Work around some issues in python2 due to its piss-poor unicode implementation ( can be seen, for example, by using ↑ as a value in a tag ). For proper unicode support you should use python3, but this at least makes python2 more on-par with python3.

Author: kata198

AdvancedHTMLParser/SpecialAttributes.py

@@ -7,7 +7,7 @@
 
 from collections import OrderedDict
 
-from .utils import escapeQuotes
+from .utils import escapeQuotes, tostr
 
 __all__ = ('SpecialAttributesDict', 'AttributeNode', 'AttributeNodeMap', 'StyleAttribute' )
 
@@ -214,7 +214,7 @@ def X__setitem__(self, name, value):
 
 
     def __str__(self):
-        return '[ %s ]' %(' '.join([str(self.getNamedItem(name)) for name in self._attributesDict.keys()]))
+        return '[ %s ]' %(' '.join([tostr(self.getNamedItem(name)) for name in self._attributesDict.keys()]))
 
 
 
@@ -232,7 +232,7 @@ def __init__(self, styleValue, tag=None):
             @param styleValue <str> - A style string ( like "display: none; padding-top: 5px" )
         '''
         if isinstance(styleValue, StyleAttribute):
-            styleValue = str(styleValue)
+            styleValue = tostr(styleValue)
 
         self._styleValue = styleValue
         self._styleDict = StyleAttribute.styleToDict(styleValue)

AdvancedHTMLParser/Tags.py

@@ -10,7 +10,7 @@
 from .constants import PREFORMATTED_TAGS, IMPLICIT_SELF_CLOSING_TAGS, TAG_NAMES_TO_ADDITIONAL_ATTRIBUTES, COMMON_JAVASCRIPT_ATTRIBUTES, ALL_JAVASCRIPT_EVENT_ATTRIBUTES, TAG_ITEM_BINARY_ATTRIBUTES, TAG_ITEM_ATTRIBUTE_LINKS, TAG_ITEM_ATTRIBUTES_SPECIAL_VALUES, TAG_ITEM_CHANGE_NAME_FROM_ATTR, TAG_ITEM_CHANGE_NAME_FROM_ITEM
 from .SpecialAttributes import SpecialAttributesDict, StyleAttribute, AttributeNodeMap
 
-from .utils import escapeQuotes
+from .utils import escapeQuotes, tostr
 
 __all__ = ('AdvancedTag', 'uniqueTags', 'TagCollection', 'FilterableTagCollection', 'toggleAttributesDOM', 'isTextNode', 'isTagNode')
 
@@ -143,7 +143,7 @@ def __setattr__(self, name, value):
                     self.setAttribute(name, "")
                 return value
 
-            self.setAttribute(name, str(value))
+            self.setAttribute(name, tostr(value))
             return value
 
         if name == 'style' and not isinstance(value, StyleAttribute):
@@ -855,7 +855,7 @@ def getStartTag(self):
         '''
         attributeString = []
         for name, val in self._attributes.items():
-            val = str(val)
+            val = tostr(val)
             if val:
                 val = escapeQuotes(val)
                 attributeString.append('%s="%s"' %(name, val) )
@@ -932,7 +932,7 @@ def getAttributesList(self):
 
                 This is suitable for passing back into AdvancedTag when creating a new tag.
         '''
-        return [ (str(name)[:], str(value)[:]) for name, value in self._attributes.items() ]
+        return [ (tostr(name)[:], tostr(value)[:]) for name, value in self._attributes.items() ]
 
     def getAttributesDict(self):
         '''
@@ -944,7 +944,7 @@ def getAttributesDict(self):
               @return <dict ( str(name), str(value) )> - A dict of attrName to attrValue , all as strings and copies.
         '''
 
-        return { str(name)[:] : str(value)[:] for name, value in self._attributes.items() }
+        return { tostr(name)[:] : tostr(value)[:] for name, value in self._attributes.items() }
 
     def setAttribute(self, attrName, attrValue):
         '''

AdvancedHTMLParser/utils.py

@@ -1,10 +1,11 @@
 # Copyright (c) 2015, 2017 Tim Savannah  under terms of LGPLv3
 #  Some misc utils and regular expressions
 
+import sys
 import re
 
 __all__ = ('IE_CONDITIONAL_PATTERN', 'END_HTML', 'START_HTML', 'DOCTYPE_MATCH', 
-    'stripIEConditionals', 'addStartTag', 'escapeQuotes', 'unescapeQuotes')
+    'stripIEConditionals', 'addStartTag', 'escapeQuotes', 'unescapeQuotes', 'tostr')
 
 IE_CONDITIONAL_PATTERN = re.compile('[<][!][-][-][ \t\r\n]*[\[][ \t\r\n]*if.*-->', re.MULTILINE)
 
@@ -70,4 +71,12 @@ def unescapeQuotes(value):
     '''
     return value.replace('&quot;', '"')
 
-
+if sys.version_info.major < 3:
+    def tostr(value):
+        if not isinstance(value, (str, unicode)):
+            value = unicode(value)
+        return value
+else:
+    def tostr(value):
+        return str(value)
+