aboutsummaryrefslogtreecommitdiff
path: root/youtube_dl/extractor/common.py
diff options
context:
space:
mode:
authorJaime Marquínez Ferrándiz <jaime.marquinez.ferrandiz@gmail.com>2013-07-12 19:00:19 +0200
committerJaime Marquínez Ferrándiz <jaime.marquinez.ferrandiz@gmail.com>2013-07-12 22:12:04 +0200
commit46720279c28afb646b6ac19bcb11e85bb4bea726 (patch)
treec38b786e0176511bfbe5295a8a73582b121018fc /youtube_dl/extractor/common.py
parentd8269e1dfbe06d4b373211e54476bcd326f69abe (diff)
InfoExtractor: add some helper methods to extract OpenGraph info
Diffstat (limited to 'youtube_dl/extractor/common.py')
-rw-r--r--youtube_dl/extractor/common.py24
1 files changed, 24 insertions, 0 deletions
diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py
index 1bd5538ca..0a0c4047d 100644
--- a/youtube_dl/extractor/common.py
+++ b/youtube_dl/extractor/common.py
@@ -257,6 +257,30 @@ class InfoExtractor(object):
return (username, password)
+ # Helper functions for extracting OpenGraph info
+ @staticmethod
+ def _og_regex(property):
+ return r'<meta.+?property=[\'"]og:%s[\'"].+?content=(?:"(.+?)"|\'(.+?)\')' % property
+
+ def _og_search_property(self, property, html, name=None, **kargs):
+ if name is None:
+ name = 'OpenGraph %s' % property
+ return self._html_search_regex(self._og_regex(property), html, name, **kargs)
+
+ def _og_search_thumbnail(self, html, **kargs):
+ return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
+
+ def _og_search_description(self, html, **kargs):
+ return self._og_search_property('description', html, fatal=False, **kargs)
+
+ def _og_search_title(self, html, **kargs):
+ return self._og_search_property('title', html, **kargs)
+
+ def _og_search_video_url(self, html, name='video url', **kargs):
+ return self._html_search_regex([self._og_regex('video:secure_url'),
+ self._og_regex('video')],
+ html, name, **kargs)
+
class SearchInfoExtractor(InfoExtractor):
"""
Base class for paged search queries extractors.