X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/099764c0c91f4ad7db03d9347798f8619383ea7e..f8df414a4abcde0ddd39325dac26ca071d2d15c6:/youtube_dl/extractor/globo.py diff --git a/youtube_dl/extractor/globo.py b/youtube_dl/extractor/globo.py index 66ca379..dc7b266 100644 --- a/youtube_dl/extractor/globo.py +++ b/youtube_dl/extractor/globo.py @@ -2,6 +2,7 @@ from __future__ import unicode_literals import random +import re import math from .common import InfoExtractor @@ -13,66 +14,63 @@ from ..compat import ( from ..utils import ( ExtractorError, float_or_none, + int_or_none, + orderedSet, + str_or_none, ) class GloboIE(InfoExtractor): - _VALID_URL = 'https?://.+?\.globo\.com/(?P.+)' + _VALID_URL = r'(?:globo:|https?://.+?\.globo\.com/(?:[^/]+/)*(?:v/(?:[^/]+/)?|videos/))(?P\d{7,})' _API_URL_TEMPLATE = 'http://api.globovideos.com/videos/%s/playlist' - _SECURITY_URL_TEMPLATE = 'http://security.video.globo.com/videos/%s/hash?player=flash&version=2.9.9.50&resource_id=%s' - - _VIDEOID_REGEXES = [ - r'\bdata-video-id="(\d+)"', - r'\bdata-player-videosids="(\d+)"', - r']+\bid="(\d+)"', - ] + _SECURITY_URL_TEMPLATE = 'http://security.video.globo.com/videos/%s/hash?player=flash&version=17.0.0.132&resource_id=%s' _RESIGN_EXPIRATION = 86400 - _TESTS = [ - { - 'url': 'http://globotv.globo.com/sportv/futebol-nacional/v/os-gols-de-atletico-mg-3-x-2-santos-pela-24a-rodada-do-brasileirao/3654973/', - 'md5': '03ebf41cb7ade43581608b7d9b71fab0', - 'info_dict': { - 'id': '3654973', - 'ext': 'mp4', - 'title': 'Os gols de Atlético-MG 3 x 2 Santos pela 24ª rodada do Brasileirão', - 'duration': 251.585, - 'uploader': 'SporTV', - 'uploader_id': 698, - 'like_count': int, - } + _TESTS = [{ + 'url': 'http://g1.globo.com/carros/autoesporte/videos/t/exclusivos-do-g1/v/mercedes-benz-gla-passa-por-teste-de-colisao-na-europa/3607726/', + 'md5': 'b3ccc801f75cd04a914d51dadb83a78d', + 'info_dict': { + 'id': '3607726', + 'ext': 'mp4', + 'title': 'Mercedes-Benz GLA passa por teste de colisão na Europa', + 'duration': 103.204, + 'uploader': 'Globo.com', + 'uploader_id': '265', }, - { - 'url': 'http://g1.globo.com/carros/autoesporte/videos/t/exclusivos-do-g1/v/mercedes-benz-gla-passa-por-teste-de-colisao-na-europa/3607726/', - 'md5': 'b3ccc801f75cd04a914d51dadb83a78d', - 'info_dict': { - 'id': '3607726', - 'ext': 'mp4', - 'title': 'Mercedes-Benz GLA passa por teste de colisão na Europa', - 'duration': 103.204, - 'uploader': 'Globo.com', - 'uploader_id': 265, - 'like_count': int, - } + }, { + 'url': 'http://globoplay.globo.com/v/4581987/', + 'md5': 'f36a1ecd6a50da1577eee6dd17f67eff', + 'info_dict': { + 'id': '4581987', + 'ext': 'mp4', + 'title': 'Acidentes de trânsito estão entre as maiores causas de queda de energia em SP', + 'duration': 137.973, + 'uploader': 'Rede Globo', + 'uploader_id': '196', }, - { - 'url': 'http://g1.globo.com/jornal-nacional/noticia/2014/09/novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes.html', - 'md5': '307fdeae4390ccfe6ba1aa198cf6e72b', - 'info_dict': { - 'id': '3652183', - 'ext': 'mp4', - 'title': 'Receita Federal explica como vai fiscalizar bagagens de quem retorna ao Brasil de avião', - 'duration': 110.711, - 'uploader': 'Rede Globo', - 'uploader_id': 196, - 'like_count': int, - } - }, - ] - - class MD5(): + }, { + 'url': 'http://canalbrasil.globo.com/programas/sangue-latino/videos/3928201.html', + 'only_matching': True, + }, { + 'url': 'http://globosatplay.globo.com/globonews/v/4472924/', + 'only_matching': True, + }, { + 'url': 'http://globotv.globo.com/t/programa/v/clipe-sexo-e-as-negas-adeus/3836166/', + 'only_matching': True, + }, { + '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/', + 'only_matching': True, + }, { + 'url': 'http://canaloff.globo.com/programas/desejar-profundo/videos/4518560.html', + 'only_matching': True, + }, { + 'url': 'globo:3607726', + 'only_matching': True, + }] + + class MD5(object): HEX_FORMAT_LOWERCASE = 0 HEX_FORMAT_UPPERCASE = 1 BASE64_PAD_CHARACTER_DEFAULT_COMPLIANCE = '' @@ -339,23 +337,15 @@ class GloboIE(InfoExtractor): def _real_extract(self, url): video_id = self._match_id(url) - webpage = self._download_webpage(url, video_id) - video_id = self._search_regex(self._VIDEOID_REGEXES, webpage, 'video id') - video = self._download_json( self._API_URL_TEMPLATE % video_id, video_id)['videos'][0] title = video['title'] - duration = float_or_none(video['duration'], 1000) - like_count = video['likes'] - uploader = video['channel'] - uploader_id = video['channel_id'] formats = [] - for resource in video['resources']: resource_id = resource.get('_id') - if not resource_id: + if not resource_id or resource_id.endswith('manifest'): continue security = self._download_json( @@ -381,20 +371,86 @@ class GloboIE(InfoExtractor): signed_md5 = self.MD5.b64_md5(received_md5 + compat_str(sign_time) + padding) signed_hash = hash_code + compat_str(received_time) + received_random + compat_str(sign_time) + padding + signed_md5 - formats.append({ - 'url': '%s?h=%s&k=%s' % (resource['url'], signed_hash, 'flash'), - 'format_id': resource_id, - 'height': resource['height'] - }) + resource_url = resource['url'] + signed_url = '%s?h=%s&k=%s' % (resource_url, signed_hash, 'flash') + if resource_id.endswith('m3u8') or resource_url.endswith('.m3u8'): + formats.extend(self._extract_m3u8_formats( + signed_url, resource_id, 'mp4', entry_protocol='m3u8_native', + m3u8_id='hls', fatal=False)) + else: + formats.append({ + 'url': signed_url, + 'format_id': 'http-%s' % resource_id, + 'height': int_or_none(resource.get('height')), + }) self._sort_formats(formats) + duration = float_or_none(video.get('duration'), 1000) + uploader = video.get('channel') + uploader_id = str_or_none(video.get('channel_id')) + return { 'id': video_id, 'title': title, 'duration': duration, 'uploader': uploader, 'uploader_id': uploader_id, - 'like_count': like_count, 'formats': formats - } \ No newline at end of file + } + + +class GloboArticleIE(InfoExtractor): + _VALID_URL = r'https?://.+?\.globo\.com/(?:[^/]+/)*(?P[^/.]+)(?:\.html)?' + + _VIDEOID_REGEXES = [ + r'\bdata-video-id=["\'](\d{7,})', + r'\bdata-player-videosids=["\'](\d{7,})', + r'\bvideosIDs\s*:\s*["\']?(\d{7,})', + r'\bdata-id=["\'](\d{7,})', + r']+\bid=["\'](\d{7,})', + ] + + _TESTS = [{ + 'url': 'http://g1.globo.com/jornal-nacional/noticia/2014/09/novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes.html', + 'info_dict': { + 'id': 'novidade-na-fiscalizacao-de-bagagem-pela-receita-provoca-discussoes', + 'title': 'Novidade na fiscalização de bagagem pela Receita provoca discussões', + 'description': 'md5:c3c4b4d4c30c32fce460040b1ac46b12', + }, + 'playlist_count': 1, + }, { + 'url': 'http://g1.globo.com/pr/parana/noticia/2016/09/mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato.html', + 'info_dict': { + 'id': 'mpf-denuncia-lula-marisa-e-mais-seis-na-operacao-lava-jato', + 'title': "Lula era o 'comandante máximo' do esquema da Lava Jato, diz MPF", + 'description': 'md5:8aa7cc8beda4dc71cc8553e00b77c54c', + }, + 'playlist_count': 6, + }, { + 'url': 'http://gq.globo.com/Prazeres/Poder/noticia/2015/10/all-o-desafio-assista-ao-segundo-capitulo-da-serie.html', + 'only_matching': True, + }, { + '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', + 'only_matching': True, + }, { + 'url': 'http://oglobo.globo.com/rio/a-amizade-entre-um-entregador-de-farmacia-um-piano-19946271', + 'only_matching': True, + }] + + @classmethod + def suitable(cls, url): + return False if GloboIE.suitable(url) else super(GloboArticleIE, cls).suitable(url) + + def _real_extract(self, url): + display_id = self._match_id(url) + webpage = self._download_webpage(url, display_id) + video_ids = [] + for video_regex in self._VIDEOID_REGEXES: + video_ids.extend(re.findall(video_regex, webpage)) + entries = [ + self.url_result('globo:%s' % video_id, GloboIE.ie_key()) + for video_id in orderedSet(video_ids)] + title = self._og_search_title(webpage, fatal=False) + description = self._html_search_meta('description', webpage) + return self.playlist_result(entries, display_id, title, description)