X-Git-Url: https://git.phdru.name/?a=blobdiff_plain;f=Robots%2Fparse_html.py;h=b30c66458eaa32960b5badce9bc849a2082213d7;hb=76e0bd7defe7af4337fb1857141b7ef3e9f3f634;hp=57ad6a0cf69c89247281700915ced9854a3ef8ba;hpb=bf97248fba49413a16c41152bdb0722ce157164c;p=bookmarks_db.git

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 57ad6a0..b30c664 100755
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -5,20 +5,12 @@
    Written by BroytMann. Copyright (C) 1997-2007 PhiloSoft Design
 """
 
+import codecs
 
-import sys
-current_charset = sys.getdefaultencoding()
-if current_charset == "ascii":
-   try:
-      import locale
-   except ImportError:
-      pass
-   else:
-      current_charset = locale.getpreferredencoding()
-current_charset = current_charset.replace("windows-", "cp").lower()
+from m_lib.defenc import default_encoding
+current_charset = default_encoding.replace("windows-", "cp")
 DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
 
-
 from HTMLParser import HTMLParseError
 from m_lib.net.www.html import HTMLParser as _HTMLParser
 
@@ -33,6 +25,7 @@ class HTMLParser(_HTMLParser):
       self.meta_charset = 0
       self.title = ''
       self.refresh = ''
+      self.icon = None
 
    def end_head(self):
       raise HTMLHeadDone()
@@ -78,15 +71,13 @@ class HTMLParser(_HTMLParser):
       for attrname, value in attrs:
          if value:
             value = value.strip().lower()
-            if (attrname == 'rel') and (value in ('icon', 'shortcut icon')):
+            if (attrname == 'rel') and (value.lower() in ('icon', 'shortcut icon')):
                has_icon = True
             elif attrname == 'href':
                href = value
 
       if has_icon:
          self.icon = href
-      else:
-         self.icon = None
 
 
 import re
@@ -103,6 +94,12 @@ def recode_entities(title, charset):
 
 
 def parse_html(filename, charset=None, log=None):
+   if charset:
+      try:
+         codecs.lookup(charset) # In case of unknown charset...
+      except (ValueError, LookupError):
+         charset = None         # ...try charset from HTML
+
    infile = open(filename, 'r')
    parser = HTMLParser(charset)