w/youtube-dl
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

[generic] Prefer enclosures over links in RSS feeds

Browse source
This commit is contained in:
Bastian de Groot 2018-04-29 17:14:37 +02:00 committed by Sergey M
parent 12b0d4e0e1
commit 01aec84880
1 changed files with 18 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

16
youtube_dl/extractor/generic.py
View file

@ -191,6 +191,16 @@ class GenericIE(InfoExtractor):
'title': 'pdv_maddow_netcast_m4v-02-27-2015-201624', 'title': 'pdv_maddow_netcast_m4v-02-27-2015-201624',
} }
}, },
# RSS feed with enclosures and unsupported link URLs
{
'url': 'http://www.hellointernet.fm/podcast?format=rss',
'info_dict': {
'id': 'http://www.hellointernet.fm/podcast?format=rss',
'description': 'CGP Grey and Brady Haran talk about YouTube, life, work, whatever.',
'title': 'Hello Internet',
},
'playlist_mincount': 100,
},
# SMIL from http://videolectures.net/promogram_igor_mekjavic_eng # SMIL from http://videolectures.net/promogram_igor_mekjavic_eng
{ {
'url': 'http://videolectures.net/promogram_igor_mekjavic_eng/video/1/smil.xml', 'url': 'http://videolectures.net/promogram_igor_mekjavic_eng/video/1/smil.xml',
@ -2026,14 +2036,16 @@ class GenericIE(InfoExtractor):
entries = [] entries = []
for it in doc.findall('./channel/item'): for it in doc.findall('./channel/item'):
next_url = xpath_text(it, 'link', fatal=False) next_url = None
if not next_url:
enclosure_nodes = it.findall('./enclosure') enclosure_nodes = it.findall('./enclosure')
for e in enclosure_nodes: for e in enclosure_nodes:
next_url = e.attrib.get('url') next_url = e.attrib.get('url')
if next_url: if next_url:
break break
if not next_url:
next_url = xpath_text(it, 'link', fatal=False)
if not next_url: if not next_url:
continue continue