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
   8 import re
   9
  10 from ..compat import (
  11     compat_urllib_parse_urlencode,
  12     compat_urlparse,
  13 )
  14 from ..utils import (
  15     js_to_json,
  16 )
  17
  18
  19 class WeiboIE(InfoExtractor):
  20     _VALID_URL = r'https?://weibo\.com/[0-9]+/(?P<id>[a-zA-Z0-9]+)'
  21     _TEST = {
  22         'url': 'https://weibo.com/6275294458/Fp6RGfbff?type=comment',
  23         'info_dict': {
  24             'id': 'Fp6RGfbff',
  25             'ext': 'mp4',
  26             'title': 'You should have servants to massage you,... 来自Hosico_猫 - 微博',
  27         }
  28     }
  29
  30     def _real_extract(self, url):
  31         video_id = self._match_id(url)
  32         # to get Referer url for genvisitor
  33         webpage, urlh = self._download_webpage_handle(url, video_id, note="first visit the page")
  34
  35         visitor_url = urlh.geturl()
  36         headers = {
  37             'Referer': visitor_url
  38         }
  39
  40         fp = {
  41             "os": "2",
  42             "browser": "Gecko57,0,0,0",
  43             "fonts": "undefined",
  44             "screenInfo": "1440*900*24",
  45             "plugins": ""
  46         }
  47         data = compat_urllib_parse_urlencode({
  48             "cb": "gen_callback",
  49             "fp": json.dumps(fp),
  50         }).encode()
  51
  52         genvisitor_url = 'https://passport.weibo.com/visitor/genvisitor'
  53         webpage, _ = self._download_webpage_handle(genvisitor_url, video_id, data=data, headers=headers, note="gen visitor")
  54
  55         p = webpage.split("&&")[1]  # split "gen_callback && gen_callback(...)"
  56         i1 = p.find('{')
  57         i2 = p.rfind('}')
  58         j = p[i1:i2 + 1]  # get JSON object
  59         d = json.loads(j)
  60         tid = d["data"]["tid"]
  61         cnfd = "%03d" % d["data"]["confidence"]
  62
  63         query = {
  64             'a': 'incarnate',
  65             't': tid,
  66             'w': 2,
  67             'c': cnfd,
  68             'cb': 'cross_domain',
  69             'from': 'weibo',
  70             '_rand': random.random()
  71         }
  72         gencallback_url = "https://passport.weibo.com/visitor/visitor"
  73         self._download_webpage_handle(gencallback_url, video_id, note="gen callback", query=query)
  74
  75         webpage, _ = self._download_webpage_handle(url, video_id, note="retry to visit the page")
  76
  77         title = self._html_search_regex(r'<title>(.+?)</title>', webpage, 'title')
  78
  79         video_sources_text = self._search_regex(r'video-sources=\\\"(.+?)\"', webpage, 'video_sources')
  80
  81         video_formats = compat_urlparse.parse_qs(video_sources_text)
  82
  83         formats = []
  84         supported_resolutions = ('720', '480')
  85         for res in supported_resolutions:
  86             f = video_formats.get(res)
  87             if isinstance(f, list):
  88                 if len(f) > 0:
  89                     vid_url = f[0]
  90                     formats.append({
  91                         'url': vid_url,
  92                         'format': 'mp4',
  93                         'height': int(res),
  94                     })
  95         self._sort_formats(formats)
  96         uploader = self._og_search_property('nick-name', webpage, 'uploader', default=None)
  97         return {
  98             'id': video_id,
  99             'title': title,
 100             'uploader': uploader,
 101             'formats': formats
 102         }
 103
 104
 105 class WeiboMobileIE(InfoExtractor):
 106     _VALID_URL = r'https?://m\.weibo\.cn/status/(?P<id>[0-9]+)(\?.+)?'
 107     _TEST = {
 108         'url': 'https://m.weibo.cn/status/4189191225395228?wm=3333_2001&sourcetype=weixin&featurecode=newtitle&from=singlemessage&isappinstalled=0',
 109         'info_dict': {
 110             'id': '4189191225395228',
 111             'ext': 'mp4',
 112             'title': '午睡当然是要甜甜蜜蜜的啦',
 113             'uploader': '柴犬柴犬'
 114         }
 115     }
 116
 117     def _real_extract(self, url):
 118         video_id = self._match_id(url)
 119         # to get Referer url for genvisitor
 120         webpage, _ = self._download_webpage_handle(url, video_id, note="visit the page")
 121         js_code = self._search_regex(r'var\s+\$render_data\s*=\s*\[({.*})\]\[0\] \|\| {};', webpage, 'js_code', flags=re.DOTALL)
 122         weibo_info = self._parse_json(js_code, video_id, transform_source=js_to_json)
 123         page_info = weibo_info['status']['page_info']
 124         title = weibo_info.get('status').get('status_title')
 125         uploader = weibo_info.get('status').get('user').get('screen_name')
 126
 127         return {
 128             'id': video_id,
 129             'title': title,
 130             'uploader': uploader,
 131             'url': page_info['media_info']['stream_url']
 132         }