WIP; Backported features from stable branch

buriy · buriy · commit 5855beb32a5a · 2014-09-22T15:29:49.000+07:00
diff --git a/readability/readability.py b/readability/readability.py
@@ -435,6 +435,7 @@ def sanitize(self, node, candidates):
                 for kind in ['p', 'img', 'li', 'a', 'embed', 'input']:
                     counts[kind] = len(el.findall('.//%s' % kind))
                 counts["li"] -= 100
+                counts["input"] -= len(el.findall('.//input[@type="hidden"]'))
 
                 # Count the text length excluding any surrounding whitespace
                 content_length = text_length(el)
@@ -590,6 +591,9 @@ def main():
     parser.add_option('-b', '--browser', default=None, action='store_true', help="open in browser")
     parser.add_option('-l', '--log', default=None, help="use filename for logs (appended)")
     parser.add_option('-u', '--url', default=None, help="use URL instead of a local file")
+    parser.add_option('-s', '--show-xpath', default=None, help="show xpath")
+    parser.add_option('-x', '--xpath', default=None, help="use xpath")
+    parser.add_option('-t', '--support-text', default=None, help="use this support text")
     parser.add_option('-p', '--positive-keywords', default=None, help="positive keywords (separated with comma)", action='store')
     parser.add_option('-n', '--negative-keywords', default=None, help="negative keywords (separated with comma)", action='store')
     (options, args) = parser.parse_args()
diff --git a/setup.py b/setup.py
@@ -12,7 +12,7 @@
 
 setup(
     name="readability-lxml",
-    version="0.4.0.b1",
+    version="0.5.0.2",
     author="Yuri Baburov",
     author_email="burchik@gmail.com",
     description="fast python port of arc90's readability tool",
