[nba] rewrite extractor

2 files changed, 385 insertions, 104 deletions

diff --git a/youtube_dl/extractor/extractors.py b/youtube_dl/extractor/extractors.py
index c19633122..c44746e57 100644
--- a/youtube_dl/extractor/extractors.py
+++ b/youtube_dl/extractor/extractors.py
@@ -679,7 +679,14 @@ from .nationalgeographic import (
     NationalGeographicTVIE,
 )
 from .naver import NaverIE
-from .nba import NBAIE
+from .nba import (
+    NBAWatchEmbedIE,
+    NBAWatchIE,
+    NBAWatchCollectionIE,
+    NBAEmbedIE,
+    NBAIE,
+    NBAChannelIE,
+)
 from .nbc import (
     CSNNEIE,
     NBCIE,
diff --git a/youtube_dl/extractor/nba.py b/youtube_dl/extractor/nba.py
index be295a7a3..fbc7adaf4 100644
--- a/youtube_dl/extractor/nba.py
+++ b/youtube_dl/extractor/nba.py
@@ -5,33 +5,137 @@ import re
 
 from .turner import TurnerBaseIE
 from ..compat import (
-    compat_urllib_parse_urlencode,
-    compat_urlparse,
+    compat_parse_qs,
+    compat_str,
+    compat_urllib_parse_unquote,
+    compat_urllib_parse_urlparse,
 )
 from ..utils import (
+    int_or_none,
+    merge_dicts,
     OnDemandPagedList,
-    remove_start,
+    parse_duration,
+    parse_iso8601,
+    try_get,
+    update_url_query,
+    urljoin,
 )
 
 
-class NBAIE(TurnerBaseIE):
-    _VALID_URL = r'https?://(?:watch\.|www\.)?nba\.com/(?P<path>(?:[^/]+/)+(?P<id>[^?]*?))/?(?:/index\.html)?(?:\?.*)?$'
+class NBACVPBaseIE(TurnerBaseIE):
+    def _extract_nba_cvp_info(self, path, video_id, fatal=False):
+        return self._extract_cvp_info(
+            'http://secure.nba.com/%s' % path, video_id, {
+                'default': {
+                    'media_src': 'http://nba.cdn.turner.com/nba/big',
+                },
+                'm3u8': {
+                    'media_src': 'http://nbavod-f.akamaihd.net',
+                },
+            }, fatal=fatal)
+
+
+class NBAWatchBaseIE(NBACVPBaseIE):
+    _VALID_URL_BASE = r'https?://(?:(?:www\.)?nba\.com(?:/watch)?|watch\.nba\.com)/'
+
+    def _extract_video(self, filter_key, filter_value):
+        video = self._download_json(
+            'https://neulionscnbav2-a.akamaihd.net/solr/nbad_program/usersearch',
+            filter_value, query={
+                'fl': 'description,image,name,pid,releaseDate,runtime,tags,seoName',
+                'q': filter_key + ':' + filter_value,
+                'wt': 'json',
+            })['response']['docs'][0]
+
+        video_id = str(video['pid'])
+        title = video['name']
+
+        formats = []
+        m3u8_url = (self._download_json(
+            'https://watch.nba.com/service/publishpoint', video_id, query={
+                'type': 'video',
+                'format': 'json',
+                'id': video_id,
+            }, headers={
+                'User-Agent': 'Mozilla/5.0 (iPhone; CPU iPhone OS 11_0_1 like Mac OS X) AppleWebKit/604.1.38 (KHTML, like Gecko) Version/11.0 Mobile/15A402 Safari/604.1',
+            }, fatal=False) or {}).get('path')
+        if m3u8_url:
+            m3u8_formats = self._extract_m3u8_formats(
+                re.sub(r'_(?:pc|iphone)\.', '.', m3u8_url), video_id, 'mp4',
+                'm3u8_native', m3u8_id='hls', fatal=False)
+            formats.extend(m3u8_formats)
+            for f in m3u8_formats:
+                http_f = f.copy()
+                http_f.update({
+                    'format_id': http_f['format_id'].replace('hls-', 'http-'),
+                    'protocol': 'http',
+                    'url': http_f['url'].replace('.m3u8', ''),
+                })
+                formats.append(http_f)
+
+        info = {
+            'id': video_id,
+            'title': title,
+            'thumbnail': urljoin('https://nbadsdmt.akamaized.net/media/nba/nba/thumbs/', video.get('image')),
+            'description': video.get('description'),
+            'duration': int_or_none(video.get('runtime')),
+            'timestamp': parse_iso8601(video.get('releaseDate')),
+            'tags': video.get('tags'),
+        }
+
+        seo_name = video.get('seoName')
+        if seo_name and re.search(r'\d{4}/\d{2}/\d{2}/', seo_name):
+            base_path = ''
+            if seo_name.startswith('teams/'):
+                base_path += seo_name.split('/')[1] + '/'
+            base_path += 'video/'
+            cvp_info = self._extract_nba_cvp_info(
+                base_path + seo_name + '.xml', video_id, False)
+            if cvp_info:
+                formats.extend(cvp_info['formats'])
+                info = merge_dicts(info, cvp_info)
+
+        self._sort_formats(formats)
+        info['formats'] = formats
+        return info
+
+
+class NBAWatchEmbedIE(NBAWatchBaseIE):
+    IENAME = 'nba:watch:embed'
+    _VALID_URL = NBAWatchBaseIE._VALID_URL_BASE + r'embed\?.*?\bid=(?P<id>\d+)'
+    _TESTS = [{
+        'url': 'http://watch.nba.com/embed?id=659395',
+        'md5': 'b7e3f9946595f4ca0a13903ce5edd120',
+        'info_dict': {
+            'id': '659395',
+            'ext': 'mp4',
+            'title': 'Mix clip: More than 7 points of  Joe Ingles, Luc Mbah a Moute, Blake Griffin and 6 more in Utah Jazz vs. the Clippers, 4/15/2017',
+            'description': 'Mix clip: More than 7 points of  Joe Ingles, Luc Mbah a Moute, Blake Griffin and 6 more in Utah Jazz vs. the Clippers, 4/15/2017',
+            'timestamp': 1492228800,
+            'upload_date': '20170415',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        return self._extract_video('pid', video_id)
+
+
+class NBAWatchIE(NBAWatchBaseIE):
+    IE_NAME = 'nba:watch'
+    _VALID_URL = NBAWatchBaseIE._VALID_URL_BASE + r'(?:nba/)?video/(?P<id>.+?(?=/index\.html)|(?:[^/]+/)*[^/?#&]+)'
     _TESTS = [{
         'url': 'http://www.nba.com/video/games/nets/2012/12/04/0021200253-okc-bkn-recap.nba/index.html',
-        'md5': '9e7729d3010a9c71506fd1248f74e4f4',
+        'md5': '9d902940d2a127af3f7f9d2f3dc79c96',
         'info_dict': {
-            'id': '0021200253-okc-bkn-recap',
+            'id': '70946',
             'ext': 'mp4',
             'title': 'Thunder vs. Nets',
             'description': 'Kevin Durant scores 32 points and dishes out six assists as the Thunder beat the Nets in Brooklyn.',
             'duration': 181,
-            'timestamp': 1354638466,
+            'timestamp': 1354597200,
             'upload_date': '20121204',
         },
-        'params': {
-            # m3u8 download
-            'skip_download': True,
-        },
     }, {
         'url': 'http://www.nba.com/video/games/hornets/2014/12/05/0021400276-nyk-cha-play5.nba/',
         'only_matching': True,
@@ -39,116 +143,286 @@ class NBAIE(TurnerBaseIE):
         'url': 'http://watch.nba.com/video/channels/playoffs/2015/05/20/0041400301-cle-atl-recap.nba',
         'md5': 'b2b39b81cf28615ae0c3360a3f9668c4',
         'info_dict': {
-            'id': 'channels/playoffs/2015/05/20/0041400301-cle-atl-recap.nba',
+            'id': '330865',
             'ext': 'mp4',
             'title': 'Hawks vs. Cavaliers Game 1',
             'description': 'md5:8094c3498d35a9bd6b1a8c396a071b4d',
             'duration': 228,
-            'timestamp': 1432134543,
-            'upload_date': '20150520',
+            'timestamp': 1432094400,
+            'upload_date': '20150521',
         },
-        'expected_warnings': ['Unable to download f4m manifest'],
     }, {
-        'url': 'http://www.nba.com/clippers/news/doc-rivers-were-not-trading-blake',
-        'info_dict': {
-            'id': 'teams/clippers/2016/02/17/1455672027478-Doc_Feb16_720.mov-297324',
-            'ext': 'mp4',
-            'title': 'Practice: Doc Rivers - 2/16/16',
-            'description': 'Head Coach Doc Rivers addresses the media following practice.',
-            'upload_date': '20160216',
-            'timestamp': 1455672000,
-        },
-        'params': {
-            # m3u8 download
-            'skip_download': True,
-        },
-        'expected_warnings': ['Unable to download f4m manifest'],
+        'url': 'http://watch.nba.com/nba/video/channels/nba_tv/2015/06/11/YT_go_big_go_home_Game4_061115',
+        'only_matching': True,
     }, {
-        'url': 'http://www.nba.com/timberwolves/wiggins-shootaround#',
-        'info_dict': {
-            'id': 'timberwolves',
-            'title': 'Shootaround Access - Dec. 12 | Andrew Wiggins',
-        },
-        'playlist_count': 30,
-        'params': {
-            # Download the whole playlist takes too long time
-            'playlist_items': '1-30',
-        },
+        # only CVP mp4 format available
+        'url': 'https://watch.nba.com/video/teams/cavaliers/2012/10/15/sloan121015mov-2249106',
+        'only_matching': True,
     }, {
-        'url': 'http://www.nba.com/timberwolves/wiggins-shootaround#',
+        'url': 'https://watch.nba.com/video/top-100-dunks-from-the-2019-20-season?plsrc=nba&collection=2019-20-season-highlights',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url)
+        collection_id = compat_parse_qs(compat_urllib_parse_urlparse(url).query).get('collection', [None])[0]
+        if collection_id:
+            if self._downloader.params.get('noplaylist'):
+                self.to_screen('Downloading just video %s because of --no-playlist' % display_id)
+            else:
+                self.to_screen('Downloading playlist %s - add --no-playlist to just download video' % collection_id)
+                return self.url_result(
+                    'https://www.nba.com/watch/list/collection/' + collection_id,
+                    NBAWatchCollectionIE.ie_key(), collection_id)
+        return self._extract_video('seoName', display_id)
+
+
+class NBAWatchCollectionIE(NBAWatchBaseIE):
+    IE_NAME = 'nba:watch:collection'
+    _VALID_URL = NBAWatchBaseIE._VALID_URL_BASE + r'list/collection/(?P<id>[^/?#&]+)'
+    _TESTS = [{
+        'url': 'https://watch.nba.com/list/collection/season-preview-2020',
         'info_dict': {
-            'id': 'teams/timberwolves/2014/12/12/Wigginsmp4-3462601',
-            'ext': 'mp4',
-            'title': 'Shootaround Access - Dec. 12 | Andrew Wiggins',
-            'description': 'Wolves rookie Andrew Wiggins addresses the media after Friday\'s shootaround.',
-            'upload_date': '20141212',
-            'timestamp': 1418418600,
+            'id': 'season-preview-2020',
         },
-        'params': {
-            'noplaylist': True,
-            # m3u8 download
-            'skip_download': True,
-        },
-        'expected_warnings': ['Unable to download f4m manifest'],
+        'playlist_mincount': 43,
     }]
+    _PAGE_SIZE = 100
 
-    _PAGE_SIZE = 30
+    def _fetch_page(self, collection_id, page):
+        page += 1
+        videos = self._download_json(
+            'https://content-api-prod.nba.com/public/1/endeavor/video-list/collection/' + collection_id,
+            collection_id, 'Downloading page %d JSON metadata' % page, query={
+                'count': self._PAGE_SIZE,
+                'page': page,
+            })['results']['videos']
+        for video in videos:
+            program = video.get('program') or {}
+            seo_name = program.get('seoName') or program.get('slug')
+            if not seo_name:
+                continue
+            yield {
+                '_type': 'url',
+                'id': program.get('id'),
+                'title': program.get('title') or video.get('title'),
+                'url': 'https://www.nba.com/watch/video/' + seo_name,
+                'thumbnail': video.get('image'),
+                'description': program.get('description') or video.get('description'),
+                'duration': parse_duration(program.get('runtimeHours')),
+                'timestamp': parse_iso8601(video.get('releaseDate')),
+            }
 
-    def _fetch_page(self, team, video_id, page):
-        search_url = 'http://searchapp2.nba.com/nba-search/query.jsp?' + compat_urllib_parse_urlencode({
-            'type': 'teamvideo',
-            'start': page * self._PAGE_SIZE + 1,
-            'npp': (page + 1) * self._PAGE_SIZE + 1,
-            'sort': 'recent',
-            'output': 'json',
-            'site': team,
-        })
-        results = self._download_json(
-            search_url, video_id, note='Download page %d of playlist data' % page)['results'][0]
-        for item in results:
-            yield self.url_result(compat_urlparse.urljoin('http://www.nba.com/', item['url']))
-
-    def _extract_playlist(self, orig_path, video_id, webpage):
-        team = orig_path.split('/')[0]
-
-        if self._downloader.params.get('noplaylist'):
-            self.to_screen('Downloading just video because of --no-playlist')
-            video_path = self._search_regex(
-                r'nbaVideoCore\.firstVideo\s*=\s*\'([^\']+)\';', webpage, 'video path')
-            video_url = 'http://www.nba.com/%s/video/%s' % (team, video_path)
-            return self.url_result(video_url)
-
-        self.to_screen('Downloading playlist - add --no-playlist to just download video')
-        playlist_title = self._og_search_title(webpage, fatal=False)
+    def _real_extract(self, url):
+        collection_id = self._match_id(url)
         entries = OnDemandPagedList(
-            functools.partial(self._fetch_page, team, video_id),
+            functools.partial(self._fetch_page, collection_id),
             self._PAGE_SIZE)
+        return self.playlist_result(entries, collection_id)
 
-        return self.playlist_result(entries, team, playlist_title)
 
-    def _real_extract(self, url):
-        path, video_id = re.match(self._VALID_URL, url).groups()
-        orig_path = path
-        if path.startswith('nba/'):
-            path = path[3:]
+class NBABaseIE(NBACVPBaseIE):
+    _VALID_URL_BASE = r'''(?x)
+        https?://(?:www\.)?nba\.com/
+            (?P<team>
+                blazers|
+                bucks|
+                bulls|
+                cavaliers|
+                celtics|
+                clippers|
+                grizzlies|
+                hawks|
+                heat|
+                hornets|
+                jazz|
+                kings|
+                knicks|
+                lakers|
+                magic|
+                mavericks|
+                nets|
+                nuggets|
+                pacers|
+                pelicans|
+                pistons|
+                raptors|
+                rockets|
+                sixers|
+                spurs|
+                suns|
+                thunder|
+                timberwolves|
+                warriors|
+                wizards
+            )
+        (?:/play\#)?/'''
+    _CHANNEL_PATH_REGEX = r'video/channel|series'
 
-        if 'video/' not in path:
-            webpage = self._download_webpage(url, video_id)
-            path = remove_start(self._search_regex(r'data-videoid="([^"]+)"', webpage, 'video id'), '/')
+    def _embed_url_result(self, team, content_id):
+        return self.url_result(update_url_query(
+            'https://secure.nba.com/assets/amp/include/video/iframe.html', {
+                'contentId': content_id,
+                'team': team,
+            }), NBAEmbedIE.ie_key())
 
-            if path == '{{id}}':
-                return self._extract_playlist(orig_path, video_id, webpage)
+    def _call_api(self, team, content_id, query, resource):
+        return self._download_json(
+            'https://api.nba.net/2/%s/video,imported_video,wsc/' % team,
+            content_id, 'Download %s JSON metadata' % resource,
+            query=query, headers={
+                'accessToken': 'internal|bb88df6b4c2244e78822812cecf1ee1b',
+            })['response']['result']
 
-            # See prepareContentId() of pkgCvp.js
-            if path.startswith('video/teams'):
-                path = 'video/channels/proxy/' + path[6:]
+    def _extract_video(self, video, team, extract_all=True):
+        video_id = compat_str(video['nid'])
+        team = video['brand']
 
-        return self._extract_cvp_info(
-            'http://www.nba.com/%s.xml' % path, video_id, {
-                'default': {
-                    'media_src': 'http://nba.cdn.turner.com/nba/big',
-                },
-                'm3u8': {
-                    'media_src': 'http://nbavod-f.akamaihd.net',
-                },
+        info = {
+            'id': video_id,
+            'title': video.get('title') or video.get('headline') or video['shortHeadline'],
+            'description': video.get('description'),
+            'timestamp': parse_iso8601(video.get('published')),
+        }
+
+        subtitles = {}
+        captions = try_get(video, lambda x: x['videoCaptions']['sidecars'], dict) or {}
+        for caption_url in captions.values():
+            subtitles.setdefault('en', []).append({'url': caption_url})
+
+        formats = []
+        mp4_url = video.get('mp4')
+        if mp4_url:
+            formats.append({
+                'url': mp4_url,
             })
+
+        if extract_all:
+            source_url = video.get('videoSource')
+            if source_url and not source_url.startswith('s3://') and self._is_valid_url(source_url, video_id, 'source'):
+                formats.append({
+                    'format_id': 'source',
+                    'url': source_url,
+                    'preference': 1,
+                })
+
+            m3u8_url = video.get('m3u8')
+            if m3u8_url:
+                if '.akamaihd.net/i/' in m3u8_url:
+                    formats.extend(self._extract_akamai_formats(
+                        m3u8_url, video_id, {'http': 'pmd.cdn.turner.com'}))
+                else:
+                    formats.extend(self._extract_m3u8_formats(
+                        m3u8_url, video_id, 'mp4',
+                        'm3u8_native', m3u8_id='hls', fatal=False))
+
+            content_xml = video.get('contentXml')
+            if team and content_xml:
+                cvp_info = self._extract_nba_cvp_info(
+                    team + content_xml, video_id, fatal=False)
+                if cvp_info:
+                    formats.extend(cvp_info['formats'])
+                    subtitles = self._merge_subtitles(subtitles, cvp_info['subtitles'])
+                    info = merge_dicts(info, cvp_info)
+
+            self._sort_formats(formats)
+        else:
+            info.update(self._embed_url_result(team, video['videoId']))
+
+        info.update({
+            'formats': formats,
+            'subtitles': subtitles,
+        })
+
+        return info
+
+    def _real_extract(self, url):
+        team, display_id = re.match(self._VALID_URL, url).groups()
+        if '/play#/' in url:
+            display_id = compat_urllib_parse_unquote(display_id)
+        else:
+            webpage = self._download_webpage(url, display_id)
+            display_id = self._search_regex(
+                self._CONTENT_ID_REGEX + r'\s*:\s*"([^"]+)"', webpage, 'video id')
+        return self._extract_url_results(team, display_id)
+
+
+class NBAEmbedIE(NBABaseIE):
+    IENAME = 'nba:embed'
+    _VALID_URL = r'https?://secure\.nba\.com/assets/amp/include/video/(?:topI|i)frame\.html\?.*?\bcontentId=(?P<id>[^?#&]+)'
+    _TESTS = [{
+        'url': 'https://secure.nba.com/assets/amp/include/video/topIframe.html?contentId=teams/bulls/2020/12/04/3478774/1607105587854-20201204_SCHEDULE_RELEASE_FINAL_DRUPAL-3478774&team=bulls&adFree=false&profile=71&videoPlayerName=TAMPCVP&baseUrl=&videoAdsection=nba.com_mobile_web_teamsites_chicagobulls&ampEnv=',
+        'only_matching': True,
+    }, {
+        'url': 'https://secure.nba.com/assets/amp/include/video/iframe.html?contentId=2016/10/29/0021600027boschaplay7&adFree=false&profile=71&team=&videoPlayerName=LAMPCVP',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        qs = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
+        content_id = qs['contentId'][0]
+        team = qs.get('team', [None])[0]
+        if not team:
+            return self.url_result(
+                'https://watch.nba.com/video/' + content_id, NBAWatchIE.ie_key())
+        video = self._call_api(team, content_id, {'videoid': content_id}, 'video')[0]
+        return self._extract_video(video, team)
+
+
+class NBAIE(NBABaseIE):
+    IENAME = 'nba'
+    _VALID_URL = NBABaseIE._VALID_URL_BASE + '(?!%s)video/(?P<id>(?:[^/]+/)*[^/?#&]+)' % NBABaseIE._CHANNEL_PATH_REGEX
+    _TESTS = [{
+        'url': 'https://www.nba.com/bulls/video/teams/bulls/2020/12/04/3478774/1607105587854-20201204schedulereleasefinaldrupal-3478774',
+        'info_dict': {
+            'id': '45039',
+            'ext': 'mp4',
+            'title': 'AND WE BACK.',
+            'description': 'Part 1 of our 2020-21 schedule is here! Watch our games on NBC Sports Chicago.',
+            'duration': 94,
+            'timestamp': 1607112000,
+            'upload_date': '20201218',
+        },
+    }, {
+        'url': 'https://www.nba.com/bucks/play#/video/teams%2Fbucks%2F2020%2F12%2F17%2F64860%2F1608252863446-Op_Dream_16x9-64860',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.nba.com/bucks/play#/video/wsc%2Fteams%2F2787C911AA1ACD154B5377F7577CCC7134B2A4B0',
+        'only_matching': True,
+    }]
+    _CONTENT_ID_REGEX = r'videoID'
+
+    def _extract_url_results(self, team, content_id):
+        return self._embed_url_result(team, content_id)
+
+
+class NBAChannelIE(NBABaseIE):
+    IENAME = 'nba:channel'
+    _VALID_URL = NBABaseIE._VALID_URL_BASE + '(?:%s)/(?P<id>[^/?#&]+)' % NBABaseIE._CHANNEL_PATH_REGEX
+    _TESTS = [{
+        'url': 'https://www.nba.com/blazers/video/channel/summer_league',
+        'info_dict': {
+            'title': 'Summer League',
+        },
+        'playlist_mincount': 138,
+    }, {
+        'url': 'https://www.nba.com/bucks/play#/series/On%20This%20Date',
+        'only_matching': True,
+    }]
+    _CONTENT_ID_REGEX = r'videoSubCategory'
+    _PAGE_SIZE = 100
+
+    def _fetch_page(self, team, channel, page):
+        results = self._call_api(team, channel, {
+            'channels': channel,
+            'count': self._PAGE_SIZE,
+            'offset': page * self._PAGE_SIZE,
+        }, 'page %d' % (page + 1))
+        for video in results:
+            yield self._extract_video(video, team, False)
+
+    def _extract_url_results(self, team, content_id):
+        entries = OnDemandPagedList(
+            functools.partial(self._fetch_page, team, content_id),
+            self._PAGE_SIZE)
+        return self.playlist_result(entries, playlist_title=content_id)