Charset was guessed if it is not from META and not from HTTP.

[bookmarks_db.git] / Robots / parse_html.py

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 0d7b20b5ce82b88d596b1cb25aac90084a997735..ec93ea5acb0903e9557bc370559a0a0ebee4dd6e 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -24,7 +24,7 @@ parsers.append(parse_html)
 
 
 import re
-from htmlentitydefs import entitydefs
+from htmlentitydefs import name2codepoint
 
 entity_re = re.compile("(&\w+;)")
 num_entity_re = re.compile("(&#[0-9]+;)")
@@ -32,20 +32,22 @@ 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):
-         part = entitydefs.get(part[1:-1], part)
+      if part not in ("&", "<", ">", "&quote;") and \
+            entity_re.match(part):
+         part = unichr(name2codepoint.get(part[1:-1], part)).encode(charset)
       output.append(part)
+   title = ''.join(output)
 
-   output2 = []
-   for part in num_entity_re.split(''.join(output)):
+   output = []
+   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
-      output2.append(part)
+      output.append(part)
 
-   return ''.join(output2)
+   return ''.join(output)
 
 
 def parse_html(filename, charset=None, log=None):
@@ -55,10 +57,13 @@ def parse_html(filename, charset=None, log=None):
       except (ValueError, LookupError):
          charset = None         # ...try charset from HTML
 
-   for p in parsers:
-      charsets = [universal_charset, DEFAULT_CHARSET]
+   charsets = [universal_charset, DEFAULT_CHARSET]
+   if charset:
+      charset = charset.lower().replace("windows-", "cp")
       if charset not in charsets:
          charsets.insert(0, charset)
+
+   for p in parsers:
       parser = None
       for c in charsets:
          try:
@@ -79,13 +84,13 @@ def parse_html(filename, charset=None, log=None):
          parser.charset = DEFAULT_CHARSET
 
    if parser.charset:
-      parser.charset = parser.charset.replace("windows-", "cp").lower()
+      parser.charset = parser.charset.lower().replace("windows-", "cp")
 
    if parser.charset and (parser.charset <> universal_charset):
       try:
          if parser.meta_charset:
             if log: log("   META charset   : %s" % parser.charset)
-         else:
+         elif (not charset) or charset <> parser.charset:
             if log: log("   guessed charset: %s" % parser.charset)
          if log: log("   current charset: %s" % universal_charset)
          if log: log("   title          : %s" % title)