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 25b3bb04c55f654c9620bf4682fc0a6cec313eb0..3e20a5516a775eb8e425bdf932831bbcb559641f 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -1,83 +1,20 @@
 #! /usr/bin/env python
 """
-   HTML Parser
+   HTML Parsers wrapper
 
    Written by BroytMann. Copyright (C) 1997-2007 PhiloSoft Design
 """
 
+import codecs
+
 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
-
-
-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 = ''
-      self.icon = None
-
-   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
-
-
-   def do_link(self, attrs):
-      has_icon = False
-      href = None
-
-      for attrname, value in attrs:
-         if value:
-            value = value.strip().lower()
-            if (attrname == 'rel') and (value in ('icon', 'shortcut icon')):
-               has_icon = True
-            elif attrname == 'href':
-               href = value
-
-      if has_icon:
-         self.icon = href
-      else:
-         self.icon = None
+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
@@ -100,22 +37,7 @@ def parse_html(filename, charset=None, log=None):
       except (ValueError, LookupError):
          charset = None         # ...try charset from HTML
 
-   infile = open(filename, 'r')
-   parser = HTMLParser(charset)
-
-   for line in infile:
-      try:
-         parser.feed(line)
-      except (HTMLParseError, HTMLHeadDone):
-         break
-
-   infile.close()
-
-   try:
-      parser.close()
-   except (HTMLParseError, HTMLHeadDone):
-      pass
-
+   parser = _parse_html(filename, charset)
    title = parser.title
 
    if not parser.charset:
@@ -145,7 +67,8 @@ def parse_html(filename, charset=None, log=None):
 
 
 if __name__ == '__main__':
-   parser = parse_html(sys.argv[1])
+   import sys
+   parser = parse_html(sys.argv[1], current_charset)
    print parser.charset
    print parser.title
    print parser.refresh