_ 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
   9 from hashlib import sha1
  10 from math import pow, sqrt, floor
  11 from .common import InfoExtractor
  12 from ..compat import (
  13     compat_etree_fromstring,
  14     compat_urllib_parse_urlencode,
  15     compat_urllib_request,
  16     compat_urlparse,
  17 )
  18 from ..utils import (
  19     ExtractorError,
  20     bytes_to_intlist,
  21     intlist_to_bytes,
  22     int_or_none,
  23     lowercase_escape,
  24     remove_end,
  25     sanitized_Request,
  26     unified_strdate,
  27     urlencode_postdata,
  28     xpath_text,
  29     extract_attributes,
  30 )
  31 from ..aes import (
  32     aes_cbc_decrypt,
  33 )
  34
  35
  36 class CrunchyrollBaseIE(InfoExtractor):
  37     _NETRC_MACHINE = 'crunchyroll'
  38
  39     def _login(self):
  40         (username, password) = self._get_login_info()
  41         if username is None:
  42             return
  43         self.report_login()
  44         login_url = 'https://www.crunchyroll.com/?a=formhandler'
  45         data = urlencode_postdata({
  46             'formname': 'RpcApiUser_Login',
  47             'name': username,
  48             'password': password,
  49         })
  50         login_request = sanitized_Request(login_url, data)
  51         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  52         self._download_webpage(login_request, None, False, 'Wrong login info')
  53
  54     def _real_initialize(self):
  55         self._login()
  56
  57     def _download_webpage(self, url_or_request, *args, **kwargs):
  58         request = (url_or_request if isinstance(url_or_request, compat_urllib_request.Request)
  59                    else sanitized_Request(url_or_request))
  60         # Accept-Language must be set explicitly to accept any language to avoid issues
  61         # similar to https://github.com/rg3/youtube-dl/issues/6797.
  62         # Along with IP address Crunchyroll uses Accept-Language to guess whether georestriction
  63         # should be imposed or not (from what I can see it just takes the first language
  64         # ignoring the priority and requires it to correspond the IP). By the way this causes
  65         # Crunchyroll to not work in georestriction cases in some browsers that don't place
  66         # the locale lang first in header. However allowing any language seems to workaround the issue.
  67         request.add_header('Accept-Language', '*')
  68         return super(CrunchyrollBaseIE, self)._download_webpage(request, *args, **kwargs)
  69
  70     @staticmethod
  71     def _add_skip_wall(url):
  72         parsed_url = compat_urlparse.urlparse(url)
  73         qs = compat_urlparse.parse_qs(parsed_url.query)
  74         # Always force skip_wall to bypass maturity wall, namely 18+ confirmation message:
  75         # > This content may be inappropriate for some people.
  76         # > Are you sure you want to continue?
  77         # since it's not disabled by default in crunchyroll account's settings.
  78         # See https://github.com/rg3/youtube-dl/issues/7202.
  79         qs['skip_wall'] = ['1']
  80         return compat_urlparse.urlunparse(
  81             parsed_url._replace(query=compat_urllib_parse_urlencode(qs, True)))
  82
  83
  84 class CrunchyrollIE(CrunchyrollBaseIE):
  85     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.(?:com|fr)/(?:media(?:-|/\?id=)|[^/]*/[^/?&]*?)(?P<video_id>[0-9]+))(?:[/?&]|$)'
  86     _TESTS = [{
  87         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
  88         'info_dict': {
  89             'id': '645513',
  90             'ext': 'flv',
  91             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
  92             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
  93             'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
  94             'uploader': 'Yomiuri Telecasting Corporation (YTV)',
  95             'upload_date': '20131013',
  96             'url': 're:(?!.*&amp)',
  97         },
  98         'params': {
  99             # rtmp
 100             'skip_download': True,
 101         },
 102     }, {
 103         'url': 'http://www.crunchyroll.com/media-589804/culture-japan-1',
 104         'info_dict': {
 105             'id': '589804',
 106             'ext': 'flv',
 107             'title': 'Culture Japan Episode 1 – Rebuilding Japan after the 3.11',
 108             'description': 'md5:2fbc01f90b87e8e9137296f37b461c12',
 109             'thumbnail': 're:^https?://.*\.jpg$',
 110             'uploader': 'Danny Choo Network',
 111             'upload_date': '20120213',
 112         },
 113         'params': {
 114             # rtmp
 115             'skip_download': True,
 116         },
 117     }, {
 118         'url': 'http://www.crunchyroll.com/rezero-starting-life-in-another-world-/episode-5-the-morning-of-our-promise-is-still-distant-702409',
 119         'info_dict': {
 120             'id': '702409',
 121             'ext': 'mp4',
 122             'title': 'Re:ZERO -Starting Life in Another World- Episode 5 – The Morning of Our Promise Is Still Distant',
 123             'description': 'md5:97664de1ab24bbf77a9c01918cb7dca9',
 124             'thumbnail': 're:^https?://.*\.jpg$',
 125             'uploader': 'TV TOKYO',
 126             'upload_date': '20160508',
 127         },
 128         'params': {
 129             # m3u8 download
 130             'skip_download': True,
 131         },
 132     }, {
 133         'url': 'http://www.crunchyroll.fr/girl-friend-beta/episode-11-goodbye-la-mode-661697',
 134         'only_matching': True,
 135     }, {
 136         # geo-restricted (US), 18+ maturity wall, non-premium available
 137         'url': 'http://www.crunchyroll.com/cosplay-complex-ova/episode-1-the-birth-of-the-cosplay-club-565617',
 138         'only_matching': True,
 139     }]
 140
 141     _FORMAT_IDS = {
 142         '360': ('60', '106'),
 143         '480': ('61', '106'),
 144         '720': ('62', '106'),
 145         '1080': ('80', '108'),
 146     }
 147
 148     def _decrypt_subtitles(self, data, iv, id):
 149         data = bytes_to_intlist(base64.b64decode(data.encode('utf-8')))
 150         iv = bytes_to_intlist(base64.b64decode(iv.encode('utf-8')))
 151         id = int(id)
 152
 153         def obfuscate_key_aux(count, modulo, start):
 154             output = list(start)
 155             for _ in range(count):
 156                 output.append(output[-1] + output[-2])
 157             # cut off start values
 158             output = output[2:]
 159             output = list(map(lambda x: x % modulo + 33, output))
 160             return output
 161
 162         def obfuscate_key(key):
 163             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
 164             num2 = (num1 ^ key) << 5
 165             num3 = key ^ num1
 166             num4 = num3 ^ (num3 >> 3) ^ num2
 167             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
 168             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
 169             # Extend 160 Bit hash to 256 Bit
 170             return shaHash + [0] * 12
 171
 172         key = obfuscate_key(id)
 173
 174         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
 175         return zlib.decompress(decrypted_data)
 176
 177     def _convert_subtitles_to_srt(self, sub_root):
 178         output = ''
 179
 180         for i, event in enumerate(sub_root.findall('./events/event'), 1):
 181             start = event.attrib['start'].replace('.', ',')
 182             end = event.attrib['end'].replace('.', ',')
 183             text = event.attrib['text'].replace('\\N', '\n')
 184             output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
 185         return output
 186
 187     def _convert_subtitles_to_ass(self, sub_root):
 188         output = ''
 189
 190         def ass_bool(strvalue):
 191             assvalue = '0'
 192             if strvalue == '1':
 193                 assvalue = '-1'
 194             return assvalue
 195
 196         output = '[Script Info]\n'
 197         output += 'Title: %s\n' % sub_root.attrib['title']
 198         output += 'ScriptType: v4.00+\n'
 199         output += 'WrapStyle: %s\n' % sub_root.attrib['wrap_style']
 200         output += 'PlayResX: %s\n' % sub_root.attrib['play_res_x']
 201         output += 'PlayResY: %s\n' % sub_root.attrib['play_res_y']
 202         output += """ScaledBorderAndShadow: yes
 203
 204 [V4+ Styles]
 205 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
 206 """
 207         for style in sub_root.findall('./styles/style'):
 208             output += 'Style: ' + style.attrib['name']
 209             output += ',' + style.attrib['font_name']
 210             output += ',' + style.attrib['font_size']
 211             output += ',' + style.attrib['primary_colour']
 212             output += ',' + style.attrib['secondary_colour']
 213             output += ',' + style.attrib['outline_colour']
 214             output += ',' + style.attrib['back_colour']
 215             output += ',' + ass_bool(style.attrib['bold'])
 216             output += ',' + ass_bool(style.attrib['italic'])
 217             output += ',' + ass_bool(style.attrib['underline'])
 218             output += ',' + ass_bool(style.attrib['strikeout'])
 219             output += ',' + style.attrib['scale_x']
 220             output += ',' + style.attrib['scale_y']
 221             output += ',' + style.attrib['spacing']
 222             output += ',' + style.attrib['angle']
 223             output += ',' + style.attrib['border_style']
 224             output += ',' + style.attrib['outline']
 225             output += ',' + style.attrib['shadow']
 226             output += ',' + style.attrib['alignment']
 227             output += ',' + style.attrib['margin_l']
 228             output += ',' + style.attrib['margin_r']
 229             output += ',' + style.attrib['margin_v']
 230             output += ',' + style.attrib['encoding']
 231             output += '\n'
 232
 233         output += """
 234 [Events]
 235 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
 236 """
 237         for event in sub_root.findall('./events/event'):
 238             output += 'Dialogue: 0'
 239             output += ',' + event.attrib['start']
 240             output += ',' + event.attrib['end']
 241             output += ',' + event.attrib['style']
 242             output += ',' + event.attrib['name']
 243             output += ',' + event.attrib['margin_l']
 244             output += ',' + event.attrib['margin_r']
 245             output += ',' + event.attrib['margin_v']
 246             output += ',' + event.attrib['effect']
 247             output += ',' + event.attrib['text']
 248             output += '\n'
 249
 250         return output
 251
 252     def _extract_subtitles(self, subtitle):
 253         sub_root = compat_etree_fromstring(subtitle)
 254         return [{
 255             'ext': 'srt',
 256             'data': self._convert_subtitles_to_srt(sub_root),
 257         }, {
 258             'ext': 'ass',
 259             'data': self._convert_subtitles_to_ass(sub_root),
 260         }]
 261
 262     def _get_subtitles(self, video_id, webpage):
 263         subtitles = {}
 264         for sub_id, sub_name in re.findall(r'\bssid=([0-9]+)"[^>]+?\btitle="([^"]+)', webpage):
 265             sub_page = self._download_webpage(
 266                 'http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id=' + sub_id,
 267                 video_id, note='Downloading subtitles for ' + sub_name)
 268             id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
 269             iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
 270             data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
 271             if not id or not iv or not data:
 272                 continue
 273             subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
 274             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 275             if not lang_code:
 276                 continue
 277             subtitles[lang_code] = self._extract_subtitles(subtitle)
 278         return subtitles
 279
 280     def _real_extract(self, url):
 281         mobj = re.match(self._VALID_URL, url)
 282         video_id = mobj.group('video_id')
 283
 284         if mobj.group('prefix') == 'm':
 285             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 286             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 287         else:
 288             webpage_url = 'http://www.' + mobj.group('url')
 289
 290         webpage = self._download_webpage(self._add_skip_wall(webpage_url), video_id, 'Downloading webpage')
 291         note_m = self._html_search_regex(
 292             r'<div class="showmedia-trailer-notice">(.+?)</div>',
 293             webpage, 'trailer-notice', default='')
 294         if note_m:
 295             raise ExtractorError(note_m)
 296
 297         mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
 298         if mobj:
 299             msg = json.loads(mobj.group('msg'))
 300             if msg.get('type') == 'error':
 301                 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
 302
 303         if 'To view this, please log in to verify you are 18 or older.' in webpage:
 304             self.raise_login_required()
 305
 306         video_title = self._html_search_regex(
 307             r'(?s)<h1[^>]*>((?:(?!<h1).)*?<span[^>]+itemprop=["\']title["\'][^>]*>(?:(?!<h1).)+?)</h1>',
 308             webpage, 'video_title')
 309         video_title = re.sub(r' {2,}', ' ', video_title)
 310         video_description = self._html_search_regex(
 311             r'<script[^>]*>\s*.+?\[media_id=%s\].+?"description"\s*:\s*"([^"]+)' % video_id,
 312             webpage, 'description', default=None)
 313         if video_description:
 314             video_description = lowercase_escape(video_description.replace(r'\r\n', '\n'))
 315         video_upload_date = self._html_search_regex(
 316             [r'<div>Availability for free users:(.+?)</div>', r'<div>[^<>]+<span>\s*(.+?\d{4})\s*</span></div>'],
 317             webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
 318         if video_upload_date:
 319             video_upload_date = unified_strdate(video_upload_date)
 320         video_uploader = self._html_search_regex(
 321             r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', webpage,
 322             'video_uploader', fatal=False)
 323
 324         available_fmts = []
 325         for a, fmt in re.findall(r'(<a[^>]+token=["\']showmedia\.([0-9]{3,4})p["\'][^>]+>)', webpage):
 326             attrs = extract_attributes(a)
 327             href = attrs.get('href')
 328             if href and '/freetrial' in href:
 329                 continue
 330             available_fmts.append(fmt)
 331         if not available_fmts:
 332             for p in (r'token=["\']showmedia\.([0-9]{3,4})p"', r'showmedia\.([0-9]{3,4})p'):
 333                 available_fmts = re.findall(p, webpage)
 334                 if available_fmts:
 335                     break
 336         video_encode_ids = []
 337         formats = []
 338         for fmt in available_fmts:
 339             stream_quality, stream_format = self._FORMAT_IDS[fmt]
 340             video_format = fmt + 'p'
 341             streamdata_req = sanitized_Request(
 342                 'http://www.crunchyroll.com/xml/?req=RpcApiVideoPlayer_GetStandardConfig&media_id=%s&video_format=%s&video_quality=%s'
 343                 % (video_id, stream_format, stream_quality),
 344                 compat_urllib_parse_urlencode({'current_page': url}).encode('utf-8'))
 345             streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 346             streamdata = self._download_xml(
 347                 streamdata_req, video_id,
 348                 note='Downloading media info for %s' % video_format)
 349             stream_info = streamdata.find('./{default}preload/stream_info')
 350             video_encode_id = xpath_text(stream_info, './video_encode_id')
 351             if video_encode_id in video_encode_ids:
 352                 continue
 353             video_encode_ids.append(video_encode_id)
 354
 355             video_file = xpath_text(stream_info, './file')
 356             if not video_file:
 357                 continue
 358             if video_file.startswith('http'):
 359                 formats.extend(self._extract_m3u8_formats(
 360                     video_file, video_id, 'mp4', entry_protocol='m3u8_native',
 361                     m3u8_id='hls', fatal=False))
 362                 continue
 363
 364             video_url = xpath_text(stream_info, './host')
 365             if not video_url:
 366                 continue
 367             metadata = stream_info.find('./metadata')
 368             format_info = {
 369                 'format': video_format,
 370                 'format_id': video_format,
 371                 'height': int_or_none(xpath_text(metadata, './height')),
 372                 'width': int_or_none(xpath_text(metadata, './width')),
 373             }
 374
 375             if '.fplive.net/' in video_url:
 376                 video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip())
 377                 parsed_video_url = compat_urlparse.urlparse(video_url)
 378                 direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace(
 379                     netloc='v.lvlt.crcdn.net',
 380                     path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_file.split(':')[-1])))
 381                 if self._is_valid_url(direct_video_url, video_id, video_format):
 382                     format_info.update({
 383                         'url': direct_video_url,
 384                     })
 385                     formats.append(format_info)
 386                     continue
 387
 388             format_info.update({
 389                 'url': video_url,
 390                 'play_path': video_file,
 391                 'ext': 'flv',
 392             })
 393             formats.append(format_info)
 394         self._sort_formats(formats)
 395
 396         metadata = self._download_xml(
 397             'http://www.crunchyroll.com/xml', video_id,
 398             note='Downloading media info', query={
 399                 'req': 'RpcApiVideoPlayer_GetMediaMetadata',
 400                 'media_id': video_id,
 401             })
 402
 403         subtitles = self.extract_subtitles(video_id, webpage)
 404
 405         return {
 406             'id': video_id,
 407             'title': video_title,
 408             'description': video_description,
 409             'thumbnail': xpath_text(metadata, 'episode_image_url'),
 410             'uploader': video_uploader,
 411             'upload_date': video_upload_date,
 412             'series': xpath_text(metadata, 'series_title'),
 413             'episode': xpath_text(metadata, 'episode_title'),
 414             'episode_number': int_or_none(xpath_text(metadata, 'episode_number')),
 415             'subtitles': subtitles,
 416             'formats': formats,
 417         }
 418
 419
 420 class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE):
 421     IE_NAME = 'crunchyroll:playlist'
 422     _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\-]+))/?(?:\?|$)'
 423
 424     _TESTS = [{
 425         'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 426         'info_dict': {
 427             'id': 'a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 428             'title': 'A Bridge to the Starry Skies - Hoshizora e Kakaru Hashi'
 429         },
 430         'playlist_count': 13,
 431     }, {
 432         # geo-restricted (US), 18+ maturity wall, non-premium available
 433         'url': 'http://www.crunchyroll.com/cosplay-complex-ova',
 434         'info_dict': {
 435             'id': 'cosplay-complex-ova',
 436             'title': 'Cosplay Complex OVA'
 437         },
 438         'playlist_count': 3,
 439         'skip': 'Georestricted',
 440     }, {
 441         # geo-restricted (US), 18+ maturity wall, non-premium will be available since 2015.11.14
 442         'url': 'http://www.crunchyroll.com/ladies-versus-butlers?skip_wall=1',
 443         'only_matching': True,
 444     }]
 445
 446     def _real_extract(self, url):
 447         show_id = self._match_id(url)
 448
 449         webpage = self._download_webpage(self._add_skip_wall(url), show_id)
 450         title = self._html_search_regex(
 451             r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>',
 452             webpage, 'title')
 453         episode_paths = re.findall(
 454             r'(?s)<li id="showview_videos_media_[0-9]+"[^>]+>.*?<a href="([^"]+)"',
 455             webpage)
 456         entries = [
 457             self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll')
 458             for ep in episode_paths
 459         ]
 460         entries.reverse()
 461
 462         return {
 463             '_type': 'playlist',
 464             'id': show_id,
 465             'title': title,
 466             'entries': entries,
 467         }