_ 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 import json
   9 import random as rnd
  10
  11 class WeiboIE(InfoExtractor):
  12     _VALID_URL = r'https?://weibo\.com/[0-9]+/(?P<id>[a-zA-Z0-9]+)'
  13     _TEST = {
  14             'url': 'https://weibo.com/6275294458/Fp6RGfbff?from=page_1005056275294458_profile&wvr=6&mod=weibotime&type=comment',
  15             'md5': 'TODO: md5 sum of the first 10241 bytes of the video file (use --test)',
  16             'info_dict': {
  17                 'id': '42',
  18                 'ext': 'mp4',
  19                 'title': 'Video title goes here',
  20                 'thumbnail': r're:^https?://.*\.jpg$',
  21                 # TODO more properties, either as:
  22                 # * A value
  23                 # * MD5 checksum; start the string with md5:
  24                 # * A regular expression; start the string with re:
  25                 # * Any Python type (for example int or float)
  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         print("webpage", webpage)
  61
  62         p = webpage.split("&&")[1] # split "gen_callback && gen_callback(...)"
  63         i1 = p.find('{')
  64         i2 = p.rfind('}')
  65         j = p[i1:i2+1] # get JSON object
  66         d = json.loads(j)
  67         tid = d["data"]["tid"]
  68         cnfd = "%03d" % d["data"]["confidence"]
  69
  70         param = urlencode({
  71             'a': 'incarnate',
  72             't': tid,
  73             'w': 2,
  74             'c': cnfd,
  75             'cb': 'cross_domain',
  76             'from': 'weibo',
  77             '_rand': rnd.random()
  78             })
  79         gencallback_url = "https://passport.weibo.com/visitor/visitor?" + param
  80         webpage,urlh = self._download_webpage_handle(gencallback_url, video_id, note="gen callback")
  81         print("webpage", webpage)
  82
  83         webpage,urlh = self._download_webpage_handle(url, video_id, headers=headers, note="retry to visit the page")
  84         print("webpage", webpage)
  85
  86         # TODO more code goes here, for example ...
  87         title = self._html_search_regex(r'<title>(.+?)</title>', webpage, 'title')
  88
  89         video_sources = self._search_regex(r'video-sources=(.+?)', webpage, 'video_sources')
  90         print("video_sources:", video_sources)
  91         return {
  92                 'id': video_id,
  93                 'title': title,
  94                 'description': self._og_search_description(webpage),
  95                 'uploader': self._search_regex(r'<div[^>]+id="uploader"[^>]*>([^<]+)<', webpage, 'uploader', fatal=False),
  96                 # TODO more properties (see youtube_dl/extractor/common.py)
  97                 }