Some sites put TITLE in HTML without HEAD.

[bookmarks_db.git] / Robots / parse_html_beautifulsoup.py

diff --git a/Robots/parse_html_beautifulsoup.py b/Robots/parse_html_beautifulsoup.py
index c8e8f5db54af91c96a439584d09ade66b094de46..209486c4054f015a179d4469ccaa5adc3f8bbe8f 100644 (file)
--- a/Robots/parse_html_beautifulsoup.py
+++ b/Robots/parse_html_beautifulsoup.py
@@ -4,8 +4,10 @@
    Written by BroytMann. Copyright (C) 2007 PhiloSoft Design
 """
 
+import re
+from sgmllib import SGMLParser, SGMLParseError
 from HTMLParser import HTMLParser
-from BeautifulSoup import BeautifulSoup
+from BeautifulSoup import BeautifulSoup, CData
 
 
 class BSoupParser(HTMLParser):
@@ -18,38 +20,71 @@ class BSoupParser(HTMLParser):
       self.icon = icon
 
 
+# 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
+         declaration as a CData object."""
+         j = None
+         if self.rawdata[i:i+9] == '<![CDATA[':
+              k = self.rawdata.find(']]>', i)
+              if k == -1:
+                  k = len(self.rawdata)
+              data = self.rawdata[i+9:k]
+              j = k+3
+              self._toStringSubclass(data, CData)
+         else:
+             try:
+                 j = SGMLParser.parse_declaration(self, i)
+             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)
+                 if match:
+                     toHandle = self.rawdata[i:match.end()]
+                 else:
+                     toHandle = self.rawdata[i:]
+                 self.handle_data(toHandle)
+                 j = i + len(toHandle)
+         return j
+
+
 def parse_html(filename, charset=None):
    infile = open(filename, 'r')
-   root = BeautifulSoup(infile, fromEncoding=charset)
-   infile.close()
+   try:
+      root = BadDeclParser(infile, fromEncoding=charset)
+   except TypeError:
+      return None
+   finally:
+      infile.close()
 
-   _charset = root.originalEncoding
    try:
-      title = root.html.head.title.string.encode(_charset)
+      head = root.html.head
    except AttributeError:
-      title = ''
+      return None
 
+   if head is None:
+      head = root.html # Some sites put TITLE in HTML without HEAD
+
+   _charset = root.originalEncoding
    try:
-      meta = root.html.head.find(_find_refresh, recursive=False)
+      title = head.title.string.encode(_charset)
    except AttributeError:
-      refresh = None
+      title = '' # HEAD but no TITLE
+
+   meta = head.find(_find_refresh, recursive=False)
+   if meta:
+      refresh = meta.get("content")
    else:
-      if meta:
-         refresh = meta.get("content")
-      else:
-         refresh = None
+      refresh = None
 
-   try:
-      meta = root.html.head.find(_find_icon, recursive=False)
-   except AttributeError:
-      icon = None
+   meta = head.find(_find_icon, recursive=False)
+   if meta:
+      icon = meta.get("href")
    else:
-      if meta:
-         icon = meta.get("href")
-      else:
-         icon = None
+      icon = None
 
-   return BSoupParser(_charset, _charset == charset, title, refresh, icon)
+   return BSoupParser(_charset, _charset != charset, title, refresh, icon)
 
 def _find_refresh(Tag):
    return (Tag.name == "meta") and \