|
|
- # coding: utf-8
- from __future__ import unicode_literals
-
- import re
-
- from .common import InfoExtractor
- from ..utils import (
- ExtractorError,
- HEADRequest,
- )
-
-
- class AparatIE(InfoExtractor):
- _VALID_URL = r'^https?://(?:www\.)?aparat\.com/(?:v/|video/video/embed/videohash/)(?P<id>[a-zA-Z0-9]+)'
-
- _TEST = {
- 'url': 'http://www.aparat.com/v/wP8On',
- 'md5': '6714e0af7e0d875c5a39c4dc4ab46ad1',
- 'info_dict': {
- 'id': 'wP8On',
- 'ext': 'mp4',
- 'title': 'تیم گلکسی 11 - زومیت',
- 'age_limit': 0,
- },
- # 'skip': 'Extremely unreliable',
- }
-
- def _real_extract(self, url):
- video_id = self._match_id(url)
-
- # Note: There is an easier-to-parse configuration at
- # http://www.aparat.com/video/video/config/videohash/%video_id
- # but the URL in there does not work
- embed_url = ('http://www.aparat.com/video/video/embed/videohash/' +
- video_id + '/vt/frame')
- webpage = self._download_webpage(embed_url, video_id)
-
- video_urls = [video_url.replace('\\/', '/') for video_url in re.findall(
- r'(?:fileList\[[0-9]+\]\s*=|"file"\s*:)\s*"([^"]+)"', webpage)]
- for i, video_url in enumerate(video_urls):
- req = HEADRequest(video_url)
- res = self._request_webpage(
- req, video_id, note='Testing video URL %d' % i, errnote=False)
- if res:
- break
- else:
- raise ExtractorError('No working video URLs found')
-
- title = self._search_regex(r'\s+title:\s*"([^"]+)"', webpage, 'title')
- thumbnail = self._search_regex(
- r'image:\s*"([^"]+)"', webpage, 'thumbnail', fatal=False)
-
- return {
- 'id': video_id,
- 'title': title,
- 'url': video_url,
- 'ext': 'mp4',
- 'thumbnail': thumbnail,
- 'age_limit': self._family_friendly_search(webpage),
- }
|