_ Git - youtube-dl/blob - youtube_dl/extractor/crunchyroll.py

   1 # encoding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import base64
   6 import zlib
   7
   8 from hashlib import sha1
   9 from math import pow, sqrt, floor
  10 from .common import InfoExtractor
  11 from ..utils import (
  12     ExtractorError,
  13     compat_urllib_parse,
  14     compat_urllib_request,
  15     bytes_to_intlist,
  16     intlist_to_bytes,
  17     unified_strdate,
  18     clean_html,
  19 )
  20 from ..aes import (
  21     aes_cbc_decrypt,
  22     inc,
  23 )
  24
  25
  26 class CrunchyrollIE(InfoExtractor):
  27     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.com/(?:[^/]*/[^/?&]*?|media/\?id=)(?P<video_id>[0-9]+))(?:[/?&]|$)'
  28     _TEST = {
  29         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
  30         #'md5': 'b1639fd6ddfaa43788c85f6d1dddd412',
  31         'info_dict': {
  32             'id': '645513',
  33             'ext': 'flv',
  34             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
  35             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
  36             'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
  37             'uploader': 'Yomiuri Telecasting Corporation (YTV)',
  38             'upload_date': '20131013',
  39         },
  40         'params': {
  41             # rtmp
  42             'skip_download': True,
  43         },
  44     }
  45
  46     _FORMAT_IDS = {
  47         '360': ('60', '106'),
  48         '480': ('61', '106'),
  49         '720': ('62', '106'),
  50         '1080': ('80', '108'),
  51     }
  52
  53     def _decrypt_subtitles(self, data, iv, id):
  54         data = bytes_to_intlist(data)
  55         iv = bytes_to_intlist(iv)
  56         id = int(id)
  57
  58         def obfuscate_key_aux(count, modulo, start):
  59             output = list(start)
  60             for _ in range(count):
  61                 output.append(output[-1] + output[-2])
  62             # cut off start values
  63             output = output[2:]
  64             output = list(map(lambda x: x % modulo + 33, output))
  65             return output
  66
  67         def obfuscate_key(key):
  68             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
  69             num2 = (num1 ^ key) << 5
  70             num3 = key ^ num1
  71             num4 = num3 ^ (num3 >> 3) ^ num2
  72             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
  73             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
  74             # Extend 160 Bit hash to 256 Bit
  75             return shaHash + [0] * 12
  76
  77         key = obfuscate_key(id)
  78         class Counter:
  79             __value = iv
  80             def next_value(self):
  81                 temp = self.__value
  82                 self.__value = inc(self.__value)
  83                 return temp
  84         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
  85         return zlib.decompress(decrypted_data)
  86
  87     def _convert_subtitles_to_srt(self, subtitles):
  88         i = 1
  89         output = ''
  90         for start, end, text in re.findall(r'<event [^>]*?start="([^"]+)" [^>]*?end="([^"]+)" [^>]*?text="([^"]+)"[^>]*?>', subtitles):
  91             start = start.replace('.', ',')
  92             end = end.replace('.', ',')
  93             text = clean_html(text)
  94             text = text.replace('\\N', '\n')
  95             if not text:
  96                 continue
  97             output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
  98             i += 1
  99         return output
 100
 101     def _real_extract(self,url):
 102         mobj = re.match(self._VALID_URL, url)
 103         video_id = mobj.group('video_id')
 104
 105         if mobj.group('prefix') == 'm':
 106             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 107             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 108         else:
 109             webpage_url = 'http://www.' + mobj.group('url')
 110
 111         webpage = self._download_webpage(webpage_url, video_id, 'Downloading webpage')
 112         note_m = self._html_search_regex(r'<div class="showmedia-trailer-notice">(.+?)</div>', webpage, 'trailer-notice', default='')
 113         if note_m:
 114             raise ExtractorError(note_m)
 115
 116         video_title = self._html_search_regex(r'<h1[^>]*>(.+?)</h1>', webpage, 'video_title', flags=re.DOTALL)
 117         video_title = re.sub(r' {2,}', ' ', video_title)
 118         video_description = self._html_search_regex(r'"description":"([^"]+)', webpage, 'video_description', default='')
 119         if not video_description:
 120             video_description = None
 121         video_upload_date = self._html_search_regex(r'<div>Availability for free users:(.+?)</div>', webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
 122         if video_upload_date:
 123             video_upload_date = unified_strdate(video_upload_date)
 124         video_uploader = self._html_search_regex(r'<div>\s*Publisher:(.+?)</div>', webpage, 'video_uploader', fatal=False, flags=re.DOTALL)
 125
 126         playerdata_url = compat_urllib_parse.unquote(self._html_search_regex(r'"config_url":"([^"]+)', webpage, 'playerdata_url'))
 127         playerdata_req = compat_urllib_request.Request(playerdata_url)
 128         playerdata_req.data = compat_urllib_parse.urlencode({'current_page': webpage_url})
 129         playerdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 130         playerdata = self._download_webpage(playerdata_req, video_id, note='Downloading media info')
 131
 132         stream_id = self._search_regex(r'<media_id>([^<]+)', playerdata, 'stream_id')
 133         video_thumbnail = self._search_regex(r'<episode_image_url>([^<]+)', playerdata, 'thumbnail', fatal=False)
 134
 135         formats = []
 136         for fmt in re.findall(r'\?p([0-9]{3,4})=1', webpage):
 137             stream_quality, stream_format = self._FORMAT_IDS[fmt]
 138             video_format = fmt+'p'
 139             streamdata_req = compat_urllib_request.Request('http://www.crunchyroll.com/xml/')
 140             # urlencode doesn't work!
 141             streamdata_req.data = 'req=RpcApiVideoEncode%5FGetStreamInfo&video%5Fencode%5Fquality='+stream_quality+'&media%5Fid='+stream_id+'&video%5Fformat='+stream_format
 142             streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 143             streamdata_req.add_header('Content-Length', str(len(streamdata_req.data)))
 144             streamdata = self._download_webpage(streamdata_req, video_id, note='Downloading media info for '+video_format)
 145             video_url = self._search_regex(r'<host>([^<]+)', streamdata, 'video_url')
 146             video_play_path = self._search_regex(r'<file>([^<]+)', streamdata, 'video_play_path')
 147             formats.append({
 148                 'url': video_url,
 149                 'play_path':   video_play_path,
 150                 'ext': 'flv',
 151                 'format': video_format,
 152                 'format_id': video_format,
 153             })
 154
 155         subtitles = {}
 156         for sub_id, sub_name in re.findall(r'\?ssid=([0-9]+)" title="([^"]+)', webpage):
 157             sub_page = self._download_webpage('http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id='+sub_id,\
 158                                               video_id, note='Downloading subtitles for '+sub_name)
 159             id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
 160             iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
 161             data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
 162             if not id or not iv or not data:
 163                 continue
 164             id = int(id)
 165             iv = base64.b64decode(iv)
 166             data = base64.b64decode(data)
 167
 168             subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
 169             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 170             if not lang_code:
 171                 continue
 172             subtitles[lang_code] = self._convert_subtitles_to_srt(subtitle)
 173
 174         return {
 175             'id':          video_id,
 176             'title':       video_title,
 177             'description': video_description,
 178             'thumbnail':   video_thumbnail,
 179             'uploader':    video_uploader,
 180             'upload_date': video_upload_date,
 181             'subtitles':   subtitles,
 182             'formats':     formats,
 183         }