X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/fe979149c83b5a935f7d28baf75848a9137316fd..1f17a37b9b95db09a420a1f52cf18723ce4eb8b5:/youtube_dl/extractor/brightcove.py

diff --git a/youtube_dl/extractor/brightcove.py b/youtube_dl/extractor/brightcove.py
index ef560b5..f045050 100644
--- a/youtube_dl/extractor/brightcove.py
+++ b/youtube_dl/extractor/brightcove.py
@@ -1,10 +1,11 @@
-# encoding: utf-8
+# coding: utf-8
 from __future__ import unicode_literals
 
 import re
 import json
 
 from .common import InfoExtractor
+from .adobepass import AdobePassIE
 from ..compat import (
     compat_etree_fromstring,
     compat_parse_qs,
@@ -17,6 +18,7 @@ from ..compat import (
 from ..utils import (
     determine_ext,
     ExtractorError,
+    extract_attributes,
     find_xpath_attr,
     fix_xml_ampersands,
     float_or_none,
@@ -26,6 +28,8 @@ from ..utils import (
     unescapeHTML,
     unsmuggle_url,
     update_url_query,
+    clean_html,
+    mimetype2ext,
 )
 
 
@@ -90,6 +94,7 @@ class BrightcoveLegacyIE(InfoExtractor):
                 'description': 'md5:363109c02998fee92ec02211bd8000df',
                 'uploader': 'National Ballet of Canada',
             },
+            'skip': 'Video gone',
         },
         {
             # test flv videos served by akamaihd.net
@@ -106,9 +111,10 @@ class BrightcoveLegacyIE(InfoExtractor):
                 'upload_date': '20140827',
                 'uploader_id': '710858724001',
             },
+            'skip': 'Video gone',
         },
         {
-            # playlist test
+            # playlist with 'videoList'
             # from http://support.brightcove.com/en/video-cloud/docs/playlist-support-single-video-players
             'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=3550052898001&playerKey=AQ%7E%7E%2CAAABmA9XpXk%7E%2C-Kp7jNgisre1fG5OdqpAFUTcs0lP_ZoL',
             'info_dict': {
@@ -117,6 +123,21 @@ class BrightcoveLegacyIE(InfoExtractor):
             },
             'playlist_mincount': 7,
         },
+        {
+            # playlist with 'playlistTab' (https://github.com/rg3/youtube-dl/issues/9965)
+            'url': 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=AQ%7E%7E,AAABXlLMdok%7E,NJ4EoMlZ4rZdx9eU1rkMVd8EaYPBBUlg',
+            'info_dict': {
+                'id': '1522758701001',
+                'title': 'Lesson 08',
+            },
+            'playlist_mincount': 10,
+        },
+        {
+            # playerID inferred from bcpid
+            # from http://www.un.org/chinese/News/story.asp?NewsID=27724
+            'url': 'https://link.brightcove.com/services/player/bcpid1722935254001/?bctid=5360463607001&autoStart=false&secureConnections=true&width=650&height=350',
+            'only_matching': True,  # Tested in GenericIE
+        }
     ]
     FLV_VCODECS = {
         1: 'SORENSON',
@@ -167,7 +188,7 @@ class BrightcoveLegacyIE(InfoExtractor):
 
         params = {}
 
-        playerID = find_param('playerID')
+        playerID = find_param('playerID') or find_param('playerId')
         if playerID is None:
             raise ExtractorError('Cannot find player ID')
         params['playerID'] = playerID
@@ -179,6 +200,16 @@ class BrightcoveLegacyIE(InfoExtractor):
         # These fields hold the id of the video
         videoPlayer = find_param('@videoPlayer') or find_param('videoId') or find_param('videoID') or find_param('@videoList')
         if videoPlayer is not None:
+            if isinstance(videoPlayer, list):
+                videoPlayer = videoPlayer[0]
+            videoPlayer = videoPlayer.strip()
+            # UUID is also possible for videoPlayer (e.g.
+            # http://www.popcornflix.com/hoodies-vs-hooligans/7f2d2b87-bbf2-4623-acfb-ea942b4f01dd
+            # or http://www8.hp.com/cn/zh/home.html)
+            if not (re.match(
+                    r'^(?:\d+|[\da-fA-F]{8}-?[\da-fA-F]{4}-?[\da-fA-F]{4}-?[\da-fA-F]{4}-?[\da-fA-F]{12})$',
+                    videoPlayer) or videoPlayer.startswith('ref:')):
+                return None
             params['@videoPlayer'] = videoPlayer
         linkBase = find_param('linkBaseURL')
         if linkBase is not None:
@@ -192,7 +223,7 @@ class BrightcoveLegacyIE(InfoExtractor):
         #   // build Brightcove <object /> XML
         # }
         m = re.search(
-            r'''(?x)customBC.\createVideo\(
+            r'''(?x)customBC\.createVideo\(
                 .*?                                                  # skipping width and height
                 ["\'](?P<playerID>\d+)["\']\s*,\s*                   # playerID
                 ["\'](?P<playerKey>AQ[^"\']{48})[^"\']*["\']\s*,\s*  # playerKey begins with AQ and is 50 characters
@@ -220,13 +251,16 @@ class BrightcoveLegacyIE(InfoExtractor):
         """Return a list of all Brightcove URLs from the webpage """
 
         url_m = re.search(
-            r'<meta\s+property=[\'"]og:video[\'"]\s+content=[\'"](https?://(?:secure|c)\.brightcove.com/[^\'"]+)[\'"]',
-            webpage)
+            r'''(?x)
+                <meta\s+
+                    (?:property|itemprop)=([\'"])(?:og:video|embedURL)\1[^>]+
+                    content=([\'"])(?P<url>https?://(?:secure|c)\.brightcove.com/(?:(?!\2).)+)\2
+            ''', webpage)
         if url_m:
-            url = unescapeHTML(url_m.group(1))
+            url = unescapeHTML(url_m.group('url'))
             # Some sites don't add it, we can't download with this url, for example:
             # http://www.ktvu.com/videos/news/raw-video-caltrain-releases-video-of-man-almost/vCTZdY/
-            if 'playerKey' in url or 'videoId' in url:
+            if 'playerKey' in url or 'videoId' in url or 'idVideo' in url:
                 return [url]
 
         matches = re.findall(
@@ -239,15 +273,19 @@ class BrightcoveLegacyIE(InfoExtractor):
         if matches:
             return list(filter(None, [cls._build_brighcove_url(m) for m in matches]))
 
-        return list(filter(None, [
-            cls._build_brighcove_url_from_js(custom_bc)
-            for custom_bc in re.findall(r'(customBC\.createVideo\(.+?\);)', webpage)]))
+        matches = re.findall(r'(customBC\.createVideo\(.+?\);)', webpage)
+        if matches:
+            return list(filter(None, [
+                cls._build_brighcove_url_from_js(custom_bc)
+                for custom_bc in matches]))
+        return [src for _, src in re.findall(
+            r'<iframe[^>]+src=([\'"])((?:https?:)?//link\.brightcove\.com/services/player/(?!\1).+)\1', webpage)]
 
     def _real_extract(self, url):
         url, smuggled_data = unsmuggle_url(url, {})
 
         # Change the 'videoId' and others field to '@videoPlayer'
-        url = re.sub(r'(?<=[?&])(videoI(d|D)|bctid)', '%40videoPlayer', url)
+        url = re.sub(r'(?<=[?&])(videoI(d|D)|idVideo|bctid)', '%40videoPlayer', url)
         # Change bckey (used by bcove.me urls) to playerKey
         url = re.sub(r'(?<=[?&])bckey', 'playerKey', url)
         mobj = re.match(self._VALID_URL, url)
@@ -258,6 +296,10 @@ class BrightcoveLegacyIE(InfoExtractor):
         if videoPlayer:
             # We set the original url as the default 'Referer' header
             referer = smuggled_data.get('Referer', url)
+            if 'playerID' not in query:
+                mobj = re.search(r'/bcpid(\d+)', url)
+                if mobj is not None:
+                    query['playerID'] = [mobj.group(1)]
             return self._get_video_info(
                 videoPlayer[0], query, referer=referer)
         elif 'playerKey' in query:
@@ -298,13 +340,19 @@ class BrightcoveLegacyIE(InfoExtractor):
             info_url, player_key, 'Downloading playlist information')
 
         json_data = json.loads(playlist_info)
-        if 'videoList' not in json_data:
+        if 'videoList' in json_data:
+            playlist_info = json_data['videoList']
+            playlist_dto = playlist_info['mediaCollectionDTO']
+        elif 'playlistTabs' in json_data:
+            playlist_info = json_data['playlistTabs']
+            playlist_dto = playlist_info['lineupListDTO']['playlistDTOs'][0]
+        else:
             raise ExtractorError('Empty playlist')
-        playlist_info = json_data['videoList']
-        videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']]
+
+        videos = [self._extract_video_info(video_info) for video_info in playlist_dto['videoDTOs']]
 
         return self.playlist_result(videos, playlist_id='%s' % playlist_info['id'],
-                                    playlist_title=playlist_info['mediaCollectionDTO']['displayName'])
+                                    playlist_title=playlist_dto['displayName'])
 
     def _extract_video_info(self, video_info):
         video_id = compat_str(video_info['id'])
@@ -401,7 +449,7 @@ class BrightcoveLegacyIE(InfoExtractor):
         return info
 
 
-class BrightcoveNewIE(InfoExtractor):
+class BrightcoveNewIE(AdobePassIE):
     IE_NAME = 'brightcove:new'
     _VALID_URL = r'https?://players\.brightcove\.net/(?P<account_id>\d+)/(?P<player_id>[^/]+)_(?P<embed>[^/]+)/index\.html\?.*videoId=(?P<video_id>\d+|ref:[^&]+)'
     _TESTS = [{
@@ -416,7 +464,7 @@ class BrightcoveNewIE(InfoExtractor):
             'timestamp': 1441391203,
             'upload_date': '20150904',
             'uploader_id': '929656772001',
-            'formats': 'mincount:22',
+            'formats': 'mincount:20',
         },
     }, {
         # with rtmp streams
@@ -430,7 +478,7 @@ class BrightcoveNewIE(InfoExtractor):
             'timestamp': 1433556729,
             'upload_date': '20150606',
             'uploader_id': '4036320279001',
-            'formats': 'mincount:41',
+            'formats': 'mincount:39',
         },
         'params': {
             # m3u8 download
@@ -451,17 +499,18 @@ class BrightcoveNewIE(InfoExtractor):
     }]
 
     @staticmethod
-    def _extract_url(webpage):
-        urls = BrightcoveNewIE._extract_urls(webpage)
+    def _extract_url(ie, webpage):
+        urls = BrightcoveNewIE._extract_urls(ie, webpage)
         return urls[0] if urls else None
 
     @staticmethod
-    def _extract_urls(webpage):
+    def _extract_urls(ie, webpage):
         # Reference:
         # 1. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideoiniframe
-        # 2. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideousingjavascript
-        # 3. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/embed-in-page.html
-        # 4. https://support.brightcove.com/en/video-cloud/docs/dynamically-assigning-videos-player
+        # 2. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#tag
+        # 3. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideousingjavascript
+        # 4. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/in-page-embed-player-implementation.html
+        # 5. https://support.brightcove.com/en/video-cloud/docs/dynamically-assigning-videos-player
 
         entries = []
 
@@ -470,72 +519,67 @@ class BrightcoveNewIE(InfoExtractor):
                 r'<iframe[^>]+src=(["\'])((?:https?:)?//players\.brightcove\.net/\d+/[^/]+/index\.html.+?)\1', webpage):
             entries.append(url if url.startswith('http') else 'http:' + url)
 
-        # Look for embed_in_page embeds [2]
-        for video_id, account_id, player_id, embed in re.findall(
-                # According to examples from [3] it's unclear whether video id
-                # may be optional and what to do when it is
-                # According to [4] data-video-id may be prefixed with ref:
-                r'''(?sx)
-                    <video[^>]+
-                        data-video-id=["\'](\d+|ref:[^"\']+)["\'][^>]*>.*?
-                    </video>.*?
-                    <script[^>]+
-                        src=["\'](?:https?:)?//players\.brightcove\.net/
-                        (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
+        # Look for <video> tags [2] and embed_in_page embeds [3]
+        # [2] looks like:
+        for video, script_tag, account_id, player_id, embed in re.findall(
+                r'''(?isx)
+                    (<video\s+[^>]*\bdata-video-id\s*=\s*['"]?[^>]+>)
+                    (?:.*?
+                        (<script[^>]+
+                            src=["\'](?:https?:)?//players\.brightcove\.net/
+                            (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
+                        )
+                    )?
                 ''', webpage):
-            entries.append(
-                'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s'
-                % (account_id, player_id, embed, video_id))
+            attrs = extract_attributes(video)
 
-        return entries
+            # According to examples from [4] it's unclear whether video id
+            # may be optional and what to do when it is
+            video_id = attrs.get('data-video-id')
+            if not video_id:
+                continue
 
-    def _real_extract(self, url):
-        account_id, player_id, embed, video_id = re.match(self._VALID_URL, url).groups()
+            account_id = account_id or attrs.get('data-account')
+            if not account_id:
+                continue
 
-        webpage = self._download_webpage(
-            'http://players.brightcove.net/%s/%s_%s/index.min.js'
-            % (account_id, player_id, embed), video_id)
+            player_id = player_id or attrs.get('data-player') or 'default'
+            embed = embed or attrs.get('data-embed') or 'default'
 
-        policy_key = None
+            bc_url = 'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s' % (
+                account_id, player_id, embed, video_id)
 
-        catalog = self._search_regex(
-            r'catalog\(({.+?})\);', webpage, 'catalog', default=None)
-        if catalog:
-            catalog = self._parse_json(
-                js_to_json(catalog), video_id, fatal=False)
-            if catalog:
-                policy_key = catalog.get('policyKey')
+            # Some brightcove videos may be embedded with video tag only and
+            # without script tag or any mentioning of brightcove at all. Such
+            # embeds are considered ambiguous since they are matched based only
+            # on data-video-id and data-account attributes and in the wild may
+            # not be brightcove embeds at all. Let's check reconstructed
+            # brightcove URLs in case of such embeds and only process valid
+            # ones. By this we ensure there is indeed a brightcove embed.
+            if not script_tag and not ie._is_valid_url(
+                    bc_url, video_id, 'possible brightcove video'):
+                continue
 
-        if not policy_key:
-            policy_key = self._search_regex(
-                r'policyKey\s*:\s*(["\'])(?P<pk>.+?)\1',
-                webpage, 'policy key', group='pk')
+            entries.append(bc_url)
 
-        api_url = 'https://edge.api.brightcove.com/playback/v1/accounts/%s/videos/%s' % (account_id, video_id)
-        try:
-            json_data = self._download_json(api_url, video_id, headers={
-                'Accept': 'application/json;pk=%s' % policy_key
-            })
-        except ExtractorError as e:
-            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
-                json_data = self._parse_json(e.cause.read().decode(), video_id)[0]
-                raise ExtractorError(
-                    json_data.get('message') or json_data['error_code'], expected=True)
-            raise
+        return entries
 
+    def _parse_brightcove_metadata(self, json_data, video_id):
         title = json_data['name'].strip()
 
         formats = []
         for source in json_data.get('sources', []):
             container = source.get('container')
-            source_type = source.get('type')
+            ext = mimetype2ext(source.get('type'))
             src = source.get('src')
-            if source_type == 'application/x-mpegURL' or container == 'M2TS':
+            if ext == 'ism' or container == 'WVM':
+                continue
+            elif ext == 'm3u8' or container == 'M2TS':
                 if not src:
                     continue
                 formats.extend(self._extract_m3u8_formats(
                     src, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
-            elif source_type == 'application/dash+xml':
+            elif ext == 'mpd':
                 if not src:
                     continue
                 formats.extend(self._extract_mpd_formats(src, video_id, 'dash', fatal=False))
@@ -551,7 +595,7 @@ class BrightcoveNewIE(InfoExtractor):
                     'tbr': tbr,
                     'filesize': int_or_none(source.get('size')),
                     'container': container,
-                    'ext': container.lower(),
+                    'ext': ext or container.lower(),
                 }
                 if width == 0 and height == 0:
                     f.update({
@@ -585,6 +629,13 @@ class BrightcoveNewIE(InfoExtractor):
                         'format_id': build_format_id('rtmp'),
                     })
                 formats.append(f)
+
+        errors = json_data.get('errors')
+        if not formats and errors:
+            error = errors[0]
+            raise ExtractorError(
+                error.get('message') or error.get('error_subcode') or error['error_code'], expected=True)
+
         self._sort_formats(formats)
 
         subtitles = {}
@@ -594,15 +645,76 @@ class BrightcoveNewIE(InfoExtractor):
                     'url': text_track['src'],
                 })
 
+        is_live = False
+        duration = float_or_none(json_data.get('duration'), 1000)
+        if duration is not None and duration <= 0:
+            is_live = True
+
         return {
             'id': video_id,
-            'title': title,
-            'description': json_data.get('description'),
+            'title': self._live_title(title) if is_live else title,
+            'description': clean_html(json_data.get('description')),
             'thumbnail': json_data.get('thumbnail') or json_data.get('poster'),
-            'duration': float_or_none(json_data.get('duration'), 1000),
+            'duration': duration,
             'timestamp': parse_iso8601(json_data.get('published_at')),
-            'uploader_id': account_id,
+            'uploader_id': json_data.get('account_id'),
             'formats': formats,
             'subtitles': subtitles,
             'tags': json_data.get('tags', []),
+            'is_live': is_live,
         }
+
+    def _real_extract(self, url):
+        url, smuggled_data = unsmuggle_url(url, {})
+        self._initialize_geo_bypass(smuggled_data.get('geo_countries'))
+
+        account_id, player_id, embed, video_id = re.match(self._VALID_URL, url).groups()
+
+        webpage = self._download_webpage(
+            'http://players.brightcove.net/%s/%s_%s/index.min.js'
+            % (account_id, player_id, embed), video_id)
+
+        policy_key = None
+
+        catalog = self._search_regex(
+            r'catalog\(({.+?})\);', webpage, 'catalog', default=None)
+        if catalog:
+            catalog = self._parse_json(
+                js_to_json(catalog), video_id, fatal=False)
+            if catalog:
+                policy_key = catalog.get('policyKey')
+
+        if not policy_key:
+            policy_key = self._search_regex(
+                r'policyKey\s*:\s*(["\'])(?P<pk>.+?)\1',
+                webpage, 'policy key', group='pk')
+
+        api_url = 'https://edge.api.brightcove.com/playback/v1/accounts/%s/videos/%s' % (account_id, video_id)
+        try:
+            json_data = self._download_json(api_url, video_id, headers={
+                'Accept': 'application/json;pk=%s' % policy_key
+            })
+        except ExtractorError as e:
+            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
+                json_data = self._parse_json(e.cause.read().decode(), video_id)[0]
+                message = json_data.get('message') or json_data['error_code']
+                if json_data.get('error_subcode') == 'CLIENT_GEO':
+                    self.raise_geo_restricted(msg=message)
+                raise ExtractorError(message, expected=True)
+            raise
+
+        errors = json_data.get('errors')
+        if errors and errors[0].get('error_subcode') == 'TVE_AUTH':
+            custom_fields = json_data['custom_fields']
+            tve_token = self._extract_mvpd_auth(
+                smuggled_data['source_url'], video_id,
+                custom_fields['bcadobepassrequestorid'],
+                custom_fields['bcadobepassresourceid'])
+            json_data = self._download_json(
+                api_url, video_id, headers={
+                    'Accept': 'application/json;pk=%s' % policy_key
+                }, query={
+                    'tveToken': tve_token,
+                })
+
+        return self._parse_brightcove_metadata(json_data, video_id)