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
   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,
  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     lowercase_escape,
  25     remove_end,
  26     unified_strdate,
  27     urlencode_postdata,
  28     xpath_text,
  29 )
  30 from ..aes import (
  31     aes_cbc_decrypt,
  32 )
  33
  34
  35 class CrunchyrollBaseIE(InfoExtractor):
  36     _NETRC_MACHINE = 'crunchyroll'
  37
  38     def _login(self):
  39         (username, password) = self._get_login_info()
  40         if username is None:
  41             return
  42         self.report_login()
  43         login_url = 'https://www.crunchyroll.com/?a=formhandler'
  44         data = urlencode_postdata({
  45             'formname': 'RpcApiUser_Login',
  46             'name': username,
  47             'password': password,
  48         })
  49         login_request = compat_urllib_request.Request(login_url, data)
  50         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  51         self._download_webpage(login_request, None, False, 'Wrong login info')
  52
  53     def _real_initialize(self):
  54         self._login()
  55
  56     def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None):
  57         request = (url_or_request if isinstance(url_or_request, compat_urllib_request.Request)
  58                    else compat_urllib_request.Request(url_or_request))
  59         # Accept-Language must be set explicitly to accept any language to avoid issues
  60         # similar to https://github.com/rg3/youtube-dl/issues/6797.
  61         # Along with IP address Crunchyroll uses Accept-Language to guess whether georestriction
  62         # should be imposed or not (from what I can see it just takes the first language
  63         # ignoring the priority and requires it to correspond the IP). By the way this causes
  64         # Crunchyroll to not work in georestriction cases in some browsers that don't place
  65         # the locale lang first in header. However allowing any language seems to workaround the issue.
  66         request.add_header('Accept-Language', '*')
  67         return super(CrunchyrollBaseIE, self)._download_webpage(
  68             request, video_id, note, errnote, fatal, tries, timeout, encoding)
  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.fr/girl-friend-beta/episode-11-goodbye-la-mode-661697',
 119         'only_matching': True,
 120     }, {
 121         # geo-restricted (US), 18+ maturity wall, non-premium available
 122         'url': 'http://www.crunchyroll.com/cosplay-complex-ova/episode-1-the-birth-of-the-cosplay-club-565617',
 123         'only_matching': True,
 124     }]
 125
 126     _FORMAT_IDS = {
 127         '360': ('60', '106'),
 128         '480': ('61', '106'),
 129         '720': ('62', '106'),
 130         '1080': ('80', '108'),
 131     }
 132
 133     def _decrypt_subtitles(self, data, iv, id):
 134         data = bytes_to_intlist(base64.b64decode(data.encode('utf-8')))
 135         iv = bytes_to_intlist(base64.b64decode(iv.encode('utf-8')))
 136         id = int(id)
 137
 138         def obfuscate_key_aux(count, modulo, start):
 139             output = list(start)
 140             for _ in range(count):
 141                 output.append(output[-1] + output[-2])
 142             # cut off start values
 143             output = output[2:]
 144             output = list(map(lambda x: x % modulo + 33, output))
 145             return output
 146
 147         def obfuscate_key(key):
 148             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
 149             num2 = (num1 ^ key) << 5
 150             num3 = key ^ num1
 151             num4 = num3 ^ (num3 >> 3) ^ num2
 152             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
 153             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
 154             # Extend 160 Bit hash to 256 Bit
 155             return shaHash + [0] * 12
 156
 157         key = obfuscate_key(id)
 158
 159         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
 160         return zlib.decompress(decrypted_data)
 161
 162     def _convert_subtitles_to_srt(self, sub_root):
 163         output = ''
 164
 165         for i, event in enumerate(sub_root.findall('./events/event'), 1):
 166             start = event.attrib['start'].replace('.', ',')
 167             end = event.attrib['end'].replace('.', ',')
 168             text = event.attrib['text'].replace('\\N', '\n')
 169             output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
 170         return output
 171
 172     def _convert_subtitles_to_ass(self, sub_root):
 173         output = ''
 174
 175         def ass_bool(strvalue):
 176             assvalue = '0'
 177             if strvalue == '1':
 178                 assvalue = '-1'
 179             return assvalue
 180
 181         output = '[Script Info]\n'
 182         output += 'Title: %s\n' % sub_root.attrib["title"]
 183         output += 'ScriptType: v4.00+\n'
 184         output += 'WrapStyle: %s\n' % sub_root.attrib["wrap_style"]
 185         output += 'PlayResX: %s\n' % sub_root.attrib["play_res_x"]
 186         output += 'PlayResY: %s\n' % sub_root.attrib["play_res_y"]
 187         output += """ScaledBorderAndShadow: yes
 188
 189 [V4+ Styles]
 190 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
 191 """
 192         for style in sub_root.findall('./styles/style'):
 193             output += 'Style: ' + style.attrib["name"]
 194             output += ',' + style.attrib["font_name"]
 195             output += ',' + style.attrib["font_size"]
 196             output += ',' + style.attrib["primary_colour"]
 197             output += ',' + style.attrib["secondary_colour"]
 198             output += ',' + style.attrib["outline_colour"]
 199             output += ',' + style.attrib["back_colour"]
 200             output += ',' + ass_bool(style.attrib["bold"])
 201             output += ',' + ass_bool(style.attrib["italic"])
 202             output += ',' + ass_bool(style.attrib["underline"])
 203             output += ',' + ass_bool(style.attrib["strikeout"])
 204             output += ',' + style.attrib["scale_x"]
 205             output += ',' + style.attrib["scale_y"]
 206             output += ',' + style.attrib["spacing"]
 207             output += ',' + style.attrib["angle"]
 208             output += ',' + style.attrib["border_style"]
 209             output += ',' + style.attrib["outline"]
 210             output += ',' + style.attrib["shadow"]
 211             output += ',' + style.attrib["alignment"]
 212             output += ',' + style.attrib["margin_l"]
 213             output += ',' + style.attrib["margin_r"]
 214             output += ',' + style.attrib["margin_v"]
 215             output += ',' + style.attrib["encoding"]
 216             output += '\n'
 217
 218         output += """
 219 [Events]
 220 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
 221 """
 222         for event in sub_root.findall('./events/event'):
 223             output += 'Dialogue: 0'
 224             output += ',' + event.attrib["start"]
 225             output += ',' + event.attrib["end"]
 226             output += ',' + event.attrib["style"]
 227             output += ',' + event.attrib["name"]
 228             output += ',' + event.attrib["margin_l"]
 229             output += ',' + event.attrib["margin_r"]
 230             output += ',' + event.attrib["margin_v"]
 231             output += ',' + event.attrib["effect"]
 232             output += ',' + event.attrib["text"]
 233             output += '\n'
 234
 235         return output
 236
 237     def _extract_subtitles(self, subtitle):
 238         sub_root = compat_etree_fromstring(subtitle)
 239         return [{
 240             'ext': 'srt',
 241             'data': self._convert_subtitles_to_srt(sub_root),
 242         }, {
 243             'ext': 'ass',
 244             'data': self._convert_subtitles_to_ass(sub_root),
 245         }]
 246
 247     def _get_subtitles(self, video_id, webpage):
 248         subtitles = {}
 249         for sub_id, sub_name in re.findall(r'\bssid=([0-9]+)"[^>]+?\btitle="([^"]+)', webpage):
 250             sub_page = self._download_webpage(
 251                 'http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id=' + sub_id,
 252                 video_id, note='Downloading subtitles for ' + sub_name)
 253             id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
 254             iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
 255             data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
 256             if not id or not iv or not data:
 257                 continue
 258             subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
 259             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 260             if not lang_code:
 261                 continue
 262             subtitles[lang_code] = self._extract_subtitles(subtitle)
 263         return subtitles
 264
 265     def _real_extract(self, url):
 266         mobj = re.match(self._VALID_URL, url)
 267         video_id = mobj.group('video_id')
 268
 269         if mobj.group('prefix') == 'm':
 270             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 271             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 272         else:
 273             webpage_url = 'http://www.' + mobj.group('url')
 274
 275         webpage = self._download_webpage(self._add_skip_wall(webpage_url), video_id, 'Downloading webpage')
 276         note_m = self._html_search_regex(
 277             r'<div class="showmedia-trailer-notice">(.+?)</div>',
 278             webpage, 'trailer-notice', default='')
 279         if note_m:
 280             raise ExtractorError(note_m)
 281
 282         mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
 283         if mobj:
 284             msg = json.loads(mobj.group('msg'))
 285             if msg.get('type') == 'error':
 286                 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
 287
 288         if 'To view this, please log in to verify you are 18 or older.' in webpage:
 289             self.raise_login_required()
 290
 291         video_title = self._html_search_regex(
 292             r'(?s)<h1[^>]*>((?:(?!<h1).)*?<span[^>]+itemprop=["\']title["\'][^>]*>(?:(?!<h1).)+?)</h1>',
 293             webpage, 'video_title')
 294         video_title = re.sub(r' {2,}', ' ', video_title)
 295         video_description = self._html_search_regex(
 296             r'<script[^>]*>\s*.+?\[media_id=%s\].+?"description"\s*:\s*"([^"]+)' % video_id,
 297             webpage, 'description', default=None)
 298         if video_description:
 299             video_description = lowercase_escape(video_description.replace(r'\r\n', '\n'))
 300         video_upload_date = self._html_search_regex(
 301             [r'<div>Availability for free users:(.+?)</div>', r'<div>[^<>]+<span>\s*(.+?\d{4})\s*</span></div>'],
 302             webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
 303         if video_upload_date:
 304             video_upload_date = unified_strdate(video_upload_date)
 305         video_uploader = self._html_search_regex(
 306             r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', webpage,
 307             'video_uploader', fatal=False)
 308
 309         playerdata_url = compat_urllib_parse_unquote(self._html_search_regex(r'"config_url":"([^"]+)', webpage, 'playerdata_url'))
 310         playerdata_req = compat_urllib_request.Request(playerdata_url)
 311         playerdata_req.data = compat_urllib_parse.urlencode({'current_page': webpage_url})
 312         playerdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 313         playerdata = self._download_webpage(playerdata_req, video_id, note='Downloading media info')
 314
 315         stream_id = self._search_regex(r'<media_id>([^<]+)', playerdata, 'stream_id')
 316         video_thumbnail = self._search_regex(r'<episode_image_url>([^<]+)', playerdata, 'thumbnail', fatal=False)
 317
 318         formats = []
 319         for fmt in re.findall(r'showmedia\.([0-9]{3,4})p', webpage):
 320             stream_quality, stream_format = self._FORMAT_IDS[fmt]
 321             video_format = fmt + 'p'
 322             streamdata_req = compat_urllib_request.Request(
 323                 'http://www.crunchyroll.com/xml/?req=RpcApiVideoPlayer_GetStandardConfig&media_id=%s&video_format=%s&video_quality=%s'
 324                 % (stream_id, stream_format, stream_quality),
 325                 compat_urllib_parse.urlencode({'current_page': url}).encode('utf-8'))
 326             streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 327             streamdata = self._download_xml(
 328                 streamdata_req, video_id,
 329                 note='Downloading media info for %s' % video_format)
 330             stream_info = streamdata.find('./{default}preload/stream_info')
 331             video_url = stream_info.find('./host').text
 332             video_play_path = stream_info.find('./file').text
 333             metadata = stream_info.find('./metadata')
 334             format_info = {
 335                 'format': video_format,
 336                 'format_id': video_format,
 337                 'height': int_or_none(xpath_text(metadata, './height')),
 338                 'width': int_or_none(xpath_text(metadata, './width')),
 339             }
 340
 341             if '.fplive.net/' in video_url:
 342                 video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip())
 343                 parsed_video_url = compat_urlparse.urlparse(video_url)
 344                 direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace(
 345                     netloc='v.lvlt.crcdn.net',
 346                     path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_play_path.split(':')[-1])))
 347                 if self._is_valid_url(direct_video_url, video_id, video_format):
 348                     format_info.update({
 349                         'url': direct_video_url,
 350                     })
 351                     formats.append(format_info)
 352                     continue
 353
 354             format_info.update({
 355                 'url': video_url,
 356                 'play_path': video_play_path,
 357                 'ext': 'flv',
 358             })
 359             formats.append(format_info)
 360
 361         subtitles = self.extract_subtitles(video_id, webpage)
 362
 363         return {
 364             'id': video_id,
 365             'title': video_title,
 366             'description': video_description,
 367             'thumbnail': video_thumbnail,
 368             'uploader': video_uploader,
 369             'upload_date': video_upload_date,
 370             'subtitles': subtitles,
 371             'formats': formats,
 372         }
 373
 374
 375 class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE):
 376     IE_NAME = "crunchyroll:playlist"
 377     _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\-]+))/?(?:\?|$)'
 378
 379     _TESTS = [{
 380         'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 381         'info_dict': {
 382             'id': 'a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 383             'title': 'A Bridge to the Starry Skies - Hoshizora e Kakaru Hashi'
 384         },
 385         'playlist_count': 13,
 386     }, {
 387         # geo-restricted (US), 18+ maturity wall, non-premium available
 388         'url': 'http://www.crunchyroll.com/cosplay-complex-ova',
 389         'info_dict': {
 390             'id': 'cosplay-complex-ova',
 391             'title': 'Cosplay Complex OVA'
 392         },
 393         'playlist_count': 3,
 394         'skip': 'Georestricted',
 395     }, {
 396         # geo-restricted (US), 18+ maturity wall, non-premium will be available since 2015.11.14
 397         'url': 'http://www.crunchyroll.com/ladies-versus-butlers?skip_wall=1',
 398         'only_matching': True,
 399     }]
 400
 401     def _real_extract(self, url):
 402         show_id = self._match_id(url)
 403
 404         webpage = self._download_webpage(self._add_skip_wall(url), show_id)
 405         title = self._html_search_regex(
 406             r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>',
 407             webpage, 'title')
 408         episode_paths = re.findall(
 409             r'(?s)<li id="showview_videos_media_[0-9]+"[^>]+>.*?<a href="([^"]+)"',
 410             webpage)
 411         entries = [
 412             self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll')
 413             for ep in episode_paths
 414         ]
 415         entries.reverse()
 416
 417         return {
 418             '_type': 'playlist',
 419             'id': show_id,
 420             'title': title,
 421             'entries': entries,
 422         }