Support <img> srcset attribute parsing.

doc/changelog.txt

@@ -6,6 +6,7 @@ Fixes:
 
 Features:
 - checking: Parse and check HTTP Link: headers.
+- checking: Support parsing of the <img> srcset attribute.
 
 
 9.1 "Don Jon" (released 30.3.2014)

linkcheck/htmlutil/linkparse.py

@@ -46,7 +46,7 @@ LinkTags = {
     'html':     [u'manifest'], # HTML5
     'iframe':   [u'src', u'longdesc'],
     'ilayer':   [u'background'],
-    'img':      [u'src', u'lowsrc', u'longdesc', u'usemap'],
+    'img':      [u'src', u'lowsrc', u'longdesc', u'usemap', u'srcset'],
     'input':    [u'src', u'usemap', u'formaction'],
     'ins':      [u'cite'],
     'isindex':  [u'action'],
@@ -269,6 +269,10 @@ class LinkFinder (TagFinder):
         elif attr == u'archive':
             for url in value.split(u','):
                 self.found_url(url, name, base)
+        elif attr == u'srcset':
+            for img_candidate in value.split(u','):
+                url = img_candidate.split()[0]
+                self.found_url(url, name, base)
         else:
             self.found_url(value, name, base)
 

tests/test_linkparser.py

@@ -75,6 +75,11 @@ class TestLinkparser (unittest.TestCase):
         url = u" alink "
         self._test_one_link(content % url, url)
 
+    def test_img_srcset_parsing(self):
+        content = u'<img srcset="%s 1x">'
+        url = u"imagesmall.jpg"
+        self._test_one_link(content % url, url)
+
     def test_form_parsing(self):
         # Test <form action> parsing
         content = u'<form action="%s">'