]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/crunchyroll.py
8d5b69f68d3ddb345dc67487db998cf164b2765c
[youtubedl] / youtube_dl / extractor / crunchyroll.py
1 # coding: 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_urlencode,
15 compat_urllib_request,
16 compat_urlparse,
17 )
18 from ..utils import (
19 ExtractorError,
20 bytes_to_intlist,
21 intlist_to_bytes,
22 int_or_none,
23 lowercase_escape,
24 remove_end,
25 sanitized_Request,
26 unified_strdate,
27 urlencode_postdata,
28 xpath_text,
29 extract_attributes,
30 )
31 from ..aes import (
32 aes_cbc_decrypt,
33 )
34
35
36 class CrunchyrollBaseIE(InfoExtractor):
37 _LOGIN_URL = 'https://www.crunchyroll.com/login'
38 _LOGIN_FORM = 'login_form'
39 _NETRC_MACHINE = 'crunchyroll'
40
41 def _login(self):
42 (username, password) = self._get_login_info()
43 if username is None:
44 return
45
46 login_page = self._download_webpage(
47 self._LOGIN_URL, None, 'Downloading login page')
48
49 def is_logged(webpage):
50 return '<title>Redirecting' in webpage
51
52 # Already logged in
53 if is_logged(login_page):
54 return
55
56 login_form_str = self._search_regex(
57 r'(?P<form><form[^>]+?id=(["\'])%s\2[^>]*>)' % self._LOGIN_FORM,
58 login_page, 'login form', group='form')
59
60 post_url = extract_attributes(login_form_str).get('action')
61 if not post_url:
62 post_url = self._LOGIN_URL
63 elif not post_url.startswith('http'):
64 post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
65
66 login_form = self._form_hidden_inputs(self._LOGIN_FORM, login_page)
67
68 login_form.update({
69 'login_form[name]': username,
70 'login_form[password]': password,
71 })
72
73 response = self._download_webpage(
74 post_url, None, 'Logging in', 'Wrong login info',
75 data=urlencode_postdata(login_form),
76 headers={'Content-Type': 'application/x-www-form-urlencoded'})
77
78 # Successful login
79 if is_logged(response):
80 return
81
82 error = self._html_search_regex(
83 '(?s)<ul[^>]+class=["\']messages["\'][^>]*>(.+?)</ul>',
84 response, 'error message', default=None)
85 if error:
86 raise ExtractorError('Unable to login: %s' % error, expected=True)
87
88 raise ExtractorError('Unable to log in')
89
90 def _real_initialize(self):
91 self._login()
92
93 def _download_webpage(self, url_or_request, *args, **kwargs):
94 request = (url_or_request if isinstance(url_or_request, compat_urllib_request.Request)
95 else sanitized_Request(url_or_request))
96 # Accept-Language must be set explicitly to accept any language to avoid issues
97 # similar to https://github.com/rg3/youtube-dl/issues/6797.
98 # Along with IP address Crunchyroll uses Accept-Language to guess whether georestriction
99 # should be imposed or not (from what I can see it just takes the first language
100 # ignoring the priority and requires it to correspond the IP). By the way this causes
101 # Crunchyroll to not work in georestriction cases in some browsers that don't place
102 # the locale lang first in header. However allowing any language seems to workaround the issue.
103 request.add_header('Accept-Language', '*')
104 return super(CrunchyrollBaseIE, self)._download_webpage(request, *args, **kwargs)
105
106 @staticmethod
107 def _add_skip_wall(url):
108 parsed_url = compat_urlparse.urlparse(url)
109 qs = compat_urlparse.parse_qs(parsed_url.query)
110 # Always force skip_wall to bypass maturity wall, namely 18+ confirmation message:
111 # > This content may be inappropriate for some people.
112 # > Are you sure you want to continue?
113 # since it's not disabled by default in crunchyroll account's settings.
114 # See https://github.com/rg3/youtube-dl/issues/7202.
115 qs['skip_wall'] = ['1']
116 return compat_urlparse.urlunparse(
117 parsed_url._replace(query=compat_urllib_parse_urlencode(qs, True)))
118
119
120 class CrunchyrollIE(CrunchyrollBaseIE):
121 _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.(?:com|fr)/(?:media(?:-|/\?id=)|[^/]*/[^/?&]*?)(?P<video_id>[0-9]+))(?:[/?&]|$)'
122 _TESTS = [{
123 'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
124 'info_dict': {
125 'id': '645513',
126 'ext': 'flv',
127 'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
128 'description': 'md5:2d17137920c64f2f49981a7797d275ef',
129 'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
130 'uploader': 'Yomiuri Telecasting Corporation (YTV)',
131 'upload_date': '20131013',
132 'url': 're:(?!.*&amp)',
133 },
134 'params': {
135 # rtmp
136 'skip_download': True,
137 },
138 }, {
139 'url': 'http://www.crunchyroll.com/media-589804/culture-japan-1',
140 'info_dict': {
141 'id': '589804',
142 'ext': 'flv',
143 'title': 'Culture Japan Episode 1 – Rebuilding Japan after the 3.11',
144 'description': 'md5:2fbc01f90b87e8e9137296f37b461c12',
145 'thumbnail': 're:^https?://.*\.jpg$',
146 'uploader': 'Danny Choo Network',
147 'upload_date': '20120213',
148 },
149 'params': {
150 # rtmp
151 'skip_download': True,
152 },
153 'skip': 'Video gone',
154 }, {
155 'url': 'http://www.crunchyroll.com/rezero-starting-life-in-another-world-/episode-5-the-morning-of-our-promise-is-still-distant-702409',
156 'info_dict': {
157 'id': '702409',
158 'ext': 'mp4',
159 'title': 'Re:ZERO -Starting Life in Another World- Episode 5 – The Morning of Our Promise Is Still Distant',
160 'description': 'md5:97664de1ab24bbf77a9c01918cb7dca9',
161 'thumbnail': 're:^https?://.*\.jpg$',
162 'uploader': 'TV TOKYO',
163 'upload_date': '20160508',
164 },
165 'params': {
166 # m3u8 download
167 'skip_download': True,
168 },
169 }, {
170 'url': 'http://www.crunchyroll.fr/girl-friend-beta/episode-11-goodbye-la-mode-661697',
171 'only_matching': True,
172 }, {
173 # geo-restricted (US), 18+ maturity wall, non-premium available
174 'url': 'http://www.crunchyroll.com/cosplay-complex-ova/episode-1-the-birth-of-the-cosplay-club-565617',
175 'only_matching': True,
176 }]
177
178 _FORMAT_IDS = {
179 '360': ('60', '106'),
180 '480': ('61', '106'),
181 '720': ('62', '106'),
182 '1080': ('80', '108'),
183 }
184
185 def _decrypt_subtitles(self, data, iv, id):
186 data = bytes_to_intlist(base64.b64decode(data.encode('utf-8')))
187 iv = bytes_to_intlist(base64.b64decode(iv.encode('utf-8')))
188 id = int(id)
189
190 def obfuscate_key_aux(count, modulo, start):
191 output = list(start)
192 for _ in range(count):
193 output.append(output[-1] + output[-2])
194 # cut off start values
195 output = output[2:]
196 output = list(map(lambda x: x % modulo + 33, output))
197 return output
198
199 def obfuscate_key(key):
200 num1 = int(floor(pow(2, 25) * sqrt(6.9)))
201 num2 = (num1 ^ key) << 5
202 num3 = key ^ num1
203 num4 = num3 ^ (num3 >> 3) ^ num2
204 prefix = intlist_to_bytes(obfuscate_key_aux(20, 97, (1, 2)))
205 shaHash = bytes_to_intlist(sha1(prefix + str(num4).encode('ascii')).digest())
206 # Extend 160 Bit hash to 256 Bit
207 return shaHash + [0] * 12
208
209 key = obfuscate_key(id)
210
211 decrypted_data = intlist_to_bytes(aes_cbc_decrypt(data, key, iv))
212 return zlib.decompress(decrypted_data)
213
214 def _convert_subtitles_to_srt(self, sub_root):
215 output = ''
216
217 for i, event in enumerate(sub_root.findall('./events/event'), 1):
218 start = event.attrib['start'].replace('.', ',')
219 end = event.attrib['end'].replace('.', ',')
220 text = event.attrib['text'].replace('\\N', '\n')
221 output += '%d\n%s --> %s\n%s\n\n' % (i, start, end, text)
222 return output
223
224 def _convert_subtitles_to_ass(self, sub_root):
225 output = ''
226
227 def ass_bool(strvalue):
228 assvalue = '0'
229 if strvalue == '1':
230 assvalue = '-1'
231 return assvalue
232
233 output = '[Script Info]\n'
234 output += 'Title: %s\n' % sub_root.attrib['title']
235 output += 'ScriptType: v4.00+\n'
236 output += 'WrapStyle: %s\n' % sub_root.attrib['wrap_style']
237 output += 'PlayResX: %s\n' % sub_root.attrib['play_res_x']
238 output += 'PlayResY: %s\n' % sub_root.attrib['play_res_y']
239 output += """ScaledBorderAndShadow: no
240
241 [V4+ Styles]
242 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
243 """
244 for style in sub_root.findall('./styles/style'):
245 output += 'Style: ' + style.attrib['name']
246 output += ',' + style.attrib['font_name']
247 output += ',' + style.attrib['font_size']
248 output += ',' + style.attrib['primary_colour']
249 output += ',' + style.attrib['secondary_colour']
250 output += ',' + style.attrib['outline_colour']
251 output += ',' + style.attrib['back_colour']
252 output += ',' + ass_bool(style.attrib['bold'])
253 output += ',' + ass_bool(style.attrib['italic'])
254 output += ',' + ass_bool(style.attrib['underline'])
255 output += ',' + ass_bool(style.attrib['strikeout'])
256 output += ',' + style.attrib['scale_x']
257 output += ',' + style.attrib['scale_y']
258 output += ',' + style.attrib['spacing']
259 output += ',' + style.attrib['angle']
260 output += ',' + style.attrib['border_style']
261 output += ',' + style.attrib['outline']
262 output += ',' + style.attrib['shadow']
263 output += ',' + style.attrib['alignment']
264 output += ',' + style.attrib['margin_l']
265 output += ',' + style.attrib['margin_r']
266 output += ',' + style.attrib['margin_v']
267 output += ',' + style.attrib['encoding']
268 output += '\n'
269
270 output += """
271 [Events]
272 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
273 """
274 for event in sub_root.findall('./events/event'):
275 output += 'Dialogue: 0'
276 output += ',' + event.attrib['start']
277 output += ',' + event.attrib['end']
278 output += ',' + event.attrib['style']
279 output += ',' + event.attrib['name']
280 output += ',' + event.attrib['margin_l']
281 output += ',' + event.attrib['margin_r']
282 output += ',' + event.attrib['margin_v']
283 output += ',' + event.attrib['effect']
284 output += ',' + event.attrib['text']
285 output += '\n'
286
287 return output
288
289 def _extract_subtitles(self, subtitle):
290 sub_root = compat_etree_fromstring(subtitle)
291 return [{
292 'ext': 'srt',
293 'data': self._convert_subtitles_to_srt(sub_root),
294 }, {
295 'ext': 'ass',
296 'data': self._convert_subtitles_to_ass(sub_root),
297 }]
298
299 def _get_subtitles(self, video_id, webpage):
300 subtitles = {}
301 for sub_id, sub_name in re.findall(r'\bssid=([0-9]+)"[^>]+?\btitle="([^"]+)', webpage):
302 sub_page = self._download_webpage(
303 'http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id=' + sub_id,
304 video_id, note='Downloading subtitles for ' + sub_name)
305 id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
306 iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
307 data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
308 if not id or not iv or not data:
309 continue
310 subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
311 lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
312 if not lang_code:
313 continue
314 subtitles[lang_code] = self._extract_subtitles(subtitle)
315 return subtitles
316
317 def _real_extract(self, url):
318 mobj = re.match(self._VALID_URL, url)
319 video_id = mobj.group('video_id')
320
321 if mobj.group('prefix') == 'm':
322 mobile_webpage = self._download_webpage(url, video_id, 'Downloading mobile webpage')
323 webpage_url = self._search_regex(r'<link rel="canonical" href="([^"]+)" />', mobile_webpage, 'webpage_url')
324 else:
325 webpage_url = 'http://www.' + mobj.group('url')
326
327 webpage = self._download_webpage(self._add_skip_wall(webpage_url), video_id, 'Downloading webpage')
328 note_m = self._html_search_regex(
329 r'<div class="showmedia-trailer-notice">(.+?)</div>',
330 webpage, 'trailer-notice', default='')
331 if note_m:
332 raise ExtractorError(note_m)
333
334 mobj = re.search(r'Page\.messaging_box_controller\.addItems\(\[(?P<msg>{.+?})\]\)', webpage)
335 if mobj:
336 msg = json.loads(mobj.group('msg'))
337 if msg.get('type') == 'error':
338 raise ExtractorError('crunchyroll returned error: %s' % msg['message_body'], expected=True)
339
340 if 'To view this, please log in to verify you are 18 or older.' in webpage:
341 self.raise_login_required()
342
343 video_title = self._html_search_regex(
344 r'(?s)<h1[^>]*>((?:(?!<h1).)*?<span[^>]+itemprop=["\']title["\'][^>]*>(?:(?!<h1).)+?)</h1>',
345 webpage, 'video_title')
346 video_title = re.sub(r' {2,}', ' ', video_title)
347 video_description = self._html_search_regex(
348 r'<script[^>]*>\s*.+?\[media_id=%s\].+?"description"\s*:\s*"([^"]+)' % video_id,
349 webpage, 'description', default=None)
350 if video_description:
351 video_description = lowercase_escape(video_description.replace(r'\r\n', '\n'))
352 video_upload_date = self._html_search_regex(
353 [r'<div>Availability for free users:(.+?)</div>', r'<div>[^<>]+<span>\s*(.+?\d{4})\s*</span></div>'],
354 webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
355 if video_upload_date:
356 video_upload_date = unified_strdate(video_upload_date)
357 video_uploader = self._html_search_regex(
358 r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', webpage,
359 'video_uploader', fatal=False)
360
361 available_fmts = []
362 for a, fmt in re.findall(r'(<a[^>]+token=["\']showmedia\.([0-9]{3,4})p["\'][^>]+>)', webpage):
363 attrs = extract_attributes(a)
364 href = attrs.get('href')
365 if href and '/freetrial' in href:
366 continue
367 available_fmts.append(fmt)
368 if not available_fmts:
369 for p in (r'token=["\']showmedia\.([0-9]{3,4})p"', r'showmedia\.([0-9]{3,4})p'):
370 available_fmts = re.findall(p, webpage)
371 if available_fmts:
372 break
373 video_encode_ids = []
374 formats = []
375 for fmt in available_fmts:
376 stream_quality, stream_format = self._FORMAT_IDS[fmt]
377 video_format = fmt + 'p'
378 streamdata_req = sanitized_Request(
379 'http://www.crunchyroll.com/xml/?req=RpcApiVideoPlayer_GetStandardConfig&media_id=%s&video_format=%s&video_quality=%s'
380 % (video_id, stream_format, stream_quality),
381 compat_urllib_parse_urlencode({'current_page': url}).encode('utf-8'))
382 streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
383 streamdata = self._download_xml(
384 streamdata_req, video_id,
385 note='Downloading media info for %s' % video_format)
386 stream_info = streamdata.find('./{default}preload/stream_info')
387 video_encode_id = xpath_text(stream_info, './video_encode_id')
388 if video_encode_id in video_encode_ids:
389 continue
390 video_encode_ids.append(video_encode_id)
391
392 video_file = xpath_text(stream_info, './file')
393 if not video_file:
394 continue
395 if video_file.startswith('http'):
396 formats.extend(self._extract_m3u8_formats(
397 video_file, video_id, 'mp4', entry_protocol='m3u8_native',
398 m3u8_id='hls', fatal=False))
399 continue
400
401 video_url = xpath_text(stream_info, './host')
402 if not video_url:
403 continue
404 metadata = stream_info.find('./metadata')
405 format_info = {
406 'format': video_format,
407 'format_id': video_format,
408 'height': int_or_none(xpath_text(metadata, './height')),
409 'width': int_or_none(xpath_text(metadata, './width')),
410 }
411
412 if '.fplive.net/' in video_url:
413 video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip())
414 parsed_video_url = compat_urlparse.urlparse(video_url)
415 direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace(
416 netloc='v.lvlt.crcdn.net',
417 path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_file.split(':')[-1])))
418 if self._is_valid_url(direct_video_url, video_id, video_format):
419 format_info.update({
420 'url': direct_video_url,
421 })
422 formats.append(format_info)
423 continue
424
425 format_info.update({
426 'url': video_url,
427 'play_path': video_file,
428 'ext': 'flv',
429 })
430 formats.append(format_info)
431 self._sort_formats(formats)
432
433 metadata = self._download_xml(
434 'http://www.crunchyroll.com/xml', video_id,
435 note='Downloading media info', query={
436 'req': 'RpcApiVideoPlayer_GetMediaMetadata',
437 'media_id': video_id,
438 })
439
440 subtitles = self.extract_subtitles(video_id, webpage)
441
442 return {
443 'id': video_id,
444 'title': video_title,
445 'description': video_description,
446 'thumbnail': xpath_text(metadata, 'episode_image_url'),
447 'uploader': video_uploader,
448 'upload_date': video_upload_date,
449 'series': xpath_text(metadata, 'series_title'),
450 'episode': xpath_text(metadata, 'episode_title'),
451 'episode_number': int_or_none(xpath_text(metadata, 'episode_number')),
452 'subtitles': subtitles,
453 'formats': formats,
454 }
455
456
457 class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE):
458 IE_NAME = 'crunchyroll:playlist'
459 _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\-]+))/?(?:\?|$)'
460
461 _TESTS = [{
462 'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
463 'info_dict': {
464 'id': 'a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi',
465 'title': 'A Bridge to the Starry Skies - Hoshizora e Kakaru Hashi'
466 },
467 'playlist_count': 13,
468 }, {
469 # geo-restricted (US), 18+ maturity wall, non-premium available
470 'url': 'http://www.crunchyroll.com/cosplay-complex-ova',
471 'info_dict': {
472 'id': 'cosplay-complex-ova',
473 'title': 'Cosplay Complex OVA'
474 },
475 'playlist_count': 3,
476 'skip': 'Georestricted',
477 }, {
478 # geo-restricted (US), 18+ maturity wall, non-premium will be available since 2015.11.14
479 'url': 'http://www.crunchyroll.com/ladies-versus-butlers?skip_wall=1',
480 'only_matching': True,
481 }]
482
483 def _real_extract(self, url):
484 show_id = self._match_id(url)
485
486 webpage = self._download_webpage(self._add_skip_wall(url), show_id)
487 title = self._html_search_regex(
488 r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>',
489 webpage, 'title')
490 episode_paths = re.findall(
491 r'(?s)<li id="showview_videos_media_[0-9]+"[^>]+>.*?<a href="([^"]+)"',
492 webpage)
493 entries = [
494 self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll')
495 for ep in episode_paths
496 ]
497 entries.reverse()
498
499 return {
500 '_type': 'playlist',
501 'id': show_id,
502 'title': title,
503 'entries': entries,
504 }