[openload] Fix extraction (closes #10408)

This commit is contained in:
Yen Chi Hsuan 2016-12-09 02:15:16 +08:00
parent f43795e56b
commit 6c20a0bb99
No known key found for this signature in database
GPG Key ID: 7F902A182457CA23
2 changed files with 14 additions and 77 deletions

View File

@ -1,6 +1,7 @@
version <unreleased> version <unreleased>
Extractors Extractors
* [openload] Fix extraction (#10408)
+ [thisoldhouse] Recognize /tv-episode/ URLs (#11271) + [thisoldhouse] Recognize /tv-episode/ URLs (#11271)
version 2016.12.01 version 2016.12.01

View File

@ -1,21 +1,12 @@
# coding: utf-8 # coding: utf-8
from __future__ import unicode_literals, division from __future__ import unicode_literals
import re
from .common import InfoExtractor from .common import InfoExtractor
from ..compat import ( from ..compat import compat_chr
compat_chr,
compat_ord,
)
from ..utils import ( from ..utils import (
determine_ext, determine_ext,
ExtractorError, ExtractorError,
) )
from ..jsinterp import (
JSInterpreter,
_NAME_RE
)
class OpenloadIE(InfoExtractor): class OpenloadIE(InfoExtractor):
@ -62,44 +53,6 @@ class OpenloadIE(InfoExtractor):
'only_matching': True, 'only_matching': True,
}] }]
def openload_decode(self, txt):
symbol_dict = {
'(゚Д゚) [゚Θ゚]': '_',
'(゚Д゚) [゚ω゚ノ]': 'a',
'(゚Д゚) [゚Θ゚ノ]': 'b',
'(゚Д゚) [\'c\']': 'c',
'(゚Д゚) [゚ー゚ノ]': 'd',
'(゚Д゚) [゚Д゚ノ]': 'e',
'(゚Д゚) [1]': 'f',
'(゚Д゚) [\'o\']': 'o',
'(o゚ー゚o)': 'u',
'(゚Д゚) [\'c\']': 'c',
'((゚ー゚) + (o^_^o))': '7',
'((o^_^o) +(o^_^o) +(c^_^o))': '6',
'((゚ー゚) + (゚Θ゚))': '5',
'(-~3)': '4',
'(-~-~1)': '3',
'(-~1)': '2',
'(-~0)': '1',
'((c^_^o)-(c^_^o))': '0',
}
delim = '(゚Д゚)[゚ε゚]+'
end_token = '(゚Д゚)[゚o゚]'
symbols = '|'.join(map(re.escape, symbol_dict.keys()))
txt = re.sub('(%s)\+\s?' % symbols, lambda m: symbol_dict[m.group(1)], txt)
ret = ''
for aacode in re.findall(r'{0}\+\s?{1}(.*?){0}'.format(re.escape(end_token), re.escape(delim)), txt):
for aachar in aacode.split(delim):
if aachar.isdigit():
ret += compat_chr(int(aachar, 8))
else:
m = re.match(r'^u([\da-f]{4})$', aachar)
if m:
ret += compat_chr(int(m.group(1), 16))
else:
self.report_warning("Cannot decode: %s" % aachar)
return ret
def _real_extract(self, url): def _real_extract(self, url):
video_id = self._match_id(url) video_id = self._match_id(url)
webpage = self._download_webpage('https://openload.co/embed/%s/' % video_id, video_id) webpage = self._download_webpage('https://openload.co/embed/%s/' % video_id, video_id)
@ -107,36 +60,20 @@ def _real_extract(self, url):
if 'File not found' in webpage or 'deleted by the owner' in webpage: if 'File not found' in webpage or 'deleted by the owner' in webpage:
raise ExtractorError('File not found', expected=True) raise ExtractorError('File not found', expected=True)
# The following decryption algorithm is written by @yokrysty and ol_id = self._search_regex(
# declared to be freely used in youtube-dl '<span[^>]+id="[a-zA-Z0-9]+x"[^>]*>([0-9]+)</span>',
# See https://github.com/rg3/youtube-dl/issues/10408 webpage, 'openload ID')
enc_data = self._html_search_regex(
r'<span[^>]*>([^<]+)</span>\s*<span[^>]*>[^<]+</span>\s*<span[^>]+id="streamurl"',
webpage, 'encrypted data')
enc_code = self._html_search_regex(r'<script[^>]+>(゚ω゚[^<]+)</script>', first_two_chars = int(float(ol_id[0:][:2]))
webpage, 'encrypted code') urlcode = ''
num = 2
js_code = self.openload_decode(enc_code) while num < len(ol_id):
jsi = JSInterpreter(js_code) urlcode += compat_chr(int(float(ol_id[num:][:3])) -
first_two_chars * int(float(ol_id[num + 3:][:2])))
num += 5
m_offset_fun = self._search_regex(r'slice\(0\s*-\s*(%s)\(\)' % _NAME_RE, js_code, 'javascript offset function') video_url = 'https://openload.co/stream/' + urlcode
m_diff_fun = self._search_regex(r'charCodeAt\(0\)\s*\+\s*(%s)\(\)' % _NAME_RE, js_code, 'javascript diff function')
offset = jsi.call_function(m_offset_fun)
diff = jsi.call_function(m_diff_fun)
video_url_chars = []
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) - offset:
j += diff
video_url_chars += compat_chr(j)
video_url = 'https://openload.co/stream/%s?mime=true' % ''.join(video_url_chars)
title = self._og_search_title(webpage, default=None) or self._search_regex( title = self._og_search_title(webpage, default=None) or self._search_regex(
r'<span[^>]+class=["\']title["\'][^>]*>([^<]+)', webpage, r'<span[^>]+class=["\']title["\'][^>]*>([^<]+)', webpage,
@ -155,5 +92,4 @@ def _real_extract(self, url):
'ext': determine_ext(title), 'ext': determine_ext(title),
'subtitles': subtitles, 'subtitles': subtitles,
} }
return info_dict return info_dict