Back-port ITV extractor from yt-dlp PR #1776

Includes commit 6fbcc16, fix by staubichsauger & pukkandan

youtube_dl/extractor/itv.py

@@ -2,11 +2,13 @@
 from __future__ import unicode_literals
 
 import json
-import re
 
 from .common import InfoExtractor
 from .brightcove import BrightcoveNewIE
+
+from ..compat import compat_str
 from ..utils import (
+    base_url,
     clean_html,
     determine_ext,
     extract_attributes,
@@ -18,6 +20,8 @@ from ..utils import (
     smuggle_url,
     try_get,
     url_or_none,
+    url_basename,
+    urljoin,
 )
 
 
@@ -33,7 +37,24 @@ class ITVIE(InfoExtractor):
             'description': 'md5:4d7159af53ebd5b36e8b3ec82a41fdb4',
             'series': 'Plebs',
             'season_number': 1,
-            'episode_number': 1
+            'episode_number': 1,
+            'thumbnail': r're:https?://hubimages\.itv\.com/episode/2_1873_0002'
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
+    }, {
+        'url': 'https://www.itv.com/hub/the-jonathan-ross-show/2a1166a0209',
+        'info_dict': {
+            'id': '2a1166a0209',
+            'ext': 'mp4',
+            'title': 'The Jonathan Ross Show - Series 17 - Episode 8',
+            'description': 'md5:3023dcdd375db1bc9967186cdb3f1399',
+            'series': 'The Jonathan Ross Show',
+            'episode_number': 8,
+            'season_number': 17,
+            'thumbnail': r're:https?://hubimages\.itv\.com/episode/2(?:_\d{4}){2}'
         },
         'params': {
             # m3u8 download
@@ -53,38 +74,16 @@ class ITVIE(InfoExtractor):
         'only_matching': True,
     }]
 
-    def _real_extract(self, url):
+    def _generate_api_headers(self, hmac):
-        video_id = self._match_id(url)
+        return merge_dicts({
-        webpage = self._download_webpage(url, video_id)
-        params = extract_attributes(self._search_regex(
-            r'(?s)(<[^>]+id="video"[^>]*>)', webpage, 'params'))
-        variants = self._parse_json(
-            try_get(params, lambda x: x['data-video-variants'], str) or '{}',
-            video_id, fatal=False)
-        platform_tag_video = None
-        featureset_video = None
-        for platform_tag, featuresets in variants.items():
-            for featureset in featuresets:
-                if (isinstance(featureset, list) and len(featureset) >= 2
-                        and featureset[0] == 'hls' and featureset[1] == 'aes'):
-                    platform_tag_video = platform_tag
-                    featureset_video = featureset
-                    break
-        if not platform_tag_video or not featureset_video:
-            raise ExtractorError(
-                '[%s] %s: %s' % (self.IE_NAME, video_id, 'No downloads available'),
-                expected=True)
-
-        ios_playlist_url = params.get('data-video-playlist') or params['data-video-id']
-        hmac = params['data-video-hmac']
-        headers = self.geo_verification_headers()
-        headers.update({
             'Accept': 'application/vnd.itv.vod.playlist.v2+json',
             'Content-Type': 'application/json',
             'hmac': hmac.upper(),
-        })
+        }, self.geo_verification_headers())
-        ios_playlist = self._download_json(
+
-            ios_playlist_url, video_id, data=json.dumps({
+    def _call_api(self, video_id, playlist_url, headers, platform_tag, featureset, fatal=True):
+        return self._download_json(
+            playlist_url, video_id, data=json.dumps({
                 'user': {
                     'itvUserId': '',
                     'entitlements': [],
@@ -105,15 +104,62 @@ class ITVIE(InfoExtractor):
                 },
                 'variantAvailability': {
                     'featureset': {
-                        'min': featureset_video,
+                        'min': featureset,
-                        'max': featureset_video
+                        'max': featureset
                     },
-                    'platformTag': platform_tag_video
+                    'platformTag': platform_tag
                 }
-            }).encode(), headers=headers)
+            }).encode(), headers=headers, fatal=fatal)
+
+    def _get_subtitles(self, video_id, variants, ios_playlist_url, headers, *args, **kwargs):
+        subtitles = {}
+        # Prefer last matching featureset
+        # See: https://github.com/yt-dlp/yt-dlp/issues/986
+        platform_tag_subs, featureset_subs = next(
+            ((platform_tag, featureset)
+             for platform_tag, featuresets in reversed(tuple(variants.items())) for featureset in featuresets
+             if try_get(featureset, lambda x: x[2]) == 'outband-webvtt'),
+            (None, None))
+
+        if platform_tag_subs and featureset_subs:
+            subs_playlist = self._call_api(
+                video_id, ios_playlist_url, headers, platform_tag_subs, featureset_subs, fatal=False)
+            subs = try_get(subs_playlist, lambda x: x['Playlist']['Video']['Subtitles'], list) or []
+            for sub in subs:
+                if not isinstance(sub, dict):
+                    continue
+                href = url_or_none(sub.get('Href'))
+                if not href:
+                    continue
+                subtitles.setdefault('en', []).append({'url': href})
+        return subtitles
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        params = extract_attributes(self._search_regex(
+            r'(?s)(<[^>]+id="video"[^>]*>)', webpage, 'params'))
+        variants = self._parse_json(
+            try_get(params, lambda x: x['data-video-variants'], compat_str) or '{}',
+            video_id, fatal=False)
+        # Prefer last matching featureset
+        # See: https://github.com/yt-dlp/yt-dlp/issues/986
+        platform_tag_video, featureset_video = next(
+            ((platform_tag, featureset)
+             for platform_tag, featuresets in reversed(tuple(variants.items())) for featureset in featuresets
+             if set(try_get(featureset, lambda x: x[:2]) or {}) == {'aes', 'hls'}),
+            (None, None))
+        if not platform_tag_video or not featureset_video:
+            raise ExtractorError('No downloads available', expected=True, video_id=video_id)
+
+        ios_playlist_url = params.get('data-video-playlist') or params['data-video-id']
+        headers = self._generate_api_headers(params['data-video-hmac'])
+        headers['Referer'] = url
+        ios_playlist = self._call_api(
+            video_id, ios_playlist_url, headers, platform_tag_video, featureset_video)
+
         video_data = try_get(ios_playlist, lambda x: x['Playlist']['Video'], dict) or {}
         ios_base_url = video_data.get('Base')
-
         formats = []
         for media_file in (video_data.get('MediaFiles') or []):
             href = media_file.get('Href')
@@ -131,60 +177,6 @@ class ITVIE(InfoExtractor):
                     'url': href,
                 })
         self._sort_formats(formats)
-
-        subtitles = {}
-        platform_tag_subs = None
-        featureset_subs = None
-        for platform_tag, featuresets in variants.items():
-            for featureset in featuresets:
-                if (isinstance(featureset, list) and len(featureset) >= 3
-                        and featureset[2] == 'outband-webvtt'):
-                    platform_tag_subs = platform_tag
-                    featureset_subs = featureset
-                    break
-        if not platform_tag_subs or not featureset_subs:
-            self.report_warning('%s: %s' % (video_id, 'No subtitles available'))
-        else:
-            subs_playlist = self._download_json(
-                ios_playlist_url, video_id, data=json.dumps({
-                    'user': {
-                        'itvUserId': '',
-                        'entitlements': [],
-                        'token': ''
-                    },
-                    'device': {
-                        'manufacturer': 'Safari',
-                        'model': '5',
-                        'os': {
-                            'name': 'Windows NT',
-                            'version': '6.1',
-                            'type': 'desktop'
-                        }
-                    },
-                    'client': {
-                        'version': '4.1',
-                        'id': 'browser'
-                    },
-                    'variantAvailability': {
-                        'featureset': {
-                            'min': featureset_subs,
-                            'max': featureset_subs
-                        },
-                        'platformTag': platform_tag_subs
-                    }
-                }).encode(), headers=headers)
-            subs = try_get(subs_playlist, lambda x: x['Playlist']['Video']['Subtitles'], list) or []
-            for sub in subs:
-                if not isinstance(sub, dict):
-                    continue
-                href = url_or_none(sub.get('Href'))
-                if not href:
-                    continue
-                subtitles.setdefault('en', []).append({
-                    'url': href,
-                    'ext': determine_ext(href, 'vtt'),
-                })
-
         info = self._search_json_ld(webpage, video_id, default={})
         if not info:
             json_ld = self._parse_json(self._search_regex(
@@ -198,36 +190,73 @@ class ITVIE(InfoExtractor):
                         info = self._json_ld(item, video_id, fatal=False) or {}
                         break
 
+        thumbnails = []
+        thumbnail_url = try_get(params, lambda x: x['data-video-posterframe'], compat_str)
+        if thumbnail_url:
+            thumbnails.extend([{
+                'url': thumbnail_url.format(width=1920, height=1080, quality=100, blur=0, bg='false'),
+                'width': 1920,
+                'height': 1080,
+            }, {
+                'url': urljoin(base_url(thumbnail_url), url_basename(thumbnail_url)),
+                'preference': -2
+            }])
+
+        thumbnail_url = self._html_search_meta(['og:image', 'twitter:image'], webpage, default=None)
+        if thumbnail_url:
+            thumbnails.append({
+                'url': thumbnail_url,
+            })
+        self._remove_duplicate_formats(thumbnails)
+
         return merge_dicts({
             'id': video_id,
             'title': self._html_search_meta(['og:title', 'twitter:title'], webpage),
             'formats': formats,
-            'subtitles': subtitles,
+            'subtitles': self.extract_subtitles(video_id, variants, ios_playlist_url, headers),
             'duration': parse_duration(video_data.get('Duration')),
             'description': clean_html(get_element_by_class('episode-info__synopsis', webpage)),
+            'thumbnails': thumbnails
         }, info)
 
 
 class ITVBTCCIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?itv\.com/btcc/(?:[^/]+/)*(?P<id>[^/?#&]+)'
+    _VALID_URL = r'https?://(?:www\.)?itv\.com/(?:news|btcc)/(?:[^/]+/)*(?P<id>[^/?#&]+)'
-    _TEST = {
+    _TESTS = [{
-        'url': 'http://www.itv.com/btcc/races/btcc-2018-all-the-action-from-brands-hatch',
+        'url': 'https://www.itv.com/btcc/articles/btcc-2019-brands-hatch-gp-race-action',
         'info_dict': {
-            'id': 'btcc-2018-all-the-action-from-brands-hatch',
+            'id': 'btcc-2019-brands-hatch-gp-race-action',
-            'title': 'BTCC 2018: All the action from Brands Hatch',
+            'title': 'BTCC 2019: Brands Hatch GP race action',
         },
-        'playlist_mincount': 9,
+        'playlist_count': 12,
-    }
+    }, {
-    BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/1582188683001/HkiHLnNRx_default/index.html?videoId=%s'
+        'url': 'https://www.itv.com/news/2021-10-27/i-have-to-protect-the-country-says-rishi-sunak-as-uk-faces-interest-rate-hike',
+        'info_dict': {
+            'id': 'i-have-to-protect-the-country-says-rishi-sunak-as-uk-faces-interest-rate-hike',
+            'title': 'md5:6ef054dd9f069330db3dcc66cb772d32'
+        },
+        'playlist_count': 4
+    }]
+    BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/%s/%s_default/index.html?videoId=%s'
 
     def _real_extract(self, url):
         playlist_id = self._match_id(url)
 
         webpage = self._download_webpage(url, playlist_id)
 
-        entries = [
+        json_map = try_get(self._parse_json(self._html_search_regex(
-            self.url_result(
+            '(?s)<script[^>]+id=[\'"]__NEXT_DATA__[^>]*>([^<]+)</script>', webpage, 'json_map'), playlist_id),
-                smuggle_url(self.BRIGHTCOVE_URL_TEMPLATE % video_id, {
+            lambda x: x['props']['pageProps']['article']['body']['content']) or []
+
+        entries = []
+        for video in json_map:
+            if not any(video['data'].get(attr) == 'Brightcove' for attr in ('name', 'type')):
+                continue
+            video_id = video['data']['id']
+            account_id = video['data']['accountId']
+            player_id = video['data']['playerId']
+            entries.append(self.url_result(
+                smuggle_url(self.BRIGHTCOVE_URL_TEMPLATE % (account_id, player_id, video_id), {
                     # ITV does not like some GB IP ranges, so here are some
                     # IP blocks it accepts
                     'geo_ip_blocks': [
@@ -235,8 +264,7 @@ class ITVBTCCIE(InfoExtractor):
                     ],
                     'referrer': url,
                 }),
-                ie=BrightcoveNewIE.ie_key(), video_id=video_id)
+                ie=BrightcoveNewIE.ie_key(), video_id=video_id))
-            for video_id in re.findall(r'data-video-id=["\'](\d+)', webpage)]
 
         title = self._og_search_title(webpage, fatal=False)