Recode HTML entities.

git-svn-id: file:///home/phd/archive/SVN/bookmarks_db/trunk@168 fdd5c36f-1aea-0310-aeeb-c58d7e2b6c23

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 9f924c49597a3f4436fd477363ee7cd1fe1b5c32..1125dcc7170378d4b83f316e14a29465bb6451ed 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -24,19 +24,29 @@ parsers.append(parse_html)
 
 
 import re
-entity_re = re.compile("(&#[0-9]+;)")
+from htmlentitydefs import entitydefs
+
+entity_re = re.compile("(&\w+;)")
+num_entity_re = re.compile("(&#[0-9]+;)")
 
 def recode_entities(title, charset):
    output = []
-   for part in entity_re.split(title):
-      if entity_re.match(part):
+
+   for part in num_entity_re.split(title):
+      if num_entity_re.match(part):
          try:
             part = unichr(int(part[2:-1])).encode(charset)
          except UnicodeEncodeError:
             pass # Leave the entity as is
       output.append(part)
 
-   return ''.join(output)
+   output2 = []
+   for part in entity_re.split(''.join(output)):
+      if entity_re.match(part):
+         part = entitydefs.get(part[1:-1], part)
+      output2.append(part)
+
+   return ''.join(output2)
 
 
 def parse_html(filename, charset=None, log=None):