yt-dlp/youtube_dl/extractor/weibo.py

134 lines
4.3 KiB
Python
Raw Normal View History

# coding: utf-8
from __future__ import unicode_literals
from .common import InfoExtractor
import json
import random
2017-12-26 08:18:35 +00:00
import re
2017-12-26 08:28:47 +00:00
from ..compat import (
compat_urlparse,
2017-12-26 08:28:47 +00:00
)
2017-12-26 08:18:35 +00:00
from ..utils import (
js_to_json,
2017-12-29 17:11:30 +00:00
strip_jsonp,
urlencode_postdata,
2017-12-26 08:18:35 +00:00
)
2017-12-26 08:38:51 +00:00
class WeiboIE(InfoExtractor):
_VALID_URL = r'https?://weibo\.com/[0-9]+/(?P<id>[a-zA-Z0-9]+)'
_TEST = {
2017-12-26 08:38:51 +00:00
'url': 'https://weibo.com/6275294458/Fp6RGfbff?type=comment',
'info_dict': {
'id': 'Fp6RGfbff',
'ext': 'mp4',
'title': 'You should have servants to massage you,... 来自Hosico_猫 - 微博',
}
}
def _real_extract(self, url):
video_id = self._match_id(url)
# to get Referer url for genvisitor
webpage, urlh = self._download_webpage_handle(url, video_id, note="first visit the page")
visitor_url = urlh.geturl()
headers = {
'Referer': visitor_url
}
fp = {
"os": "2",
"browser": "Gecko57,0,0,0",
"fonts": "undefined",
"screenInfo": "1440*900*24",
"plugins": ""
}
data = urlencode_postdata({
"cb": "gen_callback",
"fp": json.dumps(fp),
})
genvisitor_url = 'https://passport.weibo.com/visitor/genvisitor'
webpage = self._download_webpage(genvisitor_url, video_id, data=data, headers=headers, note="gen visitor")
2017-12-29 17:11:30 +00:00
p = strip_jsonp(webpage)
i1 = p.find('{')
i2 = p.rfind('}')
2017-12-26 08:38:51 +00:00
j = p[i1:i2 + 1] # get JSON object
d = json.loads(j)
tid = d["data"]["tid"]
cnfd = "%03d" % d["data"]["confidence"]
query = {
'a': 'incarnate',
't': tid,
'w': 2,
'c': cnfd,
'cb': 'cross_domain',
'from': 'weibo',
'_rand': random.random()
}
gencallback_url = "https://passport.weibo.com/visitor/visitor"
self._download_webpage(gencallback_url, video_id, note="gen callback", query=query)
webpage = self._download_webpage(url, video_id, note="retry to visit the page")
title = self._html_search_regex(r'<title>(.+?)</title>', webpage, 'title')
video_sources_text = self._search_regex(r'video-sources=\\\"(.+?)\"', webpage, 'video_sources')
2017-12-26 08:38:51 +00:00
video_formats = compat_urlparse.parse_qs(video_sources_text)
2017-12-11 07:56:54 +00:00
formats = []
supported_resolutions = ('720', '480')
2017-12-11 07:56:54 +00:00
for res in supported_resolutions:
f = video_formats.get(res)
if isinstance(f, list):
if len(f) > 0:
vid_url = f[0]
formats.append({
'url': vid_url,
'format': 'mp4',
'height': int(res),
2017-12-26 08:38:51 +00:00
})
2017-12-11 07:56:54 +00:00
self._sort_formats(formats)
2017-12-26 08:38:51 +00:00
uploader = self._og_search_property('nick-name', webpage, 'uploader', default=None)
return {
2017-12-26 08:38:51 +00:00
'id': video_id,
'title': title,
'uploader': uploader,
'formats': formats
}
2017-12-26 08:18:35 +00:00
class WeiboMobileIE(InfoExtractor):
_VALID_URL = r'https?://m\.weibo\.cn/status/(?P<id>[0-9]+)(\?.+)?'
2017-12-26 08:18:35 +00:00
_TEST = {
2017-12-26 08:38:51 +00:00
'url': 'https://m.weibo.cn/status/4189191225395228?wm=3333_2001&sourcetype=weixin&featurecode=newtitle&from=singlemessage&isappinstalled=0',
'info_dict': {
'id': '4189191225395228',
'ext': 'mp4',
'title': '午睡当然是要甜甜蜜蜜的啦',
'uploader': '柴犬柴犬'
}
}
2017-12-26 08:18:35 +00:00
def _real_extract(self, url):
video_id = self._match_id(url)
# to get Referer url for genvisitor
webpage = self._download_webpage(url, video_id, note="visit the page")
2017-12-26 08:38:51 +00:00
js_code = self._search_regex(r'var\s+\$render_data\s*=\s*\[({.*})\]\[0\] \|\| {};', webpage, 'js_code', flags=re.DOTALL)
2017-12-26 08:18:35 +00:00
weibo_info = self._parse_json(js_code, video_id, transform_source=js_to_json)
page_info = weibo_info.get('status').get('page_info')
title = weibo_info.get('status').get('status_title')
uploader = weibo_info.get('status').get('user').get('screen_name')
2017-12-26 08:18:35 +00:00
return {
2017-12-26 08:38:51 +00:00
'id': video_id,
'title': title,
'uploader': uploader,
'url': page_info['media_info']['stream_url']
2017-12-26 08:38:51 +00:00
}