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     sanitized_Request,
  27     unified_strdate,
  28     urlencode_postdata,
  29     xpath_text,
  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, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None):
  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(
  69             request, video_id, note, errnote, fatal, tries, timeout, encoding)
  70
  71     @staticmethod
  72     def _add_skip_wall(url):
  73         parsed_url = compat_urlparse.urlparse(url)
  74         qs = compat_urlparse.parse_qs(parsed_url.query)
  75         # Always force skip_wall to bypass maturity wall, namely 18+ confirmation message:
  76         # > This content may be inappropriate for some people.
  77         # > Are you sure you want to continue?
  78         # since it's not disabled by default in crunchyroll account's settings.
  79         # See https://github.com/rg3/youtube-dl/issues/7202.
  80         qs['skip_wall'] = ['1']
  81         return compat_urlparse.urlunparse(
  82             parsed_url._replace(query=compat_urllib_parse.urlencode(qs, True)))
  83
  84
  85 class CrunchyrollIE(CrunchyrollBaseIE):
  86     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.(?:com|fr)/(?:media(?:-|/\?id=)|[^/]*/[^/?&]*?)(?P<video_id>[0-9]+))(?:[/?&]|$)'
  87     _TESTS = [{
  88         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
  89         'info_dict': {
  90             'id': '645513',
  91             'ext': 'flv',
  92             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
  93             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
  94             'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
  95             'uploader': 'Yomiuri Telecasting Corporation (YTV)',
  96             'upload_date': '20131013',
  97             'url': 're:(?!.*&amp)',
  98         },
  99         'params': {
 100             # rtmp
 101             'skip_download': True,
 102         },
 103     }, {
 104         'url': 'http://www.crunchyroll.com/media-589804/culture-japan-1',
 105         'info_dict': {
 106             'id': '589804',
 107             'ext': 'flv',
 108             'title': 'Culture Japan Episode 1 – Rebuilding Japan after the 3.11',
 109             'description': 'md5:2fbc01f90b87e8e9137296f37b461c12',
 110             'thumbnail': 're:^https?://.*\.jpg$',
 111             'uploader': 'Danny Choo Network',
 112             'upload_date': '20120213',
 113         },
 114         'params': {
 115             # rtmp
 116             'skip_download': True,
 117         },
 118     }, {
 119         'url': 'http://www.crunchyroll.fr/girl-friend-beta/episode-11-goodbye-la-mode-661697',
 120         'only_matching': True,
 121     }, {
 122         # geo-restricted (US), 18+ maturity wall, non-premium available
 123         'url': 'http://www.crunchyroll.com/cosplay-complex-ova/episode-1-the-birth-of-the-cosplay-club-565617',
 124         'only_matching': True,
 125     }]
 126
 127     _FORMAT_IDS = {
 128         '360': ('60', '106'),
 129         '480': ('61', '106'),
 130         '720': ('62', '106'),
 131         '1080': ('80', '108'),
 132     }
 133
 134     def _decrypt_subtitles(self, data, iv, id):
 135         data = bytes_to_intlist(base64.b64decode(data.encode('utf-8')))
 136         iv = bytes_to_intlist(base64.b64decode(iv.encode('utf-8')))
 137         id = int(id)
 138
 139         def obfuscate_key_aux(count, modulo, start):
 140             output = list(start)
 141             for _ in range(count):
 142                 output.append(output[-1] + output[-2])
 143             # cut off start values
 144             output = output[2:]
 145             output = list(map(lambda x: x % modulo + 33, output))
 146             return output
 147
 148         def obfuscate_key(key):
 149             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
 150             num2 = (num1 ^ key) << 5
 151             num3 = key ^ num1
 152             num4 = num3 ^ (num3 >> 3) ^ num2
 153             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
 154             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
 155             # Extend 160 Bit hash to 256 Bit
 156             return shaHash + [0] * 12
 157
 158         key = obfuscate_key(id)
 159
 160         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
 161         return zlib.decompress(decrypted_data)
 162
 163     def _convert_subtitles_to_srt(self, sub_root):
 164         output = ''
 165
 166         for i, event in enumerate(sub_root.findall('./events/event'), 1):
 167             start = event.attrib['start'].replace('.', ',')
 168             end = event.attrib['end'].replace('.', ',')
 169             text = event.attrib['text'].replace('\\N', '\n')
 170             output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
 171         return output
 172
 173     def _convert_subtitles_to_ass(self, sub_root):
 174         output = ''
 175
 176         def ass_bool(strvalue):
 177             assvalue = '0'
 178             if strvalue == '1':
 179                 assvalue = '-1'
 180             return assvalue
 181
 182         output = '[Script Info]\n'
 183         output += 'Title: %s\n' % sub_root.attrib["title"]
 184         output += 'ScriptType: v4.00+\n'
 185         output += 'WrapStyle: %s\n' % sub_root.attrib["wrap_style"]
 186         output += 'PlayResX: %s\n' % sub_root.attrib["play_res_x"]
 187         output += 'PlayResY: %s\n' % sub_root.attrib["play_res_y"]
 188         output += """ScaledBorderAndShadow: yes
 189
 190 [V4+ Styles]
 191 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
 192 """
 193         for style in sub_root.findall('./styles/style'):
 194             output += 'Style: ' + style.attrib["name"]
 195             output += ',' + style.attrib["font_name"]
 196             output += ',' + style.attrib["font_size"]
 197             output += ',' + style.attrib["primary_colour"]
 198             output += ',' + style.attrib["secondary_colour"]
 199             output += ',' + style.attrib["outline_colour"]
 200             output += ',' + style.attrib["back_colour"]
 201             output += ',' + ass_bool(style.attrib["bold"])
 202             output += ',' + ass_bool(style.attrib["italic"])
 203             output += ',' + ass_bool(style.attrib["underline"])
 204             output += ',' + ass_bool(style.attrib["strikeout"])
 205             output += ',' + style.attrib["scale_x"]
 206             output += ',' + style.attrib["scale_y"]
 207             output += ',' + style.attrib["spacing"]
 208             output += ',' + style.attrib["angle"]
 209             output += ',' + style.attrib["border_style"]
 210             output += ',' + style.attrib["outline"]
 211             output += ',' + style.attrib["shadow"]
 212             output += ',' + style.attrib["alignment"]
 213             output += ',' + style.attrib["margin_l"]
 214             output += ',' + style.attrib["margin_r"]
 215             output += ',' + style.attrib["margin_v"]
 216             output += ',' + style.attrib["encoding"]
 217             output += '\n'
 218
 219         output += """
 220 [Events]
 221 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
 222 """
 223         for event in sub_root.findall('./events/event'):
 224             output += 'Dialogue: 0'
 225             output += ',' + event.attrib["start"]
 226             output += ',' + event.attrib["end"]
 227             output += ',' + event.attrib["style"]
 228             output += ',' + event.attrib["name"]
 229             output += ',' + event.attrib["margin_l"]
 230             output += ',' + event.attrib["margin_r"]
 231             output += ',' + event.attrib["margin_v"]
 232             output += ',' + event.attrib["effect"]
 233             output += ',' + event.attrib["text"]
 234             output += '\n'
 235
 236         return output
 237
 238     def _extract_subtitles(self, subtitle):
 239         sub_root = compat_etree_fromstring(subtitle)
 240         return [{
 241             'ext': 'srt',
 242             'data': self._convert_subtitles_to_srt(sub_root),
 243         }, {
 244             'ext': 'ass',
 245             'data': self._convert_subtitles_to_ass(sub_root),
 246         }]
 247
 248     def _get_subtitles(self, video_id, webpage):
 249         subtitles = {}
 250         for sub_id, sub_name in re.findall(r'\bssid=([0-9]+)"[^>]+?\btitle="([^"]+)', webpage):
 251             sub_page = self._download_webpage(
 252                 'http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id=' + sub_id,
 253                 video_id, note='Downloading subtitles for ' + sub_name)
 254             id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
 255             iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
 256             data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
 257             if not id or not iv or not data:
 258                 continue
 259             subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
 260             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 261             if not lang_code:
 262                 continue
 263             subtitles[lang_code] = self._extract_subtitles(subtitle)
 264         return subtitles
 265
 266     def _real_extract(self, url):
 267         mobj = re.match(self._VALID_URL, url)
 268         video_id = mobj.group('video_id')
 269
 270         if mobj.group('prefix') == 'm':
 271             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 272             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 273         else:
 274             webpage_url = 'http://www.' + mobj.group('url')
 275
 276         webpage = self._download_webpage(self._add_skip_wall(webpage_url), video_id, 'Downloading webpage')
 277         note_m = self._html_search_regex(
 278             r'<div class="showmedia-trailer-notice">(.+?)</div>',
 279             webpage, 'trailer-notice', default='')
 280         if note_m:
 281             raise ExtractorError(note_m)
 282
 283         mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
 284         if mobj:
 285             msg = json.loads(mobj.group('msg'))
 286             if msg.get('type') == 'error':
 287                 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
 288
 289         if 'To view this, please log in to verify you are 18 or older.' in webpage:
 290             self.raise_login_required()
 291
 292         video_title = self._html_search_regex(
 293             r'(?s)<h1[^>]*>((?:(?!<h1).)*?<span[^>]+itemprop=["\']title["\'][^>]*>(?:(?!<h1).)+?)</h1>',
 294             webpage, 'video_title')
 295         video_title = re.sub(r' {2,}', ' ', video_title)
 296         video_description = self._html_search_regex(
 297             r'<script[^>]*>\s*.+?\[media_id=%s\].+?"description"\s*:\s*"([^"]+)' % video_id,
 298             webpage, 'description', default=None)
 299         if video_description:
 300             video_description = lowercase_escape(video_description.replace(r'\r\n', '\n'))
 301         video_upload_date = self._html_search_regex(
 302             [r'<div>Availability for free users:(.+?)</div>', r'<div>[^<>]+<span>\s*(.+?\d{4})\s*</span></div>'],
 303             webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
 304         if video_upload_date:
 305             video_upload_date = unified_strdate(video_upload_date)
 306         video_uploader = self._html_search_regex(
 307             r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', webpage,
 308             'video_uploader', fatal=False)
 309
 310         playerdata_url = compat_urllib_parse_unquote(self._html_search_regex(r'"config_url":"([^"]+)', webpage, 'playerdata_url'))
 311         playerdata_req = sanitized_Request(playerdata_url)
 312         playerdata_req.data = compat_urllib_parse.urlencode({'current_page': webpage_url})
 313         playerdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 314         playerdata = self._download_webpage(playerdata_req, video_id, note='Downloading media info')
 315
 316         stream_id = self._search_regex(r'<media_id>([^<]+)', playerdata, 'stream_id')
 317         video_thumbnail = self._search_regex(r'<episode_image_url>([^<]+)', playerdata, 'thumbnail', fatal=False)
 318
 319         formats = []
 320         for fmt in re.findall(r'showmedia\.([0-9]{3,4})p', webpage):
 321             stream_quality, stream_format = self._FORMAT_IDS[fmt]
 322             video_format = fmt + 'p'
 323             streamdata_req = sanitized_Request(
 324                 'http://www.crunchyroll.com/xml/?req=RpcApiVideoPlayer_GetStandardConfig&media_id=%s&video_format=%s&video_quality=%s'
 325                 % (stream_id, stream_format, stream_quality),
 326                 compat_urllib_parse.urlencode({'current_page': url}).encode('utf-8'))
 327             streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 328             streamdata = self._download_xml(
 329                 streamdata_req, video_id,
 330                 note='Downloading media info for %s' % video_format)
 331             stream_info = streamdata.find('./{default}preload/stream_info')
 332             video_url = stream_info.find('./host').text
 333             video_play_path = stream_info.find('./file').text
 334             metadata = stream_info.find('./metadata')
 335             format_info = {
 336                 'format': video_format,
 337                 'format_id': video_format,
 338                 'height': int_or_none(xpath_text(metadata, './height')),
 339                 'width': int_or_none(xpath_text(metadata, './width')),
 340             }
 341
 342             if '.fplive.net/' in video_url:
 343                 video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip())
 344                 parsed_video_url = compat_urlparse.urlparse(video_url)
 345                 direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace(
 346                     netloc='v.lvlt.crcdn.net',
 347                     path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_play_path.split(':')[-1])))
 348                 if self._is_valid_url(direct_video_url, video_id, video_format):
 349                     format_info.update({
 350                         'url': direct_video_url,
 351                     })
 352                     formats.append(format_info)
 353                     continue
 354
 355             format_info.update({
 356                 'url': video_url,
 357                 'play_path': video_play_path,
 358                 'ext': 'flv',
 359             })
 360             formats.append(format_info)
 361
 362         subtitles = self.extract_subtitles(video_id, webpage)
 363
 364         return {
 365             'id': video_id,
 366             'title': video_title,
 367             'description': video_description,
 368             'thumbnail': video_thumbnail,
 369             'uploader': video_uploader,
 370             'upload_date': video_upload_date,
 371             'subtitles': subtitles,
 372             'formats': formats,
 373         }
 374
 375
 376 class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE):
 377     IE_NAME = "crunchyroll:playlist"
 378     _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\-]+))/?(?:\?|$)'
 379
 380     _TESTS = [{
 381         'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 382         'info_dict': {
 383             'id': 'a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 384             'title': 'A Bridge to the Starry Skies - Hoshizora e Kakaru Hashi'
 385         },
 386         'playlist_count': 13,
 387     }, {
 388         # geo-restricted (US), 18+ maturity wall, non-premium available
 389         'url': 'http://www.crunchyroll.com/cosplay-complex-ova',
 390         'info_dict': {
 391             'id': 'cosplay-complex-ova',
 392             'title': 'Cosplay Complex OVA'
 393         },
 394         'playlist_count': 3,
 395         'skip': 'Georestricted',
 396     }, {
 397         # geo-restricted (US), 18+ maturity wall, non-premium will be available since 2015.11.14
 398         'url': 'http://www.crunchyroll.com/ladies-versus-butlers?skip_wall=1',
 399         'only_matching': True,
 400     }]
 401
 402     def _real_extract(self, url):
 403         show_id = self._match_id(url)
 404
 405         webpage = self._download_webpage(self._add_skip_wall(url), show_id)
 406         title = self._html_search_regex(
 407             r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>',
 408             webpage, 'title')
 409         episode_paths = re.findall(
 410             r'(?s)<li id="showview_videos_media_[0-9]+"[^>]+>.*?<a href="([^"]+)"',
 411             webpage)
 412         entries = [
 413             self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll')
 414             for ep in episode_paths
 415         ]
 416         entries.reverse()
 417
 418         return {
 419             '_type': 'playlist',
 420             'id': show_id,
 421             'title': title,
 422             'entries': entries,
 423         }