Merge pull request #328 from ytdl-org/master

[pull] master from ytdl-org:master

youtube_dl/extractor/brightcove.py

@@ -5,32 +5,34 @@ import base64
 import re
 import struct
 
-from .common import InfoExtractor
 from .adobepass import AdobePassIE
+from .common import InfoExtractor
 from ..compat import (
     compat_etree_fromstring,
+    compat_HTTPError,
     compat_parse_qs,
     compat_urllib_parse_urlparse,
     compat_urlparse,
     compat_xml_parse_error,
-    compat_HTTPError,
 )
 from ..utils import (
-    ExtractorError,
+    clean_html,
     extract_attributes,
+    ExtractorError,
     find_xpath_attr,
     fix_xml_ampersands,
     float_or_none,
-    js_to_json,
     int_or_none,
+    js_to_json,
+    mimetype2ext,
     parse_iso8601,
     smuggle_url,
+    str_or_none,
     unescapeHTML,
     unsmuggle_url,
-    update_url_query,
-    clean_html,
-    mimetype2ext,
     UnsupportedError,
+    update_url_query,
+    url_or_none,
 )
 
 
@@ -553,10 +555,16 @@ class BrightcoveNewIE(AdobePassIE):
 
         subtitles = {}
         for text_track in json_data.get('text_tracks', []):
-            if text_track.get('src'):
+            if text_track.get('kind') != 'captions':
-                subtitles.setdefault(text_track.get('srclang'), []).append({
+                continue
-                    'url': text_track['src'],
+            text_track_url = url_or_none(text_track.get('src'))
-                })
+            if not text_track_url:
+                continue
+            lang = (str_or_none(text_track.get('srclang'))
+                    or str_or_none(text_track.get('label')) or 'en').lower()
+            subtitles.setdefault(lang, []).append({
+                'url': text_track_url,
+            })
 
         is_live = False
         duration = float_or_none(json_data.get('duration'), 1000)

youtube_dl/extractor/twitch.py

@@ -21,6 +21,8 @@ from ..utils import (
     orderedSet,
     parse_duration,
     parse_iso8601,
+    qualities,
+    str_or_none,
     try_get,
     unified_timestamp,
     update_url_query,
@@ -50,8 +52,14 @@ class TwitchBaseIE(InfoExtractor):
 
     def _call_api(self, path, item_id, *args, **kwargs):
         headers = kwargs.get('headers', {}).copy()
-        headers['Client-ID'] = self._CLIENT_ID
+        headers.update({
-        kwargs['headers'] = headers
+            'Accept': 'application/vnd.twitchtv.v5+json; charset=UTF-8',
+            'Client-ID': self._CLIENT_ID,
+        })
+        kwargs.update({
+            'headers': headers,
+            'expected_status': (400, 410),
+        })
         response = self._download_json(
             '%s/%s' % (self._API_BASE, path), item_id,
             *args, **compat_kwargs(kwargs))
@@ -186,12 +194,27 @@ class TwitchItemBaseIE(TwitchBaseIE):
             is_live = False
         else:
             is_live = None
+        _QUALITIES = ('small', 'medium', 'large')
+        quality_key = qualities(_QUALITIES)
+        thumbnails = []
+        preview = info.get('preview')
+        if isinstance(preview, dict):
+            for thumbnail_id, thumbnail_url in preview.items():
+                thumbnail_url = url_or_none(thumbnail_url)
+                if not thumbnail_url:
+                    continue
+                if thumbnail_id not in _QUALITIES:
+                    continue
+                thumbnails.append({
+                    'url': thumbnail_url,
+                    'preference': quality_key(thumbnail_id),
+                })
         return {
             'id': info['_id'],
             'title': info.get('title') or 'Untitled Broadcast',
             'description': info.get('description'),
             'duration': int_or_none(info.get('length')),
-            'thumbnail': info.get('preview'),
+            'thumbnails': thumbnails,
             'uploader': info.get('channel', {}).get('display_name'),
             'uploader_id': info.get('channel', {}).get('name'),
             'timestamp': parse_iso8601(info.get('recorded_at')),
@@ -583,10 +606,18 @@ class TwitchStreamIE(TwitchBaseIE):
                 else super(TwitchStreamIE, cls).suitable(url))
 
     def _real_extract(self, url):
-        channel_id = self._match_id(url)
+        channel_name = self._match_id(url)
+
+        access_token = self._call_api(
+            'api/channels/%s/access_token' % channel_name, channel_name,
+            'Downloading access token JSON')
+
+        token = access_token['token']
+        channel_id = compat_str(self._parse_json(
+            token, channel_name)['channel_id'])
 
         stream = self._call_api(
-            'kraken/streams/%s?stream_type=all' % channel_id.lower(),
+            'kraken/streams/%s?stream_type=all' % channel_id,
             channel_id, 'Downloading stream JSON').get('stream')
 
         if not stream:
@@ -596,11 +627,9 @@ class TwitchStreamIE(TwitchBaseIE):
         # (e.g. http://www.twitch.tv/TWITCHPLAYSPOKEMON) that will lead to constructing
         # an invalid m3u8 URL. Working around by use of original channel name from stream
         # JSON and fallback to lowercase if it's not available.
-        channel_id = stream.get('channel', {}).get('name') or channel_id.lower()
+        channel_name = try_get(
-
+            stream, lambda x: x['channel']['name'],
-        access_token = self._call_api(
+            compat_str) or channel_name.lower()
-            'api/channels/%s/access_token' % channel_id, channel_id,
-            'Downloading channel access token')
 
         query = {
             'allow_source': 'true',
@@ -611,11 +640,11 @@ class TwitchStreamIE(TwitchBaseIE):
             'playlist_include_framerate': 'true',
             'segment_preference': '4',
             'sig': access_token['sig'].encode('utf-8'),
-            'token': access_token['token'].encode('utf-8'),
+            'token': token.encode('utf-8'),
         }
         formats = self._extract_m3u8_formats(
             '%s/api/channel/hls/%s.m3u8?%s'
-            % (self._USHER_BASE, channel_id, compat_urllib_parse_urlencode(query)),
+            % (self._USHER_BASE, channel_name, compat_urllib_parse_urlencode(query)),
             channel_id, 'mp4')
         self._prefer_source(formats)
 
@@ -638,8 +667,8 @@ class TwitchStreamIE(TwitchBaseIE):
             })
 
         return {
-            'id': compat_str(stream['_id']),
+            'id': str_or_none(stream.get('_id')) or channel_id,
-            'display_id': channel_id,
+            'display_id': channel_name,
             'title': title,
             'description': description,
             'thumbnails': thumbnails,