Merge pull request #174 from ytdl-org/master

[pull] master from ytdl-org:master

test/test_utils.py

@@ -19,6 +19,7 @@ from youtube_dl.utils import (
     age_restricted,
     args_to_str,
     encode_base_n,
+    caesar,
     clean_html,
     date_from_str,
     DateRange,
@@ -69,6 +70,7 @@ from youtube_dl.utils import (
     remove_start,
     remove_end,
     remove_quotes,
+    rot47,
     shell_quote,
     smuggle_url,
     str_to_int,
@@ -1369,6 +1371,20 @@ Line 1
         self.assertRaises(ValueError, encode_base_n, 0, 70)
         self.assertRaises(ValueError, encode_base_n, 0, 60, custom_table)
 
+    def test_caesar(self):
+        self.assertEqual(caesar('ace', 'abcdef', 2), 'cea')
+        self.assertEqual(caesar('cea', 'abcdef', -2), 'ace')
+        self.assertEqual(caesar('ace', 'abcdef', -2), 'eac')
+        self.assertEqual(caesar('eac', 'abcdef', 2), 'ace')
+        self.assertEqual(caesar('ace', 'abcdef', 0), 'ace')
+        self.assertEqual(caesar('xyz', 'abcdef', 2), 'xyz')
+        self.assertEqual(caesar('abc', 'acegik', 2), 'ebg')
+        self.assertEqual(caesar('ebg', 'acegik', -2), 'abc')
+
+    def test_rot47(self):
+        self.assertEqual(rot47('youtube-dl'), r'J@FEF36\5=')
+        self.assertEqual(rot47('YOUTUBE-DL'), r'*~&%&qt\s{')
+
     def test_urshift(self):
         self.assertEqual(urshift(3, 1), 1)
         self.assertEqual(urshift(-3, 1), 2147483646)

youtube_dl/extractor/shared.py

@@ -1,13 +1,18 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
-from ..compat import compat_b64decode
+from ..compat import (
+    compat_b64decode,
+    compat_urllib_parse_unquote_plus,
+)
 from ..utils import (
     determine_ext,
     ExtractorError,
     int_or_none,
+    js_to_json,
     KNOWN_EXTENSIONS,
     parse_filesize,
+    rot47,
     url_or_none,
     urlencode_postdata,
 )
@@ -112,16 +117,22 @@ class VivoIE(SharedBaseIE):
             webpage, 'filesize', fatal=False))
 
     def _extract_video_url(self, webpage, video_id, url):
-        def decode_url(encoded_url):
+        def decode_url_old(encoded_url):
             return compat_b64decode(encoded_url).decode('utf-8')
 
-        stream_url = url_or_none(decode_url(self._search_regex(
+        stream_url = self._search_regex(
             r'data-stream\s*=\s*(["\'])(?P<url>(?:(?!\1).)+)\1', webpage,
-            'stream url', default=None, group='url')))
+            'stream url', default=None, group='url')
+        if stream_url:
+            stream_url = url_or_none(decode_url_old(stream_url))
         if stream_url:
             return stream_url
-        return self._parse_json(
+
+        def decode_url(encoded_url):
+            return rot47(compat_urllib_parse_unquote_plus(encoded_url))
+
+        return decode_url(self._parse_json(
             self._search_regex(
-                r'InitializeStream\s*\(\s*(["\'])(?P<url>(?:(?!\1).)+)\1',
-                webpage, 'stream', group='url'),
-            video_id, transform_source=decode_url)[0]
+                r'(?s)InitializeStream\s*\(\s*({.+?})\s*\)\s*;', webpage,
+                'stream'),
+            video_id, transform_source=js_to_json)['source'])

youtube_dl/utils.py

@@ -5383,6 +5383,19 @@ def decode_packed_codes(code):
         obfucasted_code)
 
 
+def caesar(s, alphabet, shift):
+    if shift == 0:
+        return s
+    l = len(alphabet)
+    return ''.join(
+        alphabet[(alphabet.index(c) + shift) % l] if c in alphabet else c
+        for c in s)
+
+
+def rot47(s):
+    return caesar(s, r'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
+
+
 def parse_m3u8_attributes(attrib):
     info = {}
     for (key, val) in re.findall(r'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib):