Merge tag 'upstream/2015.11.27.1'

[youtubedl] / youtube_dl / extractor / viki.py

diff --git a/youtube_dl/extractor/viki.py b/youtube_dl/extractor/viki.py
index 944901e1482a666ae90cc5e1c0f86e325ec2aecc..a63c236177eb6f232cc544f3b8406cbe934d7d5c 100644 (file)
--- a/youtube_dl/extractor/viki.py
+++ b/youtube_dl/extractor/viki.py
@@ -1,21 +1,106 @@
+# coding: utf-8

 from __future__ import unicode_literals
 
 from __future__ import unicode_literals
 

-import re
+import json
+import time
+import hmac
+import hashlib
+import itertools

 
 

+from .common import InfoExtractor

 from ..utils import (
     ExtractorError,
 from ..utils import (
     ExtractorError,

-    unescapeHTML,
-    unified_strdate,
-    US_RATINGS,
+    int_or_none,
+    parse_age_limit,
+    parse_iso8601,
+    sanitized_Request,

 )
 )

-from .subtitles import SubtitlesInfoExtractor

 
 
 
 

-class VikiIE(SubtitlesInfoExtractor):
-    IE_NAME = 'viki'
+class VikiBaseIE(InfoExtractor):
+    _VALID_URL_BASE = r'https?://(?:www\.)?viki\.(?:com|net|mx|jp|fr)/'
+    _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_VERSION = '2.2.5.1428709186'
+    _APP_SECRET = '-$iJ}@p7!G@SyU/je1bEyWg}upLu-6V6-Lg9VD(]siH,r.,m-r|ulZ,U4LC/SeR)'
+
+    _NETRC_MACHINE = 'viki'
+
+    _token = None
+
+    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)
+        if self._token:
+            query += '&token=%s' % self._token
+        sig = hmac.new(
+            self._APP_SECRET.encode('ascii'),
+            query.encode('ascii'),
+            hashlib.sha1
+        ).hexdigest()
+        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, post_data=None):
+        resp = self._download_json(
+            self._prepare_call(path, timestamp, post_data), video_id, note)
+
+        error = resp.get('error')
+        if error:
+            if error == 'invalid timestamp':
+                resp = self._download_json(
+                    self._prepare_call(path, int(resp['current_timestamp']), post_data),
+                    video_id, '%s (retry)' % note)
+                error = resp.get('error')
+            if error:
+                self._raise_error(resp['error'])
+
+        return resp
+
+    def _raise_error(self, error):
+        raise ExtractorError(
+            '%s returned error: %s' % (self.IE_NAME, error),
+            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')

 
 

-    _VALID_URL = r'^https?://(?:www\.)?viki\.com/videos/(?P<id>[0-9]+v)'
-    _TEST = {
+    @staticmethod
+    def dict_selection(dict_obj, preferred_key):
+        if preferred_key in dict_obj:
+            return dict_obj.get(preferred_key)
+
+        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'
+    _VALID_URL = r'%s(?:videos|player)/(?P<id>[0-9]+v)' % VikiBaseIE._VALID_URL_BASE
+    _TESTS = [{

         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
         'info_dict': {
             'id': '1023585v',
         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
         'info_dict': {
             'id': '1023585v',


@@ -27,70 +112,225 @@ class VikiIE(SubtitlesInfoExtractor):
             'age_limit': 13,
         },
         'skip': 'Blocked in the US',
             'age_limit': 13,
         },
         'skip': 'Blocked in the US',

-    }
+    }, {
+        # clip
+        'url': 'http://www.viki.com/videos/1067139v-the-avengers-age-of-ultron-press-conference',
+        'md5': '86c0b5dbd4d83a6611a79987cc7a1989',
+        'info_dict': {
+            'id': '1067139v',
+            'ext': 'mp4',
+            'title': "'The Avengers: Age of Ultron' Press Conference",
+            'description': 'md5:d70b2f9428f5488321bfe1db10d612ea',
+            'duration': 352,
+            'timestamp': 1430380829,
+            'upload_date': '20150430',
+            'uploader': 'Arirang TV',
+            'like_count': int,
+            'age_limit': 0,
+        }
+    }, {
+        'url': 'http://www.viki.com/videos/1048879v-ankhon-dekhi',
+        'info_dict': {
+            'id': '1048879v',
+            'ext': 'mp4',
+            'title': 'Ankhon Dekhi',
+            'duration': 6512,
+            'timestamp': 1408532356,
+            'upload_date': '20140820',
+            'uploader': 'Spuul',
+            'like_count': int,
+            'age_limit': 13,
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        }
+    }, {
+        # episode
+        'url': 'http://www.viki.com/videos/44699v-boys-over-flowers-episode-1',
+        'md5': '190f3ef426005ba3a080a63325955bc3',
+        'info_dict': {
+            'id': '44699v',
+            'ext': 'mp4',
+            'title': 'Boys Over Flowers - Episode 1',
+            'description': 'md5:52617e4f729c7d03bfd4bcbbb6e946f2',
+            'duration': 4155,
+            'timestamp': 1270496524,
+            'upload_date': '20100405',
+            'uploader': 'group8',
+            'like_count': int,
+            'age_limit': 13,
+        }
+    }, {
+        # youtube external
+        'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1',
+        'md5': '216d1afdc0c64d1febc1e9f2bd4b864b',
+        'info_dict': {
+            'id': '50562v',
+            'ext': 'mp4',
+            'title': 'Poor Nastya [COMPLETE] - Episode 1',
+            'description': '',
+            'duration': 607,
+            'timestamp': 1274949505,
+            'upload_date': '20101213',
+            'uploader': 'ad14065n',
+            'uploader_id': 'ad14065n',
+            'like_count': int,
+            'age_limit': 13,
+        }
+    }, {
+        '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',
+        },
+    }]

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

-        webpage = self._download_webpage(url, video_id)
-        title = self._og_search_title(webpage)
-        description = self._og_search_description(webpage)
-        thumbnail = self._og_search_thumbnail(webpage)
-
-        uploader_m = re.search(
-            r'<strong>Broadcast Network: </strong>\s*([^<]*)<', webpage)
-        if uploader_m is None:
-            uploader = None
-        else:
-            uploader = uploader_m.group(1).strip()
-
-        rating_str = self._html_search_regex(
-            r'<strong>Rating: </strong>\s*([^<]*)<', webpage,
-            'rating information', default='').strip()
-        age_limit = US_RATINGS.get(rating_str)
-
-        info_url = 'http://www.viki.com/player5_fragment/%s?action=show&controller=videos' % video_id
-        info_webpage = self._download_webpage(
-            info_url, video_id, note='Downloading info page')
-        if re.match(r'\s*<div\s+class="video-error', info_webpage):
-            raise ExtractorError(
-                'Video %s is blocked from your location.' % video_id,
-                expected=True)
-        video_url = self._html_search_regex(
-            r'<source[^>]+src="([^"]+)"', info_webpage, 'video URL')
-
-        upload_date_str = self._html_search_regex(
-            r'"created_at":"([^"]+)"', info_webpage, 'upload date')
-        upload_date = (
-            unified_strdate(upload_date_str)
-            if upload_date_str is not None
-            else None
-        )
-
-        # subtitles
-        video_subtitles = self.extract_subtitles(video_id, info_webpage)
-        if self._downloader.params.get('listsubtitles', False):
-            self._list_available_subtitles(video_id, info_webpage)
-            return
+        video = self._call_api(
+            'videos/%s.json' % video_id, video_id, 'Downloading video JSON')

 
 

-        return {
+        title = self.dict_selection(video.get('titles', {}), 'en')
+        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', {})
+            container_title = self.dict_selection(container_titles, 'en')
+            title = '%s - %s' % (container_title, title)
+
+        description = self.dict_selection(video.get('descriptions', {}), 'en')
+
+        duration = int_or_none(video.get('duration'))
+        timestamp = parse_iso8601(video.get('created_at'))
+        uploader = video.get('author')
+        like_count = int_or_none(video.get('likes', {}).get('count'))
+        age_limit = parse_age_limit(video.get('rating'))
+
+        thumbnails = []
+        for thumbnail_id, thumbnail in video.get('images', {}).items():
+            thumbnails.append({
+                'id': thumbnail_id,
+                'url': thumbnail.get('url'),
+            })
+
+        subtitles = {}
+        for subtitle_lang, _ in video.get('subtitle_completions', {}).items():
+            subtitles[subtitle_lang] = [{
+                'ext': subtitles_format,
+                'url': self._prepare_call(
+                    'videos/%s/subtitles/%s.%s' % (video_id, subtitle_lang, subtitles_format)),
+            } for subtitles_format in ('srt', 'vtt')]
+
+        result = {

             'id': video_id,
             'title': title,
             'id': video_id,
             'title': title,

-            'url': video_url,

             'description': description,
             'description': description,

-            'thumbnail': thumbnail,
-            'age_limit': age_limit,
+            'duration': duration,
+            'timestamp': timestamp,

             'uploader': uploader,
             'uploader': uploader,

-            'subtitles': video_subtitles,
-            'upload_date': upload_date,
+            'like_count': like_count,
+            'age_limit': age_limit,
+            'thumbnails': thumbnails,
+            'subtitles': subtitles,

         }
 
         }
 

-    def _get_available_subtitles(self, video_id, info_webpage):
-        res = {}
-        for sturl_html in re.findall(r'<track src="([^"]+)"/>', info_webpage):
-            sturl = unescapeHTML(sturl_html)
-            m = re.search(r'/(?P<lang>[a-z]+)\.vtt', sturl)
-            if not m:
-                continue
-            res[m.group('lang')] = sturl
-        return res
+        streams = self._call_api(
+            'videos/%s/streams.json' % video_id, video_id,
+            'Downloading video streams JSON')
+
+        if 'external' in streams:
+            result.update({
+                '_type': 'url_transparent',
+                'url': streams['external']['url'],
+            })
+            return result
+
+        formats = []
+        for format_id, stream_dict in streams.items():
+            height = int_or_none(self._search_regex(
+                r'^(\d+)[pP]$', format_id, 'height', default=None))
+            for protocol, format_dict in stream_dict.items():
+                if format_id == 'm3u8':
+                    formats = self._extract_m3u8_formats(
+                        format_dict['url'], video_id, 'mp4', m3u8_id='m3u8-%s' % protocol)
+                else:
+                    formats.append({
+                        'url': format_dict['url'],
+                        'format_id': '%s-%s' % (format_id, protocol),
+                        'height': height,
+                    })
+        self._sort_formats(formats)
+
+        result['formats'] = formats
+        return result
+
+
+class VikiChannelIE(VikiBaseIE):
+    IE_NAME = 'viki:channel'
+    _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
+    _TESTS = [{
+        'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
+        'info_dict': {
+            'id': '50c',
+            'title': 'Boys Over Flowers',
+            'description': 'md5:ecd3cff47967fe193cff37c0bec52790',
+        },
+        'playlist_count': 70,
+    }, {
+        'url': 'http://www.viki.com/tv/1354c-poor-nastya-complete',
+        'info_dict': {
+            'id': '1354c',
+            'title': 'Poor Nastya [COMPLETE]',
+            'description': 'md5:05bf5471385aa8b21c18ad450e350525',
+        },
+        'playlist_count': 127,
+    }, {
+        'url': 'http://www.viki.com/news/24569c-showbiz-korea',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.viki.com/artists/2141c-shinee',
+        'only_matching': True,
+    }]
+
+    _PER_PAGE = 25
+
+    def _real_extract(self, url):
+        channel_id = self._match_id(url)
+
+        channel = self._call_api(
+            'containers/%s.json' % channel_id, channel_id,
+            'Downloading channel JSON')
+
+        title = self.dict_selection(channel['titles'], 'en')
+
+        description = self.dict_selection(channel['descriptions'], 'en')
+
+        entries = []
+        for video_type in ('episodes', 'clips', 'movies'):
+            for page_num in itertools.count(1):
+                page = self._call_api(
+                    'containers/%s/%s.json?per_page=%d&sort=number&direction=asc&with_paging=true&page=%d'
+                    % (channel_id, video_type, self._PER_PAGE, page_num), channel_id,
+                    'Downloading %s JSON page #%d' % (video_type, page_num))
+                for video in page['response']:
+                    video_id = video['id']
+                    entries.append(self.url_result(
+                        'http://www.viki.com/videos/%s' % video_id, 'Viki'))
+                if not page['pagination']['next']:
+                    break
+
+        return self.playlist_result(entries, channel_id, title, description)