_ Git - youtube-dl/blob - youtube_dl/extractor/weibo.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5
   6 from urllib.request import Request
   7 from urllib.parse import urlencode
   8 from urllib import parse
   9 import json
  10 import random as rnd
  11 from os import path
  12 import re
  13
  14 from ..utils import (
  15     js_to_json,
  16 )
  17
  18 class WeiboIE(InfoExtractor):
  19     _VALID_URL = r'https?://weibo\.com/[0-9]+/(?P<id>[a-zA-Z0-9]+)'
  20     _TEST = {
  21             'url': 'https://weibo.com/6275294458/Fp6RGfbff?type=comment',
  22             'info_dict': {
  23                 'id': 'Fp6RGfbff',
  24                 'ext': 'mp4',
  25                 'title': 'You should have servants to massage you,... 来自Hosico_猫 - 微博',
  26                 }
  27             }
  28
  29     def _real_extract(self, url):
  30         video_id = self._match_id(url)
  31         headers = {
  32             'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8',
  33             'Accept-Encoding': 'gzip, deflate, br',
  34             'Accept-Language': 'en,zh-CN;q=0.9,zh;q=0.8',
  35             'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_13_1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.84 Safari/537.36',
  36             'Upgrade-Insecure-Requests': '1',
  37         }
  38         # to get Referer url for genvisitor
  39         webpage,urlh = self._download_webpage_handle(url, video_id, headers=headers, note="first visit the page")
  40
  41         visitor_url = urlh.geturl()
  42
  43         data = urlencode({
  44             "cb": "gen_callback",
  45             "fp": '{"os":"2","browser":"Gecko57,0,0,0","fonts":"undefined","screenInfo":"1440*900*24","plugins":""}',
  46             }).encode()
  47         headers = {
  48                 'Accept-Encoding': 'gzip, deflate, br',
  49                 'Accept': '*/*',
  50                 'Referer': visitor_url,
  51                 }
  52
  53         r_genvisitor = Request(
  54             'https://passport.weibo.com/visitor/genvisitor',
  55             data = data,
  56             headers = headers,
  57             method = 'POST'
  58             )
  59         webpage,urlh = self._download_webpage_handle(r_genvisitor, video_id, note="gen visitor")
  60
  61         p = webpage.split("&&")[1] # split "gen_callback && gen_callback(...)"
  62         i1 = p.find('{')
  63         i2 = p.rfind('}')
  64         j = p[i1:i2+1] # get JSON object
  65         d = json.loads(j)
  66         tid = d["data"]["tid"]
  67         cnfd = "%03d" % d["data"]["confidence"]
  68
  69         param = urlencode({
  70             'a': 'incarnate',
  71             't': tid,
  72             'w': 2,
  73             'c': cnfd,
  74             'cb': 'cross_domain',
  75             'from': 'weibo',
  76             '_rand': rnd.random()
  77             })
  78         gencallback_url = "https://passport.weibo.com/visitor/visitor?" + param
  79         webpage,urlh = self._download_webpage_handle(gencallback_url, video_id, note="gen callback")
  80
  81         webpage,urlh = self._download_webpage_handle(url, video_id, headers=headers, note="retry to visit the page")
  82
  83         # TODO more code goes here, for example ...
  84         title = self._html_search_regex(r'<title>(.+?)</title>', webpage, 'title')
  85
  86         video_sources_text = self._search_regex("video-sources=\\\\\"(.+?)\"", webpage, 'video_sources')
  87
  88         video_formats = parse.parse_qs(video_sources_text)
  89
  90         formats = []
  91         supported_resolutions = ['720', '480']
  92         for res in supported_resolutions:
  93             f = video_formats.get(res)
  94             if isinstance(f, list):
  95                 if len(f) > 0:
  96                     vid_url = f[0]
  97                     formats.append({
  98                         'url': vid_url,
  99                         'format': 'mp4',
 100                         'height': int(res),
 101                         })
 102         self._sort_formats(formats)
 103         uploader = self._og_search_property('nick-name', webpage, 'uploader', default = None)
 104         return {
 105                 'id': video_id,
 106                 'title': title,
 107                 'uploader': uploader,
 108                 'formats': formats
 109                 # TODO more properties (see youtube_dl/extractor/common.py)
 110                 }
 111
 112 class WeiboMobileIE(InfoExtractor):
 113     _VALID_URL = r'https?://m.weibo.cn/status/(?P<id>[0-9]+)(\?.+)?'
 114     _TEST = {
 115             'url': 'https://m.weibo.cn/status/4189191225395228?wm=3333_2001&sourcetype=weixin&featurecode=newtitle&from=singlemessage&isappinstalled=0',
 116             'info_dict': {
 117                 'id': '4189191225395228',
 118                 'ext': 'mp4',
 119                 'title': '午睡当然是要甜甜蜜蜜的啦',
 120                 'uploader': '柴犬柴犬'
 121                 }
 122             }
 123
 124     def _real_extract(self, url):
 125         video_id = self._match_id(url)
 126         headers = {
 127             'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8',
 128             'Accept-Encoding': 'gzip, deflate, br',
 129             'Accept-Language': 'en,zh-CN;q=0.9,zh;q=0.8',
 130             'Upgrade-Insecure-Requests': '1',
 131         }
 132         # to get Referer url for genvisitor
 133         webpage,urlh = self._download_webpage_handle(url, video_id, headers=headers, note="visit the page")
 134         js_code = self._search_regex(r'var\s+\$render_data\s*=\s*\[({.*})\]\[0\] \|\| {};', webpage, 'js_code', flags = re.DOTALL)
 135         weibo_info = self._parse_json(js_code, video_id, transform_source=js_to_json)
 136         page_info = weibo_info['status']['page_info']
 137         title = weibo_info['status']['status_title']
 138         format = {
 139             'url': page_info['media_info']['stream_url'],
 140             'format': 'mp4',
 141                   }
 142         formats = [format]
 143         uploader = weibo_info['status']['user']['screen_name']
 144
 145         return {
 146                 'id': video_id,
 147                 'title': title,
 148                 'uploader': uploader,
 149                 'formats': formats
 150                 # TODO more properties (see youtube_dl/extractor/common.py)
 151                 }