Make sure lxml unesacpes xml names properly

jgraham · jgraham · commit 6cb006a08c18 · 2010-01-03T19:05:14.000+01:00
diff --git a/src/html5lib/ihatexml.py b/src/html5lib/ihatexml.py
@@ -72,18 +72,14 @@ def listToRegexpStr(charList):
     rv = []
     for item in charList:
         if item[0] == item[1]:
-           rv.append(intToUnicodeStr(item[0]))
+           rv.append(unichar(item[0]))
         else:
-            rv.append(intToUnicodeStr(item[0]) + "-" + intToUnicodeStr(item[1]))
-    return "[%s]"%"|".join(rv)
+            rv.append(unichar(item[0]) + "-" + unichar(item[1]))
+    return "[%s]"%"".join(rv)
 
 def hexToInt(hex_str):
     return int(hex_str, 16)
 
-def intToUnicodeStr(intValue):
-    #There must be a better (non-evil) way to do this
-    return escapeRegexp(eval(r"u'\u%s'"%hex(intValue)[2:].rjust(4, "0")))
-
 def escapeRegexp(string):
     specialCharacters = (".", "^", "$", "*", "+", "?", "{", "}",
                           "[", "]", "|", "(", ")", "-")
diff --git a/src/html5lib/treebuilders/etree_lxml.py b/src/html5lib/treebuilders/etree_lxml.py
@@ -203,12 +203,12 @@ def __init__(self, name, namespace):
                 self._attributes = Attributes(self)
 
             def _setName(self, name):
-                self._name = filter.coerceElement(name)                
+                self._name = filter.coerceElement(name)
                 self._element.tag = self._getETreeTag(
                     self._name, self._namespace)
         
             def _getName(self):
-                return self._name
+                return filter.fromXmlName(self._name)
         
             name = property(_getName, _setName)
 
