"""
HTML Parser using BeautifulSoup
- Written by BroytMann. Copyright (C) 2007, 2008 PhiloSoft Design
+ Written by Broytman. Copyright (C) 2007-2010 PhiloSoft Design
"""
import re
from sgmllib import SGMLParser, SGMLParseError
-from HTMLParser import HTMLParser
from BeautifulSoup import BeautifulSoup, CData
-
-
-class BSoupParser(HTMLParser):
- def __init__(self, charset, meta, title, refresh, icon):
- object.__init__(self)
- self.charset = charset
- self.meta_charset = meta
- self.title = title
- self.refresh = refresh
- self.icon = icon
+from parse_html_util import HTMLParser
# http://groups.google.com/group/beautifulsoup/browse_thread/thread/69093cb0d3a3cf63
except SGMLParseError:
# Could not parse the DOCTYPE declaration
# Try to just skip the actual declaration
- match = re.search(r'<!DOCTYPE([^>]*?)>', self.rawdata[i:], re.MULTILINE)
+ match = re.search(r'<!DOCTYPE([^>]*?)>', self.rawdata[i:], re.MULTILINE|re.IGNORECASE)
if match:
toHandle = self.rawdata[i:match.end()]
else:
return j
-def parse_html(filename, charset=None):
+def _parse_html(filename, charset):
infile = open(filename, 'r')
try:
- root = BadDeclParser(infile, fromEncoding=charset)
+ return BadDeclParser(infile, fromEncoding=charset)
except TypeError:
return None
finally:
infile.close()
- try:
- head = root.html.head
- except AttributeError:
+def parse_html(filename, charset=None, log=None):
+ root = _parse_html(filename, charset)
+ if root is None:
return None
- if head is None:
- head = root.html # Some sites put TITLE in HTML without HEAD
-
_charset = root.originalEncoding
- try:
- title = head.title.string.encode(_charset)
- except AttributeError:
- title = '' # HEAD but no TITLE
+ if _charset in ("ISO-8859-2", "windows-1252", "MacCyrillic"): # Replace default
+ _charset = DEFAULT_CHARSET
+ root = _parse_html(filename, _charset)
+ if root is None:
+ return None
+
+ html = root.html
+ if html is None:
+ html = root
- if (not title) and (head is not root.html):
+ head = html.head
+ if head is None:
+ head = html # Some sites put TITLE in HTML without HEAD
+
+ title = head.title
+ if (title is None) and (html is not head):
# Some sites put TITLE in HTML outside of HEAD
+ title = html.title
+
+ if title is None:
+ # Lookup TITLE in the root
+ title = root.title
+ if title is None:
+ return None
+
+ if title.string:
+ title = title.string
+ else:
+ parts = []
+ for part in title:
+ if not isinstance(part, basestring):
+ part = unicode(part)
+ parts.append(part.strip())
+ title = ''.join(parts)
+
+ meta = head.find(_find_contenttype, recursive=False)
+ if meta:
try:
- title = root.html.title.string.encode(_charset)
- except AttributeError:
- title = '' # no TITLE in HTML too
+ meta_content = meta.get("content")
+ if meta_content:
+ __charset = meta_content.lower().split('charset=')[1].split(';')[0]
+ else:
+ __charset = False
+ except IndexError: # No charset in the META Content-Type
+ meta_charset = False
+ else:
+ meta_charset = _charset == __charset
+ else:
+ meta_charset = False
+
+ if charset or meta_charset:
+ title = title.encode(charset or meta_charset)
meta = head.find(_find_refresh, recursive=False)
if meta:
else:
icon = None
- return BSoupParser(_charset, _charset != charset, title, refresh, icon)
+ 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_refresh(Tag):
return (Tag.name == "meta") and \