commit 0ba9e3ca2233d018d695bac4eebe0e34043a7ec9
parent 4b537629143c8f51c5814c650227971c438b12e1
Author: remitamine <remitamine@gmail.com>
Date: Thu, 28 Apr 2016 17:44:33 +0100
[viewster] extract formats for videos with multiple audios/subtitles
Diffstat:
1 file changed, 78 insertions(+), 65 deletions(-)
diff --git a/youtube_dl/extractor/viewster.py b/youtube_dl/extractor/viewster.py
@@ -78,11 +78,11 @@ class ViewsterIE(InfoExtractor):
_ACCEPT_HEADER = 'application/json, text/javascript, */*; q=0.01'
- def _download_json(self, url, video_id, note='Downloading JSON metadata', fatal=True):
+ def _download_json(self, url, video_id, note='Downloading JSON metadata', fatal=True, query={}):
request = sanitized_Request(url)
request.add_header('Accept', self._ACCEPT_HEADER)
request.add_header('Auth-token', self._AUTH_TOKEN)
- return super(ViewsterIE, self)._download_json(request, video_id, note, fatal=fatal)
+ return super(ViewsterIE, self)._download_json(request, video_id, note, fatal=fatal, query=query)
def _real_extract(self, url):
video_id = self._match_id(url)
@@ -117,72 +117,85 @@ class ViewsterIE(InfoExtractor):
return self.playlist_result(entries, video_id, title, description)
formats = []
- manifest_url = None
- m3u8_formats = []
- for media_type in ('application/f4m+xml', 'application/x-mpegURL', 'video/mp4'):
- media = self._download_json(
- 'https://public-api.viewster.com/movies/%s/video?mediaType=%s'
- % (entry_id, compat_urllib_parse.quote(media_type)),
- video_id, 'Downloading %s JSON' % media_type, fatal=False)
- if not media:
- continue
- video_url = media.get('Uri')
- if not video_url:
- continue
- ext = determine_ext(video_url)
- if ext == 'f4m':
- manifest_url = video_url
- video_url += '&' if '?' in video_url else '?'
- video_url += 'hdcore=3.2.0&plugin=flowplayer-3.2.0.1'
- formats.extend(self._extract_f4m_formats(
- video_url, video_id, f4m_id='hds'))
- elif ext == 'm3u8':
- manifest_url = video_url
- m3u8_formats = self._extract_m3u8_formats(
- video_url, video_id, 'mp4', m3u8_id='hls',
- fatal=False) # m3u8 sometimes fail
- if m3u8_formats:
- formats.extend(m3u8_formats)
- else:
- qualities_basename = self._search_regex(
- '/([^/]+)\.csmil/',
- manifest_url, 'qualities basename', default=None)
- if not qualities_basename:
+ for language_set in info.get('LanguageSets', []):
+ manifest_url = None
+ m3u8_formats = []
+ audio = language_set.get('Audio') or ''
+ subtitle = language_set.get('Subtitle') or ''
+ base_format_id = audio
+ if subtitle:
+ base_format_id += '-%s' % subtitle
+
+ def concat(suffix, sep='-'):
+ return (base_format_id + '%s%s' % (sep, suffix)) if base_format_id else suffix
+
+ for media_type in ('application/f4m+xml', 'application/x-mpegURL', 'video/mp4'):
+ media = self._download_json(
+ 'https://public-api.viewster.com/movies/%s/video' % entry_id,
+ video_id, 'Downloading %s JSON' % concat(media_type, ' '), fatal=False, query={
+ 'mediaType': media_type,
+ 'language': audio,
+ 'subtitle': subtitle,
+ })
+ if not media:
continue
- QUALITIES_RE = r'((,\d+k)+,?)'
- qualities = self._search_regex(
- QUALITIES_RE, qualities_basename,
- 'qualities', default=None)
- if not qualities:
+ video_url = media.get('Uri')
+ if not video_url:
continue
- qualities = list(map(lambda q: int(q[:-1]), qualities.strip(',').split(',')))
- qualities.sort()
- http_template = re.sub(QUALITIES_RE, r'%dk', qualities_basename)
- http_url_basename = url_basename(video_url)
- if m3u8_formats:
- self._sort_formats(m3u8_formats)
- m3u8_formats = list(filter(
- lambda f: f.get('vcodec') != 'none' and f.get('resolution') != 'multiple',
- m3u8_formats))
- if len(qualities) == len(m3u8_formats):
- for q, m3u8_format in zip(qualities, m3u8_formats):
- f = m3u8_format.copy()
- f.update({
- 'url': video_url.replace(http_url_basename, http_template % q),
- 'format_id': f['format_id'].replace('hls', 'http'),
- 'protocol': 'http',
- })
- formats.append(f)
+ ext = determine_ext(video_url)
+ if ext == 'f4m':
+ manifest_url = video_url
+ video_url += '&' if '?' in video_url else '?'
+ video_url += 'hdcore=3.2.0&plugin=flowplayer-3.2.0.1'
+ formats.extend(self._extract_f4m_formats(
+ video_url, video_id, f4m_id=concat('hds')))
+ elif ext == 'm3u8':
+ manifest_url = video_url
+ m3u8_formats = self._extract_m3u8_formats(
+ video_url, video_id, 'mp4', m3u8_id=concat('hls'),
+ fatal=False) # m3u8 sometimes fail
+ if m3u8_formats:
+ formats.extend(m3u8_formats)
else:
- for q in qualities:
- formats.append({
- 'url': video_url.replace(http_url_basename, http_template % q),
- 'ext': 'mp4',
- 'format_id': 'http-%d' % q,
- 'tbr': q,
- })
-
- if not formats and not info.get('LanguageSets') and not info.get('VODSettings'):
+ qualities_basename = self._search_regex(
+ '/([^/]+)\.csmil/',
+ manifest_url, 'qualities basename', default=None)
+ if not qualities_basename:
+ continue
+ QUALITIES_RE = r'((,\d+k)+,?)'
+ qualities = self._search_regex(
+ QUALITIES_RE, qualities_basename,
+ 'qualities', default=None)
+ if not qualities:
+ continue
+ qualities = list(map(lambda q: int(q[:-1]), qualities.strip(',').split(',')))
+ qualities.sort()
+ http_template = re.sub(QUALITIES_RE, r'%dk', qualities_basename)
+ http_url_basename = url_basename(video_url)
+ if m3u8_formats:
+ self._sort_formats(m3u8_formats)
+ m3u8_formats = list(filter(
+ lambda f: f.get('vcodec') != 'none' and f.get('resolution') != 'multiple',
+ m3u8_formats))
+ if len(qualities) == len(m3u8_formats):
+ for q, m3u8_format in zip(qualities, m3u8_formats):
+ f = m3u8_format.copy()
+ f.update({
+ 'url': video_url.replace(http_url_basename, http_template % q),
+ 'format_id': f['format_id'].replace('hls', 'http'),
+ 'protocol': 'http',
+ })
+ formats.append(f)
+ else:
+ for q in qualities:
+ formats.append({
+ 'url': video_url.replace(http_url_basename, http_template % q),
+ 'ext': 'mp4',
+ 'format_id': 'http-%d' % q,
+ 'tbr': q,
+ })
+
+ if not formats and not info.get('VODSettings'):
self.raise_geo_restricted()
self._sort_formats(formats)
