+ _charset = root.originalEncoding
+ 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
+
+ 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 not None:
+ if title.string:
+ title = title.string.encode(_charset)
+ else:
+ parts = []
+ for part in title:
+ if not isinstance(part, basestring):
+ part = unicode(part)
+ parts.append(part.strip())
+ title = ''.join(parts).encode(_charset)
+
+ meta = head.find(_find_contenttype, recursive=False)
+ if meta:
+ try:
+ __charset = meta.get("content").lower().split('charset=')[1].split(';')[0]
+ except IndexError: # No charset in the META Content-Type
+ meta_charset = False
+ else:
+ meta_charset = _charset == __charset
+ else:
+ meta_charset = False
+
+ meta = head.find(_find_refresh, recursive=False)