Git-Mirrors/brozzler
1
0
Fork 0
mirror of https://github.com/internetarchive/brozzler.git synced 2025-02-24 00:29:53 -05:00
Code Issues Actions Packages Projects Releases Wiki Activity

delay instagram youtube-dl captures; collapse if block

Browse Source
This commit is contained in:
Barbara Miller 2019-10-08 17:10:18 -07:00
parent 464562461c
commit c62c9f9063
1 changed files with 3 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
brozzler/ydl.py
View File

@ -154,7 +154,7 @@ def _build_youtube_dl(worker, destdir, site):
if ie_result.get('_type') == 'playlist':
self.logger.info(
'extractor %r found playlist in %s', ie.IE_NAME, url)
if ie.IE_NAME == 'youtube:playlist':
if ie.IE_NAME in {'youtube:playlist', 'soundcloud:user', 'instagram:user'}:
# At this point ie_result['entries'] is an iterator that
# will fetch more metadata from youtube to list all the
# videos. We unroll that iterator here partly because
@ -163,17 +163,9 @@ def _build_youtube_dl(worker, destdir, site):
ie_result['entries_no_dl'] = list(ie_result['entries'])
ie_result['entries'] = []
self.logger.info(
'not downloading %s videos from this youtube '
'not downloading %s media files from this '
'playlist because we expect to capture them from '
'individual watch pages',
len(ie_result['entries_no_dl']))
elif ie.IE_NAME == 'soundcloud:user':
ie_result['entries_no_dl'] = list(ie_result['entries'])
ie_result['entries'] = []
self.logger.info(
'not downloading %s tracks from this soundcloud '
'user page because we expect to capture them from '
'individual track pages',
'individual watch/track/detail pages',
len(ie_result['entries_no_dl']))
else:
self.logger.info(