#!/usr/bin/env python # -*- coding: utf-8 -*- from ..common import * from ..extractor import VideoExtractor from json import loads from urllib.parse import urlsplit from os.path import dirname import re class MGTV(VideoExtractor): name = "芒果 (MGTV)" # Last updated: 2016-11-13 stream_types = [ {'id': 'hd', 'container': 'ts', 'video_profile': '超清'}, {'id': 'sd', 'container': 'ts', 'video_profile': '高清'}, {'id': 'ld', 'container': 'ts', 'video_profile': '标清'}, ] id_dic = {i['video_profile']:(i['id']) for i in stream_types} api_endpoint = 'http://pcweb.api.mgtv.com/player/video?video_id={video_id}' @staticmethod def get_vid_from_url(url): """Extracts video ID from URL. """ vid = match1(url, 'https?://www.mgtv.com/(?:b|l)/\d+/(\d+).html') if not vid: vid = match1(url, 'https?://www.mgtv.com/hz/bdpz/\d+/(\d+).html') return vid #---------------------------------------------------------------------- @staticmethod def get_mgtv_real_url(url): """str->list of str Give you the real URLs.""" content = loads(get_content(url)) m3u_url = content['info'] split = urlsplit(m3u_url) base_url = "{scheme}://{netloc}{path}/".format(scheme = split[0], netloc = split[1], path = dirname(split[2])) content = get_content(content['info']) #get the REAL M3U url, maybe to be changed later? segment_list = [] segments_size = 0 for i in content.split(): if not i.startswith('#'): #not the best way, better we use the m3u8 package segment_list.append(base_url + i) # use ext-info for fast size calculate elif i.startswith('#EXT-MGTV-File-SIZE:'): segments_size += int(i[i.rfind(':')+1:]) return m3u_url, segments_size, segment_list def download_playlist_by_url(self, url, **kwargs): pass def prepare(self, **kwargs): if self.url: self.vid = self.get_vid_from_url(self.url) content = get_content(self.api_endpoint.format(video_id = self.vid)) content = loads(content) self.title = content['data']['info']['title'] domain = content['data']['stream_domain'][0] #stream_available = [i['name'] for i in content['data']['stream']] stream_available = {} for i in content['data']['stream']: stream_available[i['name']] = i['url'] for s in self.stream_types: if s['video_profile'] in stream_available.keys(): quality_id = self.id_dic[s['video_profile']] url = stream_available[s['video_profile']] url = domain + re.sub( r'(\&arange\=\d+)', '', url) #Un-Hum m3u8_url, m3u8_size, segment_list_this = self.get_mgtv_real_url(url) stream_fileid_list = [] for i in segment_list_this: stream_fileid_list.append(os.path.basename(i).split('.')[0]) #make pieces pieces = [] for i in zip(stream_fileid_list, segment_list_this): pieces.append({'fileid': i[0], 'segs': i[1],}) self.streams[quality_id] = { 'container': s['container'], 'video_profile': s['video_profile'], 'size': m3u8_size, 'pieces': pieces, 'm3u8_url': m3u8_url } if not kwargs['info_only']: self.streams[quality_id]['src'] = segment_list_this def extract(self, **kwargs): if 'stream_id' in kwargs and kwargs['stream_id']: # Extract the stream stream_id = kwargs['stream_id'] if stream_id not in self.streams: log.e('[Error] Invalid video format.') log.e('Run \'-i\' command with no specific video format to view all available formats.') exit(2) else: # Extract stream with the best quality stream_id = self.streams_sorted[0]['id'] def download(self, **kwargs): if 'stream_id' in kwargs and kwargs['stream_id']: stream_id = kwargs['stream_id'] else: stream_id = 'null' # print video info only if 'info_only' in kwargs and kwargs['info_only']: if stream_id != 'null': if 'index' not in kwargs: self.p(stream_id) else: self.p_i(stream_id) else: # Display all available streams if 'index' not in kwargs: self.p([]) else: stream_id = self.streams_sorted[0]['id'] if 'id' in self.streams_sorted[0] else self.streams_sorted[0]['itag'] self.p_i(stream_id) # default to use the best quality if stream_id == 'null': stream_id = self.streams_sorted[0]['id'] stream_info = self.streams[stream_id] if not kwargs['info_only']: if player: # with m3u8 format because some video player can process urls automatically (e.g. mpv) launch_player(player, [stream_info['m3u8_url']]) else: download_urls(stream_info['src'], self.title, stream_info['container'], stream_info['size'], output_dir=kwargs['output_dir'], merge=kwargs.get('merge', True)) # av=stream_id in self.dash_streams) site = MGTV() download = site.download_by_url download_playlist = site.download_playlist_by_url