etc
/
yt-dlp
mirror of https://github.com/yt-dlp/yt-dlp.git
0
0
Fork 0
Code Issues Projects Releases Wiki Activity

[Arnes] Add extractor (added support for private videos)

This commit is contained in:
JSubelj 2023-09-22 13:36:55 +02:00
parent 3eb1503ce5
commit 1a66bcbfba
1 changed files with 87 additions and 57 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

144
yt_dlp/extractor/arnes.py
View File

@ -1,20 +1,25 @@
import binascii
import hashlib
import json
import os
import random
import time
from datetime import datetime
from .common import InfoExtractor
from ..compat import (
compat_parse_qs,
compat_urllib_parse_urlparse,
)
from ..dependencies import Cryptodome
from ..utils import (
format_field,
float_or_none,
int_or_none,
parse_iso8601,
remove_start,
ExtractorError
)
from ..dependencies import Cryptodome
from ..networking import Request
import time, hashlib, random, os, json, binascii
from datetime import datetime
class ArnesIE(InfoExtractor):
IE_NAME = 'video.arnes.si'
@ -39,6 +44,29 @@ class ArnesIE(InfoExtractor):
'view_count': int,
'tags': ['linearna_algebra'],
'start_time': 10,
'thumbnail': 'https://video.arnes.si/attachments/video/a1/a1qrWTOQfVoU/transcoded/image/ZfOtoJ9CVl7AKcZIpK3jTEvB.480p.jpg'
}
}, {
# one needs to be registered to get this url on: https://bsf.si/sl/film/dan-ljubezni-epizoda-10/
'url': 'https://video-4.arnes.si/embed/4zdg767pl9sv?accessToken=ZONPEE6G6ZM36WYPAIC5M2BL&hideRelated=1',
'only_matching': True,
'md5': '84c1e19403a4e50fa2394451675563f3',
'info_dict': {
'id': '4zdg767pl9sv',
'ext': 'mp4',
'title': 'Dan ljubezni E10 - nova',
'thumbnail': 'https://video-4.arnes.si/attachments/video/4z/4zdg767pl9sv/transcoded/image/Q1gZUVXgDgZZ.1nvm5msy0xkg.480p.jpg',
'description': 'film',
'license': 'CC_BY_NC_SA',
'creator': 'Video Filmoteka',
'timestamp': 1681422237,
'channel': 'Video Filmoteka',
'channel_id': 'd6rb30mw368k',
'channel_url': 'https://video-4.arnes.si/?channel=d6rb30mw368k',
'duration': 651.6,
'view_count': 44,
'upload_date': '20230413',
'tags': []
}
}, {
'url': 'https://video.arnes.si/api/asset/s1YjnV7hadlC/play.mp4',
@ -57,86 +85,87 @@ class ArnesIE(InfoExtractor):
'only_matching': True,
}]
_PARSED_URL=None
_ACCESS_TOKEN=None
_PARSED_URL = None
_ACCESS_TOKEN = None
_TIME_OF_LAST_VID_ACCESS_COOKIE_CHANGED = 0
# 10s was found when watching how requests are send and when cookie changes
_TIME_BETWEEN_NEW_COOKIE = 10
def _generate_video_access_cookie(self):
key_text = "Xmw1MSIlpZYdyy1DlqIl"
key = hashlib.sha256(key_text.encode("utf-8")).digest()
# Text of key that is then sha256ed gathered from main-es2015.62cf37eb23d6552e6e54.js
# before video_access_cookie is added
key_text = 'Xmw1MSIlpZYdyy1DlqIl'
key = hashlib.sha256(key_text.encode('utf-8')).digest()
iv = os.urandom(12)
# iv = bytearray([177, 70, 39, 213, 109, 49, 245, 24, 92, 85, 25, 187])
time_component = int(datetime.now().timestamp()*1000)+random.randint(0,1000)
# time_component = 1695271197003
time_component = int(datetime.now().timestamp() * 1000) + random.randint(0, 1000)
token_data = f"{self._ACCESS_TOKEN}|{time_component}"
token_data = f'{self._ACCESS_TOKEN}|{time_component}'
encoded_token = token_data.encode('utf-8')
encrypted_data, tag = Cryptodome.AES.new(key, Cryptodome.AES.MODE_GCM, iv).encrypt_and_digest(encoded_token)
base16encrypted_data = binascii.hexlify(iv).decode() + binascii.hexlify(encrypted_data+tag).decode()
base16encrypted_data = binascii.hexlify(iv).decode() + binascii.hexlify(encrypted_data + tag).decode()
return base16encrypted_data
def _generate_video_test_cookie(self):
# Generate a random timestamp
current_time = str(int(time.time()) + random.randint(1, 10000))
return current_time
def _set_access_cookie(self):
vid_access_cookie=self._generate_video_access_cookie()
self._set_cookie(self._PARSED_URL.netloc,"video_access_cookie",vid_access_cookie)
"""
Generates a new video access cookie and sets it.
This function should be executed between each downloaded segment.
"""
if self._ACCESS_TOKEN is None:
return
current_time = time.time()
if current_time - self._TIME_OF_LAST_VID_ACCESS_COOKIE_CHANGED > self._TIME_BETWEEN_NEW_COOKIE:
vid_access_cookie = self._generate_video_access_cookie()
self._set_cookie(self._PARSED_URL.netloc, 'video_access_cookie', vid_access_cookie)
self._TIME_OF_LAST_VID_ACCESS_COOKIE_CHANGED = current_time
self.write_debug("New video_access_cookie generated.")
def _fragment_hook_before_download(self, frag_filename, frag_info_dict, ctx):
self._set_access_cookie()
def _real_extract(self, url):
video_id = self._match_id(url)
self._PARSED_URL = compat_urllib_parse_urlparse(url)
at_list = compat_parse_qs(self._PARSED_URL.query).get('accessToken')
self._ACCESS_TOKEN = at_list[0] if at_list is not None and len(at_list) != 0 else None
base_url = f"{self._PARSED_URL.scheme}://{self._PARSED_URL.netloc}"
base_url = f'{self._PARSED_URL.scheme}://{self._PARSED_URL.netloc}'
access_token_query = f'?accessToken={self._ACCESS_TOKEN}' if self._ACCESS_TOKEN is not None else '?'
self._ACCESS_TOKEN = compat_parse_qs(self._PARSED_URL.query).get("accessToken")[0]
video_res = self._downloader.urlopen(base_url + '/api/public/video/' + video_id + f"?accessToken={self._ACCESS_TOKEN}")
header_date = video_res.headers["Date"]
date_object = datetime.strptime(header_date, '%a, %d %b %Y %H:%M:%S %Z')
timestamp = int(date_object.timestamp()*1000)
vid_access_cookie = self._generate_video_access_cookie()
video_res = self._downloader.urlopen(f'{base_url}/api/public/video/{video_id}{access_token_query}')
video_res = json.loads(video_res.read().decode())
#self._download_json(base_url + '/api/public/video/' + video_id + f"?accessToken={self._ACCESS_TOKEN}", video_id)
video = video_res['data']
vid_test_cookie = self._generate_video_test_cookie()
req_headers = {'Cookie': f'video_test_cookie={vid_test_cookie}; video_access_cookie={vid_access_cookie}'}
check_cookies_request = Request(base_url + f'/api/public/video/embed-support/{vid_test_cookie}',
headers = req_headers)
check_cookies_data = self._download_json(check_cookies_request, video_id)
if check_cookies_data['status'] != 'OK' or check_cookies_data['data'] != True:
raise ExtractorError('Check cookies request did not succeed.')
title = video['title']
formats = []
for media in (video.get('media') or []):
media_url = media.get('url')
if not media_url:
continue
formats.append({
'url': base_url + media_url,
'format_id': remove_start(media.get('format'), 'FORMAT_'),
'format_note': media.get('formatTranslation'),
'width': int_or_none(media.get('width')),
'height': int_or_none(media.get('height')),
})
channel = video.get('channel') or {}
channel_id = channel.get('url')
thumbnail = video.get('thumbnailUrl')
m3u8_url = base_url+video.get('hls').get('url') + f'?accessToken={self._ACCESS_TOKEN}'
self._set_cookie(self._PARSED_URL.netloc,"video_access_cookie",vid_access_cookie)
formats = self._extract_m3u8_formats(m3u8_url,video_id)
formats = []
if video.get('hls') is None:
for media in (video.get('media') or []):
media_url = media.get('url')
if not media_url:
continue
formats.append({
'url': base_url + media_url + access_token_query,
'format_id': remove_start(media.get('format'), 'FORMAT_'),
'format_note': media.get('formatTranslation'),
'width': int_or_none(media.get('width')),
'height': int_or_none(media.get('height')),
})
else:
m3u8_url = f'{base_url}{video.get("hls").get("url")}{access_token_query}'
self._set_access_cookie()
formats = self._extract_m3u8_formats(m3u8_url, video_id)
return {
'id': video_id,
'title': title,
@ -155,5 +184,6 @@ class ArnesIE(InfoExtractor):
'tags': video.get('hashtags'),
'start_time': int_or_none(compat_parse_qs(
compat_urllib_parse_urlparse(url).query).get('t', [None])[0]),
'_execute_before_each_fragment': self._set_access_cookie
'_fragment_hook_before_dl': self._fragment_hook_before_download,
'_test': self._fragment_hook_before_download
}