Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/crunchyroll.py

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