If sys.getdefaultencoding() returns "ascii" - use

locale.getpreferredencoding() to get user's charset.
Convert charset names to the canonical form:
replace "windows-" by "cp", and make the name lowercase.

git-svn-id: file:///home/phd/archive/SVN/bookmarks_db/trunk@51 fdd5c36f-1aea-0310-aeeb-c58d7e2b6c23

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index ffacbee521ddecd0553008a158db163d07487503..350ad789b5fa34592db37ce6e097de98e8b8dd6a 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -2,13 +2,21 @@
 """
    HTML Parser
 
-   Written by BroytMann, Jun 2002 - May 2003. Copyright (C) 1997-2003 PhiloSoft Design
+   Written by BroytMann. Copyright (C) 1997-2005 PhiloSoft Design
 """
 
 
 import sys
 current_charset = sys.getdefaultencoding()
-DEFAULT_CHARSET = "windows-1251" # Stupid default for Russian Cyrillic
+if current_charset == "ascii":
+   try:
+      import locale
+   except ImportError:
+      pass
+   else:
+      current_charset = locale.getpreferredencoding()
+current_charset = current_charset.replace("windows-", "cp").lower()
+DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
 
 
 from HTMLParser import HTMLParseError
@@ -96,14 +104,14 @@ def parse_html(filename, charset=None, log=None):
    title = parser.title
 
    if not parser.charset:
-      ascii = 1
-      for c in title:
-         if not (32 <= ord(c) <= 127): # non-ASCII character
-            ascii = 0
-            break
-      if not ascii:
+      try:
+         unicode(title, "ascii")
+      except UnicodeDecodeError:
          parser.charset = DEFAULT_CHARSET
 
+   if parser.charset:
+      parser.charset = parser.charset.replace("windows-", "cp").lower()
+
    if parser.charset and (parser.charset <> current_charset):
       try:
          if parser.meta_charset: