]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/vlive.py
Start new release.
[youtubedl] / youtube_dl / extractor / vlive.py
index 86c1cb5ef5d352608416a653bea5ef9f74c4287a..64d0224e6f92779c68e2170564c2b1ad459269ed 100644 (file)
 # coding: utf-8
 from __future__ import unicode_literals
 
-import hmac
-from hashlib import sha1
-from base64 import b64encode
-from time import time
+import re
+import time
+import itertools
 
 from .common import InfoExtractor
+from ..compat import (
+    compat_urllib_parse_urlencode,
+    compat_str,
+)
 from ..utils import (
+    dict_get,
     ExtractorError,
-    determine_ext
+    float_or_none,
+    int_or_none,
+    remove_start,
+    try_get,
+    urlencode_postdata,
 )
-from ..compat import compat_urllib_parse
 
 
 class VLiveIE(InfoExtractor):
     IE_NAME = 'vlive'
-    # www.vlive.tv/video/ links redirect to m.vlive.tv/video/ for mobile devices
-    _VALID_URL = r'https?://(?:(www|m)\.)?vlive\.tv/video/(?P<id>[0-9]+)'
-    _TEST = {
-        'url': 'http://m.vlive.tv/video/1326',
+    _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/video/(?P<id>[0-9]+)'
+    _TESTS = [{
+        'url': 'http://www.vlive.tv/video/1326',
         'md5': 'cc7314812855ce56de70a06a27314983',
         'info_dict': {
             'id': '1326',
             'ext': 'mp4',
-            'title': '[V] Girl\'s Day\'s Broadcast',
-            'creator': 'Girl\'s Day',
+            'title': "[V LIVE] Girl's Day's Broadcast",
+            'creator': "Girl's Day",
+            'view_count': int,
         },
-    }
-    _SECRET = 'rFkwZet6pqk1vQt6SxxUkAHX7YL3lmqzUMrU4IDusTo4jEBdtOhNfT4BYYAdArwH'
+    }, {
+        '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)
 
         webpage = self._download_webpage(
-            'http://m.vlive.tv/video/%s' % video_id,
-            video_id, note='Download video page')
+            'http://www.vlive.tv/video/%s' % video_id, video_id)
+
+        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_')
+
+        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:
+                status = 'COMING_SOON'
+
+        if status == 'LIVE_END':
+            raise ExtractorError('Uploading for replay. Please wait...',
+                                 expected=True)
+        elif status == 'COMING_SOON':
+            raise ExtractorError('Coming soon!', expected=True)
+        elif status == 'CANCELED':
+            raise ExtractorError('We are sorry, '
+                                 'but the live broadcast has been canceled.',
+                                 expected=True)
+        else:
+            raise ExtractorError('Unknown status %s' % status)
+
+    def _get_common_fields(self, webpage):
         title = self._og_search_title(webpage)
-        thumbnail = self._og_search_thumbnail(webpage)
         creator = self._html_search_regex(
-            r'<span[^>]+class="name">([^<>]+)</span>', webpage, 'creator')
-
-        url = 'http://global.apis.naver.com/globalV/globalV/vod/%s/playinfo?' % video_id
-        msgpad = '%.0f' % (time() * 1000)
-        md = b64encode(
-            hmac.new(self._SECRET.encode('ascii'),
-                     (url[:255] + msgpad).encode('ascii'), sha1).digest()
-        )
-        url += '&' + compat_urllib_parse.urlencode({'msgpad': msgpad, 'md': md})
-        playinfo = self._download_json(url, video_id, 'Downloading video json')
+            r'<div[^>]+class="info_area"[^>]*>\s*<a\s+[^>]*>([^<]+)',
+            webpage, 'creator', fatal=False)
+        thumbnail = self._og_search_thumbnail(webpage)
+        return {
+            'title': title,
+            'creator': creator,
+            'thumbnail': thumbnail,
+        }
 
-        if playinfo.get('message', '') != 'success':
-            raise ExtractorError(playinfo.get('message', 'JSON request unsuccessful'))
+    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'
+            })
 
-        if not playinfo.get('result'):
-            raise ExtractorError('No videos found.')
+        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 playinfo['result'].get('videos', {}).get('list', []):
-            formats.append({
-                'url': vid['source'],
-                'ext': 'mp4',
-                'abr': vid.get('bitrate', {}).get('audio'),
-                'vbr': vid.get('bitrate', {}).get('video'),
-                'format_id': vid['encodingOption']['name'],
-                'height': vid.get('height'),
-                'width': vid.get('width'),
-            })
+        for vid in live_params.get('resolutions', []):
+            formats.extend(self._extract_m3u8_formats(
+                vid['cdnUrl'], video_id, 'mp4',
+                m3u8_id=vid.get('name'),
+                fatal=False, live=True))
+        self._sort_formats(formats)
+
+        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(
+            'http://global.apis.naver.com/rmcnmv/rmcnmv/vod_play_videoInfo.json?%s'
+            % compat_urllib_parse_urlencode({
+                'videoId': long_video_id,
+                'key': key,
+                'ptc': 'http',
+                'doct': 'json',  # document type (xml or json)
+                'cpt': 'vtt',  # captions type (vtt or ttml)
+            }), video_id)
+
+        formats = [{
+            'url': vid['source'],
+            'format_id': vid.get('encodingOption', {}).get('name'),
+            'abr': float_or_none(vid.get('bitrate', {}).get('audio')),
+            'vbr': float_or_none(vid.get('bitrate', {}).get('video')),
+            'width': int_or_none(vid.get('encodingOption', {}).get('width')),
+            'height': int_or_none(vid.get('encodingOption', {}).get('height')),
+            'filesize': int_or_none(vid.get('size')),
+        } for vid in playinfo.get('videos', {}).get('list', []) if vid.get('source')]
         self._sort_formats(formats)
 
+        view_count = int_or_none(playinfo.get('meta', {}).get('count'))
+
         subtitles = {}
-        for caption in playinfo['result'].get('captions', {}).get('list', []):
-            subtitles[caption['language']] = [
-                {'ext': determine_ext(caption['source'], default_ext='vtt'),
-                 'url': caption['source']}]
+        for caption in playinfo.get('captions', {}).get('list', []):
+            lang = dict_get(caption, ('locale', 'language', 'country', 'label'))
+            if lang and caption.get('source'):
+                subtitles[lang] = [{
+                    'ext': 'vtt',
+                    'url': caption['source']}]
 
-        return {
+        info = self._get_common_fields(webpage)
+        info.update({
             'id': video_id,
-            'title': title,
-            'creator': creator,
-            'thumbnail': thumbnail,
             '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)