Raphaël G. Git Repositories - youtubedl/blob - 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     clean_html,
  17     ExtractorError,
  18     intlist_to_bytes,
  19     get_element_by_id,
  20     js_to_json,
  21     int_or_none,
  22     long_to_bytes,
  23     pkcs1pad,
  24     remove_end,
  25 )
  26
  27
  28 class DaisukiIE(InfoExtractor):
  29     _VALID_URL = r'https?://(?:www\.)?daisuki\.net/[^/]+/[^/]+/[^/]+/watch\.[^.]+\.(?P<id>\d+)\.html'
  30
  31     _TEST = {
  32         'url': 'http://www.daisuki.net/tw/en/anime/watch.TheIdolMasterCG.11213.html',
  33         'info_dict': {
  34             'id': '11213',
  35             'ext': 'mp4',
  36             'title': '#01 Who is in the pumpkin carriage? - THE IDOLM@STER CINDERELLA GIRLS',
  37             'subtitles': {
  38                 'mul': [{
  39                     'ext': 'ttml',
  40                 }],
  41             },
  42             'creator': 'BANDAI NAMCO Entertainment',
  43         },
  44         'params': {
  45             'skip_download': True,  # AES-encrypted HLS stream
  46         },
  47     }
  48
  49     # The public key in PEM format can be found in clientlibs_anime_watch.min.js
  50     _RSA_KEY = (0xc5524c25e8e14b366b3754940beeb6f96cb7e2feef0b932c7659a0c5c3bf173d602464c2df73d693b513ae06ff1be8f367529ab30bf969c5640522181f2a0c51ea546ae120d3d8d908595e4eff765b389cde080a1ef7f1bbfb07411cc568db73b7f521cedf270cbfbe0ddbc29b1ac9d0f2d8f4359098caffee6d07915020077d, 65537)
  51
  52     def _real_extract(self, url):
  53         video_id = self._match_id(url)
  54
  55         webpage = self._download_webpage(url, video_id)
  56
  57         flashvars = self._parse_json(self._search_regex(
  58             r'(?s)var\s+flashvars\s*=\s*({.+?});', webpage, 'flashvars'),
  59             video_id, transform_source=js_to_json)
  60
  61         iv = [0] * 16
  62
  63         data = {}
  64         for key in ('device_cd', 'mv_id', 'ss1_prm', 'ss2_prm', 'ss3_prm', 'ss_id'):
  65             data[key] = flashvars.get(key, '')
  66
  67         encrypted_rtn = None
  68
  69         # Some AES keys are rejected. Try it with different AES keys
  70         for idx in range(5):
  71             aes_key = [random.randint(0, 254) for _ in range(32)]
  72             padded_aeskey = intlist_to_bytes(pkcs1pad(aes_key, 128))
  73
  74             n, e = self._RSA_KEY
  75             encrypted_aeskey = long_to_bytes(pow(bytes_to_long(padded_aeskey), e, n))
  76             init_data = self._download_json('http://www.daisuki.net/bin/bgn/init', video_id, query={
  77                 's': flashvars.get('s', ''),
  78                 'c': flashvars.get('ss3_prm', ''),
  79                 'e': url,
  80                 'd': base64.b64encode(intlist_to_bytes(aes_cbc_encrypt(
  81                     bytes_to_intlist(json.dumps(data)),
  82                     aes_key, iv))).decode('ascii'),
  83                 'a': base64.b64encode(encrypted_aeskey).decode('ascii'),
  84             }, note='Downloading JSON metadata' + (' (try #%d)' % (idx + 1) if idx > 0 else ''))
  85
  86             if 'rtn' in init_data:
  87                 encrypted_rtn = init_data['rtn']
  88                 break
  89
  90             self._sleep(5, video_id)
  91
  92         if encrypted_rtn is None:
  93             raise ExtractorError('Failed to fetch init data')
  94
  95         rtn = self._parse_json(
  96             intlist_to_bytes(aes_cbc_decrypt(bytes_to_intlist(
  97                 base64.b64decode(encrypted_rtn)),
  98                 aes_key, iv)).decode('utf-8').rstrip('\0'),
  99             video_id)
 100
 101         formats = self._extract_m3u8_formats(
 102             rtn['play_url'], video_id, ext='mp4', entry_protocol='m3u8_native')
 103
 104         title = remove_end(self._og_search_title(webpage), ' - DAISUKI')
 105
 106         creator = self._html_search_regex(
 107             r'Creator\s*:\s*([^<]+)', webpage, 'creator', fatal=False)
 108
 109         subtitles = {}
 110         caption_url = rtn.get('caption_url')
 111         if caption_url:
 112             # mul: multiple languages
 113             subtitles['mul'] = [{
 114                 'url': caption_url,
 115                 'ext': 'ttml',
 116             }]
 117
 118         return {
 119             'id': video_id,
 120             'title': title,
 121             'formats': formats,
 122             'subtitles': subtitles,
 123             'creator': creator,
 124         }
 125
 126
 127 class DaisukiPlaylistIE(InfoExtractor):
 128     _VALID_URL = r'https?://(?:www\.)daisuki\.net/[^/]+/[^/]+/[^/]+/detail\.(?P<id>[a-zA-Z0-9]+)\.html'
 129
 130     _TEST = {
 131         'url': 'http://www.daisuki.net/tw/en/anime/detail.TheIdolMasterCG.html',
 132         'info_dict': {
 133             'id': 'TheIdolMasterCG',
 134             'title': 'THE IDOLM@STER CINDERELLA GIRLS',
 135             'description': 'md5:0f2c028a9339f7a2c7fbf839edc5c5d8',
 136         },
 137         'playlist_count': 26,
 138     }
 139
 140     def _real_extract(self, url):
 141         playlist_id = self._match_id(url)
 142
 143         webpage = self._download_webpage(url, playlist_id)
 144
 145         episode_pattern = r'''(?sx)
 146             <img[^>]+delay="[^"]+/(\d+)/movie\.jpg".+?
 147             <p[^>]+class=".*?\bepisodeNumber\b.*?">(?:<a[^>]+>)?([^<]+)'''
 148         entries = [{
 149             '_type': 'url_transparent',
 150             'url': url.replace('detail', 'watch').replace('.html', '.' + movie_id + '.html'),
 151             'episode_id': episode_id,
 152             'episode_number': int_or_none(episode_id),
 153         } for movie_id, episode_id in re.findall(episode_pattern, webpage)]
 154
 155         playlist_title = remove_end(
 156             self._og_search_title(webpage, fatal=False), ' - Anime - DAISUKI')
 157         playlist_description = clean_html(get_element_by_id('synopsisTxt', webpage))
 158
 159         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)