Test meta charset by looking in META HTTP-Equiv.

git-svn-id: file:///home/phd/archive/SVN/bookmarks_db/trunk@187 fdd5c36f-1aea-0310-aeeb-c58d7e2b6c23

diff --git a/Robots/parse_html_beautifulsoup.py b/Robots/parse_html_beautifulsoup.py
index c7263fe64b478769e931ab978e4e382ed2233d3c..c45af2b0221a61f441cf5165fc2a25f6fe978bb5 100644 (file)
--- a/Robots/parse_html_beautifulsoup.py
+++ b/Robots/parse_html_beautifulsoup.py
@@ -83,6 +83,13 @@ def parse_html(filename, charset=None):
       except AttributeError:
          title = '' # no TITLE in HTML too
 
+   meta = head.find(_find_contenttype, recursive=False)
+   if meta:
+      __charset = meta.get("content").lower().split('charset=')[1].split(';')[0]
+      meta_charset = _charset == __charset
+   else:
+      meta_charset = False
+
    meta = head.find(_find_refresh, recursive=False)
    if meta:
       refresh = meta.get("content")
@@ -95,7 +102,11 @@ 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 \