]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/viewster.py
New upstream version 2018.11.07
[youtubedl] / youtube_dl / extractor / viewster.py
index 7cf930d699fd75ad564ead644a4371d401df137a..6e318479c866288dc877fdf4731a3486308ac46c 100644 (file)
@@ -1,11 +1,11 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
 from ..compat import (
     compat_HTTPError,
-    compat_urllib_request,
-    compat_urllib_parse,
     compat_urllib_parse_unquote,
 )
 from ..utils import (
@@ -13,7 +13,9 @@ from ..utils import (
     ExtractorError,
     int_or_none,
     parse_iso8601,
+    sanitized_Request,
     HEADRequest,
+    url_basename,
 )
 
 
@@ -75,16 +77,18 @@ 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):
-        request = compat_urllib_request.Request(url)
+    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)
         # Get 'api_token' cookie
-        self._request_webpage(HEADRequest('http://www.viewster.com/'), video_id)
+        self._request_webpage(
+            HEADRequest('http://www.viewster.com/'),
+            video_id, headers=self.geo_verification_headers())
         cookies = self._get_cookies('http://www.viewster.com/')
         self._AUTH_TOKEN = compat_urllib_parse_unquote(cookies['api_token'].value)
 
@@ -114,51 +118,92 @@ class ViewsterIE(InfoExtractor):
             return self.playlist_result(entries, video_id, title, description)
 
         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:
+        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
+
+            medias = self._download_json(
+                'https://public-api.viewster.com/movies/%s/videos' % entry_id,
+                video_id, fatal=False, query={
+                    'mediaTypes': ['application/f4m+xml', 'application/x-mpegURL', 'video/mp4'],
+                    'language': audio,
+                    'subtitle': subtitle,
+                })
+            if not medias:
                 continue
-            ext = determine_ext(video_url)
-            if ext == 'f4m':
-                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':
-                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:
-                format_id = media.get('Bitrate')
-                f = {
-                    'url': video_url,
-                    'format_id': 'mp4-%s' % format_id,
-                    'height': int_or_none(media.get('Height')),
-                    'width': int_or_none(media.get('Width')),
-                    'preference': 1,
-                }
-                if format_id and not f['height']:
-                    f['height'] = int_or_none(self._search_regex(
-                        r'^(\d+)[pP]$', format_id, 'height', default=None))
-                formats.append(f)
-
-        if not formats and not info.get('LanguageSets') and not info.get('VODSettings'):
+            for media in medias:
+                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=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:
+                    qualities_basename = self._search_regex(
+                        r'/([^/]+)\.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', 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)
 
-        synopsis = info.get('Synopsis', {})
+        synopsis = info.get('Synopsis') or {}
         # Prefer title outside synopsis since it's less messy
         title = (info.get('Title') or synopsis['Title']).strip()
-        description = synopsis.get('Detailed') or info.get('Synopsis', {}).get('Short')
+        description = synopsis.get('Detailed') or (info.get('Synopsis') or {}).get('Short')
         duration = int_or_none(info.get('Duration'))
         timestamp = parse_iso8601(info.get('ReleaseDate'))