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     _LOGGED_IN = False
  27     _NETRC_MACHINE = 'globo'
  28     _TESTS = [{
  29         'url': 'http://g1.globo.com/carros/autoesporte/videos/t/exclusivos-do-g1/v/mercedes-benz-gla-passa-por-teste-de-colisao-na-europa/3607726/',
  30         'md5': 'b3ccc801f75cd04a914d51dadb83a78d',
  31         'info_dict': {
  32             'id': '3607726',
  33             'ext': 'mp4',
  34             'title': 'Mercedes-Benz GLA passa por teste de colisão na Europa',
  35             'duration': 103.204,
  36             'uploader': 'Globo.com',
  37             'uploader_id': '265',
  38         },
  39     }, {
  40         'url': 'http://globoplay.globo.com/v/4581987/',
  41         'md5': 'f36a1ecd6a50da1577eee6dd17f67eff',
  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     }, {
  51         'url': 'http://canalbrasil.globo.com/programas/sangue-latino/videos/3928201.html',
  52         'only_matching': True,
  53     }, {
  54         'url': 'http://globosatplay.globo.com/globonews/v/4472924/',
  55         'only_matching': True,
  56     }, {
  57         'url': 'http://globotv.globo.com/t/programa/v/clipe-sexo-e-as-negas-adeus/3836166/',
  58         'only_matching': True,
  59     }, {
  60         '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/',
  61         'only_matching': True,
  62     }, {
  63         'url': 'http://canaloff.globo.com/programas/desejar-profundo/videos/4518560.html',
  64         'only_matching': True,
  65     }, {
  66         'url': 'globo:3607726',
  67         'only_matching': True,
  68     }]
  69
  70     def _real_initialize(self):
  71         if self._LOGGED_IN:
  72             return
  73
  74         email, password = self._get_login_info()
  75         if email is None:
  76             return
  77
  78         try:
  79             self._download_json(
  80                 'https://login.globo.com/api/authentication', None, data=json.dumps({
  81                     'payload': {
  82                         'email': email,
  83                         'password': password,
  84                         'serviceId': 4654,
  85                     },
  86                 }).encode(), headers={
  87                     'Content-Type': 'application/json; charset=utf-8',
  88                 })
  89         except ExtractorError as e:
  90             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
  91                 resp = self._parse_json(e.cause.read(), None)
  92                 raise ExtractorError(resp.get('userMessage') or resp['id'], expected=True)
  93             raise
  94         self._LOGGED_IN = True
  95
  96     def _real_extract(self, url):
  97         video_id = self._match_id(url)
  98
  99         video = self._download_json(
 100             'http://api.globovideos.com/videos/%s/playlist' % video_id,
 101             video_id)['videos'][0]
 102
 103         title = video['title']
 104
 105         formats = []
 106         for resource in video['resources']:
 107             resource_id = resource.get('_id')
 108             resource_url = resource.get('url')
 109             if not resource_id or not resource_url:
 110                 continue
 111
 112             security = self._download_json(
 113                 'http://security.video.globo.com/videos/%s/hash' % video_id,
 114                 video_id, 'Downloading security hash for %s' % resource_id, query={
 115                     'player': 'flash',
 116                     'version': '17.0.0.132',
 117                     'resource_id': resource_id,
 118                 })
 119
 120             security_hash = security.get('hash')
 121             if not security_hash:
 122                 message = security.get('message')
 123                 if message:
 124                     raise ExtractorError(
 125                         '%s returned error: %s' % (self.IE_NAME, message), expected=True)
 126                 continue
 127
 128             hash_code = security_hash[:2]
 129             received_time = security_hash[2:12]
 130             received_random = security_hash[12:22]
 131             received_md5 = security_hash[22:]
 132
 133             sign_time = compat_str(int(received_time) + 86400)
 134             padding = '%010d' % random.randint(1, 10000000000)
 135
 136             md5_data = (received_md5 + sign_time + padding + '0xFF01DD').encode()
 137             signed_md5 = base64.urlsafe_b64encode(hashlib.md5(md5_data).digest()).decode().strip('=')
 138             signed_hash = hash_code + received_time + received_random + sign_time + padding + signed_md5
 139
 140             signed_url = '%s?h=%s&k=%s' % (resource_url, signed_hash, 'flash')
 141             if resource_id.endswith('m3u8') or resource_url.endswith('.m3u8'):
 142                 formats.extend(self._extract_m3u8_formats(
 143                     signed_url, resource_id, 'mp4', entry_protocol='m3u8_native',
 144                     m3u8_id='hls', fatal=False))
 145             elif resource_id.endswith('mpd') or resource_url.endswith('.mpd'):
 146                 formats.extend(self._extract_mpd_formats(
 147                     signed_url, resource_id, mpd_id='dash', fatal=False))
 148             elif resource_id.endswith('manifest') or resource_url.endswith('/manifest'):
 149                 formats.extend(self._extract_ism_formats(
 150                     signed_url, resource_id, ism_id='mss', fatal=False))
 151             else:
 152                 formats.append({
 153                     'url': signed_url,
 154                     'format_id': 'http-%s' % resource_id,
 155                     'height': int_or_none(resource.get('height')),
 156                 })
 157
 158         self._sort_formats(formats)
 159
 160         duration = float_or_none(video.get('duration'), 1000)
 161         uploader = video.get('channel')
 162         uploader_id = str_or_none(video.get('channel_id'))
 163
 164         return {
 165             'id': video_id,
 166             'title': title,
 167             'duration': duration,
 168             'uploader': uploader,
 169             'uploader_id': uploader_id,
 170             'formats': formats
 171         }
 172
 173
 174 class GloboArticleIE(InfoExtractor):
 175     _VALID_URL = r'https?://.+?\.globo\.com/(?:[^/]+/)*(?P<id>[^/.]+)(?:\.html)?'
 176
 177     _VIDEOID_REGEXES = [
 178         r'\bdata-video-id=["\'](\d{7,})',
 179         r'\bdata-player-videosids=["\'](\d{7,})',
 180         r'\bvideosIDs\s*:\s*["\']?(\d{7,})',
 181         r'\bdata-id=["\'](\d{7,})',
 182         r'<div[^>]+\bid=["\'](\d{7,})',
 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
 212     @classmethod
 213     def suitable(cls, url):
 214         return False if GloboIE.suitable(url) else super(GloboArticleIE, cls).suitable(url)
 215
 216     def _real_extract(self, url):
 217         display_id = self._match_id(url)
 218         webpage = self._download_webpage(url, display_id)
 219         video_ids = []
 220         for video_regex in self._VIDEOID_REGEXES:
 221             video_ids.extend(re.findall(video_regex, webpage))
 222         entries = [
 223             self.url_result('globo:%s' % video_id, GloboIE.ie_key())
 224             for video_id in orderedSet(video_ids)]
 225         title = self._og_search_title(webpage, fatal=False)
 226         description = self._html_search_meta('description', webpage)
 227         return self.playlist_result(entries, display_id, title, description)