X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;ds=sidebyside;f=youtube_dl%2Fextractor%2Fopenload.py;h=fc7ff43a62ba4a1a294d368d0b8fbaeb9abdb2ed;hb=fd5c4aab5958a2a086072488913cc190ff028bc3;hp=71021d573f1fbe7af75b278a9ed519cb73f02517;hpb=2bfeee69b976fe049761dd3012e30b637ee05a58;p=youtube-dl

diff --git a/youtube_dl/extractor/openload.py b/youtube_dl/extractor/openload.py
index 71021d573..fc7ff43a6 100644
--- a/youtube_dl/extractor/openload.py
+++ b/youtube_dl/extractor/openload.py
@@ -5,88 +5,108 @@ import re
 
 from .common import InfoExtractor
 from ..compat import compat_chr
-from ..utils import encode_base_n
+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)|oload\.tv)/(?: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': 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',
+        '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,
+    }, {
+        '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) + '"'
+    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('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]'),
+        ol_id = self._search_regex(
+            '<span[^>]+id="[^"]+"[^>]*>([0-9A-Za-z]+)</span>',
+            webpage, 'openload ID')
 
-            ('o', '(ﾟДﾟ) [\'o\']'),
-            ('u', '(oﾟｰﾟo)'),
-            ('c', '(ﾟДﾟ) [\'c\']'),
+        first_char = int(ol_id[0])
+        urlcode = []
+        num = 1
 
-            ('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)
+        while num < len(ol_id):
+            i = ord(ol_id[num])
+            key = 0
+            if i <= 90:
+                key = i - 65
+            elif i >= 97:
+                key = 25 + i - 97
+            urlcode.append((key, compat_chr(int(ol_id[num + 2:num + 5]) // int(ol_id[num + 1]) - first_char)))
+            num += 5
 
-    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/' + ''.join(
+            [value for _, value in sorted(urlcode, key=lambda x: x[0])])
+
+        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)
 
-        video_url = self._search_regex(
-            r'return\s+"(https?://[^"]+)"', self.openload_decode(code), 'video URL')
+        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': 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, 'mp4'),
+            'subtitles': subtitles,
         }
+        return info_dict