Calculate if the charset came from HTTP or from HTML meta.

[bookmarks_db.git] / Robots / parse_html.py

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index a534e21fa3a13379c9ef85801b027dbfec8669d4..3e20a5516a775eb8e425bdf932831bbcb559641f 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -1,66 +1,20 @@
-#! /usr/local/bin/python -O
+#! /usr/bin/env python
 """
-   HTML Parser
+   HTML Parsers wrapper
 
-   Written by BroytMann, Jun 2002 - May 2003. Copyright (C) 1997-2003 PhiloSoft Design
+   Written by BroytMann. Copyright (C) 1997-2007 PhiloSoft Design
 """
 
+import codecs
 
-import sys
-current_charset = sys.getdefaultencoding()
-DEFAULT_CHARSET = "windows-1251" # Stupid default for Russian Cyrillic
+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.www.html import HTMLParser as _HTMLParser
-
-
-class HTMLHeadDone(Exception): pass
-
-
-class HTMLParser(_HTMLParser):
-   def __init__(self, charset=None):
-      _HTMLParser.__init__(self)
-      self.charset = charset
-      self.meta_charset = 0
-      self.title = ''
-      self.refresh = ''
-
-   def end_head(self):
-      raise HTMLHeadDone()
-
-
-   def do_meta(self, attrs):
-      http_equiv = ""
-      content = ""
-
-      for attrname, value in attrs:
-         if value:
-            value = value.strip()
-            if attrname == 'http-equiv':
-               http_equiv = value.lower()
-            elif attrname == 'content':
-               content = value
-
-      if (not self.charset) and (http_equiv == "content-type"):
-         try:
-            # extract charset from "text/html; foo; charset=UTF-8; bar;"
-            self.charset = content.lower().split('charset=')[1].split(';')[0]
-            self.meta_charset = 1 # Remember that the charset was retrieved from
-                                  # META tag, not from the Content-Type header
-         except IndexError:
-            pass
-
-      if http_equiv == "refresh":
-         self.refresh = content
-
-
-   def start_title(self, attrs):
-      self.accumulator = ''
-
-   def end_title(self):
-      if not self.title: # use only the first title
-         self.title = self.accumulator
+try:
+   from parse_html_beautifulsoup import parse_html as _parse_html
+except ImportError:
+   from parse_html_htmlparser import parse_html as _parse_html
 
 
 import re
@@ -77,33 +31,24 @@ def recode_entities(title, charset):
 
 
 def parse_html(filename, charset=None, log=None):
-   infile = open(filename, 'r')
-   parser = HTMLParser(charset)
-
-   for line in infile:
+   if charset:
       try:
-         parser.feed(line)
-      except (HTMLParseError, HTMLHeadDone):
-         break
-
-   infile.close()
-
-   try:
-      parser.close()
-   except (HTMLParseError, HTMLHeadDone):
-      pass
+         codecs.lookup(charset) # In case of unknown charset...
+      except (ValueError, LookupError):
+         charset = None         # ...try charset from HTML
 
+   parser = _parse_html(filename, charset)
    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:
@@ -123,7 +68,8 @@ def parse_html(filename, charset=None, log=None):
 
 if __name__ == '__main__':
    import sys
-   parser = parse_html(sys.argv[1])
+   parser = parse_html(sys.argv[1], current_charset)
    print parser.charset
    print parser.title
    print parser.refresh
+   print parser.icon