- 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
- _charset = DEFAULT_CHARSET
- root = _parse_html(html_text, _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
- 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:
- 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 not meta_charset:
- meta = head.find(_find_charset, recursive=False)
- if meta:
- meta_content = meta.get("charset")
- if meta_content:
- meta_charset = _charset = meta_content.lower()
-
- if title and (_charset or meta_charset):
- title = title.encode(_charset or meta_charset)
-
- meta = head.find(_find_refresh, recursive=False)
- if meta:
- refresh = meta.get("content")
- else:
- refresh = None
-
- meta = head.find(_find_icon, recursive=False)
- if meta:
- icon = meta.get("href")
- else:
- icon = None
-
- if (title is None) and (refresh is None) and (icon is None):
- return None
- return HTMLParser(_charset, meta_charset, title, refresh, icon)
+ 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 with default and re-parse
+ _charset = DEFAULT_CHARSET
+ root = _parse_html(html_text, _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
+ else:
+ parts = []
+ for part in title:
+ if not isinstance(part, string_type):
+ part = part.decode()
+ parts.append(part.strip())
+ title = ''.join(parts)
+
+ meta = head.find(_find_contenttype, recursive=False)
+ if meta:
+ try:
+ 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 not meta_charset:
+ meta = head.find(_find_charset, recursive=False)
+ if meta:
+ meta_content = meta.get("charset")
+ if meta_content:
+ meta_charset = _charset = meta_content.lower()
+
+ #if title and (_charset or meta_charset):
+ # title = title.encode(_charset or meta_charset)
+
+ meta = head.find(_find_refresh, recursive=False)
+ if meta:
+ refresh = meta.get("content")
+ else:
+ refresh = None
+
+ meta = head.find(_find_icon, recursive=False)
+ if meta:
+ icon = meta.get("href")
+ else:
+ icon = None
+
+ if (title is None) and (refresh is None) and (icon is None):
+ return None
+ return HTMLParser(_charset, meta_charset, title, refresh, icon)
+