_ 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
  13 class WeiboIE(InfoExtractor):
  14     _VALID_URL = r'https?://weibo\.com/[0-9]+/(?P<id>[a-zA-Z0-9]+)'
  15     _TEST = {
  16             'url': 'https://weibo.com/6275294458/Fp6RGfbff?type=comment',
  17             'info_dict': {
  18                 'id': 'Fp6RGfbff',
  19                 'ext': 'mp4',
  20                 'title': 'You should have servants to massage you,... 来自Hosico_猫 - 微博',
  21                 }
  22             }
  23
  24     def _real_extract(self, url):
  25         video_id = self._match_id(url)
  26         headers = {
  27             'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8',
  28             'Accept-Encoding': 'gzip, deflate, br',
  29             'Accept-Language': 'en,zh-CN;q=0.9,zh;q=0.8',
  30             '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',
  31             'Upgrade-Insecure-Requests': '1',
  32         }
  33         # to get Referer url for genvisitor
  34         webpage,urlh = self._download_webpage_handle(url, video_id, headers=headers, note="first visit the page")
  35
  36         visitor_url = urlh.geturl()
  37
  38         data = urlencode({
  39             "cb": "gen_callback",
  40             "fp": '{"os":"2","browser":"Gecko57,0,0,0","fonts":"undefined","screenInfo":"1440*900*24","plugins":""}',
  41             }).encode()
  42         headers = {
  43                 'Accept-Encoding': 'gzip, deflate, br',
  44                 'Accept': '*/*',
  45                 'Referer': visitor_url,
  46                 }
  47
  48         r_genvisitor = Request(
  49             'https://passport.weibo.com/visitor/genvisitor',
  50             data = data,
  51             headers = headers,
  52             method = 'POST'
  53             )
  54         webpage,urlh = self._download_webpage_handle(r_genvisitor, video_id, note="gen visitor")
  55
  56         p = webpage.split("&&")[1] # split "gen_callback && gen_callback(...)"
  57         i1 = p.find('{')
  58         i2 = p.rfind('}')
  59         j = p[i1:i2+1] # get JSON object
  60         d = json.loads(j)
  61         tid = d["data"]["tid"]
  62         cnfd = "%03d" % d["data"]["confidence"]
  63
  64         param = urlencode({
  65             'a': 'incarnate',
  66             't': tid,
  67             'w': 2,
  68             'c': cnfd,
  69             'cb': 'cross_domain',
  70             'from': 'weibo',
  71             '_rand': rnd.random()
  72             })
  73         gencallback_url = "https://passport.weibo.com/visitor/visitor?" + param
  74         webpage,urlh = self._download_webpage_handle(gencallback_url, video_id, note="gen callback")
  75
  76         webpage,urlh = self._download_webpage_handle(url, video_id, headers=headers, note="retry to visit the page")
  77
  78         # TODO more code goes here, for example ...
  79         title = self._html_search_regex(r'<title>(.+?)</title>', webpage, 'title')
  80
  81         video_sources_text = self._search_regex("video-sources=\\\\\"(.+?)\"", webpage, 'video_sources')
  82
  83         video_formats = parse.parse_qs(video_sources_text)
  84
  85         formats = []
  86         supported_resolutions = ['720', '480']
  87         for res in supported_resolutions:
  88             f = video_formats.get(res)
  89             if isinstance(f, list):
  90                 if len(f) > 0:
  91                     vid_url = f[0]
  92                     formats.append({
  93                         'url': vid_url,
  94                         'format': 'mp4',
  95                         'height': int(res),
  96                         })
  97         self._sort_formats(formats)
  98         uploader = self._og_search_property('nick-name', webpage, 'uploader', default = None)
  99         return {
 100                 'id': video_id,
 101                 'title': title,
 102                 'uploader': uploader,
 103                 'formats': formats
 104                 # TODO more properties (see youtube_dl/extractor/common.py)
 105                 }