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 (
  12     compat_HTTPError,
  13     compat_str,
  14 )
  15 from ..utils import (
  16     ExtractorError,
  17     float_or_none,
  18     int_or_none,
  19     orderedSet,
  20     str_or_none,
  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         'md5': 'b3ccc801f75cd04a914d51dadb83a78d',
  30         'info_dict': {
  31             'id': '3607726',
  32             'ext': 'mp4',
  33             'title': 'Mercedes-Benz GLA passa por teste de colisão na Europa',
  34             'duration': 103.204,
  35             'uploader': 'Globo.com',
  36             'uploader_id': '265',
  37         },
  38     }, {
  39         'url': 'http://globoplay.globo.com/v/4581987/',
  40         'md5': 'f36a1ecd6a50da1577eee6dd17f67eff',
  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     }, {
  50         'url': 'http://canalbrasil.globo.com/programas/sangue-latino/videos/3928201.html',
  51         'only_matching': True,
  52     }, {
  53         'url': 'http://globosatplay.globo.com/globonews/v/4472924/',
  54         'only_matching': True,
  55     }, {
  56         'url': 'http://globotv.globo.com/t/programa/v/clipe-sexo-e-as-negas-adeus/3836166/',
  57         'only_matching': True,
  58     }, {
  59         '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/',
  60         'only_matching': True,
  61     }, {
  62         'url': 'http://canaloff.globo.com/programas/desejar-profundo/videos/4518560.html',
  63         'only_matching': True,
  64     }, {
  65         'url': 'globo:3607726',
  66         'only_matching': True,
  67     }]
  68
  69     def _real_initialize(self):
  70         email, password = self._get_login_info()
  71         if email is None:
  72             return
  73
  74         try:
  75             glb_id = (self._download_json(
  76                 'https://login.globo.com/api/authentication', None, data=json.dumps({
  77                     'payload': {
  78                         'email': email,
  79                         'password': password,
  80                         'serviceId': 4654,
  81                     },
  82                 }).encode(), headers={
  83                     'Content-Type': 'application/json; charset=utf-8',
  84                 }) or {}).get('glbId')
  85             if glb_id:
  86                 self._set_cookie('.globo.com', 'GLBID', glb_id)
  87         except ExtractorError as e:
  88             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
  89                 resp = self._parse_json(e.cause.read(), None)
  90                 raise ExtractorError(resp.get('userMessage') or resp['id'], expected=True)
  91             raise
  92
  93     def _real_extract(self, url):
  94         video_id = self._match_id(url)
  95
  96         video = self._download_json(
  97             'http://api.globovideos.com/videos/%s/playlist' % video_id,
  98             video_id)['videos'][0]
  99         if video.get('encrypted') is True:
 100             raise ExtractorError('This video is DRM protected.', expected=True)
 101
 102         title = video['title']
 103
 104         formats = []
 105         subtitles = {}
 106         for resource in video['resources']:
 107             resource_id = resource.get('_id')
 108             resource_url = resource.get('url')
 109             resource_type = resource.get('type')
 110             if not resource_url or (resource_type == 'media' and not resource_id) or resource_type not in ('subtitle', 'media'):
 111                 continue
 112
 113             if resource_type == 'subtitle':
 114                 subtitles.setdefault(resource.get('language') or 'por', []).append({
 115                     'url': resource_url,
 116                 })
 117                 continue
 118
 119             security = self._download_json(
 120                 'http://security.video.globo.com/videos/%s/hash' % video_id,
 121                 video_id, 'Downloading security hash for %s' % resource_id, query={
 122                     'player': 'desktop',
 123                     'version': '5.19.1',
 124                     'resource_id': resource_id,
 125                 })
 126
 127             security_hash = security.get('hash')
 128             if not security_hash:
 129                 message = security.get('message')
 130                 if message:
 131                     raise ExtractorError(
 132                         '%s returned error: %s' % (self.IE_NAME, message), expected=True)
 133                 continue
 134
 135             hash_code = security_hash[:2]
 136             padding = '%010d' % random.randint(1, 10000000000)
 137             if hash_code in ('04', '14'):
 138                 received_time = security_hash[3:13]
 139                 received_md5 = security_hash[24:]
 140                 hash_prefix = security_hash[:23]
 141             elif hash_code in ('02', '12', '03', '13'):
 142                 received_time = security_hash[2:12]
 143                 received_md5 = security_hash[22:]
 144                 padding += '1'
 145                 hash_prefix = '05' + security_hash[:22]
 146
 147             padded_sign_time = compat_str(int(received_time) + 86400) + padding
 148             md5_data = (received_md5 + padded_sign_time + '0xAC10FD').encode()
 149             signed_md5 = base64.urlsafe_b64encode(hashlib.md5(md5_data).digest()).decode().strip('=')
 150             signed_hash = hash_prefix + padded_sign_time + signed_md5
 151             signed_url = '%s?h=%s&k=html5&a=%s&u=%s' % (resource_url, signed_hash, 'F' if video.get('subscriber_only') else 'A', security.get('user') or '')
 152
 153             if resource_id.endswith('m3u8') or resource_url.endswith('.m3u8'):
 154                 formats.extend(self._extract_m3u8_formats(
 155                     signed_url, resource_id, 'mp4', entry_protocol='m3u8_native',
 156                     m3u8_id='hls', fatal=False))
 157             elif resource_id.endswith('mpd') or resource_url.endswith('.mpd'):
 158                 formats.extend(self._extract_mpd_formats(
 159                     signed_url, resource_id, mpd_id='dash', fatal=False))
 160             elif resource_id.endswith('manifest') or resource_url.endswith('/manifest'):
 161                 formats.extend(self._extract_ism_formats(
 162                     signed_url, resource_id, ism_id='mss', fatal=False))
 163             else:
 164                 formats.append({
 165                     'url': signed_url,
 166                     'format_id': 'http-%s' % resource_id,
 167                     'height': int_or_none(resource.get('height')),
 168                 })
 169
 170         self._sort_formats(formats)
 171
 172         duration = float_or_none(video.get('duration'), 1000)
 173         uploader = video.get('channel')
 174         uploader_id = str_or_none(video.get('channel_id'))
 175
 176         return {
 177             'id': video_id,
 178             'title': title,
 179             'duration': duration,
 180             'uploader': uploader,
 181             'uploader_id': uploader_id,
 182             'formats': formats,
 183             'subtitles': subtitles,
 184         }
 185
 186
 187 class GloboArticleIE(InfoExtractor):
 188     _VALID_URL = r'https?://.+?\.globo\.com/(?:[^/]+/)*(?P<id>[^/.]+)(?:\.html)?'
 189
 190     _VIDEOID_REGEXES = [
 191         r'\bdata-video-id=["\'](\d{7,})',
 192         r'\bdata-player-videosids=["\'](\d{7,})',
 193         r'\bvideosIDs\s*:\s*["\']?(\d{7,})',
 194         r'\bdata-id=["\'](\d{7,})',
 195         r'<div[^>]+\bid=["\'](\d{7,})',
 196     ]
 197
 198     _TESTS = [{
 199         'url': 'http://g1.globo.com/jornal-nacional/noticia/2014/09/novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes.html',
 200         'info_dict': {
 201             'id': 'novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes',
 202             'title': 'Novidade na fiscalização de bagagem pela Receita provoca discussões',
 203             'description': 'md5:c3c4b4d4c30c32fce460040b1ac46b12',
 204         },
 205         'playlist_count': 1,
 206     }, {
 207         'url': 'http://g1.globo.com/pr/parana/noticia/2016/09/mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato.html',
 208         'info_dict': {
 209             'id': 'mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato',
 210             'title': "Lula era o 'comandante máximo' do esquema da Lava Jato, diz MPF",
 211             'description': 'md5:8aa7cc8beda4dc71cc8553e00b77c54c',
 212         },
 213         'playlist_count': 6,
 214     }, {
 215         'url': 'http://gq.globo.com/Prazeres/Poder/noticia/2015/10/all-o-desafio-assista-ao-segundo-capitulo-da-serie.html',
 216         'only_matching': True,
 217     }, {
 218         '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',
 219         'only_matching': True,
 220     }, {
 221         'url': 'http://oglobo.globo.com/rio/a-amizade-entre-um-entregador-de-farmacia-um-piano-19946271',
 222         'only_matching': True,
 223     }]
 224
 225     @classmethod
 226     def suitable(cls, url):
 227         return False if GloboIE.suitable(url) else super(GloboArticleIE, cls).suitable(url)
 228
 229     def _real_extract(self, url):
 230         display_id = self._match_id(url)
 231         webpage = self._download_webpage(url, display_id)
 232         video_ids = []
 233         for video_regex in self._VIDEOID_REGEXES:
 234             video_ids.extend(re.findall(video_regex, webpage))
 235         entries = [
 236             self.url_result('globo:%s' % video_id, GloboIE.ie_key())
 237             for video_id in orderedSet(video_ids)]
 238         title = self._og_search_title(webpage, fatal=False)
 239         description = self._html_search_meta('description', webpage)
 240         return self.playlist_result(entries, display_id, title, description)