[InfoExtractor] Handle unquoted values in OpenGraph searches

This commit is contained in:
dirkf 2023-02-14 02:47:09 +00:00
parent 6f8c2635a5
commit 42b098dd79
2 changed files with 3 additions and 1 deletions

View file

@ -62,6 +62,7 @@ class TestInfoExtractor(unittest.TestCase):
<meta name="og:test1" content='foo > < bar'/>
<meta name="og:test2" content="foo >//< bar"/>
<meta property=og-test3 content='Ill-formatted opengraph'/>
<meta property=og:test4 content=unquoted-value/>
'''
self.assertEqual(ie._og_search_title(html), 'Foo')
self.assertEqual(ie._og_search_description(html), 'Some video\'s description ')
@ -74,6 +75,7 @@ class TestInfoExtractor(unittest.TestCase):
self.assertEqual(ie._og_search_property(('test0', 'test1'), html), 'foo > < bar')
self.assertRaises(RegexNotFoundError, ie._og_search_property, 'test0', html, None, fatal=True)
self.assertRaises(RegexNotFoundError, ie._og_search_property, ('test0', 'test00'), html, None, fatal=True)
self.assertEqual(ie._og_search_property('test4', html), 'unquoted-value')
def test_html_search_meta(self):
ie = self.ie

View file

@ -1087,7 +1087,7 @@ class InfoExtractor(object):
# Helper functions for extracting OpenGraph info
@staticmethod
def _og_regexes(prop):
content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))'
content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?)(?=\s|/?>))'
property_re = (r'(?:name|property)=(?:\'og[:-]%(prop)s\'|"og[:-]%(prop)s"|\s*og[:-]%(prop)s\b)'
% {'prop': re.escape(prop)})
template = r'<meta[^>]+?%s[^>]+?%s'