X-Git-Url: https://git.phdru.name/?a=blobdiff_plain;f=Robots%2Fparse_html.py;h=d3870fa210b8c3bdeefdac3c45d05aa13ec6cc1a;hb=cbe95b9700da2b5b75fa07312253736d5447d318;hp=350ad789b5fa34592db37ce6e097de98e8b8dd6a;hpb=62977582003da9f434cd4fd377837935b6f0520c;p=bookmarks_db.git

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 350ad78..d3870fa 100755
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -1,74 +1,26 @@
-#! /usr/local/bin/python -O
+#! /usr/bin/env python
 """
-   HTML Parser
+   HTML Parsers wrapper
 
-   Written by BroytMann. Copyright (C) 1997-2005 PhiloSoft Design
+   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
 
+parsers = []
+try:
+   from parse_html_beautifulsoup import parse_html
+except ImportError:
+   pass
+else:
+   parsers.append(parse_html)
 
-from HTMLParser import HTMLParseError
-from m_lib.net.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
+from parse_html_htmlparser import parse_html
+parsers.append(parse_html)
 
 
 import re
@@ -78,31 +30,30 @@ def recode_entities(title, charset):
    output = []
    for part in entity_re.split(title):
       if entity_re.match(part):
-         part = unichr(int(part[2:-1])).encode(charset, "replace")
+         try:
+            part = unichr(int(part[2:-1])).encode(charset)
+         except UnicodeEncodeError:
+            pass # Leave the entity as is
       output.append(part)
 
    return ''.join(output)
 
 
 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()
+         codecs.lookup(charset) # In case of unknown charset...
+      except (ValueError, LookupError):
+         charset = None         # ...try charset from HTML
 
-   try:
-      parser.close()
-   except (HTMLParseError, HTMLHeadDone):
-      pass
+   for p in parsers:
+      parser = p(filename, charset)
+      if parser:
+         break
+      else:
+         if log: log("Parser %s.%s failed, trying next one." % (p.__module__, p.__name__))
 
    title = parser.title
-
    if not parser.charset:
       try:
          unicode(title, "ascii")
@@ -117,21 +68,36 @@ def parse_html(filename, charset=None, log=None):
          if parser.meta_charset:
             if log: log("   META charset   : %s" % parser.charset)
          else:
-            if log: log("   charset        : %s" % parser.charset)
-         if log: log("   title          : %s" % title)
-         title = unicode(title, parser.charset, "replace").encode(current_charset, "replace")
+            if log: log("   HTTP charset   : %s" % parser.charset)
          if log: log("   current charset: %s" % current_charset)
+         if log: log("   title          : %s" % title)
+         save_title = title
+         try:
+            title = unicode(title, parser.charset).encode(current_charset)
+         except UnicodeError:
+            if parser.meta_charset and parser.charset.endswith("1252") and \
+                  not DEFAULT_CHARSET.endswith("1252") and (DEFAULT_CHARSET <> current_charset):
+               parser.charset = DEFAULT_CHARSET
+               if log: log("   incorrect conversion from cp1252, converting from %s" % DEFAULT_CHARSET)
+               title = unicode(save_title, DEFAULT_CHARSET, "replace").encode(current_charset, "replace")
+            else:
+               title = unicode(title, parser.charset, "replace").encode(current_charset, "replace")
          if log: log("   converted title: %s" % title)
       except LookupError:
          if log: log("   unknown charset: `%s' or `%s'" % (parser.charset, current_charset))
 
-   parser.title = recode_entities(title, current_charset)
+   title = recode_entities(title, current_charset)
+   parts = [s.strip() for s in title.replace('\r', '').split('\n')]
+   title = ' '.join([s for s in parts if s])
+   if log: log("   final title    : %s" % title)
+   parser.title = title
    return parser
 
 
 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