]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/crunchyroll.py
Prepare to upload
[youtubedl] / youtube_dl / extractor / crunchyroll.py
index 8bdaf0c2c5af7aa1a562033487a16867c3304f19..85a9a577f645395d6edde83e7b72a1f4001561f8 100644 (file)
@@ -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 '<title>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:(?!.*&amp)',
@@ -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(
@@ -427,95 +460,152 @@ 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'(?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)<h\d[^>]+id=["\']showmedia_about_episode_num[^>]+>.+?</h\d>\s*<h4>\s*Season (\d+)',
@@ -525,7 +615,8 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
             '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)