Add check_gn_headers.py to find missing headers in GN

build/check_gn_headers.py

Index: build/check_gn_headers.py
diff --git a/build/check_gn_headers.py b/build/check_gn_headers.py
new file mode 100755
index 0000000000000000000000000000000000000000..3724b61d37d375d892a7fe32fa9f87f7c192f61e
--- /dev/null
+++ b/build/check_gn_headers.py
@@ -0,0 +1,83 @@
+#!/usr/bin/env python
+# Copyright 2016 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+"""Find header files missing in GN.
+
+This script gets all the header files from ninja_deps, which is from the true
+dependency generated by the compiler, and report if they don't exist in GN.
+"""
+
+import argparse
+import json
+import os
+import re
+import subprocess
+import sys
+
+
+IGNORE_REGEX = re.compile('(^(third_party|build)\/)|(.*/gen/.*)')

[stgao, 2016/11/18 02:17:28]

Just curious, why we ignore third_party?

[wychen, 2016/11/18 03:35:40]

I guess we shouldn't.

+
+
+def get_d_headers(out_dir):
+  """Return all the header files from ninja_deps"""
+  all_headers = set()
+
+  out = subprocess.check_output(['ninja', '-C', out_dir, '-t', 'deps'])
+  is_valid = False
+  for line in out.split('\n'):
+    if line.startswith('    '):
+      if is_valid and line.endswith('.h'):

[stgao, 2016/11/18 02:17:28]

Some header files are with extension .hh

[wychen, 2016/11/18 03:35:40]

Done.

+        f = os.path.normpath(os.path.join(out_dir, line.strip()))

[stgao, 2016/11/18 02:17:28]

IIRC, normpath is platform-dependent, so path separator is '\\' on windows while
it is '/' on Linux.
This might cause some issue with the regex used here.

[wychen, 2016/11/18 03:35:40]

Done.

+        if not IGNORE_REGEX.match(f):
+          all_headers.add(f)
+    else:
+      is_valid = line.endswith('(VALID)')
+
+  return all_headers
+
+
+def get_gn_headers(out_dir):
+  """Return all the header files from GN"""
+  all_headers = set()
+
+  subprocess.check_call(['gn', 'gen', out_dir, '--ide=json'])
+
+  gn = json.load(open(os.path.join(out_dir, 'project.json')))
+  for target, properties in gn['targets'].iteritems():
+    if 'sources' in properties:
+      for f in properties['sources']:

[stgao, 2016/11/18 02:17:28]

nit: properties.get('sources', [])?

+        if f.endswith('.h'):
+          all_headers.add(os.path.normpath('.' + f))

[stgao, 2016/11/18 02:17:28]

same here for normpath

[wychen, 2016/11/18 03:35:40]

Done.

+
+  return all_headers
+
+
+def main():
+  parser = argparse.ArgumentParser()
+  parser.add_argument('--out-dir', default='out/Debug')
+  parser.add_argument('--whitelist')
+  parser.add_argument('args', nargs=argparse.REMAINDER)
+
+  args, extras = parser.parse_known_args()
+
+  d = get_d_headers(args.out_dir)
+  gn = get_gn_headers(args.out_dir)
+  missing = d - gn
+
+  if args.whitelist:
+    whitelist = set(open(args.whitelist).read().split('\n'))
+    missing -= whitelist
+
+  if len(missing) > 0:
+    for i in sorted(missing):
+      print(i)
+    return 1
+
+if __name__ == '__main__':
+  try:
+    sys.exit(main())
+  except KeyboardInterrupt:
+    sys.stderr.write('interrupted\n')
+    sys.exit(1)