# coding: utf-8
-from __future__ import unicode_literals
-
-import re
+from __future__ import unicode_literals, division
from .common import InfoExtractor
-from ..compat import compat_chr
-from ..utils import encode_base_n
+from ..compat import (
+ compat_chr,
+ compat_ord,
+)
+from ..utils import (
+ determine_ext,
+ ExtractorError,
+)
class OpenloadIE(InfoExtractor):
- _VALID_URL = r'https://openload.co/f/(?P<id>[a-zA-Z0-9]+)'
+ _VALID_URL = r'https://openload.(?:co|io)/(?:f|embed)/(?P<id>[a-zA-Z0-9-_]+)'
- _TEST = {
+ _TESTS = [{
'url': 'https://openload.co/f/kUEfGclsU9o',
'md5': 'bf1c059b004ebc7a256f89408e65c36e',
'info_dict': {
'id': 'kUEfGclsU9o',
'ext': 'mp4',
'title': 'skyrim_no-audio_1080.mp4',
+ 'thumbnail': 're:^https?://.*\.jpg$',
},
- }
+ }, {
+ 'url': 'https://openload.co/embed/kUEfGclsU9o/skyrim_no-audio_1080.mp4',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://openload.io/f/ZAn6oz-VZGE/',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://openload.co/f/_-ztPaZtMhM/',
+ 'only_matching': True,
+ }, {
+ # unavailable via https://openload.co/f/Sxz5sADo82g/, different layout
+ # for title and ext
+ 'url': 'https://openload.co/embed/Sxz5sADo82g/',
+ 'only_matching': True,
+ }]
- @staticmethod
- def openload_level2_debase(m):
- radix, num = int(m.group(1)) + 27, int(m.group(2))
- return '"' + encode_base_n(num, radix) + '"'
+ def _real_extract(self, url):
+ video_id = self._match_id(url)
+ webpage = self._download_webpage('https://openload.co/embed/%s/' % video_id, video_id)
- @classmethod
- def openload_level2(cls, txt):
- # The function name is ǃ \u01c3
- # Using escaped unicode literals does not work in Python 3.2
- return re.sub(r'ǃ\((\d+),(\d+)\)', cls.openload_level2_debase, txt, re.UNICODE).replace('"+"', '')
+ if 'File not found' in webpage or 'deleted by the owner' in webpage:
+ raise ExtractorError('File not found', expected=True)
- # Openload uses a variant of aadecode
- # openload_decode and related functions are originally written by
- # vitas@matfyz.cz and released with public domain
- # See https://github.com/rg3/youtube-dl/issues/8489
- @classmethod
- def openload_decode(cls, txt):
- symbol_table = [
- ('_', '(゚Д゚) [゚Θ゚]'),
- ('a', '(゚Д゚) [゚ω゚ノ]'),
- ('b', '(゚Д゚) [゚Θ゚ノ]'),
- ('c', '(゚Д゚) [\'c\']'),
- ('d', '(゚Д゚) [゚ー゚ノ]'),
- ('e', '(゚Д゚) [゚Д゚ノ]'),
- ('f', '(゚Д゚) [1]'),
+ # The following decryption algorithm is written by @yokrysty and
+ # declared to be freely used in youtube-dl
+ # See https://github.com/rg3/youtube-dl/issues/10408
+ enc_data = self._html_search_regex(
+ r'<span[^>]+id="hiddenurl"[^>]*>([^<]+)</span>', webpage, 'encrypted data')
- ('o', '(゚Д゚) [\'o\']'),
- ('u', '(o゚ー゚o)'),
- ('c', '(゚Д゚) [\'c\']'),
+ video_url_chars = []
- ('7', '((゚ー゚) + (o^_^o))'),
- ('6', '((o^_^o) +(o^_^o) +(c^_^o))'),
- ('5', '((゚ー゚) + (゚Θ゚))'),
- ('4', '(-~3)'),
- ('3', '(-~-~1)'),
- ('2', '(-~1)'),
- ('1', '(-~0)'),
- ('0', '((c^_^o)-(c^_^o))'),
- ]
- delim = '(゚Д゚)[゚ε゚]+'
- ret = ''
- for aachar in txt.split(delim):
- for val, pat in symbol_table:
- aachar = aachar.replace(pat, val)
- aachar = aachar.replace('+ ', '')
- m = re.match(r'^\d+', aachar)
- if m:
- ret += compat_chr(int(m.group(0), 8))
- else:
- m = re.match(r'^u([\da-f]+)', aachar)
- if m:
- ret += compat_chr(int(m.group(1), 16))
- return cls.openload_level2(ret)
+ for idx, c in enumerate(enc_data):
+ j = compat_ord(c)
+ if j >= 33 and j <= 126:
+ j = ((j + 14) % 94) + 33
+ if idx == len(enc_data) - 1:
+ j += 3
+ video_url_chars += compat_chr(j)
- def _real_extract(self, url):
- video_id = self._match_id(url)
- webpage = self._download_webpage(url, video_id)
- code = self._search_regex(
- r'<video[^>]+>\s*<script[^>]+>([^<]+)</script>',
- webpage, 'JS code')
+ video_url = 'https://openload.co/stream/%s?mime=true' % ''.join(video_url_chars)
- video_url = self._search_regex(
- r'return\s+"(https?://[^"]+)"', self.openload_decode(code), 'video URL')
+ title = self._og_search_title(webpage, default=None) or self._search_regex(
+ r'<span[^>]+class=["\']title["\'][^>]*>([^<]+)', webpage,
+ 'title', default=None) or self._html_search_meta(
+ 'description', webpage, 'title', fatal=True)
return {
'id': video_id,
- 'title': self._og_search_title(webpage),
+ 'title': title,
+ 'thumbnail': self._og_search_thumbnail(webpage, default=None),
'url': video_url,
+ # Seems all videos have extensions in their titles
+ 'ext': determine_ext(title),
}