mirror of
https://github.com/ytdl-org/youtube-dl.git
synced 2025-08-22 22:34:01 -07:00
[coverapi] Add new extractor
This commit is contained in:
parent
211cbfd5d4
commit
f6805e9447
2 changed files with 78 additions and 0 deletions
77
youtube_dl/extractor/coverapi.py
Normal file
77
youtube_dl/extractor/coverapi.py
Normal file
|
@ -0,0 +1,77 @@
|
||||||
|
# coding: utf-8
|
||||||
|
from __future__ import unicode_literals
|
||||||
|
from .common import InfoExtractor
|
||||||
|
from ..compat import compat_urllib_parse_urlparse
|
||||||
|
from ..utils import (
|
||||||
|
sanitized_Request,
|
||||||
|
urlencode_postdata,
|
||||||
|
)
|
||||||
|
|
||||||
|
|
||||||
|
class CoverApiIE(InfoExtractor):
|
||||||
|
_VALID_URL = r'https?://(?:www\.)?coverapi\.store/embed/(?P<id>[0-9a-zA-Z]+)(?:/)?'
|
||||||
|
_TESTS = [{
|
||||||
|
'url': 'https://coverapi.store/embed/tt3915174/',
|
||||||
|
'md5': '0ed4bc62f04c48ab3f6760179ec87f27',
|
||||||
|
'info_dict': {
|
||||||
|
'id': 'tt3915174',
|
||||||
|
'title': 'Ο Παπουτσωμένος Γάτος: Η τελευταία επιθυμία / Puss in Boots: The Last Wish (2022)',
|
||||||
|
'ext': 'mp4'
|
||||||
|
}
|
||||||
|
}, {
|
||||||
|
'url': 'https://coverapi.store/embed/tt23177868',
|
||||||
|
'md5': '3040b5715b6d50103e4b4dc71e8901b4',
|
||||||
|
'info_dict': {
|
||||||
|
'id': 'tt23177868',
|
||||||
|
'title': 'GR Audio',
|
||||||
|
'ext': 'mp4'
|
||||||
|
}
|
||||||
|
}]
|
||||||
|
|
||||||
|
def _real_extract(self, url):
|
||||||
|
video_id = self._match_id(url)
|
||||||
|
html = self._download_webpage(url, video_id)
|
||||||
|
news_id = self._search_regex(r"news_id: '(\d+)'", html, 'news_id')
|
||||||
|
parsed_url = compat_urllib_parse_urlparse(url)
|
||||||
|
base_url = parsed_url.scheme + '://' + parsed_url.netloc
|
||||||
|
controller_path = self._search_regex(r"url:\s*'([^']*)'", html, 'controller_path')
|
||||||
|
controller_url = base_url + controller_path
|
||||||
|
controller_data = urlencode_postdata({'mod': 'players', 'news_id': news_id})
|
||||||
|
controller_request = sanitized_Request(controller_url, controller_data)
|
||||||
|
controller_request.add_header('Content-type', 'application/x-www-form-urlencoded, charset=UTF-8')
|
||||||
|
controller_response = self._download_json(controller_request, video_id)
|
||||||
|
html5 = controller_response.get('html5')
|
||||||
|
file_url = self._search_regex(r'file:"(.*?)"', html5, 'url')
|
||||||
|
if "playlists" in file_url:
|
||||||
|
playlist_path = file_url
|
||||||
|
playlist_url = base_url + playlist_path
|
||||||
|
playlist_request = sanitized_Request(playlist_url)
|
||||||
|
playlist_response = self._download_json(playlist_request, video_id)
|
||||||
|
playlist = playlist_response.get('playlist', [])
|
||||||
|
video_url = ''
|
||||||
|
video_title = ''
|
||||||
|
for element in playlist:
|
||||||
|
if 'audio' in element.get('comment').lower():
|
||||||
|
video_url = element.get('file')
|
||||||
|
video_title = element.get('comment')
|
||||||
|
break
|
||||||
|
elif 'subtitles' in element.get('comment').lower():
|
||||||
|
video_url = element.get('file')
|
||||||
|
video_title = element.get('comment')
|
||||||
|
break
|
||||||
|
if video_url == '' and playlist:
|
||||||
|
video_url = playlist[0].get('file')
|
||||||
|
return {
|
||||||
|
'id': video_id,
|
||||||
|
'url': video_url,
|
||||||
|
'title': video_title,
|
||||||
|
'ext': 'mp4'
|
||||||
|
}
|
||||||
|
else:
|
||||||
|
video_title = self._search_regex(r'title:\'(.*?)\'', html5, 'title')
|
||||||
|
return {
|
||||||
|
'id': video_id,
|
||||||
|
'url': file_url,
|
||||||
|
'title': video_title,
|
||||||
|
'ext': 'mp4'
|
||||||
|
}
|
|
@ -258,6 +258,7 @@ from .commonprotocols import (
|
||||||
from .condenast import CondeNastIE
|
from .condenast import CondeNastIE
|
||||||
from .contv import CONtvIE
|
from .contv import CONtvIE
|
||||||
from .corus import CorusIE
|
from .corus import CorusIE
|
||||||
|
from .coverapi import CoverApiIE
|
||||||
from .cpac import (
|
from .cpac import (
|
||||||
CPACIE,
|
CPACIE,
|
||||||
CPACPlaylistIE,
|
CPACPlaylistIE,
|
||||||
|
|
Loading…
Add table
Add a link
Reference in a new issue