yt_dlp/extractor/globo.py

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