import re
from sgmllib import SGMLParser, SGMLParseError
from BeautifulSoup import BeautifulSoup, CData
+
from .bkmk_ph_util import HTMLParser
+from compat import string_type
-DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
+DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
# http://groups.google.com/group/beautifulsoup/browse_thread/thread/69093cb0d3a3cf63
+
+
class BadDeclParser(BeautifulSoup):
def parse_declaration(self, i):
"""Treat a bogus SGML declaration as raw data. Treat a CDATA
except TypeError:
return None
+
def parse_html(html_text, charset=None, log=None):
+ if not html_text:
+ return None
root = _parse_html(html_text, charset)
if root is None:
return None
_charset = root.originalEncoding
- if _charset in ("ISO-8859-2", "windows-1252", "MacCyrillic"): # Replace default
+ if _charset in ("ISO-8859-2", "windows-1252", "MacCyrillic"):
+ # Replace with default and re-parse
_charset = DEFAULT_CHARSET
root = _parse_html(html_text, _charset)
if root is None:
head = html.head
if head is None:
- head = html # Some sites put TITLE in HTML without HEAD
+ head = html # Some sites put TITLE in HTML without HEAD
title = head.title
if (title is None) and (html is not head):
else:
parts = []
for part in title:
- if not isinstance(part, basestring):
- part = unicode(part)
+ if not isinstance(part, string_type):
+ part = part.decode()
parts.append(part.strip())
title = ''.join(parts)
try:
meta_content = meta.get("content")
if meta_content:
- __charset = meta_content.lower().split('charset=')[1].split(';')[0]
+ __charset = meta_content.lower().split('charset=')[1].\
+ split(';')[0]
else:
__charset = False
- except IndexError: # No charset in the META Content-Type
+ except IndexError: # No charset in the META Content-Type
meta_charset = False
else:
meta_charset = _charset == __charset
if meta_content:
meta_charset = _charset = meta_content.lower()
- if title and (_charset or meta_charset):
- title = title.encode(_charset or meta_charset)
+ #if title and (_charset or meta_charset):
+ # title = title.encode(_charset or meta_charset)
meta = head.find(_find_refresh, recursive=False)
if meta:
return None
return HTMLParser(_charset, meta_charset, title, refresh, icon)
+
def _find_contenttype(Tag):
return (Tag.name == "meta") and \
(Tag.get("http-equiv", '').lower() == "content-type")
+
def _find_charset(Tag):
return (Tag.name == "meta") and Tag.get("charset", '')
+
def _find_refresh(Tag):
return (Tag.name == "meta") and \
(Tag.get("http-equiv", '').lower() == "refresh")
+
def _find_icon(Tag):
return (Tag.name == "link") and \
(Tag.get("rel", '').lower() in ('icon', 'shortcut icon'))