yt_dlp/extractor/arcpublishing.py

   1 import re
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     extract_attributes,
   6     int_or_none,
   7     join_nonempty,
   8     parse_iso8601,
   9     try_get,
  10 )
  11
  12
  13 class ArcPublishingIE(InfoExtractor):
  14     _UUID_REGEX = r'[\da-f]{8}-(?:[\da-f]{4}-){3}[\da-f]{12}'
  15     _VALID_URL = rf'arcpublishing:(?P<org>[a-z]+):(?P<id>{_UUID_REGEX})'
  16     _TESTS = [{
  17         # https://www.adn.com/politics/2020/11/02/video-senate-candidates-campaign-in-anchorage-on-eve-of-election-day/
  18         'url': 'arcpublishing:adn:8c99cb6e-b29c-4bc9-9173-7bf9979225ab',
  19         'only_matching': True,
  20     }, {
  21         # https://www.bostonglobe.com/video/2020/12/30/metro/footage-released-showing-officer-talking-about-striking-protesters-with-car/
  22         'url': 'arcpublishing:bostonglobe:232b7ae6-7d73-432d-bc0a-85dbf0119ab1',
  23         'only_matching': True,
  24     }, {
  25         # https://www.actionnewsjax.com/video/live-stream/
  26         'url': 'arcpublishing:cmg:cfb1cf1b-3ab5-4d1b-86c5-a5515d311f2a',
  27         'only_matching': True,
  28     }, {
  29         # https://elcomercio.pe/videos/deportes/deporte-total-futbol-peruano-seleccion-peruana-la-valorizacion-de-los-peruanos-en-el-exterior-tras-un-2020-atipico-nnav-vr-video-noticia/
  30         'url': 'arcpublishing:elcomercio:27a7e1f8-2ec7-4177-874f-a4feed2885b3',
  31         'only_matching': True,
  32     }, {
  33         # https://www.clickondetroit.com/video/community/2020/05/15/events-surrounding-woodward-dream-cruise-being-canceled/
  34         'url': 'arcpublishing:gmg:c8793fb2-8d44-4242-881e-2db31da2d9fe',
  35         'only_matching': True,
  36     }, {
  37         # https://www.wabi.tv/video/2020/12/30/trenton-company-making-equipment-pfizer-covid-vaccine/
  38         'url': 'arcpublishing:gray:0b0ba30e-032a-4598-8810-901d70e6033e',
  39         'only_matching': True,
  40     }, {
  41         # https://www.lateja.cr/el-mundo/video-china-aprueba-con-condiciones-su-primera/dfcbfa57-527f-45ff-a69b-35fe71054143/video/
  42         'url': 'arcpublishing:gruponacion:dfcbfa57-527f-45ff-a69b-35fe71054143',
  43         'only_matching': True,
  44     }, {
  45         # https://www.fifthdomain.com/video/2018/03/09/is-america-vulnerable-to-a-cyber-attack/
  46         'url': 'arcpublishing:mco:aa0ca6fe-1127-46d4-b32c-be0d6fdb8055',
  47         'only_matching': True,
  48     }, {
  49         # https://www.vl.no/kultur/2020/12/09/en-melding-fra-en-lytter-endret-julelista-til-lewi-bergrud/
  50         'url': 'arcpublishing:mentormedier:47a12084-650b-4011-bfd0-3699b6947b2d',
  51         'only_matching': True,
  52     }, {
  53         # https://www.14news.com/2020/12/30/whiskey-theft-caught-camera-henderson-liquor-store/
  54         'url': 'arcpublishing:raycom:b89f61f8-79fa-4c09-8255-e64237119bf7',
  55         'only_matching': True,
  56     }, {
  57         # https://www.theglobeandmail.com/world/video-ethiopian-woman-who-became-symbol-of-integration-in-italy-killed-on/
  58         'url': 'arcpublishing:tgam:411b34c1-8701-4036-9831-26964711664b',
  59         'only_matching': True,
  60     }, {
  61         # https://www.pilotonline.com/460f2931-8130-4719-8ea1-ffcb2d7cb685-132.html
  62         'url': 'arcpublishing:tronc:460f2931-8130-4719-8ea1-ffcb2d7cb685',
  63         'only_matching': True,
  64     }]
  65     _POWA_DEFAULTS = [
  66         (['cmg', 'prisa'], '%s-config-prod.api.cdn.arcpublishing.com/video'),
  67         ([
  68             'adn', 'advancelocal', 'answers', 'bonnier', 'bostonglobe', 'demo',
  69             'gmg', 'gruponacion', 'infobae', 'mco', 'nzme', 'pmn', 'raycom',
  70             'spectator', 'tbt', 'tgam', 'tronc', 'wapo', 'wweek',
  71         ], 'video-api-cdn.%s.arcpublishing.com/api'),
  72     ]
  73
  74     @classmethod
  75     def _extract_embed_urls(cls, url, webpage):
  76         entries = []
  77         # https://arcpublishing.atlassian.net/wiki/spaces/POWA/overview
  78         for powa_el in re.findall(rf'(<div[^>]+class="[^"]*\bpowa\b[^"]*"[^>]+data-uuid="{ArcPublishingIE._UUID_REGEX}"[^>]*>)', webpage):
  79             powa = extract_attributes(powa_el) or {}
  80             org = powa.get('data-org')
  81             uuid = powa.get('data-uuid')
  82             if org and uuid:
  83                 entries.append(f'arcpublishing:{org}:{uuid}')
  84         return entries
  85
  86     def _real_extract(self, url):
  87         org, uuid = self._match_valid_url(url).groups()
  88         for orgs, tmpl in self._POWA_DEFAULTS:
  89             if org in orgs:
  90                 base_api_tmpl = tmpl
  91                 break
  92         else:
  93             base_api_tmpl = '%s-prod-cdn.video-api.arcpublishing.com/api'
  94         if org == 'wapo':
  95             org = 'washpost'
  96         video = self._download_json(
  97             'https://%s/v1/ansvideos/findByUuid' % (base_api_tmpl % org),
  98             uuid, query={'uuid': uuid})[0]
  99         title = video['headlines']['basic']
 100         is_live = video.get('status') == 'live'
 101
 102         urls = []
 103         formats = []
 104         for s in video.get('streams', []):
 105             s_url = s.get('url')
 106             if not s_url or s_url in urls:
 107                 continue
 108             urls.append(s_url)
 109             stream_type = s.get('stream_type')
 110             if stream_type == 'smil':
 111                 smil_formats = self._extract_smil_formats(
 112                     s_url, uuid, fatal=False)
 113                 for f in smil_formats:
 114                     if f['url'].endswith('/cfx/st'):
 115                         f['app'] = 'cfx/st'
 116                         if not f['play_path'].startswith('mp4:'):
 117                             f['play_path'] = 'mp4:' + f['play_path']
 118                         if isinstance(f['tbr'], float):
 119                             f['vbr'] = f['tbr'] * 1000
 120                             del f['tbr']
 121                             f['format_id'] = 'rtmp-%d' % f['vbr']
 122                 formats.extend(smil_formats)
 123             elif stream_type in ('ts', 'hls'):
 124                 m3u8_formats = self._extract_m3u8_formats(
 125                     s_url, uuid, 'mp4', live=is_live, m3u8_id='hls', fatal=False)
 126                 if all(f.get('acodec') == 'none' for f in m3u8_formats):
 127                     continue
 128                 for f in m3u8_formats:
 129                     height = f.get('height')
 130                     if not height:
 131                         continue
 132                     vbr = self._search_regex(
 133                         r'[_x]%d[_-](\d+)' % height, f['url'], 'vbr', default=None)
 134                     if vbr:
 135                         f['vbr'] = int(vbr)
 136                 formats.extend(m3u8_formats)
 137             else:
 138                 vbr = int_or_none(s.get('bitrate'))
 139                 formats.append({
 140                     'format_id': join_nonempty(stream_type, vbr),
 141                     'vbr': vbr,
 142                     'width': int_or_none(s.get('width')),
 143                     'height': int_or_none(s.get('height')),
 144                     'filesize': int_or_none(s.get('filesize')),
 145                     'url': s_url,
 146                     'quality': -10,
 147                 })
 148
 149         subtitles = {}
 150         for subtitle in (try_get(video, lambda x: x['subtitles']['urls'], list) or []):
 151             subtitle_url = subtitle.get('url')
 152             if subtitle_url:
 153                 subtitles.setdefault('en', []).append({'url': subtitle_url})
 154
 155         return {
 156             'id': uuid,
 157             'title': title,
 158             'thumbnail': try_get(video, lambda x: x['promo_image']['url']),
 159             'description': try_get(video, lambda x: x['subheadlines']['basic']),
 160             'formats': formats,
 161             'duration': int_or_none(video.get('duration'), 100),
 162             'timestamp': parse_iso8601(video.get('created_date')),
 163             'subtitles': subtitles,
 164             'is_live': is_live,
 165         }