/loose.py new file mode 100644 index 000000000..e6765cdeb --- /dev/null +++ b/gitdb/gitdb/db/loose.py @@ -0,0 +1,268 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +from contextlib import suppress + +from gitdb.db.base import ( + FileDBBase, + ObjectDBR, + ObjectDBW +) + +from gitdb.exc import ( + BadObject, + AmbiguousObjectName +) + +from gitdb.stream import ( + DecompressMemMapReader, + FDCompressedSha1Writer, + FDStream, + Sha1Writer +) + +from gitdb.base import ( + OStream, + OInfo +) + +from gitdb.util import ( + file_contents_ro_filepath, + ENOENT, + hex_to_bin, + bin_to_hex, + exists, + chmod, + isfile, + remove, + rename, + dirname, + basename, + join +) + +from gitdb.fun import ( + chunk_size, + loose_object_header_info, + write_object, + stream_copy +) + +from gitdb.utils.encoding import force_bytes + +import tempfile +import os +import sys +import time + + +__all__ = ('LooseObjectDB', ) + + +class LooseObjectDB(FileDBBase, ObjectDBR, ObjectDBW): + + """A database which operates on loose object files""" + + # CONFIGURATION + # chunks in which data will be copied between streams + stream_chunk_size = chunk_size + + # On windows we need to keep it writable, otherwise it cannot be removed + # either + new_objects_mode = int("444", 8) + if os.name == 'nt': + new_objects_mode = int("644", 8) + + def __init__(self, root_path): + super().__init__(root_path) + self._hexsha_to_file = dict() + # Additional Flags - might be set to 0 after the first failure + # Depending on the root, this might work for some mounts, for others not, which + # is why it is per instance + self._fd_open_flags = getattr(os, 'O_NOATIME', 0) + + #{ Interface + def object_path(self, hexsha): + """ + :return: path at which the object with the given hexsha would be stored, + relative to the database root""" + return join(hexsha[:2], hexsha[2:]) + + def readable_db_object_path(self, hexsha): + """ + :return: readable object path to the object identified by hexsha + :raise BadObject: If the object file does not exist""" + with suppress(KeyError): + return self._hexsha_to_file[hexsha] + # END ignore cache misses + + # try filesystem + path = self.db_path(self.object_path(hexsha)) + if exists(path): + self._hexsha_to_file[hexsha] = path + return path + # END handle cache + raise BadObject(hexsha) + + def partial_to_complete_sha_hex(self, partial_hexsha): + """:return: 20 byte binary sha1 string which matches the given name uniquely + :param name: hexadecimal partial name (bytes or ascii string) + :raise AmbiguousObjectName: + :raise BadObject: """ + candidate = None + for binsha in self.sha_iter(): + if bin_to_hex(binsha).startswith(force_bytes(partial_hexsha)): + # it can't ever find the same object twice + if candidate is not None: + raise AmbiguousObjectName(partial_hexsha) + candidate = binsha + # END for each object + if candidate is None: + raise BadObject(partial_hexsha) + return candidate + + #} END interface + + def _map_loose_object(self, sha): + """ + :return: memory map of that file to allow random read access + :raise BadObject: if object could not be located""" + db_path = self.db_path(self.object_path(bin_to_hex(sha))) + try: + return file_contents_ro_filepath(db_path, flags=self._fd_open_flags) + except OSError as e: + if e.errno != ENOENT: + # try again without noatime + try: + return file_contents_ro_filepath(db_path) + except OSError as new_e: + raise BadObject(sha) from new_e + # didn't work because of our flag, don't try it again + self._fd_open_flags = 0 + else: + raise BadObject(sha) from e + # END handle error + # END exception handling + + def set_ostream(self, stream): + """:raise TypeError: if the stream does not support the Sha1Writer interface""" + if stream is not None and not isinstance(stream, Sha1Writer): + raise TypeError("Output stream musst support the %s interface" % Sha1Writer.__name__) + return super().set_ostream(stream) + + def info(self, sha): + m = self._map_loose_object(sha) + try: + typ, size = loose_object_header_info(m) + return OInfo(sha, typ, size) + finally: + if hasattr(m, 'close'): + m.close() + # END assure release of system resources + + def stream(self, sha): + m = self._map_loose_object(sha) + type, size, stream = DecompressMemMapReader.new(m, close_on_deletion=True) + return OStream(sha, type, size, stream) + + def has_object(self, sha): + try: + self.readable_db_object_path(bin_to_hex(sha)) + return True + except BadObject: + return False + # END check existence + + def store(self, istream): + """note: The sha we produce will be hex by nature""" + tmp_path = None + writer = self.ostream() + if writer is None: + # open a tmp file to write the data to + fd, tmp_path = tempfile.mkstemp(prefix='obj', dir=self._root_path) + + if istream.binsha is None: + writer = FDCompressedSha1Writer(fd) + else: + writer = FDStream(fd) + # END handle direct stream copies + # END handle custom writer + + try: + try: + if istream.binsha is not None: + # copy as much as possible, the actual uncompressed item size might + # be smaller than the compressed version + stream_copy(istream.read, writer.write, sys.maxsize, self.stream_chunk_size) + else: + # write object with header, we have to make a new one + write_object(istream.type, istream.size, istream.read, writer.write, + chunk_size=self.stream_chunk_size) + # END handle direct stream copies + finally: + if tmp_path: + writer.close() + # END assure target stream is closed + except: + if tmp_path: + remove(tmp_path) + raise + # END assure tmpfile removal on error + + hexsha = None + if istream.binsha: + hexsha = istream.hexsha + else: + hexsha = writer.sha(as_hex=True) + # END handle sha + + if tmp_path: + obj_path = self.db_path(self.object_path(hexsha)) + obj_dir = dirname(obj_path) + os.makedirs(obj_dir, exist_ok=True) + # END handle destination directory + # rename onto existing doesn't work on NTFS + if isfile(obj_path): + remove(tmp_path) + else: + rename(tmp_path, obj_path) + # end rename only if needed + + # Ensure rename is actually done and file is stable + # Retry up to 14 times - quadratic wait & retry in ms. + # The total maximum wait time is 1000ms, which should be vastly enough for the + # OS to return and commit the file to disk. + for backoff_ms in [1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121, 144, 169, 181]: + with suppress(PermissionError): + # make sure its readable for all ! It started out as rw-- tmp file + # but needs to be rwrr + chmod(obj_path, self.new_objects_mode) + break + time.sleep(backoff_ms / 1000.0) + else: + raise PermissionError( + "Impossible to apply `chmod` to file {}".format(obj_path) + ) + + # END handle dry_run + + istream.binsha = hex_to_bin(hexsha) + return istream + + def sha_iter(self): + # find all files which look like an object, extract sha from there + for root, dirs, files in os.walk(self.root_path()): + root_base = basename(root) + if len(root_base) != 2: + continue + + for f in files: + if len(f) != 38: + continue + yield hex_to_bin(root_base + f) + # END for each file + # END for each walk iteration + + def size(self): + return len(tuple(self.sha_iter())) diff --git a/gitdb/gitdb/db/mem.py b/gitdb/gitdb/db/mem.py new file mode 100644 index 000000000..d4772fdb5 --- /dev/null +++ b/gitdb/gitdb/db/mem.py @@ -0,0 +1,110 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Contains the MemoryDatabase implementation""" +from gitdb.db.loose import LooseObjectDB +from gitdb.db.base import ( + ObjectDBR, + ObjectDBW +) + +from gitdb.base import ( + OStream, + IStream, +) + +from gitdb.exc import ( + BadObject, + UnsupportedOperation +) + +from gitdb.stream import ( + ZippedStoreShaWriter, + DecompressMemMapReader, +) + +from io import BytesIO + +__all__ = ("MemoryDB", ) + + +class MemoryDB(ObjectDBR, ObjectDBW): + + """A memory database stores everything to memory, providing fast IO and object + retrieval. It should be used to buffer results and obtain SHAs before writing + it to the actual physical storage, as it allows to query whether object already + exists in the target storage before introducing actual IO""" + + def __init__(self): + super().__init__() + self._db = LooseObjectDB("path/doesnt/matter") + + # maps 20 byte shas to their OStream objects + self._cache = dict() + + def set_ostream(self, stream): + raise UnsupportedOperation("MemoryDB's always stream into memory") + + def store(self, istream): + zstream = ZippedStoreShaWriter() + self._db.set_ostream(zstream) + + istream = self._db.store(istream) + zstream.close() # close to flush + zstream.seek(0) + + # don't provide a size, the stream is written in object format, hence the + # header needs decompression + decomp_stream = DecompressMemMapReader(zstream.getvalue(), close_on_deletion=False) + self._cache[istream.binsha] = OStream(istream.binsha, istream.type, istream.size, decomp_stream) + + return istream + + def has_object(self, sha): + return sha in self._cache + + def info(self, sha): + # we always return streams, which are infos as well + return self.stream(sha) + + def stream(self, sha): + try: + ostream = self._cache[sha] + # rewind stream for the next one to read + ostream.stream.seek(0) + return ostream + except KeyError as e: + raise BadObject(sha) from e + # END exception handling + + def size(self): + return len(self._cache) + + def sha_iter(self): + return self._cache.keys() + + #{ Interface + def stream_copy(self, sha_iter, odb): + """Copy the streams as identified by sha's yielded by sha_iter into the given odb + The streams will be copied directly + **Note:** the object will only be written if it did not exist in the target db + + :return: amount of streams actually copied into odb. If smaller than the amount + of input shas, one or more objects did already exist in odb""" + count = 0 + for sha in sha_iter: + if odb.has_object(sha): + continue + # END check object existence + + ostream = self.stream(sha) + # compressed data including header + sio = BytesIO(ostream.stream.data()) + istream = IStream(ostream.type, ostream.size, sio, sha) + + odb.store(istream) + count += 1 + # END for each sha + return count + #} END interface diff --git a/gitdb/gitdb/db/pack.py b/gitdb/gitdb/db/pack.py new file mode 100644 index 000000000..274ea5990 --- /dev/null +++ b/gitdb/gitdb/db/pack.py @@ -0,0 +1,206 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Module containing a database to deal with packs""" +from gitdb.db.base import ( + FileDBBase, + ObjectDBR, + CachingDB +) + +from gitdb.util import LazyMixin + +from gitdb.exc import ( + BadObject, + UnsupportedOperation, + AmbiguousObjectName +) + +from gitdb.pack import PackEntity + +from functools import reduce + +import os +import glob + +__all__ = ('PackedDB', ) + +#{ Utilities + + +class PackedDB(FileDBBase, ObjectDBR, CachingDB, LazyMixin): + + """A database operating on a set of object packs""" + + # sort the priority list every N queries + # Higher values are better, performance tests don't show this has + # any effect, but it should have one + _sort_interval = 500 + + def __init__(self, root_path): + super().__init__(root_path) + # list of lists with three items: + # * hits - number of times the pack was hit with a request + # * entity - Pack entity instance + # * sha_to_index - PackIndexFile.sha_to_index method for direct cache query + # self._entities = list() # lazy loaded list + self._hit_count = 0 # amount of hits + self._st_mtime = 0 # last modification data of our root path + + def _set_cache_(self, attr): + if attr == '_entities': + self._entities = list() + self.update_cache(force=True) + # END handle entities initialization + + def _sort_entities(self): + self._entities.sort(key=lambda l: l[0], reverse=True) + + def _pack_info(self, sha): + """:return: tuple(entity, index) for an item at the given sha + :param sha: 20 or 40 byte sha + :raise BadObject: + **Note:** This method is not thread-safe, but may be hit in multi-threaded + operation. The worst thing that can happen though is a counter that + was not incremented, or the list being in wrong order. So we safe + the time for locking here, lets see how that goes""" + # presort ? + if self._hit_count % self._sort_interval == 0: + self._sort_entities() + # END update sorting + + for item in self._entities: + index = item[2](sha) + if index is not None: + item[0] += 1 # one hit for you + self._hit_count += 1 # general hit count + return (item[1], index) + # END index found in pack + # END for each item + + # no hit, see whether we have to update packs + # NOTE: considering packs don't change very often, we safe this call + # and leave it to the super-caller to trigger that + raise BadObject(sha) + + #{ Object DB Read + + def has_object(self, sha): + try: + self._pack_info(sha) + return True + except BadObject: + return False + # END exception handling + + def info(self, sha): + entity, index = self._pack_info(sha) + return entity.info_at_index(index) + + def stream(self, sha): + entity, index = self._pack_info(sha) + return entity.stream_at_index(index) + + def sha_iter(self): + for entity in self.entities(): + index = entity.index() + sha_by_index = index.sha + for index in range(index.size()): + yield sha_by_index(index) + # END for each index + # END for each entity + + def size(self): + sizes = [item[1].index().size() for item in self._entities] + return reduce(lambda x, y: x + y, sizes, 0) + + #} END object db read + + #{ object db write + + def store(self, istream): + """Storing individual objects is not feasible as a pack is designed to + hold multiple objects. Writing or rewriting packs for single objects is + inefficient""" + raise UnsupportedOperation() + + #} END object db write + + #{ Interface + + def update_cache(self, force=False): + """ + Update our cache with the actually existing packs on disk. Add new ones, + and remove deleted ones. We keep the unchanged ones + + :param force: If True, the cache will be updated even though the directory + does not appear to have changed according to its modification timestamp. + :return: True if the packs have been updated so there is new information, + False if there was no change to the pack database""" + stat = os.stat(self.root_path()) + if not force and stat.st_mtime <= self._st_mtime: + return False + # END abort early on no change + self._st_mtime = stat.st_mtime + + # packs are supposed to be prefixed with pack- by git-convention + # get all pack files, figure out what changed + pack_files = set(glob.glob(os.path.join(self.root_path(), "pack-*.pack"))) + our_pack_files = {item[1].pack().path() for item in self._entities} + + # new packs + for pack_file in (pack_files - our_pack_files): + # init the hit-counter/priority with the size, a good measure for hit- + # probability. Its implemented so that only 12 bytes will be read + entity = PackEntity(pack_file) + self._entities.append([entity.pack().size(), entity, entity.index().sha_to_index]) + # END for each new packfile + + # removed packs + for pack_file in (our_pack_files - pack_files): + del_index = -1 + for i, item in enumerate(self._entities): + if item[1].pack().path() == pack_file: + del_index = i + break + # END found index + # END for each entity + assert del_index != -1 + del(self._entities[del_index]) + # END for each removed pack + + # reinitialize prioritiess + self._sort_entities() + return True + + def entities(self): + """:return: list of pack entities operated upon by this database""" + return [item[1] for item in self._entities] + + def partial_to_complete_sha(self, partial_binsha, canonical_length): + """:return: 20 byte sha as inferred by the given partial binary sha + :param partial_binsha: binary sha with less than 20 bytes + :param canonical_length: length of the corresponding canonical representation. + It is required as binary sha's cannot display whether the original hex sha + had an odd or even number of characters + :raise AmbiguousObjectName: + :raise BadObject: """ + candidate = None + for item in self._entities: + item_index = item[1].index().partial_sha_to_index(partial_binsha, canonical_length) + if item_index is not None: + sha = item[1].index().sha(item_index) + if candidate and candidate != sha: + raise AmbiguousObjectName(partial_binsha) + candidate = sha + # END handle full sha could be found + # END for each entity + + if candidate: + return candidate + + # still not found ? + raise BadObject(partial_binsha) + + #} END interface diff --git a/gitdb/gitdb/db/ref.py b/gitdb/gitdb/db/ref.py new file mode 100644 index 000000000..5536db0f2 --- /dev/null +++ b/gitdb/gitdb/db/ref.py @@ -0,0 +1,81 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +from gitdb.db.base import ( + CompoundDB, +) + +__all__ = ('ReferenceDB', ) + + +class ReferenceDB(CompoundDB): + + """A database consisting of database referred to in a file""" + + # Configuration + # Specifies the object database to use for the paths found in the alternates + # file. If None, it defaults to the GitDB + ObjectDBCls = None + + def __init__(self, ref_file): + super().__init__() + self._ref_file = ref_file + + def _set_cache_(self, attr): + if attr == '_dbs': + self._dbs = list() + self._update_dbs_from_ref_file() + else: + super()._set_cache_(attr) + # END handle attrs + + def _update_dbs_from_ref_file(self): + dbcls = self.ObjectDBCls + if dbcls is None: + # late import + from gitdb.db.git import GitDB + dbcls = GitDB + # END get db type + + # try to get as many as possible, don't fail if some are unavailable + ref_paths = list() + try: + with open(self._ref_file, 'r', encoding="utf-8") as f: + ref_paths = [l.strip() for l in f] + except OSError: + pass + # END handle alternates + + ref_paths_set = set(ref_paths) + cur_ref_paths_set = {db.root_path() for db in self._dbs} + + # remove existing + for path in (cur_ref_paths_set - ref_paths_set): + for i, db in enumerate(self._dbs[:]): + if db.root_path() == path: + del(self._dbs[i]) + continue + # END del matching db + # END for each path to remove + + # add new + # sort them to maintain order + added_paths = sorted(ref_paths_set - cur_ref_paths_set, key=lambda p: ref_paths.index(p)) + for path in added_paths: + try: + db = dbcls(path) + # force an update to verify path + if isinstance(db, CompoundDB): + db.databases() + # END verification + self._dbs.append(db) + except Exception: + # ignore invalid paths or issues + pass + # END for each path to add + + def update_cache(self, force=False): + # re-read alternates and update databases + self._update_dbs_from_ref_file() + return super().update_cache(force) diff --git a/gitdb/gitdb/exc.py b/gitdb/gitdb/exc.py new file mode 100644 index 000000000..752dafdbb --- /dev/null +++ b/gitdb/gitdb/exc.py @@ -0,0 +1,57 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Module with common exceptions""" +from gitdb.util import to_hex_sha + +__all__ = [ + 'AmbiguousObjectName', + 'BadName', + 'BadObject', + 'BadObjectType', + 'InvalidDBRoot', + 'ODBError', + 'ParseError', + 'UnsupportedOperation', + 'to_hex_sha', +] + +class ODBError(Exception): + """All errors thrown by the object database""" + + +class InvalidDBRoot(ODBError): + """Thrown if an object database cannot be initialized at the given path""" + + +class BadObject(ODBError): + """The object with the given SHA does not exist. Instantiate with the + failed sha""" + + def __str__(self): + return "BadObject: %s" % to_hex_sha(self.args[0]) + + +class BadName(ODBError): + """A name provided to rev_parse wasn't understood""" + + def __str__(self): + return "Ref '%s' did not resolve to an object" % self.args[0] + + +class ParseError(ODBError): + """Thrown if the parsing of a file failed due to an invalid format""" + + +class AmbiguousObjectName(ODBError): + """Thrown if a possibly shortened name does not uniquely represent a single object + in the database""" + + +class BadObjectType(ODBError): + """The object had an unsupported type""" + + +class UnsupportedOperation(ODBError): + """Thrown if the given operation cannot be supported by the object database""" diff --git a/gitdb/gitdb/fun.py b/gitdb/gitdb/fun.py new file mode 100644 index 000000000..a272e5caa --- /dev/null +++ b/gitdb/gitdb/fun.py @@ -0,0 +1,704 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Contains basic c-functions which usually contain performance critical code +Keeping this code separate from the beginning makes it easier to out-source +it into c later, if required""" + +import zlib +from gitdb.util import byte_ord +decompressobj = zlib.decompressobj + +import mmap +from itertools import islice +from functools import reduce + +from gitdb.const import NULL_BYTE, BYTE_SPACE +from gitdb.utils.encoding import force_text +from gitdb.typ import ( + str_blob_type, + str_commit_type, + str_tree_type, + str_tag_type, +) + +from io import StringIO + +# INVARIANTS +OFS_DELTA = 6 +REF_DELTA = 7 +delta_types = (OFS_DELTA, REF_DELTA) + +type_id_to_type_map = { + 0: b'', # EXT 1 + 1: str_commit_type, + 2: str_tree_type, + 3: str_blob_type, + 4: str_tag_type, + 5: b'', # EXT 2 + OFS_DELTA: "OFS_DELTA", # OFFSET DELTA + REF_DELTA: "REF_DELTA" # REFERENCE DELTA +} + +type_to_type_id_map = { + str_commit_type: 1, + str_tree_type: 2, + str_blob_type: 3, + str_tag_type: 4, + "OFS_DELTA": OFS_DELTA, + "REF_DELTA": REF_DELTA, +} + +# used when dealing with larger streams +chunk_size = 1000 * mmap.PAGESIZE + +__all__ = ('is_loose_object', 'loose_object_header_info', 'msb_size', 'pack_object_header_info', + 'write_object', 'loose_object_header', 'stream_copy', 'apply_delta_data', + 'is_equal_canonical_sha', 'connect_deltas', 'DeltaChunkList', 'create_pack_object_header') + + +#{ Structures + +def _set_delta_rbound(d, size): + """Truncate the given delta to the given size + :param size: size relative to our target offset, may not be 0, must be smaller or equal + to our size + :return: d""" + d.ts = size + + # NOTE: data is truncated automatically when applying the delta + # MUST NOT DO THIS HERE + return d + + +def _move_delta_lbound(d, bytes): + """Move the delta by the given amount of bytes, reducing its size so that its + right bound stays static + :param bytes: amount of bytes to move, must be smaller than delta size + :return: d""" + if bytes == 0: + return + + d.to += bytes + d.so += bytes + d.ts -= bytes + if d.data is not None: + d.data = d.data[bytes:] + # END handle data + + return d + + +def delta_duplicate(src): + return DeltaChunk(src.to, src.ts, src.so, src.data) + + +def delta_chunk_apply(dc, bbuf, write): + """Apply own data to the target buffer + :param bbuf: buffer providing source bytes for copy operations + :param write: write method to call with data to write""" + if dc.data is None: + # COPY DATA FROM SOURCE + write(bbuf[dc.so:dc.so + dc.ts]) + else: + # APPEND DATA + # what's faster: if + 4 function calls or just a write with a slice ? + # Considering data can be larger than 127 bytes now, it should be worth it + if dc.ts < len(dc.data): + write(dc.data[:dc.ts]) + else: + write(dc.data) + # END handle truncation + # END handle chunk mode + + +class DeltaChunk: + + """Represents a piece of a delta, it can either add new data, or copy existing + one from a source buffer""" + __slots__ = ( + 'to', # start offset in the target buffer in bytes + 'ts', # size of this chunk in the target buffer in bytes + 'so', # start offset in the source buffer in bytes or None + 'data', # chunk of bytes to be added to the target buffer, + # DeltaChunkList to use as base, or None + ) + + def __init__(self, to, ts, so, data): + self.to = to + self.ts = ts + self.so = so + self.data = data + + def __repr__(self): + return "DeltaChunk(%i, %i, %s, %s)" % (self.to, self.ts, self.so, self.data or "") + + #{ Interface + + def rbound(self): + return self.to + self.ts + + def has_data(self): + """:return: True if the instance has data to add to the target stream""" + return self.data is not None + + #} END interface + + +def _closest_index(dcl, absofs): + """:return: index at which the given absofs should be inserted. The index points + to the DeltaChunk with a target buffer absofs that equals or is greater than + absofs. + **Note:** global method for performance only, it belongs to DeltaChunkList""" + lo = 0 + hi = len(dcl) + while lo < hi: + mid = (lo + hi) / 2 + dc = dcl[mid] + if dc.to > absofs: + hi = mid + elif dc.rbound() > absofs or dc.to == absofs: + return mid + else: + lo = mid + 1 + # END handle bound + # END for each delta absofs + return len(dcl) - 1 + + +def delta_list_apply(dcl, bbuf, write): + """Apply the chain's changes and write the final result using the passed + write function. + :param bbuf: base buffer containing the base of all deltas contained in this + list. It will only be used if the chunk in question does not have a base + chain. + :param write: function taking a string of bytes to write to the output""" + for dc in dcl: + delta_chunk_apply(dc, bbuf, write) + # END for each dc + + +def delta_list_slice(dcl, absofs, size, ndcl): + """:return: Subsection of this list at the given absolute offset, with the given + size in bytes. + :return: None""" + cdi = _closest_index(dcl, absofs) # delta start index + cd = dcl[cdi] + slen = len(dcl) + lappend = ndcl.append + + if cd.to != absofs: + tcd = DeltaChunk(cd.to, cd.ts, cd.so, cd.data) + _move_delta_lbound(tcd, absofs - cd.to) + tcd.ts = min(tcd.ts, size) + lappend(tcd) + size -= tcd.ts + cdi += 1 + # END lbound overlap handling + + while cdi < slen and size: + # are we larger than the current block + cd = dcl[cdi] + if cd.ts <= size: + lappend(DeltaChunk(cd.to, cd.ts, cd.so, cd.data)) + size -= cd.ts + else: + tcd = DeltaChunk(cd.to, cd.ts, cd.so, cd.data) + tcd.ts = size + lappend(tcd) + size -= tcd.ts + break + # END hadle size + cdi += 1 + # END for each chunk + + +class DeltaChunkList(list): + + """List with special functionality to deal with DeltaChunks. + There are two types of lists we represent. The one was created bottom-up, working + towards the latest delta, the other kind was created top-down, working from the + latest delta down to the earliest ancestor. This attribute is queryable + after all processing with is_reversed.""" + + __slots__ = tuple() + + def rbound(self): + """:return: rightmost extend in bytes, absolute""" + if len(self) == 0: + return 0 + return self[-1].rbound() + + def lbound(self): + """:return: leftmost byte at which this chunklist starts""" + if len(self) == 0: + return 0 + return self[0].to + + def size(self): + """:return: size of bytes as measured by our delta chunks""" + return self.rbound() - self.lbound() + + def apply(self, bbuf, write): + """Only used by public clients, internally we only use the global routines + for performance""" + return delta_list_apply(self, bbuf, write) + + def compress(self): + """Alter the list to reduce the amount of nodes. Currently we concatenate + add-chunks + :return: self""" + slen = len(self) + if slen < 2: + return self + i = 0 + + first_data_index = None + while i < slen: + dc = self[i] + i += 1 + if dc.data is None: + if first_data_index is not None and i - 2 - first_data_index > 1: + # if first_data_index is not None: + nd = StringIO() # new data + so = self[first_data_index].to # start offset in target buffer + for x in range(first_data_index, i - 1): + xdc = self[x] + nd.write(xdc.data[:xdc.ts]) + # END collect data + + del(self[first_data_index:i - 1]) + buf = nd.getvalue() + self.insert(first_data_index, DeltaChunk(so, len(buf), 0, buf)) + + slen = len(self) + i = first_data_index + 1 + + # END concatenate data + first_data_index = None + continue + # END skip non-data chunks + + if first_data_index is None: + first_data_index = i - 1 + # END iterate list + + # if slen_orig != len(self): + # print "INFO: Reduced delta list len to %f %% of former size" % ((float(len(self)) / slen_orig) * 100) + return self + + def check_integrity(self, target_size=-1): + """Verify the list has non-overlapping chunks only, and the total size matches + target_size + :param target_size: if not -1, the total size of the chain must be target_size + :raise AssertionError: if the size doesn't match""" + if target_size > -1: + assert self[-1].rbound() == target_size + assert reduce(lambda x, y: x + y, (d.ts for d in self), 0) == target_size + # END target size verification + + if len(self) < 2: + return + + # check data + for dc in self: + assert dc.ts > 0 + if dc.has_data(): + assert len(dc.data) >= dc.ts + # END for each dc + + left = islice(self, 0, len(self) - 1) + right = iter(self) + right.next() + # this is very pythonic - we might have just use index based access here, + # but this could actually be faster + for lft, rgt in zip(left, right): + assert lft.rbound() == rgt.to + assert lft.to + lft.ts == rgt.to + # END for each pair + + +class TopdownDeltaChunkList(DeltaChunkList): + + """Represents a list which is generated by feeding its ancestor streams one by + one""" + __slots__ = tuple() + + def connect_with_next_base(self, bdcl): + """Connect this chain with the next level of our base delta chunklist. + The goal in this game is to mark as many of our chunks rigid, hence they + cannot be changed by any of the upcoming bases anymore. Once all our + chunks are marked like that, we can stop all processing + :param bdcl: data chunk list being one of our bases. They must be fed in + consecutively and in order, towards the earliest ancestor delta + :return: True if processing was done. Use it to abort processing of + remaining streams if False is returned""" + nfc = 0 # number of frozen chunks + dci = 0 # delta chunk index + slen = len(self) # len of self + ccl = list() # temporary list + while dci < slen: + dc = self[dci] + dci += 1 + + # all add-chunks which are already topmost don't need additional processing + if dc.data is not None: + nfc += 1 + continue + # END skip add chunks + + # copy chunks + # integrate the portion of the base list into ourselves. Lists + # dont support efficient insertion ( just one at a time ), but for now + # we live with it. Internally, its all just a 32/64bit pointer, and + # the portions of moved memory should be smallish. Maybe we just rebuild + # ourselves in order to reduce the amount of insertions ... + del(ccl[:]) + delta_list_slice(bdcl, dc.so, dc.ts, ccl) + + # move the target bounds into place to match with our chunk + ofs = dc.to - dc.so + for cdc in ccl: + cdc.to += ofs + # END update target bounds + + if len(ccl) == 1: + self[dci - 1] = ccl[0] + else: + # maybe try to compute the expenses here, and pick the right algorithm + # It would normally be faster than copying everything physically though + # TODO: Use a deque here, and decide by the index whether to extend + # or extend left ! + post_dci = self[dci:] + del(self[dci - 1:]) # include deletion of dc + self.extend(ccl) + self.extend(post_dci) + + slen = len(self) + dci += len(ccl) - 1 # deleted dc, added rest + + # END handle chunk replacement + # END for each chunk + + if nfc == slen: + return False + # END handle completeness + return True + + +#} END structures + +#{ Routines + +def is_loose_object(m): + """ + :return: True the file contained in memory map m appears to be a loose object. + Only the first two bytes are needed""" + b0, b1 = map(ord, m[:2]) + word = (b0 << 8) + b1 + return b0 == 0x78 and (word % 31) == 0 + + +def loose_object_header_info(m): + """ + :return: tuple(type_string, uncompressed_size_in_bytes) the type string of the + object as well as its uncompressed size in bytes. + :param m: memory map from which to read the compressed object data""" + decompress_size = 8192 # is used in cgit as well + hdr = decompressobj().decompress(m, decompress_size) + type_name, size = hdr[:hdr.find(NULL_BYTE)].split(BYTE_SPACE) + + return type_name, int(size) + + +def pack_object_header_info(data): + """ + :return: tuple(type_id, uncompressed_size_in_bytes, byte_offset) + The type_id should be interpreted according to the ``type_id_to_type_map`` map + The byte-offset specifies the start of the actual zlib compressed datastream + :param m: random-access memory, like a string or memory map""" + c = byte_ord(data[0]) # first byte + i = 1 # next char to read + type_id = (c >> 4) & 7 # numeric type + size = c & 15 # starting size + s = 4 # starting bit-shift size + while c & 0x80: + c = byte_ord(data[i]) + i += 1 + size += (c & 0x7f) << s + s += 7 + # END character loop + # end performance at expense of maintenance ... + return (type_id, size, i) + + +def create_pack_object_header(obj_type, obj_size): + """ + :return: string defining the pack header comprised of the object type + and its incompressed size in bytes + + :param obj_type: pack type_id of the object + :param obj_size: uncompressed size in bytes of the following object stream""" + c = 0 # 1 byte + hdr = bytearray() # output string + + c = (obj_type << 4) | (obj_size & 0xf) + obj_size >>= 4 + while obj_size: + hdr.append(c | 0x80) + c = obj_size & 0x7f + obj_size >>= 7 + # END until size is consumed + hdr.append(c) + # end handle interpreter + return hdr + + +def msb_size(data, offset=0): + """ + :return: tuple(read_bytes, size) read the msb size from the given random + access data starting at the given byte offset""" + size = 0 + i = 0 + l = len(data) + hit_msb = False + while i < l: + c = data[i + offset] + size |= (c & 0x7f) << i * 7 + i += 1 + if not c & 0x80: + hit_msb = True + break + # END check msb bit + # END while in range + # end performance ... + if not hit_msb: + raise AssertionError("Could not find terminating MSB byte in data stream") + return i + offset, size + + +def loose_object_header(type, size): + """ + :return: bytes representing the loose object header, which is immediately + followed by the content stream of size 'size'""" + return ('%s %i\0' % (force_text(type), size)).encode('ascii') + + +def write_object(type, size, read, write, chunk_size=chunk_size): + """ + Write the object as identified by type, size and source_stream into the + target_stream + + :param type: type string of the object + :param size: amount of bytes to write from source_stream + :param read: read method of a stream providing the content data + :param write: write method of the output stream + :param close_target_stream: if True, the target stream will be closed when + the routine exits, even if an error is thrown + :return: The actual amount of bytes written to stream, which includes the header and a trailing newline""" + tbw = 0 # total num bytes written + + # WRITE HEADER: type SP size NULL + tbw += write(loose_object_header(type, size)) + tbw += stream_copy(read, write, size, chunk_size) + + return tbw + + +def stream_copy(read, write, size, chunk_size): + """ + Copy a stream up to size bytes using the provided read and write methods, + in chunks of chunk_size + + **Note:** its much like stream_copy utility, but operates just using methods""" + dbw = 0 # num data bytes written + + # WRITE ALL DATA UP TO SIZE + while True: + cs = min(chunk_size, size - dbw) + # NOTE: not all write methods return the amount of written bytes, like + # mmap.write. Its bad, but we just deal with it ... perhaps its not + # even less efficient + # data_len = write(read(cs)) + # dbw += data_len + data = read(cs) + data_len = len(data) + dbw += data_len + write(data) + if data_len < cs or dbw == size: + break + # END check for stream end + # END duplicate data + return dbw + + +def connect_deltas(dstreams): + """ + Read the condensed delta chunk information from dstream and merge its information + into a list of existing delta chunks + + :param dstreams: iterable of delta stream objects, the delta to be applied last + comes first, then all its ancestors in order + :return: DeltaChunkList, containing all operations to apply""" + tdcl = None # topmost dcl + + dcl = tdcl = TopdownDeltaChunkList() + for dsi, ds in enumerate(dstreams): + # print "Stream", dsi + db = ds.read() + delta_buf_size = ds.size + + # read header + i, base_size = msb_size(db) + i, target_size = msb_size(db, i) + + # interpret opcodes + tbw = 0 # amount of target bytes written + while i < delta_buf_size: + c = ord(db[i]) + i += 1 + if c & 0x80: + cp_off, cp_size = 0, 0 + if (c & 0x01): + cp_off = ord(db[i]) + i += 1 + if (c & 0x02): + cp_off |= (ord(db[i]) << 8) + i += 1 + if (c & 0x04): + cp_off |= (ord(db[i]) << 16) + i += 1 + if (c & 0x08): + cp_off |= (ord(db[i]) << 24) + i += 1 + if (c & 0x10): + cp_size = ord(db[i]) + i += 1 + if (c & 0x20): + cp_size |= (ord(db[i]) << 8) + i += 1 + if (c & 0x40): + cp_size |= (ord(db[i]) << 16) + i += 1 + + if not cp_size: + cp_size = 0x10000 + + rbound = cp_off + cp_size + if (rbound < cp_size or + rbound > base_size): + break + + dcl.append(DeltaChunk(tbw, cp_size, cp_off, None)) + tbw += cp_size + elif c: + # NOTE: in C, the data chunks should probably be concatenated here. + # In python, we do it as a post-process + dcl.append(DeltaChunk(tbw, c, 0, db[i:i + c])) + i += c + tbw += c + else: + raise ValueError("unexpected delta opcode 0") + # END handle command byte + # END while processing delta data + + dcl.compress() + + # merge the lists ! + if dsi > 0: + if not tdcl.connect_with_next_base(dcl): + break + # END handle merge + + # prepare next base + dcl = DeltaChunkList() + # END for each delta stream + + return tdcl + + +def apply_delta_data(src_buf, src_buf_size, delta_buf, delta_buf_size, write): + """ + Apply data from a delta buffer using a source buffer to the target file + + :param src_buf: random access data from which the delta was created + :param src_buf_size: size of the source buffer in bytes + :param delta_buf_size: size for the delta buffer in bytes + :param delta_buf: random access delta data + :param write: write method taking a chunk of bytes + + **Note:** transcribed to python from the similar routine in patch-delta.c""" + i = 0 + db = delta_buf + while i < delta_buf_size: + c = db[i] + i += 1 + if c & 0x80: + cp_off, cp_size = 0, 0 + if (c & 0x01): + cp_off = db[i] + i += 1 + if (c & 0x02): + cp_off |= (db[i] << 8) + i += 1 + if (c & 0x04): + cp_off |= (db[i] << 16) + i += 1 + if (c & 0x08): + cp_off |= (db[i] << 24) + i += 1 + if (c & 0x10): + cp_size = db[i] + i += 1 + if (c & 0x20): + cp_size |= (db[i] << 8) + i += 1 + if (c & 0x40): + cp_size |= (db[i] << 16) + i += 1 + + if not cp_size: + cp_size = 0x10000 + + rbound = cp_off + cp_size + if (rbound < cp_size or + rbound > src_buf_size): + break + write(src_buf[cp_off:cp_off + cp_size]) + elif c: + write(db[i:i + c]) + i += c + else: + raise ValueError("unexpected delta opcode 0") + # END handle command byte + # END while processing delta data + + # yes, lets use the exact same error message that git uses :) + assert i == delta_buf_size, "delta replay has gone wild" + + +def is_equal_canonical_sha(canonical_length, match, sha1): + """ + :return: True if the given lhs and rhs 20 byte binary shas + The comparison will take the canonical_length of the match sha into account, + hence the comparison will only use the last 4 bytes for uneven canonical representations + :param match: less than 20 byte sha + :param sha1: 20 byte sha""" + binary_length = canonical_length // 2 + if match[:binary_length] != sha1[:binary_length]: + return False + + if canonical_length - binary_length and \ + (byte_ord(match[-1]) ^ byte_ord(sha1[len(match) - 1])) & 0xf0: + return False + # END handle uneven canonnical length + return True + +#} END routines + + +try: + from gitdb_speedups._perf import connect_deltas +except ImportError: + pass diff --git a/gitdb/gitdb/pack.py b/gitdb/gitdb/pack.py new file mode 100644 index 000000000..2e0707947 --- /dev/null +++ b/gitdb/gitdb/pack.py @@ -0,0 +1,1028 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Contains PackIndexFile and PackFile implementations""" +import zlib + +from gitdb.exc import ( + BadObject, + AmbiguousObjectName, + UnsupportedOperation, + ParseError +) + +from gitdb.util import ( + mman, + LazyMixin, + unpack_from, + bin_to_hex, + byte_ord, +) + +from gitdb.fun import ( + create_pack_object_header, + pack_object_header_info, + is_equal_canonical_sha, + type_id_to_type_map, + write_object, + stream_copy, + chunk_size, + delta_types, + OFS_DELTA, + REF_DELTA, + msb_size +) + +try: + from gitdb_speedups._perf import PackIndexFile_sha_to_index +except ImportError: + pass +# END try c module + +from gitdb.base import ( # Amazing ! + OInfo, + OStream, + OPackInfo, + OPackStream, + ODeltaStream, + ODeltaPackInfo, + ODeltaPackStream, +) + +from gitdb.stream import ( + DecompressMemMapReader, + DeltaApplyReader, + Sha1Writer, + NullStream, + FlexibleSha1Writer +) + +from struct import pack +from binascii import crc32 + +from gitdb.const import NULL_BYTE + +import tempfile +import array +import os +import sys + +__all__ = ('PackIndexFile', 'PackFile', 'PackEntity') + + +#{ Utilities + +def pack_object_at(cursor, offset, as_stream): + """ + :return: Tuple(abs_data_offset, PackInfo|PackStream) + an object of the correct type according to the type_id of the object. + If as_stream is True, the object will contain a stream, allowing the + data to be read decompressed. + :param data: random accessible data containing all required information + :parma offset: offset in to the data at which the object information is located + :param as_stream: if True, a stream object will be returned that can read + the data, otherwise you receive an info object only""" + data = cursor.use_region(offset).buffer() + type_id, uncomp_size, data_rela_offset = pack_object_header_info(data) + total_rela_offset = None # set later, actual offset until data stream begins + delta_info = None + + # OFFSET DELTA + if type_id == OFS_DELTA: + i = data_rela_offset + c = byte_ord(data[i]) + i += 1 + delta_offset = c & 0x7f + while c & 0x80: + c = byte_ord(data[i]) + i += 1 + delta_offset += 1 + delta_offset = (delta_offset << 7) + (c & 0x7f) + # END character loop + delta_info = delta_offset + total_rela_offset = i + # REF DELTA + elif type_id == REF_DELTA: + total_rela_offset = data_rela_offset + 20 + delta_info = data[data_rela_offset:total_rela_offset] + # BASE OBJECT + else: + # assume its a base object + total_rela_offset = data_rela_offset + # END handle type id + abs_data_offset = offset + total_rela_offset + if as_stream: + stream = DecompressMemMapReader(data[total_rela_offset:], False, uncomp_size) + if delta_info is None: + return abs_data_offset, OPackStream(offset, type_id, uncomp_size, stream) + else: + return abs_data_offset, ODeltaPackStream(offset, type_id, uncomp_size, delta_info, stream) + else: + if delta_info is None: + return abs_data_offset, OPackInfo(offset, type_id, uncomp_size) + else: + return abs_data_offset, ODeltaPackInfo(offset, type_id, uncomp_size, delta_info) + # END handle info + # END handle stream + + +def write_stream_to_pack(read, write, zstream, base_crc=None): + """Copy a stream as read from read function, zip it, and write the result. + Count the number of written bytes and return it + :param base_crc: if not None, the crc will be the base for all compressed data + we consecutively write and generate a crc32 from. If None, no crc will be generated + :return: tuple(no bytes read, no bytes written, crc32) crc might be 0 if base_crc + was false""" + br = 0 # bytes read + bw = 0 # bytes written + want_crc = base_crc is not None + crc = 0 + if want_crc: + crc = base_crc + # END initialize crc + + while True: + chunk = read(chunk_size) + br += len(chunk) + compressed = zstream.compress(chunk) + bw += len(compressed) + write(compressed) # cannot assume return value + + if want_crc: + crc = crc32(compressed, crc) + # END handle crc + + if len(chunk) != chunk_size: + break + # END copy loop + + compressed = zstream.flush() + bw += len(compressed) + write(compressed) + if want_crc: + crc = crc32(compressed, crc) + # END handle crc + + return (br, bw, crc) + + +#} END utilities + + +class IndexWriter: + + """Utility to cache index information, allowing to write all information later + in one go to the given stream + **Note:** currently only writes v2 indices""" + __slots__ = '_objs' + + def __init__(self): + self._objs = list() + + def append(self, binsha, crc, offset): + """Append one piece of object information""" + self._objs.append((binsha, crc, offset)) + + def write(self, pack_sha, write): + """Write the index file using the given write method + :param pack_sha: binary sha over the whole pack that we index + :return: sha1 binary sha over all index file contents""" + # sort for sha1 hash + self._objs.sort(key=lambda o: o[0]) + + sha_writer = FlexibleSha1Writer(write) + sha_write = sha_writer.write + sha_write(PackIndexFile.index_v2_signature) + sha_write(pack(">L", PackIndexFile.index_version_default)) + + # fanout + tmplist = list((0,) * 256) # fanout or list with 64 bit offsets + for t in self._objs: + tmplist[byte_ord(t[0][0])] += 1 + # END prepare fanout + for i in range(255): + v = tmplist[i] + sha_write(pack('>L', v)) + tmplist[i + 1] += v + # END write each fanout entry + sha_write(pack('>L', tmplist[255])) + + # sha1 ordered + # save calls, that is push them into c + sha_write(b''.join(t[0] for t in self._objs)) + + # crc32 + for t in self._objs: + sha_write(pack('>L', t[1] & 0xffffffff)) + # END for each crc + + tmplist = list() + # offset 32 + for t in self._objs: + ofs = t[2] + if ofs > 0x7fffffff: + tmplist.append(ofs) + ofs = 0x80000000 + len(tmplist) - 1 + # END handle 64 bit offsets + sha_write(pack('>L', ofs & 0xffffffff)) + # END for each offset + + # offset 64 + for ofs in tmplist: + sha_write(pack(">Q", ofs)) + # END for each offset + + # trailer + assert(len(pack_sha) == 20) + sha_write(pack_sha) + sha = sha_writer.sha(as_hex=False) + write(sha) + return sha + + +class PackIndexFile(LazyMixin): + + """A pack index provides offsets into the corresponding pack, allowing to find + locations for offsets faster.""" + + # Dont use slots as we dynamically bind functions for each version, need a dict for this + # The slots you see here are just to keep track of our instance variables + # __slots__ = ('_indexpath', '_fanout_table', '_cursor', '_version', + # '_sha_list_offset', '_crc_list_offset', '_pack_offset', '_pack_64_offset') + + # used in v2 indices + _sha_list_offset = 8 + 1024 + index_v2_signature = b'\xfftOc' + index_version_default = 2 + + def __init__(self, indexpath): + super().__init__() + self._indexpath = indexpath + + def close(self): + mman.force_map_handle_removal_win(self._indexpath) + self._cursor = None + + def _set_cache_(self, attr): + if attr == "_packfile_checksum": + self._packfile_checksum = self._cursor.map()[-40:-20] + elif attr == "_cursor": + # Note: We don't lock the file when reading as we cannot be sure + # that we can actually write to the location - it could be a read-only + # alternate for instance + self._cursor = mman.make_cursor(self._indexpath).use_region() + # We will assume that the index will always fully fit into memory ! + if mman.window_size() > 0 and self._cursor.file_size() > mman.window_size(): + raise AssertionError("The index file at %s is too large to fit into a mapped window (%i > %i). This is a limitation of the implementation" % ( + self._indexpath, self._cursor.file_size(), mman.window_size())) + # END assert window size + else: + # now its time to initialize everything - if we are here, someone wants + # to access the fanout table or related properties + + # CHECK VERSION + mmap = self._cursor.map() + self._version = (mmap[:4] == self.index_v2_signature and 2) or 1 + if self._version == 2: + version_id = unpack_from(">L", mmap, 4)[0] + assert version_id == self._version, "Unsupported index version: %i" % version_id + # END assert version + + # SETUP FUNCTIONS + # setup our functions according to the actual version + for fname in ('entry', 'offset', 'sha', 'crc'): + setattr(self, fname, getattr(self, "_%s_v%i" % (fname, self._version))) + # END for each function to initialize + + # INITIALIZE DATA + # byte offset is 8 if version is 2, 0 otherwise + self._initialize() + # END handle attributes + + #{ Access V1 + + def _entry_v1(self, i): + """:return: tuple(offset, binsha, 0)""" + return unpack_from(">L20s", self._cursor.map(), 1024 + i * 24) + (0, ) + + def _offset_v1(self, i): + """see ``_offset_v2``""" + return unpack_from(">L", self._cursor.map(), 1024 + i * 24)[0] + + def _sha_v1(self, i): + """see ``_sha_v2``""" + base = 1024 + (i * 24) + 4 + return self._cursor.map()[base:base + 20] + + def _crc_v1(self, i): + """unsupported""" + return 0 + + #} END access V1 + + #{ Access V2 + def _entry_v2(self, i): + """:return: tuple(offset, binsha, crc)""" + return (self._offset_v2(i), self._sha_v2(i), self._crc_v2(i)) + + def _offset_v2(self, i): + """:return: 32 or 64 byte offset into pack files. 64 byte offsets will only + be returned if the pack is larger than 4 GiB, or 2^32""" + offset = unpack_from(">L", self._cursor.map(), self._pack_offset + i * 4)[0] + + # if the high-bit is set, this indicates that we have to lookup the offset + # in the 64 bit region of the file. The current offset ( lower 31 bits ) + # are the index into it + if offset & 0x80000000: + offset = unpack_from(">Q", self._cursor.map(), self._pack_64_offset + (offset & ~0x80000000) * 8)[0] + # END handle 64 bit offset + + return offset + + def _sha_v2(self, i): + """:return: sha at the given index of this file index instance""" + base = self._sha_list_offset + i * 20 + return self._cursor.map()[base:base + 20] + + def _crc_v2(self, i): + """:return: 4 bytes crc for the object at index i""" + return unpack_from(">L", self._cursor.map(), self._crc_list_offset + i * 4)[0] + + #} END access V2 + + #{ Initialization + + def _initialize(self): + """initialize base data""" + self._fanout_table = self._read_fanout((self._version == 2) * 8) + + if self._version == 2: + self._crc_list_offset = self._sha_list_offset + self.size() * 20 + self._pack_offset = self._crc_list_offset + self.size() * 4 + self._pack_64_offset = self._pack_offset + self.size() * 4 + # END setup base + + def _read_fanout(self, byte_offset): + """Generate a fanout table from our data""" + d = self._cursor.map() + out = list() + append = out.append + for i in range(256): + append(unpack_from('>L', d, byte_offset + i * 4)[0]) + # END for each entry + return out + + #} END initialization + + #{ Properties + def version(self): + return self._version + + def size(self): + """:return: amount of objects referred to by this index""" + return self._fanout_table[255] + + def path(self): + """:return: path to the packindexfile""" + return self._indexpath + + def packfile_checksum(self): + """:return: 20 byte sha representing the sha1 hash of the pack file""" + return self._cursor.map()[-40:-20] + + def indexfile_checksum(self): + """:return: 20 byte sha representing the sha1 hash of this index file""" + return self._cursor.map()[-20:] + + def offsets(self): + """:return: sequence of all offsets in the order in which they were written + + **Note:** return value can be random accessed, but may be immmutable""" + if self._version == 2: + # read stream to array, convert to tuple + a = array.array('I') # 4 byte unsigned int, long are 8 byte on 64 bit it appears + a.frombytes(self._cursor.map()[self._pack_offset:self._pack_64_offset]) + + # networkbyteorder to something array likes more + if sys.byteorder == 'little': + a.byteswap() + return a + else: + return tuple(self.offset(index) for index in range(self.size())) + # END handle version + + def sha_to_index(self, sha): + """ + :return: index usable with the ``offset`` or ``entry`` method, or None + if the sha was not found in this pack index + :param sha: 20 byte sha to lookup""" + first_byte = byte_ord(sha[0]) + get_sha = self.sha + lo = 0 # lower index, the left bound of the bisection + if first_byte != 0: + lo = self._fanout_table[first_byte - 1] + hi = self._fanout_table[first_byte] # the upper, right bound of the bisection + + # bisect until we have the sha + while lo < hi: + mid = (lo + hi) // 2 + mid_sha = get_sha(mid) + if sha < mid_sha: + hi = mid + elif sha == mid_sha: + return mid + else: + lo = mid + 1 + # END handle midpoint + # END bisect + return None + + def partial_sha_to_index(self, partial_bin_sha, canonical_length): + """ + :return: index as in `sha_to_index` or None if the sha was not found in this + index file + :param partial_bin_sha: an at least two bytes of a partial binary sha as bytes + :param canonical_length: length of the original hexadecimal representation of the + given partial binary sha + :raise AmbiguousObjectName:""" + if len(partial_bin_sha) < 2: + raise ValueError("Require at least 2 bytes of partial sha") + + assert isinstance(partial_bin_sha, bytes), "partial_bin_sha must be bytes" + first_byte = byte_ord(partial_bin_sha[0]) + + get_sha = self.sha + lo = 0 # lower index, the left bound of the bisection + if first_byte != 0: + lo = self._fanout_table[first_byte - 1] + hi = self._fanout_table[first_byte] # the upper, right bound of the bisection + + # fill the partial to full 20 bytes + filled_sha = partial_bin_sha + NULL_BYTE * (20 - len(partial_bin_sha)) + + # find lowest + while lo < hi: + mid = (lo + hi) // 2 + mid_sha = get_sha(mid) + if filled_sha < mid_sha: + hi = mid + elif filled_sha == mid_sha: + # perfect match + lo = mid + break + else: + lo = mid + 1 + # END handle midpoint + # END bisect + + if lo < self.size(): + cur_sha = get_sha(lo) + if is_equal_canonical_sha(canonical_length, partial_bin_sha, cur_sha): + next_sha = None + if lo + 1 < self.size(): + next_sha = get_sha(lo + 1) + if next_sha and next_sha == cur_sha: + raise AmbiguousObjectName(partial_bin_sha) + return lo + # END if we have a match + # END if we found something + return None + + if 'PackIndexFile_sha_to_index' in globals(): + # NOTE: Its just about 25% faster, the major bottleneck might be the attr + # accesses + def sha_to_index(self, sha): + return PackIndexFile_sha_to_index(self, sha) + # END redefine heavy-hitter with c version + + #} END properties + + +class PackFile(LazyMixin): + + """A pack is a file written according to the Version 2 for git packs + + As we currently use memory maps, it could be assumed that the maximum size of + packs therefore is 32 bit on 32 bit systems. On 64 bit systems, this should be + fine though. + + **Note:** at some point, this might be implemented using streams as well, or + streams are an alternate path in the case memory maps cannot be created + for some reason - one clearly doesn't want to read 10GB at once in that + case""" + + __slots__ = ('_packpath', '_cursor', '_size', '_version') + pack_signature = 0x5041434b # 'PACK' + pack_version_default = 2 + + # offset into our data at which the first object starts + first_object_offset = 3 * 4 # header bytes + footer_size = 20 # final sha + + def __init__(self, packpath): + self._packpath = packpath + + def close(self): + mman.force_map_handle_removal_win(self._packpath) + self._cursor = None + + def _set_cache_(self, attr): + # we fill the whole cache, whichever attribute gets queried first + self._cursor = mman.make_cursor(self._packpath).use_region() + + # read the header information + type_id, self._version, self._size = unpack_from(">LLL", self._cursor.map(), 0) + + # TODO: figure out whether we should better keep the lock, or maybe + # add a .keep file instead ? + if type_id != self.pack_signature: + raise ParseError("Invalid pack signature: %i" % type_id) + + def _iter_objects(self, start_offset, as_stream=True): + """Handle the actual iteration of objects within this pack""" + c = self._cursor + content_size = c.file_size() - self.footer_size + cur_offset = start_offset or self.first_object_offset + + null = NullStream() + while cur_offset < content_size: + data_offset, ostream = pack_object_at(c, cur_offset, True) + # scrub the stream to the end - this decompresses the object, but yields + # the amount of compressed bytes we need to get to the next offset + + stream_copy(ostream.read, null.write, ostream.size, chunk_size) + assert ostream.stream._br == ostream.size + cur_offset += (data_offset - ostream.pack_offset) + ostream.stream.compressed_bytes_read() + + # if a stream is requested, reset it beforehand + # Otherwise return the Stream object directly, its derived from the + # info object + if as_stream: + ostream.stream.seek(0) + yield ostream + # END until we have read everything + + #{ Pack Information + + def size(self): + """:return: The amount of objects stored in this pack""" + return self._size + + def version(self): + """:return: the version of this pack""" + return self._version + + def data(self): + """ + :return: read-only data of this pack. It provides random access and usually + is a memory map. + :note: This method is unsafe as it returns a window into a file which might be larger than than the actual window size""" + # can use map as we are starting at offset 0. Otherwise we would have to use buffer() + return self._cursor.use_region().map() + + def checksum(self): + """:return: 20 byte sha1 hash on all object sha's contained in this file""" + return self._cursor.use_region(self._cursor.file_size() - 20).buffer()[:] + + def path(self): + """:return: path to the packfile""" + return self._packpath + #} END pack information + + #{ Pack Specific + + def collect_streams(self, offset): + """ + :return: list of pack streams which are required to build the object + at the given offset. The first entry of the list is the object at offset, + the last one is either a full object, or a REF_Delta stream. The latter + type needs its reference object to be locked up in an ODB to form a valid + delta chain. + If the object at offset is no delta, the size of the list is 1. + :param offset: specifies the first byte of the object within this pack""" + out = list() + c = self._cursor + while True: + ostream = pack_object_at(c, offset, True)[1] + out.append(ostream) + if ostream.type_id == OFS_DELTA: + offset = ostream.pack_offset - ostream.delta_info + else: + # the only thing we can lookup are OFFSET deltas. Everything + # else is either an object, or a ref delta, in the latter + # case someone else has to find it + break + # END handle type + # END while chaining streams + return out + + #} END pack specific + + #{ Read-Database like Interface + + def info(self, offset): + """Retrieve information about the object at the given file-absolute offset + + :param offset: byte offset + :return: OPackInfo instance, the actual type differs depending on the type_id attribute""" + return pack_object_at(self._cursor, offset or self.first_object_offset, False)[1] + + def stream(self, offset): + """Retrieve an object at the given file-relative offset as stream along with its information + + :param offset: byte offset + :return: OPackStream instance, the actual type differs depending on the type_id attribute""" + return pack_object_at(self._cursor, offset or self.first_object_offset, True)[1] + + def stream_iter(self, start_offset=0): + """ + :return: iterator yielding OPackStream compatible instances, allowing + to access the data in the pack directly. + :param start_offset: offset to the first object to iterate. If 0, iteration + starts at the very first object in the pack. + + **Note:** Iterating a pack directly is costly as the datastream has to be decompressed + to determine the bounds between the objects""" + return self._iter_objects(start_offset, as_stream=True) + + #} END Read-Database like Interface + + +class PackEntity(LazyMixin): + + """Combines the PackIndexFile and the PackFile into one, allowing the + actual objects to be resolved and iterated""" + + __slots__ = ('_index', # our index file + '_pack', # our pack file + '_offset_map' # on demand dict mapping one offset to the next consecutive one + ) + + IndexFileCls = PackIndexFile + PackFileCls = PackFile + + def __init__(self, pack_or_index_path): + """Initialize ourselves with the path to the respective pack or index file""" + basename, ext = os.path.splitext(pack_or_index_path) + self._index = self.IndexFileCls("%s.idx" % basename) # PackIndexFile instance + self._pack = self.PackFileCls("%s.pack" % basename) # corresponding PackFile instance + + def close(self): + self._index.close() + self._pack.close() + + def _set_cache_(self, attr): + # currently this can only be _offset_map + # TODO: make this a simple sorted offset array which can be bisected + # to find the respective entry, from which we can take a +1 easily + # This might be slower, but should also be much lighter in memory ! + offsets_sorted = sorted(self._index.offsets()) + last_offset = len(self._pack.data()) - self._pack.footer_size + assert offsets_sorted, "Cannot handle empty indices" + + offset_map = None + if len(offsets_sorted) == 1: + offset_map = {offsets_sorted[0]: last_offset} + else: + iter_offsets = iter(offsets_sorted) + iter_offsets_plus_one = iter(offsets_sorted) + next(iter_offsets_plus_one) + consecutive = zip(iter_offsets, iter_offsets_plus_one) + + offset_map = dict(consecutive) + + # the last offset is not yet set + offset_map[offsets_sorted[-1]] = last_offset + # END handle offset amount + self._offset_map = offset_map + + def _sha_to_index(self, sha): + """:return: index for the given sha, or raise""" + index = self._index.sha_to_index(sha) + if index is None: + raise BadObject(sha) + return index + + def _iter_objects(self, as_stream): + """Iterate over all objects in our index and yield their OInfo or OStream instences""" + _sha = self._index.sha + _object = self._object + for index in range(self._index.size()): + yield _object(_sha(index), as_stream, index) + # END for each index + + def _object(self, sha, as_stream, index=-1): + """:return: OInfo or OStream object providing information about the given sha + :param index: if not -1, its assumed to be the sha's index in the IndexFile""" + # its a little bit redundant here, but it needs to be efficient + if index < 0: + index = self._sha_to_index(sha) + if sha is None: + sha = self._index.sha(index) + # END assure sha is present ( in output ) + offset = self._index.offset(index) + type_id, uncomp_size, data_rela_offset = pack_object_header_info(self._pack._cursor.use_region(offset).buffer()) + if as_stream: + if type_id not in delta_types: + packstream = self._pack.stream(offset) + return OStream(sha, packstream.type, packstream.size, packstream.stream) + # END handle non-deltas + + # produce a delta stream containing all info + # To prevent it from applying the deltas when querying the size, + # we extract it from the delta stream ourselves + streams = self.collect_streams_at_offset(offset) + dstream = DeltaApplyReader.new(streams) + + return ODeltaStream(sha, dstream.type, None, dstream) + else: + if type_id not in delta_types: + return OInfo(sha, type_id_to_type_map[type_id], uncomp_size) + # END handle non-deltas + + # deltas are a little tougher - unpack the first bytes to obtain + # the actual target size, as opposed to the size of the delta data + streams = self.collect_streams_at_offset(offset) + buf = streams[0].read(512) + offset, src_size = msb_size(buf) + offset, target_size = msb_size(buf, offset) + + # collect the streams to obtain the actual object type + if streams[-1].type_id in delta_types: + raise BadObject(sha, "Could not resolve delta object") + return OInfo(sha, streams[-1].type, target_size) + # END handle stream + + #{ Read-Database like Interface + + def info(self, sha): + """Retrieve information about the object identified by the given sha + + :param sha: 20 byte sha1 + :raise BadObject: + :return: OInfo instance, with 20 byte sha""" + return self._object(sha, False) + + def stream(self, sha): + """Retrieve an object stream along with its information as identified by the given sha + + :param sha: 20 byte sha1 + :raise BadObject: + :return: OStream instance, with 20 byte sha""" + return self._object(sha, True) + + def info_at_index(self, index): + """As ``info``, but uses a PackIndexFile compatible index to refer to the object""" + return self._object(None, False, index) + + def stream_at_index(self, index): + """As ``stream``, but uses a PackIndexFile compatible index to refer to the + object""" + return self._object(None, True, index) + + #} END Read-Database like Interface + + #{ Interface + + def pack(self): + """:return: the underlying pack file instance""" + return self._pack + + def index(self): + """:return: the underlying pack index file instance""" + return self._index + + def is_valid_stream(self, sha, use_crc=False): + """ + Verify that the stream at the given sha is valid. + + :param use_crc: if True, the index' crc is run over the compressed stream of + the object, which is much faster than checking the sha1. It is also + more prone to unnoticed corruption or manipulation. + :param sha: 20 byte sha1 of the object whose stream to verify + whether the compressed stream of the object is valid. If it is + a delta, this only verifies that the delta's data is valid, not the + data of the actual undeltified object, as it depends on more than + just this stream. + If False, the object will be decompressed and the sha generated. It must + match the given sha + + :return: True if the stream is valid + :raise UnsupportedOperation: If the index is version 1 only + :raise BadObject: sha was not found""" + if use_crc: + if self._index.version() < 2: + raise UnsupportedOperation("Version 1 indices do not contain crc's, verify by sha instead") + # END handle index version + + index = self._sha_to_index(sha) + offset = self._index.offset(index) + next_offset = self._offset_map[offset] + crc_value = self._index.crc(index) + + # create the current crc value, on the compressed object data + # Read it in chunks, without copying the data + crc_update = zlib.crc32 + pack_data = self._pack.data() + cur_pos = offset + this_crc_value = 0 + while cur_pos < next_offset: + rbound = min(cur_pos + chunk_size, next_offset) + size = rbound - cur_pos + this_crc_value = crc_update(pack_data[cur_pos:cur_pos + size], this_crc_value) + cur_pos += size + # END window size loop + + # crc returns signed 32 bit numbers, the AND op forces it into unsigned + # mode ... wow, sneaky, from dulwich. + return (this_crc_value & 0xffffffff) == crc_value + else: + shawriter = Sha1Writer() + stream = self._object(sha, as_stream=True) + # write a loose object, which is the basis for the sha + write_object(stream.type, stream.size, stream.read, shawriter.write) + + assert shawriter.sha(as_hex=False) == sha + return shawriter.sha(as_hex=False) == sha + # END handle crc/sha verification + + def info_iter(self): + """ + :return: Iterator over all objects in this pack. The iterator yields + OInfo instances""" + return self._iter_objects(as_stream=False) + + def stream_iter(self): + """ + :return: iterator over all objects in this pack. The iterator yields + OStream instances""" + return self._iter_objects(as_stream=True) + + def collect_streams_at_offset(self, offset): + """ + As the version in the PackFile, but can resolve REF deltas within this pack + For more info, see ``collect_streams`` + + :param offset: offset into the pack file at which the object can be found""" + streams = self._pack.collect_streams(offset) + + # try to resolve the last one if needed. It is assumed to be either + # a REF delta, or a base object, as OFFSET deltas are resolved by the pack + if streams[-1].type_id == REF_DELTA: + stream = streams[-1] + while stream.type_id in delta_types: + if stream.type_id == REF_DELTA: + # smmap can return memory view objects, which can't be compared as buffers/bytes can ... + if isinstance(stream.delta_info, memoryview): + sindex = self._index.sha_to_index(stream.delta_info.tobytes()) + else: + sindex = self._index.sha_to_index(stream.delta_info) + if sindex is None: + break + stream = self._pack.stream(self._index.offset(sindex)) + streams.append(stream) + else: + # must be another OFS DELTA - this could happen if a REF + # delta we resolve previously points to an OFS delta. Who + # would do that ;) ? We can handle it though + stream = self._pack.stream(stream.delta_info) + streams.append(stream) + # END handle ref delta + # END resolve ref streams + # END resolve streams + + return streams + + def collect_streams(self, sha): + """ + As ``PackFile.collect_streams``, but takes a sha instead of an offset. + Additionally, ref_delta streams will be resolved within this pack. + If this is not possible, the stream will be left alone, hence it is adivsed + to check for unresolved ref-deltas and resolve them before attempting to + construct a delta stream. + + :param sha: 20 byte sha1 specifying the object whose related streams you want to collect + :return: list of streams, first being the actual object delta, the last being + a possibly unresolved base object. + :raise BadObject:""" + return self.collect_streams_at_offset(self._index.offset(self._sha_to_index(sha))) + + @classmethod + def write_pack(cls, object_iter, pack_write, index_write=None, + object_count=None, zlib_compression=zlib.Z_BEST_SPEED): + """ + Create a new pack by putting all objects obtained by the object_iterator + into a pack which is written using the pack_write method. + The respective index is produced as well if index_write is not Non. + + :param object_iter: iterator yielding odb output objects + :param pack_write: function to receive strings to write into the pack stream + :param indx_write: if not None, the function writes the index file corresponding + to the pack. + :param object_count: if you can provide the amount of objects in your iteration, + this would be the place to put it. Otherwise we have to pre-iterate and store + all items into a list to get the number, which uses more memory than necessary. + :param zlib_compression: the zlib compression level to use + :return: tuple(pack_sha, index_binsha) binary sha over all the contents of the pack + and over all contents of the index. If index_write was None, index_binsha will be None + + **Note:** The destination of the write functions is up to the user. It could + be a socket, or a file for instance + + **Note:** writes only undeltified objects""" + objs = object_iter + if not object_count: + if not isinstance(object_iter, (tuple, list)): + objs = list(object_iter) + # END handle list type + object_count = len(objs) + # END handle object + + pack_writer = FlexibleSha1Writer(pack_write) + pwrite = pack_writer.write + ofs = 0 # current offset into the pack file + index = None + wants_index = index_write is not None + + # write header + pwrite(pack('>LLL', PackFile.pack_signature, PackFile.pack_version_default, object_count)) + ofs += 12 + + if wants_index: + index = IndexWriter() + # END handle index header + + actual_count = 0 + for obj in objs: + actual_count += 1 + crc = 0 + + # object header + hdr = create_pack_object_header(obj.type_id, obj.size) + if index_write: + crc = crc32(hdr) + else: + crc = None + # END handle crc + pwrite(hdr) + + # data stream + zstream = zlib.compressobj(zlib_compression) + ostream = obj.stream + br, bw, crc = write_stream_to_pack(ostream.read, pwrite, zstream, base_crc=crc) + assert(br == obj.size) + if wants_index: + index.append(obj.binsha, crc, ofs) + # END handle index + + ofs += len(hdr) + bw + if actual_count == object_count: + break + # END abort once we are done + # END for each object + + if actual_count != object_count: + raise ValueError( + "Expected to write %i objects into pack, but received only %i from iterators" % (object_count, actual_count)) + # END count assertion + + # write footer + pack_sha = pack_writer.sha(as_hex=False) + assert len(pack_sha) == 20 + pack_write(pack_sha) + ofs += len(pack_sha) # just for completeness ;) + + index_sha = None + if wants_index: + index_sha = index.write(pack_sha, index_write) + # END handle index + + return pack_sha, index_sha + + @classmethod + def create(cls, object_iter, base_dir, object_count=None, zlib_compression=zlib.Z_BEST_SPEED): + """Create a new on-disk entity comprised of a properly named pack file and a properly named + and corresponding index file. The pack contains all OStream objects contained in object iter. + :param base_dir: directory which is to contain the files + :return: PackEntity instance initialized with the new pack + + **Note:** for more information on the other parameters see the write_pack method""" + pack_fd, pack_path = tempfile.mkstemp('', 'pack', base_dir) + index_fd, index_path = tempfile.mkstemp('', 'index', base_dir) + pack_write = lambda d: os.write(pack_fd, d) + index_write = lambda d: os.write(index_fd, d) + + pack_binsha, index_binsha = cls.write_pack(object_iter, pack_write, index_write, object_count, zlib_compression) + os.close(pack_fd) + os.close(index_fd) + + fmt = "pack-%s.%s" + new_pack_path = os.path.join(base_dir, fmt % (bin_to_hex(pack_binsha), 'pack')) + new_index_path = os.path.join(base_dir, fmt % (bin_to_hex(pack_binsha), 'idx')) + os.rename(pack_path, new_pack_path) + os.rename(index_path, new_index_path) + + return cls(new_pack_path) + + #} END interface diff --git a/gitdb/gitdb/stream.py b/gitdb/gitdb/stream.py new file mode 100644 index 000000000..7c59f0b3c --- /dev/null +++ b/gitdb/gitdb/stream.py @@ -0,0 +1,755 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ + +from io import BytesIO + +import mmap +import os +import sys +import zlib + +from gitdb.fun import ( + msb_size, + stream_copy, + apply_delta_data, + connect_deltas, + delta_types +) + +from gitdb.util import ( + allocate_memory, + LazyMixin, + make_sha, + write, + close, +) + +from gitdb.const import NULL_BYTE, BYTE_SPACE +from gitdb.utils.encoding import force_bytes + +has_perf_mod = False +try: + from gitdb_speedups._perf import apply_delta as c_apply_delta + has_perf_mod = True +except ImportError: + pass + +__all__ = ('DecompressMemMapReader', 'FDCompressedSha1Writer', 'DeltaApplyReader', + 'Sha1Writer', 'FlexibleSha1Writer', 'ZippedStoreShaWriter', 'FDCompressedSha1Writer', + 'FDStream', 'NullStream') + + +#{ RO Streams + +class DecompressMemMapReader(LazyMixin): + + """Reads data in chunks from a memory map and decompresses it. The client sees + only the uncompressed data, respective file-like read calls are handling on-demand + buffered decompression accordingly + + A constraint on the total size of bytes is activated, simulating + a logical file within a possibly larger physical memory area + + To read efficiently, you clearly don't want to read individual bytes, instead, + read a few kilobytes at least. + + **Note:** The chunk-size should be carefully selected as it will involve quite a bit + of string copying due to the way the zlib is implemented. Its very wasteful, + hence we try to find a good tradeoff between allocation time and number of + times we actually allocate. An own zlib implementation would be good here + to better support streamed reading - it would only need to keep the mmap + and decompress it into chunks, that's all ... """ + __slots__ = ('_m', '_zip', '_buf', '_buflen', '_br', '_cws', '_cwe', '_s', '_close', + '_cbr', '_phi') + + max_read_size = 512 * 1024 # currently unused + + def __init__(self, m, close_on_deletion, size=None): + """Initialize with mmap for stream reading + :param m: must be content data - use new if you have object data and no size""" + self._m = m + self._zip = zlib.decompressobj() + self._buf = None # buffer of decompressed bytes + self._buflen = 0 # length of bytes in buffer + if size is not None: + self._s = size # size of uncompressed data to read in total + self._br = 0 # num uncompressed bytes read + self._cws = 0 # start byte of compression window + self._cwe = 0 # end byte of compression window + self._cbr = 0 # number of compressed bytes read + self._phi = False # is True if we parsed the header info + self._close = close_on_deletion # close the memmap on deletion ? + + def _set_cache_(self, attr): + assert attr == '_s' + # only happens for size, which is a marker to indicate we still + # have to parse the header from the stream + self._parse_header_info() + + def __del__(self): + self.close() + + def _parse_header_info(self): + """If this stream contains object data, parse the header info and skip the + stream to a point where each read will yield object content + + :return: parsed type_string, size""" + # read header + # should really be enough, cgit uses 8192 I believe + # And for good reason !! This needs to be that high for the header to be read correctly in all cases + maxb = 8192 + self._s = maxb + hdr = self.read(maxb) + hdrend = hdr.find(NULL_BYTE) + typ, size = hdr[:hdrend].split(BYTE_SPACE) + size = int(size) + self._s = size + + # adjust internal state to match actual header length that we ignore + # The buffer will be depleted first on future reads + self._br = 0 + hdrend += 1 + self._buf = BytesIO(hdr[hdrend:]) + self._buflen = len(hdr) - hdrend + + self._phi = True + + return typ, size + + #{ Interface + + @classmethod + def new(self, m, close_on_deletion=False): + """Create a new DecompressMemMapReader instance for acting as a read-only stream + This method parses the object header from m and returns the parsed + type and size, as well as the created stream instance. + + :param m: memory map on which to operate. It must be object data ( header + contents ) + :param close_on_deletion: if True, the memory map will be closed once we are + being deleted""" + inst = DecompressMemMapReader(m, close_on_deletion, 0) + typ, size = inst._parse_header_info() + return typ, size, inst + + def data(self): + """:return: random access compatible data we are working on""" + return self._m + + def close(self): + """Close our underlying stream of compressed bytes if this was allowed during initialization + :return: True if we closed the underlying stream + :note: can be called safely + """ + if self._close: + if hasattr(self._m, 'close'): + self._m.close() + self._close = False + # END handle resource freeing + + def compressed_bytes_read(self): + """ + :return: number of compressed bytes read. This includes the bytes it + took to decompress the header ( if there was one )""" + # ABSTRACT: When decompressing a byte stream, it can be that the first + # x bytes which were requested match the first x bytes in the loosely + # compressed datastream. This is the worst-case assumption that the reader + # does, it assumes that it will get at least X bytes from X compressed bytes + # in call cases. + # The caveat is that the object, according to our known uncompressed size, + # is already complete, but there are still some bytes left in the compressed + # stream that contribute to the amount of compressed bytes. + # How can we know that we are truly done, and have read all bytes we need + # to read ? + # Without help, we cannot know, as we need to obtain the status of the + # decompression. If it is not finished, we need to decompress more data + # until it is finished, to yield the actual number of compressed bytes + # belonging to the decompressed object + # We are using a custom zlib module for this, if its not present, + # we try to put in additional bytes up for decompression if feasible + # and check for the unused_data. + + # Only scrub the stream forward if we are officially done with the + # bytes we were to have. + if self._br == self._s and not self._zip.unused_data: + # manipulate the bytes-read to allow our own read method to continue + # but keep the window at its current position + self._br = 0 + if hasattr(self._zip, 'status'): + while self._zip.status == zlib.Z_OK: + self.read(mmap.PAGESIZE) + # END scrub-loop custom zlib + else: + # pass in additional pages, until we have unused data + while not self._zip.unused_data and self._cbr != len(self._m): + self.read(mmap.PAGESIZE) + # END scrub-loop default zlib + # END handle stream scrubbing + + # reset bytes read, just to be sure + self._br = self._s + # END handle stream scrubbing + + # unused data ends up in the unconsumed tail, which was removed + # from the count already + return self._cbr + + #} END interface + + def seek(self, offset, whence=getattr(os, 'SEEK_SET', 0)): + """Allows to reset the stream to restart reading + :raise ValueError: If offset and whence are not 0""" + if offset != 0 or whence != getattr(os, 'SEEK_SET', 0): + raise ValueError("Can only seek to position 0") + # END handle offset + + self._zip = zlib.decompressobj() + self._br = self._cws = self._cwe = self._cbr = 0 + if self._phi: + self._phi = False + del(self._s) # trigger header parsing on first access + # END skip header + + def read(self, size=-1): + if size < 1: + size = self._s - self._br + else: + size = min(size, self._s - self._br) + # END clamp size + + if size == 0: + return b'' + # END handle depletion + + # deplete the buffer, then just continue using the decompress object + # which has an own buffer. We just need this to transparently parse the + # header from the zlib stream + dat = b'' + if self._buf: + if self._buflen >= size: + # have enough data + dat = self._buf.read(size) + self._buflen -= size + self._br += size + return dat + else: + dat = self._buf.read() # ouch, duplicates data + size -= self._buflen + self._br += self._buflen + + self._buflen = 0 + self._buf = None + # END handle buffer len + # END handle buffer + + # decompress some data + # Abstract: zlib needs to operate on chunks of our memory map ( which may + # be large ), as it will otherwise and always fill in the 'unconsumed_tail' + # attribute which possible reads our whole map to the end, forcing + # everything to be read from disk even though just a portion was requested. + # As this would be a nogo, we workaround it by passing only chunks of data, + # moving the window into the memory map along as we decompress, which keeps + # the tail smaller than our chunk-size. This causes 'only' the chunk to be + # copied once, and another copy of a part of it when it creates the unconsumed + # tail. We have to use it to hand in the appropriate amount of bytes during + # the next read. + # + # Decompress in a loop until we have produced `size` bytes or run out + # of progress. Iteration (instead of recursion) keeps the call bounded + # for streams that consume many input bytes per produced output byte + # (e.g. zlib stored blocks of length zero); the previous recursive + # form blew the stack on inputs > ~1500 empty blocks (issue #120 + # follow-up). + dcompdat = b'' + while True: + tail = self._zip.unconsumed_tail + remaining = size - len(dcompdat) + if tail: + # move the window, make it as large as size demands. For code-clarity, + # we just take the chunk from our map again instead of reusing the unconsumed + # tail. The latter one would save some memory copying, but we could end up + # with not getting enough data uncompressed, so we had to sort that out as well. + # Now we just assume the worst case, hence the data is uncompressed and the window + # needs to be as large as the uncompressed bytes we want to read. + self._cws = self._cwe - len(tail) + self._cwe = self._cws + remaining + else: + cws = self._cws + self._cws = self._cwe + self._cwe = cws + remaining + # END handle tail + + # if window is too small, make it larger so zip can decompress something + if self._cwe - self._cws < 8: + self._cwe = self._cws + 8 + # END adjust winsize + + # takes a slice, but doesn't copy the data, it says ... + indata = self._m[self._cws:self._cwe] + + # get the actual window end to be sure we don't use it for computations + self._cwe = self._cws + len(indata) + chunk = self._zip.decompress(indata, remaining) + # update the amount of compressed bytes read + # We feed possibly overlapping chunks, which is why the unconsumed tail + # has to be taken into consideration, as well as the unused data + # if we hit the end of the stream + # NOTE: Behavior changed in PY2.7 onward, which requires special handling to make the tests work properly. + # They are thorough, and I assume it is truly working. + # Why is this logic as convoluted as it is ? Please look at the table in + # https://github.com/gitpython-developers/gitdb/issues/19 to learn about the test-results. + # Basically, on py2.6, you want to use branch 1, whereas on all other python version, the second branch + # will be the one that works. + # However, the zlib VERSIONs as well as the platform check is used to further match the entries in the + # table in the github issue. This is it ... it was the only way I could make this work everywhere. + # IT's CERTAINLY GOING TO BITE US IN THE FUTURE ... . + if getattr(zlib, 'ZLIB_RUNTIME_VERSION', zlib.ZLIB_VERSION) in ('1.2.7', '1.2.5') and not sys.platform == 'darwin': + unused_datalen = len(self._zip.unconsumed_tail) + else: + unused_datalen = len(self._zip.unconsumed_tail) + len(self._zip.unused_data) + # # end handle very special case ... + + consumed = len(indata) - unused_datalen + self._cbr += consumed + self._br += len(chunk) + if chunk: + if not isinstance(dcompdat, bytearray): + dcompdat = bytearray(dcompdat) + dcompdat.extend(chunk) + + # Stop when we have enough or there is no path to more output. + # `chunk` may legitimately be empty mid-stream when zlib is + # consuming header / dictionary frames; in that case we keep + # iterating as long as we are still feeding zlib new bytes + # (consumed > 0) and zlib has not flagged end-of-stream. The + # compressed_bytes_read() scrub loop drives this same code with + # _br manipulated to 0 past zip EOF; it terminates here because + # `getattr(_zip, 'eof', False)` is True or no compressed bytes + # are consumed. The empty-block recursion attack from issue #120 + # follow-up is bounded by the iteration; each empty block does + # consume input, so the loop walks the stream forward a constant + # amount per iteration without growing the call stack. + if len(dcompdat) >= size or self._br >= self._s: + break + zip_eof = getattr(self._zip, 'eof', False) + if not chunk and (zip_eof or len(indata) == 0 or consumed == 0): + break + # END iterative decompress + + if dat: + dcompdat = dat + dcompdat + # END prepend our cached data + + return dcompdat + + +class DeltaApplyReader(LazyMixin): + + """A reader which dynamically applies pack deltas to a base object, keeping the + memory demands to a minimum. + + The size of the final object is only obtainable once all deltas have been + applied, unless it is retrieved from a pack index. + + The uncompressed Delta has the following layout (MSB being a most significant + bit encoded dynamic size): + + * MSB Source Size - the size of the base against which the delta was created + * MSB Target Size - the size of the resulting data after the delta was applied + * A list of one byte commands (cmd) which are followed by a specific protocol: + + * cmd & 0x80 - copy delta_data[offset:offset+size] + + * Followed by an encoded offset into the delta data + * Followed by an encoded size of the chunk to copy + + * cmd & 0x7f - insert + + * insert cmd bytes from the delta buffer into the output stream + + * cmd == 0 - invalid operation ( or error in delta stream ) + """ + __slots__ = ( + "_bstream", # base stream to which to apply the deltas + "_dstreams", # tuple of delta stream readers + "_mm_target", # memory map of the delta-applied data + "_size", # actual number of bytes in _mm_target + "_br" # number of bytes read + ) + + #{ Configuration + k_max_memory_move = 250 * 1000 * 1000 + #} END configuration + + def __init__(self, stream_list): + """Initialize this instance with a list of streams, the first stream being + the delta to apply on top of all following deltas, the last stream being the + base object onto which to apply the deltas""" + assert len(stream_list) > 1, "Need at least one delta and one base stream" + + self._bstream = stream_list[-1] + self._dstreams = tuple(stream_list[:-1]) + self._br = 0 + + def _set_cache_too_slow_without_c(self, attr): + # the direct algorithm is fastest and most direct if there is only one + # delta. Also, the extra overhead might not be worth it for items smaller + # than X - definitely the case in python, every function call costs + # huge amounts of time + # if len(self._dstreams) * self._bstream.size < self.k_max_memory_move: + if len(self._dstreams) == 1: + return self._set_cache_brute_(attr) + + # Aggregate all deltas into one delta in reverse order. Hence we take + # the last delta, and reverse-merge its ancestor delta, until we receive + # the final delta data stream. + dcl = connect_deltas(self._dstreams) + + # call len directly, as the (optional) c version doesn't implement the sequence + # protocol + if dcl.rbound() == 0: + self._size = 0 + self._mm_target = allocate_memory(0) + return + # END handle empty list + + self._size = dcl.rbound() + self._mm_target = allocate_memory(self._size) + + bbuf = allocate_memory(self._bstream.size) + stream_copy(self._bstream.read, bbuf.write, self._bstream.size, 256 * mmap.PAGESIZE) + + # APPLY CHUNKS + write = self._mm_target.write + dcl.apply(bbuf, write) + + self._mm_target.seek(0) + + def _set_cache_brute_(self, attr): + """If we are here, we apply the actual deltas""" + # TODO: There should be a special case if there is only one stream + # Then the default-git algorithm should perform a tad faster, as the + # delta is not peaked into, causing less overhead. + buffer_info_list = list() + max_target_size = 0 + for dstream in self._dstreams: + buf = dstream.read(512) # read the header information + X + offset, src_size = msb_size(buf) + offset, target_size = msb_size(buf, offset) + buffer_info_list.append((buf[offset:], offset, src_size, target_size)) + max_target_size = max(max_target_size, target_size) + # END for each delta stream + + # sanity check - the first delta to apply should have the same source + # size as our actual base stream + base_size = self._bstream.size + target_size = max_target_size + + # if we have more than 1 delta to apply, we will swap buffers, hence we must + # assure that all buffers we use are large enough to hold all the results + if len(self._dstreams) > 1: + base_size = target_size = max(base_size, max_target_size) + # END adjust buffer sizes + + # Allocate private memory map big enough to hold the first base buffer + # We need random access to it + bbuf = allocate_memory(base_size) + stream_copy(self._bstream.read, bbuf.write, base_size, 256 * mmap.PAGESIZE) + + # allocate memory map large enough for the largest (intermediate) target + # We will use it as scratch space for all delta ops. If the final + # target buffer is smaller than our allocated space, we just use parts + # of it upon return. + tbuf = allocate_memory(target_size) + + # for each delta to apply, memory map the decompressed delta and + # work on the op-codes to reconstruct everything. + # For the actual copying, we use a seek and write pattern of buffer + # slices. + final_target_size = None + for (dbuf, offset, src_size, target_size), dstream in zip(reversed(buffer_info_list), reversed(self._dstreams)): + # allocate a buffer to hold all delta data - fill in the data for + # fast access. We do this as we know that reading individual bytes + # from our stream would be slower than necessary ( although possible ) + # The dbuf buffer contains commands after the first two MSB sizes, the + # offset specifies the amount of bytes read to get the sizes. + ddata = allocate_memory(dstream.size - offset) + ddata.write(dbuf) + # read the rest from the stream. The size we give is larger than necessary + stream_copy(dstream.read, ddata.write, dstream.size, 256 * mmap.PAGESIZE) + + ####################################################################### + if 'c_apply_delta' in globals(): + c_apply_delta(bbuf, ddata, tbuf) + else: + apply_delta_data(bbuf, src_size, ddata, len(ddata), tbuf.write) + ####################################################################### + + # finally, swap out source and target buffers. The target is now the + # base for the next delta to apply + bbuf, tbuf = tbuf, bbuf + bbuf.seek(0) + tbuf.seek(0) + final_target_size = target_size + # END for each delta to apply + + # its already seeked to 0, constrain it to the actual size + # NOTE: in the end of the loop, it swaps buffers, hence our target buffer + # is not tbuf, but bbuf ! + self._mm_target = bbuf + self._size = final_target_size + + #{ Configuration + if not has_perf_mod: + _set_cache_ = _set_cache_brute_ + else: + _set_cache_ = _set_cache_too_slow_without_c + + #} END configuration + + def read(self, count=0): + bl = self._size - self._br # bytes left + if count < 1 or count > bl: + count = bl + # NOTE: we could check for certain size limits, and possibly + # return buffers instead of strings to prevent byte copying + data = self._mm_target.read(count) + self._br += len(data) + return data + + def seek(self, offset, whence=getattr(os, 'SEEK_SET', 0)): + """Allows to reset the stream to restart reading + + :raise ValueError: If offset and whence are not 0""" + if offset != 0 or whence != getattr(os, 'SEEK_SET', 0): + raise ValueError("Can only seek to position 0") + # END handle offset + self._br = 0 + self._mm_target.seek(0) + + #{ Interface + + @classmethod + def new(cls, stream_list): + """ + Convert the given list of streams into a stream which resolves deltas + when reading from it. + + :param stream_list: two or more stream objects, first stream is a Delta + to the object that you want to resolve, followed by N additional delta + streams. The list's last stream must be a non-delta stream. + + :return: Non-Delta OPackStream object whose stream can be used to obtain + the decompressed resolved data + :raise ValueError: if the stream list cannot be handled""" + if len(stream_list) < 2: + raise ValueError("Need at least two streams") + # END single object special handling + + if stream_list[-1].type_id in delta_types: + raise ValueError( + "Cannot resolve deltas if there is no base object stream, last one was type: %s" % stream_list[-1].type) + # END check stream + return cls(stream_list) + + #} END interface + + #{ OInfo like Interface + + @property + def type(self): + return self._bstream.type + + @property + def type_id(self): + return self._bstream.type_id + + @property + def size(self): + """:return: number of uncompressed bytes in the stream""" + return self._size + + #} END oinfo like interface + + +#} END RO streams + + +#{ W Streams + +class Sha1Writer: + + """Simple stream writer which produces a sha whenever you like as it degests + everything it is supposed to write""" + __slots__ = "sha1" + + def __init__(self): + self.sha1 = make_sha() + + #{ Stream Interface + + def write(self, data): + """:raise IOError: If not all bytes could be written + :param data: byte object + :return: length of incoming data""" + + self.sha1.update(data) + + return len(data) + + # END stream interface + + #{ Interface + + def sha(self, as_hex=False): + """:return: sha so far + :param as_hex: if True, sha will be hex-encoded, binary otherwise""" + if as_hex: + return self.sha1.hexdigest() + return self.sha1.digest() + + #} END interface + + +class FlexibleSha1Writer(Sha1Writer): + + """Writer producing a sha1 while passing on the written bytes to the given + write function""" + __slots__ = 'writer' + + def __init__(self, writer): + Sha1Writer.__init__(self) + self.writer = writer + + def write(self, data): + Sha1Writer.write(self, data) + self.writer(data) + + +class ZippedStoreShaWriter(Sha1Writer): + + """Remembers everything someone writes to it and generates a sha""" + __slots__ = ('buf', 'zip') + + def __init__(self): + Sha1Writer.__init__(self) + self.buf = BytesIO() + self.zip = zlib.compressobj(zlib.Z_BEST_SPEED) + + def __getattr__(self, attr): + return getattr(self.buf, attr) + + def write(self, data): + alen = Sha1Writer.write(self, data) + self.buf.write(self.zip.compress(data)) + + return alen + + def close(self): + self.buf.write(self.zip.flush()) + + def seek(self, offset, whence=getattr(os, 'SEEK_SET', 0)): + """Seeking currently only supports to rewind written data + Multiple writes are not supported""" + if offset != 0 or whence != getattr(os, 'SEEK_SET', 0): + raise ValueError("Can only seek to position 0") + # END handle offset + self.buf.seek(0) + + def getvalue(self): + """:return: string value from the current stream position to the end""" + return self.buf.getvalue() + + +class FDCompressedSha1Writer(Sha1Writer): + + """Digests data written to it, making the sha available, then compress the + data and write it to the file descriptor + + **Note:** operates on raw file descriptors + **Note:** for this to work, you have to use the close-method of this instance""" + __slots__ = ("fd", "sha1", "zip") + + # default exception + exc = IOError("Failed to write all bytes to filedescriptor") + + def __init__(self, fd): + super().__init__() + self.fd = fd + self.zip = zlib.compressobj(zlib.Z_BEST_SPEED) + + #{ Stream Interface + + def write(self, data): + """:raise IOError: If not all bytes could be written + :return: length of incoming data""" + self.sha1.update(data) + cdata = self.zip.compress(data) + bytes_written = write(self.fd, cdata) + + if bytes_written != len(cdata): + raise self.exc + + return len(data) + + def close(self): + remainder = self.zip.flush() + if write(self.fd, remainder) != len(remainder): + raise self.exc + return close(self.fd) + + #} END stream interface + + +class FDStream: + + """A simple wrapper providing the most basic functions on a file descriptor + with the fileobject interface. Cannot use os.fdopen as the resulting stream + takes ownership""" + __slots__ = ("_fd", '_pos') + + def __init__(self, fd): + self._fd = fd + self._pos = 0 + + def write(self, data): + self._pos += len(data) + os.write(self._fd, data) + + def read(self, count=0): + if count == 0: + count = os.path.getsize(self._filepath) + # END handle read everything + + bytes = os.read(self._fd, count) + self._pos += len(bytes) + return bytes + + def fileno(self): + return self._fd + + def tell(self): + return self._pos + + def close(self): + close(self._fd) + + +class NullStream: + + """A stream that does nothing but providing a stream interface. + Use it like /dev/null""" + __slots__ = tuple() + + def read(self, size=0): + return '' + + def close(self): + pass + + def write(self, data): + return len(data) + + +#} END W streams diff --git a/gitdb/gitdb/test/__init__.py b/gitdb/gitdb/test/__init__.py new file mode 100644 index 000000000..03bd406be --- /dev/null +++ b/gitdb/gitdb/test/__init__.py @@ -0,0 +1,4 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ diff --git a/gitdb/gitdb/test/db/__init__.py b/gitdb/gitdb/test/db/__init__.py new file mode 100644 index 000000000..03bd406be --- /dev/null +++ b/gitdb/gitdb/test/db/__init__.py @@ -0,0 +1,4 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ diff --git a/gitdb/gitdb/test/db/lib.py b/gitdb/gitdb/test/db/lib.py new file mode 100644 index 000000000..408dd8c66 --- /dev/null +++ b/gitdb/gitdb/test/db/lib.py @@ -0,0 +1,128 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Base classes for object db testing""" +from gitdb.test.lib import ( + with_rw_directory, + with_packs_rw, + fixture_path, + TestBase +) + +from gitdb.stream import ( + Sha1Writer, + ZippedStoreShaWriter +) + +from gitdb.base import ( + IStream, + OStream, + OInfo +) + +from gitdb.exc import BadObject +from gitdb.typ import str_blob_type + +from io import BytesIO + +from struct import pack + + +__all__ = ('TestDBBase', 'with_rw_directory', 'with_packs_rw', 'fixture_path') + + +class TestDBBase(TestBase): + + """Base class providing testing routines on databases""" + + # data + two_lines = b'1234\nhello world' + all_data = (two_lines, ) + + def _assert_object_writing_simple(self, db): + # write a bunch of objects and query their streams and info + null_objs = db.size() + ni = 250 + for i in range(ni): + data = pack(">L", i) + istream = IStream(str_blob_type, len(data), BytesIO(data)) + new_istream = db.store(istream) + assert new_istream is istream + assert db.has_object(istream.binsha) + + info = db.info(istream.binsha) + assert isinstance(info, OInfo) + assert info.type == istream.type and info.size == istream.size + + stream = db.stream(istream.binsha) + assert isinstance(stream, OStream) + assert stream.binsha == info.binsha and stream.type == info.type + assert stream.read() == data + # END for each item + + assert db.size() == null_objs + ni + shas = list(db.sha_iter()) + assert len(shas) == db.size() + assert len(shas[0]) == 20 + + def _assert_object_writing(self, db): + """General tests to verify object writing, compatible to ObjectDBW + **Note:** requires write access to the database""" + # start in 'dry-run' mode, using a simple sha1 writer + ostreams = (ZippedStoreShaWriter, None) + for ostreamcls in ostreams: + for data in self.all_data: + dry_run = ostreamcls is not None + ostream = None + if ostreamcls is not None: + ostream = ostreamcls() + assert isinstance(ostream, Sha1Writer) + # END create ostream + + prev_ostream = db.set_ostream(ostream) + assert type(prev_ostream) in ostreams or prev_ostream in ostreams + istream = IStream(str_blob_type, len(data), BytesIO(data)) + + # store returns same istream instance, with new sha set + my_istream = db.store(istream) + sha = istream.binsha + assert my_istream is istream + assert db.has_object(sha) != dry_run + assert len(sha) == 20 + + # verify data - the slow way, we want to run code + if not dry_run: + info = db.info(sha) + assert str_blob_type == info.type + assert info.size == len(data) + + ostream = db.stream(sha) + assert ostream.read() == data + assert ostream.type == str_blob_type + assert ostream.size == len(data) + else: + self.assertRaises(BadObject, db.info, sha) + self.assertRaises(BadObject, db.stream, sha) + + # DIRECT STREAM COPY + # our data hase been written in object format to the StringIO + # we passed as output stream. No physical database representation + # was created. + # Test direct stream copy of object streams, the result must be + # identical to what we fed in + ostream.seek(0) + istream.stream = ostream + assert istream.binsha is not None + prev_sha = istream.binsha + + db.set_ostream(ZippedStoreShaWriter()) + db.store(istream) + assert istream.binsha == prev_sha + new_ostream = db.ostream() + + # note: only works as long our store write uses the same compression + # level, which is zip_best + assert ostream.getvalue() == new_ostream.getvalue() + # END for each data set + # END for each dry_run mode diff --git a/gitdb/gitdb/test/db/test_git.py b/gitdb/gitdb/test/db/test_git.py new file mode 100644 index 000000000..73ac1a08c --- /dev/null +++ b/gitdb/gitdb/test/db/test_git.py @@ -0,0 +1,53 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +import os +from gitdb.test.db.lib import ( + TestDBBase, + with_rw_directory +) +from gitdb.exc import BadObject +from gitdb.db import GitDB +from gitdb.base import OStream, OInfo +from gitdb.util import bin_to_hex + + +class TestGitDB(TestDBBase): + + def test_reading(self): + gdb = GitDB(os.path.join(self.gitrepopath, 'objects')) + + # we have packs and loose objects, alternates doesn't necessarily exist + assert 1 < len(gdb.databases()) < 4 + + # access should be possible + gitdb_sha = next(gdb.sha_iter()) + assert isinstance(gdb.info(gitdb_sha), OInfo) + assert isinstance(gdb.stream(gitdb_sha), OStream) + ni = 50 + assert gdb.size() >= ni + sha_list = list(gdb.sha_iter()) + assert len(sha_list) == gdb.size() + sha_list = sha_list[:ni] # speed up tests ... + + # This is actually a test for compound functionality, but it doesn't + # have a separate test module + # test partial shas + # this one as uneven and quite short + gitdb_sha_hex = bin_to_hex(gitdb_sha) + assert gdb.partial_to_complete_sha_hex(gitdb_sha_hex[:5]) == gitdb_sha + + # mix even/uneven hexshas + for i, binsha in enumerate(sha_list): + assert gdb.partial_to_complete_sha_hex(bin_to_hex(binsha)[:8 - (i % 2)]) == binsha + # END for each sha + + self.assertRaises(BadObject, gdb.partial_to_complete_sha_hex, "0000") + + @with_rw_directory + def test_writing(self, path): + gdb = GitDB(path) + + # its possible to write objects + self._assert_object_writing(gdb) diff --git a/gitdb/gitdb/test/db/test_loose.py b/gitdb/gitdb/test/db/test_loose.py new file mode 100644 index 000000000..295e2eecd --- /dev/null +++ b/gitdb/gitdb/test/db/test_loose.py @@ -0,0 +1,36 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +from gitdb.test.db.lib import ( + TestDBBase, + with_rw_directory +) +from gitdb.db import LooseObjectDB +from gitdb.exc import BadObject +from gitdb.util import bin_to_hex + + +class TestLooseDB(TestDBBase): + + @with_rw_directory + def test_basics(self, path): + ldb = LooseObjectDB(path) + + # write data + self._assert_object_writing(ldb) + + # verify sha iteration and size + shas = list(ldb.sha_iter()) + assert shas and len(shas[0]) == 20 + + assert len(shas) == ldb.size() + + # verify find short object + long_sha = bin_to_hex(shas[-1]) + for short_sha in (long_sha[:20], long_sha[:5]): + assert bin_to_hex(ldb.partial_to_complete_sha_hex(short_sha)) == long_sha + # END for each sha + + self.assertRaises(BadObject, ldb.partial_to_complete_sha_hex, '0000') + # raises if no object could be found diff --git a/gitdb/gitdb/test/db/test_mem.py b/gitdb/gitdb/test/db/test_mem.py new file mode 100644 index 000000000..882e54fe5 --- /dev/null +++ b/gitdb/gitdb/test/db/test_mem.py @@ -0,0 +1,34 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +from gitdb.test.db.lib import ( + TestDBBase, + with_rw_directory +) +from gitdb.db import ( + MemoryDB, + LooseObjectDB +) + + +class TestMemoryDB(TestDBBase): + + @with_rw_directory + def test_writing(self, path): + mdb = MemoryDB() + + # write data + self._assert_object_writing_simple(mdb) + + # test stream copy + ldb = LooseObjectDB(path) + assert ldb.size() == 0 + num_streams_copied = mdb.stream_copy(mdb.sha_iter(), ldb) + assert num_streams_copied == mdb.size() + + assert ldb.size() == mdb.size() + for sha in mdb.sha_iter(): + assert ldb.has_object(sha) + assert ldb.stream(sha).read() == mdb.stream(sha).read() + # END verify objects where copied and are equal diff --git a/gitdb/gitdb/test/db/test_pack.py b/gitdb/gitdb/test/db/test_pack.py new file mode 100644 index 000000000..bd07906da --- /dev/null +++ b/gitdb/gitdb/test/db/test_pack.py @@ -0,0 +1,87 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +from gitdb.test.db.lib import ( + TestDBBase, + with_rw_directory, + with_packs_rw +) +from gitdb.db import PackedDB + +from gitdb.exc import BadObject, AmbiguousObjectName +from gitdb.util import mman + +import os +import random +import sys + +import pytest + +class TestPackDB(TestDBBase): + + @with_rw_directory + @with_packs_rw + def test_writing(self, path): + if sys.platform == "win32": + pytest.skip("FIXME: Currently fail on windows") + + pdb = PackedDB(path) + + # on demand, we init our pack cache + num_packs = len(pdb.entities()) + assert pdb._st_mtime != 0 + + # test pack directory changed: + # packs removed - rename a file, should affect the glob + pack_path = pdb.entities()[0].pack().path() + new_pack_path = pack_path + "renamed" + if sys.platform == "win32": + # While using this function, we are not allowed to have any handle + # to this path, which is currently not the case. The pack caching + # does still have a handle :-( + mman.force_map_handle_removal_win(pack_path) + os.rename(pack_path, new_pack_path) + + pdb.update_cache(force=True) + assert len(pdb.entities()) == num_packs - 1 + + # packs added + os.rename(new_pack_path, pack_path) + pdb.update_cache(force=True) + assert len(pdb.entities()) == num_packs + + # bang on the cache + # access the Entities directly, as there is no iteration interface + # yet ( or required for now ) + sha_list = list(pdb.sha_iter()) + assert len(sha_list) == pdb.size() + + # hit all packs in random order + random.shuffle(sha_list) + + for sha in sha_list: + pdb.info(sha) + pdb.stream(sha) + # END for each sha to query + + # test short finding - be a bit more brutal here + max_bytes = 19 + min_bytes = 2 + num_ambiguous = 0 + for i, sha in enumerate(sha_list): + short_sha = sha[:max((i % max_bytes), min_bytes)] + try: + assert pdb.partial_to_complete_sha(short_sha, len(short_sha) * 2) == sha + except AmbiguousObjectName: + num_ambiguous += 1 + pass # valid, we can have short objects + # END exception handling + # END for each sha to find + + # we should have at least one ambiguous, considering the small sizes + # but in our pack, there is no ambiguous ... + # assert num_ambiguous + + # non-existing + self.assertRaises(BadObject, pdb.partial_to_complete_sha, b'\0\0', 4) diff --git a/gitdb/gitdb/test/db/test_ref.py b/gitdb/gitdb/test/db/test_ref.py new file mode 100644 index 000000000..0816e649f --- /dev/null +++ b/gitdb/gitdb/test/db/test_ref.py @@ -0,0 +1,58 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +from gitdb.test.db.lib import ( + TestDBBase, + with_rw_directory, +) +from gitdb.db import ReferenceDB + +from gitdb.util import ( + NULL_BIN_SHA, + hex_to_bin +) + +import os + + +class TestReferenceDB(TestDBBase): + + def make_alt_file(self, alt_path, alt_list): + """Create an alternates file which contains the given alternates. + The list can be empty""" + with open(alt_path, "wb") as alt_file: + for alt in alt_list: + alt_file.write(alt.encode("utf-8") + b"\n") + + @with_rw_directory + def test_writing(self, path): + alt_path = os.path.join(path, 'alternates') + rdb = ReferenceDB(alt_path) + assert len(rdb.databases()) == 0 + assert rdb.size() == 0 + assert len(list(rdb.sha_iter())) == 0 + + # try empty, non-existing + assert not rdb.has_object(NULL_BIN_SHA) + + # setup alternate file + # add two, one is invalid + own_repo_path = os.path.join(self.gitrepopath, 'objects') # use own repo + self.make_alt_file(alt_path, [own_repo_path, "invalid/path"]) + rdb.update_cache() + assert len(rdb.databases()) == 1 + + # we should now find a default revision of ours + gitdb_sha = next(rdb.sha_iter()) + assert rdb.has_object(gitdb_sha) + + # remove valid + self.make_alt_file(alt_path, ["just/one/invalid/path"]) + rdb.update_cache() + assert len(rdb.databases()) == 0 + + # add valid + self.make_alt_file(alt_path, [own_repo_path]) + rdb.update_cache() + assert len(rdb.databases()) == 1 diff --git a/gitdb/gitdb/test/fixtures/objects/7b/b839852ed5e3a069966281bb08d50012fb309b b/gitdb/gitdb/test/fixtures/objects/7b/b839852ed5e3a069966281bb08d50012fb309b new file mode 100644 index 000000000..021c2db34 Binary files /dev/null and b/gitdb/gitdb/test/fixtures/objects/7b/b839852ed5e3a069966281bb08d50012fb309b differ diff --git a/gitdb/gitdb/test/fixtures/objects/88/8401851f15db0eed60eb1bc29dec5ddcace911 b/gitdb/gitdb/test/fixtures/objects/88/8401851f15db0eed60eb1bc29dec5ddcace911 new file mode 100644 index 000000000..d60aeeff7 Binary files /dev/null and b/gitdb/gitdb/test/fixtures/objects/88/8401851f15db0eed60eb1bc29dec5ddcace911 differ diff --git a/gitdb/gitdb/test/fixtures/packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.idx b/gitdb/gitdb/test/fixtures/packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.idx new file mode 100644 index 000000000..fda5969bc Binary files /dev/null and b/gitdb/gitdb/test/fixtures/packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.idx differ diff --git a/gitdb/gitdb/test/fixtures/packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.pack b/gitdb/gitdb/test/fixtures/packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.pack new file mode 100644 index 000000000..a3209d2be Binary files /dev/null and b/gitdb/gitdb/test/fixtures/packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.pack differ diff --git a/gitdb/gitdb/test/fixtures/packs/pack-a2bf8e71d8c18879e499335762dd95119d93d9f1.idx b/gitdb/gitdb/test/fixtures/packs/pack-a2bf8e71d8c18879e499335762dd95119d93d9f1.idx new file mode 100644 index 000000000..a7d6c7177 Binary files /dev/null and b/gitdb/gitdb/test/fixtures/packs/pack-a2bf8e71d8c18879e499335762dd95119d93d9f1.idx differ diff --git a/gitdb/gitdb/test/fixtures/packs/pack-a2bf8e71d8c18879e499335762dd95119d93d9f1.pack b/gitdb/gitdb/test/fixtures/packs/pack-a2bf8e71d8c18879e499335762dd95119d93d9f1.pack new file mode 100644 index 000000000..955c424c7 Binary files /dev/null and b/gitdb/gitdb/test/fixtures/packs/pack-a2bf8e71d8c18879e499335762dd95119d93d9f1.pack differ diff --git a/gitdb/gitdb/test/fixtures/packs/pack-c0438c19fb16422b6bbcce24387b3264416d485b.idx b/gitdb/gitdb/test/fixtures/packs/pack-c0438c19fb16422b6bbcce24387b3264416d485b.idx new file mode 100644 index 000000000..87c635f48 Binary files /dev/null and b/gitdb/gitdb/test/fixtures/packs/pack-c0438c19fb16422b6bbcce24387b3264416d485b.idx differ diff --git a/gitdb/gitdb/test/fixtures/packs/pack-c0438c19fb16422b6bbcce24387b3264416d485b.pack b/gitdb/gitdb/test/fixtures/packs/pack-c0438c19fb16422b6bbcce24387b3264416d485b.pack new file mode 100644 index 000000000..a69b28ac6 Binary files /dev/null and b/gitdb/gitdb/test/fixtures/packs/pack-c0438c19fb16422b6bbcce24387b3264416d485b.pack differ diff --git a/gitdb/gitdb/test/lib.py b/gitdb/gitdb/test/lib.py new file mode 100644 index 000000000..8e602342c --- /dev/null +++ b/gitdb/gitdb/test/lib.py @@ -0,0 +1,192 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Utilities used in ODB testing""" +from gitdb import OStream + +import sys +import random +from array import array + +from io import BytesIO + +import glob +import unittest +import tempfile +import shutil +import os +import gc +import logging +from functools import wraps + + +#{ Bases + +class TestBase(unittest.TestCase): + """Base class for all tests + + TestCase providing access to readonly repositories using the following member variables. + + * gitrepopath + + * read-only base path of the git source repository, i.e. .../git/.git + """ + + #{ Invvariants + k_env_git_repo = "GITDB_TEST_GIT_REPO_BASE" + #} END invariants + + @classmethod + def setUpClass(cls): + try: + super().setUpClass() + except AttributeError: + pass + + cls.gitrepopath = os.environ.get(cls.k_env_git_repo) + if not cls.gitrepopath: + logging.info( + "You can set the %s environment variable to a .git repository of your choice - defaulting to the gitdb repository", cls.k_env_git_repo) + ospd = os.path.dirname + cls.gitrepopath = os.path.join(ospd(ospd(ospd(__file__))), '.git') + # end assure gitrepo is set + assert cls.gitrepopath.endswith('.git') + + +#} END bases + +#{ Decorators + +def with_rw_directory(func): + """Create a temporary directory which can be written to, remove it if the + test succeeds, but leave it otherwise to aid additional debugging""" + + def wrapper(self): + path = tempfile.mktemp(prefix=func.__name__) + os.mkdir(path) + keep = False + try: + try: + return func(self, path) + except Exception: + sys.stderr.write(f"Test {type(self).__name__}.{func.__name__} failed, output is at {path!r}\n") + keep = True + raise + finally: + # Need to collect here to be sure all handles have been closed. It appears + # a windows-only issue. In fact things should be deleted, as well as + # memory maps closed, once objects go out of scope. For some reason + # though this is not the case here unless we collect explicitly. + if not keep: + gc.collect() + shutil.rmtree(path) + # END handle exception + # END wrapper + + wrapper.__name__ = func.__name__ + return wrapper + + +def with_packs_rw(func): + """Function that provides a path into which the packs for testing should be + copied. Will pass on the path to the actual function afterwards""" + + def wrapper(self, path): + src_pack_glob = fixture_path('packs/*') + copy_files_globbed(src_pack_glob, path, hard_link_ok=True) + return func(self, path) + # END wrapper + + wrapper.__name__ = func.__name__ + return wrapper + +#} END decorators + +#{ Routines + + +def fixture_path(relapath=''): + """:return: absolute path into the fixture directory + :param relapath: relative path into the fixtures directory, or '' + to obtain the fixture directory itself""" + return os.path.join(os.path.dirname(__file__), 'fixtures', relapath) + + +def copy_files_globbed(source_glob, target_dir, hard_link_ok=False): + """Copy all files found according to the given source glob into the target directory + :param hard_link_ok: if True, hard links will be created if possible. Otherwise + the files will be copied""" + for src_file in glob.glob(source_glob): + if hard_link_ok and hasattr(os, 'link'): + target = os.path.join(target_dir, os.path.basename(src_file)) + try: + os.link(src_file, target) + except OSError: + shutil.copy(src_file, target_dir) + # END handle cross device links ( and resulting failure ) + else: + shutil.copy(src_file, target_dir) + # END try hard link + # END for each file to copy + + +def make_bytes(size_in_bytes, randomize=False): + """:return: string with given size in bytes + :param randomize: try to produce a very random stream""" + actual_size = size_in_bytes // 4 + producer = range(actual_size) + if randomize: + producer = list(producer) + random.shuffle(producer) + # END randomize + a = array('i', producer) + return a.tobytes() + + +def make_object(type, data): + """:return: bytes resembling an uncompressed object""" + odata = "blob %i\0" % len(data) + return odata.encode("ascii") + data + + +def make_memory_file(size_in_bytes, randomize=False): + """:return: tuple(size_of_stream, stream) + :param randomize: try to produce a very random stream""" + d = make_bytes(size_in_bytes, randomize) + return len(d), BytesIO(d) + +#} END routines + +#{ Stream Utilities + + +class DummyStream: + + def __init__(self): + self.was_read = False + self.bytes = 0 + self.closed = False + + def read(self, size): + self.was_read = True + self.bytes = size + + def close(self): + self.closed = True + + def _assert(self): + assert self.was_read + + +class DeriveTest(OStream): + + def __init__(self, sha, type, size, stream, *args, **kwargs): + self.myarg = kwargs.pop('myarg') + self.args = args + + def _assert(self): + assert self.args + assert self.myarg + +#} END stream utilitiess diff --git a/gitdb/gitdb/test/performance/__init__.py b/gitdb/gitdb/test/performance/__init__.py new file mode 100644 index 000000000..e69de29bb diff --git a/gitdb/gitdb/test/performance/lib.py b/gitdb/gitdb/test/performance/lib.py new file mode 100644 index 000000000..36916ede3 --- /dev/null +++ b/gitdb/gitdb/test/performance/lib.py @@ -0,0 +1,17 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Contains library functions""" +from gitdb.test.lib import TestBase + + + +#{ Base Classes + +class TestBigRepoR(TestBase): + """A placeholder in case we want to add additional functionality to all performance test-cases + """ + + +#} END base classes diff --git a/gitdb/gitdb/test/performance/test_pack.py b/gitdb/gitdb/test/performance/test_pack.py new file mode 100644 index 000000000..fc3d3349f --- /dev/null +++ b/gitdb/gitdb/test/performance/test_pack.py @@ -0,0 +1,127 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Performance tests for object store""" + +from gitdb.test.performance.lib import ( + TestBigRepoR +) + +from gitdb import ( + MemoryDB, + GitDB, + IStream, +) +from gitdb.typ import str_blob_type +from gitdb.exc import UnsupportedOperation +from gitdb.db.pack import PackedDB + +import sys +import os +from time import time + + +class TestPackedDBPerformance(TestBigRepoR): + + def test_pack_random_access(self): + pdb = PackedDB(os.path.join(self.gitrepopath, "objects/pack")) + + # sha lookup + st = time() + sha_list = list(pdb.sha_iter()) + elapsed = time() - st + ns = len(sha_list) + print("PDB: looked up %i shas by index in %f s ( %f shas/s )" % (ns, elapsed, ns / (elapsed or 1)), file=sys.stderr) + + # sha lookup: best-case and worst case access + pdb_pack_info = pdb._pack_info + # END shuffle shas + st = time() + for sha in sha_list: + pdb_pack_info(sha) + # END for each sha to look up + elapsed = time() - st + + # discard cache + del(pdb._entities) + pdb.entities() + print("PDB: looked up %i sha in %i packs in %f s ( %f shas/s )" % + (ns, len(pdb.entities()), elapsed, ns / (elapsed or 1)), file=sys.stderr) + # END for each random mode + + # query info and streams only + max_items = 10000 # can wait longer when testing memory + for pdb_fun in (pdb.info, pdb.stream): + st = time() + for sha in sha_list[:max_items]: + pdb_fun(sha) + elapsed = time() - st + print("PDB: Obtained %i object %s by sha in %f s ( %f items/s )" % + (max_items, pdb_fun.__name__.upper(), elapsed, max_items / (elapsed or 1)), file=sys.stderr) + # END for each function + + # retrieve stream and read all + max_items = 5000 + pdb_stream = pdb.stream + total_size = 0 + st = time() + for sha in sha_list[:max_items]: + stream = pdb_stream(sha) + read_len = len(stream.read()) + assert read_len == stream.size + total_size += stream.size + elapsed = time() - st + total_kib = total_size / 1000 + print("PDB: Obtained %i streams by sha and read all bytes totallying %i KiB ( %f KiB / s ) in %f s ( %f streams/s )" % + (max_items, total_kib, total_kib / (elapsed or 1), elapsed, max_items / (elapsed or 1)), file=sys.stderr) + + def test_loose_correctness(self): + """based on the pack(s) of our packed object DB, we will just copy and verify all objects in the back + into the loose object db (memory). + This should help finding dormant issues like this one https://github.com/gitpython-developers/GitPython/issues/220 + faster + :note: It doesn't seem this test can find the issue unless the given pack contains highly compressed + data files, like archives.""" + from gitdb.util import bin_to_hex + pdb = GitDB(os.path.join(self.gitrepopath, 'objects')) + mdb = MemoryDB() + for c, sha in enumerate(pdb.sha_iter()): + ostream = pdb.stream(sha) + # the issue only showed on larger files which are hardly compressible ... + if ostream.type != str_blob_type: + continue + istream = IStream(ostream.type, ostream.size, ostream.stream) + mdb.store(istream) + assert istream.binsha == sha, "Failed on object %s" % bin_to_hex(sha).decode('ascii') + # this can fail ... sometimes, so the packs dataset should be huge + assert len(mdb.stream(sha).read()) == ostream.size + + if c and c % 1000 == 0: + print("Verified %i loose object compression/decompression cycles" % c, file=sys.stderr) + mdb._cache.clear() + # end for each sha to copy + + def test_correctness(self): + pdb = PackedDB(os.path.join(self.gitrepopath, "objects/pack")) + # disabled for now as it used to work perfectly, checking big repositories takes a long time + print("Endurance run: verify streaming of objects (crc and sha)", file=sys.stderr) + for crc in range(2): + count = 0 + st = time() + for entity in pdb.entities(): + pack_verify = entity.is_valid_stream + sha_by_index = entity.index().sha + for index in range(entity.index().size()): + try: + assert pack_verify(sha_by_index(index), use_crc=crc) + count += 1 + except UnsupportedOperation: + pass + # END ignore old indices + # END for each index + # END for each entity + elapsed = time() - st + print("PDB: verified %i objects (crc=%i) in %f s ( %f objects/s )" % + (count, crc, elapsed, count / (elapsed or 1)), file=sys.stderr) + # END for each verify mode diff --git a/gitdb/gitdb/test/performance/test_pack_streaming.py b/gitdb/gitdb/test/performance/test_pack_streaming.py new file mode 100644 index 000000000..80c798bbc --- /dev/null +++ b/gitdb/gitdb/test/performance/test_pack_streaming.py @@ -0,0 +1,81 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Specific test for pack streams only""" + +from gitdb.test.performance.lib import ( + TestBigRepoR +) + +from gitdb.db.pack import PackedDB +from gitdb.stream import NullStream +from gitdb.pack import PackEntity + +import os +import sys +from time import time + + +class CountedNullStream(NullStream): + __slots__ = '_bw' + + def __init__(self): + self._bw = 0 + + def bytes_written(self): + return self._bw + + def write(self, d): + self._bw += NullStream.write(self, d) + + +class TestPackStreamingPerformance(TestBigRepoR): + + def test_pack_writing(self): + # see how fast we can write a pack from object streams. + # This will not be fast, as we take time for decompressing the streams as well + ostream = CountedNullStream() + pdb = PackedDB(os.path.join(self.gitrepopath, "objects/pack")) + + ni = 1000 + count = 0 + st = time() + for sha in pdb.sha_iter(): + count += 1 + pdb.stream(sha) + if count == ni: + break + # END gather objects for pack-writing + elapsed = time() - st + print("PDB Streaming: Got %i streams by sha in in %f s ( %f streams/s )" % + (ni, elapsed, ni / (elapsed or 1)), file=sys.stderr) + + st = time() + PackEntity.write_pack((pdb.stream(sha) for sha in pdb.sha_iter()), ostream.write, object_count=ni) + elapsed = time() - st + total_kb = ostream.bytes_written() / 1000 + print(sys.stderr, "PDB Streaming: Wrote pack of size %i kb in %f s (%f kb/s)" % + (total_kb, elapsed, total_kb / (elapsed or 1)), sys.stderr) + + def test_stream_reading(self): + # raise SkipTest() + pdb = PackedDB(os.path.join(self.gitrepopath, "objects/pack")) + + # streaming only, meant for --with-profile runs + ni = 5000 + count = 0 + pdb_stream = pdb.stream + total_size = 0 + st = time() + for sha in pdb.sha_iter(): + if count == ni: + break + stream = pdb_stream(sha) + stream.read() + total_size += stream.size + count += 1 + elapsed = time() - st + total_kib = total_size / 1000 + print(sys.stderr, "PDB Streaming: Got %i streams by sha and read all bytes totallying %i KiB ( %f KiB / s ) in %f s ( %f streams/s )" % + (ni, total_kib, total_kib / (elapsed or 1), elapsed, ni / (elapsed or 1)), sys.stderr) diff --git a/gitdb/gitdb/test/performance/test_stream.py b/gitdb/gitdb/test/performance/test_stream.py new file mode 100644 index 000000000..fb10871ae --- /dev/null +++ b/gitdb/gitdb/test/performance/test_stream.py @@ -0,0 +1,106 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Performance data streaming performance""" + +from gitdb.test.performance.lib import TestBigRepoR +from gitdb.db import LooseObjectDB +from gitdb import IStream + +from gitdb.util import bin_to_hex, remove +from gitdb.fun import chunk_size + +from time import time +import os +import sys + + +from gitdb.test.lib import ( + make_memory_file, + with_rw_directory, +) + + +#{ Utilities +def read_chunked_stream(stream): + total = 0 + while True: + chunk = stream.read(chunk_size) + total += len(chunk) + if len(chunk) < chunk_size: + break + # END read stream loop + assert total == stream.size + return stream + + +#} END utilities + +class TestObjDBPerformance(TestBigRepoR): + + large_data_size_bytes = 1000 * 1000 * 50 # some MiB should do it + moderate_data_size_bytes = 1000 * 1000 * 1 # just 1 MiB + + @with_rw_directory + def test_large_data_streaming(self, path): + ldb = LooseObjectDB(path) + string_ios = list() # list of streams we previously created + + # serial mode + for randomize in range(2): + desc = (randomize and 'random ') or '' + print("Creating %s data ..." % desc, file=sys.stderr) + st = time() + size, stream = make_memory_file(self.large_data_size_bytes, randomize) + elapsed = time() - st + print("Done (in %f s)" % elapsed, file=sys.stderr) + string_ios.append(stream) + + # writing - due to the compression it will seem faster than it is + st = time() + sha = ldb.store(IStream('blob', size, stream)).binsha + elapsed_add = time() - st + assert ldb.has_object(sha) + db_file = ldb.readable_db_object_path(bin_to_hex(sha)) + fsize_kib = os.path.getsize(db_file) / 1000 + + size_kib = size / 1000 + print("Added %i KiB (filesize = %i KiB) of %s data to loose odb in %f s ( %f Write KiB / s)" % + (size_kib, fsize_kib, desc, elapsed_add, size_kib / (elapsed_add or 1)), file=sys.stderr) + + # reading all at once + st = time() + ostream = ldb.stream(sha) + shadata = ostream.read() + elapsed_readall = time() - st + + stream.seek(0) + assert shadata == stream.getvalue() + print("Read %i KiB of %s data at once from loose odb in %f s ( %f Read KiB / s)" % + (size_kib, desc, elapsed_readall, size_kib / (elapsed_readall or 1)), file=sys.stderr) + + # reading in chunks of 1 MiB + cs = 512 * 1000 + chunks = list() + st = time() + ostream = ldb.stream(sha) + while True: + data = ostream.read(cs) + chunks.append(data) + if len(data) < cs: + break + # END read in chunks + elapsed_readchunks = time() - st + + stream.seek(0) + assert b''.join(chunks) == stream.getvalue() + + cs_kib = cs / 1000 + print("Read %i KiB of %s data in %i KiB chunks from loose odb in %f s ( %f Read KiB / s)" % + (size_kib, desc, cs_kib, elapsed_readchunks, size_kib / (elapsed_readchunks or 1)), file=sys.stderr) + + # del db file so we keep something to do + ostream = None # To release the file handle (win) + remove(db_file) + # END for each randomization factor diff --git a/gitdb/gitdb/test/test_base.py b/gitdb/gitdb/test/test_base.py new file mode 100644 index 000000000..17906c9c2 --- /dev/null +++ b/gitdb/gitdb/test/test_base.py @@ -0,0 +1,105 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Test for object db""" +from gitdb.test.lib import ( + TestBase, + DummyStream, + DeriveTest, +) + +from gitdb import ( + OInfo, + OPackInfo, + ODeltaPackInfo, + OStream, + OPackStream, + ODeltaPackStream, + IStream +) +from gitdb.util import ( + NULL_BIN_SHA +) + +from gitdb.typ import ( + str_blob_type +) + + +class TestBaseTypes(TestBase): + + def test_streams(self): + # test info + sha = NULL_BIN_SHA + s = 20 + blob_id = 3 + + info = OInfo(sha, str_blob_type, s) + assert info.binsha == sha + assert info.type == str_blob_type + assert info.type_id == blob_id + assert info.size == s + + # test pack info + # provides type_id + pinfo = OPackInfo(0, blob_id, s) + assert pinfo.type == str_blob_type + assert pinfo.type_id == blob_id + assert pinfo.pack_offset == 0 + + dpinfo = ODeltaPackInfo(0, blob_id, s, sha) + assert dpinfo.type == str_blob_type + assert dpinfo.type_id == blob_id + assert dpinfo.delta_info == sha + assert dpinfo.pack_offset == 0 + + # test ostream + stream = DummyStream() + ostream = OStream(*(info + (stream, ))) + assert ostream.stream is stream + ostream.read(15) + stream._assert() + assert stream.bytes == 15 + ostream.read(20) + assert stream.bytes == 20 + + # test packstream + postream = OPackStream(*(pinfo + (stream, ))) + assert postream.stream is stream + postream.read(10) + stream._assert() + assert stream.bytes == 10 + + # test deltapackstream + dpostream = ODeltaPackStream(*(dpinfo + (stream, ))) + assert dpostream.stream is stream + dpostream.read(5) + stream._assert() + assert stream.bytes == 5 + + # derive with own args + DeriveTest(sha, str_blob_type, s, stream, 'mine', myarg=3)._assert() + + # test istream + istream = IStream(str_blob_type, s, stream) + assert istream.binsha == None + istream.binsha = sha + assert istream.binsha == sha + + assert len(istream.binsha) == 20 + assert len(istream.hexsha) == 40 + + assert istream.size == s + istream.size = s * 2 + assert istream.size == s * 2 + assert istream.type == str_blob_type + istream.type = "something" + assert istream.type == "something" + assert istream.stream is stream + istream.stream = None + assert istream.stream is None + + assert istream.error is None + istream.error = Exception() + assert isinstance(istream.error, Exception) diff --git a/gitdb/gitdb/test/test_example.py b/gitdb/gitdb/test/test_example.py new file mode 100644 index 000000000..3b4c9084b --- /dev/null +++ b/gitdb/gitdb/test/test_example.py @@ -0,0 +1,43 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Module with examples from the tutorial section of the docs""" +import os +from gitdb.test.lib import TestBase +from gitdb import IStream +from gitdb.db import LooseObjectDB + +from io import BytesIO + + +class TestExamples(TestBase): + + def test_base(self): + ldb = LooseObjectDB(os.path.join(self.gitrepopath, 'objects')) + + for sha1 in ldb.sha_iter(): + oinfo = ldb.info(sha1) + ostream = ldb.stream(sha1) + assert oinfo[:3] == ostream[:3] + + assert len(ostream.read()) == ostream.size + assert ldb.has_object(oinfo.binsha) + # END for each sha in database + # assure we close all files + try: + del(ostream) + del(oinfo) + except UnboundLocalError: + pass + # END ignore exception if there are no loose objects + + data = b"my data" + istream = IStream("blob", len(data), BytesIO(data)) + + # the object does not yet have a sha + assert istream.binsha is None + ldb.store(istream) + # now the sha is set + assert len(istream.binsha) == 20 + assert ldb.has_object(istream.binsha) diff --git a/gitdb/gitdb/test/test_pack.py b/gitdb/gitdb/test/test_pack.py new file mode 100644 index 000000000..e72348228 --- /dev/null +++ b/gitdb/gitdb/test/test_pack.py @@ -0,0 +1,249 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Test everything about packs reading and writing""" +from gitdb.test.lib import ( + TestBase, + with_rw_directory, + fixture_path +) + +from gitdb.stream import DeltaApplyReader + +from gitdb.pack import ( + PackEntity, + PackIndexFile, + PackFile +) + +from gitdb.base import ( + OInfo, + OStream, +) + +from gitdb.fun import delta_types +from gitdb.exc import UnsupportedOperation +from gitdb.util import to_bin_sha + +import pytest + +import os +import tempfile + + +#{ Utilities +def bin_sha_from_filename(filename): + return to_bin_sha(os.path.splitext(os.path.basename(filename))[0][5:]) +#} END utilities + + +class TestPack(TestBase): + + packindexfile_v1 = (fixture_path('packs/pack-c0438c19fb16422b6bbcce24387b3264416d485b.idx'), 1, 67) + packindexfile_v2 = (fixture_path('packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.idx'), 2, 30) + packindexfile_v2_3_ascii = (fixture_path('packs/pack-a2bf8e71d8c18879e499335762dd95119d93d9f1.idx'), 2, 42) + packfile_v2_1 = (fixture_path('packs/pack-c0438c19fb16422b6bbcce24387b3264416d485b.pack'), 2, packindexfile_v1[2]) + packfile_v2_2 = (fixture_path('packs/pack-11fdfa9e156ab73caae3b6da867192221f2089c2.pack'), 2, packindexfile_v2[2]) + packfile_v2_3_ascii = ( + fixture_path('packs/pack-a2bf8e71d8c18879e499335762dd95119d93d9f1.pack'), 2, packindexfile_v2_3_ascii[2]) + + def _assert_index_file(self, index, version, size): + assert index.packfile_checksum() != index.indexfile_checksum() + assert len(index.packfile_checksum()) == 20 + assert len(index.indexfile_checksum()) == 20 + assert index.version() == version + assert index.size() == size + assert len(index.offsets()) == size + + # get all data of all objects + for oidx in range(index.size()): + sha = index.sha(oidx) + assert oidx == index.sha_to_index(sha) + + entry = index.entry(oidx) + assert len(entry) == 3 + + assert entry[0] == index.offset(oidx) + assert entry[1] == sha + assert entry[2] == index.crc(oidx) + + # verify partial sha + for l in (4, 8, 11, 17, 20): + assert index.partial_sha_to_index(sha[:l], l * 2) == oidx + + # END for each object index in indexfile + self.assertRaises(ValueError, index.partial_sha_to_index, "\0", 2) + + def _assert_pack_file(self, pack, version, size): + assert pack.version() == 2 + assert pack.size() == size + assert len(pack.checksum()) == 20 + + num_obj = 0 + for obj in pack.stream_iter(): + num_obj += 1 + info = pack.info(obj.pack_offset) + stream = pack.stream(obj.pack_offset) + + assert info.pack_offset == stream.pack_offset + assert info.type_id == stream.type_id + assert hasattr(stream, 'read') + + # it should be possible to read from both streams + assert obj.read() == stream.read() + + streams = pack.collect_streams(obj.pack_offset) + assert streams + + # read the stream + try: + dstream = DeltaApplyReader.new(streams) + except ValueError: + # ignore these, old git versions use only ref deltas, + # which we haven't resolved ( as we are without an index ) + # Also ignore non-delta streams + continue + # END get deltastream + + # read all + data = dstream.read() + assert len(data) == dstream.size + + # test seek + dstream.seek(0) + assert dstream.read() == data + + # read chunks + # NOTE: the current implementation is safe, it basically transfers + # all calls to the underlying memory map + + # END for each object + assert num_obj == size + + def test_pack_index(self): + # check version 1 and 2 + for indexfile, version, size in (self.packindexfile_v1, self.packindexfile_v2): + index = PackIndexFile(indexfile) + self._assert_index_file(index, version, size) + # END run tests + + def test_pack(self): + # there is this special version 3, but apparently its like 2 ... + for packfile, version, size in (self.packfile_v2_3_ascii, self.packfile_v2_1, self.packfile_v2_2): + pack = PackFile(packfile) + self._assert_pack_file(pack, version, size) + # END for each pack to test + + @with_rw_directory + def test_pack_entity(self, rw_dir): + pack_objs = list() + for packinfo, indexinfo in ((self.packfile_v2_1, self.packindexfile_v1), + (self.packfile_v2_2, self.packindexfile_v2), + (self.packfile_v2_3_ascii, self.packindexfile_v2_3_ascii)): + packfile, version, size = packinfo + indexfile, version, size = indexinfo + entity = PackEntity(packfile) + assert entity.pack().path() == packfile + assert entity.index().path() == indexfile + pack_objs.extend(entity.stream_iter()) + + count = 0 + for info, stream in zip(entity.info_iter(), entity.stream_iter()): + count += 1 + assert info.binsha == stream.binsha + assert len(info.binsha) == 20 + assert info.type_id == stream.type_id + assert info.size == stream.size + + # we return fully resolved items, which is implied by the sha centric access + assert not info.type_id in delta_types + + # try all calls + assert len(entity.collect_streams(info.binsha)) + oinfo = entity.info(info.binsha) + assert isinstance(oinfo, OInfo) + assert oinfo.binsha is not None + ostream = entity.stream(info.binsha) + assert isinstance(ostream, OStream) + assert ostream.binsha is not None + + # verify the stream + try: + assert entity.is_valid_stream(info.binsha, use_crc=True) + except UnsupportedOperation: + pass + # END ignore version issues + assert entity.is_valid_stream(info.binsha, use_crc=False) + # END for each info, stream tuple + assert count == size + + # END for each entity + + # pack writing - write all packs into one + # index path can be None + pack_path1 = tempfile.mktemp('', "pack1", rw_dir) + pack_path2 = tempfile.mktemp('', "pack2", rw_dir) + index_path = tempfile.mktemp('', 'index', rw_dir) + iteration = 0 + + def rewind_streams(): + for obj in pack_objs: + obj.stream.seek(0) + # END utility + for ppath, ipath, num_obj in zip((pack_path1, pack_path2), + (index_path, None), + (len(pack_objs), None)): + iwrite = None + if ipath: + ifile = open(ipath, 'wb') + iwrite = ifile.write + # END handle ip + + # make sure we rewind the streams ... we work on the same objects over and over again + if iteration > 0: + rewind_streams() + # END rewind streams + iteration += 1 + + with open(ppath, 'wb') as pfile: + pack_sha, index_sha = PackEntity.write_pack(pack_objs, pfile.write, iwrite, object_count=num_obj) + assert os.path.getsize(ppath) > 100 + + # verify pack + pf = PackFile(ppath) + assert pf.size() == len(pack_objs) + assert pf.version() == PackFile.pack_version_default + assert pf.checksum() == pack_sha + pf.close() + + # verify index + if ipath is not None: + ifile.close() + assert os.path.getsize(ipath) > 100 + idx = PackIndexFile(ipath) + assert idx.version() == PackIndexFile.index_version_default + assert idx.packfile_checksum() == pack_sha + assert idx.indexfile_checksum() == index_sha + assert idx.size() == len(pack_objs) + idx.close() + # END verify files exist + # END for each packpath, indexpath pair + + # verify the packs thoroughly + rewind_streams() + entity = PackEntity.create(pack_objs, rw_dir) + count = 0 + for info in entity.info_iter(): + count += 1 + for use_crc in range(2): + assert entity.is_valid_stream(info.binsha, use_crc) + # END for each crc mode + # END for each info + assert count == len(pack_objs) + entity.close() + + def test_pack_64(self): + # TODO: hex-edit a pack helping us to verify that we can handle 64 byte offsets + # of course without really needing such a huge pack + pytest.skip('not implemented') diff --git a/gitdb/gitdb/test/test_stream.py b/gitdb/gitdb/test/test_stream.py new file mode 100644 index 000000000..f36b06b96 --- /dev/null +++ b/gitdb/gitdb/test/test_stream.py @@ -0,0 +1,200 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Test for object db""" + +from gitdb.test.lib import ( + TestBase, + DummyStream, + make_bytes, + make_object, + fixture_path +) + +from gitdb import ( + DecompressMemMapReader, + FDCompressedSha1Writer, + LooseObjectDB, + Sha1Writer, + MemoryDB, + IStream, +) +from gitdb.util import hex_to_bin + +import zlib +from gitdb.typ import ( + str_blob_type +) + +import tempfile +import os +from io import BytesIO + + +class TestStream(TestBase): + + """Test stream classes""" + + data_sizes = (15, 10000, 1000 * 1024 + 512) + + def _assert_stream_reader(self, stream, cdata, rewind_stream=lambda s: None): + """Make stream tests - the orig_stream is seekable, allowing it to be + rewound and reused + :param cdata: the data we expect to read from stream, the contents + :param rewind_stream: function called to rewind the stream to make it ready + for reuse""" + ns = 10 + assert len(cdata) > ns - 1, "Data must be larger than %i, was %i" % (ns, len(cdata)) + + # read in small steps + ss = len(cdata) // ns + for i in range(ns): + data = stream.read(ss) + chunk = cdata[i * ss:(i + 1) * ss] + assert data == chunk + # END for each step + rest = stream.read() + if rest: + assert rest == cdata[-len(rest):] + # END handle rest + + if isinstance(stream, DecompressMemMapReader): + assert len(stream.data()) == stream.compressed_bytes_read() + # END handle special type + + rewind_stream(stream) + + # read everything + rdata = stream.read() + assert rdata == cdata + + if isinstance(stream, DecompressMemMapReader): + assert len(stream.data()) == stream.compressed_bytes_read() + # END handle special type + + def test_decompress_reader(self): + for close_on_deletion in range(2): + for with_size in range(2): + for ds in self.data_sizes: + cdata = make_bytes(ds, randomize=False) + + # zdata = zipped actual data + # cdata = original content data + + # create reader + if with_size: + # need object data + zdata = zlib.compress(make_object(str_blob_type, cdata)) + typ, size, reader = DecompressMemMapReader.new(zdata, close_on_deletion) + assert size == len(cdata) + assert typ == str_blob_type + + # even if we don't set the size, it will be set automatically on first read + test_reader = DecompressMemMapReader(zdata, close_on_deletion=False) + assert test_reader._s == len(cdata) + else: + # here we need content data + zdata = zlib.compress(cdata) + reader = DecompressMemMapReader(zdata, close_on_deletion, len(cdata)) + assert reader._s == len(cdata) + # END get reader + + self._assert_stream_reader(reader, cdata, lambda r: r.seek(0)) + + # put in a dummy stream for closing + dummy = DummyStream() + reader._m = dummy + + assert not dummy.closed + del(reader) + assert dummy.closed == close_on_deletion + # END for each datasize + # END whether size should be used + # END whether stream should be closed when deleted + + def test_sha_writer(self): + writer = Sha1Writer() + assert 2 == writer.write(b"hi") + assert len(writer.sha(as_hex=1)) == 40 + assert len(writer.sha(as_hex=0)) == 20 + + # make sure it does something ;) + prev_sha = writer.sha() + writer.write(b"hi again") + assert writer.sha() != prev_sha + + def test_compressed_writer(self): + for ds in self.data_sizes: + fd, path = tempfile.mkstemp() + ostream = FDCompressedSha1Writer(fd) + data = make_bytes(ds, randomize=False) + + # for now, just a single write, code doesn't care about chunking + assert len(data) == ostream.write(data) + ostream.close() + + # its closed already + self.assertRaises(OSError, os.close, fd) + + # read everything back, compare to data we zip + fd = os.open(path, os.O_RDONLY | getattr(os, 'O_BINARY', 0)) + written_data = os.read(fd, os.path.getsize(path)) + assert len(written_data) == os.path.getsize(path) + os.close(fd) + assert written_data == zlib.compress(data, 1) # best speed + + os.remove(path) + # END for each os + + def test_decompress_reader_special_case(self): + odb = LooseObjectDB(fixture_path('objects')) + mdb = MemoryDB() + for sha in (b'888401851f15db0eed60eb1bc29dec5ddcace911', + b'7bb839852ed5e3a069966281bb08d50012fb309b',): + ostream = odb.stream(hex_to_bin(sha)) + + # if there is a bug, we will be missing one byte exactly ! + data = ostream.read() + assert len(data) == ostream.size + + # Putting it back in should yield nothing new - after all, we have + dump = mdb.store(IStream(ostream.type, ostream.size, BytesIO(data))) + assert dump.hexsha == sha + # end for each loose object sha to test + + def test_decompress_reader_chunked_read_does_not_terminate_early(self): + """Regression test for #120: read(N) must not return b'' before EOF. + + zlib can consume input without producing decompressed output (e.g. + while ingesting block headers). The reader's internal recursion + previously bailed on any empty zip output, so a caller reading in + small chunks via the standard `while chunk := stream.read(N)` idiom + would terminate at the first empty chunk -- before the actual end + of the uncompressed stream. + """ + # Highly compressible data exposes the bug because each zlib chunk + # spans many uncompressed bytes -- intermediate decompress() calls + # often return empty while consuming input. + data = b"hello world! " * 1000 + zdata = zlib.compress(data) + + # Loop with a small chunk size to force many internal read/decompression + # iterations before EOF. + for chunk_size in (1, 4, 16, 64): + reader = DecompressMemMapReader( + zdata, close_on_deletion=False, size=len(data) + ) + out = bytearray() + while True: + chunk = reader.read(chunk_size) + if not chunk: + break + out.extend(chunk) + assert bytes(out) == data, ( + f"chunk_size={chunk_size}: got {len(out)}/{len(data)} bytes" + ) + assert reader._br == reader._s, ( + f"chunk_size={chunk_size}: stream stopped at " + f"{reader._br}/{reader._s}" + ) diff --git a/gitdb/gitdb/test/test_util.py b/gitdb/gitdb/test/test_util.py new file mode 100644 index 000000000..166b33c3a --- /dev/null +++ b/gitdb/gitdb/test/test_util.py @@ -0,0 +1,100 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Test for object db""" +import tempfile +import os + +from gitdb.test.lib import TestBase +from gitdb.util import ( + to_hex_sha, + to_bin_sha, + NULL_HEX_SHA, + LockedFD +) + + +class TestUtils(TestBase): + + def test_basics(self): + assert to_hex_sha(NULL_HEX_SHA) == NULL_HEX_SHA + assert len(to_bin_sha(NULL_HEX_SHA)) == 20 + assert to_hex_sha(to_bin_sha(NULL_HEX_SHA)) == NULL_HEX_SHA.encode("ascii") + + def _cmp_contents(self, file_path, data): + # raise if data from file at file_path + # does not match data string + with open(file_path, "rb") as fp: + assert fp.read() == data.encode("ascii") + + def test_lockedfd(self): + my_file = tempfile.mktemp() + orig_data = "hello" + new_data = "world" + with open(my_file, "wb") as my_file_fp: + my_file_fp.write(orig_data.encode("ascii")) + + try: + lfd = LockedFD(my_file) + lockfilepath = lfd._lockfilepath() + + # cannot end before it was started + self.assertRaises(AssertionError, lfd.rollback) + self.assertRaises(AssertionError, lfd.commit) + + # open for writing + assert not os.path.isfile(lockfilepath) + wfd = lfd.open(write=True) + assert lfd._fd is wfd + assert os.path.isfile(lockfilepath) + + # write data and fail + os.write(wfd, new_data.encode("ascii")) + lfd.rollback() + assert lfd._fd is None + self._cmp_contents(my_file, orig_data) + assert not os.path.isfile(lockfilepath) + + # additional call doesn't fail + lfd.commit() + lfd.rollback() + + # test reading + lfd = LockedFD(my_file) + rfd = lfd.open(write=False) + assert os.read(rfd, len(orig_data)) == orig_data.encode("ascii") + + assert os.path.isfile(lockfilepath) + # deletion rolls back + del(lfd) + assert not os.path.isfile(lockfilepath) + + # write data - concurrently + lfd = LockedFD(my_file) + olfd = LockedFD(my_file) + assert not os.path.isfile(lockfilepath) + wfdstream = lfd.open(write=True, stream=True) # this time as stream + assert os.path.isfile(lockfilepath) + # another one fails + self.assertRaises(IOError, olfd.open) + + wfdstream.write(new_data.encode("ascii")) + lfd.commit() + assert not os.path.isfile(lockfilepath) + self._cmp_contents(my_file, new_data) + + # could test automatic _end_writing on destruction + finally: + os.remove(my_file) + # END final cleanup + + # try non-existing file for reading + lfd = LockedFD(tempfile.mktemp()) + try: + lfd.open(write=False) + except OSError: + assert not os.path.exists(lfd._lockfilepath()) + else: + self.fail("expected OSError") + # END handle exceptions diff --git a/gitdb/gitdb/typ.py b/gitdb/gitdb/typ.py new file mode 100644 index 000000000..314db50a7 --- /dev/null +++ b/gitdb/gitdb/typ.py @@ -0,0 +1,10 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +"""Module containing information about types known to the database""" + +str_blob_type = b'blob' +str_commit_type = b'commit' +str_tree_type = b'tree' +str_tag_type = b'tag' diff --git a/gitdb/gitdb/util.py b/gitdb/gitdb/util.py new file mode 100644 index 000000000..bb6d8797a --- /dev/null +++ b/gitdb/gitdb/util.py @@ -0,0 +1,398 @@ +# Copyright (C) 2010, 2011 Sebastian Thiel (byronimo@gmail.com) and contributors +# +# This module is part of GitDB and is released under +# the New BSD License: https://opensource.org/license/bsd-3-clause/ +import binascii +import os +import mmap +import sys +import time +import errno + +from io import BytesIO + +from smmap import ( + StaticWindowMapManager, + SlidingWindowMapManager, + SlidingWindowMapBuffer +) + +# initialize our global memory manager instance +# Use it to free cached (and unused) resources. +mman = SlidingWindowMapManager() +# END handle mman + +import hashlib + +try: + from struct import unpack_from +except ImportError: + from struct import unpack, calcsize + __calcsize_cache = dict() + + def unpack_from(fmt, data, offset=0): + try: + size = __calcsize_cache[fmt] + except KeyError: + size = calcsize(fmt) + __calcsize_cache[fmt] = size + # END exception handling + return unpack(fmt, data[offset: offset + size]) + # END own unpack_from implementation + + +#{ Aliases + +hex_to_bin = binascii.a2b_hex +bin_to_hex = binascii.b2a_hex + +# errors +ENOENT = errno.ENOENT + +# os shortcuts +exists = os.path.exists +mkdir = os.mkdir +chmod = os.chmod +isdir = os.path.isdir +isfile = os.path.isfile +rename = os.rename +dirname = os.path.dirname +basename = os.path.basename +join = os.path.join +read = os.read +write = os.write +close = os.close +fsync = os.fsync + + +def _retry(func, *args, **kwargs): + # Wrapper around functions, that are problematic on "Windows". Sometimes + # the OS or someone else has still a handle to the file + if sys.platform == "win32": + for _ in range(10): + try: + return func(*args, **kwargs) + except Exception: + time.sleep(0.1) + return func(*args, **kwargs) + else: + return func(*args, **kwargs) + + +def remove(*args, **kwargs): + return _retry(os.remove, *args, **kwargs) + + +# Backwards compatibility imports +from gitdb.const import ( + NULL_BIN_SHA, + NULL_HEX_SHA +) + +#} END Aliases + +#{ compatibility stuff ... + + +class _RandomAccessBytesIO: + + """Wrapper to provide required functionality in case memory maps cannot or may + not be used. This is only really required in python 2.4""" + __slots__ = '_sio' + + def __init__(self, buf=''): + self._sio = BytesIO(buf) + + def __getattr__(self, attr): + return getattr(self._sio, attr) + + def __len__(self): + return len(self.getvalue()) + + def __getitem__(self, i): + return self.getvalue()[i] + + def __getslice__(self, start, end): + return self.getvalue()[start:end] + + +def byte_ord(b): + """ + Return the integer representation of the byte string. This supports Python + 3 byte arrays as well as standard strings. + """ + try: + return ord(b) + except TypeError: + return b + +#} END compatibility stuff ... + +#{ Routines + + +def make_sha(source=b''): + """A python2.4 workaround for the sha/hashlib module fiasco + + **Note** From the dulwich project """ + try: + return hashlib.sha1(source) + except NameError: + import sha + sha1 = sha.sha(source) + return sha1 + + +def allocate_memory(size): + """:return: a file-protocol accessible memory block of the given size""" + if size == 0: + return _RandomAccessBytesIO(b'') + # END handle empty chunks gracefully + + try: + return mmap.mmap(-1, size) # read-write by default + except OSError: + # setup real memory instead + # this of course may fail if the amount of memory is not available in + # one chunk - would only be the case in python 2.4, being more likely on + # 32 bit systems. + return _RandomAccessBytesIO(b"\0" * size) + # END handle memory allocation + + +def file_contents_ro(fd, stream=False, allow_mmap=True): + """:return: read-only contents of the file represented by the file descriptor fd + + :param fd: file descriptor opened for reading + :param stream: if False, random access is provided, otherwise the stream interface + is provided. + :param allow_mmap: if True, its allowed to map the contents into memory, which + allows large files to be handled and accessed efficiently. The file-descriptor + will change its position if this is False""" + try: + if allow_mmap: + # supports stream and random access + try: + return mmap.mmap(fd, 0, access=mmap.ACCESS_READ) + except OSError: + # python 2.4 issue, 0 wants to be the actual size + return mmap.mmap(fd, os.fstat(fd).st_size, access=mmap.ACCESS_READ) + # END handle python 2.4 + except OSError: + pass + # END exception handling + + # read manually + contents = os.read(fd, os.fstat(fd).st_size) + if stream: + return _RandomAccessBytesIO(contents) + return contents + + +def file_contents_ro_filepath(filepath, stream=False, allow_mmap=True, flags=0): + """Get the file contents at filepath as fast as possible + + :return: random access compatible memory of the given filepath + :param stream: see ``file_contents_ro`` + :param allow_mmap: see ``file_contents_ro`` + :param flags: additional flags to pass to os.open + :raise OSError: If the file could not be opened + + **Note** for now we don't try to use O_NOATIME directly as the right value needs to be + shared per database in fact. It only makes a real difference for loose object + databases anyway, and they use it with the help of the ``flags`` parameter""" + fd = os.open(filepath, os.O_RDONLY | getattr(os, 'O_BINARY', 0) | flags) + try: + return file_contents_ro(fd, stream, allow_mmap) + finally: + close(fd) + # END assure file is closed + + +def sliding_ro_buffer(filepath, flags=0): + """ + :return: a buffer compatible object which uses our mapped memory manager internally + ready to read the whole given filepath""" + return SlidingWindowMapBuffer(mman.make_cursor(filepath), flags=flags) + + +def to_hex_sha(sha): + """:return: hexified version of sha""" + if len(sha) == 40: + return sha + return bin_to_hex(sha) + + +def to_bin_sha(sha): + if len(sha) == 20: + return sha + return hex_to_bin(sha) + + +#} END routines + + +#{ Utilities + +class LazyMixin: + + """ + Base class providing an interface to lazily retrieve attribute values upon + first access. If slots are used, memory will only be reserved once the attribute + is actually accessed and retrieved the first time. All future accesses will + return the cached value as stored in the Instance's dict or slot. + """ + + __slots__ = tuple() + + def __getattr__(self, attr): + """ + Whenever an attribute is requested that we do not know, we allow it + to be created and set. Next time the same attribute is requested, it is simply + returned from our dict/slots. """ + self._set_cache_(attr) + # will raise in case the cache was not created + return object.__getattribute__(self, attr) + + def _set_cache_(self, attr): + """ + This method should be overridden in the derived class. + It should check whether the attribute named by attr can be created + and cached. Do nothing if you do not know the attribute or call your subclass + + The derived class may create as many additional attributes as it deems + necessary in case a git command returns more information than represented + in the single attribute.""" + pass + + +class LockedFD: + + """ + This class facilitates a safe read and write operation to a file on disk. + If we write to 'file', we obtain a lock file at 'file.lock' and write to + that instead. If we succeed, the lock file will be renamed to overwrite + the original file. + + When reading, we obtain a lock file, but to prevent other writers from + succeeding while we are reading the file. + + This type handles error correctly in that it will assure a consistent state + on destruction. + + **note** with this setup, parallel reading is not possible""" + __slots__ = ("_filepath", '_fd', '_write') + + def __init__(self, filepath): + """Initialize an instance with the givne filepath""" + self._filepath = filepath + self._fd = None + self._write = None # if True, we write a file + + def __del__(self): + # will do nothing if the file descriptor is already closed + if self._fd is not None: + self.rollback() + + def _lockfilepath(self): + return "%s.lock" % self._filepath + + def open(self, write=False, stream=False): + """ + Open the file descriptor for reading or writing, both in binary mode. + + :param write: if True, the file descriptor will be opened for writing. Other + wise it will be opened read-only. + :param stream: if True, the file descriptor will be wrapped into a simple stream + object which supports only reading or writing + :return: fd to read from or write to. It is still maintained by this instance + and must not be closed directly + :raise IOError: if the lock could not be retrieved + :raise OSError: If the actual file could not be opened for reading + + **note** must only be called once""" + if self._write is not None: + raise AssertionError("Called %s multiple times" % self.open) + + self._write = write + + # try to open the lock file + binary = getattr(os, 'O_BINARY', 0) + lockmode = os.O_WRONLY | os.O_CREAT | os.O_EXCL | binary + try: + fd = os.open(self._lockfilepath(), lockmode, int("600", 8)) + if not write: + os.close(fd) + else: + self._fd = fd + # END handle file descriptor + except OSError as e: + raise OSError("Lock at %r could not be obtained" % self._lockfilepath()) from e + # END handle lock retrieval + + # open actual file if required + if self._fd is None: + # we could specify exclusive here, as we obtained the lock anyway + try: + self._fd = os.open(self._filepath, os.O_RDONLY | binary) + except: + # assure we release our lockfile + remove(self._lockfilepath()) + raise + # END handle lockfile + # END open descriptor for reading + + if stream: + # need delayed import + from gitdb.stream import FDStream + return FDStream(self._fd) + else: + return self._fd + # END handle stream + + def commit(self): + """When done writing, call this function to commit your changes into the + actual file. + The file descriptor will be closed, and the lockfile handled. + + **Note** can be called multiple times""" + self._end_writing(successful=True) + + def rollback(self): + """Abort your operation without any changes. The file descriptor will be + closed, and the lock released. + + **Note** can be called multiple times""" + self._end_writing(successful=False) + + def _end_writing(self, successful=True): + """Handle the lock according to the write mode """ + if self._write is None: + raise AssertionError("Cannot end operation if it wasn't started yet") + + if self._fd is None: + return + + os.close(self._fd) + self._fd = None + + lockfile = self._lockfilepath() + if self._write and successful: + # on windows, rename does not silently overwrite the existing one + if sys.platform == "win32": + if isfile(self._filepath): + remove(self._filepath) + # END remove if exists + # END win32 special handling + os.rename(lockfile, self._filepath) + + # assure others can at least read the file - the tmpfile left it at rw-- + # We may also write that file, on windows that boils down to a remove- + # protection as well + chmod(self._filepath, int("644", 8)) + else: + # just delete the file so far, we failed + remove(lockfile) + # END successful handling + +#} END utilities diff --git a/gitdb/gitdb/utils/__init__.py b/gitdb/gitdb/utils/__init__.py new file mode 100644 index 000000000..e69de29bb diff --git a/gitdb/gitdb/utils/encoding.py b/gitdb/gitdb/utils/encoding.py new file mode 100644 index 000000000..b534ef763 --- /dev/null +++ b/gitdb/gitdb/utils/encoding.py @@ -0,0 +1,18 @@ +def force_bytes(data, encoding="utf-8"): + if isinstance(data, bytes): + return data + + if isinstance(data, str): + return data.encode(encoding) + + return data + + +def force_text(data, encoding="utf-8"): + if isinstance(data, str): + return data + + if isinstance(data, bytes): + return data.decode(encoding) + + return str(data, encoding) diff --git a/gitdb/requirements.txt b/gitdb/requirements.txt new file mode 100644 index 000000000..1b2e11db7 --- /dev/null +++ b/gitdb/requirements.txt @@ -0,0 +1 @@ +smmap>=3.0.1,<6 diff --git a/gitdb/setup.py b/gitdb/setup.py new file mode 100755 index 000000000..924f026b8 --- /dev/null +++ b/gitdb/setup.py @@ -0,0 +1,46 @@ +from setuptools import setup + +# NOTE: This is currently duplicated from the gitdb.__init__ module, because +# that's just how you write a setup.py (nobody reads this stuff out of the +# module) + +__author__ = "Sebastian Thiel" +__contact__ = "byronimo@gmail.com" +__homepage__ = "https://github.com/gitpython-developers/GitPython/tree/main/gitdb" +version_info = (4, 0, 12) +__version__ = '.'.join(str(i) for i in version_info) + +setup( + name="gitdb", + version=__version__, + description="Git Object Database", + author=__author__, + author_email=__contact__, + url=__homepage__, + packages=('gitdb', 'gitdb.db', 'gitdb.utils', 'gitdb.test'), + license="BSD License", + zip_safe=False, + install_requires=['smmap>=3.0.1,<6'], + long_description="""GitDB is a pure-Python git object database""", + python_requires='>=3.7', + # See https://pypi.python.org/pypi?%3Aaction=list_classifiers + classifiers=[ + "Development Status :: 5 - Production/Stable", + "Environment :: Console", + "Intended Audience :: Developers", + "Operating System :: OS Independent", + "Operating System :: POSIX", + "Operating System :: Microsoft :: Windows", + "Operating System :: MacOS :: MacOS X", + "Programming Language :: Python", + "Programming Language :: Python :: 3", + "Programming Language :: Python :: 3.7", + "Programming Language :: Python :: 3.8", + "Programming Language :: Python :: 3.9", + "Programming Language :: Python :: 3.10", + "Programming Language :: Python :: 3.11", + "Programming Language :: Python :: 3.12", + "Programming Language :: Python :: 3.13", + "Programming Language :: Python :: 3 :: Only", + ] +) diff --git a/init-tests-after-clone.sh b/init-tests-after-clone.sh index bfada01b0..db3540857 100755 --- a/init-tests-after-clone.sh +++ b/init-tests-after-clone.sh @@ -40,6 +40,11 @@ fi git tag __testing_point__ # The tests need a branch called master. +# +# If master is locally absent but more than one remote has it, checkout fails +# by default even if all remotes agree, and we fall back to creating it at +# HEAD. The reflog we populate below then traces HEAD's history rather than +# a remote master's, but master is reset to __testing_point__ either way. git checkout master -- || git checkout -b master # The tests need a reflog history on the master branch. @@ -50,9 +55,6 @@ git reset --hard HEAD~1 # Point the master branch where we started, so we test the correct code. git reset --hard __testing_point__ -# The tests need submodules, including a submodule with a submodule. -git submodule update --init --recursive - # The tests need some version tags. Try to get them even in forks. This fetches # other objects too. So, locally, we always do it, for a consistent experience. if ! ci || no_version_tags; then diff --git a/pyproject.toml b/pyproject.toml index 090972eed..b7c437bf3 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -3,7 +3,7 @@ requires = ["setuptools"] build-backend = "setuptools.build_meta" [tool.pytest.ini_options] -addopts = "--cov=git --cov-report=term -ra" +addopts = "--cov=git --cov-report=term -rfEX" filterwarnings = "ignore::DeprecationWarning" python_files = "test_*.py" tmp_path_retention_policy = "failed" @@ -13,13 +13,12 @@ testpaths = "test" # Space separated list of paths from root e.g test tests doc # --cov-report term-missing # to terminal with line numbers # --cov-report html:path # html file at path # --maxfail # number of errors before giving up -# -rfE # default test summary: list fail and error +# -rfEX # test summary: list failures, errors, and unexpected passes; omit expected failures # -ra # test summary: list all non-passing (fail, error, skip, xfail, xpass) # --ignore-glob=**/gitdb/* # ignore glob paths # filterwarnings ignore::WarningType # ignores those warnings [tool.mypy] -python_version = "3.8" files = ["git/", "test/deprecation/"] disallow_untyped_defs = true no_implicit_optional = true @@ -29,11 +28,28 @@ warn_unreachable = true implicit_reexport = true # strict = true # TODO: Remove when 'gitdb' is fully annotated. -exclude = ["^git/ext/gitdb"] +exclude = ["^(gitdb|smmap)/"] [[tool.mypy.overrides]] -module = "gitdb.*" +module = ["gitdb", "gitdb.*"] +follow_imports = "skip" ignore_missing_imports = true +[tool.basedpyright] +typeCheckingMode = "standard" +pythonVersion = "3.7" +include = [ + "git", + "test/deprecation", +] +extraPaths = [ + "gitdb", + "smmap", +] +exclude = [ + "gitdb", + "smmap", +] + [tool.coverage.run] source = ["git"] @@ -47,6 +63,8 @@ line-length = 120 # Exclude a variety of commonly ignored directories. exclude = [ "git/ext/", + "gitdb/", + "smmap/", "build", "dist", ] @@ -60,16 +78,14 @@ lint.select = [ # "UP", # See: https://docs.astral.sh/ruff/rules/#pyupgrade-up ] lint.extend-select = [ - # "A", # See: https://pypi.org/project/flake8-builtins - "B", # See: https://pypi.org/project/flake8-bugbear - "C4", # See: https://pypi.org/project/flake8-comprehensions - "TCH004", # See: https://docs.astral.sh/ruff/rules/runtime-import-in-type-checking-block/ + # "A", # See: https://pypi.org/project/flake8-builtins + "B", # See: https://pypi.org/project/flake8-bugbear + "C4", # See: https://pypi.org/project/flake8-comprehensions + "TC004", # See: https://docs.astral.sh/ruff/rules/runtime-import-in-type-checking-block/ ] lint.ignore = [ - "E203", # Whitespace before ':' - "E731", # Do not assign a `lambda` expression, use a `def` + # If it becomes necessary to ignore any rules, list them here. ] -lint.ignore-init-module-imports = true lint.unfixable = [ "F401", # Module imported but unused ] diff --git a/requirements-dev.txt b/requirements-dev.txt index f626644af..066b192b8 100644 --- a/requirements-dev.txt +++ b/requirements-dev.txt @@ -1,8 +1,8 @@ --r requirements.txt --r test-requirements.txt - -# For additional local testing/linting - to be added elsewhere eventually. -ruff -shellcheck -pytest-icdiff -# pytest-profiling +-r requirements.txt +-r test-requirements.txt + +# For additional local testing/linting - to be added elsewhere eventually. +ruff >=0.8 +shellcheck +pytest-icdiff +# pytest-profiling diff --git a/requirements.txt b/requirements.txt index 7159416a9..61d8403b0 100644 --- a/requirements.txt +++ b/requirements.txt @@ -1,2 +1,2 @@ gitdb>=4.0.1,<5 -typing-extensions>=3.7.4.3;python_version<"3.8" +typing-extensions>=3.10.0.2;python_version<"3.10" diff --git a/setup.py b/setup.py index f28fedb85..0b5293685 100755 --- a/setup.py +++ b/setup.py @@ -71,7 +71,7 @@ def _stamp_version(filename: str) -> None: author_email="byronimo@gmail.com, mtrier@gmail.com", license="BSD-3-Clause", url="https://github.com/gitpython-developers/GitPython", - packages=find_packages(exclude=["test", "test.*"]), + packages=find_packages(include=["git", "git.*"]), include_package_data=True, package_dir={"git": "git"}, python_requires=">=3.7", @@ -95,7 +95,6 @@ def _stamp_version(filename: str) -> None: # "Development Status :: 7 - Inactive", "Environment :: Console", "Intended Audience :: Developers", - "License :: OSI Approved :: BSD License", "Operating System :: OS Independent", "Operating System :: POSIX", "Operating System :: Microsoft :: Windows", @@ -109,5 +108,8 @@ def _stamp_version(filename: str) -> None: "Programming Language :: Python :: 3.10", "Programming Language :: Python :: 3.11", "Programming Language :: Python :: 3.12", + "Programming Language :: Python :: 3.13", + "Programming Language :: Python :: 3.14", + "Programming Language :: Python :: 3.15", ], ) diff --git a/smmap/.coveragerc b/smmap/.coveragerc new file mode 100644 index 000000000..e61d27ba9 --- /dev/null +++ b/smmap/.coveragerc @@ -0,0 +1,10 @@ +[run] +source = smmap + +; to make nosetests happy +[report] +omit = + */yaml* + */tests/* + */python?.?/* + */site-packages/nose/* \ No newline at end of file diff --git a/smmap/.gitignore b/smmap/.gitignore new file mode 100644 index 000000000..85cc74fdf --- /dev/null +++ b/smmap/.gitignore @@ -0,0 +1,12 @@ +*.pyc +build/ +.coverage +coverage +cover/ +dist/ +MANIFEST +.tox +*.egg-info +.noseids +*.sublime-workspace +/env/ diff --git a/smmap/LICENSE b/smmap/LICENSE new file mode 100644 index 000000000..710010f1f --- /dev/null +++ b/smmap/LICENSE @@ -0,0 +1,30 @@ +Copyright (C) 2010, 2011 Sebastian Thiel and contributors +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions +are met: + +* Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + +* Redistributions in binary form must reproduce the above copyright +notice, this list of conditions and the following disclaimer in the +documentation and/or other materials provided with the distribution. + +* Neither the name of the async project nor the names of +its contributors may be used to endorse or promote products derived +from this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED +TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR +PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF +LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING +NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + diff --git a/smmap/MANIFEST.in b/smmap/MANIFEST.in new file mode 100644 index 000000000..399a207e6 --- /dev/null +++ b/smmap/MANIFEST.in @@ -0,0 +1,2 @@ +# Include the license file +include LICENSE diff --git a/smmap/Makefile b/smmap/Makefile new file mode 100644 index 000000000..ad0912ffe --- /dev/null +++ b/smmap/Makefile @@ -0,0 +1,7 @@ +.PHONY: all clean + +all: + @grep -Ee '^[a-z].*:' Makefile | cut -d: -f1 | grep -vF all + +clean: + rm -rf build/ dist/ .eggs/ .tox/ diff --git a/smmap/README.md b/smmap/README.md new file mode 100644 index 000000000..748f0f0db --- /dev/null +++ b/smmap/README.md @@ -0,0 +1,82 @@ +## Motivation + +When reading from many possibly large files in a fashion similar to random access, it is usually the fastest and most efficient to use memory maps. + +Although memory maps have many advantages, they represent a very limited system resource as every map uses one file descriptor, whose amount is limited per process. On 32 bit systems, the amount of memory you can have mapped at a time is naturally limited to theoretical 4GB of memory, which may not be enough for some applications. + + +## Limitations + +* **System resources (file-handles) are likely to be leaked!** This is due to the library authors reliance on a deterministic `__del__()` destructor. +* The memory access is read-only by design. + + +## Overview + +![Dependency packages](https://github.com/gitpython-developers/GitPython/actions/workflows/dependencies.yml/badge.svg) + +Smmap wraps an interface around mmap and tracks the mapped files as well as the amount of clients who use it. If the system runs out of resources, or if a memory limit is reached, it will automatically unload unused maps to allow continued operation. + +To allow processing large files even on 32 bit systems, it allows only portions of the file to be mapped. Once the user reads beyond the mapped region, smmap will automatically map the next required region, unloading unused regions using a LRU algorithm. + +Although the library can be used most efficiently with its native interface, a Buffer implementation is provided to hide these details behind a simple string-like interface. + +For performance critical 64 bit applications, a simplified version of memory mapping is provided which always maps the whole file, but still provides the benefit of unloading unused mappings on demand. + + + +## Prerequisites + +* Python 3.7+ +* OSX, Windows or Linux + +The package was tested on all of the previously mentioned configurations. + +## Installing smmap + +[![Documentation Status](https://readthedocs.org/projects/smmap/badge/?version=latest)](https://readthedocs.org/projects/smmap/?badge=latest) + +Its easiest to install smmap using the [pip](http://www.pip-installer.org/en/latest) program: + +```bash +$ pip install smmap +``` + +As the command will install smmap in your respective python distribution, you will most likely need root permissions to authorize the required changes. + +If you have downloaded the source archive, the package can be installed by running the `setup.py` script: + +```bash +$ python setup.py install +``` + +It is advised to have a look at the **Usage Guide** for a brief introduction on the different database implementations. + + + +## Homepage and Links + +The project is maintained in the GitPython repository at https://github.com/gitpython-developers/GitPython/tree/main/smmap . + +The latest source can be cloned from github as well: + +* https://github.com/gitpython-developers/GitPython.git + + +For support, please use the git-python mailing list: + +* http://groups.google.com/group/git-python + + +Issues can be filed on github: + +* https://github.com/gitpython-developers/GitPython/issues + +A link to the pypi page related to this repository: + +* https://pypi.org/project/smmap/ + + +## License Information + +*smmap* is licensed under the New BSD License. diff --git a/smmap/SECURITY.md b/smmap/SECURITY.md new file mode 100644 index 000000000..9e0c0d16a --- /dev/null +++ b/smmap/SECURITY.md @@ -0,0 +1,3 @@ +# Security Policy + +See [GitPython](https://github.com/gitpython-developers/GitPython/blob/main/SECURITY.md). Vulnerabilities found in `smmap` can be reported there. diff --git a/smmap/doc/.gitignore b/smmap/doc/.gitignore new file mode 100644 index 000000000..32060acdd --- /dev/null +++ b/smmap/doc/.gitignore @@ -0,0 +1,2 @@ +build +*.version_info diff --git a/smmap/doc/Makefile b/smmap/doc/Makefile new file mode 100644 index 000000000..675ec2094 --- /dev/null +++ b/smmap/doc/Makefile @@ -0,0 +1,89 @@ +# Makefile for Sphinx documentation +# + +# You can set these variables from the command line. +SPHINXOPTS = +SPHINXBUILD = sphinx-build +PAPER = +BUILDDIR = build + +# Internal variables. +PAPEROPT_a4 = -D latex_paper_size=a4 +PAPEROPT_letter = -D latex_paper_size=letter +ALLSPHINXOPTS = -d $(BUILDDIR)/doctrees $(PAPEROPT_$(PAPER)) $(SPHINXOPTS) source + +.PHONY: help clean html dirhtml pickle json htmlhelp qthelp latex changes linkcheck doctest + +help: + @echo "Please use \`make ' where is one of" + @echo " html to make standalone HTML files" + @echo " dirhtml to make HTML files named index.html in directories" + @echo " pickle to make pickle files" + @echo " json to make JSON files" + @echo " htmlhelp to make HTML files and a HTML help project" + @echo " qthelp to make HTML files and a qthelp project" + @echo " latex to make LaTeX files, you can set PAPER=a4 or PAPER=letter" + @echo " changes to make an overview of all changed/added/deprecated items" + @echo " linkcheck to check all external links for integrity" + @echo " doctest to run all doctests embedded in the documentation (if enabled)" + +clean: + -rm -rf $(BUILDDIR)/* + +html: + $(SPHINXBUILD) -b html $(ALLSPHINXOPTS) $(BUILDDIR)/html + @echo + @echo "Build finished. The HTML pages are in $(BUILDDIR)/html." + +dirhtml: + $(SPHINXBUILD) -b dirhtml $(ALLSPHINXOPTS) $(BUILDDIR)/dirhtml + @echo + @echo "Build finished. The HTML pages are in $(BUILDDIR)/dirhtml." + +pickle: + $(SPHINXBUILD) -b pickle $(ALLSPHINXOPTS) $(BUILDDIR)/pickle + @echo + @echo "Build finished; now you can process the pickle files." + +json: + $(SPHINXBUILD) -b json $(ALLSPHINXOPTS) $(BUILDDIR)/json + @echo + @echo "Build finished; now you can process the JSON files." + +htmlhelp: + $(SPHINXBUILD) -b htmlhelp $(ALLSPHINXOPTS) $(BUILDDIR)/htmlhelp + @echo + @echo "Build finished; now you can run HTML Help Workshop with the" \ + ".hhp project file in $(BUILDDIR)/htmlhelp." + +qthelp: + $(SPHINXBUILD) -b qthelp $(ALLSPHINXOPTS) $(BUILDDIR)/qthelp + @echo + @echo "Build finished; now you can run "qcollectiongenerator" with the" \ + ".qhcp project file in $(BUILDDIR)/qthelp, like this:" + @echo "# qcollectiongenerator $(BUILDDIR)/qthelp/smmap.qhcp" + @echo "To view the help file:" + @echo "# assistant -collectionFile $(BUILDDIR)/qthelp/smmap.qhc" + +latex: + $(SPHINXBUILD) -b latex $(ALLSPHINXOPTS) $(BUILDDIR)/latex + @echo + @echo "Build finished; the LaTeX files are in $(BUILDDIR)/latex." + @echo "Run \`make all-pdf' or \`make all-ps' in that directory to" \ + "run these through (pdf)latex." + +changes: + $(SPHINXBUILD) -b changes $(ALLSPHINXOPTS) $(BUILDDIR)/changes + @echo + @echo "The overview file is in $(BUILDDIR)/changes." + +linkcheck: + $(SPHINXBUILD) -b linkcheck $(ALLSPHINXOPTS) $(BUILDDIR)/linkcheck + @echo + @echo "Link check complete; look for any errors in the above output " \ + "or in $(BUILDDIR)/linkcheck/output.txt." + +doctest: + $(SPHINXBUILD) -b doctest $(ALLSPHINXOPTS) $(BUILDDIR)/doctest + @echo "Testing of doctests in the sources finished, look at the " \ + "results in $(BUILDDIR)/doctest/output.txt." diff --git a/smmap/doc/make.bat b/smmap/doc/make.bat new file mode 100644 index 000000000..6900a2a46 --- /dev/null +++ b/smmap/doc/make.bat @@ -0,0 +1,113 @@ +@ECHO OFF + +REM Command file for Sphinx documentation + +set SPHINXBUILD=sphinx-build +set BUILDDIR=build +set ALLSPHINXOPTS=-d %BUILDDIR%/doctrees %SPHINXOPTS% source +if NOT "%PAPER%" == "" ( + set ALLSPHINXOPTS=-D latex_paper_size=%PAPER% %ALLSPHINXOPTS% +) + +if "%1" == "" goto help + +if "%1" == "help" ( + :help + echo.Please use `make ^` where ^ is one of + echo. html to make standalone HTML files + echo. dirhtml to make HTML files named index.html in directories + echo. pickle to make pickle files + echo. json to make JSON files + echo. htmlhelp to make HTML files and a HTML help project + echo. qthelp to make HTML files and a qthelp project + echo. latex to make LaTeX files, you can set PAPER=a4 or PAPER=letter + echo. changes to make an overview over all changed/added/deprecated items + echo. linkcheck to check all external links for integrity + echo. doctest to run all doctests embedded in the documentation if enabled + goto end +) + +if "%1" == "clean" ( + for /d %%i in (%BUILDDIR%\*) do rmdir /q /s %%i + del /q /s %BUILDDIR%\* + goto end +) + +if "%1" == "html" ( + %SPHINXBUILD% -b html %ALLSPHINXOPTS% %BUILDDIR%/html + echo. + echo.Build finished. The HTML pages are in %BUILDDIR%/html. + goto end +) + +if "%1" == "dirhtml" ( + %SPHINXBUILD% -b dirhtml %ALLSPHINXOPTS% %BUILDDIR%/dirhtml + echo. + echo.Build finished. The HTML pages are in %BUILDDIR%/dirhtml. + goto end +) + +if "%1" == "pickle" ( + %SPHINXBUILD% -b pickle %ALLSPHINXOPTS% %BUILDDIR%/pickle + echo. + echo.Build finished; now you can process the pickle files. + goto end +) + +if "%1" == "json" ( + %SPHINXBUILD% -b json %ALLSPHINXOPTS% %BUILDDIR%/json + echo. + echo.Build finished; now you can process the JSON files. + goto end +) + +if "%1" == "htmlhelp" ( + %SPHINXBUILD% -b htmlhelp %ALLSPHINXOPTS% %BUILDDIR%/htmlhelp + echo. + echo.Build finished; now you can run HTML Help Workshop with the ^ +.hhp project file in %BUILDDIR%/htmlhelp. + goto end +) + +if "%1" == "qthelp" ( + %SPHINXBUILD% -b qthelp %ALLSPHINXOPTS% %BUILDDIR%/qthelp + echo. + echo.Build finished; now you can run "qcollectiongenerator" with the ^ +.qhcp project file in %BUILDDIR%/qthelp, like this: + echo.^> qcollectiongenerator %BUILDDIR%\qthelp\smmap.qhcp + echo.To view the help file: + echo.^> assistant -collectionFile %BUILDDIR%\qthelp\smmap.ghc + goto end +) + +if "%1" == "latex" ( + %SPHINXBUILD% -b latex %ALLSPHINXOPTS% %BUILDDIR%/latex + echo. + echo.Build finished; the LaTeX files are in %BUILDDIR%/latex. + goto end +) + +if "%1" == "changes" ( + %SPHINXBUILD% -b changes %ALLSPHINXOPTS% %BUILDDIR%/changes + echo. + echo.The overview file is in %BUILDDIR%/changes. + goto end +) + +if "%1" == "linkcheck" ( + %SPHINXBUILD% -b linkcheck %ALLSPHINXOPTS% %BUILDDIR%/linkcheck + echo. + echo.Link check complete; look for any errors in the above output ^ +or in %BUILDDIR%/linkcheck/output.txt. + goto end +) + +if "%1" == "doctest" ( + %SPHINXBUILD% -b doctest %ALLSPHINXOPTS% %BUILDDIR%/doctest + echo. + echo.Testing of doctests in the sources finished, look at the ^ +results in %BUILDDIR%/doctest/output.txt. + goto end +) + +:end diff --git a/smmap/doc/source/api.rst b/smmap/doc/source/api.rst new file mode 100644 index 000000000..2e2dac41a --- /dev/null +++ b/smmap/doc/source/api.rst @@ -0,0 +1,34 @@ +.. _api-label: + +############# +API Reference +############# + +*********************** +Mapped Memory Managers +*********************** + +.. automodule:: smmap.mman + :members: + :undoc-members: + +******* +Buffers +******* + +.. automodule:: smmap.buf + :members: + :undoc-members: + +********* +Utilities +********* + +.. automodule:: smmap.util + :members: + :undoc-members: + + + + + diff --git a/smmap/doc/source/changes.rst b/smmap/doc/source/changes.rst new file mode 100644 index 000000000..faed6085d --- /dev/null +++ b/smmap/doc/source/changes.rst @@ -0,0 +1,108 @@ +######### +Changelog +######### + +****** +v5.0.3 +****** + +- declare support for Python 3.13 + +For more, see https://github.com/gitpython-developers/smmap/compare/v5.0.3...v5.0.2 + +****** +v5.0.2 +****** + +- remove a usage of mktemp + +****** +v5.0.1 +****** + +- Added support for Python 3.12 + +****** +v6.0.0 +****** + +YANKED + +- Dropped support 3.6 and 3.7 +- Declared support for Python 3.11 and 3.12 + +****** +v5.0.0 +****** + +- Dropped support 3.5 +- Added support for Python 3.10 + +****** +v4.0.0 +****** + +- Dropped support for Python 2.7 and 3.4 +- Added support for Python 3.7, 3.8, and 3.9 +- Removed unused exc.MemoryManagerError and exc.RegionCollectionError + +****** +v3.0.5 +****** + +- Restored Python 2 support removed in v3.0.2 +- Changed release signature key to 27C50E7F590947D7273A741E85194C08421980C9. + See https://keybase.io/byronbates for proof of ownership. + +****** +v3.0.4 +****** + +- Signed release (with correct key this time) + +****** +v3.0.2 +****** + +- Signed release +- Switched to GitHub Actions for CI + +****** +v3.0.1 +****** +- Switched back to the smmap package name on PyPI and fixed the smmap2 mirror package + (`#44 `_) +- Fixed setup.py ``long_description`` rendering + (`#40 `_) + +********** +v0.9.0 +********** +- Fixed issue with resources never being freed as mmaps were never closed. +- Client counting is now done manually, instead of relying on pyton's reference count + +********** +v0.8.5 +********** +- Fixed Python 3.0-3.3 regression, which also causes smmap to become about 3 times slower depending on the code path. It's related to this bug (http://bugs.python.org/issue15958), which was fixed in python 3.4 + +********** +v0.8.4 +********** +- Fixed Python 3 performance regression + +********** +v0.8.3 +********** +- Cleaned up code and assured it works sufficiently well with python 3 + +********** +v0.8.1 +********** +- A single bugfix + +********** +v0.8.0 +********** + +- Initial Release diff --git a/smmap/doc/source/conf.py b/smmap/doc/source/conf.py new file mode 100644 index 000000000..55dfc5c46 --- /dev/null +++ b/smmap/doc/source/conf.py @@ -0,0 +1,194 @@ +# +# smmap documentation build configuration file, created by +# sphinx-quickstart on Wed Jun 8 15:14:25 2011. +# +# This file is execfile()d with the current directory set to its containing dir. +# +# Note that not all possible configuration values are present in this +# autogenerated file. +# +# All configuration values have a default; values that are commented out +# serve to show the default. + +import sys +import os + +# If extensions (or modules to document with autodoc) are in another directory, +# add these directories to sys.path here. If the directory is relative to the +# documentation root, use os.path.abspath to make it absolute, like shown here. +sys.path.append(os.path.abspath('../../')) + +# -- General configuration ----------------------------------------------------- + +# Add any Sphinx extension module names here, as strings. They can be extensions +# coming with Sphinx (named 'sphinx.ext.*') or your custom ones. +extensions = ['sphinx.ext.autodoc', 'sphinx.ext.todo'] + +# Add any paths that contain templates here, relative to this directory. +templates_path = ['.templates'] + +# The suffix of source filenames. +source_suffix = '.rst' + +# The encoding of source files. +#source_encoding = 'utf-8' + +# The master toctree document. +master_doc = 'index' + +# General information about the project. +project = 'smmap' +copyright = '2011, Sebastian Thiel' + +# The version info for the project you're documenting, acts as replacement for +# |version| and |release|, also used in various other places throughout the +# built documents. +# +# The short X.Y version. +version = '0.8.0' +# The full version, including alpha/beta/rc tags. +release = '0.8.0' + +# The language for content autogenerated by Sphinx. Refer to documentation +# for a list of supported languages. +#language = None + +# There are two options for replacing |today|: either, you set today to some +# non-false value, then it is used: +#today = '' +# Else, today_fmt is used as the format for a strftime call. +#today_fmt = '%B %d, %Y' + +# List of documents that shouldn't be included in the build. +#unused_docs = [] + +# List of directories, relative to source directory, that shouldn't be searched +# for source files. +exclude_trees = [] + +# The reST default role (used for this markup: `text`) to use for all documents. +#default_role = None + +# If true, '()' will be appended to :func: etc. cross-reference text. +#add_function_parentheses = True + +# If true, the current module name will be prepended to all description +# unit titles (such as .. function::). +#add_module_names = True + +# If true, sectionauthor and moduleauthor directives will be shown in the +# output. They are ignored by default. +#show_authors = False + +# The name of the Pygments (syntax highlighting) style to use. +pygments_style = 'sphinx' + +# A list of ignored prefixes for module index sorting. +#modindex_common_prefix = [] + + +# -- Options for HTML output --------------------------------------------------- + +# The theme to use for HTML and HTML Help pages. Major themes that come with +# Sphinx are currently 'default' and 'sphinxdoc'. +html_theme = 'default' + +# Theme options are theme-specific and customize the look and feel of a theme +# further. For a list of options available for each theme, see the +# documentation. +#html_theme_options = {} + +# Add any paths that contain custom themes here, relative to this directory. +#html_theme_path = [] + +# The name for this set of Sphinx documents. If None, it defaults to +# " v documentation". +#html_title = None + +# A shorter title for the navigation bar. Default is the same as html_title. +#html_short_title = None + +# The name of an image file (relative to this directory) to place at the top +# of the sidebar. +#html_logo = None + +# The name of an image file (within the static path) to use as favicon of the +# docs. This file should be a Windows icon file (.ico) being 16x16 or 32x32 +# pixels large. +#html_favicon = None + +# Add any paths that contain custom static files (such as style sheets) here, +# relative to this directory. They are copied after the builtin static files, +# so a file named "default.css" will overwrite the builtin "default.css". +html_static_path = ['.static'] + +# If not '', a 'Last updated on:' timestamp is inserted at every page bottom, +# using the given strftime format. +#html_last_updated_fmt = '%b %d, %Y' + +# If true, SmartyPants will be used to convert quotes and dashes to +# typographically correct entities. +#html_use_smartypants = True + +# Custom sidebar templates, maps document names to template names. +#html_sidebars = {} + +# Additional templates that should be rendered to pages, maps page names to +# template names. +#html_additional_pages = {} + +# If false, no module index is generated. +#html_use_modindex = True + +# If false, no index is generated. +#html_use_index = True + +# If true, the index is split into individual pages for each letter. +#html_split_index = False + +# If true, links to the reST sources are added to the pages. +#html_show_sourcelink = True + +# If true, an OpenSearch description file will be output, and all pages will +# contain a tag referring to it. The value of this option must be the +# base URL from which the finished HTML is served. +#html_use_opensearch = '' + +# If nonempty, this is the file name suffix for HTML files (e.g. ".xhtml"). +#html_file_suffix = '' + +# Output file base name for HTML help builder. +htmlhelp_basename = 'smmapdoc' + + +# -- Options for LaTeX output -------------------------------------------------- + +# The paper size ('letter' or 'a4'). +#latex_paper_size = 'letter' + +# The font size ('10pt', '11pt' or '12pt'). +#latex_font_size = '10pt' + +# Grouping the document tree into LaTeX files. List of tuples +# (source start file, target name, title, author, documentclass [howto/manual]). +latex_documents = [ + ('index', 'smmap.tex', 'smmap Documentation', + 'Sebastian Thiel', 'manual'), +] + +# The name of an image file (relative to this directory) to place at the top of +# the title page. +#latex_logo = None + +# For "manual" documents, if this is true, then toplevel headings are parts, +# not chapters. +#latex_use_parts = False + +# Additional stuff for the LaTeX preamble. +#latex_preamble = '' + +# Documents to append as an appendix to all manuals. +#latex_appendices = [] + +# If false, no module index is generated. +#latex_use_modindex = True diff --git a/smmap/doc/source/index.rst b/smmap/doc/source/index.rst new file mode 100644 index 000000000..d25ef8244 --- /dev/null +++ b/smmap/doc/source/index.rst @@ -0,0 +1,26 @@ +.. smmap documentation master file, created by + sphinx-quickstart on Wed Jun 8 15:14:25 2011. + You can adapt this file completely to your liking, but it should at least + contain the root `toctree` directive. + +Welcome to smmap's documentation! +================================= +**smmap** is a pure python implementation of a sliding memory map to help unifying memory mapped access on 32 and 64 bit systems and to help managing resources more efficiently. + +Contents: + +.. toctree:: + :maxdepth: 2 + + intro + tutorial + api + changes + +Indices and tables +================== + +* :ref:`genindex` +* :ref:`modindex` +* :ref:`search` + diff --git a/smmap/doc/source/intro.rst b/smmap/doc/source/intro.rst new file mode 100644 index 000000000..109fec247 --- /dev/null +++ b/smmap/doc/source/intro.rst @@ -0,0 +1,72 @@ +########### +Motivation +########### +When reading from many possibly large files in a fashion similar to random access, it is usually the fastest and most efficient to use memory maps. + +Although memory maps have many advantages, they represent a very limited system resource as every map uses one file descriptor, whose amount is limited per process. On 32 bit systems, the amount of memory you can have mapped at a time is naturally limited to theoretical 4GB of memory, which may not be enough for some applications. + +######## +Overview +######## + +Smmap wraps an interface around mmap and tracks the mapped files as well as the amount of clients who use it. If the system runs out of resources, or if a memory limit is reached, it will automatically unload unused maps to allow continued operation. + +To allow processing large files even on 32 bit systems, it allows only portions of the file to be mapped. Once the user reads beyond the mapped region, smmap will automatically map the next required region, unloading unused regions using a LRU algorithm. + +Although the library can be used most efficiently with its native interface, a Buffer implementation is provided to hide these details behind a simple string-like interface. + +For performance critical 64 bit applications, a simplified version of memory mapping is provided which always maps the whole file, but still provides the benefit of unloading unused mappings on demand. + +############# +Prerequisites +############# +* Python 3.7+ +* OSX, Windows or Linux + +The package was tested on all of the previously mentioned configurations. + +########### +Limitations +########### +* The memory access is read-only by design. + +################ +Installing smmap +################ +Its easiest to install smmap using the *pip* program:: + + $ pip install smmap + +As the command will install smmap in your respective python distribution, you will most likely need root permissions to authorize the required changes. + +If you have downloaded the source archive, the package can be installed by running the ``setup.py`` script:: + + $ python setup.py install + +It is advised to have a look at the :ref:`Usage Guide ` for a brief introduction on the different database implementations. + +################## +Homepage and Links +################## +The project is home on github at `https://github.com/gitpython-developers/smmap `_. + +The latest source can be cloned from github as well: + + * git://github.com/gitpython-developers/smmap.git + + +For support, please use the git-python mailing list: + + * http://groups.google.com/group/git-python + + +Issues can be filed on github: + + * https://github.com/gitpython-developers/smmap/issues + +################### +License Information +################### +*smmap* is licensed under the New BSD License. + +.. _pip: http://www.pip-installer.org/en/latest/ diff --git a/smmap/doc/source/tutorial.rst b/smmap/doc/source/tutorial.rst new file mode 100644 index 000000000..917b24594 --- /dev/null +++ b/smmap/doc/source/tutorial.rst @@ -0,0 +1,118 @@ +.. _tutorial-label: + +########### +Usage Guide +########### +This text briefly introduces you to the basic design decisions and accompanying classes. + +****** +Design +****** +Per application, there is *MemoryManager* which is held as static instance and used throughout the application. It can be configured to keep your resources within certain limits. + +To access mapped regions, you require a cursor. Cursors point to exactly one file and serve as handles into it. As long as it exists, the respective memory region will remain available. + +For convenience, a buffer implementation is provided which handles cursors and resource allocation behind its simple buffer like interface. + +*************** +Memory Managers +*************** +There are two types of memory managers, one uses *static* windows, the other one uses *sliding* windows. A window is a region of a file mapped into memory. Although the names might be somewhat misleading as technically windows are always static, the *sliding* version will allocate relatively small windows whereas the *static* version will always map the whole file. + +The *static* manager does nothing more than keeping a client count on the respective memory maps which always map the whole file, which allows to make some assumptions that can lead to simplified data access and increased performance, but reduces the compatibility to 32 bit systems or giant files. + +The *sliding* memory manager therefore should be the default manager when preparing an application for handling huge amounts of data on 32 bit and 64 bit platforms:: + + import smmap + # This instance should be globally available in your application + # It is configured to be well suitable for 32-bit or 64 bit applications. + mman = smmap.SlidingWindowMapManager() + + # the manager provides much useful information about its current state + # like the amount of open file handles or the amount of mapped memory + mman.num_file_handles() + mman.mapped_memory_size() + # and many more ... + + +Cursors +******* +*Cursors* are handles that point onto a window, i.e. a region of a file mapped into memory. From them you may obtain a buffer through which the data of that window can actually be accessed:: + + import smmap.test.lib + fc = smmap.test.lib.FileCreator(1024*1024*8, "test_file") + + # obtain a cursor to access some file. + c = mman.make_cursor(fc.path) + + # the cursor is now associated with the file, but not yet usable + assert c.is_associated() + assert not c.is_valid() + + # before you can use the cursor, you have to specify a window you want to + # access. The following just says you want as much data as possible starting + # from offset 0. + # To be sure your region could be mapped, query for validity + assert c.use_region().is_valid() # use_region returns self + + # once a region was mapped, you must query its dimension regularly + # to assure you don't try to access its buffer out of its bounds + assert c.size() + c.buffer()[0] # first byte + c.buffer()[1:10] # first 9 bytes + c.buffer()[c.size()-1] # last byte + + # its recommended not to create big slices when feeding the buffer + # into consumers (e.g. struct or zlib). + # Instead, either give the buffer directly, or use pythons buffer command. + buffer(c.buffer(), 1, 9) # first 9 bytes without copying them + + # you can query absolute offsets, and check whether an offset is included + # in the cursor's data. + assert c.ofs_begin() < c.ofs_end() + assert c.includes_ofs(100) + + # If you are over out of bounds with one of your region requests, the + # cursor will be come invalid. It cannot be used in that state + assert not c.use_region(fc.size, 100).is_valid() + # map as much as possible after skipping the first 100 bytes + assert c.use_region(100).is_valid() + + # You can explicitly free cursor resources by unusing the cursor's region + c.unuse_region() + assert not c.is_valid() + + +Now you would have to write your algorithms around this interface to properly slide through huge amounts of data. + +Alternatively you can use a convenience interface. + +******* +Buffers +******* +To make first use easier, at the expense of performance, there is a Buffer implementation which uses a cursor underneath. + +With it, you can access all data in a possibly huge file without having to take care of setting the cursor to different regions yourself:: + + # Create a default buffer which can operate on the whole file + buf = smmap.SlidingWindowMapBuffer(mman.make_cursor(fc.path)) + + # you can use it right away + assert buf.cursor().is_valid() + + buf[0] # access the first byte + buf[-1] # access the last ten bytes on the file + buf[-10:]# access the last ten bytes + + # If you want to keep the instance between different accesses, use the + # dedicated methods + buf.end_access() + assert not buf.cursor().is_valid() # you cannot use the buffer anymore + assert buf.begin_access(offset=10) # start using the buffer at an offset + + # it will stop using resources automatically once it goes out of scope + +Disadvantages +************* +Buffers cannot be used in place of strings or maps, hence you have to slice them to have valid input for the sorts of struct and zlib. A slice means a lot of data handling overhead which makes buffers slower compared to using cursors directly. + diff --git a/smmap/setup.cfg b/smmap/setup.cfg new file mode 100644 index 000000000..38c535ba5 --- /dev/null +++ b/smmap/setup.cfg @@ -0,0 +1,2 @@ +[flake8] +exclude = .tox,.venv,build,dist,doc diff --git a/smmap/setup.py b/smmap/setup.py new file mode 100755 index 000000000..585f0e76c --- /dev/null +++ b/smmap/setup.py @@ -0,0 +1,52 @@ +#!/usr/bin/env python +import os + +try: + from setuptools import setup, find_packages +except ImportError: + from ez_setup import use_setuptools + use_setuptools() + from setuptools import setup, find_packages + +import smmap + +if os.path.exists("README.md"): + with open("README.md", encoding="utf-8") as readme: + long_description = readme.read().replace("\r\n", "\n") +else: + long_description = "See https://github.com/gitpython-developers/GitPython/tree/main/smmap" + +setup( + name="smmap", + version=smmap.__version__, + description="A pure Python implementation of a sliding window memory map manager", + author=smmap.__author__, + author_email=smmap.__contact__, + url=smmap.__homepage__, + platforms=["any"], + license="BSD-3-Clause", + packages=find_packages(), + zip_safe=True, + python_requires=">=3.7", + classifiers=[ + "Development Status :: 5 - Production/Stable", + "Environment :: Console", + "Intended Audience :: Developers", + "Operating System :: OS Independent", + "Operating System :: POSIX", + "Operating System :: Microsoft :: Windows", + "Operating System :: MacOS :: MacOS X", + "Programming Language :: Python", + "Programming Language :: Python :: 3", + "Programming Language :: Python :: 3.7", + "Programming Language :: Python :: 3.8", + "Programming Language :: Python :: 3.9", + "Programming Language :: Python :: 3.10", + "Programming Language :: Python :: 3.11", + "Programming Language :: Python :: 3.12", + "Programming Language :: Python :: 3.13", + "Programming Language :: Python :: 3 :: Only", + ], + long_description=long_description, + long_description_content_type='text/markdown', +) diff --git a/smmap/smmap/__init__.py b/smmap/smmap/__init__.py new file mode 100644 index 000000000..79ec0674d --- /dev/null +++ b/smmap/smmap/__init__.py @@ -0,0 +1,11 @@ +"""Initialize the smmap package""" + +__author__ = "Sebastian Thiel" +__contact__ = "byronimo@gmail.com" +__homepage__ = "https://github.com/gitpython-developers/GitPython/tree/main/smmap" +version_info = (5, 0, 3) +__version__ = '.'.join(str(i) for i in version_info) + +# make everything available in root package for convenience +from .mman import * +from .buf import * diff --git a/smmap/smmap/buf.py b/smmap/smmap/buf.py new file mode 100644 index 000000000..731b0644b --- /dev/null +++ b/smmap/smmap/buf.py @@ -0,0 +1,143 @@ +"""Module with a simple buffer implementation using the memory manager""" +import sys + +__all__ = ["SlidingWindowMapBuffer"] + + +class SlidingWindowMapBuffer: + + """A buffer like object which allows direct byte-wise object and slicing into + memory of a mapped file. The mapping is controlled by the provided cursor. + + The buffer is relative, that is if you map an offset, index 0 will map to the + first byte at the offset you used during initialization or begin_access + + **Note:** Although this type effectively hides the fact that there are mapped windows + underneath, it can unfortunately not be used in any non-pure python method which + needs a buffer or string""" + __slots__ = ( + '_c', # our cursor + '_size', # our supposed size + ) + + def __init__(self, cursor=None, offset=0, size=sys.maxsize, flags=0): + """Initialize the instance to operate on the given cursor. + :param cursor: if not None, the associated cursor to the file you want to access + If None, you have call begin_access before using the buffer and provide a cursor + :param offset: absolute offset in bytes + :param size: the total size of the mapping. Defaults to the maximum possible size + From that point on, the __len__ of the buffer will be the given size or the file size. + If the size is larger than the mappable area, you can only access the actually available + area, although the length of the buffer is reported to be your given size. + Hence it is in your own interest to provide a proper size ! + :param flags: Additional flags to be passed to os.open + :raise ValueError: if the buffer could not achieve a valid state""" + self._c = cursor + if cursor and not self.begin_access(cursor, offset, size, flags): + raise ValueError("Failed to allocate the buffer - probably the given offset is out of bounds") + # END handle offset + + def __del__(self): + self.end_access() + + def __enter__(self): + return self + + def __exit__(self, exc_type, exc_value, traceback): + self.end_access() + + def __len__(self): + return self._size + + def __getitem__(self, i): + if isinstance(i, slice): + return self.__getslice__(i.start or 0, i.stop or self._size) + c = self._c + assert c.is_valid() + if i < 0: + i = self._size + i + if not c.includes_ofs(i): + c.use_region(i, 1) + # END handle region usage + return c.buffer()[i - c.ofs_begin()] + + def __getslice__(self, i, j): + c = self._c + # fast path, slice fully included - safes a concatenate operation and + # should be the default + assert c.is_valid() + if i < 0: + i = self._size + i + if j == sys.maxsize: + j = self._size + if j < 0: + j = self._size + j + if (c.ofs_begin() <= i) and (j < c.ofs_end()): + b = c.ofs_begin() + return c.buffer()[i - b:j - b] + else: + l = j - i # total length + ofs = i + # It's fastest to keep tokens and join later, especially in py3, which was 7 times slower + # in the previous iteration of this code + md = list() + while l: + c.use_region(ofs, l) + assert c.is_valid() + d = c.buffer()[:l] + ofs += len(d) + l -= len(d) + # Make sure we don't keep references, as c.use_region() might attempt to free resources, but + # can't unless we use pure bytes + if hasattr(d, 'tobytes'): + d = d.tobytes() + md.append(d) + # END while there are bytes to read + return b''.join(md) + # END fast or slow path + #{ Interface + + def begin_access(self, cursor=None, offset=0, size=sys.maxsize, flags=0): + """Call this before the first use of this instance. The method was already + called by the constructor in case sufficient information was provided. + + For more information no the parameters, see the __init__ method + :param path: if cursor is None the existing one will be used. + :return: True if the buffer can be used""" + if cursor: + self._c = cursor + # END update our cursor + + # reuse existing cursors if possible + if self._c is not None and self._c.is_associated(): + res = self._c.use_region(offset, size, flags).is_valid() + if res: + # if given size is too large or default, we computer a proper size + # If its smaller, we assume the combination between offset and size + # as chosen by the user is correct and use it ! + # If not, the user is in trouble. + if size > self._c.file_size(): + size = self._c.file_size() - offset + # END handle size + self._size = size + # END set size + return res + # END use our cursor + return False + + def end_access(self): + """Call this method once you are done using the instance. It is automatically + called on destruction, and should be called just in time to allow system + resources to be freed. + + Once you called end_access, you must call begin access before reusing this instance!""" + self._size = 0 + if self._c is not None: + self._c.unuse_region() + # END unuse region + + def cursor(self): + """:return: the currently set cursor which provides access to the data""" + return self._c + + #}END interface diff --git a/smmap/smmap/mman.py b/smmap/smmap/mman.py new file mode 100644 index 000000000..873f687fd --- /dev/null +++ b/smmap/smmap/mman.py @@ -0,0 +1,588 @@ +"""Module containing a memory memory manager which provides a sliding window on a number of memory mapped files""" +from .util import ( + MapWindow, + MapRegion, + MapRegionList, + is_64_bit, +) + +import sys +from functools import reduce + +__all__ = ["StaticWindowMapManager", "SlidingWindowMapManager", "WindowCursor"] +#{ Utilities + +#}END utilities + + +class WindowCursor: + + """ + Pointer into the mapped region of the memory manager, keeping the map + alive until it is destroyed and no other client uses it. + + Cursors should not be created manually, but are instead returned by the SlidingWindowMapManager + + **Note:**: The current implementation is suited for static and sliding window managers, but it also means + that it must be suited for the somewhat quite different sliding manager. It could be improved, but + I see no real need to do so.""" + __slots__ = ( + '_manager', # the manager keeping all file regions + '_rlist', # a regions list with regions for our file + '_region', # our current class:`MapRegion` or None + '_ofs', # relative offset from the actually mapped area to our start area + '_size' # maximum size we should provide + ) + + def __init__(self, manager=None, regions=None): + self._manager = manager + self._rlist = regions + self._region = None + self._ofs = 0 + self._size = 0 + + def __del__(self): + self._destroy() + + def __enter__(self): + return self + + def __exit__(self, exc_type, exc_value, traceback): + self._destroy() + + def _destroy(self): + """Destruction code to decrement counters""" + self.unuse_region() + + if self._rlist is not None: + # Actual client count, which doesn't include the reference kept by the manager, nor ours + # as we are about to be deleted + try: + if len(self._rlist) == 0: + # Free all resources associated with the mapped file + self._manager._fdict.pop(self._rlist.path_or_fd()) + # END remove regions list from manager + except (TypeError, KeyError): + # sometimes, during shutdown, getrefcount is None. Its possible + # to re-import it, however, its probably better to just ignore + # this python problem (for now). + # The next step is to get rid of the error prone getrefcount altogether. + pass + # END exception handling + # END handle regions + + def _copy_from(self, rhs): + """Copy all data from rhs into this instance, handles usage count""" + self._manager = rhs._manager + self._rlist = type(rhs._rlist)(rhs._rlist) + self._region = rhs._region + self._ofs = rhs._ofs + self._size = rhs._size + + for region in self._rlist: + region.increment_client_count() + + if self._region is not None: + self._region.increment_client_count() + # END handle regions + + def __copy__(self): + """copy module interface""" + cpy = type(self)() + cpy._copy_from(self) + return cpy + + #{ Interface + def assign(self, rhs): + """Assign rhs to this instance. This is required in order to get a real copy. + Alternatively, you can copy an existing instance using the copy module""" + self._destroy() + self._copy_from(rhs) + + def use_region(self, offset=0, size=0, flags=0): + """Assure we point to a window which allows access to the given offset into the file + + :param offset: absolute offset in bytes into the file + :param size: amount of bytes to map. If 0, all available bytes will be mapped + :param flags: additional flags to be given to os.open in case a file handle is initially opened + for mapping. Has no effect if a region can actually be reused. + :return: this instance - it should be queried for whether it points to a valid memory region. + This is not the case if the mapping failed because we reached the end of the file + + **Note:**: The size actually mapped may be smaller than the given size. If that is the case, + either the file has reached its end, or the map was created between two existing regions""" + need_region = True + man = self._manager + fsize = self._rlist.file_size() + size = min(size or fsize, man.window_size() or fsize) # clamp size to window size + + if self._region is not None: + if self._region.includes_ofs(offset): + need_region = False + else: + self.unuse_region() + # END handle existing region + # END check existing region + + # offset too large ? + if offset >= fsize: + return self + # END handle offset + + if need_region: + self._region = man._obtain_region(self._rlist, offset, size, flags, False) + self._region.increment_client_count() + # END need region handling + + self._ofs = offset - self._region._b + self._size = min(size, self._region.ofs_end() - offset) + + return self + + def unuse_region(self): + """Unuse the current region. Does nothing if we have no current region + + **Note:** the cursor unuses the region automatically upon destruction. It is recommended + to un-use the region once you are done reading from it in persistent cursors as it + helps to free up resource more quickly""" + if self._region is not None: + self._region.increment_client_count(-1) + self._region = None + # note: should reset ofs and size, but we spare that for performance. Its not + # allowed to query information if we are not valid ! + + def buffer(self): + """Return a buffer object which allows access to our memory region from our offset + to the window size. Please note that it might be smaller than you requested when calling use_region() + + **Note:** You can only obtain a buffer if this instance is_valid() ! + + **Note:** buffers should not be cached passed the duration of your access as it will + prevent resources from being freed even though they might not be accounted for anymore !""" + return memoryview(self._region.buffer())[self._ofs:self._ofs+self._size] + + def map(self): + """ + :return: the underlying raw memory map. Please not that the offset and size is likely to be different + to what you set as offset and size. Use it only if you are sure about the region it maps, which is the whole + file in case of StaticWindowMapManager""" + return self._region.map() + + def is_valid(self): + """:return: True if we have a valid and usable region""" + return self._region is not None + + def is_associated(self): + """:return: True if we are associated with a specific file already""" + return self._rlist is not None + + def ofs_begin(self): + """:return: offset to the first byte pointed to by our cursor + + **Note:** only if is_valid() is True""" + return self._region._b + self._ofs + + def ofs_end(self): + """:return: offset to one past the last available byte""" + # unroll method calls for performance ! + return self._region._b + self._ofs + self._size + + def size(self): + """:return: amount of bytes we point to""" + return self._size + + def region(self): + """:return: our mapped region, or None if nothing is mapped yet + :raise AssertionError: if we have no current region. This is only useful for debugging""" + return self._region + + def includes_ofs(self, ofs): + """:return: True if the given absolute offset is contained in the cursors + current region + + **Note:** cursor must be valid for this to work""" + # unroll methods + return (self._region._b + self._ofs) <= ofs < (self._region._b + self._ofs + self._size) + + def file_size(self): + """:return: size of the underlying file""" + return self._rlist.file_size() + + def path_or_fd(self): + """:return: path or file descriptor of the underlying mapped file""" + return self._rlist.path_or_fd() + + def path(self): + """:return: path of the underlying mapped file + :raise ValueError: if attached path is not a path""" + if isinstance(self._rlist.path_or_fd(), int): + raise ValueError("Path queried although mapping was applied to a file descriptor") + # END handle type + return self._rlist.path_or_fd() + + def fd(self): + """:return: file descriptor used to create the underlying mapping. + + **Note:** it is not required to be valid anymore + :raise ValueError: if the mapping was not created by a file descriptor""" + if isinstance(self._rlist.path_or_fd(), str): + raise ValueError("File descriptor queried although mapping was generated from path") + # END handle type + return self._rlist.path_or_fd() + + #} END interface + + +class StaticWindowMapManager: + + """Provides a manager which will produce single size cursors that are allowed + to always map the whole file. + + Clients must be written to specifically know that they are accessing their data + through a StaticWindowMapManager, as they otherwise have to deal with their window size. + + These clients would have to use a SlidingWindowMapBuffer to hide this fact. + + This type will always use a maximum window size, and optimize certain methods to + accommodate this fact""" + + __slots__ = [ + '_fdict', # mapping of path -> StorageHelper (of some kind + '_window_size', # maximum size of a window + '_max_memory_size', # maximum amount of memory we may allocate + '_max_handle_count', # maximum amount of handles to keep open + '_memory_size', # currently allocated memory size + '_handle_count', # amount of currently allocated file handles + ] + + #{ Configuration + MapRegionListCls = MapRegionList + MapWindowCls = MapWindow + MapRegionCls = MapRegion + WindowCursorCls = WindowCursor + #} END configuration + + _MB_in_bytes = 1024 * 1024 + + def __init__(self, window_size=0, max_memory_size=0, max_open_handles=sys.maxsize): + """initialize the manager with the given parameters. + :param window_size: if -1, a default window size will be chosen depending on + the operating system's architecture. It will internally be quantified to a multiple of the page size + If 0, the window may have any size, which basically results in mapping the whole file at one + :param max_memory_size: maximum amount of memory we may map at once before releasing mapped regions. + If 0, a viable default will be set depending on the system's architecture. + It is a soft limit that is tried to be kept, but nothing bad happens if we have to over-allocate + :param max_open_handles: if not maxint, limit the amount of open file handles to the given number. + Otherwise the amount is only limited by the system itself. If a system or soft limit is hit, + the manager will free as many handles as possible""" + self._fdict = dict() + self._window_size = window_size + self._max_memory_size = max_memory_size + self._max_handle_count = max_open_handles + self._memory_size = 0 + self._handle_count = 0 + + if window_size < 0: + coeff = 64 + if is_64_bit(): + coeff = 1024 + # END handle arch + self._window_size = coeff * self._MB_in_bytes + # END handle max window size + + if max_memory_size == 0: + coeff = 1024 + if is_64_bit(): + coeff = 8192 + # END handle arch + self._max_memory_size = coeff * self._MB_in_bytes + # END handle max memory size + + #{ Internal Methods + + def _collect_lru_region(self, size): + """Unmap the region which was least-recently used and has no client + :param size: size of the region we want to map next (assuming its not already mapped partially or full + if 0, we try to free any available region + :return: Amount of freed regions + + .. Note:: + We don't raise exceptions anymore, in order to keep the system working, allowing temporary overallocation. + If the system runs out of memory, it will tell. + + .. TODO:: + implement a case where all unusued regions are discarded efficiently. + Currently its only brute force + """ + num_found = 0 + while (size == 0) or (self._memory_size + size > self._max_memory_size): + lru_region = None + lru_list = None + for regions in self._fdict.values(): + for region in regions: + # check client count - if it's 1, it's just us + if (region.client_count() == 1 and + (lru_region is None or region._uc < lru_region._uc)): + lru_region = region + lru_list = regions + # END update lru_region + # END for each region + # END for each regions list + + if lru_region is None: + break + # END handle region not found + + num_found += 1 + del(lru_list[lru_list.index(lru_region)]) + lru_region.increment_client_count(-1) + self._memory_size -= lru_region.size() + self._handle_count -= 1 + # END while there is more memory to free + return num_found + + def _obtain_region(self, a, offset, size, flags, is_recursive): + """Utility to create a new region - for more information on the parameters, + see MapCursor.use_region. + :param a: A regions (a)rray + :return: The newly created region""" + if self._memory_size + size > self._max_memory_size: + self._collect_lru_region(size) + # END handle collection + + r = None + if a: + assert len(a) == 1 + r = a[0] + else: + try: + r = self.MapRegionCls(a.path_or_fd(), 0, sys.maxsize, flags) + except Exception: + # apparently we are out of system resources or hit a limit + # As many more operations are likely to fail in that condition ( + # like reading a file from disk, etc) we free up as much as possible + # As this invalidates our insert position, we have to recurse here + if is_recursive: + # we already tried this, and still have no success in obtaining + # a mapping. This is an exception, so we propagate it + raise + # END handle existing recursion + self._collect_lru_region(0) + return self._obtain_region(a, offset, size, flags, True) + # END handle exceptions + + self._handle_count += 1 + self._memory_size += r.size() + a.append(r) + # END handle array + + assert r.includes_ofs(offset) + return r + + #}END internal methods + + #{ Interface + def make_cursor(self, path_or_fd): + """ + :return: a cursor pointing to the given path or file descriptor. + It can be used to map new regions of the file into memory + + **Note:** if a file descriptor is given, it is assumed to be open and valid, + but may be closed afterwards. To refer to the same file, you may reuse + your existing file descriptor, but keep in mind that new windows can only + be mapped as long as it stays valid. This is why the using actual file paths + are preferred unless you plan to keep the file descriptor open. + + **Note:** file descriptors are problematic as they are not necessarily unique, as two + different files opened and closed in succession might have the same file descriptor id. + + **Note:** Using file descriptors directly is faster once new windows are mapped as it + prevents the file to be opened again just for the purpose of mapping it.""" + regions = self._fdict.get(path_or_fd) + if regions is None: + regions = self.MapRegionListCls(path_or_fd) + self._fdict[path_or_fd] = regions + # END obtain region for path + return self.WindowCursorCls(self, regions) + + def collect(self): + """Collect all available free-to-collect mapped regions + :return: Amount of freed handles""" + return self._collect_lru_region(0) + + def num_file_handles(self): + """:return: amount of file handles in use. Each mapped region uses one file handle""" + return self._handle_count + + def num_open_files(self): + """Amount of opened files in the system""" + return reduce(lambda x, y: x + y, (1 for rlist in self._fdict.values() if len(rlist) > 0), 0) + + def window_size(self): + """:return: size of each window when allocating new regions""" + return self._window_size + + def mapped_memory_size(self): + """:return: amount of bytes currently mapped in total""" + return self._memory_size + + def max_file_handles(self): + """:return: maximum amount of handles we may have opened""" + return self._max_handle_count + + def max_mapped_memory_size(self): + """:return: maximum amount of memory we may allocate""" + return self._max_memory_size + + #} END interface + + #{ Special Purpose Interface + + def force_map_handle_removal_win(self, base_path): + """ONLY AVAILABLE ON WINDOWS + On windows removing files is not allowed if anybody still has it opened. + If this process is ourselves, and if the whole process uses this memory + manager (as far as the parent framework is concerned) we can enforce + closing all memory maps whose path matches the given base path to + allow the respective operation after all. + The respective system must NOT access the closed memory regions anymore ! + This really may only be used if you know that the items which keep + the cursors alive will not be using it anymore. They need to be recreated ! + :return: Amount of closed handles + + **Note:** does nothing on non-windows platforms""" + if sys.platform != 'win32': + return + # END early bailout + + num_closed = 0 + for path, rlist in self._fdict.items(): + if path.startswith(base_path): + for region in rlist: + region.release() + num_closed += 1 + # END path matches + # END for each path + return num_closed + #} END special purpose interface + + +class SlidingWindowMapManager(StaticWindowMapManager): + + """Maintains a list of ranges of mapped memory regions in one or more files and allows to easily + obtain additional regions assuring there is no overlap. + Once a certain memory limit is reached globally, or if there cannot be more open file handles + which result from each mmap call, the least recently used, and currently unused mapped regions + are unloaded automatically. + + **Note:** currently not thread-safe ! + + **Note:** in the current implementation, we will automatically unload windows if we either cannot + create more memory maps (as the open file handles limit is hit) or if we have allocated more than + a safe amount of memory already, which would possibly cause memory allocations to fail as our address + space is full.""" + + __slots__ = tuple() + + def __init__(self, window_size=-1, max_memory_size=0, max_open_handles=sys.maxsize): + """Adjusts the default window size to -1""" + super().__init__(window_size, max_memory_size, max_open_handles) + + def _obtain_region(self, a, offset, size, flags, is_recursive): + # bisect to find an existing region. The c++ implementation cannot + # do that as it uses a linked list for regions. + r = None + lo = 0 + hi = len(a) + while lo < hi: + mid = (lo + hi) // 2 + ofs = a[mid]._b + if ofs <= offset: + if a[mid].includes_ofs(offset): + r = a[mid] + break + # END have region + lo = mid + 1 + else: + hi = mid + # END handle position + # END while bisecting + + if r is None: + window_size = self._window_size + left = self.MapWindowCls(0, 0) + mid = self.MapWindowCls(offset, size) + right = self.MapWindowCls(a.file_size(), 0) + + # we want to honor the max memory size, and assure we have anough + # memory available + # Save calls ! + if self._memory_size + window_size > self._max_memory_size: + self._collect_lru_region(window_size) + # END handle collection + + # we assume the list remains sorted by offset + insert_pos = 0 + len_regions = len(a) + if len_regions == 1: + if a[0]._b <= offset: + insert_pos = 1 + # END maintain sort + else: + # find insert position + insert_pos = len_regions + for i, region in enumerate(a): + if region._b > offset: + insert_pos = i + break + # END if insert position is correct + # END for each region + # END obtain insert pos + + # adjust the actual offset and size values to create the largest + # possible mapping + if insert_pos == 0: + if len_regions: + right = self.MapWindowCls.from_region(a[insert_pos]) + # END adjust right side + else: + if insert_pos != len_regions: + right = self.MapWindowCls.from_region(a[insert_pos]) + # END adjust right window + left = self.MapWindowCls.from_region(a[insert_pos - 1]) + # END adjust surrounding windows + + mid.extend_left_to(left, window_size) + mid.extend_right_to(right, window_size) + mid.align() + + # it can happen that we align beyond the end of the file + if mid.ofs_end() > right.ofs: + mid.size = right.ofs - mid.ofs + # END readjust size + + # insert new region at the right offset to keep the order + try: + if self._handle_count >= self._max_handle_count: + raise Exception + # END assert own imposed max file handles + r = self.MapRegionCls(a.path_or_fd(), mid.ofs, mid.size, flags) + except Exception: + # apparently we are out of system resources or hit a limit + # As many more operations are likely to fail in that condition ( + # like reading a file from disk, etc) we free up as much as possible + # As this invalidates our insert position, we have to recurse here + if is_recursive: + # we already tried this, and still have no success in obtaining + # a mapping. This is an exception, so we propagate it + raise + # END handle existing recursion + self._collect_lru_region(0) + return self._obtain_region(a, offset, size, flags, True) + # END handle exceptions + + self._handle_count += 1 + self._memory_size += r.size() + a.insert(insert_pos, r) + # END create new region + return r diff --git a/smmap/smmap/test/__init__.py b/smmap/smmap/test/__init__.py new file mode 100644 index 000000000..e69de29bb diff --git a/smmap/smmap/test/lib.py b/smmap/smmap/test/lib.py new file mode 100644 index 000000000..b15b0ec6a --- /dev/null +++ b/smmap/smmap/test/lib.py @@ -0,0 +1,72 @@ +"""Provide base classes for the test system""" +from unittest import TestCase +import os +import tempfile + +__all__ = ['TestBase', 'FileCreator'] + + +#{ Utilities + +class FileCreator: + + """A instance which creates a temporary file with a prefix and a given size + and provides this info to the user. + Once it gets deleted, it will remove the temporary file as well.""" + __slots__ = ("_size", "_path") + + def __init__(self, size, prefix=''): + assert size, "Require size to be larger 0" + + self._size = size + + with tempfile.NamedTemporaryFile("wb", prefix=prefix, delete=False) as file: + self._path = file.name + file.seek(size - 1) + file.write(b'1') + + assert os.path.getsize(self.path) == size + + def __del__(self): + try: + os.remove(self.path) + except OSError: + pass + # END exception handling + + def __enter__(self): + return self + + def __exit__(self, exc_type, exc_value, traceback): + self.__del__() + + @property + def path(self): + return self._path + + @property + def size(self): + return self._size + +#} END utilities + + +class TestBase(TestCase): + + """Foundation used by all tests""" + + #{ Configuration + k_window_test_size = 1000 * 1000 * 8 + 5195 + #} END configuration + + #{ Overrides + @classmethod + def setUpAll(cls): + # nothing for now + pass + + # END overrides + + #{ Interface + + #} END interface diff --git a/smmap/smmap/test/test_buf.py b/smmap/smmap/test/test_buf.py new file mode 100644 index 000000000..f0a86fb64 --- /dev/null +++ b/smmap/smmap/test/test_buf.py @@ -0,0 +1,126 @@ +from .lib import TestBase, FileCreator + +from smmap.mman import ( + SlidingWindowMapManager, + StaticWindowMapManager +) +from smmap.buf import SlidingWindowMapBuffer + +from random import randint +from time import time +import sys +import os + + +man_optimal = SlidingWindowMapManager() +man_worst_case = SlidingWindowMapManager( + window_size=TestBase.k_window_test_size // 100, + max_memory_size=TestBase.k_window_test_size // 3, + max_open_handles=15) +static_man = StaticWindowMapManager() + + +class TestBuf(TestBase): + + def test_basics(self): + with FileCreator(self.k_window_test_size, "buffer_test") as fc: + + # invalid paths fail upon construction + c = man_optimal.make_cursor(fc.path) + self.assertRaises(ValueError, SlidingWindowMapBuffer, type(c)()) # invalid cursor + self.assertRaises(ValueError, SlidingWindowMapBuffer, c, fc.size) # offset too large + + buf = SlidingWindowMapBuffer() # can create uninitialized buffers + assert buf.cursor() is None + + # can call end access any time + buf.end_access() + buf.end_access() + assert len(buf) == 0 + + # begin access can revive it, if the offset is suitable + offset = 100 + assert buf.begin_access(c, fc.size) == False + assert buf.begin_access(c, offset) == True + assert len(buf) == fc.size - offset + assert buf.cursor().is_valid() + + # empty begin access keeps it valid on the same path, but alters the offset + assert buf.begin_access() == True + assert len(buf) == fc.size + assert buf.cursor().is_valid() + + # simple access + with open(fc.path, 'rb') as fp: + data = fp.read() + assert data[offset] == buf[0] + assert data[offset:offset * 2] == buf[0:offset] + + # negative indices, partial slices + assert buf[-1] == buf[len(buf) - 1] + assert buf[-10:] == buf[len(buf) - 10:len(buf)] + + # end access makes its cursor invalid + buf.end_access() + assert not buf.cursor().is_valid() + assert buf.cursor().is_associated() # but it remains associated + + # an empty begin access fixes it up again + assert buf.begin_access() == True and buf.cursor().is_valid() + del(buf) # ends access automatically + del(c) + + assert man_optimal.num_file_handles() == 1 + + # PERFORMANCE + # blast away with random access and a full mapping - we don't want to + # exaggerate the manager's overhead, but measure the buffer overhead + # We do it once with an optimal setting, and with a worse manager which + # will produce small mappings only ! + max_num_accesses = 100 + fd = os.open(fc.path, os.O_RDONLY) + for item in (fc.path, fd): + for manager, man_id in ((man_optimal, 'optimal'), + (man_worst_case, 'worst case'), + (static_man, 'static optimal')): + buf = SlidingWindowMapBuffer(manager.make_cursor(item)) + assert manager.num_file_handles() == 1 + for access_mode in range(2): # single, multi + num_accesses_left = max_num_accesses + num_bytes = 0 + fsize = fc.size + + st = time() + buf.begin_access() + while num_accesses_left: + num_accesses_left -= 1 + if access_mode: # multi + ofs_start = randint(0, fsize) + ofs_end = randint(ofs_start, fsize) + d = buf[ofs_start:ofs_end] + assert len(d) == ofs_end - ofs_start + assert d == data[ofs_start:ofs_end] + num_bytes += len(d) + del d + else: + pos = randint(0, fsize) + assert buf[pos] == data[pos] + num_bytes += 1 + # END handle mode + # END handle num accesses + + buf.end_access() + assert manager.num_file_handles() + assert manager.collect() + assert manager.num_file_handles() == 0 + elapsed = max(time() - st, 0.001) # prevent zero division errors on windows + mb = float(1000 * 1000) + mode_str = (access_mode and "slice") or "single byte" + print("%s: Made %i random %s accesses to buffer created from %s reading a total of %f mb in %f s (%f mb/s)" + % (man_id, max_num_accesses, mode_str, type(item), num_bytes / mb, elapsed, (num_bytes / mb) / elapsed), + file=sys.stderr) + # END handle access mode + del buf + # END for each manager + # END for each input + os.close(fd) diff --git a/smmap/smmap/test/test_mman.py b/smmap/smmap/test/test_mman.py new file mode 100644 index 000000000..7a5f4092c --- /dev/null +++ b/smmap/smmap/test/test_mman.py @@ -0,0 +1,224 @@ +from .lib import TestBase, FileCreator + +from smmap.mman import ( + WindowCursor, + SlidingWindowMapManager, + StaticWindowMapManager +) +from smmap.util import align_to_mmap + +from random import randint +from time import time +import os +import sys +from copy import copy + + +class TestMMan(TestBase): + + def test_cursor(self): + with FileCreator(self.k_window_test_size, "cursor_test") as fc: + man = SlidingWindowMapManager() + ci = WindowCursor(man) # invalid cursor + assert not ci.is_valid() + assert not ci.is_associated() + assert ci.size() == 0 # this is cached, so we can query it in invalid state + + cv = man.make_cursor(fc.path) + assert not cv.is_valid() # no region mapped yet + assert cv.is_associated() # but it know where to map it from + assert cv.file_size() == fc.size + assert cv.path() == fc.path + + # copy module + cio = copy(cv) + assert not cio.is_valid() and cio.is_associated() + + # assign method + assert not ci.is_associated() + ci.assign(cv) + assert not ci.is_valid() and ci.is_associated() + + # unuse non-existing region is fine + cv.unuse_region() + cv.unuse_region() + + # destruction is fine (even multiple times) + cv._destroy() + WindowCursor(man)._destroy() + + def test_memory_manager(self): + slide_man = SlidingWindowMapManager() + static_man = StaticWindowMapManager() + + for man in (static_man, slide_man): + assert man.num_file_handles() == 0 + assert man.num_open_files() == 0 + winsize_cmp_val = 0 + if isinstance(man, StaticWindowMapManager): + winsize_cmp_val = -1 + # END handle window size + assert man.window_size() > winsize_cmp_val + assert man.mapped_memory_size() == 0 + assert man.max_mapped_memory_size() > 0 + + # collection doesn't raise in 'any' mode + man._collect_lru_region(0) + # doesn't raise if we are within the limit + man._collect_lru_region(10) + + # doesn't fail if we over-allocate + assert man._collect_lru_region(sys.maxsize) == 0 + + # use a region, verify most basic functionality + with FileCreator(self.k_window_test_size, "manager_test") as fc: + fd = os.open(fc.path, os.O_RDONLY) + try: + for item in (fc.path, fd): + c = man.make_cursor(item) + assert c.path_or_fd() is item + assert c.use_region(10, 10).is_valid() + assert c.ofs_begin() == 10 + assert c.size() == 10 + with open(fc.path, 'rb') as fp: + assert c.buffer()[:] == fp.read(20)[10:] + + if isinstance(item, int): + self.assertRaises(ValueError, c.path) + else: + self.assertRaises(ValueError, c.fd) + # END handle value error + # END for each input + finally: + os.close(fd) + # END for each manasger type + + def test_memman_operation(self): + # test more access, force it to actually unmap regions + with FileCreator(self.k_window_test_size, "manager_operation_test") as fc: + with open(fc.path, 'rb') as fp: + data = fp.read() + fd = os.open(fc.path, os.O_RDONLY) + try: + max_num_handles = 15 + # small_size = + for mtype, args in ((StaticWindowMapManager, (0, fc.size // 3, max_num_handles)), + (SlidingWindowMapManager, (fc.size // 100, fc.size // 3, max_num_handles)),): + for item in (fc.path, fd): + assert len(data) == fc.size + + # small windows, a reasonable max memory. Not too many regions at once + man = mtype(window_size=args[0], max_memory_size=args[1], max_open_handles=args[2]) + c = man.make_cursor(item) + + # still empty (more about that is tested in test_memory_manager() + assert man.num_open_files() == 0 + assert man.mapped_memory_size() == 0 + + base_offset = 5000 + # window size is 0 for static managers, hence size will be 0. We take that into consideration + size = man.window_size() // 2 + assert c.use_region(base_offset, size).is_valid() + rr = c.region() + assert rr.client_count() == 2 # the manager and the cursor and us + + assert man.num_open_files() == 1 + assert man.num_file_handles() == 1 + assert man.mapped_memory_size() == rr.size() + + # assert c.size() == size # the cursor may overallocate in its static version + assert c.ofs_begin() == base_offset + assert rr.ofs_begin() == 0 # it was aligned and expanded + if man.window_size(): + # but isn't larger than the max window (aligned) + assert rr.size() == align_to_mmap(man.window_size(), True) + else: + assert rr.size() == fc.size + # END ignore static managers which dont use windows and are aligned to file boundaries + + assert c.buffer()[:] == data[base_offset:base_offset + (size or c.size())] + + # obtain second window, which spans the first part of the file - it is a still the same window + nsize = (size or fc.size) - 10 + assert c.use_region(0, nsize).is_valid() + assert c.region() == rr + assert man.num_file_handles() == 1 + assert c.size() == nsize + assert c.ofs_begin() == 0 + assert c.buffer()[:] == data[:nsize] + + # map some part at the end, our requested size cannot be kept + overshoot = 4000 + base_offset = fc.size - (size or c.size()) + overshoot + assert c.use_region(base_offset, size).is_valid() + if man.window_size(): + assert man.num_file_handles() == 2 + assert c.size() < size + assert c.region() is not rr # old region is still available, but has not cursor ref anymore + assert rr.client_count() == 1 # only held by manager + else: + assert c.size() < fc.size + # END ignore static managers which only have one handle per file + rr = c.region() + assert rr.client_count() == 2 # manager + cursor + assert rr.ofs_begin() < c.ofs_begin() # it should have extended itself to the left + assert rr.ofs_end() <= fc.size # it cannot be larger than the file + assert c.buffer()[:] == data[base_offset:base_offset + (size or c.size())] + + # unising a region makes the cursor invalid + c.unuse_region() + assert not c.is_valid() + if man.window_size(): + # but doesn't change anything regarding the handle count - we cache it and only + # remove mapped regions if we have to + assert man.num_file_handles() == 2 + # END ignore this for static managers + + # iterate through the windows, verify data contents + # this will trigger map collection after a while + max_random_accesses = 5000 + num_random_accesses = max_random_accesses + memory_read = 0 + st = time() + + # cache everything to get some more performance + includes_ofs = c.includes_ofs + max_mapped_memory_size = man.max_mapped_memory_size() + max_file_handles = man.max_file_handles() + mapped_memory_size = man.mapped_memory_size + num_file_handles = man.num_file_handles + while num_random_accesses: + num_random_accesses -= 1 + base_offset = randint(0, fc.size - 1) + + # precondition + if man.window_size(): + assert max_mapped_memory_size >= mapped_memory_size() + # END statistics will overshoot, which is fine + assert max_file_handles >= num_file_handles() + assert c.use_region(base_offset, (size or c.size())).is_valid() + csize = c.size() + assert c.buffer()[:] == data[base_offset:base_offset + csize] + memory_read += csize + + assert includes_ofs(base_offset) + assert includes_ofs(base_offset + csize - 1) + assert not includes_ofs(base_offset + csize) + # END while we should do an access + elapsed = max(time() - st, 0.001) # prevent zero division errors on windows + mb = float(1000 * 1000) + print("%s: Read %i mb of memory with %i random on cursor initialized with %s accesses in %fs (%f mb/s)\n" + % (mtype, memory_read / mb, max_random_accesses, type(item), elapsed, (memory_read / mb) / elapsed), + file=sys.stderr) + + # an offset as large as the size doesn't work ! + assert not c.use_region(fc.size, size).is_valid() + + # collection - it should be able to collect all + assert man.num_file_handles() + assert man.collect() + assert man.num_file_handles() == 0 + # END for each item + # END for each manager type + finally: + os.close(fd) diff --git a/smmap/smmap/test/test_tutorial.py b/smmap/smmap/test/test_tutorial.py new file mode 100644 index 000000000..31c272abb --- /dev/null +++ b/smmap/smmap/test/test_tutorial.py @@ -0,0 +1,75 @@ +from .lib import TestBase + + +class TestTutorial(TestBase): + + def test_example(self): + # Memory Managers + ################## + import smmap + # This instance should be globally available in your application + # It is configured to be well suitable for 32-bit or 64 bit applications. + mman = smmap.SlidingWindowMapManager() + + # the manager provides much useful information about its current state + # like the amount of open file handles or the amount of mapped memory + assert mman.num_file_handles() == 0 + assert mman.mapped_memory_size() == 0 + # and many more ... + + # Cursors + ########## + import smmap.test.lib + with smmap.test.lib.FileCreator(1024 * 1024 * 8, "test_file") as fc: + # obtain a cursor to access some file. + c = mman.make_cursor(fc.path) + + # the cursor is now associated with the file, but not yet usable + assert c.is_associated() + assert not c.is_valid() + + # before you can use the cursor, you have to specify a window you want to + # access. The following just says you want as much data as possible starting + # from offset 0. + # To be sure your region could be mapped, query for validity + assert c.use_region().is_valid() # use_region returns self + + # once a region was mapped, you must query its dimension regularly + # to assure you don't try to access its buffer out of its bounds + assert c.size() + c.buffer()[0] # first byte + c.buffer()[1:10] # first 9 bytes + c.buffer()[c.size() - 1] # last byte + + # you can query absolute offsets, and check whether an offset is included + # in the cursor's data. + assert c.ofs_begin() < c.ofs_end() + assert c.includes_ofs(100) + + # If you are over out of bounds with one of your region requests, the + # cursor will be come invalid. It cannot be used in that state + assert not c.use_region(fc.size, 100).is_valid() + # map as much as possible after skipping the first 100 bytes + assert c.use_region(100).is_valid() + + # You can explicitly free cursor resources by unusing the cursor's region + c.unuse_region() + assert not c.is_valid() + + # Buffers + ######### + # Create a default buffer which can operate on the whole file + buf = smmap.SlidingWindowMapBuffer(mman.make_cursor(fc.path)) + + # you can use it right away + assert buf.cursor().is_valid() + + buf[0] # access the first byte + buf[-1] # access the last ten bytes on the file + buf[-10:] # access the last ten bytes + + # If you want to keep the instance between different accesses, use the + # dedicated methods + buf.end_access() + assert not buf.cursor().is_valid() # you cannot use the buffer anymore + assert buf.begin_access(offset=10) # start using the buffer at an offset diff --git a/smmap/smmap/test/test_util.py b/smmap/smmap/test/test_util.py new file mode 100644 index 000000000..e6ac10f35 --- /dev/null +++ b/smmap/smmap/test/test_util.py @@ -0,0 +1,105 @@ +from .lib import TestBase, FileCreator + +from smmap.util import ( + MapWindow, + MapRegion, + MapRegionList, + ALLOCATIONGRANULARITY, + is_64_bit, + align_to_mmap +) + +import os +import sys + + +class TestMMan(TestBase): + + def test_window(self): + wl = MapWindow(0, 1) # left + wc = MapWindow(1, 1) # center + wc2 = MapWindow(10, 5) # another center + wr = MapWindow(8000, 50) # right + + assert wl.ofs_end() == 1 + assert wc.ofs_end() == 2 + assert wr.ofs_end() == 8050 + + # extension does nothing if already in place + maxsize = 100 + wc.extend_left_to(wl, maxsize) + assert wc.ofs == 1 and wc.size == 1 + wl.extend_right_to(wc, maxsize) + wl.extend_right_to(wc, maxsize) + assert wl.ofs == 0 and wl.size == 1 + + # an actual left extension + pofs_end = wc2.ofs_end() + wc2.extend_left_to(wc, maxsize) + assert wc2.ofs == wc.ofs_end() and pofs_end == wc2.ofs_end() + + # respects maxsize + wc.extend_right_to(wr, maxsize) + assert wc.ofs == 1 and wc.size == maxsize + wc.extend_right_to(wr, maxsize) + assert wc.ofs == 1 and wc.size == maxsize + + # without maxsize + wc.extend_right_to(wr, sys.maxsize) + assert wc.ofs_end() == wr.ofs and wc.ofs == 1 + + # extend left + wr.extend_left_to(wc2, maxsize) + wr.extend_left_to(wc2, maxsize) + assert wr.size == maxsize + + wr.extend_left_to(wc2, sys.maxsize) + assert wr.ofs == wc2.ofs_end() + + wc.align() + assert wc.ofs == 0 and wc.size == align_to_mmap(wc.size, True) + + def test_region(self): + with FileCreator(self.k_window_test_size, "window_test") as fc: + half_size = fc.size // 2 + rofs = align_to_mmap(4200, False) + rfull = MapRegion(fc.path, 0, fc.size) + rhalfofs = MapRegion(fc.path, rofs, fc.size) + rhalfsize = MapRegion(fc.path, 0, half_size) + + # offsets + assert rfull.ofs_begin() == 0 and rfull.size() == fc.size + assert rfull.ofs_end() == fc.size # if this method works, it works always + + assert rhalfofs.ofs_begin() == rofs and rhalfofs.size() == fc.size - rofs + assert rhalfsize.ofs_begin() == 0 and rhalfsize.size() == half_size + + assert rfull.includes_ofs(0) and rfull.includes_ofs(fc.size - 1) and rfull.includes_ofs(half_size) + assert not rfull.includes_ofs(-1) and not rfull.includes_ofs(sys.maxsize) + + # auto-refcount + assert rfull.client_count() == 1 + rfull2 = rfull + assert rfull.client_count() == 1, "no auto-counting" + + # window constructor + w = MapWindow.from_region(rfull) + assert w.ofs == rfull.ofs_begin() and w.ofs_end() == rfull.ofs_end() + + def test_region_list(self): + with FileCreator(100, "sample_file") as fc: + fd = os.open(fc.path, os.O_RDONLY) + try: + for item in (fc.path, fd): + ml = MapRegionList(item) + + assert len(ml) == 0 + assert ml.path_or_fd() == item + assert ml.file_size() == fc.size + finally: + os.close(fd) + + def test_util(self): + assert isinstance(is_64_bit(), bool) # just call it + assert align_to_mmap(1, False) == 0 + assert align_to_mmap(1, True) == ALLOCATIONGRANULARITY diff --git a/smmap/smmap/util.py b/smmap/smmap/util.py new file mode 100644 index 000000000..fbb387211 --- /dev/null +++ b/smmap/smmap/util.py @@ -0,0 +1,222 @@ +"""Module containing a memory memory manager which provides a sliding window on a number of memory mapped files""" +import os +import sys + +from mmap import mmap, ACCESS_READ +from mmap import ALLOCATIONGRANULARITY + +__all__ = ["align_to_mmap", "is_64_bit", + "MapWindow", "MapRegion", "MapRegionList", "ALLOCATIONGRANULARITY"] + +#{ Utilities + + +def align_to_mmap(num, round_up): + """ + Align the given integer number to the closest page offset, which usually is 4096 bytes. + + :param round_up: if True, the next higher multiple of page size is used, otherwise + the lower page_size will be used (i.e. if True, 1 becomes 4096, otherwise it becomes 0) + :return: num rounded to closest page""" + res = (num // ALLOCATIONGRANULARITY) * ALLOCATIONGRANULARITY + if round_up and (res != num): + res += ALLOCATIONGRANULARITY + # END handle size + return res + + +def is_64_bit(): + """:return: True if the system is 64 bit. Otherwise it can be assumed to be 32 bit""" + return sys.maxsize > (1 << 32) - 1 + +#}END utilities + + +#{ Utility Classes + +class MapWindow: + + """Utility type which is used to snap windows towards each other, and to adjust their size""" + __slots__ = ( + 'ofs', # offset into the file in bytes + 'size' # size of the window in bytes + ) + + def __init__(self, offset, size): + self.ofs = offset + self.size = size + + def __repr__(self): + return "MapWindow(%i, %i)" % (self.ofs, self.size) + + @classmethod + def from_region(cls, region): + """:return: new window from a region""" + return cls(region._b, region.size()) + + def ofs_end(self): + return self.ofs + self.size + + def align(self): + """Assures the previous window area is contained in the new one""" + nofs = align_to_mmap(self.ofs, 0) + self.size += self.ofs - nofs # keep size constant + self.ofs = nofs + self.size = align_to_mmap(self.size, 1) + + def extend_left_to(self, window, max_size): + """Adjust the offset to start where the given window on our left ends if possible, + but don't make yourself larger than max_size. + The resize will assure that the new window still contains the old window area""" + rofs = self.ofs - window.ofs_end() + nsize = rofs + self.size + rofs -= nsize - min(nsize, max_size) + self.ofs -= rofs + self.size += rofs + + def extend_right_to(self, window, max_size): + """Adjust the size to make our window end where the right window begins, but don't + get larger than max_size""" + self.size = min(self.size + (window.ofs - self.ofs_end()), max_size) + + +class MapRegion: + + """Defines a mapped region of memory, aligned to pagesizes + + **Note:** deallocates used region automatically on destruction""" + __slots__ = [ + '_b', # beginning of mapping + '_mf', # mapped memory chunk (as returned by mmap) + '_uc', # total amount of usages + '_size', # cached size of our memory map + '__weakref__' + ] + + #{ Configuration + #} END configuration + + def __init__(self, path_or_fd, ofs, size, flags=0): + """Initialize a region, allocate the memory map + :param path_or_fd: path to the file to map, or the opened file descriptor + :param ofs: **aligned** offset into the file to be mapped + :param size: if size is larger then the file on disk, the whole file will be + allocated the the size automatically adjusted + :param flags: additional flags to be given when opening the file. + :raise Exception: if no memory can be allocated""" + self._b = ofs + self._size = 0 + self._uc = 0 + + if isinstance(path_or_fd, int): + fd = path_or_fd + else: + fd = os.open(path_or_fd, os.O_RDONLY | getattr(os, 'O_BINARY', 0) | flags) + # END handle fd + + try: + kwargs = dict(access=ACCESS_READ, offset=ofs) + corrected_size = size + sizeofs = ofs + + # have to correct size, otherwise (instead of the c version) it will + # bark that the size is too large ... many extra file accesses because + # if this ... argh ! + actual_size = min(os.fstat(fd).st_size - sizeofs, corrected_size) + self._mf = mmap(fd, actual_size, **kwargs) + # END handle memory mode + + self._size = len(self._mf) + finally: + if isinstance(path_or_fd, str): + os.close(fd) + # END only close it if we opened it + # END close file handle + # We assume the first one to use us keeps us around + self.increment_client_count() + + def __repr__(self): + return "MapRegion<%i, %i>" % (self._b, self.size()) + + #{ Interface + + def buffer(self): + """:return: a buffer containing the memory""" + return self._mf + + def map(self): + """:return: a memory map containing the memory""" + return self._mf + + def ofs_begin(self): + """:return: absolute byte offset to the first byte of the mapping""" + return self._b + + def size(self): + """:return: total size of the mapped region in bytes""" + return self._size + + def ofs_end(self): + """:return: Absolute offset to one byte beyond the mapping into the file""" + return self._b + self._size + + def includes_ofs(self, ofs): + """:return: True if the given offset can be read in our mapped region""" + return self._b <= ofs < self._b + self._size + + def client_count(self): + """:return: number of clients currently using this region""" + return self._uc + + def increment_client_count(self, ofs = 1): + """Adjust the usage count by the given positive or negative offset. + If usage count equals 0, we will auto-release our resources + :return: True if we released resources, False otherwise. In the latter case, we can still be used""" + self._uc += ofs + assert self._uc > -1, "Increments must match decrements, usage counter negative: %i" % self._uc + + if self.client_count() == 0: + self.release() + return True + else: + return False + # end handle release + + def release(self): + """Release all resources this instance might hold. Must only be called if there usage_count() is zero""" + self._mf.close() + + #} END interface + + +class MapRegionList(list): + + """List of MapRegion instances associating a path with a list of regions.""" + __slots__ = ( + '_path_or_fd', # path or file descriptor which is mapped by all our regions + '_file_size' # total size of the file we map + ) + + def __new__(cls, path): + return super().__new__(cls) + + def __init__(self, path_or_fd): + self._path_or_fd = path_or_fd + self._file_size = None + + def path_or_fd(self): + """:return: path or file descriptor we are attached to""" + return self._path_or_fd + + def file_size(self): + """:return: size of file we manager""" + if self._file_size is None: + if isinstance(self._path_or_fd, str): + self._file_size = os.stat(self._path_or_fd).st_size + else: + self._file_size = os.fstat(self._path_or_fd).st_size + # END handle path type + # END update file size + return self._file_size + +#} END utility classes diff --git a/smmap/tox.ini b/smmap/tox.ini new file mode 100644 index 000000000..f7205c5a3 --- /dev/null +++ b/smmap/tox.ini @@ -0,0 +1,17 @@ +# Tox (http://tox.testrun.org/) is a tool for running tests +# in multiple virtualenvs. This configuration file will run the +# test suite on all supported python versions. To use it, "pip install tox" +# and then run "tox" from this directory. + +[tox] +envlist = flake8, py{37, 38, 39, 310, 311, 312} + +[testenv] +commands = {envpython} -m pytest --cov smmap --cov-report xml {posargs} +deps = + pytest + pytest-cov + +[testenv:flake8] +commands = flake8 {posargs} +deps = flake8 diff --git a/test-requirements.txt b/test-requirements.txt index 75e9e81fa..e2443825c 100644 --- a/test-requirements.txt +++ b/test-requirements.txt @@ -1,7 +1,8 @@ coverage[toml] +basedpyright==1.39.9 ; python_version >= "3.9" and sys_platform != "cygwin" ddt >= 1.1.1, != 1.4.3 mock ; python_version < "3.8" -mypy +mypy==1.18.2 ; python_version >= "3.9" # pin mypy version to avoid new errors pre-commit pytest >= 7.3.1 pytest-cov diff --git a/test/deprecation/test_remote_progress.py b/test/deprecation/test_remote_progress.py new file mode 100644 index 000000000..a65739d31 --- /dev/null +++ b/test/deprecation/test_remote_progress.py @@ -0,0 +1,19 @@ +# This module is part of GitPython and is released under the +# 3-Clause BSD License: https://opensource.org/license/bsd-3-clause/ + +"""Tests for static characteristics of remote progress helpers.""" + +from typing import AnyStr + +from git import RemoteProgress + + +def test_parse_progress_line_override_can_return_non_none() -> None: + """Subclass overrides may return a value without violating base typing.""" + + class ReturningProgress(RemoteProgress): + def _parse_progress_line(self, line: AnyStr) -> str: + super()._parse_progress_line(line) + return "handled" + + assert ReturningProgress()._parse_progress_line("unmatched progress") == "handled" diff --git a/test/deprecation/test_types.py b/test/deprecation/test_types.py index f97375a85..d3c6af645 100644 --- a/test/deprecation/test_types.py +++ b/test/deprecation/test_types.py @@ -36,7 +36,7 @@ def test_can_access_lit_commit_ish_but_it_is_not_usable() -> None: assert 'Literal["commit", "tag"]' in message, "Has new definition." assert "GitObjectTypeString" in message, "Has new type name for old definition." - _: Lit_commit_ish = "commit" # type: ignore[valid-type] + _: Lit_commit_ish = "commit" # It should be as documented (even though deliberately unusable in static checks). assert Lit_commit_ish == Literal["commit", "tag"] diff --git a/test/fixtures/git_config_with_empty_quotes b/test/fixtures/git_config_with_empty_quotes new file mode 100644 index 000000000..f11fe4248 --- /dev/null +++ b/test/fixtures/git_config_with_empty_quotes @@ -0,0 +1,2 @@ +[core] + filemode = "" diff --git a/test/fixtures/git_config_with_extra_whitespace b/test/fixtures/git_config_with_extra_whitespace new file mode 100644 index 000000000..0f727cb5d --- /dev/null +++ b/test/fixtures/git_config_with_extra_whitespace @@ -0,0 +1,2 @@ +[init] + defaultBranch = trunk diff --git a/test/fixtures/git_config_with_quotes b/test/fixtures/git_config_with_quotes new file mode 100644 index 000000000..40e6710d9 --- /dev/null +++ b/test/fixtures/git_config_with_quotes @@ -0,0 +1,3 @@ +[user] + name = "Cody Veal" + email = "cveal05@gmail.com" diff --git a/test/fixtures/git_config_with_quotes_escapes b/test/fixtures/git_config_with_quotes_escapes new file mode 100644 index 000000000..33332c221 --- /dev/null +++ b/test/fixtures/git_config_with_quotes_escapes @@ -0,0 +1,9 @@ +[custom] + hasnewline = "first\nsecond" + hasbackslash = "foo\\bar" + hasquote = "ab\"cd" + hastrailingbackslash = "word\\" + hasunrecognized = "p\qrs" + hasunescapedquotes = "ab"cd"e" + ordinary = "hello world" + unquoted = good evening diff --git a/test/fixtures/git_config_with_quotes_whitespace_inside b/test/fixtures/git_config_with_quotes_whitespace_inside new file mode 100644 index 000000000..c6014cc61 --- /dev/null +++ b/test/fixtures/git_config_with_quotes_whitespace_inside @@ -0,0 +1,2 @@ +[core] + commentString = "# " diff --git a/test/fixtures/git_config_with_quotes_whitespace_outside b/test/fixtures/git_config_with_quotes_whitespace_outside new file mode 100644 index 000000000..4b1615a51 --- /dev/null +++ b/test/fixtures/git_config_with_quotes_whitespace_outside @@ -0,0 +1,2 @@ +[init] + defaultBranch = "trunk" diff --git a/test/fixtures/index_extended_flags b/test/fixtures/index_extended_flags new file mode 100644 index 000000000..f03713b68 Binary files /dev/null and b/test/fixtures/index_extended_flags differ diff --git a/test/lib/helper.py b/test/lib/helper.py index 5d91447ea..58923eaef 100644 --- a/test/lib/helper.py +++ b/test/lib/helper.py @@ -10,6 +10,7 @@ "with_rw_directory", "with_rw_repo", "with_rw_and_rw_remote_repo", + "PathLikeMock", "TestBase", "VirtualEnvironment", "TestCase", @@ -17,15 +18,20 @@ "skipIf", "GIT_REPO", "GIT_DAEMON_PORT", + "xfail_if_raises", + "symlinks_supported", + "requires_symlinks", ] import contextlib +from dataclasses import dataclass from functools import wraps import gc import io import logging import os import os.path as osp +from stat import S_ISLNK, ST_MODE import subprocess import sys import tempfile @@ -33,8 +39,10 @@ import time import unittest import venv +from typing import Union, Type, Tuple import gitdb +import pytest from git.util import rmtree, cwd @@ -49,6 +57,15 @@ _logger = logging.getLogger(__name__) + +@dataclass +class PathLikeMock: + path: str + + def __fspath__(self) -> str: + return self.path + + # { Routines @@ -76,7 +93,7 @@ def __init__(self, input_string): self.stdout = io.BytesIO(input_string) self.stderr = io.BytesIO() - def wait(self): + def wait(self, stderr=None): return 0 poll = wait @@ -149,7 +166,7 @@ def repo_creator(self): os.chdir(rw_repo.working_dir) try: return func(self, rw_repo) - except: # noqa: E722 B001 + except: # noqa: E722 _logger.info("Keeping repo after failure: %s", repo_dir) repo_dir = None raise @@ -309,7 +326,7 @@ def remote_repo_creator(self): with cwd(rw_repo.working_dir): try: return func(self, rw_repo, rw_daemon_repo) - except: # noqa: E722 B001 + except: # noqa: E722 _logger.info( "Keeping repos after failure: \n rw_repo_dir: %s \n rw_daemon_repo_dir: %s", rw_repo_dir, @@ -368,7 +385,13 @@ def _small_repo_url(self): """:return: A path to a small, clonable repository""" from git.cmd import Git - return Git.polish_url(osp.join(self.rorepo.working_tree_dir, "git/ext/gitdb/gitdb/ext/smmap")) + return Git.polish_url(self._dependency_repo_dirs["smmap"]) + + def _gitdb_repo_url(self): + """:return: A local gitdb repository reconstructed from merged history""" + from git.cmd import Git + + return Git.polish_url(self._dependency_repo_dirs["gitdb"]) @classmethod def setUpClass(cls): @@ -380,11 +403,26 @@ def setUpClass(cls): gc.collect() cls.rorepo = Repo(GIT_REPO) + cls._dependency_repo_root = tempfile.mkdtemp(prefix="gitpython-dependency-repos-") + cls._dependency_repo_dirs = {} + for name, rev in { + "gitdb": "2da3232f9d58e7761e384ac6d32f7b1ed77a74a2", + "smmap": "8ce61ad5cc4016bffaf25080bc0d69b3acbe8555", + }.items(): + path = osp.join(cls._dependency_repo_root, name) + repo = cls.rorepo.clone(path, shared=True, no_checkout=True) + repo.create_head("master", repo.commit(rev), force=True).checkout() + if name == "smmap": + repo.create_tag("v0.8.1", ref="master~10", message="Test fixture tag", force=True) + repo.git.gc() + repo.close() + cls._dependency_repo_dirs[name] = path @classmethod def tearDownClass(cls): cls.rorepo.git.clear_cache() cls.rorepo.git = None + rmtree(cls._dependency_repo_root) def _make_file(self, rela_path, data, repo=None): """ @@ -415,9 +453,15 @@ def __init__(self, env_dir, *, with_pip): if with_pip: # The upgrade_deps parameter to venv.create is 3.9+ only, so do it this way. - command = [self.python, "-m", "pip", "install", "--upgrade", "pip"] - if sys.version_info < (3, 12): - command.append("setuptools") + command = [ + self.python, + "-m", + "pip", + "install", + "--upgrade", + "pip", + 'setuptools; python_version<"3.12"', + ] subprocess.check_output(command) @property @@ -448,3 +492,49 @@ def _executable(self, basename): if osp.isfile(path) or osp.islink(path): return path raise RuntimeError(f"no regular file or symlink {path!r}") + + +def symlinks_supported() -> bool: + """Check whether this process can actually create a symlink. + + On Windows the platform alone doesn't decide it: creating a symlink needs either + Developer Mode or SeCreateSymbolicLinkPrivilege, and an unprivileged process gets + OSError (WinError 1314) instead. + """ + with tempfile.TemporaryDirectory(prefix="gitpython-symlink-check-") as temp_dir: + link_path = osp.join(temp_dir, "link") + try: + os.symlink("missing-target", link_path) + except (NotImplementedError, OSError): + return False + return S_ISLNK(os.lstat(link_path)[ST_MODE]) + + +requires_symlinks = pytest.mark.skipif( + not symlinks_supported(), + reason="symlinks are unavailable, or need privileges this process doesn't have", +) + + +@contextlib.contextmanager +def xfail_if_raises( + condition: bool, + *, + raises: Union[Type[BaseException], Tuple[Type[BaseException], ...]], + reason: str = "", + strict: bool = False, +): + """Approximates the behavior of @pytest.mark.xfail(..., raises=...) as a context + manager that can be used within a test, such as when the condition is complex or has + side effects + + One difference is it will not report XPASS if the test passes, but setting `strict` + simulates it by raising an exception""" + try: + yield + except raises: + if condition: + pytest.xfail(reason) + raise + if strict and condition: + pytest.fail("[XPASS(strict)] " + reason) diff --git a/test/test_autointerrupt.py b/test/test_autointerrupt.py new file mode 100644 index 000000000..645ec402c --- /dev/null +++ b/test/test_autointerrupt.py @@ -0,0 +1,33 @@ +from git.cmd import Git + + +class _DummyProc: + """Minimal stand-in for subprocess.Popen used to exercise AutoInterrupt. + + We deliberately raise AttributeError from terminate() to simulate interpreter + shutdown on Windows where subprocess internals (e.g. subprocess._winapi) may + already be torn down. + """ + + stdin = None + stdout = None + stderr = None + + def poll(self): + return None + + def terminate(self): + raise AttributeError("TerminateProcess") + + def wait(self): # pragma: no cover - should not be reached in this test + raise AssertionError("wait() should not be called if terminate() fails") + + +def test_autointerrupt_terminate_ignores_attributeerror(): + ai = Git.AutoInterrupt(_DummyProc(), args=["git", "rev-list"]) + + # Should not raise, even if terminate() triggers AttributeError. + ai._terminate() + + # Ensure the reference is cleared to avoid repeated attempts. + assert ai.proc is None diff --git a/test/test_clone.py b/test/test_clone.py index 126ef0063..5af67613a 100644 --- a/test/test_clone.py +++ b/test/test_clone.py @@ -1,12 +1,24 @@ # This module is part of GitPython and is released under the # 3-Clause BSD License: https://opensource.org/license/bsd-3-clause/ +import os +import os.path as osp +import pathlib +import sys +import tempfile +from unittest import skip +from unittest import mock + +from git import Git, GitCommandError, Repo +from git.exc import UnsafeOptionError, UnsafeProtocolError + +from test.lib import TestBase, with_rw_directory, with_rw_repo, PathLikeMock + from pathlib import Path import re import git - -from test.lib import TestBase, with_rw_directory +import pytest class TestClone(TestBase): @@ -29,3 +41,393 @@ def test_checkout_in_non_empty_dir(self, rw_dir): ) else: self.fail("GitCommandError not raised") + + @with_rw_directory + def test_clone_from_pathlib(self, rw_dir): + original_repo = Repo.init(osp.join(rw_dir, "repo")) + + Repo.clone_from(pathlib.Path(original_repo.git_dir), pathlib.Path(rw_dir) / "clone_pathlib") + + @with_rw_directory + def test_clone_from_pathlike(self, rw_dir): + original_repo = Repo.init(osp.join(rw_dir, "repo")) + Repo.clone_from(PathLikeMock(original_repo.git_dir), PathLikeMock(os.path.join(rw_dir, "clone_pathlike"))) + + @with_rw_directory + def test_clone_from_pathlib_withConfig(self, rw_dir): + original_repo = Repo.init(osp.join(rw_dir, "repo")) + + cloned = Repo.clone_from( + original_repo.git_dir, + pathlib.Path(rw_dir) / "clone_pathlib_withConfig", + multi_options=[ + "--recurse-submodules=repo", + "--config core.filemode=false", + "--config submodule.repo.update=checkout", + "--config filter.lfs.clean='git-lfs clean -- %f'", + ], + allow_unsafe_options=True, + ) + + self.assertEqual(cloned.config_reader().get_value("submodule", "active"), "repo") + self.assertEqual(cloned.config_reader().get_value("core", "filemode"), False) + self.assertEqual(cloned.config_reader().get_value('submodule "repo"', "update"), "checkout") + self.assertEqual( + cloned.config_reader().get_value('filter "lfs"', "clean"), + "git-lfs clean -- %f", + ) + + def test_clone_from_with_path_contains_unicode(self): + with tempfile.TemporaryDirectory() as tmpdir: + unicode_dir_name = "\u0394" + path_with_unicode = os.path.join(tmpdir, unicode_dir_name) + os.makedirs(path_with_unicode) + + try: + Repo.clone_from( + url=self._small_repo_url(), + to_path=path_with_unicode, + # Local clones hardlink the reconstructed smmap repository's + # read-only metadata, which Windows cannot remove at cleanup. + no_hardlinks=True, + ) + except UnicodeEncodeError: + self.fail("Raised UnicodeEncodeError") + + @with_rw_directory + @skip( + """The referenced repository was removed, and one needs to set up a new + password controlled repo under the org's control.""" + ) + def test_leaking_password_in_clone_logs(self, rw_dir): + password = "fakepassword1234" + try: + Repo.clone_from( + url="https://fakeuser:{}@fakerepo.example.com/testrepo".format(password), + to_path=rw_dir, + ) + except GitCommandError as err: + assert password not in str(err), "The error message '%s' should not contain the password" % err + # Working example from a blank private project. + Repo.clone_from( + url="https://gitlab+deploy-token-392045:mLWhVus7bjLsy8xj8q2V@gitlab.com/mercierm/test_git_python", + to_path=rw_dir, + ) + + @with_rw_repo("HEAD") + def test_clone_unsafe_options(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + tmp_file = tmp_dir / "pwn" + unsafe_options = [ + f"--upload-pack='touch {tmp_file}'", + f"--upl='touch {tmp_file}'", + f"-u 'touch {tmp_file}'", + f"-utouch {tmp_file}; false", + f"-futouch${{IFS}}{tmp_file}; false", + f"-qutouch${{IFS}}{tmp_file}; false", + "--config=protocol.ext.allow=always", + "--conf=protocol.ext.allow=always", + "-c protocol.ext.allow=always", + "-cprotocol.ext.allow=always", + "-vcprotocol.ext.allow=always", + f"--template={tmp_dir}", + f"--bundle-uri=file://{tmp_dir}", + ] + for unsafe_option in unsafe_options: + with self.assertRaises(UnsafeOptionError): + rw_repo.clone(tmp_dir, multi_options=[unsafe_option]) + assert not tmp_file.exists() + + unsafe_options = [ + {"upload-pack": f"touch {tmp_file}"}, + {"upload_pack": f"touch {tmp_file}"}, + {"upl": f"touch {tmp_file}"}, + {"u": f"touch {tmp_file}"}, + {"config": "protocol.ext.allow=always"}, + {"conf": "protocol.ext.allow=always"}, + {"c": "protocol.ext.allow=always"}, + {"template": tmp_dir}, + {"bundle_uri": f"file://{tmp_dir}"}, + ] + for unsafe_option in unsafe_options: + with self.assertRaises(UnsafeOptionError): + rw_repo.clone(tmp_dir, **unsafe_option) + assert not tmp_file.exists() + + @with_rw_repo("HEAD") + def test_clone_unsafe_options_abbreviated(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + tmp_file = tmp_dir / "pwn" + unsafe_options = [ + f"--upl='touch {tmp_file}'", + f"--upload-pac='touch {tmp_file}'", + "--conf=protocol.ext.allow=always", + ] + for unsafe_option in unsafe_options: + with self.assertRaises(UnsafeOptionError): + rw_repo.clone(tmp_dir, multi_options=[unsafe_option]) + assert not tmp_file.exists() + + unsafe_kwargs = [ + {"upl": f"touch {tmp_file}"}, + {"upload_pac": f"touch {tmp_file}"}, + {"conf": "protocol.ext.allow=always"}, + ] + for unsafe_option in unsafe_kwargs: + with self.assertRaises(UnsafeOptionError): + rw_repo.clone(tmp_dir, **unsafe_option) + assert not tmp_file.exists() + + @with_rw_repo("HEAD") + def test_clone_unsafe_options_are_checked_after_splitting_multi_options(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + payload = "--single-branch --config protocol.ext.allow=always" + + with self.assertRaises(UnsafeOptionError): + rw_repo.clone(tmp_dir, multi_options=[payload]) + + @pytest.mark.xfail( + sys.platform == "win32", + reason=( + "File not created. A separate Windows command may be needed. This and the " + "currently passing test test_clone_unsafe_options must be adjusted in the " + "same way. Until then, test_clone_unsafe_options is unreliable on Windows." + ), + raises=AssertionError, + ) + @with_rw_repo("HEAD") + def test_clone_unsafe_options_allowed(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + tmp_file = tmp_dir / "pwn" + unsafe_options = [ + f"--upload-pack='touch {tmp_file}'", + f"-u 'touch {tmp_file}'", + ] + for i, unsafe_option in enumerate(unsafe_options): + destination = tmp_dir / str(i) + assert not tmp_file.exists() + # The options will be allowed, but the command will fail. + with self.assertRaises(GitCommandError): + rw_repo.clone(destination, multi_options=[unsafe_option], allow_unsafe_options=True) + assert tmp_file.exists() + tmp_file.unlink() + + unsafe_options = [ + "--config=protocol.ext.allow=always", + "-c protocol.ext.allow=always", + ] + for i, unsafe_option in enumerate(unsafe_options): + destination = tmp_dir / str(i) + assert not destination.exists() + rw_repo.clone(destination, multi_options=[unsafe_option], allow_unsafe_options=True) + assert destination.exists() + + @with_rw_repo("HEAD") + def test_clone_safe_options(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + options = [ + "--depth=1", + "--single-branch", + "--origin upload", + "-q", + "-oupstream", + ] + for option in options: + destination = tmp_dir / option + assert not destination.exists() + rw_repo.clone(destination, multi_options=[option]) + assert destination.exists() + + @with_rw_repo("HEAD") + def test_clone_from_unsafe_options(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + tmp_file = tmp_dir / "pwn" + unsafe_options = [ + f"--upload-pack='touch {tmp_file}'", + f"-u 'touch {tmp_file}'", + f"-utouch {tmp_file}; false", + f"-futouch${{IFS}}{tmp_file}; false", + f"-qutouch${{IFS}}{tmp_file}; false", + "--config=protocol.ext.allow=always", + "-c protocol.ext.allow=always", + "-cprotocol.ext.allow=always", + "-vcprotocol.ext.allow=always", + ] + for unsafe_option in unsafe_options: + with self.assertRaises(UnsafeOptionError): + Repo.clone_from(rw_repo.working_dir, tmp_dir, multi_options=[unsafe_option]) + assert not tmp_file.exists() + + unsafe_options = [ + {"upload-pack": f"touch {tmp_file}"}, + {"upload_pack": f"touch {tmp_file}"}, + {"u": f"touch {tmp_file}"}, + {"config": "protocol.ext.allow=always"}, + {"c": "protocol.ext.allow=always"}, + ] + for unsafe_option in unsafe_options: + with self.assertRaises(UnsafeOptionError): + Repo.clone_from(rw_repo.working_dir, tmp_dir, **unsafe_option) + assert not tmp_file.exists() + + @with_rw_repo("HEAD") + def test_clone_from_unsafe_options_are_checked_after_splitting_multi_options(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + payload = "--single-branch --config protocol.ext.allow=always" + + with self.assertRaises(UnsafeOptionError): + Repo.clone_from(rw_repo.working_dir, tmp_dir, multi_options=[payload]) + + @pytest.mark.xfail( + sys.platform == "win32", + reason=( + "File not created. A separate Windows command may be needed. This and the " + "currently passing test test_clone_from_unsafe_options must be adjusted in the " + "same way. Until then, test_clone_from_unsafe_options is unreliable on Windows." + ), + raises=AssertionError, + ) + @with_rw_repo("HEAD") + def test_clone_from_unsafe_options_allowed(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + tmp_file = tmp_dir / "pwn" + unsafe_options = [ + f"--upload-pack='touch {tmp_file}'", + f"-u 'touch {tmp_file}'", + ] + for i, unsafe_option in enumerate(unsafe_options): + destination = tmp_dir / str(i) + assert not tmp_file.exists() + # The options will be allowed, but the command will fail. + with self.assertRaises(GitCommandError): + Repo.clone_from( + rw_repo.working_dir, destination, multi_options=[unsafe_option], allow_unsafe_options=True + ) + assert tmp_file.exists() + tmp_file.unlink() + + unsafe_options = [ + "--config=protocol.ext.allow=always", + "-c protocol.ext.allow=always", + ] + for i, unsafe_option in enumerate(unsafe_options): + destination = tmp_dir / str(i) + assert not destination.exists() + Repo.clone_from( + rw_repo.working_dir, destination, multi_options=[unsafe_option], allow_unsafe_options=True + ) + assert destination.exists() + + @with_rw_repo("HEAD") + def test_clone_from_safe_options(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + options = [ + "--depth=1", + "--single-branch", + "-q", + ] + for option in options: + destination = tmp_dir / option + assert not destination.exists() + Repo.clone_from(rw_repo.common_dir, destination, multi_options=[option]) + assert destination.exists() + + def test_clone_from_unsafe_protocol(self): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + tmp_file = tmp_dir / "pwn" + urls = [ + f"ext::sh -c touch% {tmp_file}", + "fd::17/foo", + ] + for url in urls: + with self.assertRaises(UnsafeProtocolError): + Repo.clone_from(url, tmp_dir / "repo") + assert not tmp_file.exists() + + def test_clone_from_does_not_expand_environment_variables_in_url(self): + urls = [ + "https://example.com/$GITPYTHON_TEST_SECRET/repo.git", + "https://example.com/${GITPYTHON_TEST_SECRET}/repo.git", + "https://example.com/%GITPYTHON_TEST_SECRET%/repo.git", + ] + with mock.patch.dict(os.environ, {"GITPYTHON_TEST_SECRET": "sensitive-value"}): + for url in urls: + with mock.patch.object(Git, "_call_process", side_effect=RuntimeError) as call_process: + with self.assertRaises(RuntimeError): + Repo.clone_from(url, "unused") + + assert call_process.call_args[0][3] == url + + @with_rw_directory + def test_clone_from_does_not_expand_environment_variables_in_stored_url(self, rw_dir): + url = pathlib.Path(rw_dir) / "$GITPYTHON_TEST_SECRET" / "source" + Git().init(url) + + with mock.patch.dict(os.environ, {"GITPYTHON_TEST_SECRET": "sensitive-value"}): + cloned = Repo.clone_from(url, pathlib.Path(rw_dir) / "clone") + + assert cloned.remotes.origin.url == Git.polish_url(str(url), expand_vars=False) + + def test_clone_from_checks_polished_url_for_unsafe_protocol(self): + with mock.patch.object(Git, "polish_url", return_value="ext::command"): + with mock.patch.object(Git, "_call_process") as call_process: + with self.assertRaises(UnsafeProtocolError): + Repo.clone_from("$GITPYTHON_TEST_URL", "unused") + + call_process.assert_not_called() + + def test_polish_url_does_not_expand_environment_variables_for_cygwin(self): + urls = ["$GITPYTHON_TEST_SECRET/repo", "user@example.com:$GITPYTHON_TEST_SECRET/repo"] + with mock.patch.dict(os.environ, {"GITPYTHON_TEST_SECRET": "sensitive-value"}): + for url in urls: + assert Git.polish_url(url, is_cygwin=True, expand_vars=False) == url + + def test_clone_from_unsafe_protocol_allowed(self): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + tmp_file = tmp_dir / "pwn" + urls = [ + f"ext::sh -c touch% {tmp_file}", + "fd::/foo", + ] + for url in urls: + # The URL will be allowed into the command, but the command will + # fail since we don't have that protocol enabled in the Git config file. + with self.assertRaises(GitCommandError): + Repo.clone_from(url, tmp_dir / "repo", allow_unsafe_protocols=True) + assert not tmp_file.exists() + + def test_clone_from_unsafe_protocol_allowed_and_enabled(self): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = pathlib.Path(tdir) + tmp_file = tmp_dir / "pwn" + urls = [ + f"ext::sh -c touch% {tmp_file}", + ] + allow_ext = [ + "--config=protocol.ext.allow=always", + ] + for url in urls: + # The URL will be allowed into the command, and the protocol is enabled, + # but the command will fail since it can't read from the remote repo. + assert not tmp_file.exists() + with self.assertRaises(GitCommandError): + Repo.clone_from( + url, + tmp_dir / "repo", + multi_options=allow_ext, + allow_unsafe_protocols=True, + allow_unsafe_options=True, + ) + assert tmp_file.exists() + tmp_file.unlink() diff --git a/test/test_commit.py b/test/test_commit.py index 37c66e3e7..431269b29 100644 --- a/test/test_commit.py +++ b/test/test_commit.py @@ -6,6 +6,7 @@ import copy from datetime import datetime from io import BytesIO +import tempfile import os.path as osp import re import sys @@ -15,6 +16,7 @@ from gitdb import IStream from git import Actor, Commit, Repo +from git.exc import UnsafeOptionError from git.objects.util import tzoffset, utc from git.repo.fun import touch @@ -274,20 +276,36 @@ def test_iteration(self): assert ltd_commits and len(ltd_commits) < len(all_commits) # Show commits of multiple paths, resulting in a union of commits. - less_ltd_commits = list(Commit.iter_items(self.rorepo, "master", paths=("CHANGES", "AUTHORS"))) + less_ltd_commits = list(Commit.iter_items(self.rorepo, "HEAD", paths=("CHANGES", "AUTHORS"))) assert len(ltd_commits) < len(less_ltd_commits) class Child(Commit): def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs) - child_commits = list(Child.iter_items(self.rorepo, "master", paths=("CHANGES", "AUTHORS"))) + child_commits = list(Child.iter_items(self.rorepo, "HEAD", paths=("CHANGES", "AUTHORS"))) assert type(child_commits[0]) is Child def test_iter_items(self): # pretty not allowed. self.assertRaises(ValueError, Commit.iter_items, self.rorepo, "master", pretty="raw") + def test_iter_items_rejects_unsafe_revision(self): + with tempfile.TemporaryDirectory() as tdir: + marker = osp.join(tdir, "pwn") + self.assertRaises(UnsafeOptionError, Commit.iter_items, self.rorepo, f"--output={marker}") + + def test_iter_items_rejects_unsafe_options(self): + with tempfile.TemporaryDirectory() as tdir: + marker = osp.join(tdir, "pwn") + with self.assertRaises(UnsafeOptionError): + list(Commit.iter_items(self.rorepo, "HEAD", output=marker)) + + def test_count_rejects_unsafe_options(self): + with tempfile.NamedTemporaryFile() as marker: + with self.assertRaises(UnsafeOptionError): + self.rorepo.head.commit.count(output=marker.name) + def test_rev_list_bisect_all(self): """ 'git rev-list --bisect-all' returns additional information @@ -309,6 +327,11 @@ def test_rev_list_bisect_all(self): for sha1, commit in zip(expected_ids, commits): self.assertEqual(sha1, commit.hexsha) + def test_iter_from_invalid_process_or_stream(self): + for source, error in ((Mock(wait=Mock(), stdout=None), ValueError), (object(), TypeError)): + with self.assertRaises(error): + list(Commit._iter_from_process_or_stream(self.rorepo, source)) + @with_rw_directory def test_ambiguous_arg_iteration(self, rw_dir): rw_repo = Repo.init(osp.join(rw_dir, "test_ambiguous_arg")) @@ -518,7 +541,7 @@ def test_trailers(self): ), ] for msg in msgs: - commit = copy.copy(self.rorepo.commit("master")) + commit = copy.copy(self.rorepo.commit("HEAD")) commit.message = msg assert commit.trailers_list == [ (KEY_1, VALUE_1_1), @@ -541,13 +564,13 @@ def test_trailers(self): ] for msg in msgs: - commit = copy.copy(self.rorepo.commit("master")) + commit = copy.copy(self.rorepo.commit("HEAD")) commit.message = msg assert commit.trailers_list == [] assert commit.trailers_dict == {} # Check that only the last key value paragraph is evaluated. - commit = copy.copy(self.rorepo.commit("master")) + commit = copy.copy(self.rorepo.commit("HEAD")) commit.message = f"Subject\n\nMultiline\nBody\n\n{KEY_1}: {VALUE_1_1}\n\n{KEY_2}: {VALUE_2}\n" assert commit.trailers_list == [(KEY_2, VALUE_2)] assert commit.trailers_dict == {KEY_2: [VALUE_2]} @@ -566,3 +589,142 @@ def test_commit_co_authors(self): Actor("test_user_2", "another_user-email@github.com"), Actor("test_user_3", "test_user_3@github.com"), ] + + @with_rw_directory + def test_create_from_tree_with_trailers_dict(self, rw_dir): + """Test that create_from_tree supports adding trailers via a dict.""" + rw_repo = Repo.init(osp.join(rw_dir, "test_trailers_dict")) + path = osp.join(str(rw_repo.working_tree_dir), "hello.txt") + touch(path) + rw_repo.index.add([path]) + tree = rw_repo.index.write_tree() + + trailers = {"Issue": "123", "Signed-off-by": "Test User "} + commit = Commit.create_from_tree( + rw_repo, + tree, + "Test commit with trailers", + head=True, + trailers=trailers, + ) + + assert "Issue: 123" in commit.message + assert "Signed-off-by: Test User " in commit.message + assert commit.trailers_dict == { + "Issue": ["123"], + "Signed-off-by": ["Test User "], + } + + @with_rw_directory + def test_create_from_tree_with_trailers_list(self, rw_dir): + """Test that create_from_tree supports adding trailers via a list of tuples.""" + rw_repo = Repo.init(osp.join(rw_dir, "test_trailers_list")) + path = osp.join(str(rw_repo.working_tree_dir), "hello.txt") + touch(path) + rw_repo.index.add([path]) + tree = rw_repo.index.write_tree() + + trailers = [ + ("Signed-off-by", "Alice "), + ("Signed-off-by", "Bob "), + ("Issue", "456"), + ] + commit = Commit.create_from_tree( + rw_repo, + tree, + "Test commit with multiple trailers", + head=True, + trailers=trailers, + ) + + assert "Signed-off-by: Alice " in commit.message + assert "Signed-off-by: Bob " in commit.message + assert "Issue: 456" in commit.message + assert commit.trailers_dict == { + "Signed-off-by": ["Alice ", "Bob "], + "Issue": ["456"], + } + + @with_rw_directory + def test_create_from_tree_with_non_utf8_trailers(self, rw_dir): + """Test that trailer creation and parsing respect the configured commit encoding.""" + rw_repo = Repo.init(osp.join(rw_dir, "test_trailers_non_utf8")) + with rw_repo.config_writer() as writer: + writer.set_value("i18n", "commitencoding", "ISO-8859-1") + + path = osp.join(str(rw_repo.working_tree_dir), "hello.txt") + touch(path) + rw_repo.index.add([path]) + tree = rw_repo.index.write_tree() + + commit = Commit.create_from_tree( + rw_repo, + tree, + "RÊsumÊ", + head=True, + trailers={"Reviewed-by": "AndrÊ "}, + ) + + assert commit.encoding == "ISO-8859-1" + assert "RÊsumÊ" in commit.message + assert "Reviewed-by: AndrÊ " in commit.message + assert commit.trailers_list == [("Reviewed-by", "AndrÊ ")] + + @with_rw_directory + def test_trailers_list_with_non_utf8_message_bytes(self, rw_dir): + """Test that trailer parsing handles non-UTF-8 commit message bytes.""" + rw_repo = Repo.init(osp.join(rw_dir, "test_trailers_non_utf8_bytes")) + with rw_repo.config_writer() as writer: + writer.set_value("i18n", "commitencoding", "ISO-8859-1") + + path = osp.join(str(rw_repo.working_tree_dir), "hello.txt") + touch(path) + rw_repo.index.add([path]) + tree = rw_repo.index.write_tree() + + commit = Commit.create_from_tree( + rw_repo, + tree, + "RÊsumÊ", + head=True, + trailers={"Reviewed-by": "AndrÊ "}, + ) + + bytes_commit = Commit( + rw_repo, + commit.binsha, + message=commit.message.encode(commit.encoding), + encoding=commit.encoding, + ) + + assert bytes_commit.trailers_list == [("Reviewed-by", "AndrÊ ")] + + def test_interpret_trailers_encodes_before_launching_process(self): + """Test that encoding failures happen before spawning interpret-trailers.""" + repo = Mock() + repo.git = Mock() + repo.git.GIT_PYTHON_GIT_EXECUTABLE = "git" + + with self.assertRaises(UnicodeEncodeError): + Commit._interpret_trailers(repo, "Euro: â‚Ŧ", ["--parse"], encoding="ISO-8859-1") + + repo.git.execute.assert_not_called() + + @with_rw_directory + def test_index_commit_with_trailers(self, rw_dir): + """Test that IndexFile.commit() supports adding trailers.""" + rw_repo = Repo.init(osp.join(rw_dir, "test_index_trailers")) + path = osp.join(str(rw_repo.working_tree_dir), "hello.txt") + touch(path) + rw_repo.index.add([path]) + + trailers = {"Reviewed-by": "Reviewer "} + commit = rw_repo.index.commit( + "Test index commit with trailers", + trailers=trailers, + ) + + assert "Reviewed-by: Reviewer " in commit.message + assert commit.trailers_dict == { + "Reviewed-by": ["Reviewer "], + } diff --git a/test/test_config.py b/test/test_config.py index 92997422d..fd0d347a4 100644 --- a/test/test_config.py +++ b/test/test_config.py @@ -14,7 +14,7 @@ from git import GitConfigParser from git.config import _OMD, cp -from git.util import rmfile +from git.util import cwd, rmfile from test.lib import SkipTest, TestCase, fixture_path, with_rw_directory @@ -150,6 +150,154 @@ def test_config_value_with_trailing_new_line(self): git_config = GitConfigParser(config_file) git_config.read() # This should not throw an exception + @with_rw_directory + def test_set_value_rejects_config_injection(self, rw_dir): + config_path = osp.join(rw_dir, "config") + payload = "foo\n[core]\nhooksPath=/tmp/hooks" + + with GitConfigParser(config_path, read_only=False) as git_config: + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.set_value("user", "name", payload) + + with GitConfigParser(config_path, read_only=True) as git_config: + self.assertFalse(git_config.has_section("user")) + self.assertFalse(git_config.has_section("core")) + + @with_rw_directory + def test_set_value_rejects_unsafe_section_and_option_names(self, rw_dir): + config_path = osp.join(rw_dir, "config") + bad_keys = ("user]\n[core", "user]\r[core", "user]\x00[core") + + with GitConfigParser(config_path, read_only=False) as git_config: + git_config.add_section("user") + for bad_key in bad_keys: + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.add_section(bad_key) + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.set(bad_key, "hooksPath", "/tmp/hooks") + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.set("user", bad_key, "/tmp/hooks") + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.set_value(bad_key, "hooksPath", "/tmp/hooks") + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.set_value("user", bad_key, "/tmp/hooks") + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.add_value(bad_key, "hooksPath", "/tmp/hooks") + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.add_value("user", bad_key, "/tmp/hooks") + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.rename_section("user", bad_key) + + git_config.set_value("user", "name", "safe") + + with GitConfigParser(config_path, read_only=True) as git_config: + self.assertEqual(git_config.get_value("user", "name"), "safe") + self.assertFalse(git_config.has_section("core")) + + @with_rw_directory + def test_writer_rejects_invalid_option_names(self, rw_dir): + config_path = osp.join(rw_dir, "config") + bad_options = ( + "name=value", + "name#comment", + "name;comment", + "name with space", + "name\twith-tab", + "name[section", + "name]section", + "name:colon", + 'name"quote', + "name\\escape", + ) + + with GitConfigParser(config_path, read_only=False) as git_config: + git_config.add_section("user") + for bad_option in bad_options: + with pytest.raises(ValueError, match="option name"): + git_config.set("user", bad_option, "unsafe") + with pytest.raises(ValueError, match="option name"): + git_config.set_value("user", bad_option, "unsafe") + with pytest.raises(ValueError, match="option name"): + git_config.add_value("user", bad_option, "unsafe") + + git_config.set_value("user", "safe-option1", "safe") + git_config.set_value("user", "safe_option2", "safe") + git_config.set_value("user", "3safe_option", "safe") + git_config.set_value("user", "safe.option3", "safe") + + with GitConfigParser(config_path, read_only=True) as git_config: + self.assertEqual(git_config.get_value("user", "safe-option1"), "safe") + self.assertEqual(git_config.get_value("user", "safe_option2"), "safe") + self.assertEqual(git_config.get_value("user", "3safe_option"), "safe") + self.assertEqual(git_config.get_value("user", "safe.option3"), "safe") + + @with_rw_directory + def test_writer_rejects_unquoted_section_terminators(self, rw_dir): + config_path = osp.join(rw_dir, "config") + bad_sections = ( + "user] [other", + 'user"] [other"', + 'submodule "docs"] [other', + 'submodule "docs] [other', + 'submodule "docs] [other\\', + ) + safe_section = 'submodule "docs]archive"' + + with GitConfigParser(config_path, read_only=False) as git_config: + git_config.add_section("user") + for bad_section in bad_sections: + with pytest.raises(ValueError, match="section name"): + git_config.add_section(bad_section) + with pytest.raises(ValueError, match="section name"): + git_config.set(bad_section, "name", "value") + with pytest.raises(ValueError, match="section name"): + git_config.set_value(bad_section, "name", "value") + with pytest.raises(ValueError, match="section name"): + git_config.add_value(bad_section, "name", "value") + with pytest.raises(ValueError, match="section name"): + git_config.rename_section("user", bad_section) + + git_config.set_value("user", "name", "safe") + git_config.set_value(safe_section, "name", "safe") + self.assertEqual(git_config.get_value(safe_section, "name"), "safe") + + # A closing bracket inside a quoted subsection name is data, not a section terminator. + with open(config_path, "rb") as config_file: + self.assertIn( + b'[submodule "docs]archive"]\n', + config_file.read(), + "a closing bracket within a quoted subsection name should be preserved", + ) + + # Reparse the file to verify that rejected names did not inject an [other] section. + with GitConfigParser(config_path, read_only=True) as git_config: + self.assertEqual( + git_config.get_value("user", "name"), + "safe", + "rejected section names corrupted the existing section", + ) + self.assertFalse(git_config.has_section("other"), "an unsafe section name injected an [other] section") + + @with_rw_directory + def test_set_and_add_value_reject_unsafe_value_characters(self, rw_dir): + config_path = osp.join(rw_dir, "config") + bad_values = ("foo\rbar", "foo\nbar", "foo\x00bar", b"foo\nbar") + + with GitConfigParser(config_path, read_only=False) as git_config: + git_config.add_section("user") + for bad_value in bad_values: + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.set("user", "name", bad_value) + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.set_value("user", "name", bad_value) + with pytest.raises(ValueError, match="CR, LF, or NUL"): + git_config.add_value("user", "name", bad_value) + + git_config.set_value("user", "name", "safe") + + with GitConfigParser(config_path, read_only=True) as git_config: + self.assertEqual(git_config.get_value("user", "name"), "safe") + def test_base(self): path_repo = fixture_path("git_config") path_global = fixture_path("git_config_global") @@ -246,15 +394,81 @@ def check_test_value(cr, value): with GitConfigParser(fpa, read_only=True) as cr: check_test_value(cr, tv) - @pytest.mark.xfail( - sys.platform == "win32", - reason='Second config._has_includes() assertion fails (for "config is included if path is matching git_dir")', - raises=AssertionError, - ) + @with_rw_directory + def test_config_relative_path_include(self, rw_dir): + included_path = osp.join(rw_dir, "included") + with GitConfigParser(included_path, read_only=False) as cw: + cw.set_value("included", "value", "included") + + config_path = osp.join(rw_dir, "config") + with GitConfigParser(config_path, read_only=False) as cw: + cw.set_value("include", "path", "included") + + if osp.splitdrive(config_path)[0] != osp.splitdrive(os.getcwd())[0]: + pytest.skip("The temporary directory and checkout are on different drives") + + relative_config_path = osp.relpath(config_path) + with GitConfigParser(relative_config_path, read_only=True) as cr: + assert cr.get_value("included", "value") == "included" + + @pytest.mark.skipif(os.name != "nt", reason="Specifically for Windows drive-rooted paths.") + @with_rw_directory + def test_config_drive_rooted_path_include(self, rw_dir): + with cwd(rw_dir): + included_path = osp.join(rw_dir, "included") + with GitConfigParser(included_path, read_only=False) as cw: + cw.set_value("included", "value", "included") + + _drive, rooted_included_path = osp.splitdrive(included_path) + config_path = osp.join(rw_dir, "config") + with GitConfigParser(config_path, read_only=False) as cw: + cw.set_value("include", "path", rooted_included_path) + + with GitConfigParser(config_path, read_only=True) as cr: + assert cr.get_value("included", "value") == "included" + + @with_rw_directory + def test_multiple_include_paths_with_same_key(self, rw_dir): + """Test that multiple 'path' entries under [include] are all respected. + + Regression test for https://github.com/gitpython-developers/GitPython/issues/2099. + Git config allows multiple ``path`` values under ``[include]``, e.g.:: + + [include] + path = file1 + path = file2 + + Previously only one of these was included because _OMD.items() returns + only the last value for each key. + """ + # Create two config files to be included. + fp_inc1 = osp.join(rw_dir, "inc1.cfg") + fp_inc2 = osp.join(rw_dir, "inc2.cfg") + fp_main = osp.join(rw_dir, "main.cfg") + + with GitConfigParser(fp_inc1, read_only=False) as cw: + cw.set_value("user", "name", "from-inc1") + + with GitConfigParser(fp_inc2, read_only=False) as cw: + cw.set_value("core", "bar", "from-inc2") + + # Write a config with two path entries under a single [include] section. + # We write it manually because set_value would overwrite the key. + with open(fp_main, "w") as f: + f.write("[include]\n") + f.write(f"\tpath = {fp_inc1}\n") + f.write(f"\tpath = {fp_inc2}\n") + + with GitConfigParser(fp_main, read_only=True) as cr: + # Both included files should be loaded. + assert cr.get_value("user", "name") == "from-inc1" + assert cr.get_value("core", "bar") == "from-inc2" + @with_rw_directory def test_conditional_includes_from_git_dir(self, rw_dir): # Initiate repository path. git_dir = osp.join(rw_dir, "target1", "repo1") + git_dir_pattern = git_dir.replace("\\", "/") os.makedirs(git_dir) # Initiate mocked repository. @@ -266,6 +480,7 @@ def test_conditional_includes_from_git_dir(self, rw_dir): template = '[includeIf "{}:{}"]\n path={}\n' with open(path1, "w") as stream: + # on Windows, this writes a backslash pattern. stream.write(template.format("gitdir", git_dir, path2)) # Ensure that config is ignored if no repo is set. @@ -273,14 +488,26 @@ def test_conditional_includes_from_git_dir(self, rw_dir): assert not config._has_includes() assert config._included_paths() == [] - # Ensure that config is included if path is matching git_dir. + # Git uses forward slashes in gitdir patterns on every platform: + # backslashes escape the next pattern character rather than separate + # path components. On Windows, GitPython therefore normalizes git_dir + # to forward slashes but leaves this backslash pattern unchanged, so + # the two do not match and no path is included. + with GitConfigParser(path1, repo=repo, merge_includes=False) as config: + expected_paths = [] if sys.platform == "win32" else [("path", path2)] + assert config._included_paths() == expected_paths + + # Ensure that Git's forward-slash syntax matches native Windows paths. + with open(path1, "w") as stream: + stream.write(template.format("gitdir", git_dir_pattern, path2)) + with GitConfigParser(path1, repo=repo) as config: assert config._has_includes() assert config._included_paths() == [("path", path2)] # Ensure that config is ignored if case is incorrect. with open(path1, "w") as stream: - stream.write(template.format("gitdir", git_dir.upper(), path2)) + stream.write(template.format("gitdir", git_dir_pattern.upper(), path2)) with GitConfigParser(path1, repo=repo) as config: assert not config._has_includes() @@ -288,7 +515,7 @@ def test_conditional_includes_from_git_dir(self, rw_dir): # Ensure that config is included if case is ignored. with open(path1, "w") as stream: - stream.write(template.format("gitdir/i", git_dir.upper(), path2)) + stream.write(template.format("gitdir/i", git_dir_pattern.upper(), path2)) with GitConfigParser(path1, repo=repo) as config: assert config._has_includes() @@ -318,6 +545,20 @@ def test_conditional_includes_from_git_dir(self, rw_dir): assert config._has_includes() assert config._included_paths() == [("path", path2)] + @with_rw_directory + def test_conditional_includes_do_not_treat_backslashes_as_separators(self, rw_dir): + git_dir = osp.join(rw_dir, "target", "repo") + repo = mock.Mock(git_dir=git_dir) + config_path = osp.join(rw_dir, "config") + included_path = osp.join(rw_dir, "included") + pattern = git_dir.replace("\\", "/").replace("/target/repo", R"/target\repo") + + with open(config_path, "w") as stream: + stream.write(f'[includeIf "gitdir:{pattern}"]\n path={included_path}\n') + + with GitConfigParser(config_path, repo=repo, merge_includes=False) as config: + assert config._included_paths() == [] + @with_rw_directory def test_conditional_includes_from_branch_name(self, rw_dir): # Initiate mocked branch. @@ -373,6 +614,41 @@ def test_conditional_includes_from_branch_name_error(self, rw_dir): assert not config._has_includes() assert config._included_paths() == [] + @with_rw_directory + def test_conditional_includes_remote_url(self, rw_dir): + # Initiate mocked repository. + repo = mock.Mock() + repo.remotes = [mock.Mock(url="https://github.com/foo/repo")] + + # Initiate config files. + path1 = osp.join(rw_dir, "config1") + path2 = osp.join(rw_dir, "config2") + template = '[includeIf "hasconfig:remote.*.url:{}"]\n path={}\n' + + # Ensure that config with hasconfig and full url is correct. + with open(path1, "w") as stream: + stream.write(template.format("https://github.com/foo/repo", path2)) + + with GitConfigParser(path1, repo=repo) as config: + assert config._has_includes() + assert config._included_paths() == [("path", path2)] + + # Ensure that config with hasconfig and incorrect url is incorrect. + with open(path1, "w") as stream: + stream.write(template.format("incorrect", path2)) + + with GitConfigParser(path1, repo=repo) as config: + assert not config._has_includes() + assert config._included_paths() == [] + + # Ensure that config with hasconfig and url using glob pattern is correct. + with open(path1, "w") as stream: + stream.write(template.format("**/**github.com*/**", path2)) + + with GitConfigParser(path1, repo=repo) as config: + assert config._has_includes() + assert config._included_paths() == [("path", path2)] + def test_rename(self): file_obj = self._to_memcache(fixture_path("git_config")) with GitConfigParser(file_obj, read_only=False, merge_includes=False) as cw: @@ -391,13 +667,17 @@ def test_complex_aliases(self): with GitConfigParser(file_obj, read_only=False) as w_config: self.assertEqual( w_config.get("alias", "rbi"), - '"!g() { git rebase -i origin/${1:-master} ; } ; g"', + "!g() { git rebase -i origin/${1:-master} ; } ; g", ) self.assertEqual( file_obj.getvalue(), self._to_memcache(fixture_path(".gitconfig")).getvalue(), ) + def test_config_with_extra_whitespace(self): + cr = GitConfigParser(fixture_path("git_config_with_extra_whitespace"), read_only=True) + self.assertEqual(cr.get("init", "defaultBranch"), "trunk") + def test_empty_config_value(self): cr = GitConfigParser(fixture_path("git_config_with_empty_value"), read_only=True) @@ -406,6 +686,44 @@ def test_empty_config_value(self): with self.assertRaises(cp.NoOptionError): cr.get_value("color", "ui") + def test_config_with_quotes(self): + cr = GitConfigParser(fixture_path("git_config_with_quotes"), read_only=True) + + self.assertEqual(cr.get("user", "name"), "Cody Veal") + self.assertEqual(cr.get("user", "email"), "cveal05@gmail.com") + + def test_config_with_empty_quotes(self): + cr = GitConfigParser(fixture_path("git_config_with_empty_quotes"), read_only=True) + self.assertEqual(cr.get("core", "filemode"), "", "quotes can form a literal empty string as value") + + def test_config_with_quotes_with_literal_whitespace(self): + cr = GitConfigParser(fixture_path("git_config_with_quotes_whitespace_inside"), read_only=True) + self.assertEqual(cr.get("core", "commentString"), "# ") + + def test_config_with_quotes_with_whitespace_outside_value(self): + cr = GitConfigParser(fixture_path("git_config_with_quotes_whitespace_outside"), read_only=True) + self.assertEqual(cr.get("init", "defaultBranch"), "trunk") + + def test_config_with_quotes_containing_escapes(self): + """For now just suppress quote removal. But it would be good to interpret most of these.""" + cr = GitConfigParser(fixture_path("git_config_with_quotes_escapes"), read_only=True) + + # These can eventually be supported by substituting the represented character. + self.assertEqual(cr.get("custom", "hasnewline"), R'"first\nsecond"') + self.assertEqual(cr.get("custom", "hasbackslash"), R'"foo\\bar"') + self.assertEqual(cr.get("custom", "hasquote"), R'"ab\"cd"') + self.assertEqual(cr.get("custom", "hastrailingbackslash"), R'"word\\"') + self.assertEqual(cr.get("custom", "hasunrecognized"), R'"p\qrs"') + + # It is less obvious whether and what to eventually do with this. + self.assertEqual(cr.get("custom", "hasunescapedquotes"), '"ab"cd"e"') + + # Cases where quote removal is clearly safe should happen even after those. + self.assertEqual(cr.get("custom", "ordinary"), "hello world") + + # Cases without quotes should still parse correctly even after those, too. + self.assertEqual(cr.get("custom", "unquoted"), "good evening") + def test_get_values_works_without_requiring_any_other_calls_first(self): file_obj = self._to_memcache(fixture_path("git_config_multiple")) cr = GitConfigParser(file_obj, read_only=True) diff --git a/test/test_diff.py b/test/test_diff.py index 612fbd9e0..d5e14f3de 100644 --- a/test/test_diff.py +++ b/test/test_diff.py @@ -14,6 +14,7 @@ from git import NULL_TREE, Diff, DiffIndex, Diffable, GitCommandError, Repo, Submodule from git.cmd import Git +from git.exc import UnsafeOptionError from test.lib import StringProcessAdapter, TestBase, fixture, with_rw_directory @@ -91,6 +92,26 @@ def test_diff_with_staged_file(self, rw_dir): "This should work, but doesn't right now ... it's OK", ) + @with_rw_directory + def test_patch_diff_ignores_mnemonic_prefix(self, rw_dir): + repo = Repo.init(rw_dir) + filepath = osp.join(rw_dir, "file.txt") + with open(filepath, "w") as stream: + stream.write("initial\n") + repo.index.add([filepath]) + repo.index.commit("initial") + + with open(filepath, "w") as stream: + stream.write("changed\n") + repo.git.config("diff.mnemonicPrefix", "true") + + self.assertEqual(len(repo.head.commit.diff(None, create_patch=True)), 1) + self.assertEqual(len(repo.index.diff(None, create_patch=True)), 1) + + repo.index.add([filepath]) + self.assertEqual(len(repo.index.diff(NULL_TREE, create_patch=True)), 1) + self.assertEqual(repo.git.config("--get", "diff.mnemonicPrefix"), "true") + def test_list_from_string_new_mode(self): output = StringProcessAdapter(fixture("diff_new_mode")) diffs = Diff._index_from_patch_format(self.rorepo, output) @@ -352,6 +373,39 @@ def test_diff_submodule(self): self.assertIsInstance(diff.a_blob.size, int) self.assertIsInstance(diff.b_blob.size, int) + def test_diff_rejects_unsafe_output_options(self): + commit = self.rorepo.head.commit + + commit.diff(S="needle") + + calls = ( + lambda target: commit.diff(output=target), + lambda target: commit.diff(other=f"--output={target}"), + lambda target: commit.diff(O=target), + lambda target: commit.diff(orderfile=target), + lambda target: commit.diff(other=f"--orderfile={target}"), + lambda target: commit.diff(other=f"-pO{target}"), + lambda target: commit.diff(other=f"-uO{target}"), + lambda target: commit.diff(other=f"-DO{target}"), + lambda target: self.rorepo.index.diff(NULL_TREE, output=target), + lambda target: self.rorepo.index.diff(f"--output={target}"), + lambda target: self.rorepo.index.diff(NULL_TREE, O=target), + lambda target: self.rorepo.index.diff(NULL_TREE, orderfile=target), + lambda target: self.rorepo.index.diff(f"--orderfile={target}"), + lambda target: self.rorepo.index.diff(f"-pO{target}"), + lambda target: self.rorepo.index.diff(f"-uO{target}"), + lambda target: self.rorepo.index.diff(f"-DO{target}"), + ) + for index, call in enumerate(calls): + target = osp.join(self.repo_dir, f"diff-output-{index}") + with self.assertRaises(UnsafeOptionError): + call(target) + self.assertFalse(osp.exists(target)) + + allowed_target = osp.join(self.repo_dir, "allowed-diff-output") + commit.diff(output=allowed_target, allow_unsafe_options=True) + self.assertTrue(osp.isfile(allowed_target)) + def test_diff_interface(self): """Test a few variations of the main diff routine.""" assertion_map = {} diff --git a/test/test_docs.py b/test/test_docs.py index cc0bbf26a..0810d954d 100644 --- a/test/test_docs.py +++ b/test/test_docs.py @@ -6,12 +6,9 @@ import gc import os import os.path -import sys - -import pytest from test.lib import TestBase -from test.lib.helper import with_rw_directory +from test.lib.helper import with_rw_directory, with_rw_repo class Tutorials(TestBase): @@ -75,21 +72,20 @@ def test_init_repo_object(self, rw_dir): # heads, tags and references # heads are branches in git-speak # [8-test_init_repo_object] - self.assertEqual( - repo.head.ref, - repo.heads.master, # head is a sym-ref pointing to master. - "It's ok if TC not running from `master`.", - ) + active_branch = repo.active_branch + self.assertEqual(repo.head.ref, active_branch) # HEAD is a sym-ref pointing to the active branch. self.assertEqual(repo.tags["0.3.5"], repo.tag("refs/tags/0.3.5")) # You can access tags in various ways too. - self.assertEqual(repo.refs.master, repo.heads["master"]) # .refs provides all refs, i.e. heads... + self.assertEqual(repo.refs[active_branch.name], repo.heads[active_branch.name]) # .refs provides all refs... if "TRAVIS" not in os.environ: - self.assertEqual(repo.refs["origin/master"], repo.remotes.origin.refs.master) # ... remotes ... + remote_branch = next(ref for ref in repo.remotes.origin.refs if ref.remote_head != "HEAD") + self.assertEqual(repo.refs[remote_branch.name], remote_branch) # ...remotes... self.assertEqual(repo.refs["0.3.5"], repo.tags["0.3.5"]) # ... and tags. # ![8-test_init_repo_object] # Create a new head/branch. # [9-test_init_repo_object] + original_branch = cloned_repo.active_branch new_branch = cloned_repo.create_head("feature") # Create a new branch ... assert cloned_repo.active_branch != new_branch # which wasn't checked out yet ... self.assertEqual(new_branch.commit, cloned_repo.active_branch.commit) # pointing to the checked-out commit. @@ -149,10 +145,10 @@ def update(self, op_code, cur_count, max_count=None, message=""): assert origin.exists() for fetch_info in origin.fetch(progress=MyProgressPrinter()): print("Updated %s to %s" % (fetch_info.ref, fetch_info.commit)) - # Create a local branch at the latest fetched master. We specify the name - # statically, but you have all information to do it programmatically as well. - bare_master = bare_repo.create_head("master", origin.refs.master) - bare_repo.head.set_reference(bare_master) + # Create a local branch at one of the remote's branches. + remote_branch = next(ref for ref in origin.refs if ref.remote_head != "HEAD") + bare_branch = bare_repo.create_head(remote_branch.remote_head, remote_branch) + bare_repo.head.set_reference(bare_branch) assert not bare_repo.delete_remote(origin).exists() # push and pull behave very similarly. # ![12-test_init_repo_object] @@ -165,35 +161,39 @@ def update(self, op_code, cur_count, max_count=None, message=""): new_file_path = os.path.join(cloned_repo.working_tree_dir, "my-new-file") open(new_file_path, "wb").close() # Create new file in working tree. cloned_repo.index.add([new_file_path]) # Add it to the index. - # Commit the changes to deviate masters history. + # Commit the changes to deviate from the original branch's history. cloned_repo.index.commit("Added a new file in the past - for later merge") # Prepare a merge. - master = cloned_repo.heads.master # Right-hand side is ahead of us, in the future. - merge_base = cloned_repo.merge_base(new_branch, master) # Allows for a three-way merge. - cloned_repo.index.merge_tree(master, base=merge_base) # Write the merge result into index. + merge_base = cloned_repo.merge_base(new_branch, original_branch) # Allows for a three-way merge. + cloned_repo.index.merge_tree(original_branch, base=merge_base) # Write the merge result into index. cloned_repo.index.commit( "Merged past and now into future ;)", - parent_commits=(new_branch.commit, master.commit), + parent_commits=(new_branch.commit, original_branch.commit), ) - # Now new_branch is ahead of master, which probably should be checked out and reset softly. + # Now new_branch is ahead of the original branch, which probably should be checked out and reset softly. # Note that all these operations didn't touch the working tree, as we managed it ourselves. # This definitely requires you to know what you are doing! :) assert os.path.basename(new_file_path) in new_branch.commit.tree # New file is now in tree. - master.commit = new_branch.commit # Let master point to most recent commit. - cloned_repo.head.reference = master # We adjusted just the reference, not the working tree or index. + original_branch.commit = new_branch.commit # Let the original branch point to the most recent commit. + cloned_repo.head.reference = original_branch # We adjusted just the reference, not the working tree or index. # ![13-test_init_repo_object] # submodules # [14-test_init_repo_object] - # Create a new submodule and check it out on the spot, setup to track master - # branch of `bare_repo`. As our GitPython repository has submodules already that - # point to GitHub, make sure we don't interact with them. + # Create a new submodule and check it out on the spot, set up to track the + # default branch of `bare_repo`. As our GitPython repository has submodules + # already that point to GitHub, make sure we don't interact with them. for sm in cloned_repo.submodules: assert not sm.remove().exists() # after removal, the sm doesn't exist anymore - sm = cloned_repo.create_submodule("mysubrepo", "path/to/subrepo", url=bare_repo.git_dir, branch="master") + sm = cloned_repo.create_submodule( + "mysubrepo", + "path/to/subrepo", + url=bare_repo.git_dir, + branch=bare_branch.name, + ) # .gitmodules was written and added to the index, which is now being committed. cloned_repo.index.commit("Added submodule") @@ -478,18 +478,20 @@ def test_references_and_objects(self, rw_dir): repo.git.clear_cache() - @pytest.mark.xfail( - sys.platform == "cygwin", - reason="Cygwin GitPython can't find SHA for submodule", - raises=ValueError, - ) - def test_submodules(self): + @with_rw_repo("c15a6e1923a14bc760851913858a3942a4193cdb") + def test_submodules(self, repo): # [1-test_submodules] - repo = self.rorepo sms = repo.submodules assert len(sms) == 1 sm = sms[0] + with sm.config_writer() as writer: + writer.set_value("url", self._gitdb_repo_url()) + sm.update(recursive=False) + child = sm.children()[0] + with child.config_writer() as writer: + writer.set_value("url", self._small_repo_url()) + child.update() self.assertEqual(sm.name, "gitdb") # GitPython has gitdb as its one and only (direct) submodule... self.assertEqual(sm.children()[0].name, "smmap") # ...which has smmap as its one and only submodule. diff --git a/test/test_fixture_health.py b/test/test_fixture_health.py new file mode 100644 index 000000000..8301001c4 --- /dev/null +++ b/test/test_fixture_health.py @@ -0,0 +1,70 @@ +# This module is part of GitPython and is released under the +# 3-Clause BSD License: https://opensource.org/license/bsd-3-clause/ + +"""Verify that the source repository is usable by git.""" + +import subprocess +from pathlib import Path + +import pytest + +REPO_ROOT = Path(__file__).resolve().parent.parent + +FIXTURE_DIRS = [pytest.param(REPO_ROOT, id="repo_root")] + + +def test_source_tree_has_no_gitlinks() -> None: + """All formerly external dependencies are regular monorepo directories.""" + if not (REPO_ROOT / ".git").exists(): + pytest.skip(f"{REPO_ROOT} is not a git checkout") + try: + result = subprocess.run( + ["git", "ls-files", "--stage"], + cwd=REPO_ROOT, + capture_output=True, + text=True, + check=True, + ) + except FileNotFoundError: + pytest.skip("git is not installed or not on PATH") + gitlinks = [line for line in result.stdout.splitlines() if line.startswith("160000 ")] + assert not gitlinks, "Unexpected submodules:\n" + "\n".join(gitlinks) + + +@pytest.mark.parametrize("fixture_dir", FIXTURE_DIRS) +def test_fixture_dir_is_trusted_by_git(fixture_dir: Path) -> None: + """git accepts ``fixture_dir`` as its own repository owned by a trusted user. + + Run ``git -C rev-parse --show-toplevel`` and assert it + succeeds and reports ``fixture_dir`` itself as the toplevel. Failure + typically means the directory's on-disk ownership doesn't match the + running user and the CI workflow's ``safe.directory`` list is missing + an entry that would override the check. + """ + if not (fixture_dir / ".git").exists(): + pytest.skip(f"{fixture_dir} is not a git checkout") + try: + result = subprocess.run( + ["git", "-C", str(fixture_dir), "rev-parse", "--show-toplevel"], + capture_output=True, + text=True, + check=False, + ) + except FileNotFoundError: + pytest.skip("git is not installed or not on PATH") + assert result.returncode == 0, ( + f"git refuses to operate in {fixture_dir}.\n" + f"stderr: {result.stderr.strip()}\n" + "The directory's owner doesn't match the running user and no " + "`safe.directory` entry overrides the check. On CI, the " + "workflow's `safe.directory` list typically needs an entry for " + "this path. Locally, this is unexpected and usually indicates " + "an ownership problem worth investigating." + ) + reported = Path(result.stdout.strip()) + assert reported.samefile(fixture_dir), ( + f"git reports the toplevel as {reported}, " + f"not as {fixture_dir} itself. " + "This usually means the directory is not an initialized git " + "repository (its `.git` marker may be stale or pointing elsewhere)." + ) diff --git a/test/test_fun.py b/test/test_fun.py index b8593b400..38f506ff4 100644 --- a/test/test_fun.py +++ b/test/test_fun.py @@ -243,6 +243,7 @@ def test_tree_traversal(self): B_old = self.rorepo.tree("1f66cfbbce58b4b552b041707a12d437cc5f400a") # old base tree # Two very different trees. + entries = traverse_trees_recursive(odb, [B_old.binsha, H.binsha], "") self._assert_tree_entries(entries, 2) @@ -251,7 +252,10 @@ def test_tree_traversal(self): self._assert_tree_entries(oentries, 2) # Single tree. - is_no_tree = lambda i, d: i.type != "tree" + + def is_no_tree(i, _d): + return i.type != "tree" + entries = traverse_trees_recursive(odb, [B.binsha], "") assert len(entries) == len(list(B.traverse(predicate=is_no_tree))) self._assert_tree_entries(entries, 1) @@ -302,3 +306,7 @@ def test_linked_worktree_traversal(self, rw_dir): def test_tree_entries_from_data_with_failing_name_decode_py3(self): r = tree_entries_from_data(b"100644 \x9f\0aaa") assert r == [(b"aaa", 33188, "\udc9f")], r + + def test_tree_entries_from_bytearray(self): + r = tree_entries_from_data(bytearray(b"100644 name\0abcdefghijklmnopqrst")) + assert r == [(b"abcdefghijklmnopqrst", 33188, "name")], r diff --git a/test/test_git.py b/test/test_git.py index 274511f8d..a88d980fb 100644 --- a/test/test_git.py +++ b/test/test_git.py @@ -6,6 +6,7 @@ import contextlib import gc import inspect +import io import logging import os import os.path as osp @@ -16,6 +17,7 @@ import subprocess import sys import tempfile +import time from unittest import skipUnless if sys.version_info >= (3, 8): @@ -26,6 +28,7 @@ import ddt from git import Git, GitCommandError, GitCommandNotFound, Repo, cmd, refresh +from git.exc import UnsafeOptionError from git.util import cwd, finalize_process from test.lib import TestBase, fixture_path, with_rw_directory @@ -153,6 +156,102 @@ def test_it_transforms_kwargs_into_git_command_arguments(self): res = self.git.transform_kwargs(**{"s": True, "t": True}) self.assertEqual({"-s", "-t"}, set(res)) + def test_check_unsafe_options_normalizes_kwargs(self): + cases = [ + (["upload_pack"], ["--upload-pack"]), + (["receive_pack"], ["--receive-pack"]), + (["exec"], ["--exec"]), + (["u"], ["-u"]), + (["c"], ["-c"]), + (["--upload-pack=/tmp/helper"], ["--upload-pack"]), + (["--config core.filemode=false"], ["--config"]), + (["--upl=/tmp/helper"], ["--upload-pack"]), + (["conf"], ["--config"]), + (["--out=/tmp/output"], ["--output"]), + ] + + for options, unsafe_options in cases: + with self.assertRaises(UnsafeOptionError): + Git.check_unsafe_options(options=options, unsafe_options=unsafe_options) + + def test_check_unsafe_options_does_not_treat_short_options_as_abbreviations(self): + unsafe_options = ["--upload-pack"] + + Git.check_unsafe_options(options=["-u", "u", "-upl"], unsafe_options=unsafe_options) + with self.assertRaises(UnsafeOptionError): + Git.check_unsafe_options(options=["--u"], unsafe_options=unsafe_options) + + def test_check_unsafe_options_does_not_treat_option_values_as_abbreviations(self): + Git.check_unsafe_options(options=["--branch", "conf"], unsafe_options=["--config"]) + + def test_check_unsafe_options_rejects_joined_unsafe_short_options(self): + cases = [ + (["-utouch /tmp/pwn"], ["-u"]), + (["-futouch /tmp/pwn"], ["-u"]), + (["-qutouch${IFS}/tmp/pwn"], ["-u"]), + (["-cprotocol.ext.allow=always"], ["-c"]), + (["-vcprotocol.ext.allow=always"], ["-c"]), + ] + + for options, unsafe_options in cases: + with self.assertRaises(UnsafeOptionError): + Git.check_unsafe_options(options=options, unsafe_options=unsafe_options) + + def test_check_unsafe_options_allows_attached_safe_short_option_values(self): + unsafe_options = ["--upload-pack", "-u", "--config", "-c"] + + Git.check_unsafe_options(options=["-oupstream", "-bcurrent"], unsafe_options=unsafe_options) + + def test_option_candidates_ignore_untransformed_kwargs(self): + options = Git._option_candidates( + kwargs={ + "output": None, + "upload_pack": False, + "exec": [], + "config": [None, False], + "max_count": 1, + } + ) + + self.assertEqual(options, ["--max-count"]) + + def test_option_candidates_include_falsey_non_boolean_values(self): + kwargs = {"pathspec_from_file": 0} + candidates = Git._option_candidates(kwargs=kwargs) + + self.assertEqual(candidates, ["--pathspec-from-file"]) + self.assertEqual(self.git.transform_kwargs(**kwargs), ["--pathspec-from-file=0"]) + with self.assertRaises(UnsafeOptionError): + Git.check_unsafe_options( + options=candidates, + unsafe_options=Git.unsafe_git_pathspec_from_file_options, + ) + + def test_option_candidates_include_split_single_char_option_values(self): + cases = [ + ({"n": "--upload-pack=helper"}, ["-n", "--upload-pack=helper"], ["--upload-pack"]), + ({"g": ("safe", "--out=target")}, ["-g", "--out=target"], ["--output"]), + ] + + for kwargs, candidates, unsafe_options in cases: + self.assertEqual(Git._option_candidates(kwargs=kwargs), candidates) + with self.assertRaises(UnsafeOptionError): + Git.check_unsafe_options(options=candidates, unsafe_options=unsafe_options) + + self.assertEqual(self.git.transform_kwargs(n="--upload-pack=helper"), ["-n", "--upload-pack=helper"]) + + unsplit_kwargs = {"n": "--upload-pack=helper", "split_single_char_options": False} + self.assertEqual(self.git.transform_kwargs(**unsplit_kwargs), ["-n--upload-pack=helper"]) + self.assertEqual(Git._option_candidates(kwargs=unsplit_kwargs), ["-n--upload-pack=helper"]) + + def test_option_candidates_include_joined_single_char_option_values(self): + kwargs = {"n": "uhelper", "split_single_char_options": False} + candidates = Git._option_candidates(kwargs=kwargs) + + self.assertEqual(candidates, ["-nuhelper"]) + with self.assertRaises(UnsafeOptionError): + Git.check_unsafe_options(options=candidates, unsafe_options=["-u"]) + _shell_cases = ( # value_in_call, value_from_class, expected_popen_arg (None, False, False), @@ -201,6 +300,48 @@ def test_it_logs_istream_summary_for_stdin(self, case): def test_it_executes_git_and_returns_result(self): self.assertRegex(self.git.execute(["git", "version"]), r"^git version [\d\.]{2}.*$") + def test_it_output_stream_with_stdout_is_false(self): + temp_stream = io.BytesIO() + self.git.execute( + ["git", "version"], + output_stream=temp_stream, + with_stdout=False, + ) + self.assertEqual(temp_stream.tell(), 0) + + @skipUnless(sys.platform != "win32", "kill_after_timeout is not supported on Windows") + def test_it_honors_kill_after_timeout_with_output_stream(self): + output_stream = io.BytesIO() + command = [ + sys.executable, + "-c", + "import sys, time; sys.stdout.write('started\\n'); sys.stdout.flush(); sys.stdout.close(); time.sleep(60)", + ] + + started = time.monotonic() + status, _, stderr = self.git.execute( + command, + output_stream=output_stream, + kill_after_timeout=1, + with_exceptions=False, + with_extended_output=True, + ) + + self.assertLess(time.monotonic() - started, 5) + self.assertNotEqual(status, 0) + self.assertEqual(output_stream.getvalue(), b"started\n") + self.assertIn("Timeout: the command", stderr) + + def test_it_executes_git_without_stdout_redirect(self): + returncode, stdout, stderr = self.git.execute( + ["git", "version"], + with_extended_output=True, + with_stdout=False, + ) + self.assertEqual(returncode, 0) + self.assertIsNone(stdout) + self.assertIsNotNone(stderr) + @ddt.data( # chdir_to_repo, shell, command, use_shell_impostor (False, False, ["git", "version"], False), @@ -298,6 +439,24 @@ def test_it_raises_proper_exception_with_output_stream(self): with self.assertRaises(GitCommandError): self.git.checkout("non-existent-branch", output_stream=tmp_file) + def test_it_redacts_authorization_extra_header_from_error(self): + token = "fake-token-1234" + command = [ + "git", + "-c", + "http.extraHeader=Authorization: Bearer %s" % token, + "rev-parse", + "--verify", + "refs/does-not-exist", + ] + + with self.assertRaises(GitCommandError) as context: + self.git.execute(command) + + message = str(context.exception) + self.assertNotIn(token, message) + self.assertIn("http.extraHeader=Authorization: *****", message) + def test_it_accepts_environment_variables(self): filename = fixture_path("ls_tree_empty") with open(filename, "r") as fh: @@ -747,7 +906,7 @@ def test_environment(self, rw_dir): path = osp.join(rw_dir, "failing-script.sh") with open(path, "wt") as stream: - stream.write("#!/usr/bin/env sh\n" "echo FOO\n") + stream.write("#!/usr/bin/env sh\necho FOO\n") os.chmod(path, 0o777) rw_repo = Repo.init(osp.join(rw_dir, "repo")) @@ -773,6 +932,7 @@ def stderr_handler(line): cmdline = [ sys.executable, + "-S", # Keep any `CoverageWarning` messages out of the subprocess stderr. fixture_path("cat_file.py"), str(fixture_path("issue-301_stderr")), ] diff --git a/test/test_index.py b/test/test_index.py index c586a0b5a..469a9dc17 100644 --- a/test/test_index.py +++ b/test/test_index.py @@ -16,34 +16,31 @@ import subprocess import sys import tempfile +from unittest import mock from gitdb.base import IStream import ddt import pytest -from git import BlobFilter, Diff, Git, IndexFile, Object, Repo, Tree +from git import BlobFilter, Diff, Git, IndexFile, NULL_TREE, Object, Repo, Tree +from git.diff import NULL_TREE_SHA from git.exc import ( CheckoutError, GitCommandError, HookExecutionError, InvalidGitRepositoryError, UnmergedEntriesError, + UnsafeOptionError, ) -from git.index.fun import hook_path, run_commit_hook +from git.index.fun import _git_for_windows_bash, _which_from_path, hook_path, run_commit_hook from git.index.typ import BaseIndexEntry, IndexEntry from git.index.util import TemporaryFileSwap from git.objects import Blob from git.util import Actor, cwd, hex_to_bin, rmtree -from test.lib import ( - TestBase, - VirtualEnvironment, - fixture, - fixture_path, - with_rw_directory, - with_rw_repo, -) +from test.lib import TestBase, VirtualEnvironment, fixture, fixture_path, with_rw_directory, with_rw_repo, PathLikeMock +from test.lib.helper import symlinks_supported, xfail_if_raises HOOKS_SHEBANG = "#!/usr/bin/env sh\n" @@ -193,6 +190,41 @@ def _make_hook(git_dir, name, content, make_exec=True): @ddt.ddt class TestIndex(TestBase): + @with_rw_repo("HEAD") + def test_checkout_rejects_unsafe_prefix(self, rw_repo): + with tempfile.TemporaryDirectory() as target: + with self.assertRaises(UnsafeOptionError): + rw_repo.index.checkout(prefix=f"{target}/") + + rw_repo.index.checkout(prefix=f"{target}/", allow_unsafe_options=True) + self.assertTrue(osp.isfile(osp.join(target, "CHANGES"))) + + @with_rw_repo("HEAD") + def test_remove_rejects_pathspec_from_file(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + pathspecs = Path(tdir) / "pathspecs" + pathspecs.write_bytes(b"unmatched-path-one\nunmatched-path-two") + for option_name in ("pathspec_from_file", "pathspec_from"): + with self.assertRaises(UnsafeOptionError): + rw_repo.index.remove( + [], + pathspec_file_nul=True, + **{option_name: str(pathspecs)}, + ) + + @with_rw_repo("HEAD") + def test_remove_allows_explicit_pathspec_from_file(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + pathspecs = Path(tdir) / "pathspecs" + pathspecs.write_bytes(b"CHANGES\0") + removed = rw_repo.index.remove( + [], + pathspec_from_file=str(pathspecs), + pathspec_file_nul=True, + allow_unsafe_options=True, + ) + assert "CHANGES" in removed + def __init__(self, *args): super().__init__(*args) self._reset_progress() @@ -308,6 +340,22 @@ def add_bad_blob(): except Exception as ex: assert "index.lock' could not be obtained" not in str(ex) + @with_rw_repo("0.1.6") + def test_read_tree_methods_reject_index_output(self, rw_repo): + output_path = (Path(rw_repo.working_tree_dir) / "alternate-index").as_posix() + unsafe_option = f"--index-output={output_path}" + + with pytest.raises(UnsafeOptionError): + IndexFile.from_tree(rw_repo, unsafe_option) + with pytest.raises(UnsafeOptionError): + IndexFile.from_tree(rw_repo, "HEAD", index_output=output_path) + with pytest.raises(UnsafeOptionError): + rw_repo.index.reset(unsafe_option) + with pytest.raises(UnsafeOptionError): + rw_repo.index.merge_tree(unsafe_option) + with pytest.raises(UnsafeOptionError): + rw_repo.index.merge_tree("HEAD", base=unsafe_option) + @with_rw_repo("0.1.6") def test_index_file_from_tree(self, rw_repo): common_ancestor_sha = "5117c9c8a4d3af19a9958677e45cda9269de1541" @@ -329,7 +377,10 @@ def test_index_file_from_tree(self, rw_repo): assert len([e for e in three_way_index.entries.values() if e.stage != 0]) # ITERATE BLOBS - merge_required = lambda t: t[0] != 0 + + def merge_required(t): + return t[0] != 0 + merge_blobs = list(three_way_index.iter_blobs(merge_required)) assert merge_blobs assert merge_blobs[0][0] in (1, 2, 3) @@ -544,6 +595,39 @@ def test_index_file_diffing(self, rw_repo): rval = index.checkout("lib") assert len(list(rval)) > 1 + @with_rw_directory + def test_index_file_diff_null_tree_with_initial_index(self, rw_dir): + repo = Repo.init(rw_dir) + filename = ".gitkeep" + file_path = osp.join(repo.working_tree_dir, filename) + with open(file_path, "w") as fp: + fp.write("# Initial file\n") + + index = repo.index + index.add([filename]) + index.write() + + index = IndexFile(repo) + self.assertEqual(len(index.diff(None)), 0) + + diff = index.diff(NULL_TREE) + self.assertEqual(len(diff), 1) + self.assertEqual(diff[0].change_type, "A") + assert diff[0].new_file + self.assertEqual(diff[0].b_path, filename) + + self.assertEqual(len(index.diff(NULL_TREE, paths=filename)), 1) + self.assertEqual(len(index.diff(NULL_TREE_SHA, paths=filename)), 1) + self.assertEqual(len(index.diff(NULL_TREE, paths="missing")), 0) + + patch = index.diff(NULL_TREE, create_patch=True) + self.assertEqual(len(patch), 1) + self.assertIn(b"+# Initial file", patch[0].diff) + + with self.assertRaises(GitCommandError) as exc_info: + index.diff(NULL_TREE, bogus_option=True) + self.assertIn("usage: git diff", exc_info.exception.stderr) + def _count_existing(self, repo, files): """Return count of files that actually exist in the repository directory.""" existing = 0 @@ -555,365 +639,369 @@ def _count_existing(self, repo, files): # END num existing helper - @pytest.mark.xfail( - sys.platform == "win32" and Git().config("core.symlinks") == "true", - reason="Assumes symlinks are not created on Windows and opens a symlink to a nonexistent target.", - raises=FileNotFoundError, - ) @with_rw_repo("0.1.6") def test_index_mutation(self, rw_repo): - index = rw_repo.index - num_entries = len(index.entries) - cur_head = rw_repo.head - - uname = "Thomas MÃŧller" - umail = "sd@company.com" - with rw_repo.config_writer() as writer: - writer.set_value("user", "name", uname) - writer.set_value("user", "email", umail) - self.assertEqual(writer.get_value("user", "name"), uname) - - # Remove all of the files, provide a wild mix of paths, BaseIndexEntries, - # IndexEntries. - def mixed_iterator(): - count = 0 - for entry in index.entries.values(): - type_id = count % 5 - if type_id == 0: # path (str) - yield entry.path - elif type_id == 1: # path (PathLike) - yield Path(entry.path) - elif type_id == 2: # blob - yield Blob(rw_repo, entry.binsha, entry.mode, entry.path) - elif type_id == 3: # BaseIndexEntry - yield BaseIndexEntry(entry[:4]) - elif type_id == 4: # IndexEntry - yield entry - else: - raise AssertionError("Invalid Type") - count += 1 - # END for each entry - - # END mixed iterator - deleted_files = index.remove(mixed_iterator(), working_tree=False) - assert deleted_files - self.assertEqual(self._count_existing(rw_repo, deleted_files), len(deleted_files)) - self.assertEqual(len(index.entries), 0) - - # Reset the index to undo our changes. - index.reset() - self.assertEqual(len(index.entries), num_entries) - - # Remove with working copy. - deleted_files = index.remove(mixed_iterator(), working_tree=True) - assert deleted_files - self.assertEqual(self._count_existing(rw_repo, deleted_files), 0) - - # Reset everything. - index.reset(working_tree=True) - self.assertEqual(self._count_existing(rw_repo, deleted_files), len(deleted_files)) - - # Invalid type. - self.assertRaises(TypeError, index.remove, [1]) - - # Absolute path. - deleted_files = index.remove([osp.join(rw_repo.working_tree_dir, "lib")], r=True) - assert len(deleted_files) > 1 - self.assertRaises(ValueError, index.remove, ["/doesnt/exists"]) - - # TEST COMMITTING - # Commit changed index. - cur_commit = cur_head.commit - commit_message = "commit default head by Frèderic Çauflâ‚Ŧ" - - new_commit = index.commit(commit_message, head=False) - assert cur_commit != new_commit - self.assertEqual(new_commit.author.name, uname) - self.assertEqual(new_commit.author.email, umail) - self.assertEqual(new_commit.committer.name, uname) - self.assertEqual(new_commit.committer.email, umail) - self.assertEqual(new_commit.message, commit_message) - self.assertEqual(new_commit.parents[0], cur_commit) - self.assertEqual(len(new_commit.parents), 1) - self.assertEqual(cur_head.commit, cur_commit) - - # Commit with other actor. - cur_commit = cur_head.commit - - my_author = Actor("Frèderic Çauflâ‚Ŧ", "author@example.com") - my_committer = Actor("Committing Frèderic Çauflâ‚Ŧ", "committer@example.com") - commit_actor = index.commit(commit_message, author=my_author, committer=my_committer) - assert cur_commit != commit_actor - self.assertEqual(commit_actor.author.name, "Frèderic Çauflâ‚Ŧ") - self.assertEqual(commit_actor.author.email, "author@example.com") - self.assertEqual(commit_actor.committer.name, "Committing Frèderic Çauflâ‚Ŧ") - self.assertEqual(commit_actor.committer.email, "committer@example.com") - self.assertEqual(commit_actor.message, commit_message) - self.assertEqual(commit_actor.parents[0], cur_commit) - self.assertEqual(len(new_commit.parents), 1) - self.assertEqual(cur_head.commit, commit_actor) - self.assertEqual(cur_head.log()[-1].actor, my_committer) - - # Commit with author_date and commit_date. - cur_commit = cur_head.commit - commit_message = "commit with dates by Avinash Sajjanshetty" - - new_commit = index.commit( - commit_message, - author_date="2006-04-07T22:13:13", - commit_date="2005-04-07T22:13:13", - ) - assert cur_commit != new_commit - print(new_commit.authored_date, new_commit.committed_date) - self.assertEqual(new_commit.message, commit_message) - self.assertEqual(new_commit.authored_date, 1144447993) - self.assertEqual(new_commit.committed_date, 1112911993) - - # Same index, no parents. - commit_message = "index without parents" - commit_no_parents = index.commit(commit_message, parent_commits=[], head=True) - self.assertEqual(commit_no_parents.message, commit_message) - self.assertEqual(len(commit_no_parents.parents), 0) - self.assertEqual(cur_head.commit, commit_no_parents) - - # same index, multiple parents. - commit_message = "Index with multiple parents\n commit with another line" - commit_multi_parent = index.commit(commit_message, parent_commits=(commit_no_parents, new_commit)) - self.assertEqual(commit_multi_parent.message, commit_message) - self.assertEqual(len(commit_multi_parent.parents), 2) - self.assertEqual(commit_multi_parent.parents[0], commit_no_parents) - self.assertEqual(commit_multi_parent.parents[1], new_commit) - self.assertEqual(cur_head.commit, commit_multi_parent) - - # Re-add all files in lib. - # Get the lib folder back on disk, but get an index without it. - index.reset(new_commit.parents[0], working_tree=True).reset(new_commit, working_tree=False) - lib_file_path = osp.join("lib", "git", "__init__.py") - assert (lib_file_path, 0) not in index.entries - assert osp.isfile(osp.join(rw_repo.working_tree_dir, lib_file_path)) - - # Directory. - entries = index.add(["lib"], fprogress=self._fprogress_add) - self._assert_entries(entries) - self._assert_fprogress(entries) - assert len(entries) > 1 - - # Glob. - entries = index.reset(new_commit).add([osp.join("lib", "git", "*.py")], fprogress=self._fprogress_add) - self._assert_entries(entries) - self._assert_fprogress(entries) - self.assertEqual(len(entries), 14) - - # Same file. - entries = index.reset(new_commit).add( - [osp.join(rw_repo.working_tree_dir, "lib", "git", "head.py")] * 2, - fprogress=self._fprogress_add, - ) - self._assert_entries(entries) - self.assertEqual(entries[0].mode & 0o644, 0o644) - # Would fail, test is too primitive to handle this case. - # self._assert_fprogress(entries) - self._reset_progress() - self.assertEqual(len(entries), 2) - - # Missing path. - self.assertRaises(OSError, index.reset(new_commit).add, ["doesnt/exist/must/raise"]) - - # Blob from older revision overrides current index revision. - old_blob = new_commit.parents[0].tree.blobs[0] - entries = index.reset(new_commit).add([old_blob], fprogress=self._fprogress_add) - self._assert_entries(entries) - self._assert_fprogress(entries) - self.assertEqual(index.entries[(old_blob.path, 0)].hexsha, old_blob.hexsha) - self.assertEqual(len(entries), 1) - - # Mode 0 not allowed. - null_hex_sha = Diff.NULL_HEX_SHA - null_bin_sha = b"\0" * 20 - self.assertRaises( - ValueError, - index.reset(new_commit).add, - [BaseIndexEntry((0, null_bin_sha, 0, "doesntmatter"))], - ) - - # Add new file. - new_file_relapath = "my_new_file" - self._make_file(new_file_relapath, "hello world", rw_repo) - entries = index.reset(new_commit).add( - [BaseIndexEntry((0o10644, null_bin_sha, 0, new_file_relapath))], - fprogress=self._fprogress_add, - ) - self._assert_entries(entries) - self._assert_fprogress(entries) - self.assertEqual(len(entries), 1) - self.assertNotEqual(entries[0].hexsha, null_hex_sha) - - # Add symlink. - if sys.platform != "win32": - for target in ("/etc/nonexisting", "/etc/passwd", "/etc"): - basename = "my_real_symlink" - - link_file = osp.join(rw_repo.working_tree_dir, basename) - os.symlink(target, link_file) - entries = index.reset(new_commit).add([link_file], fprogress=self._fprogress_add) - self._assert_entries(entries) - self._assert_fprogress(entries) - self.assertEqual(len(entries), 1) - self.assertTrue(S_ISLNK(entries[0].mode)) - self.assertTrue(S_ISLNK(index.entries[index.entry_key("my_real_symlink", 0)].mode)) - - # We expect only the target to be written. - self.assertEqual( - index.repo.odb.stream(entries[0].binsha).read().decode("ascii"), - target, - ) - - os.remove(link_file) - # END for each target - # END real symlink test - - # Add fake symlink and assure it checks out as a symlink. - fake_symlink_relapath = "my_fake_symlink" - link_target = "/etc/that" - fake_symlink_path = self._make_file(fake_symlink_relapath, link_target, rw_repo) - fake_entry = BaseIndexEntry((0o120000, null_bin_sha, 0, fake_symlink_relapath)) - entries = index.reset(new_commit).add([fake_entry], fprogress=self._fprogress_add) - self._assert_entries(entries) - self._assert_fprogress(entries) - assert entries[0].hexsha != null_hex_sha - self.assertEqual(len(entries), 1) - self.assertTrue(S_ISLNK(entries[0].mode)) - - # Check that this also works with an alternate method. - full_index_entry = IndexEntry.from_base(BaseIndexEntry((0o120000, entries[0].binsha, 0, entries[0].path))) - entry_key = index.entry_key(full_index_entry) - index.reset(new_commit) - - assert entry_key not in index.entries - index.entries[entry_key] = full_index_entry - index.write() - index.update() # Force reread of entries. - new_entry = index.entries[entry_key] - assert S_ISLNK(new_entry.mode) + with xfail_if_raises( + sys.platform == "win32" and (Git().config("core.symlinks") == "true" or symlinks_supported()), + raises=(FileNotFoundError, GitCommandError), + reason="Assumes symlinks are not created on Windows and opens a symlink to a nonexistent target.", + ): + index = rw_repo.index + num_entries = len(index.entries) + cur_head = rw_repo.head + + uname = "Thomas MÃŧller" + umail = "sd@company.com" + with rw_repo.config_writer() as writer: + writer.set_value("user", "name", uname) + writer.set_value("user", "email", umail) + self.assertEqual(writer.get_value("user", "name"), uname) + + # Remove all of the files, provide a wild mix of paths, BaseIndexEntries, + # IndexEntries. + def mixed_iterator(): + count = 0 + for entry in index.entries.values(): + type_id = count % 5 + if type_id == 0: # path (str) + yield entry.path + elif type_id == 1: # path (PathLike) + yield Path(entry.path) + elif type_id == 2: # path mock (PathLike) + yield PathLikeMock(entry.path) + elif type_id == 3: # path mock in a blob + yield Blob(rw_repo, entry.binsha, entry.mode, entry.path) + elif type_id == 4: # blob + yield Blob(rw_repo, entry.binsha, entry.mode, entry.path) + elif type_id == 5: # BaseIndexEntry + yield BaseIndexEntry(entry[:4]) + elif type_id == 6: # IndexEntry + yield entry + else: + raise AssertionError("Invalid Type") + count += 1 + # END for each entry + + # END mixed iterator + deleted_files = index.remove(mixed_iterator(), working_tree=False) + assert deleted_files + self.assertEqual(self._count_existing(rw_repo, deleted_files), len(deleted_files)) + self.assertEqual(len(index.entries), 0) + + # Reset the index to undo our changes. + index.reset() + self.assertEqual(len(index.entries), num_entries) + + # Remove with working copy. + deleted_files = index.remove(mixed_iterator(), working_tree=True) + assert deleted_files + self.assertEqual(self._count_existing(rw_repo, deleted_files), 0) + + # Reset everything. + index.reset(working_tree=True) + self.assertEqual(self._count_existing(rw_repo, deleted_files), len(deleted_files)) + + # Invalid type. + self.assertRaises(TypeError, index.remove, [1]) + + # Absolute path. + deleted_files = index.remove([osp.join(rw_repo.working_tree_dir, "lib")], r=True) + assert len(deleted_files) > 1 + self.assertRaises(ValueError, index.remove, ["/doesnt/exists"]) + + # TEST COMMITTING + # Commit changed index. + cur_commit = cur_head.commit + commit_message = "commit default head by Frèderic Çauflâ‚Ŧ" + + new_commit = index.commit(commit_message, head=False) + assert cur_commit != new_commit + self.assertEqual(new_commit.author.name, uname) + self.assertEqual(new_commit.author.email, umail) + self.assertEqual(new_commit.committer.name, uname) + self.assertEqual(new_commit.committer.email, umail) + self.assertEqual(new_commit.message, commit_message) + self.assertEqual(new_commit.parents[0], cur_commit) + self.assertEqual(len(new_commit.parents), 1) + self.assertEqual(cur_head.commit, cur_commit) + + # Commit with other actor. + cur_commit = cur_head.commit + + my_author = Actor("Frèderic Çauflâ‚Ŧ", "author@example.com") + my_committer = Actor("Committing Frèderic Çauflâ‚Ŧ", "committer@example.com") + commit_actor = index.commit(commit_message, author=my_author, committer=my_committer) + assert cur_commit != commit_actor + self.assertEqual(commit_actor.author.name, "Frèderic Çauflâ‚Ŧ") + self.assertEqual(commit_actor.author.email, "author@example.com") + self.assertEqual(commit_actor.committer.name, "Committing Frèderic Çauflâ‚Ŧ") + self.assertEqual(commit_actor.committer.email, "committer@example.com") + self.assertEqual(commit_actor.message, commit_message) + self.assertEqual(commit_actor.parents[0], cur_commit) + self.assertEqual(len(new_commit.parents), 1) + self.assertEqual(cur_head.commit, commit_actor) + self.assertEqual(cur_head.log()[-1].actor, my_committer) + + # Commit with author_date and commit_date. + cur_commit = cur_head.commit + commit_message = "commit with dates by Avinash Sajjanshetty" + + new_commit = index.commit( + commit_message, + author_date="2006-04-07T22:13:13", + commit_date="2005-04-07T22:13:13", + ) + assert cur_commit != new_commit + print(new_commit.authored_date, new_commit.committed_date) + self.assertEqual(new_commit.message, commit_message) + self.assertEqual(new_commit.authored_date, 1144447993) + self.assertEqual(new_commit.committed_date, 1112911993) + + # Same index, no parents. + commit_message = "index without parents" + commit_no_parents = index.commit(commit_message, parent_commits=[], head=True) + self.assertEqual(commit_no_parents.message, commit_message) + self.assertEqual(len(commit_no_parents.parents), 0) + self.assertEqual(cur_head.commit, commit_no_parents) + + # same index, multiple parents. + commit_message = "Index with multiple parents\n commit with another line" + commit_multi_parent = index.commit(commit_message, parent_commits=(commit_no_parents, new_commit)) + self.assertEqual(commit_multi_parent.message, commit_message) + self.assertEqual(len(commit_multi_parent.parents), 2) + self.assertEqual(commit_multi_parent.parents[0], commit_no_parents) + self.assertEqual(commit_multi_parent.parents[1], new_commit) + self.assertEqual(cur_head.commit, commit_multi_parent) + + # Re-add all files in lib. + # Get the lib folder back on disk, but get an index without it. + index.reset(new_commit.parents[0], working_tree=True).reset(new_commit, working_tree=False) + lib_file_path = osp.join("lib", "git", "__init__.py") + assert (lib_file_path, 0) not in index.entries + assert osp.isfile(osp.join(rw_repo.working_tree_dir, lib_file_path)) + + # Directory. + entries = index.add(["lib"], fprogress=self._fprogress_add) + self._assert_entries(entries) + self._assert_fprogress(entries) + assert len(entries) > 1 + + # Glob. + entries = index.reset(new_commit).add([osp.join("lib", "git", "*.py")], fprogress=self._fprogress_add) + self._assert_entries(entries) + self._assert_fprogress(entries) + self.assertEqual(len(entries), 14) + + # Same file. + entries = index.reset(new_commit).add( + [osp.join(rw_repo.working_tree_dir, "lib", "git", "head.py")] * 2, + fprogress=self._fprogress_add, + ) + self._assert_entries(entries) + self.assertEqual(entries[0].mode & 0o644, 0o644) + # Would fail, test is too primitive to handle this case. + # self._assert_fprogress(entries) + self._reset_progress() + self.assertEqual(len(entries), 2) + + # Missing path. + self.assertRaises(OSError, index.reset(new_commit).add, ["doesnt/exist/must/raise"]) + + # Blob from older revision overrides current index revision. + old_blob = new_commit.parents[0].tree.blobs[0] + entries = index.reset(new_commit).add([old_blob], fprogress=self._fprogress_add) + self._assert_entries(entries) + self._assert_fprogress(entries) + self.assertEqual(index.entries[(old_blob.path, 0)].hexsha, old_blob.hexsha) + self.assertEqual(len(entries), 1) + + # Mode 0 not allowed. + null_hex_sha = Diff.NULL_HEX_SHA + null_bin_sha = b"\0" * 20 + self.assertRaises( + ValueError, + index.reset(new_commit).add, + [BaseIndexEntry((0, null_bin_sha, 0, "doesntmatter"))], + ) - # A tree created from this should contain the symlink. - tree = index.write_tree() - assert fake_symlink_relapath in tree - index.write() # Flush our changes for the checkout. - - # Check out the fake link, should be a link then. - assert not S_ISLNK(os.stat(fake_symlink_path)[ST_MODE]) - os.remove(fake_symlink_path) - index.checkout(fake_symlink_path) - - # On Windows, we currently assume we will never get symlinks. - if sys.platform == "win32": - # Symlinks should contain the link as text (which is what a - # symlink actually is). - with open(fake_symlink_path, "rt") as fd: - self.assertEqual(fd.read(), link_target) - else: - self.assertTrue(S_ISLNK(os.lstat(fake_symlink_path)[ST_MODE])) - - # TEST RENAMING - def assert_mv_rval(rval): - for source, dest in rval: - assert not osp.exists(source) and osp.exists(dest) - # END for each renamed item - - # END move assertion utility - - self.assertRaises(ValueError, index.move, ["just_one_path"]) - # Try to move a file onto an existing file. - files = ["AUTHORS", "LICENSE"] - self.assertRaises(GitCommandError, index.move, files) - - # Again, with force. - assert_mv_rval(index.move(files, f=True)) - - # Move files into a directory - dry run. - paths = ["LICENSE", "VERSION", "doc"] - rval = index.move(paths, dry_run=True) - self.assertEqual(len(rval), 2) - assert osp.exists(paths[0]) - - # Again, no dry run. - rval = index.move(paths) - assert_mv_rval(rval) - - # Move dir into dir. - rval = index.move(["doc", "test"]) - assert_mv_rval(rval) - - # TEST PATH REWRITING - ###################### - count = [0] - - def rewriter(entry): - rval = str(count[0]) - count[0] += 1 - return rval - - # END rewriter - - def make_paths(): - """Help out the test by yielding two existing paths and one new path.""" - yield "CHANGES" - yield "ez_setup.py" - yield index.entries[index.entry_key("README", 0)] - yield index.entries[index.entry_key(".gitignore", 0)] - - for fid in range(3): - fname = "newfile%i" % fid - with open(fname, "wb") as fd: - fd.write(b"abcd") - yield Blob(rw_repo, Blob.NULL_BIN_SHA, 0o100644, fname) - # END for each new file - - # END path producer - paths = list(make_paths()) - self._assert_entries(index.add(paths, path_rewriter=rewriter)) - - for filenum in range(len(paths)): - assert index.entry_key(str(filenum), 0) in index.entries - - # TEST RESET ON PATHS - ###################### - arela = "aa" - brela = "bb" - afile = self._make_file(arela, "adata", rw_repo) - bfile = self._make_file(brela, "bdata", rw_repo) - akey = index.entry_key(arela, 0) - bkey = index.entry_key(brela, 0) - keys = (akey, bkey) - absfiles = (afile, bfile) - files = (arela, brela) - - for fkey in keys: - assert fkey not in index.entries - - index.add(files, write=True) - nc = index.commit("2 files committed", head=False) - - for fkey in keys: - assert fkey in index.entries - - # Just the index. - index.reset(paths=(arela, afile)) - assert akey not in index.entries - assert bkey in index.entries - - # Now with working tree - files on disk as well as entries must be recreated. - rw_repo.head.commit = nc - for absfile in absfiles: - os.remove(absfile) - - index.reset(working_tree=True, paths=files) - - for fkey in keys: - assert fkey in index.entries - for absfile in absfiles: - assert osp.isfile(absfile) + # Add new file. + new_file_relapath = "my_new_file" + self._make_file(new_file_relapath, "hello world", rw_repo) + entries = index.reset(new_commit).add( + [BaseIndexEntry((0o10644, null_bin_sha, 0, new_file_relapath))], + fprogress=self._fprogress_add, + ) + self._assert_entries(entries) + self._assert_fprogress(entries) + self.assertEqual(len(entries), 1) + self.assertNotEqual(entries[0].hexsha, null_hex_sha) + + # Add symlink. + if sys.platform != "win32": + for target in ("/etc/nonexisting", "/etc/passwd", "/etc"): + basename = "my_real_symlink" + + link_file = osp.join(rw_repo.working_tree_dir, basename) + os.symlink(target, link_file) + entries = index.reset(new_commit).add([link_file], fprogress=self._fprogress_add) + self._assert_entries(entries) + self._assert_fprogress(entries) + self.assertEqual(len(entries), 1) + self.assertTrue(S_ISLNK(entries[0].mode)) + self.assertTrue(S_ISLNK(index.entries[index.entry_key("my_real_symlink", 0)].mode)) + + # We expect only the target to be written. + self.assertEqual( + index.repo.odb.stream(entries[0].binsha).read().decode("ascii"), + target, + ) + + os.remove(link_file) + # END for each target + # END real symlink test + + # Add fake symlink and assure it checks out as a symlink. + fake_symlink_relapath = "my_fake_symlink" + link_target = "/etc/that" + fake_symlink_path = self._make_file(fake_symlink_relapath, link_target, rw_repo) + fake_entry = BaseIndexEntry((0o120000, null_bin_sha, 0, fake_symlink_relapath)) + entries = index.reset(new_commit).add([fake_entry], fprogress=self._fprogress_add) + self._assert_entries(entries) + self._assert_fprogress(entries) + assert entries[0].hexsha != null_hex_sha + self.assertEqual(len(entries), 1) + self.assertTrue(S_ISLNK(entries[0].mode)) + + # Check that this also works with an alternate method. + full_index_entry = IndexEntry.from_base(BaseIndexEntry((0o120000, entries[0].binsha, 0, entries[0].path))) + entry_key = index.entry_key(full_index_entry) + index.reset(new_commit) + + assert entry_key not in index.entries + index.entries[entry_key] = full_index_entry + index.write() + index.update() # Force reread of entries. + new_entry = index.entries[entry_key] + assert S_ISLNK(new_entry.mode) + + # A tree created from this should contain the symlink. + tree = index.write_tree() + assert fake_symlink_relapath in tree + index.write() # Flush our changes for the checkout. + + # Check out the fake link, should be a link then. + assert not S_ISLNK(os.stat(fake_symlink_path)[ST_MODE]) + os.remove(fake_symlink_path) + index.checkout(fake_symlink_path) + + # On Windows, we currently assume we will never get symlinks. + if sys.platform == "win32": + # Symlinks should contain the link as text (which is what a + # symlink actually is). + with open(fake_symlink_path, "rt") as fd: + self.assertEqual(fd.read(), link_target) + else: + self.assertTrue(S_ISLNK(os.lstat(fake_symlink_path)[ST_MODE])) + + # TEST RENAMING + def assert_mv_rval(rval): + for source, dest in rval: + assert not osp.exists(source) and osp.exists(dest) + # END for each renamed item + + # END move assertion utility + + self.assertRaises(ValueError, index.move, ["just_one_path"]) + # Try to move a file onto an existing file. + files = ["AUTHORS", "LICENSE"] + self.assertRaises(GitCommandError, index.move, files) + + # Again, with force. + assert_mv_rval(index.move(files, f=True)) + + # Move files into a directory - dry run. + paths = ["LICENSE", "VERSION", "doc"] + rval = index.move(paths, dry_run=True) + self.assertEqual(len(rval), 2) + assert osp.exists(paths[0]) + + # Again, no dry run. + rval = index.move(paths) + assert_mv_rval(rval) + + # Move dir into dir. + rval = index.move(["doc", "test"]) + assert_mv_rval(rval) + + # TEST PATH REWRITING + ###################### + count = [0] + + def rewriter(entry): + rval = str(count[0]) + count[0] += 1 + return rval + + # END rewriter + + def make_paths(): + """Help out the test by yielding two existing paths and one new path.""" + yield "CHANGES" + yield "ez_setup.py" + yield index.entries[index.entry_key("README", 0)] + yield index.entries[index.entry_key(".gitignore", 0)] + + for fid in range(3): + fname = "newfile%i" % fid + with open(fname, "wb") as fd: + fd.write(b"abcd") + yield Blob(rw_repo, Blob.NULL_BIN_SHA, 0o100644, fname) + # END for each new file + + # END path producer + paths = list(make_paths()) + self._assert_entries(index.add(paths, path_rewriter=rewriter)) + + for filenum in range(len(paths)): + assert index.entry_key(str(filenum), 0) in index.entries + + # TEST RESET ON PATHS + ###################### + arela = "aa" + brela = "bb" + afile = self._make_file(arela, "adata", rw_repo) + bfile = self._make_file(brela, "bdata", rw_repo) + akey = index.entry_key(arela, 0) + bkey = index.entry_key(brela, 0) + keys = (akey, bkey) + absfiles = (afile, bfile) + files = (arela, brela) + + for fkey in keys: + assert fkey not in index.entries + + index.add(files, write=True) + nc = index.commit("2 files committed", head=False) + + for fkey in keys: + assert fkey in index.entries + + # Just the index. + index.reset(paths=(arela, afile)) + assert akey not in index.entries + assert bkey in index.entries + + # Now with working tree - files on disk as well as entries must be recreated. + rw_repo.head.commit = nc + for absfile in absfiles: + os.remove(absfile) + + index.reset(working_tree=True, paths=files) + + for fkey in keys: + assert fkey in index.entries + for absfile in absfiles: + assert osp.isfile(absfile) @with_rw_repo("HEAD") def test_compare_write_tree(self, rw_repo): @@ -1015,6 +1103,65 @@ class Mocked: rel = index._to_relative_path(path) self.assertEqual(rel, os.path.relpath(path, root)) + def test__to_relative_path_absolute_trailing_slash(self): + repo_root = os.path.join(osp.abspath(os.sep), "directory1", "repo_root") + + class Mocked: + bare = False + git_dir = repo_root + working_tree_dir = repo_root + + repo = Mocked() + path = os.path.join(repo_root, f"directory2{os.sep}") + index = IndexFile(repo) + + expected_path = "directory2/" + actual_path = index._to_relative_path(path) + self.assertEqual(expected_path, actual_path) + + @pytest.mark.skipif(sys.platform != "win32", reason="Specifically for Windows.") + def test__to_relative_path_windows_unc_share_root(self): + for repo_root in [R"\\server\share", R"\\?\UNC\server\share"]: + with self.subTest(repo_root=repo_root): + repo = mock.Mock(bare=False, git_dir=repo_root, working_tree_dir=repo_root) + index = IndexFile(repo) + + self.assertEqual(index._to_relative_path(repo_root), ".") + + @pytest.mark.skipif(sys.platform != "win32", reason="Specifically for Windows.") + @with_rw_directory + def test__to_relative_path_windows_path_kinds(self, rw_dir): + repo_root = osp.join(rw_dir, "repo") + os.makedirs(osp.join(repo_root, "nested")) + + class Mocked: + bare = False + git_dir = osp.join(repo_root, ".git") + working_tree_dir = repo_root + + index = IndexFile(Mocked()) + inside_path = osp.join(repo_root, "nested", "file") + _drive, rooted_inside_path = osp.splitdrive(inside_path) + + self.assertEqual(index._to_relative_path(inside_path), "nested/file") + self.assertEqual(index._to_relative_path(rooted_inside_path), "nested/file") + self.assertEqual(index._to_relative_path(rooted_inside_path.replace("\\", "/")), "nested/file") + self.assertEqual(index._to_relative_path(PathLikeMock(inside_path)), "nested/file") + self.assertEqual(index._to_relative_path(inside_path.upper()), "NESTED/FILE") + self.assertRaises(ValueError, index._to_relative_path, osp.join(repo_root + "-other", "file")) + self.assertRaises(ValueError, index._to_relative_path, osp.join("..", "outside")) + self.assertRaises(ValueError, index._to_relative_path, f"{osp.splitdrive(repo_root)[0]}relative") + self.assertRaises(ValueError, index._to_relative_path, R"Z:\outside") + self.assertRaises(ValueError, index._to_relative_path, R"\\server\share\outside") + self.assertRaises(ValueError, index._to_relative_path, R"\\?\C:\outside") + + Mocked.bare = True + bare_index = IndexFile(Mocked()) + self.assertRaises(InvalidGitRepositoryError, bare_index._to_relative_path, rooted_inside_path) + self.assertRaises( + InvalidGitRepositoryError, bare_index._to_relative_path, f"{osp.splitdrive(repo_root)[0]}relative" + ) + @pytest.mark.xfail( type(_win_bash_status) is WinBashStatus.Absent, reason="Can't run a hook on Windows without bash.exe.", @@ -1029,9 +1176,99 @@ class Mocked: def test_run_commit_hook(self, rw_repo): index = rw_repo.index _make_hook(index.repo.git_dir, "fake-hook", "echo 'ran fake hook' >output.txt") - run_commit_hook("fake-hook", index) - output = Path(rw_repo.git_dir, "output.txt").read_text(encoding="utf-8") - self.assertEqual(output, "ran fake hook\n") + output = Path(rw_repo.git_dir, "output.txt") + with mock.patch.object(Repo, "config_level", ("repository",)): + with mock.patch.object(Git, "execute", side_effect=AssertionError("hook lookup must not run git")): + run_commit_hook("fake-hook", index) + self.assertEqual(output.read_text(encoding="utf-8"), "ran fake hook\n") + + output.unlink() + with index.repo.config_writer() as writer: + writer.set_value("core", "hooksPath", "") + run_commit_hook("fake-hook", index) + + self.assertEqual(output.read_text(encoding="utf-8"), "ran fake hook\n") + + @with_rw_directory + def test_run_commit_hook_outside_worktree_on_windows(self, rw_dir): + root = Path(rw_dir).resolve() + repo = Repo.init(root / "repo") + hooks_dir = root / "hooks" + _make_hook(root, "fake-hook", "exit 0") + system_root = root / "Windows" + system_bash = system_root / "System32" / "bash.exe" + git_executable = root / "Git" / "cmd" / "git.exe" + git_bash = root / "Git" / "bin" / "bash.exe" + for executable in (system_bash, git_executable, git_bash): + executable.parent.mkdir(parents=True) + executable.touch() + executable.chmod(0o755) + with repo.config_writer() as writer: + writer.set_value("core", "hooksPath", str(hooks_dir)) + + # Model a normal Windows PATH: System32 (containing the WSL launcher) comes + # before Git's cmd directory, while Git's Bash is not itself on PATH. This + # exercises both Git-installation discovery and shell selection without + # mocking either resolver's answer. + with mock.patch("git.index.fun.sys.platform", "win32"), mock.patch.object( + Git, "GIT_PYTHON_GIT_EXECUTABLE", "git" + ), mock.patch.dict(os.environ, {"SystemRoot": str(system_root)}), mock.patch( + "git.index.fun.os.get_exec_path", return_value=["", str(system_bash.parent), str(git_executable.parent)] + ): + with mock.patch("git.index.fun.safer_popen") as popen, mock.patch("git.index.fun.handle_process_output"): + popen.return_value.returncode = 0 + run_commit_hook("fake-hook", repo.index) + + command = popen.call_args[0][0] + self.assertEqual(command, [str(git_bash), "../hooks/fake-hook"]) + + @with_rw_directory + def test_windows_bash_lookup_respects_explicit_current_directory_in_path(self, rw_dir): + root = Path(rw_dir).resolve() + bash = root / "bash.exe" + bash.touch() + bash.chmod(0o755) + + # An explicitly listed directory is trusted PATH configuration, even when + # it happens to be the current directory. This differs from an empty entry, + # which Windows requires PATH lookup to ignore. + with cwd(root), mock.patch("git.index.fun.os.get_exec_path", return_value=[str(root)]): + self.assertEqual(_which_from_path("bash.exe"), str(bash)) + + @with_rw_directory + def test_windows_bash_lookup_from_explicit_git_bin(self, rw_dir): + git_root = Path(rw_dir).resolve() / "Git" + git_executable = git_root / "bin" / "git.exe" + bash = git_root / "bin" / "bash.exe" + git_executable.parent.mkdir(parents=True) + for executable in (git_executable, bash): + executable.touch() + executable.chmod(0o755) + + # A relative executable containing a directory is resolved by CreateProcess + # from the parent process cwd, not the separately supplied child cwd. Enter the + # temporary root first because Windows cannot express a relative path between + # drives, and CI may keep the checkout and its temporary directory on different + # drives. + with cwd(Path(rw_dir).resolve()): + relative_git = osp.relpath(git_executable, os.curdir) + with mock.patch.object(Git, "GIT_PYTHON_GIT_EXECUTABLE", relative_git): + self.assertEqual(_git_for_windows_bash(), str(bash)) + + @with_rw_directory + def test_windows_bash_lookup_ignores_custom_git_executable(self, rw_dir): + root = Path(rw_dir).resolve() + for directory_name in ("cmd", "bin"): + executable = root / directory_name / "mygit.exe" + bash = root / "bin" / "bash.exe" + executable.parent.mkdir(parents=True, exist_ok=True) + bash.parent.mkdir(parents=True, exist_ok=True) + executable.touch() + bash.touch() + executable.chmod(0o755) + bash.chmod(0o755) + with mock.patch.object(Git, "GIT_PYTHON_GIT_EXECUTABLE", str(executable)): + self.assertIsNone(_git_for_windows_bash()) @ddt.data((False,), (True,)) @with_rw_directory @@ -1090,6 +1327,32 @@ def test_pre_commit_hook_success(self, rw_repo): _make_hook(index.repo.git_dir, "pre-commit", "exit 0") index.commit("This should not fail") + @pytest.mark.xfail( + type(_win_bash_status) is WinBashStatus.Absent, + reason="Can't run a hook on Windows without bash.exe.", + raises=HookExecutionError, + ) + @pytest.mark.xfail( + type(_win_bash_status) is WinBashStatus.WslNoDistro, + reason="Currently uses the bash.exe of WSL, even with no WSL distro installed", + raises=HookExecutionError, + ) + @with_rw_repo("HEAD") + def test_pre_commit_hook_respects_core_hooks_path(self, rw_repo): + index = rw_repo.index + hooks_dir = Path(index.repo.working_dir, "custom-hooks") + hooks_dir.mkdir() + hp = hooks_dir / "pre-commit" + hp.write_text(HOOKS_SHEBANG + "echo 'ran custom hook' >custom-hook-output.txt", encoding="utf-8") + os.chmod(hp, 0o744) + + with index.repo.config_writer() as writer: + writer.set_value("core", "hooksPath", "custom-hooks") + + index.commit("This should run the custom hook") + output = Path(rw_repo.working_dir, "custom-hook-output.txt").read_text(encoding="utf-8") + self.assertEqual(output, "ran custom hook\n") + @pytest.mark.xfail( type(_win_bash_status) is WinBashStatus.WslNoDistro, reason="Currently uses the bash.exe of WSL, even with no WSL distro installed", @@ -1173,7 +1436,7 @@ def test_commit_msg_hook_fail(self, rw_repo): raise AssertionError("Should have caught a HookExecutionError") @with_rw_repo("HEAD") - def test_index_add_pathlike(self, rw_repo): + def test_index_add_pathlib(self, rw_repo): git_dir = Path(rw_repo.git_dir) file = git_dir / "file.txt" @@ -1181,6 +1444,15 @@ def test_index_add_pathlike(self, rw_repo): rw_repo.index.add(file) + @with_rw_repo("HEAD") + def test_index_add_pathlike(self, rw_repo): + git_dir = Path(rw_repo.git_dir) + + file = git_dir / "file.txt" + file.touch() + + rw_repo.index.add(PathLikeMock(str(file))) + @with_rw_repo("HEAD") def test_index_add_non_normalized_path(self, rw_repo): git_dir = Path(rw_repo.git_dir) @@ -1193,6 +1465,48 @@ def test_index_add_non_normalized_path(self, rw_repo): rw_repo.index.add(non_normalized_path) + def test_index_file_v3(self): + index = IndexFile(self.rorepo, fixture_path("index_extended_flags")) + assert index.entries + assert index.version == 3 + assert len(index.entries) == 4 + assert index.entries[("init.t", 0)].skip_worktree + + # Write the data - it must match the original. + with tempfile.NamedTemporaryFile() as tmpfile: + index.write(tmpfile.name) + assert Path(tmpfile.name).read_bytes() == Path(fixture_path("index_extended_flags")).read_bytes() + + @with_rw_directory + def test_index_file_v3_with_git_command(self, tmp_dir): + tmp_dir = Path(tmp_dir) + with cwd(tmp_dir): + git = Git(tmp_dir) + git.init() + + file = tmp_dir / "file.txt" + file.write_text("hello") + git.add("--intent-to-add", "file.txt") # intent-to-add sets extended flag + + repo = Repo(tmp_dir) + index = repo.index + + assert len(index.entries) == 1 + assert index.version == 3 + entry = list(index.entries.values())[0] + assert entry.path == "file.txt" + assert entry.intent_to_add + + file2 = tmp_dir / "file2.txt" + file2.write_text("world") + index.add(["file2.txt"]) + index.write() + + status_str = git.status(porcelain=True) + status_lines = status_str.splitlines() + assert " A file.txt" in status_lines + assert "A file2.txt" in status_lines + class TestIndexUtils: @pytest.mark.parametrize("file_path_type", [str, Path]) diff --git a/test/test_installation.py b/test/test_installation.py index ae6472e98..e8956d5cf 100644 --- a/test/test_installation.py +++ b/test/test_installation.py @@ -2,59 +2,34 @@ # 3-Clause BSD License: https://opensource.org/license/bsd-3-clause/ import ast +import functools import os import subprocess -from test.lib import TestBase, VirtualEnvironment, with_rw_directory +from test.lib import TestBase, VirtualEnvironment, requires_symlinks, with_rw_directory class TestInstallation(TestBase): + @requires_symlinks @with_rw_directory def test_installation(self, rw_dir): - venv = self._set_up_venv(rw_dir) + venv, run = self._set_up_venv(rw_dir) - result = subprocess.run( - [venv.pip, "install", "."], - stdout=subprocess.PIPE, - cwd=venv.sources, - ) - self.assertEqual( - 0, - result.returncode, - msg=result.stderr or result.stdout or "Can't install project", - ) + for project in ("./smmap", "./gitdb", "."): + result = run([venv.pip, "install", project]) + self._check_result(result, f"Can't install {project}") - result = subprocess.run( - [venv.python, "-c", "import git"], - stdout=subprocess.PIPE, - cwd=venv.sources, - ) - self.assertEqual( - 0, - result.returncode, - msg=result.stderr or result.stdout or "Self-test failed", - ) + result = run([venv.python, "-c", "import git"]) + self._check_result(result, "Self-test failed") - result = subprocess.run( - [venv.python, "-c", "import gitdb; import smmap"], - stdout=subprocess.PIPE, - cwd=venv.sources, - ) - self.assertEqual( - 0, - result.returncode, - msg=result.stderr or result.stdout or "Dependencies not installed", - ) + result = run([venv.python, "-c", "import gitdb; import smmap"]) + self._check_result(result, "Dependencies not installed") # Even IF gitdb or any other dependency is supplied during development by # inserting its location into PYTHONPATH or otherwise patched into sys.path, # make sure it is not wrongly inserted as the *first* entry. - result = subprocess.run( - [venv.python, "-c", "import sys; import git; print(sys.path)"], - stdout=subprocess.PIPE, - cwd=venv.sources, - ) - syspath = result.stdout.decode("utf-8").splitlines()[0] + result = run([venv.python, "-c", "import sys; import git; print(sys.path)"]) + syspath = result.stdout.splitlines()[0] syspath = ast.literal_eval(syspath) self.assertEqual( "", @@ -64,10 +39,37 @@ def test_installation(self, rw_dir): @staticmethod def _set_up_venv(rw_dir): + # Initialize the virtual environment. venv = VirtualEnvironment(rw_dir, with_pip=True) + + # Make its src directory a symlink to our own top-level source tree. os.symlink( os.path.dirname(os.path.dirname(__file__)), venv.sources, target_is_directory=True, ) - return venv + + # Create a convenience function to run commands in it. + run = functools.partial( + subprocess.run, + stdout=subprocess.PIPE, + stderr=subprocess.PIPE, + universal_newlines=True, + cwd=venv.sources, + env={**os.environ, "PYTHONWARNINGS": "error"}, + ) + + return venv, run + + def _check_result(self, result, failure_summary): + self.assertEqual( + 0, + result.returncode, + msg=self._prepare_failure_message(result, failure_summary), + ) + + @staticmethod + def _prepare_failure_message(result, failure_summary): + stdout = result.stdout.rstrip() + stderr = result.stderr.rstrip() + return f"{failure_summary}\n\nstdout:\n{stdout}\n\nstderr:\n{stderr}" diff --git a/test/test_quick_doc.py b/test/test_quick_doc.py index 4ef75f4aa..98658e02f 100644 --- a/test/test_quick_doc.py +++ b/test/test_quick_doc.py @@ -173,7 +173,7 @@ def test_cloned_repo_object(self, local_dir): # [15-test_cloned_repo_object] def print_files_from_git(root, level=0): for entry in root: - print(f'{"-" * 4 * level}| {entry.path}, {entry.type}') + print(f"{'-' * 4 * level}| {entry.path}, {entry.type}") if entry.type == "tree": print_files_from_git(entry, level + 1) diff --git a/test/test_refs.py b/test/test_refs.py index 08096e69e..9a3f58c7b 100644 --- a/test/test_refs.py +++ b/test/test_refs.py @@ -3,6 +3,7 @@ # This module is part of GitPython and is released under the # 3-Clause BSD License: https://opensource.org/license/bsd-3-clause/ +import contextlib from itertools import chain import os.path as osp from pathlib import Path @@ -18,17 +19,31 @@ RefLog, Reference, RemoteReference, + Repo, SymbolicReference, TagReference, ) +from git.exc import UnsafeOptionError from git.objects.tag import TagObject import git.refs as refs from git.util import Actor -from test.lib import TestBase, with_rw_repo +from test.lib import TestBase, requires_symlinks, with_rw_repo, PathLikeMock class TestRefs(TestBase): + @contextlib.contextmanager + def _repo_with_initial_commit(self, base_dir): + repo_dir = base_dir / "repo" + repo = Repo.init(repo_dir) + (repo_dir / "file.txt").write_text("initial\n", encoding="utf-8") + repo.index.add(["file.txt"]) + repo.index.commit("initial") + try: + yield repo + finally: + repo.git.clear_cache() + def test_from_path(self): # Should be able to create any reference directly. for ref_type in (Reference, Head, TagReference, RemoteReference): @@ -43,6 +58,52 @@ def test_from_path(self): self.assertRaises(ValueError, TagReference, self.rorepo, "refs/invalid/tag") # Works without path check. TagReference(self.rorepo, "refs/invalid/tag", check_path=False) + # Check remoteness + assert Reference(self.rorepo, "refs/remotes/origin").is_remote() + + @with_rw_repo("HEAD") + def test_tag_create_rejects_unsafe_file_options(self, rw_repo): + with tempfile.NamedTemporaryFile("w", encoding="utf-8") as message: + message.write("private tag message") + message.flush() + for index, option in enumerate(({"F": message.name}, {"file": message.name})): + with self.assertRaises(UnsafeOptionError): + TagReference.create(rw_repo, f"unsafe-{index}", **option) + + for args in ( + ("unsafe-reference", f"--file={message.name}"), + (f"--file={message.name}", "HEAD"), + (f"-eF{message.name}", "HEAD"), + (f"-iF{message.name}", "HEAD"), + ): + with self.assertRaises(UnsafeOptionError): + TagReference.create(rw_repo, *args) + + with self.assertRaises(UnsafeOptionError): + TagReference.create(rw_repo, "unsafe-ref-kwarg", ref=f"--file={message.name}") + + tag = TagReference.create(rw_repo, "legacy-ref", ref="HEAD", allow_unsafe_options=True) + self.assertEqual(tag.commit, rw_repo.head.commit) + + tag = TagReference.create(rw_repo, "allowed-file", F=message.name, allow_unsafe_options=True) + self.assertEqual(tag.tag.message, "private tag message") + + def test_from_pathlike(self): + # Should be able to create any reference directly. + for ref_type in (Reference, Head, TagReference, RemoteReference): + for name in ("rela_name", "path/rela_name"): + full_path = ref_type.to_full_path(PathLikeMock(name)) + instance = ref_type.from_path(self.rorepo, PathLikeMock(full_path)) + assert isinstance(instance, ref_type) + # END for each name + # END for each type + + # Invalid path. + self.assertRaises(ValueError, TagReference, self.rorepo, "refs/invalid/tag") + # Works without path check. + TagReference(self.rorepo, PathLikeMock("refs/invalid/tag"), check_path=False) + # Check remoteness + assert Reference(self.rorepo, PathLikeMock("refs/remotes/origin")).is_remote() def test_tag_base(self): tag_object_refs = [] @@ -212,10 +273,55 @@ def test_orig_head(self): @with_rw_repo("0.1.6") def test_head_checkout_detached_head(self, rw_repo): - res = rw_repo.remotes.origin.refs.master.checkout() + res = rw_repo.remotes.origin.refs.HEAD.reference.checkout() assert isinstance(res, SymbolicReference) assert res.name == "HEAD" + @with_rw_repo("HEAD") + def test_head_checkout_rejects_pathspec_from_file(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + pathspecs = Path(tdir) / "pathspecs" + pathspecs.write_bytes(b"unmatched-path-one\nunmatched-path-two") + for option_name in ("pathspec_from_file", "pathspec_from"): + with self.assertRaises(UnsafeOptionError): + rw_repo.active_branch.checkout( + pathspec_file_nul=True, + **{option_name: str(pathspecs)}, + ) + + @with_rw_repo("HEAD") + def test_head_reset_rejects_pathspec_from_file(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + pathspecs = Path(tdir) / "pathspecs" + pathspecs.write_bytes(b"unmatched-path-one\nunmatched-path-two") + for option_name in ("pathspec_from_file", "pathspec_from"): + with self.assertRaises(UnsafeOptionError): + rw_repo.head.reset( + pathspec_file_nul=True, + **{option_name: str(pathspecs)}, + ) + for option_name in ("--pathspec-from-file", "--pathspec-from"): + with self.assertRaises(UnsafeOptionError): + rw_repo.head.reset( + f"{option_name}={pathspecs}", + pathspec_file_nul=True, + ) + + @with_rw_repo("HEAD") + def test_head_commands_allow_explicit_pathspec_from_file(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + pathspecs = Path(tdir) / "pathspecs" + pathspecs.write_bytes(b"CHANGES\0") + options = { + "pathspec_from_file": str(pathspecs), + "pathspec_file_nul": True, + "allow_unsafe_options": True, + } + head = rw_repo.head + branch = rw_repo.active_branch + assert head.reset(**options) is head + assert branch.checkout(**options) == branch + @with_rw_repo("0.1.6") def test_head_reset(self, rw_repo): cur_head = rw_repo.head @@ -615,7 +721,7 @@ def test_dereference_recursive(self): assert SymbolicReference.dereference_recursive(self.rorepo, "HEAD") def test_reflog(self): - assert isinstance(self.rorepo.heads.master.log(), RefLog) + assert isinstance(self.rorepo.active_branch.log(), RefLog) def test_refs_outside_repo(self): # Create a file containing a valid reference outside the repository. Attempting @@ -629,6 +735,113 @@ def test_refs_outside_repo(self): ref_file_name = Path(ref_file.name).name self.assertRaises(BadName, self.rorepo.commit, f"../../{ref_file_name}") + def test_reference_create_rejects_path_traversal(self): + with tempfile.TemporaryDirectory() as tmp_dir: + base_dir = Path(tmp_dir) + with self._repo_with_initial_commit(base_dir) as repo: + outside_path = base_dir / "outside_write.txt" + + self.assertRaises(ValueError, Reference.create, repo, "../../../outside_write.txt", "HEAD") + assert not outside_path.exists() + + def test_symbolic_reference_create_rejects_path_traversal(self): + with tempfile.TemporaryDirectory() as tmp_dir: + base_dir = Path(tmp_dir) + with self._repo_with_initial_commit(base_dir) as repo: + outside_path = base_dir / "outside_write.txt" + + self.assertRaises(ValueError, SymbolicReference.create, repo, "../../outside_write.txt", "HEAD") + assert not outside_path.exists() + + def test_symbolic_reference_set_reference_rejects_path_traversal(self): + with tempfile.TemporaryDirectory() as tmp_dir: + base_dir = Path(tmp_dir) + with self._repo_with_initial_commit(base_dir) as repo: + outside_path = base_dir / "outside_write.txt" + + self.assertRaises(ValueError, SymbolicReference(repo, "../../outside_write.txt").set_reference, "HEAD") + assert not outside_path.exists() + + def test_symbolic_reference_rename_rejects_path_traversal(self): + with tempfile.TemporaryDirectory() as tmp_dir: + base_dir = Path(tmp_dir) + with self._repo_with_initial_commit(base_dir) as repo: + outside_path = base_dir / "outside_move.txt" + ref = SymbolicReference.create(repo, "SAFE_RENAME_SOURCE", "HEAD") + + self.assertRaises(ValueError, ref.rename, "../../outside_move.txt") + assert not outside_path.exists() + assert Path(ref.abspath).is_file() + + def test_symbolic_reference_delete_rejects_path_traversal(self): + with tempfile.TemporaryDirectory() as tmp_dir: + base_dir = Path(tmp_dir) + with self._repo_with_initial_commit(base_dir) as repo: + outside_path = base_dir / "outside_delete.txt" + outside_path.write_text("do not delete\n", encoding="utf-8") + + self.assertRaises(ValueError, SymbolicReference.delete, repo, "../../outside_delete.txt") + assert outside_path.read_text(encoding="utf-8") == "do not delete\n" + + def test_symbolic_reference_log_append_rejects_path_traversal(self): + with tempfile.TemporaryDirectory() as tmp_dir: + base_dir = Path(tmp_dir) + with self._repo_with_initial_commit(base_dir) as repo: + outside_path = base_dir / "outside_reflog.txt" + + ref = SymbolicReference(repo, "../../../outside_reflog.txt") + self.assertRaises( + ValueError, ref.log_append, Commit.NULL_BIN_SHA, "do not write", repo.head.commit.binsha + ) + assert not outside_path.exists() + + @requires_symlinks + def test_symbolic_reference_set_reference_rejects_symlink_escape(self): + with tempfile.TemporaryDirectory() as tmp_dir: + base_dir = Path(tmp_dir) + with self._repo_with_initial_commit(base_dir) as repo: + outside_dir = base_dir / "outside_refs" + outside_dir.mkdir() + outside_path = outside_dir / "escaped" + + refs_heads_dir = Path(repo.common_dir) / "refs" / "heads" + refs_heads_dir.mkdir(parents=True, exist_ok=True) + symlink_path = refs_heads_dir / "link_out" + symlink_path.symlink_to(outside_dir, target_is_directory=True) + if osp.realpath(symlink_path / "escaped") == osp.abspath(symlink_path / "escaped"): + self.skipTest("realpath does not resolve directory symlinks on this platform") + + ref = SymbolicReference(repo, "refs/heads/link_out/escaped") + self.assertRaises(ValueError, ref.set_reference, "HEAD") + assert not outside_path.exists() + + def test_remote_reference_delete_cleanup_rejects_path_traversal(self): + with tempfile.TemporaryDirectory() as tmp_dir: + base_dir = Path(tmp_dir) + git_dir = base_dir / "repo" / ".git" + git_dir.mkdir(parents=True) + outside_path = base_dir / "outside_remote_delete.txt" + outside_path.write_text("do not delete\n", encoding="utf-8") + + class GitStub: + branch_called = False + + def branch(self, *args): + self.branch_called = True + + class RepoStub: + pass + + repo = RepoStub() + repo.git = GitStub() + repo.common_dir = str(git_dir) + repo.git_dir = str(git_dir) + ref = RemoteReference(repo, "../../outside_remote_delete.txt", check_path=False) + + self.assertRaises(ValueError, RemoteReference.delete, repo, ref) + assert not repo.git.branch_called + assert outside_path.read_text(encoding="utf-8") == "do not delete\n" + def test_validity_ref_names(self): """Ensure ref names are checked for validity. diff --git a/test/test_remote.py b/test/test_remote.py index 5ddb41bc0..e1793214c 100644 --- a/test/test_remote.py +++ b/test/test_remote.py @@ -4,12 +4,13 @@ # 3-Clause BSD License: https://opensource.org/license/bsd-3-clause/ import gc +import os import os.path as osp from pathlib import Path import random import sys import tempfile -from unittest import skipIf +from unittest import mock, skipIf import pytest @@ -44,7 +45,7 @@ class TestRemoteProgress(RemoteProgress): __slots__ = ("_seen_lines", "_stages_per_op", "_num_progress_messages") - def __init__(self): + def __init__(self) -> None: super().__init__() self._seen_lines = [] self._stages_per_op = {} @@ -103,6 +104,9 @@ def assert_received_message(self): assert self._num_progress_messages +TestRemoteProgress.__test__ = False # type: ignore + + class TestRemote(TestBase): def tearDown(self): gc.collect() @@ -682,9 +686,22 @@ def test_multiple_urls(self, rw_repo): # Will raise fatal: Will not delete all non-push URLs. self.assertRaises(GitCommandError, remote.delete_url, test3) + @with_rw_repo("HEAD", bare=False) + def test_create_does_not_expand_environment_variables_in_url(self, rw_repo): + url = "https://example.com/${GITPYTHON_TEST_SECRET}/repo.git" + with mock.patch.dict(os.environ, {"GITPYTHON_TEST_SECRET": "sensitive-value"}): + remote = rw_repo.create_remote("untrusted", url) + + assert remote.url == url + def test_fetch_error(self): rem = self.rorepo.remote("origin") - with self.assertRaisesRegex(GitCommandError, "[Cc]ouldn't find remote ref __BAD_REF__"): + msg = ( + r"[Cc]ouldn't find remote ref __BAD_REF__|" + r"could not read Username|" + r"expected flush after ref listing" + ) + with self.assertRaisesRegex(GitCommandError, msg): rem.fetch("__BAD_REF__") @with_rw_repo("0.1.6", bare=False) @@ -824,7 +841,11 @@ def test_fetch_unsafe_options(self, rw_repo): remote = rw_repo.remote("origin") tmp_dir = Path(tdir) tmp_file = tmp_dir / "pwn" - unsafe_options = [{"upload-pack": f"touch {tmp_file}"}] + unsafe_options = [ + {"upload-pack": f"touch {tmp_file}"}, + {"upload_pack": f"touch {tmp_file}"}, + {"upload_p": f"touch {tmp_file}"}, + ] for unsafe_option in unsafe_options: with self.assertRaises(UnsafeOptionError): remote.fetch(**unsafe_option) @@ -892,7 +913,11 @@ def test_pull_unsafe_options(self, rw_repo): remote = rw_repo.remote("origin") tmp_dir = Path(tdir) tmp_file = tmp_dir / "pwn" - unsafe_options = [{"upload-pack": f"touch {tmp_file}"}] + unsafe_options = [ + {"upload-pack": f"touch {tmp_file}"}, + {"upload_pack": f"touch {tmp_file}"}, + {"upload_p": f"touch {tmp_file}"}, + ] for unsafe_option in unsafe_options: with self.assertRaises(UnsafeOptionError): remote.pull(**unsafe_option) @@ -961,10 +986,11 @@ def test_push_unsafe_options(self, rw_repo): tmp_dir = Path(tdir) tmp_file = tmp_dir / "pwn" unsafe_options = [ - { - "receive-pack": f"touch {tmp_file}", - "exec": f"touch {tmp_file}", - } + {"receive-pack": f"touch {tmp_file}"}, + {"receive_pack": f"touch {tmp_file}"}, + {"receive_p": f"touch {tmp_file}"}, + {"exec": f"touch {tmp_file}"}, + {"exe": f"touch {tmp_file}"}, ] for unsafe_option in unsafe_options: assert not tmp_file.exists() @@ -988,10 +1014,9 @@ def test_push_unsafe_options_allowed(self, rw_repo): tmp_dir = Path(tdir) tmp_file = tmp_dir / "pwn" unsafe_options = [ - { - "receive-pack": f"touch {tmp_file}", - "exec": f"touch {tmp_file}", - } + {"receive-pack": f"touch {tmp_file}"}, + {"receive_pack": f"touch {tmp_file}"}, + {"exec": f"touch {tmp_file}"}, ] for unsafe_option in unsafe_options: # The options will be allowed, but the command will fail. @@ -1001,6 +1026,50 @@ def test_push_unsafe_options_allowed(self, rw_repo): assert tmp_file.exists() tmp_file.unlink() + @with_rw_repo("HEAD") + def test_ls_remote_unsafe_options(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = Path(tdir) + tmp_file = tmp_dir / "pwn" + unsafe_options = [ + {"upload-pack": f"touch {tmp_file}"}, + {"upload_pack": f"touch {tmp_file}"}, + {"upl": f"touch {tmp_file}"}, + {"exec": f"touch {tmp_file}"}, + ] + for unsafe_option in unsafe_options: + with self.assertRaises(UnsafeOptionError): + rw_repo.git.ls_remote(".", **unsafe_option) + with self.assertRaises(UnsafeOptionError): + rw_repo.git.ls_remote([f"--upload-pack={tmp_file}"], ".") + with self.assertRaises(UnsafeOptionError): + rw_repo.git.ls_remote([f"--upl={tmp_file}"], ".") + with self.assertRaises(UnsafeOptionError): + rw_repo.git.ls_remote(f"--upload-pack={tmp_file}", ".") + with self.assertRaises(UnsafeOptionError): + rw_repo.git.ls_remote(f"--upl={tmp_file}", ".") + with self.assertRaises(UnsafeOptionError): + rw_repo.git.ls_remote(f"--exec={tmp_file}", ".") + with self.assertRaises(UnsafeOptionError): + rw_repo.git.ls_remote("--upload-pack", "touch", ".") + with self.assertRaises(UnsafeOptionError): + rw_repo.git.ls_remote("--refs", ".", upl=f"touch {tmp_file}") + + def test_ls_remote_allows_operand_named_like_unsafe_option(self): + with mock.patch.object(Git, "_call_process") as git: + Git().ls_remote("upload-pack") + git.assert_called_once() + + @with_rw_repo("HEAD") + def test_ls_remote_unsafe_options_allowed(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = Path(tdir) + tmp_file = tmp_dir / "pwn" + unsafe_options = [{"upload-pack": f"touch {tmp_file}"}] + for unsafe_option in unsafe_options: + with self.assertRaises(GitCommandError): + rw_repo.git.ls_remote(".", **unsafe_option, allow_unsafe_options=True) + @with_rw_and_rw_remote_repo("0.1.6") def test_fetch_unsafe_branch_name(self, rw_repo, remote_repo): # Create branch with a name containing a NBSP diff --git a/test/test_repo.py b/test/test_repo.py index bfa1bbb78..1dfec951a 100644 --- a/test/test_repo.py +++ b/test/test_repo.py @@ -14,7 +14,8 @@ import pickle import sys import tempfile -from unittest import mock, skip +from unittest import mock +from pathlib import Path import pytest @@ -36,11 +37,13 @@ Submodule, Tree, ) -from git.exc import BadObject, UnsafeOptionError, UnsafeProtocolError +from git.exc import UnsafeOptionError +from git.exc import UnsafeProtocolError +from git.exc import BadObject from git.repo.fun import touch from git.util import bin_to_hex, cwd, cygpath, join_path_native, rmfile, rmtree -from test.lib import TestBase, fixture, with_rw_directory, with_rw_repo +from test.lib import TestBase, fixture, requires_symlinks, with_rw_directory, with_rw_repo, PathLikeMock def iter_flatten(lol): @@ -79,6 +82,33 @@ def test_new_should_raise_on_invalid_repo_location(self): with tempfile.TemporaryDirectory() as tdir: self.assertRaises(InvalidGitRepositoryError, Repo, tdir) + def test_init_rejects_unsafe_options(self): + with tempfile.TemporaryDirectory() as tdir: + template_dir = osp.join(tdir, "template") + os.mkdir(template_dir) + unsafe_options = [ + {"template": template_dir}, + {"templa": template_dir}, + {"separate_git_dir": osp.join(tdir, "git-dir")}, + {"separate_git_di": osp.join(tdir, "git-dir")}, + ] + for index, kwargs in enumerate(unsafe_options): + repo_dir = osp.join(tdir, f"repo-{index}") + with self.assertRaises(UnsafeOptionError): + Repo.init(repo_dir, **kwargs) + assert not osp.exists(repo_dir) + + def test_init_allows_explicitly_unsafe_options(self): + with tempfile.TemporaryDirectory() as tdir: + template_dir = osp.join(tdir, "template") + os.mkdir(template_dir) + repo = Repo.init( + osp.join(tdir, "repo"), + template=template_dir, + allow_unsafe_options=True, + ) + assert repo.git_dir + @with_rw_directory def test_new_should_raise_on_invalid_repo_location_within_repo(self, rw_dir): repo_dir = osp.join(rw_dir, "repo") @@ -105,6 +135,11 @@ def test_repo_creation_pathlib(self, rw_repo): r_from_gitdir = Repo(pathlib.Path(rw_repo.git_dir)) self.assertEqual(r_from_gitdir.git_dir, rw_repo.git_dir) + @with_rw_repo("0.3.2.1") + def test_repo_creation_pathlike(self, rw_repo): + r_from_gitdir = Repo(PathLikeMock(rw_repo.git_dir)) + self.assertEqual(r_from_gitdir.git_dir, rw_repo.git_dir) + def test_description(self): txt = "Test repository" self.rorepo.description = txt @@ -120,8 +155,9 @@ def test_heads_should_populate_head_data(self): self.assertIsInstance(head.commit, Commit) # END for each head - self.assertIsInstance(self.rorepo.heads.master, Head) - self.assertIsInstance(self.rorepo.heads["master"], Head) + active_branch = self.rorepo.active_branch + self.assertIsInstance(self.rorepo.heads[active_branch.name], Head) + self.assertEqual(self.rorepo.heads[active_branch.name], active_branch) def test_tree_from_revision(self): tree = self.rorepo.tree("0.1.6") @@ -140,6 +176,23 @@ def test_commit_from_revision(self): self.assertEqual(commit.type, "commit") self.assertEqual(self.rorepo.commit(commit), commit) + @with_rw_directory + def test_commit_from_tag_starting_with_at(self, rw_dir): + repo = Repo.init(rw_dir) + with repo.config_writer() as writer: + writer.set_value("user", "name", "GitPython Tests") + writer.set_value("user", "email", "gitpython@example.com") + + tracked_file = Path(rw_dir) / "hello.txt" + tracked_file.write_text("hello") + repo.index.add([str(tracked_file)]) + commit = repo.index.commit("init") + repo.create_tag("@foo") + + self.assertEqual(repo.tags["@foo"].commit, commit) + self.assertEqual(repo.commit("@"), commit) + self.assertEqual(repo.commit("@foo"), commit) + def test_commits(self): mc = 10 commits = list(self.rorepo.iter_commits("0.1.6", max_count=mc)) @@ -208,290 +261,23 @@ def test_clone_from_keeps_env(self, rw_dir): self.assertEqual(environment, cloned.git.environment()) + @pytest.mark.skipif(os.name != "nt", reason="Specifically for Windows drive-rooted paths.") @with_rw_directory - def test_date_format(self, rw_dir): - repo = Repo.init(osp.join(rw_dir, "repo")) - # @-timestamp is the format used by git commit hooks. - repo.index.commit("Commit messages", commit_date="@1400000000 +0000") - - @with_rw_directory - def test_clone_from_pathlib(self, rw_dir): - original_repo = Repo.init(osp.join(rw_dir, "repo")) - - Repo.clone_from(original_repo.git_dir, pathlib.Path(rw_dir) / "clone_pathlib") - - @with_rw_directory - def test_clone_from_pathlib_withConfig(self, rw_dir): + def test_clone_from_drive_rooted_destination(self, rw_dir): original_repo = Repo.init(osp.join(rw_dir, "repo")) + with cwd(rw_dir): + destination = osp.join(rw_dir, "clone") + _drive, rooted_destination = osp.splitdrive(destination) - cloned = Repo.clone_from( - original_repo.git_dir, - pathlib.Path(rw_dir) / "clone_pathlib_withConfig", - multi_options=[ - "--recurse-submodules=repo", - "--config core.filemode=false", - "--config submodule.repo.update=checkout", - "--config filter.lfs.clean='git-lfs clean -- %f'", - ], - allow_unsafe_options=True, - ) - - self.assertEqual(cloned.config_reader().get_value("submodule", "active"), "repo") - self.assertEqual(cloned.config_reader().get_value("core", "filemode"), False) - self.assertEqual(cloned.config_reader().get_value('submodule "repo"', "update"), "checkout") - self.assertEqual( - cloned.config_reader().get_value('filter "lfs"', "clean"), - "git-lfs clean -- %f", - ) + cloned = Repo.clone_from(original_repo.git_dir, rooted_destination) - def test_clone_from_with_path_contains_unicode(self): - with tempfile.TemporaryDirectory() as tmpdir: - unicode_dir_name = "\u0394" - path_with_unicode = os.path.join(tmpdir, unicode_dir_name) - os.makedirs(path_with_unicode) - - try: - Repo.clone_from( - url=self._small_repo_url(), - to_path=path_with_unicode, - ) - except UnicodeEncodeError: - self.fail("Raised UnicodeEncodeError") + assert osp.samefile(cloned.working_tree_dir, destination) @with_rw_directory - @skip( - """The referenced repository was removed, and one needs to set up a new - password controlled repo under the org's control.""" - ) - def test_leaking_password_in_clone_logs(self, rw_dir): - password = "fakepassword1234" - try: - Repo.clone_from( - url="https://fakeuser:{}@fakerepo.example.com/testrepo".format(password), - to_path=rw_dir, - ) - except GitCommandError as err: - assert password not in str(err), "The error message '%s' should not contain the password" % err - # Working example from a blank private project. - Repo.clone_from( - url="https://gitlab+deploy-token-392045:mLWhVus7bjLsy8xj8q2V@gitlab.com/mercierm/test_git_python", - to_path=rw_dir, - ) - - @with_rw_repo("HEAD") - def test_clone_unsafe_options(self, rw_repo): - with tempfile.TemporaryDirectory() as tdir: - tmp_dir = pathlib.Path(tdir) - tmp_file = tmp_dir / "pwn" - unsafe_options = [ - f"--upload-pack='touch {tmp_file}'", - f"-u 'touch {tmp_file}'", - "--config=protocol.ext.allow=always", - "-c protocol.ext.allow=always", - ] - for unsafe_option in unsafe_options: - with self.assertRaises(UnsafeOptionError): - rw_repo.clone(tmp_dir, multi_options=[unsafe_option]) - assert not tmp_file.exists() - - unsafe_options = [ - {"upload-pack": f"touch {tmp_file}"}, - {"u": f"touch {tmp_file}"}, - {"config": "protocol.ext.allow=always"}, - {"c": "protocol.ext.allow=always"}, - ] - for unsafe_option in unsafe_options: - with self.assertRaises(UnsafeOptionError): - rw_repo.clone(tmp_dir, **unsafe_option) - assert not tmp_file.exists() - - @pytest.mark.xfail( - sys.platform == "win32", - reason=( - "File not created. A separate Windows command may be needed. This and the " - "currently passing test test_clone_unsafe_options must be adjusted in the " - "same way. Until then, test_clone_unsafe_options is unreliable on Windows." - ), - raises=AssertionError, - ) - @with_rw_repo("HEAD") - def test_clone_unsafe_options_allowed(self, rw_repo): - with tempfile.TemporaryDirectory() as tdir: - tmp_dir = pathlib.Path(tdir) - tmp_file = tmp_dir / "pwn" - unsafe_options = [ - f"--upload-pack='touch {tmp_file}'", - f"-u 'touch {tmp_file}'", - ] - for i, unsafe_option in enumerate(unsafe_options): - destination = tmp_dir / str(i) - assert not tmp_file.exists() - # The options will be allowed, but the command will fail. - with self.assertRaises(GitCommandError): - rw_repo.clone(destination, multi_options=[unsafe_option], allow_unsafe_options=True) - assert tmp_file.exists() - tmp_file.unlink() - - unsafe_options = [ - "--config=protocol.ext.allow=always", - "-c protocol.ext.allow=always", - ] - for i, unsafe_option in enumerate(unsafe_options): - destination = tmp_dir / str(i) - assert not destination.exists() - rw_repo.clone(destination, multi_options=[unsafe_option], allow_unsafe_options=True) - assert destination.exists() - - @with_rw_repo("HEAD") - def test_clone_safe_options(self, rw_repo): - with tempfile.TemporaryDirectory() as tdir: - tmp_dir = pathlib.Path(tdir) - options = [ - "--depth=1", - "--single-branch", - "-q", - ] - for option in options: - destination = tmp_dir / option - assert not destination.exists() - rw_repo.clone(destination, multi_options=[option]) - assert destination.exists() - - @with_rw_repo("HEAD") - def test_clone_from_unsafe_options(self, rw_repo): - with tempfile.TemporaryDirectory() as tdir: - tmp_dir = pathlib.Path(tdir) - tmp_file = tmp_dir / "pwn" - unsafe_options = [ - f"--upload-pack='touch {tmp_file}'", - f"-u 'touch {tmp_file}'", - "--config=protocol.ext.allow=always", - "-c protocol.ext.allow=always", - ] - for unsafe_option in unsafe_options: - with self.assertRaises(UnsafeOptionError): - Repo.clone_from(rw_repo.working_dir, tmp_dir, multi_options=[unsafe_option]) - assert not tmp_file.exists() - - unsafe_options = [ - {"upload-pack": f"touch {tmp_file}"}, - {"u": f"touch {tmp_file}"}, - {"config": "protocol.ext.allow=always"}, - {"c": "protocol.ext.allow=always"}, - ] - for unsafe_option in unsafe_options: - with self.assertRaises(UnsafeOptionError): - Repo.clone_from(rw_repo.working_dir, tmp_dir, **unsafe_option) - assert not tmp_file.exists() - - @pytest.mark.xfail( - sys.platform == "win32", - reason=( - "File not created. A separate Windows command may be needed. This and the " - "currently passing test test_clone_from_unsafe_options must be adjusted in the " - "same way. Until then, test_clone_from_unsafe_options is unreliable on Windows." - ), - raises=AssertionError, - ) - @with_rw_repo("HEAD") - def test_clone_from_unsafe_options_allowed(self, rw_repo): - with tempfile.TemporaryDirectory() as tdir: - tmp_dir = pathlib.Path(tdir) - tmp_file = tmp_dir / "pwn" - unsafe_options = [ - f"--upload-pack='touch {tmp_file}'", - f"-u 'touch {tmp_file}'", - ] - for i, unsafe_option in enumerate(unsafe_options): - destination = tmp_dir / str(i) - assert not tmp_file.exists() - # The options will be allowed, but the command will fail. - with self.assertRaises(GitCommandError): - Repo.clone_from( - rw_repo.working_dir, destination, multi_options=[unsafe_option], allow_unsafe_options=True - ) - assert tmp_file.exists() - tmp_file.unlink() - - unsafe_options = [ - "--config=protocol.ext.allow=always", - "-c protocol.ext.allow=always", - ] - for i, unsafe_option in enumerate(unsafe_options): - destination = tmp_dir / str(i) - assert not destination.exists() - Repo.clone_from( - rw_repo.working_dir, destination, multi_options=[unsafe_option], allow_unsafe_options=True - ) - assert destination.exists() - - @with_rw_repo("HEAD") - def test_clone_from_safe_options(self, rw_repo): - with tempfile.TemporaryDirectory() as tdir: - tmp_dir = pathlib.Path(tdir) - options = [ - "--depth=1", - "--single-branch", - "-q", - ] - for option in options: - destination = tmp_dir / option - assert not destination.exists() - Repo.clone_from(rw_repo.common_dir, destination, multi_options=[option]) - assert destination.exists() - - def test_clone_from_unsafe_protocol(self): - with tempfile.TemporaryDirectory() as tdir: - tmp_dir = pathlib.Path(tdir) - tmp_file = tmp_dir / "pwn" - urls = [ - f"ext::sh -c touch% {tmp_file}", - "fd::17/foo", - ] - for url in urls: - with self.assertRaises(UnsafeProtocolError): - Repo.clone_from(url, tmp_dir / "repo") - assert not tmp_file.exists() - - def test_clone_from_unsafe_protocol_allowed(self): - with tempfile.TemporaryDirectory() as tdir: - tmp_dir = pathlib.Path(tdir) - tmp_file = tmp_dir / "pwn" - urls = [ - f"ext::sh -c touch% {tmp_file}", - "fd::/foo", - ] - for url in urls: - # The URL will be allowed into the command, but the command will - # fail since we don't have that protocol enabled in the Git config file. - with self.assertRaises(GitCommandError): - Repo.clone_from(url, tmp_dir / "repo", allow_unsafe_protocols=True) - assert not tmp_file.exists() - - def test_clone_from_unsafe_protocol_allowed_and_enabled(self): - with tempfile.TemporaryDirectory() as tdir: - tmp_dir = pathlib.Path(tdir) - tmp_file = tmp_dir / "pwn" - urls = [ - f"ext::sh -c touch% {tmp_file}", - ] - allow_ext = [ - "--config=protocol.ext.allow=always", - ] - for url in urls: - # The URL will be allowed into the command, and the protocol is enabled, - # but the command will fail since it can't read from the remote repo. - assert not tmp_file.exists() - with self.assertRaises(GitCommandError): - Repo.clone_from( - url, - tmp_dir / "repo", - multi_options=allow_ext, - allow_unsafe_protocols=True, - allow_unsafe_options=True, - ) - assert tmp_file.exists() - tmp_file.unlink() + def test_date_format(self, rw_dir): + repo = Repo.init(osp.join(rw_dir, "repo")) + # @-timestamp is the format used by git commit hooks. + repo.index.commit("Commit messages", commit_date="@1400000000 +0000") @with_rw_repo("HEAD") def test_max_chunk_size(self, repo): @@ -578,15 +364,37 @@ def test_daemon_export(self): self.rorepo.daemon_export = orig_val self.assertEqual(self.rorepo.daemon_export, orig_val) + def test_daemon_export_without_git_dir(self): + repo = Repo.__new__(Repo) + repo.git_dir = None + self.assertFalse(repo._get_daemon_export()) + repo._set_daemon_export(True) + def test_alternates(self): cur_alternates = self.rorepo.alternates - # empty alternates - self.rorepo.alternates = [] - self.assertEqual(self.rorepo.alternates, []) + try: + # Empty alternates. + self.rorepo.alternates = [] + self.assertEqual(self.rorepo.alternates, []) + alts = ["other/location", "this/location"] + self.rorepo.alternates = alts + self.assertEqual(alts, self.rorepo.alternates) + finally: + self.rorepo.alternates = cur_alternates + + @with_rw_directory + def test_alternates_use_common_dir(self, rw_dir): + common_dir = osp.join(rw_dir, "common") + git_dir = osp.join(rw_dir, "worktrees", "linked") + os.makedirs(osp.join(common_dir, "objects", "info")) + os.makedirs(osp.join(git_dir, "objects", "info")) + repo = mock.Mock(common_dir=common_dir, git_dir=git_dir) + alts = ["other/location", "this/location"] - self.rorepo.alternates = alts - self.assertEqual(alts, self.rorepo.alternates) - self.rorepo.alternates = cur_alternates + Repo._set_alternates(repo, alts) + + self.assertEqual(Repo._get_alternates(repo), alts) + self.assertFalse(osp.exists(osp.join(git_dir, "objects", "info", "alternates"))) def test_repr(self): assert repr(self.rorepo).startswith("= (2, 48, 0): + wt_add_kwargs["relative_paths"] = True + rw_master.git.worktree("add", worktree_path, branch.name, **wt_add_kwargs) # This ensures that we can read the repo's gitdir correctly. repo = Repo(worktree_path) @@ -1326,6 +1261,35 @@ def test_git_work_tree_dotgit(self, rw_dir): self.assertIsInstance(repo.heads["aaaaaaaa"], Head) + def test_git_work_tree_dotgit_relative(self): + """Check that we find .git as a worktree file containing a relative path + and find the worktree based on it.""" + if Git().version_info[:3] < (2, 48, 0): + pytest.skip("relative worktree feature unsupported, needs git 2.48.0 or later") + # this class inherits from TestCase so we can't use pytest.mark.parametrize on + # test_git_work_tree_dotgit; delegate instead + self.test_git_work_tree_dotgit(use_relative_paths=True) + + @pytest.mark.skipif(os.name != "nt", reason="Specifically for Windows drive-rooted paths.") + @with_rw_directory + def test_linked_worktree_drive_rooted_gitdir_file(self, rw_dir): + with cwd(rw_dir): + main_repo = self.rorepo.clone(join_path_native(rw_dir, "main_repo")) + branch = main_repo.create_head("drive-rooted-worktree") + worktree_path = join_path_native(rw_dir, "worktree_repo") + main_repo.git.worktree("add", worktree_path, branch.name) + + linked_repo = Repo(worktree_path) + gitdir_file = osp.join(linked_repo.git_dir, "gitdir") + worktree_gitfile = osp.join(worktree_path, ".git") + _drive, rooted_worktree_gitfile = osp.splitdrive(worktree_gitfile) + with open(gitdir_file, "w") as stream: + stream.write(rooted_worktree_gitfile) + + reopened_repo = Repo(linked_repo.git_dir) + + assert osp.samefile(reopened_repo.working_tree_dir, worktree_path) + @with_rw_directory def test_git_work_tree_env(self, rw_dir): """Check that we yield to GIT_WORK_TREE.""" @@ -1350,6 +1314,42 @@ def test_git_work_tree_env(self, rw_dir): self.assertEqual(r.working_tree_dir, repo_dir) self.assertEqual(r.working_dir, repo_dir) + @with_rw_directory + def test_git_work_tree_env_in_linked_worktree(self, rw_dir): + """Check that Repo() autodiscovers a linked worktree when GIT_DIR is set.""" + git = Git(rw_dir) + if git.version_info[:3] < (2, 5, 1): + raise RuntimeError("worktree feature unsupported (test needs git 2.5.1 or later)") + + rw_master = self.rorepo.clone(join_path_native(rw_dir, "master_repo")) + branch = rw_master.create_head("bbbbbbbb") + worktree_path = join_path_native(rw_dir, "worktree_repo") + if Git.is_cygwin(): + worktree_path = cygpath(worktree_path) + + rw_master.git.worktree("add", worktree_path, branch.name) + + git_dir = Git(worktree_path).rev_parse("--git-dir") + + patched_env = dict(os.environ) + patched_env["GIT_DIR"] = git_dir + patched_env.pop("GIT_WORK_TREE", None) + patched_env.pop("GIT_COMMON_DIR", None) + + with mock.patch.dict(os.environ, patched_env, clear=True): + old_cwd = os.getcwd() + try: + os.chdir(worktree_path) + + explicit = Repo(os.getcwd()) + autodiscovered = Repo() + + self.assertTrue(osp.samefile(explicit.working_tree_dir, worktree_path)) + self.assertTrue(osp.samefile(autodiscovered.working_tree_dir, worktree_path)) + self.assertTrue(osp.samefile(autodiscovered.working_tree_dir, explicit.working_tree_dir)) + finally: + os.chdir(old_cwd) + @with_rw_directory def test_rebasing(self, rw_dir): r = Repo.init(rw_dir) @@ -1436,6 +1436,7 @@ def test_ignored_items_reported(self): ["included_file.txt", "ignored_file.txt", "included_dir/file.txt", "ignored_dir/file.txt"] ) == ["ignored_file.txt", "ignored_dir/file.txt"] + @requires_symlinks def test_ignored_raises_error_w_symlink(self): with tempfile.TemporaryDirectory() as tdir: tmp_dir = pathlib.Path(tdir) diff --git a/test/test_rev_parse.py b/test/test_rev_parse.py new file mode 100644 index 000000000..b00347668 --- /dev/null +++ b/test/test_rev_parse.py @@ -0,0 +1,159 @@ +# Copyright (C) 2026 Michael Trier (mtrier@gmail.com) and contributors +# +# This module is part of GitPython and is released under the +# 3-Clause BSD License: https://opensource.org/license/bsd-3-clause/ + +from pathlib import Path + +import pytest + +from git import Repo +from git.refs import RemoteReference +from git.refs import SymbolicReference +from gitdb.exc import BadName + + +def _write(repo, path, content): + full_path = Path(repo.working_tree_dir) / path + full_path.parent.mkdir(parents=True, exist_ok=True) + full_path.write_text(content) + repo.index.add([str(full_path)]) + + +@pytest.fixture +def rev_parse_repo(tmp_path): + repo = Repo.init(tmp_path) + with repo.config_writer() as writer: + writer.set_value("user", "name", "GitPython Tests") + writer.set_value("user", "email", "gitpython@example.com") + + _write(repo, "README.md", "root\n") + _write(repo, "CHANGES", "root changes\n") + _write(repo, "dir/file.txt", "root file\n") + root = repo.index.commit("root commit") + repo.create_tag("ann", ref=root, message="annotated tag") + + _write(repo, "README.md", "release\n") + release = repo.index.commit("release candidate") + repo.create_tag("v1.0", ref=release) + main = repo.active_branch + + _write(repo, "side.txt", "side\n") + side_commit = repo.index.commit("side branch", parent_commits=[root], head=False, skip_hooks=True) + repo.create_head("side", side_commit) + + merge = repo.index.commit("merge side", parent_commits=[release, side_commit], skip_hooks=True) + repo.head.log_append(side_commit.binsha, "checkout: moving from side to main", merge.binsha) + + repo.create_head("aaaaaaaa", merge) + repo.create_tag("@foo", ref=merge) + + return { + "repo": repo, + "root": root, + "release": release, + "side": side_commit, + "merge": merge, + "main": main, + } + + +def test_rev_parse_names_hex_and_describe_forms(rev_parse_repo): + repo = rev_parse_repo["repo"] + release = rev_parse_repo["release"] + merge = rev_parse_repo["merge"] + + assert repo.rev_parse("@") == merge + assert repo.rev_parse("@foo") == merge + assert repo.rev_parse("aaaaaaaa") == merge + assert repo.rev_parse(merge.hexsha[:7]) == merge + describe_name = "anything-9-g%s" % merge.hexsha[:7] + assert repo.rev_parse("v1.0-1-g%s" % merge.hexsha[:7]) == merge + assert repo.rev_parse(describe_name) == merge + assert repo.rev_parse("%s-dirty" % merge.hexsha[:7]) == merge + + repo.create_tag(describe_name, ref=release) + assert repo.rev_parse(describe_name) == release + + +def test_rev_parse_navigation_and_peeling(rev_parse_repo): + repo = rev_parse_repo["repo"] + root = rev_parse_repo["root"] + release = rev_parse_repo["release"] + side = rev_parse_repo["side"] + merge = rev_parse_repo["merge"] + tag = repo.rev_parse("ann") + + assert repo.rev_parse("HEAD^0") == merge + assert repo.rev_parse("HEAD~0") == merge + assert repo.rev_parse("HEAD^1") == release + assert repo.rev_parse("HEAD^2") == side + assert repo.rev_parse("HEAD~") == release + assert repo.rev_parse("HEAD^^") == root + + assert tag.type == "tag" + assert repo.rev_parse("ann^{object}") == tag + assert repo.rev_parse("ann^{tag}") == tag + assert repo.rev_parse("ann^{}") == root + assert repo.rev_parse("ann^{commit}") == root + assert repo.rev_parse("HEAD^{tree}") == merge.tree + with pytest.raises(ValueError): + repo.rev_parse("HEAD^{/}") + + +def test_rev_parse_tree_and_index_paths(rev_parse_repo): + repo = rev_parse_repo["repo"] + merge = rev_parse_repo["merge"] + + assert repo.rev_parse("HEAD:") == merge.tree + assert repo.rev_parse("HEAD:README.md") == merge.tree["README.md"] + assert repo.rev_parse("HEAD^{tree}:README.md") == merge.tree["README.md"] + assert repo.rev_parse(":README.md").binsha == merge.tree["README.md"].binsha + assert repo.rev_parse(":0:README.md").binsha == merge.tree["README.md"].binsha + + +def test_rev_parse_reflog_selectors(rev_parse_repo): + repo = rev_parse_repo["repo"] + merge = rev_parse_repo["merge"] + side = rev_parse_repo["side"] + main = rev_parse_repo["main"] + release = rev_parse_repo["release"] + + assert repo.rev_parse("@{0}") == merge + assert repo.rev_parse("@{+0}") == merge + assert repo.rev_parse("@{1}") == release + assert repo.rev_parse("%s@{0}" % main.name) == merge + assert repo.rev_parse("@{-1}") == side + + SymbolicReference.create(repo, "refs/remotes/origin/%s" % main.name, merge) + main.set_tracking_branch(RemoteReference(repo, "refs/remotes/origin/%s" % main.name)) + assert repo.rev_parse("%s@{upstream}" % main.name) == merge + + +def test_rev_parse_commit_message_search(rev_parse_repo): + repo = rev_parse_repo["repo"] + release = rev_parse_repo["release"] + merge = rev_parse_repo["merge"] + + assert repo.rev_parse(":/release") == release + assert repo.rev_parse("HEAD^{/release}") == release + assert repo.rev_parse("HEAD^{/!-release}") == merge + + +def test_rev_parse_rejects_invalid_object_specs(rev_parse_repo): + repo = rev_parse_repo["repo"] + + with pytest.raises(ValueError): + repo.rev_parse(":") + with pytest.raises(ValueError): + repo.rev_parse(":/") + with pytest.raises(ValueError): + repo.rev_parse(":/[") + with pytest.raises(ValueError): + repo.rev_parse("HEAD^{/[}") + with pytest.raises(ValueError): + repo.rev_parse("@{-0}") + with pytest.raises(ValueError): + repo.rev_parse("HEAD^{invalid}") + with pytest.raises(BadName): + repo.rev_parse(":missing") diff --git a/test/test_submodule.py b/test/test_submodule.py index d88f9dab0..287986059 100644 --- a/test/test_submodule.py +++ b/test/test_submodule.py @@ -26,9 +26,9 @@ from git.objects.submodule.base import Submodule from git.objects.submodule.root import RootModule, RootUpdateProgress from git.repo.fun import find_submodule_git_dir, touch -from git.util import HIDE_WINDOWS_KNOWN_ERRORS, join_path_native, to_native_path_linux +from git.util import HIDE_WINDOWS_KNOWN_ERRORS, cwd, join_path_native, to_native_path_linux -from test.lib import TestBase, with_rw_directory, with_rw_repo +from test.lib import TestBase, with_rw_directory, with_rw_repo, PathLikeMock @contextlib.contextmanager @@ -58,6 +58,7 @@ def update(self, op, cur_count, max_count, message=""): print(op, cur_count, max_count, message) +TestRootProgress.__test__ = False prog = TestRootProgress() @@ -123,7 +124,7 @@ def _do_base_tests(self, rwrepo): else: with sm.config_writer() as writer: # For faster checkout, set the url to the local path. - new_smclone_path = Git.polish_url(osp.join(self.rorepo.working_tree_dir, sm.path)) + new_smclone_path = self._gitdb_repo_url() writer.set_value("url", new_smclone_path) writer.release() assert sm.config_reader().get_value("url") == new_smclone_path @@ -174,6 +175,10 @@ def _do_base_tests(self, rwrepo): sma = Submodule.add(rwrepo, sm.name, sm.path) assert sma.path == sm.path + # Adding existing as pathlike + sma = Submodule.add(rwrepo, sm.name, PathLikeMock(sm.path)) + assert sma.path == sm.path + # No url and no module at path fails. self.assertRaises(ValueError, Submodule.add, rwrepo, "newsubm", "pathtorepo", url=None) @@ -234,7 +239,7 @@ def _do_base_tests(self, rwrepo): # Adjust the path of the submodules module to point to the local # destination. - new_csmclone_path = Git.polish_url(osp.join(self.rorepo.working_tree_dir, sm.path, csm.path)) + new_csmclone_path = self._small_repo_url() with csm.config_writer() as writer: writer.set_value("url", new_csmclone_path) assert csm.url == new_csmclone_path @@ -475,11 +480,6 @@ def test_base_rw(self, rwrepo): def test_base_bare(self, rwrepo): self._do_base_tests(rwrepo) - @pytest.mark.xfail( - sys.platform == "cygwin", - reason="Cygwin GitPython can't find submodule SHA", - raises=ValueError, - ) @pytest.mark.xfail( HIDE_WINDOWS_KNOWN_ERRORS, reason=( @@ -491,15 +491,15 @@ def test_base_bare(self, rwrepo): @with_rw_repo(k_subm_current, bare=False) def test_root_module(self, rwrepo): # Can query everything without problems. - rm = RootModule(self.rorepo) - assert rm.module() is self.rorepo + rm = RootModule(rwrepo) + assert rm.module() is rwrepo # Try attributes. rm.binsha rm.mode rm.path assert rm.name == rm.k_root_name - assert rm.parent_commit == self.rorepo.head.commit + assert rm.parent_commit == rwrepo.head.commit rm.url rm.branch @@ -508,13 +508,12 @@ def test_root_module(self, rwrepo): with rm.config_writer(): pass - # Deep traversal gitdb / async. - rsmsp = [sm.path for sm in rm.traverse()] - assert len(rsmsp) >= 2 # gitdb and async [and smmap], async being a child of gitdb. - # Cannot set the parent commit as root module's path didn't exist. self.assertRaises(ValueError, rm.set_parent_commit, "HEAD") + with mock.patch.object(RootModule, "list_items", side_effect=ValueError("boom")): + rm.update(keep_going=True) + # TEST UPDATE ############# # Set up a commit that removes existing, adds new and modifies existing @@ -533,7 +532,7 @@ def test_root_module(self, rwrepo): # Ensure we clone from a local source. with sm.config_writer() as writer: - writer.set_value("url", Git.polish_url(osp.join(self.rorepo.working_tree_dir, sm.path))) + writer.set_value("url", self._gitdb_repo_url()) # Dry-run does nothing. sm.update(recursive=False, dry_run=True, progress=prog) @@ -568,7 +567,7 @@ def test_root_module(self, rwrepo): # ============== nsmn = "newsubmodule" nsmp = "submrepo" - subrepo_url = Git.polish_url(osp.join(self.rorepo.working_tree_dir, rsmsp[0], rsmsp[1])) + subrepo_url = self._small_repo_url() nsm = Submodule.add(rwrepo, nsmn, nsmp, url=subrepo_url) csmadded = rwrepo.index.commit("Added submodule").hexsha # Make sure we don't keep the repo reference. nsm.set_parent_commit(csmadded) @@ -633,7 +632,7 @@ def test_root_module(self, rwrepo): # ...to the first repository. This way we have a fast checkout, and a completely # different repository at the different url. nsm.set_parent_commit(csmremoved) - nsmurl = Git.polish_url(osp.join(self.rorepo.working_tree_dir, rsmsp[0])) + nsmurl = self._gitdb_repo_url() with nsm.config_writer() as writer: writer.set_value("url", nsmurl) csmpathchange = rwrepo.index.commit("changed url") @@ -713,6 +712,166 @@ def test_iter_items_from_invalid_hash(self): next(it) self.assertIsNone(ctx.exception.value) + @with_rw_directory + def test_deinit_calls_git_submodule(self, rwdir): + repo = git.Repo.init(rwdir) + submodule = Submodule(repo, b"\0" * 20, name="module", path="module") + + with mock.patch.object(Git, "submodule", create=True) as git_submodule: + submodule.deinit() + + git_submodule.assert_called_once_with("deinit", "--", submodule.path) + git_submodule.reset_mock() + + submodule.deinit(force=True) + + git_submodule.assert_called_once_with("deinit", "--force", "--", submodule.path) + + @with_rw_directory + def test_update_after_deinit(self, rwdir): + source_path = osp.join(rwdir, "source") + source_repo = git.Repo.init(source_path) + touch(osp.join(source_path, "file")) + source_repo.index.add(["file"]) + source_repo.index.commit("initial commit") + + parent_path = osp.join(rwdir, "parent") + parent_repo = git.Repo.init(parent_path) + submodule = parent_repo.create_submodule("module", "module", source_path) + parent_repo.index.commit("add submodule") + + submodule.deinit() + assert not submodule.module_exists() + assert osp.isdir(osp.join(parent_repo.git_dir, "modules", submodule.name)) + + submodule.update() + + assert submodule.module_exists() + assert submodule.module().head.commit == source_repo.head.commit + assert osp.isfile(osp.join(submodule.abspath, "file")) + + @with_rw_directory + def test_update_to_latest_revision_after_deinit_fetches_remote(self, rwdir): + source_path = osp.join(rwdir, "source") + source_repo = git.Repo.init(source_path) + source_repo.git.commit(m="initial commit", allow_empty=True) + + parent_repo = git.Repo.init(osp.join(rwdir, "parent")) + submodule = parent_repo.create_submodule("module", "module", source_path) + parent_repo.index.commit("add submodule") + submodule.deinit() + + touch(osp.join(source_path, "new-file")) + source_repo.index.add(["new-file"]) + source_repo.index.commit("advance remote") + + submodule.update(to_latest_revision=True) + + assert submodule.module().head.commit == source_repo.head.commit + assert osp.isfile(osp.join(submodule.abspath, "new-file")) + + @with_rw_directory + def test_update_after_deinit_fetches_new_gitlink_commit(self, rwdir): + source_path = osp.join(rwdir, "source") + source_repo = git.Repo.init(source_path) + source_repo.git.commit(m="initial commit", allow_empty=True) + + parent_repo = git.Repo.init(osp.join(rwdir, "parent")) + submodule = parent_repo.create_submodule("module", "module", source_path) + parent_repo.index.commit("add submodule") + submodule.deinit() + + touch(osp.join(source_path, "new-file")) + source_repo.index.add(["new-file"]) + source_repo.index.commit("advance remote") + parent_repo.git.update_index( + "--cacheinfo", + f"160000,{source_repo.head.commit.hexsha},{submodule.path}", + ) + parent_repo.index.commit("advance submodule") + + submodule = parent_repo.submodule(submodule.name) + submodule.update() + + assert submodule.module().head.commit == source_repo.head.commit + assert osp.isfile(osp.join(submodule.abspath, "new-file")) + + @with_rw_directory + def test_update_after_deinit_refuses_non_empty_checkout(self, rwdir): + source_path = osp.join(rwdir, "source") + source_repo = git.Repo.init(source_path) + tracked_file = osp.join(source_path, "file") + with open(tracked_file, "w") as fp: + fp.write("submodule content") + source_repo.index.add(["file"]) + source_repo.index.commit("initial commit") + + parent_repo = git.Repo.init(osp.join(rwdir, "parent")) + submodule = parent_repo.create_submodule("module", "module", source_path) + parent_repo.index.commit("add submodule") + submodule.deinit() + + checkout_file = osp.join(submodule.abspath, "file") + with open(checkout_file, "w") as fp: + fp.write("user content") + + with pytest.raises(OSError, match="does already exist and is non-empty"): + submodule.update() + + with open(checkout_file) as fp: + assert fp.read() == "user content" + assert not osp.exists(osp.join(submodule.abspath, ".git")) + + @with_rw_directory + def test_update_after_deinit_without_init_does_not_restore_checkout(self, rwdir): + source_path = osp.join(rwdir, "source") + source_repo = git.Repo.init(source_path) + source_repo.git.commit(m="initial commit", allow_empty=True) + + parent_repo = git.Repo.init(osp.join(rwdir, "parent")) + submodule = parent_repo.create_submodule("module", "module", source_path) + parent_repo.index.commit("add submodule") + submodule.deinit() + + assert submodule.update(init=False) is submodule + assert not submodule.module_exists() + assert not osp.exists(osp.join(submodule.abspath, ".git")) + + @with_rw_directory + def test_dry_run_update_after_deinit_does_not_restore_checkout(self, rwdir): + source_path = osp.join(rwdir, "source") + source_repo = git.Repo.init(source_path) + source_repo.git.commit(m="initial commit", allow_empty=True) + + parent_repo = git.Repo.init(osp.join(rwdir, "parent")) + submodule = parent_repo.create_submodule("module", "module", source_path) + parent_repo.index.commit("add submodule") + submodule.deinit() + + assert submodule.update(dry_run=True) is submodule + assert not submodule.module_exists() + assert not osp.exists(osp.join(submodule.abspath, ".git")) + + @with_rw_directory + def test_update_after_deinit_restores_nested_checkout(self, rwdir): + source_path = osp.join(rwdir, "source") + source_repo = git.Repo.init(source_path) + touch(osp.join(source_path, "file")) + source_repo.index.add(["file"]) + source_repo.index.commit("initial commit") + + parent_repo = git.Repo.init(osp.join(rwdir, "parent")) + submodule = parent_repo.create_submodule("nested/module", "deps/module", source_path) + parent_repo.index.commit("add nested submodule") + submodule.deinit() + + submodule.update() + + module_repo = submodule.module() + assert module_repo.head.commit == source_repo.head.commit + assert osp.isfile(osp.join(submodule.abspath, "file")) + assert osp.samefile(module_repo.git_dir, osp.join(parent_repo.git_dir, "modules", submodule.name)) + @with_rw_repo(k_no_subm_tag, bare=False) def test_first_submodule(self, rwrepo): assert len(list(rwrepo.iter_submodules())) == 0 @@ -753,6 +912,97 @@ def test_add_empty_repo(self, rwdir): ) # END for each checkout mode + @with_rw_directory + @_patch_git_config("protocol.file.allow", "always") + def test_update_submodule_with_relative_path(self, rwdir): + repo_path = osp.join(rwdir, "parent") + repo = git.Repo.init(repo_path) + module_repo_path = osp.join(rwdir, "module") + module_repo = git.Repo.init(module_repo_path) + module_repo.git.commit(m="test", allow_empty=True) + repo.git.submodule("add", "../module", "module") + repo.index.commit("add submodule") + + cloned_repo_path = osp.join(rwdir, "cloned_repo") + cloned_repo = git.Repo.clone_from(repo_path, cloned_repo_path) + + cloned_repo.submodule_update(init=True, recursive=True) + + @with_rw_directory + @_patch_git_config("protocol.file.allow", "always") + def test_update_rejects_parent_component_in_name(self, rwdir): + source = git.Repo.init(osp.join(rwdir, "source")) + source.git.commit(m="initial commit", allow_empty=True) + + parent = git.Repo.init(osp.join(rwdir, "parent")) + parent.git.submodule("add", source.working_tree_dir, "module") + parent.index.commit("add submodule") + modules_file = Path(parent.working_tree_dir) / ".gitmodules" + modules_file.write_text( + modules_file.read_text().replace('submodule "module"', 'submodule "../../../escaped/module"') + ) + parent.index.add([".gitmodules"]) + parent.index.commit("change submodule name") + + clone = git.Repo.clone_from(parent.working_tree_dir, osp.join(rwdir, "clone")) + with pytest.raises(ValueError, match="submodule name"): + clone.submodules[0].update(init=True) + assert not osp.exists(osp.join(rwdir, "escaped")) + + Path(rwdir, "escaped").mkdir() + git.Repo.clone_from( + source.working_tree_dir, + osp.join(clone.working_tree_dir, "module"), + separate_git_dir=osp.join(rwdir, "escaped", "module"), + ) + with pytest.raises(ValueError, match="submodule name"): + clone.submodules[0].update(init=True) + + invalid_names = ( + "", + "..", + "../module", + R"..\module", + "nested/../module", + R"nested\..\module", + "/module", + R"\module", + "C:module", + R"C:\module", + ) + for name in invalid_names: + with pytest.raises(ValueError, match="submodule name"): + Submodule._module_abspath(clone, "module", name) + + with mock.patch.object(Submodule, "_need_gitfile_submodules", return_value=False): + with pytest.raises(ValueError, match="submodule name"): + Submodule._module_abspath(clone, "module", "../module") + + @with_rw_directory + @_patch_git_config("protocol.file.allow", "always") + def test_root_update_keeps_going_after_invalid_submodule_name(self, rwdir): + source = git.Repo.init(osp.join(rwdir, "source")) + source.git.commit(m="initial commit", allow_empty=True) + + parent = git.Repo.init(osp.join(rwdir, "parent")) + parent.git.submodule("add", source.working_tree_dir, "invalid") + parent.git.submodule("add", source.working_tree_dir, "valid") + modules_file = Path(parent.working_tree_dir) / ".gitmodules" + modules_file.write_text(modules_file.read_text().replace('submodule "invalid"', 'submodule "../invalid"')) + parent.index.add([".gitmodules"]) + parent.index.commit("add submodules") + + clone = git.Repo.clone_from(parent.working_tree_dir, osp.join(rwdir, "clone")) + assert [sm.name for sm in clone.submodules] == ["../invalid", "valid"] + + clone.submodule_update(keep_going=True) + + assert os.listdir(osp.join(clone.working_tree_dir, "invalid")) == [] + assert not clone.submodule("../invalid").module_exists() + assert clone.submodule("valid").module_exists() + + clone.submodule("../invalid").update(recursive=True, keep_going=True) + @with_rw_directory @_patch_git_config("protocol.file.allow", "always") def test_list_only_valid_submodules(self, rwdir): @@ -916,7 +1166,7 @@ def assert_exists(sm, value=True): csm.repo.index.commit("Have to commit submodule change for algorithm to pick it up") assert csm.url == "bar" - self.assertRaises( + self.assertRaises( # noqa: B017 Exception, rsm.update, recursive=True, @@ -990,6 +1240,7 @@ def test_rename(self, rwdir): # garbage collector detailed in https://github.com/python/cpython/issues/97922.) if sys.platform == "win32" and sys.version_info >= (3, 12): gc.collect() + gc.collect() # Some finalizer scenarios need two collections, at least in theory. new_path = "renamed/myname" assert sm.move(new_path).name == new_path @@ -1083,6 +1334,61 @@ class Repo: msg = '_to_relative_path should be "submodule_path" but was "%s"' % relative_path assert relative_path == "submodule_path", msg + @with_rw_directory + def test_to_relative_path_is_confined_to_parent(self, rwdir): + class Repo: + working_tree_dir = osp.join(rwdir, "parent") + + os.makedirs(osp.join(Repo.working_tree_dir, "nested")) + + assert Submodule._to_relative_path(Repo(), osp.join(Repo.working_tree_dir, "nested")) == "nested" + assert Submodule._to_relative_path(Repo(), osp.join("nested", "..", "module")) == "module" + self.assertRaises(ValueError, Submodule._to_relative_path, Repo(), Repo.working_tree_dir) + self.assertRaises(ValueError, Submodule._to_relative_path, Repo(), osp.join("..", "outside")) + self.assertRaises( + ValueError, + Submodule._to_relative_path, + Repo(), + osp.join(Repo.working_tree_dir + "-other", "module"), + ) + + @skipUnless(sys.platform == "win32", "Specifically for Windows.") + @with_rw_directory + def test_to_relative_path_windows_path_kinds(self, rwdir): + class Repo: + working_tree_dir = osp.join(rwdir, "parent") + + inside_path = osp.join(Repo.working_tree_dir, "nested", "module") + _drive, rooted_inside_path = osp.splitdrive(inside_path) + + assert Submodule._to_relative_path(Repo(), rooted_inside_path) == "nested/module" + assert Submodule._to_relative_path(Repo(), rooted_inside_path.replace("\\", "/")) == "nested/module" + assert Submodule._to_relative_path(Repo(), PathLikeMock(inside_path)) == "nested/module" + self.assertRaises( + ValueError, + Submodule._to_relative_path, + Repo(), + f"{osp.splitdrive(Repo.working_tree_dir)[0]}relative", + ) + self.assertRaises(ValueError, Submodule._to_relative_path, Repo(), R"Z:\outside") + self.assertRaises(ValueError, Submodule._to_relative_path, Repo(), R"\\server\share\outside") + self.assertRaises(ValueError, Submodule._to_relative_path, Repo(), R"\\?\C:\outside") + + @skipUnless(sys.platform == "win32", "Specifically for Windows.") + @with_rw_directory + def test_find_submodule_git_dir_with_drive_rooted_gitfile(self, rwdir): + with cwd(rwdir): + module = git.Repo.init(osp.join(rwdir, "module")) + _drive, rooted_git_dir = osp.splitdrive(module.git_dir) + gitfile = osp.join(rwdir, "gitfile") + with open(gitfile, "w") as stream: + stream.write(f"gitdir: {rooted_git_dir}") + + resolved_git_dir = find_submodule_git_dir(gitfile) + + assert resolved_git_dir is not None + assert osp.samefile(resolved_git_dir, module.git_dir) + @pytest.mark.xfail( reason="for some unknown reason the assertion fails, even though it in fact is working in more common setup", raises=AssertionError, @@ -1178,6 +1484,20 @@ def test_add_no_clone_multi_options_argument(self, rwdir): with self.assertRaises(cp.NoOptionError): sm_config.get_value("core", "eol") + @with_rw_directory + def test_add_does_not_expand_environment_variables_in_url(self, rwdir): + parent = git.Repo.init(osp.join(rwdir, "parent")) + url = osp.join(rwdir, "${GITPYTHON_TEST_SECRET}") + source = git.Repo.init(url) + Path(source.working_tree_dir, "file").write_text("content") + source.index.add(["file"]) + source.index.commit("initial") + + with mock.patch.dict(os.environ, {"GITPYTHON_TEST_SECRET": "sensitive-value"}): + sm = Submodule.add(parent, "new", "new", url) + + assert sm.url == Git.polish_url(url, expand_vars=False) + @with_rw_repo("HEAD") def test_submodule_add_unsafe_url(self, rw_repo): with tempfile.TemporaryDirectory() as tdir: @@ -1310,6 +1630,17 @@ def test_submodule_update_unsafe_options(self, rw_repo): submodule.update(clone_multi_options=[unsafe_option]) assert not tmp_file.exists() + @with_rw_repo("HEAD") + def test_submodule_update_unsafe_options_are_checked_after_splitting_multi_options(self, rw_repo): + with tempfile.TemporaryDirectory() as tdir: + tmp_dir = Path(tdir) + payload = "--single-branch --config protocol.ext.allow=always" + submodule = Submodule(rw_repo, b"\0" * 20, name="new", path="new", url=str(tmp_dir)) + + with self.assertRaises(UnsafeOptionError): + submodule.update(clone_multi_options=[payload]) + assert not submodule.module_exists() + @with_rw_repo("HEAD") def test_submodule_update_unsafe_options_allowed(self, rw_repo): with tempfile.TemporaryDirectory() as tdir: @@ -1334,3 +1665,23 @@ def test_submodule_update_unsafe_options_allowed(self, rw_repo): for unsafe_option in unsafe_options: with self.assertRaises(GitCommandError): submodule.update(clone_multi_options=[unsafe_option], allow_unsafe_options=True) + + @with_rw_directory + @_patch_git_config("protocol.file.allow", "always") + def test_submodule_update_relative_url(self, rwdir): + parent_path = osp.join(rwdir, "parent") + parent_repo = git.Repo.init(parent_path) + submodule_path = osp.join(rwdir, "module") + submodule_repo = git.Repo.init(submodule_path) + submodule_repo.git.commit(m="initial commit", allow_empty=True) + + parent_repo.git.submodule("add", "../module", "module") + parent_repo.index.commit("add submodule with relative URL") + + cloned_path = osp.join(rwdir, "cloned_repo") + cloned_repo = git.Repo.clone_from(parent_path, cloned_path) + + cloned_repo.submodule_update(init=True, recursive=True) + + has_module = any(sm.name == "module" for sm in cloned_repo.submodules) + assert has_module, "Relative submodule was not updated properly" diff --git a/test/test_tree.py b/test/test_tree.py index 73158113d..629fd4d32 100644 --- a/test/test_tree.py +++ b/test/test_tree.py @@ -8,10 +8,14 @@ from pathlib import Path import subprocess +import pytest + from git.objects import Blob, Tree +from git.repo import Repo from git.util import cwd from test.lib import TestBase, with_rw_directory +from .lib.helper import PathLikeMock, with_rw_repo class TestTree(TestBase): @@ -126,12 +130,18 @@ def test_traverse(self): assert len(list(root)) == len(list(root.traverse(depth=1))) # Only choose trees. - trees_only = lambda i, d: i.type == "tree" + + def trees_only(i, _d): + return i.type == "tree" + trees = list(root.traverse(predicate=trees_only)) assert len(trees) == len([i for i in root.traverse() if trees_only(i, 0)]) # Test prune. - lib_folder = lambda t, d: t.path == "lib" + + def lib_folder(t, _d): + return t.path == "lib" + pruned_trees = list(root.traverse(predicate=trees_only, prune=lib_folder)) assert len(pruned_trees) < len(trees) @@ -155,3 +165,57 @@ def test_traverse(self): assert root[item.path] == item == root / item.path # END for each item assert found_slash + + @with_rw_repo("0.3.2.1") + def test_repo_lookup_string_path(self, rw_repo): + repo = Repo(rw_repo.git_dir) + blob = repo.tree() / ".gitignore" + assert isinstance(blob, Blob) + assert blob.hexsha == "787b3d442a113b78e343deb585ab5531eb7187fa" + + @with_rw_repo("0.3.2.1") + def test_repo_lookup_pathlike_path(self, rw_repo): + repo = Repo(rw_repo.git_dir) + blob = repo.tree() / PathLikeMock(".gitignore") + assert isinstance(blob, Blob) + assert blob.hexsha == "787b3d442a113b78e343deb585ab5531eb7187fa" + + @with_rw_repo("0.3.2.1") + def test_repo_lookup_invalid_string_path(self, rw_repo): + repo = Repo(rw_repo.git_dir) + with pytest.raises(KeyError): + repo.tree() / "doesnotexist" + + @with_rw_repo("0.3.2.1") + def test_repo_lookup_invalid_pathlike_path(self, rw_repo): + repo = Repo(rw_repo.git_dir) + with pytest.raises(KeyError): + repo.tree() / PathLikeMock("doesnotexist") + + @with_rw_repo("0.3.2.1") + def test_repo_lookup_nested_string_path(self, rw_repo): + repo = Repo(rw_repo.git_dir) + blob = repo.tree() / "git/__init__.py" + assert isinstance(blob, Blob) + assert blob.hexsha == "d87dcbdbb65d2782e14eea27e7f833a209c052f3" + + @with_rw_repo("0.3.2.1") + def test_repo_lookup_nested_pathlike_path(self, rw_repo): + repo = Repo(rw_repo.git_dir) + blob = repo.tree() / PathLikeMock("git/__init__.py") + assert isinstance(blob, Blob) + assert blob.hexsha == "d87dcbdbb65d2782e14eea27e7f833a209c052f3" + + @with_rw_repo("0.3.2.1") + def test_repo_lookup_folder_string_path(self, rw_repo): + repo = Repo(rw_repo.git_dir) + tree = repo.tree() / "git" + assert isinstance(tree, Tree) + assert tree.hexsha == "ec8ae429156d65afde4bbb3455570193b56f0977" + + @with_rw_repo("0.3.2.1") + def test_repo_lookup_folder_pathlike_path(self, rw_repo): + repo = Repo(rw_repo.git_dir) + tree = repo.tree() / PathLikeMock("git") + assert isinstance(tree, Tree) + assert tree.hexsha == "ec8ae429156d65afde4bbb3455570193b56f0977" diff --git a/test/test_util.py b/test/test_util.py index dad2f3dcd..cf4299d5a 100644 --- a/test/test_util.py +++ b/test/test_util.py @@ -34,12 +34,13 @@ LockFile, cygpath, decygpath, + is_cygwin_git, get_user_id, remove_password_if_present, rmtree, ) -from test.lib import TestBase, with_rw_repo +from test.lib import TestBase, requires_symlinks, with_rw_repo @pytest.fixture @@ -112,7 +113,8 @@ def test_deletes_dir_with_readonly_files(self, tmp_path): sys.platform == "cygwin", reason="Cygwin can't set the permissions that make the test meaningful.", ) - def test_avoids_changing_permissions_outside_tree(self, tmp_path): + @requires_symlinks + def test_avoids_changing_permissions_outside_tree(self, tmp_path, request): # Automatically works on Windows, but on Unix requires either special handling # or refraining from attempting to fix PermissionError by making chmod calls. @@ -124,9 +126,32 @@ def test_avoids_changing_permissions_outside_tree(self, tmp_path): dir2 = tmp_path / "dir2" dir2.mkdir() - (dir2 / "symlink").symlink_to(dir1 / "file") + symlink = dir2 / "symlink" + symlink.symlink_to(dir1 / "file") dir2.chmod(stat.S_IRUSR | stat.S_IXUSR) + def preen_dir2(): + """Don't leave unwritable directories behind. + + pytest has difficulties cleaning up after the fact on some platforms, + e.g., macOS, and whines incessantly until the issue is resolved--regardless + of the pytest session. + """ + rwx = stat.S_IRUSR | stat.S_IWUSR | stat.S_IXUSR + if not dir2.exists(): + return + if symlink.exists(): + try: + # Try lchmod first, if the platform supports it. + symlink.lchmod(rwx) + except NotImplementedError: + # The platform (probably win32) doesn't support lchmod; fall back to chmod. + symlink.chmod(rwx) + dir2.chmod(rwx) + rmtree(dir2) + + request.addfinalizer(preen_dir2) + try: rmtree(dir2) except PermissionError: @@ -349,6 +374,24 @@ def test_decygpath(self, wpath, cpath): assert wcpath == wpath.replace("/", "\\"), cpath +class TestIsCygwinGit: + """Tests for :func:`is_cygwin_git`""" + + def test_on_path_executable(self): + # Currently we assume tests run on Cygwin use Cygwin git. See #533 and #1455 for background. + if sys.platform == "cygwin": + assert is_cygwin_git("git") + else: + assert not is_cygwin_git("git") + + def test_none_executable(self): + assert not is_cygwin_git(None) + + def test_with_missing_uname(self): + """Test for handling when `uname` isn't in the same directory as `git`""" + assert not is_cygwin_git("/bogus_path/git") + + class _Member: """A member of an IterableList.""" @@ -365,6 +408,10 @@ def __repr__(self): class TestUtils(TestBase): """Tests for most utilities in :mod:`git.util`.""" + @pytest.mark.skipif(os.name != "nt", reason="Specifically for Windows drive-rooted paths.") + def test_cygpath_drive_rooted_path(self): + assert cygpath(R"\directory\file") == "/directory/file" + def test_it_should_dashify(self): self.assertEqual("this-is-my-argument", dashify("this_is_my_argument")) self.assertEqual("foo", dashify("foo")) @@ -629,6 +676,7 @@ def test_pickle_tzoffset(self): def test_remove_password_from_command_line(self): username = "fakeuser" password = "fakepassword1234" + authorization = "Bearer fake-token-1234" url_with_user_and_pass = "https://{}:{}@fakerepo.example.com/testrepo".format(username, password) url_with_user = "https://{}@fakerepo.example.com/testrepo".format(username) url_with_pass = "https://:{}@fakerepo.example.com/testrepo".format(password) @@ -639,6 +687,7 @@ def test_remove_password_from_command_line(self): cmd_3 = ["git", "clone", "-v", url_with_pass] cmd_4 = ["git", "clone", "-v", url_without_user_or_pass] cmd_5 = ["no", "url", "in", "this", "one"] + cmd_6 = ["git", "-c", "http.extraHeader=Authorization: %s" % authorization, "fetch"] redacted_cmd_1 = remove_password_if_present(cmd_1) assert username not in " ".join(redacted_cmd_1) @@ -658,3 +707,7 @@ def test_remove_password_from_command_line(self): assert cmd_4 == remove_password_if_present(cmd_4) assert cmd_5 == remove_password_if_present(cmd_5) + + redacted_cmd_6 = remove_password_if_present(cmd_6) + assert authorization not in " ".join(redacted_cmd_6) + assert "http.extraHeader=Authorization: *****" in redacted_cmd_6 diff --git a/tox.ini b/tox.ini index fc62fa587..fc3f821d6 100644 --- a/tox.ini +++ b/tox.ini @@ -1,6 +1,6 @@ [tox] requires = tox>=4 -env_list = py{37,38,39,310,311,312}, ruff, format, mypy, html, misc +env_list = py{37,38,39,310,311,312,313,314,315}, ruff, format, mypy, html, misc [testenv] description = Run unit tests