_ 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 import json
   7 import random as rnd
   8 import re
   9
  10 from ..compat import (
  11     compat_urllib_parse_urlencode as urlencode,
  12     compat_urllib_request as request,
  13     compat_urlparse as parse,
  14 )
  15 from ..utils import (
  16     js_to_json,
  17 )
  18
  19
  20 class WeiboIE(InfoExtractor):
  21     _VALID_URL = r'https?://weibo\.com/[0-9]+/(?P<id>[a-zA-Z0-9]+)'
  22     _TEST = {
  23         'url': 'https://weibo.com/6275294458/Fp6RGfbff?type=comment',
  24         'info_dict': {
  25             'id': 'Fp6RGfbff',
  26             'ext': 'mp4',
  27             'title': 'You should have servants to massage you,... 来自Hosico_猫 - 微博',
  28         }
  29     }
  30
  31     def _real_extract(self, url):
  32         video_id = self._match_id(url)
  33         headers = {
  34             'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8',
  35             'Accept-Encoding': 'gzip, deflate, br',
  36             'Accept-Language': 'en,zh-CN;q=0.9,zh;q=0.8',
  37             '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',
  38             'Upgrade-Insecure-Requests': '1',
  39         }
  40         # to get Referer url for genvisitor
  41         webpage, urlh = self._download_webpage_handle(url, video_id, headers=headers, note="first visit the page")
  42
  43         visitor_url = urlh.geturl()
  44
  45         data = urlencode({
  46             "cb": "gen_callback",
  47             "fp": '{"os":"2","browser":"Gecko57,0,0,0","fonts":"undefined","screenInfo":"1440*900*24","plugins":""}',
  48         }).encode()
  49         headers = {
  50             'Accept-Encoding': 'gzip, deflate, br',
  51             'Accept': '*/*',
  52             'Referer': visitor_url,
  53         }
  54
  55         r_genvisitor = request.Request(
  56             'https://passport.weibo.com/visitor/genvisitor',
  57             data=data,
  58             headers=headers,
  59         )
  60         webpage, urlh = self._download_webpage_handle(r_genvisitor, video_id, note="gen visitor")
  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
  82         webpage, urlh = self._download_webpage_handle(url, video_id, headers=headers, note="retry to visit the page")
  83
  84         # TODO more code goes here, for example ...
  85         title = self._html_search_regex(r'<title>(.+?)</title>', webpage, 'title')
  86
  87         video_sources_text = self._search_regex(r'video-sources=\\\"(.+?)\"', webpage, 'video_sources')
  88
  89         video_formats = parse.parse_qs(video_sources_text)
  90
  91         formats = []
  92         supported_resolutions = ['720', '480']
  93         for res in supported_resolutions:
  94             f = video_formats.get(res)
  95             if isinstance(f, list):
  96                 if len(f) > 0:
  97                     vid_url = f[0]
  98                     formats.append({
  99                         'url': vid_url,
 100                         'format': 'mp4',
 101                         'height': int(res),
 102                     })
 103         self._sort_formats(formats)
 104         uploader = self._og_search_property('nick-name', webpage, 'uploader', default=None)
 105         return {
 106             'id': video_id,
 107             'title': title,
 108             'uploader': uploader,
 109             'formats': formats
 110             # TODO more properties (see youtube_dl/extractor/common.py)
 111         }
 112
 113
 114 class WeiboMobileIE(InfoExtractor):
 115     _VALID_URL = r'https?://m.weibo.cn/status/(?P<id>[0-9]+)(\?.+)?'
 116     _TEST = {
 117         'url': 'https://m.weibo.cn/status/4189191225395228?wm=3333_2001&sourcetype=weixin&featurecode=newtitle&from=singlemessage&isappinstalled=0',
 118         'info_dict': {
 119             'id': '4189191225395228',
 120             'ext': 'mp4',
 121             'title': '午睡当然是要甜甜蜜蜜的啦',
 122             'uploader': '柴犬柴犬'
 123         }
 124     }
 125
 126     def _real_extract(self, url):
 127         video_id = self._match_id(url)
 128         headers = {
 129             'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8',
 130             'Accept-Encoding': 'gzip, deflate, br',
 131             'Accept-Language': 'en,zh-CN;q=0.9,zh;q=0.8',
 132             'Upgrade-Insecure-Requests': '1',
 133         }
 134         # to get Referer url for genvisitor
 135         webpage, urlh = self._download_webpage_handle(url, video_id, headers=headers, note="visit the page")
 136         js_code = self._search_regex(r'var\s+\$render_data\s*=\s*\[({.*})\]\[0\] \|\| {};', webpage, 'js_code', flags=re.DOTALL)
 137         weibo_info = self._parse_json(js_code, video_id, transform_source=js_to_json)
 138         page_info = weibo_info['status']['page_info']
 139         title = weibo_info['status']['status_title']
 140         format = {
 141             'url': page_info['media_info']['stream_url'],
 142             'format': 'mp4',
 143         }
 144         formats = [format]
 145         uploader = weibo_info['status']['user']['screen_name']
 146
 147         return {
 148             'id': video_id,
 149             'title': title,
 150             'uploader': uploader,
 151             'formats': formats
 152             # TODO more properties (see youtube_dl/extractor/common.py)
 153         }