Merge pull request EFForg#6679 from Hainish/fix-validate-filenames-test

cooperq · web-flow · commit 8a838314df6c · 2016-08-30T11:49:56.000-07:00
Ensure ruleset filename validation happens on ruleset PRs
diff --git a/test/travis.sh b/test/travis.sh
@@ -55,6 +55,7 @@ if [ "$RULESETS_CHANGED" ]; then
     echo >&2 "Performing comprehensive coverage test."
     docker_build
     docker run --rm -ti -v $(pwd):/opt httpse python utils/trivial-validate.py
+    docker run --rm -ti -v $(pwd):/opt httpse python utils/ruleset_filenames_validate.py
     docker run --rm -ti -v $(pwd):/opt httpse bash -c "test/rules.sh"
   fi
 
diff --git a/utils/make-json.py b/utils/make-json.py
@@ -4,18 +4,16 @@
 # The JSON DB is used by the Firefox addon.
 #
 
-import glob
 import locale
 import json
 import os
-import re
-import sqlite3
-import subprocess
 import sys
 
 import collections
 from lxml import etree
 
+from ruleset_filenames_validate import validate_filenames
+
 # Explicitly set locale so sorting order for filenames is consistent.
 # This is important for deterministic builds.
 # https://trac.torproject.org/projects/tor/ticket/11630#comment:20
@@ -35,28 +33,7 @@
 xpath_host = etree.XPath("/ruleset/target/@host")
 xpath_ruleset = etree.XPath("/ruleset")
 
-# Sort filenames so output is deterministic.
-filenames = sorted(glob.glob('src/chrome/content/rules/*'))
-
-counted_lowercase_names = collections.Counter([name.lower() for name in filenames])
-most_common_entry = counted_lowercase_names.most_common(1)[0]
-if most_common_entry[1] > 1:
-    dupe_filename = re.compile(re.escape(most_common_entry[0]), re.IGNORECASE)
-    print("%s failed case-insensitivity testing." % filter(dupe_filename.match, filenames))
-    print("Rules exist with identical case-insensitive names, which breaks some filesystems.")
-    sys.exit(1)
-
-for fi in filenames:
-    basename = fi.split(os.path.sep)[-1]
-    if basename == '00README' or basename == 'make-trivial-rule' or basename == 'default.rulesets':
-        continue
-
-    if " " in fi:
-        print("%s failed validity: Rule filenames cannot contain spaces" % (fi))
-        sys.exit(1)
-    if not fi.endswith('.xml'):
-        print("%s failed validity: Rule filenames must end in .xml" % (fi))
-        sys.exit(1)
+for fi in validate_filenames():
 
     try:
         tree = etree.parse(fi, parser)
diff --git a/utils/ruleset_filenames_validate.py b/utils/ruleset_filenames_validate.py
@@ -0,0 +1,40 @@
+#!/usr/bin/env python2.7
+# 
+# Validates and provides a generator for ruleset filenames
+#
+
+import glob
+import os
+import re
+import sys
+
+import collections
+
+def validate_filenames():
+    # Sort filenames so output is deterministic.
+    filenames = sorted(glob.glob('src/chrome/content/rules/*'))
+
+    counted_lowercase_names = collections.Counter([name.lower() for name in filenames])
+    most_common_entry = counted_lowercase_names.most_common(1)[0]
+    if most_common_entry[1] > 1:
+        dupe_filename = re.compile(re.escape(most_common_entry[0]), re.IGNORECASE)
+        print("%s failed case-insensitivity testing." % filter(dupe_filename.match, filenames))
+        print("Rules exist with identical case-insensitive names, which breaks some filesystems.")
+        sys.exit(1)
+
+    for fi in filenames:
+        basename = fi.split(os.path.sep)[-1]
+        if basename == '00README' or basename == 'make-trivial-rule' or basename == 'default.rulesets':
+            continue
+
+        if " " in fi:
+            print("%s failed validity: Rule filenames cannot contain spaces" % (fi))
+            sys.exit(1)
+        if not fi.endswith('.xml'):
+            print("%s failed validity: Rule filenames must end in .xml" % (fi))
+            sys.exit(1)
+
+        yield fi
+
+if __name__ == "__main__":
+    [ fi for fi in validate_filenames() ]
