Do not even open non-image files; report the content type of a non-image file.

[bookmarks_db.git] / Robots / parse_html.py

diff --git a/Robots/parse_html.py b/Robots/parse_html.py
index 90b777b4e97a2754c2bfc33d2e4f9ff966f6d0b5..b30c66458eaa32960b5badce9bc849a2082213d7 100755 (executable)
--- a/Robots/parse_html.py
+++ b/Robots/parse_html.py
@@ -71,15 +71,13 @@ class HTMLParser(_HTMLParser):
       for attrname, value in attrs:
          if value:
             value = value.strip().lower()
-            if (attrname == 'rel') and (value in ('icon', 'shortcut icon')):
+            if (attrname == 'rel') and (value.lower() in ('icon', 'shortcut icon')):
                has_icon = True
             elif attrname == 'href':
                href = value
 
       if has_icon:
          self.icon = href
-      else:
-         self.icon = None
 
 
 import re
@@ -147,6 +145,7 @@ def parse_html(filename, charset=None, log=None):
 
 
 if __name__ == '__main__':
+   import sys
    parser = parse_html(sys.argv[1])
    print parser.charset
    print parser.title