Wrong filename on file upload

url/third_party/mozilla/url_parse.cc

 /* Based on nsURLParsers.cc from Mozilla
  * -------------------------------------
  * The contents of this file are subject to the Mozilla Public License Version
  * 1.1 (the "License"); you may not use this file except in compliance with
  * the License. You may obtain a copy of the License at
  * http://www.mozilla.org/MPL/
  *
  * Software distributed under the License is distributed on an "AS IS" basis,
  * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
  * for the specific language governing rights and limitations under the
@@ skipping 607 matching lines @@
   // Handle empty paths: they have no file names.
   if (!path.is_nonempty()) {
     file_name->reset();
     return;
   }
 
   // Search backwards for a parameter, which is a normally unused field in a
   // URL delimited by a semicolon. We parse the parameter as part of the
   // path, but here, we don't want to count it. The last semicolon is the
   // parameter. The path should start with a slash, so we don't need to check
   // the first one.

[asanka, 2014/09/17 20:56:25]

Update the comment?

I'd suggest adding a reference to http://tools.ietf.org/html/rfc3986#section-3.3
which states some caveats about the assumptions we are making here.

Also, the now obsolete http://tools.ietf.org/html/rfc2396#section-3.3 might be
relevant since the concept of a path segment specific parameter existed in RFC
2396 but not in RFC 3986. The latter just mentions that the ';' and ','
characters are sometimes used as delimiters for segment specific parameters.
However a ';' is no longer considered a reserved character.

Twitter uses a ':' as a delimiter for path segment parameters. Should we handle
that too? (https://code.google.com/p/chromium/issues/detail?id=172529)

[brettw, 2014/10/14 18:28:55]

Yes, this needs updating. The whole thing should be rewritten. This loop is no
longer searching backwards for the parameter, but finding the range between the
last slash and the following semicolon.
No, this has nothing to do with how some sites may decide to use certain
characters in a URL.

   int file_end = path.end();
-  for (int i = path.end() - 1; i > path.begin; i--) {
+  for (int i = path.end() - 1; i >= path.begin; i--) {
     if (spec[i] == ';') {
       file_end = i;
-      break;
-    }
-  }
-
-  // Now search backwards from the filename end to the previous slash
-  // to find the beginning of the filename.
-  for (int i = file_end - 1; i >= path.begin; i--) {
-    if (IsURLSlash(spec[i])) {
+    } else if (IsURLSlash(spec[i])) {
       // File name is everything following this character to the end
       *file_name = MakeRange(i + 1, file_end);
       return;
     }
   }
 
   // No slash found, this means the input was degenerate (generally paths
   // will start with a slash). Let's call everything the file name.
   *file_name = MakeRange(path.begin, file_end);
   return;
@@ skipping 284 matching lines @@
 }
 
 void ParseAfterScheme(const base::char16* spec,
                       int spec_len,
                       int after_scheme,
                       Parsed* parsed) {
   DoParseAfterScheme(spec, spec_len, after_scheme, parsed);
 }
 
 }  // namespace url