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