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