Made tagNameState faster (saves ~3% overall)

philiptaylor · philiptaylor · commit c6f09403a5fe · 2008-12-18T16:57:28.000Z
--HG--
extra : convert_revision : svn%3Aacbfec75-9323-0410-a652-858a13e371e0/trunk%401244
diff --git a/src/html5lib/tokenizer.py b/src/html5lib/tokenizer.py
@@ -489,9 +489,6 @@ def tagNameState(self):
         data = self.stream.char()
         if data in spaceCharacters:
             self.state = self.states["beforeAttributeName"]
-        elif data in asciiLetters:
-            self.currentToken["name"] += data +\
-              self.stream.charsUntil(asciiLetters, True)
         elif data == u">":
             self.emitCurrentToken()
         elif data is EOF:
@@ -503,6 +500,8 @@ def tagNameState(self):
                 self.state = self.states["beforeAttributeName"]
         else:
             self.currentToken["name"] += data
+            # (Don't use charsUntil here, because tag names are
+            # very short and it's faster to not do anything fancy)
         return True
 
     def beforeAttributeNameState(self):
