_ 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 json
   6 import base64
   7 import zlib
   8 import xml.etree.ElementTree
   9
  10 from hashlib import sha1
  11 from math import pow, sqrt, floor
  12 from .common import InfoExtractor
  13 from ..compat import (
  14     compat_urllib_parse,
  15     compat_urllib_request,
  16 )
  17 from ..utils import (
  18     ExtractorError,
  19     bytes_to_intlist,
  20     intlist_to_bytes,
  21     unified_strdate,
  22     urlencode_postdata,
  23 )
  24 from ..aes import (
  25     aes_cbc_decrypt,
  26     inc,
  27 )
  28
  29
  30 class CrunchyrollIE(InfoExtractor):
  31     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.(?:com|fr)/(?:[^/]*/[^/?&]*?|media/\?id=)(?P<video_id>[0-9]+))(?:[/?&]|$)'
  32     _NETRC_MACHINE = 'crunchyroll'
  33     _TESTS = [{
  34         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
  35         'info_dict': {
  36             'id': '645513',
  37             'ext': 'flv',
  38             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
  39             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
  40             'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
  41             'uploader': 'Yomiuri Telecasting Corporation (YTV)',
  42             'upload_date': '20131013',
  43             'url': 're:(?!.*&amp)',
  44         },
  45         'params': {
  46             # rtmp
  47             'skip_download': True,
  48         },
  49     }, {
  50         'url': 'http://www.crunchyroll.fr/girl-friend-beta/episode-11-goodbye-la-mode-661697',
  51         'only_matching': True,
  52     }]
  53
  54     _FORMAT_IDS = {
  55         '360': ('60', '106'),
  56         '480': ('61', '106'),
  57         '720': ('62', '106'),
  58         '1080': ('80', '108'),
  59     }
  60
  61     def _login(self):
  62         (username, password) = self._get_login_info()
  63         if username is None:
  64             return
  65         self.report_login()
  66         login_url = 'https://www.crunchyroll.com/?a=formhandler'
  67         data = urlencode_postdata({
  68             'formname': 'RpcApiUser_Login',
  69             'name': username,
  70             'password': password,
  71         })
  72         login_request = compat_urllib_request.Request(login_url, data)
  73         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  74         self._download_webpage(login_request, None, False, 'Wrong login info')
  75
  76     def _real_initialize(self):
  77         self._login()
  78
  79     def _decrypt_subtitles(self, data, iv, id):
  80         data = bytes_to_intlist(data)
  81         iv = bytes_to_intlist(iv)
  82         id = int(id)
  83
  84         def obfuscate_key_aux(count, modulo, start):
  85             output = list(start)
  86             for _ in range(count):
  87                 output.append(output[-1] + output[-2])
  88             # cut off start values
  89             output = output[2:]
  90             output = list(map(lambda x: x % modulo + 33, output))
  91             return output
  92
  93         def obfuscate_key(key):
  94             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
  95             num2 = (num1 ^ key) << 5
  96             num3 = key ^ num1
  97             num4 = num3 ^ (num3 >> 3) ^ num2
  98             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
  99             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
 100             # Extend 160 Bit hash to 256 Bit
 101             return shaHash + [0] * 12
 102
 103         key = obfuscate_key(id)
 104
 105         class Counter:
 106             __value = iv
 107
 108             def next_value(self):
 109                 temp = self.__value
 110                 self.__value = inc(self.__value)
 111                 return temp
 112         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
 113         return zlib.decompress(decrypted_data)
 114
 115     def _convert_subtitles_to_srt(self, sub_root):
 116         output = ''
 117
 118         for i, event in enumerate(sub_root.findall('./events/event'), 1):
 119             start = event.attrib['start'].replace('.', ',')
 120             end = event.attrib['end'].replace('.', ',')
 121             text = event.attrib['text'].replace('\\N', '\n')
 122             output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
 123         return output
 124
 125     def _convert_subtitles_to_ass(self, sub_root):
 126         output = ''
 127
 128         def ass_bool(strvalue):
 129             assvalue = '0'
 130             if strvalue == '1':
 131                 assvalue = '-1'
 132             return assvalue
 133
 134         output = '[Script Info]\n'
 135         output += 'Title: %s\n' % sub_root.attrib["title"]
 136         output += 'ScriptType: v4.00+\n'
 137         output += 'WrapStyle: %s\n' % sub_root.attrib["wrap_style"]
 138         output += 'PlayResX: %s\n' % sub_root.attrib["play_res_x"]
 139         output += 'PlayResY: %s\n' % sub_root.attrib["play_res_y"]
 140         output += """ScaledBorderAndShadow: yes
 141
 142 [V4+ Styles]
 143 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
 144 """
 145         for style in sub_root.findall('./styles/style'):
 146             output += 'Style: ' + style.attrib["name"]
 147             output += ',' + style.attrib["font_name"]
 148             output += ',' + style.attrib["font_size"]
 149             output += ',' + style.attrib["primary_colour"]
 150             output += ',' + style.attrib["secondary_colour"]
 151             output += ',' + style.attrib["outline_colour"]
 152             output += ',' + style.attrib["back_colour"]
 153             output += ',' + ass_bool(style.attrib["bold"])
 154             output += ',' + ass_bool(style.attrib["italic"])
 155             output += ',' + ass_bool(style.attrib["underline"])
 156             output += ',' + ass_bool(style.attrib["strikeout"])
 157             output += ',' + style.attrib["scale_x"]
 158             output += ',' + style.attrib["scale_y"]
 159             output += ',' + style.attrib["spacing"]
 160             output += ',' + style.attrib["angle"]
 161             output += ',' + style.attrib["border_style"]
 162             output += ',' + style.attrib["outline"]
 163             output += ',' + style.attrib["shadow"]
 164             output += ',' + style.attrib["alignment"]
 165             output += ',' + style.attrib["margin_l"]
 166             output += ',' + style.attrib["margin_r"]
 167             output += ',' + style.attrib["margin_v"]
 168             output += ',' + style.attrib["encoding"]
 169             output += '\n'
 170
 171         output += """
 172 [Events]
 173 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
 174 """
 175         for event in sub_root.findall('./events/event'):
 176             output += 'Dialogue: 0'
 177             output += ',' + event.attrib["start"]
 178             output += ',' + event.attrib["end"]
 179             output += ',' + event.attrib["style"]
 180             output += ',' + event.attrib["name"]
 181             output += ',' + event.attrib["margin_l"]
 182             output += ',' + event.attrib["margin_r"]
 183             output += ',' + event.attrib["margin_v"]
 184             output += ',' + event.attrib["effect"]
 185             output += ',' + event.attrib["text"]
 186             output += '\n'
 187
 188         return output
 189
 190     def _get_subtitles(self, video_id, webpage):
 191         subtitles = {}
 192         for sub_id, sub_name in re.findall(r'\?ssid=([0-9]+)" title="([^"]+)', webpage):
 193             sub_page = self._download_webpage(
 194                 'http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id=' + sub_id,
 195                 video_id, note='Downloading subtitles for ' + sub_name)
 196             id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
 197             iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
 198             data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
 199             if not id or not iv or not data:
 200                 continue
 201             id = int(id)
 202             iv = base64.b64decode(iv)
 203             data = base64.b64decode(data)
 204
 205             subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
 206             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 207             if not lang_code:
 208                 continue
 209             sub_root = xml.etree.ElementTree.fromstring(subtitle)
 210             subtitles[lang_code] = [
 211                 {
 212                     'ext': 'srt',
 213                     'data': self._convert_subtitles_to_srt(sub_root),
 214                 },
 215                 {
 216                     'ext': 'ass',
 217                     'data': self._convert_subtitles_to_ass(sub_root),
 218                 },
 219             ]
 220         return subtitles
 221
 222     def _real_extract(self, url):
 223         mobj = re.match(self._VALID_URL, url)
 224         video_id = mobj.group('video_id')
 225
 226         if mobj.group('prefix') == 'm':
 227             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 228             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 229         else:
 230             webpage_url = 'http://www.' + mobj.group('url')
 231
 232         webpage = self._download_webpage(webpage_url, video_id, 'Downloading webpage')
 233         note_m = self._html_search_regex(r'<div class="showmedia-trailer-notice">(.+?)</div>', webpage, 'trailer-notice', default='')
 234         if note_m:
 235             raise ExtractorError(note_m)
 236
 237         mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
 238         if mobj:
 239             msg = json.loads(mobj.group('msg'))
 240             if msg.get('type') == 'error':
 241                 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
 242
 243         video_title = self._html_search_regex(r'<h1[^>]*>(.+?)</h1>', webpage, 'video_title', flags=re.DOTALL)
 244         video_title = re.sub(r' {2,}', ' ', video_title)
 245         video_description = self._html_search_regex(r'"description":"([^"]+)', webpage, 'video_description', default='')
 246         if not video_description:
 247             video_description = None
 248         video_upload_date = self._html_search_regex(r'<div>Availability for free users:(.+?)</div>', webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
 249         if video_upload_date:
 250             video_upload_date = unified_strdate(video_upload_date)
 251         video_uploader = self._html_search_regex(r'<div>\s*Publisher:(.+?)</div>', webpage, 'video_uploader', fatal=False, flags=re.DOTALL)
 252
 253         playerdata_url = compat_urllib_parse.unquote(self._html_search_regex(r'"config_url":"([^"]+)', webpage, 'playerdata_url'))
 254         playerdata_req = compat_urllib_request.Request(playerdata_url)
 255         playerdata_req.data = compat_urllib_parse.urlencode({'current_page': webpage_url})
 256         playerdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 257         playerdata = self._download_webpage(playerdata_req, video_id, note='Downloading media info')
 258
 259         stream_id = self._search_regex(r'<media_id>([^<]+)', playerdata, 'stream_id')
 260         video_thumbnail = self._search_regex(r'<episode_image_url>([^<]+)', playerdata, 'thumbnail', fatal=False)
 261
 262         formats = []
 263         for fmt in re.findall(r'showmedia\.([0-9]{3,4})p', webpage):
 264             stream_quality, stream_format = self._FORMAT_IDS[fmt]
 265             video_format = fmt + 'p'
 266             streamdata_req = compat_urllib_request.Request('http://www.crunchyroll.com/xml/')
 267             # urlencode doesn't work!
 268             streamdata_req.data = 'req=RpcApiVideoEncode%5FGetStreamInfo&video%5Fencode%5Fquality=' + stream_quality + '&media%5Fid=' + stream_id + '&video%5Fformat=' + stream_format
 269             streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 270             streamdata_req.add_header('Content-Length', str(len(streamdata_req.data)))
 271             streamdata = self._download_xml(
 272                 streamdata_req, video_id,
 273                 note='Downloading media info for %s' % video_format)
 274             video_url = streamdata.find('.//host').text
 275             video_play_path = streamdata.find('.//file').text
 276             formats.append({
 277                 'url': video_url,
 278                 'play_path': video_play_path,
 279                 'ext': 'flv',
 280                 'format': video_format,
 281                 'format_id': video_format,
 282             })
 283
 284         subtitles = self.extract_subtitles(video_id, webpage)
 285
 286         return {
 287             'id': video_id,
 288             'title': video_title,
 289             'description': video_description,
 290             'thumbnail': video_thumbnail,
 291             'uploader': video_uploader,
 292             'upload_date': video_upload_date,
 293             'subtitles': subtitles,
 294             'formats': formats,
 295         }
 296
 297
 298 class CrunchyrollShowPlaylistIE(InfoExtractor):
 299     IE_NAME = "crunchyroll:playlist"
 300     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.com/(?!(?:news|anime-news|library|forum|launchcalendar|lineup|store|comics|freetrial|login))(?P<id>[\w\-]+))/?$'
 301
 302     _TESTS = [{
 303         'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 304         'info_dict': {
 305             'id': 'a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 306             'title': 'A Bridge to the Starry Skies - Hoshizora e Kakaru Hashi'
 307         },
 308         'playlist_count': 13,
 309     }]
 310
 311     def _real_extract(self, url):
 312         show_id = self._match_id(url)
 313
 314         webpage = self._download_webpage(url, show_id)
 315         title = self._html_search_regex(
 316             r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>',
 317             webpage, 'title')
 318         episode_paths = re.findall(
 319             r'(?s)<li id="showview_videos_media_[0-9]+"[^>]+>.*?<a href="([^"]+)"',
 320             webpage)
 321         entries = [
 322             self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll')
 323             for ep in episode_paths
 324         ]
 325         entries.reverse()
 326
 327         return {
 328             '_type': 'playlist',
 329             'id': show_id,
 330             'title': title,
 331             'entries': entries,
 332         }