]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/adn.py
debian/copyright: use spaces rather than tabs to start continuation lines.
[youtubedl] / youtube_dl / extractor / adn.py
index 66caf6a818278796648dd4496da95c661fcdb36c..c95ad2173522455d925b3248065ce914a35e3f0c 100644 (file)
@@ -2,19 +2,27 @@
 from __future__ import unicode_literals
 
 import base64
+import binascii
 import json
 import os
+import random
 
 from .common import InfoExtractor
 from ..aes import aes_cbc_decrypt
-from ..compat import compat_ord
+from ..compat import (
+    compat_b64decode,
+    compat_ord,
+)
 from ..utils import (
     bytes_to_intlist,
+    bytes_to_long,
     ExtractorError,
     float_or_none,
     intlist_to_bytes,
-    srt_subtitles_timecode,
+    long_to_bytes,
+    pkcs1pad,
     strip_or_none,
+    urljoin,
 )
 
 
@@ -31,48 +39,72 @@ class ADNIE(InfoExtractor):
             'description': 'md5:2f7b5aa76edbc1a7a92cedcda8a528d5',
         }
     }
+    _BASE_URL = 'http://animedigitalnetwork.fr'
+    _RSA_KEY = (0xc35ae1e4356b65a73b551493da94b8cb443491c0aa092a357a5aee57ffc14dda85326f42d716e539a34542a0d3f363adf16c5ec222d713d5997194030ee2e4f0d1fb328c01a81cf6868c090d50de8e169c6b13d1675b9eeed1cbc51e1fffca9b38af07f37abd790924cd3bee59d0257cfda4fe5f3f0534877e21ce5821447d1b, 65537)
+    _POS_ALIGN_MAP = {
+        'start': 1,
+        'end': 3,
+    }
+    _LINE_ALIGN_MAP = {
+        'middle': 8,
+        'end': 4,
+    }
+
+    @staticmethod
+    def _ass_subtitles_timecode(seconds):
+        return '%01d:%02d:%02d.%02d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 100)
 
     def _get_subtitles(self, sub_path, video_id):
         if not sub_path:
             return None
 
         enc_subtitles = self._download_webpage(
-            'http://animedigitalnetwork.fr/' + sub_path,
-            video_id, fatal=False)
+            urljoin(self._BASE_URL, sub_path),
+            video_id, 'Downloading subtitles location', fatal=False) or '{}'
+        subtitle_location = (self._parse_json(enc_subtitles, video_id, fatal=False) or {}).get('location')
+        if subtitle_location:
+            enc_subtitles = self._download_webpage(
+                urljoin(self._BASE_URL, subtitle_location),
+                video_id, 'Downloading subtitles data', fatal=False,
+                headers={'Origin': 'https://animedigitalnetwork.fr'})
         if not enc_subtitles:
             return None
 
         # http://animedigitalnetwork.fr/components/com_vodvideo/videojs/adn-vjs.min.js
         dec_subtitles = intlist_to_bytes(aes_cbc_decrypt(
-            bytes_to_intlist(base64.b64decode(enc_subtitles[24:])),
-            bytes_to_intlist(b'\nd\xaf\xd2J\xd0\xfc\xe1\xfc\xdf\xb61\xe8\xe1\xf0\xcc'),
-            bytes_to_intlist(base64.b64decode(enc_subtitles[:24]))
+            bytes_to_intlist(compat_b64decode(enc_subtitles[24:])),
+            bytes_to_intlist(binascii.unhexlify(self._K + '4b8ef13ec1872730')),
+            bytes_to_intlist(compat_b64decode(enc_subtitles[:24]))
         ))
         subtitles_json = self._parse_json(
-            dec_subtitles[:-compat_ord(dec_subtitles[-1])],
+            dec_subtitles[:-compat_ord(dec_subtitles[-1])].decode(),
             None, fatal=False)
         if not subtitles_json:
             return None
 
         subtitles = {}
         for sub_lang, sub in subtitles_json.items():
-            srt = ''
-            for num, current in enumerate(sub):
-                start, end, text = (
+            ssa = '''[Script Info]
+ScriptType:V4.00
+[V4 Styles]
+Format: Name,Fontname,Fontsize,PrimaryColour,SecondaryColour,TertiaryColour,BackColour,Bold,Italic,BorderStyle,Outline,Shadow,Alignment,MarginL,MarginR,MarginV,AlphaLevel,Encoding
+Style: Default,Arial,18,16777215,16777215,16777215,0,-1,0,1,1,0,2,20,20,20,0,0
+[Events]
+Format: Marked,Start,End,Style,Name,MarginL,MarginR,MarginV,Effect,Text'''
+            for current in sub:
+                start, end, text, line_align, position_align = (
                     float_or_none(current.get('startTime')),
                     float_or_none(current.get('endTime')),
-                    current.get('text'))
+                    current.get('text'), current.get('lineAlign'),
+                    current.get('positionAlign'))
                 if start is None or end is None or text is None:
                     continue
-                srt += os.linesep.join(
-                    (
-                        '%d' % num,
-                        '%s --> %s' % (
-                            srt_subtitles_timecode(start),
-                            srt_subtitles_timecode(end)),
-                        text,
-                        os.linesep,
-                    ))
+                alignment = self._POS_ALIGN_MAP.get(position_align, 2) + self._LINE_ALIGN_MAP.get(line_align, 0)
+                ssa += os.linesep + 'Dialogue: Marked=0,%s,%s,Default,,0,0,0,,%s%s' % (
+                    self._ass_subtitles_timecode(start),
+                    self._ass_subtitles_timecode(end),
+                    '{\\a%d}' % alignment if alignment != 2 else '',
+                    text.replace('\n', '\\N').replace('<i>', '{\\i1}').replace('</i>', '{\\i0}'))
 
             if sub_lang == 'vostf':
                 sub_lang = 'fr'
@@ -80,8 +112,8 @@ class ADNIE(InfoExtractor):
                 'ext': 'json',
                 'data': json.dumps(sub),
             }, {
-                'ext': 'srt',
-                'data': srt,
+                'ext': 'ssa',
+                'data': ssa,
             }])
         return subtitles
 
@@ -89,7 +121,15 @@ class ADNIE(InfoExtractor):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
         player_config = self._parse_json(self._search_regex(
-            r'playerConfig\s*=\s*({.+});', webpage, 'player config'), video_id)
+            r'playerConfig\s*=\s*({.+});', webpage,
+            'player config', default='{}'), video_id, fatal=False)
+        if not player_config:
+            config_url = urljoin(self._BASE_URL, self._search_regex(
+                r'(?:id="player"|class="[^"]*adn-player-container[^"]*")[^>]+data-url="([^"]+)"',
+                webpage, 'config url'))
+            player_config = self._download_json(
+                config_url, video_id,
+                'Downloading player config JSON metadata')['player']
 
         video_info = {}
         video_info_str = self._search_regex(
@@ -101,14 +141,44 @@ class ADNIE(InfoExtractor):
 
         options = player_config.get('options') or {}
         metas = options.get('metas') or {}
-        title = metas.get('title') or video_info['title']
         links = player_config.get('links') or {}
+        sub_path = player_config.get('subtitles')
+        error = None
+        if not links:
+            links_url = player_config.get('linksurl') or options['videoUrl']
+            token = options['token']
+            self._K = ''.join([random.choice('0123456789abcdef') for _ in range(16)])
+            message = bytes_to_intlist(json.dumps({
+                'k': self._K,
+                'e': 60,
+                't': token,
+            }))
+            padded_message = intlist_to_bytes(pkcs1pad(message, 128))
+            n, e = self._RSA_KEY
+            encrypted_message = long_to_bytes(pow(bytes_to_long(padded_message), e, n))
+            authorization = base64.b64encode(encrypted_message).decode()
+            links_data = self._download_json(
+                urljoin(self._BASE_URL, links_url), video_id,
+                'Downloading links JSON metadata', headers={
+                    'Authorization': 'Bearer ' + authorization,
+                })
+            links = links_data.get('links') or {}
+            metas = metas or links_data.get('meta') or {}
+            sub_path = sub_path or links_data.get('subtitles') or \
+                'index.php?option=com_vodapi&task=subtitles.getJSON&format=json&id=' + video_id
+            sub_path += '&token=' + token
+            error = links_data.get('error')
+        title = metas.get('title') or video_info['title']
 
         formats = []
         for format_id, qualities in links.items():
-            for load_balancer_url in qualities.values():
+            if not isinstance(qualities, dict):
+                continue
+            for quality, load_balancer_url in qualities.items():
                 load_balancer_data = self._download_json(
-                    load_balancer_url, video_id, fatal=False) or {}
+                    load_balancer_url, video_id,
+                    'Downloading %s %s JSON metadata' % (format_id, quality),
+                    fatal=False) or {}
                 m3u8_url = load_balancer_data.get('location')
                 if not m3u8_url:
                     continue
@@ -119,7 +189,8 @@ class ADNIE(InfoExtractor):
                     for f in m3u8_formats:
                         f['language'] = 'fr'
                 formats.extend(m3u8_formats)
-        error = options.get('error')
+        if not error:
+            error = options.get('error')
         if not formats and error:
             raise ExtractorError('%s said: %s' % (self.IE_NAME, error), expected=True)
         self._sort_formats(formats)
@@ -130,7 +201,7 @@ class ADNIE(InfoExtractor):
             'description': strip_or_none(metas.get('summary') or video_info.get('resume')),
             'thumbnail': video_info.get('image'),
             'formats': formats,
-            'subtitles': self.extract_subtitles(player_config.get('subtitles'), video_id),
+            'subtitles': self.extract_subtitles(sub_path, video_id),
             'episode': metas.get('subtitle') or video_info.get('videoTitle'),
             'series': video_info.get('playlistTitle'),
         }