Modify w3c import script to be able to run over a whole repo.

Tools/Scripts/webkitpy/w3c/test_parser.py

 #!/usr/bin/env python
 
 # Copyright (C) 2013 Adobe Systems Incorporated. All rights reserved.
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions
 # are met:
 #
 # 1. Redistributions of source code must retain the above
 #    copyright notice, this list of conditions and the following
@@ skipping 28 matching lines @@
         self.options = options
         self.filename = filename
         self.host = Host()
         self.filesystem = self.host.filesystem
 
         self.test_doc = None
         self.ref_doc = None
         self.load_file(filename)
 
     def load_file(self, filename):
-        if self.filesystem.exists(filename):
-            self.test_doc = Parser(self.filesystem.read_binary_file(filename))
+        if self.filesystem.isfile(filename):
+            try:
+                self.test_doc = Parser(self.filesystem.read_binary_file(filename))
+            except:
+                # FIXME: Figure out what to do if we can't parse the file.
+                print "Error: failed to parse %s" % filename
+                self.test_doc is None
         else:
+            if self.filesystem.isdir(filename):
+                # FIXME: Figure out what is triggering this and what to do about it.
+                print "Error: trying to load %s, which is a directory" % filename
             self.test_doc = None
         self.ref_doc = None
 
     def analyze_test(self, test_contents=None, ref_contents=None):
         """ Analyzes a file to determine if it's a test, what type of test, and what reference or support files it requires. Returns all of the test info """
 
         test_info = None
 
         if test_contents is None and self.test_doc is None:
             return test_info
 
         if test_contents is not None:
             self.test_doc = Parser(test_contents)
 
         if ref_contents is not None:
             self.ref_doc = Parser(ref_contents)
 
         # First check if it's a reftest
 
         matches = self.reference_links_of_type('match') + self.reference_links_of_type('mismatch')
         if matches:
             if len(matches) > 1:
                 print 'Warning: Webkit does not support multiple references. Importing the first ref defined in ' + self.filesystem.basename(self.filename)
 
-            ref_file = self.filesystem.join(self.filesystem.dirname(self.filename), matches[0]['href'])
+            try:
+                ref_file = self.filesystem.join(self.filesystem.dirname(self.filename), matches[0]['href'])
+            except KeyError as e:
+                # FIXME: Figure out what to do w/ invalid test files.
+                print "Error: %s is has a reference link but is missing the 'href'" % (self.filesystem)

[ojan, 2013/05/28 23:45:54]

s/is has/has/

+                return None
+
             if self.ref_doc is None:
                 self.ref_doc = self.load_file(ref_file)
 
             test_info = {'test': self.filename, 'reference': ref_file}
 
             # If the ref file path is relative, we need to check it for
             # relative paths also because when it lands in WebKit, it will be
             # moved down into the test dir.
             #
             # Note: The test files themselves are not checked for support files
@@ skipping 46 matching lines @@
 
         src_paths = [src_tag['src'] for src_tag in elements_with_src_attributes]
         href_paths = [href_tag['href'] for href_tag in elements_with_href_attributes]
 
         paths = src_paths + href_paths + urls
         for path in paths:
             if not(path.startswith('http:')) and not(path.startswith('mailto:')):
                 support_files.append(path)
 
         return support_files