X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/d9d7cd0e85dc712461d9185db9df9d6c900a573b..54d5fa9938f088c9610aaf81f3a90df4c2b1f956:/youtube_dl/extractor/crunchyroll.py diff --git a/youtube_dl/extractor/crunchyroll.py b/youtube_dl/extractor/crunchyroll.py index 2ffa4a7..85a9a57 100644 --- a/youtube_dl/extractor/crunchyroll.py +++ b/youtube_dl/extractor/crunchyroll.py @@ -3,13 +3,15 @@ from __future__ import unicode_literals import re import json -import base64 import zlib from hashlib import sha1 from math import pow, sqrt, floor from .common import InfoExtractor +from .vrv import VRVIE from ..compat import ( + compat_b64decode, + compat_etree_Element, compat_etree_fromstring, compat_urllib_parse_urlencode, compat_urllib_request, @@ -18,6 +20,8 @@ from ..compat import ( from ..utils import ( ExtractorError, bytes_to_intlist, + extract_attributes, + float_or_none, intlist_to_bytes, int_or_none, lowercase_escape, @@ -26,7 +30,6 @@ from ..utils import ( unified_strdate, urlencode_postdata, xpath_text, - extract_attributes, ) from ..aes import ( aes_cbc_decrypt, @@ -38,8 +41,18 @@ class CrunchyrollBaseIE(InfoExtractor): _LOGIN_FORM = 'login_form' _NETRC_MACHINE = 'crunchyroll' + def _call_rpc_api(self, method, video_id, note=None, data=None): + data = data or {} + data['req'] = 'RpcApi' + method + data = compat_urllib_parse_urlencode(data).encode('utf-8') + return self._download_xml( + 'https://www.crunchyroll.com/xml/', + video_id, note, fatal=False, data=data, headers={ + 'Content-Type': 'application/x-www-form-urlencoded', + }) + def _login(self): - (username, password) = self._get_login_info() + username, password = self._get_login_info() if username is None: return @@ -47,7 +60,7 @@ class CrunchyrollBaseIE(InfoExtractor): self._LOGIN_URL, None, 'Downloading login page') def is_logged(webpage): - return 'Redirecting' in webpage + return 'href="/logout"' in webpage # Already logged in if is_logged(login_page): @@ -90,19 +103,6 @@ class CrunchyrollBaseIE(InfoExtractor): def _real_initialize(self): self._login() - def _download_webpage(self, url_or_request, *args, **kwargs): - request = (url_or_request if isinstance(url_or_request, compat_urllib_request.Request) - else sanitized_Request(url_or_request)) - # Accept-Language must be set explicitly to accept any language to avoid issues - # similar to https://github.com/rg3/youtube-dl/issues/6797. - # Along with IP address Crunchyroll uses Accept-Language to guess whether georestriction - # should be imposed or not (from what I can see it just takes the first language - # ignoring the priority and requires it to correspond the IP). By the way this causes - # Crunchyroll to not work in georestriction cases in some browsers that don't place - # the locale lang first in header. However allowing any language seems to workaround the issue. - request.add_header('Accept-Language', '*') - return super(CrunchyrollBaseIE, self)._download_webpage(request, *args, **kwargs) - @staticmethod def _add_skip_wall(url): parsed_url = compat_urlparse.urlparse(url) @@ -111,14 +111,15 @@ class CrunchyrollBaseIE(InfoExtractor): # > This content may be inappropriate for some people. # > Are you sure you want to continue? # since it's not disabled by default in crunchyroll account's settings. - # See https://github.com/rg3/youtube-dl/issues/7202. + # See https://github.com/ytdl-org/youtube-dl/issues/7202. qs['skip_wall'] = ['1'] return compat_urlparse.urlunparse( parsed_url._replace(query=compat_urllib_parse_urlencode(qs, True))) -class CrunchyrollIE(CrunchyrollBaseIE): - _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.(?:com|fr)/(?:media(?:-|/\?id=)|[^/]*/[^/?&]*?)(?P<video_id>[0-9]+))(?:[/?&]|$)' +class CrunchyrollIE(CrunchyrollBaseIE, VRVIE): + IE_NAME = 'crunchyroll' + _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.(?:com|fr)/(?:media(?:-|/\?id=)|(?:[^/]*/){1,2}[^/?&]*?)(?P<video_id>[0-9]+))(?:[/?&]|$)' _TESTS = [{ 'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513', 'info_dict': { @@ -126,7 +127,7 @@ class CrunchyrollIE(CrunchyrollBaseIE): 'ext': 'mp4', 'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!', 'description': 'md5:2d17137920c64f2f49981a7797d275ef', - 'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg', + 'thumbnail': r're:^https?://.*\.jpg$', 'uploader': 'Yomiuri Telecasting Corporation (YTV)', 'upload_date': '20131013', 'url': 're:(?!.*&)', @@ -199,7 +200,7 @@ class CrunchyrollIE(CrunchyrollBaseIE): 'info_dict': { 'id': '535080', 'ext': 'mp4', - 'title': '11eyes Episode 1 – Piros éjszaka - Red Night', + 'title': '11eyes Episode 1 – Red Night ~ Piros éjszaka', 'description': 'Kakeru and Yuka are thrown into an alternate nightmarish world they call "Red Night".', 'uploader': 'Marvelous AQL Inc.', 'upload_date': '20091021', @@ -240,6 +241,12 @@ class CrunchyrollIE(CrunchyrollBaseIE): # Just test metadata extraction 'skip_download': True, }, + }, { + 'url': 'http://www.crunchyroll.com/media-723735', + 'only_matching': True, + }, { + 'url': 'https://www.crunchyroll.com/en-gb/mob-psycho-100/episode-2-urban-legends-encountering-rumors-780921', + 'only_matching': True, }] _FORMAT_IDS = { @@ -249,9 +256,22 @@ class CrunchyrollIE(CrunchyrollBaseIE): '1080': ('80', '108'), } + def _download_webpage(self, url_or_request, *args, **kwargs): + request = (url_or_request if isinstance(url_or_request, compat_urllib_request.Request) + else sanitized_Request(url_or_request)) + # Accept-Language must be set explicitly to accept any language to avoid issues + # similar to https://github.com/ytdl-org/youtube-dl/issues/6797. + # Along with IP address Crunchyroll uses Accept-Language to guess whether georestriction + # should be imposed or not (from what I can see it just takes the first language + # ignoring the priority and requires it to correspond the IP). By the way this causes + # Crunchyroll to not work in georestriction cases in some browsers that don't place + # the locale lang first in header. However allowing any language seems to workaround the issue. + request.add_header('Accept-Language', '*') + return super(CrunchyrollBaseIE, self)._download_webpage(request, *args, **kwargs) + def _decrypt_subtitles(self, data, iv, id): - data = bytes_to_intlist(base64.b64decode(data.encode('utf-8'))) - iv = bytes_to_intlist(base64.b64decode(iv.encode('utf-8'))) + data = bytes_to_intlist(compat_b64decode(data)) + iv = bytes_to_intlist(compat_b64decode(iv)) id = int(id) def obfuscate_key_aux(count, modulo, start): @@ -365,15 +385,19 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text def _get_subtitles(self, video_id, webpage): subtitles = {} for sub_id, sub_name in re.findall(r'\bssid=([0-9]+)"[^>]+?\btitle="([^"]+)', webpage): - sub_page = self._download_webpage( - 'http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id=' + sub_id, - video_id, note='Downloading subtitles for ' + sub_name) - id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False) - iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False) - data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False) - if not id or not iv or not data: + sub_doc = self._call_rpc_api( + 'Subtitle_GetXml', video_id, + 'Downloading subtitles for ' + sub_name, data={ + 'subtitle_script_id': sub_id, + }) + if not isinstance(sub_doc, compat_etree_Element): + continue + sid = sub_doc.get('id') + iv = xpath_text(sub_doc, 'iv', 'subtitle iv') + data = xpath_text(sub_doc, 'data', 'subtitle data') + if not sid or not iv or not data: continue - subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8') + subtitle = self._decrypt_subtitles(data, iv, sid).decode('utf-8') lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False) if not lang_code: continue @@ -408,13 +432,22 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text if 'To view this, please log in to verify you are 18 or older.' in webpage: self.raise_login_required() + media = self._parse_json(self._search_regex( + r'vilos\.config\.media\s*=\s*({.+?});', + webpage, 'vilos media', default='{}'), video_id) + media_metadata = media.get('metadata') or {} + + language = self._search_regex( + r'(?:vilos\.config\.player\.language|LOCALE)\s*=\s*(["\'])(?P<lang>(?:(?!\1).)+)\1', + webpage, 'language', default=None, group='lang') + video_title = self._html_search_regex( r'(?s)<h1[^>]*>((?:(?!<h1).)*?<span[^>]+itemprop=["\']title["\'][^>]*>(?:(?!<h1).)+?)</h1>', webpage, 'video_title') video_title = re.sub(r' {2,}', ' ', video_title) - video_description = self._parse_json(self._html_search_regex( + video_description = (self._parse_json(self._html_search_regex( r'<script[^>]*>\s*.+?\[media_id=%s\].+?({.+?"description"\s*:.+?})\);' % video_id, - webpage, 'description', default='{}'), video_id).get('description') + webpage, 'description', default='{}'), video_id) or media_metadata).get('description') if video_description: video_description = lowercase_escape(video_description.replace(r'\r\n', '\n')) video_upload_date = self._html_search_regex( @@ -427,105 +460,163 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text [r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', r'<div>\s*Publisher:\s*<span>\s*(.+?)\s*</span>\s*</div>'], webpage, 'video_uploader', fatal=False) - available_fmts = [] - for a, fmt in re.findall(r'(<a[^>]+token=["\']showmedia\.([0-9]{3,4})p["\'][^>]+>)', webpage): - attrs = extract_attributes(a) - href = attrs.get('href') - if href and '/freetrial' in href: - continue - available_fmts.append(fmt) - if not available_fmts: - for p in (r'token=["\']showmedia\.([0-9]{3,4})p"', r'showmedia\.([0-9]{3,4})p'): - available_fmts = re.findall(p, webpage) - if available_fmts: - break - video_encode_ids = [] formats = [] - for fmt in available_fmts: - stream_quality, stream_format = self._FORMAT_IDS[fmt] - video_format = fmt + 'p' - streamdata_req = sanitized_Request( - 'http://www.crunchyroll.com/xml/?req=RpcApiVideoPlayer_GetStandardConfig&media_id=%s&video_format=%s&video_quality=%s' - % (video_id, stream_format, stream_quality), - compat_urllib_parse_urlencode({'current_page': url}).encode('utf-8')) - streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded') - streamdata = self._download_xml( - streamdata_req, video_id, - note='Downloading media info for %s' % video_format) - stream_info = streamdata.find('./{default}preload/stream_info') - video_encode_id = xpath_text(stream_info, './video_encode_id') - if video_encode_id in video_encode_ids: - continue - video_encode_ids.append(video_encode_id) - - video_file = xpath_text(stream_info, './file') - if not video_file: - continue - if video_file.startswith('http'): - formats.extend(self._extract_m3u8_formats( - video_file, video_id, 'mp4', entry_protocol='m3u8_native', - m3u8_id='hls', fatal=False)) - continue + for stream in media.get('streams', []): + audio_lang = stream.get('audio_lang') + hardsub_lang = stream.get('hardsub_lang') + vrv_formats = self._extract_vrv_formats( + stream.get('url'), video_id, stream.get('format'), + audio_lang, hardsub_lang) + for f in vrv_formats: + if not hardsub_lang: + f['preference'] = 1 + language_preference = 0 + if audio_lang == language: + language_preference += 1 + if hardsub_lang == language: + language_preference += 1 + if language_preference: + f['language_preference'] = language_preference + formats.extend(vrv_formats) + if not formats: + available_fmts = [] + for a, fmt in re.findall(r'(<a[^>]+token=["\']showmedia\.([0-9]{3,4})p["\'][^>]+>)', webpage): + attrs = extract_attributes(a) + href = attrs.get('href') + if href and '/freetrial' in href: + continue + available_fmts.append(fmt) + if not available_fmts: + for p in (r'token=["\']showmedia\.([0-9]{3,4})p"', r'showmedia\.([0-9]{3,4})p'): + available_fmts = re.findall(p, webpage) + if available_fmts: + break + if not available_fmts: + available_fmts = self._FORMAT_IDS.keys() + video_encode_ids = [] + + for fmt in available_fmts: + stream_quality, stream_format = self._FORMAT_IDS[fmt] + video_format = fmt + 'p' + stream_infos = [] + streamdata = self._call_rpc_api( + 'VideoPlayer_GetStandardConfig', video_id, + 'Downloading media info for %s' % video_format, data={ + 'media_id': video_id, + 'video_format': stream_format, + 'video_quality': stream_quality, + 'current_page': url, + }) + if isinstance(streamdata, compat_etree_Element): + stream_info = streamdata.find('./{default}preload/stream_info') + if stream_info is not None: + stream_infos.append(stream_info) + stream_info = self._call_rpc_api( + 'VideoEncode_GetStreamInfo', video_id, + 'Downloading stream info for %s' % video_format, data={ + 'media_id': video_id, + 'video_format': stream_format, + 'video_encode_quality': stream_quality, + }) + if isinstance(stream_info, compat_etree_Element): + stream_infos.append(stream_info) + for stream_info in stream_infos: + video_encode_id = xpath_text(stream_info, './video_encode_id') + if video_encode_id in video_encode_ids: + continue + video_encode_ids.append(video_encode_id) + + video_file = xpath_text(stream_info, './file') + if not video_file: + continue + if video_file.startswith('http'): + formats.extend(self._extract_m3u8_formats( + video_file, video_id, 'mp4', entry_protocol='m3u8_native', + m3u8_id='hls', fatal=False)) + continue + + video_url = xpath_text(stream_info, './host') + if not video_url: + continue + metadata = stream_info.find('./metadata') + format_info = { + 'format': video_format, + 'height': int_or_none(xpath_text(metadata, './height')), + 'width': int_or_none(xpath_text(metadata, './width')), + } + + if '.fplive.net/' in video_url: + video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip()) + parsed_video_url = compat_urlparse.urlparse(video_url) + direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace( + netloc='v.lvlt.crcdn.net', + path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_file.split(':')[-1]))) + if self._is_valid_url(direct_video_url, video_id, video_format): + format_info.update({ + 'format_id': 'http-' + video_format, + 'url': direct_video_url, + }) + formats.append(format_info) + continue - video_url = xpath_text(stream_info, './host') - if not video_url: - continue - metadata = stream_info.find('./metadata') - format_info = { - 'format': video_format, - 'format_id': video_format, - 'height': int_or_none(xpath_text(metadata, './height')), - 'width': int_or_none(xpath_text(metadata, './width')), - } - - if '.fplive.net/' in video_url: - video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip()) - parsed_video_url = compat_urlparse.urlparse(video_url) - direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace( - netloc='v.lvlt.crcdn.net', - path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_file.split(':')[-1]))) - if self._is_valid_url(direct_video_url, video_id, video_format): format_info.update({ - 'url': direct_video_url, + 'format_id': 'rtmp-' + video_format, + 'url': video_url, + 'play_path': video_file, + 'ext': 'flv', }) formats.append(format_info) - continue - - format_info.update({ - 'url': video_url, - 'play_path': video_file, - 'ext': 'flv', - }) - formats.append(format_info) - self._sort_formats(formats) + self._sort_formats(formats, ('preference', 'language_preference', 'height', 'width', 'tbr', 'fps')) - metadata = self._download_xml( - 'http://www.crunchyroll.com/xml', video_id, - note='Downloading media info', query={ - 'req': 'RpcApiVideoPlayer_GetMediaMetadata', + metadata = self._call_rpc_api( + 'VideoPlayer_GetMediaMetadata', video_id, + note='Downloading media info', data={ 'media_id': video_id, }) - subtitles = self.extract_subtitles(video_id, webpage) + subtitles = {} + for subtitle in media.get('subtitles', []): + subtitle_url = subtitle.get('url') + if not subtitle_url: + continue + subtitles.setdefault(subtitle.get('language', 'enUS'), []).append({ + 'url': subtitle_url, + 'ext': subtitle.get('format', 'ass'), + }) + if not subtitles: + subtitles = self.extract_subtitles(video_id, webpage) # webpage provide more accurate data than series_title from XML series = self._html_search_regex( - r'id=["\']showmedia_about_episode_num[^>]+>\s*<a[^>]+>([^<]+)', + r'(?s)<h\d[^>]+\bid=["\']showmedia_about_episode_num[^>]+>(.+?)</h\d', webpage, 'series', fatal=False) - season = xpath_text(metadata, 'series_title') - episode = xpath_text(metadata, 'episode_title') - episode_number = int_or_none(xpath_text(metadata, 'episode_number')) + season = episode = episode_number = duration = thumbnail = None + + if isinstance(metadata, compat_etree_Element): + season = xpath_text(metadata, 'series_title') + episode = xpath_text(metadata, 'episode_title') + episode_number = int_or_none(xpath_text(metadata, 'episode_number')) + duration = float_or_none(media_metadata.get('duration'), 1000) + thumbnail = xpath_text(metadata, 'episode_image_url') + + if not episode: + episode = media_metadata.get('title') + if not episode_number: + episode_number = int_or_none(media_metadata.get('episode_number')) + if not thumbnail: + thumbnail = media_metadata.get('thumbnail', {}).get('url') season_number = int_or_none(self._search_regex( - r'(?s)<h4[^>]+id=["\']showmedia_about_episode_num[^>]+>.+?</h4>\s*<h4>\s*Season (\d+)', + r'(?s)<h\d[^>]+id=["\']showmedia_about_episode_num[^>]+>.+?</h\d>\s*<h4>\s*Season (\d+)', webpage, 'season number', default=None)) return { 'id': video_id, 'title': video_title, 'description': video_description, - 'thumbnail': xpath_text(metadata, 'episode_image_url'), + 'duration': duration, + 'thumbnail': thumbnail, 'uploader': video_uploader, 'upload_date': video_upload_date, 'series': series, @@ -540,7 +631,7 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE): IE_NAME = 'crunchyroll:playlist' - _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\-]+))/?(?:\?|$)' + _VALID_URL = r'https?://(?:(?P<prefix>www|m)\.)?(?P<url>crunchyroll\.com/(?!(?:news|anime-news|library|forum|launchcalendar|lineup|store|comics|freetrial|login|media-\d+))(?P<id>[\w\-]+))/?(?:\?|$)' _TESTS = [{ 'url': 'http://www.crunchyroll.com/a-bridge-to-the-starry-skies-hoshizora-e-kakaru-hashi', @@ -570,9 +661,8 @@ class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE): webpage = self._download_webpage( self._add_skip_wall(url), show_id, headers=self.geo_verification_headers()) - title = self._html_search_regex( - r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>', - webpage, 'title') + title = self._html_search_meta('name', webpage, default=None) + episode_paths = re.findall( r'(?s)<li id="showview_videos_media_(\d+)"[^>]+>.*?<a href="([^"]+)"', webpage)