[yandexdisk] fix extraction(closes #17861)(closes #27131)

This commit is contained in:
Remita Amine 2020-12-30 13:43:56 +01:00
parent bdd044e67b
commit bd18824c2a

View file

@ -1,19 +1,40 @@
# coding: utf-8
from __future__ import unicode_literals
import json
from .common import InfoExtractor
from ..compat import compat_str
from ..compat import compat_HTTPError
from ..utils import (
determine_ext,
ExtractorError,
float_or_none,
int_or_none,
try_get,
urlencode_postdata,
mimetype2ext,
parse_iso8601,
urljoin,
)
class YandexDiskIE(InfoExtractor):
_VALID_URL = r'https?://yadi\.sk/[di]/(?P<id>[^/?#&]+)'
_VALID_URL = r'''(?x)https?://
(?:
(?:www\.)?yadi\.sk|
disk\.yandex\.
(?:
az|
by|
co(?:m(?:\.(?:am|ge|tr))?|\.il)|
ee|
fr|
k[gz]|
l[tv]|
md|
t[jm]|
u[az]|
ru
)
)/(?:[di]/|public.*?\bhash=)(?P<id>[^/?#&]+)'''
_TESTS = [{
'url': 'https://yadi.sk/i/VdOeDou8eZs6Y',
@ -25,94 +46,97 @@ class YandexDiskIE(InfoExtractor):
'duration': 168.6,
'uploader': 'y.botova',
'uploader_id': '300043621',
'timestamp': 1421396809,
'upload_date': '20150116',
'view_count': int,
},
}, {
'url': 'https://yadi.sk/d/h3WAXvDS3Li3Ce',
'only_matching': True,
}, {
'url': 'https://yadi.sk/public?hash=5DZ296JK9GWCLp02f6jrObjnctjRxMs8L6%2B%2FuhNqk38%3D',
'only_matching': True,
}]
def _real_extract(self, url):
video_id = self._match_id(url)
status = self._download_webpage(
'https://disk.yandex.com/auth/status', video_id, query={
'urlOrigin': url,
'source': 'public',
'md5': 'false',
})
try:
resource = self._download_json(
'https://cloud-api.yandex.net/v1/disk/public/resources',
video_id, query={'public_key': url})
except ExtractorError as e:
if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
error_description = self._parse_json(
e.cause.read().decode(), video_id)['description']
raise ExtractorError(error_description, expected=True)
raise
sk = self._search_regex(
r'(["\'])sk(?:External)?\1\s*:\s*(["\'])(?P<value>(?:(?!\2).)+)\2',
status, 'sk', group='value')
title = resource['name']
public_url = resource.get('public_url')
if public_url:
video_id = self._match_id(public_url)
webpage = self._download_webpage(url, video_id)
self._set_cookie('yadi.sk', 'yandexuid', '0')
models = self._parse_json(
self._search_regex(
r'<script[^>]+id=["\']models-client[^>]+>\s*(\[.+?\])\s*</script',
webpage, 'video JSON'),
video_id)
data = next(
model['data'] for model in models
if model.get('model') == 'resource')
video_hash = data['id']
title = data['name']
models = self._download_json(
'https://disk.yandex.com/models/', video_id,
data=urlencode_postdata({
'_model.0': 'videoInfo',
'id.0': video_hash,
'_model.1': 'do-get-resource-url',
'id.1': video_hash,
'version': '13.6',
'sk': sk,
}), query={'_m': 'videoInfo'})['models']
videos = try_get(models, lambda x: x[0]['data']['videos'], list) or []
source_url = try_get(
models, lambda x: x[1]['data']['file'], compat_str)
def call_api(action):
return (self._download_json(
urljoin(url, '/public/api/') + action, video_id, data=json.dumps({
'hash': url,
# obtain sk if needed from call_api('check-auth') while
# the yandexuid cookie is set and sending an empty JSON object
'sk': 'ya6b52f8c6b12abe91a66d22d3a31084b'
}).encode(), headers={
'Content-Type': 'text/plain',
}, fatal=False) or {}).get('data') or {}
formats = []
source_url = resource.get('file')
if not source_url:
source_url = call_api('download-url').get('url')
if source_url:
formats.append({
'url': source_url,
'format_id': 'source',
'ext': determine_ext(title, 'mp4'),
'ext': determine_ext(title, mimetype2ext(resource.get('mime_type')) or 'mp4'),
'quality': 1,
'filesize': int_or_none(resource.get('size'))
})
for video in videos:
video_streams = call_api('get-video-streams')
for video in (video_streams.get('videos') or []):
format_url = video.get('url')
if not format_url:
continue
if determine_ext(format_url) == 'm3u8':
if video.get('dimension') == 'adaptive':
formats.extend(self._extract_m3u8_formats(
format_url, video_id, 'mp4', entry_protocol='m3u8_native',
format_url, video_id, 'mp4', 'm3u8_native',
m3u8_id='hls', fatal=False))
else:
size = video.get('size') or {}
height = int_or_none(size.get('height'))
format_id = 'hls'
if height:
format_id += '-%dp' % height
formats.append({
'ext': 'mp4',
'format_id': format_id,
'height': height,
'protocol': 'm3u8_native',
'url': format_url,
'width': int_or_none(size.get('width')),
})
self._sort_formats(formats)
duration = float_or_none(try_get(
models, lambda x: x[0]['data']['duration']), 1000)
uploader = try_get(
data, lambda x: x['user']['display_name'], compat_str)
uploader_id = try_get(
data, lambda x: x['user']['uid'], compat_str)
view_count = int_or_none(try_get(
data, lambda x: x['meta']['views_counter']))
owner = resource.get('owner') or {}
return {
'id': video_id,
'title': title,
'duration': duration,
'uploader': uploader,
'uploader_id': uploader_id,
'view_count': view_count,
'duration': float_or_none(video_streams.get('duration'), 1000),
'uploader': owner.get('display_name'),
'uploader_id': owner.get('uid'),
'view_count': int_or_none(resource.get('views_count')),
'timestamp': parse_iso8601(resource.get('created')),
'formats': formats,
}