Accept CSS url()s without quotes.

grit/format/html_inline.py

 #!/usr/bin/env python
 # Copyright (c) 2012 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.
 
 """Flattens a HTML file by inlining its external resources.
 
 This is a small script that takes a HTML file, looks for src attributes
 and inlines the specified file, producing one HTML file with no external
 dependencies. It recursively inlines the included files.
@@ skipping 63 matching lines @@
   filename = filename.replace(DIST_SUBSTR , distribution)
   filepath = os.path.join(base_path, filename)
   inlined_files.add(filepath)
 
   if names_only:
     return ""
 
   mimetype = mimetypes.guess_type(filename)[0] or 'text/plain'
   inline_data = base64.standard_b64encode(util.ReadFile(filepath, util.BINARY))
 
-  prefix = src_match.string[src_match.start():src_match.start('filename')-1]
-  return "%s\"data:%s;base64,%s\"" % (prefix, mimetype, inline_data)
+  prefix = src_match.string[src_match.start():src_match.start('filename')]
+  suffix = src_match.string[src_match.end('filename'):src_match.end()]
+  if prefix and prefix[-1] in '\'"':
+    prefix = prefix[:-1]
+  if suffix and suffix[0] in '\'"':
+    suffix = suffix[1:]
+  return '%s"data:%s;base64,%s"%s' % (prefix, mimetype, inline_data, suffix)
 
 
 class InlinedData:
   """Helper class holding the results from DoInline().
 
   Holds the inlined data and the set of filenames of all the inlined
   files.
   """
   def __init__(self, inlined_data, inlined_files):
     self.inlined_data = inlined_data
@@ skipping 134 matching lines @@
     inlined_files.add(filepath)
     # When resolving CSS files we need to pass in the path so that relative URLs
     # can be resolved.
     return '<style>%s</style>' % InlineCSSText(
         util.ReadFile(filepath, util.BINARY), filepath)
 
   def InlineCSSImages(text, filepath=input_filepath):
     """Helper function that inlines external images in CSS backgrounds."""
     # Replace contents of url() for css attributes: content, background,
     # or *-image.
-    return re.sub('(?:content|background|[\w-]*-image):[^;]*' +
-                  '(?:url\((?:\'|\")([^"\'\)\(]*)(?:\'|\")\)|' +
+    return re.sub('(content|background|[\w-]*-image):[^;]*' +
+                  '(url\((?P<quote1>"|\'|)[^"\'()]*(?P=quote1)\)|' +
                       'image-set\(' +
-                          '([ ]*url\((?:\'|\")([^"\'\)\(]*)(?:\'|\")\)' +
-                              '[ ]*[0-9.]*x[ ]*(,[ ]*)?)*\))',
+                          '([ ]*url\((?P<quote2>"|\'|)[^"\'()]*(?P=quote2)\)' +
+                              '[ ]*[0-9.]*x[ ]*(,[ ]*)?)+\))',
                   lambda m: InlineCSSUrls(m, filepath),
                   text)
 
   def InlineCSSUrls(src_match, filepath=input_filepath):
     """Helper function that inlines each url on a CSS image rule match."""
     # Replace contents of url() references in matches.
-    return re.sub('url\((?:\'|\")(?P<filename>[^"\'\)\(]*)(?:\'|\")',
+    return re.sub('url\((?P<quote>"|\'|)(?P<filename>[^"\'()]*)(?P=quote)\)',
                   lambda m: SrcReplace(m, filepath),
                   src_match.group(0))
 
 
 
   flat_text = util.ReadFile(input_filename, util.BINARY)
 
   if not allow_external_script:
     # We need to inline css and js before we inline images so that image
     # references gets inlined in the css and js
@@ skipping 91 matching lines @@
 
 def main():
   if len(sys.argv) <= 2:
     print "Flattens a HTML file by inlining its external resources.\n"
     print "html_inline.py inputfile outputfile"
   else:
     InlineToFile(sys.argv[1], sys.argv[2], None)
 
 if __name__ == '__main__':
   main()