[ITV] Overhaul ITV extractor (#30266)

* support ITVX URLs (thanks Vangelis66)
* support legacy ITV Hub URLs
* include extraction fix 4c57dd2 from sleaux-meaux 3 May 2021
* include extraction fix 6fbcc16, fix by staubichsauger & pukkandan
* work-around duration parsing pending fix to utils.parse_duration
* apply default vanilla UA for pages and media to avoid site blocking
* also detect and report `Episode not found` instead of generic 404
* rework ITVBTCCIE with geo-block detection, best effort geo-restriction handling, news article support
* fix tests

youtube_dl/extractor/itv.py

@@ -3,123 +3,266 @@ from __future__ import unicode_literals
 
 import json
 import re
+import sys
 
 from .common import InfoExtractor
 from .brightcove import BrightcoveNewIE
+from ..compat import (
+    compat_HTTPError,
+    compat_integer_types,
+    compat_kwargs,
+    compat_urlparse,
+)
 from ..utils import (
     clean_html,
     determine_ext,
+    error_to_compat_str,
     extract_attributes,
-    get_element_by_class,
+    ExtractorError,
-    JSON_LD_RE,
+    get_element_by_attribute,
+    int_or_none,
     merge_dicts,
     parse_duration,
+    parse_iso8601,
+    remove_start,
     smuggle_url,
+    strip_or_none,
+    traverse_obj,
     url_or_none,
+    urljoin,
 )
 
 
-class ITVIE(InfoExtractor):
+class ITVBaseIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?itv\.com/hub/[^/]+/(?P<id>[0-9a-zA-Z]+)'
+
-    _GEO_COUNTRIES = ['GB']
+    def _search_nextjs_data(self, webpage, video_id, **kw):
+        transform_source = kw.pop('transform_source', None)
+        fatal = kw.pop('fatal', True)
+        return self._parse_json(
+            self._search_regex(
+                r'''<script\b[^>]+\bid=('|")__NEXT_DATA__\1[^>]*>(?P<js>[^<]+)</script>''',
+                webpage, 'next.js data', group='js', fatal=fatal, **kw),
+            video_id, transform_source=transform_source, fatal=fatal)
+
+    def __handle_request_webpage_error(self, err, video_id=None, errnote=None, fatal=True):
+        if errnote is False:
+            return False
+        if errnote is None:
+            errnote = 'Unable to download webpage'
+
+        errmsg = '%s: %s' % (errnote, error_to_compat_str(err))
+        if fatal:
+            raise ExtractorError(errmsg, sys.exc_info()[2], cause=err, video_id=video_id)
+        else:
+            self._downloader.report_warning(errmsg)
+            return False
+
+    @staticmethod
+    def _vanilla_ua_header():
+        return {'User-agent': 'Mozilla/5.0'}
+
+    def _download_webpage_handle(self, url, video_id, *args, **kwargs):
+        # specialised to (a) use vanilla UA (b) detect geo-block
+        params = self._downloader.params
+        nkwargs = {}
+        if (
+                'user_agent' not in params
+                and not any(re.match(r'(?i)user-agent\s*:', h)
+                            for h in (params.get('headers') or []))
+                and 'User-agent' not in (kwargs.get('headers') or {})):
+
+            kwargs.setdefault('headers', {})
+            kwargs['headers'] = self._vanilla_ua_header()
+            nkwargs = kwargs
+        if kwargs.get('expected_status') is not None:
+            exp = kwargs['expected_status']
+            if isinstance(exp, compat_integer_types):
+                exp = [exp]
+            if isinstance(exp, (list, tuple)) and 403 not in exp:
+                kwargs['expected_status'] = [403]
+                kwargs['expected_status'].extend(exp)
+                nkwargs = kwargs
+        else:
+            kwargs['expected_status'] = 403
+            nkwargs = kwargs
+
+        if nkwargs:
+            kwargs = compat_kwargs(kwargs)
+
+        ret = super(ITVBaseIE, self)._download_webpage_handle(url, video_id, *args, **kwargs)
+        if ret is False:
+            return ret
+        webpage, urlh = ret
+
+        if urlh.getcode() == 403:
+            # geo-block error is like this, with an unnecessary 'Of':
+            # '{\n  "Message" : "Request Originated Outside Of Allowed Geographic Region",\
+            # \n  "TransactionId" : "oas-magni-475082-xbYF0W"\n}'
+            if '"Request Originated Outside Of Allowed Geographic Region"' in webpage:
+                self.raise_geo_restricted(countries=['GB'])
+            ret = self.__handle_request_webpage_error(
+                compat_HTTPError(urlh.geturl(), 403, 'HTTP Error 403: Forbidden', urlh.headers, urlh),
+                fatal=kwargs.get('fatal'))
+
+        return ret
+
+
+class ITVIE(ITVBaseIE):
+    _VALID_URL = r'https?://(?:www\.)?itv\.com/(?:(?P<w>watch)|hub)/[^/]+/(?(w)[\w-]+/)(?P<id>\w+)'
+    _IE_DESC = 'ITVX'
     _TESTS = [{
+        'note': 'Hub URLs redirect to ITVX',
         'url': 'https://www.itv.com/hub/liar/2a4547a0012',
-        'info_dict': {
+        'only_matching': True,
-            'id': '2a4547a0012',
-            'ext': 'mp4',
-            'title': 'Liar - Series 2 - Episode 6',
-            'description': 'md5:d0f91536569dec79ea184f0a44cca089',
-            'series': 'Liar',
-            'season_number': 2,
-            'episode_number': 6,
-        },
-        'params': {
-            # m3u8 download
-            'skip_download': True,
-        },
     }, {
-        # unavailable via data-playlist-url
+        'note': 'Hub page unavailable via data-playlist-url (404 now)',
         'url': 'https://www.itv.com/hub/through-the-keyhole/2a2271a0033',
         'only_matching': True,
     }, {
-        # InvalidVodcrid
+        'note': 'Hub page with InvalidVodcrid (404 now)',
         'url': 'https://www.itv.com/hub/james-martins-saturday-morning/2a5159a0034',
         'only_matching': True,
     }, {
-        # ContentUnavailable
+        'note': 'Hub page with ContentUnavailable (404 now)',
         'url': 'https://www.itv.com/hub/whos-doing-the-dishes/2a2898a0024',
         'only_matching': True,
-    }]
+    }, {
+        'note': 'ITVX, or itvX, show',
+        'url': 'https://www.itv.com/watch/vera/1a7314/1a7314a0014',
+        'md5': 'bd0ad666b2c058fffe7d036785880064',
+        'info_dict': {
+            'id': '1a7314a0014',
+            'ext': 'mp4',
+            'title': 'Vera - Series 3 - Episode 4 - Prodigal Son',
+            'description': 'Vera and her team investigate the fatal stabbing of an ex-Met police officer outside a busy Newcastle nightclub - but there aren\'t many clues.',
+            'timestamp': 1653591600,
+            'upload_date': '20220526',
+            'uploader': 'ITVX',
+            'thumbnail': r're:https://\w+\.itv\.com/images/(?:\w+/)+\d+x\d+\?',
+            'duration': 5340.8,
+            'age_limit': 16,
+            'series': 'Vera',
+            'series_number': 3,
+            'episode': 'Prodigal Son',
+            'episode_number': 4,
+            'channel': 'ITV3',
+            'categories': list,
+        },
+        'params': {
+            # m3u8 download
+            # 'skip_download': True,
+        },
+        'skip': 'only available in UK',
+    }, {
+        'note': 'Latest ITV news bulletin: details change daily',
+        'url': 'https://www.itv.com/watch/news/varies-but-is-not-checked/6js5d0f',
+        'info_dict': {
+            'id': '6js5d0f',
+            'ext': 'mp4',
+            'title': r're:The latest ITV News headlines - \S.+',
+            'description': r'''re:.* today's top stories from the ITV News team.$''',
+            'timestamp': int,
+            'upload_date': r're:2\d\d\d(?:0[1-9]|1[0-2])(?:[012][1-9]|3[01])',
+            'uploader': 'ITVX',
+            'thumbnail': r're:https://images\.ctfassets\.net/(?:\w+/)+[\w.]+\.(?:jpg|png)',
+            'duration': float,
+            'age_limit': None,
+        },
+        'params': {
+            # variable download
+            # 'skip_download': True,
+        },
+        'skip': 'only available in UK',
+    }
+    ]
+
+    def _og_extract(self, webpage, require_title=False):
+        return {
+            'title': self._og_search_title(webpage, fatal=require_title),
+            'description': self._og_search_description(webpage, default=None),
+            'thumbnail': self._og_search_thumbnail(webpage, default=None),
+            'uploader': self._og_search_property('site_name', webpage, default=None),
+        }
 
     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'))
 
-        ios_playlist_url = params.get('data-video-playlist') or params['data-video-id']
+        webpage = self._download_webpage(url, video_id)
-        hmac = params['data-video-hmac']
+
+        # now quite different params!
+        params = extract_attributes(self._search_regex(
+            r'''(<[^>]+\b(?:class|data-testid)\s*=\s*("|')genie-container\2[^>]*>)''',
+            webpage, 'params'))
+
+        ios_playlist_url = traverse_obj(
+            params, 'data-video-id', 'data-video-playlist',
+            get_all=False, expected_type=url_or_none)
+
         headers = self.geo_verification_headers()
         headers.update({
             'Accept': 'application/vnd.itv.vod.playlist.v2+json',
             'Content-Type': 'application/json',
-            'hmac': hmac.upper(),
         })
         ios_playlist = self._download_json(
             ios_playlist_url, video_id, data=json.dumps({
                 'user': {
-                    'itvUserId': '',
                     'entitlements': [],
-                    'token': ''
                 },
                 'device': {
-                    'manufacturer': 'Safari',
+                    'manufacturer': 'Mobile Safari',
-                    'model': '5',
+                    'model': '5.1',
                     'os': {
-                        'name': 'Windows NT',
+                        'name': 'iOS',
-                        'version': '6.1',
+                        'version': '5.0',
-                        'type': 'desktop'
+                        'type': ' mobile'
                     }
                 },
                 'client': {
                     'version': '4.1',
-                    'id': 'browser'
+                    'id': 'browser',
+                    'supportsAdPods': True,
+                    'service': 'itv.x',
+                    'appversion': '2.43.28',
                 },
                 'variantAvailability': {
+                    'player': 'hls',
                     'featureset': {
                         'min': ['hls', 'aes', 'outband-webvtt'],
                         'max': ['hls', 'aes', 'outband-webvtt']
                     },
-                    'platformTag': 'dotcom'
+                    'platformTag': 'mobile'
                 }
             }).encode(), headers=headers)
         video_data = ios_playlist['Playlist']['Video']
-        ios_base_url = video_data.get('Base')
+        ios_base_url = traverse_obj(video_data, 'Base', expected_type=url_or_none)
+
+        media_url = (
+            (lambda u: url_or_none(urljoin(ios_base_url, u)))
+            if ios_base_url else url_or_none)
 
         formats = []
-        for media_file in (video_data.get('MediaFiles') or []):
+        for media_file in traverse_obj(video_data, 'MediaFiles', expected_type=list) or []:
-            href = media_file.get('Href')
+            href = traverse_obj(media_file, 'Href', expected_type=media_url)
             if not href:
                 continue
-            if ios_base_url:
-                href = ios_base_url + href
             ext = determine_ext(href)
             if ext == 'm3u8':
                 formats.extend(self._extract_m3u8_formats(
-                    href, video_id, 'mp4', entry_protocol='m3u8_native',
+                    href, video_id, 'mp4', entry_protocol='m3u8',
                     m3u8_id='hls', fatal=False))
+
             else:
                 formats.append({
                     'url': href,
                 })
         self._sort_formats(formats)
+        for f in formats:
+            f.setdefault('http_headers', {})
+            f['http_headers'].update(self._vanilla_ua_header())
 
         subtitles = {}
-        subs = video_data.get('Subtitles') or []
+        for sub in traverse_obj(video_data, 'Subtitles', expected_type=list) or []:
-        for sub in subs:
+            href = traverse_obj(sub, 'Href', expected_type=url_or_none)
-            if not isinstance(sub, dict):
-                continue
-            href = url_or_none(sub.get('Href'))
             if not href:
                 continue
             subtitles.setdefault('en', []).append({
@@ -127,59 +270,132 @@ class ITVIE(InfoExtractor):
                 'ext': determine_ext(href, 'vtt'),
             })
 
-        info = self._search_json_ld(webpage, video_id, default={})
+        next_data = self._search_nextjs_data(webpage, video_id, fatal=False, default='{}')
-        if not info:
+        video_data.update(traverse_obj(next_data, ('props', 'pageProps', ('title', 'episode')), expected_type=dict)[0] or {})
-            json_ld = self._parse_json(self._search_regex(
+        title = traverse_obj(video_data, 'headerTitle', 'episodeTitle')
-                JSON_LD_RE, webpage, 'JSON-LD', '{}',
+        info = self._og_extract(webpage, require_title=not title)
-                group='json_ld'), video_id, fatal=False)
+        tn = info.pop('thumbnail', None)
-            if json_ld and json_ld.get('@type') == 'BreadcrumbList':
+        if tn:
-                for ile in (json_ld.get('itemListElement:') or []):
+            info['thumbnails'] = [{'url': tn}]
-                    item = ile.get('item:') or {}
+
-                    if item.get('@type') == 'TVEpisode':
+        # num. episode title
-                        item['@context'] = 'http://schema.org'
+        num_ep_title = video_data.get('numberedEpisodeTitle')
-                        info = self._json_ld(item, video_id, fatal=False) or {}
+        if not num_ep_title:
-                        break
+            num_ep_title = clean_html(get_element_by_attribute('data-testid', 'episode-hero-description-strong', webpage))
+            num_ep_title = num_ep_title and num_ep_title.rstrip(' -')
+        ep_title = strip_or_none(
+            video_data.get('episodeTitle')
+            or (num_ep_title.split('.', 1)[-1] if num_ep_title else None))
+        title = title or re.sub(r'\s+-\s+ITVX$', '', info['title'])
+        if ep_title and ep_title != title:
+            title = title + ' - ' + ep_title
+
+        def get_thumbnails():
+            tns = []
+            for w, x in (traverse_obj(video_data, ('imagePresets'), expected_type=dict) or {}).items():
+                if isinstance(x, dict):
+                    for y, z in x.items():
+                        tns.append({'id': w + '_' + y, 'url': z})
+            return tns or None
+
+        video_str = lambda *x: traverse_obj(
+            video_data, *x, get_all=False, expected_type=strip_or_none)
 
         return merge_dicts({
             'id': video_id,
-            'title': self._html_search_meta(['og:title', 'twitter:title'], webpage),
+            'title': title,
             'formats': formats,
             'subtitles': subtitles,
-            'duration': parse_duration(video_data.get('Duration')),
+            # parsing hh:mm:ss:nnn not yet patched
-            'description': clean_html(get_element_by_class('episode-info__synopsis', webpage)),
+            'duration': parse_duration(re.sub(r'(\d{2})(:)(\d{3}$)', r'\1.\3', video_data.get('Duration') or '')),
+            'description': video_str('synopsis'),
+            'timestamp': traverse_obj(video_data, 'broadcastDateTime', 'dateTime', expected_type=parse_iso8601),
+            'thumbnails': get_thumbnails(),
+            'series': video_str('showTitle', 'programmeTitle'),
+            'series_number': int_or_none(video_data.get('seriesNumber')),
+            'episode': ep_title,
+            'episode_number': int_or_none((num_ep_title or '').split('.')[0]),
+            'channel': video_str('channel'),
+            'categories': traverse_obj(video_data, ('categories', 'formatted'), expected_type=list),
+            'age_limit': {False: 16, True: 0}.get(video_data.get('isChildrenCategory')),
         }, info)
 
 
-class ITVBTCCIE(InfoExtractor):
+class ITVBTCCIE(ITVBaseIE):
-    _VALID_URL = r'https?://(?:www\.)?itv\.com/btcc/(?:[^/]+/)*(?P<id>[^/?#&]+)'
+    _VALID_URL = r'https?://(?:www\.)?itv\.com/(?!(?:watch|hub)/)(?:[^/]+/)+(?P<id>[^/?#&]+)'
-    _TEST = {
+    _IE_DESC = 'ITV articles: News, British Touring Car Championship'
-        'url': 'http://www.itv.com/btcc/races/btcc-2018-all-the-action-from-brands-hatch',
+    _TESTS = [{
+        'note': 'British Touring Car Championship',
+        'url': 'https://www.itv.com/btcc/articles/btcc-2018-all-the-action-from-brands-hatch',
         'info_dict': {
             'id': 'btcc-2018-all-the-action-from-brands-hatch',
             'title': 'BTCC 2018: All the action from Brands Hatch',
         },
         'playlist_mincount': 9,
-    }
+    }, {
-    BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/1582188683001/HkiHLnNRx_default/index.html?videoId=%s'
+        'note': 'redirects to /btcc/articles/...',
+        'url': 'http://www.itv.com/btcc/races/btcc-2018-all-the-action-from-brands-hatch',
+        'only_matching': True,
+    }, {
+        'note': 'news article',
+        'url': 'https://www.itv.com/news/wales/2020-07-23/sean-fletcher-shows-off-wales-coastline-in-new-itv-series-as-british-tourists-opt-for-staycations',
+        'info_dict': {
+            'id': 'sean-fletcher-shows-off-wales-coastline-in-new-itv-series-as-british-tourists-opt-for-staycations',
+            'title': '''Sean Fletcher on why Wales' coastline should be your 'staycation' destination | ITV News''',
+        },
+        'playlist_mincount': 1,
+    }]
+
+    # should really be a class var of the BC IE
+    BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/%s/%s_default/index.html?videoId=%s'
+    BRIGHTCOVE_ACCOUNT = '1582188683001'
+    BRIGHTCOVE_PLAYER = 'HkiHLnNRx'
 
     def _real_extract(self, url):
         playlist_id = self._match_id(url)
 
-        webpage = self._download_webpage(url, playlist_id)
+        webpage, urlh = self._download_webpage_handle(url, playlist_id)
+        link = compat_urlparse.urlparse(urlh.geturl()).path.strip('/')
 
-        entries = [
+        next_data = self._search_nextjs_data(webpage, playlist_id, fatal=False, default='{}')
-            self.url_result(
+        path_prefix = compat_urlparse.urlparse(next_data.get('assetPrefix') or '').path.strip('/')
-                smuggle_url(self.BRIGHTCOVE_URL_TEMPLATE % video_id, {
+        link = remove_start(link, path_prefix).strip('/')
+
+        content = traverse_obj(
+            next_data, ('props', 'pageProps', Ellipsis),
+            expected_type=lambda x: x if x['link'] == link else None,
+            get_all=False, default={})
+        content = traverse_obj(
+            content, ('body', 'content', Ellipsis, 'data'),
+            expected_type=lambda x: x if x.get('name') == 'Brightcove' or x.get('type') == 'Brightcove' else None)
+
+        contraband = {
             # ITV does not like some GB IP ranges, so here are some
             # IP blocks it accepts
             'geo_ip_blocks': [
                 '193.113.0.0/16', '54.36.162.0/23', '159.65.16.0/21'
             ],
-                    'referrer': url,
+            'referrer': urlh.geturl(),
-                }),
+        }
+
+        def entries():
+
+            for data in content or []:
+                video_id = data.get('id')
+                if not video_id:
+                    continue
+                account = data.get('accountId') or self.BRIGHTCOVE_ACCOUNT
+                player = data.get('playerId') or self.BRIGHTCOVE_PLAYER
+                yield self.url_result(
+                    smuggle_url(self.BRIGHTCOVE_URL_TEMPLATE % (account, player, video_id), contraband),
+                    ie=BrightcoveNewIE.ie_key(), video_id=video_id)
+
+            # obsolete ?
+            for video_id in re.findall(r'''data-video-id=["'](\d+)''', webpage):
+                yield self.url_result(
+                    smuggle_url(self.BRIGHTCOVE_URL_TEMPLATE % (self.BRIGHTCOVE_ACCOUNT, self.BRIGHTCOVE_PLAYER, video_id), contraband),
                     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)
 
-        return self.playlist_result(entries, playlist_id, title)
+        return self.playlist_result(entries(), playlist_id, title)