Changeset 361 for trunk/genshi/input.py

Timestamp:

Oct 13, 2006, 3:42:38 PM (17 years ago)

Author:

cmlenz

Message:

Fix a bug in the XML parser, where attributes containing HTML entity references would get pulled out of the attribute value, and instead added as a text node just before the associated start tag. Thanks to Hamish Lawson for ​pointing out the problem.

File:

• trunk/genshi/input.py (modified) (5 diffs)

trunk/genshi/input.py

@@ -22 +22 @@
 from StringIO import StringIO
 
-from genshi.core import Attrs, QName, Stream
+from genshi.core import Attrs, QName, Stream, stripentities
 from genshi.core import DOCTYPE, START, END, START_NS, END_NS, TEXT, \
                         START_CDATA, END_CDATA, PI, COMMENT
@@ -71 +71 @@
     END root
     """
+
+    _entitydefs = ['<!ENTITY %s "&#%d;">' % (name, value) for name, value in
+                   htmlentitydefs.name2codepoint.items()]
+    _external_dtd = '\n'.join(_entitydefs)
 
     def __init__(self, source, filename=None):
@@ -101 +105 @@
         # (in _handle_other)
         parser.DefaultHandler = self._handle_other
+        parser.SetParamEntityParsing(expat.XML_PARAM_ENTITY_PARSING_ALWAYS)
         parser.UseForeignDTD()
+        parser.ExternalEntityRefHandler = self._build_foreign
 
         # Location reporting is only support in Python >= 2.4
@@ -142 +148 @@
         return iter(self.parse())
 
+    def _build_foreign(self, context, base, sysid, pubid):
+        parser = self.expat.ExternalEntityParserCreate(context)
+        parser.ParseFile(StringIO(self._external_dtd))
+        return 1
+
     def _enqueue(self, kind, data=None, pos=None):
         if pos is None:
@@ -293 +304 @@
             if value is None:
                 value = name
-            fixed_attrib.append((name, unicode(value)))
+            fixed_attrib.append((name, unicode(stripentities(value))))
 
         self._enqueue(START, (QName(tag), Attrs(fixed_attrib)))