diff --git a/.github/FUNDING.yml b/.github/FUNDING.yml new file mode 100644 index 00000000..4d68e318 --- /dev/null +++ b/.github/FUNDING.yml @@ -0,0 +1,3 @@ +# These are supported funding model platforms + +github: harelba diff --git a/.github/workflows/build-and-package.yaml b/.github/workflows/build-and-package.yaml new file mode 100644 index 00000000..37add713 --- /dev/null +++ b/.github/workflows/build-and-package.yaml @@ -0,0 +1,528 @@ +name: BuildAndPackage + +on: + push: + tags: + - "v*" + branches: master + pull_request: + branches: master + paths-ignore: + - "*.md" + - "*.markdown" + - "mkdocs/**/*" + tags-ignore: + - "*" + +jobs: + version_info: + runs-on: ubuntu-18.04 + steps: + - name: Checkout + uses: actions/checkout@v2 + - id: vars + run: | + set -x -e + + echo "github event ref is ${{ github.ref }}" + + if [ "x${{ startsWith(github.ref, 'refs/tags/v') }}" == "xtrue" ] + then + echo "Trigger was a version tag - ${{ github.ref }}" + echo ::set-output name=q_version::${GITHUB_REF#refs/tags/v} + echo ::set-output name=is_release::true + else + # For testing version propagation inside the PR + echo "Either branch of a non-version tag - setting version to 0.0.0" + echo ::set-output name=q_version::0.0.0 + echo ::set-output name=is_release::false + fi + + outputs: + q_version: ${{ steps.vars.outputs.q_version }} + is_release: ${{ steps.vars.outputs.is_release }} + + check_version_info: + runs-on: ubuntu-18.04 + needs: version_info + steps: + - name: test q_version + run: | + set -e -x + + echo "outputs: ${{ toJson(needs.version_info) }}" + + create-man: + runs-on: ubuntu-18.04 + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Install Ruby + uses: ruby/setup-ruby@v1 + with: + ruby-version: '2.6' + - name: Create man page + run: | + set -x -e + gem install ronn + + ronn doc/USAGE.markdown + # Must be gzipped, otherwise debian does not install it + gzip doc/USAGE + - name: Upload man page + uses: actions/upload-artifact@v1.0.0 + with: + name: q-man-page + path: doc/USAGE.gz + + build-linux: + runs-on: ubuntu-18.04 + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Cache pyox + uses: actions/cache@v2 + with: + path: | + ~/.cache/pyoxidizer + key: ${{ runner.os }}-pyox + - name: Install pyoxidizer + run: | + set -e -x + + sudo apt-get update + sudo apt-get install -y zip sqlite3 rpm + + curl -o pyoxidizer.zip -L "https://github.com/indygreg/PyOxidizer/releases/download/pyoxidizer%2F0.17/pyoxidizer-0.17.0-linux_x86_64.zip" + unzip pyoxidizer.zip + chmod +x ./pyoxidizer + - name: Create Q Executable - Linux + run: | + set -e -x + + ./pyoxidizer build --release + + export Q_EXECUTABLE=./build/x86_64-unknown-linux-gnu/release/install/q + chmod 755 $Q_EXECUTABLE + + seq 1 100 | $Q_EXECUTABLE -c 1 "select sum(c1),count(*) from -" -S test.sqlite + + mkdir -p packages/linux/ + cp $Q_EXECUTABLE packages/linux/linux-q + - name: Upload Linux Executable + uses: actions/upload-artifact@v1.0.0 + with: + name: linux-q + path: packages/linux/linux-q + + test-linux: + needs: build-linux + runs-on: ubuntu-18.04 + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Install Python for Testing + uses: actions/setup-python@v2 + with: + python-version: '3.8.12' + architecture: 'x64' + - name: Prepare Testing + run: | + set -e -x + + pip3 install -r test-requirements.txt + - name: Download Linux Executable + uses: actions/download-artifact@v2 + with: + name: linux-q + - name: Run Tests on Linux Executable + run: | + set -x -e + + find ./ -ls + + chmod 755 ./linux-q + + Q_EXECUTABLE=`pwd`/linux-q Q_SKIP_EXECUTABLE_VALIDATION=true ./run-tests.sh -v + + package-linux-deb: + needs: [test-linux, create-man, version_info] + runs-on: ubuntu-18.04 + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Install Ruby + uses: ruby/setup-ruby@v1 + with: + ruby-version: '2.6' + - name: Downoad man page + uses: actions/download-artifact@v2 + with: + name: q-man-page + - name: Download Linux Executable + uses: actions/download-artifact@v2 + with: + name: linux-q + - name: Build DEB Package + run: | + set -e -x + + mkdir -p packages/linux/ + + find ./ -ls + + chmod 755 ./linux-q + + export q_version=${{ needs.version_info.outputs.q_version }} + + gem install fpm + cp dist/fpm-config ~/.fpm + fpm -s dir -t deb --deb-use-file-permissions -p packages/linux/q-text-as-data-${q_version}-1.x86_64.deb --version ${q_version} ./linux-q=/usr/bin/q USAGE.gz=/usr/share/man/man1/q.1.gz + - name: Upload DEB Package + uses: actions/upload-artifact@v1.0.0 + with: + name: q-text-as-data-${{ needs.version_info.outputs.q_version }}-1.x86_64.deb + path: packages/linux/q-text-as-data-${{ needs.version_info.outputs.q_version }}-1.x86_64.deb + + test-deb-packaging: + runs-on: ubuntu-18.04 + needs: [package-linux-deb, version_info] + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Download DEB + uses: actions/download-artifact@v2 + with: + name: q-text-as-data-${{ needs.version_info.outputs.q_version }}-1.x86_64.deb + - name: Install Python for Testing + uses: actions/setup-python@v2 + with: + python-version: '3.8.12' + architecture: 'x64' + - name: Prepare Testing + run: | + set -e -x + + pip3 install -r test-requirements.txt + - name: Test DEB Package Installation + run: ./dist/test-using-deb.sh ./q-text-as-data-${{ needs.version_info.outputs.q_version }}-1.x86_64.deb + + package-linux-rpm: + needs: [test-linux, create-man, version_info] + runs-on: ubuntu-18.04 + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Install Ruby + uses: ruby/setup-ruby@v1 + with: + ruby-version: '2.6' + - name: Download man page + uses: actions/download-artifact@v2 + with: + name: q-man-page + - name: Download Linux Executable + uses: actions/download-artifact@v2 + with: + name: linux-q + - name: Build RPM Package + run: | + set -e -x + + mkdir -p packages/linux + + + chmod 755 ./linux-q + + export q_version=${{ needs.version_info.outputs.q_version }} + + gem install fpm + cp dist/fpm-config ~/.fpm + fpm -s dir -t rpm --rpm-use-file-permissions -p packages/linux/q-text-as-data-${q_version}.x86_64.rpm --version ${q_version} ./linux-q=/usr/bin/q USAGE.gz=/usr/share/man/man1/q.1.gz + - name: Upload RPM Package + uses: actions/upload-artifact@v1.0.0 + with: + name: q-text-as-data-${{ needs.version_info.outputs.q_version }}.x86_64.rpm + path: packages/linux/q-text-as-data-${{ needs.version_info.outputs.q_version }}.x86_64.rpm + + test-rpm-packaging: + runs-on: ubuntu-18.04 + needs: [package-linux-rpm, version_info] + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Download RPM + uses: actions/download-artifact@v2 + with: + name: q-text-as-data-${{ needs.version_info.outputs.q_version }}.x86_64.rpm + - name: Retest using RPM + run: ./dist/test-using-rpm.sh ./q-text-as-data-${{ needs.version_info.outputs.q_version }}.x86_64.rpm + + build-mac: + runs-on: macos-11 + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Cache pyox + uses: actions/cache@v2 + with: + path: | + ~/.cache/pyoxidizer + key: ${{ runner.os }}-pyox + - name: Install pyoxidizer + run: | + set -e -x + + curl -o pyoxidizer.zip -L "https://github.com/indygreg/PyOxidizer/releases/download/pyoxidizer%2F0.17/pyoxidizer-0.17.0-macos-universal.zip" + unzip pyoxidizer.zip + mv macos-universal/pyoxidizer ./pyoxidizer + + chmod +x ./pyoxidizer + - name: Create Q Executable - Mac + run: | + set -e -x + + ./pyoxidizer build --release + + export Q_EXECUTABLE=./build/x86_64-apple-darwin/release/install/q + chmod 755 $Q_EXECUTABLE + + seq 1 100 | $Q_EXECUTABLE -c 1 "select sum(c1),count(*) from -" -S test.sqlite + + mkdir -p packages/macos/ + cp $Q_EXECUTABLE packages/macos/macos-q + - name: Upload MacOS Executable + uses: actions/upload-artifact@v1.0.0 + with: + name: macos-q + path: packages/macos/macos-q + + test-mac: + needs: build-mac + runs-on: macos-11 + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Install Python for Testing + uses: actions/setup-python@v2 + with: + python-version: '3.8.12' + architecture: 'x64' + - name: Prepare Testing + run: | + set -e -x + + pip3 install wheel + + pip3 install -r test-requirements.txt + - name: Download MacOS Executable + uses: actions/download-artifact@v2 + with: + name: macos-q + - name: Run Tests on MacOS Executable + run: | + set -e -x + + chmod 755 ./macos-q + + Q_EXECUTABLE=`pwd`/macos-q Q_SKIP_EXECUTABLE_VALIDATION=true ./run-tests.sh -v + + not-package-mac: + # create-man is not needed, as it's generated inside the brew formula independently + needs: [test-mac] + runs-on: macos-11 + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Not Packaging Mac + run: | + echo "homebrew mac cannot be packaged from the source code itself, due to the package build process of homebrew. See https://github.com/harelba/homebrew-q" + + not-test-mac-packaging: + needs: not-package-mac + runs-on: macos-11 + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Not Testing Mac Packaging + run: | + echo "homebrew mac packaging cannot be tested here, due to the package build process of homebrew. See https://github.com/harelba/homebrew-q" + + build-windows: + runs-on: windows-latest + needs: version_info + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Install MSVC build tools + uses: ilammy/msvc-dev-cmd@v1 + - name: Install Python + uses: actions/setup-python@v2 + with: + python-version: '3.8.10' + architecture: 'x64' + - name: Install pyoxidizer + shell: bash + run: | + set -x -e + + python3 -V + pip3 -V + + pip3 install pyoxidizer + - name: Create Q Executable - Windows + shell: bash + run: | + set -e -x + + pyoxidizer build --release --var Q_VERSION ${{ needs.version_info.outputs.q_version }} + + export Q_EXECUTABLE=./build/x86_64-pc-windows-msvc/release/install/q + chmod 755 $Q_EXECUTABLE + + seq 1 100 | $Q_EXECUTABLE -c 1 "select sum(c1),count(*) from -" -S test.sqlite + + mkdir -p packages/windows/ + cp $Q_EXECUTABLE packages/windows/win-q.exe + + find ./ -ls + - name: Upload Linux Executable + uses: actions/upload-artifact@v1.0.0 + with: + name: win-q.exe + path: packages/windows/win-q.exe + + not-really-test-windows: + needs: build-windows + runs-on: windows-latest + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Install Python for Testing + uses: actions/setup-python@v2 + with: + python-version: '3.8' + architecture: 'x64' + - name: Download Windows Executable + uses: actions/download-artifact@v2 + with: + name: win-q.exe + - name: Not-Really-Test Windows + shell: bash + continue-on-error: true + run: | + echo "Tests are not compatible with Windows (path separators, tmp folder names etc.). Only a sanity wil be tested" + + chmod +x ./win-q.exe + + seq 1 10000 | ./win-q.exe -c 1 "select sum(c1),count(*) from -" -S some-db.sqlite + + package-windows: + needs: [create-man, not-really-test-windows, version_info] + runs-on: windows-latest + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Install MSVC build tools + uses: ilammy/msvc-dev-cmd@v1 + - name: Install Python + uses: actions/setup-python@v2 + with: + python-version: '3.8.10' + architecture: 'x64' + - name: Install pyoxidizer + shell: bash + run: | + set -x -e + + python3 -V + pip3 -V + + pip3 install pyoxidizer + - name: Create Q MSI - Windows + shell: bash + run: | + set -e -x + + pyoxidizer build --release msi_installer --var Q_VERSION ${{ needs.version_info.outputs.q_version }} + + export Q_MSI=./build/x86_64-pc-windows-msvc/release/msi_installer/q-text-as-data-${{ needs.version_info.outputs.q_version }}.msi + chmod 755 $Q_MSI + + mkdir -p packages/windows/ + cp $Q_MSI packages/windows/q-text-as-data-${{ needs.version_info.outputs.q_version }}.msi + + - name: Upload Windows MSI + uses: actions/upload-artifact@v1.0.0 + with: + name: q-text-as-data-${{ needs.version_info.outputs.q_version }}.msi + path: packages/windows/q-text-as-data-${{ needs.version_info.outputs.q_version }}.msi + + test-windows-packaging: + needs: [package-windows, version_info] + runs-on: windows-latest + steps: + - name: Checkout + uses: actions/checkout@v2 + - name: Download Windows Package + uses: actions/download-artifact@v2 + with: + name: q-text-as-data-${{ needs.version_info.outputs.q_version }}.msi + - name: Test Install of MSI + continue-on-error: true + shell: powershell + run: | + $process = Start-Process msiexec.exe -ArgumentList "/i q-text-as-data-${{ needs.version_info.outputs.q_version }}.msi -l* msi-install.log /norestart /quiet" -PassThru -Wait + $process.ExitCode + gc msi-install.log + + exit $process.ExitCode + - name: Test Uninstall of MSI + continue-on-error: true + shell: powershell + run: | + $process = Start-Process msiexec.exe -ArgumentList "/u q-text-as-data-${{ needs.version_info.outputs.q_version }}.msi /norestart /quiet" -PassThru -Wait + $process.ExitCode + exit $process.ExitCode + + perform-prerelease: + # We'd like artifacts to be uploaded regardless of tests succeeded or not, + # this is why the dependency here is not on test-X-packaging jobs + needs: [package-linux-deb, package-linux-rpm, not-package-mac, package-windows, version_info] + runs-on: ubuntu-latest + if: needs.version_info.outputs.is_release == 'false' + steps: + - name: Download All Artifacts + uses: actions/download-artifact@v2 + with: + path: artifacts/ + - name: Timestamp pre-release + run: | + set -e -x + + echo "Workflow finished at $(date)" >> artifacts/workflow-finish-time.txt + - name: Create pre-release + uses: "marvinpinto/action-automatic-releases@v1.2.1" + with: + repo_token: "${{ secrets.GITHUB_TOKEN }}" + automatic_release_tag: "latest" + prerelease: true + title: "Next Release Development Build" + files: | + artifacts/**/* + + perform-release: + needs: [not-test-mac-packaging, test-deb-packaging, test-rpm-packaging, test-windows-packaging, version_info] + runs-on: ubuntu-latest + if: needs.version_info.outputs.is_release == 'true' + steps: + - name: Download All Artifacts + uses: actions/download-artifact@v2 + with: + path: artifacts/ + - uses: "marvinpinto/action-automatic-releases@v1.2.1" + with: + repo_token: "${{ secrets.GITHUB_TOKEN }}" + prerelease: false + files: | + artifacts/**/* diff --git a/.gitignore b/.gitignore index a1e04869..eb3a2d28 100644 --- a/.gitignore +++ b/.gitignore @@ -11,4 +11,15 @@ win_build packages .idea/ dist/windows/ -dist/ +generated-site/ +benchmark_data.tar.gz +_benchmark_data/ +q.egg-info/ +.pytest_cache/ +*.qsql +htmlcov/ +*.sqlite +*.tar.gz +.coverage +.DS_Store +*.egg diff --git a/.travis.yml b/.travis.yml deleted file mode 100644 index 60d000eb..00000000 --- a/.travis.yml +++ /dev/null @@ -1,5 +0,0 @@ -language: python -python: - - "2.6" - - "2.7" -script: test/test-all diff --git a/QSQL-NOTES.md b/QSQL-NOTES.md new file mode 100644 index 00000000..5f61f83a --- /dev/null +++ b/QSQL-NOTES.md @@ -0,0 +1,106 @@ + +## Major changes and additions in the new `3.x` version +This is the list of new/changed functionality in this version. Large changes, please make sure to read the details if you're already using q. + +* **Automatic Immutable Caching** - Automatic caching of data files (into `.qsql` files), with huge speedups for medium/large files. Enabled through `-C readwrite` or `-C read` +* **Direct querying of standard sqlite databases** - Just use it as a table name in the query. Format is `select ... from :::`, or just `` if the database contains only one table. Multiple separate sqlite databases are fully supported in the same query. +* **Direct querying of the `qsql` cache files** - The user can query directly from the `qsql` files, removing the need for the original files. Just use `select ... from .qsql`. Please wait until the non-beta version is out before thinking about deleting any of your original files... +* **Revamped `.qrc` mechanism** - allows opting-in to caching without specifying it in every query. By default, caching is **disabled**, for backward compatibility and for finding usability issues. +* **Save-to-db is now reusable for queries** - `--save-db-to-disk` option (`-S`) has been enhanced to match the new capabilities. You can query the resulting file directly through q, using the method mentioned above (it's just a standard sqlite database). +* **Only python3 is supported from now on** - Shouldn't be an issue, since q is a self-contained binary executable which has its own python embedded in it. Internally, q is now packaged with Python 3.8. After everything cools down, I'll probably bump this to 3.9/3.10. +* **Minimal Linux Version Bumped** - Works with CentOS 8, Ubuntu 18.04+, Debian 10+. Currently only for x86_64. Depends on glibc version 2.25+. Haven't tested it on other architectures. Issuing other architectures will be possible later on +* **Completely revamped binary packaging** - Using [pyoxidizer](https://github.com/indygreg/PyOxidizer) + +The following sections provide the details of each of the new functionalities in this major version. + +## Automatic caching of data files +Speeding up subsequent reads from the same file by several orders of magnitude by automatically creating an immutable cache file for each tabular text file. + +For example, reading a 0.9GB file with 1M rows and 100 columns without caching takes ~50 seconds. When the cache exists, querying the same file will take around ~1-2 seconds. Obviously, the cache can be used in order to perform any query and not just the original query that was used for creating the cache. + +When caching is enabled, the cache is created on the first read of a file, and used automatically when reading it in other queries. A separate cache is being created for each file that is being used, allowing reuse in multiple use-cases. For example, if two csv files each have their own cache file from previous queries, then running a query that JOINs these two files would use the caches as well (without loading the data into memory), speeding it up considerably. + +The tradeoff for using cache files is disk space - A new file with the postfix `.qsql` is created and automatically detected and used in queries as needed. This file is essentially a standard sqlite file (with some additional metadata tables), and can be used directly by any standard sqlite tool later on. + +For backward compatibility, the caching option is not turned on by default. You'd need to use the new `-C ` to determine the caching mode. Available options are as follows: +* `none` - The default, provides the original q's behaviour without caching +* `read` - Only reads cache files if they exists, but doesn't create any new ones +* `readwrite` - Uses cache files if they exists, or creates new ones if they don't. Writing new cache files doesn't interfere with the actual run of the query, so this option can be used in order to dynamically create the cache files if they don't exist + +Content signatures are being stored in the caches, allowing to detect a state where the original file has been modified after the cache has been created. q will issue an error if this happens. For now, just delete the `.qsql` file in order to recreate the cache. In the future, another `-C` option would be added to automatically recreate the updated cache in such a case. Notice that the content signature contains various q flags which affect parsing, so make sure to use the same parameters to q when performing the queries, otherwise q will issue an error. + +Notice that when running with `-A`, the cache is not written, even when `-C` is set to `readwrite`. This is due to the fact that `-A` does not really read the entire content of the files. For now, if you'd like to just prepare the cache without running the actual query, you can run it with a `select 1` query or something, although in terms of speed it will mostly not matter. If there's demand for adding an explicit `prepare caches only` option, I'll consider adding it. + +## Revamped `.qrc` mechanism +Adding `-C ` for each query can be cumbersome at some point, so the `.qrc` file has been revamped for easy addition of default parameters. + +For example, if you want the caching behaviour to be `read` all the time, then just add a `~/.qrc` file, and set the following in it: +``` +[options] +caching_mode=read +``` + +All other flags and parameters to q can be controlled by the `.qrc` file. To see the proper names for each parameter, run `q --dump-defaults` and it will dump a default `.qrc` file that contains all parameters to `stdout`. + +## Direct querying of standard sqlite databases +q now supports direct querying of standard sqlite databases. The syntax for accessing a table inside an sqlite database is `:::`. A query can contain any mix of sqlite files, qsql files or regular delimited files. + +For example, this command joins two tables from two separate sqlite databases: +``` +$ q "select count(*) from mydatabase1.sqlite:::mytable1 a left join mydatabase2.sqlite:::mytable2 b on (a.c1 = b.c1)" +``` + +Running queries on sqlite databases does not usually entail loading the data into memory. Databases are attached to a virtual database and queried directly from disk. This means that querying speed is practically identical to standard sqlite access. This is also true when multiple sqlite databases are used in a single query. The same mechanism is being used by q whenever it uses a qsql file (either directly or as a cache of a delimited fild). + +sqlite itself does have a pre-compiled limit of the number of databases that can be attached simultanously. If this limit is reached, then q will attach as many databases as possible, and then continue processing by loading additional tables into memory in order to execute the query. The standard limit in sqlite3 (unless compiled specifically with another limit) is 10 databases. This allows q to access as many as 8 user databases without having to load any data into memory (2 databases are always used for q's internal logic). Using more databases in a single query than this pre-compiled sqlite limit would slow things down, since some of the data would go into memory, but the query should still provide correct results. + +Whenever the sqlite database file contains only one table, the table name part can be ommitted, and the user can specify only the sqlite-filename as the table name. For example, querying an sqlite database `mydatabase.sqlite` that only has one table `mytable` is possible with `q "SELECT ... FROM mydatabase.sqlite"`. There's no need to specify the table name in this case. + +Since `.qsql` files are also standard sqlite files, they can be queried directly as well. This allows the user to actually delete the original CSV file and use the caches as if they were the original files. For example: + +``` +$ q "select count(*) from myfile.csv.qsql" +``` + +Notice that there's no need to write the `:::` as part of the table name, since `qsql` files that are created as caches contain only one table (e.g. the table matching the original file). + +Running a query that uses an sqlite/qsql database without specifying a table name will fail if there is more than one table in the database, showing the list of existing tables. This can be used in order to detect which tables exist in the database without resorting to other tools. For example: +``` +$ q "select * from chinook.db:::blah" +Table blah could not be found in sqlite file chinook.db . Existing table names: albums,sqlite_sequence,artists,customers,employees,genres,invoices,invoice_items,media_types,playlists,playlist_track,tracks,sqlite_stat1 +``` + +## Storing source data into a disk database +The `-S` option (`--save-db-to-disk`) has been modified to match the new capabilities. It works with all types of input tables/files, and writes the output database as a standard sqlite database. I've considered making the output a multi-table `qsql` file (e.g. with the additional metadata that q uses), but some things still need to be ironed out in order to make these qsql files work seamlessly with all other aspects of q. This will probably happen in the next version. + +This database can be accessed directly by q later on, by providing `:::` as the table name in the query. The table names that are chosen match the original file names, but go through the following process: +* The names are normalised in order to by compatible with sqlite restrictions (e.g. `x.csv` is normalised to `x_dot_csv`) +* duplicate table names are de-deduped by adding `_` to their names (e.g. two different csv files in separate folders which both have the name `companies` will be written to the file as `companies` and `companies_2`) + +This table-name normalisation happens also inside `.qsql` cache files, but in most cases there won't be any need to know these table names, since q automatically detects table names for databases which have a single-table. + +## File-concatenation and wildcard-matching features - Breaking change +File concatenation using '+' has been removed in this version, which is a breaking change. + +This was a controversial feature anyway, and can be done using standard SQL relatively easily. It also complicated the caching implementation significantly, and it seemed that it was not worth it. If there's demand for bringing this feature back, please write to me and I'll consider re-adding it. + +If you have a case of using file concatenation, you can use the following SQL instead: +``` +# Instead of writing +$ q "select * from myfile1+myfile2" +# Use the following: +$ q "select * from (select * from myfile1 UNION ALL select * from myfile2)" +``` + +This will provide the same results, but the error checking is a bit less robust, so be mindful on whether you're performing the right query on the right files. + +Conceptually, this is similar to wildcard matching (e.g. `select * from myfolder/myfile*`), but I have decided to leave wildcard-matching intact, since it seems to be a more common use-case. Cache creation and use is limited for now when using wildcards. Use the same method as described above for file concatenation if you wanna make sure that caches are being used. + +After this version is fully stabilised, I'll make more efforts to consolidate wildcard (and perhaps concatenation) to fully utilise caching seamlessly. + +## Code runs only on python 3 +Removed the dual py2/py3 support. Since q is packaged as a self-contained executable, along with python 3.8 itself, then this is not needed anymore. + +Users which for some reason still use q's main source code file directly and use python 2 would need to stay with the latest 2.0.19 release. In some next version, q's code structure is going to change significantly anyway in order to become a standard python module, so using the main source code file directly would not be possible. + +If you are such a user, and this decision hurts you considerably, please ping me. diff --git a/README.markdown b/README.markdown index 8106c9f2..4d5edd5a 100644 --- a/README.markdown +++ b/README.markdown @@ -1,33 +1,61 @@ -[![Build Status](https://travis-ci.org/harelba/q.svg?branch=master)](https://travis-ci.org/harelba/q) +[![Build and Package](https://github.com/harelba/q/workflows/BuildAndPackage/badge.svg?branch=master)](https://github.com/harelba/q/actions?query=branch%3Amaster) # q - Text as Data -q is a command line tool that allows direct execution of SQL-like queries on CSVs/TSVs (and any other tabular text files). +q's purpose is to bring SQL expressive power to the Linux command line and to provide easy access to text as actual data. -q treats ordinary files as database tables, and supports all SQL constructs, such as WHERE, GROUP BY, JOINs etc. It supports automatic column name and column type detection, and provides full support for multiple encodings. +q allows the following: -q's web site is [http://harelba.github.io/q/](http://harelba.github.io/q/). It contains everything you need to download and use q in no time. +* Performing SQL-like statements directly on tabular text data, auto-caching the data in order to accelerate additional querying on the same file. +* Performing SQL statements directly on multi-file sqlite3 databases, without having to merge them or load them into memory -## Installation. -Extremely simple. +The following table shows the impact of using caching: + +| Rows | Columns | File Size | Query time without caching | Query time with caching | Speed Improvement | +|:---------:|:-------:|:---------:|:--------------------------:|:-----------------------:|:-----------------:| +| 5,000,000 | 100 | 4.8GB | 4 minutes, 47 seconds | 1.92 seconds | x149 | +| 1,000,000 | 100 | 983MB | 50.9 seconds | 0.461 seconds | x110 | +| 1,000,000 | 50 | 477MB | 27.1 seconds | 0.272 seconds | x99 | +| 100,000 | 100 | 99MB | 5.2 seconds | 0.141 seconds | x36 | +| 100,000 | 50 | 48MB | 2.7 seconds | 0.105 seconds | x25 | + +Notice that for the current version, caching is **not enabled** by default, since the caches take disk space. Use `-C readwrite` or `-C read` to enable it for a query, or add `caching_mode` to `.qrc` to set a new default. + +q's web site is [https://harelba.github.io/q/](https://harelba.github.io/q/) or [https://q.textasdata.wiki](https://q.textasdata.wiki) It contains everything you need to download and use q immediately. + + +## Usage Examples +q treats ordinary files as database tables, and supports all SQL constructs, such as `WHERE`, `GROUP BY`, `JOIN`s, etc. It supports automatic column name and type detection, and provides full support for multiple character encodings. + +Here are some example commands to get the idea: -Instructions for all OSs are [here](http://harelba.github.io/q/install.html). +```bash +$ q "SELECT COUNT(*) FROM ./clicks_file.csv WHERE c3 > 32.3" -## Examples +$ ps -ef | q -H "SELECT UID, COUNT(*) cnt FROM - GROUP BY UID ORDER BY cnt DESC LIMIT 3" -![blah](doc/basic-examples.png) +$ q "select count(*) from some_db.sqlite3:::albums a left join another_db.sqlite3:::tracks t on (a.album_id = t.album_id)" +``` -Go [here](http://harelba.github.io/q/examples.html) for more examples. +Detailed examples are in [here](https://harelba.github.io/q/#examples) -## Python API -A development branch for exposing q's capabilities as a Python module can be viewed here, along with examples of the alpha version of the API.
Existing functionality as a command-line tool will not be affected by this. Your input will be most appreciated. +## Installation. +**New Major Version `3.1.6` is out with a lot of significant additions.** + +Instructions for all OSs are [here](https://harelba.github.io/q/#installation). -## Change log -Click [here](http://harelba.github.io/q/changelog.html) to see the change log. +The previous version `2.0.19` Can still be downloaded from [here](https://github.com/harelba/q/releases/tag/2.0.19) ## Contact Any feedback/suggestions/complaints regarding this tool would be much appreciated. Contributions are most welcome as well, of course. -Harel Ben-Attia, harelba@gmail.com, [@harelba](https://twitter.com/harelba) on Twitter +Linkedin: [Harel Ben Attia](https://www.linkedin.com/in/harelba/) + +Twitter [@harelba](https://twitter.com/harelba) + +Email [harelba@gmail.com](mailto:harelba@gmail.com) + +q on twitter: [#qtextasdata](https://twitter.com/hashtag/qtextasdata?src=hashtag_click) + +Patreon: [harelba](https://www.patreon.com/harelba) - All the money received is donated to the [Center for the Prevention and Treatment of Domestic Violence](https://www.gov.il/he/departments/bureaus/molsa-almab-ramla) in my hometown - Ramla, Israel. -q on twitter: #qtextasdata diff --git a/benchmark-config.sh b/benchmark-config.sh new file mode 100644 index 00000000..b293d9a2 --- /dev/null +++ b/benchmark-config.sh @@ -0,0 +1,3 @@ +#!/bin/bash + +BENCHMARK_PYTHON_VERSIONS=(3.8.5) diff --git a/bin/q b/bin/q deleted file mode 100755 index 0e3c2dfe..00000000 --- a/bin/q +++ /dev/null @@ -1,1873 +0,0 @@ -#!/usr/bin/env python2 - -# Copyright (C) 2012-2018 Harel Ben-Attia -# -# This program is free software; you can redistribute it and/or modify -# it under the terms of the GNU General Public License as published by -# the Free Software Foundation; either version 3, or (at your option) -# any later version. -# -# This program is distributed in the hope that it will be useful, -# but WITHOUT ANY WARRANTY; without even the implied warranty of -# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -# GNU General Public License for more details (doc/LICENSE contains -# a copy of it) -# -# -# Name : q (With respect to The Q Continuum) -# Author : Harel Ben Attia - harelba@gmail.com, harelba @ github, @harelba on twitter -# Requires : python with sqlite3 (standard in python>=2.6) -# -# -# q allows performing SQL-like statements on tabular text data. -# -# Its purpose is to bring SQL expressive power to manipulating text data using the Linux command line. -# -# Full Documentation and details in http://harelba.github.io/q/ -# -# Run with --help for command line details -# -q_version = "1.8" - -__all__ = [ 'QTextAsData' ] - -import os -import sys -import sqlite3 -import gzip -import glob -from optparse import OptionParser,OptionGroup -import traceback as tb -import codecs -import locale -import time -import re -from ConfigParser import ConfigParser -import traceback -import csv -import hashlib -import uuid -import cStringIO -import math - -DEBUG = False - -def get_stdout_encoding(encoding_override=None): - if encoding_override is not None and encoding_override != 'none': - return encoding_override - - if sys.stdout.isatty(): - return sys.stdout.encoding - else: - return locale.getpreferredencoding() - -SHOW_SQL = False - -def sha1(data): - if not isinstance(data,str) and not isinstance(data,unicode): - return hashlib.sha1(str(data)).hexdigest() - return hashlib.sha1(data).hexdigest() - -def regexp(regular_expression, data): - if data is not None: - if not isinstance(data, str) and not isinstance(data, unicode): - data = str(data) - return re.search(regular_expression, data) is not None - else: - return False - -class Sqlite3DBResults(object): - def __init__(self,query_column_names,results): - self.query_column_names = query_column_names - self.results = results - -def percentile(l, p): - # TODO Alpha implementation, need to provide multiple interpolation methods, and add tests - if not l: - return None - k = p*(len(l) - 1) - f = math.floor(k) - c = math.ceil(k) - if c == f: - return l[int(k)] - return (c-k) * l[int(f)] + (k-f) * l[int(c)] - -class StrictPercentile(object): - def __init__(self): - self.values = [] - self.p = None - def step(self,value,p): - if self.p is None: - self.p = p - self.values.append(value) - - def finalize(self): - if len(self.values) == 0 or (self.p < 0 or self.p > 1): - return None - else: - return percentile(sorted(self.values),self.p) - -class Sqlite3DB(object): - - def __init__(self, show_sql=SHOW_SQL): - self.show_sql = show_sql - self.conn = sqlite3.connect(':memory:') - self.last_temp_table_id = 10000 - self.cursor = self.conn.cursor() - self.type_names = { - str: 'TEXT', int: 'INT', long : 'INT' , float: 'FLOAT', None: 'TEXT'} - self.numeric_column_types = set([int, long, float]) - self.add_user_functions() - - def done(self): - self.conn.commit() - - def store_db_to_disk_standard(self,sqlite_db_filename,table_names_mapping): - new_db = sqlite3.connect(sqlite_db_filename,isolation_level=None) - c = new_db.cursor() - for s in self.conn.iterdump(): - c.execute(s) - results = c.fetchall() - #print "executed %s results %s " % (s,results) - for source_filename_str,tn in table_names_mapping.iteritems(): - c.execute('alter table `%s` rename to `%s`' % (tn, source_filename_str)) - new_db.close() - - def store_db_to_disk_fast(self,sqlite_db_filename,table_names_mapping): - try: - import sqlitebck - except ImportError, e: - msg = "sqlitebck python module cannot be found - fast store to disk cannot be performed. Note that for now, sqlitebck is not packaged as part of q. In order to use the fast method, you need to manually `pip install sqlitebck` into your python environment. We obviously consider this as a bug and it will be fixed once proper packaging will be done, making the fast method the standard one." - raise MissingSqliteBckModuleException(msg) - - new_db = sqlite3.connect(sqlite_db_filename) - sqlitebck.copy(self.conn,new_db) - c = new_db.cursor() - for source_filename_str,tn in table_names_mapping.iteritems(): - c.execute('alter table `%s` rename to `%s`' % (tn, source_filename_str)) - new_db.close() - - def store_db_to_disk(self,sqlite_db_filename,table_names_mapping,method='standard'): - if method == 'standard': - self.store_db_to_disk_standard(sqlite_db_filename,table_names_mapping) - elif method == 'fast': - self.store_db_to_disk_fast(sqlite_db_filename,table_names_mapping) - else: - raise ValueError('Unknown store-db-to-disk method %s' % method) - - def add_user_functions(self): - self.conn.create_function("regexp", 2, regexp) - self.conn.create_function("sha1", 1, sha1) - self.conn.create_aggregate("percentile",2,StrictPercentile) - - def is_numeric_type(self, column_type): - return column_type in self.numeric_column_types - - def update_many(self, sql, params): - try: - if self.show_sql: - print sql, " params: " + str(params) - self.cursor.executemany(sql, params) - finally: - pass # cursor.close() - - def execute_and_fetch(self, q): - try: - if self.show_sql: - print repr(q) - self.cursor.execute(q) - if self.cursor.description is not None: - # we decode the column names, so they can be encoded to any output format later on - query_column_names = [c[0].decode('utf-8') for c in self.cursor.description] - else: - query_column_names = None - result = self.cursor.fetchall() - finally: - pass # cursor.close() - return Sqlite3DBResults(query_column_names,result) - - def _get_as_list_str(self, l): - return ",".join(['"%s"' % x.replace('"', '""') for x in l]) - - def _get_col_values_as_list_str(self, col_vals, col_types): - result = [] - for col_val, col_type in zip(col_vals, col_types): - if col_val == '' and col_type is not str: - col_val = "null" - else: - if col_val is not None: - if "'" in col_val: - col_val = col_val.replace("'", "''") - col_val = "'" + col_val + "'" - else: - col_val = "null" - - result.append(col_val) - return ",".join(result) - - def generate_insert_row(self, table_name, column_names): - col_names_str = self._get_as_list_str(column_names) - question_marks = ", ".join(["?" for i in range(0, len(column_names))]) - return 'INSERT INTO %s (%s) VALUES (%s)' % (table_name, col_names_str, question_marks) - - def generate_begin_transaction(self): - return "BEGIN TRANSACTION" - - def generate_end_transaction(self): - return "COMMIT" - - # Get a list of column names so order will be preserved (Could have used OrderedDict, but - # then we would need python 2.7) - def generate_create_table(self, table_name, column_names, column_dict): - # Convert dict from python types to db types - column_name_to_db_type = dict( - (n, self.type_names[t]) for n, t in column_dict.iteritems()) - column_defs = ','.join(['"%s" %s' % ( - n.replace('"', '""'), column_name_to_db_type[n]) for n in column_names]) - return 'CREATE TABLE %s (%s)' % (table_name, column_defs) - - def generate_temp_table_name(self): - self.last_temp_table_id += 1 - tn = "temp_table_%s" % self.last_temp_table_id - return tn - - def generate_drop_table(self, table_name): - return "DROP TABLE %s" % table_name - - def drop_table(self, table_name): - return self.execute_and_fetch(self.generate_drop_table(table_name)) - -class CouldNotConvertStringToNumericValueException(Exception): - - def __init__(self, msg): - self.msg = msg - - def __str(self): - return repr(self.msg) - -class ColumnMaxLengthLimitExceededException(Exception): - - def __init__(self, msg): - self.msg = msg - - def __str(self): - return repr(self.msg) - -class MissingSqliteBckModuleException(Exception): - - def __init__(self, msg): - self.msg = msg - - def __str(self): - return repr(self.msg) - - -class CouldNotParseInputException(Exception): - - def __init__(self, msg): - self.msg = msg - - def __str(self): - return repr(self.msg) - -class BadHeaderException(Exception): - - def __init__(self, msg): - self.msg = msg - - def __str(self): - return repr(self.msg) - -class EncodedQueryException(Exception): - - def __init__(self, msg): - self.msg = msg - - def __str(self): - return repr(self.msg) - - -class CannotUnzipStdInException(Exception): - - def __init__(self): - pass - -class UniversalNewlinesExistException(Exception): - - def __init__(self): - pass - -class UnprovidedStdInException(Exception): - - def __init__(self): - pass - -class EmptyDataException(Exception): - - def __init__(self): - pass - -class MissingHeaderException(Exception): - - def __init__(self,msg): - self.msg = msg - - -class FileNotFoundException(Exception): - - def __init__(self, msg): - self.msg = msg - - def __str(self): - return repr(self.msg) - - -class ColumnCountMismatchException(Exception): - - def __init__(self, msg): - self.msg = msg - - def __str(self): - return repr(self.msg) - -class StrictModeColumnCountMismatchException(Exception): - - def __init__(self,expected_col_count,actual_col_count): - self.expected_col_count = expected_col_count - self.actual_col_count = actual_col_count - -class FluffyModeColumnCountMismatchException(Exception): - - def __init__(self,expected_col_count,actual_col_count): - self.expected_col_count = expected_col_count - self.actual_col_count = actual_col_count - - -# Simplistic Sql "parsing" class... We'll eventually require a real SQL parser which will provide us with a parse tree -# -# A "qtable" is a filename which behaves like an SQL table... - - -class Sql(object): - - def __init__(self, sql): - # Currently supports only standard SELECT statements - - # Holds original SQL - self.sql = sql - # Holds sql parts - self.sql_parts = sql.split() - - # Set of qtable names - self.qtable_names = set() - # Dict from qtable names to their positions in sql_parts. Value here is a *list* of positions, - # since it is possible that the same qtable_name (file) is referenced in multiple positions - # and we don't want the database table to be recreated for each - # reference - self.qtable_name_positions = {} - # Dict from qtable names to their effective (actual database) table - # names - self.qtable_name_effective_table_names = {} - - self.query_column_names = None - - # Go over all sql parts - idx = 0 - while idx < len(self.sql_parts): - # Get the part string - part = self.sql_parts[idx] - # If it's a FROM or a JOIN - if part.upper() in ['FROM', 'JOIN']: - # and there is nothing after it, - if idx == len(self.sql_parts) - 1: - # Just fail - raise Exception( - 'FROM/JOIN is missing a table name after it') - - qtable_name = self.sql_parts[idx + 1] - # Otherwise, the next part contains the qtable name. In most cases the next part will be only the qtable name. - # We handle one special case here, where this is a subquery as a column: "SELECT (SELECT ... FROM qtable),100 FROM ...". - # In that case, there will be an ending paranthesis as part of the name, and we want to handle this case gracefully. - # This is obviously a hack of a hack :) Just until we have - # complete parsing capabilities - if ')' in qtable_name: - leftover = qtable_name[qtable_name.index(')'):] - self.sql_parts.insert(idx + 2, leftover) - qtable_name = qtable_name[:qtable_name.index(')')] - self.sql_parts[idx + 1] = qtable_name - - self.qtable_names.add(qtable_name) - - if qtable_name not in self.qtable_name_positions.keys(): - self.qtable_name_positions[qtable_name] = [] - - self.qtable_name_positions[qtable_name].append(idx + 1) - idx += 2 - else: - idx += 1 - - def set_effective_table_name(self, qtable_name, effective_table_name): - if qtable_name not in self.qtable_names: - raise Exception("Unknown qtable %s" % qtable_name) - if qtable_name in self.qtable_name_effective_table_names.keys(): - raise Exception( - "Already set effective table name for qtable %s" % qtable_name) - - self.qtable_name_effective_table_names[ - qtable_name] = effective_table_name - - def get_effective_sql(self,original_names=False): - if len(filter(lambda x: x is None, self.qtable_name_effective_table_names)) != 0: - raise Exception('There are qtables without effective tables') - - effective_sql = [x for x in self.sql_parts] - - for qtable_name, positions in self.qtable_name_positions.iteritems(): - for pos in positions: - if not original_names: - effective_sql[pos] = self.qtable_name_effective_table_names[ - qtable_name] - else: - effective_sql[pos] = "`%s`" % qtable_name - - return " ".join(effective_sql) - - def get_qtable_name_effective_table_names(self): - return self.qtable_name_effective_table_names - - def execute_and_fetch(self, db): - db_results_obj = db.execute_and_fetch(self.get_effective_sql()) - return db_results_obj - - -class LineSplitter(object): - - def __init__(self, delimiter, expected_column_count): - self.delimiter = delimiter - self.expected_column_count = expected_column_count - if delimiter is not None: - escaped_delimiter = re.escape(delimiter) - self.split_regexp = re.compile('(?:%s)+' % escaped_delimiter) - else: - self.split_regexp = re.compile(r'\s+') - - def split(self, line): - if line and line[-1] == '\n': - line = line[:-1] - return self.split_regexp.split(line, max_split=self.expected_column_count) - - -class TableColumnInferer(object): - - def __init__(self, mode, expected_column_count, input_delimiter, skip_header=False,disable_column_type_detection=False): - self.inferred = False - self.mode = mode - self.rows = [] - self.skip_header = skip_header - self.header_row = None - self.expected_column_count = expected_column_count - self.input_delimiter = input_delimiter - self.disable_column_type_detection = disable_column_type_detection - - def analyze(self, col_vals): - if self.inferred: - raise Exception("Already inferred columns") - - if self.skip_header and self.header_row is None: - self.header_row = col_vals - else: - self.rows.append(col_vals) - - if len(self.rows) < 100: - return False - - self.do_analysis() - return True - - def force_analysis(self): - # This method is called whenever there is no more data, and an analysis needs - # to be performed immediately, regardless of the amount of sample data that has - # been collected - self.do_analysis() - - def determine_type_of_value(self, value): - if self.disable_column_type_detection: - return str - - if value is not None: - value = value.strip() - if value == '' or value is None: - return None - - try: - i = int(value) - if type(i) == long: - return long - else: - return int - except: - pass - - try: - f = float(value) - return float - except: - pass - - return str - - def determine_type_of_value_list(self, value_list): - type_list = [self.determine_type_of_value(v) for v in value_list] - all_types = set(type_list) - if len(set(type_list)) == 1: - # all the sample lines are of the same type - return type_list[0] - else: - # check for the number of types without nulls, - type_list_without_nulls = filter( - lambda x: x is not None, type_list) - # If all the sample lines are of the same type, - if len(set(type_list_without_nulls)) == 1: - # return it - return type_list_without_nulls[0] - else: - return str - - def do_analysis(self): - if self.mode == 'strict': - self._do_strict_analysis() - elif self.mode in ['relaxed', 'fluffy']: - self._do_relaxed_analysis() - else: - raise Exception('Unknown parsing mode %s' % self.mode) - - if self.column_count == 1 and self.expected_column_count != 1: - print >>sys.stderr, "Warning: column count is one - did you provide the correct delimiter?" - - self.infer_column_types() - - self.infer_column_names() - - def validate_column_names(self, value_list): - column_name_errors = [] - for v in value_list: - if v is None: - # we allow column names to be None, in relaxed mode it'll be filled with default names. - # RLRL - continue - if ',' in v: - column_name_errors.append( - (v, "Column name cannot contain commas")) - continue - if self.input_delimiter in v: - column_name_errors.append( - (v, "Column name cannot contain the input delimiter. Please make sure you've set the correct delimiter")) - continue - if '\n' in v: - column_name_errors.append( - (v, "Column name cannot contain newline")) - continue - if v != v.strip(): - column_name_errors.append( - (v, "Column name contains leading/trailing spaces")) - continue - try: - v.encode("utf-8", "strict").decode("utf-8") - except: - column_name_errors.append( - (v, "Column name must be UTF-8 Compatible")) - continue - # We're checking for column duplication for each field in order to be able to still provide it along with other errors - if len(filter(lambda x: x == v,value_list)) > 1: - entry = (v, "Column name is duplicated") - # Don't duplicate the error report itself - if entry not in column_name_errors: - column_name_errors.append(entry) - continue - nul_index = v.find("\x00") - if nul_index >= 0: - column_name_errors.append( - (v, "Column name cannot contain NUL")) - continue - t = self.determine_type_of_value(v) - if t != str: - column_name_errors.append((v, "Column name must be a string")) - return column_name_errors - - def infer_column_names(self): - if self.header_row is not None: - column_name_errors = self.validate_column_names(self.header_row) - if len(column_name_errors) > 0: - raise BadHeaderException("Header must contain only strings and not numbers or empty strings: '%s'\n%s" % ( - ",".join(self.header_row), "\n".join(["'%s': %s" % (x, y) for x, y in column_name_errors]))) - - # use header row in order to name columns - if len(self.header_row) < self.column_count: - if self.mode == 'strict': - raise ColumnCountMismatchException("Strict mode. Header row contains less columns than expected column count(%s vs %s)" % ( - len(self.header_row), self.column_count)) - elif self.mode in ['relaxed', 'fluffy']: - # in relaxed mode, add columns to fill the missing ones - self.header_row = self.header_row + \ - ['c%s' % (x + len(self.header_row) + 1) - for x in xrange(self.column_count - len(self.header_row))] - elif len(self.header_row) > self.column_count: - if self.mode == 'strict': - raise ColumnCountMismatchException("Strict mode. Header row contains more columns than expected column count (%s vs %s)" % ( - len(self.header_row), self.column_count)) - elif self.mode in ['relaxed', 'fluffy']: - # In relaxed mode, just cut the extra column names - self.header_row = self.header_row[:self.column_count] - self.column_names = self.header_row - else: - # Column names are cX starting from 1 - self.column_names = ['c%s' % (i + 1) - for i in range(self.column_count)] - - def _do_relaxed_analysis(self): - column_count_list = [len(col_vals) for col_vals in self.rows] - - if len(self.rows) == 0: - self.column_count = 0 - else: - if self.expected_column_count is not None: - self.column_count = self.expected_column_count - else: - # If not specified, we'll take the largest row in the sample rows - self.column_count = max(column_count_list) - - def get_column_count_summary(self, column_count_list): - counts = {} - for column_count in column_count_list: - counts[column_count] = counts.get(column_count, 0) + 1 - return ", ".join(["%s rows with %s columns" % (v, k) for k, v in counts.iteritems()]) - - def _do_strict_analysis(self): - column_count_list = [len(col_vals) for col_vals in self.rows] - - if len(set(column_count_list)) != 1: - raise ColumnCountMismatchException('Strict mode. Column Count is expected to identical. Multiple column counts exist at the first part of the file. Try to check your delimiter, or change to relaxed mode. Details: %s' % ( - self.get_column_count_summary(column_count_list))) - - self.column_count = len(self.rows[0]) - - if self.expected_column_count is not None and self.column_count != self.expected_column_count: - raise ColumnCountMismatchException('Strict mode. Column count is expected to be %s but is %s' % ( - self.expected_column_count, self.column_count)) - - self.infer_column_types() - - def infer_column_types(self): - self.column_types = [] - self.column_types2 = [] - for column_number in xrange(self.column_count): - column_value_list = [ - row[column_number] if column_number < len(row) else None for row in self.rows] - column_type = self.determine_type_of_value_list(column_value_list) - self.column_types.append(column_type) - - column_value_list2 = [row[column_number] if column_number < len( - row) else None for row in self.rows[1:]] - column_type2 = self.determine_type_of_value_list( - column_value_list2) - self.column_types2.append(column_type2) - - comparison = map( - lambda x: x[0] == x[1], zip(self.column_types, self.column_types2)) - if False in comparison and not self.skip_header: - number_of_column_types = len(set(self.column_types)) - if number_of_column_types == 1 and list(set(self.column_types))[0] == str: - print >>sys.stderr, 'Warning - There seems to be header line in the file, but -H has not been specified. All fields will be detected as text fields, and the header line will appear as part of the data' - - def get_column_dict(self): - return dict(zip(self.column_names, self.column_types)) - - def get_column_count(self): - return self.column_count - - def get_column_names(self): - return self.column_names - - def get_column_types(self): - return self.column_types - - -def encoded_csv_reader(encoding, f, dialect, **kwargs): - try: - csv_reader = csv.reader(f, dialect, **kwargs) - if encoding is not None and encoding != 'none': - for row in csv_reader: - yield [unicode(x, encoding) for x in row] - else: - for row in csv_reader: - yield row - except ValueError,e: - if e.message is not None and e.message.startswith('could not convert string to'): - raise CouldNotConvertStringToNumericValueException(e.message) - else: - raise CouldNotParseInputException(str(e)) - except Exception,e: - if str(e).startswith("field larger than field limit"): - raise ColumnMaxLengthLimitExceededException(str(e)) - elif 'universal-newline' in str(e): - raise UniversalNewlinesExistException() - else: - raise - -def normalized_filename(filename): - if filename == '-': - return 'stdin' - else: - return filename - -class TableCreatorState(object): - NEW = 'NEW' - INITIALIZED = 'INITIALIZED' - ANALYZED = 'ANALYZED' - FULLY_READ = 'FULLY_READ' - -class MaterializedFileState(object): - def __init__(self,filename,f,encoding,dialect,is_stdin): - self.filename = filename - self.lines_read = 0 - self.f = f - self.encoding = encoding - self.dialect = dialect - self.is_stdin = is_stdin - self.skipped_bom = False - - def read_file_using_csv(self): - # This is a hack for utf-8 with BOM encoding in order to skip the BOM. python's csv module - # has a bug which prevents fixing it using the proper encoding, and it has been encountered by - # multiple people. - if self.encoding == 'utf-8-sig' and self.lines_read == 0 and not self.skipped_bom: - try: - BOM = self.f.read(3) - if BOM != '\xef\xbb\xbf': - raise Exception('Value of BOM is not as expected - Value is "%s"' % str(BOM)) - except Exception,e: - raise Exception('Tried to skip BOM for "utf-8-sig" encoding and failed. Error message is ' + str(e)) - csv_reader = encoded_csv_reader(self.encoding, self.f, dialect=self.dialect) - try: - for col_vals in csv_reader: - self.lines_read += 1 - yield col_vals - except ColumnMaxLengthLimitExceededException,e: - msg = "Column length is larger than the maximum. Offending file is '%s' - Line is %s, counting from 1 (encoding %s). The line number is the raw line number of the file, ignoring whether there's a header or not" % (self.filename,self.lines_read + 1,self.encoding) - raise ColumnMaxLengthLimitExceededException(msg) - except UniversalNewlinesExistException,e2: - # No need to translate the exception, but we want it to be explicitly defined here for clarity - raise UniversalNewlinesExistException() - - def close(self): - if self.f != sys.stdin: - self.f.close() - -class TableCreator(object): - - def __init__(self, db, filenames_str, line_splitter, skip_header=False, gzipped=False, with_universal_newlines=False, encoding='UTF-8', mode='fluffy', expected_column_count=None, input_delimiter=None,disable_column_type_detection=False, - stdin_file=None,stdin_filename='-'): - self.db = db - self.filenames_str = filenames_str - self.skip_header = skip_header - self.gzipped = gzipped - self.table_created = False - self.line_splitter = line_splitter - self.encoding = encoding - self.mode = mode - self.expected_column_count = expected_column_count - self.input_delimiter = input_delimiter - self.stdin_file = stdin_file - self.stdin_filename = stdin_filename - self.with_universal_newlines = with_universal_newlines - - self.column_inferer = TableColumnInferer( - mode, expected_column_count, input_delimiter, skip_header,disable_column_type_detection) - - # Filled only after table population since we're inferring the table - # creation data - self.table_name = None - - self.pre_creation_rows = [] - self.buffered_inserts = [] - self.effective_column_names = None - - # Column type indices for columns that contain numeric types. Lazily initialized - # so column inferer can do its work before this information is needed - self.numeric_column_indices = None - - self.materialized_file_list = self.materialize_file_list() - self.materialized_file_dict = {} - - self.state = TableCreatorState.NEW - - def materialize_file_list(self): - materialized_file_list = [] - - # Get the list of filenames - filenames = self.filenames_str.split("+") - - # for each filename (or pattern) - for fileglob in filenames: - # Allow either stdin or a glob match - if fileglob == self.stdin_filename: - materialized_file_list.append(self.stdin_filename) - else: - materialized_file_list += glob.glob(fileglob) - - # If there are no files to go over, - if len(materialized_file_list) == 0: - raise FileNotFoundException( - "No files matching '%s' have been found" % self.filenames_str) - - return materialized_file_list - - def get_table_name(self): - return self.table_name - - def open_file(self,filename): - # TODO Support universal newlines for gzipped and stdin data as well - - # Check if it's standard input or a file - if filename == self.stdin_filename: - if self.stdin_file is None: - raise UnprovidedStdInException() - f = self.stdin_file - if self.gzipped: - raise CannotUnzipStdInException() - else: - if self.gzipped or filename.endswith('.gz'): - f = gzip.GzipFile(fileobj=file(filename,'rb')) - else: - if self.with_universal_newlines: - file_opening_mode = 'rbU' - else: - file_opening_mode = 'rb' - f = file(filename,file_opening_mode) - return f - - def _pre_populate(self,dialect): - # For each match - for filename in self.materialized_file_list: - if filename in self.materialized_file_dict.keys(): - continue - - f = self.open_file(filename) - - is_stdin = filename == self.stdin_filename - - mfs = MaterializedFileState(filename,f,self.encoding,dialect,is_stdin) - self.materialized_file_dict[filename] = mfs - - def _populate(self,dialect,stop_after_analysis=False): - total_data_lines_read = 0 - - # For each match - for filename in self.materialized_file_list: - mfs = self.materialized_file_dict[filename] - - try: - try: - for col_vals in mfs.read_file_using_csv(): - self._insert_row(col_vals) - if stop_after_analysis and self.column_inferer.inferred: - return - if mfs.lines_read == 0 and self.skip_header: - raise MissingHeaderException("Header line is expected but missing in file %s" % filename) - - total_data_lines_read += mfs.lines_read - (1 if self.skip_header else 0) - except StrictModeColumnCountMismatchException,e: - raise ColumnCountMismatchException( - 'Strict mode - Expected %s columns instead of %s columns in file %s row %s. Either use relaxed/fluffy modes or check your delimiter' % ( - e.expected_col_count, e.actual_col_count, normalized_filename(mfs.filename), mfs.lines_read)) - except FluffyModeColumnCountMismatchException,e: - raise ColumnCountMismatchException( - 'Deprecated fluffy mode - Too many columns in file %s row %s (%s fields instead of %s fields). Consider moving to either relaxed or strict mode' % ( - normalized_filename(mfs.filename), mfs.lines_read, e.actual_col_count, e.expected_col_count)) - finally: - if not stop_after_analysis: - mfs.close() - self._flush_inserts() - - if not self.table_created: - self.column_inferer.force_analysis() - self._do_create_table() - - - if total_data_lines_read == 0: - raise EmptyDataException() - - def populate(self,dialect,stop_after_analysis=False): - if self.state == TableCreatorState.NEW: - self._pre_populate(dialect) - self.state = TableCreatorState.INITIALIZED - - if self.state == TableCreatorState.INITIALIZED: - self._populate(dialect,stop_after_analysis=True) - self.state = TableCreatorState.ANALYZED - - if stop_after_analysis: - return - - if self.state == TableCreatorState.ANALYZED: - self._populate(dialect,stop_after_analysis=False) - self.state = TableCreatorState.FULLY_READ - return - - def _flush_pre_creation_rows(self): - for i, col_vals in enumerate(self.pre_creation_rows): - if self.skip_header and i == 0: - # skip header line - continue - self._insert_row(col_vals) - self._flush_inserts() - self.pre_creation_rows = [] - - def _insert_row(self, col_vals): - # If table has not been created yet - if not self.table_created: - # Try to create it along with another "example" line of data - self.try_to_create_table(col_vals) - - # If the table is still not created, then we don't have enough data, just - # store the data and return - if not self.table_created: - self.pre_creation_rows.append(col_vals) - return - - - # The table already exists, so we can just add a new row - self._insert_row_i(col_vals) - - def initialize_numeric_column_indices_if_needed(self): - # Lazy initialization of numeric column indices - if self.numeric_column_indices is None: - column_types = self.column_inferer.get_column_types() - self.numeric_column_indices = [idx for idx, column_type in enumerate( - column_types) if self.db.is_numeric_type(column_type)] - - def nullify_values_if_needed(self, col_vals): - new_vals = col_vals[:] - col_count = len(col_vals) - for i in self.numeric_column_indices: - if i >= col_count: - continue - v = col_vals[i] - if v == '': - new_vals[i] = None - return new_vals - - def normalize_col_vals(self, col_vals): - # Make sure that numeric column indices are initializd - self.initialize_numeric_column_indices_if_needed() - - col_vals = self.nullify_values_if_needed(col_vals) - - expected_col_count = self.column_inferer.get_column_count() - actual_col_count = len(col_vals) - if self.mode == 'strict': - if actual_col_count != expected_col_count: - raise StrictModeColumnCountMismatchException(expected_col_count,actual_col_count) - return col_vals - - # in all non strict mode, we add dummy data to missing columns - - if actual_col_count < expected_col_count: - col_vals = col_vals + \ - [None for x in xrange(expected_col_count - actual_col_count)] - - # in relaxed mode, we merge all extra columns to the last column value - if self.mode == 'relaxed': - if actual_col_count > expected_col_count: - xxx = col_vals[:expected_col_count - 1] + \ - [self.input_delimiter.join([v if v is not None else '' for v in - col_vals[expected_col_count - 1:]])] - return xxx - else: - return col_vals - - if self.mode == 'fluffy': - if actual_col_count > expected_col_count: - raise FluffyModeColumnCountMismatchException(expected_col_count,actual_col_count) - return col_vals - - raise Exception("Unidentified parsing mode %s" % self.mode) - - def _insert_row_i(self, col_vals): - col_vals = self.normalize_col_vals(col_vals) - - if self.effective_column_names is None: - self.effective_column_names = self.column_inferer.column_names[:len(col_vals)] - - if len(self.effective_column_names) > 0: - self.buffered_inserts.append(col_vals) - else: - self.buffered_inserts.append([""]) - - if len(self.buffered_inserts) < 5000: - return - self._flush_inserts() - - def _flush_inserts(self): - # print self.db.execute_and_fetch(self.db.generate_begin_transaction()) - - # If the table is still not created, then we don't have enough data - if not self.table_created: - return - - if len(self.buffered_inserts) > 0: - insert_row_stmt = self.db.generate_insert_row( - self.table_name, self.effective_column_names) - - self.db.update_many(insert_row_stmt, self.buffered_inserts) - # print self.db.execute_and_fetch(self.db.generate_end_transaction()) - self.buffered_inserts = [] - - def try_to_create_table(self, col_vals): - if self.table_created: - raise Exception('Table is already created') - - # Add that line to the column inferer - result = self.column_inferer.analyze(col_vals) - # If inferer succeeded, - if result: - self._do_create_table() - else: - pass # We don't have enough information for creating the table yet - - def _do_create_table(self): - # Then generate a temp table name - self.table_name = self.db.generate_temp_table_name() - # Get the column definition dict from the inferer - column_dict = self.column_inferer.get_column_dict() - - # Guard against empty tables (instead of preventing the creation, just create with a dummy column) - if len(column_dict) == 0: - column_dict = { 'dummy_column_for_empty_tables' : str } - ordered_column_names = [ 'dummy_column_for_empty_tables' ] - else: - ordered_column_names = self.column_inferer.get_column_names() - - # Create the CREATE TABLE statement - create_table_stmt = self.db.generate_create_table( - self.table_name, ordered_column_names, column_dict) - # And create the table itself - self.db.execute_and_fetch(create_table_stmt) - # Mark the table as created - self.table_created = True - self._flush_pre_creation_rows() - - def drop_table(self): - if self.table_created: - self.db.drop_table(self.table_name) - - -def determine_max_col_lengths(m,output_field_quoting_func,output_delimiter): - if len(m) == 0: - return [] - max_lengths = [0 for x in xrange(0, len(m[0]))] - for row_index in xrange(0, len(m)): - for col_index in xrange(0, len(m[0])): - new_len = len(unicode(output_field_quoting_func(output_delimiter,m[row_index][col_index]))) - if new_len > max_lengths[col_index]: - max_lengths[col_index] = new_len - return max_lengths - -def print_credentials(): - print >>sys.stderr,"q version %s" % q_version - print >>sys.stderr,"Copyright (C) 2012-2017 Harel Ben-Attia (harelba@gmail.com, @harelba on twitter)" - print >>sys.stderr,"http://harelba.github.io/q/" - print >>sys.stderr - -class QWarning(object): - def __init__(self,exception,msg): - self.exception = exception - self.msg = msg - -class QError(object): - def __init__(self,exception,msg,errorcode): - self.exception = exception - self.msg = msg - self.errorcode = errorcode - self.traceback = traceback.format_exc() - -class QDataLoad(object): - def __init__(self,filename,start_time,end_time): - self.filename = filename - self.start_time = start_time - self.end_time = end_time - - def duration(self): - return self.end_time - self.start_time - - def __str__(self): - return "DataLoad<'%s' at %s (took %4.3f seconds)>" % (self.filename,self.start_time,self.duration()) - __repr__ = __str__ - -class QMaterializedFile(object): - def __init__(self,filename,is_stdin): - self.filename = filename - self.is_stdin = is_stdin - - def __str__(self): - return "QMaterializedFile" % (self.filename,self.is_stdin) - __repr__ = __str__ - -class QTableStructure(object): - def __init__(self,filenames_str,materialized_files,column_names,column_types): - self.filenames_str = filenames_str - self.materialized_files = materialized_files - self.column_names = column_names - self.column_types = column_types - - def __str__(self): - return "QTableStructure" % ( - self.filenames_str,len(self.materialized_files.keys()),self.column_names,self.column_types) - __repr__ = __str__ - -class QMetadata(object): - def __init__(self,table_structures=[],output_column_name_list=None,data_loads=[]): - self.table_structures = table_structures - self.output_column_name_list = output_column_name_list - self.data_loads = data_loads - - def __str__(self): - return "QMetadata 0: - s.append("warning_count=%s" % len(self.warnings)) - if self.data is not None: - s.append("row_count=%s" % len(self.data)) - else: - s.append("row_count=None") - if self.metadata is not None: - s.append("metadata=<%s>" % self.metadata) - else: - s.append("metadata=None") - return "QOutput<%s>" % ",".join(s) - __repr__ = __str__ - -class QInputParams(object): - def __init__(self,skip_header=False, - delimiter=' ',input_encoding='UTF-8',gzipped_input=False,with_universal_newlines=False,parsing_mode='relaxed', - expected_column_count=None,keep_leading_whitespace_in_values=False, - disable_double_double_quoting=False,disable_escaped_double_quoting=False, - disable_column_type_detection=False, - input_quoting_mode='minimal',stdin_file=None,stdin_filename='-', - max_column_length_limit=131072): - self.skip_header = skip_header - self.delimiter = delimiter - self.input_encoding = input_encoding - self.gzipped_input = gzipped_input - self.with_universal_newlines = with_universal_newlines - self.parsing_mode = parsing_mode - self.expected_column_count = expected_column_count - self.keep_leading_whitespace_in_values = keep_leading_whitespace_in_values - self.disable_double_double_quoting = disable_double_double_quoting - self.disable_escaped_double_quoting = disable_escaped_double_quoting - self.input_quoting_mode = input_quoting_mode - self.disable_column_type_detection = disable_column_type_detection - self.max_column_length_limit = max_column_length_limit - - def merged_with(self,input_params): - params = QInputParams(**self.__dict__) - if input_params is not None: - params.__dict__.update(**input_params.__dict__) - return params - - def __str__(self): - return "QInputParams<%s>" % str(self.__dict__) - - def __repr__(self): - return "QInputParams(...)" - -class QTextAsData(object): - def __init__(self,default_input_params=QInputParams()): - self.default_input_params = default_input_params - - self.table_creators = {} - - # Create DB object - self.db = Sqlite3DB() - - input_quoting_modes = { 'minimal' : csv.QUOTE_MINIMAL, - 'all' : csv.QUOTE_ALL, - # nonnumeric is not supported for input quoting modes, since we determine the data types - # ourselves instead of letting the csv module try to identify the types - 'none' : csv.QUOTE_NONE } - - def determine_proper_dialect(self,input_params): - - input_quoting_mode_csv_numeral = QTextAsData.input_quoting_modes[input_params.input_quoting_mode] - - if input_params.keep_leading_whitespace_in_values: - skip_initial_space = False - else: - skip_initial_space = True - - dialect = {'skipinitialspace': skip_initial_space, - 'delimiter': input_params.delimiter, 'quotechar': '"' } - dialect['quoting'] = input_quoting_mode_csv_numeral - dialect['doublequote'] = input_params.disable_double_double_quoting - - if input_params.disable_escaped_double_quoting: - dialect['escapechar'] = '\\' - - return dialect - - def get_dialect_id(self,filename): - return 'q_dialect_%s' % filename - - def _load_data(self,filename,input_params=QInputParams(),stdin_file=None,stdin_filename='-',stop_after_analysis=False): - start_time = time.time() - - q_dialect = self.determine_proper_dialect(input_params) - dialect_id = self.get_dialect_id(filename) - csv.register_dialect(dialect_id, **q_dialect) - - csv.field_size_limit(input_params.max_column_length_limit) - - # Create a line splitter - line_splitter = LineSplitter(input_params.delimiter, input_params.expected_column_count) - - # reuse already loaded data, except for stdin file data (stdin file data will always - # be reloaded and overwritten) - if filename in self.table_creators.keys() and filename != stdin_filename: - return None - - # Create the matching database table and populate it - table_creator = TableCreator( - self.db, filename, line_splitter, input_params.skip_header, input_params.gzipped_input, input_params.with_universal_newlines,input_params.input_encoding, - mode=input_params.parsing_mode, expected_column_count=input_params.expected_column_count, - input_delimiter=input_params.delimiter,disable_column_type_detection=input_params.disable_column_type_detection, - stdin_file = stdin_file,stdin_filename = stdin_filename) - - table_creator.populate(dialect_id,stop_after_analysis) - - self.table_creators[filename] = table_creator - - return QDataLoad(filename,start_time,time.time()) - - def load_data(self,filename,input_params=QInputParams(),stop_after_analysis=False): - self._load_data(filename,input_params,stop_after_analysis=stop_after_analysis) - - def load_data_from_string(self,filename,str_data,input_params=QInputParams(),stop_after_analysis=False): - sf = cStringIO.StringIO(str_data) - try: - self._load_data(filename,input_params,stdin_file=sf,stdin_filename=filename,stop_after_analysis=stop_after_analysis) - finally: - if sf is not None: - sf.close() - - def _ensure_data_is_loaded(self,sql_object,input_params,stdin_file,stdin_filename='-',stop_after_analysis=False): - data_loads = [] - - # Get each "table name" which is actually the file name - for filename in sql_object.qtable_names: - data_load = self._load_data(filename,input_params,stdin_file=stdin_file,stdin_filename=stdin_filename,stop_after_analysis=stop_after_analysis) - if data_load is not None: - data_loads.append(data_load) - - return data_loads - - def materialize_sql_object(self,sql_object): - for filename in sql_object.qtable_names: - sql_object.set_effective_table_name(filename,self.table_creators[filename].table_name) - - def _execute(self,query_str,input_params=None,stdin_file=None,stdin_filename='-',stop_after_analysis=False,save_db_to_disk_filename=None,save_db_to_disk_method=None): - warnings = [] - error = None - data_loads = [] - table_structures = [] - - db_results_obj = None - - effective_input_params = self.default_input_params.merged_with(input_params) - - if type(query_str) != unicode: - try: - # Heuristic attempt to auto convert the query to unicode before failing - query_str = query_str.decode('utf-8') - except: - error = QError(EncodedQueryException(''),"Query should be in unicode. Please make sure to provide a unicode literal string or decode it using proper the character encoding.",91) - return QOutput(error = error) - - # Create SQL statement - sql_object = Sql('%s' % query_str) - - try: - load_start_time = time.time() - data_loads += self._ensure_data_is_loaded(sql_object,effective_input_params,stdin_file=stdin_file,stdin_filename=stdin_filename,stop_after_analysis=stop_after_analysis) - - table_structures = self._create_table_structures_list() - - self.materialize_sql_object(sql_object) - - if save_db_to_disk_filename is not None: - self.db.done() - dump_start_time = time.time() - print >>sys.stderr,"Data has been loaded in %4.3f seconds" % (dump_start_time - load_start_time) - print >>sys.stderr,"Saving data to db file %s" % save_db_to_disk_filename - self.db.store_db_to_disk(save_db_to_disk_filename,sql_object.get_qtable_name_effective_table_names(),save_db_to_disk_method) - print >>sys.stderr,"Data has been saved into %s . Saving has taken %4.3f seconds" % (save_db_to_disk_filename,time.time()-dump_start_time) - print >>sys.stderr,"Query to run on the database: %s;" % sql_object.get_effective_sql(True) - # TODO Propagate dump results using a different output class instead of an empty one - - return QOutput() - - # Execute the query and fetch the data - db_results_obj = sql_object.execute_and_fetch(self.db) - - return QOutput( - data = db_results_obj.results, - metadata = QMetadata( - table_structures=table_structures, - output_column_name_list=db_results_obj.query_column_names, - data_loads=data_loads), - warnings = warnings, - error = error) - - except EmptyDataException,e: - warnings.append(QWarning(e,"Warning - data is empty")) - except MissingHeaderException,e: - error = QError(e,e.msg,117) - except FileNotFoundException, e: - error = QError(e,e.msg,30) - except sqlite3.OperationalError, e: - msg = str(e) - error = QError(e,"query error: %s" % msg,1) - if "no such column" in msg and effective_input_params.skip_header: - warnings.append(QWarning(e,'Warning - There seems to be a "no such column" error, and -H (header line) exists. Please make sure that you are using the column names from the header line and not the default (cXX) column names')) - except ColumnCountMismatchException, e: - error = QError(e,e.msg,2) - except (UnicodeDecodeError, UnicodeError), e: - error = QError(e,"Cannot decode data. Try to change the encoding by setting it using the -e parameter. Error:%s" % e,3) - except BadHeaderException, e: - error = QError(e,"Bad header row: %s" % e.msg,35) - except CannotUnzipStdInException,e: - error = QError(e,"Cannot decompress standard input. Pipe the input through zcat in order to decompress.",36) - except UniversalNewlinesExistException,e: - error = QError(e,"Data contains universal newlines. Run q with -U to use universal newlines. Please note that q still doesn't support universal newlines for .gz files or for stdin. Route the data through a regular file to use -U.",103) - except UnprovidedStdInException,e: - error = QError(e,"Standard Input must be provided in order to use it as a table",61) - except CouldNotConvertStringToNumericValueException,e: - error = QError(e,"Could not convert string to a numeric value. Did you use `-w nonnumeric` with unquoted string values? Error: %s" % e.msg,58) - except CouldNotParseInputException,e: - error = QError(e,"Could not parse the input. Please make sure to set the proper -w input-wrapping parameter for your input, and that you use the proper input encoding (-e). Error: %s" % e.msg,59) - except ColumnMaxLengthLimitExceededException,e: - error = QError(e,e.msg,31) - except MissingSqliteBckModuleException, e: - error = QError(e,e.msg,79) - except KeyboardInterrupt,e: - warnings.append(QWarning(e,"Interrupted")) - except Exception, e: - error = QError(e,repr(e),199) - - return QOutput(warnings = warnings,error = error , metadata=QMetadata(table_structures=table_structures,data_loads = data_loads)) - - def execute(self,query_str,input_params=None,stdin_file=None,stdin_filename='-',save_db_to_disk_filename=None,save_db_to_disk_method=None): - return self._execute(query_str,input_params,stdin_file,stdin_filename,stop_after_analysis=False,save_db_to_disk_filename=save_db_to_disk_filename,save_db_to_disk_method=save_db_to_disk_method) - - def unload(self): - - for filename,table_creator in self.table_creators.iteritems(): - try: - table_creator.drop_table() - except: - # Support no-table select queries - pass - self.table_creators = {} - - def _create_materialized_files(self,table_creator): - d = table_creator.materialized_file_dict - m = {} - for filename,mfs in d.iteritems(): - m[filename] = QMaterializedFile(filename,mfs.is_stdin) - return m - - def _create_table_structures_list(self): - table_structures = [] - for filename,table_creator in self.table_creators.iteritems(): - column_names = table_creator.column_inferer.get_column_names() - column_types = [self.db.type_names[table_creator.column_inferer.get_column_dict()[k]].lower() for k in column_names] - materialized_files = self._create_materialized_files(table_creator) - table_structure = QTableStructure(table_creator.filenames_str,materialized_files,column_names,column_types) - table_structures.append(table_structure) - return table_structures - - def analyze(self,query_str,input_params=None,stdin_file=None,stdin_filename='-'): - q_output = self._execute(query_str,input_params,stdin_file,stdin_filename,stop_after_analysis=True) - - return q_output - -def escape_double_quotes_if_needed(v): - x = v.replace('"','""') - return x - -def quote_none_func(output_delimiter,v): - return v - -def quote_minimal_func(output_delimiter,v): - if v is None: - return v - t = type(v) - if t == str or t == unicode and ((output_delimiter in v) or ('"' in v)): - return '"%s"' % (escape_double_quotes_if_needed(v)) - return v; - -def quote_nonnumeric_func(output_delimiter,v): - if v is None: - return v - if type(v) == str or type(v) == unicode: - return '"%s"' % (escape_double_quotes_if_needed(v)) - return v; - -def quote_all_func(output_delimiter,v): - if type(v) == str or type(v) == unicode: - return '"%s"' % (escape_double_quotes_if_needed(v)) - else: - return '"%s"' % v - -class QOutputParams(object): - def __init__(self, - delimiter=' ', - beautify=False, - output_quoting_mode='minimal', - formatting=None, - output_header=False): - self.delimiter = delimiter - self.beautify = beautify - self.output_quoting_mode = output_quoting_mode - self.formatting = formatting - self.output_header = output_header - - def __str__(self): - return "QOutputParams<%s>" % str(self.__dict__) - - def __repr__(self): - return "QOutputParams(...)" - -class QOutputPrinter(object): - output_quoting_modes = { 'minimal' : quote_minimal_func, - 'all' : quote_all_func, - 'nonnumeric' : quote_nonnumeric_func, - 'none' : quote_none_func } - - def __init__(self,output_params,show_tracebacks=False): - self.output_params = output_params - self.show_tracebacks = show_tracebacks - - self.output_field_quoting_func = QOutputPrinter.output_quoting_modes[output_params.output_quoting_mode] - - def print_errors_and_warnings(self,f,results): - if results.status == 'error': - error = results.error - print >>f,error.msg - if self.show_tracebacks: - print >>f,error.traceback - - for warning in results.warnings: - print >>f,"%s" % warning.msg - - def print_analysis(self,f_out,f_err,results): - self.print_errors_and_warnings(f_err,results) - - if results.metadata is None: - return - - if results.metadata.table_structures is None: - return - - for table_structure in results.metadata.table_structures: - print >>f_out,"Table for file: %s" % normalized_filename(table_structure.filenames_str) - for n,t in zip(table_structure.column_names,table_structure.column_types): - print >>f_out," `%s` - %s" % (n,t) - - def print_output(self,f_out,f_err,results): - try: - self._print_output(f_out,f_err,results) - except (UnicodeEncodeError, UnicodeError), e: - print >>f_err, "Cannot encode data. Error:%s" % e - sys.exit(3) - except IOError, e: - if e.errno == 32: - # broken pipe, that's ok - pass - else: - # don't miss other problems for now - raise - except KeyboardInterrupt: - pass - - def _print_output(self,f_out,f_err,results): - self.print_errors_and_warnings(f_err,results) - - data = results.data - - if data is None: - return - - # If the user requested beautifying the output - if self.output_params.beautify: - if self.output_params.output_header: - data_with_possible_headers = data + [tuple(results.metadata.output_column_name_list)] - else: - data_with_possible_headers = data - max_lengths = determine_max_col_lengths(data_with_possible_headers,self.output_field_quoting_func,self.output_params.delimiter) - - if self.output_params.formatting: - formatting_dict = dict( - [(x.split("=")[0], x.split("=")[1]) for x in self.output_params.formatting.split(",")]) - else: - formatting_dict = None - - try: - if self.output_params.output_header and results.metadata.output_column_name_list is not None: - data.insert(0,results.metadata.output_column_name_list) - for rownum, row in enumerate(data): - row_str = [] - skip_formatting = rownum == 0 and self.output_params.output_header - for i, col in enumerate(row): - if formatting_dict is not None and str(i + 1) in formatting_dict.keys() and not skip_formatting: - fmt_str = formatting_dict[str(i + 1)] - else: - if self.output_params.beautify: - fmt_str = "%%-%ss" % max_lengths[i] - else: - fmt_str = "%s" - - if col is not None: - row_str.append(fmt_str % self.output_field_quoting_func(self.output_params.delimiter,col)) - else: - row_str.append(fmt_str % "") - - f_out.write(self.output_params.delimiter.join(row_str) + "\n") - except (UnicodeEncodeError, UnicodeError), e: - print >>sys.stderr, "Cannot encode data. Error:%s" % e - sys.exit(3) - except TypeError,e: - print >>sys.stderr, "Error while formatting output: %s" % e - sys.exit(4) - except IOError, e: - if e.errno == 32: - # broken pipe, that's ok - pass - else: - # don't miss other problem for now - raise - except KeyboardInterrupt: - pass - - try: - # Prevent python bug when order of pipe shutdowns is reversed - f_out.flush() - except IOError, e: - pass - -def run_standalone(): - p = ConfigParser() - p.read([os.path.expanduser('~/.qrc'), '.qrc']) - - def get_option_with_default(p, option_type, option, default): - if not p.has_option('options', option): - return default - if option_type == 'boolean': - return p.getboolean('options', option) - elif option_type == 'int': - return p.getint('options', option) - elif option_type == 'string': - return p.get('options', option) - elif option_type == 'escaped_string': - return p.get('options', option).decode('string-escape') - else: - raise Exception("Unknown option type") - - default_beautify = get_option_with_default(p, 'boolean', 'beautify', False) - default_gzipped = get_option_with_default(p, 'boolean', 'gzipped', False) - default_delimiter = get_option_with_default( - p, 'escaped_string', 'delimiter', None) - default_output_delimiter = get_option_with_default( - p, 'escaped_string', 'output_delimiter', None) - default_skip_header = get_option_with_default(p, 'int', 'skip_header', 0) - default_formatting = get_option_with_default(p, 'string', 'formatting', None) - default_encoding = get_option_with_default(p, 'string', 'encoding', 'UTF-8') - default_output_encoding = get_option_with_default(p, 'string', 'encoding', None) - default_query_encoding = get_option_with_default(p, 'string', 'query_encoding', locale.getpreferredencoding()) - default_output_header = get_option_with_default(p, 'string', 'output_header', False) - - parser = OptionParser(usage=""" - q allows performing SQL-like statements on tabular text data. - - Its purpose is to bring SQL expressive power to manipulating text data using the Linux command line. - - Basic usage is q "" where table names are just regular file names (Use - to read from standard input) - When the input contains a header row, use -H, and column names will be set according to the header row content. If there isn't a header row, then columns will automatically be named c1..cN. - - Column types are detected automatically. Use -A in order to see the column name/type analysis. - - Delimiter can be set using the -d (or -t) option. Output delimiter can be set using -D - - All sqlite3 SQL constructs are supported. - - Examples: - - Example 1: ls -ltrd * | q "select c1,count(1) from - group by c1" - This example would print a count of each unique permission string in the current folder. - - Example 2: seq 1 1000 | q "select avg(c1),sum(c1) from -" - This example would provide the average and the sum of the numbers in the range 1 to 1000 - - Example 3: sudo find /tmp -ls | q "select c5,c6,sum(c7)/1024.0/1024 as total from - group by c5,c6 order by total desc" - This example will output the total size in MB per user+group in the /tmp subtree - - - See the help or https://github.com/harelba/q/ for more details. - """) - - #----------------------------------------------- - parser.add_option("-v", "--version", dest="version", default=False, action="store_true", - help="Print version") - parser.add_option("-V", "--verbose", dest="verbose", default=False, action="store_true", - help="Print debug info in case of problems") - parser.add_option("-S", "--save-db-to-disk", dest="save_db_to_disk_filename", default=None, - help="Save database to an sqlite database file") - parser.add_option("", "--save-db-to-disk-method", dest="save_db_to_disk_method", default='standard', - help="Method to use to save db to disk. 'standard' does not require any deps, 'fast' currenty requires manually running `pip install sqlitebck` on your python installation. Once packing issues are solved, the fast method will be the default.") - #----------------------------------------------- - input_data_option_group = OptionGroup(parser,"Input Data Options") - input_data_option_group.add_option("-H", "--skip-header", dest="skip_header", default=default_skip_header, action="store_true", - help="Skip header row. This has been changed from earlier version - Only one header row is supported, and the header row is used for column naming") - input_data_option_group.add_option("-d", "--delimiter", dest="delimiter", default=default_delimiter, - help="Field delimiter. If none specified, then space is used as the delimiter.") - input_data_option_group.add_option("-t", "--tab-delimited", dest="tab_delimited", default=False, action="store_true", - help="Same as -d . Just a shorthand for handling standard tab delimited file You can use $'\\t' if you want (this is how Linux expects to provide tabs in the command line") - input_data_option_group.add_option("-e", "--encoding", dest="encoding", default=default_encoding, - help="Input file encoding. Defaults to UTF-8. set to none for not setting any encoding - faster, but at your own risk...") - input_data_option_group.add_option("-z", "--gzipped", dest="gzipped", default=default_gzipped, action="store_true", - help="Data is gzipped. Useful for reading from stdin. For files, .gz means automatic gunzipping") - input_data_option_group.add_option("-A", "--analyze-only", dest="analyze_only", action='store_true', - help="Analyze sample input and provide information about data types") - input_data_option_group.add_option("-m", "--mode", dest="mode", default="relaxed", - help="Data parsing mode. fluffy, relaxed and strict. In strict mode, the -c column-count parameter must be supplied as well") - input_data_option_group.add_option("-c", "--column-count", dest="column_count", default=None, - help="Specific column count when using relaxed or strict mode") - input_data_option_group.add_option("-k", "--keep-leading-whitespace", dest="keep_leading_whitespace_in_values", default=False, action="store_true", - help="Keep leading whitespace in values. Default behavior strips leading whitespace off values, in order to provide out-of-the-box usability for simple use cases. If you need to preserve whitespace, use this flag.") - input_data_option_group.add_option("--disable-double-double-quoting", dest="disable_double_double_quoting", default=True, action="store_false", - help="Disable support for double double-quoting for escaping the double quote character. By default, you can use \"\" inside double quoted fields to escape double quotes. Mainly for backward compatibility.") - input_data_option_group.add_option("--disable-escaped-double-quoting", dest="disable_escaped_double_quoting", default=True, action="store_false", - help="Disable support for escaped double-quoting for escaping the double quote character. By default, you can use \\\" inside double quoted fields to escape double quotes. Mainly for backward compatibility.") - input_data_option_group.add_option("--as-text", dest="disable_column_type_detection", default=False, action="store_true", - help="Don't detect column types - All columns will be treated as text columns") - input_data_option_group.add_option("-w","--input-quoting-mode",dest="input_quoting_mode",default="minimal", - help="Input quoting mode. Possible values are all, minimal and none. Note the slightly misleading parameter name, and see the matching -W parameter for output quoting.") - input_data_option_group.add_option("-M","--max-column-length-limit",dest="max_column_length_limit",default=131072, - help="Sets the maximum column length.") - input_data_option_group.add_option("-U","--with-universal-newlines",dest="with_universal_newlines",default=False,action="store_true", - help="Expect universal newlines in the data. Limitation: -U works only with regular files for now, stdin or .gz files are not supported yet.") - parser.add_option_group(input_data_option_group) - #----------------------------------------------- - output_data_option_group = OptionGroup(parser,"Output Options") - output_data_option_group.add_option("-D", "--output-delimiter", dest="output_delimiter", default=default_output_delimiter, - help="Field delimiter for output. If none specified, then the -d delimiter is used if present, or space if no delimiter is specified") - output_data_option_group.add_option("-T", "--tab-delimited-output", dest="tab_delimited_output", default=False, action="store_true", - help="Same as -D . Just a shorthand for outputting tab delimited output. You can use -D $'\\t' if you want.") - output_data_option_group.add_option("-O", "--output-header", dest="output_header", default=default_output_header, action="store_true",help="Output header line. Output column-names are determined from the query itself. Use column aliases in order to set your column names in the query. For example, 'select name FirstName,value1/value2 MyCalculation from ...'. This can be used even if there was no header in the input.") - output_data_option_group.add_option("-b", "--beautify", dest="beautify", default=default_beautify, action="store_true", - help="Beautify output according to actual values. Might be slow...") - output_data_option_group.add_option("-f", "--formatting", dest="formatting", default=default_formatting, - help="Output-level formatting, in the format X=fmt,Y=fmt etc, where X,Y are output column numbers (e.g. 1 for first SELECT column etc.") - output_data_option_group.add_option("-E", "--output-encoding", dest="output_encoding", default=default_output_encoding, - help="Output encoding. Defaults to 'none', leading to selecting the system/terminal encoding") - output_data_option_group.add_option("-W","--output-quoting-mode",dest="output_quoting_mode",default="minimal", - help="Output quoting mode. Possible values are all, minimal, nonnumeric and none. Note the slightly misleading parameter name, and see the matching -w parameter for input quoting.") - parser.add_option_group(output_data_option_group) - #----------------------------------------------- - query_option_group = OptionGroup(parser,"Query Related Options") - query_option_group.add_option("-q", "--query-filename", dest="query_filename", default=None, - help="Read query from the provided filename instead of the command line, possibly using the provided query encoding (using -Q).") - query_option_group.add_option("-Q", "--query-encoding", dest="query_encoding", default=default_query_encoding, - help="query text encoding. Experimental. Please send your feedback on this") - parser.add_option_group(query_option_group) - #----------------------------------------------- - - (options, args) = parser.parse_args() - - if options.version: - print_credentials() - sys.exit(0) - - if len(args) == 0 and options.query_filename is None: - print_credentials() - print >>sys.stderr,"Must provide at least one query in the command line, or through a file with the -q parameter" - sys.exit(1) - - if options.query_filename is not None: - if len(args) != 0: - print >>sys.stderr,"Can't provide both a query file and a query on the command line" - sys.exit(1) - try: - f = file(options.query_filename) - query_strs = [f.read()] - f.close() - except: - print >>sys.stderr,"Could not read query from file %s" % options.query_filename - sys.exit(1) - else: - query_strs = args - - if options.query_encoding is not None and options.query_encoding != 'none': - try: - for idx in range(len(query_strs)): - query_strs[idx] = query_strs[idx].decode(options.query_encoding).strip() - - if len(query_strs[idx]) == 0: - print >>sys.stderr,"Query cannot be empty (query number %s)" % (idx+1) - sys.exit(1) - - except Exception,e: - print >>sys.stderr,"Could not decode query number %s using the provided query encoding (%s)" % (idx+1,options.query_encoding) - sys.exit(3) - - if options.mode not in ['fluffy', 'relaxed', 'strict']: - print >>sys.stderr, "Parsing mode can be one of fluffy, relaxed or strict" - sys.exit(13) - - output_encoding = get_stdout_encoding(options.output_encoding) - try: - STDOUT = codecs.getwriter(output_encoding)(sys.stdout) - except: - print >>sys.stderr,"Could not create output stream using output encoding %s" % (output_encoding) - sys.exit(200) - - # If the user flagged for a tab-delimited file then set the delimiter to tab - if options.tab_delimited: - options.delimiter = '\t' - - if options.tab_delimited_output: - options.output_delimiter = '\t' - - if options.delimiter is None: - options.delimiter = ' ' - elif len(options.delimiter) != 1: - print >>sys.stderr, "Delimiter must be one character only" - sys.exit(5) - - if options.input_quoting_mode not in QTextAsData.input_quoting_modes.keys(): - print >>sys.stderr,"Input quoting mode can only be one of %s. It cannot be set to '%s'" % (",".join(QTextAsData.input_quoting_modes.keys()),options.input_quoting_mode) - sys.exit(55) - - if options.output_quoting_mode not in QOutputPrinter.output_quoting_modes.keys(): - print >>sys.stderr,"Output quoting mode can only be one of %s. It cannot be set to '%s'" % (",".join(QOutputPrinter.output_quoting_modes.keys()),options.input_quoting_mode) - sys.exit(56) - - if options.column_count is not None: - expected_column_count = int(options.column_count) - else: - # infer automatically - expected_column_count = None - - if options.encoding != 'none': - try: - codecs.lookup(options.encoding) - except LookupError: - print >>sys.stderr, "Encoding %s could not be found" % options.encoding - sys.exit(10) - - if options.output_delimiter: - # If output delimiter is specified, then we use it - output_delimiter = options.output_delimiter - else: - # Otherwise, - if options.delimiter: - # if an input delimiter is specified, then we use it as the output as - # well - options.output_delimiter = options.delimiter - else: - # if no input delimiter is specified, then we use space as the default - # (since no input delimiter means any whitespace) - options.output_delimiter = " " - - try: - max_column_length_limit = int(options.max_column_length_limit) - if max_column_length_limit < 1: - raise Exception() - except: - print >> sys.stderr, "Max column length limit must be a positive integer (%s)" % max_column_length_limit - sys.exit(31) - - if options.save_db_to_disk_filename is not None: - if options.analyze_only: - print >>sys.stderr,"Cannot save database to disk when running with -A (analyze-only) option." - sys.exit(119) - - print >>sys.stderr,"Going to save data into a disk database: %s" % options.save_db_to_disk_filename - if os.path.exists(options.save_db_to_disk_filename): - print >> sys.stderr, "Disk database file %s already exists." % options.save_db_to_disk_filename - sys.exit(77) - - if options.save_db_to_disk_method is not None: - if options.save_db_to_disk_method not in ['standard','fast']: - print >>sys.stderr,"save-db-to-disk method should be either standard or fast (%s)" % options.save_db_to_disk_method - sys.exit(78) - - default_input_params = QInputParams(skip_header=options.skip_header, - delimiter=options.delimiter, - input_encoding=options.encoding, - gzipped_input=options.gzipped, - with_universal_newlines=options.with_universal_newlines, - parsing_mode=options.mode, - expected_column_count=expected_column_count, - keep_leading_whitespace_in_values=options.keep_leading_whitespace_in_values, - disable_double_double_quoting=options.disable_double_double_quoting, - disable_escaped_double_quoting=options.disable_escaped_double_quoting, - input_quoting_mode=options.input_quoting_mode, - disable_column_type_detection=options.disable_column_type_detection, - max_column_length_limit=max_column_length_limit) - q_engine = QTextAsData(default_input_params=default_input_params) - - output_params = QOutputParams( - delimiter=options.output_delimiter, - beautify=options.beautify, - output_quoting_mode=options.output_quoting_mode, - formatting=options.formatting, - output_header=options.output_header) - q_output_printer = QOutputPrinter(output_params,show_tracebacks=options.verbose) - - for query_str in query_strs: - if options.analyze_only: - q_output = q_engine.analyze(query_str,stdin_file=sys.stdin) - q_output_printer.print_analysis(STDOUT,sys.stderr,q_output) - else: - q_output = q_engine.execute(query_str,stdin_file=sys.stdin,save_db_to_disk_filename=options.save_db_to_disk_filename,save_db_to_disk_method=options.save_db_to_disk_method) - q_output_printer.print_output(STDOUT,sys.stderr,q_output) - - if q_output.status == 'error': - sys.exit(q_output.error.errorcode) - - q_engine.unload() - - sys.exit(0) - - -if __name__ == '__main__': - run_standalone() diff --git a/bin/q.py b/bin/q.py new file mode 100755 index 00000000..2a2f6771 --- /dev/null +++ b/bin/q.py @@ -0,0 +1,3640 @@ +#!/usr/bin/env python3 +# -*- coding: utf-8 -*- + +# Copyright (C) 2012-2021 Harel Ben-Attia +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 3, or (at your option) +# any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details (doc/LICENSE contains +# a copy of it) +# +# +# Name : q (With respect to The Q Continuum) +# Author : Harel Ben-Attia - harelba@gmail.com, harelba @ github, @harelba on twitter +# +# +# q allows performing SQL-like statements on tabular text data. +# +# Its purpose is to bring SQL expressive power to manipulating text data using the Linux command line. +# +# Full Documentation and details in https://harelba.github.io/q/ +# +# Run with --help for command line details +# +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +from collections import OrderedDict +from sqlite3.dbapi2 import OperationalError +from uuid import uuid4 + +q_version = '3.1.6' + +#__all__ = [ 'QTextAsData' ] + +import os +import sys +import sqlite3 +import glob +from argparse import ArgumentParser +import codecs +import locale +import time +import re +from six.moves import configparser, range, filter +import traceback +import csv +import uuid +import math +import six +import io +import json +import datetime +import hashlib + +if six.PY2: + assert False, 'Python 2 is not longer supported by q' + +long = int +unicode = six.text_type + +DEBUG = bool(os.environ.get('Q_DEBUG', None)) or '-V' in sys.argv +SQL_DEBUG = False + +if DEBUG: + def xprint(*args,**kwargs): + print(datetime.datetime.utcnow().isoformat()," DEBUG ",*args,file=sys.stderr,**kwargs) + + def iprint(*args,**kwargs): + print(datetime.datetime.utcnow().isoformat()," INFO ",*args,file=sys.stderr,**kwargs) + + def sqlprint(*args,**kwargs): + pass +else: + def xprint(*args,**kwargs): pass + def iprint(*args,**kwargs): pass + def sqlprint(*args,**kwargs): pass + +if SQL_DEBUG: + def sqlprint(*args,**kwargs): + print(datetime.datetime.utcnow().isoformat(), " SQL ", *args, file=sys.stderr, **kwargs) + + +def get_stdout_encoding(encoding_override=None): + if encoding_override is not None and encoding_override != 'none': + return encoding_override + + if sys.stdout.isatty(): + return sys.stdout.encoding + else: + return locale.getpreferredencoding() + +SHOW_SQL = False + +sha_algorithms = { + 1 : hashlib.sha1, + 224: hashlib.sha224, + 256: hashlib.sha256, + 386: hashlib.sha384, + 512: hashlib.sha512 +} + +def sha(data,algorithm,encoding): + try: + f = sha_algorithms[algorithm] + return f(six.text_type(data).encode(encoding)).hexdigest() + except Exception as e: + print(e) + +# For backward compatibility only (doesn't handle encoding well enough) +def sha1(data): + return hashlib.sha1(six.text_type(data).encode('utf-8')).hexdigest() + +# TODO Add caching of compiled regexps - Will be added after benchmarking capability is baked in +def regexp(regular_expression, data): + if data is not None: + if not isinstance(data, str) and not isinstance(data, unicode): + data = str(data) + return re.search(regular_expression, data) is not None + else: + return False + +def regexp_extract(regular_expression, data,group_number): + if data is not None: + if not isinstance(data, str) and not isinstance(data, unicode): + data = str(data) + m = re.search(regular_expression, data) + if m is not None: + return m.groups()[group_number] + else: + return False + +def md5(data,encoding): + m = hashlib.md5() + m.update(six.text_type(data).encode(encoding)) + return m.hexdigest() + +def sqrt(data): + return math.sqrt(data) + +def power(data,p): + return data**p + +def file_ext(data): + if data is None: + return None + + return os.path.splitext(data)[1] + +def file_folder(data): + if data is None: + return None + return os.path.split(data)[0] + +def file_basename(data): + if data is None: + return None + return os.path.split(data)[1] + +def file_basename_no_ext(data): + if data is None: + return None + + return os.path.split(os.path.splitext(data)[0])[-1] + +def percentile(l, p): + # TODO Alpha implementation, need to provide multiple interpolation methods, and add tests + if not l: + return None + k = p*(len(l) - 1) + f = math.floor(k) + c = math.ceil(k) + if c == f: + return l[int(k)] + return (c-k) * l[int(f)] + (k-f) * l[int(c)] + +# TODO Streaming Percentile to prevent memory consumption blowup for large datasets +class StrictPercentile(object): + def __init__(self): + self.values = [] + self.p = None + + def step(self,value,p): + if self.p is None: + self.p = p + self.values.append(value) + + def finalize(self): + if len(self.values) == 0 or (self.p < 0 or self.p > 1): + return None + else: + return percentile(sorted(self.values),self.p) + +class StdevPopulation(object): + def __init__(self): + self.M = 0.0 + self.S = 0.0 + self.k = 0 + + def step(self, value): + try: + # Ignore nulls + if value is None: + return + val = float(value) # if fails, skips this iteration, which also ignores nulls + tM = self.M + self.k += 1 + self.M += ((val - tM) / self.k) + self.S += ((val - tM) * (val - self.M)) + except ValueError: + # TODO propagate udf errors to console + raise Exception("Data is not numeric when calculating stddev (%s)" % value) + + def finalize(self): + if self.k <= 1: # avoid division by zero + return None + else: + return math.sqrt(self.S / (self.k)) + +class StdevSample(object): + def __init__(self): + self.M = 0.0 + self.S = 0.0 + self.k = 0 + + def step(self, value): + try: + # Ignore nulls + if value is None: + return + val = float(value) # if fails, skips this iteration, which also ignores nulls + tM = self.M + self.k += 1 + self.M += ((val - tM) / self.k) + self.S += ((val - tM) * (val - self.M)) + except ValueError: + # TODO propagate udf errors to console + raise Exception("Data is not numeric when calculating stddev (%s)" % value) + + def finalize(self): + if self.k <= 1: # avoid division by zero + return None + else: + return math.sqrt(self.S / (self.k-1)) + +class FunctionType(object): + REGULAR = 1 + AGG = 2 + +class UserFunctionDef(object): + def __init__(self,func_type,name,usage,description,func_or_obj,param_count): + self.func_type = func_type + self.name = name + self.usage = usage + self.description = description + self.func_or_obj = func_or_obj + self.param_count = param_count + +user_functions = [ + UserFunctionDef(FunctionType.REGULAR, + "regexp","regexp(,) = <1|0>", + "Find regexp in string expression. Returns 1 if found or 0 if not", + regexp, + 2), + UserFunctionDef(FunctionType.REGULAR, + "regexp_extract","regexp_extract(,,group_number) = ", + "Get regexp capture group content", + regexp_extract, + 3), + UserFunctionDef(FunctionType.REGULAR, + "sha","sha(,,) = ", + "Calculate sha of some expression. Algorithm can be one of 1,224,256,384,512. For now encoding must be manually provided. Will use the input encoding automatically in the future.", + sha, + 3), + UserFunctionDef(FunctionType.REGULAR, + "sha1","sha1() = ", + "Exists for backward compatibility only, since it doesn't handle encoding properly. Calculates sha1 of some expression", + sha1, + 1), + UserFunctionDef(FunctionType.REGULAR, + "md5","md5(,) = ", + "Calculate md5 of expression. Returns a hex-string of the result. Currently requires to manually provide the encoding of the data. Will be taken automatically from the input encoding in the future.", + md5, + 2), + UserFunctionDef(FunctionType.REGULAR, + "sqrt","sqrt() = ", + "Calculate the square root of the expression", + sqrt, + 1), + UserFunctionDef(FunctionType.REGULAR, + "power","power(,) = ", + "Raise expr1 to the power of expr2", + power, + 2), + UserFunctionDef(FunctionType.REGULAR, + "file_ext","file_ext() = ", + "Get the extension of a filename", + file_ext, + 1), + UserFunctionDef(FunctionType.REGULAR, + "file_folder","file_folder() = ", + "Get the folder part of a filename", + file_folder, + 1), + UserFunctionDef(FunctionType.REGULAR, + "file_basename","file_basename() = ", + "Get the basename of a filename, including extension if any", + file_basename, + 1), + UserFunctionDef(FunctionType.REGULAR, + "file_basename_no_ext","file_basename_no_ext() = ", + "Get the basename of a filename, without the extension if there is one", + file_basename_no_ext, + 1), + UserFunctionDef(FunctionType.AGG, + "percentile","percentile(,) = ", + "Calculate the strict percentile of a set of a values.", + StrictPercentile, + 2), + UserFunctionDef(FunctionType.AGG, + "stddev_pop","stddev_pop() = ", + "Calculate the population standard deviation of a set of values", + StdevPopulation, + 1), + UserFunctionDef(FunctionType.AGG, + "stddev_sample","stddev_sample() = ", + "Calculate the sample standard deviation of a set of values", + StdevSample, + 1) +] + +def print_user_functions(): + for udf in user_functions: + print("Function: %s" % udf.name) + print(" Usage: %s" % udf.usage) + print(" Description: %s" % udf.description) + +class Sqlite3DBResults(object): + def __init__(self,query_column_names,results): + self.query_column_names = query_column_names + self.results = results + + def __str__(self): + return "Sqlite3DBResults" % (len(self.results),str(self.query_column_names)) + __repr__ = __str__ + +def get_sqlite_type_affinity(sqlite_type): + sqlite_type = sqlite_type.upper() + if 'INT' in sqlite_type: + return 'INTEGER' + elif 'CHAR' in sqlite_type or 'TEXT' in sqlite_type or 'CLOB' in sqlite_type: + return 'TEXT' + elif 'BLOB' in sqlite_type: + return 'BLOB' + elif 'REAL' in sqlite_type or 'FLOA' in sqlite_type or 'DOUB' in sqlite_type: + return 'REAL' + else: + return 'NUMERIC' + +def sqlite_type_to_python_type(sqlite_type): + SQLITE_AFFINITY_TO_PYTHON_TYPE_NAMES = { + 'INTEGER': long, + 'TEXT': unicode, + 'BLOB': bytes, + 'REAL': float, + 'NUMERIC': float + } + return SQLITE_AFFINITY_TO_PYTHON_TYPE_NAMES[get_sqlite_type_affinity(sqlite_type)] + + +class Sqlite3DB(object): + # TODO Add metadata table with qsql file version + + QCATALOG_TABLE_NAME = '_qcatalog' + NUMERIC_COLUMN_TYPES = {int, long, float} + PYTHON_TO_SQLITE_TYPE_NAMES = { str: 'TEXT', int: 'INT', long : 'INT' , float: 'REAL', None: 'TEXT' } + + + def __str__(self): + return "Sqlite3DB" % self.sqlite_db_url + __repr__ = __str__ + + def __init__(self, db_id, sqlite_db_url, sqlite_db_filename, create_qcatalog, show_sql=SHOW_SQL): + self.show_sql = show_sql + self.create_qcatalog = create_qcatalog + + self.db_id = db_id + # TODO Is this needed anymore? + self.sqlite_db_filename = sqlite_db_filename + self.sqlite_db_url = sqlite_db_url + self.conn = sqlite3.connect(self.sqlite_db_url, uri=True) + self.last_temp_table_id = 10000 + self.cursor = self.conn.cursor() + self.add_user_functions() + + if create_qcatalog: + self.create_qcatalog_table() + else: + xprint('Not creating qcatalog for db_id %s' % db_id) + + def retrieve_all_table_names(self): + return [x[0] for x in self.execute_and_fetch("select tbl_name from sqlite_master where type='table'").results] + + def get_sqlite_table_info(self,table_name): + return self.execute_and_fetch('PRAGMA table_info(%s)' % table_name).results + + def get_sqlite_database_list(self): + return self.execute_and_fetch('pragma database_list').results + + def find_new_table_name(self,planned_table_name): + existing_table_names = self.retrieve_all_table_names() + + possible_indices = range(1,1000) + + for index in possible_indices: + if index == 1: + suffix = '' + else: + suffix = '_%s' % index + + table_name_attempt = '%s%s' % (planned_table_name,suffix) + + if table_name_attempt not in existing_table_names: + xprint("Found free table name %s in db %s for planned table name %s" % (table_name_attempt,self.db_id,planned_table_name)) + return table_name_attempt + + # TODO Add test for this + raise Exception('Cannot find free table name in db %s for planned table name %s' % (self.db_id,planned_table_name)) + + def create_qcatalog_table(self): + if not self.qcatalog_table_exists(): + xprint("qcatalog table does not exist. Creating it") + r = self.conn.execute("""CREATE TABLE %s ( + qcatalog_entry_id text not null primary key, + content_signature_key text, + temp_table_name text, + content_signature text, + creation_time text, + source_type text, + source text)""" % self.QCATALOG_TABLE_NAME).fetchall() + else: + xprint("qcatalog table already exists. No need to create it") + + def qcatalog_table_exists(self): + return sqlite_table_exists(self.conn,self.QCATALOG_TABLE_NAME) + + def calculate_content_signature_key(self,content_signature): + assert type(content_signature) == OrderedDict + pp = json.dumps(content_signature,sort_keys=True) + xprint("Calculating content signature for:",pp,six.b(pp)) + return hashlib.sha1(six.b(pp)).hexdigest() + + def add_to_qcatalog_table(self, temp_table_name, content_signature, creation_time,source_type, source): + assert source is not None + assert source_type is not None + content_signature_key = self.calculate_content_signature_key(content_signature) + xprint("db_id: %s Adding to qcatalog table: %s. Calculated signature key %s" % (self.db_id, temp_table_name,content_signature_key)) + r = self.execute_and_fetch( + 'INSERT INTO %s (qcatalog_entry_id,content_signature_key, temp_table_name,content_signature,creation_time,source_type,source) VALUES (?,?,?,?,?,?,?)' % self.QCATALOG_TABLE_NAME, + (str(uuid4()),content_signature_key,temp_table_name,json.dumps(content_signature),creation_time,source_type,source)) + # Ensure transaction is completed + self.conn.commit() + + def get_from_qcatalog(self, content_signature): + content_signature_key = self.calculate_content_signature_key(content_signature) + xprint("Finding table in db_id %s that matches content signature key %s" % (self.db_id,content_signature_key)) + + field_names = ["content_signature_key", "temp_table_name", "content_signature", "creation_time","source_type","source","qcatalog_entry_id"] + + q = "SELECT %s FROM %s where content_signature_key = ?" % (",".join(field_names),self.QCATALOG_TABLE_NAME) + r = self.execute_and_fetch(q,(content_signature_key,)) + + if r is None: + return None + + if len(r.results) == 0: + return None + + if len(r.results) > 1: + raise Exception("Bug - Exactly one result should have been provided: %s" % str(r.results)) + + d = dict(zip(field_names,r.results[0])) + return d + + def get_from_qcatalog_using_table_name(self, temp_table_name): + xprint("getting from qcatalog using table name") + + field_names = ["content_signature", "temp_table_name","creation_time","source_type","source","content_signature_key","qcatalog_entry_id"] + + q = "SELECT %s FROM %s where temp_table_name = ?" % (",".join(field_names),self.QCATALOG_TABLE_NAME) + xprint("Query from qcatalog %s params %s" % (q,str(temp_table_name,))) + r = self.execute_and_fetch(q,(temp_table_name,)) + xprint("results: ",r.results) + + if r is None: + return None + + if len(r.results) == 0: + return None + + if len(r.results) > 1: + raise Exception("Bug - Exactly one result should have been provided: %s" % str(r.results)) + + d = dict(zip(field_names,r.results[0])) + # content_signature should be the first in the list of field_names + cs = OrderedDict(json.loads(r.results[0][0])) + if self.calculate_content_signature_key(cs) != d['content_signature_key']: + raise Exception('Table contains an invalid entry - content signature key is not matching the actual content signature') + return d + + def get_all_from_qcatalog(self): + xprint("getting from qcatalog using table name") + + field_names = ["temp_table_name", "content_signature", "creation_time","source_type","source","qcatalog_entry_id"] + + q = "SELECT %s FROM %s" % (",".join(field_names),self.QCATALOG_TABLE_NAME) + xprint("Query from qcatalog %s" % q) + r = self.execute_and_fetch(q) + + if r is None: + return None + + def convert(res): + d = dict(zip(field_names, res)) + cs = OrderedDict(json.loads(res[1])) + d['content_signature_key'] = self.calculate_content_signature_key(cs) + return d + + rr = [convert(r) for r in r.results] + + return rr + + def done(self): + xprint("Closing database %s" % self.db_id) + try: + self.conn.commit() + self.conn.close() + xprint("Database %s closed" % self.db_id) + except Exception as e: + xprint("Could not close database %s" % self.db_id) + raise + + def add_user_functions(self): + for udf in user_functions: + if type(udf.func_or_obj) == type(object): + self.conn.create_aggregate(udf.name,udf.param_count,udf.func_or_obj) + elif type(udf.func_or_obj) == type(md5): + self.conn.create_function(udf.name,udf.param_count,udf.func_or_obj) + else: + raise Exception("Invalid user function definition %s" % str(udf)) + + def is_numeric_type(self, column_type): + return column_type in Sqlite3DB.NUMERIC_COLUMN_TYPES + + def update_many(self, sql, params): + try: + sqlprint(sql, " params: " + str(params)) + self.cursor.executemany(sql, params) + _ = self.cursor.fetchall() + finally: + pass # cursor.close() + + def execute_and_fetch(self, q,params = None): + try: + try: + if self.show_sql: + print(repr(q)) + if params is None: + r = self.cursor.execute(q) + else: + r = self.cursor.execute(q,params) + if self.cursor.description is not None: + # we decode the column names, so they can be encoded to any output format later on + query_column_names = [c[0] for c in self.cursor.description] + else: + query_column_names = None + result = self.cursor.fetchall() + finally: + pass # cursor.close() + except OperationalError as e: + raise SqliteOperationalErrorException("Failed executing sqlite query %s with params %s . error: %s" % (q,params,str(e)),e) + return Sqlite3DBResults(query_column_names,result) + + def _get_as_list_str(self, l): + return ",".join(['"%s"' % x.replace('"', '""') for x in l]) + + def generate_insert_row(self, table_name, column_names): + col_names_str = self._get_as_list_str(column_names) + question_marks = ", ".join(["?" for i in range(0, len(column_names))]) + return 'INSERT INTO %s (%s) VALUES (%s)' % (table_name, col_names_str, question_marks) + + # Get a list of column names so order will be preserved (Could have used OrderedDict, but + # then we would need python 2.7) + def generate_create_table(self, table_name, column_names, column_dict): + # Convert dict from python types to db types + column_name_to_db_type = dict( + (n, Sqlite3DB.PYTHON_TO_SQLITE_TYPE_NAMES[t]) for n, t in six.iteritems(column_dict)) + column_defs = ','.join(['"%s" %s' % ( + n.replace('"', '""'), column_name_to_db_type[n]) for n in column_names]) + return 'CREATE TABLE %s (%s)' % (table_name, column_defs) + + def generate_temp_table_name(self): + # WTF - From my own past mutable-self + self.last_temp_table_id += 1 + tn = "temp_table_%s" % self.last_temp_table_id + return tn + + def generate_drop_table(self, table_name): + return "DROP TABLE %s" % table_name + + def drop_table(self, table_name): + return self.execute_and_fetch(self.generate_drop_table(table_name)) + + def attach_and_copy_table(self, from_db, relevant_table,stop_after_analysis): + xprint("Attaching %s into db %s and copying table %s into it" % (from_db,self,relevant_table)) + temp_db_id = 'temp_db_id' + q = "attach '%s' as %s" % (from_db.sqlite_db_url,temp_db_id) + xprint("Attach query: %s" % q) + c = self.execute_and_fetch(q) + + new_temp_table_name = 'temp_table_%s' % (self.last_temp_table_id + 1) + fully_qualified_table_name = '%s.%s' % (temp_db_id,relevant_table) + + if stop_after_analysis: + limit = ' limit 100' + else: + limit = '' + + copy_query = 'create table %s as select * from %s %s' % (new_temp_table_name,fully_qualified_table_name,limit) + copy_results = self.execute_and_fetch(copy_query) + xprint("Copied %s.%s into %s in db_id %s. Results %s" % (temp_db_id,relevant_table,new_temp_table_name,self.db_id,copy_results)) + self.last_temp_table_id += 1 + + xprint("Copied table into %s. Detaching db that was attached temporarily" % self.db_id) + + q = "detach database %s" % temp_db_id + xprint("detach query: %s" % q) + c = self.execute_and_fetch(q) + xprint(c) + return new_temp_table_name + + +class CouldNotConvertStringToNumericValueException(Exception): + + def __init__(self, msg): + self.msg = msg + + def __str(self): + return repr(self.msg) + +class SqliteOperationalErrorException(Exception): + + def __init__(self, msg,original_error): + self.msg = msg + self.original_error = original_error + + def __str(self): + return repr(self.msg) + "//" + repr(self.original_error) + +class IncorrectDefaultValueException(Exception): + + def __init__(self, option_type,option,actual_value): + self.option_type = option_type + self.option = option + self.actual_value = actual_value + + def __str__(self): + return repr(self) + +class NonExistentTableNameInQsql(Exception): + + def __init__(self, qsql_filename,table_name,existing_table_names): + self.qsql_filename = qsql_filename + self.table_name = table_name + self.existing_table_names = existing_table_names + +class NonExistentTableNameInSqlite(Exception): + + def __init__(self, qsql_filename,table_name,existing_table_names): + self.qsql_filename = qsql_filename + self.table_name = table_name + self.existing_table_names = existing_table_names + +class TooManyTablesInQsqlException(Exception): + + def __init__(self, qsql_filename,existing_table_names): + self.qsql_filename = qsql_filename + self.existing_table_names = existing_table_names + +class NoTableInQsqlExcption(Exception): + + def __init__(self, qsql_filename): + self.qsql_filename = qsql_filename + +class TooManyTablesInSqliteException(Exception): + + def __init__(self, qsql_filename,existing_table_names): + self.qsql_filename = qsql_filename + self.existing_table_names = existing_table_names + +class NoTablesInSqliteException(Exception): + + def __init__(self, sqlite_filename): + self.sqlite_filename = sqlite_filename + +class ColumnMaxLengthLimitExceededException(Exception): + + def __init__(self, msg): + self.msg = msg + + def __str(self): + return repr(self.msg) + +class CouldNotParseInputException(Exception): + + def __init__(self, msg): + self.msg = msg + + def __str(self): + return repr(self.msg) + +class BadHeaderException(Exception): + + def __init__(self, msg): + self.msg = msg + + def __str(self): + return repr(self.msg) + +class EncodedQueryException(Exception): + + def __init__(self, msg): + self.msg = msg + + def __str(self): + return repr(self.msg) + + +class CannotUnzipDataStreamException(Exception): + + def __init__(self): + pass + +class UniversalNewlinesExistException(Exception): + + def __init__(self): + pass + +class EmptyDataException(Exception): + + def __init__(self): + pass + +class MissingHeaderException(Exception): + + def __init__(self,msg): + self.msg = msg + +class InvalidQueryException(Exception): + + def __init__(self,msg): + self.msg = msg + +class TooManyAttachedDatabasesException(Exception): + + def __init__(self,msg): + self.msg = msg + +class FileNotFoundException(Exception): + + def __init__(self, msg): + self.msg = msg + + def __str(self): + return repr(self.msg) + +class UnknownFileTypeException(Exception): + + def __init__(self, msg): + self.msg = msg + + def __str(self): + return repr(self.msg) + + +class ColumnCountMismatchException(Exception): + + def __init__(self, msg): + self.msg = msg + +class ContentSignatureNotFoundException(Exception): + + def __init__(self, msg): + self.msg = msg + +class StrictModeColumnCountMismatchException(Exception): + + def __init__(self,atomic_fn, expected_col_count,actual_col_count,lines_read): + self.atomic_fn = atomic_fn + self.expected_col_count = expected_col_count + self.actual_col_count = actual_col_count + self.lines_read = lines_read + +class FluffyModeColumnCountMismatchException(Exception): + + def __init__(self,atomic_fn, expected_col_count,actual_col_count,lines_read): + self.atomic_fn = atomic_fn + self.expected_col_count = expected_col_count + self.actual_col_count = actual_col_count + self.lines_read = lines_read + +class ContentSignatureDiffersException(Exception): + + def __init__(self,original_filename, other_filename, filenames_str,key,source_value,signature_value): + self.original_filename = original_filename + self.other_filename = other_filename + self.filenames_str = filenames_str + self.key = key + self.source_value = source_value + self.signature_value = signature_value + + +class ContentSignatureDataDiffersException(Exception): + + def __init__(self,msg): + self.msg = msg + + +class InvalidQSqliteFileException(Exception): + + def __init__(self,msg): + self.msg = msg + + +class MaximumSourceFilesExceededException(Exception): + + def __init__(self,msg): + self.msg = msg + + + +# Simplistic Sql "parsing" class... We'll eventually require a real SQL parser which will provide us with a parse tree +# +# A "qtable" is a filename which behaves like an SQL table... +class Sql(object): + + def __init__(self, sql, data_streams): + # Currently supports only standard SELECT statements + + # Holds original SQL + self.sql = sql + # Holds sql parts + self.sql_parts = sql.split() + self.data_streams = data_streams + + self.qtable_metadata_dict = OrderedDict() + + # Set of qtable names + self.qtable_names = [] + # Dict from qtable names to their positions in sql_parts. Value here is a *list* of positions, + # since it is possible that the same qtable_name (file) is referenced in multiple positions + # and we don't want the database table to be recreated for each + # reference + self.qtable_name_positions = {} + # Dict from qtable names to their effective (actual database) table + # names + self.qtable_name_effective_table_names = {} + + self.query_column_names = None + + # Go over all sql parts + idx = 0 + while idx < len(self.sql_parts): + # Get the part string + part = self.sql_parts[idx] + # If it's a FROM or a JOIN + if part.upper() in ['FROM', 'JOIN']: + # and there is nothing after it, + if idx == len(self.sql_parts) - 1: + # Just fail + raise InvalidQueryException( + 'FROM/JOIN is missing a table name after it') + + qtable_name = self.sql_parts[idx + 1] + # Otherwise, the next part contains the qtable name. In most cases the next part will be only the qtable name. + # We handle one special case here, where this is a subquery as a column: "SELECT (SELECT ... FROM qtable),100 FROM ...". + # In that case, there will be an ending paranthesis as part of the name, and we want to handle this case gracefully. + # This is obviously a hack of a hack :) Just until we have + # complete parsing capabilities + if ')' in qtable_name: + leftover = qtable_name[qtable_name.index(')'):] + self.sql_parts.insert(idx + 2, leftover) + qtable_name = qtable_name[:qtable_name.index(')')] + self.sql_parts[idx + 1] = qtable_name + + if qtable_name[0] != '(': + normalized_qtable_name = self.normalize_qtable_name(qtable_name) + xprint("Normalized qtable name for %s is %s" % (qtable_name,normalized_qtable_name)) + self.qtable_names += [normalized_qtable_name] + + if normalized_qtable_name not in self.qtable_name_positions.keys(): + self.qtable_name_positions[normalized_qtable_name] = [] + + self.qtable_name_positions[normalized_qtable_name].append(idx + 1) + self.sql_parts[idx + 1] = normalized_qtable_name + idx += 2 + else: + idx += 1 + else: + idx += 1 + xprint("Final sql parts: %s" % self.sql_parts) + + def normalize_qtable_name(self,qtable_name): + if self.data_streams.is_data_stream(qtable_name): + return qtable_name + + if ':::' in qtable_name: + qsql_filename, table_name = qtable_name.split(":::", 1) + return '%s:::%s' % (os.path.realpath(os.path.abspath(qsql_filename)),table_name) + else: + return os.path.realpath(os.path.abspath(qtable_name)) + + def set_effective_table_name(self, qtable_name, effective_table_name): + if qtable_name in self.qtable_name_effective_table_names.keys(): + if self.qtable_name_effective_table_names[qtable_name] != effective_table_name: + raise Exception( + "Already set effective table name for qtable %s. Trying to change the effective table name from %s to %s" % + (qtable_name,self.qtable_name_effective_table_names[qtable_name],effective_table_name)) + + xprint("Setting effective table name for %s - effective table name is set to %s" % (qtable_name,effective_table_name)) + self.qtable_name_effective_table_names[ + qtable_name] = effective_table_name + + def get_effective_sql(self,table_name_mapping=None): + if len(list(filter(lambda x: x is None, self.qtable_name_effective_table_names))) != 0: + assert False, 'There are qtables without effective tables' + + effective_sql = [x for x in self.sql_parts] + + xprint("Effective table names",self.qtable_name_effective_table_names) + for qtable_name, positions in six.iteritems(self.qtable_name_positions): + xprint("Positions for qtable name %s are %s" % (qtable_name,positions)) + for pos in positions: + if table_name_mapping is not None: + x = self.qtable_name_effective_table_names[qtable_name] + effective_sql[pos] = table_name_mapping[x] + else: + effective_sql[pos] = self.qtable_name_effective_table_names[qtable_name] + + return " ".join(effective_sql) + + def get_qtable_name_effective_table_names(self): + return self.qtable_name_effective_table_names + + def execute_and_fetch(self, db): + x = self.get_effective_sql() + xprint("Final query: %s" % x) + db_results_obj = db.execute_and_fetch(x) + return db_results_obj + + def materialize_using(self,loaded_table_structures_dict): + xprint("Materializing sql object: %s" % str(self.qtable_names)) + xprint("loaded table structures dict %s" % loaded_table_structures_dict) + for qtable_name in self.qtable_names: + table_structure = loaded_table_structures_dict[qtable_name] + + table_name_in_disk_db = table_structure.get_table_name_for_querying() + + effective_table_name = '%s.%s' % (table_structure.db_id, table_name_in_disk_db) + + # for a single file - no need to create a union, just use the table name + self.set_effective_table_name(qtable_name, effective_table_name) + xprint("Materialized filename %s to effective table name %s" % (qtable_name,effective_table_name)) + + +class TableColumnInferer(object): + + def __init__(self, input_params): + self.inferred = False + self.mode = input_params.parsing_mode + self.rows = [] + self.skip_header = input_params.skip_header + self.header_row = None + self.header_row_filename = None + self.expected_column_count = input_params.expected_column_count + self.input_delimiter = input_params.delimiter + self.disable_column_type_detection = input_params.disable_column_type_detection + + def _generate_content_signature(self): + return OrderedDict({ + "inferred": self.inferred, + "mode": self.mode, + "rows": "\n".join([",".join(x) for x in self.rows]), + "skip_header": self.skip_header, + "header_row": self.header_row, + "expected_column_count": self.expected_column_count, + "input_delimiter": self.input_delimiter, + "disable_column_type_detection": self.disable_column_type_detection + }) + + def analyze(self, filename, col_vals): + if self.inferred: + assert False, "Already inferred columns" + + if self.skip_header and self.header_row is None: + self.header_row = col_vals + self.header_row_filename = filename + else: + self.rows.append(col_vals) + + if len(self.rows) < 100: + return False + + self.do_analysis() + return True + + def force_analysis(self): + # This method is called whenever there is no more data, and an analysis needs + # to be performed immediately, regardless of the amount of sample data that has + # been collected + self.do_analysis() + + def determine_type_of_value(self, value): + if self.disable_column_type_detection: + return str + + if value is not None: + value = value.strip() + if value == '' or value is None: + return None + + try: + i = int(value) + if type(i) == long: + return long + else: + return int + except: + pass + + try: + f = float(value) + return float + except: + pass + + return str + + def determine_type_of_value_list(self, value_list): + type_list = [self.determine_type_of_value(v) for v in value_list] + all_types = set(type_list) + if len(set(type_list)) == 1: + # all the sample lines are of the same type + return type_list[0] + else: + # check for the number of types without nulls, + type_list_without_nulls = list(filter( + lambda x: x is not None, type_list)) + # If all the sample lines are of the same type, + if len(set(type_list_without_nulls)) == 1: + # return it + return type_list_without_nulls[0] + else: + # If there are only two types, one float an one int, then choose a float type + if len(set(type_list_without_nulls)) == 2 and float in type_list_without_nulls and int in type_list_without_nulls: + return float + return str + + def do_analysis(self): + if self.mode == 'strict': + self._do_strict_analysis() + elif self.mode in ['relaxed']: + self._do_relaxed_analysis() + else: + raise Exception('Unknown parsing mode %s' % self.mode) + + if self.column_count == 1 and self.expected_column_count != 1 and self.expected_column_count is not None: + print(f"Warning: column count is one (expected column count is {self.expected_column_count} - did you provide the correct delimiter?", file=sys.stderr) + + self.infer_column_types() + self.infer_column_names() + self.inferred = True + + def validate_column_names(self, value_list): + column_name_errors = [] + for v in value_list: + if v is None: + # we allow column names to be None, in relaxed mode it'll be filled with default names. + # RLRL + continue + if ',' in v: + column_name_errors.append( + (v, "Column name cannot contain commas")) + continue + if self.input_delimiter in v: + column_name_errors.append( + (v, "Column name cannot contain the input delimiter. Please make sure you've set the correct delimiter")) + continue + if '\n' in v: + column_name_errors.append( + (v, "Column name cannot contain newline")) + continue + if v != v.strip(): + column_name_errors.append( + (v, "Column name contains leading/trailing spaces")) + continue + try: + v.encode("utf-8", "strict").decode("utf-8") + except: + column_name_errors.append( + (v, "Column name must be UTF-8 Compatible")) + continue + # We're checking for column duplication for each field in order to be able to still provide it along with other errors + if len(list(filter(lambda x: x == v,value_list))) > 1: + entry = (v, "Column name is duplicated") + # Don't duplicate the error report itself + if entry not in column_name_errors: + column_name_errors.append(entry) + continue + nul_index = v.find("\x00") + if nul_index >= 0: + column_name_errors.append( + (v, "Column name cannot contain NUL")) + continue + t = self.determine_type_of_value(v) + if t != str: + column_name_errors.append((v, "Column name must be a string")) + return column_name_errors + + def infer_column_names(self): + if self.header_row is not None: + column_name_errors = self.validate_column_names(self.header_row) + if len(column_name_errors) > 0: + raise BadHeaderException("Header must contain only strings and not numbers or empty strings: '%s'\n%s" % ( + ",".join(self.header_row), "\n".join(["'%s': %s" % (x, y) for x, y in column_name_errors]))) + + # use header row in order to name columns + if len(self.header_row) < self.column_count: + if self.mode == 'strict': + raise ColumnCountMismatchException("Strict mode. Header row contains less columns than expected column count(%s vs %s)" % ( + len(self.header_row), self.column_count)) + elif self.mode in ['relaxed']: + # in relaxed mode, add columns to fill the missing ones + self.header_row = self.header_row + \ + ['c%s' % (x + len(self.header_row) + 1) + for x in range(self.column_count - len(self.header_row))] + elif len(self.header_row) > self.column_count: + if self.mode == 'strict': + raise ColumnCountMismatchException("Strict mode. Header row contains more columns than expected column count (%s vs %s)" % ( + len(self.header_row), self.column_count)) + elif self.mode in ['relaxed']: + # In relaxed mode, just cut the extra column names + self.header_row = self.header_row[:self.column_count] + self.column_names = self.header_row + else: + # Column names are cX starting from 1 + self.column_names = ['c%s' % (i + 1) + for i in range(self.column_count)] + + def _do_relaxed_analysis(self): + column_count_list = [len(col_vals) for col_vals in self.rows] + + if len(self.rows) == 0: + if self.header_row is None: + self.column_count = 0 + else: + self.column_count = len(self.header_row) + else: + if self.expected_column_count is not None: + self.column_count = self.expected_column_count + else: + # If not specified, we'll take the largest row in the sample rows + self.column_count = max(column_count_list) + + def get_column_count_summary(self, column_count_list): + counts = {} + for column_count in column_count_list: + counts[column_count] = counts.get(column_count, 0) + 1 + return six.u(", ").join([six.u("{} rows with {} columns".format(v, k)) for k, v in six.iteritems(counts)]) + + def _do_strict_analysis(self): + column_count_list = [len(col_vals) for col_vals in self.rows] + + if len(set(column_count_list)) != 1: + raise ColumnCountMismatchException('Strict mode. Column Count is expected to identical. Multiple column counts exist at the first part of the file. Try to check your delimiter, or change to relaxed mode. Details: %s' % ( + self.get_column_count_summary(column_count_list))) + + self.column_count = len(self.rows[0]) + + if self.expected_column_count is not None and self.column_count != self.expected_column_count: + raise ColumnCountMismatchException('Strict mode. Column count is expected to be %s but is %s' % ( + self.expected_column_count, self.column_count)) + + self.infer_column_types() + + def infer_column_types(self): + assert self.column_count > -1 + self.column_types = [] + self.column_types2 = [] + for column_number in range(self.column_count): + column_value_list = [ + row[column_number] if column_number < len(row) else None for row in self.rows] + column_type = self.determine_type_of_value_list(column_value_list) + self.column_types.append(column_type) + + column_value_list2 = [row[column_number] if column_number < len( + row) else None for row in self.rows[1:]] + column_type2 = self.determine_type_of_value_list( + column_value_list2) + self.column_types2.append(column_type2) + + comparison = map( + lambda x: x[0] == x[1], zip(self.column_types, self.column_types2)) + if False in comparison and not self.skip_header: + number_of_column_types = len(set(self.column_types)) + if number_of_column_types == 1 and list(set(self.column_types))[0] == str: + print('Warning - There seems to be header line in the file, but -H has not been specified. All fields will be detected as text fields, and the header line will appear as part of the data', file=sys.stderr) + + def get_column_dict(self): + return OrderedDict(zip(self.column_names, self.column_types)) + + def get_column_count(self): + return self.column_count + + def get_column_names(self): + return self.column_names + + def get_column_types(self): + return self.column_types + + +def py3_encoded_csv_reader(encoding, f, dialect,row_data_only=False,**kwargs): + try: + xprint("f is %s" % str(f)) + xprint("dialect is %s" % dialect) + csv_reader = csv.reader(f, dialect, **kwargs) + + if row_data_only: + for row in csv_reader: + yield row + else: + for row in csv_reader: + yield (f.filename(),f.isfirstline(),row) + + except UnicodeDecodeError as e1: + raise CouldNotParseInputException(e1) + except ValueError as e: + # TODO Add test for this + if str(e) is not None and str(e).startswith('could not convert string to'): + raise CouldNotConvertStringToNumericValueException(str(e)) + else: + raise CouldNotParseInputException(str(e)) + except Exception as e: + if str(e).startswith("field larger than field limit"): + raise ColumnMaxLengthLimitExceededException(str(e)) + elif 'universal-newline' in str(e): + raise UniversalNewlinesExistException() + else: + raise + +encoded_csv_reader = py3_encoded_csv_reader + +def normalized_filename(filename): + return filename + +class TableCreatorState(object): + INITIALIZED = 'INITIALIZED' + ANALYZED = 'ANALYZED' + FULLY_READ = 'FULLY_READ' + +class MaterializedStateType(object): + UNKNOWN = 'unknown' + DELIMITED_FILE = 'delimited-file' + QSQL_FILE = 'qsql-file' + SQLITE_FILE = 'sqlite-file' + DATA_STREAM = 'data-stream' + +class TableSourceType(object): + DELIMITED_FILE = 'file' + DELIMITED_FILE_WITH_UNUSED_QSQL = 'file-with-unused-qsql' + QSQL_FILE = 'qsql-file' + QSQL_FILE_WITH_ORIGINAL = 'qsql-file-with-original' + SQLITE_FILE = 'sqlite-file' + DATA_STREAM = 'data-stream' + +def skip_BOM(f): + try: + BOM = f.buffer.read(3) + + if BOM != six.b('\xef\xbb\xbf'): + # TODO Add test for this (propagates to try:except) + raise Exception('Value of BOM is not as expected - Value is "%s"' % str(BOM)) + except Exception as e: + # TODO Add a test for this + raise Exception('Tried to skip BOM for "utf-8-sig" encoding and failed. Error message is ' + str(e)) + +def detect_qtable_name_source_info(qtable_name,data_streams,read_caching_enabled): + data_stream = data_streams.get_for_filename(qtable_name) + xprint("Found data stream %s" % data_stream) + + if data_stream is not None: + return MaterializedStateType.DATA_STREAM, TableSourceType.DATA_STREAM,(data_stream,) + + if ':::' in qtable_name: + qsql_filename, table_name = qtable_name.split(":::", 1) + if not os.path.exists(qsql_filename): + raise FileNotFoundException("Could not find file %s" % qsql_filename) + + if is_qsql_file(qsql_filename): + return MaterializedStateType.QSQL_FILE, TableSourceType.QSQL_FILE, (qsql_filename, table_name,) + if is_sqlite_file(qsql_filename): + return MaterializedStateType.SQLITE_FILE, TableSourceType.SQLITE_FILE, (qsql_filename, table_name,) + raise UnknownFileTypeException("Cannot detect the type of table %s" % qtable_name) + else: + if is_qsql_file(qtable_name): + return MaterializedStateType.QSQL_FILE, TableSourceType.QSQL_FILE, (qtable_name, None) + if is_sqlite_file(qtable_name): + return MaterializedStateType.SQLITE_FILE, TableSourceType.SQLITE_FILE, (qtable_name, None) + matching_qsql_file_candidate = qtable_name + '.qsql' + + table_source_type = TableSourceType.DELIMITED_FILE + if is_qsql_file(matching_qsql_file_candidate): + if read_caching_enabled: + xprint("Found matching qsql file for original file %s (matching file %s) and read caching is enabled. Using it" % (qtable_name,matching_qsql_file_candidate)) + return MaterializedStateType.QSQL_FILE, TableSourceType.QSQL_FILE_WITH_ORIGINAL, (matching_qsql_file_candidate, None) + else: + xprint("Found matching qsql file for original file %s (matching file %s), but read caching is disabled. Not using it" % (qtable_name,matching_qsql_file_candidate)) + table_source_type = TableSourceType.DELIMITED_FILE_WITH_UNUSED_QSQL + + + return MaterializedStateType.DELIMITED_FILE,table_source_type ,(qtable_name, None) + + +def is_sqlite_file(filename): + if not os.path.exists(filename): + return False + + f = open(filename,'rb') + magic = f.read(16) + f.close() + return magic == six.b("SQLite format 3\x00") + +def sqlite_table_exists(cursor,table_name): + results = cursor.execute("select count(*) from sqlite_master where type='table' and tbl_name == '%s'" % table_name).fetchall() + return results[0][0] == 1 + +def is_qsql_file(filename): + if not is_sqlite_file(filename): + return False + + db = Sqlite3DB('check_qsql_db',filename,filename,create_qcatalog=False) + qcatalog_exists = db.qcatalog_table_exists() + db.done() + return qcatalog_exists + +def normalize_filename_to_table_name(filename): + xprint("Normalizing filename %s" % filename) + if filename[0].isdigit(): + xprint("Filename starts with a digit, adding prefix") + filename = 't_%s' % filename + if filename.lower().endswith(".qsql"): + filename = filename[:-5] + elif filename.lower().endswith('.sqlite'): + filename = filename[:-7] + elif filename.lower().endswith('.sqlite3'): + filename = filename[:-8] + return filename.replace("-","_dash_").replace(".","_dot_").replace('?','_qm_').replace("/","_slash_").replace("\\","_backslash_").replace(":","_colon_").replace(" ","_space_").replace("+","_plus_") + +def validate_content_signature(original_filename, source_signature,other_filename, content_signature,scope=None,dump=False): + if dump: + xprint("Comparing: source value: %s target value: %s" % (source_signature,content_signature)) + + s = "%s vs %s:" % (original_filename,other_filename) + if scope is None: + scope = [] + for k in source_signature: + if type(source_signature[k]) == OrderedDict: + validate_content_signature(original_filename, source_signature[k],other_filename, content_signature[k],scope + [k]) + else: + if k not in content_signature: + raise ContentSignatureDataDiffersException("%s Content Signatures differ. %s is missing from content signature" % (s,k)) + if source_signature[k] != content_signature[k]: + if k == 'rows': + raise ContentSignatureDataDiffersException("%s Content Signatures differ at %s.%s (actual analysis data differs)" % (s,".".join(scope),k)) + else: + raise ContentSignatureDiffersException(original_filename, other_filename, original_filename,".".join(scope + [k]),source_signature[k],content_signature[k]) + +class DelimitedFileReader(object): + def __init__(self,atomic_fns, input_params, dialect, f = None,external_f_name = None): + if f is not None: + assert len(atomic_fns) == 0 + + self.atomic_fns = atomic_fns + self.input_params = input_params + self.dialect = dialect + + self.f = f + self.lines_read = 0 + self.file_number = -1 + + self.skipped_bom = False + + self.is_open = f is not None + + self.external_f = f is not None + self.external_f_name = external_f_name + + def get_lines_read(self): + return self.lines_read + + def get_size_hash(self): + if self.atomic_fns is None or len(self.atomic_fns) == 0: + return "data-stream-size" + else: + return ",".join(map(str,[os.stat(atomic_fn).st_size for atomic_fn in self.atomic_fns])) + + def get_last_modification_time_hash(self): + if self.atomic_fns is None or len(self.atomic_fns) == 0: + return "data stream-lmt" + else: + x = ",".join(map(lambda x: ':%s:' % x,[os.stat(x).st_mtime_ns for x in self.atomic_fns])) + res = hashlib.sha1(six.b(x)).hexdigest() + '///' + x + xprint("Hash of last modification time is %s" % res) + return res + + def open_file(self): + if self.external_f: + xprint("External f has been provided. No need to open the file") + return + + # TODO Support universal newlines for gzipped and stdin data as well + + xprint("XX Opening file %s" % ",".join(self.atomic_fns)) + import fileinput + + def q_openhook(filename, mode): + if self.input_params.gzipped_input or filename.endswith('.gz'): + import gzip + f = gzip.open(filename,mode='rt',encoding=self.input_params.input_encoding) + else: + if six.PY3: + if self.input_params.with_universal_newlines: + f = io.open(filename, 'rU', newline=None, encoding=self.input_params.input_encoding) + else: + f = io.open(filename, 'r', newline=None, encoding=self.input_params.input_encoding) + else: + if self.input_params.with_universal_newlines: + file_opening_mode = 'rbU' + else: + file_opening_mode = 'rb' + f = open(filename, file_opening_mode) + + if self.input_params.input_encoding == 'utf-8-sig' and not self.skipped_bom: + skip_BOM(f) + + return f + + f = fileinput.input(self.atomic_fns,mode='rb',openhook=q_openhook) + + self.f = f + self.is_open = True + xprint("Actually opened file %s" % self.f) + return f + + def close_file(self): + if not self.is_open: + # TODO Convert to assertion + raise Exception("Bug - file should already be open: %s" % ",".join(self.atomic_fns)) + + self.f.close() + xprint("XX Closed file %s" % ",".join(self.atomic_fns)) + + def generate_rows(self): + csv_reader = encoded_csv_reader(self.input_params.input_encoding, self.f, dialect=self.dialect,row_data_only=self.external_f) + try: + # TODO Some order with regard to separating data-streams for actual files + if self.external_f: + for col_vals in csv_reader: + self.lines_read += 1 + yield self.external_f_name,0, self.lines_read == 0, col_vals + else: + for file_name,is_first_line,col_vals in csv_reader: + if is_first_line: + self.file_number = self.file_number + 1 + self.lines_read += 1 + yield file_name,self.file_number,is_first_line,col_vals + except ColumnMaxLengthLimitExceededException as e: + msg = "Column length is larger than the maximum. Offending file is '%s' - Line is %s, counting from 1 (encoding %s). The line number is the raw line number of the file, ignoring whether there's a header or not" % (",".join(self.atomic_fns),self.lines_read + 1,self.input_params.input_encoding) + raise ColumnMaxLengthLimitExceededException(msg) + except UniversalNewlinesExistException as e2: + # No need to translate the exception, but we want it to be explicitly defined here for clarity + raise UniversalNewlinesExistException() + +class MaterializedState(object): + def __init__(self, table_source_type,qtable_name, engine_id): + xprint("Creating new MS: %s %s" % (id(self), qtable_name)) + + self.table_source_type = table_source_type + + self.qtable_name = qtable_name + self.engine_id = engine_id + + self.db_to_use = None + self.db_id = None + + self.source_type = None + self.source = None + + self.mfs_structure = None + + self.start_time = None + self.end_time = None + self.duration = None + + self.effective_table_name = None + + + def get_materialized_state_type(self): + return MaterializedStateType.UNKNOWN + + def get_planned_table_name(self): + assert False, 'not implemented' + + def autodetect_table_name(self): + xprint("Autodetecting table name. db_to_use=%s" % self.db_to_use) + existing_table_names = self.db_to_use.retrieve_all_table_names() + xprint("Existing table names: %s" % existing_table_names) + + possible_indices = range(1,1000) + + for index in possible_indices: + if index == 1: + suffix = '' + else: + suffix = '_%s' % index + + table_name_attempt = '%s%s' % (self.get_planned_table_name(),suffix) + xprint("Table name attempt: index=%s name=%s" % (index,table_name_attempt)) + + if table_name_attempt not in existing_table_names: + xprint("Found free table name %s for source type %s source %s" % (table_name_attempt,self.source_type,self.source)) + return table_name_attempt + + raise Exception('Cannot find free table name for source type %s source %s' % (self.source_type,self.source)) + + def initialize(self): + self.start_time = time.time() + + def finalize(self): + self.end_time = time.time() + self.duration = self.end_time - self.start_time + + def choose_db_to_use(self,forced_db_to_use=None,stop_after_analysis=False): + assert False, 'not implemented' + + def make_data_available(self,stop_after_analysis): + assert False, 'not implemented' + +class MaterializedDelimitedFileState(MaterializedState): + def __init__(self, table_source_type,qtable_name, input_params, dialect_id,engine_id,target_table_name=None): + super().__init__(table_source_type,qtable_name,engine_id) + + self.input_params = input_params + self.dialect_id = dialect_id + self.target_table_name = target_table_name + + self.content_signature = None + + self.atomic_fns = None + + self.can_store_as_cached = None + + def get_materialized_state_type(self): + return MaterializedStateType.DELIMITED_FILE + + def initialize(self): + super(MaterializedDelimitedFileState, self).initialize() + + self.atomic_fns = self.materialize_file_list(self.qtable_name) + self.delimited_file_reader = DelimitedFileReader(self.atomic_fns,self.input_params,self.dialect_id) + + self.source_type = self.table_source_type + self.source = ",".join(self.atomic_fns) + + return + + def materialize_file_list(self,qtable_name): + materialized_file_list = [] + + unfound_files = [] + # First check if the file exists without globbing. This will ensure that we don't support non-existent files + if os.path.exists(qtable_name): + # If it exists, then just use it + found_files = [qtable_name] + else: + # If not, then try with globs (and sort for predictability) + found_files = list(sorted(glob.glob(qtable_name))) + # If no files + if len(found_files) == 0: + unfound_files += [qtable_name] + materialized_file_list += found_files + + # If there are no files to go over, + if len(unfound_files) == 1: + raise FileNotFoundException( + "No files matching '%s' have been found" % unfound_files[0]) + elif len(unfound_files) > 1: + # TODO Add test for this + raise FileNotFoundException( + "The following files have not been found for table %s: %s" % (qtable_name,",".join(unfound_files))) + + # deduplicate with matching qsql files + filtered_file_list = list(filter(lambda x: not x.endswith('.qsql'),materialized_file_list)) + xprint("Filtered qsql files from glob search. Original file count: %s new file count: %s" % (len(materialized_file_list),len(filtered_file_list))) + + l = len(filtered_file_list) + # If this proves to be a problem for users in terms of usability, then we'll just materialize the files + # into the adhoc db, as with the db attach limit of sqlite + if l > 500: + msg = "Maximum source files for table must be 500. Table is name is %s Number of actual files is %s" % (qtable_name,l) + raise MaximumSourceFilesExceededException(msg) + + absolute_path_list = [os.path.abspath(x) for x in filtered_file_list] + return absolute_path_list + + def choose_db_to_use(self,forced_db_to_use=None,stop_after_analysis=False): + if forced_db_to_use is not None: + self.db_id = forced_db_to_use.db_id + self.db_to_use = forced_db_to_use + self.can_store_as_cached = False + assert self.target_table_name is None + self.target_table_name = self.autodetect_table_name() + return + + self.can_store_as_cached = True + + self.db_id = '%s' % self._generate_db_name(self.atomic_fns[0]) + xprint("Database id is %s" % self.db_id) + self.db_to_use = Sqlite3DB(self.db_id, 'file:%s?mode=memory&cache=shared' % self.db_id, 'memory<%s>' % self.db_id,create_qcatalog=True) + + if self.target_table_name is None: + self.target_table_name = self.autodetect_table_name() + + + def __analyze_delimited_file(self,database_info): + xprint("Analyzing delimited file") + if self.target_table_name is not None: + target_sqlite_table_name = self.target_table_name + else: + assert False + + xprint("Target sqlite table name is %s" % target_sqlite_table_name) + # Create the matching database table and populate it + table_creator = TableCreator(self.qtable_name, self.delimited_file_reader,self.input_params, sqlite_db=database_info.sqlite_db, + target_sqlite_table_name=target_sqlite_table_name) + table_creator.perform_analyze(self.dialect_id) + xprint("after perform_analyze") + self.content_signature = table_creator._generate_content_signature() + + now = datetime.datetime.utcnow().isoformat() + + database_info.sqlite_db.add_to_qcatalog_table(target_sqlite_table_name, + self.content_signature, + now, + self.source_type, + self.source) + return table_creator + + def _generate_disk_db_filename(self, filenames_str): + fn = '%s.qsql' % (os.path.abspath(filenames_str).replace("+","__")) + return fn + + + def _get_should_read_from_cache(self, disk_db_filename): + disk_db_file_exists = os.path.exists(disk_db_filename) + + should_read_from_cache = self.input_params.read_caching and disk_db_file_exists + + return should_read_from_cache + + def calculate_should_read_from_cache(self): + # TODO cache filename is chosen according to first filename only, which makes multi-file (glob) caching difficult + # cache writing is blocked for now in these cases. Will be added in the future (see save_cache_to_disk_if_needed) + disk_db_filename = self._generate_disk_db_filename(self.atomic_fns[0]) + should_read_from_cache = self._get_should_read_from_cache(disk_db_filename) + xprint("should read from cache %s" % should_read_from_cache) + return disk_db_filename,should_read_from_cache + + def get_planned_table_name(self): + return normalize_filename_to_table_name(os.path.basename(self.atomic_fns[0])) + + def make_data_available(self,stop_after_analysis): + xprint("In make_data_available. db_id %s db_to_use %s" % (self.db_id,self.db_to_use)) + assert self.db_id is not None + + disk_db_filename, should_read_from_cache = self.calculate_should_read_from_cache() + xprint("disk_db_filename=%s should_read_from_cache=%s" % (disk_db_filename,should_read_from_cache)) + + database_info = DatabaseInfo(self.db_id,self.db_to_use, needs_closing=True) + xprint("db %s (%s) has been added to the database list" % (self.db_id, self.db_to_use)) + + self.delimited_file_reader.open_file() + + table_creator = self.__analyze_delimited_file(database_info) + + self.mfs_structure = MaterializedStateTableStructure(self.qtable_name, self.atomic_fns, self.db_id, + table_creator.column_inferer.get_column_names(), + table_creator.column_inferer.get_column_types(), + None, + self.target_table_name, + self.source_type, + self.source, + self.get_planned_table_name()) + + content_signature = table_creator.content_signature + content_signature_key = self.db_to_use.calculate_content_signature_key(content_signature) + xprint("table creator signature key: %s" % content_signature_key) + + relevant_table = self.db_to_use.get_from_qcatalog(content_signature)['temp_table_name'] + + if not stop_after_analysis: + table_creator.perform_read_fully(self.dialect_id) + + self.save_cache_to_disk_if_needed(disk_db_filename, table_creator) + + + self.delimited_file_reader.close_file() + + return database_info, relevant_table + + def save_cache_to_disk_if_needed(self, disk_db_filename, table_creator): + if len(self.atomic_fns) > 1: + xprint("Cannot save cache for multi-files for now, deciding auto-naming for cache is challenging. Will be added in the future.") + return + + effective_write_caching = self.input_params.write_caching + if effective_write_caching: + if self.can_store_as_cached: + assert self.table_source_type != TableSourceType.DELIMITED_FILE_WITH_UNUSED_QSQL + xprint("Going to write file cache for %s. Disk filename is %s" % (",".join(self.atomic_fns), disk_db_filename)) + self._store_qsql(table_creator.sqlite_db, disk_db_filename) + else: + xprint("Database has been provided externally. Skipping storing a cached version of the data") + + def _store_qsql(self, source_sqlite_db, disk_db_filename): + xprint("Storing data as disk db") + disk_db_conn = sqlite3.connect(disk_db_filename) + with disk_db_conn: + source_sqlite_db.conn.backup(disk_db_conn) + xprint("Written db to disk: disk db filename %s" % (disk_db_filename)) + disk_db_conn.close() + + def _generate_db_name(self, qtable_name): + return 'e_%s_fn_%s' % (self.engine_id,normalize_filename_to_table_name(qtable_name)) + + +class MaterialiedDataStreamState(MaterializedDelimitedFileState): + def __init__(self, table_source_type, qtable_name, input_params, dialect_id, engine_id, data_stream, stream_target_db): ## should pass adhoc_db + assert data_stream is not None + + super().__init__(table_source_type, qtable_name, input_params, dialect_id, engine_id,target_table_name=None) + + self.data_stream = data_stream + + self.stream_target_db = stream_target_db + + self.target_table_name = None + + def get_planned_table_name(self): + return 'data_stream_%s' % (normalize_filename_to_table_name(self.source)) + + def get_materialized_state_type(self): + return MaterializedStateType.DATA_STREAM + + def initialize(self): + self.start_time = time.time() + if self.input_params.gzipped_input: + raise CannotUnzipDataStreamException() + + self.source_type = self.table_source_type + self.source = self.data_stream.stream_id + + self.delimited_file_reader = DelimitedFileReader([], self.input_params, self.dialect_id, f=self.data_stream.stream,external_f_name=self.source) + + def choose_db_to_use(self,forced_db_to_use=None,stop_after_analysis=False): + assert forced_db_to_use is None + + self.db_id = self.stream_target_db.db_id + self.db_to_use = self.stream_target_db + + self.target_table_name = self.autodetect_table_name() + + return + + def calculate_should_read_from_cache(self): + # No disk_db_filename, and no reading from cache when reading a datastream + return None, False + + def finalize(self): + super(MaterialiedDataStreamState, self).finalize() + + def save_cache_to_disk_if_needed(self, disk_db_filename, table_creator): + xprint("Saving to cache is disabled for data streams") + return + + +class MaterializedSqliteState(MaterializedState): + def __init__(self,table_source_type,qtable_name,sqlite_filename,table_name, engine_id): + super(MaterializedSqliteState, self).__init__(table_source_type,qtable_name,engine_id) + self.sqlite_filename = sqlite_filename + self.table_name = table_name + + self.table_name_autodetected = None + + def initialize(self): + super(MaterializedSqliteState, self).initialize() + + self.table_name_autodetected = False + if self.table_name is None: + self.table_name = self.autodetect_table_name() + self.table_name_autodetected = True + return + + self.validate_table_name() + + def get_planned_table_name(self): + if self.table_name_autodetected: + return normalize_filename_to_table_name(os.path.basename(self.qtable_name)) + else: + return self.table_name + + + def autodetect_table_name(self): + db = Sqlite3DB('temp_db','file:%s?immutable=1' % self.sqlite_filename,self.sqlite_filename,create_qcatalog=False) + try: + table_names = list(sorted(db.retrieve_all_table_names())) + if len(table_names) == 1: + return table_names[0] + elif len(table_names) == 0: + raise NoTablesInSqliteException(self.sqlite_filename) + else: + raise TooManyTablesInSqliteException(self.sqlite_filename,table_names) + finally: + db.done() + + def validate_table_name(self): + db = Sqlite3DB('temp_db', 'file:%s?immutable=1' % self.sqlite_filename, self.sqlite_filename, + create_qcatalog=False) + try: + table_names = list(db.retrieve_all_table_names()) + if self.table_name.lower() not in map(lambda x:x.lower(),table_names): + raise NonExistentTableNameInSqlite(self.sqlite_filename, self.table_name, table_names) + finally: + db.done() + + def finalize(self): + super(MaterializedSqliteState, self).finalize() + + def get_materialized_state_type(self): + return MaterializedStateType.SQLITE_FILE + + def _generate_qsql_only_db_name__temp(self, filenames_str): + return 'e_%s_fn_%s' % (self.engine_id,hashlib.sha1(six.b(filenames_str)).hexdigest()) + + def choose_db_to_use(self,forced_db_to_use=None,stop_after_analysis=False): + self.source = self.sqlite_filename + self.source_type = self.table_source_type + + self.db_id = '%s' % self._generate_qsql_only_db_name__temp(self.qtable_name) + + x = 'file:%s?immutable=1' % self.sqlite_filename + self.db_to_use = Sqlite3DB(self.db_id, x, self.sqlite_filename,create_qcatalog=False) + + if forced_db_to_use: + xprint("Forced sqlite db_to_use %s" % forced_db_to_use) + new_table_name = forced_db_to_use.attach_and_copy_table(self.db_to_use,self.table_name,stop_after_analysis) + self.table_name = new_table_name + self.db_id = forced_db_to_use.db_id + self.db_to_use = forced_db_to_use + + return + + def make_data_available(self,stop_after_analysis): + xprint("db %s (%s) has been added to the database list" % (self.db_id, self.db_to_use)) + + database_info,relevant_table = DatabaseInfo(self.db_id,self.db_to_use, needs_closing=True), self.table_name + + column_names, column_types, sqlite_column_types = self._extract_information() + + self.mfs_structure = MaterializedStateTableStructure(self.qtable_name, [self.qtable_name], self.db_id, + column_names, column_types, sqlite_column_types, + self.table_name, + self.source_type,self.source, + self.get_planned_table_name()) + return database_info, relevant_table + + def _extract_information(self): + table_list = self.db_to_use.retrieve_all_table_names() + if len(table_list) == 1: + table_name = table_list[0][0] + xprint("Only one table in sqlite database, choosing it: %s" % table_name) + else: + # self.table_name has either beein autodetected, or validated as an existing table up the stack + table_name = self.table_name + xprint("Multiple tables in sqlite file. Using provided table name %s" % self.table_name) + + table_info = self.db_to_use.get_sqlite_table_info(table_name) + xprint('Table info is %s' % table_info) + column_names = list(map(lambda x: x[1], table_info)) + sqlite_column_types = list(map(lambda x: x[2].lower(),table_info)) + column_types = list(map(lambda x: sqlite_type_to_python_type(x[2]), table_info)) + xprint("Column names and types for table %s: %s" % (table_name, list(zip(column_names, zip(sqlite_column_types,column_types))))) + self.content_signature = OrderedDict() + + return column_names, column_types, sqlite_column_types + + +class MaterializedQsqlState(MaterializedState): + def __init__(self,table_source_type,qtable_name,qsql_filename,table_name, engine_id,input_params,dialect_id): + super(MaterializedQsqlState, self).__init__(table_source_type,qtable_name,engine_id) + self.qsql_filename = qsql_filename + self.table_name = table_name + + # These are for cases where the qsql file is just a cache and the original is still there, used for content + # validation + self.input_params = input_params + self.dialect_id = dialect_id + + self.table_name_autodetected = None + + def initialize(self): + super(MaterializedQsqlState, self).initialize() + + self.table_name_autodetected = False + if self.table_name is None: + self.table_name = self.autodetect_table_name() + self.table_name_autodetected = True + return + + self.validate_table_name() + + def get_planned_table_name(self): + if self.table_name_autodetected: + return normalize_filename_to_table_name(os.path.basename(self.qtable_name)) + else: + return self.table_name + + + def autodetect_table_name(self): + db = Sqlite3DB('temp_db','file:%s?immutable=1' % self.qsql_filename,self.qsql_filename,create_qcatalog=False) + assert db.qcatalog_table_exists() + try: + qcatalog_entries = db.get_all_from_qcatalog() + if len(qcatalog_entries) == 0: + raise NoTableInQsqlExcption(self.qsql_filename) + elif len(qcatalog_entries) == 1: + return qcatalog_entries[0]['temp_table_name'] + else: + # TODO Add a test for this + table_names = list(sorted([x['temp_table_name'] for x in qcatalog_entries])) + raise TooManyTablesInQsqlException(self.qsql_filename,table_names) + finally: + db.done() + + def validate_table_name(self): + db = Sqlite3DB('temp_db', 'file:%s?immutable=1' % self.qsql_filename, self.qsql_filename, + create_qcatalog=False) + assert db.qcatalog_table_exists() + try: + entry = db.get_from_qcatalog_using_table_name(self.table_name) + if entry is None: + qcatalog_entries = db.get_all_from_qcatalog() + table_names = list(sorted([x['temp_table_name'] for x in qcatalog_entries])) + raise NonExistentTableNameInQsql(self.qsql_filename,self.table_name,table_names) + finally: + db.done() + + def finalize(self): + super(MaterializedQsqlState, self).finalize() + + def get_materialized_state_type(self): + return MaterializedStateType.QSQL_FILE + + def _generate_qsql_only_db_name__temp(self, filenames_str): + return 'e_%s_fn_%s' % (self.engine_id,hashlib.sha1(six.b(filenames_str)).hexdigest()) + + def choose_db_to_use(self,forced_db_to_use=None,stop_after_analysis=False): + self.source = self.qsql_filename + self.source_type = self.table_source_type + + self.db_id = '%s' % self._generate_qsql_only_db_name__temp(self.qtable_name) + + x = 'file:%s?immutable=1' % self.qsql_filename + self.db_to_use = Sqlite3DB(self.db_id, x, self.qsql_filename,create_qcatalog=False) + + if forced_db_to_use: + xprint("Forced qsql to use forced_db: %s" % forced_db_to_use) + + # TODO RLRL Move query to Sqlite3DB + all_table_names = [(x[0],x[1]) for x in self.db_to_use.execute_and_fetch("select content_signature_key,temp_table_name from %s" % self.db_to_use.QCATALOG_TABLE_NAME).results] + csk,t = list(filter(lambda x: x[1] == self.table_name,all_table_names))[0] + xprint("Copying table %s from db_id %s" % (t,self.db_id)) + d = self.db_to_use.get_from_qcatalog_using_table_name(t) + + new_table_name = forced_db_to_use.attach_and_copy_table(self.db_to_use,self.table_name,stop_after_analysis) + + xprint("CS",d['content_signature']) + cs = OrderedDict(json.loads(d['content_signature'])) + forced_db_to_use.add_to_qcatalog_table(new_table_name, cs, d['creation_time'], + d['source_type'], d['source']) + + self.table_name = new_table_name + self.db_id = forced_db_to_use.db_id + self.db_to_use = forced_db_to_use + + return + + def make_data_available(self,stop_after_analysis): + xprint("db %s (%s) has been added to the database list" % (self.db_id, self.db_to_use)) + + database_info,relevant_table = self._read_table_from_cache(stop_after_analysis) + + column_names, column_types, sqlite_column_types = self._extract_information() + + self.mfs_structure = MaterializedStateTableStructure(self.qtable_name, [self.qtable_name], self.db_id, + column_names, column_types, sqlite_column_types, + self.table_name, + self.source_type,self.source, + self.get_planned_table_name()) + return database_info, relevant_table + + def _extract_information(self): + assert self.db_to_use.qcatalog_table_exists() + table_info = self.db_to_use.get_sqlite_table_info(self.table_name) + xprint('table_name=%s Table info is %s' % (self.table_name,table_info)) + + x = self.db_to_use.get_from_qcatalog_using_table_name(self.table_name) + + column_names = list(map(lambda x: x[1], table_info)) + sqlite_column_types = list(map(lambda x: x[2].lower(),table_info)) + column_types = list(map(lambda x: sqlite_type_to_python_type(x[2]), table_info)) + self.content_signature = OrderedDict( + **json.loads(x['content_signature'])) + xprint('Inferred column names and types from qsql: %s' % list(zip(column_names, zip(sqlite_column_types,column_types)))) + + return column_names, column_types, sqlite_column_types + + def _backing_original_file_exists(self): + return '%s.qsql' % self.qtable_name == self.qsql_filename + + def _read_table_from_cache(self, stop_after_analysis): + if self._backing_original_file_exists(): + xprint("Found a matching source file for qsql file with qtable name %s. Checking content signature by creating a temp MFDS + analysis" % self.qtable_name) + mdfs = MaterializedDelimitedFileState(TableSourceType.DELIMITED_FILE,self.qtable_name,self.input_params,self.dialect_id,self.engine_id,target_table_name=None) + mdfs.initialize() + mdfs.choose_db_to_use(forced_db_to_use=None,stop_after_analysis=stop_after_analysis) + _,_ = mdfs.make_data_available(stop_after_analysis=True) + + original_file_content_signature = mdfs.content_signature + original_file_content_signature_key = self.db_to_use.calculate_content_signature_key(original_file_content_signature) + + qcatalog_entry = self.db_to_use.get_from_qcatalog_using_table_name(self.table_name) + + if qcatalog_entry is None: + raise Exception('missing content signature!') + + xprint("Actual Signature Key: %s Expected Signature Key: %s" % (qcatalog_entry['content_signature_key'],original_file_content_signature_key)) + actual_content_signature = json.loads(qcatalog_entry['content_signature']) + + xprint("Validating content signatures: original %s vs qsql %s" % (original_file_content_signature,actual_content_signature)) + validate_content_signature(self.qtable_name, original_file_content_signature, self.qsql_filename, actual_content_signature,dump=True) + mdfs.finalize() + return DatabaseInfo(self.db_id,self.db_to_use, needs_closing=True), self.table_name + + +class MaterializedStateTableStructure(object): + def __init__(self,qtable_name, atomic_fns, db_id, column_names, python_column_types, sqlite_column_types, table_name_for_querying,source_type,source,planned_table_name): + self.qtable_name = qtable_name + self.atomic_fns = atomic_fns + self.db_id = db_id + self.column_names = column_names + self.python_column_types = python_column_types + self.table_name_for_querying = table_name_for_querying + self.source_type = source_type + self.source = source + self.planned_table_name = planned_table_name + + if sqlite_column_types is not None: + self.sqlite_column_types = sqlite_column_types + else: + self.sqlite_column_types = [Sqlite3DB.PYTHON_TO_SQLITE_TYPE_NAMES[t].lower() for t in python_column_types] + + def get_table_name_for_querying(self): + return self.table_name_for_querying + + def __str__(self): + return "MaterializedStateTableStructure<%s>" % self.__dict__ + __repr__ = __str__ + +class TableCreator(object): + def __str__(self): + return "TableCreator<%s>" % str(self) + __repr__ = __str__ + + def __init__(self, qtable_name, delimited_file_reader,input_params,sqlite_db=None,target_sqlite_table_name=None): + + self.qtable_name = qtable_name + self.delimited_file_reader = delimited_file_reader + + self.db_id = sqlite_db.db_id + + self.sqlite_db = sqlite_db + self.target_sqlite_table_name = target_sqlite_table_name + + self.skip_header = input_params.skip_header + self.gzipped = input_params.gzipped_input + self.table_created = False + + self.encoding = input_params.input_encoding + self.mode = input_params.parsing_mode + self.expected_column_count = input_params.expected_column_count + self.input_delimiter = input_params.delimiter + self.with_universal_newlines = input_params.with_universal_newlines + + self.column_inferer = TableColumnInferer(input_params) + + self.pre_creation_rows = [] + self.buffered_inserts = [] + self.effective_column_names = None + + # Column type indices for columns that contain numeric types. Lazily initialized + # so column inferer can do its work before this information is needed + self.numeric_column_indices = None + + self.state = TableCreatorState.INITIALIZED + + self.content_signature = None + + def _generate_content_signature(self): + if self.state != TableCreatorState.ANALYZED: + # TODO Change to assertion + raise Exception('Bug - Wrong state %s. Table needs to be analyzed before a content signature can be calculated' % self.state) + + size = self.delimited_file_reader.get_size_hash() + last_modification_time = self.delimited_file_reader.get_last_modification_time_hash() + + m = OrderedDict({ + "_signature_version": "v1", + "skip_header": self.skip_header, + "gzipped": self.gzipped, + "with_universal_newlines": self.with_universal_newlines, + "encoding": self.encoding, + "mode": self.mode, + "expected_column_count": self.expected_column_count, + "input_delimiter": self.input_delimiter, + "inferer": self.column_inferer._generate_content_signature(), + "original_file_size": size, + "last_modification_time": last_modification_time + }) + + return m + + def validate_extra_header_if_needed(self, file_number, filename,col_vals): + xprint("HHX validate",file_number,filename,col_vals) + if not self.skip_header: + xprint("No need to validate header") + return False + + if file_number == 0: + xprint("First file, no need to validate extra header") + return False + + header_already_exists = self.column_inferer.header_row is not None + + if header_already_exists: + xprint("Validating extra header") + if tuple(self.column_inferer.header_row) != tuple(col_vals): + raise BadHeaderException("Extra header '{}' in file '{}' mismatches original header '{}' from file '{}'. Table name is '{}'".format( + ",".join(col_vals),filename, + ",".join(self.column_inferer.header_row), + self.column_inferer.header_row_filename, + self.qtable_name)) + xprint("header already exists: %s" % self.column_inferer.header_row) + else: + xprint("Header doesn't already exist") + + return header_already_exists + + def _populate(self,dialect,stop_after_analysis=False): + total_data_lines_read = 0 + try: + try: + for file_name,file_number,is_first_line,col_vals in self.delimited_file_reader.generate_rows(): + if is_first_line: + if self.validate_extra_header_if_needed(file_number,file_name,col_vals): + continue + self._insert_row(file_name, col_vals) + if stop_after_analysis: + if self.column_inferer.inferred: + xprint("Stopping after analysis") + return + if self.delimited_file_reader.get_lines_read() == 0 and self.skip_header: + raise MissingHeaderException("Header line is expected but missing in file %s" % ",".join(self.delimited_file_reader.atomic_fns)) + + total_data_lines_read += self.delimited_file_reader.lines_read - (1 if self.skip_header else 0) + xprint("Total Data lines read %s" % total_data_lines_read) + except StrictModeColumnCountMismatchException as e: + raise ColumnCountMismatchException( + 'Strict mode - Expected %s columns instead of %s columns in file %s row %s. Either use relaxed modes or check your delimiter' % ( + e.expected_col_count, e.actual_col_count, normalized_filename(e.atomic_fn), e.lines_read)) + except FluffyModeColumnCountMismatchException as e: + raise ColumnCountMismatchException( + 'Deprecated fluffy mode - Too many columns in file %s row %s (%s fields instead of %s fields). Consider moving to either relaxed or strict mode' % ( + normalized_filename(e.atomic_fn), e.lines_read, e.actual_col_count, e.expected_col_count)) + finally: + self._flush_inserts() + + if not self.table_created: + self.column_inferer.force_analysis() + self._do_create_table(self.qtable_name) + + self.sqlite_db.conn.commit() + + def perform_analyze(self, dialect): + xprint("Analyzing... %s" % dialect) + if self.state == TableCreatorState.INITIALIZED: + self._populate(dialect,stop_after_analysis=True) + self.state = TableCreatorState.ANALYZED + + self.content_signature = self._generate_content_signature() + content_signature_key = self.sqlite_db.calculate_content_signature_key(self.content_signature) + xprint("Setting content signature after analysis: %s" % content_signature_key) + else: + # TODO Convert to assertion + raise Exception('Bug - Wrong state %s' % self.state) + + def perform_read_fully(self, dialect): + if self.state == TableCreatorState.ANALYZED: + self._populate(dialect,stop_after_analysis=False) + self.state = TableCreatorState.FULLY_READ + else: + # TODO Convert to assertion + raise Exception('Bug - Wrong state %s' % self.state) + + def _flush_pre_creation_rows(self, filename): + for i, col_vals in enumerate(self.pre_creation_rows): + if self.skip_header and i == 0: + # skip header line + continue + self._insert_row(filename, col_vals) + self._flush_inserts() + self.pre_creation_rows = [] + + def _insert_row(self, filename, col_vals): + # If table has not been created yet + if not self.table_created: + # Try to create it along with another "example" line of data + self.try_to_create_table(filename, col_vals) + + # If the table is still not created, then we don't have enough data, just + # store the data and return + if not self.table_created: + self.pre_creation_rows.append(col_vals) + return + + + # The table already exists, so we can just add a new row + self._insert_row_i(col_vals) + + def initialize_numeric_column_indices_if_needed(self): + # Lazy initialization of numeric column indices + if self.numeric_column_indices is None: + column_types = self.column_inferer.get_column_types() + self.numeric_column_indices = [idx for idx, column_type in enumerate( + column_types) if self.sqlite_db.is_numeric_type(column_type)] + + def nullify_values_if_needed(self, col_vals): + new_vals = col_vals[:] + col_count = len(col_vals) + for i in self.numeric_column_indices: + if i >= col_count: + continue + v = col_vals[i] + if v == '': + new_vals[i] = None + return new_vals + + def normalize_col_vals(self, col_vals): + # Make sure that numeric column indices are initializd + self.initialize_numeric_column_indices_if_needed() + + col_vals = self.nullify_values_if_needed(col_vals) + + expected_col_count = self.column_inferer.get_column_count() + actual_col_count = len(col_vals) + if self.mode == 'strict': + if actual_col_count != expected_col_count: + raise StrictModeColumnCountMismatchException(",".join(self.delimited_file_reader.atomic_fns), expected_col_count,actual_col_count,self.delimited_file_reader.get_lines_read()) + return col_vals + + # in all non strict mode, we add dummy data to missing columns + + if actual_col_count < expected_col_count: + col_vals = col_vals + \ + [None for x in range(expected_col_count - actual_col_count)] + + # in relaxed mode, we merge all extra columns to the last column value + if self.mode == 'relaxed': + if actual_col_count > expected_col_count: + xxx = col_vals[:expected_col_count - 1] + \ + [self.input_delimiter.join([v if v is not None else '' for v in + col_vals[expected_col_count - 1:]])] + return xxx + else: + return col_vals + + assert False, "Unidentified parsing mode %s" % self.mode + + def _insert_row_i(self, col_vals): + col_vals = self.normalize_col_vals(col_vals) + + if self.effective_column_names is None: + self.effective_column_names = self.column_inferer.column_names[:len(col_vals)] + + if len(self.effective_column_names) > 0: + self.buffered_inserts.append(col_vals) + else: + self.buffered_inserts.append([""]) + + if len(self.buffered_inserts) < 5000: + return + self._flush_inserts() + + def _flush_inserts(self): + # If the table is still not created, then we don't have enough data + if not self.table_created: + return + + if len(self.buffered_inserts) > 0: + insert_row_stmt = self.sqlite_db.generate_insert_row( + self.target_sqlite_table_name, self.effective_column_names) + + self.sqlite_db.update_many(insert_row_stmt, self.buffered_inserts) + self.buffered_inserts = [] + + def try_to_create_table(self, filename, col_vals): + if self.table_created: + # TODO Convert to assertion + raise Exception('Table is already created') + + # Add that line to the column inferer + result = self.column_inferer.analyze(filename, col_vals) + # If inferer succeeded, + if result: + self._do_create_table(filename) + else: + pass # We don't have enough information for creating the table yet + + def _do_create_table(self,filename): + # Get the column definition dict from the inferer + column_dict = self.column_inferer.get_column_dict() + + # Guard against empty tables (instead of preventing the creation, just create with a dummy column) + if len(column_dict) == 0: + column_dict = { 'dummy_column_for_empty_tables' : str } + ordered_column_names = [ 'dummy_column_for_empty_tables' ] + else: + ordered_column_names = self.column_inferer.get_column_names() + + # Create the CREATE TABLE statement + create_table_stmt = self.sqlite_db.generate_create_table( + self.target_sqlite_table_name, ordered_column_names, column_dict) + # And create the table itself + self.sqlite_db.execute_and_fetch(create_table_stmt) + # Mark the table as created + self.table_created = True + self._flush_pre_creation_rows(filename) + + +def determine_max_col_lengths(m,output_field_quoting_func,output_delimiter): + if len(m) == 0: + return [] + max_lengths = [0 for x in range(0, len(m[0]))] + for row_index in range(0, len(m)): + for col_index in range(0, len(m[0])): + # TODO Optimize this + new_len = len("{}".format(output_field_quoting_func(output_delimiter,m[row_index][col_index]))) + if new_len > max_lengths[col_index]: + max_lengths[col_index] = new_len + return max_lengths + +def print_credentials(): + print("q version %s" % q_version, file=sys.stderr) + print("Python: %s" % " // ".join([str(x).strip() for x in sys.version.split("\n")]), file=sys.stderr) + print("Copyright (C) 2012-2021 Harel Ben-Attia (harelba@gmail.com, @harelba on twitter)", file=sys.stderr) + print("https://harelba.github.io/q/", file=sys.stderr) + print(file=sys.stderr) + +class QWarning(object): + def __init__(self,exception,msg): + self.exception = exception + self.msg = msg + +class QError(object): + def __init__(self,exception,msg,errorcode): + self.exception = exception + self.msg = msg + self.errorcode = errorcode + self.traceback = traceback.format_exc() + + def __str__(self): + return "QError" % (self.errorcode,self.msg,self.exception,str(self.traceback)) + __repr__ = __str__ + +class QMetadata(object): + def __init__(self,table_structures={},new_table_structures={},output_column_name_list=None): + self.table_structures = table_structures + self.new_table_structures = new_table_structures + self.output_column_name_list = output_column_name_list + + def __str__(self): + return "QMetadata<%s" % (self.__dict__) + __repr__ = __str__ + +class QOutput(object): + def __init__(self,data=None,metadata=None,warnings=[],error=None): + self.data = data + self.metadata = metadata + + self.warnings = warnings + self.error = error + if error is None: + self.status = 'ok' + else: + self.status = 'error' + + def __str__(self): + s = [] + s.append('status=%s' % self.status) + if self.error is not None: + s.append("error=%s" % self.error.msg) + if len(self.warnings) > 0: + s.append("warning_count=%s" % len(self.warnings)) + if self.data is not None: + s.append("row_count=%s" % len(self.data)) + else: + s.append("row_count=None") + if self.metadata is not None: + s.append("metadata=<%s>" % self.metadata) + else: + s.append("metadata=None") + return "QOutput<%s>" % ",".join(s) + __repr__ = __str__ + +class QInputParams(object): + def __init__(self,skip_header=False, + delimiter=' ',input_encoding='UTF-8',gzipped_input=False,with_universal_newlines=False,parsing_mode='relaxed', + expected_column_count=None,keep_leading_whitespace_in_values=False, + disable_double_double_quoting=False,disable_escaped_double_quoting=False, + disable_column_type_detection=False, + input_quoting_mode='minimal',stdin_file=None,stdin_filename='-', + max_column_length_limit=131072, + read_caching=False, + write_caching=False, + max_attached_sqlite_databases = 10): + self.skip_header = skip_header + self.delimiter = delimiter + self.input_encoding = input_encoding + self.gzipped_input = gzipped_input + self.with_universal_newlines = with_universal_newlines + self.parsing_mode = parsing_mode + self.expected_column_count = expected_column_count + self.keep_leading_whitespace_in_values = keep_leading_whitespace_in_values + self.disable_double_double_quoting = disable_double_double_quoting + self.disable_escaped_double_quoting = disable_escaped_double_quoting + self.input_quoting_mode = input_quoting_mode + self.disable_column_type_detection = disable_column_type_detection + self.max_column_length_limit = max_column_length_limit + self.read_caching = read_caching + self.write_caching = write_caching + self.max_attached_sqlite_databases = max_attached_sqlite_databases + + def merged_with(self,input_params): + params = QInputParams(**self.__dict__) + if input_params is not None: + params.__dict__.update(**input_params.__dict__) + return params + + def __str__(self): + return "QInputParams<%s>" % str(self.__dict__) + + def __repr__(self): + return "QInputParams(...)" + +class DataStream(object): + # TODO Can stream-id be removed? + def __init__(self,stream_id,filename,stream): + self.stream_id = stream_id + self.filename = filename + self.stream = stream + + def __str__(self): + return "QDataStream" % (self.stream_id,self.filename,self.stream) + __repr__ = __str__ + + +class DataStreams(object): + def __init__(self, data_streams_dict): + assert type(data_streams_dict) == dict + self.validate(data_streams_dict) + self.data_streams_dict = data_streams_dict + + def validate(self,d): + for k in d: + v = d[k] + if type(k) != str or type(v) != DataStream: + raise Exception('Bug - Invalid dict: %s' % str(d)) + + def get_for_filename(self, filename): + xprint("Data streams dict is %s. Trying to find %s" % (self.data_streams_dict,filename)) + x = self.data_streams_dict.get(filename) + return x + + def is_data_stream(self,filename): + return filename in self.data_streams_dict + +class DatabaseInfo(object): + def __init__(self,db_id,sqlite_db,needs_closing): + self.db_id = db_id + self.sqlite_db = sqlite_db + self.needs_closing = needs_closing + + def __str__(self): + return "DatabaseInfo" % (self.sqlite_db,self.needs_closing) + __repr__ = __str__ + +class QTextAsData(object): + def __init__(self,default_input_params=QInputParams(),data_streams_dict=None): + self.engine_id = str(uuid.uuid4()).replace("-","_") + + self.default_input_params = default_input_params + xprint("Default input params: %s" % self.default_input_params) + + self.loaded_table_structures_dict = OrderedDict() + self.databases = OrderedDict() + + if data_streams_dict is not None: + self.data_streams = DataStreams(data_streams_dict) + else: + self.data_streams = DataStreams({}) + + # Create DB object + self.query_level_db_id = 'query_e_%s' % self.engine_id + self.query_level_db = Sqlite3DB(self.query_level_db_id, + 'file:%s?mode=memory&cache=shared' % self.query_level_db_id,'',create_qcatalog=True) + self.adhoc_db_id = 'adhoc_e_%s' % self.engine_id + self.adhoc_db_name = 'file:%s?mode=memory&cache=shared' % self.adhoc_db_id + self.adhoc_db = Sqlite3DB(self.adhoc_db_id,self.adhoc_db_name,'',create_qcatalog=True) + self.query_level_db.conn.execute("attach '%s' as %s" % (self.adhoc_db_name,self.adhoc_db_id)) + + self.add_db_to_database_list(DatabaseInfo(self.query_level_db_id,self.query_level_db,needs_closing=True)) + self.add_db_to_database_list(DatabaseInfo(self.adhoc_db_id,self.adhoc_db,needs_closing=True)) + + def done(self): + xprint("Inside done: Database list is %s" % self.databases) + for db_id in reversed(self.databases.keys()): + database_info = self.databases[db_id] + if database_info.needs_closing: + xprint("Gonna close database %s - %s" % (db_id,self.databases[db_id])) + self.databases[db_id].sqlite_db.done() + xprint("Database %s has been closed" % db_id) + else: + xprint("No need to close database %s" % db_id) + xprint("Closed all databases") + + input_quoting_modes = { 'minimal' : csv.QUOTE_MINIMAL, + 'all' : csv.QUOTE_ALL, + # nonnumeric is not supported for input quoting modes, since we determine the data types + # ourselves instead of letting the csv module try to identify the types + 'none' : csv.QUOTE_NONE } + + def determine_proper_dialect(self,input_params): + + input_quoting_mode_csv_numeral = QTextAsData.input_quoting_modes[input_params.input_quoting_mode] + + if input_params.keep_leading_whitespace_in_values: + skip_initial_space = False + else: + skip_initial_space = True + + dialect = {'skipinitialspace': skip_initial_space, + 'delimiter': input_params.delimiter, 'quotechar': '"' } + dialect['quoting'] = input_quoting_mode_csv_numeral + dialect['doublequote'] = input_params.disable_double_double_quoting + + if input_params.disable_escaped_double_quoting: + dialect['escapechar'] = '\\' + + return dialect + + def get_dialect_id(self,filename): + return 'q_dialect_%s' % filename + + def _open_files_and_get_mfss(self,qtable_name,input_params,dialect): + materialized_file_dict = OrderedDict() + + materialized_state_type,table_source_type,source_info = detect_qtable_name_source_info(qtable_name,self.data_streams,read_caching_enabled=input_params.read_caching) + xprint("Detected source type %s source info %s" % (materialized_state_type,source_info)) + + if materialized_state_type == MaterializedStateType.DATA_STREAM: + (data_stream,) = source_info + ms = MaterialiedDataStreamState(table_source_type,qtable_name,input_params,dialect,self.engine_id,data_stream,stream_target_db=self.adhoc_db) + effective_qtable_name = data_stream.stream_id + elif materialized_state_type == MaterializedStateType.QSQL_FILE: + (qsql_filename,table_name) = source_info + ms = MaterializedQsqlState(table_source_type,qtable_name, qsql_filename=qsql_filename, table_name=table_name, + engine_id=self.engine_id, input_params=input_params, dialect_id=dialect) + effective_qtable_name = '%s:::%s' % (qsql_filename, table_name) + elif materialized_state_type == MaterializedStateType.SQLITE_FILE: + (sqlite_filename,table_name) = source_info + ms = MaterializedSqliteState(table_source_type,qtable_name, sqlite_filename=sqlite_filename, table_name=table_name, + engine_id=self.engine_id) + effective_qtable_name = '%s:::%s' % (sqlite_filename, table_name) + elif materialized_state_type == MaterializedStateType.DELIMITED_FILE: + (source_qtable_name,_) = source_info + ms = MaterializedDelimitedFileState(table_source_type,source_qtable_name, input_params, dialect, self.engine_id) + effective_qtable_name = source_qtable_name + else: + assert False, "Unknown file type for qtable %s should have exited with an exception" % (qtable_name) + + assert effective_qtable_name not in materialized_file_dict + materialized_file_dict[effective_qtable_name] = ms + + xprint("MS dict: %s" % str(materialized_file_dict)) + + return list([item for item in materialized_file_dict.values()]) + + def _load_mfs(self,mfs,input_params,dialect_id,stop_after_analysis): + xprint("Loading MFS:", mfs) + + materialized_state_type = mfs.get_materialized_state_type() + xprint("Detected materialized state type for %s: %s" % (mfs.qtable_name,materialized_state_type)) + + mfs.initialize() + + if not materialized_state_type in [MaterializedStateType.DATA_STREAM]: + if stop_after_analysis or self.should_copy_instead_of_attach(input_params): + xprint("Should copy instead of attaching. Forcing db to use to adhoc db") + forced_db_to_use = self.adhoc_db + else: + forced_db_to_use = None + else: + forced_db_to_use = None + + mfs.choose_db_to_use(forced_db_to_use,stop_after_analysis) + xprint("Chosen db to use: source %s source_type %s db_id %s db_to_use %s" % (mfs.source,mfs.source_type,mfs.db_id,mfs.db_to_use)) + + database_info,relevant_table = mfs.make_data_available(stop_after_analysis) + + if not self.is_adhoc_db(mfs.db_to_use) and not self.should_copy_instead_of_attach(input_params): + if not self.already_attached_to_query_level_db(mfs.db_to_use): + self.attach_to_db(mfs.db_to_use, self.query_level_db) + self.add_db_to_database_list(database_info) + else: + xprint("DB %s is already attached to query level db. No need to attach it again.") + + mfs.finalize() + + xprint("MFS Loaded") + + return mfs.source,mfs.source_type + + def add_db_to_database_list(self,database_info): + db_id = database_info.db_id + assert db_id is not None + assert database_info.sqlite_db is not None + if db_id in self.databases: + # TODO Convert to assertion + if id(database_info.sqlite_db) != id(self.databases[db_id].sqlite_db): + raise Exception('Bug - database already in database list: db_id %s: old %s new %s' % (db_id,self.databases[db_id],database_info)) + else: + return + self.databases[db_id] = database_info + + def is_adhoc_db(self,db_to_use): + return db_to_use.db_id == self.adhoc_db_id + + def should_copy_instead_of_attach(self,input_params): + attached_database_count = len(self.query_level_db.get_sqlite_database_list()) + x = attached_database_count >= input_params.max_attached_sqlite_databases + xprint("should_copy_instead_of_attach: attached_database_count=%s should_copy=%s" % (attached_database_count,x)) + return x + + def _load_data(self,qtable_name,input_params=QInputParams(),stop_after_analysis=False): + xprint("Attempting to load data for materialized file names %s" % qtable_name) + + q_dialect = self.determine_proper_dialect(input_params) + xprint("Dialect is %s" % q_dialect) + dialect_id = self.get_dialect_id(qtable_name) + csv.register_dialect(dialect_id, **q_dialect) + + xprint("qtable metadata for loading is %s" % qtable_name) + mfss = self._open_files_and_get_mfss(qtable_name, + input_params, + dialect_id) + assert len(mfss) == 1, "one MS now encapsulated an entire table" + mfs = mfss[0] + + xprint("MFS to load: %s" % mfs) + + if qtable_name in self.loaded_table_structures_dict.keys(): + xprint("Atomic filename %s found. no need to load" % qtable_name) + return None + + xprint("qtable %s not found - loading" % qtable_name) + + + self._load_mfs(mfs, input_params, dialect_id, stop_after_analysis) + xprint("Loaded: source-type %s source %s mfs_structure %s" % (mfs.source_type, mfs.source, mfs.mfs_structure)) + + assert qtable_name not in self.loaded_table_structures_dict, "loaded_table_structures_dict has been changed to have a non-list value" + self.loaded_table_structures_dict[qtable_name] = mfs.mfs_structure + + return mfs.mfs_structure + + def already_attached_to_query_level_db(self,db_to_attach): + attached_dbs = list(map(lambda x:x[1],self.query_level_db.get_sqlite_database_list())) + return db_to_attach.db_id in attached_dbs + + def attach_to_db(self, target_db, source_db): + q = "attach '%s' as %s" % (target_db.sqlite_db_url,target_db.db_id) + xprint("Attach query: %s" % q) + try: + c = source_db.execute_and_fetch(q) + except SqliteOperationalErrorException as e: + if 'too many attached databases' in str(e): + raise TooManyAttachedDatabasesException('There are too many attached databases. Use a proper --max-attached-sqlite-databases parameter which is below the maximum. Original error: %s' % str(e)) + except Exception as e1: + raise + + def detach_from_db(self, target_db, source_db): + q = "detach %s" % (target_db.db_id) + xprint("Detach query: %s" % q) + try: + c = source_db.execute_and_fetch(q) + except Exception as e1: + raise + + def load_data(self,filename,input_params=QInputParams(),stop_after_analysis=False): + return self._load_data(filename,input_params,stop_after_analysis=stop_after_analysis) + + def _ensure_data_is_loaded_for_sql(self,sql_object,input_params,data_streams=None,stop_after_analysis=False): + xprint("Ensuring Data load") + new_table_structures = OrderedDict() + + # For each "table name" + for qtable_name in sql_object.qtable_names: + tss = self._load_data(qtable_name,input_params,stop_after_analysis=stop_after_analysis) + if tss is not None: + xprint("New Table Structures:",new_table_structures) + assert qtable_name not in new_table_structures, "new_table_structures was changed not to contain a list as a value" + new_table_structures[qtable_name] = tss + + return new_table_structures + + def materialize_query_level_db(self,save_db_to_disk_filename,sql_object): + # TODO More robust creation - Create the file in a separate folder and move it to the target location only after success + + materialized_db = Sqlite3DB("materialized","file:%s" % save_db_to_disk_filename,save_db_to_disk_filename,create_qcatalog=False) + table_name_mapping = OrderedDict() + + # For each table in the query + effective_table_names = sql_object.get_qtable_name_effective_table_names() + + for i, qtable_name in enumerate(effective_table_names): + # table name, in the format db_id.table_name + effective_table_name_for_qtable_name = effective_table_names[qtable_name] + + source_db_id, actual_table_name_in_db = effective_table_name_for_qtable_name.split(".", 1) + # The DatabaseInfo instance for this db + source_database = self.databases[source_db_id] + if source_db_id != self.query_level_db_id: + self.attach_to_db(source_database.sqlite_db,materialized_db) + + ts = self.loaded_table_structures_dict[qtable_name] + proposed_new_table_name = ts.planned_table_name + xprint("Proposed table name is %s" % proposed_new_table_name) + + new_table_name = materialized_db.find_new_table_name(proposed_new_table_name) + + xprint("Materializing",source_db_id,actual_table_name_in_db,"as",new_table_name) + # Copy the table into the materialized database + xx = materialized_db.execute_and_fetch('CREATE TABLE %s AS SELECT * FROM %s' % (new_table_name,effective_table_name_for_qtable_name)) + + table_name_mapping[effective_table_name_for_qtable_name] = new_table_name + + # TODO RLRL Preparation for writing materialized database as a qsql file + # if source_database.sqlite_db.qcatalog_table_exists(): + # qcatalog_entry = source_database.sqlite_db.get_from_qcatalog_using_table_name(actual_table_name_in_db) + # # TODO RLRL Encapsulate dictionary transform inside qcatalog access methods + # materialized_db.add_to_qcatalog_table(new_table_name,OrderedDict(json.loads(qcatalog_entry['content_signature'])), + # qcatalog_entry['creation_time'], + # qcatalog_entry['source_type'], + # qcatalog_entry['source_type']) + # xprint("PQX Added to qcatalog",source_db_id,actual_table_name_in_db,'as',new_table_name) + # else: + # xprint("PQX Skipped adding to qcatalog",source_db_id,actual_table_name_in_db) + + if source_db_id != self.query_level_db: + self.detach_from_db(source_database.sqlite_db,materialized_db) + + return table_name_mapping + + def validate_query(self,sql_object,table_structures): + + for qtable_name in sql_object.qtable_names: + relevant_table_structures = [table_structures[qtable_name]] + + column_names = None + column_types = None + for ts in relevant_table_structures: + names = ts.column_names + types = ts.python_column_types + xprint("Comparing column names: %s with %s" % (column_names,names)) + if column_names is None: + column_names = names + else: + if column_names != names: + raise BadHeaderException("Column names differ for table %s: %s vs %s" % ( + qtable_name, ",".join(column_names), ",".join(names))) + + xprint("Comparing column types: %s with %s" % (column_types,types)) + if column_types is None: + column_types = types + else: + if column_types != types: + raise BadHeaderException("Column types differ for table %s: %s vs %s" % ( + qtable_name, ",".join(column_types), ",".join(types))) + + xprint("All column names match for qtable name %s: column names: %s column types: %s" % (ts.qtable_name,column_names,column_types)) + + xprint("Query validated") + + def _execute(self,query_str,input_params=None,data_streams=None,stop_after_analysis=False,save_db_to_disk_filename=None): + warnings = [] + error = None + table_structures = [] + + db_results_obj = None + + effective_input_params = self.default_input_params.merged_with(input_params) + + if type(query_str) != unicode: + try: + # Heuristic attempt to auto convert the query to unicode before failing + query_str = query_str.decode('utf-8') + except: + error = QError(EncodedQueryException(''),"Query should be in unicode. Please make sure to provide a unicode literal string or decode it using proper the character encoding.",91) + return QOutput(error = error) + + + try: + # Create SQL statement + sql_object = Sql('%s' % query_str, self.data_streams) + + load_start_time = time.time() + iprint("Going to ensure data is loaded. Currently loaded tables: %s" % str(self.loaded_table_structures_dict)) + new_table_structures = self._ensure_data_is_loaded_for_sql(sql_object,effective_input_params,data_streams,stop_after_analysis=stop_after_analysis) + iprint("Ensured data is loaded. loaded tables: %s" % self.loaded_table_structures_dict) + + self.validate_query(sql_object,self.loaded_table_structures_dict) + + iprint("Query validated") + + sql_object.materialize_using(self.loaded_table_structures_dict) + + iprint("Materialized sql object") + + if save_db_to_disk_filename is not None: + xprint("Saving query data to disk") + dump_start_time = time.time() + table_name_mapping = self.materialize_query_level_db(save_db_to_disk_filename,sql_object) + print("Data has been saved into %s . Saving has taken %4.3f seconds" % (save_db_to_disk_filename,time.time()-dump_start_time), file=sys.stderr) + effective_sql = sql_object.get_effective_sql(table_name_mapping) + print("Query to run on the database: %s;" % effective_sql, file=sys.stderr) + command_line = 'echo "%s" | sqlite3 %s' % (effective_sql,save_db_to_disk_filename) + print("You can run the query directly from the command line using the following command: %s" % command_line, file=sys.stderr) + + # TODO Propagate dump results using a different output class instead of an empty one + return QOutput() + + # Ensure that adhoc db is not in the middle of a transaction + self.adhoc_db.conn.commit() + + all_databases = self.query_level_db.get_sqlite_database_list() + xprint("Query level db: databases %s" % all_databases) + + # Execute the query and fetch the data + db_results_obj = sql_object.execute_and_fetch(self.query_level_db) + iprint("Query executed") + + if len(db_results_obj.results) == 0: + warnings.append(QWarning(None, "Warning - data is empty")) + + return QOutput( + data = db_results_obj.results, + metadata = QMetadata( + table_structures=self.loaded_table_structures_dict, + new_table_structures=new_table_structures, + output_column_name_list=db_results_obj.query_column_names), + warnings = warnings, + error = error) + except InvalidQueryException as e: + error = QError(e,str(e),118) + except MissingHeaderException as e: + error = QError(e,e.msg,117) + except FileNotFoundException as e: + error = QError(e,e.msg,30) + except SqliteOperationalErrorException as e: + xprint("Sqlite Operational error: %s" % e) + msg = str(e.original_error) + error = QError(e,"query error: %s" % msg,1) + if "no such column" in msg and effective_input_params.skip_header: + warnings.append(QWarning(e,'Warning - There seems to be a "no such column" error, and -H (header line) exists. Please make sure that you are using the column names from the header line and not the default (cXX) column names. Another issue might be that the file contains a BOM. Files that are encoded with UTF8 and contain a BOM can be read by specifying `-e utf-9-sig` in the command line. Support for non-UTF8 encoding will be provided in the future.')) + except ColumnCountMismatchException as e: + error = QError(e,e.msg,2) + except (UnicodeDecodeError, UnicodeError) as e: + error = QError(e,"Cannot decode data. Try to change the encoding by setting it using the -e parameter. Error:%s" % e,3) + except BadHeaderException as e: + error = QError(e,"Bad header row: %s" % e.msg,35) + except CannotUnzipDataStreamException as e: + error = QError(e,"Cannot decompress standard input. Pipe the input through zcat in order to decompress.",36) + except UniversalNewlinesExistException as e: + error = QError(e,"Data contains universal newlines. Run q with -U to use universal newlines. Please note that q still doesn't support universal newlines for .gz files or for stdin. Route the data through a regular file to use -U.",103) + # deprecated, but shouldn't be used: error = QError(e,"Standard Input must be provided in order to use it as a table",61) + except CouldNotConvertStringToNumericValueException as e: + error = QError(e,"Could not convert string to a numeric value. Did you use `-w nonnumeric` with unquoted string values? Error: %s" % e.msg,58) + except CouldNotParseInputException as e: + error = QError(e,"Could not parse the input. Please make sure to set the proper -w input-wrapping parameter for your input, and that you use the proper input encoding (-e). Error: %s" % e.msg,59) + except ColumnMaxLengthLimitExceededException as e: + error = QError(e,e.msg,31) + # deprecated, but shouldn't be used: error = QError(e,e.msg,79) + except ContentSignatureDiffersException as e: + error = QError(e,"%s vs %s: Content Signatures for table %s differ at %s (source value '%s' disk signature value '%s')" % + (e.original_filename,e.other_filename,e.filenames_str,e.key,e.source_value,e.signature_value),80) + except ContentSignatureDataDiffersException as e: + error = QError(e,e.msg,81) + except MaximumSourceFilesExceededException as e: + error = QError(e,e.msg,82) + except ContentSignatureNotFoundException as e: + error = QError(e,e.msg,83) + except NonExistentTableNameInQsql as e: + msg = "Table %s could not be found in qsql file %s . Existing table names: %s" % (e.table_name,e.qsql_filename,",".join(e.existing_table_names)) + error = QError(e,msg,84) + except NonExistentTableNameInSqlite as e: + msg = "Table %s could not be found in sqlite file %s . Existing table names: %s" % (e.table_name,e.qsql_filename,",".join(e.existing_table_names)) + error = QError(e,msg,85) + except TooManyTablesInQsqlException as e: + msg = "Could not autodetect table name in qsql file. Existing Tables %s" % ",".join(e.existing_table_names) + error = QError(e,msg,86) + except NoTableInQsqlExcption as e: + msg = "Could not autodetect table name in qsql file. File contains no record of a table" + error = QError(e,msg,97) + except TooManyTablesInSqliteException as e: + msg = "Could not autodetect table name in sqlite file %s . Existing tables: %s" % (e.qsql_filename,",".join(e.existing_table_names)) + error = QError(e,msg,87) + except NoTablesInSqliteException as e: + msg = "sqlite file %s has no tables" % e.sqlite_filename + error = QError(e,msg,88) + except TooManyAttachedDatabasesException as e: + msg = str(e) + error = QError(e,msg,89) + except UnknownFileTypeException as e: + msg = str(e) + error = QError(e,msg,95) + except KeyboardInterrupt as e: + warnings.append(QWarning(e,"Interrupted")) + except Exception as e: + global DEBUG + if DEBUG: + xprint(traceback.format_exc()) + error = QError(e,repr(e),199) + + return QOutput(data=None,warnings = warnings,error = error , metadata=QMetadata(table_structures=self.loaded_table_structures_dict,new_table_structures=self.loaded_table_structures_dict,output_column_name_list=[])) + + def execute(self,query_str,input_params=None,save_db_to_disk_filename=None): + r = self._execute(query_str,input_params,stop_after_analysis=False,save_db_to_disk_filename=save_db_to_disk_filename) + return r + + def unload(self): + # TODO This would fail, since table structures are just value objects now. Will be fixed as part of making q a full python module + for qtable_name,table_creator in six.iteritems(self.loaded_table_structures_dict): + try: + table_creator.drop_table() + except: + # Support no-table select queries + pass + self.loaded_table_structures_dict = OrderedDict() + + def analyze(self,query_str,input_params=None,data_streams=None): + q_output = self._execute(query_str,input_params,data_streams=data_streams,stop_after_analysis=True) + + return q_output + +def escape_double_quotes_if_needed(v): + x = v.replace(six.u('"'), six.u('""')) + return x + +def quote_none_func(output_delimiter,v): + return v + +def quote_minimal_func(output_delimiter,v): + if v is None: + return v + t = type(v) + if (t == str or t == unicode) and ((output_delimiter in v) or ('\n' in v) or ('"' in v)): + return six.u('"{}"').format(escape_double_quotes_if_needed(v)) + return v + +def quote_nonnumeric_func(output_delimiter,v): + if v is None: + return v + if type(v) == str or type(v) == unicode: + return six.u('"{}"').format(escape_double_quotes_if_needed(v)) + return v + +def quote_all_func(output_delimiter,v): + if type(v) == str or type(v) == unicode: + return six.u('"{}"').format(escape_double_quotes_if_needed(v)) + else: + return six.u('"{}"').format(v) + +class QOutputParams(object): + def __init__(self, + delimiter=' ', + beautify=False, + output_quoting_mode='minimal', + formatting=None, + output_header=False, + encoding=None): + self.delimiter = delimiter + self.beautify = beautify + self.output_quoting_mode = output_quoting_mode + self.formatting = formatting + self.output_header = output_header + self.encoding = encoding + + def __str__(self): + return "QOutputParams<%s>" % str(self.__dict__) + + def __repr__(self): + return "QOutputParams(...)" + +class QOutputPrinter(object): + output_quoting_modes = { 'minimal' : quote_minimal_func, + 'all' : quote_all_func, + 'nonnumeric' : quote_nonnumeric_func, + 'none' : quote_none_func } + + def __init__(self,output_params,show_tracebacks=False): + self.output_params = output_params + self.show_tracebacks = show_tracebacks + + self.output_field_quoting_func = QOutputPrinter.output_quoting_modes[output_params.output_quoting_mode] + + def print_errors_and_warnings(self,f,results): + if results.status == 'error': + error = results.error + print(error.msg, file=f) + if self.show_tracebacks: + print(error.traceback, file=f) + + for warning in results.warnings: + print("%s" % warning.msg, file=f) + + def print_analysis(self,f_out,f_err,results): + self.print_errors_and_warnings(f_err,results) + + if results.metadata is None: + return + + if results.metadata.table_structures is None: + return + + for qtable_name in results.metadata.table_structures: + table_structures = results.metadata.table_structures[qtable_name] + print("Table: %s" % qtable_name,file=f_out) + print(" Sources:",file=f_out) + dl = results.metadata.new_table_structures[qtable_name] + print(" source_type: %s source: %s" % (dl.source_type,dl.source),file=f_out) + print(" Fields:",file=f_out) + for n,t in zip(table_structures.column_names,table_structures.sqlite_column_types): + print(" `%s` - %s" % (n,t), file=f_out) + + def print_output(self,f_out,f_err,results): + try: + self._print_output(f_out,f_err,results) + except (UnicodeEncodeError, UnicodeError) as e: + print("Cannot encode data. Error:%s" % e, file=f_err) + sys.exit(3) + except IOError as e: + if e.errno == 32: + # broken pipe, that's ok + pass + else: + # don't miss other problems for now + raise + except KeyboardInterrupt: + pass + + def _print_output(self,f_out,f_err,results): + self.print_errors_and_warnings(f_err,results) + + data = results.data + + if data is None: + return + + # If the user requested beautifying the output + if self.output_params.beautify: + if self.output_params.output_header: + data_with_possible_headers = data + [tuple(results.metadata.output_column_name_list)] + else: + data_with_possible_headers = data + max_lengths = determine_max_col_lengths(data_with_possible_headers,self.output_field_quoting_func,self.output_params.delimiter) + + if self.output_params.formatting: + formatting_dict = dict( + [(x.split("=")[0], x.split("=")[1]) for x in self.output_params.formatting.split(",")]) + else: + formatting_dict = {} + + try: + if self.output_params.output_header and results.metadata.output_column_name_list is not None: + data.insert(0,results.metadata.output_column_name_list) + for rownum, row in enumerate(data): + row_str = [] + skip_formatting = rownum == 0 and self.output_params.output_header + for i, col in enumerate(row): + if str(i + 1) in formatting_dict.keys() and not skip_formatting: + fmt_str = formatting_dict[str(i + 1)] + else: + if self.output_params.beautify: + fmt_str = six.u("{{0:<{}}}").format(max_lengths[i]) + else: + fmt_str = six.u("{}") + + if col is not None: + xx = self.output_field_quoting_func(self.output_params.delimiter,col) + row_str.append(fmt_str.format(xx)) + else: + row_str.append(fmt_str.format("")) + + + xxxx = six.u(self.output_params.delimiter).join(row_str) + six.u("\n") + f_out.write(xxxx) + except (UnicodeEncodeError, UnicodeError) as e: + print("Cannot encode data. Error:%s" % e, file=sys.stderr) + sys.exit(3) + except TypeError as e: + print("Error while formatting output: %s" % e, file=sys.stderr) + sys.exit(4) + except IOError as e: + if e.errno == 32: + # broken pipe, that's ok + pass + else: + # don't miss other problem for now + raise + except KeyboardInterrupt: + pass + + try: + # Prevent python bug when order of pipe shutdowns is reversed + f_out.flush() + except IOError as e: + pass + +def get_option_with_default(p, option_type, option, default): + try: + if not p.has_option('options', option): + return default + if p.get('options',option) == 'None': + return None + if option_type == 'boolean': + r = p.getboolean('options', option) + return r + elif option_type == 'int': + r = p.getint('options', option) + return r + elif option_type == 'string': + r = p.get('options', option) + return r + else: + raise Exception("Unknown option type %s " % option_type) + except ValueError as e: + raise IncorrectDefaultValueException(option_type,option,p.get("options",option)) + +QRC_FILENAME_ENVVAR = 'QRC_FILENAME' + +def dump_default_values_as_qrc(parser,exclusions): + m = parser.parse_args([]).__dict__ + m.pop('leftover') + print("[options]",file=sys.stdout) + for k in sorted(m.keys()): + if k not in exclusions: + print("%s=%s" % (k,m[k]),file=sys.stdout) + +USAGE_TEXT = """ + q + + Example Execution for a delimited file: + + q "select * from myfile.csv" + + Example Execution for an sqlite3 database: + + q "select * from mydatabase.sqlite:::my_table_name" + + or + + q "select * from mydatabase.sqlite" + + if the database file contains only one table + + Auto-caching of delimited files can be activated through `-C readwrite` (writes new caches if needed) or `-C read` (only reads existing cache files) + + Setting the default caching mode (`-C`) can be done by writing a `~/.qrc` file. See docs for more info. + +q's purpose is to bring SQL expressive power to the Linux command line and to provide easy access to text as actual data. + +q allows the following: + +* Performing SQL-like statements directly on tabular text data, auto-caching the data in order to accelerate additional querying on the same file +* Performing SQL statements directly on multi-file sqlite3 databases, without having to merge them or load them into memory + +Changing the default values for parameters can be done by creating a `~/.qrc` file. Run q with `--dump-defaults` in order to dump a default `.qrc` file into stdout. + +See https://github.com/harelba/q for more details. + +""" + +def run_standalone(): + sqlite3.enable_callback_tracebacks(True) + + p, qrc_filename = parse_qrc_file() + + args, options, parser = initialize_command_line_parser(p, qrc_filename) + + dump_defaults_and_stop__if_needed(options, parser) + + dump_version_and_stop__if_needed(options) + + STDOUT, default_input_params, q_output_printer, query_strs = parse_options(args, options) + + data_streams_dict = initialize_default_data_streams() + + q_engine = QTextAsData(default_input_params=default_input_params,data_streams_dict=data_streams_dict) + + execute_queries(STDOUT, options, q_engine, q_output_printer, query_strs) + + q_engine.done() + + sys.exit(0) + + +def dump_version_and_stop__if_needed(options): + if options.version: + print_credentials() + sys.exit(0) + + +def dump_defaults_and_stop__if_needed(options, parser): + if options.dump_defaults: + dump_default_values_as_qrc(parser, ['dump-defaults', 'version']) + sys.exit(0) + + +def execute_queries(STDOUT, options, q_engine, q_output_printer, query_strs): + for query_str in query_strs: + if options.analyze_only: + q_output = q_engine.analyze(query_str) + q_output_printer.print_analysis(STDOUT, sys.stderr, q_output) + else: + q_output = q_engine.execute(query_str, save_db_to_disk_filename=options.save_db_to_disk_filename) + q_output_printer.print_output(STDOUT, sys.stderr, q_output) + + if q_output.status == 'error': + sys.exit(q_output.error.errorcode) + + +def initialize_command_line_parser(p, qrc_filename): + try: + default_verbose = get_option_with_default(p, 'boolean', 'verbose', False) + default_save_db_to_disk = get_option_with_default(p, 'string', 'save_db_to_disk_filename', None) + default_caching_mode = get_option_with_default(p, 'string', 'caching_mode', 'none') + + default_skip_header = get_option_with_default(p, 'boolean', 'skip_header', False) + default_delimiter = get_option_with_default(p, 'string', 'delimiter', None) + default_pipe_delimited = get_option_with_default(p, 'boolean', 'pipe_delimited', False) + default_tab_delimited = get_option_with_default(p, 'boolean', 'tab_delimited', False) + default_encoding = get_option_with_default(p, 'string', 'encoding', 'UTF-8') + default_gzipped = get_option_with_default(p, 'boolean', 'gzipped', False) + default_analyze_only = get_option_with_default(p, 'boolean', 'analyze_only', False) + default_mode = get_option_with_default(p, 'string', 'mode', "relaxed") + default_column_count = get_option_with_default(p, 'string', 'column_count', None) + default_keep_leading_whitespace_in_values = get_option_with_default(p, 'boolean', + 'keep_leading_whitespace_in_values', False) + default_disable_double_double_quoting = get_option_with_default(p, 'boolean', 'disable_double_double_quoting', + True) + default_disable_escaped_double_quoting = get_option_with_default(p, 'boolean', 'disable_escaped_double_quoting', + True) + default_disable_column_type_detection = get_option_with_default(p, 'boolean', 'disable_column_type_detection', + False) + default_input_quoting_mode = get_option_with_default(p, 'string', 'input_quoting_mode', 'minimal') + default_max_column_length_limit = get_option_with_default(p, 'int', 'max_column_length_limit', 131072) + default_with_universal_newlines = get_option_with_default(p, 'boolean', 'with_universal_newlines', False) + + default_output_delimiter = get_option_with_default(p, 'string', 'output_delimiter', None) + default_pipe_delimited_output = get_option_with_default(p, 'boolean', 'pipe_delimited_output', False) + default_tab_delimited_output = get_option_with_default(p, 'boolean', 'tab_delimited_output', False) + default_output_header = get_option_with_default(p, 'boolean', 'output_header', False) + default_beautify = get_option_with_default(p, 'boolean', 'beautify', False) + default_formatting = get_option_with_default(p, 'string', 'formatting', None) + default_output_encoding = get_option_with_default(p, 'string', 'output_encoding', 'none') + default_output_quoting_mode = get_option_with_default(p, 'string', 'output_quoting_mode', 'minimal') + default_list_user_functions = get_option_with_default(p, 'boolean', 'list_user_functions', False) + default_overwrite_qsql = get_option_with_default(p, 'boolean', 'overwrite_qsql', False) + + default_query_filename = get_option_with_default(p, 'string', 'query_filename', None) + default_query_encoding = get_option_with_default(p, 'string', 'query_encoding', locale.getpreferredencoding()) + default_max_attached_sqlite_databases = get_option_with_default(p,'int','max_attached_sqlite_databases', 10) + except IncorrectDefaultValueException as e: + print("Incorrect value '%s' for option %s in .qrc file %s (option type is %s)" % ( + e.actual_value, e.option, qrc_filename, e.option_type)) + sys.exit(199) + parser = ArgumentParser(prog="q",usage=USAGE_TEXT) + parser.add_argument("-v", "--version", action="store_true", help="Print version") + parser.add_argument("-V", "--verbose", default=default_verbose, action="store_true", + help="Print debug info in case of problems") + parser.add_argument("-S", "--save-db-to-disk", dest="save_db_to_disk_filename", default=default_save_db_to_disk, + help="Save database to an sqlite database file") + parser.add_argument("-C", "--caching-mode", default=default_caching_mode, + help="Choose the autocaching mode (none/read/readwrite). Autocaches files to disk db so further queries will be faster. Caching is done to a side-file with the same name of the table, but with an added extension .qsql") + parser.add_argument("--dump-defaults", action="store_true", + help="Dump all default values for parameters and exit. Can be used in order to make sure .qrc file content is being read properly.") + parser.add_argument("--max-attached-sqlite-databases", default=default_max_attached_sqlite_databases,type=int, + help="Set the maximum number of concurrently-attached sqlite dbs. This is a compile time definition of sqlite. q's performance will slow down once this limit is reached for a query, since it will perform table copies in order to avoid that limit.") + # ----------------------------------------------- + input_data_option_group = parser.add_argument_group("Input Data Options") + input_data_option_group.add_argument("-H", "--skip-header", default=default_skip_header, + action="store_true", + help="Skip header row. This has been changed from earlier version - Only one header row is supported, and the header row is used for column naming") + input_data_option_group.add_argument("-d", "--delimiter", default=default_delimiter, + help="Field delimiter. If none specified, then space is used as the delimiter.") + input_data_option_group.add_argument("-p", "--pipe-delimited", default=default_pipe_delimited, + action="store_true", + help="Same as -d '|'. Added for convenience and readability") + input_data_option_group.add_argument("-t", "--tab-delimited", default=default_tab_delimited, + action="store_true", + help="Same as -d . Just a shorthand for handling standard tab delimited file You can use $'\\t' if you want (this is how Linux expects to provide tabs in the command line") + input_data_option_group.add_argument("-e", "--encoding", default=default_encoding, + help="Input file encoding. Defaults to UTF-8. set to none for not setting any encoding - faster, but at your own risk...") + input_data_option_group.add_argument("-z", "--gzipped", default=default_gzipped, action="store_true", + help="Data is gzipped. Useful for reading from stdin. For files, .gz means automatic gunzipping") + input_data_option_group.add_argument("-A", "--analyze-only", default=default_analyze_only, + action='store_true', + help="Analyze sample input and provide information about data types") + input_data_option_group.add_argument("-m", "--mode", default=default_mode, + help="Data parsing mode. fluffy, relaxed and strict. In strict mode, the -c column-count parameter must be supplied as well") + input_data_option_group.add_argument("-c", "--column-count", default=default_column_count, + help="Specific column count when using relaxed or strict mode") + input_data_option_group.add_argument("-k", "--keep-leading-whitespace", dest="keep_leading_whitespace_in_values", + default=default_keep_leading_whitespace_in_values, action="store_true", + help="Keep leading whitespace in values. Default behavior strips leading whitespace off values, in order to provide out-of-the-box usability for simple use cases. If you need to preserve whitespace, use this flag.") + input_data_option_group.add_argument("--disable-double-double-quoting", + default=default_disable_double_double_quoting, action="store_false", + help="Disable support for double double-quoting for escaping the double quote character. By default, you can use \"\" inside double quoted fields to escape double quotes. Mainly for backward compatibility.") + input_data_option_group.add_argument("--disable-escaped-double-quoting", + default=default_disable_escaped_double_quoting, action="store_false", + help="Disable support for escaped double-quoting for escaping the double quote character. By default, you can use \\\" inside double quoted fields to escape double quotes. Mainly for backward compatibility.") + input_data_option_group.add_argument("--as-text", dest="disable_column_type_detection", + default=default_disable_column_type_detection, action="store_true", + help="Don't detect column types - All columns will be treated as text columns") + input_data_option_group.add_argument("-w", "--input-quoting-mode", + default=default_input_quoting_mode, + help="Input quoting mode. Possible values are all, minimal and none. Note the slightly misleading parameter name, and see the matching -W parameter for output quoting.") + input_data_option_group.add_argument("-M", "--max-column-length-limit", + default=default_max_column_length_limit, + help="Sets the maximum column length.") + input_data_option_group.add_argument("-U", "--with-universal-newlines", + default=default_with_universal_newlines, action="store_true", + help="Expect universal newlines in the data. Limitation: -U works only with regular files for now, stdin or .gz files are not supported yet.") + # ----------------------------------------------- + output_data_option_group = parser.add_argument_group("Output Options") + output_data_option_group.add_argument("-D", "--output-delimiter", + default=default_output_delimiter, + help="Field delimiter for output. If none specified, then the -d delimiter is used if present, or space if no delimiter is specified") + output_data_option_group.add_argument("-P", "--pipe-delimited-output", + default=default_pipe_delimited_output, action="store_true", + help="Same as -D '|'. Added for convenience and readability.") + output_data_option_group.add_argument("-T", "--tab-delimited-output", + default=default_tab_delimited_output, action="store_true", + help="Same as -D . Just a shorthand for outputting tab delimited output. You can use -D $'\\t' if you want.") + output_data_option_group.add_argument("-O", "--output-header", default=default_output_header, + action="store_true", + help="Output header line. Output column-names are determined from the query itself. Use column aliases in order to set your column names in the query. For example, 'select name FirstName,value1/value2 MyCalculation from ...'. This can be used even if there was no header in the input.") + output_data_option_group.add_argument("-b", "--beautify", default=default_beautify, + action="store_true", + help="Beautify output according to actual values. Might be slow...") + output_data_option_group.add_argument("-f", "--formatting", default=default_formatting, + help="Output-level formatting, in the format X=fmt,Y=fmt etc, where X,Y are output column numbers (e.g. 1 for first SELECT column etc.") + output_data_option_group.add_argument("-E", "--output-encoding", + default=default_output_encoding, + help="Output encoding. Defaults to 'none', leading to selecting the system/terminal encoding") + output_data_option_group.add_argument("-W", "--output-quoting-mode", + default=default_output_quoting_mode, + help="Output quoting mode. Possible values are all, minimal, nonnumeric and none. Note the slightly misleading parameter name, and see the matching -w parameter for input quoting.") + output_data_option_group.add_argument("-L", "--list-user-functions", + default=default_list_user_functions, action="store_true", + help="List all user functions") + parser.add_argument("--overwrite-qsql", default=default_overwrite_qsql, + help="When used, qsql files (both caches and store-to-db) will be overwritten if they already exist. Use with care.") + # ----------------------------------------------- + query_option_group = parser.add_argument_group("Query Related Options") + query_option_group.add_argument("-q", "--query-filename", default=default_query_filename, + help="Read query from the provided filename instead of the command line, possibly using the provided query encoding (using -Q).") + query_option_group.add_argument("-Q", "--query-encoding", default=default_query_encoding, + help="query text encoding. Experimental. Please send your feedback on this") + # ----------------------------------------------- + parser.add_argument('leftover', nargs='*') + args = parser.parse_args() + return args.leftover, args, parser + + +def parse_qrc_file(): + p = configparser.ConfigParser() + if QRC_FILENAME_ENVVAR in os.environ: + qrc_filename = os.environ[QRC_FILENAME_ENVVAR] + if qrc_filename != 'None': + xprint("qrc filename is %s" % qrc_filename) + if os.path.exists(qrc_filename): + p.read([os.environ[QRC_FILENAME_ENVVAR]]) + else: + print('QRC_FILENAME env var exists, but cannot find qrc file at %s' % qrc_filename, file=sys.stderr) + sys.exit(244) + else: + pass # special handling of 'None' env var value for QRC_FILENAME. Allows to eliminate the default ~/.qrc reading + else: + qrc_filename = os.path.expanduser('~/.qrc') + p.read([qrc_filename, '.qrc']) + return p, qrc_filename + + +def initialize_default_data_streams(): + data_streams_dict = { + '-': DataStream('stdin', '-', sys.stdin) + } + return data_streams_dict + + +def parse_options(args, options): + if options.list_user_functions: + print_user_functions() + sys.exit(0) + if len(args) == 0 and options.query_filename is None: + print_credentials() + print("Must provide at least one query in the command line, or through a file with the -q parameter", + file=sys.stderr) + sys.exit(1) + if options.query_filename is not None: + if len(args) != 0: + print("Can't provide both a query file and a query on the command line", file=sys.stderr) + sys.exit(1) + try: + f = open(options.query_filename, 'rb') + query_strs = [f.read()] + f.close() + except: + print("Could not read query from file %s" % options.query_filename, file=sys.stderr) + sys.exit(1) + else: + if sys.stdin.encoding is not None: + query_strs = [x.encode(sys.stdin.encoding) for x in args] + else: + query_strs = args + if options.query_encoding is not None and options.query_encoding != 'none': + try: + for idx in range(len(query_strs)): + query_strs[idx] = query_strs[idx].decode(options.query_encoding).strip() + + if len(query_strs[idx]) == 0: + print("Query cannot be empty (query number %s)" % (idx + 1), file=sys.stderr) + sys.exit(1) + + except Exception as e: + print("Could not decode query number %s using the provided query encoding (%s)" % ( + idx + 1, options.query_encoding), file=sys.stderr) + sys.exit(3) + ### + if options.mode not in ['relaxed', 'strict']: + print("Parsing mode can either be relaxed or strict", file=sys.stderr) + sys.exit(13) + output_encoding = get_stdout_encoding(options.output_encoding) + try: + if six.PY3: + STDOUT = codecs.getwriter(output_encoding)(sys.stdout.buffer) + else: + STDOUT = codecs.getwriter(output_encoding)(sys.stdout) + except: + print("Could not create output stream using output encoding %s" % (output_encoding), file=sys.stderr) + sys.exit(200) + # If the user flagged for a tab-delimited file then set the delimiter to tab + if options.tab_delimited: + if options.delimiter is not None and options.delimiter != '\t': + print("Warning: -t parameter overrides -d parameter (%s)" % options.delimiter, file=sys.stderr) + options.delimiter = '\t' + # If the user flagged for a pipe-delimited file then set the delimiter to pipe + if options.pipe_delimited: + if options.delimiter is not None and options.delimiter != '|': + print("Warning: -p parameter overrides -d parameter (%s)" % options.delimiter, file=sys.stderr) + options.delimiter = '|' + if options.delimiter is None: + options.delimiter = ' ' + elif len(options.delimiter) != 1: + print("Delimiter must be one character only", file=sys.stderr) + sys.exit(5) + if options.tab_delimited_output: + if options.output_delimiter is not None and options.output_delimiter != '\t': + print("Warning: -T parameter overrides -D parameter (%s)" % options.output_delimiter, file=sys.stderr) + options.output_delimiter = '\t' + if options.pipe_delimited_output: + if options.output_delimiter is not None and options.output_delimiter != '|': + print("Warning: -P parameter overrides -D parameter (%s)" % options.output_delimiter, file=sys.stderr) + options.output_delimiter = '|' + if options.output_delimiter: + # If output delimiter is specified, then we use it + options.output_delimiter = options.output_delimiter + else: + # Otherwise, + if options.delimiter: + # if an input delimiter is specified, then we use it as the output as + # well + options.output_delimiter = options.delimiter + else: + # if no input delimiter is specified, then we use space as the default + # (since no input delimiter means any whitespace) + options.output_delimiter = " " + try: + max_column_length_limit = int(options.max_column_length_limit) + except: + print("Max column length limit must be an integer larger than 2 (%s)" % options.max_column_length_limit, + file=sys.stderr) + sys.exit(31) + if max_column_length_limit < 3: + print("Maximum column length must be larger than 2",file=sys.stderr) + sys.exit(31) + + csv.field_size_limit(max_column_length_limit) + xprint("Max column length limit is %s" % options.max_column_length_limit) + + if options.input_quoting_mode not in list(QTextAsData.input_quoting_modes.keys()): + print("Input quoting mode can only be one of %s. It cannot be set to '%s'" % ( + ",".join(sorted(QTextAsData.input_quoting_modes.keys())), options.input_quoting_mode), file=sys.stderr) + sys.exit(55) + if options.output_quoting_mode not in list(QOutputPrinter.output_quoting_modes.keys()): + print("Output quoting mode can only be one of %s. It cannot be set to '%s'" % ( + ",".join(QOutputPrinter.output_quoting_modes.keys()), options.input_quoting_mode), file=sys.stderr) + sys.exit(56) + if options.column_count is not None: + expected_column_count = int(options.column_count) + if expected_column_count < 1 or expected_column_count > int(options.max_column_length_limit): + print("Column count must be between 1 and %s" % int(options.max_column_length_limit),file=sys.stderr) + sys.exit(90) + else: + # infer automatically + expected_column_count = None + if options.encoding != 'none': + try: + codecs.lookup(options.encoding) + except LookupError: + print("Encoding %s could not be found" % options.encoding, file=sys.stderr) + sys.exit(10) + if options.save_db_to_disk_filename is not None: + if options.analyze_only: + print("Cannot save database to disk when running with -A (analyze-only) option.", file=sys.stderr) + sys.exit(119) + + print("Going to save data into a disk database: %s" % options.save_db_to_disk_filename, file=sys.stderr) + if os.path.exists(options.save_db_to_disk_filename): + print("Disk database file %s already exists." % options.save_db_to_disk_filename, file=sys.stderr) + sys.exit(77) + # sys.exit(78) Deprecated, but shouldn't be reused + if options.caching_mode not in ['none', 'read', 'readwrite']: + print("caching mode must be none,read or readwrite",file=sys.stderr) + sys.exit(85) + read_caching = options.caching_mode in ['read', 'readwrite'] + write_caching = options.caching_mode in ['readwrite'] + + if options.max_attached_sqlite_databases <= 3: + print("Max attached sqlite databases must be larger than 3") + sys.exit(99) + + default_input_params = QInputParams(skip_header=options.skip_header, + delimiter=options.delimiter, + input_encoding=options.encoding, + gzipped_input=options.gzipped, + with_universal_newlines=options.with_universal_newlines, + parsing_mode=options.mode, + expected_column_count=expected_column_count, + keep_leading_whitespace_in_values=options.keep_leading_whitespace_in_values, + disable_double_double_quoting=options.disable_double_double_quoting, + disable_escaped_double_quoting=options.disable_escaped_double_quoting, + input_quoting_mode=options.input_quoting_mode, + disable_column_type_detection=options.disable_column_type_detection, + max_column_length_limit=max_column_length_limit, + read_caching=read_caching, + write_caching=write_caching, + max_attached_sqlite_databases=options.max_attached_sqlite_databases) + + output_params = QOutputParams( + delimiter=options.output_delimiter, + beautify=options.beautify, + output_quoting_mode=options.output_quoting_mode, + formatting=options.formatting, + output_header=options.output_header, + encoding=output_encoding) + q_output_printer = QOutputPrinter(output_params, show_tracebacks=DEBUG) + + return STDOUT, default_input_params, q_output_printer, query_strs + + +if __name__ == '__main__': + run_standalone() diff --git a/bin/qtextasdata.py b/bin/qtextasdata.py deleted file mode 120000 index ea0c8a85..00000000 --- a/bin/qtextasdata.py +++ /dev/null @@ -1 +0,0 @@ -q \ No newline at end of file diff --git a/build-deb-builder-container b/build-deb-builder-container deleted file mode 100755 index abd021a6..00000000 --- a/build-deb-builder-container +++ /dev/null @@ -1,10 +0,0 @@ -#!/bin/bash - -if [ $# -ne 1 ]; -then - echo "Usage: $(basename $0) " - exit 1 -fi -VERSION_TAG="$1" - -docker build -f dist/deb-builder-Dockerfile -t q-text-as-data-deb-builder:${VERSION_TAG} . diff --git a/build-rpm-builder-container b/build-rpm-builder-container deleted file mode 100755 index 4788f19e..00000000 --- a/build-rpm-builder-container +++ /dev/null @@ -1,10 +0,0 @@ -#!/bin/bash - -if [ $# -ne 1 ]; -then - echo "Usage: $(basename $0) " - exit 1 -fi -VERSION_TAG="$1" - -docker build -f dist/rpm-builder-Dockerfile -t q-text-as-data-rpm-builder:${VERSION_TAG} . diff --git a/conftest.py b/conftest.py new file mode 100644 index 00000000..2a7272c8 --- /dev/null +++ b/conftest.py @@ -0,0 +1,5 @@ +#!/usr/bin/env python + +# Required so pytest can find files properly + + diff --git a/create-windows-setup-instructions b/create-windows-setup-instructions deleted file mode 100644 index 9ff5fbef..00000000 --- a/create-windows-setup-instructions +++ /dev/null @@ -1,76 +0,0 @@ - -# Alpha version - process is working, instructions not fully tested for minor details yet. - -Instructions for creating a windows package for q: - -The installation is based on a wine docker container. - -mkdir -p dist/windows - -pushd dist/windows - -fetch all files from https://github.com/harelba/packages-for-q/tree/master/artifactory-for-packaging into that folder. - -tar xvzf PyInstaller-2.1.tar.gz - -Fix pyinstaller to work around a bug by running the following command (use gsed for osx or sed for linux): -gsed -i '1587s/^.*$/ if tpl[2] in ["BINARY", "DATA"]:/' ./PyInstaller-2.1/PyInstaller/build.py - -popd - - -d=`pwd` -cid1=`docker run -d -v ${d}:/q -e VNC_PASSWORD=newPW -p 5900:5900 suchja/x11server` -cid2=`docker run -d --rm -i --link ${cid1}:xserver --volumes-from ${cid1} suchja/wine:latest /bin/bash` - -sleep 1 - -function kill_container { - tmp=`docker kill ${cid1} ${cid2}` -} -trap kill_container EXIT - -docker exec -it ${cid2} /bin/bash - - -inside the docker container prompt: - - export DISPLAY=xserver:0 - - wine wineboot --init - - cd ~/.wine/dosdevices/ - - ln -s /q "q:" - - wine msiexec /i q:\\dist\\windows\\python-2.7.13.msi /q - - wine q:\\dist\\windows\\pywin32-219.win32-py2.7.exe - - mkdir ~/.wine/drive_c/q-build-environment - - cp -r /q/dist/windows/PyInstaller-2.1 ~/.wine/drive_c/q-build-environment/ - - cd /q/dist - - wine q:\\dist\\windows\\nsis-2.46-setup.exe - install to c:\q-build-environment\nsis - - - wine c:\\python27\\python.exe c:\\q-build-environment\\PyInstaller-2.1\\pyinstaller.py -F --distpath=win_output --workpath=win_build q:\\bin\\q - - ### Don't forget to change the version in the command below: - - wine c:\\q-build-environment\\nsis\\makensis.exe -DVERSION=1.6.2.0 q:\\dist\\q-TextAsData-with-path.nsi - -run the installation file and check that the install works properly: - - wine setup.exe - -move the installation to the packages folder: - - mv /q/dist/setup.exe /q/packages/setup-q-.exe - -exit from the docker container - - - diff --git a/dist/AddToPath.nsh b/dist/AddToPath.nsh deleted file mode 100644 index b961a1f4..00000000 --- a/dist/AddToPath.nsh +++ /dev/null @@ -1,440 +0,0 @@ -!ifndef _AddToPath_nsh -!define _AddToPath_nsh - -!verbose 3 -!include "WinMessages.NSH" -!verbose 4 - -!ifndef WriteEnvStr_RegKey - !ifdef ALL_USERS - !define WriteEnvStr_RegKey \ - 'HKLM "SYSTEM\CurrentControlSet\Control\Session Manager\Environment"' - !else - !define WriteEnvStr_RegKey 'HKCU "Environment"' - !endif -!endif - -; AddToPath - Adds the given dir to the search path. -; Input - head of the stack -; Note - Win9x systems requires reboot - -Function AddToPath - Exch $0 - Push $1 - Push $2 - Push $3 - - # don't add if the path doesn't exist - IfFileExists "$0\*.*" "" AddToPath_done - - ReadEnvStr $1 PATH - Push "$1;" - Push "$0;" - Call StrStr - Pop $2 - StrCmp $2 "" "" AddToPath_done - Push "$1;" - Push "$0\;" - Call StrStr - Pop $2 - StrCmp $2 "" "" AddToPath_done - GetFullPathName /SHORT $3 $0 - Push "$1;" - Push "$3;" - Call StrStr - Pop $2 - StrCmp $2 "" "" AddToPath_done - Push "$1;" - Push "$3\;" - Call StrStr - Pop $2 - StrCmp $2 "" "" AddToPath_done - - Call IsNT - Pop $1 - StrCmp $1 1 AddToPath_NT - ; Not on NT - StrCpy $1 $WINDIR 2 - FileOpen $1 "$1\autoexec.bat" a - FileSeek $1 -1 END - FileReadByte $1 $2 - IntCmp $2 26 0 +2 +2 # DOS EOF - FileSeek $1 -1 END # write over EOF - FileWrite $1 "$\r$\nSET PATH=%PATH%;$3$\r$\n" - FileClose $1 - SetRebootFlag true - Goto AddToPath_done - - AddToPath_NT: - ReadRegStr $1 ${WriteEnvStr_RegKey} "PATH" - StrCmp $1 "" AddToPath_NTdoIt - Push $1 - Call Trim - Pop $1 - StrCpy $0 "$1;$0" - AddToPath_NTdoIt: - WriteRegExpandStr ${WriteEnvStr_RegKey} "PATH" $0 - SendMessage ${HWND_BROADCAST} ${WM_WININICHANGE} 0 "STR:Environment" /TIMEOUT=5000 - - AddToPath_done: - Pop $3 - Pop $2 - Pop $1 - Pop $0 -FunctionEnd - -; RemoveFromPath - Remove a given dir from the path -; Input: head of the stack - -Function un.RemoveFromPath - Exch $0 - Push $1 - Push $2 - Push $3 - Push $4 - Push $5 - Push $6 - - IntFmt $6 "%c" 26 # DOS EOF - - Call un.IsNT - Pop $1 - StrCmp $1 1 unRemoveFromPath_NT - ; Not on NT - StrCpy $1 $WINDIR 2 - FileOpen $1 "$1\autoexec.bat" r - GetTempFileName $4 - FileOpen $2 $4 w - GetFullPathName /SHORT $0 $0 - StrCpy $0 "SET PATH=%PATH%;$0" - Goto unRemoveFromPath_dosLoop - - unRemoveFromPath_dosLoop: - FileRead $1 $3 - StrCpy $5 $3 1 -1 # read last char - StrCmp $5 $6 0 +2 # if DOS EOF - StrCpy $3 $3 -1 # remove DOS EOF so we can compare - StrCmp $3 "$0$\r$\n" unRemoveFromPath_dosLoopRemoveLine - StrCmp $3 "$0$\n" unRemoveFromPath_dosLoopRemoveLine - StrCmp $3 "$0" unRemoveFromPath_dosLoopRemoveLine - StrCmp $3 "" unRemoveFromPath_dosLoopEnd - FileWrite $2 $3 - Goto unRemoveFromPath_dosLoop - unRemoveFromPath_dosLoopRemoveLine: - SetRebootFlag true - Goto unRemoveFromPath_dosLoop - - unRemoveFromPath_dosLoopEnd: - FileClose $2 - FileClose $1 - StrCpy $1 $WINDIR 2 - Delete "$1\autoexec.bat" - CopyFiles /SILENT $4 "$1\autoexec.bat" - Delete $4 - Goto unRemoveFromPath_done - - unRemoveFromPath_NT: - ReadRegStr $1 ${WriteEnvStr_RegKey} "PATH" - StrCpy $5 $1 1 -1 # copy last char - StrCmp $5 ";" +2 # if last char != ; - StrCpy $1 "$1;" # append ; - Push $1 - Push "$0;" - Call un.StrStr ; Find `$0;` in $1 - Pop $2 ; pos of our dir - StrCmp $2 "" unRemoveFromPath_done - ; else, it is in path - # $0 - path to add - # $1 - path var - StrLen $3 "$0;" - StrLen $4 $2 - StrCpy $5 $1 -$4 # $5 is now the part before the path to remove - StrCpy $6 $2 "" $3 # $6 is now the part after the path to remove - StrCpy $3 $5$6 - - StrCpy $5 $3 1 -1 # copy last char - StrCmp $5 ";" 0 +2 # if last char == ; - StrCpy $3 $3 -1 # remove last char - - WriteRegExpandStr ${WriteEnvStr_RegKey} "PATH" $3 - SendMessage ${HWND_BROADCAST} ${WM_WININICHANGE} 0 "STR:Environment" /TIMEOUT=5000 - - unRemoveFromPath_done: - Pop $6 - Pop $5 - Pop $4 - Pop $3 - Pop $2 - Pop $1 - Pop $0 -FunctionEnd - - - -; AddToEnvVar - Adds the given value to the given environment var -; Input - head of the stack $0 environement variable $1=value to add -; Note - Win9x systems requires reboot - -Function AddToEnvVar - - Exch $1 ; $1 has environment variable value - Exch - Exch $0 ; $0 has environment variable name - - DetailPrint "Adding $1 to $0" - Push $2 - Push $3 - Push $4 - - - ReadEnvStr $2 $0 - Push "$2;" - Push "$1;" - Call StrStr - Pop $3 - StrCmp $3 "" "" AddToEnvVar_done - - Push "$2;" - Push "$1\;" - Call StrStr - Pop $3 - StrCmp $3 "" "" AddToEnvVar_done - - - Call IsNT - Pop $2 - StrCmp $2 1 AddToEnvVar_NT - ; Not on NT - StrCpy $2 $WINDIR 2 - FileOpen $2 "$2\autoexec.bat" a - FileSeek $2 -1 END - FileReadByte $2 $3 - IntCmp $3 26 0 +2 +2 # DOS EOF - FileSeek $2 -1 END # write over EOF - FileWrite $2 "$\r$\nSET $0=%$0%;$4$\r$\n" - FileClose $2 - SetRebootFlag true - Goto AddToEnvVar_done - - AddToEnvVar_NT: - ReadRegStr $2 ${WriteEnvStr_RegKey} $0 - StrCpy $3 $2 1 -1 # copy last char - StrCmp $3 ";" 0 +2 # if last char == ; - StrCpy $2 $2 -1 # remove last char - StrCmp $2 "" AddToEnvVar_NTdoIt - StrCpy $1 "$2;$1" - AddToEnvVar_NTdoIt: - WriteRegExpandStr ${WriteEnvStr_RegKey} $0 $1 - SendMessage ${HWND_BROADCAST} ${WM_WININICHANGE} 0 "STR:Environment" /TIMEOUT=5000 - - AddToEnvVar_done: - Pop $4 - Pop $3 - Pop $2 - Pop $0 - Pop $1 - -FunctionEnd - -; RemoveFromEnvVar - Remove a given value from a environment var -; Input: head of the stack - -Function un.RemoveFromEnvVar - - Exch $1 ; $1 has environment variable value - Exch - Exch $0 ; $0 has environment variable name - - DetailPrint "Removing $1 from $0" - Push $2 - Push $3 - Push $4 - Push $5 - Push $6 - Push $7 - - IntFmt $7 "%c" 26 # DOS EOF - - Call un.IsNT - Pop $2 - StrCmp $2 1 unRemoveFromEnvVar_NT - ; Not on NT - StrCpy $2 $WINDIR 2 - FileOpen $2 "$2\autoexec.bat" r - GetTempFileName $5 - FileOpen $3 $5 w - GetFullPathName /SHORT $1 $1 - StrCpy $1 "SET $0=%$0%;$1" - Goto unRemoveFromEnvVar_dosLoop - - unRemoveFromEnvVar_dosLoop: - FileRead $2 $4 - StrCpy $6 $4 1 -1 # read last char - StrCmp $6 $7 0 +2 # if DOS EOF - StrCpy $4 $4 -1 # remove DOS EOF so we can compare - StrCmp $4 "$1$\r$\n" unRemoveFromEnvVar_dosLoopRemoveLine - StrCmp $4 "$1$\n" unRemoveFromEnvVar_dosLoopRemoveLine - StrCmp $4 "$1" unRemoveFromEnvVar_dosLoopRemoveLine - StrCmp $4 "" unRemoveFromEnvVar_dosLoopEnd - FileWrite $3 $4 - Goto unRemoveFromEnvVar_dosLoop - unRemoveFromEnvVar_dosLoopRemoveLine: - SetRebootFlag true - Goto unRemoveFromEnvVar_dosLoop - - unRemoveFromEnvVar_dosLoopEnd: - FileClose $3 - FileClose $2 - StrCpy $2 $WINDIR 2 - Delete "$2\autoexec.bat" - CopyFiles /SILENT $5 "$2\autoexec.bat" - Delete $5 - Goto unRemoveFromEnvVar_done - - unRemoveFromEnvVar_NT: - ReadRegStr $2 ${WriteEnvStr_RegKey} $0 - StrCpy $6 $2 1 -1 # copy last char - StrCmp $6 ";" +2 # if last char != ; - StrCpy $2 "$2;" # append ; - Push $2 - Push "$1;" - Call un.StrStr ; Find `$1;` in $2 - Pop $3 ; pos of our dir - StrCmp $3 "" unRemoveFromEnvVar_done - ; else, it is in path - # $1 - path to add - # $2 - path var - StrLen $4 "$1;" - StrLen $5 $3 - StrCpy $6 $2 -$5 # $6 is now the part before the path to remove - StrCpy $7 $3 "" $4 # $7 is now the part after the path to remove - StrCpy $4 $6$7 - - StrCpy $6 $4 1 -1 # copy last char - StrCmp $6 ";" 0 +2 # if last char == ; - StrCpy $4 $4 -1 # remove last char - - WriteRegExpandStr ${WriteEnvStr_RegKey} $0 $4 - - ; delete reg value if null - StrCmp $4 "" 0 +2 # if null delete reg - DeleteRegValue ${WriteEnvStr_RegKey} $0 - - SendMessage ${HWND_BROADCAST} ${WM_WININICHANGE} 0 "STR:Environment" /TIMEOUT=5000 - - unRemoveFromEnvVar_done: - Pop $7 - Pop $6 - Pop $5 - Pop $4 - Pop $3 - Pop $2 - Pop $1 - Pop $0 -FunctionEnd - - - - -!ifndef IsNT_KiCHiK -!define IsNT_KiCHiK - -########################################### -# Utility Functions # -########################################### - -; IsNT -; no input -; output, top of the stack = 1 if NT or 0 if not -; -; Usage: -; Call IsNT -; Pop $R0 -; ($R0 at this point is 1 or 0) - -!macro IsNT un -Function ${un}IsNT - Push $0 - ReadRegStr $0 HKLM "SOFTWARE\Microsoft\Windows NT\CurrentVersion" CurrentVersion - StrCmp $0 "" 0 IsNT_yes - ; we are not NT. - Pop $0 - Push 0 - Return - - IsNT_yes: - ; NT!!! - Pop $0 - Push 1 -FunctionEnd -!macroend -!insertmacro IsNT "" -!insertmacro IsNT "un." - -!endif ; IsNT_KiCHiK - -; StrStr -; input, top of stack = string to search for -; top of stack-1 = string to search in -; output, top of stack (replaces with the portion of the string remaining) -; modifies no other variables. -; -; Usage: -; Push "this is a long ass string" -; Push "ass" -; Call StrStr -; Pop $R0 -; ($R0 at this point is "ass string") - -!macro StrStr un -Function ${un}StrStr -Exch $R1 ; st=haystack,old$R1, $R1=needle - Exch ; st=old$R1,haystack - Exch $R2 ; st=old$R1,old$R2, $R2=haystack - Push $R3 - Push $R4 - Push $R5 - StrLen $R3 $R1 - StrCpy $R4 0 - ; $R1=needle - ; $R2=haystack - ; $R3=len(needle) - ; $R4=cnt - ; $R5=tmp - loop: - StrCpy $R5 $R2 $R3 $R4 - StrCmp $R5 $R1 done - StrCmp $R5 "" done - IntOp $R4 $R4 + 1 - Goto loop -done: - StrCpy $R1 $R2 "" $R4 - Pop $R5 - Pop $R4 - Pop $R3 - Pop $R2 - Exch $R1 -FunctionEnd -!macroend -!insertmacro StrStr "" -!insertmacro StrStr "un." - -Function Trim ; Added by Pelaca - Exch $R1 - Push $R2 -Loop: - StrCpy $R2 "$R1" 1 -1 - StrCmp "$R2" " " RTrim - StrCmp "$R2" "$\n" RTrim - StrCmp "$R2" "$\r" RTrim - StrCmp "$R2" ";" RTrim - GoTo Done -RTrim: - StrCpy $R1 "$R1" -1 - Goto Loop -Done: - Pop $R2 - Exch $R1 -FunctionEnd - -!endif ; _AddToPath_nsh diff --git a/dist/create-deb b/dist/create-deb deleted file mode 100755 index 0017ff35..00000000 --- a/dist/create-deb +++ /dev/null @@ -1,11 +0,0 @@ -#!/bin/bash - -if [ $# -ne 1 ]; -then - echo 'create-deb ' - exit 1 -fi - -command -v alien &>/dev/null || { echo >&2 "alien needs to be installed."; exit 1; } - -alien -d -k $1 diff --git a/dist/create-rpm b/dist/create-rpm deleted file mode 100755 index 7d86661f..00000000 --- a/dist/create-rpm +++ /dev/null @@ -1,53 +0,0 @@ -#!/bin/bash - -# -# Commit tag and Version number should be provided as input in the command line -# -# - -if [ $# -ne 1 ]; -then - echo 'create-rpm ' - exit 1 -fi - -command -v ronn &>/dev/null || { echo >&2 "ronn needs to be installed."; exit 1; } -command -v rpmbuild &>/dev/null || { echo >&2 "rpmbuild needs to be installed."; exit 1; } - -set -e - -base_folder=`readlink -m $(dirname $0)` - -pushd ${base_folder} >/dev/null - -rpm_build_area=${base_folder}/rpm_build_area -rm -rf ${rpm_build_area:=/non-existent-dir} -mkdir -p ${rpm_build_area}/{SOURCES,SPECS,BUILD,RPMS,SRPMS,BUILDROOT} - -echo RPM build area is in ${rpm_build_area} - -VERSION=$1 -REAL_PACKAGE_NAME=q -RPM_PACKAGE_NAME=q-text-as-data - -FULL_NAME_FOLDER=${RPM_PACKAGE_NAME}-${VERSION} - -if [ ! -e ${RPM_PACKAGE_NAME}.spec.template ]; -then - echo "spec template does not exist. can't continue" - exit 1 -fi - -curl -o ${rpm_build_area}/SOURCES/q.tar.gz -L -R "https://github.com/harelba/q/tarball/${VERSION}" -mkdir -p ${rpm_build_area}/SOURCES -pushd ${rpm_build_area}/SOURCES >/dev/null -tar xvzf ./q.tar.gz --strip-components=1 -rm -vf ./q.tar.gz -popd >/dev/null -find ${rpm_build_area}/ -ls - -cat ${RPM_PACKAGE_NAME}.spec.template | sed "s/VERSION_PLACEHOLDER/$VERSION/g" > ${rpm_build_area}/SPECS/${RPM_PACKAGE_NAME}.spec - -rpmbuild -v --define "_topdir ${rpm_build_area}" -ba ${rpm_build_area}/SPECS/${RPM_PACKAGE_NAME}.spec - -popd >/dev/null diff --git a/dist/deb-builder-Dockerfile b/dist/deb-builder-Dockerfile deleted file mode 100644 index 7ff7b08f..00000000 --- a/dist/deb-builder-Dockerfile +++ /dev/null @@ -1,8 +0,0 @@ - -FROM ubuntu:12.04 - -RUN apt-get update && apt-get install -y alien - -ENTRYPOINT "/bin/bash" - - diff --git a/dist/fpm-config b/dist/fpm-config new file mode 100644 index 00000000..12adbf73 --- /dev/null +++ b/dist/fpm-config @@ -0,0 +1,7 @@ +-s dir +--name q-text-as-data +--license GPLv3 +--architecture x86_64 +--description "q allows to perform SQL-like statements on tabular text data." +--url https://github.com/harelba/q +--maintainer "Harel Ben-Attia " diff --git a/dist/q-TextAsData-with-path.nsi b/dist/q-TextAsData-with-path.nsi deleted file mode 100644 index f2bec1e4..00000000 --- a/dist/q-TextAsData-with-path.nsi +++ /dev/null @@ -1,182 +0,0 @@ -############################################################################################ -# NSIS Installation Script created by NSIS Quick Setup Script Generator v1.09.18 -# Entirely Edited with NullSoft Scriptable Installation System -# by Vlasis K. Barkas aka Red Wine red_wine@freemail.gr Sep 2006 -############################################################################################ - -!define APP_NAME "q-TextAsData" -!define COMP_NAME "harelba" -!define WEB_SITE "http://harelba.github.io/q/" -# REQUIRED TO BE DEFINED EXTERNALLY !define VERSION "1.5.0.0" -!define COPYRIGHT "Harel Ben-Attia @ 2012-2014" -!define DESCRIPTION "Application" -!define LICENSE_TXT "..\doc\LICENSE" -!define INSTALLER_NAME "setup.exe" -!define MAIN_APP_EXE "q.exe" -!define INSTALL_TYPE "SetShellVarContext all" -!define REG_ROOT "HKLM" -!define REG_APP_PATH "Software\Microsoft\Windows\CurrentVersion\App Paths\${MAIN_APP_EXE}" -!define UNINSTALL_PATH "Software\Microsoft\Windows\CurrentVersion\Uninstall\${APP_NAME}" - -!ifndef VERSION - Abort ; VERSION has to be defined externally -!endif -###################################################################### - -VIProductVersion "${VERSION}" -VIAddVersionKey "ProductName" "${APP_NAME}" -VIAddVersionKey "CompanyName" "${COMP_NAME}" -VIAddVersionKey "LegalCopyright" "${COPYRIGHT}" -VIAddVersionKey "FileDescription" "${DESCRIPTION}" -VIAddVersionKey "FileVersion" "${VERSION}" - -###################################################################### - -SetCompressor ZLIB -Name "${APP_NAME}" -Caption "${APP_NAME}" -OutFile "${INSTALLER_NAME}" -BrandingText "${APP_NAME}" -XPStyle on -InstallDirRegKey "${REG_ROOT}" "${REG_APP_PATH}" "" -InstallDir "$PROGRAMFILES\q-TextAsData" - -###################################################################### - -!include "AddToPath.nsh" - - -!include "MUI.nsh" - -!define MUI_ABORTWARNING -!define MUI_UNABORTWARNING - -!insertmacro MUI_PAGE_WELCOME - -!ifdef LICENSE_TXT -!insertmacro MUI_PAGE_LICENSE "${LICENSE_TXT}" -!endif - -!ifdef REG_START_MENU -!define MUI_STARTMENUPAGE_NODISABLE -!define MUI_STARTMENUPAGE_DEFAULTFOLDER "q-TextAsData" -!define MUI_STARTMENUPAGE_REGISTRY_ROOT "${REG_ROOT}" -!define MUI_STARTMENUPAGE_REGISTRY_KEY "${UNINSTALL_PATH}" -!define MUI_STARTMENUPAGE_REGISTRY_VALUENAME "${REG_START_MENU}" -!insertmacro MUI_PAGE_STARTMENU Application $SM_Folder -!endif - -!insertmacro MUI_PAGE_INSTFILES - -!insertmacro MUI_PAGE_FINISH - -!insertmacro MUI_UNPAGE_CONFIRM - -!insertmacro MUI_UNPAGE_INSTFILES - -!insertmacro MUI_UNPAGE_FINISH - -!insertmacro MUI_LANGUAGE "English" - -###################################################################### - -Section -MainProgram -${INSTALL_TYPE} -SetOverwrite ifnewer -SetOutPath "$INSTDIR" -File "q:\dist\win_output\q.exe" - -Push $INSTDIR -Call AddToPath -SectionEnd - -###################################################################### - -Section -Icons_Reg -SetOutPath "$INSTDIR" -WriteUninstaller "$INSTDIR\uninstall.exe" - -!ifdef REG_START_MENU -!insertmacro MUI_STARTMENU_WRITE_BEGIN Application -CreateDirectory "$SMPROGRAMS\$SM_Folder" -CreateShortCut "$SMPROGRAMS\$SM_Folder\${APP_NAME}.lnk" "$INSTDIR\${MAIN_APP_EXE}" -!ifdef WEB_SITE -WriteIniStr "$INSTDIR\${APP_NAME} website.url" "InternetShortcut" "URL" "${WEB_SITE}" -CreateShortCut "$SMPROGRAMS\$SM_Folder\${APP_NAME} Website.lnk" "$INSTDIR\${APP_NAME} website.url" -!endif -!insertmacro MUI_STARTMENU_WRITE_END -!endif - -!ifndef REG_START_MENU -CreateDirectory "$SMPROGRAMS\q-TextAsData" -CreateShortCut "$SMPROGRAMS\q-TextAsData\${APP_NAME}.lnk" "$INSTDIR\${MAIN_APP_EXE}" -!ifdef WEB_SITE -WriteIniStr "$INSTDIR\${APP_NAME} website.url" "InternetShortcut" "URL" "${WEB_SITE}" -CreateShortCut "$SMPROGRAMS\q-TextAsData\${APP_NAME} Website.lnk" "$INSTDIR\${APP_NAME} website.url" -!endif -!endif - -CreateShortCut "$SMPROGRAMS\q-TextAsData\Uninstall ${APP_NAME}.lnk" "$INSTDIR\uninstall.exe" - -WriteRegStr ${REG_ROOT} "${REG_APP_PATH}" "" "$INSTDIR\${MAIN_APP_EXE}" -WriteRegStr ${REG_ROOT} "${UNINSTALL_PATH}" "DisplayName" "${APP_NAME}" -WriteRegStr ${REG_ROOT} "${UNINSTALL_PATH}" "UninstallString" "$INSTDIR\uninstall.exe" -WriteRegStr ${REG_ROOT} "${UNINSTALL_PATH}" "DisplayIcon" "$INSTDIR\${MAIN_APP_EXE}" -WriteRegStr ${REG_ROOT} "${UNINSTALL_PATH}" "DisplayVersion" "${VERSION}" -WriteRegStr ${REG_ROOT} "${UNINSTALL_PATH}" "Publisher" "${COMP_NAME}" - -!ifdef WEB_SITE -WriteRegStr ${REG_ROOT} "${UNINSTALL_PATH}" "URLInfoAbout" "${WEB_SITE}" -!endif -SectionEnd - -###################################################################### - -Section Uninstall - -Push $INSTDIR -Call un.RemoveFromPath - -${INSTALL_TYPE} -Delete "$INSTDIR\select.pyd" -Delete "$INSTDIR\unicodedata.pyd" -Delete "$INSTDIR\library.zip" -Delete "$INSTDIR\bz2.pyd" -Delete "$INSTDIR\sqlite3.dll" -Delete "$INSTDIR\q.exe" -Delete "$INSTDIR\w9xpopen.exe" -Delete "$INSTDIR\python27.dll" -Delete "$INSTDIR\_sqlite3.pyd" -Delete "$INSTDIR\_hashlib.pyd" - - -Delete "$INSTDIR\uninstall.exe" -!ifdef WEB_SITE -Delete "$INSTDIR\${APP_NAME} website.url" -!endif - -RmDir "$INSTDIR" - -!ifdef REG_START_MENU -!insertmacro MUI_STARTMENU_GETFOLDER "Application" $SM_Folder -Delete "$SMPROGRAMS\$SM_Folder\${APP_NAME}.lnk" -!ifdef WEB_SITE -Delete "$SMPROGRAMS\$SM_Folder\${APP_NAME} Website.lnk" -!endif -RmDir "$SMPROGRAMS\$SM_Folder" -!endif - -!ifndef REG_START_MENU -Delete "$SMPROGRAMS\q-TextAsData\${APP_NAME}.lnk" -!ifdef WEB_SITE -Delete "$SMPROGRAMS\q-TextAsData\${APP_NAME} Website.lnk" -!endif -RmDir "$SMPROGRAMS\q-TextAsData" -!endif - -DeleteRegKey ${REG_ROOT} "${REG_APP_PATH}" -DeleteRegKey ${REG_ROOT} "${UNINSTALL_PATH}" -SectionEnd - -###################################################################### - diff --git a/dist/q-text-as-data.spec.template b/dist/q-text-as-data.spec.template deleted file mode 100644 index ad3d0c2f..00000000 --- a/dist/q-text-as-data.spec.template +++ /dev/null @@ -1,64 +0,0 @@ -%global _enable_debug_package 0 -%global debug_package %{nil} -%global __os_install_post %{nil} - -Name: q-text-as-data -Version: VERSION_PLACEHOLDER -Release: 1%{?dist} -Summary: q - Text as Data - -Group: Applications/Text -License: GPLv3 -URL: https://github.com/harelba/q -BuildArch: noarch - -%description -q allows to perform SQL-like statements on tabular text data. - - -%prep -cd %{_topdir}/BUILD -cp -vrf %{_topdir}/SOURCES/* %{_topdir}/BUILD/ -chmod -Rf a+rX,u+w,g-w,o-w %{_topdir}/BUILD/ - -%build -cd %{_topdir}/BUILD -ronn doc/USAGE.markdown - -%install -rm -vrf ${RPM_BUILD_ROOT}/ -install -d -m 0755 ${RPM_BUILD_ROOT}%{_bindir} -install -d -m 0755 ${RPM_BUILD_ROOT}%{_datadir}/q-text-as-data -install -Dm 0644 bin/q ${RPM_BUILD_ROOT}%{_datadir}/q-text-as-data/ -ln -s %{_datadir}/q-text-as-data/q ${RPM_BUILD_ROOT}%{_bindir}/q -install -d -m 0755 ${RPM_BUILD_ROOT}%{_mandir}/man1/ -install -m 0644 doc/USAGE ${RPM_BUILD_ROOT}%{_mandir}/man1/q.1 -gzip ${RPM_BUILD_ROOT}%{_mandir}/man1/q.1 - -%files -%defattr(-,root,root,-) -%{_bindir}/q -%doc README.markdown doc/AUTHORS doc/IMPLEMENTATION.markdown doc/LICENSE doc/RATIONALE.markdown doc/THANKS doc/USAGE.markdown -%attr(755,root,root) %{_datadir}/q-text-as-data/q -%{_datadir}/q-text-as-data -%doc %_mandir/man1/q.1.gz - -%changelog -*Wed Apr 05 2017 Harel Ben-Attia 1.6.0-1 -- Moved RPM building to be dockerized -- Removed the need for providing commit hashes -*Fri Dec 12 2014 Harel Ben-Attia 1.5.0-1 -- Moved stuff from create-rpm script into the rpm spec itself -*Sat Jun 14 2014 Harel Ben-Attia 1.4.0-1 -- Changed RPM package name to q-text-as-data -- Fixed RPM creation logic after folder restructuring -- Man page is now taken directly from USAGE.markdown - -* Mon Mar 03 2014 Harel Ben-Attia 1.3.0-1 -- Version 1.3.0 packaging - -* Thu Feb 20 2014 Harel Ben-Attia 1.1.7-1 -- Added man page - -* Wed Feb 19 2014 Jens Neu 1.1.5-1 -- initial release diff --git a/dist/rpm-builder-Dockerfile b/dist/rpm-builder-Dockerfile deleted file mode 100644 index dafcd1e9..00000000 --- a/dist/rpm-builder-Dockerfile +++ /dev/null @@ -1,12 +0,0 @@ - -FROM centos:centos6 - -RUN yum install -y which curl gcc make rpm rpm-build - -RUN curl -sSL https://get.rvm.io | bash - -RUN /bin/bash -l -c "rvm install 2.4.1" && /bin/bash -l -c "gem install ronn" - -ENTRYPOINT "/bin/bash" - - diff --git a/dist/test-rpm-inside-container.sh b/dist/test-rpm-inside-container.sh new file mode 100755 index 00000000..4fc43f4b --- /dev/null +++ b/dist/test-rpm-inside-container.sh @@ -0,0 +1,9 @@ +#!/bin/bash +set -x +set -e + +yum install -y python38 sqlite perl gcc python3-devel sqlite-devel +pip3 install -r test-requirements.txt + +rpm -i $1 +Q_EXECUTABLE=q Q_SKIP_EXECUTABLE_VALIDATION=true ./run-tests.sh -v diff --git a/dist/test-using-deb.sh b/dist/test-using-deb.sh new file mode 100755 index 00000000..d2c8d3b4 --- /dev/null +++ b/dist/test-using-deb.sh @@ -0,0 +1,8 @@ +#!/bin/bash + +set -x +set -e + +sudo dpkg -i $1 +Q_EXECUTABLE=q Q_SKIP_EXECUTABLE_VALIDATION=true ./run-tests.sh -v + diff --git a/dist/test-using-rpm.sh b/dist/test-using-rpm.sh new file mode 100755 index 00000000..84d6a9d5 --- /dev/null +++ b/dist/test-using-rpm.sh @@ -0,0 +1,8 @@ +#!/bin/bash + +set -x +set -e + +RPM_LOCATION=$1 + +docker run -i -v `pwd`:/q-sources -w /q-sources centos:8 /bin/bash -e -x ./dist/test-rpm-inside-container.sh ${RPM_LOCATION} diff --git a/dist/update-mac-homebrew-instructions b/dist/update-mac-homebrew-instructions deleted file mode 100644 index 0f3cf4df..00000000 --- a/dist/update-mac-homebrew-instructions +++ /dev/null @@ -1,51 +0,0 @@ - -Instructions to bump the version of q in homebrew - -1. Sync your fork with homebrew original fork - - git checkout master - - git pull upstream master - - git push origin master - -2. Create a branch for the version bump - - git checkout -b q - -3. Edit the file Library/Formula/q.rb. - - a. Change the url to the new tar.gz file - - b. Change the sha256 checksum - - $ curl -sL "https://github.com/harelba/q/archive/.tar.gz" | shasum -a 256 - - - - $ Change the checksum in q.rb's "sha256" line to the new checksum - - c. Verify by running the following: - - $ brew fetch -s q - - The output should show the new SHA256 without any warning - -4. Check the diff - - git diff | vi - - -5. Commit the change as "q " - -6. Push it as a separate branch to your repository: - - git push origin q - -7. Go the homebrew fork in github, and press "compare and pull-request" on the just-pushed branch - -8. Review the diff and make sure that the pull-request is from / to homebrew/master (should be the default) - -9. Press "create pull-request" button - -10. If everything is fine, you'll be confirmed after several hours - - diff --git a/doc/USAGE.markdown b/doc/USAGE.markdown index bc43021a..f0750864 100644 --- a/doc/USAGE.markdown +++ b/doc/USAGE.markdown @@ -3,13 +3,39 @@ ## SYNOPSIS `q ` - Simplest execution is `q "SELECT * FROM myfile"` which prints the entire file. + Example Execution for a delimited file: + q "select * from myfile.csv" + + Example Execution for an sqlite3 database: + + q "select * from mydatabase.sqlite:::my_table_name" + + or + + q "select * from mydatabase.sqlite" + + if the database file contains only one table + + Auto-caching of delimited files can be activated through `-C readwrite` (writes new caches if needed) or `-C read` (only reads existing cache files) + + Setting the default caching mode (`-C`) can be done by writing a `~/.qrc` file. See docs for more info. + ## DESCRIPTION -q allows performing SQL-like statements on tabular text data. Its purpose is to bring SQL expressive power to the Linux command line and to provide easy access to text as actual data. +q's purpose is to bring SQL expressive power to the Linux command line and to provide easy access to text as actual data. + +q allows the following: + +* Performing SQL-like statements directly on tabular text data, auto-caching the data in order to accelerate additional querying on the same file +* Performing SQL statements directly on multi-file sqlite3 databases, without having to merge them or load them into memory Query should be an SQL-like query which contains filenames instead of table names (or - for stdin). The query itself should be provided as one parameter to the tool (i.e. enclosed in quotes). +The following filename types are supported: + +* Delimited-file filenames, including relative/absolute paths +* sqlite3 database filenames, with an additional `:::` for accessing a specific table. If a database contains only one table, then denoting the table name is not needed. Examples: `mydatabase.sqlite3:::users_table` or `my_single_table_database.sqlite`. + Use `-H` to signify that the input contains a header line. Column names will be detected automatically in that case, and can be used in the query. If this option is not provided, columns will be named cX, starting with 1 (e.g. q "SELECT c3,c8 from ..."). Use `-d` to specify the input delimiter. @@ -20,7 +46,7 @@ Please note that column names that include spaces need to be used in the query w Query/Input/Output encodings are fully supported (and q tries to provide out-of-the-box usability in that area). Please use `-e`,`-E` and `-Q` to control encoding if needed. -All sqlite3 SQL constructs are supported, including joins across files (use an alias for each table). +All sqlite3 SQL constructs are supported, including joins across files (use an alias for each table), with the exception of CTE (for now). See https://github.com/harelba/q for more details. @@ -31,7 +57,7 @@ Any standard SQL expression, condition (both WHERE and HAVING), GROUP BY, ORDER JOINs are supported and Subqueries are supported in the WHERE clause, but unfortunately not in the FROM clause for now. Use table aliases when performing JOINs. -The SQL syntax itself is sqlite's syntax. For details look at http://www.sqlite.org/lang.html or search the net for examples. +The SQL syntax itself is sqlite's syntax. For details look at https://www.sqlite.org/lang.html or search the net for examples. **NOTE:** Full type detection is implemented, so there is no need for any casting or anything. @@ -44,6 +70,27 @@ q can also get some runtime flags. The following parameters can be used, all opt Options: -h, --help show this help message and exit -v, --version Print version + -V, --verbose Print debug info in case of problems + -S SAVE_DB_TO_DISK_FILENAME, --save-db-to-disk=SAVE_DB_TO_DISK_FILENAME + Save database to an sqlite database file + -C CACHING_MODE, --caching-mode=CACHING_MODE + Choose the autocaching mode (none/read/readwrite). + Autocaches files to disk db so further queries will be + faster. Caching is done to a side-file with the same + name of the table, but with an added extension .qsql + --dump-defaults Dump all default values for parameters and exit. Can + be used in order to make sure .qrc file content is + being read properly. + --max-attached-sqlite-databases=MAX_ATTACHED_SQLITE_DATABASES + Set the maximum number of concurrently-attached sqlite + dbs. This is a compile time definition of sqlite. q's + performance will slow down once this limit is reached + for a query, since it will perform table copies in + order to avoid that limit. + --overwrite-qsql=OVERWRITE_QSQL + When used, qsql files (both caches and store-to-db) + will be overwritten if they already exist. Use with + care. Input Data Options: -H, --skip-header Skip header row. This has been changed from earlier @@ -52,6 +99,8 @@ Options: -d DELIMITER, --delimiter=DELIMITER Field delimiter. If none specified, then space is used as the delimiter. + -p, --pipe-delimited + Same as -d '|'. Added for convenience and readability -t, --tab-delimited Same as -d . Just a shorthand for handling standard tab delimited file You can use $'\t' if you @@ -87,16 +136,26 @@ Options: escaping the double quote character. By default, you can use \" inside double quoted fields to escape double quotes. Mainly for backward compatibility. + --as-text Don't detect column types - All columns will be + treated as text columns -w INPUT_QUOTING_MODE, --input-quoting-mode=INPUT_QUOTING_MODE Input quoting mode. Possible values are all, minimal and none. Note the slightly misleading parameter name, and see the matching -W parameter for output quoting. + -M MAX_COLUMN_LENGTH_LIMIT, --max-column-length-limit=MAX_COLUMN_LENGTH_LIMIT + Sets the maximum column length. + -U, --with-universal-newlines + Expect universal newlines in the data. Limitation: -U + works only with regular files for now, stdin or .gz + files are not supported yet. Output Options: -D OUTPUT_DELIMITER, --output-delimiter=OUTPUT_DELIMITER Field delimiter for output. If none specified, then the -d delimiter is used if present, or space if no delimiter is specified + -P, --pipe-delimited-output + Same as -D '|'. Added for convenience and readability. -T, --tab-delimited-output Same as -D . Just a shorthand for outputting tab delimited output. You can use -D $'\t' if you want. @@ -121,6 +180,8 @@ Options: nonnumeric and none. Note the slightly misleading parameter name, and see the matching -w parameter for input quoting. + -L, --list-user-functions + List all user functions Query Related Options: -q QUERY_FILENAME, --query-filename=QUERY_FILENAME @@ -130,31 +191,26 @@ Options: -Q QUERY_ENCODING, --query-encoding=QUERY_ENCODING query text encoding. Experimental. Please send your feedback on this -```` +``` ### Table names The table names are the actual file names that you want to read from. Path names are allowed. Use "-" if you want to read from stdin (e.g. `q "SELECT * FROM -"`) -Multiple files can be concatenated by using one of both of the following ways: - -* Separating the filenames with a + sign: `SELECT * FROM datafile1+datafile2+datefile3`. -* Using glob matching: `SELECT * FROM mydata*.dat` +Wildcard matches are supported - For example: `SELECT ... FROM ... mydata*.dat` Files with .gz extension are considered to be gzipped and decompressed on the fly. ### Parsing Modes -q supports multiple parsing modes: +q supports two parsing modes: * `relaxed` - This is the default mode. It tries to lean towards simplicity of use. When a row doesn't contains enough columns, they'll be filled with nulls, and when there are too many, the extra values will be merged to the last column. Defining the number of expected columns in this mode is done using the `-c` parameter. If it is not provided, then the number of columns is detected automatically (In most use cases, there is no need to specify `-c`) * `strict` - Strict mode is for hardcore csv/tsv parsing. Whenever a row doesn't contain the proper number of columns, processing will stop. `-c` must be provided when using this mode -* `fluffy` - This mode should not be used, and is just some kind of "backward compatible" parsing mode which was used by q previously. It's left as a separate parsing mode on purpose, in order to accommodate existing users. If you are such a user, please open a bug for your use case, and I'll see how I can incorporate it into the other modes. It is reasonable to say that this mode will be removed in the future. ### Output formatting option The format of F is as a list of X=f separated by commas, where X is a column number and f is a python format: * X - column number - This is the SELECTed column (or expression) number, not the one from the original table. E.g, 1 is the first SELECTed column, 3 is the third SELECTed column. -* f - A python formatting string - See http://docs.python.org/release/2.4.4/lib/typesseq-strings.html for details if needed. -** Example: `-f 3=%-10s,5=%4.3f,1=%x` +* f - A python formatting string such as {} - See https://www.w3schools.com/python/ref_string_format.asp for details if needed. ## EXAMPLES Example 1: `ls -ltrd * | q "select c1,count(1) from - group by c1"` @@ -181,7 +237,7 @@ Harel Ben-Attia (harelba@gmail.com) Any feedback/suggestions/complaints regarding this tool would be much appreciated. Contributions are most welcome as well, of course. ## COPYRIGHT -Copyright (C) 2012--2014 Harel Ben Attia +Copyright (C) 2012--2021 Harel Ben Attia This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 3, or (at your option) any later version. diff --git a/doc/basic-examples.png b/doc/basic-examples.png deleted file mode 100644 index 6943d478..00000000 Binary files a/doc/basic-examples.png and /dev/null differ diff --git a/mkdocs/README.md b/mkdocs/README.md new file mode 100644 index 00000000..147e71aa --- /dev/null +++ b/mkdocs/README.md @@ -0,0 +1,24 @@ + +# Generate web site + +# mkdocs folder under project root +$ `cd mkdocs` + +* create a pyenv virtual environment + +$ `pip install -r requirements.txt` + +$ `./generate-web-site.sh` (static files will be generated into `./generated-site`) + +$ `git checkout gh-pages` + +$ `cd ../` # back to project root + +$ `scp -r mkdocs/generated-site/* ./` + +$ `git add` all modified files + +* commit to git + +$ `git push origin gh-pages` + diff --git a/mkdocs/docs/.DS_Store b/mkdocs/docs/.DS_Store new file mode 100644 index 00000000..6f61d6dc Binary files /dev/null and b/mkdocs/docs/.DS_Store differ diff --git a/mkdocs/docs/about.md b/mkdocs/docs/about.md new file mode 100644 index 00000000..60662ef5 --- /dev/null +++ b/mkdocs/docs/about.md @@ -0,0 +1,15 @@ +# About + +### Linkedin: [Harel Ben Attia](https://www.linkedin.com/in/harelba/) + +### Twitter [@harelba](https://twitter.com/harelba) + +### Email [harelba@gmail.com](mailto:harelba@gmail.com) + +### Patreon [harelba](https://www.patreon.com/harelba) +All the money received is donated to the [Center for the Prevention and Treatment of Domestic Violence](https://www.gov.il/he/departments/bureaus/molsa-almab-ramla) in my hometown - Ramla, Israel. + +Become a Patron! + +### Chinese translation [jinzhencheng@outlook.com](mailto:jinzhencheng@outlook.com) + diff --git a/mkdocs/docs/fsg9b9b1.txt b/mkdocs/docs/fsg9b9b1.txt new file mode 100644 index 00000000..e69de29b diff --git a/mkdocs/docs/google0efeb4ff0a886e81.html b/mkdocs/docs/google0efeb4ff0a886e81.html new file mode 100644 index 00000000..24947e8a --- /dev/null +++ b/mkdocs/docs/google0efeb4ff0a886e81.html @@ -0,0 +1 @@ +google-site-verification: google0efeb4ff0a886e81.html \ No newline at end of file diff --git a/mkdocs/docs/img/bg_hr.png b/mkdocs/docs/img/bg_hr.png new file mode 100644 index 00000000..7973bd69 Binary files /dev/null and b/mkdocs/docs/img/bg_hr.png differ diff --git a/mkdocs/docs/img/blacktocat.png b/mkdocs/docs/img/blacktocat.png new file mode 100644 index 00000000..6e264fe5 Binary files /dev/null and b/mkdocs/docs/img/blacktocat.png differ diff --git a/mkdocs/docs/img/icon_download.png b/mkdocs/docs/img/icon_download.png new file mode 100644 index 00000000..a2a287f6 Binary files /dev/null and b/mkdocs/docs/img/icon_download.png differ diff --git a/mkdocs/docs/img/q-logo.png b/mkdocs/docs/img/q-logo.png new file mode 100644 index 00000000..ac9599c3 Binary files /dev/null and b/mkdocs/docs/img/q-logo.png differ diff --git a/mkdocs/docs/img/q-logo1.ico b/mkdocs/docs/img/q-logo1.ico new file mode 100644 index 00000000..427a0d96 Binary files /dev/null and b/mkdocs/docs/img/q-logo1.ico differ diff --git a/mkdocs/docs/img/q-logo1.png b/mkdocs/docs/img/q-logo1.png new file mode 100644 index 00000000..10380ecf Binary files /dev/null and b/mkdocs/docs/img/q-logo1.png differ diff --git a/mkdocs/docs/img/sprite_download.png b/mkdocs/docs/img/sprite_download.png new file mode 100644 index 00000000..f2babd57 Binary files /dev/null and b/mkdocs/docs/img/sprite_download.png differ diff --git a/mkdocs/docs/img/sprite_download3.png b/mkdocs/docs/img/sprite_download3.png new file mode 100644 index 00000000..9fd451ac Binary files /dev/null and b/mkdocs/docs/img/sprite_download3.png differ diff --git a/mkdocs/docs/img/sprite_download4.png b/mkdocs/docs/img/sprite_download4.png new file mode 100644 index 00000000..db6e5186 Binary files /dev/null and b/mkdocs/docs/img/sprite_download4.png differ diff --git a/mkdocs/docs/img/torii-favicon.ico b/mkdocs/docs/img/torii-favicon.ico new file mode 100644 index 00000000..86da98b7 Binary files /dev/null and b/mkdocs/docs/img/torii-favicon.ico differ diff --git a/mkdocs/docs/index.md b/mkdocs/docs/index.md new file mode 100644 index 00000000..3d5601ce --- /dev/null +++ b/mkdocs/docs/index.md @@ -0,0 +1,624 @@ +# q - Run SQL directly on CSV or TSV files + +[![GitHub Stars](https://img.shields.io/github/stars/harelba/q.svg?style=social&label=GitHub Stars&maxAge=600)](https://GitHub.com/harelba/q/stargazers/) +[![GitHub forks](https://img.shields.io/github/forks/harelba/q.svg?style=social&label=GitHub Forks&maxAge=600)](https://GitHub.com/harelba/q/network/) + +## Overview +q's purpose is to bring SQL expressive power to the Linux command line by providing easy access to text as actual data, and allowing direct access to multi-file sqlite3 databases. + +```bash + q +``` + +q allows the following: + +* Performing SQL-like statements directly on tabular text data, auto-caching the data in order to accelerate additional querying on the same file + +```bash + # Simple query from a file, columns are named c1...cN + q "select c1,c5 from myfile.csv" + + # -d '|' sets the input delimiter, -H says there's a header + q -d '|' -H "select my_field from myfile.delimited-file-with-pipes" + + # -C readwrite writes a cache for the csv file + q -d , -H "select my_field from myfile.csv" -C readwrite + + # -C read tells q to use the cache + q -d , -H "select my_field from myfile.csv" -C read + + # Setting the default caching mode (`-C`) can be done by writing a `~/.qrc` file +``` + +* Performing SQL statements directly on multi-file sqlite3 databases, without having to merge them or load them into memory + +```bash + q "select * from mydatabase.sqlite:::my_table_name" + + or + + q "select * from mydatabase.sqlite" + + if the database file contains only one table + + # sqlite files are autodetected, no need for any special filename extension +``` + +The following table shows the impact of using caching: + +| Rows | Columns | File Size | Query time without caching | Query time with caching | Speed Improvement | +|:---------:|:-------:|:---------:|:--------------------------:|:-----------------------:|:-----------------:| +| 5,000,000 | 100 | 4.8GB | 4 minutes, 47 seconds | 1.92 seconds | x149 | +| 1,000,000 | 100 | 983MB | 50.9 seconds | 0.461 seconds | x110 | +| 1,000,000 | 50 | 477MB | 27.1 seconds | 0.272 seconds | x99 | +| 100,000 | 100 | 99MB | 5.2 seconds | 0.141 seconds | x36 | +| 100,000 | 50 | 48MB | 2.7 seconds | 0.105 seconds | x25 | + +Notice that for the current version, caching is **not enabled** by default, since the caches take disk space. Use `-C readwrite` or `-C read` to enable it for a query, or add `caching_mode` to `.qrc` to set a new default. + +q treats ordinary files as database tables, and supports all SQL constructs, such as `WHERE`, `GROUP BY`, `JOIN`s, etc. It supports automatic column name and type detection, and provides full support for multiple character encodings. + +The new features - autocaching, direct querying of sqlite database and the use of `~/.qrc` file are described in detail in [here](https://github.com/harelba/q/blob/master/QSQL-NOTES.md). + +Download the tool using the links in the [installation](#installation) below and play with it. + +### Encodings +| | | +|:--------------------------------------:|:-----------------------------------------------:| +| 完全支持所有的字符编码 | すべての文字エンコーディングを完全にサポート | +| 모든 문자 인코딩이 완벽하게 지원됩니다 | все кодировки символов полностью поддерживаются | + +**Non-english users:** q fully supports all types of encoding. Use `-e data-encoding` to set the input data encoding, `-Q query-encoding` to set the query encoding, and use `-E output-encoding` to set the output encoding. Sensible defaults are in place for all three parameters. Please contact me if you encounter any issues and I'd be glad to help. + +**Files with BOM:** Files which contain a BOM ([Byte Order Mark](https://en.wikipedia.org/wiki/Byte_order_mark)) are not properly supported inside python's csv module. q contains a workaround that allows reading UTF8 files which contain a BOM - Use `-e utf-8-sig` for this. I plan to separate the BOM handling from the encoding itself, which would allow to support BOMs for all encodings. + +## Installation + +| Format | Instructions | Comments | +:---|:---|:---| +|[OSX](https://github.com/harelba/q/releases/download/v3.1.6/macos-q)|Run `brew install harelba/q/q` in order to install q (moved it to its own tap), or download the standalone executable directly from the link on the left|A man page is available, just run `man q`|| +|[RPM Package](https://github.com/harelba/q/releases/download/v3.1.6/q-text-as-data-3.1.6.x86_64.rpm)| run `rpm -ivh ` or `rpm -U ` if you already have an older version of q.| A man page is available for this release. Just enter `man q`.| +|[DEB Package](https://github.com/harelba/q/releases/download/v3.1.6/q-text-as-data-3.1.6-1.x86_64.deb)| Run `sudo dpkg -i `|A man page is available for this release. Just enter `man q`. Some installations don't install the man page properly for some reason. I'll fix this soon| +|[Windows Installer](https://github.com/harelba/q/releases/download/v3.1.6/q-text-as-data-3.1.6.msi)|Run the installer executable and hit next next next... q.exe will be added to the PATH so you can access it everywhere.|Windows doesn't update the PATH retroactively for open windows, so you'll need to open a new `cmd`/`bash` window after the installation is done.| +|[Source tar.gz](https://github.com/harelba/q/archive/refs/tags/v3.1.6.tar.gz)|Full source file tree for latest stable version. Note that q.py cannot be used directly anymore, as it requires python dependencies|| +|[Source zip](https://github.com/harelba/q/archive/refs/tags/v3.1.6.zip)|Full source file tree for the latest stable version. Note that q.py cannot be used directly anymore, as it requires python dependencies|| + +I will add packages for additional Linux Distributions if there's demand for it. If you're interested in another Linux distribution, please ping me. It's relatively easy to add new ones with the new packaging flow. + +The previous version `2.0.19` can be downloaded directly from [here](https://github.com/harelba/q/releases/tag/2.0.19). Please let me know if for some reason the new version is not suitable for your needs, and you're planning on using the previous one. + +## Requirements +q is packaged as a compiled standalone-executable that has no dependencies, not even python itself. This was done by using the awesome [pyoxidizer](https://github.com/indygreg/PyOxidizer) project. + + +## Examples + +This section shows example flows that highlight the main features. For more basic examples, see [here](#getting-started-examples). + +### Basic Examples: + +```bash +# Prepare some data +$ seq 1 1000000 > myfile.csv + +# Query it +$ q "select sum(c1),count(*) from myfile.csv where c1 % 3 = 0" +166666833333 333333 + +# Use q to query from stdin +$ ps -ef | q -b -H "SELECT UID, COUNT(*) cnt FROM - GROUP BY UID ORDER BY cnt DESC LIMIT 3" +501 288 +0 115 +270 17 +``` + +### Auto-caching Examples + +```bash +# (time command output has been shortened for berevity) + +# Prepare some data +$ seq 1 1000000 > myfile.csv + +# Read from the resulting file +$ time q "select sum(c1),count(*) from myfile.csv" +500000500000 1000000 +total_time=4.108 seconds + +# Running with `-C readwrite` auto-creates a cache file if there is none. The cache filename would be myfile.csv.qsql. The query runs as usual +$ time q "select sum(c1),count(*) from myfile.csv" -C readwrite +500000500000 1000000 +total_time=4.057 seconds + +# Now run with `-C read`. The query will run from the cache file and not the original. As the file gets bigger, the difference will be much more noticable +$ time q "select sum(c1),count(*) from myfile.csv" -C read +500000500000 1000000 +total_time=0.229 seconds + +# Now let's try another query on that file. Notice the short query duration. The cache is being used for any query that uses this file, and queries on multiple files that contain caches will reuse the cache as well. +$ time q "select avg(c1) from myfile.csv" -C read +500000.5 +total_time=0.217 seconds + +# You can also query the qsql file directly, as it's just a standard sqlite3 DB file (see next section for q's support of reading directly from sqlite DBs) +$ time q "select sum(c1),count(*) from myfile.csv.qsql" +500000500000 1000000 +total_time=0.226 seconds + +# Now let's delete the original csv file (be careful when deleting original data) +$ rm -vf myfile.csv + +# Running another query directly on the qsql file just works +$ time q "select sum(c1),count(*) from myfile.csv.qsql" +500000500000 1000000 +total_time=0.226 seconds + +# See the `.qrc` section below if you want to set the default `-C` (`--caching-mode`) to something other than `none` (the default) +``` + +### Direct sqlite Querying Examples + +```bash +# Download example sqlite3 database from https://www.sqlitetutorial.net/sqlite-sample-database/ and unzip it. The resulting file will be chinook.db +$ curl -L https://www.sqlitetutorial.net/wp-content/uploads/2018/03/chinook.zip | tar -xvf - + +# Now we can query the database directly, specifying the name of the table in the query (:::) +$ q "select count(*) from chinook.db:::albums" +347 + +# Let's take the top 5 longest tracks of album id 34. The -b option just beautifies the output, and -O tells q to output the column names as headers +$ q "select * from chinook.db:::tracks where albumid = '34' order by milliseconds desc limit 5" -b -O +TrackId Name AlbumId MediaTypeId GenreId Composer Milliseconds Bytes UnitPrice +407 "Só Tinha De Ser Com Você" 34 1 7 Vários 389642 13085596 0.99 +398 "Only A Dream In Rio" 34 1 7 Vários 371356 12192989 0.99 +393 "Tarde Em Itapoã" 34 1 7 Vários 313704 10344491 0.99 +401 "Momentos Que Marcam" 34 1 7 Vários 280137 9313740 0.99 +391 "Garota De Ipanema" 34 1 7 Vários 279536 9141343 0.99 + +# Let's now copy the chinook database to another file, as if it's just another different database +$ cp chinook.db another_db.db + +# Now we can run a join query between the two databases. They could have been any two different databases, using the copy of chinook is just for simplicity +# Let's get the top-5 longest albums, using albums from the first database and tracks from the second database. The track times are converted to seconds, and rounded to two digits after the decimal point. +$ q -b -O "select a.title,round(sum(t.milliseconds)/1000.0/60,2) total_album_time_seconds from chinook.db:::albums a left join another_database.db:::tracks t on (a.albumid = t.albumid) group by a.albumid order by total_album_time_seconds desc limit 5" +Title total_album_time_seconds +"Lost, Season 3" 1177.76 +"Battlestar Galactica (Classic), Season 1" 1170.23 +"Lost, Season 1" 1080.92 +"Lost, Season 2" 1054.83 +"Heroes, Season 1" 996.34 +``` + +### Analysis Examples + +```bash +# Let's create a simple CSV file without a header. Make sure to copy only the three lines, press enter, and +# then press Ctrl-D to exit so the file will be written. +$ cat > some-data-without-header.csv +harel,1,2 +ben,3,4 +attia,5,6 + + +# Let's run q on it with -A, to see the detected structure of the file. `-d ,` sets the delimiter to a comma +$ q -d , "select * from some-data-without-header.csv" -A +Table: /Users/harelben-attia/dev/harelba/q/some-data-without-header.csv + Sources: + source_type: file source: /Users/harelben-attia/dev/harelba/q/some-data-without-header.csv + Fields: + `c1` - text + `c2` - int + `c3` - int + +# Now let's create another simple CSV file, this time with a header (-H tells q to expect a header in the file) +$ cat > some-data.csv +planet_id,name,diameter_km,length_of_day_hours +1000,Earth,12756,24 +2000,Mars,6792,24.7 +3000,Jupiter,142984,9.9 + + +# Let's run q with -A to see the analysis results. +$ q -b -O -H -d , "select * from some-data.csv" -A +Table: /Users/harelben-attia/dev/harelba/q/some-data.csv + Sources: + source_type: file source: /Users/harelben-attia/dev/harelba/q/some-data.csv + Fields: + `planet_id` - int + `name` - text + `diameter_km` - int + `length_of_day_hours` - real + +# Let's run it with `-C readwrite` so a cache will be created +$ q -b -O -H -d , "select * from some-data.csv" -C readwrite +planet_id,name ,diameter_km,length_of_day_hours +1000 ,Earth ,12756 ,24.0 +2000 ,Mars ,6792 ,24.7 +3000 ,Jupiter,142984 ,9.9 + +# Running another query that uses some-data.csv with -A will now show that a qsql exists for that file. The source-type +# will be "file-with-unused-qsql". The qsql cache is not being used, since by default, q does not activate caching +# so backward compatibility is maintained +$ q -b -O -H -d , "select * from some-data.csv" -A +Table: /Users/harelben-attia/dev/harelba/q/some-data.csv + Sources: + source_type: file-with-unused-qsql source: /Users/harelben-attia/dev/harelba/q/some-data.csv + Fields: + `planet_id` - int + `name` - text + `diameter_km` - int + `length_of_day_hours` - real + +# Now let's run another query, this time with `-C read`, telling q to use the qsql caches. This time source-type will +# be "qsql-file-with-original", and the cache will be used when querying: +$ q -b -O -H -d , "select * from some-data.csv" -A -C read +Table: /Users/harelben-attia/dev/harelba/q/some-data.csv + Sources: + source_type: qsql-file-with-original source: /Users/harelben-attia/dev/harelba/q/some-data.csv.qsql + Fields: + `planet_id` - int + `name` - text + `diameter_km` - int + `length_of_day_hours` - real + +# Let's now read directly from the qsql file. Notice the change in the table name inside the query. `-C read` is not needed +# here. The source-type will be "qsql-file" +$ q -b -O -H -d , "select * from some-data.csv.qsql" -A +Table: /Users/harelben-attia/dev/harelba/q/some-data.csv.qsql + Sources: + source_type: qsql-file source: /Users/harelben-attia/dev/harelba/q/some-data.csv.qsql + Fields: + `planet_id` - int + `name` - text + `diameter_km` - int + `length_of_day_hours` - real +``` + +## Usage +Query should be an SQL-like query which contains filenames instead of table names (or - for stdin). The query itself should be provided as one parameter to the tool (i.e. enclosed in quotes). + +All sqlite3 SQL constructs are supported, including joins across files (use an alias for each table). Take a look at the [limitations](#limitations) section below for some rarely-used use cases which are not fully supported. + +q gets a full SQL query as a parameter. Remember to double-quote the query. + +Historically, q supports multiple queries on the same command-line, loading each data file only once, even if it is used by multiple queries on the same q invocation. This is still supported. However, due to the new automatic-caching capabilities, this is not really required. Activate caching, and a cache file will be automatically created for each file. q Will use the cache behind the scenes in order to speed up queries. The speed up is extremely significant, so consider using caching for large files. + +The following filename types are supported: + +* **Delimited-file filenames** - including relative/absolute paths. E.g. `./my_folder/my_file.csv` or `/var/tmp/my_file.csv` +* **sqlite3 database filenames** + * **With Multiple Tables** - Add an additional `:::` for accessing a specific table. For example `mydatabase.sqlite3:::users_table`. + * **With One Table Only** - Just specify the database filename, no need for a table name postfix. For example `my_single_table_database.sqlite`. +* **`.qsql` cache files** - q can auto-generate cache files for delimited files, and they can be queried directly as a table, since they contain only one table, as they are essentially standard sqlite datbases + +Use `-H` to signify that the input contains a header line. Column names will be detected automatically in that case, and can be used in the query. If this option is not provided, columns will be named cX, starting with 1 (e.g. `q "SELECT c3,c8 from ..."`). + +Use `-d` to specify the input delimiter. + +Column types are auto detected by the tool, no casting is needed. Note that there's a flag `--as-text` which forces all columns to be treated as text columns. + +Please note that column names that include spaces need to be used in the query with back-ticks, as per the sqlite standard. Make sure to use single-quotes around the query, so bash/zsh won't interpret the backticks. + +Query/Input/Output encodings are fully supported (and q tries to provide out-of-the-box usability in that area). Please use `-e`,`-E` and `-Q` to control encoding if needed. + +JOINs are supported and Subqueries are supported in the WHERE clause, but unfortunately not in the FROM clause for now. Use table aliases when performing JOINs. + +The SQL syntax itself is sqlite's syntax. For details look at https://www.sqlite.org/lang.html or search the net for examples. + +NOTE: When using the `-O` output header option, use column name aliases if you want to control the output column names. For example, `q -O -H "select count(*) cnt,sum(*) as mysum from -"` would output `cnt` and `mysum` as the output header column names. + +``` bash +Options: + -h, --help show this help message and exit + -v, --version Print version + -V, --verbose Print debug info in case of problems + -S SAVE_DB_TO_DISK_FILENAME, --save-db-to-disk=SAVE_DB_TO_DISK_FILENAME + Save database to an sqlite database file + -C CACHING_MODE, --caching-mode=CACHING_MODE + Choose the autocaching mode (none/read/readwrite). + Autocaches files to disk db so further queries will be + faster. Caching is done to a side-file with the same + name of the table, but with an added extension .qsql + --dump-defaults Dump all default values for parameters and exit. Can + be used in order to make sure .qrc file content is + being read properly. + --max-attached-sqlite-databases=MAX_ATTACHED_SQLITE_DATABASES + Set the maximum number of concurrently-attached sqlite + dbs. This is a compile time definition of sqlite. q's + performance will slow down once this limit is reached + for a query, since it will perform table copies in + order to avoid that limit. + --overwrite-qsql=OVERWRITE_QSQL + When used, qsql files (both caches and store-to-db) + will be overwritten if they already exist. Use with + care. + + Input Data Options: + -H, --skip-header Skip header row. This has been changed from earlier + version - Only one header row is supported, and the + header row is used for column naming + -d DELIMITER, --delimiter=DELIMITER + Field delimiter. If none specified, then space is used + as the delimiter. + -p, --pipe-delimited + Same as -d '|'. Added for convenience and readability + -t, --tab-delimited + Same as -d . Just a shorthand for handling + standard tab delimited file You can use $'\t' if you + want (this is how Linux expects to provide tabs in the + command line + -e ENCODING, --encoding=ENCODING + Input file encoding. Defaults to UTF-8. set to none + for not setting any encoding - faster, but at your own + risk... + -z, --gzipped Data is gzipped. Useful for reading from stdin. For + files, .gz means automatic gunzipping + -A, --analyze-only Analyze sample input and provide information about + data types + -m MODE, --mode=MODE + Data parsing mode. fluffy, relaxed and strict. In + strict mode, the -c column-count parameter must be + supplied as well + -c COLUMN_COUNT, --column-count=COLUMN_COUNT + Specific column count when using relaxed or strict + mode + -k, --keep-leading-whitespace + Keep leading whitespace in values. Default behavior + strips leading whitespace off values, in order to + provide out-of-the-box usability for simple use cases. + If you need to preserve whitespace, use this flag. + --disable-double-double-quoting + Disable support for double double-quoting for escaping + the double quote character. By default, you can use "" + inside double quoted fields to escape double quotes. + Mainly for backward compatibility. + --disable-escaped-double-quoting + Disable support for escaped double-quoting for + escaping the double quote character. By default, you + can use \" inside double quoted fields to escape + double quotes. Mainly for backward compatibility. + --as-text Don't detect column types - All columns will be + treated as text columns + -w INPUT_QUOTING_MODE, --input-quoting-mode=INPUT_QUOTING_MODE + Input quoting mode. Possible values are all, minimal + and none. Note the slightly misleading parameter name, + and see the matching -W parameter for output quoting. + -M MAX_COLUMN_LENGTH_LIMIT, --max-column-length-limit=MAX_COLUMN_LENGTH_LIMIT + Sets the maximum column length. + -U, --with-universal-newlines + Expect universal newlines in the data. Limitation: -U + works only with regular files for now, stdin or .gz + files are not supported yet. + + Output Options: + -D OUTPUT_DELIMITER, --output-delimiter=OUTPUT_DELIMITER + Field delimiter for output. If none specified, then + the -d delimiter is used if present, or space if no + delimiter is specified + -P, --pipe-delimited-output + Same as -D '|'. Added for convenience and readability. + -T, --tab-delimited-output + Same as -D . Just a shorthand for outputting tab + delimited output. You can use -D $'\t' if you want. + -O, --output-header + Output header line. Output column-names are determined + from the query itself. Use column aliases in order to + set your column names in the query. For example, + 'select name FirstName,value1/value2 MyCalculation + from ...'. This can be used even if there was no + header in the input. + -b, --beautify Beautify output according to actual values. Might be + slow... + -f FORMATTING, --formatting=FORMATTING + Output-level formatting, in the format X=fmt,Y=fmt + etc, where X,Y are output column numbers (e.g. 1 for + first SELECT column etc. + -E OUTPUT_ENCODING, --output-encoding=OUTPUT_ENCODING + Output encoding. Defaults to 'none', leading to + selecting the system/terminal encoding + -W OUTPUT_QUOTING_MODE, --output-quoting-mode=OUTPUT_QUOTING_MODE + Output quoting mode. Possible values are all, minimal, + nonnumeric and none. Note the slightly misleading + parameter name, and see the matching -w parameter for + input quoting. + -L, --list-user-functions + List all user functions + + Query Related Options: + -q QUERY_FILENAME, --query-filename=QUERY_FILENAME + Read query from the provided filename instead of the + command line, possibly using the provided query + encoding (using -Q). + -Q QUERY_ENCODING, --query-encoding=QUERY_ENCODING + query text encoding. Experimental. Please send your + feedback on this +``` + +### Setting the default values for parameters +It's possible to set default values for parameters which are used often by configuring them in the file `~/.qrc`. + +The file format is as follows: +```bash +[options] += +``` + +It's possible to generate a default `.qrc` file by running `q --dump-defaults` and write the output into the `.qrc` file. + +One valuable use-case for this could be setting the caching-mode to `read`. This will make q automatically use generated `.qsql` cache files if they exist. Whenever you want a cache file to be generated, just use `-C readwrite` and a `.qsql` file will be generated if it doesn't exist. + +Here's the content of the `~/.qrc` file for enabling cache reads by default: +```bash +[options] +caching_mode=read +``` + +## Getting Started Examples +This section shows some more basic examples of simple SQL constructs. + +For some more complex use-cases, see the [examples](#examples) at the beginning of the documentation. + +NOTES: + +* The `-H` flag in the examples below signifies that the file has a header row which is used for naming columns. +* The `-t` flag is just a shortcut for saying that the file is a tab-separated file (any delimiter is supported - Use the `-d` flag). +* Queries are given using upper case for clarity, but actual query keywords such as SELECT and WHERE are not really case sensitive. + +Basic Example List: + +* [Example 1 - COUNT DISTINCT values of specific field (uuid of clicks data)](#example-1) +* [Example 2 - Filter numeric data, controlling ORDERing and LIMITing output](#example-2) +* [Example 3 - Illustrate GROUP BY](#example-3) +* [Example 4 - More complex GROUP BY (group by time expression)](#example-4) +* [Example 5 - Read input from standard input](#example-5) +* [Example 6 - Use column names from header row](#example-6) +* [Example 7 - JOIN two files](#example-7) + +### Example 1 +Perform a COUNT DISTINCT values of specific field (uuid of clicks data). + +``` bash +q -H -t "SELECT COUNT(DISTINCT(uuid)) FROM ./clicks.csv" +``` +Output +``` bash +229 +``` +### Example 2 +Filter numeric data, controlling ORDERing and LIMITing output + +Note that q understands that the column is numeric and filters according to its numeric value (real numeric value comparison, not string comparison). + +``` bash +q -H -t "SELECT request_id,score FROM ./clicks.csv WHERE score > 0.7 ORDER BY score DESC LIMIT 5" +``` +Output: +``` bash +2cfab5ceca922a1a2179dc4687a3b26e 1.0 +f6de737b5aa2c46a3db3208413a54d64 0.986665809568 +766025d25479b95a224bd614141feee5 0.977105183282 +2c09058a1b82c6dbcf9dc463e73eddd2 0.703255121794 +``` + +### Example 3 +Illustrate GROUP BY + +``` bash +q -t -H "SELECT hashed_source_machine,count(*) FROM ./clicks.csv GROUP BY hashed_source_machine" +``` +Output: +``` bash +47d9087db433b9ba.domain.com 400000 +``` + +### Example 4 +More complex GROUP BY (group by time expression) + +``` bash +q -t -H "SELECT strftime('%H:%M',date_time) hour_and_minute,count(*) FROM ./clicks.csv GROUP BY hour_and_minute" +``` +Output: +``` bash +07:00 138148 +07:01 140026 +07:02 121826 +``` + +### Example 5 +Read input from standard input + +Calculates the total size per user/group in the /tmp subtree. + +``` bash +sudo find /tmp -ls | q "SELECT c5,c6,sum(c7)/1024.0/1024 AS total FROM - GROUP BY c5,c6 ORDER BY total desc" +``` +Output: +``` bash +mapred hadoop 304.00390625 +root root 8.0431451797485 +smith smith 4.34389972687 +``` + +### Example 6 +Use column names from header row + +Calculate the top 3 user ids with the largest number of owned processes, sorted in descending order. + +Note the usage of the autodetected column name UID in the query. + +``` bash +ps -ef | q -H "SELECT UID,COUNT(*) cnt FROM - GROUP BY UID ORDER BY cnt DESC LIMIT 3" +``` +Output: +``` bash +root 152 +harel 119 +avahi 2 +``` + +### Example 7 +JOIN two files + +The following command joins an ls output (exampledatafile) and a file containing rows of group-name,email (group-emails-example) and provides a row of filename,email for each of the emails of the group. For brevity of output, there is also a filter for a specific filename called ppp which is achieved using a WHERE clause. + +``` bash +q "SELECT myfiles.c8,emails.c2 FROM exampledatafile myfiles JOIN group-emails-example emails ON (myfiles.c4 = emails.c1) WHERE myfiles.c8 = 'ppp'" +``` +Output: +``` bash +ppp dip.1@otherdomain.com +ppp dip.2@otherdomain.com +``` + +You can see that the ppp filename appears twice, each time matched to one of the emails of the group dip to which it belongs. Take a look at the files `exampledatafile` and `group-emails-example` for the data. + +Column name detection is supported for JOIN scenarios as well. Just specify `-H` in the command line and make sure that the source files contain the header rows. + +## Implementation +Behind the scenes q creates a "virtual" sqlite3 database that does not contain data of its own, but attaches to multiple other databases as follows: + +* When reading delimited files or data from `stdin`, it will analyze the data and construct an in-memory "adhoc database" that contains it. This adhoc database will be attached to the virtual database +* When a delimited file has a `.qsql` cache, it will attach to that file directly, without having to read it into memory +* When querying a standard sqlite3 file, it will be attached to the virtual database to it as well, without reading it into memory. sqlite3 files are auto-detected, no need for any special filename extension + +The user query will be executed directly on the virtual database, using the attached databases. + +sqlite3 itself has a limit on the number of attached databases (usually 10). If that limit is reached, q will automatically attach databases until that limit is reached, and will load additional tables into the adhoc database's in-memory database. + +Please make sure to read the [limitations](#limitations) section as well. + +## Development + +### Tests +The code includes a test suite runnable through `run-tests.sh`. By default, it uses the python source code for running the tests. However, it is possible to provide a path to an actual executable to the tests using the `Q_EXECUTABLE` env var. This is actually being used during the build and packaging process, in order to test the resulting binary. + +## Limitations +Here's the list of known limitations. Please contact me if you have a use case that needs any of those missing capabilities. + +* Common Table Expressions (CTE) are not supported for now. Will be implemented soon - See [here](https://github.com/harelba/q/issues/67) and [here](https://github.com/harelba/q/issues/124) for details. +* `FROM ` is not supported +* Spaces in file names are not supported. Use stdin for piping the data into q, or rename the file +* Some rare cases of subqueries are not supported yet. +* Queries with more than 10 different sqlite3 databases will load some data into memory +* up to 500 tables are supported in a single query + +## Rationale +Have you ever stared at a text file on the screen, hoping it would have been a database so you could ask anything you want about it? I had that feeling many times, and I've finally understood that it's not the database that I want. It's the language - SQL. + +SQL is a declarative language for data, and as such it allows me to define what I want without caring about how exactly it's done. This is the reason SQL is so powerful, because it treats data as data and not as bits and bytes (and chars). + +The goal of this tool is to provide a bridge between the world of text files and of SQL. + +### Why aren't other Linux tools enough? +The standard Linux tools are amazing and I use them all the time, but the whole idea of Linux is mixing-and-matching the best tools for each part of job. This tool adds the declarative power of SQL to the Linux toolset, without loosing any of the other tools' benefits. In fact, I often use q together with other Linux tools, the same way I pipe awk/sed and grep together all the time. + +One additional thing to note is that many Linux tools treat text as text and not as data. In that sense, you can look at q as a meta-tool which provides access to all the data-related tools that SQL provides (e.g. expressions, ordering, grouping, aggregation etc.). + +### Philosophy +This tool has been designed with general Linux/Unix design principles in mind. If you're interested in these general design principles, read this amazing [book](http://catb.org/~esr/writings/taoup/) and specifically [this part](http://catb.org/~esr/writings/taoup/html/ch01s06.html). If you believe that the way this tool works goes strongly against any of the principles, I would love to hear your view about it. + +## Future + +* Expose python as a python module - Planned as a goal after the new version `3.x` is out + + diff --git a/mkdocs/docs/index_cn.md b/mkdocs/docs/index_cn.md new file mode 100644 index 00000000..96cba778 --- /dev/null +++ b/mkdocs/docs/index_cn.md @@ -0,0 +1,367 @@ +# q - 直接在CSV或TSV文件上运行SQL + +[![GitHub Stars](https://img.shields.io/github/stars/harelba/q.svg?style=social&label=GitHub Stars&maxAge=600)](https://GitHub.com/harelba/q/stargazers/) +[![GitHub forks](https://img.shields.io/github/forks/harelba/q.svg?style=social&label=GitHub Forks&maxAge=600)](https://GitHub.com/harelba/q/network/) + + +## 概述 +q 是一个可以运行在 CSV / TSV 文件(或其他表格式的文本文件)上运行类SQL命令的命令行工具。 + +q 将普通文本（如上述）作为数据库表，且支持所有的SQL语法如：WHERE、GROUP BY、各种JOIN等。此外，还拥有自动识别列名和列类型及广泛支持多种编码的特性。 + +``` bash +q "SELECT COUNT(*) FROM ./clicks_file.csv WHERE c3 > 32.3" +``` + +``` bash +ps -ef | q -H "SELECT UID,COUNT(*) cnt FROM - GROUP BY UID ORDER BY cnt DESC LIMIT 3" +``` + +查看[示例](#示例)或[安装](#安装)体验. + +| | | +|:--------------------------------------:|:-----------------------------------------------:| +| 完全支持所有的字符编码 | すべての文字エンコーディングを完全にサポート | +| 모든 문자 인코딩이 완벽하게 지원됩니다 | все кодировки символов полностью поддерживаются | + + +**非英语用户:** q 完全支持所有类型的字符编码。 使用 `-e data-encoding` 设置输入编码; 使用 `-Q query-encoding` 设置查询编码; 使用 `-E output-encoding` 设置输出编码; +如上三个参数均设有合理的默认值。
+ +> 如果遇到问题请与我联系，期待与你交流。 + +**含有BOM的文件:** python的csv模块并不能很好的支持含有[Byte Order Mark](https://en.wikipedia.org/wiki/Byte_order_mark) 的文件。针对该种情况，使用 `-e utf-8-sig` 命令参数可读取包含BOM的UTF8编码文件。 + +> 我们计划将BOM相关处理与编码'解耦', 这样就可以支持所有编码的BOM文件了。 + +## 安装 + +| 格式 | 说明 | 备注 | +|:---|:---|:---| +|[OSX](https://github.com/harelba/q/releases/download/2.0.19/q-x86_64-Darwin)|运行 `brew install q`| 该方式暂不支持MAN手册, 可以使用 `q --help` 查看帮助|| +|[RPM Package](https://github.com/harelba/q/releases/download/2.0.19/q-text-as-data-2.0.19-1.x86_64.rpm)| 运行 `rpm -ivh ` 如果安装过旧版则运行 `rpm -U ` | 该方式支持MAN手册，可运行`man q`查看| +|[DEB Package](https://github.com/harelba/q/releases/download/2.0.19/q-text-as-data_2.0.19-2_amd64.deb)| 运行 `sudo dpkg -i `|该方式支持MAN手册，可运行`man q`查看| +|[Windows Installer](https://github.com/harelba/q/releases/download/2.0.19/q-AMD64-Windows-installer.exe)|运行安装可执行文件，一直点击下一步、下一步... q.exe 将被添加至PATH，以便于随处运行|PATH更新后并不会即时生效，重新打开cmd命令窗口便可| +|[tar.gz](https://github.com/harelba/q/archive/2.0.19.tar.gz)|最新稳定版的所有源码文件。提示，q.py 文件不能直接使用，因为它需要python依赖|| +|[zip](https://github.com/harelba/q/archive/2.0.19.zip)|最新稳定版的所有源码文件。提示，q.py 文件不能直接使用，因为它需要python依赖|| + +**旧版本可以在这儿[下载](https://github.com/harelba/packages-for-q) 。按理说不会有人愿意用旧版本，要是你计划使用旧版，希望能与你交流。** + +## 须知 +从`2.0.9`版本开始，不需要任何外部依赖。Python(3.7)和其他所需的库包含在了安装文件中且与系统隔离。 + +## 使用 + +``` bash +q "" + + 最简单的执行语句：q "SELECT * FROM myfile" 该语句会输出文件内容 +``` + +q 支持在表格式的文本上执行类SQL命令。它的初衷是为Linux命令行附加SQL的表达力且实现对文本数据的轻松访问。 + +类SQL的查询将*文件名(或标准输入流)看作表名*。查询语句会作为命令输入的一个参数（使用引号包裹)，如果将多个文件看作一张表，可以这样写 `文件名1+文件名2....`或者使用通配符（比如：`my_files*.csv`)。 + +使用 `-H` 表示输入内容中包含表头。该情况下列名会被自动识别，如果没有指定该参数，列名将会被以`cX`命名，`X`从1开始（比如: `q "SELECT c3,c8 from ..."`) 。 + +使用 `-d` 声明输入的分隔符。 + +列类型可由工具自动识别，无需强制转换。 提示，使用`--as-text` 可以强制将所有列类型转换为文本类型。 + +依据sqlite规范，如果列名中含有空格，需要使用反引号 (即：`) 引起来。 + +完全支持查询/输入/输出的编码设置（q 力争提供一种开箱即用的方法), 可以分别使用`-Q`,`-e` 和 `-E`来指定编码设置类型。 + +支持所有的sqlite3 SQL方法，包括文件之间的 JOIN（可以为文件设置别名）操作。在下面的[限制](#限制)小节可以看到一些少有使用的、欠支持的说明。 + +### 查询 + +q 的每一个参数都是由双引号包裹的一条完整的SQL语句。所有的查询语句会依次执行，最终结果以标准输出流形式输出。 提示，在同一命令行中执行多条查询语句时，仅在执行第一条查询语句时需要耗时载入数据，其他查询语句即时执行。 + +支持所有标准SQL语法，条件（WHERE 和 HAVING）、GROUP BY、ORDER BY等。 + +在WHERE条件查询中，支持JOIN操作和子查询，但在FROM子句中并不支持。JOIN操作时，可以为文件起别名。 + +SQL语法同sqlite的语法，详情见 https://www.sqlite.org/lang.html 或上网找一些示例。 + +**注意**： + +* 支持所有类型的自动识别，无需强制转换或其他操作。 + +* 如果重命名输出列，则需要为列指定别名并使用 `-O` 声明。如: `q -O -H "select count(*) cnt,sum(*) as mysum from -"` 便会将`cnt`和`mysum`作为列名输出。 + +### 指令 + +``` bash +使用: + q 支持在表格式的文本数据上执行类SQL查询。 + + 它的初衷是为Linux命令行附加SQL的表达力且实现对文本数据的轻松访问。 + + 基本操作是 q "SQL查询语句" 表名便是文件名（使用 - 从标注输入中读取数据）。若输入内容包含表头时，可以使用 -H 指定列名。若无表头，则列将会自动命名为 c1...cN。 + + 列类型可被自动识别。可以使用 -A 命令查看每列的名称及其类型。 + + 可以使用 -d (或 -t) 指定分隔符，使用 -D 指定输出分割符。 + + 支持所有的sqlite3 SQL方法。 + + 示例: + + 例子1: ls -ltrd * | q "select c1,count(1) from - group by c1" + 上例将会输出当前目录下，所有文件的权限表达式分组及每组数量。 + + 例子2: seq 1 1000 | q "select avg(c1),sum(c1) from -" + 上例将会输出1到1000的平均数与和数。 + + 例子3: sudo find /tmp -ls | q "select c5,c6,sum(c7)/1024.0/1024 as total from - group by c5,c6 order by total desc" + 上例将会输出在/tmp目录下，相同'用户+组'的文件所占用的MB磁盘空间。 + + 更多详情见 https://github.com/harelba/q/ 或查看帮助 + +选项： + -h, --help 显示此帮助信息并退出 + -v, --version 显示版本号 + -V, --verbose 出现问题时显示调试信息 + -S SAVE_DB_TO_DISK_FILENAME, --save-db-to-disk=SAVE_DB_TO_DISK_FILENAME + 将数据库保存为一个 sqlite 数据库文件 + --save-db-to-disk-method=SAVE_DB_TO_DISK_METHOD + 保存数据库到磁盘的方法 + 'standard' 不需要任何设置 + 'fast'需要手动在python的安装目录下执行`pip install sqlitebck` + 打包的问题解决后，'fast'即被作为默认方式 + 数据相关的选项: + + -H, --skip-header 忽略表头，在早期的版本中已修改为：仅支持用于标明列名的一行表头 + -d DELIMITER, --delimiter=DELIMITER + 列分隔符，若无特别指定，默认为空格符 + -p, --pipe-delimited + 作用同 -d '|'，为了方便和可读性提供该参数 + -t, --tab-delimited + 作用同 -d ，这仅是一种简写，也可以在Linux命令行中使用$'\t' + -e ENCODING, --encoding=ENCODING + 输入文件的编码，默认是UTF-8 + -z, --gzipped 压缩数据，对于从输入流读取文件非常高效 .gz 是自动压缩后文件扩展名 + -A, --analyze-only 简单分析：各列的数据类型 + -m MODE, --mode=MODE + 数据解析模式: 松散, 宽松和严格。在严格模式下必须指定 -c + --column-count 参数。 + -c COLUMN_COUNT, --column-count=COLUMN_COUNT + 当使用宽松或严格模式时，用于指定列的数量 + -k, --keep-leading-whitespace + 保留每列前的空格。为了使其开箱即用，默认去除了列前的空格 + 如果有需要，可以指定该参数 + --disable-double-double-quoting + 禁止一对双引号的转义。默认可以使用 "" 转义双引号 + 主要为了向后兼容 + --disable-escaped-double-quoting + 禁止转义双引号 + 默认可以在双引号字段中使用 \" 进行转义 + 主要为了向后兼容 + --as-text 不识别列类型（所有列被当作文本类型） + -w INPUT_QUOTING_MODE, --input-quoting-mode=INPUT_QUOTING_MODE + 输入内容的转义模式，可选值 all、minimal、none + 该参数稍有误导性，-W 指定输出内容的转义模式 + -M MAX_COLUMN_LENGTH_LIMIT, --max-column-length-limit=MAX_COLUMN_LENGTH_LIMIT + 设置列的最大长度 + -U, --with-universal-newlines + 设置通用换行符 + -U 参数当前仅适用于常规文件，输入流或.gz类文件暂不支持 + + 输出相关的选项: + -D OUTPUT_DELIMITER, --output-delimiter=OUTPUT_DELIMITER + 输出列间的分隔符 + 若未指定，则与 -d 指定的分隔符相同；若均为指定，则默认为空格符 + -P, --pipe-delimited-output + 同 -D '|' 为了方便和可读性提供该参数 + -T, --tab-delimited-output + 同 -D 这仅是一种简写，也可以在Linux命令行中使用$'\t' + -O, --output-header + 输出表头，输出的列名是由查询中指定的别名 + 如: 'select name FirstName, value1/value2 MyCalculation + from ...' 即使输入时未指定表头仍可使用该参数。 + -b, --beautify 美化输出结果，可能较慢... + -f FORMATTING, --formatting=FORMATTING + 格式化输出列 + 如格式X=fmt，Y=fmt等，上述中的X、Y是指第几列（如：1 表示 SELECT + 的第一列) + -E OUTPUT_ENCODING, --output-encoding=OUTPUT_ENCODING + 输出内容的编码，默认是 'none'，跟随系统或终端的编码 + -W OUTPUT_QUOTING_MODE, --output-quoting-mode=OUTPUT_QUOTING_MODE + 输出内容的转义模式，可选值 all、minimal、none + 该参数稍有误导性，-w 指定输入内容的转义模式 + -L, --list-user-functions + 列出所有内置函数 + + 查询相关的参数: + -q QUERY_FILENAME, --query-filename=QUERY_FILENAME + 指定文件名，由文件中读取查询语句。 + 该操作常与查询编码（使用 -Q)一同使用 + -Q QUERY_ENCODING, --query-encoding=QUERY_ENCODING + 查询编码(包含查询语句的文件编码) + 实验性参数，对该参数的意见可反馈 +``` + +## 示例 +下述 `-H` 参数的例子，表示文件中含有表头时使用该参数。 + +`-t` 参数是指定文件以 tab 作为分隔符的缩写（可以使用 `-d` 参数指定任意分隔符）。 + +为了清楚起见，查询关键字均使用大写，实际上关键字(如 SELECT、WHERE等)对大小写并不敏感。 + +示例目录: + +* [例1 - 统计指定列唯一值的数量](#1) +* [例2 - 数值条件过滤、排序并限制输出数](#2) +* [例3 - GROUP BY简单示例](#3) +* [例4 - GROUP BY进阶示例 (以时间格式分组)](#4) +* [例5 - 标准输入流作为输入](#5) +* [例6 - 使用表头中列名](#6) +* [例7 - JOIN 两个文件](#7) + +### 例1 +对指定字段（点击数据中的uuid）执行 COUNT DISTINCT + +``` bash +q -H -t "SELECT COUNT(DISTINCT(uuid)) FROM ./clicks.csv" +``` +输出: +``` bash +229 +``` + +### 例2 +过滤数值数据、排序并限制输出数量 + +注意：q 将其看作数值类型并对其进行数值过滤(数值比较而不是字符串比较) + +``` bash +q -H -t "SELECT request_id,score FROM ./clicks.csv WHERE score > 0.7 ORDER BY score DESC LIMIT 5" +``` +输出: +``` bash +2cfab5ceca922a1a2179dc4687a3b26e 1.0 +f6de737b5aa2c46a3db3208413a54d64 0.986665809568 +766025d25479b95a224bd614141feee5 0.977105183282 +2c09058a1b82c6dbcf9dc463e73eddd2 0.703255121794 +``` + +### 例3 +GROUP BY 简单示例 + +``` bash +q -t -H "SELECT hashed_source_machine,count(*) FROM ./clicks.csv GROUP BY hashed_source_machine" +``` +输出: +``` bash +47d9087db433b9ba.domain.com 400000 +``` + +### 例4 +GROUP BY进阶示例 (以时间格式分组) + +``` bash +q -t -H "SELECT strftime('%H:%M',date_time) hour_and_minute,count(*) FROM ./clicks.csv GROUP BY hour_and_minute" +``` +输出: +``` bash +07:00 138148 +07:01 140026 +07:02 121826 +``` + +### 例5 +标准输入流作为输入 + +计算 /tmp 目录下各 user/group 的占用空间大小 + +``` bash +sudo find /tmp -ls | q "SELECT c5,c6,sum(c7)/1024.0/1024 AS total FROM - GROUP BY c5,c6 ORDER BY total desc" +``` +输出: +``` bash +mapred hadoop 304.00390625 +root root 8.0431451797485 +smith smith 4.34389972687 +``` + +### 例6 +使用表头中列名 + +计算拥有进程数最多的前3位用户名及其数量 + +注意: 该查询中自动识别了列名 + +``` bash +ps -ef | q -H "SELECT UID,COUNT(*) cnt FROM - GROUP BY UID ORDER BY cnt DESC LIMIT 3" +``` +输出: +``` bash +root 152 +harel 119 +avahi 2 +``` + +### 例7 +JOIN 两个文件 + +如下命令中JOIN一个ls命令输出内容文件（exampledatafile) 和一个包含group_name、email两列字段的文件（group-emails-example)，每一邮件组均包含filename、email列, 为了输出简便，使用WHERE条件过滤出名为 ppp 的文件 + +``` bash +q "SELECT myfiles.c8,emails.c2 FROM exampledatafile myfiles JOIN group-emails-example emails ON (myfiles.c4 = emails.c1) WHERE myfiles.c8 = 'ppp'" +``` +输出: +``` bash +ppp dip.1@otherdomain.com +ppp dip.2@otherdomain.com +``` +可以看出 ppp 文件出现了两次，每次都匹配到了它所属的dip邮件组（如例中 dip.1@... / dip2@...)，可以在 `exampledatafile` 和 `group-emails-example` 文件中查看数据。 + +JOIN 的应用场景中也支持列名识别，在查询包含表头的文件时，只需指定 `-H` 参数即可。 + +## 声明 +为了避免引用外部依赖，当前是使用由Python编写的内存数据库实现的。当前是支持 SELECT 语句及 各种JOIN （ 目前仅在 WHERE 语句中支持子查询)。 +若想对数据进一步分析，可以使用 `--save-db-to-disk` 参数，以将结果输出为 sqlite 数据库文件，然后使用 `sqlite3` 语句来执行查询操作。 + +需要提示的是，当前并没有对数据量的大小进行检测和限制 - 也就是说，需要用户自己掌控文件大小。 + +请务必阅读[限制](#限制)小节。 + +## 开发 + +### 测试 +源码中包含了测试用例，可以通过 `test/test-all` 来执行。若想要提交 PR的话，一定先确保其均执行成功。 + +## 限制 +如下罗列了一些已知的限制，若你的使用场景中需要用到以下标明的限制，请联系我。 + +* 不支持 `FROM ` +* 不支持公用表表达式(CTE) +* 不支持文件名中包含空格 (可以将文件以标准输入流的方式输入 q 或重命名文件) +* 不支持较少用到的子查询 + +## 原理 +你是否曾经盯着屏幕上的文本文件发呆，希望它要是数据库就好了，这样就可以找出自己想要的内容？我曾有过很多次，最终顿悟。我想要的不是数据库，而是 SQL。 + +SQL 是一种面向数据声明的语言，它允许自定义数据内容而无需关心其执行过程。这也正是SQL强大之处，因为它对于数据'所见即所得'，而不是将数据看作字节码。 + +本工具的目的是：在文本文件和SQL之间搭建一座桥梁。 + +### 为什么其他Linux工具不能满足需求？ +传统的Linux工具库也很酷，我也经常使用它们， 但Linux的整体理念是为任一部分搭配最好的工具。本工具为传统Linux工具集新添了 SQL 族类工具，其他工具并不会失去本来优势。 +事实上，我也经常将 q 和其他Linux工具搭配使用，就如同使用管道将 awk/sed 和 grep 搭配使用一样。 + +另外需要注意的是,许多Linux工具就将文本看作文本，而不是数据。从这个意义上来讲，可以将 q 看作提供了 SQL 功能（如：表达式、排序、分组、聚合等）的元工具。 + +### 理念 + +本工具的设计遵从了 Linux/Unix 的传统设计原则。若你对这些设计原则感兴趣，可以阅读 [这本书](http://catb.org/~esr/writings/taoup/) ，尤其是书中 [这部分](http://catb.org/~esr/writings/taoup/html/ch01s06.html) +若你认为本工具工作方式与之背道而驰，愿洗耳恭听你的建议。 + +## 展望 + +* 主要方向：将其作为python的模块公开。 在公开之前，需要对处理标准输入流做一些内部API的完善。 +* 支持分布式以提高算力。 + + + diff --git a/mkdocs/docs/js/google-analytics.js b/mkdocs/docs/js/google-analytics.js new file mode 100644 index 00000000..186332e4 --- /dev/null +++ b/mkdocs/docs/js/google-analytics.js @@ -0,0 +1,56 @@ +// Monitor all download links in GA + +var dlCnt = 0; +var tocCnt = 0; + +function GAizeDownloadLink(a) { + var url = a.href; + var x = url.indexOf("?"); + if (x != -1) { + url = url.substr(0, x); + } + var url_test = url.match(/^http.*(archive\/|releases\/)(?.*)/); + if (url_test) { + a.event_action = url_test.groups.path; + console.log("Converting download link to be GA aware: " + url + " . download path is " + a.event_action); + dlCnt = dlCnt + 1; + a.onclick = function() { + console.log("Sending GA event for link" + url); + var that = this; + gtag('event','perform download', { 'event_category': 'Downloads', 'event_label': 'Download ' + this.event_action , 'value': 1 }); + setTimeout(function() { + location.href = that.href; + }, 500); + return false; + }; + } +} + +function GAizeTOCLink(l) { + tocCnt = tocCnt + 1; + l.onclick = function() { + url_test = l.href.match(/^https?:\/\/.+(#.*)$/i); + toc_name = url_test[1]; + var that = this; + console.log("Sending GA event for toc link " + this.href); + + gtag('event','navigate', { 'event_category': 'Navigation', 'event_label': 'go to ' + toc_name, 'value': 1 }); + setTimeout(function() { + location.href = that.href; + }, 250); + return false; + }; + +} + +window.onload = function() { + var anchors = document.getElementsByTagName('a'); + for (i = 0; i < anchors.length; i++) { + GAizeDownloadLink(anchors[i]); + } + var toc_links = document.querySelectorAll('div.md-sidebar[data-md-component=toc] a.md-nav__link'); + for (i = 0; i < toc_links.length; i++) { + GAizeTOCLink(toc_links[i]); + } + console.log("Converted " + dlCnt + " download links and " + tocCnt + " TOC links to be GA aware"); +} diff --git a/mkdocs/docs/stylesheets/extra.css b/mkdocs/docs/stylesheets/extra.css new file mode 100644 index 00000000..e90c3300 --- /dev/null +++ b/mkdocs/docs/stylesheets/extra.css @@ -0,0 +1,38 @@ + +div.md-content pre { + background-color: black; + color: #41FF00; +} + +.md-typeset code pre { + background-color: black; + color: #41FF00; +} + +.md-typeset p code { + color: rgba(0,0,0,.87); +} + +.md-typeset code.bash { + color: #41FF00; +} + +.md-typeset__scrollwrap { + text-align: center; +} + +.md-typeset .headerlink { + opacity: 50%; +} + +article.md-content__inner.md-typeset>p { + text-align: left; +} + +.md-nav__link[data-md-state=blur] { + color: rgba(0.3,0.5,0.4,.4) +} + +.md-nav__link[data-md-state=current] { + font-weight: 700; +} diff --git a/mkdocs/generate-web-site.sh b/mkdocs/generate-web-site.sh new file mode 100755 index 00000000..0014729b --- /dev/null +++ b/mkdocs/generate-web-site.sh @@ -0,0 +1,3 @@ +#!/bin/bash + +mkdocs build -c -s -d ./generated-site diff --git a/mkdocs/mkdocs.yml b/mkdocs/mkdocs.yml new file mode 100644 index 00000000..473af88c --- /dev/null +++ b/mkdocs/mkdocs.yml @@ -0,0 +1,49 @@ +site_name: q - Text as Data +site_url: https://harelba.github.io/q/ +repo_url: https://github.com/harelba/q +edit_uri: "" +site_description: Text as Data - q is a command line tool that allows direct execution of SQL-like queries on CSVs/TSVs (and any other tabular text files). +site_author: Harel Ben-Attia +copyright: 'Copyright © 2012-2019 Harel Ben-Attia' +google_analytics: + - "UA-48316355-1" + - "auto" +nav: + - Home: index.md + - 首页: index_cn.md + - About: about.md +theme: + name: material + language: 'en' + palette: + primary: purple + accent: amber + fonts: + text: 'Roboto' + code: 'Roboto Mono' + favicon: 'img/q-logo1.ico' + logo: 'img/q-logo1.ico' + custom_dir: 'theme' +extra: + social: + - type: 'github' + link: 'https://github.com/harelba' + - type: 'twitter' + link: 'https://twitter.com/harelba' + - type: 'linkedin' + link: 'https://www.linkedin.com/in/harelba' +extra_css: + - 'stylesheets/extra.css' +extra_javascript: + - 'js/google-analytics.js' +markdown_extensions: + - meta + - toc: + permalink: true + - tables + - fenced_code + - admonition + # - codehilite + + + diff --git a/mkdocs/requirements.txt b/mkdocs/requirements.txt new file mode 100644 index 00000000..b9a11148 --- /dev/null +++ b/mkdocs/requirements.txt @@ -0,0 +1,28 @@ +Click==7.0 +Deprecated==1.2.7 +Jinja2==2.10.3 +Markdown==3.1.1 +MarkupSafe==1.1.1 +PyGithub==1.45 +PyJWT==1.7.1 +PyYAML==5.3 +Pygments==2.5.2 +certifi==2019.11.28 +chardet==3.0.4 +htmlmin==0.1.12 +idna==2.8 +jsmin==2.2.2 +livereload==2.6.1 +mkdocs-bootstrap4==0.1.2 +mkdocs-bootswatch==1.0 +mkdocs-git-committers-plugin==0.1.8 +mkdocs-material==4.6.0 +mkdocs-minify-plugin==0.2.1 +mkdocs==1.0.4 +pep562==1.0 +pymdown-extensions==6.2.1 +requests==2.22.0 +six==1.14.0 +tornado==6.0.3 +urllib3==1.25.8 +wrapt==1.11.2 diff --git a/mkdocs/theme/main.html b/mkdocs/theme/main.html new file mode 100644 index 00000000..8fd1489e --- /dev/null +++ b/mkdocs/theme/main.html @@ -0,0 +1,28 @@ +{% extends "base.html" %} + +{% block analytics %} + +{% set analytics = config.google_analytics %} + + +{% endblock %} diff --git a/package-release b/package-release deleted file mode 100755 index 0bbe6044..00000000 --- a/package-release +++ /dev/null @@ -1,35 +0,0 @@ -#!/bin/bash - -set -e - -base_folder=$(dirname $0) -pushd ${base_folder} >/dev/null - -if [ $# -ne 1 ]; -then - echo "Usage: $(dirname $0) " - echo - echo "Note that the git tag must be pushed to github before doing this." - exit 1 -fi -TAG="$1" - -d=`pwd` -cid1=`docker run -i -d -v ${d}:/q q-text-as-data-rpm-builder:0.1` -cid2=`docker run -i -d -v ${d}:/q q-text-as-data-deb-builder:0.1` - -function kill_container { - tmp=`docker kill ${cid1} ${cid2}` -} -trap kill_container EXIT - -rm -rvf ${base_folder}/packages -mkdir -p ${base_folder}/packages - -sleep 1 -docker exec -it ${cid1} /bin/bash -i -c "/q/dist/create-rpm ${TAG}" - -docker cp ${cid1}:/q/dist/rpm_build_area/RPMS/noarch/q-text-as-data-${TAG}-1.el6.noarch.rpm ${base_folder}/packages/q-text-as-data-${TAG}-1.noarch.rpm - -docker exec -it ${cid2} /bin/bash -i -c "cd /q/packages && alien ./q-text-as-data-${TAG}-1.noarch.rpm" - diff --git a/prepare-benchmark-env b/prepare-benchmark-env new file mode 100755 index 00000000..81bc84a1 --- /dev/null +++ b/prepare-benchmark-env @@ -0,0 +1,44 @@ +#!/bin/bash + +set -e + +eval "$(pyenv init -)" +eval "$(pyenv virtualenv-init -)" + +source benchmark-config.sh + +if [ ! -f ./benchmark_data.tar.gz ]; +then + echo benchmark data not found. downloading it + curl "https://s3.amazonaws.com/harelba-q-public/benchmark_data.tar.gz" -o ./benchmark_data.tar.gz +else + echo no need to download benchmark data +fi + +if [ ! -d ./_benchmark_data ]; +then + echo extracting benchmark data + tar xvfz benchmark_data.tar.gz + echo benchmark data is ready +else + echo no need to extract benchmark data +fi + +for ver in "${BENCHMARK_PYTHON_VERSIONS[@]}" +do + echo installing $ver + pyenv install -s $ver + + venv_name=q-benchmark-$ver + echo create venv $venv_name + pyenv virtualenv -f $ver $venv_name + echo activate venv $venv_name + pyenv activate $venv_name + pyenv version + echo installing requirements $venv_name + pip install -r ./requirements.txt + echo deactivating $venv_name + pyenv deactivate +done + + diff --git a/pyoxidizer.bzl b/pyoxidizer.bzl new file mode 100644 index 00000000..1f7fb135 --- /dev/null +++ b/pyoxidizer.bzl @@ -0,0 +1,117 @@ +# This file defines how PyOxidizer application building and packaging is +# performed. See PyOxidizer's documentation at +# https://pyoxidizer.readthedocs.io/en/stable/ for details of this +# configuration file format. + +PYTHON_VERSION = VARS.get("PYTHON_VERSION","3.8") +Q_VERSION = VARS.get("Q_VERSION","0.0.1") + +# Configuration files consist of functions which define build "targets." +# This function creates a Python executable and installs it in a destination +# directory. +def make_exe(): + dist = default_python_distribution(python_version=PYTHON_VERSION) + + policy = dist.make_python_packaging_policy() + policy.set_resource_handling_mode("classify") + policy.resources_location = "in-memory" + policy.resources_location_fallback = "filesystem-relative:Lib" + policy.allow_in_memory_shared_library_loading = False + + python_config = dist.make_python_interpreter_config() + + python_config.run_module = "bin.q" + + exe = dist.to_python_executable( + name="q", + + packaging_policy=policy, + + config=python_config, + ) + + exe.pip_install(["wheel"]) + + exe.add_python_resources(exe.pip_install(["-r", "requirements.txt"])) + exe.add_python_resources(exe.pip_install(["-e", "."])) + + exe.add_python_resources(exe.read_package_root( + path="./", + packages=["bin"], + )) + + return exe + +def make_embedded_resources(exe): + return exe.to_embedded_resources() + +def make_install(exe): + # Create an object that represents our installed application file layout. + files = FileManifest() + + # Add the generated executable to our install layout in the root directory. + files.add_python_resource(".", exe) + + return files + +def make_msi(exe): + # See the full docs for more. But this will convert your Python executable + # into a `WiXMSIBuilder` Starlark type, which will be converted to a Windows + # .msi installer when it is built. + builder = exe.to_wix_msi_builder( + # Simple identifier of your app. + "q", + # The name of your application. + "q-text-as-data", + # The version of your application. + Q_VERSION, + # The author/manufacturer of your application. + "Harel Ben-Attia" + ) + return builder + + +# Dynamically enable automatic code signing. +def register_code_signers(): + # You will need to run with `pyoxidizer build --var ENABLE_CODE_SIGNING 1` for + # this if block to be evaluated. + if not VARS.get("ENABLE_CODE_SIGNING"): + return + + # Use a code signing certificate in a .pfx/.p12 file, prompting the + # user for its path and password to open. + # pfx_path = prompt_input("path to code signing certificate file") + # pfx_password = prompt_password( + # "password for code signing certificate file", + # confirm = True + # ) + # signer = code_signer_from_pfx_file(pfx_path, pfx_password) + + # Use a code signing certificate in the Windows certificate store, specified + # by its SHA-1 thumbprint. (This allows you to use YubiKeys and other + # hardware tokens if they speak to the Windows certificate APIs.) + # sha1_thumbprint = prompt_input( + # "SHA-1 thumbprint of code signing certificate in Windows store" + # ) + # signer = code_signer_from_windows_store_sha1_thumbprint(sha1_thumbprint) + + # Choose a code signing certificate automatically from the Windows + # certificate store. + # signer = code_signer_from_windows_store_auto() + + # Activate your signer so it gets called automatically. + # signer.activate() + + +# Call our function to set up automatic code signers. +register_code_signers() + +# Tell PyOxidizer about the build targets defined above. +register_target("exe", make_exe) +register_target("resources", make_embedded_resources, depends=["exe"], default_build_script=True) +register_target("install", make_install, depends=["exe"], default=True) +register_target("msi_installer", make_msi, depends=["exe"]) + +# Resolve whatever targets the invoker of this configuration file is requesting +# be resolved. +resolve_targets() diff --git a/pytest.ini b/pytest.ini new file mode 100644 index 00000000..57ea2be4 --- /dev/null +++ b/pytest.ini @@ -0,0 +1,3 @@ +[pytest] +markers = + benchmark: Benchmark tests diff --git a/requirements.txt b/requirements.txt new file mode 100644 index 00000000..3ad7d2bf --- /dev/null +++ b/requirements.txt @@ -0,0 +1,3 @@ +six==1.11.0 +flake8==3.6.0 +setuptools<45.0.0 diff --git a/run-benchmark b/run-benchmark new file mode 100755 index 00000000..f6556508 --- /dev/null +++ b/run-benchmark @@ -0,0 +1,110 @@ +#!/bin/bash + +# Usage: ./run-benchmark.sh +set -e + +get_abs_filename() { + # $1 : relative filename + echo "$(cd "$(dirname "$1")" && pwd)/$(basename "$1")" +} + +eval "$(pyenv init -)" +eval "$(pyenv virtualenv-init -)" + +if [ "x$1" == "x" ]; +then + echo Benchmark id must be provided as a parameter + exit 1 +fi +Q_BENCHMARK_ID=$1 +shift + +if [ "x$1" == "x" ]; +then + EFFECTIVE_Q_EXECUTABLE="source-files-$(git rev-parse HEAD)" +else + ABS_Q_EXECUTABLE="$(get_abs_filename $1)" + export Q_EXECUTABLE=$ABS_Q_EXECUTABLE + if [ ! -f $ABS_Q_EXECUTABLE ] + then + echo "q executable must exist ($ABS_Q_EXECUTABLE)" + exit 1 + fi + EFFECTIVE_Q_EXECUTABLE="${ABS_Q_EXECUTABLE//\//__}" + shift +fi + +echo "Q executable to use is $EFFECTIVE_Q_EXECUTABLE" + +PYTEST_OPTIONS="$@" +echo "pytest options are $PYTEST_OPTIONS" + +mkdir -p ./test/benchmark-results + +# Must be provided to the benchmark code so it knows where to write the results to +export Q_BENCHMARK_RESULTS_FOLDER="./test/benchmark-results/${EFFECTIVE_Q_EXECUTABLE}/${Q_BENCHMARK_ID}/" +echo Benchmark results folder is $Q_BENCHMARK_RESULTS_FOLDER +mkdir -p $Q_BENCHMARK_RESULTS_FOLDER + +source benchmark-config.sh +LATEST_PYTHON_VERSION=${BENCHMARK_PYTHON_VERSIONS[${#BENCHMARK_PYTHON_VERSIONS[@]}-1]} + +ALL_FILES=() + +for ver in "${BENCHMARK_PYTHON_VERSIONS[@]}" +do +venv_name=q-benchmark-$ver +echo activating $venv_name +pyenv activate $venv_name +echo "==== testing inside $venv_name ===" +if [[ -f $Q_BENCHMARK_RESULTS_FOLDER/${venv_name}.benchmark-results ]] +then + echo "Results files for version $ver already exists skipping benchmark for this version" + continue +fi + +export Q_BENCHMARK_NAME=${venv_name} +export Q_BENCHMARK_ADDITIONAL_PARAMS="-C read" + +Q_BENCHMARK_NAME=${venv_name}-with-caching Q_BENCHMARK_DATA_DIR=./_benchmark_data_with_qsql_caches pytest -m benchmark -k test_q_matrix -v -s $PYTEST_OPTIONS +Q_BENCHMARK_NAME=${venv_name} Q_BENCHMARK_DATA_DIR=./_benchmark_data pytest -m benchmark -k test_q_matrix -v -s $PYTEST_OPTIONS + +RESULT_FILE="${Q_BENCHMARK_RESULTS_FOLDER}/$venv_name.benchmark-results" +echo "==== Done. Results are in $RESULT_FILE" +ALL_FILES[${#ALL_FILES[@]}]="$RESULT_FILE" +echo "Deactivating" +pyenv deactivate +done + +exit 0 + +pyenv activate q-benchmark-${LATEST_PYTHON_VERSION} +echo "==== testing textql ===" +if [[ -f `ls $Q_BENCHMARK_RESULTS_FOLDER/textql*.benchmark-results` ]] +then + echo "Results files for textql already exist. Skipping benchmark for textql" +else + pytest -m benchmark -k test_textql_matrix -v -s $PYTEST_OPTIONS + RESULT_FILE="textql*.benchmark-results" + ALL_FILES[${#ALL_FILES[@]}]="${Q_BENCHMARK_RESULTS_FOLDER}/$RESULT_FILE" + echo "Done. Results are in textql.benchmark-results" +fi + +echo "==== testing octosql ===" +if [[ -f $Q_BENCHMARK_RESULTS_FOLDER/octosql.benchmark-results ]] +then + echo "Results files for octosql aready exist. Skipping benchmark for octosql" +else + pytest -m benchmark -k test_octosql_matrix -v -s $PYTEST_OPTIONS + RESULT_FILE="octosql*.benchmark-results" + ALL_FILES[${#ALL_FILES[@]}]="${Q_BENCHMARK_RESULTS_FOLDER}/$RESULT_FILE" + echo "Done. Results are in octosql.benchmark-results" +fi + +summary_file="$Q_BENCHMARK_RESULTS_FOLDER/summary.benchmark-results" + +rm -vf $summary_file + +paste ${ALL_FILES[*]} > $summary_file +echo "Done. final results file is $summary_file" +pyenv deactivate diff --git a/run-coverage.sh b/run-coverage.sh new file mode 100755 index 00000000..26ee0f90 --- /dev/null +++ b/run-coverage.sh @@ -0,0 +1,18 @@ +#!/bin/bash + +set -e + +rm -vf ./htmlcov/* + +pytest -m "not benchmark" --cov --cov-report html "$@" + +function cleanup() { + kill %1 +} + +# TODO Fix + +# python -m http.server 8000 +# open http://localhost:8000/htmlcov/ + + diff --git a/run-tests.sh b/run-tests.sh new file mode 100755 index 00000000..f28e57e6 --- /dev/null +++ b/run-tests.sh @@ -0,0 +1,3 @@ +#!/bin/bash + +pytest -m 'not benchmark' "$@" diff --git a/setup.py b/setup.py new file mode 100644 index 00000000..6d0fac7e --- /dev/null +++ b/setup.py @@ -0,0 +1,31 @@ +#!/usr/bin/env python + +from setuptools import setup +import setuptools + +q_version = '3.1.6' + +with open("README.markdown", "r", encoding="utf-8") as fh: + long_description = fh.read() + +setup( + name='q', + url='https://github.com/harelba/q', + license='LICENSE', + version=q_version, + author='Harel Ben-Attia', + description="Run SQL directly on CSV or TSV files", + long_description=long_description, + long_description_content_type="text/markdown", + author_email='harelba@gmail.com', + install_requires=[ + 'six==1.11.0' + ], + package_dir={"": "bin"}, + packages=setuptools.find_packages(where="bin"), + entry_points={ + 'console_scripts': [ + 'q = bin.q:run_standalone' + ] + } +) diff --git a/test-requirements.txt b/test-requirements.txt new file mode 100644 index 00000000..7e995bae --- /dev/null +++ b/test-requirements.txt @@ -0,0 +1,3 @@ +pytest==6.2.2 +flake8==3.6.0 +six==1.11.0 diff --git a/test/BENCHMARK.md b/test/BENCHMARK.md new file mode 100644 index 00000000..c8ec866d --- /dev/null +++ b/test/BENCHMARK.md @@ -0,0 +1,187 @@ + + +NOTE: *Please don't use or publish this benchmark data yet. See below for details* + +# Update +q now provides inherent automatic caching capabilities, writing the CSV/TSV file to a `.qsql` file that sits beside the original file. After the cache exists (created as part of an initial query on a file), q knows to use it behind the scenes without changing the query itself, speeding up performance significantly. + +The following table shows the impact of using caching in q: + +| Rows | Columns | File Size | Query time without caching | Query time with caching | Speed Improvement | +|:---------:|:-------:|:---------:|:--------------------------:|:-----------------------:|:-----------------:| +| 5,000,000 | 100 | 4.8GB | 4 minutes, 47 seconds | 1.92 seconds | x149 | +| 1,000,000 | 100 | 983MB | 50.9 seconds | 0.461 seconds | x110 | +| 1,000,000 | 50 | 477MB | 27.1 seconds | 0.272 seconds | x99 | +| 100,000 | 100 | 99MB | 5.2 seconds | 0.141 seconds | x36 | +| 100,000 | 50 | 48MB | 2.7 seconds | 0.105 seconds | x25 | + + +Effectively, `.qsql` files are just standard sqlite3 files, with an additional metadata table that is used for detecting changes in the original delimited file. This means that any tool that can read sqlite3 files can use these files directly. The tradeoff is of course the additional disk usage that the cache files take. + +A good side-effect to this addition, is that q now knows how to directly query multi-file sqlite3 databases. This means that the user can query any sqlite3 database file, or the `.qsql` file itself, even when the original file doesn't exist anymore. For example: + +```bash +q "select a.*,b.* from my_file.csv.qsql a left join some-sqlite3-database:::some_table_name b on (a.id = b.id)" +``` + +NOTE: In the current version, caching is not enabled by default - Use `-C readwrite` to enable reading+writing cache files, or `-C read` to just read any existing cache files. A `~/.qrc` file can be added in order to make these options the default if you want. + +The benchmark results below reflect the peformance without the caching, e.g. directly reading the delimited files, parsing them and performing the query. + +I'll update benchmark results later on to provide cached results as well. + +# Overview +This just a preliminary benchmark, originally created for validating performance optimizations and suggestions from users, and analyzing q's move to python3. After writing it, I thought it might be interesting to test its speed against textql and octosql as well. + +The results I'm getting are somewhat surprising, to the point of me questioning them a bit, so it would be great to validate the further before finalizing the benchmark results. + +The most surprising results are as follows: +* python3 vs python2 - A huge improvement (for large files, execution times with python 3 are around 40% of the times for python 2) +* python3 vs textql (written in golang) - Seems that textql becomes slower than the python3 q version as the data sizes grows (both rows and columns) + +I would love to validate these results by having other people run the benchmark as well and send me their results. + +If you're interested, follow the instructions and run the benchmark on your machine. After the benchmark is finished, send me the final results file, along with some details about your hardware, and i'll add it to the spreadsheet. + +I've tried to make running the benchmark as seamless as possible, but there obviously might be errors/issues. Please contact me if you encounter any issue, or just open a ticket. + +# Benchmark +This is an initial version of the benchmark, along with some results. The following is compared: +* q running on multiple python versions +* textql 2.0.3 +* octosql v0.3.0 + +The specific python versions which are being tested are specified in `benchmark-config.sh`. + +This is by no means a scientific benchmark, and it only focuses on the data loading time which is the only significant factor for comparison (e.g. the query itself is a very simple count query). Also, it does not try to provide any usability comparison between q and textql/octosql, an interesting topic on its own. + +## Methodology +The idea was to compare the time sensitivity of row and column count. + +* Row counts: 1,10,100,1000,10000,100000,1000000 +* Column counts: 1,5,10,20,50,100 +* Iterations for each combination: 10 + +File sizes: +* 1M rows by 100 columns - 976MB (~1GB) - Largest file +* 1M rows by 50 columns - 477MB + +The benchmark executes simple `select count(*) from ` queries for each combination, calculating the mean and stddev of each set of iterations. The stddev is used in order to measure the validity of the results. + +The graphs below only compare the means of the results, the standard deviations are written into the google sheet itself, and can be viewed there if needed. + +Instructions on how to run the benchmark are at the bottom section of this document, after the results section. + +## Hardware +OSX Catalina on a 15" Macbook Pro from Mid 2015, with 16GB of RAM, and an internal Flash Drive of 256GB. + +## Results +(Results are automatically updated from the baseline tab in the google spreadsheet). + +Detailed results below. + +Summary: +* All python 3 versions (3.6/3.7/3.8) provide similar results across all scales. +* python 3.x provides significantly better results than python2. Improvement grows as the file size grows (20% improvement for small files, up to ~70% improvement for the largest file) +* textql seems to provide faster results than q (py3) for smaller files, up to around 30MB of data. As the size grows further, it becomes slower than q, up to 80% (74 seconds vs 41 seconds) for the largest file +* The larger the files, textql becomes slower than q-py3 (up to 80% more time than q for the largest file) +* octosql is significantly slower than both q and textql, even for small files with a low number of rows and columns + +### Data for 1M rows + +#### Run time durations for 1M rows and different column counts: +| rows | columns | File Size | python 2.7 | python 3.6 | python 3.7 | python 3.8 | textql | octosql | +|:-------: |:-------: |:---------: |:----------: |:----------: |:----------: |:----------: |:------: |:-------: | +| 1000000 | 1 | 17M | 5.15 | 4.24 | 4.08 | 3.98 | 2.90 | 49.95 | +| 1000000 | 5 | 37M | 10.68 | 5.37 | 5.26 | 5.14 | 5.88 | 54.69 | +| 1000000 | 10 | 89M | 17.56 | 7.25 | 7.15 | 7.01 | 9.69 | 65.32 | +| 1000000 | 20 | 192M | 30.28 | 10.96 | 10.78 | 10.64 | 17.34 | 83.94 | +| 1000000 | 50 | 477M | 71.56 | 21.98 | 21.59 | 21.70 | 38.57 | 158.26 | +| 1000000 | 100 | 986M | 131.86 | 41.71 | 40.82 | 41.02 | 74.62 | 289.58 | + +#### Comparison between python 3.x and python 2 run times (1M rows): +(>100% is slower than q-py2, <100% is faster than q-py2) + +| rows | columns | file size | q-py2 runtime | q-py3.6 vs q-py2 runtime | q-py3.7 vs q-py2 runtime | q-py3.8 vs q-py2 runtime | +|:-------: |:-------: |:---------: |:-------------: |:------------------------: |:------------------------: |:------------------------: | +| 1000000 | 1 | 17M | 100.00% | 82.34% | 79.34% | 77.36% | +| 1000000 | 5 | 37M | 100.00% | 50.25% | 49.22% | 48.08% | +| 1000000 | 10 | 89M | 100.00% | 41.30% | 40.69% | 39.93% | +| 1000000 | 20 | 192M | 100.00% | 36.18% | 35.59% | 35.14% | +| 1000000 | 50 | 477M | 100.00% | 30.71% | 30.17% | 30.32% | +| 1000000 | 100 | 986M | 100.00% | 31.63% | 30.96% | 31.11% | + +#### textql and octosql comparison against q-py3 run time (1M rows): +(>100% is slower than q-py3, <100% is faster than q-py3) + +| rows | columns | file size | avg q-py3 runtime | textql vs q-py3 runtime | octosql vs q-py3 runtime | +|:-------: |:-------: |:---------: |:-----------------: |:-----------------------: |:------------------------: | +| 1000000 | 1 | 17M | 100.00% | 70.67% | 1217.76% | +| 1000000 | 5 | 37M | 100.00% | 111.86% | 1040.70% | +| 1000000 | 10 | 89M | 100.00% | 135.80% | 915.28% | +| 1000000 | 20 | 192M | 100.00% | 160.67% | 777.92% | +| 1000000 | 50 | 477M | 100.00% | 177.26% | 727.40% | +| 1000000 | 100 | 986M | 100.00% | 181.19% | 703.15% | + +### Sensitivity to column count +Based on a the largest file size of 1,000,000 rows. + +![Sensitivity to column count](https://docs.google.com/spreadsheets/d/e/2PACX-1vQy9Zm4I322Tdf5uoiFFJx6Oi3Z4AMq7He3fUUtsEQVQIdTGfWgjxFD6k8PAy9wBjvFkqaG26oBgNTP/pubchart?oid=1585602598&format=image) + +### Sensitivity to line count (per column count) + +#### 1 Column Table +![1 column table](https://docs.google.com/spreadsheets/d/e/2PACX-1vQy9Zm4I322Tdf5uoiFFJx6Oi3Z4AMq7He3fUUtsEQVQIdTGfWgjxFD6k8PAy9wBjvFkqaG26oBgNTP/pubchart?oid=1119350798&format=image) + +#### 5 Column Table +![5 column table](https://docs.google.com/spreadsheets/d/e/2PACX-1vQy9Zm4I322Tdf5uoiFFJx6Oi3Z4AMq7He3fUUtsEQVQIdTGfWgjxFD6k8PAy9wBjvFkqaG26oBgNTP/pubchart?oid=599223098&format=image) + +#### 10 Column Table +![10 column table](https://docs.google.com/spreadsheets/d/e/2PACX-1vQy9Zm4I322Tdf5uoiFFJx6Oi3Z4AMq7He3fUUtsEQVQIdTGfWgjxFD6k8PAy9wBjvFkqaG26oBgNTP/pubchart?oid=82695414&format=image) + +#### 20 Column Table +![20 column table](https://docs.google.com/spreadsheets/d/e/2PACX-1vQy9Zm4I322Tdf5uoiFFJx6Oi3Z4AMq7He3fUUtsEQVQIdTGfWgjxFD6k8PAy9wBjvFkqaG26oBgNTP/pubchart?oid=1573199483&format=image) + +#### 50 Column Table +![50 column table](https://docs.google.com/spreadsheets/d/e/2PACX-1vQy9Zm4I322Tdf5uoiFFJx6Oi3Z4AMq7He3fUUtsEQVQIdTGfWgjxFD6k8PAy9wBjvFkqaG26oBgNTP/pubchart?oid=448568670&format=image) + +#### 100 Column Table +![100 column table](https://docs.google.com/spreadsheets/d/e/2PACX-1vQy9Zm4I322Tdf5uoiFFJx6Oi3Z4AMq7He3fUUtsEQVQIdTGfWgjxFD6k8PAy9wBjvFkqaG26oBgNTP/pubchart?oid=2101488258&format=image) + +## Running the benchmark +Please note that the initial run generates large files, so you'd need more than 3GB of free space available. All the generated files reside in the `_benchmark_data/` folder. + +Part of the preparation flow will download the benchmark data as needed. + +### Preparations +* Prerequisites: + * pyenv installed + * pyenv-virtualenv installed + * [`textql`](https://github.com/dinedal/textql#install) + * [`octosql`](https://github.com/cube2222/octosql#installation) + +Run `./prepare-benchmark-env` + +### Execution +Run `./run-benchmark `. + +Benchmark output files will be written to `./benchmark-results///`. + +* `benchmark-id` is the id you wanna give the benchmark. +* `q-executable` is the name of the q executable being used for the benchmark. If none has been provided through Q_EXECUTABLE, then the value will be the last commit hash. Note that there is no checking of whether the working tree is clean. + +The summary of benchmark will be written to `./benchmark-results//summary.benchmark-results`` + +By default, the benchmark will use the source python files inside the project. If you wanna run it on one of the standalone binary executable, the set Q_EXECUTABLE to the full path of the q binary. + +For anyone helping with running the benchmark, don't use this parameter for now, just test against a clean checkout of the code using `./run-benchmark `. + +## Benchmark Development info +### Running against the standalone binary +* `./run-benchmark` can accept a second parameter with the q executable. If it gets this parameter, it will use this path for running q. This provides a way to test the standalone q binaries in the new packaging format. When this parameter does not exist, the benchmark is executed directly from the source code. + +### Updating the benchmark markdown document file +The results should reside in the following [google sheet](https://docs.google.com/spreadsheets/d/1Ljr8YIJwUQ5F4wr6ATga5Aajpu1CvQp1pe52KGrLkbY/edit?usp=sharing). + +add a new tab to the google sheet, and paste the content of `summary.benchmark-results` to the new sheet. + diff --git a/test/__init__.py b/test/__init__.py new file mode 100644 index 00000000..e69de29b diff --git a/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/octosql_v0.3.0.benchmark-results b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/octosql_v0.3.0.benchmark-results new file mode 100644 index 00000000..ced04856 --- /dev/null +++ b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/octosql_v0.3.0.benchmark-results @@ -0,0 +1,48 @@ +lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 1 0.582091641426 0.0235290239617 +10 1 0.596219730377 0.0320124029461 +100 1 0.575977492332 0.0199296245316 +1000 1 0.56785056591 0.00846389017466 +10000 1 1.1466334343 0.00760108698846 +100000 1 5.49565172195 0.131791932977 +1000000 1 49.9513648033 0.443430523063 +lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 5 0.582160949707 0.0274409391571 +10 5 0.57046456337 0.0199413000359 +100 5 0.585747480392 0.0372543971623 +1000 5 0.572268772125 0.00384300349763 +10000 5 1.15530762672 0.0117990775856 +100000 5 6.10629923344 0.146711842919 +1000000 5 54.6851765394 0.315486399525 +lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 10 0.586222410202 0.0232479065914 +10 10 0.59000480175 0.0186508192447 +100 10 0.581873703003 0.0331332482772 +1000 10 0.569027900696 0.0103675493106 +10000 10 1.40067322254 0.00583352224401 +100000 10 7.30705575943 0.0165839217599 +1000000 10 65.3242264032 0.512552576414 +lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 20 0.571048212051 0.0166919396871 +10 20 0.594776701927 0.0368900941023 +100 20 0.561370825768 0.00907051791451 +1000 20 0.577527880669 0.00983965108957 +10000 20 1.90710241795 0.00757011452155 +100000 20 9.8267291069 0.127844155326 +1000000 20 83.9448960066 0.46121344046 +lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 50 0.572030115128 0.0253648479103 +10 50 0.56993534565 0.0230474303306 +100 50 0.563336873055 0.00964411866903 +1000 50 0.826378440857 0.00941629472813 +10000 50 3.27872717381 0.126592845956 +100000 50 17.890055728 0.116794666005 +1000000 50 158.262442636 0.826290454446 +lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 100 0.569358110428 0.0279801762531 +10 100 0.580981063843 0.0272341107532 +100 100 0.559471726418 0.00668155858429 +1000 100 1.08161640167 0.00698594638512 +10000 100 5.67823712826 0.0123398407167 +100000 100 32.2797194242 0.315508270241 +1000000 100 289.582628798 0.929455236817 diff --git a/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-2.7.18.benchmark-results b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-2.7.18.benchmark-results new file mode 100644 index 00000000..5b7aa05a --- /dev/null +++ b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-2.7.18.benchmark-results @@ -0,0 +1,48 @@ +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev +1 1 0.106449890137 0.002010027753 +10 1 0.106737875938 0.00224112203891 +100 1 0.107839012146 0.00102954061006 +1000 1 0.113026666641 0.00147361890226 +10000 1 0.160376381874 0.00569766179806 +100000 1 0.608236479759 0.00604026519608 +1000000 1 5.14807910919 0.0584474028762 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev +1 5 0.106719517708 0.00236752032369 +10 5 0.107823801041 0.00238873169438 +100 5 0.109785079956 0.0013047675259 +1000 5 0.120395207405 0.00207224422629 +10000 5 0.21783041954 0.00522254475716 +100000 5 1.17115747929 0.0221394865225 +1000000 5 10.6830974817 0.339822977934 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev +1 10 0.104981088638 0.00166552032929 +10 10 0.108320140839 0.00204034349199 +100 10 0.112528729439 0.00168376477305 +1000 10 0.13019015789 0.00253773120965 +10000 10 0.284891676903 0.00384009140782 +100000 10 1.84725661278 0.00860738744089 +1000000 10 17.5610994339 0.228322442172 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev +1 20 0.106477689743 0.00254429925697 +10 20 0.108580899239 0.00173704653824 +100 20 0.118750286102 0.00247623639866 +1000 20 0.146431708336 0.00249685551944 +10000 20 0.419492387772 0.00248210434668 +100000 20 3.15847921371 0.0550301268026 +1000000 20 30.279082489 0.124978814506 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev +1 50 0.105411934853 0.00171651054128 +10 50 0.109102797508 0.00111620290512 +100 50 0.135682177544 0.00196166766665 +1000 50 0.198261427879 0.00396172489054 +10000 50 0.821499919891 0.0111642692132 +100000 50 7.05980975628 0.121182371277 +1000000 50 71.5645889759 5.02009516291 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev +1 100 0.10662381649 0.00193146624495 +10 100 0.110662698746 0.00171461379583 +100 100 0.163547992706 0.00166570196628 +1000 100 0.280023741722 0.00337543024145 +10000 100 1.46053376198 0.0221691284465 +100000 100 13.2369835854 0.309375896258 +1000000 100 131.864977288 1.22415449691 diff --git a/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-3.6.4.benchmark-results b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-3.6.4.benchmark-results new file mode 100644 index 00000000..e611b7a5 --- /dev/null +++ b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-3.6.4.benchmark-results @@ -0,0 +1,48 @@ +lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev +1 1 0.10342762470245362 0.0017673875851759295 +10 1 0.10239293575286865 0.0012505611685910795 +100 1 0.10317318439483643 0.0010581783881541751 +1000 1 0.10687050819396973 0.0014050135772919004 +10000 1 0.1447664737701416 0.001841256227287192 +100000 1 0.5162809371948243 0.006962985088492867 +1000000 1 4.238853335380554 0.04834401143632507 +lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev +1 5 0.10211825370788574 0.0022568191323651568 +10 5 0.1025341272354126 0.0016446470901070106 +100 5 0.1053577184677124 0.0015298114223855884 +1000 5 0.10980842113494874 0.002536098780902228 +10000 5 0.1590113162994385 0.003123074098301634 +100000 5 0.6348223447799682 0.0082691507829872 +1000000 5 5.368562030792236 0.11628913334105236 +lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev +1 10 0.10251858234405517 0.0015963869535345293 +10 10 0.10278875827789306 0.0009920577082124496 +100 10 0.10715732574462891 0.002033320000941064 +1000 10 0.11389360427856446 0.0023603847702423973 +10000 10 0.17806434631347656 0.001114054252191835 +100000 10 0.8252989768981933 0.0037080843359275904 +1000000 10 7.252838873863221 0.029052130546213153 +lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev +1 20 0.10367965698242188 0.003661761341842434 +10 20 0.10489590167999267 0.001977141196109372 +100 20 0.11108210086822509 0.0014801173497056886 +1000 20 0.12110791206359864 0.001648524669420912 +10000 20 0.2178968906402588 0.0019298316207276716 +100000 20 1.1962245225906372 0.010541407803235559 +1000000 20 10.956057572364807 0.12677108174061705 +lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev +1 50 0.10458300113677979 0.0016367630302744722 +10 50 0.10616152286529541 0.002345135740908088 +100 50 0.12375867366790771 0.00238414904864133 +1000 50 0.14462883472442628 0.0022428030896492978 +10000 50 0.34488487243652344 0.004867441221052092 +100000 50 2.3394312858581543 0.02263239858944125 +1000000 50 21.979821610450745 0.09080404939303836 +lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev +1 100 0.10372309684753418 0.0010299126833031144 +10 100 0.10784556865692138 0.0016557634029464607 +100 100 0.14526791572570802 0.0028194506905186724 +1000 100 0.18315494060516357 0.0023585311962114673 +10000 100 0.5586131334304809 0.004808492789681402 +100000 100 4.287398314476013 0.00957500108409644 +1000000 100 41.706851434707644 0.4161526076289425 diff --git a/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-3.7.9.benchmark-results b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-3.7.9.benchmark-results new file mode 100644 index 00000000..7a1f7715 --- /dev/null +++ b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-3.7.9.benchmark-results @@ -0,0 +1,48 @@ +lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev +1 1 0.08099310398101807 0.001417385651688644 +10 1 0.0822291374206543 0.0014809900020001858 +100 1 0.08169686794281006 0.002108157069167563 +1000 1 0.08690853118896484 0.0012595326919263487 +10000 1 0.12215542793273926 0.0020152625320395434 +100000 1 0.4825761795043945 0.0050418000028856335 +1000000 1 4.084399747848511 0.027731958079814215 +lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev +1 5 0.0817826271057129 0.002665533758836163 +10 5 0.08261749744415284 0.0019205430658525572 +100 5 0.08472237586975098 0.002571239449841039 +1000 5 0.08973510265350342 0.002323797583077552 +10000 5 0.13746986389160157 0.001964971666036654 +100000 5 0.60649254322052 0.007131635266871318 +1000000 5 5.2585612535476685 0.05661789407928516 +lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev +1 10 0.08112843036651611 0.002251300165899426 +10 10 0.08175232410430908 0.0014557171018568637 +100 10 0.08572309017181397 0.0019643550214810675 +1000 10 0.09268453121185302 0.001816414236580489 +10000 10 0.15538835525512695 0.0024978076091814994 +100000 10 0.7879442930221557 0.009412516078916211 +1000000 10 7.146207928657532 0.06659760176757985 +lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev +1 20 0.08142082691192627 0.001304584466639188 +10 20 0.08197519779205323 0.0014842098503865223 +100 20 0.08949971199035645 0.0009937446141285785 +1000 20 0.09955930709838867 0.0013978961740806384 +10000 20 0.1966566801071167 0.0028489273218240147 +100000 20 1.1518636226654053 0.006410720031542237 +1000000 20 10.776052689552307 0.04739925571001746 +lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev +1 50 0.08237688541412354 0.0016494314799953837 +10 50 0.08519520759582519 0.002610550182895596 +100 50 0.10423583984375 0.0018808335751867933 +1000 50 0.12195603847503662 0.0023611894043373983 +10000 50 0.3163540124893188 0.002761333651520998 +100000 50 2.237372374534607 0.009955353920396077 +1000000 50 21.59097549915314 0.081188190530421 +lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev +1 100 0.08336784839630126 0.0013840724401561887 +10 100 0.0864112138748169 0.0017946939354350697 +100 100 0.12199611663818359 0.0013003743156634682 +1000 100 0.15871686935424806 0.0035993681064501234 +10000 100 0.5243751525878906 0.004370273273595629 +100000 100 4.175828623771667 0.016127303710583043 +1000000 100 40.82292411327362 0.12328165162380703 diff --git a/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-3.8.5.benchmark-results b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-3.8.5.benchmark-results new file mode 100644 index 00000000..ca8c87ad --- /dev/null +++ b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/q-benchmark-3.8.5.benchmark-results @@ -0,0 +1,48 @@ +lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev +1 1 0.10138180255889892 0.0017947074090971444 +10 1 0.10056869983673096 0.003442371291904885 +100 1 0.10126984119415283 0.0016392348107127808 +1000 1 0.10484635829925537 0.0019743937339163262 +10000 1 0.1400548219680786 0.0024523366133394117 +100000 1 0.4901275157928467 0.003970374711691596 +1000000 1 3.982502889633179 0.045292138461945054 +lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev +1 5 0.09946837425231933 0.0018876161478998787 +10 5 0.099178147315979 0.0014194733014858227 +100 5 0.10171806812286377 0.0017580984705406846 +1000 5 0.10602672100067138 0.002000261880840017 +10000 5 0.15207929611206056 0.0015802680033212048 +100000 5 0.609218978881836 0.006150144273259608 +1000000 5 5.13688440322876 0.03649575898109647 +lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev +1 10 0.09925477504730225 0.002168389758635997 +10 10 0.09943633079528809 0.0016154501074880502 +100 10 0.10376312732696533 0.0017275485891005433 +1000 10 0.11087138652801513 0.0016934328033239559 +10000 10 0.17246220111846924 0.0023824485659318527 +100000 10 0.7999232530593872 0.003442975393506892 +1000000 10 7.012071299552917 0.059217904448851263 +lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev +1 20 0.10027089118957519 0.0020291529595204906 +10 20 0.10038816928863525 0.001957086760826999 +100 20 0.10723590850830078 0.0013833918448622436 +1000 20 0.11735000610351562 0.0020318895390750882 +10000 20 0.21264209747314453 0.00482341642419078 +100000 20 1.1567201137542724 0.002987096441878969 +1000000 20 10.640758633613586 0.06116581724028616 +lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev +1 50 0.10066506862640381 0.002051307639276982 +10 50 0.10588631629943848 0.0035835389655972105 +100 50 0.11841504573822022 0.001608174845404568 +1000 50 0.14032282829284667 0.002640027148889162 +10000 50 0.33160474300384524 0.0027796660009712947 +100000 50 2.258401036262512 0.011041280982383895 +1000000 50 21.70080256462097 0.15897944629180621 +lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev +1 100 0.10147004127502442 0.0021285682695135768 +10 100 0.10471885204315186 0.001248479289219899 +100 100 0.13894760608673096 0.002307980025026551 +1000 100 0.17586205005645753 0.0023822296091426 +10000 100 0.5414002418518067 0.0036291866664635458 +100000 100 4.222555088996887 0.08562968951916528 +1000000 100 41.021552324295044 0.16033566363076862 diff --git a/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/summary.benchmark-results b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/summary.benchmark-results new file mode 100644 index 00000000..dcb1d280 --- /dev/null +++ b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/summary.benchmark-results @@ -0,0 +1,48 @@ +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev lines columns textql_2.0.3_mean textql_2.0.3_stddev lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 1 0.106449890137 0.002010027753 1 1 0.10342762470245362 0.0017673875851759295 1 1 0.08099310398101807 0.001417385651688644 1 1 0.10138180255889892 0.0017947074090971444 1 1 0.0196103572845 0.00207355214257 1 1 0.582091641426 0.0235290239617 +10 1 0.106737875938 0.00224112203891 10 1 0.10239293575286865 0.0012505611685910795 10 1 0.0822291374206543 0.0014809900020001858 10 1 0.10056869983673096 0.003442371291904885 10 1 0.0186784029007 0.000970810220668 10 1 0.596219730377 0.0320124029461 +100 1 0.107839012146 0.00102954061006 100 1 0.10317318439483643 0.0010581783881541751 100 1 0.08169686794281006 0.002108157069167563 100 1 0.10126984119415283 0.0016392348107127808 100 1 0.019472026825 0.00181951524514 100 1 0.575977492332 0.0199296245316 +1000 1 0.113026666641 0.00147361890226 1000 1 0.10687050819396973 0.0014050135772919004 1000 1 0.08690853118896484 0.0012595326919263487 1000 1 0.10484635829925537 0.0019743937339163262 1000 1 0.022180891037 0.00116649968967 1000 1 0.56785056591 0.00846389017466 +10000 1 0.160376381874 0.00569766179806 10000 1 0.1447664737701416 0.001841256227287192 10000 1 0.12215542793273926 0.0020152625320395434 10000 1 0.1400548219680786 0.0024523366133394117 10000 1 0.051066827774 0.0018168767618 10000 1 1.1466334343 0.00760108698846 +100000 1 0.608236479759 0.00604026519608 100000 1 0.5162809371948243 0.006962985088492867 100000 1 0.4825761795043945 0.0050418000028856335 100000 1 0.4901275157928467 0.003970374711691596 100000 1 0.307463979721 0.00246268029188 100000 1 5.49565172195 0.131791932977 +1000000 1 5.14807910919 0.0584474028762 1000000 1 4.238853335380554 0.04834401143632507 1000000 1 4.084399747848511 0.027731958079814215 1000000 1 3.982502889633179 0.045292138461945054 1000000 1 2.89862303734 0.022182722976 1000000 1 49.9513648033 0.443430523063 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev lines columns textql_2.0.3_mean textql_2.0.3_stddev lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 5 0.106719517708 0.00236752032369 1 5 0.10211825370788574 0.0022568191323651568 1 5 0.0817826271057129 0.002665533758836163 1 5 0.09946837425231933 0.0018876161478998787 1 5 0.0195286750793 0.0017840569109 1 5 0.582160949707 0.0274409391571 +10 5 0.107823801041 0.00238873169438 10 5 0.1025341272354126 0.0016446470901070106 10 5 0.08261749744415284 0.0019205430658525572 10 5 0.099178147315979 0.0014194733014858227 10 5 0.0183676958084 0.000925251595491 10 5 0.57046456337 0.0199413000359 +100 5 0.109785079956 0.0013047675259 100 5 0.1053577184677124 0.0015298114223855884 100 5 0.08472237586975098 0.002571239449841039 100 5 0.10171806812286377 0.0017580984705406846 100 5 0.0199447393417 0.000907007099218 100 5 0.585747480392 0.0372543971623 +1000 5 0.120395207405 0.00207224422629 1000 5 0.10980842113494874 0.002536098780902228 1000 5 0.08973510265350342 0.002323797583077552 1000 5 0.10602672100067138 0.002000261880840017 1000 5 0.0263328790665 0.00165486505938 1000 5 0.572268772125 0.00384300349763 +10000 5 0.21783041954 0.00522254475716 10000 5 0.1590113162994385 0.003123074098301634 10000 5 0.13746986389160157 0.001964971666036654 10000 5 0.15207929611206056 0.0015802680033212048 10000 5 0.0826982736588 0.00152451583229 10000 5 1.15530762672 0.0117990775856 +100000 5 1.17115747929 0.0221394865225 100000 5 0.6348223447799682 0.0082691507829872 100000 5 0.60649254322052 0.007131635266871318 100000 5 0.609218978881836 0.006150144273259608 100000 5 0.60660867691 0.00395761320274 100000 5 6.10629923344 0.146711842919 +1000000 5 10.6830974817 0.339822977934 1000000 5 5.368562030792236 0.11628913334105236 1000000 5 5.2585612535476685 0.05661789407928516 1000000 5 5.13688440322876 0.03649575898109647 1000000 5 5.87811236382 0.0304332294491 1000000 5 54.6851765394 0.315486399525 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev lines columns textql_2.0.3_mean textql_2.0.3_stddev lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 10 0.104981088638 0.00166552032929 1 10 0.10251858234405517 0.0015963869535345293 1 10 0.08112843036651611 0.002251300165899426 1 10 0.09925477504730225 0.002168389758635997 1 10 0.0191783189774 0.00107718516178 1 10 0.586222410202 0.0232479065914 +10 10 0.108320140839 0.00204034349199 10 10 0.10278875827789306 0.0009920577082124496 10 10 0.08175232410430908 0.0014557171018568637 10 10 0.09943633079528809 0.0016154501074880502 10 10 0.0185215950012 0.000840353961363 10 10 0.59000480175 0.0186508192447 +100 10 0.112528729439 0.00168376477305 100 10 0.10715732574462891 0.002033320000941064 100 10 0.08572309017181397 0.0019643550214810675 100 10 0.10376312732696533 0.0017275485891005433 100 10 0.0209223031998 0.00164494657684 100 10 0.581873703003 0.0331332482772 +1000 10 0.13019015789 0.00253773120965 1000 10 0.11389360427856446 0.0023603847702423973 1000 10 0.09268453121185302 0.001816414236580489 1000 10 0.11087138652801513 0.0016934328033239559 1000 10 0.0309282779694 0.00110848590345 1000 10 0.569027900696 0.0103675493106 +10000 10 0.284891676903 0.00384009140782 10000 10 0.17806434631347656 0.001114054252191835 10000 10 0.15538835525512695 0.0024978076091814994 10000 10 0.17246220111846924 0.0023824485659318527 10000 10 0.121016025543 0.00105071105139 10000 10 1.40067322254 0.00583352224401 +100000 10 1.84725661278 0.00860738744089 100000 10 0.8252989768981933 0.0037080843359275904 100000 10 0.7879442930221557 0.009412516078916211 100000 10 0.7999232530593872 0.003442975393506892 100000 10 0.987622976303 0.00699348302979 100000 10 7.30705575943 0.0165839217599 +1000000 10 17.5610994339 0.228322442172 1000000 10 7.252838873863221 0.029052130546213153 1000000 10 7.146207928657532 0.06659760176757985 1000000 10 7.012071299552917 0.059217904448851263 1000000 10 9.69240145683 0.0354453778052 1000000 10 65.3242264032 0.512552576414 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev lines columns textql_2.0.3_mean textql_2.0.3_stddev lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 20 0.106477689743 0.00254429925697 1 20 0.10367965698242188 0.003661761341842434 1 20 0.08142082691192627 0.001304584466639188 1 20 0.10027089118957519 0.0020291529595204906 1 20 0.0202306985855 0.00159619251952 1 20 0.571048212051 0.0166919396871 +10 20 0.108580899239 0.00173704653824 10 20 0.10489590167999267 0.001977141196109372 10 20 0.08197519779205323 0.0014842098503865223 10 20 0.10038816928863525 0.001957086760826999 10 20 0.0187650680542 0.000845692486156 10 20 0.594776701927 0.0368900941023 +100 20 0.118750286102 0.00247623639866 100 20 0.11108210086822509 0.0014801173497056886 100 20 0.08949971199035645 0.0009937446141285785 100 20 0.10723590850830078 0.0013833918448622436 100 20 0.0211876153946 0.000993808448942 100 20 0.561370825768 0.00907051791451 +1000 20 0.146431708336 0.00249685551944 1000 20 0.12110791206359864 0.001648524669420912 1000 20 0.09955930709838867 0.0013978961740806384 1000 20 0.11735000610351562 0.0020318895390750882 1000 20 0.0404737234116 0.00122415059261 1000 20 0.577527880669 0.00983965108957 +10000 20 0.419492387772 0.00248210434668 10000 20 0.2178968906402588 0.0019298316207276716 10000 20 0.1966566801071167 0.0028489273218240147 10000 20 0.21264209747314453 0.00482341642419078 10000 20 0.197762489319 0.00198188642677 10000 20 1.90710241795 0.00757011452155 +100000 20 3.15847921371 0.0550301268026 100000 20 1.1962245225906372 0.010541407803235559 100000 20 1.1518636226654053 0.006410720031542237 100000 20 1.1567201137542724 0.002987096441878969 100000 20 1.75432097912 0.00692372147543 100000 20 9.8267291069 0.127844155326 +1000000 20 30.279082489 0.124978814506 1000000 20 10.956057572364807 0.12677108174061705 1000000 20 10.776052689552307 0.04739925571001746 1000000 20 10.640758633613586 0.06116581724028616 1000000 20 17.3383012295 0.0410164637448 1000000 20 83.9448960066 0.46121344046 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev lines columns textql_2.0.3_mean textql_2.0.3_stddev lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 50 0.105411934853 0.00171651054128 1 50 0.10458300113677979 0.0016367630302744722 1 50 0.08237688541412354 0.0016494314799953837 1 50 0.10066506862640381 0.002051307639276982 1 50 0.0205577373505 0.00133922342068 1 50 0.572030115128 0.0253648479103 +10 50 0.109102797508 0.00111620290512 10 50 0.10616152286529541 0.002345135740908088 10 50 0.08519520759582519 0.002610550182895596 10 50 0.10588631629943848 0.0035835389655972105 10 50 0.0195438146591 0.000791630611893 10 50 0.56993534565 0.0230474303306 +100 50 0.135682177544 0.00196166766665 100 50 0.12375867366790771 0.00238414904864133 100 50 0.10423583984375 0.0018808335751867933 100 50 0.11841504573822022 0.001608174845404568 100 50 0.0246078014374 0.00108949795701 100 50 0.563336873055 0.00964411866903 +1000 50 0.198261427879 0.00396172489054 1000 50 0.14462883472442628 0.0022428030896492978 1000 50 0.12195603847503662 0.0023611894043373983 1000 50 0.14032282829284667 0.002640027148889162 1000 50 0.063302564621 0.00058195987294 1000 50 0.826378440857 0.00941629472813 +10000 50 0.821499919891 0.0111642692132 10000 50 0.34488487243652344 0.004867441221052092 10000 50 0.3163540124893188 0.002761333651520998 10000 50 0.33160474300384524 0.0027796660009712947 10000 50 0.410061001778 0.00294901155085 10000 50 3.27872717381 0.126592845956 +100000 50 7.05980975628 0.121182371277 100000 50 2.3394312858581543 0.02263239858944125 100000 50 2.237372374534607 0.009955353920396077 100000 50 2.258401036262512 0.011041280982383895 100000 50 3.87797718048 0.0123467913678 100000 50 17.890055728 0.116794666005 +1000000 50 71.5645889759 5.02009516291 1000000 50 21.979821610450745 0.09080404939303836 1000000 50 21.59097549915314 0.081188190530421 1000000 50 21.70080256462097 0.15897944629180621 1000000 50 38.5674883366 0.0602820291386 1000000 50 158.262442636 0.826290454446 +lines columns q-benchmark-2.7.18_mean q-benchmark-2.7.18_stddev lines columns q-benchmark-3.6.4_mean q-benchmark-3.6.4_stddev lines columns q-benchmark-3.7.9_mean q-benchmark-3.7.9_stddev lines columns q-benchmark-3.8.5_mean q-benchmark-3.8.5_stddev lines columns textql_2.0.3_mean textql_2.0.3_stddev lines columns octosql_v0.3.0_mean octosql_v0.3.0_stddev +1 100 0.10662381649 0.00193146624495 1 100 0.10372309684753418 0.0010299126833031144 1 100 0.08336784839630126 0.0013840724401561887 1 100 0.10147004127502442 0.0021285682695135768 1 100 0.0216581106186 0.00103280947157 1 100 0.569358110428 0.0279801762531 +10 100 0.110662698746 0.00171461379583 10 100 0.10784556865692138 0.0016557634029464607 10 100 0.0864112138748169 0.0017946939354350697 10 100 0.10471885204315186 0.001248479289219899 10 100 0.021723818779 0.000920429257416 10 100 0.580981063843 0.0272341107532 +100 100 0.163547992706 0.00166570196628 100 100 0.14526791572570802 0.0028194506905186724 100 100 0.12199611663818359 0.0013003743156634682 100 100 0.13894760608673096 0.002307980025026551 100 100 0.0299471855164 0.00130217326679 100 100 0.559471726418 0.00668155858429 +1000 100 0.280023741722 0.00337543024145 1000 100 0.18315494060516357 0.0023585311962114673 1000 100 0.15871686935424806 0.0035993681064501234 1000 100 0.17586205005645753 0.0023822296091426 1000 100 0.0996923923492 0.00155352212734 1000 100 1.08161640167 0.00698594638512 +10000 100 1.46053376198 0.0221691284465 10000 100 0.5586131334304809 0.004808492789681402 10000 100 0.5243751525878906 0.004370273273595629 10000 100 0.5414002418518067 0.0036291866664635458 10000 100 0.767001605034 0.00328944029633 10000 100 5.67823712826 0.0123398407167 +100000 100 13.2369835854 0.309375896258 100000 100 4.287398314476013 0.00957500108409644 100000 100 4.175828623771667 0.016127303710583043 100000 100 4.222555088996887 0.08562968951916528 100000 100 7.46734063625 0.0262039846119 100000 100 32.2797194242 0.315508270241 +1000000 100 131.864977288 1.22415449691 1000000 100 41.706851434707644 0.4161526076289425 1000000 100 40.82292411327362 0.12328165162380703 1000000 100 41.021552324295044 0.16033566363076862 1000000 100 74.6216712952 0.0994037504394 1000000 100 289.582628798 0.929455236817 diff --git a/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/textql_2.0.3.benchmark-results b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/textql_2.0.3.benchmark-results new file mode 100644 index 00000000..f98760cd --- /dev/null +++ b/test/benchmark-results/source-files-1443b7418b46594ad256abd9db4a7671cb251e6a/2020-09-17-v2.0.17/textql_2.0.3.benchmark-results @@ -0,0 +1,48 @@ +lines columns textql_2.0.3_mean textql_2.0.3_stddev +1 1 0.0196103572845 0.00207355214257 +10 1 0.0186784029007 0.000970810220668 +100 1 0.019472026825 0.00181951524514 +1000 1 0.022180891037 0.00116649968967 +10000 1 0.051066827774 0.0018168767618 +100000 1 0.307463979721 0.00246268029188 +1000000 1 2.89862303734 0.022182722976 +lines columns textql_2.0.3_mean textql_2.0.3_stddev +1 5 0.0195286750793 0.0017840569109 +10 5 0.0183676958084 0.000925251595491 +100 5 0.0199447393417 0.000907007099218 +1000 5 0.0263328790665 0.00165486505938 +10000 5 0.0826982736588 0.00152451583229 +100000 5 0.60660867691 0.00395761320274 +1000000 5 5.87811236382 0.0304332294491 +lines columns textql_2.0.3_mean textql_2.0.3_stddev +1 10 0.0191783189774 0.00107718516178 +10 10 0.0185215950012 0.000840353961363 +100 10 0.0209223031998 0.00164494657684 +1000 10 0.0309282779694 0.00110848590345 +10000 10 0.121016025543 0.00105071105139 +100000 10 0.987622976303 0.00699348302979 +1000000 10 9.69240145683 0.0354453778052 +lines columns textql_2.0.3_mean textql_2.0.3_stddev +1 20 0.0202306985855 0.00159619251952 +10 20 0.0187650680542 0.000845692486156 +100 20 0.0211876153946 0.000993808448942 +1000 20 0.0404737234116 0.00122415059261 +10000 20 0.197762489319 0.00198188642677 +100000 20 1.75432097912 0.00692372147543 +1000000 20 17.3383012295 0.0410164637448 +lines columns textql_2.0.3_mean textql_2.0.3_stddev +1 50 0.0205577373505 0.00133922342068 +10 50 0.0195438146591 0.000791630611893 +100 50 0.0246078014374 0.00108949795701 +1000 50 0.063302564621 0.00058195987294 +10000 50 0.410061001778 0.00294901155085 +100000 50 3.87797718048 0.0123467913678 +1000000 50 38.5674883366 0.0602820291386 +lines columns textql_2.0.3_mean textql_2.0.3_stddev +1 100 0.0216581106186 0.00103280947157 +10 100 0.021723818779 0.000920429257416 +100 100 0.0299471855164 0.00130217326679 +1000 100 0.0996923923492 0.00155352212734 +10000 100 0.767001605034 0.00328944029633 +100000 100 7.46734063625 0.0262039846119 +1000000 100 74.6216712952 0.0994037504394 diff --git a/test/test-all b/test/test-all deleted file mode 100755 index 94d1f5ab..00000000 --- a/test/test-all +++ /dev/null @@ -1,14 +0,0 @@ -#!/bin/bash - -set -e - -function return_to_original_folder() { - popd -} -trap return_to_original_folder EXIT - -pushd $(dirname $0)/ - -./test-suite - -set +e diff --git a/test/test-all.bat b/test/test-all.bat deleted file mode 100644 index 3192c9e5..00000000 --- a/test/test-all.bat +++ /dev/null @@ -1,4 +0,0 @@ -@echo off - -echo TBD - diff --git a/test/test-suite b/test/test-suite deleted file mode 100755 index b5c05956..00000000 --- a/test/test-suite +++ /dev/null @@ -1,2294 +0,0 @@ -#!/usr/bin/env python - -# -# -# Simplistic test suite for q. -# -# Currently takes into account the project folder structure for running, so it needs -# to be executed from the current folder -# -# - -import unittest -import random -import json -from json import JSONEncoder -from subprocess import PIPE, Popen, STDOUT -import sys -import os -import time -from tempfile import NamedTemporaryFile -import locale -import pprint - -sys.path.append(os.path.join(os.path.abspath(os.path.dirname(sys.argv[0])),'..','bin')) -from qtextasdata import QTextAsData,QOutput,QOutputPrinter,QInputParams - -# q uses this encoding as the default output encoding. Some of the tests use it in order to -# make sure that the output is correctly encoded -SYSTEM_ENCODING = locale.getpreferredencoding() - -def run_command(cmd_to_run): - p = Popen(cmd_to_run, stdout=PIPE, stderr=PIPE, shell=True) - o, e = p.communicate() - # remove last newline - o = o.rstrip() - e = e.strip() - # split rows - if o != '': - o = o.split(os.linesep) - else: - o = [] - if e != '': - e = e.split(os.linesep) - else: - e = [] - return (p.returncode, o, e) - -uneven_ls_output = """drwxr-xr-x 2 root root 4096 Jun 11 2012 /selinux -drwxr-xr-x 2 root root 4096 Apr 19 2013 /mnt -drwxr-xr-x 2 root root 4096 Apr 24 2013 /srv -drwx------ 2 root root 16384 Jun 21 2013 /lost+found -lrwxrwxrwx 1 root root 33 Jun 21 2013 /initrd.img.old -> /boot/initrd.img-3.8.0-19-generic -drwxr-xr-x 2 root root 4096 Jun 21 2013 /cdrom -drwxr-xr-x 3 root root 4096 Jun 21 2013 /home -lrwxrwxrwx 1 root root 29 Jun 21 2013 /vmlinuz -> boot/vmlinuz-3.8.0-19-generic -lrwxrwxrwx 1 root root 32 Jun 21 2013 /initrd.img -> boot/initrd.img-3.8.0-19-generic -""" - -find_output = """8257537 32 drwxrwxrwt 218 root root 28672 Mar 1 11:00 /tmp -8299123 4 drwxrwxr-x 2 harel harel 4096 Feb 27 10:06 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/stormdist/testTopology3fad644a-54c0-4def-b19e-77ca97941595-1-1393513576 -8263229 964 -rw-rw-r-- 1 mapred mapred 984569 Feb 27 10:06 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/stormdist/testTopology3fad644a-54c0-4def-b19e-77ca97941595-1-1393513576/stormcode.ser -8263230 4 -rw-rw-r-- 1 harel harel 1223 Feb 27 10:06 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/stormdist/testTopology3fad644a-54c0-4def-b19e-77ca97941595-1-1393513576/stormconf.ser -8299113 4 drwxrwxr-x 2 harel harel 4096 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate -8263406 4 -rw-rw-r-- 1 harel harel 2002 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514168746 -8263476 0 -rw-rw-r-- 1 harel harel 0 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514168746.version -8263607 0 -rw-rw-r-- 1 harel harel 0 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514169735.version -8263533 0 -rw-rw-r-- 1 harel harel 0 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514172733.version -8263604 0 -rw-rw-r-- 1 harel harel 0 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514175754.version -""" - -header_row = 'name,value1,value2' -sample_data_rows = ['a,1,0', 'b,2,0', 'c,,0'] -sample_data_rows_with_empty_string = ['a,aaa,0', 'b,bbb,0', 'c,,0'] -sample_data_no_header = "\n".join(sample_data_rows) + "\n" -sample_data_with_empty_string_no_header = "\n".join( - sample_data_rows_with_empty_string) + "\n" -sample_data_with_header = header_row + "\n" + sample_data_no_header -sample_data_with_missing_header_names = "name,value1\n" + sample_data_no_header - -sample_quoted_data = '''non_quoted regular_double_quoted double_double_quoted escaped_double_quoted multiline_double_double_quoted multiline_escaped_double_quoted -control-value-1 "control-value-2" control-value-3 "control-value-4" control-value-5 "control-value-6" -non-quoted-value "this is a quoted value" "this is a ""double double"" quoted value" "this is an escaped \\"quoted value\\"" "this is a double double quoted ""multiline - value""." "this is an escaped \\"multiline - value\\"." -control-value-1 "control-value-2" control-value-3 "control-value-4" control-value-5 "control-value-6" -''' - -double_double_quoted_data = '''regular_double_quoted double_double_quoted -"this is a quoted value" "this is a quoted value with ""double double quotes""" -''' - -escaped_double_quoted_data = '''regular_double_quoted escaped_double_quoted -"this is a quoted value" "this is a quoted value with \\"escaped double quotes\\"" -''' - -combined_quoted_data = '''regular_double_quoted double_double_quoted escaped_double_quoted -"this is a quoted value" "this is a quoted value with ""double double quotes""" "this is a quoted value with \\"escaped double quotes\\"" -''' - -sample_quoted_data2 = '"quoted data" 23\nunquoted-data 54' - -one_column_data = '''data without commas 1 -data without commas 2 -''' - -# Values with leading whitespace -sample_data_rows_with_spaces = ['a,1,0', ' b, 2,0', 'c,,0'] -sample_data_with_spaces_no_header = "\n".join( - sample_data_rows_with_spaces) + "\n" - -header_row_with_spaces = 'name,value 1,value2' -sample_data_with_spaces_with_header = header_row_with_spaces + \ - "\n" + sample_data_with_spaces_no_header - -long_value1 = "23683289372328372328373" -int_value = "2328372328373" -sample_data_with_long_values = "%s\n%s\n%s" % (long_value1,int_value,int_value) - -def one_column_warning(e): - return e[0].startswith('Warning: column count is one') - - -class AbstractQTestCase(unittest.TestCase): - - def create_file_with_data(self, data, encoding='utf-8'): - tmpfile = NamedTemporaryFile(delete=False) - if encoding != 'none' and encoding is not None: - tmpfile.write(data.encode(encoding)) - else: - tmpfile.write(data) - tmpfile.close() - return tmpfile - - def cleanup(self, tmpfile): - os.remove(tmpfile.name) - - def random_tmp_filename(self,prefix,postfix): - # TODO Use more robust method for this - path = '/var/tmp' - return '%s/%s-%s.%s' % (path,prefix,random.randint(0,1000000000),postfix) - -class SaveDbToDiskTests(AbstractQTestCase): - - def test_store_to_disk(self): - db_filename = self.random_tmp_filename('store-to-disk','db') - self.assertFalse(os.path.exists(db_filename)) - - retcode, o, e = run_command('seq 1 1000 | ../bin/q "select count(*) from -" -c 1 -S %s' % db_filename) - - self.assertTrue(retcode == 0) - self.assertTrue(len(o) == 0) - self.assertTrue(len(e) == 5) - self.assertTrue(e[0].startswith('Going to save data')) - self.assertTrue(db_filename in e[0]) - self.assertTrue(e[1].startswith('Data has been loaded in')) - self.assertTrue(e[2].startswith('Saving data to db file')) - self.assertTrue(e[3].startswith('Data has been saved into')) - self.assertTrue(e[4] == 'Query to run on the database: select count(*) from `-`;') - - self.assertTrue(os.path.exists(db_filename)) - - sqlite_command = """echo 'select * from `-`;' | sqlite3 %s""" % db_filename - sqlite_retcode,sqlite_o,sqlite_e = run_command(sqlite_command) - self.assertTrue(sqlite_retcode == 0) - self.assertTrue(len(sqlite_o) == 1000) - self.assertTrue(len(sqlite_e) == 0) - - os.remove(db_filename) - - def test_preventing_db_overwrite(self): - db_filename = self.random_tmp_filename('store-to-disk', 'db') - self.assertFalse(os.path.exists(db_filename)) - - retcode, o, e = run_command('seq 1 1000 | ../bin/q "select count(*) from -" -c 1 -S %s' % db_filename) - - self.assertTrue(retcode == 0) - self.assertTrue(os.path.exists(db_filename)) - - retcode2, o2, e2 = run_command('seq 1 1000 | ../bin/q "select count(*) from -" -c 1 -S %s' % db_filename) - self.assertTrue(retcode2 != 0) - self.assertTrue(e2[0].startswith('Going to save data into a disk database')) - self.assertTrue(e2[1] == 'Disk database file %s already exists.' % db_filename) - - os.remove(db_filename) - - -class BasicTests(AbstractQTestCase): - - def test_basic_aggregation(self): - retcode, o, e = run_command( - 'seq 1 10 | ../bin/q "select sum(c1),avg(c1) from -"') - self.assertTrue(retcode == 0) - self.assertTrue(len(o) == 1) - self.assertTrue(len(e) == 1) - - s = sum(xrange(1, 11)) - self.assertTrue(o[0] == '%s %s' % (s, s / 10.0)) - self.assertTrue(one_column_warning(e)) - - def test_gzipped_file(self): - tmpfile = self.create_file_with_data( - '\x1f\x8b\x08\x08\xf2\x18\x12S\x00\x03xxxxxx\x003\xe42\xe22\xe62\xe12\xe52\xe32\xe7\xb2\xe0\xb2\xe424\xe0\x02\x00\xeb\xbf\x8a\x13\x15\x00\x00\x00', encoding='none') - - cmd = '../bin/q -z "select sum(c1),avg(c1) from %s"' % tmpfile.name - - retcode, o, e = run_command(cmd) - self.assertTrue(retcode == 0) - self.assertTrue(len(o) == 1) - self.assertTrue(len(e) == 1) - - s = sum(xrange(1, 11)) - self.assertTrue(o[0] == '%s %s' % (s, s / 10.0)) - self.assertTrue(one_column_warning(e)) - - self.cleanup(tmpfile) - - def test_attempt_to_unzip_stdin(self): - tmpfile = self.create_file_with_data( - '\x1f\x8b\x08\x08\xf2\x18\x12S\x00\x03xxxxxx\x003\xe42\xe22\xe62\xe12\xe52\xe32\xe7\xb2\xe0\xb2\xe424\xe0\x02\x00\xeb\xbf\x8a\x13\x15\x00\x00\x00', encoding='none') - - cmd = 'cat %s | ../bin/q -z "select sum(c1),avg(c1) from -"' % tmpfile.name - - retcode, o, e = run_command(cmd) - self.assertTrue(retcode != 0) - self.assertTrue(len(o) == 0) - self.assertTrue(len(e) == 1) - - self.assertEquals(e[0],'Cannot decompress standard input. Pipe the input through zcat in order to decompress.') - - self.cleanup(tmpfile) - - def test_delimition_mistake_with_header(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - - cmd = '../bin/q -d " " "select * from %s" -H' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 3) - - self.assertTrue(e[0].startswith( - "Warning: column count is one - did you provide the correct delimiter")) - self.assertTrue(e[1].startswith("Bad header row")) - self.assertTrue("Column name cannot contain commas" in e[2]) - - self.cleanup(tmpfile) - - def test_regexp_int_data_handling(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - - cmd = '../bin/q -d , "select c2 from %s where regexp(\'^1\',c2)"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 1) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0],"1") - - self.cleanup(tmpfile) - - def test_regexp_null_data_handling(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - - cmd = '../bin/q -d , "select count(*) from %s where regexp(\'^\',c2)"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 1) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0],"2") - - self.cleanup(tmpfile) - - def test_select_one_column(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - - cmd = '../bin/q -d , "select c1 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - - self.assertEquals(" ".join(o), 'a b c') - - self.cleanup(tmpfile) - - def test_tab_delimition_parameter(self): - tmpfile = self.create_file_with_data( - sample_data_no_header.replace(",", "\t")) - cmd = '../bin/q -t "select c1,c2,c3 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - self.assertEquals(o[0], sample_data_rows[0].replace(",", "\t")) - self.assertEquals(o[1], sample_data_rows[1].replace(",", "\t")) - self.assertEquals(o[2], sample_data_rows[2].replace(",", "\t")) - - self.cleanup(tmpfile) - - def test_tab_delimition_parameter__with_manual_override_attempt(self): - tmpfile = self.create_file_with_data( - sample_data_no_header.replace(",", "\t")) - cmd = '../bin/q -t -d , "select c1,c2,c3 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - self.assertEquals(o[0], sample_data_rows[0].replace(",", "\t")) - self.assertEquals(o[1], sample_data_rows[1].replace(",", "\t")) - self.assertEquals(o[2], sample_data_rows[2].replace(",", "\t")) - - self.cleanup(tmpfile) - - def test_output_delimiter(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , -D "|" "select c1,c2,c3 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], sample_data_rows[0].replace(",", "|")) - self.assertEquals(o[1], sample_data_rows[1].replace(",", "|")) - self.assertEquals(o[2], sample_data_rows[2].replace(",", "|")) - - self.cleanup(tmpfile) - - def test_output_delimiter_tab_parameter(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , -T "select c1,c2,c3 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], sample_data_rows[0].replace(",", "\t")) - self.assertEquals(o[1], sample_data_rows[1].replace(",", "\t")) - self.assertEquals(o[2], sample_data_rows[2].replace(",", "\t")) - - self.cleanup(tmpfile) - - def test_output_delimiter_tab_parameter__with_manual_override_attempt(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , -T -D "|" "select c1,c2,c3 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], sample_data_rows[0].replace(",", "\t")) - self.assertEquals(o[1], sample_data_rows[1].replace(",", "\t")) - self.assertEquals(o[2], sample_data_rows[2].replace(",", "\t")) - - self.cleanup(tmpfile) - - def test_stdin_input(self): - cmd = 'printf "%s" | ../bin/q -d , "select c1,c2,c3 from -"' % sample_data_no_header - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], sample_data_rows[0]) - self.assertEquals(o[1], sample_data_rows[1]) - self.assertEquals(o[2], sample_data_rows[2]) - - def test_column_separation(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , "select c1,c2,c3 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], sample_data_rows[0]) - self.assertEquals(o[1], sample_data_rows[1]) - self.assertEquals(o[2], sample_data_rows[2]) - - self.cleanup(tmpfile) - - def test_column_analysis(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - - cmd = '../bin/q -d , "select c1 from %s" -A' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(o[0], 'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1].strip(), '`c1` - text') - self.assertEquals(o[2].strip(), '`c2` - int') - self.assertEquals(o[3].strip(), '`c3` - int') - - self.cleanup(tmpfile) - - def test_column_analysis_no_header(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - - cmd = '../bin/q -d , "select c1 from %s" -A' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(o[0], 'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1].strip(), '`c1` - text') - self.assertEquals(o[2].strip(), '`c2` - int') - self.assertEquals(o[3].strip(), '`c3` - int') - - def test_header_exception_on_numeric_header_data(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , "select * from %s" -A -H' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 3) - self.assertTrue( - 'Bad header row: Header must contain only strings' in e[0]) - self.assertTrue("Column name must be a string" in e[1]) - self.assertTrue("Column name must be a string" in e[2]) - - self.cleanup(tmpfile) - - def test_column_analysis_with_header(self): - tmpfile = self.create_file_with_data(sample_data_with_header) - cmd = '../bin/q -d , "select c1 from %s" -A -H' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(o),4) - self.assertEquals(len(e),2) - self.assertEquals(o[0], 'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1].strip(), '`name` - text') - self.assertEquals(o[2].strip(), '`value1` - int') - self.assertEquals(o[3].strip(), '`value2` - int') - self.assertEquals(e[0].strip(),'query error: no such column: c1') - self.assertTrue(e[1].startswith('Warning - There seems to be a ')) - - self.cleanup(tmpfile) - - def test_data_with_header(self): - tmpfile = self.create_file_with_data(sample_data_with_header) - cmd = '../bin/q -d , "select name from %s" -H' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(" ".join(o), "a b c") - - self.cleanup(tmpfile) - - def test_output_header_when_input_header_exists(self): - tmpfile = self.create_file_with_data(sample_data_with_header) - cmd = '../bin/q -d , "select name from %s" -H -O' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 4) - self.assertEquals(o[0],'name') - self.assertEquals(o[1],'a') - self.assertEquals(o[2],'b') - self.assertEquals(o[3],'c') - - self.cleanup(tmpfile) - - def test_generated_column_name_warning_when_header_line_exists(self): - tmpfile = self.create_file_with_data(sample_data_with_header) - cmd = '../bin/q -d , "select c3 from %s" -H' % tmpfile.name - - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 2) - self.assertTrue('no such column: c3' in e[0]) - self.assertEquals( - e[1], 'Warning - There seems to be a "no such column" error, and -H (header line) exists. Please make sure that you are using the column names from the header line and not the default (cXX) column names') - - self.cleanup(tmpfile) - - def test_column_analysis_with_unexpected_header(self): - tmpfile = self.create_file_with_data(sample_data_with_header) - cmd = '../bin/q -d , "select c1 from %s" -A' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 4) - self.assertEquals(len(e), 1) - - self.assertEquals(o[0], 'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1].strip(), '`c1` - text') - self.assertEquals(o[2].strip(), '`c2` - text') - self.assertEquals(o[3].strip(), '`c3` - text') - - self.assertEquals( - e[0], 'Warning - There seems to be header line in the file, but -H has not been specified. All fields will be detected as text fields, and the header line will appear as part of the data') - - self.cleanup(tmpfile) - - def test_empty_data(self): - tmpfile = self.create_file_with_data('') - cmd = '../bin/q -d , "select c1 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - self.assertTrue('Warning - data is empty' in e[0]) - - self.cleanup(tmpfile) - - def test_empty_data_with_header_param(self): - tmpfile = self.create_file_with_data('') - cmd = '../bin/q -d , "select c1 from %s" -H' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - m = "Header line is expected but missing in file %s" % tmpfile.name - self.assertTrue(m in e[0]) - - self.cleanup(tmpfile) - - def test_one_row_of_data_without_header_param(self): - tmpfile = self.create_file_with_data(header_row) - cmd = '../bin/q -d , "select c2 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 1) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], 'value1') - - self.cleanup(tmpfile) - - def test_one_row_of_data_with_header_param(self): - tmpfile = self.create_file_with_data(header_row) - cmd = '../bin/q -d , "select c2 from %s" -H' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - self.assertTrue('Warning - data is empty' in e[0]) - - self.cleanup(tmpfile) - - def test_dont_leading_keep_whitespace_in_values(self): - tmpfile = self.create_file_with_data(sample_data_with_spaces_no_header) - cmd = '../bin/q -d , "select c1 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 3) - - self.assertEquals(o[0], 'a') - self.assertEquals(o[1], 'b') - self.assertEquals(o[2], 'c') - - self.cleanup(tmpfile) - - def test_keep_leading_whitespace_in_values(self): - tmpfile = self.create_file_with_data(sample_data_with_spaces_no_header) - cmd = '../bin/q -d , "select c1 from %s" -k' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 3) - - self.assertEquals(o[0], 'a') - self.assertEquals(o[1], ' b') - self.assertEquals(o[2], 'c') - - self.cleanup(tmpfile) - - def test_no_impact_of_keeping_leading_whitespace_on_integers(self): - tmpfile = self.create_file_with_data(sample_data_with_spaces_no_header) - cmd = '../bin/q -d , "select c2 from %s" -k -A' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 4) - - self.assertEquals(o[0], 'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1].strip(), '`c1` - text') - self.assertEquals(o[2].strip(), '`c2` - int') - self.assertEquals(o[3].strip(), '`c3` - int') - - self.cleanup(tmpfile) - - def test_spaces_in_header_row(self): - tmpfile = self.create_file_with_data( - header_row_with_spaces + "\n" + sample_data_no_header) - cmd = '../bin/q -d , "select name,\`value 1\` from %s" -H' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 3) - - self.assertEquals(o[0], 'a,1') - self.assertEquals(o[1], 'b,2') - self.assertEquals(o[2], 'c,') - - self.cleanup(tmpfile) - - def test_column_analysis_for_spaces_in_header_row(self): - tmpfile = self.create_file_with_data( - header_row_with_spaces + "\n" + sample_data_no_header) - cmd = '../bin/q -d , "select name,\`value 1\` from %s" -H -A' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 4) - - self.assertEquals(o[0], 'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1].strip(), '`name` - text') - self.assertEquals(o[2].strip(), '`value 1` - int') - self.assertEquals(o[3].strip(), '`value2` - int') - - self.cleanup(tmpfile) - - def test_no_query_in_command_line(self): - cmd = '../bin/q -d , ""' - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 1) - self.assertEquals(len(e), 1) - self.assertEquals(len(o), 0) - - self.assertEquals(e[0],'Query cannot be empty (query number 1)') - - def test_empty_query_in_command_line(self): - cmd = '../bin/q -d , " "' - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 1) - self.assertEquals(len(e), 1) - self.assertEquals(len(o), 0) - - self.assertEquals(e[0],'Query cannot be empty (query number 1)') - - def test_failure_in_query_stops_processing_queries(self): - cmd = '../bin/q -d , "select 500" "select 300" "wrong-query" "select 8000"' - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 1) - self.assertEquals(len(e), 1) - self.assertEquals(len(o), 2) - self.assertEquals(o[0],'500') - self.assertEquals(o[1],'300') - - def test_multiple_queries_in_command_line(self): - cmd = '../bin/q -d , "select 500" "select 300+100" "select 300" "select 200"' - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 4) - - self.assertEquals(o[0],'500') - self.assertEquals(o[1],'400') - self.assertEquals(o[2],'300') - self.assertEquals(o[3],'200') - - def test_literal_calculation_query(self): - cmd = '../bin/q -d , "select 1+40/6"' - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 1) - - self.assertEquals(o[0],'7') - - def test_literal_calculation_query_float_result(self): - cmd = '../bin/q -d , "select 1+40/6.0"' - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 1) - - self.assertEquals(o[0],'7.66666666667') - - def test_use_query_file(self): - tmp_data_file = self.create_file_with_data(sample_data_with_header) - tmp_query_file = self.create_file_with_data("select name from %s" % tmp_data_file.name) - - cmd = '../bin/q -d , -q %s -H' % tmp_query_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 3) - - self.assertEquals(o[0], 'a') - self.assertEquals(o[1], 'b') - self.assertEquals(o[2], 'c') - - self.cleanup(tmp_data_file) - self.cleanup(tmp_query_file) - - def test_use_query_file_with_incorrect_query_encoding(self): - tmp_data_file = self.create_file_with_data(sample_data_with_header) - tmp_query_file = self.create_file_with_data("select name,'Hr\xc3\xa1\xc4\x8d' from %s" % tmp_data_file.name,encoding=None) - - cmd = '../bin/q -d , -q %s -H -Q ascii' % tmp_query_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,3) - self.assertEquals(len(o),0) - self.assertEquals(len(e),1) - - self.assertTrue(e[0].startswith('Could not decode query number 1 using the provided query encoding (ascii)')) - - self.cleanup(tmp_data_file) - self.cleanup(tmp_query_file) - - def test_output_header_with_non_ascii_names(self): - tmp_data_file = self.create_file_with_data(sample_data_with_header) - tmp_query_file = self.create_file_with_data("select name,'Hr\xc3\xa1\xc4\x8d' Hr\xc3\xa1\xc4\x8d from %s" % tmp_data_file.name,encoding=None) - - cmd = '../bin/q -d , -q %s -H -Q utf-8 -O' % tmp_query_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(o),4) - self.assertEquals(len(e),0) - - self.assertEquals(o[0].decode(SYSTEM_ENCODING), u'name,Hr\xe1\u010d') - self.assertEquals(o[1].decode(SYSTEM_ENCODING), u'a,Hr\xe1\u010d') - self.assertEquals(o[2].decode(SYSTEM_ENCODING), u'b,Hr\xe1\u010d') - self.assertEquals(o[3].decode(SYSTEM_ENCODING), u'c,Hr\xe1\u010d') - - self.cleanup(tmp_data_file) - self.cleanup(tmp_query_file) - - def test_use_query_file_with_query_encoding(self): - tmp_data_file = self.create_file_with_data(sample_data_with_header) - tmp_query_file = self.create_file_with_data("select name,'Hr\xc3\xa1\xc4\x8d' from %s" % tmp_data_file.name,encoding=None) - - cmd = '../bin/q -d , -q %s -H -Q utf-8' % tmp_query_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 3) - - self.assertEquals(o[0].decode(SYSTEM_ENCODING), u'a,Hr\xe1\u010d') - self.assertEquals(o[1].decode(SYSTEM_ENCODING), u'b,Hr\xe1\u010d') - self.assertEquals(o[2].decode(SYSTEM_ENCODING), u'c,Hr\xe1\u010d') - - self.cleanup(tmp_data_file) - self.cleanup(tmp_query_file) - - def test_use_query_file_and_command_line(self): - tmp_data_file = self.create_file_with_data(sample_data_with_header) - tmp_query_file = self.create_file_with_data("select name from %s" % tmp_data_file.name) - - cmd = '../bin/q -d , -q %s -H "select * from ppp"' % tmp_query_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 1) - self.assertEquals(len(e), 1) - self.assertEquals(len(o), 0) - - self.assertTrue(e[0].startswith("Can't provide both a query file and a query on the command line")) - - self.cleanup(tmp_data_file) - self.cleanup(tmp_query_file) - - def test_select_output_encoding(self): - tmp_data_file = self.create_file_with_data(sample_data_with_header) - tmp_query_file = self.create_file_with_data("select 'Hr\xc3\xa1\xc4\x8d' from %s" % tmp_data_file.name,encoding=None) - - for target_encoding in ['utf-8','ibm852']: - cmd = '../bin/q -d , -q %s -H -Q utf-8 -E %s' % (tmp_query_file.name,target_encoding) - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 3) - - self.assertEquals(o[0].decode(target_encoding), u'Hr\xe1\u010d') - self.assertEquals(o[1].decode(target_encoding), u'Hr\xe1\u010d') - self.assertEquals(o[2].decode(target_encoding), u'Hr\xe1\u010d') - - self.cleanup(tmp_data_file) - self.cleanup(tmp_query_file) - - def test_select_failed_output_encoding(self): - tmp_data_file = self.create_file_with_data(sample_data_with_header) - tmp_query_file = self.create_file_with_data("select 'Hr\xc3\xa1\xc4\x8d' from %s" % tmp_data_file.name,encoding=None) - - cmd = '../bin/q -d , -q %s -H -Q utf-8 -E ascii' % tmp_query_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 3) - self.assertEquals(len(e), 1) - self.assertEquals(len(o), 0) - - self.assertTrue(e[0].startswith('Cannot encode data')) - - self.cleanup(tmp_data_file) - self.cleanup(tmp_query_file) - - - def test_use_query_file_with_empty_query(self): - tmp_query_file = self.create_file_with_data(" ") - - cmd = '../bin/q -d , -q %s -H' % tmp_query_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 1) - self.assertEquals(len(e), 1) - self.assertEquals(len(o), 0) - - self.assertTrue(e[0].startswith("Query cannot be empty")) - - self.cleanup(tmp_query_file) - - def test_use_non_existent_query_file(self): - cmd = '../bin/q -d , -q non-existent-query-file -H' - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 1) - self.assertEquals(len(e), 1) - self.assertEquals(len(o), 0) - - self.assertTrue(e[0].startswith("Could not read query from file")) - - def test_non_quoted_values_in_quoted_data(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data) - - cmd = '../bin/q -d " " "select c1 from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),4) - - self.assertTrue(o[0],'non_quoted') - self.assertTrue(o[1],'control-value-1') - self.assertTrue(o[2],'non-quoted-value') - self.assertTrue(o[3],'control-value-1') - - self.cleanup(tmp_data_file) - - def test_regular_quoted_values_in_quoted_data(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data) - - cmd = '../bin/q -d " " "select c2 from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),4) - - self.assertTrue(o[0],'regular_double_quoted') - self.assertTrue(o[1],'control-value-2') - self.assertTrue(o[2],'this is a quoted value') - self.assertTrue(o[3],'control-value-2') - - self.cleanup(tmp_data_file) - - def test_double_double_quoted_values_in_quoted_data(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data) - - cmd = '../bin/q -d " " "select c3 from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),4) - - self.assertTrue(o[0],'double_double_quoted') - self.assertTrue(o[1],'control-value-3') - self.assertTrue(o[2],'this is a "double double" quoted value') - self.assertTrue(o[3],'control-value-3') - - self.cleanup(tmp_data_file) - - def test_escaped_double_quoted_values_in_quoted_data(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data) - - cmd = '../bin/q -d " " "select c4 from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),4) - - self.assertTrue(o[0],'escaped_double_quoted') - self.assertTrue(o[1],'control-value-4') - self.assertTrue(o[2],'this is an escaped "quoted value"') - self.assertTrue(o[3],'control-value-4') - - self.cleanup(tmp_data_file) - - def test_none_input_quoting_mode_in_relaxed_mode(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - cmd = '../bin/q -d " " -m relaxed -D , -w none -W none "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'"quoted,data",23') - self.assertEquals(o[1],'unquoted-data,54,') - - self.cleanup(tmp_data_file) - - def test_none_input_quoting_mode_in_strict_mode(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - cmd = '../bin/q -d " " -m strict -D , -w none "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode,0) - self.assertEquals(len(e),1) - self.assertEquals(len(o),0) - - self.assertTrue(e[0].startswith('Strict mode. Column Count is expected to identical')) - - self.cleanup(tmp_data_file) - - def test_minimal_input_quoting_mode(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - cmd = '../bin/q -d " " -D , -w minimal "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'quoted data,23') - self.assertEquals(o[1],'unquoted-data,54') - - self.cleanup(tmp_data_file) - - def test_all_input_quoting_mode(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - cmd = '../bin/q -d " " -D , -w all "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'quoted data,23') - self.assertEquals(o[1],'unquoted-data,54') - - self.cleanup(tmp_data_file) - - def test_incorrect_input_quoting_mode(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - cmd = '../bin/q -d " " -D , -w unknown_wrapping_mode "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode,0) - self.assertEquals(len(e),1) - self.assertEquals(len(o),0) - - self.assertTrue(e[0].startswith('Input quoting mode can only be one of all,none,minimal')) - self.assertTrue('unknown_wrapping_mode' in e[0]) - - self.cleanup(tmp_data_file) - - def test_none_output_quoting_mode(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - cmd = '../bin/q -d " " -D , -w all -W none "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'quoted data,23') - self.assertEquals(o[1],'unquoted-data,54') - - self.cleanup(tmp_data_file) - - def test_minimal_output_quoting_mode__without_need_to_quote_in_output(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - cmd = '../bin/q -d " " -D , -w all -W minimal "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'quoted data,23') - self.assertEquals(o[1],'unquoted-data,54') - - self.cleanup(tmp_data_file) - - def test_minimal_output_quoting_mode__with_need_to_quote_in_output(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - # output delimiter is set to space, so the output will contain it - cmd = '../bin/q -d " " -D " " -w all -W minimal "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'"quoted data" 23') - self.assertEquals(o[1],'unquoted-data 54') - - self.cleanup(tmp_data_file) - - def test_nonnumeric_output_quoting_mode(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - cmd = '../bin/q -d " " -D , -w all -W nonnumeric "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'"quoted data",23') - self.assertEquals(o[1],'"unquoted-data",54') - - self.cleanup(tmp_data_file) - - def test_all_output_quoting_mode(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data2) - - cmd = '../bin/q -d " " -D , -w all -W all "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'"quoted data","23"') - self.assertEquals(o[1],'"unquoted-data","54"') - - self.cleanup(tmp_data_file) - - def _internal_test_consistency_of_chaining_output_to_input(self,input_data,input_wrapping_mode,output_wrapping_mode): - - tmp_data_file = self.create_file_with_data(input_data) - - basic_cmd = '../bin/q -w %s -W %s "select * from -"' % (input_wrapping_mode,output_wrapping_mode) - chained_cmd = 'cat %s | %s | %s | %s' % (tmp_data_file.name,basic_cmd,basic_cmd,basic_cmd) - - retcode, o, e = run_command(chained_cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals("\n".join(o),input_data) - - self.cleanup(tmp_data_file) - - def test_consistency_of_chaining_minimal_wrapping_to_minimal_wrapping(self): - input_data = '"quoted data" 23\nunquoted-data 54' - self._internal_test_consistency_of_chaining_output_to_input(input_data,'minimal','minimal') - - def test_consistency_of_chaining_all_wrapping_to_all_wrapping(self): - input_data = '"quoted data" "23"\n"unquoted-data" "54"' - self._internal_test_consistency_of_chaining_output_to_input(input_data,'all','all') - - def test_utf8_with_bom_encoding(self): - utf_8_data_with_bom = '\xef\xbb\xbf"typeid","limit","apcost","date","checkpointId"\n"1","2","5","1,2,3,4,5,6,7","3000,3001,3002"\n"2","2","5","1,2,3,4,5,6,7","3003,3004,3005"\n' - tmp_data_file = self.create_file_with_data(utf_8_data_with_bom,encoding=None) - - cmd = '../bin/q -d , -H -O -e utf-8-sig "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),3) - - self.assertEquals(o[0],'typeid,limit,apcost,date,checkpointId') - self.assertEquals(o[1],'1,2,5,"1,2,3,4,5,6,7","3000,3001,3002"') - self.assertEquals(o[2],'2,2,5,"1,2,3,4,5,6,7","3003,3004,3005"') - - self.cleanup(tmp_data_file) - - def test_input_field_quoting_and_data_types_with_encoding(self): - # Checks combination of minimal input field quoting, with special characters that need to be decoded - - # Both content and proper data types are verified - data = '111,22.22,"testing text with special characters - citt\xc3\xa0 ",http://somekindofurl.com,12.13.14.15,12.1\n' - tmp_data_file = self.create_file_with_data(data,encoding='none') - - cmd = '../bin/q -d , "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),1) - - self.assertEquals(o[0].decode('utf-8'),u'111,22.22,testing text with special characters - citt\xe0 ,http://somekindofurl.com,12.13.14.15,12.1') - - cmd = '../bin/q -d , "select * from %s" -A' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),7) - - self.assertTrue(o[0].startswith('Table for file')) - self.assertEquals(o[1].strip(),'`c1` - int') - self.assertEquals(o[2].strip(),'`c2` - float') - self.assertEquals(o[3].strip(),'`c3` - text') - self.assertEquals(o[4].strip(),'`c4` - text') - self.assertEquals(o[5].strip(),'`c5` - text') - self.assertEquals(o[6].strip(),'`c6` - float') - - self.cleanup(tmp_data_file) - - def test_multiline_double_double_quoted_values_in_quoted_data(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data) - - # FIXME Need to convert \0a to proper encoding suitable for the person running the tests. - cmd = '../bin/q -d " " "select replace(c5,X\'0A\',\'::\') from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),4) - - self.assertTrue(o[0],'multiline_double_double_quoted') - self.assertTrue(o[1],'control-value-5') - self.assertTrue(o[2],'this is a double double quoted "multiline\n value".') - self.assertTrue(o[3],'control-value-5') - - self.cleanup(tmp_data_file) - - def test_multiline_escaped_double_quoted_values_in_quoted_data(self): - tmp_data_file = self.create_file_with_data(sample_quoted_data) - - # FIXME Need to convert \0a to proper encoding suitable for the person running the tests. - cmd = '../bin/q -d " " "select replace(c6,X\'0A\',\'::\') from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),4) - - self.assertTrue(o[0],'multiline_escaped_double_quoted') - self.assertTrue(o[1],'control-value-6') - self.assertTrue(o[2],'this is an escaped "multiline:: value".') - self.assertTrue(o[3],'control-value-6') - - self.cleanup(tmp_data_file) - - def test_disable_double_double_quoted_data_flag__values(self): - # This test (and flag) is meant to verify backward comptibility only. It is possible that - # this flag will be removed completely in the future - - tmp_data_file = self.create_file_with_data(double_double_quoted_data) - - cmd = '../bin/q -d " " --disable-double-double-quoting "select c2 from %s" -W none' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'double_double_quoted') - self.assertEquals(o[1],'this is a quoted value with "double') - - cmd = '../bin/q -d " " --disable-double-double-quoting "select c3 from %s" -W none' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'') - self.assertEquals(o[1],'double') - - cmd = '../bin/q -d " " --disable-double-double-quoting "select c4 from %s" -W none' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'') - self.assertEquals(o[1],'quotes"""') - - self.cleanup(tmp_data_file) - - def test_disable_escaped_double_quoted_data_flag__values(self): - # This test (and flag) is meant to verify backward comptibility only. It is possible that - # this flag will be removed completely in the future - - tmp_data_file = self.create_file_with_data(escaped_double_quoted_data) - - cmd = '../bin/q -d " " --disable-escaped-double-quoting "select c2 from %s" -W none' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'escaped_double_quoted') - self.assertEquals(o[1],'this is a quoted value with \\escaped') - - cmd = '../bin/q -d " " --disable-escaped-double-quoting "select c3 from %s" -W none' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'') - self.assertEquals(o[1],'double') - - cmd = '../bin/q -d " " --disable-escaped-double-quoting "select c4 from %s" -W none' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'') - self.assertEquals(o[1],'quotes\\""') - - self.cleanup(tmp_data_file) - - def test_combined_quoted_data_flags__number_of_columns_detected(self): - # This test (and flags) is meant to verify backward comptibility only. It is possible that - # these flags will be removed completely in the future - tmp_data_file = self.create_file_with_data(combined_quoted_data) - - cmd = '../bin/q -d " " --disable-double-double-quoting --disable-escaped-double-quoting "select * from %s" -A' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - o = o[1:] # remove the first "Table for file..." line in the output - - self.assertEquals(len(o),7) # found 7 fields - - cmd = '../bin/q -d " " --disable-escaped-double-quoting "select * from %s" -A' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - o = o[1:] # remove the first "Table for file..." line in the output - - self.assertEquals(len(o),5) # found 5 fields - - cmd = '../bin/q -d " " --disable-double-double-quoting "select * from %s" -A' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - o = o[1:] # remove the first "Table for file..." line in the output - - self.assertEquals(len(o),5) # found 5 fields - - cmd = '../bin/q -d " " "select * from %s" -A' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e),0) - o = o[1:] # remove the first "Table for file..." line in the output - - self.assertEquals(len(o),3) # found only 3 fields, which is the correct amount - - self.cleanup(tmp_data_file) - - def test_nonexistent_file(self): - cmd = '../bin/q "select * from non-existent-file"' - - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode,0) - self.assertEquals(len(o),0) - self.assertEquals(len(e),1) - - self.assertEquals(e[0],"No files matching 'non-existent-file' have been found") - - def test_default_column_max_length_parameter__short_enough(self): - huge_text = "x" * 131000 - - file_data = "a,b,c\n1,%s,3\n" % huge_text - - tmpfile = self.create_file_with_data(file_data) - - cmd = '../bin/q -H -d , "select a from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 1) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0],'1') - - self.cleanup(tmpfile) - - def test_default_column_max_length_parameter__too_long(self): - huge_text = "x" * 132000 - - file_data = "a,b,c\n1,%s,3\n" % huge_text - - tmpfile = self.create_file_with_data(file_data) - - cmd = '../bin/q -H -d , "select a from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 31) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - self.assertTrue(e[0].startswith("Column length is larger than the maximum")) - self.assertTrue(("Offending file is '%s'" % tmpfile.name) in e[0]) - self.assertTrue('Line is 2' in e[0]) - - self.cleanup(tmpfile) - - def test_column_max_length_parameter(self): - file_data = "a,b,c\nvery-long-text,2,3\n" - tmpfile = self.create_file_with_data(file_data) - - cmd = '../bin/q -H -d , -M 3 "select a from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 31) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - self.assertTrue(e[0].startswith("Column length is larger than the maximum")) - self.assertTrue(("Offending file is '%s'" % tmpfile.name) in e[0]) - self.assertTrue('Line is 2' in e[0]) - - cmd2 = '../bin/q -H -d , -M 300 -H "select a from %s"' % tmpfile.name - retcode2, o2, e2 = run_command(cmd2) - - self.assertEquals(retcode2, 0) - self.assertEquals(len(o2), 1) - self.assertEquals(len(e2), 0) - - self.assertEquals(o2[0],'very-long-text') - - self.cleanup(tmpfile) - - def test_invalid_column_max_length_parameter(self): - file_data = "a,b,c\nvery-long-text,2,3\n" - tmpfile = self.create_file_with_data(file_data) - - cmd = '../bin/q -H -d , -M 0 "select a from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 31) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - self.assertTrue(e[0].startswith('Max column length limit must be a positive integer')) - - - self.cleanup(tmpfile) - - def test_duplicate_column_name_detection(self): - file_data = "a,b,a\n10,20,30\n30,40,50" - tmpfile = self.create_file_with_data(file_data) - - cmd = '../bin/q -H -d , "select a from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 35) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 2) - - self.assertTrue(e[0].startswith('Bad header row:')) - self.assertEquals(e[1],"'a': Column name is duplicated") - - self.cleanup(tmpfile) - - -class ParsingModeTests(AbstractQTestCase): - - def test_strict_mode_column_count_mismatch_error(self): - tmpfile = self.create_file_with_data(uneven_ls_output) - cmd = '../bin/q -m strict "select count(*) from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - self.assertTrue("Column Count is expected to identical" in e[0]) - - self.cleanup(tmpfile) - - def test_strict_mode_too_large_specific_column_count(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , -m strict -c 4 "select count(*) from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - self.assertEquals( - e[0], "Strict mode. Column count is expected to be 4 but is 3") - - self.cleanup(tmpfile) - - def test_strict_mode_too_small_specific_column_count(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , -m strict -c 2 "select count(*) from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - self.assertEquals( - e[0], "Strict mode. Column count is expected to be 2 but is 3") - - self.cleanup(tmpfile) - - def test_relaxed_mode_missing_columns_in_header(self): - tmpfile = self.create_file_with_data( - sample_data_with_missing_header_names) - cmd = '../bin/q -d , -m relaxed "select count(*) from %s" -H -A' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 4) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], 'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1].strip(), '`name` - text') - self.assertEquals(o[2].strip(), '`value1` - int') - self.assertEquals(o[3].strip(), '`c3` - int') - - self.cleanup(tmpfile) - - def test_strict_mode_missing_columns_in_header(self): - tmpfile = self.create_file_with_data( - sample_data_with_missing_header_names) - cmd = '../bin/q -d , -m strict "select count(*) from %s" -H -A' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(o), 0) - self.assertEquals(len(e), 1) - - self.assertEquals( - e[0], 'Strict mode. Header row contains less columns than expected column count(2 vs 3)') - - self.cleanup(tmpfile) - - def test_output_delimiter_with_missing_fields(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , "select * from %s" -D ";"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], 'a;1;0') - self.assertEquals(o[1], 'b;2;0') - self.assertEquals(o[2], 'c;;0') - - self.cleanup(tmpfile) - - def test_handling_of_null_integers(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , "select avg(c2) from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 1) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], '1.5') - - self.cleanup(tmpfile) - - def test_empty_integer_values_converted_to_null(self): - tmpfile = self.create_file_with_data(sample_data_no_header) - cmd = '../bin/q -d , "select * from %s where c2 is null"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 1) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], 'c,,0') - - self.cleanup(tmpfile) - - def test_empty_string_values_not_converted_to_null(self): - tmpfile = self.create_file_with_data( - sample_data_with_empty_string_no_header) - cmd = '../bin/q -d , "select * from %s where c2 == %s"' % ( - tmpfile.name, "''") - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 1) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], 'c,,0') - - self.cleanup(tmpfile) - - def test_relaxed_mode_detected_columns(self): - tmpfile = self.create_file_with_data(uneven_ls_output) - cmd = '../bin/q -m relaxed "select count(*) from %s" -A' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - - table_name_row = o[0] - column_rows = o[1:] - - self.assertEquals(len(column_rows), 11) - - column_tuples = [x.strip().split(" ") for x in column_rows] - column_info = [(x[0], x[2]) for x in column_tuples] - column_names = [x[0] for x in column_tuples] - column_types = [x[2] for x in column_tuples] - - self.assertEquals(column_names, ['`c%s`' % x for x in xrange(1, 12)]) - self.assertEquals(column_types, [ - 'text', 'int', 'text', 'text', 'int', 'text', 'int', 'int', 'text', 'text', 'text']) - - self.cleanup(tmpfile) - - def test_relaxed_mode_detected_columns_with_specific_column_count(self): - tmpfile = self.create_file_with_data(uneven_ls_output) - cmd = '../bin/q -m relaxed "select count(*) from %s" -A -c 9' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - - table_name_row = o[0] - column_rows = o[1:] - - self.assertEquals(len(column_rows), 9) - - column_tuples = [x.strip().split(" ") for x in column_rows] - column_info = [(x[0], x[2]) for x in column_tuples] - column_names = [x[0] for x in column_tuples] - column_types = [x[2] for x in column_tuples] - - self.assertEquals(column_names, ['`c%s`' % x for x in xrange(1, 10)]) - self.assertEquals( - column_types, ['text', 'int', 'text', 'text', 'int', 'text', 'int', 'int', 'text']) - - self.cleanup(tmpfile) - - def test_relaxed_mode_last_column_data_with_specific_column_count(self): - tmpfile = self.create_file_with_data(uneven_ls_output) - cmd = '../bin/q -m relaxed "select c9 from %s" -c 9' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 9) - self.assertEquals(len(e), 0) - - expected_output = ["/selinux", "/mnt", "/srv", "/lost+found", '"/initrd.img.old -> /boot/initrd.img-3.8.0-19-generic"', - "/cdrom", "/home", '"/vmlinuz -> boot/vmlinuz-3.8.0-19-generic"', '"/initrd.img -> boot/initrd.img-3.8.0-19-generic"'] - - self.assertEquals(o, expected_output) - - self.cleanup(tmpfile) - - def test_1_column_warning_in_relaxed_mode(self): - tmpfile = self.create_file_with_data(one_column_data) - cmd = '../bin/q -m relaxed "select c1 from %s" -d ,' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 1) - self.assertEquals(len(o),2) - - self.assertEquals(e[0],"Warning: column count is one - did you provide the correct delimiter?") - self.assertEquals(o[0],'data without commas 1') - self.assertEquals(o[1],'data without commas 2') - - self.cleanup(tmpfile) - - def test_1_column_warning_in_strict_mode(self): - tmpfile = self.create_file_with_data(one_column_data) - cmd = '../bin/q -m relaxed "select c1 from %s" -d , -m strict' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 1) - self.assertEquals(len(o),2) - - self.assertEquals(e[0],"Warning: column count is one - did you provide the correct delimiter?") - self.assertEquals(o[0],'data without commas 1') - self.assertEquals(o[1],'data without commas 2') - - self.cleanup(tmpfile) - - - def test_1_column_warning_suppression_in_relaxed_mode_when_column_count_is_specific(self): - tmpfile = self.create_file_with_data(one_column_data) - cmd = '../bin/q -m relaxed "select c1 from %s" -d , -m relaxed -c 1' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'data without commas 1') - self.assertEquals(o[1],'data without commas 2') - - self.cleanup(tmpfile) - - def test_1_column_warning_suppression_in_strict_mode_when_column_count_is_specific(self): - tmpfile = self.create_file_with_data(one_column_data) - cmd = '../bin/q -m relaxed "select c1 from %s" -d , -m strict -c 1' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o),2) - - self.assertEquals(o[0],'data without commas 1') - self.assertEquals(o[1],'data without commas 2') - - self.cleanup(tmpfile) - - def test_fluffy_mode(self): - tmpfile = self.create_file_with_data(uneven_ls_output) - cmd = '../bin/q -m fluffy "select c9 from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 9) - self.assertEquals(len(e), 0) - - expected_output = ["/selinux", "/mnt", "/srv", "/lost+found", - "/initrd.img.old", "/cdrom", "/home", "/vmlinuz", "/initrd.img"] - - self.assertEquals(o, expected_output) - - self.cleanup(tmpfile) - - def test_fluffy_mode_column_count_mismatch(self): - data_row = "column1 column2 column3 column4" - data_list = [data_row] * 1000 - data_list[950] = "column1 column2 column3 column4 column5" - tmpfile = self.create_file_with_data("\n".join(data_list)) - - cmd = '../bin/q -m fluffy "select * from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode,0) - self.assertEquals(len(o),0) - self.assertEquals(len(e),1) - self.assertTrue(e[0].startswith("Deprecated fluffy mode")) - self.assertTrue(' row 951 ' in e[0]) - - self.cleanup(tmpfile) - - def test_strict_mode_column_count_mismatch__less_columns(self): - data_row = "column1 column2 column3 column4" - data_list = [data_row] * 1000 - data_list[750] = "column1 column3 column4" - tmpfile = self.create_file_with_data("\n".join(data_list)) - - cmd = '../bin/q -m strict "select * from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode,0) - self.assertEquals(len(o),0) - self.assertEquals(len(e),1) - self.assertTrue(e[0].startswith("Strict mode - Expected 4 columns instead of 3 columns")) - self.assertTrue(' row 751.' in e[0]) - - self.cleanup(tmpfile) - - def test_strict_mode_column_count_mismatch__more_columns(self): - data_row = "column1 column2 column3 column4" - data_list = [data_row] * 1000 - data_list[750] = "column1 column2 column3 column4 column5" - tmpfile = self.create_file_with_data("\n".join(data_list)) - - cmd = '../bin/q -m strict "select * from %s"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode,0) - self.assertEquals(len(o),0) - self.assertEquals(len(e),1) - self.assertTrue(e[0].startswith("Strict mode - Expected 4 columns instead of 5 columns")) - self.assertTrue(' row 751.' in e[0]) - - self.cleanup(tmpfile) - - - -class FormattingTests(AbstractQTestCase): - - def test_column_formatting(self): - cmd = 'seq 1 10 | ../bin/q -f 1=%4.3f,2=%4.3f "select sum(c1),avg(c1) from -" -c 1' - - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 1) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], '55.000 5.500') - - def test_column_formatting_with_output_header(self): - perl_regex = "'s/1\n/column_name\n1\n/;'" - cmd = 'seq 1 10 | perl -pe ' + perl_regex + ' | ../bin/q -f 1=%4.3f,2=%4.3f "select sum(column_name) mysum,avg(column_name) myavg from -" -c 1 -H -O' - - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 2) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], 'mysum myavg') - self.assertEquals(o[1], '55.000 5.500') - - def test_failure_to_parse_universal_newlines_without_explicit_flag(self): - data = 'permalink,company,numEmps,category,city,state,fundedDate,raisedAmt,raisedCurrency,round\rlifelock,LifeLock,,web,Tempe,AZ,1-May-07,6850000,USD,b\rlifelock,LifeLock,,web,Tempe,AZ,1-Oct-06,6000000,USD,a\rlifelock,LifeLock,,web,Tempe,AZ,1-Jan-08,25000000,USD,c\rmycityfaces,MyCityFaces,7,web,Scottsdale,AZ,1-Jan-08,50000,USD,seed\rflypaper,Flypaper,,web,Phoenix,AZ,1-Feb-08,3000000,USD,a\rinfusionsoft,Infusionsoft,105,software,Gilbert,AZ,1-Oct-07,9000000,USD,a' - tmp_data_file = self.create_file_with_data(data) - - cmd = '../bin/q -d , -H "select * from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertNotEquals(retcode, 0) - self.assertEquals(len(e), 1) - self.assertEquals(len(o), 0) - - self.assertTrue(e[0].startswith('Data contains universal newlines')) - - self.cleanup(tmp_data_file) - - def test_universal_newlines_parsing_flag(self): - expected_output = [['lifelock', 'LifeLock', '', 'web', 'Tempe', 'AZ', '1-May-07', '6850000', 'USD', 'b'], - ['lifelock', 'LifeLock', '', 'web', 'Tempe', 'AZ', '1-Oct-06', '6000000', 'USD', 'a'], - ['lifelock', 'LifeLock', '', 'web', 'Tempe', 'AZ', '1-Jan-08', '25000000', 'USD', 'c'], - ['mycityfaces', 'MyCityFaces', '7', 'web', 'Scottsdale', 'AZ', '1-Jan-08', '50000', 'USD', 'seed'], - ['flypaper', 'Flypaper', '', 'web', 'Phoenix', 'AZ', '1-Feb-08', '3000000', 'USD', 'a'], - ['infusionsoft', 'Infusionsoft', '105', 'software', 'Gilbert', 'AZ', '1-Oct-07', '9000000', 'USD', 'a']] - - data = 'permalink,company,numEmps,category,city,state,fundedDate,raisedAmt,raisedCurrency,round\rlifelock,LifeLock,,web,Tempe,AZ,1-May-07,6850000,USD,b\rlifelock,LifeLock,,web,Tempe,AZ,1-Oct-06,6000000,USD,a\rlifelock,LifeLock,,web,Tempe,AZ,1-Jan-08,25000000,USD,c\rmycityfaces,MyCityFaces,7,web,Scottsdale,AZ,1-Jan-08,50000,USD,seed\rflypaper,Flypaper,,web,Phoenix,AZ,1-Feb-08,3000000,USD,a\rinfusionsoft,Infusionsoft,105,software,Gilbert,AZ,1-Oct-07,9000000,USD,a' - tmp_data_file = self.create_file_with_data(data) - - cmd = '../bin/q -d , -H -U "select permalink,company,numEmps,category,city,state,fundedDate,raisedAmt,raisedCurrency,round from %s"' % tmp_data_file.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode,0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 6) - - actual_output = map(lambda row: row.split(","),o) - - self.assertEquals(actual_output,expected_output) - - self.cleanup(tmp_data_file) - - -class SqlTests(AbstractQTestCase): - - def test_find_example(self): - tmpfile = self.create_file_with_data(find_output) - cmd = '../bin/q "select c5,c6,sum(c7)/1024.0/1024 as total from %s group by c5,c6 order by total desc"' % tmpfile.name - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - self.assertEquals(len(e), 0) - - self.assertEquals(o[0], 'mapred mapred 0.93895816803') - self.assertEquals(o[1], 'root root 0.02734375') - self.assertEquals(o[2], 'harel harel 0.0108880996704') - - self.cleanup(tmpfile) - - def test_join_example(self): - cmd = '../bin/q "select myfiles.c8,emails.c2 from ../examples/exampledatafile myfiles join ../examples/group-emails-example emails on (myfiles.c4 = emails.c1) where myfiles.c8 = \'ppp\'"' - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 2) - - self.assertEquals(o[0], 'ppp dip.1@otherdomain.com') - self.assertEquals(o[1], 'ppp dip.2@otherdomain.com') - - def test_join_example_with_output_header(self): - cmd = '../bin/q -O "select myfiles.c8 aaa,emails.c2 bbb from ../examples/exampledatafile myfiles join ../examples/group-emails-example emails on (myfiles.c4 = emails.c1) where myfiles.c8 = \'ppp\'"' - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(o), 3) - - self.assertEquals(o[0], 'aaa bbb') - self.assertEquals(o[1], 'ppp dip.1@otherdomain.com') - self.assertEquals(o[2], 'ppp dip.2@otherdomain.com') - - def test_self_join1(self): - tmpfile = self.create_file_with_data("\n".join(["%s 9000" % i for i in range(0,10)])) - cmd = '../bin/q "select * from %s a1 join %s a2 on (a1.c1 = a2.c1)"' % (tmpfile.name,tmpfile.name) - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 10) - - self.cleanup(tmpfile) - - def test_self_join_reuses_table(self): - tmpfile = self.create_file_with_data("\n".join(["%s 9000" % i for i in range(0,10)])) - cmd = '../bin/q "select * from %s a1 join %s a2 on (a1.c1 = a2.c1)" -A' % (tmpfile.name,tmpfile.name) - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 3) - - self.assertEquals(o[0],'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1],' `c1` - int') - self.assertEquals(o[2],' `c2` - int') - - self.cleanup(tmpfile) - - def test_self_join2(self): - tmpfile1 = self.create_file_with_data("\n".join(["%s 9000" % i for i in range(0,10)])) - cmd = '../bin/q "select * from %s a1 join %s a2 on (a1.c2 = a2.c2)"' % (tmpfile1.name,tmpfile1.name) - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 10*10) - - self.cleanup(tmpfile1) - - tmpfile2 = self.create_file_with_data("\n".join(["%s 9000" % i for i in range(0,10)])) - cmd = '../bin/q "select * from %s a1 join %s a2 on (a1.c2 = a2.c2) join %s a3 on (a1.c2 = a3.c2)"' % (tmpfile2.name,tmpfile2.name,tmpfile2.name) - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 10*10*10) - - self.cleanup(tmpfile2) - - def test_disable_column_type_detection(self): - tmpfile = self.create_file_with_data('''regular_text,text_with_digits1,text_with_digits2,float_number -"regular text 1",67,"67",12.3 -"regular text 2",067,"067",22.3 -"regular text 3",123,"123",33.4 -"regular text 4",-123,"-123",0122.2 -''') - - # Check original column type detection - cmd = '../bin/q -A -d , -H "select * from %s"' % (tmpfile.name) - - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 5) - - - self.assertEquals(o[0],'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1],' `regular_text` - text') - self.assertEquals(o[2],' `text_with_digits1` - int') - self.assertEquals(o[3],' `text_with_digits2` - int') - self.assertEquals(o[4],' `float_number` - float') - - # Check column types detected when actual detection is disabled - cmd = '../bin/q -A -d , -H --as-text "select * from %s"' % (tmpfile.name) - - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 5) - - self.assertEquals(o[0],'Table for file: %s' % tmpfile.name) - self.assertEquals(o[1],' `regular_text` - text') - self.assertEquals(o[2],' `text_with_digits1` - text') - self.assertEquals(o[3],' `text_with_digits2` - text') - self.assertEquals(o[4],' `float_number` - text') - - # Get actual data with regular detection - cmd = '../bin/q -d , -H "select * from %s"' % (tmpfile.name) - - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 4) - - self.assertEquals(o[0],"regular text 1,67,67,12.3"); - self.assertEquals(o[1],"regular text 2,67,67,22.3"); - self.assertEquals(o[2],"regular text 3,123,123,33.4"); - self.assertEquals(o[3],"regular text 4,-123,-123,122.2"); - - # Get actual data without detection - cmd = '../bin/q -d , -H --as-text "select * from %s"' % (tmpfile.name) - - retcode, o, e = run_command(cmd) - - self.assertEquals(retcode, 0) - self.assertEquals(len(e), 0) - self.assertEquals(len(o), 4) - - self.assertEquals(o[0],"regular text 1,67,67,12.3"); - self.assertEquals(o[1],"regular text 2,067,067,22.3"); - self.assertEquals(o[2],"regular text 3,123,123,33.4"); - self.assertEquals(o[3],"regular text 4,-123,-123,0122.2"); - - self.cleanup(tmpfile) - - -class BasicModuleTests(AbstractQTestCase): - - def test_simple_query(self): - tmpfile = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - - q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) - r = q.execute('select * from %s' % tmpfile.name) - - self.assertTrue(r.status == 'ok') - self.assertEquals(len(r.warnings),0) - self.assertEquals(len(r.data),2) - self.assertEquals(r.metadata.output_column_name_list,['a','b','c']) - self.assertEquals(r.data,[(1,2,3),(4,5,6)]) - self.assertEquals(len(r.metadata.data_loads),1) - self.assertEquals(r.metadata.data_loads[0].filename,tmpfile.name) - - self.cleanup(tmpfile) - - def test_loaded_data_reuse(self): - tmpfile = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - - q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) - r1 = q.execute('select * from %s' % tmpfile.name) - - r2 = q.execute('select * from %s' % tmpfile.name) - - self.assertTrue(r1.status == 'ok') - self.assertEquals(len(r1.warnings),0) - self.assertEquals(len(r1.data),2) - self.assertEquals(r1.metadata.output_column_name_list,['a','b','c']) - self.assertEquals(r1.data,[(1,2,3),(4,5,6)]) - self.assertEquals(r1.metadata.data_loads[0].filename,tmpfile.name) - - self.assertTrue(r2.status == 'ok') - self.assertEquals(len(r1.metadata.data_loads),1) - self.assertEquals(r1.metadata.data_loads[0].filename,tmpfile.name) - self.assertEquals(len(r2.metadata.data_loads),0) - self.assertEquals(r2.data,r1.data) - self.assertEquals(r2.metadata.output_column_name_list,r2.metadata.output_column_name_list) - self.assertEquals(len(r2.warnings),0) - - self.cleanup(tmpfile) - - def test_stdin_injection(self): - tmpfile = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - - q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) - r = q.execute('select * from -',stdin_file=file(tmpfile.name,'rb')) - - self.assertTrue(r.status == 'ok') - self.assertEquals(len(r.warnings),0) - self.assertEquals(len(r.data),2) - self.assertEquals(r.metadata.output_column_name_list,['a','b','c']) - self.assertEquals(r.data,[(1,2,3),(4,5,6)]) - self.assertEquals(len(r.metadata.data_loads),1) - self.assertEquals(r.metadata.data_loads[0].filename,'-') - - self.cleanup(tmpfile) - - def test_named_stdin_injection(self): - tmpfile = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - - q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) - r = q.execute('select a from my_stdin_data',stdin_file=file(tmpfile.name,'rb'),stdin_filename='my_stdin_data') - - self.assertTrue(r.status == 'ok') - self.assertEquals(len(r.warnings),0) - self.assertEquals(len(r.data),2) - self.assertEquals(r.metadata.output_column_name_list,['a']) - self.assertEquals(r.data,[(1,),(4,)]) - self.assertEquals(len(r.metadata.data_loads),1) - self.assertEquals(r.metadata.data_loads[0].filename,'my_stdin_data') - - self.cleanup(tmpfile) - - def test_stdin_injection_isolation(self): - tmpfile1 = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - tmpfile2 = self.create_file_with_data("d e f\n7 8 9\n10 11 12") - - q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) - r1 = q.execute('select * from -',stdin_file=file(tmpfile1.name,'rb')) - - self.assertTrue(r1.status == 'ok') - self.assertEquals(len(r1.warnings),0) - self.assertEquals(len(r1.data),2) - self.assertEquals(r1.metadata.output_column_name_list,['a','b','c']) - self.assertEquals(r1.data,[(1,2,3),(4,5,6)]) - self.assertEquals(len(r1.metadata.data_loads),1) - self.assertEquals(r1.metadata.data_loads[0].filename,'-') - - r2 = q.execute('select * from -',stdin_file=file(tmpfile2.name,'rb')) - - self.assertTrue(r2.status == 'ok') - self.assertEquals(len(r2.warnings),0) - self.assertEquals(len(r2.data),2) - self.assertEquals(r2.metadata.output_column_name_list,['d','e','f']) - self.assertEquals(r2.data,[(7,8,9),(10,11,12)]) - # There should be another data load, even though it's the same 'filename' as before - self.assertEquals(len(r2.metadata.data_loads),1) - self.assertEquals(r2.metadata.data_loads[0].filename,'-') - - self.cleanup(tmpfile1) - self.cleanup(tmpfile2) - - def test_multiple_stdin_injection(self): - tmpfile1 = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - tmpfile2 = self.create_file_with_data("d e f\n7 8 9\n10 11 12") - - q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) - r1 = q.execute('select * from my_stdin_data1',stdin_file=file(tmpfile1.name,'rb'),stdin_filename='my_stdin_data1') - - self.assertTrue(r1.status == 'ok') - self.assertEquals(len(r1.warnings),0) - self.assertEquals(len(r1.data),2) - self.assertEquals(r1.metadata.output_column_name_list,['a','b','c']) - self.assertEquals(r1.data,[(1,2,3),(4,5,6)]) - self.assertEquals(len(r1.metadata.data_loads),1) - self.assertEquals(r1.metadata.data_loads[0].filename,'my_stdin_data1') - - r2 = q.execute('select * from my_stdin_data2',stdin_file=file(tmpfile2.name,'rb'),stdin_filename='my_stdin_data2') - - self.assertTrue(r2.status == 'ok') - self.assertEquals(len(r2.warnings),0) - self.assertEquals(len(r2.data),2) - self.assertEquals(r2.metadata.output_column_name_list,['d','e','f']) - self.assertEquals(r2.data,[(7,8,9),(10,11,12)]) - # There should be another data load, even though it's the same 'filename' as before - self.assertEquals(len(r2.metadata.data_loads),1) - self.assertEquals(r2.metadata.data_loads[0].filename,'my_stdin_data2') - - r3 = q.execute('select aa.*,bb.* from my_stdin_data1 aa join my_stdin_data2 bb') - - self.assertTrue(r3.status == 'ok') - self.assertEquals(len(r3.warnings),0) - self.assertEquals(len(r3.data),4) - self.assertEquals(r3.metadata.output_column_name_list,['a','b','c','d','e','f']) - self.assertEquals(r3.data,[(1,2,3,7,8,9),(1,2,3,10,11,12),(4,5,6,7,8,9),(4,5,6,10,11,12)]) - self.assertEquals(len(r3.metadata.data_loads),0) - - self.cleanup(tmpfile1) - self.cleanup(tmpfile2) - - def test_different_input_params_for_different_files(self): - tmpfile1 = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - tmpfile2 = self.create_file_with_data("7\t8\t9\n10\t11\t12") - - q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) - - q.load_data(tmpfile1.name,QInputParams(skip_header=True,delimiter=' ')) - q.load_data(tmpfile2.name,QInputParams(skip_header=False,delimiter='\t')) - - r = q.execute('select aa.*,bb.* from %s aa join %s bb' % (tmpfile1.name,tmpfile2.name)) - - self.assertTrue(r.status == 'ok') - self.assertEquals(len(r.warnings),0) - self.assertEquals(len(r.data),4) - self.assertEquals(r.metadata.output_column_name_list,['a','b','c','c1','c2','c3']) - self.assertEquals(r.data,[(1,2,3,7,8,9),(1,2,3,10,11,12),(4,5,6,7,8,9),(4,5,6,10,11,12)]) - self.assertEquals(len(r.metadata.data_loads),0) - - self.cleanup(tmpfile1) - self.cleanup(tmpfile2) - - def test_different_input_params_for_different_files(self): - tmpfile1 = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - tmpfile2 = self.create_file_with_data("7\t8\t9\n10\t11\t12") - - q = QTextAsData() - - q.load_data(tmpfile1.name,QInputParams(skip_header=True,delimiter=' ')) - q.load_data(tmpfile2.name,QInputParams(skip_header=False,delimiter='\t')) - - r = q.execute('select aa.*,bb.* from %s aa join %s bb' % (tmpfile1.name,tmpfile2.name)) - - self.assertTrue(r.status == 'ok') - self.assertEquals(len(r.warnings),0) - self.assertEquals(len(r.data),4) - self.assertEquals(r.metadata.output_column_name_list,['a','b','c','c1','c2','c3']) - self.assertEquals(r.data,[(1,2,3,7,8,9),(1,2,3,10,11,12),(4,5,6,7,8,9),(4,5,6,10,11,12)]) - self.assertEquals(len(r.metadata.data_loads),0) - - self.cleanup(tmpfile1) - self.cleanup(tmpfile2) - - def test_input_params_override(self): - tmpfile = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - - default_input_params = QInputParams() - - for k in default_input_params.__dict__.keys(): - setattr(default_input_params,k,'GARBAGE') - - q = QTextAsData(default_input_params) - - r = q.execute('select * from %s' % tmpfile.name) - - self.assertTrue(r.status == 'error') - - overwriting_input_params = QInputParams(skip_header=True,delimiter=' ') - - r2 = q.execute('select * from %s' % tmpfile.name,input_params=overwriting_input_params) - - self.assertTrue(r2.status == 'ok') - self.assertEquals(len(r2.warnings),0) - self.assertEquals(len(r2.data),2) - self.assertEquals(r2.metadata.output_column_name_list,['a','b','c']) - self.assertEquals(r2.data,[(1,2,3),(4,5,6)]) - self.assertEquals(len(r2.metadata.data_loads),1) - self.assertEquals(r2.metadata.data_loads[0].filename,tmpfile.name) - - self.cleanup(tmpfile) - - def test_input_params_merge(self): - input_params = QInputParams() - - for k in input_params.__dict__.keys(): - setattr(input_params,k,'GARBAGE') - - merged_input_params = input_params.merged_with(QInputParams()) - - for k in merged_input_params.__dict__.keys(): - self.assertTrue(getattr(merged_input_params,k) != 'GARBAGE') - - for k in input_params.__dict__.keys(): - self.assertTrue(getattr(merged_input_params,k) != 'GARBAGE') - - def test_table_analysis_with_syntax_error(self): - - q = QTextAsData() - - q_output = q.analyze("bad syntax") - - self.assertTrue(q_output.status == 'error') - self.assertTrue(q_output.error.msg.startswith('query error')) - - def test_execute_response(self): - tmpfile = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - - q = QTextAsData() - - q_output = q.execute("select a,c from %s" % tmpfile.name,QInputParams(skip_header=True)) - - self.assertTrue(q_output.status == 'ok') - self.assertTrue(q_output.error is None) - self.assertEquals(len(q_output.warnings),0) - self.assertEquals(len(q_output.data),2) - self.assertEquals(q_output.data,[ (1,3),(4,6) ]) - self.assertTrue(q_output.metadata is not None) - - metadata = q_output.metadata - - self.assertEquals(metadata.output_column_name_list, [ 'a','c']) - self.assertEquals(len(metadata.data_loads),1) - self.assertEquals(len(metadata.table_structures),1) - - table_structure = metadata.table_structures[0] - - self.assertEquals(table_structure.column_names,[ 'a','b','c']) - self.assertEquals(table_structure.column_types,[ 'int','int','int']) - self.assertEquals(table_structure.filenames_str,tmpfile.name) - self.assertTrue(len(table_structure.materialized_files.keys()),1) - self.assertTrue(table_structure.materialized_files[tmpfile.name].filename,tmpfile.name) - self.assertFalse(table_structure.materialized_files[tmpfile.name].is_stdin) - - self.cleanup(tmpfile) - - def test_analyze_response(self): - tmpfile = self.create_file_with_data("a b c\n1 2 3\n4 5 6") - - q = QTextAsData() - - q_output = q.analyze("select a,c from %s" % tmpfile.name,QInputParams(skip_header=True)) - - self.assertTrue(q_output.status == 'ok') - self.assertTrue(q_output.error is None) - self.assertEquals(len(q_output.warnings),0) - self.assertEquals(len(q_output.data),2) - self.assertEquals(q_output.data,[ (1,3),(4,6) ]) - self.assertTrue(q_output.metadata is not None) - - metadata = q_output.metadata - - self.assertEquals(metadata.output_column_name_list, [ 'a','c']) - self.assertEquals(len(metadata.data_loads),1) - self.assertEquals(len(metadata.table_structures),1) - - table_structure = metadata.table_structures[0] - - self.assertEquals(table_structure.column_names,[ 'a','b','c']) - self.assertEquals(table_structure.column_types,[ 'int','int','int']) - self.assertEquals(table_structure.filenames_str,tmpfile.name) - self.assertTrue(len(table_structure.materialized_files.keys()),1) - self.assertTrue(table_structure.materialized_files[tmpfile.name].filename,tmpfile.name) - self.assertFalse(table_structure.materialized_files[tmpfile.name].is_stdin) - - self.cleanup(tmpfile) - - def test_load_data_from_string(self): - input_str = 'column1,column2,column3\n' + '\n'.join(['value1,2.5,value3'] * 1000) - - q = QTextAsData() - - q.load_data_from_string('my_data',input_str,QInputParams(skip_header=True,delimiter=',')) - - q_output = q.execute('select column2,column3 from my_data') - - self.assertTrue(q_output.status == 'ok') - self.assertTrue(q_output.error is None) - self.assertEquals(len(q_output.warnings),0) - self.assertTrue(len(q_output.data),1000) - self.assertEquals(len(set(q_output.data)),1) - self.assertEquals(list(set(q_output.data))[0],(2.5,'value3')) - - metadata = q_output.metadata - - self.assertTrue(metadata.output_column_name_list,['column2','column3']) - self.assertEquals(len(metadata.data_loads),0) - self.assertTrue(len(metadata.table_structures),1) - - table_structure = metadata.table_structures[0] - - self.assertEquals(table_structure.column_names,['column1','column2','column3']) - self.assertEquals(table_structure.column_types,['text','float','text']) - self.assertEquals(table_structure.filenames_str,'my_data') - self.assertTrue(len(table_structure.materialized_files.keys()),1) - self.assertTrue(table_structure.materialized_files['my_data'].filename,'my_data') - self.assertTrue(table_structure.materialized_files['my_data'].is_stdin) - - -def suite(): - tl = unittest.TestLoader() - basic_stuff = tl.loadTestsFromTestCase(BasicTests) - parsing_mode = tl.loadTestsFromTestCase(ParsingModeTests) - sql = tl.loadTestsFromTestCase(SqlTests) - formatting = tl.loadTestsFromTestCase(FormattingTests) - basic_module_stuff = tl.loadTestsFromTestCase(BasicModuleTests) - save_db_to_disk_tests = tl.loadTestsFromTestCase(SaveDbToDiskTests) - return unittest.TestSuite([basic_module_stuff, basic_stuff, parsing_mode, sql, formatting,save_db_to_disk_tests]) - -if __name__ == '__main__': - test_runner = unittest.TextTestRunner(verbosity=2) - result = test_runner.run(suite()) - sys.exit(not result.wasSuccessful()) diff --git a/test/test_suite.py b/test/test_suite.py new file mode 100755 index 00000000..aaa7476f --- /dev/null +++ b/test/test_suite.py @@ -0,0 +1,5764 @@ +#!/usr/bin/env python3 + +# +# test suite for q. +# +# Prefer end-to-end tests, running the actual q command and testing stdout/stderr, and the return code. +# Some utilities are provided for making that easy, see other tests for examples. +# +# Q_EXECUTABLE env var can be used to inject the path of q. This allows full e2e testing of the resulting executable +# instead of just testing the python code. +# +# Tests are compatible with Linux and OSX (path separators, tmp folder, etc.). + +from __future__ import print_function + +import collections +import functools +import tempfile +import unittest +import random +import json +import uuid +from collections import OrderedDict +from json import JSONEncoder +from subprocess import PIPE, Popen, STDOUT +import sys +import os +import time +from tempfile import NamedTemporaryFile +import locale +import pprint +import six +from six.moves import range +import codecs +import itertools +from gzip import GzipFile +import pytest +import uuid +import sqlite3 +import re +import collections + +sys.path.append(os.path.join(os.path.abspath(os.path.dirname(sys.argv[0])),'..','bin')) +from bin.q import QTextAsData, QOutput, QOutputPrinter, QInputParams, DataStream, Sqlite3DB + +# q uses this encoding as the default output encoding. Some of the tests use it in order to +# make sure that the output is correctly encoded +SYSTEM_ENCODING = locale.getpreferredencoding() + +EXAMPLES = os.path.abspath(os.path.join(os.getcwd(), 'examples')) + +Q_EXECUTABLE = os.getenv('Q_EXECUTABLE', os.path.abspath('./bin/q.py')) +Q_SKIP_EXECUTABLE_VALIDATION = os.getenv('Q_SKIP_EXECUTABLE_VALIDATION','false') + +if not Q_SKIP_EXECUTABLE_VALIDATION == 'true': + Q_EXECUTABLE = os.path.abspath(Q_EXECUTABLE) + if not os.path.exists(Q_EXECUTABLE): + raise Exception("q executable must reside in {}".format(Q_EXECUTABLE)) +else: + Q_EXECUTABLE = os.getenv('Q_EXECUTABLE') + # Skip checking of executable (useful for testing that q is in the path) + pass + +DEBUG = '-v' in sys.argv +if os.environ.get('Q_DEBUG'): + DEBUG = True + +def batch(iterable, n=1): + r = [] + l = len(iterable) + for ndx in range(0, l, n): + r += [iterable[ndx:min(ndx + n, l)]] + return r + +def partition(pred, iterable): + t1, t2 = itertools.tee(iterable) + return list(itertools.filterfalse(pred, t1)), list(filter(pred, t2)) + +def run_command(cmd_to_run,env_to_inject=None): + global DEBUG + if DEBUG: + print("CMD: {}".format(cmd_to_run)) + + if env_to_inject is None: + env_to_inject = os.environ + + env = env_to_inject + + p = Popen(cmd_to_run, stdout=PIPE, stderr=PIPE, shell=True,env=env) + o, e = p.communicate() + # remove last newline + o = o.rstrip() + e = e.strip() + # split rows + if o != six.b(''): + o = o.split(six.b(os.linesep)) + else: + o = [] + if e != six.b(''): + e = e.split(six.b(os.linesep)) + else: + e = [] + + res = (p.returncode, o, e) + if DEBUG: + print("RESULT:{}".format(res)) + return res + + +uneven_ls_output = six.b("""drwxr-xr-x 2 root root 4096 Jun 11 2012 /selinux +drwxr-xr-x 2 root root 4096 Apr 19 2013 /mnt +drwxr-xr-x 2 root root 4096 Apr 24 2013 /srv +drwx------ 2 root root 16384 Jun 21 2013 /lost+found +lrwxrwxrwx 1 root root 33 Jun 21 2013 /initrd.img.old -> /boot/initrd.img-3.8.0-19-generic +drwxr-xr-x 2 root root 4096 Jun 21 2013 /cdrom +drwxr-xr-x 3 root root 4096 Jun 21 2013 /home +lrwxrwxrwx 1 root root 29 Jun 21 2013 /vmlinuz -> boot/vmlinuz-3.8.0-19-generic +lrwxrwxrwx 1 root root 32 Jun 21 2013 /initrd.img -> boot/initrd.img-3.8.0-19-generic +""") + + +find_output = six.b("""8257537 32 drwxrwxrwt 218 root root 28672 Mar 1 11:00 /tmp +8299123 4 drwxrwxr-x 2 harel harel 4096 Feb 27 10:06 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/stormdist/testTopology3fad644a-54c0-4def-b19e-77ca97941595-1-1393513576 +8263229 964 -rw-rw-r-- 1 mapred mapred 984569 Feb 27 10:06 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/stormdist/testTopology3fad644a-54c0-4def-b19e-77ca97941595-1-1393513576/stormcode.ser +8263230 4 -rw-rw-r-- 1 harel harel 1223 Feb 27 10:06 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/stormdist/testTopology3fad644a-54c0-4def-b19e-77ca97941595-1-1393513576/stormconf.ser +8299113 4 drwxrwxr-x 2 harel harel 4096 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate +8263406 4 -rw-rw-r-- 1 harel harel 2002 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514168746 +8263476 0 -rw-rw-r-- 1 harel harel 0 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514168746.version +8263607 0 -rw-rw-r-- 1 harel harel 0 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514169735.version +8263533 0 -rw-rw-r-- 1 harel harel 0 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514172733.version +8263604 0 -rw-rw-r-- 1 harel harel 0 Feb 27 10:16 /tmp/1628a3fd-b9fe-4dd1-bcdc-7eb869fe7461/supervisor/localstate/1393514175754.version +""") + + +header_row = six.b('name,value1,value2') +sample_data_rows = [six.b('a,1,0'), six.b('b,2,0'), six.b('c,,0')] +sample_data_rows_with_empty_string = [six.b('a,aaa,0'), six.b('b,bbb,0'), six.b('c,,0')] +sample_data_no_header = six.b("\n").join(sample_data_rows) + six.b("\n") +sample_data_with_empty_string_no_header = six.b("\n").join( + sample_data_rows_with_empty_string) + six.b("\n") +sample_data_with_header = header_row + six.b("\n") + sample_data_no_header +sample_data_with_missing_header_names = six.b("name,value1\n") + sample_data_no_header + +def generate_sample_data_with_header(header): + return header + six.b("\n") + sample_data_no_header + +sample_quoted_data = six.b('''non_quoted regular_double_quoted double_double_quoted escaped_double_quoted multiline_double_double_quoted multiline_escaped_double_quoted +control-value-1 "control-value-2" control-value-3 "control-value-4" control-value-5 "control-value-6" +non-quoted-value "this is a quoted value" "this is a ""double double"" quoted value" "this is an escaped \\"quoted value\\"" "this is a double double quoted ""multiline + value""." "this is an escaped \\"multiline + value\\"." +control-value-1 "control-value-2" control-value-3 "control-value-4" control-value-5 "control-value-6" +''') + +double_double_quoted_data = six.b('''regular_double_quoted double_double_quoted +"this is a quoted value" "this is a quoted value with ""double double quotes""" +''') + +escaped_double_quoted_data = six.b('''regular_double_quoted escaped_double_quoted +"this is a quoted value" "this is a quoted value with \\"escaped double quotes\\"" +''') + +combined_quoted_data = six.b('''regular_double_quoted double_double_quoted escaped_double_quoted +"this is a quoted value" "this is a quoted value with ""double double quotes""" "this is a quoted value with \\"escaped double quotes\\"" +''') + +sample_quoted_data2 = six.b('"quoted data" 23\nunquoted-data 54') + +sample_quoted_data2_with_newline = six.b('"quoted data with\na new line inside it":23\nunquoted-data:54') + +one_column_data = six.b('''data without commas 1 +data without commas 2 +''') + +# Values with leading whitespace +sample_data_rows_with_spaces = [six.b('a,1,0'), six.b(' b, 2,0'), six.b('c,,0')] +sample_data_with_spaces_no_header = six.b("\n").join( + sample_data_rows_with_spaces) + six.b("\n") + +header_row_with_spaces = six.b('name,value 1,value2') +sample_data_with_spaces_with_header = header_row_with_spaces + \ + six.b("\n") + sample_data_with_spaces_no_header + +long_value1 = "23683289372328372328373" +int_value = "2328372328373" +sample_data_with_long_values = "%s\n%s\n%s" % (long_value1,int_value,int_value) + + +def one_column_warning(e): + return e[0].startswith(six.b('Warning: column count is one')) + +def sqlite_dict_factory(cursor, row): + d = {} + for idx, col in enumerate(cursor.description): + d[col[0]] = row[idx] + return d + +class AbstractQTestCase(unittest.TestCase): + + def create_file_with_data(self, data, encoding=None,prefix=None,suffix=None,use_real_path=True): + if encoding is not None: + raise Exception('Deprecated: Encoding must be none') + tmpfile = NamedTemporaryFile(delete=False,prefix=prefix,suffix=suffix) + tmpfile.write(data) + tmpfile.close() + if use_real_path: + tmpfile.name = os.path.realpath(tmpfile.name) + return tmpfile + + def generate_tmpfile_name(self,prefix=None,suffix=None): + tmpfile = NamedTemporaryFile(delete=False,prefix=prefix,suffix=suffix) + os.remove(tmpfile.name) + return os.path.realpath(tmpfile.name) + + def arrays_to_csv_file_content(self,delimiter,header_row_list,cell_list): + all_rows = [delimiter.join(row) for row in [header_row_list] + cell_list] + return six.b("\n").join(all_rows) + + def create_qsql_file_with_content_and_return_filename(self, header_row,cell_list): + csv_content = self.arrays_to_csv_file_content(six.b(','),header_row,cell_list) + tmpfile = self.create_file_with_data(csv_content) + + cmd = '%s -d , -H "select count(*) from %s" -C readwrite' % (Q_EXECUTABLE,tmpfile.name) + r, o, e = run_command(cmd) + self.assertEqual(r, 0) + + created_qsql_filename = '%s.qsql' % tmpfile.name + self.assertTrue(os.path.exists(created_qsql_filename)) + + return created_qsql_filename + + def arrays_to_qsql_file_content(self, header_row,cell_list): + csv_content = self.arrays_to_csv_file_content(six.b(','),header_row,cell_list) + tmpfile = self.create_file_with_data(csv_content) + + cmd = '%s -d , -H "select count(*) from %s" -C readwrite' % (Q_EXECUTABLE,tmpfile.name) + r, o, e = run_command(cmd) + self.assertEqual(r, 0) + + matching_qsql_filename = '%s.qsql' % tmpfile.name + f = open(matching_qsql_filename,'rb') + qsql_file_bytes = f.read() + f.close() + + self.assertEqual(matching_qsql_filename,'%s.qsql' % tmpfile.name) + + return qsql_file_bytes + + def write_file(self,filename,data): + f = open(filename,'wb') + f.write(data) + f.close() + + def create_folder_with_files(self,filename_to_content_dict,prefix, suffix): + name = self.random_tmp_filename(prefix,suffix) + os.makedirs(name) + for filename,content in six.iteritems(filename_to_content_dict): + if os.path.sep in filename: + os.makedirs('%s/%s' % (name,os.path.split(filename)[0])) + f = open(os.path.join(name,filename),'wb') + f.write(content) + f.close() + return name + + def cleanup_folder(self,tmpfolder): + if not tmpfolder.startswith(os.path.realpath('/var/tmp')): + raise Exception('Guard against accidental folder deletions: %s' % tmpfolder) + global DEBUG + if not DEBUG: + print("should have removed tmpfolder %s. Not doing it for the sake of safety. # TODO re-add" % tmpfolder) + pass # os.remove(tmpfolder) + + def cleanup(self, tmpfile): + global DEBUG + if not DEBUG: + os.remove(tmpfile.name) + + def random_tmp_filename(self,prefix,postfix): + # TODO Use more robust method for this + path = '/var/tmp' + return os.path.realpath('%s/%s-%s.%s' % (path,prefix,random.randint(0,1000000000),postfix)) + + + +def get_sqlite_table_list(c,exclude_qcatalog=True): + if exclude_qcatalog: + r = c.execute("select tbl_name from sqlite_master where type='table' and tbl_name != '_qcatalog'").fetchall() + else: + r = c.execute("select tbl_name from sqlite_master where type='table'").fetchall() + + return r + +class SaveToSqliteTests(AbstractQTestCase): + + # Returns a folder with files and a header in each, one column named 'a' + def generate_files_in_folder(self,batch_size, file_count): + numbers = list(range(1, 1 + batch_size * file_count)) + numbers_as_text = batch([str(x) for x in numbers], n=batch_size) + + content_list = list(map(six.b, ['a\n' + "\n".join(x) + '\n' for x in numbers_as_text])) + + filename_list = list(map(lambda x: 'file-%s' % x, range(file_count))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d, 'split-files', 'sqlite-stuff') + return (tmpfolder,filename_list) + + # 11074 3.8.2021 10:53 bin/q.py "select count(*) from xxxx/file-95 left join xxxx/file-96 left join xxxx/file-97 left join xxxx/file-97 left join xxxx/file-98 left join xxxx/*" -c 1 -C readwrite -A + # # fails because it takes qsql files as well + + def test_save_glob_files_to_sqlite(self): + BATCH_SIZE = 50 + FILE_COUNT = 5 + + tmpfolder,filename_list = self.generate_files_in_folder(BATCH_SIZE,FILE_COUNT) + + output_sqlite_file = self.random_tmp_filename("x","sqlite") + + cmd = '%s -H "select count(*) from %s/*" -c 1 -S %s' % (Q_EXECUTABLE,tmpfolder,output_sqlite_file) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + c = sqlite3.connect(output_sqlite_file) + results = c.execute('select a from file_dash_0').fetchall() + self.assertEqual(len(results),BATCH_SIZE*FILE_COUNT) + self.assertEqual(sum(map(lambda x:x[0],results)),sum(range(1,BATCH_SIZE*FILE_COUNT+1))) + tables = get_sqlite_table_list(c) + self.assertEqual(len(tables),1) + + c.close() + + self.cleanup_folder(tmpfolder) + + def test_save_multiple_files_to_sqlite(self): + BATCH_SIZE = 50 + FILE_COUNT = 5 + + tmpfolder,filename_list = self.generate_files_in_folder(BATCH_SIZE,FILE_COUNT) + + output_sqlite_file = self.random_tmp_filename("x","sqlite") + + tables_as_str = " left join ".join(["%s/%s" % (tmpfolder,x) for x in filename_list]) + cmd = '%s -H "select count(*) from %s" -c 1 -S %s' % (Q_EXECUTABLE,tables_as_str,output_sqlite_file) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + c = sqlite3.connect(output_sqlite_file) + + tables = get_sqlite_table_list(c) + self.assertEqual(len(tables), FILE_COUNT) + + for i in range(FILE_COUNT): + results = c.execute('select a from file_dash_%s' % i).fetchall() + self.assertEqual(len(results),BATCH_SIZE) + self.assertEqual(sum(map(lambda x:x[0],results)),sum(range(1+i*BATCH_SIZE,1+(i+1)*BATCH_SIZE))) + + c.close() + + self.cleanup_folder(tmpfolder) + + def test_save_multiple_files_to_sqlite_without_duplicates(self): + BATCH_SIZE = 50 + FILE_COUNT = 5 + + tmpfolder,filename_list = self.generate_files_in_folder(BATCH_SIZE,FILE_COUNT) + + output_sqlite_file = self.random_tmp_filename("x","sqlite") + + tables_as_str = " left join ".join(["%s/%s" % (tmpfolder,x) for x in filename_list]) + + # duplicate the left-joins for all the files, so the query will contain each filename twice + tables_as_str = "%s left join %s" % (tables_as_str,tables_as_str) + + cmd = '%s -H "select count(*) from %s" -c 1 -S %s' % (Q_EXECUTABLE,tables_as_str,output_sqlite_file) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + c = sqlite3.connect(output_sqlite_file) + + tables = get_sqlite_table_list(c) + # total table count should still be FILE_COUNT, even with the duplications + self.assertEqual(len(tables), FILE_COUNT) + + for i in range(FILE_COUNT): + results = c.execute('select a from file_dash_%s' % i).fetchall() + self.assertEqual(len(results),BATCH_SIZE) + self.assertEqual(sum(map(lambda x:x[0],results)),sum(range(1+i*BATCH_SIZE,1+(i+1)*BATCH_SIZE))) + + c.close() + + self.cleanup_folder(tmpfolder) + + def test_sqlite_file_is_not_created_if_some_table_does_not_exist(self): + BATCH_SIZE = 50 + FILE_COUNT = 5 + + tmpfolder,filename_list = self.generate_files_in_folder(BATCH_SIZE,FILE_COUNT) + + output_sqlite_file = self.random_tmp_filename("x","sqlite") + + tables_as_str = " left join ".join(["%s/%s" % (tmpfolder,x) for x in filename_list]) + + tables_as_str = tables_as_str + ' left join %s/non_existent_table' % (tmpfolder) + + cmd = '%s -H "select count(*) from %s" -c 1 -S %s' % (Q_EXECUTABLE,tables_as_str,output_sqlite_file) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 30) + self.assertEqual(len(e), 2) + self.assertEqual(e[0],six.b("Going to save data into a disk database: %s" % output_sqlite_file)) + self.assertEqual(e[1],six.b("No files matching '%s/non_existent_table' have been found" % tmpfolder)) + + self.assertTrue(not os.path.exists(output_sqlite_file)) + + self.cleanup_folder(tmpfolder) + + def test_recurring_glob_and_separate_files_in_same_query_when_writing_to_sqlite(self): + BATCH_SIZE = 50 + FILE_COUNT = 5 + + tmpfolder,filename_list = self.generate_files_in_folder(BATCH_SIZE,FILE_COUNT) + + output_sqlite_file = self.random_tmp_filename("x","sqlite") + + tables_as_str = " left join ".join(["%s/%s" % (tmpfolder,x) for x in filename_list]) + # The same files are left-joined in the query as an additional "left join /*". This should create an additional table + # in the sqlite file, with all the data in it + cmd = '%s -H "select count(*) from %s left join %s/*" -c 1 -S %s' % (Q_EXECUTABLE,tables_as_str,tmpfolder,output_sqlite_file) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + c = sqlite3.connect(output_sqlite_file) + + tables = get_sqlite_table_list(c) + # plus the additional table from the glob + self.assertEqual(len(tables), FILE_COUNT+1) + + # check all the per-file tables + for i in range(FILE_COUNT): + results = c.execute('select a from file_dash_%s' % i).fetchall() + self.assertEqual(len(results),BATCH_SIZE) + self.assertEqual(sum(map(lambda x:x[0],results)),sum(range(1+i*BATCH_SIZE,1+(i+1)*BATCH_SIZE))) + + # ensure the glob-based table exists, with an _2 added to the name, as the original "file_dash_0" already exists in the sqlite db + results = c.execute('select a from file_dash_0_2').fetchall() + self.assertEqual(len(results),FILE_COUNT*BATCH_SIZE) + self.assertEqual(sum(map(lambda x:x[0],results)),sum(range(1,1+FILE_COUNT*BATCH_SIZE))) + c.close() + + self.cleanup_folder(tmpfolder) + + def test_empty_sqlite_handling(self): + fn = self.generate_tmpfile_name("empty",".sqlite") + + c = sqlite3.connect(fn) + c.execute('create table x (a int)').fetchall() + c.execute('drop table x').fetchall() + c.close() + + cmd = '%s "select * from %s"' % (Q_EXECUTABLE,fn) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,88) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + self.assertEqual(e[0],six.b('sqlite file %s has no tables' % fn)) + + def test_storing_to_disk_too_many_qsql_files(self): + BATCH_SIZE = 10 + MAX_ATTACHED_DBS = 5 + FILE_COUNT = MAX_ATTACHED_DBS + 4 + + numbers_as_text = batch([str(x) for x in range(1, 1 + BATCH_SIZE * FILE_COUNT)], n=BATCH_SIZE) + + content_list = map(six.b, ["\n".join(x) for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x, range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d, 'split-files', 'attach-limit') + + for fn in filename_list: + cmd = '%s -c 1 "select count(*) from %s/%s" -C readwrite' % (Q_EXECUTABLE,tmpfolder, fn) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + + output_sqlite_file = self.generate_tmpfile_name("many-sqlites",".sqlite") + + table_refs = list(['select * from %s/%s.qsql' % (tmpfolder,x) for x in filename_list]) + table_refs_str = " UNION ALL ".join(table_refs) + # Limit max attached dbs according to the parameter (must be below the hardcoded sqlite limit, which is 10 when having a standard version compiled) + cmd = '%s "select * from (%s)" -S %s --max-attached-sqlite-databases=%s' % (Q_EXECUTABLE,table_refs_str,output_sqlite_file,MAX_ATTACHED_DBS) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode,0) + self.assertEqual(len(o),0) + self.assertEqual(len(e),4) + + c = sqlite3.connect(output_sqlite_file) + tables_results = c.execute("select tbl_name from sqlite_master where type='table'").fetchall() + table_names = list(sorted([x[0] for x in tables_results])) + self.assertEqual(len(table_names),FILE_COUNT) + + for i,tn in enumerate(table_names): + self.assertEqual(tn,'file_dash_%s' % i) + + table_content = c.execute('select * from %s' % tn).fetchall() + self.assertEqual(len(table_content),BATCH_SIZE) + + cmd = '%s "select * from %s:::%s"' % (Q_EXECUTABLE,output_sqlite_file,tn) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),BATCH_SIZE) + self.assertEqual(o,list([six.b(str(x)) for x in range(1 + i*BATCH_SIZE,1+(i+1)*BATCH_SIZE)])) + + self.cleanup_folder(tmpfolder) + + def test_storing_to_disk_too_many_sqlite_files(self): + # a variation of test_storing_to_disk_too_many_qsql_files, which deletes the qcatalog file from the caches, + # so they'll be just regular sqlite files + + BATCH_SIZE = 10 + MAX_ATTACHED_DBS = 5 + FILE_COUNT = MAX_ATTACHED_DBS + 4 + + numbers_as_text = batch([str(x) for x in range(1, 1 + BATCH_SIZE * FILE_COUNT)], n=BATCH_SIZE) + + content_list = map(six.b, ["\n".join(x) for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x, range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d, 'split-files', 'attach-limit') + + for fn in filename_list: + cmd = '%s -c 1 "select count(*) from %s/%s" -C readwrite' % (Q_EXECUTABLE,tmpfolder, fn) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + + c = sqlite3.connect('%s/%s.qsql' % (tmpfolder,fn)) + c.execute('drop table _qcatalog').fetchall() + c.close() + os.rename('%s/%s.qsql' % (tmpfolder,fn),'%s/%s.sqlite' % (tmpfolder,fn)) + + output_sqlite_file = self.generate_tmpfile_name("many-sqlites",".sqlite") + + table_refs = list(['select * from %s/%s.sqlite' % (tmpfolder,x) for x in filename_list]) + table_refs_str = " UNION ALL ".join(table_refs) + # Limit max attached dbs according to the parameter (must be below the hardcoded sqlite limit, which is 10 when having a standard version compiled) + cmd = '%s "select * from (%s)" -S %s --max-attached-sqlite-databases=%s' % (Q_EXECUTABLE,table_refs_str,output_sqlite_file,MAX_ATTACHED_DBS) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode,0) + self.assertEqual(len(o),0) + self.assertEqual(len(e),4) + + c = sqlite3.connect(output_sqlite_file) + tables_results = c.execute("select tbl_name from sqlite_master where type='table'").fetchall() + table_names = list(sorted([x[0] for x in tables_results])) + self.assertEqual(len(table_names),FILE_COUNT) + + for i,tn in enumerate(table_names): + self.assertEqual(tn,'file_dash_%s' % i) + + table_content = c.execute('select * from %s' % tn).fetchall() + self.assertEqual(len(table_content),BATCH_SIZE) + + cmd = '%s "select * from %s:::%s"' % (Q_EXECUTABLE,output_sqlite_file,tn) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),BATCH_SIZE) + self.assertEqual(o,list([six.b(str(x)) for x in range(1 + i*BATCH_SIZE,1+(i+1)*BATCH_SIZE)])) + + self.cleanup_folder(tmpfolder) + + def test_storing_to_disk_too_many_sqlite_files__over_the_sqlite_limit(self): + # a variation of test_storing_to_disk_too_many_sqlite_files, but with a limit above the sqlite hardcoded limit + MAX_ATTACHED_DBS = 20 # standard sqlite limit is 10, so q should throw an error + + BATCH_SIZE = 10 + FILE_COUNT = MAX_ATTACHED_DBS + 4 + + numbers_as_text = batch([str(x) for x in range(1, 1 + BATCH_SIZE * FILE_COUNT)], n=BATCH_SIZE) + + content_list = map(six.b, ["\n".join(x) for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x, range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d, 'split-files', 'attach-limit') + + for fn in filename_list: + cmd = '%s -c 1 "select count(*) from %s/%s" -C readwrite' % (Q_EXECUTABLE,tmpfolder, fn) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + + c = sqlite3.connect('%s/%s.qsql' % (tmpfolder,fn)) + c.execute('drop table _qcatalog').fetchall() + c.close() + os.rename('%s/%s.qsql' % (tmpfolder,fn),'%s/%s.sqlite' % (tmpfolder,fn)) + + output_sqlite_file = self.generate_tmpfile_name("many-sqlites",".sqlite") + + table_refs = list(['select * from %s/%s.sqlite' % (tmpfolder,x) for x in filename_list]) + table_refs_str = " UNION ALL ".join(table_refs) + # Limit max attached dbs according to the parameter (must be below the hardcoded sqlite limit, which is 10 when having a standard version compiled) + cmd = '%s "select * from (%s)" -S %s --max-attached-sqlite-databases=%s' % (Q_EXECUTABLE,table_refs_str,output_sqlite_file,MAX_ATTACHED_DBS) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode,89) + self.assertEqual(len(o),0) + self.assertEqual(len(e),2) + self.assertTrue(e[0].startswith(six.b('Going to save data into'))) + self.assertTrue(e[1].startswith(six.b('There are too many attached databases. Use a proper --max-attached-sqlite-databases parameter which is below the maximum'))) + + self.cleanup_folder(tmpfolder) + + def test_qtable_name_normalization__starting_with_a_digit(self): + numbers = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 101)] + + header = [six.b('aa'), six.b('bb'), six.b('cc')] + + base_filename_with_digits = '010' + + new_tmp_folder = self.create_folder_with_files({ + base_filename_with_digits : self.arrays_to_csv_file_content(six.b(','),header,numbers) + },prefix='xx',suffix='digits') + + effective_filename = '%s/010' % new_tmp_folder + + output_sqlite_filename = self.generate_tmpfile_name("starting-with-digit",".sqlite") + cmd = '%s -d , -H "select count(aa),count(bb),count(cc) from %s" -S %s' % (Q_EXECUTABLE,effective_filename,output_sqlite_filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),0) + self.assertEqual(len(e),4) + + c = sqlite3.connect(output_sqlite_filename) + results = c.execute('select aa,bb,cc from t_%s' % base_filename_with_digits).fetchall() + self.assertEqual(results,list([(x,x,x) for x in range(1,101)])) + c.close() + + self.cleanup_folder(new_tmp_folder) + + def test_qtable_name_normalization(self): + x = [six.b(a) for a in map(str, range(1, 101))] + large_file_data = six.b("val\n") + six.b("\n").join(x) + tmpfile = self.create_file_with_data(large_file_data) + + tmpfile_folder = os.path.dirname(tmpfile.name) + tmpfile_basename = os.path.basename(tmpfile.name) + + cmd = 'cd %s && %s -c 1 -H -D , -O "select a.val,b.val from %s a cross join ./%s b on (a.val = b.val * 2)"' % (tmpfile_folder,Q_EXECUTABLE,tmpfile_basename,tmpfile_basename) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 51) + + evens = list(filter(lambda x: x%2 == 0,range(1,101))) + expected_result_rows = [six.b('val,val')] + [six.b('%d,%d' % (x,x / 2)) for x in evens] + self.assertEqual(o,expected_result_rows) + + def test_qtable_name_normalization2(self): + cmd = '%s "select * from"' % Q_EXECUTABLE + + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 118) + self.assertEqual(len(e), 1) + self.assertEqual(e[0],six.b('FROM/JOIN is missing a table name after it')) + + def test_qtable_name_normalization3(self): + # with a space after the from + cmd = '%s "select * from "' % Q_EXECUTABLE + + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 118) + self.assertEqual(len(e), 1) + self.assertEqual(e[0],six.b('FROM/JOIN is missing a table name after it')) + + def test_save_multiple_files_to_sqlite_while_caching_them(self): + BATCH_SIZE = 50 + FILE_COUNT = 5 + + tmpfolder,filename_list = self.generate_files_in_folder(BATCH_SIZE,FILE_COUNT) + + output_sqlite_file = self.random_tmp_filename("x","sqlite") + + tables_as_str = " left join ".join(["%s/%s" % (tmpfolder,x) for x in filename_list]) + cmd = '%s -H "select count(*) from %s" -c 1 -S %s -C readwrite' % (Q_EXECUTABLE,tables_as_str,output_sqlite_file) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + c = sqlite3.connect(output_sqlite_file) + + tables = get_sqlite_table_list(c) + self.assertEqual(len(tables), FILE_COUNT) + + for i,filename in enumerate(filename_list): + matching_table_name = 'file_dash_%s' % i + + results = c.execute('select a from %s' % matching_table_name).fetchall() + self.assertEqual(len(results),BATCH_SIZE) + self.assertEqual(sum(map(lambda x:x[0],results)),sum(range(1+i*BATCH_SIZE,1+(i+1)*BATCH_SIZE))) + + # check actual resulting qsql file for the file + cmd = '%s -c 1 -H "select a from %s/%s"' % (Q_EXECUTABLE,tmpfolder,filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), BATCH_SIZE) + self.assertEqual(sum(map(int,o)),sum(range(1+i*BATCH_SIZE,1+(i+1)*BATCH_SIZE))) + self.assertEqual(len(e), 0) + + # check analysis returns proper file-with-unused-qsql for each file, since by default `-C none` which means don't read the cache + # even if it exists + cmd = '%s -c 1 -H "select a from %s/%s" -A' % (Q_EXECUTABLE,tmpfolder,filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 5) + self.assertEqual(o,[ + six.b('Table: %s/file-%s' % (tmpfolder,i)), + six.b(' Sources:'), + six.b(' source_type: file-with-unused-qsql source: %s/file-%s' % (tmpfolder,i)), + six.b(' Fields:'), + six.b(' `a` - int') + ]) + + cmd = '%s -c 1 -H "select a from %s/%s" -A -C read' % (Q_EXECUTABLE,tmpfolder,filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 5) + self.assertEqual(o,[ + six.b('Table: %s/file-%s' % (tmpfolder,i)), + six.b(' Sources:'), + six.b(' source_type: qsql-file-with-original source: %s/file-%s.qsql' % (tmpfolder,i)), + six.b(' Fields:'), + six.b(' `a` - int') + ]) + + # check qsql file is readable directly through q + cmd = '%s -c 1 -H "select a from %s/%s.qsql"' % (Q_EXECUTABLE,tmpfolder,filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), BATCH_SIZE) + self.assertEqual(sum(map(int,o)),sum(range(1+i*BATCH_SIZE,1+(i+1)*BATCH_SIZE))) + self.assertEqual(len(e), 0) + + # check analysis returns proper qsql-with-original for each file when running directly against the qsql file + cmd = '%s -c 1 -H "select a from %s/%s.qsql" -A' % (Q_EXECUTABLE,tmpfolder,filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 5) + self.assertEqual(o,[ + six.b('Table: %s/file-%s.qsql' % (tmpfolder,i)), + six.b(' Sources:'), + six.b(' source_type: qsql-file source: %s/file-%s.qsql' % (tmpfolder,i)), + six.b(' Fields:'), + six.b(' `a` - int') + ]) + c.close() + + import glob + filename_list_with_qsql = list(map(lambda x: x+'.qsql',filename_list)) + + files_in_folder = glob.glob('%s/*' % tmpfolder) + regular_files,qsql_files = partition(lambda x: x.endswith('.qsql'),files_in_folder) + + self.assertEqual(len(files_in_folder),2*FILE_COUNT) + self.assertEqual(sorted(list(map(os.path.basename,regular_files))),sorted(list(map(os.path.basename,filename_list)))) + self.assertEqual(sorted(list(map(os.path.basename,qsql_files))),sorted(list(map(os.path.basename,filename_list_with_qsql)))) + + self.cleanup_folder(tmpfolder) + + def test_globs_ignore_matching_qsql_files(self): + BATCH_SIZE = 10 + FILE_COUNT = 5 + + tmpfolder,filename_list = self.generate_files_in_folder(BATCH_SIZE,FILE_COUNT) + + tables_as_str = " left join ".join(["%s/%s" % (tmpfolder,x) for x in filename_list]) + cmd = '%s -H "select count(*) from %s" -c 1 -C readwrite' % (Q_EXECUTABLE,tables_as_str) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b(str(pow(BATCH_SIZE,FILE_COUNT)))) + + cmd = '%s -H "select a from %s/*" -c 1 -C read' % (Q_EXECUTABLE,tmpfolder) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), BATCH_SIZE*FILE_COUNT) + self.assertEqual(len(e), 0) + self.assertEqual(sum(map(int,o)),sum(range(1,1+BATCH_SIZE*FILE_COUNT))) + + self.cleanup_folder(tmpfolder) + + def test_error_on_reading_from_multi_table_sqlite_without_explicit_table_name(self): + BATCH_SIZE = 50 + FILE_COUNT = 5 + + tmpfolder,filename_list = self.generate_files_in_folder(BATCH_SIZE,FILE_COUNT) + + output_sqlite_file = self.random_tmp_filename("x","sqlite") + + tables_as_str = " left join ".join(["%s/%s" % (tmpfolder,x) for x in filename_list]) + cmd = '%s -H "select count(*) from %s" -c 1 -S %s' % (Q_EXECUTABLE,tables_as_str,output_sqlite_file) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + + cmd = '%s -H "select count(*) from %s"' % (Q_EXECUTABLE,output_sqlite_file) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 87) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0],six.b("Could not autodetect table name in sqlite file %s . Existing tables: file_dash_0,file_dash_1,file_dash_2,file_dash_3,file_dash_4" % output_sqlite_file)) + + self.cleanup_folder(tmpfolder) + + def test_error_on_trying_to_specify_an_explicit_non_existent_qsql_file(self): + cmd = '%s -H "select count(*) from /non-existent-folder/non-existent.qsql:::mytable"' % (Q_EXECUTABLE) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 30) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0],six.b("Could not find file /non-existent-folder/non-existent.qsql")) + + def test_error_on_providing_a_non_qsql_file_when_specifying_an_explicit_table(self): + data = six.b("\x1f\x8b\x08\x00\tZ\x0ea\x00\x03\xed\x93\xdd\n\xc20\x0cF\xf3(}\x01ij\x93\xf6y:\xd9P\x10)\xb3\xbe\xbf\x9d\x1d\xbbQ\xc6\x06F\x10rn\xbe\x9b\xd0\xfc\x1c\x9a-\x88\x83\x88\x91\xd9\xbc2\xb4\xc4#\xb5\x9c1\x8e\x1czb\x8a\xd1\x19t\xdeS\x00\xc3\xf2\xa3\x01<\xee%\x8du\x94s\x1a\xfbk\xd7\xdf\x0e\xa9\x94Kz\xaf\xabe\xc3\xb0\xf2\xce\xbc\xc7\x92\x7fB\xb6\x1fv\xfd2\xf5\x1e\x81h\xa3\xff\x10'\xff\x8c\x04\x06\xc5'\x03\xf5oO\xe2=v\xf9o\xff\x9f\xd1\xa9\xff_\x90m'\xdec\x9f\x7f\x9c\xfc\xd7T\xff\x8a\xa2(\x92<\x01WY\x0c\x06\x00\x0c\x00\x00") + tmpfilename = self.random_tmp_filename('xx','yy') + f = open(tmpfilename,'wb') + f.write(data) + f.close() + + cmd = '%s -H "select count(*) from %s:::mytable1"' % (Q_EXECUTABLE,tmpfilename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 95) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0],six.b("Cannot detect the type of table %s:::mytable1" % tmpfilename)) + + def test_error_on_providing_a_non_qsql_file_when_not_specifying_an_explicit_table(self): + data = six.b("\x1f\x8b\x08\x00\tZ\x0ea\x00\x03\xed\x93\xdd\n\xc20\x0cF\xf3(}\x01ij\x93\xf6y:\xd9P\x10)\xb3\xbe\xbf\x9d\x1d\xbbQ\xc6\x06F\x10rn\xbe\x9b\xd0\xfc\x1c\x9a-\x88\x83\x88\x91\xd9\xbc2\xb4\xc4#\xb5\x9c1\x8e\x1czb\x8a\xd1\x19t\xdeS\x00\xc3\xf2\xa3\x01<\xee%\x8du\x94s\x1a\xfbk\xd7\xdf\x0e\xa9\x94Kz\xaf\xabe\xc3\xb0\xf2\xce\xbc\xc7\x92\x7fB\xb6\x1fv\xfd2\xf5\x1e\x81h\xa3\xff\x10'\xff\x8c\x04\x06\xc5'\x03\xf5oO\xe2=v\xf9o\xff\x9f\xd1\xa9\xff_\x90m'\xdec\x9f\x7f\x9c\xfc\xd7T\xff\x8a\xa2(\x92<\x01WY\x0c\x06\x00\x0c\x00\x00") + tmpfilename = self.random_tmp_filename('xx','yy') + f = open(tmpfilename,'wb') + f.write(data) + f.close() + + cmd = '%s -H "select count(*) from %s"' % (Q_EXECUTABLE,tmpfilename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 59) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertTrue(e[0].startswith(six.b("Could not parse the input. Please make sure to set the proper -w input-wrapping parameter for your input, and that you use the proper input encoding (-e). Error:"))) + +class OldSaveDbToDiskTests(AbstractQTestCase): + + def test_join_with_stdin_and_save(self): + x = [six.b(a) for a in map(str,range(1,101))] + large_file_data = six.b("val\n") + six.b("\n").join(x) + tmpfile = self.create_file_with_data(large_file_data) + tmpfile_expected_table_name = os.path.basename(tmpfile.name) + + disk_db_filename = self.random_tmp_filename('save-to-db','sqlite') + + cmd = '(echo id ; seq 1 2 10) | ' + Q_EXECUTABLE + ' -c 1 -H -O "select stdin.*,f.* from - stdin left join %s f on (stdin.id * 10 = f.val)" -S %s' % \ + (tmpfile.name,disk_db_filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + self.assertEqual(e[0],six.b('Going to save data into a disk database: %s' % disk_db_filename)) + self.assertTrue(e[1].startswith(six.b('Data has been saved into %s . Saving has taken ' % disk_db_filename))) + self.assertEqual(e[2],six.b('Query to run on the database: select stdin.*,f.* from data_stream_stdin stdin left join %s f on (stdin.id * 10 = f.val);' % \ + tmpfile_expected_table_name)) + self.assertEqual(e[3],six.b('You can run the query directly from the command line using the following command: echo "select stdin.*,f.* from data_stream_stdin stdin left join %s f on (stdin.id * 10 = f.val)" | sqlite3 %s' % + (tmpfile_expected_table_name,disk_db_filename))) + + P = re.compile(six.b("^Query to run on the database: (?P.*)$")) + m = P.search(e[2]) + query_to_run_on_db = m.groupdict()['query_to_run_on_db'] + + self.assertTrue(os.path.exists(disk_db_filename)) + + # validate disk db content natively + c = sqlite3.connect(disk_db_filename) + c.row_factory = sqlite_dict_factory + t0_results = c.execute('select * from data_stream_stdin').fetchall() + self.assertEqual(len(t0_results),5) + self.assertEqual(sorted(list(t0_results[0].keys())), ['id']) + self.assertEqual(list(map(lambda x:x['id'],t0_results)),[1,3,5,7,9]) + t1_results = c.execute('select * from %s' % tmpfile_expected_table_name).fetchall() + self.assertEqual(len(t1_results),100) + self.assertEqual(sorted(list(t1_results[0].keys())), ['val']) + self.assertEqual("\n".join(list(map(lambda x:str(x['val']),t1_results))),"\n".join(map(str,range(1,101)))) + + query_results = c.execute(query_to_run_on_db.decode('utf-8')).fetchall() + + self.assertEqual(query_results[0],{ 'id': 1 , 'val': 10}) + self.assertEqual(query_results[1],{ 'id': 3 , 'val': 30}) + self.assertEqual(query_results[2],{ 'id': 5 , 'val': 50}) + self.assertEqual(query_results[3],{ 'id': 7 , 'val': 70}) + self.assertEqual(query_results[4],{ 'id': 9 , 'val': 90}) + + self.cleanup(tmpfile) + + def test_join_with_qsql_file(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + + header = [six.b('aa'), six.b('bb'), six.b('cc')] + + new_tmp_folder = self.create_folder_with_files({ + 'some_csv_file': self.arrays_to_csv_file_content(six.b(','),header,numbers1), + 'some_qsql_database.qsql' : self.arrays_to_qsql_file_content(header,numbers2) + },prefix='xx',suffix='yy') + + effective_filename1 = '%s/some_csv_file' % new_tmp_folder + effective_filename2 = '%s/some_qsql_database.qsql' % new_tmp_folder + + cmd = Q_EXECUTABLE + ' -d , -H "select sum(large_file.aa),sum(small_file.aa) from %s large_file left join %s small_file on (small_file.aa == large_file.bb)"' % \ + (effective_filename1,effective_filename2) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('50005000,55')) + + # TODO RLRL Check if needed anymore + + # def test_creation_of_qsql_database(self): + # numbers = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + # header = [six.b('aa'), six.b('bb'), six.b('cc')] + # + # qsql_filename = self.create_qsql_file_with_content_and_return_filename(header,numbers) + # + # conn = sqlite3.connect(qsql_filename) + # qcatalog = conn.execute('select temp_table_name,source_type,source from _qcatalog').fetchall() + # print(qcatalog) + # + # cmd = '%s "select count(*) from %s" -A' % (Q_EXECUTABLE,qsql_filename) + # retcode, o, e = run_command(cmd) + # print(o) + + def test_join_with_qsql_file_and_save(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + + header = [six.b('aa'), six.b('bb'), six.b('cc')] + + saved_qsql_with_multiple_tables = self.generate_tmpfile_name(suffix='.qsql') + + new_tmp_folder = self.create_folder_with_files({ + 'some_csv_file': self.arrays_to_csv_file_content(six.b(','),header,numbers1), + 'some_qsql_database' : self.arrays_to_csv_file_content(six.b(','),header,numbers2) + },prefix='xx',suffix='yy') + cmd = '%s -d , -H "select count(*) from %s/some_qsql_database" -C readwrite' % (Q_EXECUTABLE,new_tmp_folder) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode,0) + os.remove('%s/some_qsql_database' % new_tmp_folder) + + effective_filename1 = '%s/some_csv_file' % new_tmp_folder + effective_filename2 = '%s/some_qsql_database.qsql' % new_tmp_folder + + cmd = Q_EXECUTABLE + ' -d , -H "select sum(large_file.aa),sum(small_file.aa) from %s large_file left join %s small_file on (small_file.aa == large_file.bb)" -S %s' % \ + (effective_filename1,effective_filename2,saved_qsql_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + + conn = sqlite3.connect(saved_qsql_with_multiple_tables) + c1 = conn.execute('select count(*) from some_csv_file').fetchall() + c2 = conn.execute('select count(*) from some_qsql_database').fetchall() + + self.assertEqual(c1[0][0],10000) + self.assertEqual(c2[0][0],10) + + + def test_saving_to_db_with_same_basename_files(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + + header = [six.b('aa'), six.b('bb'), six.b('cc')] + + qsql_with_multiple_tables = self.generate_tmpfile_name(suffix='.qsql') + + new_tmp_folder = self.create_folder_with_files({ + 'filename1': self.arrays_to_csv_file_content(six.b(','),header,numbers1), + 'otherfolder/filename1' : self.arrays_to_csv_file_content(six.b(','),header,numbers2) + },prefix='xx',suffix='yy') + + effective_filename1 = '%s/filename1' % new_tmp_folder + effective_filename2 = '%s/otherfolder/filename1' % new_tmp_folder + + expected_stored_table_name1 = 'filename1' + expected_stored_table_name2 = 'filename1_2' + + cmd = Q_EXECUTABLE + ' -d , -H "select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s small_file left join %s large_file on (large_file.aa == small_file.bb)" -S %s' % \ + (effective_filename1,effective_filename2,qsql_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + self.assertEqual(e[0], six.b('Going to save data into a disk database: %s' % qsql_with_multiple_tables)) + self.assertTrue(e[1].startswith(six.b('Data has been saved into %s . Saving has taken' % qsql_with_multiple_tables))) + self.assertEqual(e[2],six.b('Query to run on the database: select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s small_file left join %s large_file on (large_file.aa == small_file.bb);' % \ + (expected_stored_table_name1,expected_stored_table_name2))) + self.assertEqual(e[3],six.b('You can run the query directly from the command line using the following command: echo "select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s small_file left join %s large_file on (large_file.aa == small_file.bb)" | sqlite3 %s' % \ + (expected_stored_table_name1,expected_stored_table_name2,qsql_with_multiple_tables))) + + #self.assertTrue(False) # pxpx - need to actually test reading from the saved db file + conn = sqlite3.connect(qsql_with_multiple_tables) + c1 = conn.execute('select count(*) from filename1').fetchall() + c2 = conn.execute('select count(*) from filename1_2').fetchall() + + self.assertEqual(c1[0][0],10000) + self.assertEqual(c2[0][0],10) + + + def test_error_when_not_specifying_table_name_in_multi_table_qsql(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + + header = [six.b('aa'), six.b('bb'), six.b('cc')] + + qsql_with_multiple_tables = self.generate_tmpfile_name(suffix='.qsql') + + new_tmp_folder = self.create_folder_with_files({ + 'filename1': self.arrays_to_csv_file_content(six.b(','),header,numbers1), + 'otherfolder/filename1' : self.arrays_to_csv_file_content(six.b(','),header,numbers2) + },prefix='xx',suffix='yy') + + effective_filename1 = '%s/filename1' % new_tmp_folder + effective_filename2 = '%s/otherfolder/filename1' % new_tmp_folder + + expected_stored_table_name1 = 'filename1' + expected_stored_table_name2 = 'filename1_2' + + cmd = Q_EXECUTABLE + ' -d , -H "select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s small_file left join %s large_file on (large_file.aa == small_file.bb)" -S %s' % \ + (effective_filename1,effective_filename2,qsql_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + # Actual tests + + cmd = '%s "select count(*) from %s"' % (Q_EXECUTABLE,qsql_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 87) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + self.assertEqual(e[0],six.b('Could not autodetect table name in sqlite file %s . Existing tables: %s,%s' % (qsql_with_multiple_tables,expected_stored_table_name1,expected_stored_table_name2))) + + def test_error_when_not_specifying_table_name_in_multi_table_sqlite(self): + sqlite_with_multiple_tables = self.generate_tmpfile_name(suffix='.sqlite') + + c = sqlite3.connect(sqlite_with_multiple_tables) + c.execute('create table my_table_1 (x int, y int)').fetchall() + c.execute('create table my_table_2 (x int, y int)').fetchall() + c.close() + + cmd = '%s "select count(*) from %s"' % (Q_EXECUTABLE,sqlite_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 87) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + print(e[0]) + self.assertEqual(e[0],six.b('Could not autodetect table name in sqlite file %s . Existing tables: my_table_1,my_table_2' % sqlite_with_multiple_tables)) + + def test_querying_from_multi_table_sqlite_using_explicit_table_name(self): + sqlite_with_multiple_tables = self.generate_tmpfile_name(suffix='.sqlite') + + c = sqlite3.connect(sqlite_with_multiple_tables) + c.execute('create table my_table_1 (x int, y int)').fetchall() + c.execute('insert into my_table_1 (x,y) values (100,200),(300,400)').fetchall() + c.execute('commit').fetchall() + c.execute('create table my_table_2 (x int, y int)').fetchall() + c.close() + + cmd = '%s -d , "select * from %s:::my_table_1"' % (Q_EXECUTABLE,sqlite_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('100,200')) + self.assertEqual(o[1],six.b('300,400')) + + # Check again, this time with a different output delimiter and with explicit column names + cmd = '%s -t "select x,y from %s:::my_table_1"' % (Q_EXECUTABLE,sqlite_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('100\t200')) + self.assertEqual(o[1],six.b('300\t400')) + + + def test_error_when_specifying_nonexistent_table_name_in_multi_table_qsql(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + + header = [six.b('aa'), six.b('bb'), six.b('cc')] + + qsql_with_multiple_tables = self.generate_tmpfile_name(suffix='.qsql') + + new_tmp_folder = self.create_folder_with_files({ + 'filename1': self.arrays_to_csv_file_content(six.b(','),header,numbers1), + 'otherfolder/filename1' : self.arrays_to_csv_file_content(six.b(','),header,numbers2) + },prefix='xx',suffix='yy') + + effective_filename1 = '%s/filename1' % new_tmp_folder + effective_filename2 = '%s/otherfolder/filename1' % new_tmp_folder + + expected_stored_table_name1 = 'filename1' + expected_stored_table_name2 = 'filename1_2' + + cmd = Q_EXECUTABLE + ' -d , -H "select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s small_file left join %s large_file on (large_file.aa == small_file.bb)" -S %s' % \ + (effective_filename1,effective_filename2,qsql_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + # Actual tests + + cmd = '%s "select count(*) from %s:::non_existent_table"' % (Q_EXECUTABLE,qsql_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 85) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + self.assertEqual(e[0],six.b('Table non_existent_table could not be found in sqlite file %s . Existing table names: %s,%s' % \ + (qsql_with_multiple_tables,expected_stored_table_name1,expected_stored_table_name2))) + + def test_querying_multi_table_qsql_file(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + + header = [six.b('aa'), six.b('bb'), six.b('cc')] + + qsql_with_multiple_tables = self.generate_tmpfile_name(suffix='.qsql') + + new_tmp_folder = self.create_folder_with_files({ + 'filename1': self.arrays_to_csv_file_content(six.b(','),header,numbers1), + 'otherfolder/filename1' : self.arrays_to_csv_file_content(six.b(','),header,numbers2) + },prefix='xx',suffix='yy') + + effective_filename1 = '%s/filename1' % new_tmp_folder + effective_filename2 = '%s/otherfolder/filename1' % new_tmp_folder + + expected_stored_table_name1 = 'filename1' + expected_stored_table_name2 = 'filename1_2' + + cmd = Q_EXECUTABLE + ' -d , -H "select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s small_file left join %s large_file on (large_file.aa == small_file.bb)" -S %s' % \ + (effective_filename1,effective_filename2,qsql_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + + # Actual tests + + cmd = '%s "select count(*) from %s:::%s"' % (Q_EXECUTABLE,qsql_with_multiple_tables,expected_stored_table_name1) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('10000')) + + cmd = '%s "select count(*) from %s:::%s"' % (Q_EXECUTABLE,qsql_with_multiple_tables,expected_stored_table_name2) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('10')) + + def test_preventing_db_overwrite(self): + db_filename = self.random_tmp_filename('store-to-disk', 'db') + self.assertFalse(os.path.exists(db_filename)) + + retcode, o, e = run_command('seq 1 1000 | ' + Q_EXECUTABLE + ' "select count(*) from -" -c 1 -S %s' % db_filename) + + self.assertTrue(retcode == 0) + self.assertTrue(os.path.exists(db_filename)) + + retcode2, o2, e2 = run_command('seq 1 1000 | ' + Q_EXECUTABLE + ' "select count(*) from -" -c 1 -S %s' % db_filename) + self.assertTrue(retcode2 != 0) + self.assertTrue(e2[0].startswith(six.b('Going to save data into a disk database'))) + self.assertTrue(e2[1] == six.b('Disk database file {} already exists.'.format(db_filename))) + + os.remove(db_filename) + + +class BasicTests(AbstractQTestCase): + + def test_basic_aggregation(self): + retcode, o, e = run_command( + 'seq 1 10 | ' + Q_EXECUTABLE + ' "select sum(c1),avg(c1) from -"') + self.assertTrue(retcode == 0) + self.assertTrue(len(o) == 1) + self.assertTrue(len(e) == 0) + + s = sum(range(1, 11)) + self.assertTrue(o[0] == six.b('%s %s' % (s, s / 10.0))) + + def test_select_one_column(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + + cmd = Q_EXECUTABLE + ' -d , "select c1 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + + self.assertEqual(six.b(" ").join(o), six.b('a b c')) + + self.cleanup(tmpfile) + + def test_column_separation(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], sample_data_rows[0]) + self.assertEqual(o[1], sample_data_rows[1]) + self.assertEqual(o[2], sample_data_rows[2]) + + self.cleanup(tmpfile) + + def test_header_exception_on_numeric_header_data(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , "select * from %s" -A -H' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 3) + self.assertTrue( + six.b('Bad header row: Header must contain only strings') in e[0]) + self.assertTrue(six.b("Column name must be a string") in e[1]) + self.assertTrue(six.b("Column name must be a string") in e[2]) + + self.cleanup(tmpfile) + + def test_different_header_in_second_file(self): + folder_name = self.create_folder_with_files({ + 'file1': self.arrays_to_csv_file_content(six.b(','),[six.b('a'),six.b('b')],[[six.b(str(x)),six.b(str(x))] for x in range(1,6)]), + 'file2': self.arrays_to_csv_file_content(six.b(','),[six.b('c'),six.b('d')],[[six.b(str(x)),six.b(str(x))] for x in range(1,6)]) + },prefix="xx",suffix="aa") + + cmd = Q_EXECUTABLE + ' -d , "select * from %s/*" -H' % (folder_name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 35) + self.assertEqual(len(e),1) + self.assertEqual(e[0],six.b("Bad header row: Extra header 'c,d' in file '%s/file2' mismatches original header 'a,b' from file '%s/file1'. Table name is '%s/*'" % (folder_name,folder_name,folder_name))) + + def test_data_with_header(self): + tmpfile = self.create_file_with_data(sample_data_with_header) + cmd = Q_EXECUTABLE + ' -d , "select name from %s" -H' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(six.b(" ").join(o), six.b("a b c")) + + self.cleanup(tmpfile) + + def test_output_header_when_input_header_exists(self): + tmpfile = self.create_file_with_data(sample_data_with_header) + cmd = Q_EXECUTABLE + ' -d , "select name from %s" -H -O' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 4) + self.assertEqual(o[0],six.b('name')) + self.assertEqual(o[1],six.b('a')) + self.assertEqual(o[2],six.b('b')) + self.assertEqual(o[3],six.b('c')) + + self.cleanup(tmpfile) + + def test_generated_column_name_warning_when_header_line_exists(self): + tmpfile = self.create_file_with_data(sample_data_with_header) + cmd = Q_EXECUTABLE + ' -d , "select c3 from %s" -H' % tmpfile.name + + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 2) + self.assertTrue(six.b('no such column: c3') in e[0]) + self.assertTrue( + e[1].startswith(six.b('Warning - There seems to be a "no such column" error, and -H (header line) exists. Please make sure that you are using the column names from the header line and not the default (cXX) column names'))) + + self.cleanup(tmpfile) + + def test_empty_data(self): + tmpfile = self.create_file_with_data(six.b('')) + cmd = Q_EXECUTABLE + ' -d , "select * from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + self.assertTrue(six.b('Warning - data is empty') in e[0]) + + self.cleanup(tmpfile) + + def test_empty_data_with_header_param(self): + tmpfile = self.create_file_with_data(six.b('')) + cmd = Q_EXECUTABLE + ' -d , "select c1 from %s" -H' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + m = six.b("Header line is expected but missing in file %s" % tmpfile.name) + self.assertTrue(m in e[0]) + + self.cleanup(tmpfile) + + def test_one_row_of_data_without_header_param(self): + tmpfile = self.create_file_with_data(header_row) + cmd = Q_EXECUTABLE + ' -d , "select c2 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('value1')) + + self.cleanup(tmpfile) + + def test_one_row_of_data_with_header_param(self): + tmpfile = self.create_file_with_data(header_row) + cmd = Q_EXECUTABLE + ' -d , "select name from %s" -H' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + self.assertTrue(six.b('Warning - data is empty') in e[0]) + + self.cleanup(tmpfile) + + def test_dont_leading_keep_whitespace_in_values(self): + tmpfile = self.create_file_with_data(sample_data_with_spaces_no_header) + cmd = Q_EXECUTABLE + ' -d , "select c1 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 3) + + self.assertEqual(o[0], six.b('a')) + self.assertEqual(o[1], six.b('b')) + self.assertEqual(o[2], six.b('c')) + + self.cleanup(tmpfile) + + def test_keep_leading_whitespace_in_values(self): + tmpfile = self.create_file_with_data(sample_data_with_spaces_no_header) + cmd = Q_EXECUTABLE + ' -d , "select c1 from %s" -k' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 3) + + self.assertEqual(o[0], six.b('a')) + self.assertEqual(o[1], six.b(' b')) + self.assertEqual(o[2], six.b('c')) + + self.cleanup(tmpfile) + + def test_no_impact_of_keeping_leading_whitespace_on_integers(self): + tmpfile = self.create_file_with_data(sample_data_with_spaces_no_header) + cmd = Q_EXECUTABLE + ' -d , "select c2 from %s" -k -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + f = open("/var/tmp/XXX","wb") + f.write(six.b("\n").join(o)) + f.write(six.b("STDERR:")) + f.write(six.b("\n").join(e)) + f.close() + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 7) + + + self.assertEqual(o[0], six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1], six.b(' Sources:')) + self.assertEqual(o[2], six.b(' source_type: file source: %s') % six.b(tmpfile.name)) + self.assertEqual(o[3], six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `c1` - text')) + self.assertEqual(o[5], six.b(' `c2` - int')) + self.assertEqual(o[6], six.b(' `c3` - int')) + + + self.cleanup(tmpfile) + + def test_spaces_in_header_row(self): + tmpfile = self.create_file_with_data( + header_row_with_spaces + six.b("\n") + sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , "select name,\\`value 1\\` from %s" -H' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 3) + + self.assertEqual(o[0], six.b('a,1')) + self.assertEqual(o[1], six.b('b,2')) + self.assertEqual(o[2], six.b('c,')) + + self.cleanup(tmpfile) + + def test_no_query_in_command_line(self): + cmd = Q_EXECUTABLE + ' -d , ""' + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 1) + self.assertEqual(len(e), 1) + self.assertEqual(len(o), 0) + + self.assertEqual(e[0],six.b('Query cannot be empty (query number 1)')) + + def test_empty_query_in_command_line(self): + cmd = Q_EXECUTABLE + ' -d , " "' + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 1) + self.assertEqual(len(e), 1) + self.assertEqual(len(o), 0) + + self.assertEqual(e[0],six.b('Query cannot be empty (query number 1)')) + + def test_failure_in_query_stops_processing_queries(self): + cmd = Q_EXECUTABLE + ' -d , "select 500" "select 300" "wrong-query" "select 8000"' + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 1) + self.assertEqual(len(e), 1) + self.assertEqual(len(o), 2) + self.assertEqual(o[0],six.b('500')) + self.assertEqual(o[1],six.b('300')) + + def test_multiple_queries_in_command_line(self): + cmd = Q_EXECUTABLE + ' -d , "select 500" "select 300+100" "select 300" "select 200"' + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 4) + + self.assertEqual(o[0],six.b('500')) + self.assertEqual(o[1],six.b('400')) + self.assertEqual(o[2],six.b('300')) + self.assertEqual(o[3],six.b('200')) + + def test_literal_calculation_query(self): + cmd = Q_EXECUTABLE + ' -d , "select 1+40/6"' + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 1) + + self.assertEqual(o[0],six.b('7')) + + def test_literal_calculation_query_float_result(self): + cmd = Q_EXECUTABLE + ' -d , "select 1+40/6.0"' + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 1) + + self.assertEqual(o[0],six.b('7.666666666666667')) + + def test_use_query_file(self): + tmp_data_file = self.create_file_with_data(sample_data_with_header) + tmp_query_file = self.create_file_with_data(six.b("select name from %s" % tmp_data_file.name)) + + cmd = Q_EXECUTABLE + ' -d , -q %s -H' % tmp_query_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 3) + + self.assertEqual(o[0], six.b('a')) + self.assertEqual(o[1], six.b('b')) + self.assertEqual(o[2], six.b('c')) + + self.cleanup(tmp_data_file) + self.cleanup(tmp_query_file) + + def test_use_query_file_with_incorrect_query_encoding(self): + tmp_data_file = self.create_file_with_data(sample_data_with_header) + tmp_query_file = self.create_file_with_data(six.b("select name,'Hr\xc3\xa1\xc4\x8d' from %s" % tmp_data_file.name),encoding=None) + + cmd = Q_EXECUTABLE + ' -d , -q %s -H -Q ascii' % tmp_query_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,3) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + + self.assertTrue(e[0].startswith(six.b('Could not decode query number 1 using the provided query encoding (ascii)'))) + + self.cleanup(tmp_data_file) + self.cleanup(tmp_query_file) + + def test_output_header_with_non_ascii_names(self): + OUTPUT_ENCODING = 'utf-8' + + tmp_data_file = self.create_file_with_data(sample_data_with_header) + tmp_query_file = self.create_file_with_data(six.b("select name,'Hr\xc3\xa1\xc4\x8d' Hr\xc3\xa1\xc4\x8d from %s" % tmp_data_file.name),encoding=None) + + cmd = Q_EXECUTABLE + ' -d , -q %s -H -Q utf-8 -O -E %s' % (tmp_query_file.name,OUTPUT_ENCODING) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),4) + self.assertEqual(len(e),0) + + self.assertEqual(o[0].decode(OUTPUT_ENCODING), u'name,Hr\xe1\u010d') + self.assertEqual(o[1].decode(OUTPUT_ENCODING), u'a,Hr\xe1\u010d') + self.assertEqual(o[2].decode(OUTPUT_ENCODING), u'b,Hr\xe1\u010d') + self.assertEqual(o[3].decode(OUTPUT_ENCODING), u'c,Hr\xe1\u010d') + + self.cleanup(tmp_data_file) + self.cleanup(tmp_query_file) + + def test_use_query_file_with_query_encoding(self): + OUTPUT_ENCODING = 'utf-8' + + tmp_data_file = self.create_file_with_data(sample_data_with_header) + tmp_query_file = self.create_file_with_data(six.b("select name,'Hr\xc3\xa1\xc4\x8d' from %s" % tmp_data_file.name),encoding=None) + + cmd = Q_EXECUTABLE + ' -d , -q %s -H -Q utf-8 -E %s' % (tmp_query_file.name,OUTPUT_ENCODING) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 3) + + self.assertEqual(o[0].decode(OUTPUT_ENCODING), u'a,Hr\xe1\u010d') + self.assertEqual(o[1].decode(OUTPUT_ENCODING), u'b,Hr\xe1\u010d') + self.assertEqual(o[2].decode(OUTPUT_ENCODING), u'c,Hr\xe1\u010d') + + self.cleanup(tmp_data_file) + self.cleanup(tmp_query_file) + + def test_use_query_file_and_command_line(self): + tmp_data_file = self.create_file_with_data(sample_data_with_header) + tmp_query_file = self.create_file_with_data(six.b("select name from %s" % tmp_data_file.name)) + + cmd = Q_EXECUTABLE + ' -d , -q %s -H "select * from ppp"' % tmp_query_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 1) + self.assertEqual(len(e), 1) + self.assertEqual(len(o), 0) + + self.assertTrue(e[0].startswith(six.b("Can't provide both a query file and a query on the command line"))) + + self.cleanup(tmp_data_file) + self.cleanup(tmp_query_file) + + def test_select_output_encoding(self): + tmp_data_file = self.create_file_with_data(sample_data_with_header) + tmp_query_file = self.create_file_with_data(six.b("select 'Hr\xc3\xa1\xc4\x8d' from %s" % tmp_data_file.name),encoding=None) + + for target_encoding in ['utf-8','ibm852']: + cmd = Q_EXECUTABLE + ' -d , -q %s -H -Q utf-8 -E %s' % (tmp_query_file.name,target_encoding) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 3) + + self.assertEqual(o[0].decode(target_encoding), u'Hr\xe1\u010d') + self.assertEqual(o[1].decode(target_encoding), u'Hr\xe1\u010d') + self.assertEqual(o[2].decode(target_encoding), u'Hr\xe1\u010d') + + self.cleanup(tmp_data_file) + self.cleanup(tmp_query_file) + + def test_select_failed_output_encoding(self): + tmp_data_file = self.create_file_with_data(sample_data_with_header) + tmp_query_file = self.create_file_with_data(six.b("select 'Hr\xc3\xa1\xc4\x8d' from %s" % tmp_data_file.name),encoding=None) + + cmd = Q_EXECUTABLE + ' -d , -q %s -H -Q utf-8 -E ascii' % tmp_query_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 3) + self.assertEqual(len(e), 1) + self.assertEqual(len(o), 0) + + self.assertTrue(e[0].startswith(six.b('Cannot encode data'))) + + self.cleanup(tmp_data_file) + self.cleanup(tmp_query_file) + + + def test_use_query_file_with_empty_query(self): + tmp_query_file = self.create_file_with_data(six.b(" ")) + + cmd = Q_EXECUTABLE + ' -d , -q %s -H' % tmp_query_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 1) + self.assertEqual(len(e), 1) + self.assertEqual(len(o), 0) + + self.assertTrue(e[0].startswith(six.b("Query cannot be empty"))) + + self.cleanup(tmp_query_file) + + def test_use_non_existent_query_file(self): + cmd = Q_EXECUTABLE + ' -d , -q non-existent-query-file -H' + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 1) + self.assertEqual(len(e), 1) + self.assertEqual(len(o), 0) + + self.assertTrue(e[0].startswith(six.b("Could not read query from file"))) + + def test_nonexistent_file(self): + cmd = Q_EXECUTABLE + ' "select * from non-existent-file"' + + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode,0) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + + self.assertEqual(e[0],six.b("No files matching '%s/non-existent-file' have been found" % os.getcwd())) + + def test_default_column_max_length_parameter__short_enough(self): + huge_text = six.b("x" * 131000) + + file_data = six.b("a,b,c\n1,{},3\n".format(huge_text)) + + tmpfile = self.create_file_with_data(file_data) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b('1')) + + self.cleanup(tmpfile) + + def test_default_column_max_length_parameter__too_long(self): + huge_text = six.b("x") * 132000 + + file_data = six.b("a,b,c\n1,{},3\n".format(huge_text)) + + tmpfile = self.create_file_with_data(file_data) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 31) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + self.assertTrue(e[0].startswith(six.b("Column length is larger than the maximum"))) + self.assertTrue(six.b("Offending file is '{}'".format(tmpfile.name)) in e[0]) + self.assertTrue(six.b('Line is 2') in e[0]) + + self.cleanup(tmpfile) + + def test_column_max_length_parameter(self): + file_data = six.b("a,b,c\nvery-long-text,2,3\n") + tmpfile = self.create_file_with_data(file_data) + + cmd = Q_EXECUTABLE + ' -H -d , -M 3 "select a from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 31) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + self.assertTrue(e[0].startswith(six.b("Column length is larger than the maximum"))) + self.assertTrue((six.b("Offending file is '%s'" % tmpfile.name)) in e[0]) + self.assertTrue(six.b('Line is 2') in e[0]) + + cmd2 = Q_EXECUTABLE + ' -H -d , -M 300 -H "select a from %s"' % tmpfile.name + retcode2, o2, e2 = run_command(cmd2) + + self.assertEqual(retcode2, 0) + self.assertEqual(len(o2), 1) + self.assertEqual(len(e2), 0) + + self.assertEqual(o2[0],six.b('very-long-text')) + + self.cleanup(tmpfile) + + def test_invalid_column_max_length_parameter(self): + file_data = six.b("a,b,c\nvery-long-text,2,3\n") + tmpfile = self.create_file_with_data(file_data) + + cmd = Q_EXECUTABLE + ' -H -d , -M xx "select a from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 31) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + self.assertEqual(e[0],six.b('Max column length limit must be an integer larger than 2 (xx)')) + + self.cleanup(tmpfile) + + def test_duplicate_column_name_detection(self): + file_data = six.b("a,b,a\n10,20,30\n30,40,50") + tmpfile = self.create_file_with_data(file_data) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 35) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 2) + + self.assertTrue(e[0].startswith(six.b('Bad header row:'))) + self.assertEqual(e[1],six.b("'a': Column name is duplicated")) + + self.cleanup(tmpfile) + + def test_join_with_stdin(self): + x = [six.b(a) for a in map(str,range(1,101))] + large_file_data = six.b("val\n") + six.b("\n").join(x) + tmpfile = self.create_file_with_data(large_file_data) + + cmd = '(echo id ; seq 1 2 10) | %s -c 1 -H -O "select stdin.*,f.* from - stdin left join %s f on (stdin.id * 10 = f.val)"' % (Q_EXECUTABLE,tmpfile.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 6) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b('id val')) + self.assertEqual(o[1],six.b('1 10')) + self.assertEqual(o[2],six.b('3 30')) + self.assertEqual(o[3],six.b('5 50')) + self.assertEqual(o[4],six.b('7 70')) + self.assertEqual(o[5],six.b('9 90')) + + self.cleanup(tmpfile) + + def test_concatenated_files(self): + file_data1 = six.b("a,b,c\n10,11,12\n20,21,22") + tmpfile1 = self.create_file_with_data(file_data1) + tmpfile1_folder = os.path.dirname(tmpfile1.name) + tmpfile1_filename = os.path.basename(tmpfile1.name) + expected_cache_filename1 = os.path.join(tmpfile1_folder,tmpfile1_filename + '.qsql') + + file_data2 = six.b("a,b,c\n30,31,32\n40,41,42") + tmpfile2 = self.create_file_with_data(file_data2) + tmpfile2_folder = os.path.dirname(tmpfile2.name) + tmpfile2_filename = os.path.basename(tmpfile2.name) + expected_cache_filename2 = os.path.join(tmpfile2_folder,tmpfile2_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -O -H -d , "select * from %s UNION ALL select * from %s" -C none' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 5) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('a,b,c')) + self.assertEqual(o[1],six.b('10,11,12')) + self.assertEqual(o[2],six.b('20,21,22')) + self.assertEqual(o[3],six.b('30,31,32')) + self.assertEqual(o[4],six.b('40,41,42')) + + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + def test_out_of_range_expected_column_count(self): + cmd = '%s "select count(*) from some_table" -c -1' % Q_EXECUTABLE + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 90) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0], six.b('Column count must be between 1 and 131072')) + + def test_out_of_range_expected_column_count__with_explicit_limit(self): + cmd = '%s "select count(*) from some_table" -c -1 -M 100' % Q_EXECUTABLE + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 90) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0], six.b('Column count must be between 1 and 100')) + + def test_other_out_of_range_expected_column_count__with_explicit_limit(self): + cmd = '%s "select count(*) from some_table" -c 101 -M 100' % Q_EXECUTABLE + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 90) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0], six.b('Column count must be between 1 and 100')) + + def test_explicit_limit_of_columns__data_is_ok(self): + file_data1 = six.b("191\n192\n") + tmpfile1 = self.create_file_with_data(file_data1) + + cmd = '%s "select count(*) from %s" -c 1 -M 3' % (Q_EXECUTABLE,tmpfile1.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0], six.b('2')) + + self.cleanup(tmpfile1) + +class ManyOpenFilesTests(AbstractQTestCase): + + + def test_multi_file_header_skipping(self): + BATCH_SIZE = 50 + FILE_COUNT = 5 + + numbers = list(range(1,1+BATCH_SIZE*FILE_COUNT)) + numbers_as_text = batch([str(x) for x in numbers],n=BATCH_SIZE) + + content_list = list(map(six.b,['a\n' + "\n".join(x)+'\n' for x in numbers_as_text])) + + filename_list = list(map(lambda x: 'file-%s' % x,range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d,'split-files','multi-header') + + cmd = '%s -d , -H -c 1 "select count(a),sum(a) from %s/*" -C none' % (Q_EXECUTABLE,tmpfolder) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b("%s,%s" % (BATCH_SIZE*FILE_COUNT,sum(numbers)))) + + self.cleanup_folder(tmpfolder) + + def test_that_globs_dont_max_out_sqlite_attached_database_limits(self): + BATCH_SIZE = 50 + FILE_COUNT = 40 + + numbers_as_text = batch([str(x) for x in range(1,1+BATCH_SIZE*FILE_COUNT)],n=BATCH_SIZE) + + content_list = map(six.b,["\n".join(x)+'\n' for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x,range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d,'split-files','attach-limit') + #expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + cmd = 'cd %s && %s -c 1 "select count(*) from *" -C none --max-attached-sqlite-databases=10' % (tmpfolder,Q_EXECUTABLE) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b(str(BATCH_SIZE*FILE_COUNT))) + + self.cleanup_folder(tmpfolder) + + def test_maxing_out_max_attached_database_limits__regular_files(self): + BATCH_SIZE = 50 + FILE_COUNT = 40 + + numbers_as_text = batch([str(x) for x in range(1,1+BATCH_SIZE*FILE_COUNT)],n=BATCH_SIZE) + + content_list = map(six.b,["\n".join(x)+'\n' for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x,range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d,'split-files','attach-limit') + #expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + unioned_subquery = " UNION ALL ".join(["select * from %s/%s" % (tmpfolder,filename) for filename in filename_list]) + cmd = 'cd %s && %s -c 1 "select count(*) from (%s)" -C none --max-attached-sqlite-databases=10' % (tmpfolder,Q_EXECUTABLE,unioned_subquery) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b(str(BATCH_SIZE*FILE_COUNT))) + + self.cleanup_folder(tmpfolder) + + def test_maxing_out_max_attached_database_limits__with_qsql_files_below_attached_limit(self): + MAX_ATTACHED_SQLITE_DATABASES = 10 + + BATCH_SIZE = 50 + FILE_COUNT = MAX_ATTACHED_SQLITE_DATABASES - 1 + + numbers_as_text = batch([str(x) for x in range(1,1+BATCH_SIZE*FILE_COUNT)],n=BATCH_SIZE) + + content_list = map(six.b,["\n".join(x)+'\n' for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x,range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d,'split-files','attach-limit') + #expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + # Execute the query with -C readwrite, so all qsql files will be created + unioned_subquery = " UNION ALL ".join(["select * from %s/%s" % (tmpfolder,filename) for filename in filename_list]) + cmd = 'cd %s && %s -c 1 "select count(*) from (%s)" -C readwrite --max-attached-sqlite-databases=%s' % (tmpfolder,Q_EXECUTABLE,unioned_subquery,MAX_ATTACHED_SQLITE_DATABASES) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b(str(BATCH_SIZE*FILE_COUNT))) + + # Now execute the same query with -C readwrite, so all files will be read directly from the qsql files + cmd = 'cd %s && %s -c 1 "select count(*) from (%s)" -C readwrite' % (tmpfolder,Q_EXECUTABLE,unioned_subquery) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b(str(BATCH_SIZE*FILE_COUNT))) + + self.cleanup_folder(tmpfolder) + + def test_maxing_out_max_attached_database_limits__with_qsql_files_above_attached_limit(self): + MAX_ATTACHED_SQLITE_DATABASES = 10 + + BATCH_SIZE = 50 + # Here's the difference from test_maxing_out_max_attached_database_limits__with_qsql_files_below_attached_limit + # We're trying to cache 2 times the number of files than the number of databases that can be attached. + # Expectation is that only a part of the files will be cached + FILE_COUNT = MAX_ATTACHED_SQLITE_DATABASES * 2 + + numbers_as_text = batch([str(x) for x in range(1,1+BATCH_SIZE*FILE_COUNT)],n=BATCH_SIZE) + + content_list = map(six.b,["\n".join(x)+'\n' for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x,range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d,'split-files','attach-limit') + #expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + # Execute the query with -C readwrite, so all qsql files will be created + unioned_subquery = " UNION ALL ".join(["select * from %s/%s" % (tmpfolder,filename) for filename in filename_list]) + cmd = 'cd %s && %s -c 1 "select count(*) from (%s)" -C readwrite --max-attached-sqlite-databases=%s' % (tmpfolder,Q_EXECUTABLE,unioned_subquery,MAX_ATTACHED_SQLITE_DATABASES) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b(str(BATCH_SIZE*FILE_COUNT))) + + # Now execute the same query with -C readwrite, so all files will be read directly from the qsql files + cmd = 'cd %s && %s -c 1 "select count(*) from (%s)" -C readwrite' % (tmpfolder,Q_EXECUTABLE,unioned_subquery) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b(str(BATCH_SIZE*FILE_COUNT))) + + from glob import glob + files_in_folder = [os.path.basename(x) for x in glob('%s/*' % (tmpfolder))] + + expected_files_in_folder = filename_list + list(map(lambda x: 'file-%s.qsql' % x,range(MAX_ATTACHED_SQLITE_DATABASES-2))) + + self.assertEqual(sorted(files_in_folder),sorted(expected_files_in_folder)) + + self.cleanup_folder(tmpfolder) + + def test_maxing_out_max_attached_database_limits__with_directly_using_qsql_files(self): + MAX_ATTACHED_SQLITE_DATABASES = 10 + + BATCH_SIZE = 50 + FILE_COUNT = MAX_ATTACHED_SQLITE_DATABASES * 2 + + numbers_as_text = batch([str(x) for x in range(1,1+BATCH_SIZE*FILE_COUNT)],n=BATCH_SIZE) + + content_list = map(six.b,["\n".join(x)+'\n' for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x,range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d,'split-files','attach-limit') + #expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + # Prepare qsql for each of the files (separately, just for simplicity) + for fn in filename_list: + cmd = 'cd %s && %s -c 1 "select count(*) from %s" -C readwrite' % (tmpfolder,Q_EXECUTABLE,fn) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + # Now execute a big query which uses the created qsql files + unioned_subquery = " UNION ALL ".join(["select * from %s/%s.qsql" % (tmpfolder,filename) for filename in filename_list]) + + cmd = 'cd %s && %s -c 1 "select count(*) from (%s)" -C readwrite' % (tmpfolder,Q_EXECUTABLE,unioned_subquery) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b(str(BATCH_SIZE*FILE_COUNT))) + + self.cleanup_folder(tmpfolder) + + def test_too_many_open_files_for_one_table(self): + # Previously file opening was parallel, causing too-many-open-files + + MAX_ALLOWED_FILES = 500 + + BATCH_SIZE = 2 + FILE_COUNT = MAX_ALLOWED_FILES + 1 + + numbers_as_text = batch([str(x) for x in range(1,1+BATCH_SIZE*FILE_COUNT)],n=BATCH_SIZE) + + content_list = map(six.b,["\n".join(x) for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x,range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d,'split-files','attach-limit') + + cmd = 'cd %s && %s -c 1 "select count(*) from * where 1 = 1 or c1 != 2" -C none' % (tmpfolder,Q_EXECUTABLE) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 82) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + x = six.b('Maximum source files for table must be %s. Table is name is %s/* Number of actual files is %s' % (MAX_ALLOWED_FILES,os.path.realpath(tmpfolder),FILE_COUNT)) + print(x) + self.assertEqual(e[0],x) + + self.cleanup_folder(tmpfolder) + + def test_many_open_files_for_one_table(self): + # Previously file opening was parallel, causing too-many-open-files + + BATCH_SIZE = 2 + FILE_COUNT = 500 + + numbers_as_text = batch([str(x) for x in range(1,1+BATCH_SIZE*FILE_COUNT)],n=BATCH_SIZE) + + content_list = map(six.b,["\n".join(x) for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x,range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder = self.create_folder_with_files(d,'split-files','attach-limit') + #expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + cmd = 'cd %s && %s -c 1 "select count(*) from * where 1 = 1 or c1 != 2" -C none' % (tmpfolder,Q_EXECUTABLE) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b(str(BATCH_SIZE*FILE_COUNT))) + + self.cleanup_folder(tmpfolder) + + def test_many_open_files_for_two_tables(self): + BATCH_SIZE = 2 + FILE_COUNT = 500 + + numbers_as_text = batch([str(x) for x in range(1, 1 + BATCH_SIZE * FILE_COUNT)], n=BATCH_SIZE) + + content_list = map(six.b, ["\n".join(x) for x in numbers_as_text]) + + filename_list = list(map(lambda x: 'file-%s' % x, range(FILE_COUNT))) + d = collections.OrderedDict(zip(filename_list, content_list)) + + tmpfolder1 = self.create_folder_with_files(d, 'split-files1', 'blah') + tmpfolder2 = self.create_folder_with_files(d, 'split-files1', 'blah') + + cmd = '%s -c 1 "select count(*) from %s/* a left join %s/* b on (a.c1 = b.c1)" -C none' % ( + Q_EXECUTABLE, + tmpfolder1, + tmpfolder2) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b(str(BATCH_SIZE * FILE_COUNT))) + + self.cleanup_folder(tmpfolder1) + self.cleanup_folder(tmpfolder2) + + +class GzippingTests(AbstractQTestCase): + + def test_gzipped_file(self): + tmpfile = self.create_file_with_data( + six.b('\x1f\x8b\x08\x08\xf2\x18\x12S\x00\x03xxxxxx\x003\xe42\xe22\xe62\xe12\xe52\xe32\xe7\xb2\xe0\xb2\xe424\xe0\x02\x00\xeb\xbf\x8a\x13\x15\x00\x00\x00')) + + cmd = Q_EXECUTABLE + ' -z "select sum(c1),avg(c1) from %s"' % tmpfile.name + + retcode, o, e = run_command(cmd) + self.assertTrue(retcode == 0) + self.assertTrue(len(o) == 1) + self.assertTrue(len(e) == 0) + + s = sum(range(1, 11)) + self.assertTrue(o[0] == six.b('%s %s' % (s, s / 10.0))) + + self.cleanup(tmpfile) + + +class DelimiterTests(AbstractQTestCase): + + def test_delimition_mistake_with_header(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + + cmd = Q_EXECUTABLE + ' -d " " "select * from %s" -H' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 2) + + self.assertTrue(e[0].startswith(six.b("Bad header row"))) + self.assertTrue(six.b("Column name cannot contain commas") in e[1]) + + self.cleanup(tmpfile) + + def test_tab_delimition_parameter(self): + tmpfile = self.create_file_with_data( + sample_data_no_header.replace(six.b(","), six.b("\t"))) + cmd = Q_EXECUTABLE + ' -t "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + self.assertEqual(o[0], sample_data_rows[0].replace(six.b(","), six.b("\t"))) + self.assertEqual(o[1], sample_data_rows[1].replace(six.b(","), six.b("\t"))) + self.assertEqual(o[2], sample_data_rows[2].replace(six.b(","), six.b("\t"))) + + self.cleanup(tmpfile) + + def test_pipe_delimition_parameter(self): + tmpfile = self.create_file_with_data( + sample_data_no_header.replace(six.b(","), six.b("|"))) + cmd = Q_EXECUTABLE + ' -p "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + self.assertEqual(o[0], sample_data_rows[0].replace(six.b(","), six.b("|"))) + self.assertEqual(o[1], sample_data_rows[1].replace(six.b(","), six.b("|"))) + self.assertEqual(o[2], sample_data_rows[2].replace(six.b(","), six.b("|"))) + + self.cleanup(tmpfile) + + def test_tab_delimition_parameter__with_manual_override_attempt(self): + tmpfile = self.create_file_with_data( + sample_data_no_header.replace(six.b(","), six.b("\t"))) + cmd = Q_EXECUTABLE + ' -t -d , "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 1) + self.assertEqual(o[0], sample_data_rows[0].replace(six.b(","), six.b("\t"))) + self.assertEqual(o[1], sample_data_rows[1].replace(six.b(","), six.b("\t"))) + self.assertEqual(o[2], sample_data_rows[2].replace(six.b(","), six.b("\t"))) + self.assertEqual(e[0],six.b('Warning: -t parameter overrides -d parameter (,)')) + + self.cleanup(tmpfile) + + def test_pipe_delimition_parameter__with_manual_override_attempt(self): + tmpfile = self.create_file_with_data( + sample_data_no_header.replace(six.b(","), six.b("|"))) + cmd = Q_EXECUTABLE + ' -p -d , "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 1) + self.assertEqual(o[0], sample_data_rows[0].replace(six.b(","), six.b("|"))) + self.assertEqual(o[1], sample_data_rows[1].replace(six.b(","), six.b("|"))) + self.assertEqual(o[2], sample_data_rows[2].replace(six.b(","), six.b("|"))) + self.assertEqual(e[0],six.b('Warning: -p parameter overrides -d parameter (,)')) + + self.cleanup(tmpfile) + + def test_output_delimiter(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , -D "|" "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], sample_data_rows[0].replace(six.b(","), six.b("|"))) + self.assertEqual(o[1], sample_data_rows[1].replace(six.b(","), six.b("|"))) + self.assertEqual(o[2], sample_data_rows[2].replace(six.b(","), six.b("|"))) + + self.cleanup(tmpfile) + + def test_output_delimiter_tab_parameter(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , -T "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], sample_data_rows[0].replace(six.b(","), six.b("\t"))) + self.assertEqual(o[1], sample_data_rows[1].replace(six.b(","), six.b("\t"))) + self.assertEqual(o[2], sample_data_rows[2].replace(six.b(","), six.b("\t"))) + + self.cleanup(tmpfile) + + def test_output_delimiter_pipe_parameter(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , -P "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], sample_data_rows[0].replace(six.b(","), six.b("|"))) + self.assertEqual(o[1], sample_data_rows[1].replace(six.b(","), six.b("|"))) + self.assertEqual(o[2], sample_data_rows[2].replace(six.b(","), six.b("|"))) + + self.cleanup(tmpfile) + + def test_output_delimiter_tab_parameter__with_manual_override_attempt(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , -T -D "|" "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 1) + + self.assertEqual(o[0], sample_data_rows[0].replace(six.b(","), six.b("\t"))) + self.assertEqual(o[1], sample_data_rows[1].replace(six.b(","), six.b("\t"))) + self.assertEqual(o[2], sample_data_rows[2].replace(six.b(","), six.b("\t"))) + self.assertEqual(e[0], six.b('Warning: -T parameter overrides -D parameter (|)')) + + self.cleanup(tmpfile) + + def test_output_delimiter_pipe_parameter__with_manual_override_attempt(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , -P -D ":" "select c1,c2,c3 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 1) + + self.assertEqual(o[0], sample_data_rows[0].replace(six.b(","), six.b("|"))) + self.assertEqual(o[1], sample_data_rows[1].replace(six.b(","), six.b("|"))) + self.assertEqual(o[2], sample_data_rows[2].replace(six.b(","), six.b("|"))) + self.assertEqual(e[0],six.b('Warning: -P parameter overrides -D parameter (:)')) + + self.cleanup(tmpfile) + + +class AnalysisTests(AbstractQTestCase): + + def test_analyze_result(self): + d = "\n".join(['%s\t%s\t%s' % (x+1,x+1,x+1) for x in range(100)]) + tmpfile = self.create_file_with_data(six.b(d)) + + cmd = Q_EXECUTABLE + ' -c 1 "select count(*) from %s" -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 5) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1], six.b(' Sources:')) + self.assertEqual(o[2], six.b(' source_type: file source: %s' %(tmpfile.name))) + self.assertEqual(o[3], six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `c1` - text')) + + self.cleanup(tmpfile) + + def test_analyze_result_with_data_stream(self): + d = "\n".join(['%s\t%s\t%s' % (x+1,x+1,x+1) for x in range(100)]) + tmpfile = self.create_file_with_data(six.b(d)) + + cmd = 'cat %s | %s -c 1 "select count(*) from -" -A' % (tmpfile.name,Q_EXECUTABLE) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 5) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('Table: -')) + self.assertEqual(o[1], six.b(' Sources:')) + self.assertEqual(o[2], six.b(' source_type: data-stream source: stdin')) + self.assertEqual(o[3], six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `c1` - text')) + + self.cleanup(tmpfile) + + def test_column_analysis(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + + cmd = Q_EXECUTABLE + ' -d , "select c1 from %s" -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(o[0], six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s' % tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `c1` - text')) + self.assertEqual(o[5], six.b(' `c2` - int')) + self.assertEqual(o[6], six.b(' `c3` - int')) + + self.cleanup(tmpfile) + + def test_column_analysis_with_mixed_ints_and_floats(self): + tmpfile = self.create_file_with_data(six.b("""planet_id,name,diameter_km,length_of_day_hours\n1000,Earth,12756,24\n2000,Mars,6792,24.7\n3000,Jupiter,142984,9.9""")) + + cmd = Q_EXECUTABLE + ' -d , -H "select * from %s" -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),8) + self.assertEqual(len(e),0) + self.assertEqual(o[0], six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s' % tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `planet_id` - int')) + self.assertEqual(o[5], six.b(' `name` - text')) + self.assertEqual(o[6], six.b(' `diameter_km` - int')) + self.assertEqual(o[7], six.b(' `length_of_day_hours` - real')) + + self.cleanup(tmpfile) + + def test_column_analysis_with_mixed_ints_and_floats_and_nulls(self): + tmpfile = self.create_file_with_data(six.b("""planet_id,name,diameter_km,length_of_day_hours\n1000,Earth,12756,24\n2000,Mars,6792,24.7\n2500,Venus,,\n3000,Jupiter,142984,9.9""")) + + cmd = Q_EXECUTABLE + ' -d , -H "select * from %s" -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),8) + self.assertEqual(len(e),0) + self.assertEqual(o[0], six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s' % tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `planet_id` - int')) + self.assertEqual(o[5], six.b(' `name` - text')) + self.assertEqual(o[6], six.b(' `diameter_km` - int')) + self.assertEqual(o[7], six.b(' `length_of_day_hours` - real')) + + self.cleanup(tmpfile) + + def test_column_analysis_no_header(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + + cmd = Q_EXECUTABLE + ' -d , "select c1 from %s" -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(o[0], six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s' % tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `c1` - text')) + self.assertEqual(o[5], six.b(' `c2` - int')) + self.assertEqual(o[6], six.b(' `c3` - int')) + + def test_column_analysis_with_unexpected_header(self): + tmpfile = self.create_file_with_data(sample_data_with_header) + cmd = Q_EXECUTABLE + ' -d , "select c1 from %s" -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 7) + self.assertEqual(len(e), 1) + + self.assertEqual(o[0], six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s' % tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4],six.b(' `c1` - text')) + self.assertEqual(o[5],six.b(' `c2` - text')) + self.assertEqual(o[6],six.b(' `c3` - text')) + + self.assertEqual( + e[0], six.b('Warning - There seems to be header line in the file, but -H has not been specified. All fields will be detected as text fields, and the header line will appear as part of the data')) + + self.cleanup(tmpfile) + + def test_column_analysis_for_spaces_in_header_row(self): + tmpfile = self.create_file_with_data( + header_row_with_spaces + six.b("\n") + sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , "select name,\\`value 1\\` from %s" -H -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 7) + + self.assertEqual(o[0], six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s' % tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `name` - text')) + self.assertEqual(o[5], six.b(' `value 1` - int')) + self.assertEqual(o[6], six.b(' `value2` - int')) + + self.cleanup(tmpfile) + + def test_column_analysis_with_header(self): + tmpfile = self.create_file_with_data(sample_data_with_header) + cmd = Q_EXECUTABLE + ' -d , "select c1 from %s" -A -H' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode, 0) + self.assertEqual(len(o),7) + self.assertEqual(len(e),2) + self.assertEqual(o[0], six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s' % tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `name` - text')) + self.assertEqual(o[5], six.b(' `value1` - int')) + self.assertEqual(o[6], six.b(' `value2` - int')) + + self.assertEqual(e[0],six.b('query error: no such column: c1')) + self.assertTrue(e[1].startswith(six.b('Warning - There seems to be a '))) + + self.cleanup(tmpfile) + + + +class StdInTests(AbstractQTestCase): + + def test_stdin_input(self): + cmd = six.b('printf "%s" | ' + Q_EXECUTABLE + ' -d , "select c1,c2,c3 from -"') % sample_data_no_header + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], sample_data_rows[0]) + self.assertEqual(o[1], sample_data_rows[1]) + self.assertEqual(o[2], sample_data_rows[2]) + + def test_attempt_to_unzip_stdin(self): + tmpfile = self.create_file_with_data( + six.b('\x1f\x8b\x08\x08\xf2\x18\x12S\x00\x03xxxxxx\x003\xe42\xe22\xe62\xe12\xe52\xe32\xe7\xb2\xe0\xb2\xe424\xe0\x02\x00\xeb\xbf\x8a\x13\x15\x00\x00\x00')) + + cmd = 'cat %s | ' % tmpfile.name + Q_EXECUTABLE + ' -z "select sum(c1),avg(c1) from -"' + + retcode, o, e = run_command(cmd) + self.assertTrue(retcode != 0) + self.assertTrue(len(o) == 0) + self.assertTrue(len(e) == 1) + + self.assertEqual(e[0],six.b('Cannot decompress standard input. Pipe the input through zcat in order to decompress.')) + + self.cleanup(tmpfile) + +class QuotingTests(AbstractQTestCase): + def test_non_quoted_values_in_quoted_data(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data) + + cmd = Q_EXECUTABLE + ' -d " " "select c1 from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),4) + + self.assertTrue(o[0],'non_quoted') + self.assertTrue(o[1],'control-value-1') + self.assertTrue(o[2],'non-quoted-value') + self.assertTrue(o[3],'control-value-1') + + self.cleanup(tmp_data_file) + + def test_regular_quoted_values_in_quoted_data(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data) + + cmd = Q_EXECUTABLE + ' -d " " "select c2 from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),4) + + self.assertTrue(o[0],'regular_double_quoted') + self.assertTrue(o[1],'control-value-2') + self.assertTrue(o[2],'this is a quoted value') + self.assertTrue(o[3],'control-value-2') + + self.cleanup(tmp_data_file) + + def test_double_double_quoted_values_in_quoted_data(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data) + + cmd = Q_EXECUTABLE + ' -d " " "select c3 from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),4) + + self.assertTrue(o[0],'double_double_quoted') + self.assertTrue(o[1],'control-value-3') + self.assertTrue(o[2],'this is a "double double" quoted value') + self.assertTrue(o[3],'control-value-3') + + self.cleanup(tmp_data_file) + + def test_escaped_double_quoted_values_in_quoted_data(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data) + + cmd = Q_EXECUTABLE + ' -d " " "select c4 from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),4) + + self.assertTrue(o[0],'escaped_double_quoted') + self.assertTrue(o[1],'control-value-4') + self.assertTrue(o[2],'this is an escaped "quoted value"') + self.assertTrue(o[3],'control-value-4') + + self.cleanup(tmp_data_file) + + def test_none_input_quoting_mode_in_relaxed_mode(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + cmd = Q_EXECUTABLE + ' -d " " -m relaxed -D , -w none -W none "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('"quoted,data",23')) + self.assertEqual(o[1],six.b('unquoted-data,54,')) + + self.cleanup(tmp_data_file) + + def test_none_input_quoting_mode_in_strict_mode(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + cmd = Q_EXECUTABLE + ' -d " " -m strict -D , -w none "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode,0) + self.assertEqual(len(e),1) + self.assertEqual(len(o),0) + + self.assertTrue(e[0].startswith(six.b('Strict mode. Column Count is expected to identical'))) + + self.cleanup(tmp_data_file) + + def test_minimal_input_quoting_mode(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + cmd = Q_EXECUTABLE + ' -d " " -D , -w minimal "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('quoted data,23')) + self.assertEqual(o[1],six.b('unquoted-data,54')) + + self.cleanup(tmp_data_file) + + def test_all_input_quoting_mode(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + cmd = Q_EXECUTABLE + ' -d " " -D , -w all "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('quoted data,23')) + self.assertEqual(o[1],six.b('unquoted-data,54')) + + self.cleanup(tmp_data_file) + + def test_incorrect_input_quoting_mode(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + cmd = Q_EXECUTABLE + ' -d " " -D , -w unknown_wrapping_mode "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode,0) + self.assertEqual(len(e),1) + self.assertEqual(len(o),0) + + self.assertTrue(e[0].startswith(six.b('Input quoting mode can only be one of all,minimal,none'))) + self.assertTrue(six.b('unknown_wrapping_mode') in e[0]) + + self.cleanup(tmp_data_file) + + def test_none_output_quoting_mode(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + cmd = Q_EXECUTABLE + ' -d " " -D , -w all -W none "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('quoted data,23')) + self.assertEqual(o[1],six.b('unquoted-data,54')) + + self.cleanup(tmp_data_file) + + def test_minimal_output_quoting_mode__without_need_to_quote_in_output(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + cmd = Q_EXECUTABLE + ' -d " " -D , -w all -W minimal "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('quoted data,23')) + self.assertEqual(o[1],six.b('unquoted-data,54')) + + self.cleanup(tmp_data_file) + + def test_minimal_output_quoting_mode__with_need_to_quote_in_output_due_to_delimiter(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + # output delimiter is set to space, so the output will contain it + cmd = Q_EXECUTABLE + ' -d " " -D " " -w all -W minimal "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('"quoted data" 23')) + self.assertEqual(o[1],six.b('unquoted-data 54')) + + self.cleanup(tmp_data_file) + + def test_minimal_output_quoting_mode__with_need_to_quote_in_output_due_to_newline(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2_with_newline) + + # Delimiter is set to colon (:), so it will not be inside the data values (this will make sure that the newline is the one causing the quoting) + cmd = Q_EXECUTABLE + " -d ':' -w all -W minimal \"select c1,c2,replace(c1,'with' || x'0a' || 'a new line inside it','NEWLINE-REMOVED') from %s\"" % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),3) + + self.assertEqual(o[0],six.b('"quoted data with')) + # Notice that the third column here is not quoted, because we replaced the newline with something else + self.assertEqual(o[1],six.b('a new line inside it":23:quoted data NEWLINE-REMOVED')) + self.assertEqual(o[2],six.b('unquoted-data:54:unquoted-data')) + + self.cleanup(tmp_data_file) + + def test_nonnumeric_output_quoting_mode(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + cmd = Q_EXECUTABLE + ' -d " " -D , -w all -W nonnumeric "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('"quoted data",23')) + self.assertEqual(o[1],six.b('"unquoted-data",54')) + + self.cleanup(tmp_data_file) + + def test_all_output_quoting_mode(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data2) + + cmd = Q_EXECUTABLE + ' -d " " -D , -w all -W all "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('"quoted data","23"')) + self.assertEqual(o[1],six.b('"unquoted-data","54"')) + + self.cleanup(tmp_data_file) + + def _internal_test_consistency_of_chaining_output_to_input(self,input_data,input_wrapping_mode,output_wrapping_mode): + + tmp_data_file = self.create_file_with_data(input_data) + + basic_cmd = Q_EXECUTABLE + ' -w %s -W %s "select * from -"' % (input_wrapping_mode,output_wrapping_mode) + chained_cmd = 'cat %s | %s | %s | %s' % (tmp_data_file.name,basic_cmd,basic_cmd,basic_cmd) + + retcode, o, e = run_command(chained_cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(six.b("\n").join(o),input_data) + + self.cleanup(tmp_data_file) + + def test_consistency_of_chaining_minimal_wrapping_to_minimal_wrapping(self): + input_data = six.b('"quoted data" 23\nunquoted-data 54') + self._internal_test_consistency_of_chaining_output_to_input(input_data,'minimal','minimal') + + def test_consistency_of_chaining_all_wrapping_to_all_wrapping(self): + input_data = six.b('"quoted data" "23"\n"unquoted-data" "54"') + self._internal_test_consistency_of_chaining_output_to_input(input_data,'all','all') + + def test_input_field_quoting_and_data_types_with_encoding(self): + OUTPUT_ENCODING = 'utf-8' + + # Checks combination of minimal input field quoting, with special characters that need to be decoded - + # Both content and proper data types are verified + data = six.b('111,22.22,"testing text with special characters - citt\xc3\xa0 ",http://somekindofurl.com,12.13.14.15,12.1\n') + tmp_data_file = self.create_file_with_data(data) + + cmd = Q_EXECUTABLE + ' -d , "select * from %s" -E %s' % (tmp_data_file.name,OUTPUT_ENCODING) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),1) + + self.assertEqual(o[0].decode('utf-8'),u'111,22.22,testing text with special characters - citt\xe0 ,http://somekindofurl.com,12.13.14.15,12.1') + + cmd = Q_EXECUTABLE + ' -d , "select * from %s" -A' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),10) + + self.assertEqual(o[0],six.b('Table: %s' % tmp_data_file.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s' % tmp_data_file.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4],six.b(' `c1` - int')) + self.assertEqual(o[5],six.b(' `c2` - real')) + self.assertEqual(o[6],six.b(' `c3` - text')) + self.assertEqual(o[7],six.b(' `c4` - text')) + self.assertEqual(o[8],six.b(' `c5` - text')) + self.assertEqual(o[9],six.b(' `c6` - real')) + + self.cleanup(tmp_data_file) + + def test_multiline_double_double_quoted_values_in_quoted_data(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data) + + # FIXME Need to convert \0a to proper encoding suitable for the person running the tests. + cmd = Q_EXECUTABLE + ' -d " " "select replace(c5,X\'0A\',\'::\') from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),4) + + self.assertTrue(o[0],six.b('multiline_double_double_quoted')) + self.assertTrue(o[1],six.b('control-value-5')) + self.assertTrue(o[2],six.b('this is a double double quoted "multiline\n value".')) + self.assertTrue(o[3],six.b('control-value-5')) + + self.cleanup(tmp_data_file) + + def test_multiline_escaped_double_quoted_values_in_quoted_data(self): + tmp_data_file = self.create_file_with_data(sample_quoted_data) + + # FIXME Need to convert \0a to proper encoding suitable for the person running the tests. + cmd = Q_EXECUTABLE + ' -d " " "select replace(c6,X\'0A\',\'::\') from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),4) + + self.assertTrue(o[0],'multiline_escaped_double_quoted') + self.assertTrue(o[1],'control-value-6') + self.assertTrue(o[2],'this is an escaped "multiline:: value".') + self.assertTrue(o[3],'control-value-6') + + self.cleanup(tmp_data_file) + + def test_disable_double_double_quoted_data_flag__values(self): + # This test (and flag) is meant to verify backward comptibility only. It is possible that + # this flag will be removed completely in the future + + tmp_data_file = self.create_file_with_data(double_double_quoted_data) + + cmd = Q_EXECUTABLE + ' -d " " --disable-double-double-quoting "select c2 from %s" -W none' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('double_double_quoted')) + self.assertEqual(o[1],six.b('this is a quoted value with "double')) + + cmd = Q_EXECUTABLE + ' -d " " --disable-double-double-quoting "select c3 from %s" -W none' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('')) + self.assertEqual(o[1],six.b('double')) + + cmd = Q_EXECUTABLE + ' -d " " --disable-double-double-quoting "select c4 from %s" -W none' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('')) + self.assertEqual(o[1],six.b('quotes"""')) + + self.cleanup(tmp_data_file) + + def test_disable_escaped_double_quoted_data_flag__values(self): + # This test (and flag) is meant to verify backward comptibility only. It is possible that + # this flag will be removed completely in the future + + tmp_data_file = self.create_file_with_data(escaped_double_quoted_data) + + cmd = Q_EXECUTABLE + ' -d " " --disable-escaped-double-quoting "select c2 from %s" -W none' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('escaped_double_quoted')) + self.assertEqual(o[1],six.b('this is a quoted value with \\escaped')) + + cmd = Q_EXECUTABLE + ' -d " " --disable-escaped-double-quoting "select c3 from %s" -W none' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('')) + self.assertEqual(o[1],six.b('double')) + + cmd = Q_EXECUTABLE + ' -d " " --disable-escaped-double-quoting "select c4 from %s" -W none' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('')) + self.assertEqual(o[1],six.b('quotes\\""')) + + self.cleanup(tmp_data_file) + + def test_combined_quoted_data_flags__number_of_columns_detected(self): + # This test (and flags) is meant to verify backward comptibility only. It is possible that + # these flags will be removed completely in the future + tmp_data_file = self.create_file_with_data(combined_quoted_data) + + cmd = Q_EXECUTABLE + ' -d " " --disable-double-double-quoting --disable-escaped-double-quoting "select * from %s" -A' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + o = o[o.index(six.b(' Fields:'))+1:] + + self.assertEqual(len(o),7) # found 7 fields + + cmd = Q_EXECUTABLE + ' -d " " --disable-escaped-double-quoting "select * from %s" -A' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + o = o[o.index(six.b(' Fields:'))+1:] + + self.assertEqual(len(o),5) # found 5 fields + + cmd = Q_EXECUTABLE + ' -d " " --disable-double-double-quoting "select * from %s" -A' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + o = o[o.index(six.b(' Fields:'))+1:] + + self.assertEqual(len(o),5) # found 5 fields + + cmd = Q_EXECUTABLE + ' -d " " "select * from %s" -A' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + o = o[o.index(six.b(' Fields:'))+1:] + + self.assertEqual(len(o),3) # found only 3 fields, which is the correct amount + + self.cleanup(tmp_data_file) + + +class EncodingTests(AbstractQTestCase): + + def test_utf8_with_bom_encoding(self): + utf_8_data_with_bom = six.b('\xef\xbb\xbf"typeid","limit","apcost","date","checkpointId"\n"1","2","5","1,2,3,4,5,6,7","3000,3001,3002"\n"2","2","5","1,2,3,4,5,6,7","3003,3004,3005"\n') + tmp_data_file = self.create_file_with_data(utf_8_data_with_bom,encoding=None) + + cmd = Q_EXECUTABLE + ' -d , -H -O -e utf-8-sig "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(e),0) + self.assertEqual(len(o),3) + + self.assertEqual(o[0],six.b('typeid,limit,apcost,date,checkpointId')) + self.assertEqual(o[1],six.b('1,2,5,"1,2,3,4,5,6,7","3000,3001,3002"')) + self.assertEqual(o[2],six.b('2,2,5,"1,2,3,4,5,6,7","3003,3004,3005"')) + + self.cleanup(tmp_data_file) + + +class QrcTests(AbstractQTestCase): + + def test_explicit_qrc_filename_not_found(self): + non_existent_filename = str(uuid.uuid4()) + env_to_inject = { 'QRC_FILENAME': non_existent_filename} + cmd = Q_EXECUTABLE + ' "select 1"' + retcode, o, e = run_command(cmd, env_to_inject=env_to_inject) + + self.assertEqual(retcode, 244) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertTrue(e[0] == six.b('QRC_FILENAME env var exists, but cannot find qrc file at %s' % non_existent_filename)) + + def test_explicit_qrc_filename_that_exists(self): + tmp_qrc_file = self.create_file_with_data(six.b('''[options] +output_delimiter=| +''')) + env_to_inject = { 'QRC_FILENAME': tmp_qrc_file.name} + cmd = Q_EXECUTABLE + ' "select 1,2"' + retcode, o, e = run_command(cmd, env_to_inject=env_to_inject) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertTrue(o[0] == six.b('1|2')) + + self.cleanup(tmp_qrc_file) + + def test_all_default_options(self): + # Create a qrc file that contains all default values inside the qrc file, but with some different values than the regular defaults + tmp_qrc_file = self.create_file_with_data(six.b('''[options] +analyze_only=True +beautify=True +caching_mode=readwrite +column_count=32 +delimiter=, +disable_column_type_detection=True +disable_double_double_quoting=False +disable_escaped_double_quoting=False +encoding=ascii +formatting=xxx +gzipped=True +input_quoting_mode=all +keep_leading_whitespace_in_values=True +list_user_functions=True +max_attached_sqlite_databases=888 +max_column_length_limit=8888 +mode=strict +output_delimiter=| +output_encoding=utf-8 +output_header=True +output_quoting_mode=all +overwrite_qsql=False +pipe_delimited=True +pipe_delimited_output=True +query_encoding=ascii +query_filename=query-filename +save_db_to_disk_filename=save-db-to-disk-filename +skip_header=True +tab_delimited=True +tab_delimited_output=true +verbose=True +with_universal_newlines=True +''')) + env_to_inject = { 'QRC_FILENAME': tmp_qrc_file.name} + cmd = Q_EXECUTABLE + ' --dump-defaults' + retcode, o, e = run_command(cmd, env_to_inject=env_to_inject) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 34) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b('[options]')) + o = o[1:] + + m = {} + for r in o: + key,val = r.split(six.b("="),1) + m[key] = val + + self.assertEqual(m[six.b('analyze_only')],six.b('True')) + self.assertEqual(m[six.b('beautify')],six.b('True')) + self.assertEqual(m[six.b('caching_mode')],six.b('readwrite')) + self.assertEqual(m[six.b('column_count')],six.b('32')) + self.assertEqual(m[six.b('delimiter')],six.b(',')) + self.assertEqual(m[six.b('disable_column_type_detection')],six.b('True')) + self.assertEqual(m[six.b('disable_double_double_quoting')],six.b('False')) + self.assertEqual(m[six.b('disable_escaped_double_quoting')],six.b('False')) + self.assertEqual(m[six.b('encoding')],six.b('ascii')) + self.assertEqual(m[six.b('formatting')],six.b('xxx')) + self.assertEqual(m[six.b('gzipped')],six.b('True')) + self.assertEqual(m[six.b('input_quoting_mode')],six.b('all')) + self.assertEqual(m[six.b('keep_leading_whitespace_in_values')],six.b('True')) + self.assertEqual(m[six.b('list_user_functions')],six.b('True')) + self.assertEqual(m[six.b('max_attached_sqlite_databases')],six.b('888')) + self.assertEqual(m[six.b('max_column_length_limit')],six.b('8888')) + self.assertEqual(m[six.b('mode')],six.b('strict')) + self.assertEqual(m[six.b('output_delimiter')],six.b('|')) + self.assertEqual(m[six.b('output_encoding')],six.b('utf-8')) + self.assertEqual(m[six.b('output_header')],six.b('True')) + self.assertEqual(m[six.b('output_quoting_mode')],six.b('all')) + self.assertEqual(m[six.b('overwrite_qsql')],six.b('False')) + self.assertEqual(m[six.b('pipe_delimited')],six.b('True')) + self.assertEqual(m[six.b('pipe_delimited_output')],six.b('True')) + self.assertEqual(m[six.b('query_encoding')],six.b('ascii')) + self.assertEqual(m[six.b('query_filename')],six.b('query-filename')) + self.assertEqual(m[six.b('save_db_to_disk_filename')],six.b('save-db-to-disk-filename')) + self.assertEqual(m[six.b('skip_header')],six.b('True')) + self.assertEqual(m[six.b('tab_delimited')],six.b('True')) + self.assertEqual(m[six.b('tab_delimited_output')],six.b('True')) + self.assertEqual(m[six.b('verbose')],six.b('True')) + self.assertEqual(m[six.b('with_universal_newlines')],six.b('True')) + + self.cleanup(tmp_qrc_file) + + def test_caching_readwrite_using_qrc_file(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + tmpfile_folder = os.path.dirname(tmpfile.name) + tmpfile_filename = os.path.basename(tmpfile.name) + expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -d , "select * from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),3) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('a,1,0')) + self.assertEqual(o[1],six.b('b,2,0')) + self.assertEqual(o[2],six.b('c,,0')) + + # Ensure default does not create a cache file + self.assertTrue(not os.path.exists(expected_cache_filename)) + + tmp_qrc_file = self.create_file_with_data(six.b('''[options] +caching_mode=readwrite +''')) + env_to_inject = { 'QRC_FILENAME': tmp_qrc_file.name} + cmd = Q_EXECUTABLE + ' -d , "select * from %s"' % tmpfile.name + retcode, o, e = run_command(cmd, env_to_inject=env_to_inject) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),3) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('a,1,0')) + self.assertEqual(o[1],six.b('b,2,0')) + self.assertEqual(o[2],six.b('c,,0')) + + # Ensure that qrc file caching is being used and caching is activated (cache file should exist) + self.assertTrue(os.path.exists(expected_cache_filename)) + + self.cleanup(tmp_qrc_file) + self.cleanup(tmpfile) + + +class QsqlUsageTests(AbstractQTestCase): + + def test_concatenate_same_qsql_file_with_single_table(self): + numbers = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + + qsql_file_data = self.arrays_to_qsql_file_content([six.b('aa'), six.b('bb'), six.b('cc')], numbers) + + tmpfile = self.create_file_with_data(qsql_file_data,suffix='.qsql') + + cmd = Q_EXECUTABLE + ' -t "select count(*) from (select * from %s union all select * from %s)"' % (tmpfile.name,tmpfile.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('20000')) + + def test_query_qsql_with_single_table(self): + numbers = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + + qsql_file_data = self.arrays_to_qsql_file_content([six.b('aa'), six.b('bb'), six.b('cc')], numbers) + + tmpfile = self.create_file_with_data(qsql_file_data) + + cmd = Q_EXECUTABLE + ' -t "select sum(aa),sum(bb),sum(cc) from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('50005000\t50005000\t50005000')) + + def test_query_qsql_with_single_table_with_explicit_non_existent_tablename(self): + numbers = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + + qsql_file_data = self.arrays_to_qsql_file_content([six.b('aa'), six.b('bb'), six.b('cc')], numbers) + + tmpfile = self.create_file_with_data(qsql_file_data) + + c = sqlite3.connect(tmpfile.name) + actual_table_name = c.execute('select temp_table_name from _qcatalog').fetchall()[0][0] + c.close() + + + cmd = '%s -t "select sum(aa),sum(bb),sum(cc) from %s:::non-existent"' % (Q_EXECUTABLE,tmpfile.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 84) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + self.assertEqual(e[0],six.b('Table non-existent could not be found in qsql file %s . Existing table names: %s' % (tmpfile.name,actual_table_name))) + + def test_query_qsql_with_single_table_with_explicit_table_name(self): + numbers = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + + qsql_file_data = self.arrays_to_qsql_file_content([six.b('aa'), six.b('bb'), six.b('cc')], numbers) + + tmpfile = self.create_file_with_data(qsql_file_data) + + c = sqlite3.connect(tmpfile.name) + actual_table_name = c.execute('select temp_table_name from _qcatalog').fetchall()[0][0] + c.close() + + + cmd = '%s -t "select sum(aa),sum(bb),sum(cc) from %s:::%s"' % (Q_EXECUTABLE,tmpfile.name,actual_table_name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('50005000\t50005000\t50005000')) + + def test_query_multi_qsql_with_single_table(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + qsql_file_data1 = self.arrays_to_qsql_file_content([six.b('aa'), six.b('bb'), six.b('cc')], numbers1) + tmpfile1 = self.create_file_with_data(qsql_file_data1,suffix='.qsql') + + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + qsql_file_data2 = self.arrays_to_qsql_file_content([six.b('aa'), six.b('bb'), six.b('cc')], numbers2) + tmpfile2 = self.create_file_with_data(qsql_file_data2,suffix='.qsql') + + cmd = Q_EXECUTABLE + ' -t "select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s small_file left join %s large_file on (large_file.aa == small_file.bb)"' % (tmpfile2.name,tmpfile1.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('55\t55\t55')) + + def test_query_concatenated_qsqls_each_with_single_table(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + qsql_file_data1 = self.arrays_to_qsql_file_content([six.b('aa'), six.b('bb'), six.b('cc')], numbers1) + tmpfile1 = self.create_file_with_data(qsql_file_data1,suffix='.qsql') + + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + qsql_file_data2 = self.arrays_to_qsql_file_content([six.b('aa'), six.b('bb'), six.b('cc')], numbers2) + tmpfile2 = self.create_file_with_data(qsql_file_data2,suffix='.qsql') + + cmd = Q_EXECUTABLE + ' -t "select sum(aa),sum(bb),sum(cc) from (select * from %s union all select * from %s)"' % (tmpfile2.name,tmpfile1.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('50005055\t50005055\t50005055')) + + def test_concatenated_qsql_and_data_stream__column_names_mismatch(self): + N1 = 10000 + N2 = 100 + + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, N1 + 1)] + csv_file_data1 = self.arrays_to_csv_file_content(six.b('\t'),[six.b('aa'), six.b('bb'), six.b('cc')], numbers1) + tmpfile1 = self.create_file_with_data(csv_file_data1) + expected_cache_filename1 = '%s.qsql' % tmpfile1.name + + cmd = Q_EXECUTABLE + ' -H -t "select count(*) from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertTrue(os.path.exists(expected_cache_filename1)) + + cmd = 'seq 1 %s | %s -c 1 "select count(*) from (select * from %s UNION ALL select * from -)"' % (N2, Q_EXECUTABLE,expected_cache_filename1) + + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 1) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + self.assertEqual(e[0],six.b('query error: SELECTs to the left and right of UNION ALL do not have the same number of result columns')) + + def test_concatenated_qsql_and_data_stream(self): + N1 = 10000 + N2 = 100 + + numbers1 = [[six.b(str(i))] for i in range(1, N1 + 1)] + csv_file_data1 = self.arrays_to_csv_file_content(six.b('\t'),[six.b('c1')], numbers1) + tmpfile1 = self.create_file_with_data(csv_file_data1) + expected_cache_filename1 = '%s.qsql' % tmpfile1.name + + cmd = Q_EXECUTABLE + ' -H -t "select count(*) from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertTrue(os.path.exists(expected_cache_filename1)) + + cmd = 'seq 1 %s | %s -t -c 1 "select count(*),sum(c1) from (select * from %s UNION ALL select * from -)"' % (N2, Q_EXECUTABLE,expected_cache_filename1) + + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('%s\t%s' % (N1+N2,sum(range(1,N1+1)) + sum(range(1,N2+1))))) + + def test_concatenated_qsql_and_data_stream__explicit_table_name(self): + N1 = 10000 + N2 = 100 + + numbers1 = [[six.b(str(i))] for i in range(1, N1 + 1)] + csv_file_data1 = self.arrays_to_csv_file_content(six.b('\t'),[six.b('c1')], numbers1) + tmpfile1 = self.create_file_with_data(csv_file_data1) + tmpfile1_expected_table_name = os.path.basename(tmpfile1.name) + + expected_cache_filename1 = '%s.qsql' % tmpfile1.name + + cmd = Q_EXECUTABLE + ' -H -t "select count(*) from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertTrue(os.path.exists(expected_cache_filename1)) + + cmd = 'seq 1 %s | %s -t -c 1 "select count(*),sum(c1) from (select * from %s:::%s UNION ALL select * from -)"' % (N2, Q_EXECUTABLE,expected_cache_filename1,tmpfile1_expected_table_name) + + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('%s\t%s' % (N1+N2,sum(range(1,N1+1)) + sum(range(1,N2+1))))) + + def test_write_to_qsql__check_chosen_table_name(self): + numbers1 = [[six.b(str(i))] for i in range(1, 10001)] + csv_file_data1 = self.arrays_to_csv_file_content(six.b('\t'),[six.b('c1')], numbers1) + tmpfile1 = self.create_file_with_data(csv_file_data1) + expected_cache_filename1 = '%s.qsql' % tmpfile1.name + + cmd = Q_EXECUTABLE + ' -c 1 -H -t "select count(*) from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertTrue(os.path.exists(expected_cache_filename1)) + + c = sqlite3.connect(expected_cache_filename1) + qcatalog_entries = c.execute('select temp_table_name from _qcatalog').fetchall() + self.assertEqual(len(qcatalog_entries),1) + self.assertEqual(qcatalog_entries[0][0],os.path.basename(tmpfile1.name)) + + def test_concatenated_mixes_qsql_with_single_table_and_csv(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + csv_file_data1 = self.arrays_to_csv_file_content(six.b('\t'),[six.b('aa'), six.b('bb'), six.b('cc')], numbers1) + tmpfile1 = self.create_file_with_data(csv_file_data1) + expected_cache_filename1 = '%s.qsql' % tmpfile1.name + + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + csv_file_data2 = self.arrays_to_csv_file_content(six.b('\t'),[six.b('aa'), six.b('bb'), six.b('cc')], numbers2) + tmpfile2 = self.create_file_with_data(csv_file_data2) + expected_cache_filename2 = '%s.qsql' % tmpfile2.name + + + cmd = Q_EXECUTABLE + ' -H -t "select count(*) from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertTrue(os.path.exists(expected_cache_filename1)) + + cmd = Q_EXECUTABLE + ' -H -t "select count(*) from %s" -C readwrite' % tmpfile2.name + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertTrue(os.path.exists(expected_cache_filename2)) + + # csv and qsql files prepared. now test all four combinations + + cmd = Q_EXECUTABLE + ' -O -H -t "select count(*) cnt,sum(aa) sum_aa,sum(bb) sum_bb,sum(cc) sum_cc from (select * from %s union all select * from %s)"' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(o),2) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('cnt\tsum_aa\tsum_bb\tsum_cc')) + self.assertEqual(o[1],six.b('10010\t50005055\t50005055\t50005055')) + + cmd = Q_EXECUTABLE + ' -O -H -t "select count(*) cnt,sum(aa) sum_aa,sum(bb) sum_bb,sum(cc) sum_cc from (select * from %s union all select * from %s.qsql)"' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(o),2) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('cnt\tsum_aa\tsum_bb\tsum_cc')) + self.assertEqual(o[1],six.b('10010\t50005055\t50005055\t50005055')) + + cmd = Q_EXECUTABLE + ' -O -H -t "select count(*) cnt,sum(aa) sum_aa,sum(bb) sum_bb,sum(cc) sum_cc from (select * from %s.qsql union all select * from %s)"' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(o),2) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('cnt\tsum_aa\tsum_bb\tsum_cc')) + self.assertEqual(o[1],six.b('10010\t50005055\t50005055\t50005055')) + + cmd = Q_EXECUTABLE + ' -O -H -t "select count(*) cnt,sum(aa) sum_aa,sum(bb) sum_bb,sum(cc) sum_cc from (select * from %s.qsql union all select * from %s.qsql)"' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(o),2) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('cnt\tsum_aa\tsum_bb\tsum_cc')) + self.assertEqual(o[1],six.b('10010\t50005055\t50005055\t50005055')) + + def test_analysis_of_concatenated_mixes_qsql_with_single_table_and_csv(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + csv_file_data1 = self.arrays_to_csv_file_content(six.b('\t'),[six.b('aa'), six.b('bb'), six.b('cc')], numbers1) + tmpfile1 = self.create_file_with_data(csv_file_data1) + expected_cache_filename1 = '%s.qsql' % tmpfile1.name + + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + csv_file_data2 = self.arrays_to_csv_file_content(six.b('\t'),[six.b('aa'), six.b('bb'), six.b('cc')], numbers2) + tmpfile2 = self.create_file_with_data(csv_file_data2) + expected_cache_filename2 = '%s.qsql' % tmpfile2.name + + cmd = Q_EXECUTABLE + ' -H -t "select count(*) from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertTrue(os.path.exists(expected_cache_filename1)) + + cmd = Q_EXECUTABLE + ' -H -t "select count(*) from %s" -C readwrite' % tmpfile2.name + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertTrue(os.path.exists(expected_cache_filename2)) + + # csv and qsql files prepared + + # Test function, will be used multiple times, each time with a different combination + + def do_check(caching_mode, + file1_source_type,file1_table_postfix,file1_postfix, + file2_source_type,file2_table_postfix,file2_postfix): + cmd = '%s -C %s -O -H -t "select count(*) cnt,sum(aa) sum_aa,sum(bb) sum_bb,sum(cc) sum_cc from (select * from %s%s UNION ALL select * from %s%s)" -A' % ( + Q_EXECUTABLE, + caching_mode, + tmpfile1.name, + file1_table_postfix, + tmpfile2.name, + file2_table_postfix) + + retcode, o, e = run_command(cmd) + self.assertEqual(retcode, 0) + self.assertEqual(len(o),14) + self.assertEqual(len(e),0) + self.assertEqual(o, [ + six.b('Table: %s%s' % (tmpfile1.name,file1_table_postfix)), + six.b(' Sources:'), + six.b(' source_type: %s source: %s%s' % (file1_source_type,tmpfile1.name,file1_postfix)), + six.b(' Fields:'), + six.b(' `aa` - int'), + six.b(' `bb` - int'), + six.b(' `cc` - int'), + six.b('Table: %s%s' % (tmpfile2.name,file2_table_postfix)), + six.b(' Sources:'), + six.b(' source_type: %s source: %s%s' % (file2_source_type,tmpfile2.name,file2_postfix)), + six.b(' Fields:'), + six.b(' `aa` - int'), + six.b(' `bb` - int'), + six.b(' `cc` - int')]) + + # now test *the analysis results* of all four combinations, adding `-C read`, so the + # qsql will be used. Running with `-C none`, would have caused the qsql not to be used even if the qsql file exists + + do_check(caching_mode='read', + file1_source_type='qsql-file-with-original',file1_table_postfix='',file1_postfix='.qsql', + file2_source_type='qsql-file-with-original',file2_table_postfix='',file2_postfix='.qsql') + do_check('read', + file1_source_type='qsql-file-with-original',file1_table_postfix='',file1_postfix='.qsql', + file2_source_type='qsql-file',file2_table_postfix='.qsql',file2_postfix='.qsql') + do_check('read', + file1_source_type='qsql-file',file1_table_postfix='.qsql',file1_postfix='.qsql', + file2_source_type='qsql-file-with-original',file2_table_postfix='',file2_postfix='.qsql') + do_check('read', + file1_source_type='qsql-file',file1_table_postfix='.qsql',file1_postfix='.qsql', + file2_source_type='qsql-file',file2_table_postfix='.qsql',file2_postfix='.qsql') + + # Now test the all combinations again, this time with `-C none`, to make sure that by + # default, the qsql file is not used, and -A shows that fact + + do_check(caching_mode='none', + file1_source_type='file-with-unused-qsql',file1_table_postfix='',file1_postfix='', + file2_source_type='file-with-unused-qsql',file2_table_postfix='',file2_postfix='') + do_check('none', + file1_source_type='file-with-unused-qsql',file1_table_postfix='',file1_postfix='', + file2_source_type='qsql-file',file2_table_postfix='.qsql',file2_postfix='.qsql') + do_check('none', + file1_source_type='qsql-file',file1_table_postfix='.qsql',file1_postfix='.qsql', + file2_source_type='file-with-unused-qsql',file2_table_postfix='',file2_postfix='') + do_check('none', + file1_source_type='qsql-file',file1_table_postfix='.qsql',file1_postfix='.qsql', + file2_source_type='qsql-file',file2_table_postfix='.qsql',file2_postfix='.qsql') + + def test_mixed_qsql_with_single_table_and_csv__missing_header_parameter_for_csv(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + qsql_file_data1 = self.arrays_to_qsql_file_content([six.b('aa'), six.b('bb'), six.b('cc')], numbers1) + tmpfile1 = self.create_file_with_data(qsql_file_data1,suffix='.qsql') + + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + csv_file_data2 = self.arrays_to_csv_file_content(six.b('\t'),[six.b('aa'), six.b('bb'), six.b('cc')], numbers2) + tmpfile2 = self.create_file_with_data(csv_file_data2) + + cmd = Q_EXECUTABLE + ' -t "select sum(aa),sum(bb),sum(cc) from (select * from %s union all select * from %s)"' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 1) + self.assertEqual(e[0],six.b('Warning - There seems to be header line in the file, but -H has not been specified. All fields will be detected as text fields, and the header line will appear as part of the data')) + self.assertEqual(o[0],six.b('50005055.0\t50005055.0\t50005055.0')) + + def test_qsql_with_multiple_tables_direct_use(self): + numbers1 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 10001)] + qsql_filename1 = self.create_qsql_file_with_content_and_return_filename([six.b('aa'), six.b('bb'), six.b('cc')],numbers1) + expected_stored_table_name1 = os.path.basename(qsql_filename1)[:-5] + + numbers2 = [[six.b(str(i)), six.b(str(i)), six.b(str(i))] for i in range(1, 11)] + qsql_filename2 = self.create_qsql_file_with_content_and_return_filename([six.b('aa'), six.b('bb'), six.b('cc')],numbers2) + expected_stored_table_name2 = os.path.basename(qsql_filename2)[:-5] + + qsql_with_multiple_tables = self.generate_tmpfile_name(suffix='.qsql') + + cmd = '%s -t "select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s large_file left join %s small_file on (large_file.aa == small_file.bb)" -S %s' % \ + (Q_EXECUTABLE,qsql_filename1,qsql_filename2,qsql_with_multiple_tables) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 4) + self.assertEqual(e[0], six.b('Going to save data into a disk database: %s' % qsql_with_multiple_tables)) + self.assertTrue(e[1].startswith(six.b('Data has been saved into %s . Saving has taken' % qsql_with_multiple_tables))) + self.assertEqual(e[2],six.b('Query to run on the database: select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s large_file left join %s small_file on (large_file.aa == small_file.bb);' % \ + (expected_stored_table_name1,expected_stored_table_name2))) + self.assertEqual(e[3],six.b('You can run the query directly from the command line using the following command: echo "select sum(large_file.aa),sum(large_file.bb),sum(large_file.cc) from %s large_file left join %s small_file on (large_file.aa == small_file.bb)" | sqlite3 %s' % \ + (expected_stored_table_name1,expected_stored_table_name2,qsql_with_multiple_tables))) + + cmd = '%s -d , "select count(*) cnt,sum(aa),sum(bb),sum(cc) from %s:::%s"' % (Q_EXECUTABLE,qsql_with_multiple_tables,expected_stored_table_name1) + r, o, e = run_command(cmd) + + self.assertEqual(r,0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('10000,50005000,50005000,50005000')) + + def test_direct_use_of_sqlite_db_with_one_table(self): + tmpfile = self.create_file_with_data(six.b(''),suffix='.sqlite') + os.remove(tmpfile.name) + c = sqlite3.connect(tmpfile.name) + c.execute(' create table mytable (x int, y int)').fetchall() + c.execute(' insert into mytable (x,y) values (100,200),(300,400)').fetchall() + c.commit() + c.close() + + cmd = Q_EXECUTABLE + ' -t "select sum(x),sum(y) from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('400\t600')) + + cmd = Q_EXECUTABLE + ' -t "select sum(x),sum(y) from %s:::mytable"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('400\t600')) + + def test_direct_use_of_sqlite_db_with_one_table__nonexistent_table(self): + tmpfile = self.create_file_with_data(six.b(''),suffix='.sqlite') + os.remove(tmpfile.name) + c = sqlite3.connect(tmpfile.name) + c.execute(' create table some_numbers (x int, y int)').fetchall() + c.execute(' insert into some_numbers (x,y) values (100,200),(300,400)').fetchall() + c.commit() + c.close() + + cmd = Q_EXECUTABLE + ' -t "select sum(x),sum(y) from %s:::non_existent"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 85) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0],six.b('Table non_existent could not be found in sqlite file %s . Existing table names: some_numbers' % (tmpfile.name))) + + + def test_qsql_creation_and_direct_use(self): + numbers = [[six.b(str(i)),six.b(str(i)),six.b(str(i))] for i in range(1,10001)] + + file_data = self.arrays_to_csv_file_content(six.b('\t'),[six.b('aa'),six.b('bb'),six.b('cc')],numbers) + + tmpfile = self.create_file_with_data(file_data) + tmpfile_folder = os.path.dirname(tmpfile.name) + tmpfile_filename = os.path.basename(tmpfile.name) + expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -t "select sum(aa),sum(bb),sum(cc) from %s" -H -C readwrite' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('50005000\t50005000\t50005000')) + + self.assertTrue(os.path.exists(expected_cache_filename)) + + self.cleanup(tmpfile) + + # Get the data using a comma delimiter, to make sure that column parsing was done correctlyAdding to qcatalog table: + cmd = Q_EXECUTABLE + ' -D , "select count(*),sum(aa),sum(bb),sum(cc) from %s"' % expected_cache_filename + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('10000,50005000,50005000,50005000')) + + def test_analysis_of_qsql_direct_usage(self): + numbers = [[six.b(str(i)),six.b(str(i)),six.b(str(i))] for i in range(1,10001)] + + file_data = self.arrays_to_csv_file_content(six.b('\t'),[six.b('aa'),six.b('bb'),six.b('cc')],numbers) + + tmpfile = self.create_file_with_data(file_data) + tmpfile_folder = os.path.dirname(tmpfile.name) + tmpfile_filename = os.path.basename(tmpfile.name) + expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -t "select sum(aa),sum(bb),sum(cc) from %s" -H -C readwrite' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('50005000\t50005000\t50005000')) + + self.assertTrue(os.path.exists(expected_cache_filename)) + + self.cleanup(tmpfile) + + cmd = Q_EXECUTABLE + ' "select * from %s" -A' % expected_cache_filename + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 7) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('Table: %s' % expected_cache_filename)) + self.assertEqual(o[1],six.b(" Sources:")) + self.assertEqual(o[2],six.b(' source_type: qsql-file source: %s' % expected_cache_filename)) + self.assertEqual(o[3],six.b(" Fields:")) + self.assertEqual(o[4],six.b(' `aa` - int')) + self.assertEqual(o[5],six.b(' `bb` - int')) + self.assertEqual(o[6],six.b(' `cc` - int')) + + def test_analysis_of_qsql_direct_usage2(self): + numbers = [[six.b(str(i)),six.b(str(i)),six.b(str(i))] for i in range(1,10001)] + + file_data = self.arrays_to_csv_file_content(six.b('\t'),[six.b('aa'),six.b('bb'),six.b('cc')],numbers) + + tmpfile = self.create_file_with_data(file_data) + tmpfile_folder = os.path.dirname(tmpfile.name) + tmpfile_filename = os.path.basename(tmpfile.name) + expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -t "select sum(aa),sum(bb),sum(cc) from %s" -H -C readwrite' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('50005000\t50005000\t50005000')) + + self.assertTrue(os.path.exists(expected_cache_filename)) + + self.cleanup(tmpfile) + + cmd = Q_EXECUTABLE + ' "select * from %s" -A' % expected_cache_filename + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 7) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('Table: %s' % expected_cache_filename)) + self.assertEqual(o[1],six.b(" Sources:")) + self.assertEqual(o[2],six.b(' source_type: qsql-file source: %s' % expected_cache_filename)) + self.assertEqual(o[3],six.b(" Fields:")) + self.assertEqual(o[4],six.b(' `aa` - int')) + self.assertEqual(o[5],six.b(' `bb` - int')) + self.assertEqual(o[6],six.b(' `cc` - int')) + + def test_direct_qsql_usage_for_single_table_qsql_file(self): + disk_db_filename = self.random_tmp_filename('save-to-db','qsql') + + cmd = 'seq 1 10000 | %s -t "select sum(aa),sum(bb),sum(cc) from -" -S %s' % (Q_EXECUTABLE,disk_db_filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + + cmd = '%s -D, "select count(*),sum(c1) from %s:::data_stream_stdin"' % (Q_EXECUTABLE,disk_db_filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('10000,50005000')) + + def test_direct_qsql_usage_for_single_table_qsql_file__nonexistent_table(self): + disk_db_filename = self.random_tmp_filename('save-to-db','qsql') + + cmd = 'seq 1 10000 | %s -t "select sum(aa),sum(bb),sum(cc) from -" -S %s' % (Q_EXECUTABLE,disk_db_filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + + cmd = '%s -D, "select count(*),sum(c1) from %s:::unknown_table_name"' % (Q_EXECUTABLE,disk_db_filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 85) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + self.assertEqual(e[0],six.b('Table unknown_table_name could not be found in sqlite file %s . Existing table names: data_stream_stdin' % (disk_db_filename))) + + def test_direct_qsql_usage_from_written_data_stream(self): + disk_db_filename = self.random_tmp_filename('save-to-db','qsql') + + cmd = 'seq 1 10000 | %s -t "select sum(aa),sum(bb),sum(cc) from -" -S %s' % (Q_EXECUTABLE,disk_db_filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + + cmd = '%s -D, "select count(*),sum(c1) from %s:::data_stream_stdin"' % (Q_EXECUTABLE,disk_db_filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('10000,50005000')) + + def test_direct_qsql_self_join(self): + disk_db_filename = self.random_tmp_filename('save-to-db','qsql') + + N = 100 + cmd = 'seq 1 %s | %s -t "select count(*),sum(c1) from -" -S %s' % (N,Q_EXECUTABLE,disk_db_filename) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + + cmd = '%s -D, "select count(*),sum(a.c1),sum(b.c1) from %s:::data_stream_stdin a left join %s:::data_stream_stdin b"' % (Q_EXECUTABLE,disk_db_filename,disk_db_filename) + retcode, o, e = run_command(cmd) + + expected_sum = sum(range(1,N+1))*N + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + self.assertEqual(o[0],six.b('10000,%s,%s' % (expected_sum,expected_sum))) + + +class CachingTests(AbstractQTestCase): + + def test_cache_empty_file(self): + file_data = six.b("a,b,c") + tmpfile = self.create_file_with_data(file_data) + tmpfile_folder = os.path.dirname(tmpfile.name) + tmpfile_filename = os.path.basename(tmpfile.name) + tmpfile_expected_table_name = os.path.basename(tmpfile.name) + expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C none' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0],six.b("Warning - data is empty")) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C readwrite' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0],six.b("Warning - data is empty")) + + # After readwrite caching has been activated, the cache file is expected to exist + self.assertTrue(os.path.exists(expected_cache_filename)) + + # Read the cache file directly, to make sure it's a valid sqlite file + import sqlite3 + db = sqlite3.connect(expected_cache_filename) + table_list = db.execute("select content_signature_key,temp_table_name,content_signature,creation_time,source_type,source from _qcatalog where temp_table_name == '%s'" % (tmpfile_expected_table_name)).fetchall() + self.assertTrue(len(table_list) == 1) + table_metadata = table_list[0] + results = db.execute("select * from %s" % table_metadata[1]).fetchall() + self.assertTrue(len(results) == 0) + + self.cleanup(tmpfile) + + def test_reading_the_wrong_cache__original_file_having_different_data(self): + file_data1 = six.b("a,b,c\n10,20,30\n30,40,50") + + tmpfile1 = self.create_file_with_data(file_data1) + tmpfile1_folder = os.path.dirname(tmpfile1.name) + tmpfile1_filename = os.path.basename(tmpfile1.name) + expected_cache_filename = os.path.join(tmpfile1_folder,tmpfile1_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0], six.b('10')) + self.assertEqual(o[1], six.b('30')) + + # Ensure cache has been created + self.assertTrue(os.path.exists(expected_cache_filename)) + + # Overwrite the original file + file_data2 = six.b("a,b,c\n10,20,30\n30,40,50\n50,60,70") + self.write_file(tmpfile1.name,file_data2) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C read' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 81) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + self.assertEqual(e[0], six.b('%s vs %s.qsql: Content Signatures differ at inferer.rows (actual analysis data differs)' % \ + (tmpfile1.name,tmpfile1.name))) + + + def test_reading_the_wrong_cache__original_file_having_different_delimiter(self): + file_data1 = six.b("a,b,c\n10,20,30\n30,40,50") + + tmpfile1 = self.create_file_with_data(file_data1) + tmpfile1_folder = os.path.dirname(tmpfile1.name) + tmpfile1_filename = os.path.basename(tmpfile1.name) + expected_cache_filename = os.path.join(tmpfile1_folder,tmpfile1_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0], six.b('10')) + self.assertEqual(o[1], six.b('30')) + + # Ensure cache has been created + self.assertTrue(os.path.exists(expected_cache_filename)) + + # Overwrite the original file + file_data2 = six.b("a\tb\tc\n10\t20\t30\n30\t40\t50") + self.write_file(tmpfile1.name,file_data2) + + cmd = Q_EXECUTABLE + ' -H -t "select a from %s" -C read' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 80) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + x = six.b("%s vs %s.qsql: Content Signatures for table %s differ at input_delimiter (source value '\t' disk signature value ',')" % \ + (tmpfile1.name,tmpfile1.name,tmpfile1.name)) + self.assertEqual(e[0], x) + + def test_rename_cache_and_read_from_it(self): + # create a file, along with its qsql + file_data1 = six.b("a,b,c\n10,20,30\n30,40,50") + + tmpfile1 = self.create_file_with_data(file_data1) + tmpfile1_folder = os.path.dirname(tmpfile1.name) + tmpfile1_filename = os.path.basename(tmpfile1.name) + expected_cache_filename1 = os.path.join(tmpfile1_folder,tmpfile1_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0], six.b('10')) + self.assertEqual(o[1], six.b('30')) + # Ensure cache has been created + self.assertTrue(os.path.exists(expected_cache_filename1)) + + tmp_fn = self.generate_tmpfile_name("aa","qsql") + os.rename(expected_cache_filename1,tmp_fn) + + cmd = '%s "select a from %s"' % (Q_EXECUTABLE,tmp_fn) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0], six.b('10')) + self.assertEqual(o[1], six.b('30')) + + + def test_reading_the_wrong_cache__qsql_file_not_having_a_matching_content_signature(self): + # create a file, along with its qsql + file_data1 = six.b("a,b,c\n10,20,30\n30,40,50") + + tmpfile1 = self.create_file_with_data(file_data1) + tmpfile1_folder = os.path.dirname(tmpfile1.name) + tmpfile1_filename = os.path.basename(tmpfile1.name) + expected_cache_filename1 = os.path.join(tmpfile1_folder,tmpfile1_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0], six.b('10')) + self.assertEqual(o[1], six.b('30')) + # Ensure cache has been created + self.assertTrue(os.path.exists(expected_cache_filename1)) + + file_data2 = six.b("c,d,e\n10,20,30\n30,40,50") + + # create another file with a different header, along with its qsql + tmpfile2 = self.create_file_with_data(file_data2) + tmpfile2_folder = os.path.dirname(tmpfile2.name) + tmpfile2_filename = os.path.basename(tmpfile2.name) + expected_cache_filename2 = os.path.join(tmpfile2_folder,tmpfile2_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -d , "select c from %s" -C readwrite' % tmpfile2.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0], six.b('10')) + self.assertEqual(o[1], six.b('30')) + # Ensure cache has been created + self.assertTrue(os.path.exists(expected_cache_filename2)) + + # now take the second qsql file as if it was the first. Execution on file 1 should fail, since the qsql file + # does not really contain the table we're after + + os.remove(expected_cache_filename1) + os.rename(expected_cache_filename2,expected_cache_filename1) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C read' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 80) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + x = six.b("%s vs %s.qsql: Content Signatures for table %s differ at inferer.header_row (source value '['a', 'b', 'c']' disk signature value '['c', 'd', 'e']')" % (tmpfile1.name,tmpfile1.name,tmpfile1.name)) + self.assertEqual(e[0], x) + + def test_reading_the_wrong_cache__qsql_file_not_having_any_content_signature(self): + # create a file, along with its qsql + file_data1 = six.b("a,b,c\n10,20,30\n30,40,50") + + tmpfile1 = self.create_file_with_data(file_data1) + tmpfile1_folder = os.path.dirname(tmpfile1.name) + tmpfile1_filename = os.path.basename(tmpfile1.name) + expected_cache_filename1 = os.path.join(tmpfile1_folder,tmpfile1_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0], six.b('10')) + self.assertEqual(o[1], six.b('30')) + # Ensure cache has been created + self.assertTrue(os.path.exists(expected_cache_filename1)) + + file_data2 = six.b("c,d,e\n10,20,30\n30,40,50") + + # delete qcatalog content, so no entries will be available + c = sqlite3.connect(expected_cache_filename1) + c.execute('delete from _qcatalog').fetchall() + c.commit() + c.close() + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C read' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 97) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + self.assertEqual(e[0],six.b("Could not autodetect table name in qsql file. File contains no record of a table")) + + + def test_cache_full_flow(self): + file_data = six.b("a,b,c\n10,20,30\n30,40,50") + tmpfile = self.create_file_with_data(file_data) + tmpfile_folder = os.path.dirname(tmpfile.name) + tmpfile_filename = os.path.basename(tmpfile.name) + expected_tmpfile_table_name = tmpfile_filename + expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C none' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0],six.b('10')) + self.assertEqual(o[1],six.b('30')) + + # Ensure cache has not been created + self.assertTrue(not os.path.exists(expected_cache_filename)) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C read' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0],six.b('10')) + self.assertEqual(o[1],six.b('30')) + + # Ensure cache has not been created, as cache mode is "read" only + self.assertTrue(not os.path.exists(expected_cache_filename)) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C readwrite' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0],six.b('10')) + self.assertEqual(o[1],six.b('30')) + + # After readwrite caching has been activated, the cache file is expected to exist + self.assertTrue(os.path.exists(expected_cache_filename)) + + # Read the cache file directly, to make sure it's a valid sqlite file + db = sqlite3.connect(expected_cache_filename) + table_list = db.execute("select content_signature_key,temp_table_name,content_signature,creation_time,source_type,source from _qcatalog where temp_table_name == '%s'" % expected_tmpfile_table_name).fetchall() + self.assertTrue(len(table_list) == 1) + table_metadata = table_list[0] + results = db.execute("select * from %s" % table_metadata[1]).fetchall() + self.assertEqual(results[0],(10,20,30)) + self.assertEqual(results[1],(30,40,50)) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C read' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0],six.b('10')) + self.assertEqual(o[1],six.b('30')) + + # After readwrite caching has been activated, the cache file is expected to exist + self.assertTrue(os.path.exists(expected_cache_filename)) + + self.cleanup(tmpfile) + + def test_cache_full_flow_with_concatenated_files(self): + file_data1 = six.b("a,b,c\n10,11,12\n20,21,22") + tmpfile1 = self.create_file_with_data(file_data1) + tmpfile1_folder = os.path.dirname(tmpfile1.name) + tmpfile1_filename = os.path.basename(tmpfile1.name) + expected_cache_filename1 = os.path.join(tmpfile1_folder,tmpfile1_filename + '.qsql') + + file_data2 = six.b("a,b,c\n30,31,32\n40,41,42") + tmpfile2 = self.create_file_with_data(file_data2) + tmpfile2_folder = os.path.dirname(tmpfile2.name) + tmpfile2_filename = os.path.basename(tmpfile2.name) + expected_cache_filename2 = os.path.join(tmpfile2_folder,tmpfile2_filename + '.qsql') + + cmd = Q_EXECUTABLE + ' -O -H -d , "select * from (select * from %s UNION ALL select * from %s)" -C readwrite' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 5) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('a,b,c')) + self.assertEqual(o[1],six.b('10,11,12')) + self.assertEqual(o[2],six.b('20,21,22')) + self.assertEqual(o[3],six.b('30,31,32')) + self.assertEqual(o[4],six.b('40,41,42')) + + self.assertTrue(os.path.exists(expected_cache_filename1)) + self.assertTrue(os.path.exists(expected_cache_filename2)) + + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + + def test_analyze_result_with_cache_file(self): + file_data = six.b("a,b,c\n10,20,30\n30,40,50") + tmpfile = self.create_file_with_data(file_data) + tmpfile_folder = os.path.dirname(tmpfile.name) + tmpfile_filename = os.path.basename(tmpfile.name) + expected_cache_filename = os.path.join(tmpfile_folder,tmpfile_filename + '.qsql') + + # Ensure cache has not been created yet + self.assertTrue(not os.path.exists(expected_cache_filename)) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C readwrite' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + self.assertTrue(o[0],six.b('10')) + self.assertEqual(o[1],six.b('30')) + + # Ensure cache is now created + self.assertTrue(os.path.exists(expected_cache_filename)) + + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C read -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),7) + self.assertEqual(len(e),0) + + self.assertEqual(o[0],six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: qsql-file-with-original source: %s.qsql' % tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4],six.b(' `a` - int')) + self.assertEqual(o[5],six.b(' `b` - int')) + self.assertEqual(o[6],six.b(' `c` - int')) + + # delete the newly created cache + os.remove(expected_cache_filename) + + # Now rerun the analysis without the cache file + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C read -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o),7) + self.assertEqual(len(e),0) + + self.assertEqual(o[0],six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s' % tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4],six.b(' `a` - int')) + self.assertEqual(o[5],six.b(' `b` - int')) + self.assertEqual(o[6],six.b(' `c` - int')) + + self.cleanup(tmpfile) + + def test_partial_caching_exists(self): + file1_data = six.b("a,b,c\n10,20,30\n30,40,50\n60,70,80") + tmpfile1 = self.create_file_with_data(file1_data) + tmpfile1_folder = os.path.dirname(tmpfile1.name) + tmpfile1_filename = os.path.basename(tmpfile1.name) + expected_cache_filename1 = os.path.join(tmpfile1_folder,tmpfile1_filename + '.qsql') + + file2_data = six.b("b,x\n10,linewith10\n20,linewith20\n30,linewith30\n40,linewith40") + tmpfile2 = self.create_file_with_data(file2_data) + tmpfile2_folder = os.path.dirname(tmpfile2.name) + tmpfile2_filename = os.path.basename(tmpfile2.name) + expected_cache_filename2 = os.path.join(tmpfile2_folder,tmpfile2_filename + '.qsql') + + # Use only first file, and cache + cmd = Q_EXECUTABLE + ' -H -d , "select a from %s" -C readwrite' % tmpfile1.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + self.assertTrue(o[0],six.b('10')) + self.assertEqual(o[1],six.b('30')) + + # Ensure cache has been created for file 1 + self.assertTrue(os.path.exists(expected_cache_filename1)) + + # Use both files with read caching, one should be read from cache, the other from the file + cmd = Q_EXECUTABLE + ' -H -d , "select file1.a,file1.b,file1.c,file2.x from %s file1 left join %s file2 on (file1.b = file2.b)" -C read' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('10,20,30,linewith20')) + self.assertEqual(o[1],six.b('30,40,50,linewith40')) + self.assertEqual(o[2],six.b('60,70,80,')) + + # Ensure cache has NOT been created for file 2 + self.assertTrue(not os.path.exists(expected_cache_filename2)) + + # Now rerun the query, this time with readwrite caching, so the second file cache will be written + cmd = Q_EXECUTABLE + ' -H -d , "select file1.a,file1.b,file1.c,file2.x from %s file1 left join %s file2 on (file1.b = file2.b)" -C readwrite' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + self.assertEqual(o[0],six.b('10,20,30,linewith20')) + self.assertEqual(o[1],six.b('30,40,50,linewith40')) + self.assertEqual(o[2],six.b('60,70,80,')) + + # Ensure cache has now been created for file 2 + self.assertTrue(os.path.exists(expected_cache_filename2)) + + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + +class UserFunctionTests(AbstractQTestCase): + def test_regexp_int_data_handling(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + + cmd = Q_EXECUTABLE + ' -d , "select c2 from %s where regexp(\'^1\',c2)"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b("1")) + + self.cleanup(tmpfile) + + def test_percentile_func(self): + cmd = 'seq 1000 1999 | %s "select substr(c1,0,3),percentile(c1,0),percentile(c1,0.5),percentile(c1,1) from - group by substr(c1,0,3)" -c 1' % Q_EXECUTABLE + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 10) + self.assertEqual(len(e), 0) + + output_table = [l.split(six.b(" ")) for l in o] + group_labels = [int(row[0]) for row in output_table] + minimum_values = [float(row[1]) for row in output_table] + median_values = [float(row[2]) for row in output_table] + max_values = [float(row[3]) for row in output_table] + + base_values = list(range(1000,2000,100)) + + self.assertEqual(group_labels,list(range(10,20))) + self.assertEqual(minimum_values,base_values) + self.assertEqual(median_values,list(map(lambda x: x + 49.5,base_values))) + self.assertEqual(max_values,list(map(lambda x: x + 99,base_values))) + + def test_regexp_null_data_handling(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + + cmd = Q_EXECUTABLE + ' -d , "select count(*) from %s where regexp(\'^\',c2)"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b("2")) + + self.cleanup(tmpfile) + + def test_md5_function(self): + cmd = 'seq 1 4 | %s -c 1 -d , "select c1,md5(c1,\'utf-8\') from -"' % Q_EXECUTABLE + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),4) + self.assertEqual(len(e),0) + + self.assertEqual(tuple(o[0].split(six.b(','),1)),(six.b('1'),six.b('c4ca4238a0b923820dcc509a6f75849b'))) + self.assertEqual(tuple(o[1].split(six.b(','),1)),(six.b('2'),six.b('c81e728d9d4c2f636f067f89cc14862c'))) + self.assertEqual(tuple(o[2].split(six.b(','),1)),(six.b('3'),six.b('eccbc87e4b5ce2fe28308fd9f2a7baf3'))) + self.assertEqual(tuple(o[3].split(six.b(','),1)),(six.b('4'),six.b('a87ff679a2f3e71d9181a67b7542122c'))) + + def test_stddev_functions(self): + tmpfile = self.create_file_with_data(six.b("\n".join(map(str,[234,354,3234,123,4234,234,634,56,65])))) + + cmd = '%s -c 1 -d , "select round(stddev_pop(c1),10),round(stddev_sample(c1),10) from %s"' % (Q_EXECUTABLE,tmpfile.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + + self.assertEqual(o[0],six.b('1479.7015464838,1569.4604964764')) + + self.cleanup(tmpfile) + + def test_sqrt_function(self): + cmd = 'seq 1 5 | %s -c 1 -d , "select round(sqrt(c1),10) from -"' % Q_EXECUTABLE + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),5) + self.assertEqual(len(e),0) + + self.assertEqual(o[0],six.b('1.0')) + self.assertEqual(o[1],six.b('1.4142135624')) + self.assertEqual(o[2],six.b('1.7320508076')) + self.assertEqual(o[3],six.b('2.0')) + self.assertEqual(o[4],six.b('2.2360679775')) + + def test_power_function(self): + cmd = 'seq 1 5 | %s -c 1 -d , "select round(power(c1,2.5),10) from -"' % Q_EXECUTABLE + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),5) + self.assertEqual(len(e),0) + + self.assertEqual(o[0],six.b('1.0')) + self.assertEqual(o[1],six.b('5.6568542495')) + self.assertEqual(o[2],six.b('15.5884572681')) + self.assertEqual(o[3],six.b('32.0')) + self.assertEqual(o[4],six.b('55.9016994375')) + + def test_file_functions(self): + filenames = [ + "file1", + "file2.csv", + "/var/tmp/file3", + "/var/tmp/file4.gz", + "" + ] + data = "\n".join(filenames) + + cmd = 'echo "%s" | %s -c 1 -d , "select file_folder(c1),file_ext(c1),file_basename(c1),file_basename_no_ext(c1) from -"' % (data,Q_EXECUTABLE) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),5) + self.assertEqual(len(e),0) + self.assertEqual(o,[ + b',,file1,file1', + b',.csv,file2.csv,file2', + b'/var/tmp,,file3,file3', + b'/var/tmp,.gz,file4.gz,file4', + b',,,' + ]) + + + def test_sha1_function(self): + cmd = 'seq 1 4 | %s -c 1 -d , "select c1,sha1(c1) from -"' % Q_EXECUTABLE + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),4) + self.assertEqual(len(e),0) + + self.assertEqual(o[0],six.b('1,356a192b7913b04c54574d18c28d46e6395428ab')) + self.assertEqual(o[1],six.b('2,da4b9237bacccdf19c0760cab7aec4a8359010b0')) + self.assertEqual(o[2],six.b('3,77de68daecd823babbb58edb1c8e14d7106e83bb')) + self.assertEqual(o[3],six.b('4,1b6453892473a467d07372d45eb05abc2031647a')) + + def test_regexp_extract_function(self): + query = """ + select + regexp_extract('was ([0-9]+) seconds and ([0-9]+) ms',c1,0), + regexp_extract('was ([0-9]+) seconds and ([0-9]+) ms',c1,1), + regexp_extract('non-existent-(regexp)',c1,0) + from + - + """ + + cmd = 'echo "Duration was 322 seconds and 240 ms" | %s -c 1 -d , "%s"' % (Q_EXECUTABLE,query) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),1) + self.assertEqual(len(e),0) + + self.assertEqual(o[0],six.b('322,240,')) + + def test_sha_function(self): + cmd = 'seq 1 4 | %s -c 1 -d , "select c1,sha(c1,1,\'utf-8\') as sha1,sha(c1,224,\'utf-8\') as sha224,sha(c1,256,\'utf-8\') as sha256 from -"' % Q_EXECUTABLE + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),4) + self.assertEqual(len(e),0) + + self.assertEqual(o[0],six.b('1,356a192b7913b04c54574d18c28d46e6395428ab,e25388fde8290dc286a6164fa2d97e551b53498dcbf7bc378eb1f178,6b86b273ff34fce19d6b804eff5a3f5747ada4eaa22f1d49c01e52ddb7875b4b')) + self.assertEqual(o[1],six.b('2,da4b9237bacccdf19c0760cab7aec4a8359010b0,58b2aaa0bfae7acc021b3260e941117b529b2e69de878fd7d45c61a9,d4735e3a265e16eee03f59718b9b5d03019c07d8b6c51f90da3a666eec13ab35')) + self.assertEqual(o[2],six.b('3,77de68daecd823babbb58edb1c8e14d7106e83bb,4cfc3a1811fe40afa401b25ef7fa0379f1f7c1930a04f8755d678474,4e07408562bedb8b60ce05c1decfe3ad16b72230967de01f640b7e4729b49fce')) + self.assertEqual(o[3],six.b('4,1b6453892473a467d07372d45eb05abc2031647a,271f93f45e9b4067327ed5c8cd30a034730aaace4382803c3e1d6c2f,4b227777d4dd1fc61c6f884f48641d02b4d121d3fd328cb08b5531fcacdabf8a')) + + +class MultiHeaderTests(AbstractQTestCase): + def test_output_header_when_multiple_input_headers_exist(self): + TMPFILE_COUNT = 5 + tmpfiles = [self.create_file_with_data(sample_data_with_header) for x in range(TMPFILE_COUNT)] + + tmpfilenames = " UNION ALL ".join(map(lambda x:"select * from %s" % x.name, tmpfiles)) + + cmd = Q_EXECUTABLE + ' -d , "select name,value1,value2 from (%s) order by name" -H -O' % tmpfilenames + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), TMPFILE_COUNT*3+1) + self.assertEqual(o[0], six.b("name,value1,value2")) + + for i in range (TMPFILE_COUNT): + self.assertEqual(o[1+i],sample_data_rows[0]) + for i in range (TMPFILE_COUNT): + self.assertEqual(o[TMPFILE_COUNT+1+i],sample_data_rows[1]) + for i in range (TMPFILE_COUNT): + self.assertEqual(o[TMPFILE_COUNT*2+1+i],sample_data_rows[2]) + + for oi in o[1:]: + self.assertTrue(six.b('name') not in oi) + + for i in range(TMPFILE_COUNT): + self.cleanup(tmpfiles[i]) + + def test_output_header_when_extra_header_column_names_are_different__concatenation_replacement(self): + tmpfile1 = self.create_file_with_data(sample_data_with_header) + tmpfile2 = self.create_file_with_data(generate_sample_data_with_header(six.b('othername,value1,value2'))) + + cmd = Q_EXECUTABLE + ' -d , "select name,value1,value2 from (select * from %s union all select * from %s) order by name" -H -O' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 7) + self.assertEqual(len(e), 0) + self.assertTrue(o, [ + six.b('name,value1,value2'), + six.b('a,1,0'), + six.b('a,1,0'), + six.b('b,2,0'), + six.b('b,2,0'), + six.b('c,,0'), + six.b('c,,0') + ]) + + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + def test_output_header_when_extra_header_has_different_number_of_columns(self): + tmpfile1 = self.create_file_with_data(sample_data_with_header) + tmpfile2 = self.create_file_with_data(generate_sample_data_with_header(six.b('name,value1'))) + + cmd = Q_EXECUTABLE + ' -d , "select name,value1,value2 from (select * from %s UNION ALL select * from %s) order by name" -H -O' % (tmpfile1.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 7) + self.assertEqual(len(e), 0) + self.assertTrue(o, [ + six.b('name,value1,value2'), + six.b('a,1,0'), + six.b('a,1,0'), + six.b('b,2,0'), + six.b('b,2,0'), + six.b('c,,0'), + six.b('c,,0') + ]) + + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + +class ParsingModeTests(AbstractQTestCase): + + def test_strict_mode_column_count_mismatch_error(self): + tmpfile = self.create_file_with_data(uneven_ls_output) + cmd = Q_EXECUTABLE + ' -m strict "select count(*) from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + self.assertTrue(six.b("Column Count is expected to identical") in e[0]) + + self.cleanup(tmpfile) + + def test_strict_mode_too_large_specific_column_count(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , -m strict -c 4 "select count(*) from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + self.assertEqual( + e[0], six.b("Strict mode. Column count is expected to be 4 but is 3")) + + self.cleanup(tmpfile) + + def test_strict_mode_too_small_specific_column_count(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , -m strict -c 2 "select count(*) from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + self.assertEqual( + e[0], six.b("Strict mode. Column count is expected to be 2 but is 3")) + + self.cleanup(tmpfile) + + def test_relaxed_mode_missing_columns_in_header(self): + tmpfile = self.create_file_with_data( + sample_data_with_missing_header_names) + cmd = Q_EXECUTABLE + ' -d , -m relaxed "select count(*) from %s" -H -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 7) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0],six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s') % six.b(tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4],six.b(' `name` - text')) + self.assertEqual(o[5],six.b(' `value1` - int')) + self.assertEqual(o[6],six.b(' `c3` - int')) + + self.cleanup(tmpfile) + + def test_strict_mode_missing_columns_in_header(self): + tmpfile = self.create_file_with_data( + sample_data_with_missing_header_names) + cmd = Q_EXECUTABLE + ' -d , -m strict "select count(*) from %s" -H -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode, 0) + self.assertEqual(len(o), 0) + self.assertEqual(len(e), 1) + + self.assertEqual( + e[0], six.b('Strict mode. Header row contains less columns than expected column count(2 vs 3)')) + + self.cleanup(tmpfile) + + def test_output_delimiter_with_missing_fields(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , "select * from %s" -D ";"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('a;1;0')) + self.assertEqual(o[1], six.b('b;2;0')) + self.assertEqual(o[2], six.b('c;;0')) + + self.cleanup(tmpfile) + + def test_handling_of_null_integers(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , "select avg(c2) from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('1.5')) + + self.cleanup(tmpfile) + + def test_empty_integer_values_converted_to_null(self): + tmpfile = self.create_file_with_data(sample_data_no_header) + cmd = Q_EXECUTABLE + ' -d , "select * from %s where c2 is null"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('c,,0')) + + self.cleanup(tmpfile) + + def test_empty_string_values_not_converted_to_null(self): + tmpfile = self.create_file_with_data( + sample_data_with_empty_string_no_header) + cmd = Q_EXECUTABLE + ' -d , "select * from %s where c2 == %s"' % ( + tmpfile.name, "''") + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('c,,0')) + + self.cleanup(tmpfile) + + def test_relaxed_mode_detected_columns(self): + tmpfile = self.create_file_with_data(uneven_ls_output) + cmd = Q_EXECUTABLE + ' -m relaxed "select count(*) from %s" -A' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + + column_rows = o[o.index(six.b(' Fields:'))+1:] + + self.assertEqual(len(column_rows), 11) + + column_tuples = [x.strip().split(six.b(" ")) for x in column_rows] + column_info = [(x[0], x[2]) for x in column_tuples] + column_names = [x[0] for x in column_tuples] + column_types = [x[2] for x in column_tuples] + + self.assertEqual(column_names, [six.b('`c{}`'.format(x)) for x in range(1, 12)]) + self.assertEqual(column_types, list(map(lambda x:six.b(x),[ + 'text', 'int', 'text', 'text', 'int', 'text', 'int', 'int', 'text', 'text', 'text']))) + + self.cleanup(tmpfile) + + def test_relaxed_mode_detected_columns_with_specific_column_count(self): + tmpfile = self.create_file_with_data(uneven_ls_output) + cmd = Q_EXECUTABLE + ' -m relaxed "select count(*) from %s" -A -c 9' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + + column_rows = o[o.index(six.b(' Fields:'))+1:] + + self.assertEqual(len(column_rows), 9) + + column_tuples = [x.strip().split(six.b(" ")) for x in column_rows] + column_info = [(x[0], x[2]) for x in column_tuples] + column_names = [x[0] for x in column_tuples] + column_types = [x[2] for x in column_tuples] + + self.assertEqual(column_names, [six.b('`c{}`'.format(x)) for x in range(1, 10)]) + self.assertEqual( + column_types, list(map(lambda x:six.b(x),['text', 'int', 'text', 'text', 'int', 'text', 'int', 'int', 'text']))) + + self.cleanup(tmpfile) + + def test_relaxed_mode_last_column_data_with_specific_column_count(self): + tmpfile = self.create_file_with_data(uneven_ls_output) + cmd = Q_EXECUTABLE + ' -m relaxed "select c9 from %s" -c 9' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 9) + self.assertEqual(len(e), 0) + + expected_output = list(map(lambda x:six.b(x),["/selinux", "/mnt", "/srv", "/lost+found", '"/initrd.img.old -> /boot/initrd.img-3.8.0-19-generic"', + "/cdrom", "/home", '"/vmlinuz -> boot/vmlinuz-3.8.0-19-generic"', '"/initrd.img -> boot/initrd.img-3.8.0-19-generic"'])) + + self.assertEqual(o, expected_output) + + self.cleanup(tmpfile) + + def test_1_column_warning_in_relaxed_mode(self): + tmpfile = self.create_file_with_data(one_column_data) + cmd = Q_EXECUTABLE + ' -m relaxed "select c1 from %s" -d ,' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('data without commas 1')) + self.assertEqual(o[1],six.b('data without commas 2')) + + self.cleanup(tmpfile) + + def test_1_column_warning_in_strict_mode(self): + tmpfile = self.create_file_with_data(one_column_data) + cmd = Q_EXECUTABLE + ' -m relaxed "select c1 from %s" -d , -m strict' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('data without commas 1')) + self.assertEqual(o[1],six.b('data without commas 2')) + + self.cleanup(tmpfile) + + + def test_1_column_warning_suppression_in_relaxed_mode_when_column_count_is_specific(self): + tmpfile = self.create_file_with_data(one_column_data) + cmd = Q_EXECUTABLE + ' -m relaxed "select c1 from %s" -d , -m relaxed -c 1' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('data without commas 1')) + self.assertEqual(o[1],six.b('data without commas 2')) + + self.cleanup(tmpfile) + + def test_1_column_warning_suppression_in_strict_mode_when_column_count_is_specific(self): + tmpfile = self.create_file_with_data(one_column_data) + cmd = Q_EXECUTABLE + ' -m relaxed "select c1 from %s" -d , -m strict -c 1' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o),2) + + self.assertEqual(o[0],six.b('data without commas 1')) + self.assertEqual(o[1],six.b('data without commas 2')) + + self.cleanup(tmpfile) + + def test_fluffy_mode__as_relaxed_mode(self): + tmpfile = self.create_file_with_data(uneven_ls_output) + cmd = Q_EXECUTABLE + ' -m relaxed "select c9 from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 9) + self.assertEqual(len(e), 0) + + expected_output = list(map(lambda x:six.b(x),["/selinux", "/mnt", "/srv", "/lost+found", + "/initrd.img.old", "/cdrom", "/home", "/vmlinuz", "/initrd.img"])) + + self.assertEqual(o, expected_output) + + self.cleanup(tmpfile) + + def test_relaxed_mode_column_count_mismatch__was_previously_fluffy_mode_test(self): + data_row = six.b("column1 column2 column3 column4") + data_list = [data_row] * 1000 + data_list[950] = six.b("column1 column2 column3 column4 column5") + tmpfile = self.create_file_with_data(six.b("\n").join(data_list)) + + cmd = Q_EXECUTABLE + ' -m relaxed "select * from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + self.assertEqual(len(o),1000) + self.assertEqual(len(e),0) + self.assertEqual(o[950],six.b('column1 column2 column3 "column4 column5"')) + + self.cleanup(tmpfile) + + def test_strict_mode_column_count_mismatch__less_columns(self): + data_row = six.b("column1 column2 column3 column4") + data_list = [data_row] * 1000 + data_list[750] = six.b("column1 column3 column4") + tmpfile = self.create_file_with_data(six.b("\n").join(data_list)) + + cmd = Q_EXECUTABLE + ' -m strict "select * from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode,0) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + self.assertTrue(e[0].startswith(six.b("Strict mode - Expected 4 columns instead of 3 columns"))) + self.assertTrue(six.b(' row 751.') in e[0]) + + self.cleanup(tmpfile) + + def test_strict_mode_column_count_mismatch__more_columns(self): + data_row = six.b("column1 column2 column3 column4") + data_list = [data_row] * 1000 + data_list[750] = six.b("column1 column2 column3 column4 column5") + tmpfile = self.create_file_with_data(six.b("\n").join(data_list)) + + cmd = Q_EXECUTABLE + ' -m strict "select * from %s"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertNotEqual(retcode,0) + self.assertEqual(len(o),0) + self.assertEqual(len(e),1) + self.assertTrue(e[0].startswith(six.b("Strict mode - Expected 4 columns instead of 5 columns"))) + self.assertTrue(six.b(' row 751.') in e[0]) + + self.cleanup(tmpfile) + + +class FormattingTests(AbstractQTestCase): + + def test_column_formatting(self): + # TODO Decide if this breaking change is reasonable + #cmd = 'seq 1 10 | ' + Q_EXECUTABLE + ' -f 1=%4.3f,2=%4.3f "select sum(c1),avg(c1) from -" -c 1' + cmd = 'seq 1 10 | ' + Q_EXECUTABLE + ' -f 1={:4.3f},2={:4.3f} "select sum(c1),avg(c1) from -" -c 1' + + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 1) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('55.000 5.500')) + + def test_column_formatting_with_output_header(self): + perl_regex = "'s/1\n/column_name\n1\n/;'" + # TODO Decide if this breaking change is reasonable + #cmd = 'seq 1 10 | perl -pe ' + perl_regex + ' | ' + Q_EXECUTABLE + ' -f 1=%4.3f,2=%4.3f "select sum(column_name) mysum,avg(column_name) myavg from -" -c 1 -H -O' + cmd = 'seq 1 10 | LANG=C perl -pe ' + perl_regex + ' | ' + Q_EXECUTABLE + ' -f 1={:4.3f},2={:4.3f} "select sum(column_name) mysum,avg(column_name) myavg from -" -c 1 -H -O' + + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('mysum myavg')) + self.assertEqual(o[1], six.b('55.000 5.500')) + + def py3_test_successfuly_parse_universal_newlines_without_explicit_flag(self): + def list_as_byte_list(l): + return list(map(lambda x:six.b(x),l)) + + expected_output = list(map(lambda x:list_as_byte_list(x),[['lifelock', 'LifeLock', '', 'web', 'Tempe', 'AZ', '1-May-07', '6850000', 'USD', 'b'], + ['lifelock', 'LifeLock', '', 'web', 'Tempe', 'AZ', '1-Oct-06', '6000000', 'USD', 'a'], + ['lifelock', 'LifeLock', '', 'web', 'Tempe', 'AZ', '1-Jan-08', '25000000', 'USD', 'c'], + ['mycityfaces', 'MyCityFaces', '7', 'web', 'Scottsdale', 'AZ', '1-Jan-08', '50000', 'USD', 'seed'], + ['flypaper', 'Flypaper', '', 'web', 'Phoenix', 'AZ', '1-Feb-08', '3000000', 'USD', 'a'], + ['infusionsoft', 'Infusionsoft', '105', 'software', 'Gilbert', 'AZ', '1-Oct-07', '9000000', 'USD', 'a']])) + + data = six.b('permalink,company,numEmps,category,city,state,fundedDate,raisedAmt,raisedCurrency,round\rlifelock,LifeLock,,web,Tempe,AZ,1-May-07,6850000,USD,b\rlifelock,LifeLock,,web,Tempe,AZ,1-Oct-06,6000000,USD,a\rlifelock,LifeLock,,web,Tempe,AZ,1-Jan-08,25000000,USD,c\rmycityfaces,MyCityFaces,7,web,Scottsdale,AZ,1-Jan-08,50000,USD,seed\rflypaper,Flypaper,,web,Phoenix,AZ,1-Feb-08,3000000,USD,a\rinfusionsoft,Infusionsoft,105,software,Gilbert,AZ,1-Oct-07,9000000,USD,a') + tmp_data_file = self.create_file_with_data(data) + + cmd = Q_EXECUTABLE + ' -d , -H "select * from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 6) + + actual_output = list(map(lambda row: row.split(six.b(",")),o)) + + self.assertEqual(actual_output,expected_output) + + self.cleanup(tmp_data_file) + + test_parsing_universal_newlines_without_explicit_flag = py3_test_successfuly_parse_universal_newlines_without_explicit_flag + + def test_universal_newlines_parsing_flag(self): + def list_as_byte_list(l): + return list(map(lambda x:six.b(x),l)) + + expected_output = list(map(lambda x:list_as_byte_list(x),[['lifelock', 'LifeLock', '', 'web', 'Tempe', 'AZ', '1-May-07', '6850000', 'USD', 'b'], + ['lifelock', 'LifeLock', '', 'web', 'Tempe', 'AZ', '1-Oct-06', '6000000', 'USD', 'a'], + ['lifelock', 'LifeLock', '', 'web', 'Tempe', 'AZ', '1-Jan-08', '25000000', 'USD', 'c'], + ['mycityfaces', 'MyCityFaces', '7', 'web', 'Scottsdale', 'AZ', '1-Jan-08', '50000', 'USD', 'seed'], + ['flypaper', 'Flypaper', '', 'web', 'Phoenix', 'AZ', '1-Feb-08', '3000000', 'USD', 'a'], + ['infusionsoft', 'Infusionsoft', '105', 'software', 'Gilbert', 'AZ', '1-Oct-07', '9000000', 'USD', 'a']])) + + data = six.b('permalink,company,numEmps,category,city,state,fundedDate,raisedAmt,raisedCurrency,round\rlifelock,LifeLock,,web,Tempe,AZ,1-May-07,6850000,USD,b\rlifelock,LifeLock,,web,Tempe,AZ,1-Oct-06,6000000,USD,a\rlifelock,LifeLock,,web,Tempe,AZ,1-Jan-08,25000000,USD,c\rmycityfaces,MyCityFaces,7,web,Scottsdale,AZ,1-Jan-08,50000,USD,seed\rflypaper,Flypaper,,web,Phoenix,AZ,1-Feb-08,3000000,USD,a\rinfusionsoft,Infusionsoft,105,software,Gilbert,AZ,1-Oct-07,9000000,USD,a') + tmp_data_file = self.create_file_with_data(data) + + cmd = Q_EXECUTABLE + ' -d , -H -U "select permalink,company,numEmps,category,city,state,fundedDate,raisedAmt,raisedCurrency,round from %s"' % tmp_data_file.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode,0) + + if len(e) == 2 or len(e) == 1: + # In python 3.7, there's a deprecation warning for the 'U' file opening mode, which is ok for now + self.assertIn(len(e), [1,2]) + self.assertTrue(b"DeprecationWarning: 'U' mode is deprecated" in e[0]) + elif len(e) != 0: + # Nothing should be output to stderr in other versions + self.assertTrue(False,msg='Unidentified output in stderr') + + self.assertEqual(len(o), 6) + + actual_output = list(map(lambda row: row.split(six.b(",")),o)) + + self.assertEqual(actual_output,expected_output) + + self.cleanup(tmp_data_file) + + + +class SqlTests(AbstractQTestCase): + + def test_find_example(self): + tmpfile = self.create_file_with_data(find_output) + cmd = Q_EXECUTABLE + ' "select c5,c6,sum(c7)/1024.0/1024 as total from %s group by c5,c6 order by total desc"' % tmpfile.name + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + self.assertEqual(len(e), 0) + + self.assertEqual(o[0], six.b('mapred mapred 0.9389581680297852')) + self.assertEqual(o[1], six.b('root root 0.02734375')) + self.assertEqual(o[2], six.b('harel harel 0.010888099670410156')) + + self.cleanup(tmpfile) + + def test_join_example(self): + cmd = Q_EXECUTABLE + ' "select myfiles.c8,emails.c2 from {0}/exampledatafile myfiles join {0}/group-emails-example emails on (myfiles.c4 = emails.c1) where myfiles.c8 = \'ppp\'"'.format(EXAMPLES) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 2) + + self.assertEqual(o[0], six.b('ppp dip.1@otherdomain.com')) + self.assertEqual(o[1], six.b('ppp dip.2@otherdomain.com')) + + def test_join_example_with_output_header(self): + cmd = Q_EXECUTABLE + ' -O "select myfiles.c8 aaa,emails.c2 bbb from {0}/exampledatafile myfiles join {0}/group-emails-example emails on (myfiles.c4 = emails.c1) where myfiles.c8 = \'ppp\'"'.format(EXAMPLES) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(o), 3) + + self.assertEqual(o[0], six.b('aaa bbb')) + self.assertEqual(o[1], six.b('ppp dip.1@otherdomain.com')) + self.assertEqual(o[2], six.b('ppp dip.2@otherdomain.com')) + + def test_self_join1(self): + tmpfile = self.create_file_with_data(six.b("\n").join([six.b("{} 9000".format(i)) for i in range(0,10)])) + cmd = Q_EXECUTABLE + ' "select * from %s a1 join %s a2 on (a1.c1 = a2.c1)"' % (tmpfile.name,tmpfile.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 10) + + self.cleanup(tmpfile) + + def test_self_join_reuses_table(self): + tmpfile = self.create_file_with_data(six.b("\n").join([six.b("{} 9000".format(i)) for i in range(0,10)])) + cmd = Q_EXECUTABLE + ' "select * from %s a1 join %s a2 on (a1.c1 = a2.c1)" -A' % (tmpfile.name,tmpfile.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 6) + + self.assertEqual(o[0],six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s') % six.b(tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4],six.b(' `c1` - int')) + self.assertEqual(o[5],six.b(' `c2` - int')) + + self.cleanup(tmpfile) + + def test_self_join2(self): + tmpfile1 = self.create_file_with_data(six.b("\n").join([six.b("{} 9000".format(i)) for i in range(0,10)])) + cmd = Q_EXECUTABLE + ' "select * from %s a1 join %s a2 on (a1.c2 = a2.c2)"' % (tmpfile1.name,tmpfile1.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 10*10) + + self.cleanup(tmpfile1) + + tmpfile2 = self.create_file_with_data(six.b("\n").join([six.b("{} 9000".format(i)) for i in range(0,10)])) + cmd = Q_EXECUTABLE + ' "select * from %s a1 join %s a2 on (a1.c2 = a2.c2) join %s a3 on (a1.c2 = a3.c2)"' % (tmpfile2.name,tmpfile2.name,tmpfile2.name) + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 10*10*10) + + self.cleanup(tmpfile2) + + def test_disable_column_type_detection(self): + tmpfile = self.create_file_with_data(six.b('''regular_text,text_with_digits1,text_with_digits2,float_number +"regular text 1",67,"67",12.3 +"regular text 2",067,"067",22.3 +"regular text 3",123,"123",33.4 +"regular text 4",-123,"-123",0122.2 +''')) + + # Check original column type detection + cmd = Q_EXECUTABLE + ' -A -d , -H "select * from %s"' % (tmpfile.name) + + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 8) + + self.assertEqual(o[0],six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1], six.b(' Sources:')) + self.assertEqual(o[2], six.b(' source_type: file source: %s') % six.b(tmpfile.name)) + self.assertEqual(o[3], six.b(' Fields:')) + self.assertEqual(o[4], six.b(' `regular_text` - text')) + self.assertEqual(o[5], six.b(' `text_with_digits1` - int')) + self.assertEqual(o[6], six.b(' `text_with_digits2` - int')) + self.assertEqual(o[7], six.b(' `float_number` - real')) + + # Check column types detected when actual detection is disabled + cmd = Q_EXECUTABLE + ' -A -d , -H --as-text "select * from %s"' % (tmpfile.name) + + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 8) + + self.assertEqual(o[0],six.b('Table: %s' % tmpfile.name)) + self.assertEqual(o[1],six.b(' Sources:')) + self.assertEqual(o[2],six.b(' source_type: file source: %s') % six.b(tmpfile.name)) + self.assertEqual(o[3],six.b(' Fields:')) + self.assertEqual(o[4],six.b(' `regular_text` - text')) + self.assertEqual(o[5],six.b(' `text_with_digits1` - text')) + self.assertEqual(o[6],six.b(' `text_with_digits2` - text')) + self.assertEqual(o[7],six.b(' `float_number` - text')) + + # Get actual data with regular detection + cmd = Q_EXECUTABLE + ' -d , -H "select * from %s"' % (tmpfile.name) + + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 4) + + self.assertEqual(o[0],six.b("regular text 1,67,67,12.3")) + self.assertEqual(o[1],six.b("regular text 2,67,67,22.3")) + self.assertEqual(o[2],six.b("regular text 3,123,123,33.4")) + self.assertEqual(o[3],six.b("regular text 4,-123,-123,122.2")) + + # Get actual data without detection + cmd = Q_EXECUTABLE + ' -d , -H --as-text "select * from %s"' % (tmpfile.name) + + retcode, o, e = run_command(cmd) + + self.assertEqual(retcode, 0) + self.assertEqual(len(e), 0) + self.assertEqual(len(o), 4) + + self.assertEqual(o[0],six.b("regular text 1,67,67,12.3")) + self.assertEqual(o[1],six.b("regular text 2,067,067,22.3")) + self.assertEqual(o[2],six.b("regular text 3,123,123,33.4")) + self.assertEqual(o[3],six.b("regular text 4,-123,-123,0122.2")) + + self.cleanup(tmpfile) + + +class BasicModuleTests(AbstractQTestCase): + + def test_engine_isolation(self): + tmpfile1 = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + tmpfile2 = self.create_file_with_data(six.b("d e f\n10 20 30\n40 50 60")) + + # Run file 1 on engine 1 + q1 = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) + r = q1.execute('select * from %s' % tmpfile1.name) + print("QueryQuery",file=sys.stdout) + + self.assertTrue(r.status == 'ok') + self.assertEqual(len(r.warnings),0) + self.assertEqual(len(r.data),2) + self.assertEqual(r.metadata.output_column_name_list,['a','b','c']) + self.assertEqual(r.data,[(1,2,3),(4,5,6)]) + self.assertTrue(tmpfile1.name in r.metadata.table_structures) + self.assertTrue(tmpfile1.name in r.metadata.new_table_structures) + self.assertEqual(r.metadata.table_structures[tmpfile1.name].atomic_fns,[tmpfile1.name]) + self.assertEqual(r.metadata.table_structures[tmpfile1.name].source_type,'file') + self.assertEqual(r.metadata.table_structures[tmpfile1.name].source,tmpfile1.name) + + # run file 1 on engine 2 + q2 = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) + r2 = q2.execute('select * from %s' % tmpfile1.name) + print("QueryQuery",file=sys.stdout) + + self.assertTrue(r2.status == 'ok') + self.assertEqual(len(r2.warnings),0) + self.assertEqual(len(r2.data),2) + self.assertEqual(r2.metadata.output_column_name_list,['a','b','c']) + self.assertEqual(r2.data,[(1,2,3),(4,5,6)]) + self.assertTrue(tmpfile1.name in r2.metadata.table_structures) + self.assertTrue(tmpfile1.name in r2.metadata.new_table_structures) + self.assertEqual(r2.metadata.table_structures[tmpfile1.name].atomic_fns,[tmpfile1.name]) + self.assertEqual(r2.metadata.table_structures[tmpfile1.name].source_type,'file') + self.assertEqual(r2.metadata.table_structures[tmpfile1.name].source,tmpfile1.name) + + # run file 2 on engine 1 + r3 = q1.execute('select * from %s' % tmpfile2.name) + print("QueryQuery",file=sys.stdout) + + print(r3) + self.assertTrue(r3.status == 'ok') + self.assertEqual(len(r3.warnings),0) + self.assertEqual(len(r3.data),2) + self.assertEqual(r3.metadata.output_column_name_list,['d','e','f']) + self.assertEqual(r3.data,[(10,20,30),(40,50,60)]) + self.assertTrue(tmpfile2.name in r3.metadata.table_structures) + self.assertTrue(tmpfile2.name in r3.metadata.new_table_structures) + self.assertEqual(r3.metadata.table_structures[tmpfile2.name].atomic_fns,[tmpfile2.name]) + self.assertEqual(r3.metadata.table_structures[tmpfile2.name].source,tmpfile2.name) + self.assertEqual(r3.metadata.table_structures[tmpfile2.name].source_type,'file') + + q1.done() + q2.done() + + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + def test_simple_query(self): + tmpfile = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + + q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) + r = q.execute('select * from %s' % tmpfile.name) + + self.assertTrue(r.status == 'ok') + self.assertEqual(len(r.warnings),0) + self.assertEqual(len(r.data),2) + self.assertEqual(r.metadata.output_column_name_list,['a','b','c']) + self.assertEqual(r.data,[(1,2,3),(4,5,6)]) + self.assertTrue(tmpfile.name in r.metadata.table_structures) + self.assertTrue(tmpfile.name in r.metadata.new_table_structures) + self.assertEqual(r.metadata.table_structures[tmpfile.name].atomic_fns,[tmpfile.name]) + self.assertEqual(r.metadata.table_structures[tmpfile.name].source_type,'file') + self.assertEqual(r.metadata.table_structures[tmpfile.name].source,tmpfile.name) + + q.done() + self.cleanup(tmpfile) + + def test_loaded_data_reuse(self): + tmpfile = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + + q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) + r1 = q.execute('select * from %s' % tmpfile.name) + + r2 = q.execute('select * from %s' % tmpfile.name) + + self.assertTrue(r1.status == 'ok') + self.assertEqual(len(r1.warnings),0) + self.assertEqual(len(r1.data),2) + self.assertEqual(r1.metadata.output_column_name_list,['a','b','c']) + self.assertEqual(r1.data,[(1,2,3),(4,5,6)]) + self.assertTrue(tmpfile.name in r1.metadata.table_structures) + self.assertTrue(tmpfile.name in r1.metadata.new_table_structures) + self.assertEqual(r1.metadata.table_structures[tmpfile.name].atomic_fns,[tmpfile.name]) + self.assertEqual(r1.metadata.table_structures[tmpfile.name].source_type,'file') + self.assertEqual(r1.metadata.table_structures[tmpfile.name].source,tmpfile.name) + + self.assertTrue(r2.status == 'ok') + self.assertTrue(tmpfile.name in r2.metadata.table_structures) + self.assertTrue(tmpfile.name not in r2.metadata.new_table_structures) + self.assertEqual(r2.data,r1.data) + self.assertEqual(r2.metadata.output_column_name_list,r2.metadata.output_column_name_list) + self.assertEqual(len(r2.warnings),0) + + q.done() + + self.cleanup(tmpfile) + + def test_stdin_injection(self): + tmpfile = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + + data_streams_dict = { + '-': DataStream('stdin','-',codecs.open(tmpfile.name,'rb',encoding='utf-8')) + } + q = QTextAsData(QInputParams(skip_header=True,delimiter=' '),data_streams_dict=data_streams_dict) + r = q.execute('select * from -') + + self.assertTrue(r.status == 'ok') + self.assertEqual(len(r.warnings),0) + self.assertEqual(len(r.data),2) + self.assertEqual(r.metadata.output_column_name_list,['a','b','c']) + self.assertEqual(r.data,[(1,2,3),(4,5,6)]) + self.assertEqual(r.metadata.new_table_structures['-'],r.metadata.table_structures['-']) + self.assertEqual(r.metadata.table_structures['-'].column_names,['a','b','c']) + self.assertEqual(r.metadata.table_structures['-'].python_column_types,[int,int,int]) + self.assertEqual(r.metadata.table_structures['-'].sqlite_column_types,['int','int','int']) + self.assertEqual(r.metadata.table_structures['-'].source_type,'data-stream') + self.assertEqual(r.metadata.table_structures['-'].source,'stdin') + + q.done() + self.cleanup(tmpfile) + + def test_named_stdin_injection(self): + tmpfile = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + + data_streams_dict = { + 'my_stdin_data': DataStream('my_stdin_data','my_stdin_data',codecs.open(tmpfile.name,'rb',encoding='utf-8')) + } + + q = QTextAsData(QInputParams(skip_header=True,delimiter=' '),data_streams_dict=data_streams_dict) + r = q.execute('select a from my_stdin_data') + + self.assertTrue(r.status == 'ok') + self.assertEqual(len(r.warnings),0) + self.assertEqual(len(r.data),2) + self.assertEqual(r.metadata.output_column_name_list,['a']) + self.assertEqual(r.data,[(1,),(4,)]) + self.assertTrue('my_stdin_data' in r.metadata.table_structures) + self.assertTrue('my_stdin_data' in r.metadata.new_table_structures) + self.assertEqual(r.metadata.table_structures['my_stdin_data'].qtable_name,'my_stdin_data') + + q.done() + self.cleanup(tmpfile) + + def test_data_stream_isolation(self): + tmpfile1 = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + tmpfile2 = self.create_file_with_data(six.b("d e f\n7 8 9\n10 11 12")) + + data_streams_dict = { + 'a-': DataStream('a-','a-',codecs.open(tmpfile1.name, 'rb', encoding='utf-8')), + 'b-': DataStream('b-','b-',codecs.open(tmpfile2.name, 'rb', encoding='utf-8')) + } + + q = QTextAsData(QInputParams(skip_header=True,delimiter=' '),data_streams_dict=data_streams_dict) + r1 = q.execute('select * from a-') + + self.assertTrue(r1.status == 'ok') + self.assertEqual(len(r1.warnings),0) + self.assertEqual(len(r1.data),2) + self.assertEqual(r1.metadata.output_column_name_list,['a','b','c']) + self.assertEqual(r1.data,[(1,2,3),(4,5,6)]) + self.assertTrue('a-' in r1.metadata.table_structures) + self.assertEqual(len(r1.metadata.table_structures),1) + self.assertEqual(r1.metadata.table_structures['a-'].source_type, 'data-stream') + self.assertEqual(r1.metadata.table_structures['a-'].source, 'a-') + self.assertEqual(r1.metadata.table_structures['a-'].column_names, ['a','b','c']) + self.assertEqual(r1.metadata.table_structures['a-'].python_column_types, [int,int,int]) + self.assertEqual(r1.metadata.table_structures['a-'].sqlite_column_types, ['int','int','int']) + + r2 = q.execute('select * from b-') + + self.assertTrue(r2.status == 'ok') + self.assertEqual(len(r2.warnings),0) + self.assertEqual(len(r2.data),2) + self.assertEqual(r2.metadata.output_column_name_list,['d','e','f']) + self.assertEqual(r2.data,[(7,8,9),(10,11,12)]) + + self.assertEqual(len(r1.metadata.table_structures),2) + self.assertTrue('b-' in r1.metadata.table_structures) + self.assertEqual(r1.metadata.table_structures['b-'].source_type, 'data-stream') + self.assertEqual(r1.metadata.table_structures['b-'].source, 'b-') + self.assertEqual(r1.metadata.table_structures['b-'].column_names, ['d','e','f']) + self.assertEqual(r1.metadata.table_structures['b-'].python_column_types, [int,int,int]) + self.assertEqual(r1.metadata.table_structures['b-'].sqlite_column_types, ['int','int','int']) + + q.done() + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + def test_multiple_stdin_injection(self): + tmpfile1 = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + tmpfile2 = self.create_file_with_data(six.b("d e f\n7 8 9\n10 11 12")) + + data_streams_dict = { + 'my_stdin_data1': DataStream('my_stdin_data1','my_stdin_data1',codecs.open(tmpfile1.name,'rb',encoding='utf-8')), + 'my_stdin_data2': DataStream('my_stdin_data2','my_stdin_data2',codecs.open(tmpfile2.name,'rb',encoding='utf-8')) + } + q = QTextAsData(QInputParams(skip_header=True,delimiter=' '),data_streams_dict=data_streams_dict) + r1 = q.execute('select * from my_stdin_data1') + + self.assertTrue(r1.status == 'ok') + self.assertEqual(len(r1.warnings),0) + self.assertEqual(len(r1.data),2) + self.assertEqual(r1.metadata.output_column_name_list,['a','b','c']) + self.assertEqual(r1.data,[(1,2,3),(4,5,6)]) + self.assertTrue('my_stdin_data1' in r1.metadata.table_structures) + self.assertTrue('my_stdin_data1' in r1.metadata.new_table_structures) + self.assertEqual(r1.metadata.table_structures['my_stdin_data1'].qtable_name,'my_stdin_data1') + + r2 = q.execute('select * from my_stdin_data2') + + self.assertTrue(r2.status == 'ok') + self.assertEqual(len(r2.warnings),0) + self.assertEqual(len(r2.data),2) + self.assertEqual(r2.metadata.output_column_name_list,['d','e','f']) + self.assertEqual(r2.data,[(7,8,9),(10,11,12)]) + # There should be another data load, even though it's the same 'filename' as before + self.assertTrue('my_stdin_data2' in r2.metadata.table_structures) + self.assertTrue('my_stdin_data2' in r2.metadata.new_table_structures) + self.assertEqual(r2.metadata.table_structures['my_stdin_data2'].qtable_name,'my_stdin_data2') + + r3 = q.execute('select aa.*,bb.* from my_stdin_data1 aa join my_stdin_data2 bb') + + self.assertTrue(r3.status == 'ok') + self.assertEqual(len(r3.warnings),0) + self.assertEqual(len(r3.data),4) + self.assertEqual(r3.metadata.output_column_name_list,['a','b','c','d','e','f']) + self.assertEqual(r3.data,[(1,2,3,7,8,9),(1,2,3,10,11,12),(4,5,6,7,8,9),(4,5,6,10,11,12)]) + self.assertTrue('my_stdin_data1' in r3.metadata.table_structures) + self.assertTrue('my_stdin_data1' not in r3.metadata.new_table_structures) + + q.done() + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + def test_different_input_params_for_different_files(self): + tmpfile1 = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + tmpfile2 = self.create_file_with_data(six.b("7\t8\t9\n10\t11\t12")) + + q = QTextAsData(QInputParams(skip_header=True,delimiter=' ')) + + q.load_data(tmpfile1.name,QInputParams(skip_header=True,delimiter=' ')) + q.load_data(tmpfile2.name,QInputParams(skip_header=False,delimiter='\t')) + + r = q.execute('select aa.*,bb.* from %s aa join %s bb' % (tmpfile1.name,tmpfile2.name)) + + self.assertTrue(r.status == 'ok') + self.assertEqual(len(r.warnings),0) + self.assertEqual(len(r.data),4) + self.assertEqual(r.metadata.output_column_name_list,['a','b','c','c1','c2','c3']) + self.assertEqual(r.data,[(1,2,3,7,8,9),(1,2,3,10,11,12),(4,5,6,7,8,9),(4,5,6,10,11,12)]) + self.assertTrue(tmpfile1.name not in r.metadata.new_table_structures) + self.assertTrue(tmpfile2.name not in r.metadata.new_table_structures) + + q.done() + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + def test_different_input_params_for_different_files_2(self): + tmpfile1 = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + tmpfile2 = self.create_file_with_data(six.b("7\t8\t9\n10\t11\t12")) + + q = QTextAsData() + + q.load_data(tmpfile1.name,QInputParams(skip_header=True,delimiter=' ')) + q.load_data(tmpfile2.name,QInputParams(skip_header=False,delimiter='\t')) + + r = q.execute('select aa.*,bb.* from %s aa join %s bb' % (tmpfile1.name,tmpfile2.name)) + + self.assertTrue(r.status == 'ok') + self.assertEqual(len(r.warnings),0) + self.assertEqual(len(r.data),4) + self.assertEqual(r.metadata.output_column_name_list,['a','b','c','c1','c2','c3']) + self.assertEqual(r.data,[(1,2,3,7,8,9),(1,2,3,10,11,12),(4,5,6,7,8,9),(4,5,6,10,11,12)]) + self.assertTrue(tmpfile1.name not in r.metadata.new_table_structures) + self.assertTrue(tmpfile2.name not in r.metadata.new_table_structures) + + q.done() + self.cleanup(tmpfile1) + self.cleanup(tmpfile2) + + def test_input_params_override(self): + tmpfile = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + + default_input_params = QInputParams() + + for k in default_input_params.__dict__.keys(): + setattr(default_input_params,k,'GARBAGE') + + q = QTextAsData(default_input_params) + + r = q.execute('select * from %s' % tmpfile.name) + + self.assertTrue(r.status == 'error') + + overwriting_input_params = QInputParams(skip_header=True,delimiter=' ') + + r2 = q.execute('select * from %s' % tmpfile.name,input_params=overwriting_input_params) + + self.assertTrue(r2.status == 'ok') + self.assertEqual(len(r2.warnings),0) + self.assertEqual(len(r2.data),2) + self.assertEqual(r2.metadata.output_column_name_list,['a','b','c']) + self.assertEqual(r2.data,[(1,2,3),(4,5,6)]) + self.assertTrue(tmpfile.name in r2.metadata.table_structures) + self.assertTrue(tmpfile.name in r2.metadata.new_table_structures) + self.assertEqual(r2.metadata.table_structures[tmpfile.name].atomic_fns,[tmpfile.name]) + self.assertEqual(r2.metadata.table_structures[tmpfile.name].source,tmpfile.name) + self.assertEqual(r2.metadata.table_structures[tmpfile.name].source_type,'file') + + q.done() + self.cleanup(tmpfile) + + def test_input_params_merge(self): + input_params = QInputParams() + + for k in input_params.__dict__.keys(): + setattr(input_params,k,'GARBAGE') + + merged_input_params = input_params.merged_with(QInputParams()) + + for k in merged_input_params.__dict__.keys(): + self.assertTrue(getattr(merged_input_params,k) != 'GARBAGE') + + for k in input_params.__dict__.keys(): + self.assertTrue(getattr(merged_input_params,k) != 'GARBAGE') + + def test_table_analysis_with_syntax_error(self): + + q = QTextAsData() + + q_output = q.analyze("bad syntax") + + q.done() + self.assertTrue(q_output.status == 'error') + self.assertTrue(q_output.error.msg.startswith('query error')) + + def test_execute_response(self): + tmpfile = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + + q = QTextAsData() + + q_output = q.execute("select a,c from %s" % tmpfile.name,QInputParams(skip_header=True)) + + self.assertTrue(q_output.status == 'ok') + self.assertTrue(q_output.error is None) + self.assertEqual(len(q_output.warnings),0) + self.assertEqual(len(q_output.data),2) + self.assertEqual(q_output.data,[ (1,3),(4,6) ]) + self.assertTrue(q_output.metadata is not None) + + metadata = q_output.metadata + + self.assertEqual(metadata.output_column_name_list, [ 'a','c']) + self.assertTrue(tmpfile.name in metadata.new_table_structures) + self.assertEqual(len(metadata.table_structures),1) + + table_structure = metadata.new_table_structures[tmpfile.name] + + self.assertEqual(table_structure.column_names,[ 'a','b','c']) + self.assertEqual(table_structure.python_column_types,[ int,int,int]) + self.assertEqual(table_structure.sqlite_column_types,[ 'int','int','int']) + self.assertEqual(table_structure.qtable_name, tmpfile.name) + self.assertEqual(table_structure.atomic_fns,[tmpfile.name]) + self.assertEqual(table_structure.source_type,'file') + self.assertEqual(table_structure.source,tmpfile.name) + + q.done() + self.cleanup(tmpfile) + + def test_analyze_response(self): + tmpfile = self.create_file_with_data(six.b("a b c\n1 2 3\n4 5 6")) + + q = QTextAsData() + + q_output = q.analyze("select a,c from %s" % tmpfile.name,QInputParams(skip_header=True)) + + self.assertTrue(q_output.status == 'ok') + self.assertTrue(q_output.error is None) + self.assertEqual(len(q_output.warnings),0) + self.assertEqual(len(q_output.data),2) + self.assertEqual(q_output.data,[ (1,3),(4,6) ]) + self.assertTrue(q_output.metadata is not None) + + metadata = q_output.metadata + + self.assertEqual(metadata.output_column_name_list, [ 'a','c']) + self.assertEqual(len(metadata.table_structures),1) + self.assertTrue(tmpfile.name in metadata.new_table_structures) + + table_structure = metadata.table_structures[tmpfile.name] + + self.assertEqual(table_structure.column_names,[ 'a','b','c']) + self.assertEqual(table_structure.python_column_types,[ int,int,int]) + self.assertEqual(table_structure.sqlite_column_types,[ 'int','int','int']) + self.assertEqual(table_structure.qtable_name, tmpfile.name) + self.assertEqual(table_structure.atomic_fns,[tmpfile.name]) + self.assertEqual(table_structure.source_type,'file') + self.assertEqual(table_structure.source,tmpfile.name) + + q.done() + self.cleanup(tmpfile) + + def test_load_data_from_string_without_previous_data_load(self): + input_str = six.u('column1,column2,column3\n') + six.u('\n').join([six.u('value1,2.5,value3')] * 1000) + + + data_streams_dict = { + 'my_data': DataStream('my_data_stream_id','my_data',six.StringIO(input_str)) + } + q = QTextAsData(default_input_params=QInputParams(skip_header=True,delimiter=','),data_streams_dict=data_streams_dict) + + q_output = q.execute('select column2,column3 from my_data') + + self.assertTrue(q_output.status == 'ok') + self.assertTrue(q_output.error is None) + self.assertEqual(len(q_output.warnings),0) + self.assertTrue(len(q_output.data),1000) + self.assertEqual(len(set(q_output.data)),1) + self.assertEqual(list(set(q_output.data))[0],(2.5,'value3')) + + metadata = q_output.metadata + + self.assertTrue(metadata.output_column_name_list,['column2','column3']) + self.assertTrue('my_data' in metadata.new_table_structures) + self.assertEqual(len(metadata.table_structures),1) + + table_structure = metadata.table_structures['my_data'] + + self.assertEqual(table_structure.column_names,['column1','column2','column3']) + self.assertEqual(table_structure.sqlite_column_types,['text','real','text']) + self.assertEqual(table_structure.python_column_types,[str,float,str]) + self.assertEqual(table_structure.qtable_name, 'my_data') + self.assertEqual(table_structure.source_type, 'data-stream') + self.assertEqual(table_structure.source, 'my_data_stream_id') + + q.done() + + def test_load_data_from_string_with_previous_data_load(self): + input_str = six.u('column1,column2,column3\n') + six.u('\n').join([six.u('value1,2.5,value3')] * 1000) + + data_streams_dict = { + 'my_data': DataStream('a','my_data',six.StringIO(input_str)) + } + q = QTextAsData(default_input_params=QInputParams(skip_header=True,delimiter=','),data_streams_dict=data_streams_dict) + + dl = q.load_data('my_data',QInputParams(skip_header=True,delimiter=',')) + + q_output = q.execute('select column2,column3 from my_data') + + self.assertTrue(q_output.status == 'ok') + self.assertTrue(q_output.error is None) + self.assertEqual(len(q_output.warnings),0) + self.assertTrue(len(q_output.data),1000) + self.assertEqual(len(set(q_output.data)),1) + self.assertEqual(list(set(q_output.data))[0],(2.5,'value3')) + + metadata = q_output.metadata + + self.assertTrue(metadata.output_column_name_list,['column2','column3']) + self.assertTrue('my_data' not in metadata.new_table_structures) + self.assertEqual(len(metadata.table_structures),1) + + table_structure = metadata.table_structures['my_data'] + + self.assertEqual(table_structure.column_names,['column1','column2','column3']) + self.assertEqual(table_structure.sqlite_column_types,['text','real','text']) + self.assertEqual(table_structure.python_column_types,[str,float,str]) + self.assertEqual(table_structure.qtable_name, 'my_data') + + q.done() + + + +class BenchmarkAttemptResults(object): + def __init__(self, attempt, lines, columns, duration,return_code): + self.attempt = attempt + self.lines = lines + self.columns = columns + self.duration = duration + self.return_code = return_code + + def __str__(self): + return "{}".format(self.__dict__) + __repr__ = __str__ + +class BenchmarkResults(object): + def __init__(self, lines, columns, attempt_results, mean, stddev): + self.lines = lines + self.columns = columns + self.attempt_results = attempt_results + self.mean = mean + self.stddev = stddev + + def __str__(self): + return "{}".format(self.__dict__) + __repr__ = __str__ + +@pytest.mark.benchmark +class BenchmarkTests(AbstractQTestCase): + + BENCHMARK_DIR = os.environ.get('Q_BENCHMARK_DATA_DIR') + + def _ensure_benchmark_data_dir_exists(self): + try: + os.mkdir(BenchmarkTests.BENCHMARK_DIR) + except Exception as e: + pass + + def _create_benchmark_file_if_needed(self): + self._ensure_benchmark_data_dir_exists() + + if os.path.exists('{}/'.format(BenchmarkTests.BENCHMARK_DIR)): + return + + g = GzipFile('unit-file.csv.gz') + d = g.read().decode('utf-8') + f = open('{}/benchmark-file.csv'.format(BenchmarkTests.BENCHMARK_DIR), 'w') + for i in range(100): + f.write(d) + f.close() + + def _prepare_test_file(self, lines, columns): + + filename = '{}/_benchmark_data__lines_{}_columns_{}.csv'.format(BenchmarkTests.BENCHMARK_DIR,lines, columns) + + if os.path.exists(filename): + return filename + + c = ['c{}'.format(x + 1) for x in range(columns)] + + # write a header line + ff = open(filename,'w') + ff.write(",".join(c)) + ff.write('\n') + ff.close() + + r, o, e = run_command('head -{} {}/benchmark-file.csv | ' + Q_EXECUTABLE + ' -d , "select {} from -" >> {}'.format(lines, BenchmarkTests.BENCHMARK_DIR, ','.join(c), filename)) + self.assertEqual(r, 0) + # Create file cache as part of preparation + r, o, e = run_command(Q_EXECUTABLE + ' -C readwrite -d , "select count(*) from %s"' % filename) + self.asserEqual(r, 0) + return filename + + def _decide_result(self,attempt_results): + + failed = list(filter(lambda a: a.return_code != 0,attempt_results)) + + if len(failed) == 0: + mean = sum([x.duration for x in attempt_results]) / len(attempt_results) + sum_squared = sum([(x.duration - mean)**2 for x in attempt_results]) + ddof = 0 + pvar = sum_squared / (len(attempt_results) - ddof) + stddev = pvar ** 0.5 + else: + mean = None + stddev = None + + return BenchmarkResults( + attempt_results[0].lines, + attempt_results[0].columns, + attempt_results, + mean, + stddev + ) + + def _perform_test_performance_matrix(self,name,generate_cmd_function): + results = [] + + benchmark_results_folder = os.environ.get("Q_BENCHMARK_RESULTS_FOLDER",'') + if benchmark_results_folder == "": + raise Exception("Q_BENCHMARK_RESULTS_FOLDER must be provided as an environment variable") + + self._create_benchmark_file_if_needed() + for columns in [1, 5, 10, 20, 50, 100]: + for lines in [1, 10, 100, 1000, 10000, 100000, 1000000]: + attempt_results = [] + for attempt in range(10): + filename = self._prepare_test_file(lines, columns) + if DEBUG: + print("Testing {}".format(filename)) + t0 = time.time() + r, o, e = run_command(generate_cmd_function(filename,lines,columns)) + duration = time.time() - t0 + attempt_result = BenchmarkAttemptResults(attempt, lines, columns, duration, r) + attempt_results += [attempt_result] + if DEBUG: + print("Results: {}".format(attempt_result.__dict__)) + final_result = self._decide_result(attempt_results) + results += [final_result] + + series_fields = [six.u('lines'),six.u('columns')] + value_fields = [six.u('mean'),six.u('stddev')] + + all_fields = series_fields + value_fields + + output_filename = '{}/{}.benchmark-results'.format(benchmark_results_folder,name) + output_file = open(output_filename,'w') + for columns,g in itertools.groupby(sorted(results,key=lambda x:x.columns),key=lambda x:x.columns): + x = six.u("\t").join(series_fields + [six.u('{}_{}').format(name, f) for f in value_fields]) + print(x,file = output_file) + for result in g: + print(six.u("\t").join(map(str,[getattr(result,f) for f in all_fields])),file=output_file) + output_file.close() + + print("results have been written to : {}".format(output_filename)) + if DEBUG: + print("RESULTS FOR {}".format(name)) + print(open(output_filename,'r').read()) + + def test_q_matrix(self): + Q_BENCHMARK_NAME = os.environ.get('Q_BENCHMARK_NAME') + if Q_BENCHMARK_NAME is None: + raise Exception('Q_BENCHMARK_NAME must be provided as an env var') + + def generate_q_cmd(data_filename, line_count, column_count): + Q_BENCHMARK_ADDITIONAL_PARAMS = os.environ.get('Q_BENCHMARK_ADDITIONAL_PARAMS') or '' + additional_params = '' + additional_params = additional_params + ' ' + Q_BENCHMARK_ADDITIONAL_PARAMS + return '{} -d , {} "select count(*) from {}"'.format(Q_EXECUTABLE,additional_params, data_filename) + self._perform_test_performance_matrix(Q_BENCHMARK_NAME,generate_q_cmd) + + def _get_textql_version(self): + r,o,e = run_command("textql --version") + if r != 0: + raise Exception("Could not find textql") + if len(e) != 0: + raise Exception("Errors while getting textql version") + return o[0] + + def _get_octosql_version(self): + r,o,e = run_command("octosql --version") + if r != 0: + raise Exception("Could not find octosql") + if len(e) != 0: + raise Exception("Errors while getting octosql version") + version = re.findall('v[0-9]+\\.[0-9]+\\.[0-9]+',str(o[0],encoding='utf-8'))[0] + return version + + def test_textql_matrix(self): + def generate_textql_cmd(data_filename,line_count,column_count): + return 'textql -dlm , -sql "select count(*)" {}'.format(data_filename) + + name = 'textql_%s' % self._get_textql_version() + self._perform_test_performance_matrix(name,generate_textql_cmd) + + def test_octosql_matrix(self): + config_fn = self.random_tmp_filename('octosql', 'config') + def generate_octosql_cmd(data_filename,line_count,column_count): + j = """ +dataSources: + - name: bmdata + type: csv + config: + path: "{}" + headerRow: false + batchSize: 10000 +""".format(data_filename)[1:] + f = open(config_fn,'w') + f.write(j) + f.close() + return 'octosql -c {} -o batch-csv "select count(*) from bmdata a"'.format(config_fn) + + name = 'octosql_%s' % self._get_octosql_version() + self._perform_test_performance_matrix(name,generate_octosql_cmd) + +def suite(): + tl = unittest.TestLoader() + basic_stuff = tl.loadTestsFromTestCase(BasicTests) + parsing_mode = tl.loadTestsFromTestCase(ParsingModeTests) + sql = tl.loadTestsFromTestCase(SqlTests) + formatting = tl.loadTestsFromTestCase(FormattingTests) + basic_module_stuff = tl.loadTestsFromTestCase(BasicModuleTests) + save_db_to_disk_tests = tl.loadTestsFromTestCase(SaveDbToDiskTests) + user_functions_tests = tl.loadTestsFromTestCase(UserFunctionTests) + multi_header_tests = tl.loadTestsFromTestCase(MultiHeaderTests) + return unittest.TestSuite([basic_module_stuff, basic_stuff, parsing_mode, sql, formatting,save_db_to_disk_tests,multi_header_tests,user_functions_tests]) + +if __name__ == '__main__': + if len(sys.argv) > 1: + suite = unittest.TestSuite() + if '.' in sys.argv[1]: + c,m = sys.argv[1].split(".") + suite.addTest(globals()[c](m)) + else: + tl = unittest.TestLoader() + tc = tl.loadTestsFromTestCase(globals()[sys.argv[1]]) + suite = unittest.TestSuite([tc]) + else: + suite = suite() + + test_runner = unittest.TextTestRunner(verbosity=2) + result = test_runner.run(suite) + sys.exit(not result.wasSuccessful())