_ 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     _LOGIN_URL = 'https://www.crunchyroll.com/login'
  38     _LOGIN_FORM = 'login_form'
  39     _NETRC_MACHINE = 'crunchyroll'
  40
  41     def _login(self):
  42         (username, password) = self._get_login_info()
  43         if username is None:
  44             return
  45
  46         login_page = self._download_webpage(
  47             self._LOGIN_URL, None, 'Downloading login page')
  48
  49         login_form_str = self._search_regex(
  50             r'(?P<form><form[^>]+?id=(["\'])%s\2[^>]*>)' % self._LOGIN_FORM,
  51             login_page, 'login form', group='form')
  52
  53         post_url = extract_attributes(login_form_str).get('action')
  54         if not post_url:
  55             post_url = self._LOGIN_URL
  56         elif not post_url.startswith('http'):
  57             post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
  58
  59         login_form = self._form_hidden_inputs(self._LOGIN_FORM, login_page)
  60
  61         login_form.update({
  62             'login_form[name]': username,
  63             'login_form[password]': password,
  64         })
  65
  66         response = self._download_webpage(
  67             post_url, None, 'Logging in', 'Wrong login info',
  68             data=urlencode_postdata(login_form),
  69             headers={'Content-Type': 'application/x-www-form-urlencoded'})
  70
  71         # Successful login
  72         if '<title>Redirecting' in response:
  73             return
  74
  75         error = self._html_search_regex(
  76             '(?s)<ul[^>]+class=["\']messages["\'][^>]*>(.+?)</ul>',
  77             response, 'error message', default=None)
  78         if error:
  79             raise ExtractorError('Unable to login: %s' % error, expected=True)
  80
  81         raise ExtractorError('Unable to log in')
  82
  83     def _real_initialize(self):
  84         self._login()
  85
  86     def _download_webpage(self, url_or_request, *args, **kwargs):
  87         request = (url_or_request if isinstance(url_or_request, compat_urllib_request.Request)
  88                    else sanitized_Request(url_or_request))
  89         # Accept-Language must be set explicitly to accept any language to avoid issues
  90         # similar to https://github.com/rg3/youtube-dl/issues/6797.
  91         # Along with IP address Crunchyroll uses Accept-Language to guess whether georestriction
  92         # should be imposed or not (from what I can see it just takes the first language
  93         # ignoring the priority and requires it to correspond the IP). By the way this causes
  94         # Crunchyroll to not work in georestriction cases in some browsers that don't place
  95         # the locale lang first in header. However allowing any language seems to workaround the issue.
  96         request.add_header('Accept-Language', '*')
  97         return super(CrunchyrollBaseIE, self)._download_webpage(request, *args, **kwargs)
  98
  99     @staticmethod
 100     def _add_skip_wall(url):
 101         parsed_url = compat_urlparse.urlparse(url)
 102         qs = compat_urlparse.parse_qs(parsed_url.query)
 103         # Always force skip_wall to bypass maturity wall, namely 18+ confirmation message:
 104         # > This content may be inappropriate for some people.
 105         # > Are you sure you want to continue?
 106         # since it's not disabled by default in crunchyroll account's settings.
 107         # See https://github.com/rg3/youtube-dl/issues/7202.
 108         qs['skip_wall'] = ['1']
 109         return compat_urlparse.urlunparse(
 110             parsed_url._replace(query=compat_urllib_parse_urlencode(qs, True)))
 111
 112
 113 class CrunchyrollIE(CrunchyrollBaseIE):
 114     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.(?:com|fr)/(?:media(?:-|/\?id=)|[^/]*/[^/?&]*?)(?P<video_id>[0-9]+))(?:[/?&]|$)'
 115     _TESTS = [{
 116         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
 117         'info_dict': {
 118             'id': '645513',
 119             'ext': 'flv',
 120             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
 121             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
 122             'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
 123             'uploader': 'Yomiuri Telecasting Corporation (YTV)',
 124             'upload_date': '20131013',
 125             'url': 're:(?!.*&amp)',
 126         },
 127         'params': {
 128             # rtmp
 129             'skip_download': True,
 130         },
 131     }, {
 132         'url': 'http://www.crunchyroll.com/media-589804/culture-japan-1',
 133         'info_dict': {
 134             'id': '589804',
 135             'ext': 'flv',
 136             'title': 'Culture Japan Episode 1 – Rebuilding Japan after the 3.11',
 137             'description': 'md5:2fbc01f90b87e8e9137296f37b461c12',
 138             'thumbnail': 're:^https?://.*\.jpg$',
 139             'uploader': 'Danny Choo Network',
 140             'upload_date': '20120213',
 141         },
 142         'params': {
 143             # rtmp
 144             'skip_download': True,
 145         },
 146     }, {
 147         'url': 'http://www.crunchyroll.com/rezero-starting-life-in-another-world-/episode-5-the-morning-of-our-promise-is-still-distant-702409',
 148         'info_dict': {
 149             'id': '702409',
 150             'ext': 'mp4',
 151             'title': 'Re:ZERO -Starting Life in Another World- Episode 5 – The Morning of Our Promise Is Still Distant',
 152             'description': 'md5:97664de1ab24bbf77a9c01918cb7dca9',
 153             'thumbnail': 're:^https?://.*\.jpg$',
 154             'uploader': 'TV TOKYO',
 155             'upload_date': '20160508',
 156         },
 157         'params': {
 158             # m3u8 download
 159             'skip_download': True,
 160         },
 161     }, {
 162         'url': 'http://www.crunchyroll.fr/girl-friend-beta/episode-11-goodbye-la-mode-661697',
 163         'only_matching': True,
 164     }, {
 165         # geo-restricted (US), 18+ maturity wall, non-premium available
 166         'url': 'http://www.crunchyroll.com/cosplay-complex-ova/episode-1-the-birth-of-the-cosplay-club-565617',
 167         'only_matching': True,
 168     }]
 169
 170     _FORMAT_IDS = {
 171         '360': ('60', '106'),
 172         '480': ('61', '106'),
 173         '720': ('62', '106'),
 174         '1080': ('80', '108'),
 175     }
 176
 177     def _decrypt_subtitles(self, data, iv, id):
 178         data = bytes_to_intlist(base64.b64decode(data.encode('utf-8')))
 179         iv = bytes_to_intlist(base64.b64decode(iv.encode('utf-8')))
 180         id = int(id)
 181
 182         def obfuscate_key_aux(count, modulo, start):
 183             output = list(start)
 184             for _ in range(count):
 185                 output.append(output[-1] + output[-2])
 186             # cut off start values
 187             output = output[2:]
 188             output = list(map(lambda x: x % modulo + 33, output))
 189             return output
 190
 191         def obfuscate_key(key):
 192             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
 193             num2 = (num1 ^ key) << 5
 194             num3 = key ^ num1
 195             num4 = num3 ^ (num3 >> 3) ^ num2
 196             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
 197             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
 198             # Extend 160 Bit hash to 256 Bit
 199             return shaHash + [0] * 12
 200
 201         key = obfuscate_key(id)
 202
 203         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
 204         return zlib.decompress(decrypted_data)
 205
 206     def _convert_subtitles_to_srt(self, sub_root):
 207         output = ''
 208
 209         for i, event in enumerate(sub_root.findall('./events/event'), 1):
 210             start = event.attrib['start'].replace('.', ',')
 211             end = event.attrib['end'].replace('.', ',')
 212             text = event.attrib['text'].replace('\\N', '\n')
 213             output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
 214         return output
 215
 216     def _convert_subtitles_to_ass(self, sub_root):
 217         output = ''
 218
 219         def ass_bool(strvalue):
 220             assvalue = '0'
 221             if strvalue == '1':
 222                 assvalue = '-1'
 223             return assvalue
 224
 225         output = '[Script Info]\n'
 226         output += 'Title: %s\n' % sub_root.attrib['title']
 227         output += 'ScriptType: v4.00+\n'
 228         output += 'WrapStyle: %s\n' % sub_root.attrib['wrap_style']
 229         output += 'PlayResX: %s\n' % sub_root.attrib['play_res_x']
 230         output += 'PlayResY: %s\n' % sub_root.attrib['play_res_y']
 231         output += """ScaledBorderAndShadow: yes
 232
 233 [V4+ Styles]
 234 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
 235 """
 236         for style in sub_root.findall('./styles/style'):
 237             output += 'Style: ' + style.attrib['name']
 238             output += ',' + style.attrib['font_name']
 239             output += ',' + style.attrib['font_size']
 240             output += ',' + style.attrib['primary_colour']
 241             output += ',' + style.attrib['secondary_colour']
 242             output += ',' + style.attrib['outline_colour']
 243             output += ',' + style.attrib['back_colour']
 244             output += ',' + ass_bool(style.attrib['bold'])
 245             output += ',' + ass_bool(style.attrib['italic'])
 246             output += ',' + ass_bool(style.attrib['underline'])
 247             output += ',' + ass_bool(style.attrib['strikeout'])
 248             output += ',' + style.attrib['scale_x']
 249             output += ',' + style.attrib['scale_y']
 250             output += ',' + style.attrib['spacing']
 251             output += ',' + style.attrib['angle']
 252             output += ',' + style.attrib['border_style']
 253             output += ',' + style.attrib['outline']
 254             output += ',' + style.attrib['shadow']
 255             output += ',' + style.attrib['alignment']
 256             output += ',' + style.attrib['margin_l']
 257             output += ',' + style.attrib['margin_r']
 258             output += ',' + style.attrib['margin_v']
 259             output += ',' + style.attrib['encoding']
 260             output += '\n'
 261
 262         output += """
 263 [Events]
 264 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
 265 """
 266         for event in sub_root.findall('./events/event'):
 267             output += 'Dialogue: 0'
 268             output += ',' + event.attrib['start']
 269             output += ',' + event.attrib['end']
 270             output += ',' + event.attrib['style']
 271             output += ',' + event.attrib['name']
 272             output += ',' + event.attrib['margin_l']
 273             output += ',' + event.attrib['margin_r']
 274             output += ',' + event.attrib['margin_v']
 275             output += ',' + event.attrib['effect']
 276             output += ',' + event.attrib['text']
 277             output += '\n'
 278
 279         return output
 280
 281     def _extract_subtitles(self, subtitle):
 282         sub_root = compat_etree_fromstring(subtitle)
 283         return [{
 284             'ext': 'srt',
 285             'data': self._convert_subtitles_to_srt(sub_root),
 286         }, {
 287             'ext': 'ass',
 288             'data': self._convert_subtitles_to_ass(sub_root),
 289         }]
 290
 291     def _get_subtitles(self, video_id, webpage):
 292         subtitles = {}
 293         for sub_id, sub_name in re.findall(r'\bssid=([0-9]+)"[^>]+?\btitle="([^"]+)', webpage):
 294             sub_page = self._download_webpage(
 295                 'http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id=' + sub_id,
 296                 video_id, note='Downloading subtitles for ' + sub_name)
 297             id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
 298             iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
 299             data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
 300             if not id or not iv or not data:
 301                 continue
 302             subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
 303             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 304             if not lang_code:
 305                 continue
 306             subtitles[lang_code] = self._extract_subtitles(subtitle)
 307         return subtitles
 308
 309     def _real_extract(self, url):
 310         mobj = re.match(self._VALID_URL, url)
 311         video_id = mobj.group('video_id')
 312
 313         if mobj.group('prefix') == 'm':
 314             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 315             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 316         else:
 317             webpage_url = 'http://www.' + mobj.group('url')
 318
 319         webpage = self._download_webpage(self._add_skip_wall(webpage_url), video_id, 'Downloading webpage')
 320         note_m = self._html_search_regex(
 321             r'<div class="showmedia-trailer-notice">(.+?)</div>',
 322             webpage, 'trailer-notice', default='')
 323         if note_m:
 324             raise ExtractorError(note_m)
 325
 326         mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
 327         if mobj:
 328             msg = json.loads(mobj.group('msg'))
 329             if msg.get('type') == 'error':
 330                 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
 331
 332         if 'To view this, please log in to verify you are 18 or older.' in webpage:
 333             self.raise_login_required()
 334
 335         video_title = self._html_search_regex(
 336             r'(?s)<h1[^>]*>((?:(?!<h1).)*?<span[^>]+itemprop=["\']title["\'][^>]*>(?:(?!<h1).)+?)</h1>',
 337             webpage, 'video_title')
 338         video_title = re.sub(r' {2,}', ' ', video_title)
 339         video_description = self._html_search_regex(
 340             r'<script[^>]*>\s*.+?\[media_id=%s\].+?"description"\s*:\s*"([^"]+)' % video_id,
 341             webpage, 'description', default=None)
 342         if video_description:
 343             video_description = lowercase_escape(video_description.replace(r'\r\n', '\n'))
 344         video_upload_date = self._html_search_regex(
 345             [r'<div>Availability for free users:(.+?)</div>', r'<div>[^<>]+<span>\s*(.+?\d{4})\s*</span></div>'],
 346             webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
 347         if video_upload_date:
 348             video_upload_date = unified_strdate(video_upload_date)
 349         video_uploader = self._html_search_regex(
 350             r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', webpage,
 351             'video_uploader', fatal=False)
 352
 353         available_fmts = []
 354         for a, fmt in re.findall(r'(<a[^>]+token=["\']showmedia\.([0-9]{3,4})p["\'][^>]+>)', webpage):
 355             attrs = extract_attributes(a)
 356             href = attrs.get('href')
 357             if href and '/freetrial' in href:
 358                 continue
 359             available_fmts.append(fmt)
 360         if not available_fmts:
 361             for p in (r'token=["\']showmedia\.([0-9]{3,4})p"', r'showmedia\.([0-9]{3,4})p'):
 362                 available_fmts = re.findall(p, webpage)
 363                 if available_fmts:
 364                     break
 365         video_encode_ids = []
 366         formats = []
 367         for fmt in available_fmts:
 368             stream_quality, stream_format = self._FORMAT_IDS[fmt]
 369             video_format = fmt + 'p'
 370             streamdata_req = sanitized_Request(
 371                 'http://www.crunchyroll.com/xml/?req=RpcApiVideoPlayer_GetStandardConfig&media_id=%s&video_format=%s&video_quality=%s'
 372                 % (video_id, stream_format, stream_quality),
 373                 compat_urllib_parse_urlencode({'current_page': url}).encode('utf-8'))
 374             streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 375             streamdata = self._download_xml(
 376                 streamdata_req, video_id,
 377                 note='Downloading media info for %s' % video_format)
 378             stream_info = streamdata.find('./{default}preload/stream_info')
 379             video_encode_id = xpath_text(stream_info, './video_encode_id')
 380             if video_encode_id in video_encode_ids:
 381                 continue
 382             video_encode_ids.append(video_encode_id)
 383
 384             video_file = xpath_text(stream_info, './file')
 385             if not video_file:
 386                 continue
 387             if video_file.startswith('http'):
 388                 formats.extend(self._extract_m3u8_formats(
 389                     video_file, video_id, 'mp4', entry_protocol='m3u8_native',
 390                     m3u8_id='hls', fatal=False))
 391                 continue
 392
 393             video_url = xpath_text(stream_info, './host')
 394             if not video_url:
 395                 continue
 396             metadata = stream_info.find('./metadata')
 397             format_info = {
 398                 'format': video_format,
 399                 'format_id': video_format,
 400                 'height': int_or_none(xpath_text(metadata, './height')),
 401                 'width': int_or_none(xpath_text(metadata, './width')),
 402             }
 403
 404             if '.fplive.net/' in video_url:
 405                 video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip())
 406                 parsed_video_url = compat_urlparse.urlparse(video_url)
 407                 direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace(
 408                     netloc='v.lvlt.crcdn.net',
 409                     path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_file.split(':')[-1])))
 410                 if self._is_valid_url(direct_video_url, video_id, video_format):
 411                     format_info.update({
 412                         'url': direct_video_url,
 413                     })
 414                     formats.append(format_info)
 415                     continue
 416
 417             format_info.update({
 418                 'url': video_url,
 419                 'play_path': video_file,
 420                 'ext': 'flv',
 421             })
 422             formats.append(format_info)
 423         self._sort_formats(formats)
 424
 425         metadata = self._download_xml(
 426             'http://www.crunchyroll.com/xml', video_id,
 427             note='Downloading media info', query={
 428                 'req': 'RpcApiVideoPlayer_GetMediaMetadata',
 429                 'media_id': video_id,
 430             })
 431
 432         subtitles = self.extract_subtitles(video_id, webpage)
 433
 434         return {
 435             'id': video_id,
 436             'title': video_title,
 437             'description': video_description,
 438             'thumbnail': xpath_text(metadata, 'episode_image_url'),
 439             'uploader': video_uploader,
 440             'upload_date': video_upload_date,
 441             'series': xpath_text(metadata, 'series_title'),
 442             'episode': xpath_text(metadata, 'episode_title'),
 443             'episode_number': int_or_none(xpath_text(metadata, 'episode_number')),
 444             'subtitles': subtitles,
 445             'formats': formats,
 446         }
 447
 448
 449 class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE):
 450     IE_NAME = 'crunchyroll:playlist'
 451     _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\-]+))/?(?:\?|$)'
 452
 453     _TESTS = [{
 454         'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 455         'info_dict': {
 456             'id': 'a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 457             'title': 'A Bridge to the Starry Skies - Hoshizora e Kakaru Hashi'
 458         },
 459         'playlist_count': 13,
 460     }, {
 461         # geo-restricted (US), 18+ maturity wall, non-premium available
 462         'url': 'http://www.crunchyroll.com/cosplay-complex-ova',
 463         'info_dict': {
 464             'id': 'cosplay-complex-ova',
 465             'title': 'Cosplay Complex OVA'
 466         },
 467         'playlist_count': 3,
 468         'skip': 'Georestricted',
 469     }, {
 470         # geo-restricted (US), 18+ maturity wall, non-premium will be available since 2015.11.14
 471         'url': 'http://www.crunchyroll.com/ladies-versus-butlers?skip_wall=1',
 472         'only_matching': True,
 473     }]
 474
 475     def _real_extract(self, url):
 476         show_id = self._match_id(url)
 477
 478         webpage = self._download_webpage(self._add_skip_wall(url), show_id)
 479         title = self._html_search_regex(
 480             r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>',
 481             webpage, 'title')
 482         episode_paths = re.findall(
 483             r'(?s)<li id="showview_videos_media_[0-9]+"[^>]+>.*?<a href="([^"]+)"',
 484             webpage)
 485         entries = [
 486             self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll')
 487             for ep in episode_paths
 488         ]
 489         entries.reverse()
 490
 491         return {
 492             '_type': 'playlist',
 493             'id': show_id,
 494             'title': title,
 495             'entries': entries,
 496         }