youtube_dl/extractor/crunchyroll.py

   1 # encoding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import json
   6 import base64
   7 import zlib
   8
   9 from hashlib import sha1
  10 from math import pow, sqrt, floor
  11 from .common import InfoExtractor
  12 from ..utils import (
  13     ExtractorError,
  14     compat_urllib_parse,
  15     compat_urllib_request,
  16     bytes_to_intlist,
  17     intlist_to_bytes,
  18     unified_strdate,
  19     clean_html,
  20 )
  21 from ..aes import (
  22     aes_cbc_decrypt,
  23     inc,
  24 )
  25
  26
  27 class CrunchyrollIE(InfoExtractor):
  28     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.com/(?:[^/]*/[^/?&]*?|media/\?id=)(?P<video_id>[0-9]+))(?:[/?&]|$)'
  29     _TEST = {
  30         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
  31         #'md5': 'b1639fd6ddfaa43788c85f6d1dddd412',
  32         'info_dict': {
  33             'id': '645513',
  34             'ext': 'flv',
  35             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
  36             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
  37             'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
  38             'uploader': 'Yomiuri Telecasting Corporation (YTV)',
  39             'upload_date': '20131013',
  40         },
  41         'params': {
  42             # rtmp
  43             'skip_download': True,
  44         },
  45     }
  46
  47     _FORMAT_IDS = {
  48         '360': ('60', '106'),
  49         '480': ('61', '106'),
  50         '720': ('62', '106'),
  51         '1080': ('80', '108'),
  52     }
  53
  54     def _decrypt_subtitles(self, data, iv, id):
  55         data = bytes_to_intlist(data)
  56         iv = bytes_to_intlist(iv)
  57         id = int(id)
  58
  59         def obfuscate_key_aux(count, modulo, start):
  60             output = list(start)
  61             for _ in range(count):
  62                 output.append(output[-1] + output[-2])
  63             # cut off start values
  64             output = output[2:]
  65             output = list(map(lambda x: x % modulo + 33, output))
  66             return output
  67
  68         def obfuscate_key(key):
  69             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
  70             num2 = (num1 ^ key) << 5
  71             num3 = key ^ num1
  72             num4 = num3 ^ (num3 >> 3) ^ num2
  73             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
  74             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
  75             # Extend 160 Bit hash to 256 Bit
  76             return shaHash + [0] * 12
  77
  78         key = obfuscate_key(id)
  79         class Counter:
  80             __value = iv
  81             def next_value(self):
  82                 temp = self.__value
  83                 self.__value = inc(self.__value)
  84                 return temp
  85         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
  86         return zlib.decompress(decrypted_data)
  87
  88     def _convert_subtitles_to_srt(self, subtitles):
  89         output = ''
  90         for i, (start, end, text) in enumerate(re.findall(r'<event [^>]*?start="([^"]+)" [^>]*?end="([^"]+)" [^>]*?text="([^"]+)"[^>]*?>', subtitles), 1):
  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         return output
  99
 100     def _real_extract(self,url):
 101         mobj = re.match(self._VALID_URL, url)
 102         video_id = mobj.group('video_id')
 103
 104         if mobj.group('prefix') == 'm':
 105             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 106             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 107         else:
 108             webpage_url = 'http://www.' + mobj.group('url')
 109
 110         webpage = self._download_webpage(webpage_url, video_id, 'Downloading webpage')
 111         note_m = self._html_search_regex(r'<div class="showmedia-trailer-notice">(.+?)</div>', webpage, 'trailer-notice', default='')
 112         if note_m:
 113             raise ExtractorError(note_m)
 114
 115         mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
 116         if mobj:
 117             msg = json.loads(mobj.group('msg'))
 118             if msg.get('type') == 'error':
 119                 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
 120
 121         video_title = self._html_search_regex(r'<h1[^>]*>(.+?)</h1>', webpage, 'video_title', flags=re.DOTALL)
 122         video_title = re.sub(r' {2,}', ' ', video_title)
 123         video_description = self._html_search_regex(r'"description":"([^"]+)', webpage, 'video_description', default='')
 124         if not video_description:
 125             video_description = None
 126         video_upload_date = self._html_search_regex(r'<div>Availability for free users:(.+?)</div>', webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
 127         if video_upload_date:
 128             video_upload_date = unified_strdate(video_upload_date)
 129         video_uploader = self._html_search_regex(r'<div>\s*Publisher:(.+?)</div>', webpage, 'video_uploader', fatal=False, flags=re.DOTALL)
 130
 131         playerdata_url = compat_urllib_parse.unquote(self._html_search_regex(r'"config_url":"([^"]+)', webpage, 'playerdata_url'))
 132         playerdata_req = compat_urllib_request.Request(playerdata_url)
 133         playerdata_req.data = compat_urllib_parse.urlencode({'current_page': webpage_url})
 134         playerdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 135         playerdata = self._download_webpage(playerdata_req, video_id, note='Downloading media info')
 136
 137         stream_id = self._search_regex(r'<media_id>([^<]+)', playerdata, 'stream_id')
 138         video_thumbnail = self._search_regex(r'<episode_image_url>([^<]+)', playerdata, 'thumbnail', fatal=False)
 139
 140         formats = []
 141         for fmt in re.findall(r'\?p([0-9]{3,4})=1', webpage):
 142             stream_quality, stream_format = self._FORMAT_IDS[fmt]
 143             video_format = fmt+'p'
 144             streamdata_req = compat_urllib_request.Request('http://www.crunchyroll.com/xml/')
 145             # urlencode doesn't work!
 146             streamdata_req.data = 'req=RpcApiVideoEncode%5FGetStreamInfo&video%5Fencode%5Fquality='+stream_quality+'&media%5Fid='+stream_id+'&video%5Fformat='+stream_format
 147             streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 148             streamdata_req.add_header('Content-Length', str(len(streamdata_req.data)))
 149             streamdata = self._download_webpage(streamdata_req, video_id, note='Downloading media info for '+video_format)
 150             video_url = self._search_regex(r'<host>([^<]+)', streamdata, 'video_url')
 151             video_play_path = self._search_regex(r'<file>([^<]+)', streamdata, 'video_play_path')
 152             formats.append({
 153                 'url': video_url,
 154                 'play_path':   video_play_path,
 155                 'ext': 'flv',
 156                 'format': video_format,
 157                 'format_id': video_format,
 158             })
 159
 160         subtitles = {}
 161         for sub_id, sub_name in re.findall(r'\?ssid=([0-9]+)" title="([^"]+)', webpage):
 162             sub_page = self._download_webpage('http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id='+sub_id,\
 163                                               video_id, note='Downloading subtitles for '+sub_name)
 164             id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
 165             iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
 166             data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
 167             if not id or not iv or not data:
 168                 continue
 169             id = int(id)
 170             iv = base64.b64decode(iv)
 171             data = base64.b64decode(data)
 172
 173             subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
 174             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 175             if not lang_code:
 176                 continue
 177             subtitles[lang_code] = self._convert_subtitles_to_srt(subtitle)
 178
 179         return {
 180             'id':          video_id,
 181             'title':       video_title,
 182             'description': video_description,
 183             'thumbnail':   video_thumbnail,
 184             'uploader':    video_uploader,
 185             'upload_date': video_upload_date,
 186             'subtitles':   subtitles,
 187             'formats':     formats,
 188         }