Merge tag 'upstream/2017.02.07'

[youtubedl] / youtube_dl / extractor / openload.py

diff --git a/youtube_dl/extractor/openload.py b/youtube_dl/extractor/openload.py
index 6415b8fdcb7451e6b499d1d4d3515af2775c66c2..32289d8976dcf602839546d179c06ef224a79b20 100644 (file)
--- a/youtube_dl/extractor/openload.py
+++ b/youtube_dl/extractor/openload.py
@@ -7,14 +7,12 @@ from .common import InfoExtractor
 from ..compat import compat_chr
 from ..utils import (
     determine_ext,
-    encode_base_n,
     ExtractorError,
-    mimetype2ext,
 )
 
 
 class OpenloadIE(InfoExtractor):
-    _VALID_URL = r'https://openload.(?:co|io)/(?:f|embed)/(?P<id>[a-zA-Z0-9-_]+)'
+    _VALID_URL = r'https?://(?:openload\.(?:co|io)|oload\.tv)/(?:f|embed)/(?P<id>[a-zA-Z0-9-_]+)'
 
     _TESTS = [{
         'url': 'https://openload.co/f/kUEfGclsU9o',
@@ -23,7 +21,23 @@ class OpenloadIE(InfoExtractor):
             'id': 'kUEfGclsU9o',
             'ext': 'mp4',
             'title': 'skyrim_no-audio_1080.mp4',
-            'thumbnail': 're:^https?://.*\.jpg$',
+            'thumbnail': r're:^https?://.*\.jpg$',
+        },
+    }, {
+        'url': 'https://openload.co/embed/rjC09fkPLYs',
+        'info_dict': {
+            'id': 'rjC09fkPLYs',
+            'ext': 'mp4',
+            'title': 'movie.mp4',
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'subtitles': {
+                'en': [{
+                    'ext': 'vtt',
+                }],
+            },
+        },
+        'params': {
+            'skip_download': True,  # test subtitles only
         },
     }, {
         'url': 'https://openload.co/embed/kUEfGclsU9o/skyrim_no-audio_1080.mp4',
@@ -39,92 +53,55 @@ class OpenloadIE(InfoExtractor):
         # for title and ext
         'url': 'https://openload.co/embed/Sxz5sADo82g/',
         'only_matching': True,
+    }, {
+        'url': 'https://oload.tv/embed/KnG-kKZdcfY/',
+        '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) + '"'
-
-    @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 _extract_urls(webpage):
+        return re.findall(
+            r'<iframe[^>]+src=["\']((?:https?://)?(?:openload\.(?:co|io)|oload\.tv)/embed/[a-zA-Z0-9-_]+)',
+            webpage)
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
+        webpage = self._download_webpage('https://openload.co/embed/%s/' % video_id, video_id)
 
-        if 'File not found' in webpage:
+        if 'File not found' in webpage or 'deleted by the owner' in webpage:
             raise ExtractorError('File not found', expected=True)
 
-        code = self._search_regex(
-            r'</video>\s*</div>\s*<script[^>]+>[^>]+</script>\s*<script[^>]+>([^<]+)</script>',
-            webpage, 'JS code')
+        ol_id = self._search_regex(
+            '<span[^>]+id="[^"]+"[^>]*>([0-9]+)</span>',
+            webpage, 'openload ID')
+
+        first_three_chars = int(float(ol_id[0:][:3]))
+        fifth_char = int(float(ol_id[3:5]))
+        urlcode = ''
+        num = 5
 
-        decoded = self.openload_decode(code)
+        while num < len(ol_id):
+            urlcode += compat_chr(int(float(ol_id[num:][:3])) +
+                                  first_three_chars - fifth_char * int(float(ol_id[num + 3:][:2])))
+            num += 5
 
-        video_url = self._search_regex(
-            r'return\s+"(https?://[^"]+)"', decoded, 'video URL')
+        video_url = 'https://openload.co/stream/' + urlcode
 
         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)
 
-        ext = mimetype2ext(self._search_regex(
-            r'window\.vt\s*=\s*(["\'])(?P<mimetype>.+?)\1', decoded,
-            'mimetype', default=None, group='mimetype')) or determine_ext(
-            video_url, 'mp4')
+        entries = self._parse_html5_media_entries(url, webpage, video_id)
+        subtitles = entries[0]['subtitles'] if entries else None
 
-        return {
+        info_dict = {
             'id': video_id,
             'title': title,
-            'ext': ext,
             'thumbnail': self._og_search_thumbnail(webpage, default=None),
             'url': video_url,
+            # Seems all videos have extensions in their titles
+            'ext': determine_ext(title, 'mp4'),
+            'subtitles': subtitles,
         }
+        return info_dict