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