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     ExtractorError,
  16     float_or_none,
  17     orderedSet,
  18     str_or_none,
  19     try_get,
  20 )
  21
  22
  23 class GloboIE(InfoExtractor):
  24     _VALID_URL = r'(?:globo:|https?://.+?\.globo\.com/(?:[^/]+/)*(?:v/(?:[^/]+/)?|videos/))(?P<id>\d{7,})'
  25     _NETRC_MACHINE = 'globo'
  26     _TESTS = [{
  27         'url': 'http://g1.globo.com/carros/autoesporte/videos/t/exclusivos-do-g1/v/mercedes-benz-gla-passa-por-teste-de-colisao-na-europa/3607726/',
  28         'info_dict': {
  29             'id': '3607726',
  30             'ext': 'mp4',
  31             'title': 'Mercedes-Benz GLA passa por teste de colisão na Europa',
  32             'duration': 103.204,
  33             'uploader': 'G1',
  34             'uploader_id': '2015',
  35         },
  36         'params': {
  37             'skip_download': True,
  38         },
  39     }, {
  40         'url': 'http://globoplay.globo.com/v/4581987/',
  41         'info_dict': {
  42             'id': '4581987',
  43             'ext': 'mp4',
  44             'title': 'Acidentes de trânsito estão entre as maiores causas de queda de energia em SP',
  45             'duration': 137.973,
  46             'uploader': 'Rede Globo',
  47             'uploader_id': '196',
  48         },
  49         'params': {
  50             'skip_download': True,
  51         },
  52     }, {
  53         'url': 'http://canalbrasil.globo.com/programas/sangue-latino/videos/3928201.html',
  54         'only_matching': True,
  55     }, {
  56         'url': 'http://globosatplay.globo.com/globonews/v/4472924/',
  57         'only_matching': True,
  58     }, {
  59         'url': 'http://globotv.globo.com/t/programa/v/clipe-sexo-e-as-negas-adeus/3836166/',
  60         'only_matching': True,
  61     }, {
  62         '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/',
  63         'only_matching': True,
  64     }, {
  65         'url': 'http://canaloff.globo.com/programas/desejar-profundo/videos/4518560.html',
  66         'only_matching': True,
  67     }, {
  68         'url': 'globo:3607726',
  69         'only_matching': True,
  70     }]
  71
  72     def _real_extract(self, url):
  73         video_id = self._match_id(url)
  74
  75         video = self._download_json(
  76             'http://api.globovideos.com/videos/%s/playlist' % video_id,
  77             video_id)['videos'][0]
  78         if not self.get_param('allow_unplayable_formats') and video.get('encrypted') is True:
  79             self.report_drm(video_id)
  80
  81         title = video['title']
  82
  83         formats = []
  84         security = self._download_json(
  85             'https://playback.video.globo.com/v1/video-session', video_id, 'Downloading security hash for %s' % video_id,
  86             headers={'content-type': 'application/json'}, data=json.dumps({
  87                 "player_type": "desktop",
  88                 "video_id": video_id,
  89                 "quality": "max",
  90                 "content_protection": "widevine",
  91                 "vsid": "581b986b-4c40-71f0-5a58-803e579d5fa2",
  92                 "tz": "-3.0:00"
  93             }).encode())
  94
  95         security_hash = security['source']['token']
  96         if not security_hash:
  97             message = security.get('message')
  98             if message:
  99                 raise ExtractorError(
 100                     '%s returned error: %s' % (self.IE_NAME, message), expected=True)
 101
 102         hash_code = security_hash[:2]
 103         padding = '%010d' % random.randint(1, 10000000000)
 104         if hash_code in ('04', '14'):
 105             received_time = security_hash[3:13]
 106             received_md5 = security_hash[24:]
 107             hash_prefix = security_hash[:23]
 108         elif hash_code in ('02', '12', '03', '13'):
 109             received_time = security_hash[2:12]
 110             received_md5 = security_hash[22:]
 111             padding += '1'
 112             hash_prefix = '05' + security_hash[:22]
 113
 114         padded_sign_time = compat_str(int(received_time) + 86400) + padding
 115         md5_data = (received_md5 + padded_sign_time + '0xAC10FD').encode()
 116         signed_md5 = base64.urlsafe_b64encode(hashlib.md5(md5_data).digest()).decode().strip('=')
 117         signed_hash = hash_prefix + padded_sign_time + signed_md5
 118         source = security['source']['url_parts']
 119         resource_url = source['scheme'] + '://' + source['domain'] + source['path']
 120         signed_url = '%s?h=%s&k=html5&a=%s' % (resource_url, signed_hash, 'F' if video.get('subscriber_only') else 'A')
 121
 122         formats.extend(self._extract_m3u8_formats(
 123             signed_url, video_id, 'mp4', entry_protocol='m3u8_native', m3u8_id='hls', fatal=False))
 124         self._sort_formats(formats)
 125
 126         subtitles = {}
 127         for resource in video['resources']:
 128             if resource.get('type') == 'subtitle':
 129                 subtitles.setdefault(resource.get('language') or 'por', []).append({
 130                     'url': resource.get('url'),
 131                 })
 132         subs = try_get(security, lambda x: x['source']['subtitles'], expected_type=dict) or {}
 133         for sub_lang, sub_url in subs.items():
 134             if sub_url:
 135                 subtitles.setdefault(sub_lang or 'por', []).append({
 136                     'url': sub_url,
 137                 })
 138         subs = try_get(security, lambda x: x['source']['subtitles_webvtt'], expected_type=dict) or {}
 139         for sub_lang, sub_url in subs.items():
 140             if sub_url:
 141                 subtitles.setdefault(sub_lang or 'por', []).append({
 142                     'url': sub_url,
 143                 })
 144
 145         duration = float_or_none(video.get('duration'), 1000)
 146         uploader = video.get('channel')
 147         uploader_id = str_or_none(video.get('channel_id'))
 148
 149         return {
 150             'id': video_id,
 151             'title': title,
 152             'duration': duration,
 153             'uploader': uploader,
 154             'uploader_id': uploader_id,
 155             'formats': formats,
 156             'subtitles': subtitles,
 157         }
 158
 159
 160 class GloboArticleIE(InfoExtractor):
 161     _VALID_URL = r'https?://.+?\.globo\.com/(?:[^/]+/)*(?P<id>[^/.]+)(?:\.html)?'
 162
 163     _VIDEOID_REGEXES = [
 164         r'\bdata-video-id=["\'](\d{7,})',
 165         r'\bdata-player-videosids=["\'](\d{7,})',
 166         r'\bvideosIDs\s*:\s*["\']?(\d{7,})',
 167         r'\bdata-id=["\'](\d{7,})',
 168         r'<div[^>]+\bid=["\'](\d{7,})',
 169     ]
 170
 171     _TESTS = [{
 172         'url': 'http://g1.globo.com/jornal-nacional/noticia/2014/09/novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes.html',
 173         'info_dict': {
 174             'id': 'novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes',
 175             'title': 'Novidade na fiscalização de bagagem pela Receita provoca discussões',
 176             'description': 'md5:c3c4b4d4c30c32fce460040b1ac46b12',
 177         },
 178         'playlist_count': 1,
 179     }, {
 180         'url': 'http://g1.globo.com/pr/parana/noticia/2016/09/mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato.html',
 181         'info_dict': {
 182             'id': 'mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato',
 183             'title': "Lula era o 'comandante máximo' do esquema da Lava Jato, diz MPF",
 184             'description': 'md5:8aa7cc8beda4dc71cc8553e00b77c54c',
 185         },
 186         'playlist_count': 6,
 187     }, {
 188         'url': 'http://gq.globo.com/Prazeres/Poder/noticia/2015/10/all-o-desafio-assista-ao-segundo-capitulo-da-serie.html',
 189         'only_matching': True,
 190     }, {
 191         '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',
 192         'only_matching': True,
 193     }, {
 194         'url': 'http://oglobo.globo.com/rio/a-amizade-entre-um-entregador-de-farmacia-um-piano-19946271',
 195         'only_matching': True,
 196     }]
 197
 198     @classmethod
 199     def suitable(cls, url):
 200         return False if GloboIE.suitable(url) else super(GloboArticleIE, cls).suitable(url)
 201
 202     def _real_extract(self, url):
 203         display_id = self._match_id(url)
 204         webpage = self._download_webpage(url, display_id)
 205         video_ids = []
 206         for video_regex in self._VIDEOID_REGEXES:
 207             video_ids.extend(re.findall(video_regex, webpage))
 208         entries = [
 209             self.url_result('globo:%s' % video_id, GloboIE.ie_key())
 210             for video_id in orderedSet(video_ids)]
 211         title = self._og_search_title(webpage, fatal=False)
 212         description = self._html_search_meta('description', webpage)
 213         return self.playlist_result(entries, display_id, title, description)