[openload] Add new extractor (closes #8489)

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index b773edb3d969a6b36e015a9059fb3f44567c64ea..ee792bbe0f1d5a565921b94f973861cac46c0f41 100644 (file)
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -535,6 +535,7 @@ from .ooyala import (
     OoyalaIE,
     OoyalaExternalIE,
 )
+from .openload import OpenloadIE
 from .ora import OraTVIE
 from .orf import (
     ORFTVthekIE,

diff --git a/youtube_dl/extractor/openload.py b/youtube_dl/extractor/openload.py
new file mode 100644 (file)
index 0000000..71021d5
--- /dev/null
+++ b/youtube_dl/extractor/openload.py
@@ -0,0 +1,92 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..compat import compat_chr
+from ..utils import encode_base_n
+
+
+class OpenloadIE(InfoExtractor):
+    _VALID_URL = r'https://openload.co/f/(?P<id>[a-zA-Z0-9]+)'
+
+    _TEST = {
+        'url': 'https://openload.co/f/kUEfGclsU9o',
+        'md5': 'bf1c059b004ebc7a256f89408e65c36e',
+        'info_dict': {
+            'id': 'kUEfGclsU9o',
+            'ext': 'mp4',
+            'title': 'skyrim_no-audio_1080.mp4',
+        },
+    }
+
+    @staticmethod
+    def openload_level2_debase(m):
+        radix, num = int(m.group(1)) + 27, int(m.group(2))
+        return '"' + encode_base_n(num, radix) + '"'
+
+    @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('"+"', '')
+
+    # 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]'),
+
+            ('o', '(ﾟДﾟ) [\'o\']'),
+            ('u', '(oﾟｰﾟo)'),
+            ('c', '(ﾟДﾟ) [\'c\']'),
+
+            ('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)
+
+    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 = self._search_regex(
+            r'return\s+"(https?://[^"]+)"', self.openload_decode(code), 'video URL')
+
+        return {
+            'id': video_id,
+            'title': self._og_search_title(webpage),
+            'url': video_url,
+        }