Revert changes to vlive

youtube_dlc/extractor/vlive.py

@@ -4,48 +4,52 @@ from __future__ import unicode_literals
 import re
 import time
 import itertools
-import json
 
 from .common import InfoExtractor
 from .naver import NaverBaseIE
-from ..compat import (
+from ..compat import compat_str
-    compat_HTTPError,
-    compat_str,
-)
 from ..utils import (
     ExtractorError,
-    int_or_none,
     merge_dicts,
     try_get,
     urlencode_postdata,
 )
 
 
-class VLiveBaseIE(NaverBaseIE):
+class VLiveIE(NaverBaseIE):
-    _APP_ID = '8c6cc7b45d2568fb668be6e05b6e5a3b'
-
-
-class VLiveIE(VLiveBaseIE):
     IE_NAME = 'vlive'
-    _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/(?:video|embed)/(?P<id>[0-9]+)'
+    _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/(?:video|post)/(?P<id>(?:\d-)?[0-9]+)'
     _NETRC_MACHINE = 'vlive'
     _TESTS = [{
-        'url': 'http://www.vlive.tv/video/1326',
+        'url': 'https://www.vlive.tv/video/1326',
         'md5': 'cc7314812855ce56de70a06a27314983',
         'info_dict': {
             'id': '1326',
             'ext': 'mp4',
-            'title': "Girl's Day's Broadcast",
+            'title': "[V LIVE] Girl's Day's Broadcast",
             'creator': "Girl's Day",
             'view_count': int,
             'uploader_id': 'muploader_a',
         },
-    }, {
+    },
-        'url': 'http://www.vlive.tv/video/16937',
+        {
+        'url': 'https://vlive.tv/post/1-18244258',
+        'md5': 'cc7314812855ce56de70a06a27314983',
+        'info_dict': {
+            'id': '1326',
+            'ext': 'mp4',
+            'title': "[V LIVE] Girl's Day's Broadcast",
+            'creator': "Girl's Day",
+            'view_count': int,
+            'uploader_id': 'muploader_a',
+        },
+    },
+        {
+        'url': 'https://www.vlive.tv/video/16937',
         'info_dict': {
             'id': '16937',
             'ext': 'mp4',
-            'title': '첸백시 걍방',
+            'title': '[V LIVE] 첸백시 걍방',
             'creator': 'EXO',
             'view_count': int,
             'subtitles': 'mincount:12',
@@ -66,11 +70,12 @@ class VLiveIE(VLiveBaseIE):
             'subtitles': 'mincount:10',
         },
         'skip': 'This video is only available for CH+ subscribers',
-    }, {
-        'url': 'https://www.vlive.tv/embed/1326',
-        'only_matching': True,
     }]
 
+    @classmethod
+    def suitable(cls, url):
+        return False if VLivePlaylistIE.suitable(url) else super(VLiveIE, cls).suitable(url)
+
     def _real_initialize(self):
         self._login()
 
@@ -102,82 +107,118 @@ class VLiveIE(VLiveBaseIE):
         if not is_logged_in():
             raise ExtractorError('Unable to log in', expected=True)
 
-    def _call_api(self, path_template, video_id, fields=None):
-        query = {'appId': self._APP_ID}
-        if fields:
-            query['fields'] = fields
-        return self._download_json(
-            'https://www.vlive.tv/globalv-web/vam-web/' + path_template % video_id, video_id,
-            'Downloading %s JSON metadata' % path_template.split('/')[-1].split('-')[0],
-            headers={'Referer': 'https://www.vlive.tv/'}, query=query)
-
     def _real_extract(self, url):
-        video_id = self._match_id(url)
+        # url may match on a post or a video url with a post_id potentially matching a video_id
+        working_id = self._match_id(url)
+        webpage = self._download_webpage(url, working_id)
 
-        try:
+        PARAMS_RE = r'window\.__PRELOADED_STATE__\s*=\s*({.*});?\s*</script>'
-            post = self._call_api(
+        PARAMS_FIELD = 'params'
-                'post/v1.0/officialVideoPost-%s', video_id,
-                'author{nickname},channel{channelCode,channelName},officialVideo{commentCount,exposeStatus,likeCount,playCount,playTime,status,title,type,vodId}')
-        except ExtractorError as e:
-            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
-                self.raise_login_required(json.loads(e.cause.read().decode())['message'])
-            raise
 
-        video = post['officialVideo']
+        params = self._search_regex(
+            PARAMS_RE, webpage, PARAMS_FIELD, default='', flags=re.DOTALL)
+        params = self._parse_json(params, working_id, fatal=False)
 
-        def get_common_fields():
+        video_params = try_get(params, lambda x: x["postDetail"]["post"]["officialVideo"], dict)
-            channel = post.get('channel') or {}
-            return {
-                'title': video.get('title'),
-                'creator': post.get('author', {}).get('nickname'),
-                'channel': channel.get('channelName'),
-                'channel_id': channel.get('channelCode'),
-                'duration': int_or_none(video.get('playTime')),
-                'view_count': int_or_none(video.get('playCount')),
-                'like_count': int_or_none(video.get('likeCount')),
-                'comment_count': int_or_none(video.get('commentCount')),
-            }
 
-        video_type = video.get('type')
+        if video_params is None:
-        if video_type == 'VOD':
+            error = try_get(params, lambda x: x["postDetail"]["error"], dict)
-            inkey = self._call_api('video/v1.0/vod/%s/inkey', video_id)['inkey']
+            error_data = try_get(error, lambda x: x["data"], dict)
-            vod_id = video['vodId']
+            error_video = try_get(error_data, lambda x: x["officialVideo"], dict)
-            return merge_dicts(
+            error_msg = try_get(error, lambda x: x["message"], compat_str)
-                get_common_fields(),
+            product_type = try_get(error_data,
-                self._extract_video_info(video_id, vod_id, inkey))
+                                   [lambda x: x["officialVideo"]["productType"],
-        elif video_type == 'LIVE':
+                                    lambda x: x["board"]["boardType"]],
-            status = video.get('status')
+                                   compat_str)
-            if status == 'ON_AIR':
+
-                stream_url = self._call_api(
+            if error_video is not None:
-                    'old/v3/live/%s/playInfo',
+                if product_type in ('VLIVE_PLUS', 'VLIVE+'):
-                    video_id)['result']['adaptiveStreamUrl']
+                    self.raise_login_required('This video is only available with V LIVE+.')
-                formats = self._extract_m3u8_formats(stream_url, video_id, 'mp4')
+                elif error_msg is not None:
-                info = get_common_fields()
+                    raise ExtractorError('V LIVE reported the following error: %s' % error_msg)
-                info.update({
+                else:
-                    'title': self._live_title(video['title']),
+                    raise ExtractorError('Failed to extract video parameters.')
-                    'id': video_id,
+            elif 'post' in url:
-                    'formats': formats,
+                raise ExtractorError('Url does not appear to be a video post.', expected=True)
-                    'is_live': True,
-                })
-                return info
-            elif status == 'ENDED':
-                raise ExtractorError(
-                    'Uploading for replay. Please wait...', expected=True)
-            elif status == 'RESERVED':
-                raise ExtractorError('Coming soon!', expected=True)
-            elif video.get('exposeStatus') == 'CANCEL':
-                raise ExtractorError(
-                    'We are sorry, but the live broadcast has been canceled.',
-                    expected=True)
             else:
-                raise ExtractorError('Unknown status ' + status)
+                raise ExtractorError('Failed to extract video parameters.')
+
+        video_id = working_id if 'video' in url else str(video_params["videoSeq"])
+
+        video_type = video_params["type"]
+        if video_type in ('VOD'):
+            encoding_status = video_params["encodingStatus"]
+            if encoding_status == 'COMPLETE':
+                return self._replay(video_id, webpage, params, video_params)
+            else:
+                raise ExtractorError('VOD encoding not yet complete. Please try again later.',
+                                     expected=True)
+        elif video_type in ('LIVE'):
+            video_status = video_params["status"]
+            if video_status in ('RESERVED'):
+                raise ExtractorError('Coming soon!', expected=True)
+            elif video_status in ('ENDED', 'END'):
+                raise ExtractorError('Uploading for replay. Please wait...', expected=True)
+            else:
+                return self._live(video_id, webpage, params)
+        else:
+            raise ExtractorError('Unknown video type %s' % video_type)
+
+    def _get_common_fields(self, webpage, params):
+        title = self._og_search_title(webpage)
+        description = self._html_search_meta(
+            ['og:description', 'description', 'twitter:description'],
+            webpage, 'description', default=None)
+        creator = (try_get(params, lambda x: x["channel"]["channel"]["channelName"], compat_str)
+                   or self._search_regex(r'on (.*) channel', description or '', 'creator', fatal=False))
+        thumbnail = self._og_search_thumbnail(webpage)
+        return {
+            'title': title,
+            'creator': creator,
+            'thumbnail': thumbnail,
+        }
+
+    def _live(self, video_id, webpage, params):
+        LIVE_INFO_ENDPOINT = 'https://www.vlive.tv/globalv-web/vam-web/old/v3/live/%s/playInfo' % video_id
+        play_info = self._download_json(LIVE_INFO_ENDPOINT, video_id,
+                                        headers={"referer": "https://www.vlive.tv"})
+
+        streams = try_get(play_info, lambda x: x["result"]["streamList"], list) or []
+
+        formats = []
+        for stream in streams:
+            formats.extend(self._extract_m3u8_formats(
+                stream['serviceUrl'], video_id, 'mp4',
+                fatal=False, live=True))
+        self._sort_formats(formats)
+
+        info = self._get_common_fields(webpage, params)
+        info.update({
+            'title': self._live_title(info['title']),
+            'id': video_id,
+            'formats': formats,
+            'is_live': True,
+        })
+        return info
+
+    def _replay(self, video_id, webpage, params, video_params):
+        long_video_id = video_params["vodId"]
+
+        VOD_KEY_ENDPOINT = 'https://www.vlive.tv/globalv-web/vam-web/video/v1.0/vod/%s/inkey' % video_id
+        key_json = self._download_json(VOD_KEY_ENDPOINT, video_id,
+                                       headers={"referer": "https://www.vlive.tv"})
+        key = key_json["inkey"]
+
+        return merge_dicts(
+            self._get_common_fields(webpage, params),
+            self._extract_video_info(video_id, long_video_id, key))
 
 
-class VLiveChannelIE(VLiveBaseIE):
+class VLiveChannelIE(InfoExtractor):
     IE_NAME = 'vlive:channel'
-    _VALID_URL = r'https?://(?:channels\.vlive\.tv|(?:(?:www|m)\.)?vlive\.tv/channel)/(?P<id>[0-9A-Z]+)'
+    _VALID_URL = r'https?://(?:(?:www|m)\.)?(?:channels\.vlive\.tv/|vlive\.tv/channels?/)(?P<id>[0-9A-Z]+)'
     _TESTS = [{
-        'url': 'http://channels.vlive.tv/FCD4B',
+        'url': 'https://channels.vlive.tv/FCD4B',
         'info_dict': {
             'id': 'FCD4B',
             'title': 'MAMAMOO',
@@ -185,39 +226,63 @@ class VLiveChannelIE(VLiveBaseIE):
         'playlist_mincount': 110
     }, {
         'url': 'https://www.vlive.tv/channel/FCD4B',
-        'only_matching': True,
+        'info_dict': {
+            'id': 'FCD4B',
+            'title': 'MAMAMOO',
+        },
+        'playlist_mincount': 110
     }]
-
+    _APP_ID = '8c6cc7b45d2568fb668be6e05b6e5a3b'
-    def _call_api(self, path, channel_key_suffix, channel_value, note, query):
-        q = {
-            'app_id': self._APP_ID,
-            'channel' + channel_key_suffix: channel_value,
-        }
-        q.update(query)
-        return self._download_json(
-            'http://api.vfan.vlive.tv/vproxy/channelplus/' + path,
-            channel_value, note='Downloading ' + note, query=q)['result']
 
     def _real_extract(self, url):
         channel_code = self._match_id(url)
 
-        channel_seq = self._call_api(
+        webpage = self._download_webpage(
-            'decodeChannelCode', 'Code', channel_code,
+            'http://channels.vlive.tv/%s/video' % channel_code, channel_code)
-            'decode channel code', {})['channelSeq']
 
+        app_id = None
+
+        app_js_url = self._search_regex(
+            r'<script[^>]+src=(["\'])(?P<url>http.+?/app\.js.*?)\1',
+            webpage, 'app js', default=None, group='url')
+
+        if app_js_url:
+            app_js = self._download_webpage(
+                app_js_url, channel_code, 'Downloading app JS', fatal=False)
+            if app_js:
+                app_id = self._search_regex(
+                    r'Global\.VFAN_APP_ID\s*=\s*[\'"]([^\'"]+)[\'"]',
+                    app_js, 'app id', default=None)
+
+        app_id = app_id or self._APP_ID
+
+        channel_info = self._download_json(
+            'http://api.vfan.vlive.tv/vproxy/channelplus/decodeChannelCode',
+            channel_code, note='Downloading decode channel code',
+            query={
+                'app_id': app_id,
+                'channelCode': channel_code,
+                '_': int(time.time())
+            })
+
+        channel_seq = channel_info['result']['channelSeq']
         channel_name = None
         entries = []
 
         for page_num in itertools.count(1):
-            video_list = self._call_api(
+            video_list = self._download_json(
-                'getChannelVideoList', 'Seq', channel_seq,
+                'http://api.vfan.vlive.tv/vproxy/channelplus/getChannelVideoList',
-                'channel list page #%d' % page_num, {
+                channel_code, note='Downloading channel list page #%d' % page_num,
+                query={
+                    'app_id': app_id,
+                    'channelSeq': channel_seq,
                     # Large values of maxNumOfRows (~300 or above) may cause
                     # empty responses (see [1]), e.g. this happens for [2] that
                     # has more than 300 videos.
                     # 1. https://github.com/ytdl-org/youtube-dl/issues/13830
                     # 2. http://channels.vlive.tv/EDBF.
                     'maxNumOfRows': 100,
+                    '_': int(time.time()),
                     'pageNo': page_num
                 }
             )
@@ -225,11 +290,11 @@ class VLiveChannelIE(VLiveBaseIE):
             if not channel_name:
                 channel_name = try_get(
                     video_list,
-                    lambda x: x['channelInfo']['channelName'],
+                    lambda x: x['result']['channelInfo']['channelName'],
                     compat_str)
 
             videos = try_get(
-                video_list, lambda x: x['videoList'], list)
+                video_list, lambda x: x['result']['videoList'], list)
             if not videos:
                 break
 
@@ -247,9 +312,7 @@ class VLiveChannelIE(VLiveBaseIE):
             entries, channel_code, channel_name)
 
 
-# old extractor. Rewrite?
+class VLivePlaylistIE(InfoExtractor):
-
-class VLivePlaylistIE(VLiveBaseIE):
     IE_NAME = 'vlive:playlist'
     _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/video/(?P<video_id>[0-9]+)/playlist/(?P<id>[0-9]+)'
     _VIDEO_URL_TEMPLATE = 'http://www.vlive.tv/video/%s'