Update upstream source from tag 'upstream/2017.11.06'

[youtubedl] / youtube_dl / extractor / viki.py

diff --git a/youtube_dl/extractor/viki.py b/youtube_dl/extractor/viki.py
index 7f2fb1ca8896e29e48a41a9efddaded987ba1e96..853e5c75f431da9bbc7383c31006d29c705bb881 100644 (file)
--- a/youtube_dl/extractor/viki.py
+++ b/youtube_dl/extractor/viki.py
@@ -1,17 +1,21 @@
+# coding: utf-8

 from __future__ import unicode_literals
 
 from __future__ import unicode_literals
 

-import time
-import hmac

 import hashlib
 import hashlib

+import hmac

 import itertools
 import itertools

+import json
+import re
+import time

 
 

+from .common import InfoExtractor

 from ..utils import (
     ExtractorError,
     int_or_none,
     parse_age_limit,
     parse_iso8601,
 from ..utils import (
     ExtractorError,
     int_or_none,
     parse_age_limit,
     parse_iso8601,

+    sanitized_Request,

 )
 )

-from .common import InfoExtractor

 
 
 class VikiBaseIE(InfoExtractor):
 
 
 class VikiBaseIE(InfoExtractor):


@@ -19,31 +23,46 @@ class VikiBaseIE(InfoExtractor):
     _API_QUERY_TEMPLATE = '/v4/%sapp=%s&t=%s&site=www.viki.com'
     _API_URL_TEMPLATE = 'http://api.viki.io%s&sig=%s'
 
     _API_QUERY_TEMPLATE = '/v4/%sapp=%s&t=%s&site=www.viki.com'
     _API_URL_TEMPLATE = 'http://api.viki.io%s&sig=%s'
 

-    _APP = '65535a'
+    _APP = '100005a'

     _APP_VERSION = '2.2.5.1428709186'
     _APP_VERSION = '2.2.5.1428709186'

-    _APP_SECRET = '-$iJ}@p7!G@SyU/je1bEyWg}upLu-6V6-Lg9VD(]siH,r.,m-r|ulZ,U4LC/SeR)'
+    _APP_SECRET = 'MM_d*yP@`&1@]@!AVrXf_o-HVEnoTnm$O-ti4[G~$JDI/Dc-&piU&z&5.;:}95=Iad'
+
+    _GEO_BYPASS = False
+    _NETRC_MACHINE = 'viki'

 
 

-    def _prepare_call(self, path, timestamp=None):
+    _token = None
+
+    _ERRORS = {
+        'geo': 'Sorry, this content is not available in your region.',
+        'upcoming': 'Sorry, this content is not yet available.',
+        # 'paywall': 'paywall',
+    }
+
+    def _prepare_call(self, path, timestamp=None, post_data=None):

         path += '?' if '?' not in path else '&'
         if not timestamp:
             timestamp = int(time.time())
         query = self._API_QUERY_TEMPLATE % (path, self._APP, timestamp)
         path += '?' if '?' not in path else '&'
         if not timestamp:
             timestamp = int(time.time())
         query = self._API_QUERY_TEMPLATE % (path, self._APP, timestamp)

+        if self._token:
+            query += '&token=%s' % self._token

         sig = hmac.new(
             self._APP_SECRET.encode('ascii'),
             query.encode('ascii'),
             hashlib.sha1
         ).hexdigest()
         sig = hmac.new(
             self._APP_SECRET.encode('ascii'),
             query.encode('ascii'),
             hashlib.sha1
         ).hexdigest()

-        return self._API_URL_TEMPLATE % (query, sig)
+        url = self._API_URL_TEMPLATE % (query, sig)
+        return sanitized_Request(
+            url, json.dumps(post_data).encode('utf-8')) if post_data else url

 
 

-    def _call_api(self, path, video_id, note, timestamp=None):
+    def _call_api(self, path, video_id, note, timestamp=None, post_data=None):

         resp = self._download_json(
         resp = self._download_json(

-            self._prepare_call(path, timestamp), video_id, note)
+            self._prepare_call(path, timestamp, post_data), video_id, note)

 
         error = resp.get('error')
         if error:
             if error == 'invalid timestamp':
                 resp = self._download_json(
 
         error = resp.get('error')
         if error:
             if error == 'invalid timestamp':
                 resp = self._download_json(

-                    self._prepare_call(path, int(resp['current_timestamp'])),
+                    self._prepare_call(path, int(resp['current_timestamp']), post_data),

                     video_id, '%s (retry)' % note)
                 error = resp.get('error')
             if error:
                     video_id, '%s (retry)' % note)
                 error = resp.get('error')
             if error:


@@ -56,6 +75,47 @@ class VikiBaseIE(InfoExtractor):
             '%s returned error: %s' % (self.IE_NAME, error),
             expected=True)
 
             '%s returned error: %s' % (self.IE_NAME, error),
             expected=True)
 

+    def _check_errors(self, data):
+        for reason, status in data.get('blocking', {}).items():
+            if status and reason in self._ERRORS:
+                message = self._ERRORS[reason]
+                if reason == 'geo':
+                    self.raise_geo_restricted(msg=message)
+                raise ExtractorError('%s said: %s' % (
+                    self.IE_NAME, message), expected=True)
+
+    def _real_initialize(self):
+        self._login()
+
+    def _login(self):
+        (username, password) = self._get_login_info()
+        if username is None:
+            return
+
+        login_form = {
+            'login_id': username,
+            'password': password,
+        }
+
+        login = self._call_api(
+            'sessions.json', None,
+            'Logging in as %s' % username, post_data=login_form)
+
+        self._token = login.get('token')
+        if not self._token:
+            self.report_warning('Unable to get session token, login has probably failed')
+
+    @staticmethod
+    def dict_selection(dict_obj, preferred_key, allow_fallback=True):
+        if preferred_key in dict_obj:
+            return dict_obj.get(preferred_key)
+
+        if not allow_fallback:
+            return
+
+        filtered_dict = list(filter(None, [dict_obj.get(k) for k in dict_obj.keys()]))
+        return filtered_dict[0] if filtered_dict else None
+

 
 class VikiIE(VikiBaseIE):
     IE_NAME = 'viki'
 
 class VikiIE(VikiBaseIE):
     IE_NAME = 'viki'


@@ -101,20 +161,17 @@ class VikiIE(VikiBaseIE):
             'like_count': int,
             'age_limit': 13,
         },
             'like_count': int,
             'age_limit': 13,
         },

-        'params': {
-            # m3u8 download
-            'skip_download': True,
-        }
+        'skip': 'Blocked in the US',

     }, {
         # episode
         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
     }, {
         # episode
         'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',

-        'md5': '190f3ef426005ba3a080a63325955bc3',
+        'md5': '5fa476a902e902783ac7a4d615cdbc7a',

         'info_dict': {
             'id': '44699v',
             'ext': 'mp4',
             'title': 'Boys Over Flowers - Episode 1',
         'info_dict': {
             'id': '44699v',
             'ext': 'mp4',
             'title': 'Boys Over Flowers - Episode 1',

-            'description': 'md5:52617e4f729c7d03bfd4bcbbb6e946f2',
-            'duration': 4155,
+            'description': 'md5:b89cf50038b480b88b5b3c93589a9076',
+            'duration': 4204,

             'timestamp': 1270496524,
             'upload_date': '20100405',
             'uploader': 'group8',
             'timestamp': 1270496524,
             'upload_date': '20100405',
             'uploader': 'group8',


@@ -124,13 +181,13 @@ class VikiIE(VikiBaseIE):
     }, {
         # youtube external
         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
     }, {
         # youtube external
         'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',

-        'md5': '216d1afdc0c64d1febc1e9f2bd4b864b',
+        'md5': '63f8600c1da6f01b7640eee7eca4f1da',

         'info_dict': {
             'id': '50562v',
         'info_dict': {
             'id': '50562v',

-            'ext': 'mp4',
+            'ext': 'webm',

             'title': 'Poor Nastya [COMPLETE] - Episode 1',
             'description': '',
             'title': 'Poor Nastya [COMPLETE] - Episode 1',
             'description': '',

-            'duration': 607,
+            'duration': 606,

             'timestamp': 1274949505,
             'upload_date': '20101213',
             'uploader': 'ad14065n',
             'timestamp': 1274949505,
             'upload_date': '20101213',
             'uploader': 'ad14065n',


@@ -141,6 +198,20 @@ class VikiIE(VikiBaseIE):
     }, {
         'url': 'http://www.viki.com/player/44699v',
         'only_matching': True,
     }, {
         'url': 'http://www.viki.com/player/44699v',
         'only_matching': True,

+    }, {
+        # non-English description
+        'url': 'http://www.viki.com/videos/158036v-love-in-magic',
+        'md5': '1713ae35df5a521b31f6dc40730e7c9c',
+        'info_dict': {
+            'id': '158036v',
+            'ext': 'mp4',
+            'uploader': 'I Planet Entertainment',
+            'upload_date': '20111122',
+            'timestamp': 1321985454,
+            'description': 'md5:44b1e46619df3a072294645c770cef36',
+            'title': 'Love In Magic',
+            'age_limit': 13,
+        },

     }]
 
     def _real_extract(self, url):
     }]
 
     def _real_extract(self, url):


@@ -149,19 +220,16 @@ class VikiIE(VikiBaseIE):
         video = self._call_api(
             'videos/%s.json' % video_id, video_id, 'Downloading video JSON')
 
         video = self._call_api(
             'videos/%s.json' % video_id, video_id, 'Downloading video JSON')
 

-        title = None
-        titles = video.get('titles')
-        if titles:
-            title = titles.get('en') or titles[titles.keys()[0]]
+        self._check_errors(video)
+
+        title = self.dict_selection(video.get('titles', {}), 'en', allow_fallback=False)

         if not title:
             title = 'Episode %d' % video.get('number') if video.get('type') == 'episode' else video.get('id') or video_id
         if not title:
             title = 'Episode %d' % video.get('number') if video.get('type') == 'episode' else video.get('id') or video_id

-            container_titles = video.get('container', {}).get('titles')
-            if container_titles:
-                container_title = container_titles.get('en') or container_titles[container_titles.keys()[0]]
-                title = '%s - %s' % (container_title, title)
+            container_titles = video.get('container', {}).get('titles', {})
+            container_title = self.dict_selection(container_titles, 'en')
+            title = '%s - %s' % (container_title, title)

 
 

-        descriptions = video.get('descriptions')
-        description = descriptions.get('en') or descriptions[titles.keys()[0]] if descriptions else None
+        description = self.dict_selection(video.get('descriptions', {}), 'en')

 
         duration = int_or_none(video.get('duration'))
         timestamp = parse_iso8601(video.get('created_at'))
 
         duration = int_or_none(video.get('duration'))
         timestamp = parse_iso8601(video.get('created_at'))


@@ -210,15 +278,41 @@ class VikiIE(VikiBaseIE):
 
         formats = []
         for format_id, stream_dict in streams.items():
 
         formats = []
         for format_id, stream_dict in streams.items():

-            height = self._search_regex(
-                r'^(\d+)[pP]$', format_id, 'height', default=None)
+            height = int_or_none(self._search_regex(
+                r'^(\d+)[pP]$', format_id, 'height', default=None))

             for protocol, format_dict in stream_dict.items():
             for protocol, format_dict in stream_dict.items():

+                # rtmps URLs does not seem to work
+                if protocol == 'rtmps':
+                    continue
+                format_url = format_dict['url']

                 if format_id == 'm3u8':
                 if format_id == 'm3u8':

-                    formats = self._extract_m3u8_formats(
-                        format_dict['url'], video_id, 'mp4', m3u8_id='m3u8-%s' % protocol)
+                    m3u8_formats = self._extract_m3u8_formats(
+                        format_url, video_id, 'mp4',
+                        entry_protocol='m3u8_native',
+                        m3u8_id='m3u8-%s' % protocol, fatal=False)
+                    # Despite CODECS metadata in m3u8 all video-only formats
+                    # are actually video+audio
+                    for f in m3u8_formats:
+                        if f.get('acodec') == 'none' and f.get('vcodec') != 'none':
+                            f['acodec'] = None
+                    formats.extend(m3u8_formats)
+                elif format_url.startswith('rtmp'):
+                    mobj = re.search(
+                        r'^(?P<url>rtmp://[^/]+/(?P<app>.+?))/(?P<playpath>mp4:.+)$',
+                        format_url)
+                    if not mobj:
+                        continue
+                    formats.append({
+                        'format_id': 'rtmp-%s' % format_id,
+                        'ext': 'flv',
+                        'url': mobj.group('url'),
+                        'play_path': mobj.group('playpath'),
+                        'app': mobj.group('app'),
+                        'page_url': url,
+                    })

                 else:
                     formats.append({
                 else:
                     formats.append({

-                        'url': format_dict['url'],
+                        'url': format_url,

                         'format_id': '%s-%s' % (format_id, protocol),
                         'height': height,
                     })
                         'format_id': '%s-%s' % (format_id, protocol),
                         'height': height,
                     })


@@ -238,7 +332,7 @@ class VikiChannelIE(VikiBaseIE):
             'title': 'Boys Over Flowers',
             'description': 'md5:ecd3cff47967fe193cff37c0bec52790',
         },
             'title': 'Boys Over Flowers',
             'description': 'md5:ecd3cff47967fe193cff37c0bec52790',
         },

-        'playlist_count': 70,
+        'playlist_mincount': 71,

     }, {
         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
         'info_dict': {
     }, {
         'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
         'info_dict': {


@@ -267,11 +361,11 @@ class VikiChannelIE(VikiBaseIE):
             'containers/%s.json' % channel_id, channel_id,
             'Downloading channel JSON')
 
             'containers/%s.json' % channel_id, channel_id,
             'Downloading channel JSON')
 

-        titles = channel['titles']
-        title = titles.get('en') or titles[titles.keys()[0]]
+        self._check_errors(channel)
+
+        title = self.dict_selection(channel['titles'], 'en')

 
 

-        descriptions = channel['descriptions']
-        description = descriptions.get('en') or descriptions[descriptions.keys()[0]]
+        description = self.dict_selection(channel['descriptions'], 'en')

 
         entries = []
         for video_type in ('episodes', 'clips', 'movies'):
 
         entries = []
         for video_type in ('episodes', 'clips', 'movies'):