chrome/tools/extract_histograms.py

   1 #!/usr/bin/env python
   2 # Copyright (c) 2011 The Chromium Authors. All rights reserved.
   3 # Use of this source code is governed by a BSD-style license that can be
   4 # found in the LICENSE file.
   5
   6 """Extract UMA histogram strings from the Chrome source.
   7
   8 This program generates the list of known histograms we expect to see in
   9 the user behavior logs.  It walks the Chrome source, looking for calls
  10 to UMA histogram macros.
  11
  12 Run it from the chrome/browser directory like:
  13   extract_histograms.py > histogram_list
  14 """
  15
  16 # TODO(evanm): get all the jankometer histogram names.
  17
  18 __author__ = 'evanm (Evan Martin)'
  19
  20 import os
  21 import re
  22 import sys
  23
  24 def GrepForHistograms(path, histograms):
  25   """Grep a source file for calls to histogram macros functions.
  26
  27   Arguments:
  28     path: path to the file
  29     histograms: set of histograms to add to
  30   """
  31
  32   histogram_re = re.compile(r'HISTOGRAM_\w+\(L"(.*)"')
  33   for line in open(path):
  34     match = histogram_re.search(line)
  35     if match:
  36       histograms.add(match.group(1))
  37
  38
  39 def WalkDirectory(root_path, histograms):
  40   for path, dirs, files in os.walk(root_path):
  41     if '.svn' in dirs:
  42       dirs.remove('.svn')
  43     for file in files:
  44       ext = os.path.splitext(file)[1]
  45       if ext == '.cc':
  46         GrepForHistograms(os.path.join(path, file), histograms)
  47
  48
  49 def main(argv):
  50   histograms = set()
  51
  52   # Walk the source tree to process all .cc files.
  53   WalkDirectory('..', histograms)
  54
  55   # Print out the histograms as a sorted list.
  56   for histogram in sorted(histograms):
  57     print histogram
  58   return 0
  59
  60
  61 if '__main__' == __name__:
  62   sys.exit(main(sys.argv))