[vidio] improve metadata extraction

This commit is contained in:
Remita Amine 2021-02-01 21:35:18 +01:00
parent 65eee5a745
commit 8fa7cc387d

View File

@ -4,7 +4,13 @@ from __future__ import unicode_literals
import re
from .common import InfoExtractor
from ..utils import int_or_none
from ..utils import (
int_or_none,
parse_iso8601,
str_or_none,
strip_or_none,
try_get,
)
class VidioIE(InfoExtractor):
@ -21,57 +27,63 @@ class VidioIE(InfoExtractor):
'thumbnail': r're:^https?://.*\.jpg$',
'duration': 149,
'like_count': int,
'uploader': 'TWELVE Pic',
'timestamp': 1444902800,
'upload_date': '20151015',
'uploader_id': 'twelvepictures',
'channel': 'Cover Music Video',
'channel_id': '280236',
'view_count': int,
'dislike_count': int,
'comment_count': int,
'tags': 'count:4',
},
}, {
'url': 'https://www.vidio.com/watch/77949-south-korea-test-fires-missile-that-can-strike-all-of-the-north',
'only_matching': True,
}]
def _real_initialize(self):
self._api_key = self._download_json(
'https://www.vidio.com/auth', None, data=b'')['api_key']
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
video_id, display_id = mobj.group('id', 'display_id')
video_id, display_id = re.match(self._VALID_URL, url).groups()
data = self._download_json(
'https://api.vidio.com/videos/' + video_id, display_id, headers={
'Content-Type': 'application/vnd.api+json',
'X-API-KEY': self._api_key,
})
video = data['videos'][0]
title = video['title'].strip()
webpage = self._download_webpage(url, display_id)
title = self._og_search_title(webpage)
m3u8_url, duration, thumbnail = [None] * 3
clips = self._parse_json(
self._html_search_regex(
r'data-json-clips\s*=\s*(["\'])(?P<data>\[.+?\])\1',
webpage, 'video data', default='[]', group='data'),
display_id, fatal=False)
if clips:
clip = clips[0]
m3u8_url = clip.get('sources', [{}])[0].get('file')
duration = clip.get('clip_duration')
thumbnail = clip.get('image')
m3u8_url = m3u8_url or self._search_regex(
r'data(?:-vjs)?-clip-hls-url=(["\'])(?P<url>(?:(?!\1).)+)\1',
webpage, 'hls url', group='url')
formats = self._extract_m3u8_formats(
m3u8_url, display_id, 'mp4', entry_protocol='m3u8_native')
data['clips'][0]['hls_url'], display_id, 'mp4', 'm3u8_native')
self._sort_formats(formats)
duration = int_or_none(duration or self._search_regex(
r'data-video-duration=(["\'])(?P<duration>\d+)\1', webpage,
'duration', fatal=False, group='duration'))
thumbnail = thumbnail or self._og_search_thumbnail(webpage)
like_count = int_or_none(self._search_regex(
(r'<span[^>]+data-comment-vote-count=["\'](\d+)',
r'<span[^>]+class=["\'].*?\blike(?:__|-)count\b.*?["\'][^>]*>\s*(\d+)'),
webpage, 'like count', fatal=False))
get_first = lambda x: try_get(data, lambda y: y[x + 's'][0], dict) or {}
channel = get_first('channel')
user = get_first('user')
username = user.get('username')
get_count = lambda x: int_or_none(video.get('total_' + x))
return {
'id': video_id,
'display_id': display_id,
'title': title,
'description': self._og_search_description(webpage),
'thumbnail': thumbnail,
'duration': duration,
'like_count': like_count,
'description': strip_or_none(video.get('description')),
'thumbnail': video.get('image_url_medium'),
'duration': int_or_none(video.get('duration')),
'like_count': get_count('likes'),
'formats': formats,
'uploader': user.get('name'),
'timestamp': parse_iso8601(video.get('created_at')),
'uploader_id': username,
'uploader_url': 'https://www.vidio.com/@' + username if username else None,
'channel': channel.get('name'),
'channel_id': str_or_none(channel.get('id')),
'view_count': get_count('view_count'),
'dislike_count': get_count('dislikes'),
'comment_count': get_count('comments'),
'tags': video.get('tag_list'),
}