Merge tag 'upstream/2016.12.01'

[youtubedl] / youtube_dl / extractor / twitch.py

diff --git a/youtube_dl/extractor/twitch.py b/youtube_dl/extractor/twitch.py
index 20919774d0bcf65705e1f379a433667e2a06b7e9..77414a242d68f2309985235f0418d47c77194417 100644 (file)
--- a/youtube_dl/extractor/twitch.py
+++ b/youtube_dl/extractor/twitch.py
@@ -7,6 +7,7 @@ import random
 
 from .common import InfoExtractor
 from ..compat import (
 
 from .common import InfoExtractor
 from ..compat import (

+    compat_HTTPError,

     compat_parse_qs,
     compat_str,
     compat_urllib_parse_urlencode,
     compat_parse_qs,
     compat_str,
     compat_urllib_parse_urlencode,


@@ -14,13 +15,13 @@ from ..compat import (
     compat_urlparse,
 )
 from ..utils import (
     compat_urlparse,
 )
 from ..utils import (

+    clean_html,

     ExtractorError,
     int_or_none,
     js_to_json,
     orderedSet,
     parse_duration,
     parse_iso8601,
     ExtractorError,
     int_or_none,
     js_to_json,
     orderedSet,
     parse_duration,
     parse_iso8601,

-    sanitized_Request,

     urlencode_postdata,
 )
 
     urlencode_postdata,
 )
 


@@ -29,8 +30,9 @@ class TwitchBaseIE(InfoExtractor):
     _VALID_URL_BASE = r'https?://(?:www\.)?twitch\.tv'
 
     _API_BASE = 'https://api.twitch.tv'
     _VALID_URL_BASE = r'https?://(?:www\.)?twitch\.tv'
 
     _API_BASE = 'https://api.twitch.tv'

-    _USHER_BASE = 'http://usher.twitch.tv'
+    _USHER_BASE = 'https://usher.ttvnw.net'

     _LOGIN_URL = 'http://www.twitch.tv/login'
     _LOGIN_URL = 'http://www.twitch.tv/login'

+    _CLIENT_ID = 'jzkbprff40iqj646a697cyrvl0zt2m6'

     _NETRC_MACHINE = 'twitch'
 
     def _handle_error(self, response):
     _NETRC_MACHINE = 'twitch'
 
     def _handle_error(self, response):


@@ -42,16 +44,10 @@ class TwitchBaseIE(InfoExtractor):
                 '%s returned error: %s - %s' % (self.IE_NAME, error, response.get('message')),
                 expected=True)
 
                 '%s returned error: %s - %s' % (self.IE_NAME, error, response.get('message')),
                 expected=True)
 

-    def _download_json(self, url, video_id, note='Downloading JSON metadata'):
-        headers = {
-            'Referer': 'http://api.twitch.tv/crossdomain/receiver.html?v=2',
-            'X-Requested-With': 'XMLHttpRequest',
-        }
-        for cookie in self._downloader.cookiejar:
-            if cookie.name == 'api_token':
-                headers['Twitch-Api-Token'] = cookie.value
-        request = sanitized_Request(url, headers=headers)
-        response = super(TwitchBaseIE, self)._download_json(request, video_id, note)
+    def _call_api(self, path, item_id, note):
+        response = self._download_json(
+            '%s/%s' % (self._API_BASE, path), item_id, note,
+            headers={'Client-ID': self._CLIENT_ID})

         self._handle_error(response)
         return response
 
         self._handle_error(response)
         return response
 


@@ -63,9 +59,17 @@ class TwitchBaseIE(InfoExtractor):
         if username is None:
             return
 
         if username is None:
             return
 

+        def fail(message):
+            raise ExtractorError(
+                'Unable to login. Twitch said: %s' % message, expected=True)
+

         login_page, handle = self._download_webpage_handle(
             self._LOGIN_URL, None, 'Downloading login page')
 
         login_page, handle = self._download_webpage_handle(
             self._LOGIN_URL, None, 'Downloading login page')
 

+        # Some TOR nodes and public proxies are blocked completely
+        if 'blacklist_message' in login_page:
+            fail(clean_html(login_page))
+

         login_form = self._hidden_inputs(login_page)
 
         login_form.update({
         login_form = self._hidden_inputs(login_page)
 
         login_form.update({


@@ -82,21 +86,24 @@ class TwitchBaseIE(InfoExtractor):
         if not post_url.startswith('http'):
             post_url = compat_urlparse.urljoin(redirect_url, post_url)
 
         if not post_url.startswith('http'):
             post_url = compat_urlparse.urljoin(redirect_url, post_url)
 

-        request = sanitized_Request(
-            post_url, urlencode_postdata(login_form))
-        request.add_header('Referer', redirect_url)
-        response = self._download_webpage(
-            request, None, 'Logging in as %s' % username)
-
-        error_message = self._search_regex(
-            r'<div[^>]+class="subwindow_notice"[^>]*>([^<]+)</div>',
-            response, 'error message', default=None)
-        if error_message:
-            raise ExtractorError(
-                'Unable to login. Twitch said: %s' % error_message, expected=True)
+        headers = {'Referer': redirect_url}

 
 

-        if '>Reset your password<' in response:
-            self.report_warning('Twitch asks you to reset your password, go to https://secure.twitch.tv/reset/submit')
+        try:
+            response = self._download_json(
+                post_url, None, 'Logging in as %s' % username,
+                data=urlencode_postdata(login_form),
+                headers=headers)
+        except ExtractorError as e:
+            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 400:
+                response = self._parse_json(
+                    e.cause.read().decode('utf-8'), None)
+                fail(response['message'])
+            raise
+
+        if response.get('redirect'):
+            self._download_webpage(
+                response['redirect'], None, 'Downloading login redirect page',
+                headers=headers)

 
     def _prefer_source(self, formats):
         try:
 
     def _prefer_source(self, formats):
         try:


@@ -109,14 +116,14 @@ class TwitchBaseIE(InfoExtractor):
 
 class TwitchItemBaseIE(TwitchBaseIE):
     def _download_info(self, item, item_id):
 
 class TwitchItemBaseIE(TwitchBaseIE):
     def _download_info(self, item, item_id):

-        return self._extract_info(self._download_json(
-            '%s/kraken/videos/%s%s' % (self._API_BASE, item, item_id), item_id,
+        return self._extract_info(self._call_api(
+            'kraken/videos/%s%s' % (item, item_id), item_id,

             'Downloading %s info JSON' % self._ITEM_TYPE))
 
     def _extract_media(self, item_id):
         info = self._download_info(self._ITEM_SHORTCUT, item_id)
             'Downloading %s info JSON' % self._ITEM_TYPE))
 
     def _extract_media(self, item_id):
         info = self._download_info(self._ITEM_SHORTCUT, item_id)

-        response = self._download_json(
-            '%s/api/videos/%s%s' % (self._API_BASE, self._ITEM_SHORTCUT, item_id), item_id,
+        response = self._call_api(
+            'api/videos/%s%s' % (self._ITEM_SHORTCUT, item_id), item_id,

             'Downloading %s playlist JSON' % self._ITEM_TYPE)
         entries = []
         chunks = response['chunks']
             'Downloading %s playlist JSON' % self._ITEM_TYPE)
         entries = []
         chunks = response['chunks']


@@ -240,14 +247,15 @@ class TwitchVodIE(TwitchItemBaseIE):
             # m3u8 download
             'skip_download': True,
         },
             # m3u8 download
             'skip_download': True,
         },

+        'skip': 'HTTP Error 404: Not Found',

     }]
 
     def _real_extract(self, url):
         item_id = self._match_id(url)
 
         info = self._download_info(self._ITEM_SHORTCUT, item_id)
     }]
 
     def _real_extract(self, url):
         item_id = self._match_id(url)
 
         info = self._download_info(self._ITEM_SHORTCUT, item_id)

-        access_token = self._download_json(
-            '%s/api/vods/%s/access_token' % (self._API_BASE, item_id), item_id,
+        access_token = self._call_api(
+            'api/vods/%s/access_token' % item_id, item_id,

             'Downloading %s access token' % self._ITEM_TYPE)
 
         formats = self._extract_m3u8_formats(
             'Downloading %s access token' % self._ITEM_TYPE)
 
         formats = self._extract_m3u8_formats(


@@ -275,12 +283,12 @@ class TwitchVodIE(TwitchItemBaseIE):
 
 
 class TwitchPlaylistBaseIE(TwitchBaseIE):
 
 
 class TwitchPlaylistBaseIE(TwitchBaseIE):

-    _PLAYLIST_URL = '%s/kraken/channels/%%s/videos/?offset=%%d&limit=%%d' % TwitchBaseIE._API_BASE
+    _PLAYLIST_PATH = 'kraken/channels/%s/videos/?offset=%d&limit=%d'

     _PAGE_LIMIT = 100
 
     def _extract_playlist(self, channel_id):
     _PAGE_LIMIT = 100
 
     def _extract_playlist(self, channel_id):

-        info = self._download_json(
-            '%s/kraken/channels/%s' % (self._API_BASE, channel_id),
+        info = self._call_api(
+            'kraken/channels/%s' % channel_id,

             channel_id, 'Downloading channel info JSON')
         channel_name = info.get('display_name') or info.get('name')
         entries = []
             channel_id, 'Downloading channel info JSON')
         channel_name = info.get('display_name') or info.get('name')
         entries = []


@@ -289,8 +297,8 @@ class TwitchPlaylistBaseIE(TwitchBaseIE):
         broken_paging_detected = False
         counter_override = None
         for counter in itertools.count(1):
         broken_paging_detected = False
         counter_override = None
         for counter in itertools.count(1):

-            response = self._download_json(
-                self._PLAYLIST_URL % (channel_id, offset, limit),
+            response = self._call_api(
+                self._PLAYLIST_PATH % (channel_id, offset, limit),

                 channel_id,
                 'Downloading %s videos JSON page %s'
                 % (self._PLAYLIST_TYPE, counter_override or counter))
                 channel_id,
                 'Downloading %s videos JSON page %s'
                 % (self._PLAYLIST_TYPE, counter_override or counter))


@@ -345,7 +353,7 @@ class TwitchProfileIE(TwitchPlaylistBaseIE):
 class TwitchPastBroadcastsIE(TwitchPlaylistBaseIE):
     IE_NAME = 'twitch:past_broadcasts'
     _VALID_URL = r'%s/(?P<id>[^/]+)/profile/past_broadcasts/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
 class TwitchPastBroadcastsIE(TwitchPlaylistBaseIE):
     IE_NAME = 'twitch:past_broadcasts'
     _VALID_URL = r'%s/(?P<id>[^/]+)/profile/past_broadcasts/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE

-    _PLAYLIST_URL = TwitchPlaylistBaseIE._PLAYLIST_URL + '&broadcasts=true'
+    _PLAYLIST_PATH = TwitchPlaylistBaseIE._PLAYLIST_PATH + '&broadcasts=true'

     _PLAYLIST_TYPE = 'past broadcasts'
 
     _TEST = {
     _PLAYLIST_TYPE = 'past broadcasts'
 
     _TEST = {


@@ -389,15 +397,12 @@ class TwitchStreamIE(TwitchBaseIE):
     def _real_extract(self, url):
         channel_id = self._match_id(url)
 
     def _real_extract(self, url):
         channel_id = self._match_id(url)
 

-        stream = self._download_json(
-            '%s/kraken/streams/%s' % (self._API_BASE, channel_id), channel_id,
+        stream = self._call_api(
+            'kraken/streams/%s?stream_type=all' % channel_id, channel_id,

             'Downloading stream JSON').get('stream')
 
             'Downloading stream JSON').get('stream')
 

-        # Fallback on profile extraction if stream is offline

         if not stream:
         if not stream:

-            return self.url_result(
-                'http://www.twitch.tv/%s/profile' % channel_id,
-                'TwitchProfile', channel_id)
+            raise ExtractorError('%s is offline' % channel_id, expected=True)

 
         # Channel name may be typed if different case than the original channel name
         # (e.g. http://www.twitch.tv/TWITCHPLAYSPOKEMON) that will lead to constructing
 
         # Channel name may be typed if different case than the original channel name
         # (e.g. http://www.twitch.tv/TWITCHPLAYSPOKEMON) that will lead to constructing


@@ -405,13 +410,14 @@ class TwitchStreamIE(TwitchBaseIE):
         # JSON and fallback to lowercase if it's not available.
         channel_id = stream.get('channel', {}).get('name') or channel_id.lower()
 
         # JSON and fallback to lowercase if it's not available.
         channel_id = stream.get('channel', {}).get('name') or channel_id.lower()
 

-        access_token = self._download_json(
-            '%s/api/channels/%s/access_token' % (self._API_BASE, channel_id), channel_id,
+        access_token = self._call_api(
+            'api/channels/%s/access_token' % channel_id, channel_id,

             'Downloading channel access token')
 
         query = {
             'allow_source': 'true',
             'allow_audio_only': 'true',
             'Downloading channel access token')
 
         query = {
             'allow_source': 'true',
             'allow_audio_only': 'true',

+            'allow_spectre': 'true',

             'p': random.randint(1000000, 10000000),
             'player': 'twitchweb',
             'segment_preference': '4',
             'p': random.randint(1000000, 10000000),
             'player': 'twitchweb',
             'segment_preference': '4',


@@ -461,7 +467,7 @@ class TwitchClipsIE(InfoExtractor):
     IE_NAME = 'twitch:clips'
     _VALID_URL = r'https?://clips\.twitch\.tv/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 
     IE_NAME = 'twitch:clips'
     _VALID_URL = r'https?://clips\.twitch\.tv/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 

-    _TEST = {
+    _TESTS = [{

         'url': 'https://clips.twitch.tv/ea/AggressiveCobraPoooound',
         'md5': '761769e1eafce0ffebfb4089cb3847cd',
         'info_dict': {
         'url': 'https://clips.twitch.tv/ea/AggressiveCobraPoooound',
         'md5': '761769e1eafce0ffebfb4089cb3847cd',
         'info_dict': {


@@ -473,7 +479,11 @@ class TwitchClipsIE(InfoExtractor):
             'uploader': 'stereotype_',
             'uploader_id': 'stereotype_',
         },
             'uploader': 'stereotype_',
             'uploader_id': 'stereotype_',
         },

-    }
+    }, {
+        # multiple formats
+        'url': 'https://clips.twitch.tv/rflegendary/UninterestedBeeDAESuppy',
+        'only_matching': True,
+    }]

 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
     def _real_extract(self, url):
         video_id = self._match_id(url)


@@ -485,15 +495,27 @@ class TwitchClipsIE(InfoExtractor):
                 r'(?s)clipInfo\s*=\s*({.+?});', webpage, 'clip info'),
             video_id, transform_source=js_to_json)
 
                 r'(?s)clipInfo\s*=\s*({.+?});', webpage, 'clip info'),
             video_id, transform_source=js_to_json)
 

-        video_url = clip['clip_video_url']
-        title = clip['channel_title']
+        title = clip.get('channel_title') or self._og_search_title(webpage)
+
+        formats = [{
+            'url': option['source'],
+            'format_id': option.get('quality'),
+            'height': int_or_none(option.get('quality')),
+        } for option in clip.get('quality_options', []) if option.get('source')]
+
+        if not formats:
+            formats = [{
+                'url': clip['clip_video_url'],
+            }]
+
+        self._sort_formats(formats)

 
         return {
             'id': video_id,
 
         return {
             'id': video_id,

-            'url': video_url,

             'title': title,
             'thumbnail': self._og_search_thumbnail(webpage),
             'creator': clip.get('broadcaster_display_name') or clip.get('broadcaster_login'),
             'uploader': clip.get('curator_login'),
             'uploader_id': clip.get('curator_display_name'),
             'title': title,
             'thumbnail': self._og_search_thumbnail(webpage),
             'creator': clip.get('broadcaster_display_name') or clip.get('broadcaster_login'),
             'uploader': clip.get('curator_login'),
             'uploader_id': clip.get('curator_display_name'),

+            'formats': formats,

         }
         }