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 .subtitles import SubtitlesInfoExtractor
  13 from ..utils import (
  14     ExtractorError,
  15     compat_urllib_parse,
  16     compat_urllib_request,
  17     bytes_to_intlist,
  18     intlist_to_bytes,
  19     unified_strdate,
  20     clean_html,
  21     urlencode_postdata,
  22 )
  23 from ..aes import (
  24     aes_cbc_decrypt,
  25     inc,
  26 )
  27 from .common import InfoExtractor
  28
  29
  30 class CrunchyrollIE(SubtitlesInfoExtractor):
  31     _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.com/(?:[^/]*/[^/?&]*?|media/\?id=)(?P<video_id>[0-9]+))(?:[/?&]|$)'
  32     _TEST = {
  33         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
  34         #'md5': 'b1639fd6ddfaa43788c85f6d1dddd412',
  35         'info_dict': {
  36             'id': '645513',
  37             'ext': 'flv',
  38             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
  39             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
  40             'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
  41             'uploader': 'Yomiuri Telecasting Corporation (YTV)',
  42             'upload_date': '20131013',
  43             'url': 're:(?!.*&amp)',
  44         },
  45         'params': {
  46             # rtmp
  47             'skip_download': True,
  48         },
  49     }
  50
  51     _FORMAT_IDS = {
  52         '360': ('60', '106'),
  53         '480': ('61', '106'),
  54         '720': ('62', '106'),
  55         '1080': ('80', '108'),
  56     }
  57
  58     def _login(self):
  59         (username, password) = self._get_login_info()
  60         if username is None:
  61             return
  62         self.report_login()
  63         login_url = 'https://www.crunchyroll.com/?a=formhandler'
  64         data = urlencode_postdata({
  65             'formname': 'RpcApiUser_Login',
  66             'name': username,
  67             'password': password,
  68         })
  69         login_request = compat_urllib_request.Request(login_url, data)
  70         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
  71         self._download_webpage(login_request, None, False, 'Wrong login info')
  72
  73
  74     def _real_initialize(self):
  75         self._login()
  76
  77
  78     def _decrypt_subtitles(self, data, iv, id):
  79         data = bytes_to_intlist(data)
  80         iv = bytes_to_intlist(iv)
  81         id = int(id)
  82
  83         def obfuscate_key_aux(count, modulo, start):
  84             output = list(start)
  85             for _ in range(count):
  86                 output.append(output[-1] + output[-2])
  87             # cut off start values
  88             output = output[2:]
  89             output = list(map(lambda x: x % modulo + 33, output))
  90             return output
  91
  92         def obfuscate_key(key):
  93             num1 = int(floor(pow(2, 25) * sqrt(6.9)))
  94             num2 = (num1 ^ key) << 5
  95             num3 = key ^ num1
  96             num4 = num3 ^ (num3 >> 3) ^ num2
  97             prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
  98             shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
  99             # Extend 160 Bit hash to 256 Bit
 100             return shaHash + [0] * 12
 101
 102         key = obfuscate_key(id)
 103         class Counter:
 104             __value = iv
 105             def next_value(self):
 106                 temp = self.__value
 107                 self.__value = inc(self.__value)
 108                 return temp
 109         decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
 110         return zlib.decompress(decrypted_data)
 111
 112     def _convert_subtitles_to_srt(self, subtitles):
 113         output = ''
 114         for i, (start, end, text) in enumerate(re.findall(r'<event [^>]*?start="([^"]+)" [^>]*?end="([^"]+)" [^>]*?text="([^"]+)"[^>]*?>', subtitles), 1):
 115             start = start.replace('.', ',')
 116             end = end.replace('.', ',')
 117             text = clean_html(text)
 118             text = text.replace('\\N', '\n')
 119             if not text:
 120                 continue
 121             output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
 122         return output
 123
 124     def _convert_subtitles_to_ass(self, subtitles):
 125         output = ''
 126
 127         def ass_bool(strvalue):
 128             assvalue = '0'
 129             if strvalue == '1':
 130                 assvalue = '-1'
 131             return assvalue
 132
 133         sub_root = xml.etree.ElementTree.fromstring(subtitles)
 134         if not sub_root:
 135             return output
 136
 137         output = '[Script Info]\n'
 138         output += 'Title: %s\n' % sub_root.attrib["title"]
 139         output += 'ScriptType: v4.00+\n'
 140         output += 'WrapStyle: %s\n' % sub_root.attrib["wrap_style"]
 141         output += 'PlayResX: %s\n' % sub_root.attrib["play_res_x"]
 142         output += 'PlayResY: %s\n' % sub_root.attrib["play_res_y"]
 143         output += """ScaledBorderAndShadow: yes
 144
 145 [V4+ Styles]
 146 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
 147 """
 148         for style in sub_root.findall('./styles/style'):
 149             output += 'Style: ' + style.attrib["name"]
 150             output += ',' + style.attrib["font_name"]
 151             output += ',' + style.attrib["font_size"]
 152             output += ',' + style.attrib["primary_colour"]
 153             output += ',' + style.attrib["secondary_colour"]
 154             output += ',' + style.attrib["outline_colour"]
 155             output += ',' + style.attrib["back_colour"]
 156             output += ',' + ass_bool(style.attrib["bold"])
 157             output += ',' + ass_bool(style.attrib["italic"])
 158             output += ',' + ass_bool(style.attrib["underline"])
 159             output += ',' + ass_bool(style.attrib["strikeout"])
 160             output += ',' + style.attrib["scale_x"]
 161             output += ',' + style.attrib["scale_y"]
 162             output += ',' + style.attrib["spacing"]
 163             output += ',' + style.attrib["angle"]
 164             output += ',' + style.attrib["border_style"]
 165             output += ',' + style.attrib["outline"]
 166             output += ',' + style.attrib["shadow"]
 167             output += ',' + style.attrib["alignment"]
 168             output += ',' + style.attrib["margin_l"]
 169             output += ',' + style.attrib["margin_r"]
 170             output += ',' + style.attrib["margin_v"]
 171             output += ',' + style.attrib["encoding"]
 172             output += '\n'
 173
 174         output += """
 175 [Events]
 176 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
 177 """
 178         for event in sub_root.findall('./events/event'):
 179             output += 'Dialogue: 0'
 180             output += ',' + event.attrib["start"]
 181             output += ',' + event.attrib["end"]
 182             output += ',' + event.attrib["style"]
 183             output += ',' + event.attrib["name"]
 184             output += ',' + event.attrib["margin_l"]
 185             output += ',' + event.attrib["margin_r"]
 186             output += ',' + event.attrib["margin_v"]
 187             output += ',' + event.attrib["effect"]
 188             output += ',' + event.attrib["text"]
 189             output += '\n'
 190
 191         return output
 192
 193     def _real_extract(self,url):
 194         mobj = re.match(self._VALID_URL, url)
 195         video_id = mobj.group('video_id')
 196
 197         if mobj.group('prefix') == 'm':
 198             mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
 199             webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
 200         else:
 201             webpage_url = 'http://www.' + mobj.group('url')
 202
 203         webpage = self._download_webpage(webpage_url, video_id, 'Downloading webpage')
 204         note_m = self._html_search_regex(r'<div class="showmedia-trailer-notice">(.+?)</div>', webpage, 'trailer-notice', default='')
 205         if note_m:
 206             raise ExtractorError(note_m)
 207
 208         mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
 209         if mobj:
 210             msg = json.loads(mobj.group('msg'))
 211             if msg.get('type') == 'error':
 212                 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
 213
 214         video_title = self._html_search_regex(r'<h1[^>]*>(.+?)</h1>', webpage, 'video_title', flags=re.DOTALL)
 215         video_title = re.sub(r' {2,}', ' ', video_title)
 216         video_description = self._html_search_regex(r'"description":"([^"]+)', webpage, 'video_description', default='')
 217         if not video_description:
 218             video_description = None
 219         video_upload_date = self._html_search_regex(r'<div>Availability for free users:(.+?)</div>', webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
 220         if video_upload_date:
 221             video_upload_date = unified_strdate(video_upload_date)
 222         video_uploader = self._html_search_regex(r'<div>\s*Publisher:(.+?)</div>', webpage, 'video_uploader', fatal=False, flags=re.DOTALL)
 223
 224         playerdata_url = compat_urllib_parse.unquote(self._html_search_regex(r'"config_url":"([^"]+)', webpage, 'playerdata_url'))
 225         playerdata_req = compat_urllib_request.Request(playerdata_url)
 226         playerdata_req.data = compat_urllib_parse.urlencode({'current_page': webpage_url})
 227         playerdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 228         playerdata = self._download_webpage(playerdata_req, video_id, note='Downloading media info')
 229
 230         stream_id = self._search_regex(r'<media_id>([^<]+)', playerdata, 'stream_id')
 231         video_thumbnail = self._search_regex(r'<episode_image_url>([^<]+)', playerdata, 'thumbnail', fatal=False)
 232
 233         formats = []
 234         for fmt in re.findall(r'\?p([0-9]{3,4})=1', webpage):
 235             stream_quality, stream_format = self._FORMAT_IDS[fmt]
 236             video_format = fmt+'p'
 237             streamdata_req = compat_urllib_request.Request('http://www.crunchyroll.com/xml/')
 238             # urlencode doesn't work!
 239             streamdata_req.data = 'req=RpcApiVideoEncode%5FGetStreamInfo&video%5Fencode%5Fquality='+stream_quality+'&media%5Fid='+stream_id+'&video%5Fformat='+stream_format
 240             streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
 241             streamdata_req.add_header('Content-Length', str(len(streamdata_req.data)))
 242             streamdata = self._download_xml(
 243                 streamdata_req, video_id,
 244                 note='Downloading media info for %s' % video_format)
 245             video_url = streamdata.find('.//host').text
 246             video_play_path = streamdata.find('.//file').text
 247             formats.append({
 248                 'url': video_url,
 249                 'play_path': video_play_path,
 250                 'ext': 'flv',
 251                 'format': video_format,
 252                 'format_id': video_format,
 253             })
 254
 255         subtitles = {}
 256         sub_format = self._downloader.params.get('subtitlesformat', 'srt')
 257         for sub_id, sub_name in re.findall(r'\?ssid=([0-9]+)" title="([^"]+)', webpage):
 258             sub_page = self._download_webpage('http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id='+sub_id,\
 259                                               video_id, note='Downloading subtitles for '+sub_name)
 260             id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
 261             iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
 262             data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
 263             if not id or not iv or not data:
 264                 continue
 265             id = int(id)
 266             iv = base64.b64decode(iv)
 267             data = base64.b64decode(data)
 268
 269             subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
 270             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
 271             if not lang_code:
 272                 continue
 273             if sub_format == 'ass':
 274                 subtitles[lang_code] = self._convert_subtitles_to_ass(subtitle)
 275             else:
 276                 subtitles[lang_code] = self._convert_subtitles_to_srt(subtitle)
 277
 278         if self._downloader.params.get('listsubtitles', False):
 279             self._list_available_subtitles(video_id, subtitles)
 280             return
 281
 282         return {
 283             'id':          video_id,
 284             'title':       video_title,
 285             'description': video_description,
 286             'thumbnail':   video_thumbnail,
 287             'uploader':    video_uploader,
 288             'upload_date': video_upload_date,
 289             'subtitles':   subtitles,
 290             'formats':     formats,
 291         }
 292
 293
 294 class CrunchyrollShowPlaylistIE(InfoExtractor):
 295     IE_NAME = "crunchyroll:playlist"
 296     _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\-]+))/?$'
 297
 298     _TESTS = [{
 299         'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 300         'info_dict': {
 301             'id': 'a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
 302             'title': 'A Bridge to the Starry Skies - Hoshizora e Kakaru Hashi'
 303         },
 304         'playlist_count': 13,
 305     }]
 306
 307     def _real_extract(self, url):
 308         show_id = self._match_id(url)
 309
 310         webpage = self._download_webpage(url, show_id)
 311         title = self._html_search_regex(
 312             r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>',
 313             webpage, 'title')
 314         episode_paths = re.findall(
 315             r'(?s)<li id="showview_videos_media_[0-9]+"[^>]+>.*?<a href="([^"]+)"',
 316             webpage)
 317         entries = [
 318             self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll')
 319             for ep in episode_paths
 320         ]
 321         entries.reverse()
 322
 323         return {
 324             '_type': 'playlist',
 325             'id': show_id,
 326             'title': title,
 327             'entries': entries,
 328         }