yt_dlp/extractor/globo.py

   1 import base64
   2 import hashlib
   3 import json
   4 import random
   5 import re
   6
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_str,
  10 )
  11 from ..utils import (
  12     HEADRequest,
  13     ExtractorError,
  14     float_or_none,
  15     orderedSet,
  16     str_or_none,
  17     try_get,
  18 )
  19
  20
  21 class GloboIE(InfoExtractor):
  22     _VALID_URL = r'(?:globo:|https?://.+?\.globo\.com/(?:[^/]+/)*(?:v/(?:[^/]+/)?|videos/))(?P<id>\d{7,})'
  23     _NETRC_MACHINE = 'globo'
  24     _TESTS = [{
  25         'url': 'http://g1.globo.com/carros/autoesporte/videos/t/exclusivos-do-g1/v/mercedes-benz-gla-passa-por-teste-de-colisao-na-europa/3607726/',
  26         'info_dict': {
  27             'id': '3607726',
  28             'ext': 'mp4',
  29             'title': 'Mercedes-Benz GLA passa por teste de colisão na Europa',
  30             'duration': 103.204,
  31             'uploader': 'G1',
  32             'uploader_id': '2015',
  33         },
  34         'params': {
  35             'skip_download': True,
  36         },
  37     }, {
  38         'url': 'http://globoplay.globo.com/v/4581987/',
  39         'info_dict': {
  40             'id': '4581987',
  41             'ext': 'mp4',
  42             'title': 'Acidentes de trânsito estão entre as maiores causas de queda de energia em SP',
  43             'duration': 137.973,
  44             'uploader': 'Rede Globo',
  45             'uploader_id': '196',
  46         },
  47         'params': {
  48             'skip_download': True,
  49         },
  50     }, {
  51         'url': 'http://canalbrasil.globo.com/programas/sangue-latino/videos/3928201.html',
  52         'only_matching': True,
  53     }, {
  54         'url': 'http://globosatplay.globo.com/globonews/v/4472924/',
  55         'only_matching': True,
  56     }, {
  57         'url': 'http://globotv.globo.com/t/programa/v/clipe-sexo-e-as-negas-adeus/3836166/',
  58         'only_matching': True,
  59     }, {
  60         'url': 'http://globotv.globo.com/canal-brasil/sangue-latino/t/todos-os-videos/v/ator-e-diretor-argentino-ricado-darin-fala-sobre-utopias-e-suas-perdas/3928201/',
  61         'only_matching': True,
  62     }, {
  63         'url': 'http://canaloff.globo.com/programas/desejar-profundo/videos/4518560.html',
  64         'only_matching': True,
  65     }, {
  66         'url': 'globo:3607726',
  67         'only_matching': True,
  68     }, {
  69         'url': 'https://globoplay.globo.com/v/10248083/',
  70         'info_dict': {
  71             'id': '10248083',
  72             'ext': 'mp4',
  73             'title': 'Melhores momentos: Equador 1 x 1 Brasil pelas Eliminatórias da Copa do Mundo 2022',
  74             'duration': 530.964,
  75             'uploader': 'SporTV',
  76             'uploader_id': '698',
  77         },
  78         'params': {
  79             'skip_download': True,
  80         },
  81     }]
  82
  83     def _real_extract(self, url):
  84         video_id = self._match_id(url)
  85
  86         self._request_webpage(
  87             HEADRequest('https://globo-ab.globo.com/v2/selected-alternatives?experiments=player-isolated-experiment-02&skipImpressions=true'),
  88             video_id, 'Getting cookies')
  89
  90         video = self._download_json(
  91             'http://api.globovideos.com/videos/%s/playlist' % video_id,
  92             video_id)['videos'][0]
  93         if not self.get_param('allow_unplayable_formats') and video.get('encrypted') is True:
  94             self.report_drm(video_id)
  95
  96         title = video['title']
  97
  98         formats = []
  99         security = self._download_json(
 100             'https://playback.video.globo.com/v2/video-session', video_id, 'Downloading security hash for %s' % video_id,
 101             headers={'content-type': 'application/json'}, data=json.dumps({
 102                 "player_type": "desktop",
 103                 "video_id": video_id,
 104                 "quality": "max",
 105                 "content_protection": "widevine",
 106                 "vsid": "581b986b-4c40-71f0-5a58-803e579d5fa2",
 107                 "tz": "-3.0:00"
 108             }).encode())
 109
 110         self._request_webpage(HEADRequest(security['sources'][0]['url_template']), video_id, 'Getting locksession cookie')
 111
 112         security_hash = security['sources'][0]['token']
 113         if not security_hash:
 114             message = security.get('message')
 115             if message:
 116                 raise ExtractorError(
 117                     '%s returned error: %s' % (self.IE_NAME, message), expected=True)
 118
 119         hash_code = security_hash[:2]
 120         padding = '%010d' % random.randint(1, 10000000000)
 121         if hash_code in ('04', '14'):
 122             received_time = security_hash[3:13]
 123             received_md5 = security_hash[24:]
 124             hash_prefix = security_hash[:23]
 125         elif hash_code in ('02', '12', '03', '13'):
 126             received_time = security_hash[2:12]
 127             received_md5 = security_hash[22:]
 128             padding += '1'
 129             hash_prefix = '05' + security_hash[:22]
 130
 131         padded_sign_time = compat_str(int(received_time) + 86400) + padding
 132         md5_data = (received_md5 + padded_sign_time + '0xAC10FD').encode()
 133         signed_md5 = base64.urlsafe_b64encode(hashlib.md5(md5_data).digest()).decode().strip('=')
 134         signed_hash = hash_prefix + padded_sign_time + signed_md5
 135         source = security['sources'][0]['url_parts']
 136         resource_url = source['scheme'] + '://' + source['domain'] + source['path']
 137         signed_url = '%s?h=%s&k=html5&a=%s' % (resource_url, signed_hash, 'F' if video.get('subscriber_only') else 'A')
 138
 139         fmts, subtitles = self._extract_m3u8_formats_and_subtitles(
 140             signed_url, video_id, 'mp4', entry_protocol='m3u8_native', m3u8_id='hls', fatal=False)
 141         formats.extend(fmts)
 142         self._sort_formats(formats)
 143
 144         for resource in video['resources']:
 145             if resource.get('type') == 'subtitle':
 146                 subtitles.setdefault(resource.get('language') or 'por', []).append({
 147                     'url': resource.get('url'),
 148                 })
 149         subs = try_get(security, lambda x: x['source']['subtitles'], expected_type=dict) or {}
 150         for sub_lang, sub_url in subs.items():
 151             if sub_url:
 152                 subtitles.setdefault(sub_lang or 'por', []).append({
 153                     'url': sub_url,
 154                 })
 155         subs = try_get(security, lambda x: x['source']['subtitles_webvtt'], expected_type=dict) or {}
 156         for sub_lang, sub_url in subs.items():
 157             if sub_url:
 158                 subtitles.setdefault(sub_lang or 'por', []).append({
 159                     'url': sub_url,
 160                 })
 161
 162         duration = float_or_none(video.get('duration'), 1000)
 163         uploader = video.get('channel')
 164         uploader_id = str_or_none(video.get('channel_id'))
 165
 166         return {
 167             'id': video_id,
 168             'title': title,
 169             'duration': duration,
 170             'uploader': uploader,
 171             'uploader_id': uploader_id,
 172             'formats': formats,
 173             'subtitles': subtitles,
 174         }
 175
 176
 177 class GloboArticleIE(InfoExtractor):
 178     _VALID_URL = r'https?://.+?\.globo\.com/(?:[^/]+/)*(?P<id>[^/.]+)(?:\.html)?'
 179
 180     _VIDEOID_REGEXES = [
 181         r'\bdata-video-id=["\'](\d{7,})["\']',
 182         r'\bdata-player-videosids=["\'](\d{7,})["\']',
 183         r'\bvideosIDs\s*:\s*["\']?(\d{7,})',
 184         r'\bdata-id=["\'](\d{7,})["\']',
 185         r'<div[^>]+\bid=["\'](\d{7,})["\']',
 186         r'<bs-player[^>]+\bvideoid=["\'](\d{8,})["\']',
 187     ]
 188
 189     _TESTS = [{
 190         'url': 'http://g1.globo.com/jornal-nacional/noticia/2014/09/novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes.html',
 191         'info_dict': {
 192             'id': 'novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes',
 193             'title': 'Novidade na fiscalização de bagagem pela Receita provoca discussões',
 194             'description': 'md5:c3c4b4d4c30c32fce460040b1ac46b12',
 195         },
 196         'playlist_count': 1,
 197     }, {
 198         'url': 'http://g1.globo.com/pr/parana/noticia/2016/09/mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato.html',
 199         'info_dict': {
 200             'id': 'mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato',
 201             'title': "Lula era o 'comandante máximo' do esquema da Lava Jato, diz MPF",
 202             'description': 'md5:8aa7cc8beda4dc71cc8553e00b77c54c',
 203         },
 204         'playlist_count': 6,
 205     }, {
 206         'url': 'http://gq.globo.com/Prazeres/Poder/noticia/2015/10/all-o-desafio-assista-ao-segundo-capitulo-da-serie.html',
 207         'only_matching': True,
 208     }, {
 209         'url': 'http://gshow.globo.com/programas/tv-xuxa/O-Programa/noticia/2014/01/xuxa-e-junno-namoram-muuuito-em-luau-de-zeze-di-camargo-e-luciano.html',
 210         'only_matching': True,
 211     }, {
 212         'url': 'http://oglobo.globo.com/rio/a-amizade-entre-um-entregador-de-farmacia-um-piano-19946271',
 213         'only_matching': True,
 214     }, {
 215         'url': 'https://ge.globo.com/video/ta-na-area-como-foi-assistir-ao-jogo-do-palmeiras-que-a-globo-nao-passou-10287094.ghtml',
 216         'info_dict': {
 217             'id': 'ta-na-area-como-foi-assistir-ao-jogo-do-palmeiras-que-a-globo-nao-passou-10287094',
 218             'title': 'Tá na Área: como foi assistir ao jogo do Palmeiras que a Globo não passou',
 219             'description': 'md5:2d089d036c4c9675117d3a56f8c61739',
 220         },
 221         'playlist_count': 1,
 222     }, {
 223         'url': 'https://redeglobo.globo.com/rpc/meuparana/noticia/a-producao-de-chocolates-no-parana.ghtml',
 224         'info_dict': {
 225             'id': 'a-producao-de-chocolates-no-parana',
 226             'title': 'A produção de chocolates no Paraná',
 227             'description': 'md5:f2e3daf00ffd1dc0e9a8a6c7cfb0a89e',
 228         },
 229         'playlist_count': 2,
 230     }]
 231
 232     @classmethod
 233     def suitable(cls, url):
 234         return False if GloboIE.suitable(url) else super(GloboArticleIE, cls).suitable(url)
 235
 236     def _real_extract(self, url):
 237         display_id = self._match_id(url)
 238         webpage = self._download_webpage(url, display_id)
 239         video_ids = []
 240         for video_regex in self._VIDEOID_REGEXES:
 241             video_ids.extend(re.findall(video_regex, webpage))
 242         entries = [
 243             self.url_result('globo:%s' % video_id, GloboIE.ie_key())
 244             for video_id in orderedSet(video_ids)]
 245         title = self._og_search_title(webpage).strip()
 246         description = self._html_search_meta('description', webpage)
 247         return self.playlist_result(entries, display_id, title, description)