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