5c9ac68a02590b3eccea83cd4d9a49ffca47a2e8
[youtube-dl] / youtube_dl / extractor / daisuki.py
1 from __future__ import unicode_literals
2
3 import base64
4 import json
5 import random
6 import re
7
8 from .common import InfoExtractor
9 from ..aes import (
10     aes_cbc_decrypt,
11     aes_cbc_encrypt,
12 )
13 from ..utils import (
14     bytes_to_intlist,
15     bytes_to_long,
16     extract_attributes,
17     ExtractorError,
18     intlist_to_bytes,
19     js_to_json,
20     int_or_none,
21     long_to_bytes,
22     pkcs1pad,
23 )
24
25
26 class DaisukiMottoIE(InfoExtractor):
27     _VALID_URL = r'https?://motto\.daisuki\.net/framewatch/embed/[^/]+/(?P<id>[0-9a-zA-Z]{3})'
28
29     _TEST = {
30         'url': 'http://motto.daisuki.net/framewatch/embed/embedDRAGONBALLSUPERUniverseSurvivalsaga/V2e/760/428',
31         'info_dict': {
32             'id': 'V2e',
33             'ext': 'mp4',
34             'title': '#117 SHOWDOWN OF LOVE! ANDROIDS VS UNIVERSE 2!!',
35             'subtitles': {
36                 'mul': [{
37                     'ext': 'ttml',
38                 }],
39             },
40         },
41         'params': {
42             'skip_download': True,  # AES-encrypted HLS stream
43         },
44     }
45
46     # The public key in PEM format can be found in clientlibs_anime_watch.min.js
47     _RSA_KEY = (0xc5524c25e8e14b366b3754940beeb6f96cb7e2feef0b932c7659a0c5c3bf173d602464c2df73d693b513ae06ff1be8f367529ab30bf969c5640522181f2a0c51ea546ae120d3d8d908595e4eff765b389cde080a1ef7f1bbfb07411cc568db73b7f521cedf270cbfbe0ddbc29b1ac9d0f2d8f4359098caffee6d07915020077d, 65537)
48
49     def _real_extract(self, url):
50         video_id = self._match_id(url)
51
52         webpage = self._download_webpage(url, video_id)
53
54         flashvars = self._parse_json(self._search_regex(
55             r'(?s)var\s+flashvars\s*=\s*({.+?});', webpage, 'flashvars'),
56             video_id, transform_source=js_to_json)
57
58         iv = [0] * 16
59
60         data = {}
61         for key in ('device_cd', 'mv_id', 'ss1_prm', 'ss2_prm', 'ss3_prm', 'ss_id'):
62             data[key] = flashvars.get(key, '')
63
64         encrypted_rtn = None
65
66         # Some AES keys are rejected. Try it with different AES keys
67         for idx in range(5):
68             aes_key = [random.randint(0, 254) for _ in range(32)]
69             padded_aeskey = intlist_to_bytes(pkcs1pad(aes_key, 128))
70
71             n, e = self._RSA_KEY
72             encrypted_aeskey = long_to_bytes(pow(bytes_to_long(padded_aeskey), e, n))
73             init_data = self._download_json(
74                 'http://motto.daisuki.net/fastAPI/bgn/init/',
75                 video_id, query={
76                     's': flashvars.get('s', ''),
77                     'c': flashvars.get('ss3_prm', ''),
78                     'e': url,
79                     'd': base64.b64encode(intlist_to_bytes(aes_cbc_encrypt(
80                         bytes_to_intlist(json.dumps(data)),
81                         aes_key, iv))).decode('ascii'),
82                     'a': base64.b64encode(encrypted_aeskey).decode('ascii'),
83                 }, note='Downloading JSON metadata' + (' (try #%d)' % (idx + 1) if idx > 0 else ''))
84
85             if 'rtn' in init_data:
86                 encrypted_rtn = init_data['rtn']
87                 break
88
89             self._sleep(5, video_id)
90
91         if encrypted_rtn is None:
92             raise ExtractorError('Failed to fetch init data')
93
94         rtn = self._parse_json(
95             intlist_to_bytes(aes_cbc_decrypt(bytes_to_intlist(
96                 base64.b64decode(encrypted_rtn)),
97                 aes_key, iv)).decode('utf-8').rstrip('\0'),
98             video_id)
99
100         title = rtn['title_str']
101
102         formats = self._extract_m3u8_formats(
103             rtn['play_url'], video_id, ext='mp4', entry_protocol='m3u8_native')
104
105         subtitles = {}
106         caption_url = rtn.get('caption_url')
107         if caption_url:
108             # mul: multiple languages
109             subtitles['mul'] = [{
110                 'url': caption_url,
111                 'ext': 'ttml',
112             }]
113
114         return {
115             'id': video_id,
116             'title': title,
117             'formats': formats,
118             'subtitles': subtitles,
119         }
120
121
122 class DaisukiMottoPlaylistIE(InfoExtractor):
123     _VALID_URL = r'https?://motto\.daisuki\.net/(?P<id>information)/'
124
125     _TEST = {
126         'url': 'http://motto.daisuki.net/information/',
127         'info_dict': {
128             'title': 'DRAGON BALL SUPER',
129         },
130         'playlist_mincount': 117,
131     }
132
133     def _real_extract(self, url):
134         playlist_id = self._match_id(url)
135
136         webpage = self._download_webpage(url, playlist_id)
137
138         entries = []
139         for li in re.findall(r'(<li[^>]+?data-product_id="[a-zA-Z0-9]{3}"[^>]+>)', webpage):
140             attr = extract_attributes(li)
141             ad_id = attr.get('data-ad_id')
142             product_id = attr.get('data-product_id')
143             if ad_id and product_id:
144                 episode_id = attr.get('data-chapter')
145                 entries.append({
146                     '_type': 'url_transparent',
147                     'url': 'http://motto.daisuki.net/framewatch/embed/%s/%s/760/428' % (ad_id, product_id),
148                     'episode_id': episode_id,
149                     'episode_number': int_or_none(episode_id),
150                     'ie_key': 'DaisukiMotto',
151                 })
152
153         return self.playlist_result(entries, playlist_title='DRAGON BALL SUPER')