X-Git-Url: https://git.phdru.name/?a=blobdiff_plain;f=Robots%2Fparse_html_beautifulsoup.py;h=47ecbaf459f367a9a80d535af1add172709dcf0d;hb=0e76f1851882b99da63a7c8a9e4cdf0c4a48657f;hp=3d5c44a3f4464c699bd2b2f6090963d936944ae6;hpb=276cd9e51625c346acb375fcc1be52ea6bbc11b1;p=bookmarks_db.git

diff --git a/Robots/parse_html_beautifulsoup.py b/Robots/parse_html_beautifulsoup.py
index 3d5c44a..47ecbaf 100644
--- a/Robots/parse_html_beautifulsoup.py
+++ b/Robots/parse_html_beautifulsoup.py
@@ -1,7 +1,7 @@
 """
    HTML Parser using BeautifulSoup
 
-   Written by BroytMann. Copyright (C) 2007 PhiloSoft Design
+   Written by Broytman. Copyright (C) 2007, 2008 PhiloSoft Design
 """
 
 import re
@@ -39,7 +39,7 @@ class BadDeclParser(BeautifulSoup):
              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:
@@ -49,36 +49,65 @@ class BadDeclParser(BeautifulSoup):
          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
 
-   if not title:
-      head = root.html # Some sites put TITLE in HTML outside of HEAD
+   html = root.html
+   if html is None:
+      html = root
 
-   try:
-      title = head.title.string.encode(_charset)
-   except AttributeError:
-      title = '' # HEAD but no TITLE
+   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)
    if meta:
@@ -92,7 +121,11 @@ def parse_html(filename, charset=None):
    else:
       icon = None
 
-   return BSoupParser(_charset, _charset != charset, title, refresh, icon)
+   return BSoupParser(_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 \