Update changelog.

[youtubedl] / youtube_dl / extractor / vlive.py

diff --git a/youtube_dl/extractor/vlive.py b/youtube_dl/extractor/vlive.py
index 8d671cca767d4592a5428f7d3ad855e952df5353..64d0224e6f92779c68e2170564c2b1ad459269ed 100644 (file)
--- a/youtube_dl/extractor/vlive.py
+++ b/youtube_dl/extractor/vlive.py
@@ -2,22 +2,29 @@
 from __future__ import unicode_literals
 
 import re
 from __future__ import unicode_literals
 
 import re

+import time
+import itertools

 
 from .common import InfoExtractor
 
 from .common import InfoExtractor

+from ..compat import (
+    compat_urllib_parse_urlencode,
+    compat_str,
+)

 from ..utils import (
     dict_get,
     ExtractorError,
     float_or_none,
     int_or_none,
     remove_start,
 from ..utils import (
     dict_get,
     ExtractorError,
     float_or_none,
     int_or_none,
     remove_start,

+    try_get,
+    urlencode_postdata,

 )
 )

-from ..compat import compat_urllib_parse_urlencode

 
 
 class VLiveIE(InfoExtractor):
     IE_NAME = 'vlive'
     _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/video/(?P<id>[0-9]+)'
 
 
 class VLiveIE(InfoExtractor):
     IE_NAME = 'vlive'
     _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/video/(?P<id>[0-9]+)'

-    _TEST = {
+    _TESTS = [{

         'url': 'http://www.vlive.tv/video/1326',
         'md5': 'cc7314812855ce56de70a06a27314983',
         'info_dict': {
         'url': 'http://www.vlive.tv/video/1326',
         'md5': 'cc7314812855ce56de70a06a27314983',
         'info_dict': {


@@ -27,7 +34,24 @@ class VLiveIE(InfoExtractor):
             'creator': "Girl's Day",
             'view_count': int,
         },
             'creator': "Girl's Day",
             'view_count': int,
         },

-    }
+    }, {
+        'url': 'http://www.vlive.tv/video/16937',
+        'info_dict': {
+            'id': '16937',
+            'ext': 'mp4',
+            'title': '[V LIVE] 첸백시 걍방',
+            'creator': 'EXO',
+            'view_count': int,
+            'subtitles': 'mincount:12',
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }]
+
+    @classmethod
+    def suitable(cls, url):
+        return False if VLivePlaylistIE.suitable(url) else super(VLiveIE, cls).suitable(url)

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


@@ -35,18 +59,24 @@ class VLiveIE(InfoExtractor):
         webpage = self._download_webpage(
             'http://www.vlive.tv/video/%s' % video_id, video_id)
 
         webpage = self._download_webpage(
             'http://www.vlive.tv/video/%s' % video_id, video_id)
 

-        video_params = self._search_regex(
-            r'\bvlive\.video\.init\(([^)]+)\)',
-            webpage, 'video params')
-        status, _, _, live_params, long_video_id, key = re.split(
-            r'"\s*,\s*"', video_params)[2:8]
+        VIDEO_PARAMS_RE = r'\bvlive\.video\.init\(([^)]+)'
+        VIDEO_PARAMS_FIELD = 'video params'
+
+        params = self._parse_json(self._search_regex(
+            VIDEO_PARAMS_RE, webpage, VIDEO_PARAMS_FIELD, default=''), video_id,
+            transform_source=lambda s: '[' + s + ']', fatal=False)
+
+        if not params or len(params) < 7:
+            params = self._search_regex(
+                VIDEO_PARAMS_RE, webpage, VIDEO_PARAMS_FIELD)
+            params = [p.strip(r'"') for p in re.split(r'\s*,\s*', params)]
+
+        status, long_video_id, key = params[2], params[5], params[6]

         status = remove_start(status, 'PRODUCT_')
 
         status = remove_start(status, 'PRODUCT_')
 

-        if status == 'LIVE_ON_AIR' or status == 'BIG_EVENT_ON_AIR':
-            live_params = self._parse_json('"%s"' % live_params, video_id)
-            live_params = self._parse_json(live_params, video_id)
-            return self._live(video_id, webpage, live_params)
-        elif status == 'VOD_ON_AIR' or status == 'BIG_EVENT_INTRO':
+        if status in ('LIVE_ON_AIR', 'BIG_EVENT_ON_AIR'):
+            return self._live(video_id, webpage)
+        elif status in ('VOD_ON_AIR', 'BIG_EVENT_INTRO'):

             if long_video_id and key:
                 return self._replay(video_id, webpage, long_video_id, key)
             else:
             if long_video_id and key:
                 return self._replay(video_id, webpage, long_video_id, key)
             else:


@@ -76,7 +106,22 @@ class VLiveIE(InfoExtractor):
             'thumbnail': thumbnail,
         }
 
             'thumbnail': thumbnail,
         }
 

-    def _live(self, video_id, webpage, live_params):
+    def _live(self, video_id, webpage):
+        init_page = self._download_webpage(
+            'http://www.vlive.tv/video/init/view',
+            video_id, note='Downloading live webpage',
+            data=urlencode_postdata({'videoSeq': video_id}),
+            headers={
+                'Referer': 'http://www.vlive.tv/video/%s' % video_id,
+                'Content-Type': 'application/x-www-form-urlencoded'
+            })
+
+        live_params = self._search_regex(
+            r'"liveStreamInfo"\s*:\s*(".*"),',
+            init_page, 'live stream info')
+        live_params = self._parse_json(live_params, video_id)
+        live_params = self._parse_json(live_params, video_id)
+

         formats = []
         for vid in live_params.get('resolutions', []):
             formats.extend(self._extract_m3u8_formats(
         formats = []
         for vid in live_params.get('resolutions', []):
             formats.extend(self._extract_m3u8_formats(


@@ -85,10 +130,14 @@ class VLiveIE(InfoExtractor):
                 fatal=False, live=True))
         self._sort_formats(formats)
 
                 fatal=False, live=True))
         self._sort_formats(formats)
 

-        return dict(self._get_common_fields(webpage),
-                    id=video_id,
-                    formats=formats,
-                    is_live=True)
+        info = self._get_common_fields(webpage)
+        info.update({
+            'title': self._live_title(info['title']),
+            'id': video_id,
+            'formats': formats,
+            'is_live': True,
+        })
+        return info

 
     def _replay(self, video_id, webpage, long_video_id, key):
         playinfo = self._download_json(
 
     def _replay(self, video_id, webpage, long_video_id, key):
         playinfo = self._download_json(


@@ -116,14 +165,159 @@ class VLiveIE(InfoExtractor):
 
         subtitles = {}
         for caption in playinfo.get('captions', {}).get('list', []):
 
         subtitles = {}
         for caption in playinfo.get('captions', {}).get('list', []):

-            lang = dict_get(caption, ('language', 'locale', 'country', 'label'))
+            lang = dict_get(caption, ('locale', 'language', 'country', 'label'))

             if lang and caption.get('source'):
                 subtitles[lang] = [{
                     'ext': 'vtt',
                     'url': caption['source']}]
 
             if lang and caption.get('source'):
                 subtitles[lang] = [{
                     'ext': 'vtt',
                     'url': caption['source']}]
 

-        return dict(self._get_common_fields(webpage),
-                    id=video_id,
-                    formats=formats,
-                    view_count=view_count,
-                    subtitles=subtitles)
+        info = self._get_common_fields(webpage)
+        info.update({
+            'id': video_id,
+            'formats': formats,
+            'view_count': view_count,
+            'subtitles': subtitles,
+        })
+        return info
+
+
+class VLiveChannelIE(InfoExtractor):
+    IE_NAME = 'vlive:channel'
+    _VALID_URL = r'https?://channels\.vlive\.tv/(?P<id>[0-9A-Z]+)'
+    _TEST = {
+        'url': 'http://channels.vlive.tv/FCD4B',
+        'info_dict': {
+            'id': 'FCD4B',
+            'title': 'MAMAMOO',
+        },
+        'playlist_mincount': 110
+    }
+    _APP_ID = '8c6cc7b45d2568fb668be6e05b6e5a3b'
+
+    def _real_extract(self, url):
+        channel_code = self._match_id(url)
+
+        webpage = self._download_webpage(
+            'http://channels.vlive.tv/%s/video' % channel_code, channel_code)
+
+        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._download_json(
+                'http://api.vfan.vlive.tv/vproxy/channelplus/getChannelVideoList',
+                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/rg3/youtube-dl/issues/13830
+                    # 2. http://channels.vlive.tv/EDBF.
+                    'maxNumOfRows': 100,
+                    '_': int(time.time()),
+                    'pageNo': page_num
+                }
+            )
+
+            if not channel_name:
+                channel_name = try_get(
+                    video_list,
+                    lambda x: x['result']['channelInfo']['channelName'],
+                    compat_str)
+
+            videos = try_get(
+                video_list, lambda x: x['result']['videoList'], list)
+            if not videos:
+                break
+
+            for video in videos:
+                video_id = video.get('videoSeq')
+                if not video_id:
+                    continue
+                video_id = compat_str(video_id)
+                entries.append(
+                    self.url_result(
+                        'http://www.vlive.tv/video/%s' % video_id,
+                        ie=VLiveIE.ie_key(), video_id=video_id))
+
+        return self.playlist_result(
+            entries, channel_code, channel_name)
+
+
+class VLivePlaylistIE(InfoExtractor):
+    IE_NAME = 'vlive:playlist'
+    _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/video/(?P<video_id>[0-9]+)/playlist/(?P<id>[0-9]+)'
+    _TEST = {
+        'url': 'http://www.vlive.tv/video/22867/playlist/22912',
+        'info_dict': {
+            'id': '22912',
+            'title': 'Valentine Day Message from TWICE'
+        },
+        'playlist_mincount': 9
+    }
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id, playlist_id = mobj.group('video_id', 'id')
+
+        VIDEO_URL_TEMPLATE = 'http://www.vlive.tv/video/%s'
+        if self._downloader.params.get('noplaylist'):
+            self.to_screen(
+                'Downloading just video %s because of --no-playlist' % video_id)
+            return self.url_result(
+                VIDEO_URL_TEMPLATE % video_id,
+                ie=VLiveIE.ie_key(), video_id=video_id)
+
+        self.to_screen(
+            'Downloading playlist %s - add --no-playlist to just download video'
+            % playlist_id)
+
+        webpage = self._download_webpage(
+            'http://www.vlive.tv/video/%s/playlist/%s'
+            % (video_id, playlist_id), playlist_id)
+
+        item_ids = self._parse_json(
+            self._search_regex(
+                r'playlistVideoSeqs\s*=\s*(\[[^]]+\])', webpage,
+                'playlist video seqs'),
+            playlist_id)
+
+        entries = [
+            self.url_result(
+                VIDEO_URL_TEMPLATE % item_id, ie=VLiveIE.ie_key(),
+                video_id=compat_str(item_id))
+            for item_id in item_ids]
+
+        playlist_name = self._html_search_regex(
+            r'<div[^>]+class="[^"]*multicam_playlist[^>]*>\s*<h3[^>]+>([^<]+)',
+            webpage, 'playlist title', fatal=False)
+
+        return self.playlist_result(entries, playlist_id, playlist_name)