_ Git - youtube-dl/blob - youtube_dl/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 compat_str
  12 from ..utils import (
  13     ExtractorError,
  14     float_or_none,
  15     int_or_none,
  16     orderedSet,
  17     str_or_none,
  18 )
  19
  20
  21 class GloboIE(InfoExtractor):
  22     _VALID_URL = r'(?:globo:|https?://.+?\.globo\.com/(?:[^/]+/)*(?:v/(?:[^/]+/)?|videos/))(?P<id>\d{7,})'
  23     _LOGGED_IN = False
  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         'md5': 'b3ccc801f75cd04a914d51dadb83a78d',
  27         'info_dict': {
  28             'id': '3607726',
  29             'ext': 'mp4',
  30             'title': 'Mercedes-Benz GLA passa por teste de colisão na Europa',
  31             'duration': 103.204,
  32             'uploader': 'Globo.com',
  33             'uploader_id': '265',
  34         },
  35     }, {
  36         'url': 'http://globoplay.globo.com/v/4581987/',
  37         'md5': 'f36a1ecd6a50da1577eee6dd17f67eff',
  38         'info_dict': {
  39             'id': '4581987',
  40             'ext': 'mp4',
  41             'title': 'Acidentes de trânsito estão entre as maiores causas de queda de energia em SP',
  42             'duration': 137.973,
  43             'uploader': 'Rede Globo',
  44             'uploader_id': '196',
  45         },
  46     }, {
  47         'url': 'http://canalbrasil.globo.com/programas/sangue-latino/videos/3928201.html',
  48         'only_matching': True,
  49     }, {
  50         'url': 'http://globosatplay.globo.com/globonews/v/4472924/',
  51         'only_matching': True,
  52     }, {
  53         'url': 'http://globotv.globo.com/t/programa/v/clipe-sexo-e-as-negas-adeus/3836166/',
  54         'only_matching': True,
  55     }, {
  56         '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/',
  57         'only_matching': True,
  58     }, {
  59         'url': 'http://canaloff.globo.com/programas/desejar-profundo/videos/4518560.html',
  60         'only_matching': True,
  61     }, {
  62         'url': 'globo:3607726',
  63         'only_matching': True,
  64     }]
  65
  66     def _real_initialize(self):
  67         if self._LOGGED_IN:
  68             return
  69
  70         email, password = self._get_login_info()
  71         if email is None:
  72             return
  73
  74         self._download_json(
  75             'https://login.globo.com/api/authentication', None, data=json.dumps({
  76                 'payload': {
  77                     'email': email,
  78                     'password': password,
  79                     'serviceId': 4654,
  80                 },
  81             }).encode(), headers={
  82                 'Content-Type': 'application/json; charset=utf-8',
  83             })
  84         self._LOGGED_IN = True
  85
  86     def _real_extract(self, url):
  87         video_id = self._match_id(url)
  88
  89         video = self._download_json(
  90             'http://api.globovideos.com/videos/%s/playlist' % video_id,
  91             video_id)['videos'][0]
  92
  93         title = video['title']
  94
  95         formats = []
  96         for resource in video['resources']:
  97             resource_id = resource.get('_id')
  98             resource_url = resource.get('url')
  99             if not resource_id or not resource_url:
 100                 continue
 101
 102             security = self._download_json(
 103                 'http://security.video.globo.com/videos/%s/hash' % video_id,
 104                 video_id, 'Downloading security hash for %s' % resource_id, query={
 105                     'player': 'flash',
 106                     'version': '17.0.0.132',
 107                     'resource_id': resource_id,
 108                 })
 109
 110             security_hash = security.get('hash')
 111             if not security_hash:
 112                 message = security.get('message')
 113                 if message:
 114                     raise ExtractorError(
 115                         '%s returned error: %s' % (self.IE_NAME, message), expected=True)
 116                 continue
 117
 118             hash_code = security_hash[:2]
 119             received_time = int(security_hash[2:12])
 120             received_random = security_hash[12:22]
 121             received_md5 = security_hash[22:]
 122
 123             sign_time = received_time + 86400
 124             padding = '%010d' % random.randint(1, 10000000000)
 125
 126             md5_data = (received_md5 + str(sign_time) + padding + '0xFF01DD').encode()
 127             signed_md5 = base64.urlsafe_b64encode(hashlib.md5(md5_data).digest()).decode().strip('=')
 128             signed_hash = hash_code + compat_str(received_time) + received_random + compat_str(sign_time) + padding + signed_md5
 129
 130             signed_url = '%s?h=%s&k=%s' % (resource_url, signed_hash, 'flash')
 131             if resource_id.endswith('m3u8') or resource_url.endswith('.m3u8'):
 132                 formats.extend(self._extract_m3u8_formats(
 133                     signed_url, resource_id, 'mp4', entry_protocol='m3u8_native',
 134                     m3u8_id='hls', fatal=False))
 135             elif resource_id.endswith('mpd') or resource_url.endswith('.mpd'):
 136                 formats.extend(self._extract_mpd_formats(
 137                     signed_url, resource_id, mpd_id='dash', fatal=False))
 138             elif resource_id.endswith('manifest') or resource_url.endswith('/manifest'):
 139                 formats.extend(self._extract_ism_formats(
 140                     signed_url, resource_id, ism_id='mss', fatal=False))
 141             else:
 142                 formats.append({
 143                     'url': signed_url,
 144                     'format_id': 'http-%s' % resource_id,
 145                     'height': int_or_none(resource.get('height')),
 146                 })
 147
 148         self._sort_formats(formats)
 149
 150         duration = float_or_none(video.get('duration'), 1000)
 151         uploader = video.get('channel')
 152         uploader_id = str_or_none(video.get('channel_id'))
 153
 154         return {
 155             'id': video_id,
 156             'title': title,
 157             'duration': duration,
 158             'uploader': uploader,
 159             'uploader_id': uploader_id,
 160             'formats': formats
 161         }
 162
 163
 164 class GloboArticleIE(InfoExtractor):
 165     _VALID_URL = r'https?://.+?\.globo\.com/(?:[^/]+/)*(?P<id>[^/.]+)(?:\.html)?'
 166
 167     _VIDEOID_REGEXES = [
 168         r'\bdata-video-id=["\'](\d{7,})',
 169         r'\bdata-player-videosids=["\'](\d{7,})',
 170         r'\bvideosIDs\s*:\s*["\']?(\d{7,})',
 171         r'\bdata-id=["\'](\d{7,})',
 172         r'<div[^>]+\bid=["\'](\d{7,})',
 173     ]
 174
 175     _TESTS = [{
 176         'url': 'http://g1.globo.com/jornal-nacional/noticia/2014/09/novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes.html',
 177         'info_dict': {
 178             'id': 'novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes',
 179             'title': 'Novidade na fiscalização de bagagem pela Receita provoca discussões',
 180             'description': 'md5:c3c4b4d4c30c32fce460040b1ac46b12',
 181         },
 182         'playlist_count': 1,
 183     }, {
 184         'url': 'http://g1.globo.com/pr/parana/noticia/2016/09/mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato.html',
 185         'info_dict': {
 186             'id': 'mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato',
 187             'title': "Lula era o 'comandante máximo' do esquema da Lava Jato, diz MPF",
 188             'description': 'md5:8aa7cc8beda4dc71cc8553e00b77c54c',
 189         },
 190         'playlist_count': 6,
 191     }, {
 192         'url': 'http://gq.globo.com/Prazeres/Poder/noticia/2015/10/all-o-desafio-assista-ao-segundo-capitulo-da-serie.html',
 193         'only_matching': True,
 194     }, {
 195         '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',
 196         'only_matching': True,
 197     }, {
 198         'url': 'http://oglobo.globo.com/rio/a-amizade-entre-um-entregador-de-farmacia-um-piano-19946271',
 199         'only_matching': True,
 200     }]
 201
 202     @classmethod
 203     def suitable(cls, url):
 204         return False if GloboIE.suitable(url) else super(GloboArticleIE, cls).suitable(url)
 205
 206     def _real_extract(self, url):
 207         display_id = self._match_id(url)
 208         webpage = self._download_webpage(url, display_id)
 209         video_ids = []
 210         for video_regex in self._VIDEOID_REGEXES:
 211             video_ids.extend(re.findall(video_regex, webpage))
 212         entries = [
 213             self.url_result('globo:%s' % video_id, GloboIE.ie_key())
 214             for video_id in orderedSet(video_ids)]
 215         title = self._og_search_title(webpage, fatal=False)
 216         description = self._html_search_meta('description', webpage)
 217         return self.playlist_result(entries, display_id, title, description)