Reland of Deduplicate Monochrome locale .paks

tools/grit/pak_util.py

 #!/usr/bin/env python
 # Copyright 2017 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.
 
 """A tool for interacting with .pak files.
 
 For details on the pak file format, see:
 https://dev.chromium.org/developers/design-documents/linuxresourcesandlocalizedstrings
 """
@@ skipping 44 matching lines @@
           desc = desc[:60] + u'...'
         desc = desc.replace('\n', '\\n')
       except UnicodeDecodeError:
         pass
     sha1 = hashlib.sha1(data).hexdigest()[:10]
     line = u'Entry(id={}, len={}, sha1={}): {}'.format(
         resource_id, len(data), sha1, desc)
     print line.encode('utf-8')
 
 
+def _ListMain(args):
+  resources, _ = data_pack.ReadDataPack(args.pak_file)
+  for resource_id in sorted(resources.keys()):
+    args.output.write('%d\n' % resource_id)
+
+
 def main():
   parser = argparse.ArgumentParser(
       description=__doc__, formatter_class=argparse.RawTextHelpFormatter)
   sub_parsers = parser.add_subparsers()
 
   sub_parser = sub_parsers.add_parser('repack',
       help='Combines several .pak files into one.')
   sub_parser.add_argument('output_pak_file', help='File to create.')
   sub_parser.add_argument('input_pak_files', nargs='+',
       help='Input .pak files.')
   sub_parser.add_argument('--whitelist',
       help='Path to a whitelist used to filter output pak file resource IDs.')
   sub_parser.add_argument('--suppress-removed-key-output', action='store_true',
       help='Do not log which keys were removed by the whitelist.')
   sub_parser.set_defaults(func=_RepackMain)
 
   sub_parser = sub_parsers.add_parser('extract', help='Extracts pak file')
   sub_parser.add_argument('pak_file')
   sub_parser.add_argument('-o', '--output-dir', default='.',
                           help='Directory to extract to.')
   sub_parser.set_defaults(func=_ExtractMain)
 
   sub_parser = sub_parsers.add_parser('print',
       help='Prints all pak IDs and contents. Useful for diffing.')
   sub_parser.add_argument('pak_file')
   sub_parser.set_defaults(func=_PrintMain)
+  
+  sub_parser = sub_parsers.add_parser('list',

[agrieve, 2017/07/17 13:50:52]

nit: Might be nice to just make "print" more capable rather than adding another
almost-the-same-sounding command.

E.g.: --output would be useful for "print".

How about doing:

pak_util.py print --output=FOO --format='{id}'

Have the default for format be: 'Entry(id={id}, len={len}, sha1={sha1}):
{value:.60}'

Then print lines via fmt.format(id=resource_id, len=len(data), sha1=sha1,
value=desc) 

Having the default like this should be enough to document what keys are allowed
:)

[F, 2017/07/17 14:35:14]

Resolved offline. Name changed to "list-id"

+      help='Outputs all resource IDs to a file.')
+  sub_parser.add_argument('pak_file')
+  sub_parser.add_argument('--output', type=argparse.FileType('w'),
+      default=sys.stdout,
+      help='The resource list path to write (default stdout)')
+  sub_parser.set_defaults(func=_ListMain)
 
   if len(sys.argv) == 1:
     parser.print_help()
     sys.exit(1)
   elif len(sys.argv) == 2 and sys.argv[1] in actions:
     parser.parse_args(sys.argv[1:] + ['-h'])
     sys.exit(1)
 
   args = parser.parse_args()
   args.func(args)
 
 
 if __name__ == '__main__':
   main()