Add presubmit check to prevent package boundary violations.

tools-webrtc/check_package_boundaries.py

Index: tools-webrtc/check_package_boundaries.py
diff --git a/tools-webrtc/check_package_boundaries.py b/tools-webrtc/check_package_boundaries.py
new file mode 100644
index 0000000000000000000000000000000000000000..64cac9cfb4e40ecd42d82891910147f711b4f357
--- /dev/null
+++ b/tools-webrtc/check_package_boundaries.py
@@ -0,0 +1,126 @@
+#!/usr/bin/env python
+
+# Copyright (c) 2017 The WebRTC project authors. All Rights Reserved.
+#
+# Use of this source code is governed by a BSD-style license
+# that can be found in the LICENSE file in the root of the source
+# tree. An additional intellectual property rights grant can be found
+# in the file PATENTS.  All contributing project authors may
+# be found in the AUTHORS file in the root of the source tree.
+
+import argparse
+import logging
+import os
+import re
+import sys
+
+
+DISPLAY_LEVEL = 1
+IGNORE_LEVEL = 0
+
+# TARGET_RE matches a GN target, and extracts the target name and the contents.
+TARGET_RE = re.compile(r'\d+\$(?P<indent>\s*)\w+\("(?P<target_name>\w+)"\) {'
+                       r'(?P<target_contents>.*?)'
+                       r'\d+\$(?P=indent)}',
+                       re.MULTILINE | re.DOTALL)
+
+# SOURCES_RE matches a block of sources inside a GN target.
+SOURCES_RE = re.compile(r'sources \+?= \[(?P<sources>.*?)\]',
+                        re.MULTILINE | re.DOTALL)
+
+LOG_FORMAT = '%(message)s'
+ERROR_MESSAGE = ("{}:{} in target '{}':\n"
+                 "  Source file '{}'\n"
+                 "  crosses boundary of package '{}'.\n")
+
+
+class Logger(object):
+  def __init__(self, messages_left=None):
+    self.log_level = DISPLAY_LEVEL
+    self.messages_left = messages_left
+
+  def log(self, message):
+    if self.messages_left is not None:
+      if not self.messages_left:
+        self.log_level = IGNORE_LEVEL
+      else:
+        self.messages_left -= 1
+    logging.log(self.log_level, message)
+
+
+def _BuildSubpackagesPattern(packages, query):
+  """Returns a regular expression that matches source files inside subpackages
+  of the given query."""
+  query += '/'
+  length = len(query)
+  pattern = r'(?P<line_number>\d+)\$\s*"(?P<source_file>(?P<subpackage>'
+  pattern += '|'.join(package[length:] for package in packages
+                      if package.startswith(query))
+  pattern += r')[\w\./]*)"'
+  return re.compile(pattern)
+
+def _ReadFileAndPrependLines(file_path):
+  """Reads the contents of a file and prepends the line number to every line."""
+  with open(file_path) as f:
+    return "".join("{}${}".format(line_number, line)
+                   for line_number, line in enumerate(f, 1))
+
+def _CheckBuildFile(build_file_path, packages, logger):
+  """Iterates oven all the targets of the given BUILD.gn file, and verifies that
+  the source files referenced by it don't belong to any of it's subpackages.
+  Returns True if a package boundary violation was found.
+  """
+  found_violations = False
+  package = os.path.dirname(build_file_path)
+  subpackages_re = _BuildSubpackagesPattern(packages, package)
+
+  build_file_contents = _ReadFileAndPrependLines(build_file_path)
+  for target_match in TARGET_RE.finditer(build_file_contents):
+    target_name = target_match.group('target_name')
+    target_contents = target_match.group('target_contents')
+    for sources_match in SOURCES_RE.finditer(target_contents):
+      sources = sources_match.group('sources')
+      for subpackages_match in subpackages_re.finditer(sources):
+        subpackage = subpackages_match.group('subpackage')
+        source_file = subpackages_match.group('source_file')
+        line_number = subpackages_match.group('line_number')
+        if subpackage:
+          found_violations = True

[kjellander_webrtc, 2017/01/19 10:51:26]

Seems like a bug was found with the tests? :)

[ehmaldonado_webrtc, 2017/01/19 13:33:56]

Yup :)

+          logger.log(ERROR_MESSAGE.format(build_file_path, line_number,
+                                          target_name, source_file, subpackage))
+
+  return found_violations
+
+def main():
+  parser = argparse.ArgumentParser(
+      description='Script that checks package boundary violations in GN '
+                  'build files.')
+
+  parser.add_argument('root_dir', metavar='ROOT_DIR',
+                      help='The root directory that contains all BUILD.gn '
+                           'files to be processed.')
+  parser.add_argument('build_files', metavar='BUILD_FILE', nargs='*',
+                      help='A list of BUILD.gn files to be processed. If no '
+                           'files are given, all BUILD.gn files under ROOT_DIR '
+                           'will be processed.')
+  parser.add_argument('--max_messages', type=int, default=None,
+                      help='If set, the maximum number of violations to be '
+                           'displayed.')
+
+  args = parser.parse_args()
+
+  logging.basicConfig(format=LOG_FORMAT)
+  logging.getLogger().setLevel(DISPLAY_LEVEL)
+  logger = Logger(args.max_messages)
+
+  packages = [root for root, _, files in os.walk(args.root_dir)
+              if 'BUILD.gn' in files]
+  default_build_files = [os.path.join(package, 'BUILD.gn')
+                         for package in packages]
+
+  build_files = args.build_files or default_build_files
+  return any(_CheckBuildFile(build_file_path, packages, logger)
+             for build_file_path in build_files)
+
+if __name__ == '__main__':
+  sys.exit(main())