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_parse_unquote,
  16     compat_urllib_request,
  17     compat_urlparse,
  18 )
  19 from ..utils import (
  20     ExtractorError,
  21     bytes_to_intlist,
  22     intlist_to_bytes,
  23     int_or_none,
  24     remove_end,
  25     unified_strdate,
  26     urlencode_postdata,
  27     xpath_text,
  28 )
  29 from ..aes import (
  30     aes_cbc_decrypt,
  31 )
  32
  33
  34 class CrunchyrollBaseIE(InfoExtractor):
  35     _NETRC_MACHINE = 'crunchyroll'
  36
  37     def _login(self):
  38         (username, password) = self._get_login_info()
  39         if username is None:
  40             return
  41         self.report_login()
  42         login_url = 'https://www.crunchyroll.com/?a=formhandler'
  43         data = urlencode_postdata({
  44             'formname': 'RpcApiUser_Login',
  45             'name': username,
  46             'password': password,
  47         })
  48         login_request = compat_urllib_request.Request(login_url, data)
  49         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  50         self._download_webpage(login_request, None, False, 'Wrong login info')
  51
  52     def _real_initialize(self):
  53         self._login()
  54
  55     def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None):
  56         request = (url_or_request if isinstance(url_or_request, compat_urllib_request.Request)
  57                    else compat_urllib_request.Request(url_or_request))
  58         # Accept-Language must be set explicitly to accept any language to avoid issues
  59         # similar to https://github.com/rg3/youtube-dl/issues/6797.
  60         # Along with IP address Crunchyroll uses Accept-Language to guess whether georestriction
  61         # should be imposed or not (from what I can see it just takes the first language
  62         # ignoring the priority and requires it to correspond the IP). By the way this causes
  63         # Crunchyroll to not work in georestriction cases in some browsers that don't place
  64         # the locale lang first in header. However allowing any language seems to workaround the issue.
  65         request.add_header('Accept-Language', '*')
  66         return super(CrunchyrollBaseIE, self)._download_webpage(
  67             request, video_id, note, errnote, fatal, tries, timeout, encoding)
  68
  69     @staticmethod
  70     def _add_skip_wall(url):
  71         parsed_url = compat_urlparse.urlparse(url)
  72         qs = compat_urlparse.parse_qs(parsed_url.query)
  73         # Always force skip_wall to bypass maturity wall, namely 18+ confirmation message:
  74         # > This content may be inappropriate for some people.
  75         # > Are you sure you want to continue?
  76         # since it's not disabled by default in crunchyroll account's settings.
  77         # See https://github.com/rg3/youtube-dl/issues/7202.
  78         qs['skip_wall'] = ['1']
  79         return compat_urlparse.urlunparse(
  80             parsed_url._replace(query=compat_urllib_parse.urlencode(qs, True)))
  81
  82
  83 class CrunchyrollIE(CrunchyrollBaseIE):
  84     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.(?:com|fr)/(?:media(?:-|/\?id=)|[^/]*/[^/?&]*?)(?P<video_id>[0-9]+))(?:[/?&]|$)'
  85     _TESTS = [{
  86         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
  87         'info_dict': {
  88             'id': '645513',
  89             'ext': 'flv',
  90             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
  91             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
  92             'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
  93             'uploader': 'Yomiuri Telecasting Corporation (YTV)',
  94             'upload_date': '20131013',
  95             'url': 're:(?!.*&amp)',
  96         },
  97         'params': {
  98             # rtmp
  99             'skip_download': True,
 100         },
 101     }, {
 102         'url': 'http://www.crunchyroll.com/media-589804/culture-japan-1',
 103         'info_dict': {
 104             'id': '589804',
 105             'ext': 'flv',
 106             'title': 'Culture Japan Episode 1 – Rebuilding Japan after the 3.11',
 107             'description': 'md5:fe2743efedb49d279552926d0bd0cd9e',
 108             'thumbnail': 're:^https?://.*\.jpg$',
 109             'uploader': 'Danny Choo Network',
 110             'upload_date': '20120213',
 111         },
 112         'params': {
 113             # rtmp
 114             'skip_download': True,
 115         },
 116     }, {
 117         'url': 'http://www.crunchyroll.fr/girl-friend-beta/episode-11-goodbye-la-mode-661697',
 118         'only_matching': True,
 119     }, {
 120         # geo-restricted (US), 18+ maturity wall, non-premium available
 121         'url': 'http://www.crunchyroll.com/cosplay-complex-ova/episode-1-the-birth-of-the-cosplay-club-565617',
 122         'only_matching': True,
 123     }]
 124
 125     _FORMAT_IDS = {
 126         '360': ('60', '106'),
 127         '480': ('61', '106'),
 128         '720': ('62', '106'),
 129         '1080': ('80', '108'),
 130     }
 131
 132     def _decrypt_subtitles(self, data, iv, id):
 133         data = bytes_to_intlist(base64.b64decode(data.encode('utf-8')))
 134         iv = bytes_to_intlist(base64.b64decode(iv.encode('utf-8')))
 135         id = int(id)
 136
 137         def obfuscate_key_aux(count, modulo, start):
 138             output = list(start)
 139             for _ in range(count):
 140                 output.append(output[-1] + output[-2])
 141             # cut off start values
 142             output = output[2:]
 143             output = list(map(lambda x: x % modulo + 33, output))
 144             return output
 145
 146         def obfuscate_key(key):
 147             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
 148             num2 = (num1 ^ key) << 5
 149             num3 = key ^ num1
 150             num4 = num3 ^ (num3 >> 3) ^ num2
 151             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
 152             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
 153             # Extend 160 Bit hash to 256 Bit
 154             return shaHash + [0] * 12
 155
 156         key = obfuscate_key(id)
 157
 158         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
 159         return zlib.decompress(decrypted_data)
 160
 161     def _convert_subtitles_to_srt(self, sub_root):
 162         output = ''
 163
 164         for i, event in enumerate(sub_root.findall('./events/event'), 1):
 165             start = event.attrib['start'].replace('.', ',')
 166             end = event.attrib['end'].replace('.', ',')
 167             text = event.attrib['text'].replace('\\N', '\n')
 168             output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
 169         return output
 170
 171     def _convert_subtitles_to_ass(self, sub_root):
 172         output = ''
 173
 174         def ass_bool(strvalue):
 175             assvalue = '0'
 176             if strvalue == '1':
 177                 assvalue = '-1'
 178             return assvalue
 179
 180         output = '[Script Info]\n'
 181         output += 'Title: %s\n' % sub_root.attrib["title"]
 182         output += 'ScriptType: v4.00+\n'
 183         output += 'WrapStyle: %s\n' % sub_root.attrib["wrap_style"]
 184         output += 'PlayResX: %s\n' % sub_root.attrib["play_res_x"]
 185         output += 'PlayResY: %s\n' % sub_root.attrib["play_res_y"]
 186         output += """ScaledBorderAndShadow: yes
 187
 188 [V4+ Styles]
 189 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
 190 """
 191         for style in sub_root.findall('./styles/style'):
 192             output += 'Style: ' + style.attrib["name"]
 193             output += ',' + style.attrib["font_name"]
 194             output += ',' + style.attrib["font_size"]
 195             output += ',' + style.attrib["primary_colour"]
 196             output += ',' + style.attrib["secondary_colour"]
 197             output += ',' + style.attrib["outline_colour"]
 198             output += ',' + style.attrib["back_colour"]
 199             output += ',' + ass_bool(style.attrib["bold"])
 200             output += ',' + ass_bool(style.attrib["italic"])
 201             output += ',' + ass_bool(style.attrib["underline"])
 202             output += ',' + ass_bool(style.attrib["strikeout"])
 203             output += ',' + style.attrib["scale_x"]
 204             output += ',' + style.attrib["scale_y"]
 205             output += ',' + style.attrib["spacing"]
 206             output += ',' + style.attrib["angle"]
 207             output += ',' + style.attrib["border_style"]
 208             output += ',' + style.attrib["outline"]
 209             output += ',' + style.attrib["shadow"]
 210             output += ',' + style.attrib["alignment"]
 211             output += ',' + style.attrib["margin_l"]
 212             output += ',' + style.attrib["margin_r"]
 213             output += ',' + style.attrib["margin_v"]
 214             output += ',' + style.attrib["encoding"]
 215             output += '\n'
 216
 217         output += """
 218 [Events]
 219 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
 220 """
 221         for event in sub_root.findall('./events/event'):
 222             output += 'Dialogue: 0'
 223             output += ',' + event.attrib["start"]
 224             output += ',' + event.attrib["end"]
 225             output += ',' + event.attrib["style"]
 226             output += ',' + event.attrib["name"]
 227             output += ',' + event.attrib["margin_l"]
 228             output += ',' + event.attrib["margin_r"]
 229             output += ',' + event.attrib["margin_v"]
 230             output += ',' + event.attrib["effect"]
 231             output += ',' + event.attrib["text"]
 232             output += '\n'
 233
 234         return output
 235
 236     def _extract_subtitles(self, subtitle):
 237         sub_root = xml.etree.ElementTree.fromstring(subtitle)
 238         return [{
 239             'ext': 'srt',
 240             'data': self._convert_subtitles_to_srt(sub_root),
 241         }, {
 242             'ext': 'ass',
 243             'data': self._convert_subtitles_to_ass(sub_root),
 244         }]
 245
 246     def _get_subtitles(self, video_id, webpage):
 247         subtitles = {}
 248         for sub_id, sub_name in re.findall(r'\bssid=([0-9]+)"[^>]+?\btitle="([^"]+)', webpage):
 249             sub_page = self._download_webpage(
 250                 'http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id=' + sub_id,
 251                 video_id, note='Downloading subtitles for ' + sub_name)
 252             id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
 253             iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
 254             data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
 255             if not id or not iv or not data:
 256                 continue
 257             subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
 258             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 259             if not lang_code:
 260                 continue
 261             subtitles[lang_code] = self._extract_subtitles(subtitle)
 262         return subtitles
 263
 264     def _real_extract(self, url):
 265         mobj = re.match(self._VALID_URL, url)
 266         video_id = mobj.group('video_id')
 267
 268         if mobj.group('prefix') == 'm':
 269             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 270             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 271         else:
 272             webpage_url = 'http://www.' + mobj.group('url')
 273
 274         webpage = self._download_webpage(self._add_skip_wall(webpage_url), video_id, 'Downloading webpage')
 275         note_m = self._html_search_regex(
 276             r'<div class="showmedia-trailer-notice">(.+?)</div>',
 277             webpage, 'trailer-notice', default='')
 278         if note_m:
 279             raise ExtractorError(note_m)
 280
 281         mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
 282         if mobj:
 283             msg = json.loads(mobj.group('msg'))
 284             if msg.get('type') == 'error':
 285                 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
 286
 287         if 'To view this, please log in to verify you are 18 or older.' in webpage:
 288             self.raise_login_required()
 289
 290         video_title = self._html_search_regex(r'<h1[^>]*>(.+?)</h1>', webpage, 'video_title', flags=re.DOTALL)
 291         video_title = re.sub(r' {2,}', ' ', video_title)
 292         video_description = self._html_search_regex(r'"description":"([^"]+)', webpage, 'video_description', default='')
 293         if not video_description:
 294             video_description = None
 295         video_upload_date = self._html_search_regex(
 296             [r'<div>Availability for free users:(.+?)</div>', r'<div>[^<>]+<span>\s*(.+?\d{4})\s*</span></div>'],
 297             webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
 298         if video_upload_date:
 299             video_upload_date = unified_strdate(video_upload_date)
 300         video_uploader = self._html_search_regex(
 301             r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', webpage,
 302             'video_uploader', fatal=False)
 303
 304         playerdata_url = compat_urllib_parse_unquote(self._html_search_regex(r'"config_url":"([^"]+)', webpage, 'playerdata_url'))
 305         playerdata_req = compat_urllib_request.Request(playerdata_url)
 306         playerdata_req.data = compat_urllib_parse.urlencode({'current_page': webpage_url})
 307         playerdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 308         playerdata = self._download_webpage(playerdata_req, video_id, note='Downloading media info')
 309
 310         stream_id = self._search_regex(r'<media_id>([^<]+)', playerdata, 'stream_id')
 311         video_thumbnail = self._search_regex(r'<episode_image_url>([^<]+)', playerdata, 'thumbnail', fatal=False)
 312
 313         formats = []
 314         for fmt in re.findall(r'showmedia\.([0-9]{3,4})p', webpage):
 315             stream_quality, stream_format = self._FORMAT_IDS[fmt]
 316             video_format = fmt + 'p'
 317             streamdata_req = compat_urllib_request.Request(
 318                 'http://www.crunchyroll.com/xml/?req=RpcApiVideoPlayer_GetStandardConfig&media_id=%s&video_format=%s&video_quality=%s'
 319                 % (stream_id, stream_format, stream_quality),
 320                 compat_urllib_parse.urlencode({'current_page': url}).encode('utf-8'))
 321             streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 322             streamdata = self._download_xml(
 323                 streamdata_req, video_id,
 324                 note='Downloading media info for %s' % video_format)
 325             stream_info = streamdata.find('./{default}preload/stream_info')
 326             video_url = stream_info.find('./host').text
 327             video_play_path = stream_info.find('./file').text
 328             metadata = stream_info.find('./metadata')
 329             format_info = {
 330                 'format': video_format,
 331                 'format_id': video_format,
 332                 'height': int_or_none(xpath_text(metadata, './height')),
 333                 'width': int_or_none(xpath_text(metadata, './width')),
 334             }
 335
 336             if '.fplive.net/' in video_url:
 337                 video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip())
 338                 parsed_video_url = compat_urlparse.urlparse(video_url)
 339                 direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace(
 340                     netloc='v.lvlt.crcdn.net',
 341                     path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_play_path.split(':')[-1])))
 342                 if self._is_valid_url(direct_video_url, video_id, video_format):
 343                     format_info.update({
 344                         'url': direct_video_url,
 345                     })
 346                     formats.append(format_info)
 347                     continue
 348
 349             format_info.update({
 350                 'url': video_url,
 351                 'play_path': video_play_path,
 352                 'ext': 'flv',
 353             })
 354             formats.append(format_info)
 355
 356         subtitles = self.extract_subtitles(video_id, webpage)
 357
 358         return {
 359             'id': video_id,
 360             'title': video_title,
 361             'description': video_description,
 362             'thumbnail': video_thumbnail,
 363             'uploader': video_uploader,
 364             'upload_date': video_upload_date,
 365             'subtitles': subtitles,
 366             'formats': formats,
 367         }
 368
 369
 370 class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE):
 371     IE_NAME = "crunchyroll:playlist"
 372     _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\-]+))/?(?:\?|$)'
 373
 374     _TESTS = [{
 375         'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 376         'info_dict': {
 377             'id': 'a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 378             'title': 'A Bridge to the Starry Skies - Hoshizora e Kakaru Hashi'
 379         },
 380         'playlist_count': 13,
 381     }, {
 382         # geo-restricted (US), 18+ maturity wall, non-premium available
 383         'url': 'http://www.crunchyroll.com/cosplay-complex-ova',
 384         'info_dict': {
 385             'id': 'cosplay-complex-ova',
 386             'title': 'Cosplay Complex OVA'
 387         },
 388         'playlist_count': 3,
 389         'skip': 'Georestricted',
 390     }, {
 391         # geo-restricted (US), 18+ maturity wall, non-premium will be available since 2015.11.14
 392         'url': 'http://www.crunchyroll.com/ladies-versus-butlers?skip_wall=1',
 393         'only_matching': True,
 394     }]
 395
 396     def _real_extract(self, url):
 397         show_id = self._match_id(url)
 398
 399         webpage = self._download_webpage(self._add_skip_wall(url), show_id)
 400         title = self._html_search_regex(
 401             r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>',
 402             webpage, 'title')
 403         episode_paths = re.findall(
 404             r'(?s)<li id="showview_videos_media_[0-9]+"[^>]+>.*?<a href="([^"]+)"',
 405             webpage)
 406         entries = [
 407             self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll')
 408             for ep in episode_paths
 409         ]
 410         entries.reverse()
 411
 412         return {
 413             '_type': 'playlist',
 414             'id': show_id,
 415             'title': title,
 416             'entries': entries,
 417         }