Update html5lib from 0.9x5 to 0.9x7.

included_dependencies/html5lib/__init__.py

@@ -20,4 +20,6 @@ from .serializer import serialize
 
 __all__ = ["HTMLParser", "parse", "parseFragment", "getTreeBuilder",
            "getTreeWalker", "serialize"]
-__version__ = "0.99999"
+
+# this has to be at the top level, see how setup.py parses this
+__version__ = "0.9999999"

included_dependencies/html5lib/sanitizer.py

@@ -207,8 +207,12 @@ class HTMLSanitizerMixin(object):
                                        unescape(attrs[attr])).lower()
                 # remove replacement characters from unescaped characters
                 val_unescaped = val_unescaped.replace("\ufffd", "")
-                uri = urlparse.urlparse(val_unescaped)
-                if uri:
+                try:
+                    uri = urlparse.urlparse(val_unescaped)
+                except ValueError:
+                    uri = None
+                    del attrs[attr]
+                if uri and uri.scheme:
                     if uri.scheme not in self.allowed_protocols:
                         del attrs[attr]
                     if uri.scheme == 'data':

included_dependencies/html5lib/treewalkers/etree.py

@@ -10,7 +10,7 @@ except ImportError:
 
 import re
 
-from six import text_type
+from six import string_types
 
 from . import _base
 from ..utils import moduleFactoryFactory
@@ -58,7 +58,7 @@ def getETreeBuilder(ElementTreeImplementation):
                 return _base.COMMENT, node.text
 
             else:
-                assert type(node.tag) == text_type, type(node.tag)
+                assert isinstance(node.tag, string_types), type(node.tag)
                 # This is assumed to be an ordinary element
                 match = tag_regexp.match(node.tag)
                 if match: