add format selection for AcFun

src/you_get/extractors/acfun.py

@@ -1,114 +1,159 @@
 #!/usr/bin/env python
 
-__all__ = ['acfun_download']
-
 from ..common import *
+from ..extractor import VideoExtractor
 
-from .le import letvcloud_download_by_vu
+class AcFun(VideoExtractor):
-from .qq import qq_download_by_vid
+    name = "AcFun"
-from .sina import sina_download_by_vid
-from .tudou import tudou_download_by_iid
-from .youku import youku_download_by_vid
 
-import json
+    stream_types = [
-import re
+        {'id': '2160P', 'qualityType': '2160p'},
-import base64
+        {'id': '1080P60', 'qualityType': '1080p60'},
-import time
+        {'id': '720P60', 'qualityType': '720p60'},
+        {'id': '1080P+', 'qualityType': '1080p+'},
+        {'id': '1080P', 'qualityType': '1080p'},
+        {'id': '720P', 'qualityType': '720p'},
+        {'id': '540P', 'qualityType': '540p'},
+        {'id': '360P', 'qualityType': '360p'}
+    ]    
 
-def get_srt_json(id):
+    def prepare(self, **kwargs):
-    url = 'http://danmu.aixifan.com/V2/%s' % id
+        assert re.match(r'https?://[^\.]*\.*acfun\.[^\.]+/(\D|bangumi)/\D\D(\d+)', self.url)
-    return get_content(url)
 
-def youku_acfun_proxy(vid, sign, ref):
+        if re.match(r'https?://[^\.]*\.*acfun\.[^\.]+/\D/\D\D(\d+)', self.url):
-    endpoint = 'http://player.acfun.cn/flash_data?vid={}&ct=85&ev=3&sign={}&time={}'
+            html = get_content(self.url, headers=fake_headers)
-    url = endpoint.format(vid, sign, str(int(time.time() * 1000)))
+            json_text = match1(html, r"(?s)videoInfo\s*=\s*(\{.*?\});")
-    json_data = json.loads(get_content(url, headers=dict(referer=ref)))['data']
+            json_data = json.loads(json_text)
-    enc_text = base64.b64decode(json_data)
+            vid = json_data.get('currentVideoInfo').get('id')
-    dec_text = rc4(b'8bdc7e1a', enc_text).decode('utf8')
+            up = json_data.get('user').get('name')
-    youku_json = json.loads(dec_text)
+            self.title = json_data.get('title')
+            video_list = json_data.get('videoList')
+            if len(video_list) > 1:
+                self.title += " - " + [p.get('title') for p in video_list if p.get('id') == vid][0]
+            currentVideoInfo = json_data.get('currentVideoInfo')
+
+        elif re.match("https?://[^\.]*\.*acfun\.[^\.]+/bangumi/aa(\d+)", self.url):
+            html = get_content(self.url, headers=fake_headers)
+            tag_script = match1(html, r'<script>\s*window\.pageInfo([^<]+)</script>')
+            json_text = tag_script[tag_script.find('{') : tag_script.find('};') + 1]
+            json_data = json.loads(json_text)
+            self.title = json_data['bangumiTitle'] + " " + json_data['episodeName'] + " " + json_data['title']
+            vid = str(json_data['videoId'])
+            up = "acfun"
+            currentVideoInfo = json_data.get('currentVideoInfo')
 
-    yk_streams = {}
-    for stream in youku_json['stream']:
-        tp = stream['stream_type']
-        yk_streams[tp] = [], stream['total_size']
-        if stream.get('segs'):
-            for seg in stream['segs']:
-                yk_streams[tp][0].append(seg['url'])
         else:
-            yk_streams[tp] = stream['m3u8'], stream['total_size']
+            raise NotImplemented            
 
-    return yk_streams
+        if 'ksPlayJson' in currentVideoInfo:
+            durationMillis = currentVideoInfo['durationMillis']
+            ksPlayJson = ksPlayJson = json.loads( currentVideoInfo['ksPlayJson'] )
+            representation = ksPlayJson.get('adaptationSet')[0].get('representation')
+            stream_list = representation
 
-def acfun_download_by_vid(vid, title, output_dir='.', merge=True, info_only=False, **kwargs):
+        for stream in stream_list:
-    """str, str, str, bool, bool ->None
+            m3u8_url = stream["url"]
+            size = durationMillis * stream["avgBitrate"] / 8
+            # size = float('inf')
+            container = 'mp4'
+            stream_id = stream["qualityLabel"]
+            quality = stream["qualityType"]
             
-    Download Acfun video by vid.
+            stream_data = dict(src=m3u8_url, size=size, container=container, quality=quality)
+            self.streams[stream_id] = stream_data
 
-    Call Acfun API, decide which site to use, and pass the job to its
+        assert self.title and m3u8_url
-    extractor.
+        self.title = unescape_html(self.title)
-    """
+        self.title = escape_file_path(self.title)
+        p_title = r1('active">([^<]+)', html)
+        self.title = '%s (%s)' % (self.title, up)
+        if p_title:
+            self.title = '%s - %s' % (self.title, p_title)       
 
-    #first call the main parasing API
-    info = json.loads(get_content('http://www.acfun.cn/video/getVideo.aspx?id=' + vid, headers=fake_headers))
 
-    sourceType = info['sourceType']
+    def download(self, **kwargs):
-
+        if 'json_output' in kwargs and kwargs['json_output']:
-    #decide sourceId to know which extractor to use
+            json_output.output(self)
-    if 'sourceId' in info: sourceId = info['sourceId']
+        elif 'info_only' in kwargs and kwargs['info_only']:
-    # danmakuId = info['danmakuId']
+            if 'stream_id' in kwargs and kwargs['stream_id']:
-
+                # Display the stream
-    #call extractor decided by sourceId
+                stream_id = kwargs['stream_id']
-    if sourceType == 'sina':
+                if 'index' not in kwargs:
-        sina_download_by_vid(sourceId, title, output_dir=output_dir, merge=merge, info_only=info_only)
+                    self.p(stream_id)
-    elif sourceType == 'youku':
-        youku_download_by_vid(sourceId, title=title, output_dir=output_dir, merge=merge, info_only=info_only, **kwargs)
-    elif sourceType == 'tudou':
-        tudou_download_by_iid(sourceId, title, output_dir=output_dir, merge=merge, info_only=info_only)
-    elif sourceType == 'qq':
-        qq_download_by_vid(sourceId, title, True, output_dir=output_dir, merge=merge, info_only=info_only)
-    elif sourceType == 'letv':
-        letvcloud_download_by_vu(sourceId, '2d8c027396', title, output_dir=output_dir, merge=merge, info_only=info_only)
-    elif sourceType == 'zhuzhan':
-        #As in Jul.28.2016, Acfun is using embsig to anti hotlink so we need to pass this
-#In Mar. 2017 there is a dedicated ``acfun_proxy'' in youku cloud player
-#old code removed
-        url = 'http://www.acfun.cn/v/ac' + vid
-        yk_streams = youku_acfun_proxy(info['sourceId'], info['encode'], url)
-        seq = ['mp4hd3', 'mp4hd2', 'mp4hd', 'flvhd']
-        for t in seq:
-            if yk_streams.get(t):
-                preferred = yk_streams[t]
-                break
-#total_size in the json could be incorrect(F.I. 0)
-        size = 0
-        for url in preferred[0]:
-            _, _, seg_size = url_info(url)
-            size += seg_size
-#fallback to flvhd is not quite possible
-        if re.search(r'fid=[0-9A-Z\-]*.flv', preferred[0][0]):
-            ext = 'flv'
                 else:
+                    self.p_i(stream_id)
+            else:
+                # Display all available streams
+                if 'index' not in kwargs:
+                    self.p([])
+                else:
+                    stream_id = self.streams_sorted[0]['id'] if 'id' in self.streams_sorted[0] else self.streams_sorted[0]['itag']
+                    self.p_i(stream_id)
+
+        else:
+            if 'stream_id' in kwargs and kwargs['stream_id']:
+                # Download the stream
+                stream_id = kwargs['stream_id']
+            else:
+                stream_id = self.streams_sorted[0]['id'] if 'id' in self.streams_sorted[0] else self.streams_sorted[0]['itag']
+
+            if 'index' not in kwargs:
+                self.p(stream_id)
+            else:
+                self.p_i(stream_id)
+            if stream_id in self.streams:
+                url = self.streams[stream_id]['src']
+                ext = self.streams[stream_id]['container']
+                total_size = self.streams[stream_id]['size']
+
+
+            if ext == 'm3u8' or ext == 'm4a':
                 ext = 'mp4'
-        print_info(site_info, title, ext, size)
-        if not info_only:
-            download_urls(preferred[0], title, ext, size, output_dir=output_dir, merge=merge)
-    else:
-        raise NotImplementedError(sourceType)
 
-    if not info_only and not dry_run:
+            if not url:
-        if not kwargs['caption']:
+                log.wtf('[Failed] Cannot extract video source.')
-            print('Skipping danmaku.')
+            # For legacy main()
+            headers = {}
+            if self.ua is not None:
+                headers['User-Agent'] = self.ua
+            if self.referer is not None:
+                headers['Referer'] = self.referer
+
+            download_url_ffmpeg(url, self.title, ext, output_dir=kwargs['output_dir'], merge=kwargs['merge'])                           
+
+            if 'caption' not in kwargs or not kwargs['caption']:
+                print('Skipping captions or danmaku.')
                 return
-        try:
-            title = get_filename(title)
-            print('Downloading %s ...\n' % (title + '.cmt.json'))
-            cmt = get_srt_json(vid)
-            with open(os.path.join(output_dir, title + '.cmt.json'), 'w', encoding='utf-8') as x:
-                x.write(cmt)
-        except:
-            pass
 
-def acfun_download(url, output_dir='.', merge=True, info_only=False, **kwargs):
+            for lang in self.caption_tracks:
+                filename = '%s.%s.srt' % (get_filename(self.title), lang)
+                print('Saving %s ... ' % filename, end="", flush=True)
+                srt = self.caption_tracks[lang]
+                with open(os.path.join(kwargs['output_dir'], filename),
+                          'w', encoding='utf-8') as x:
+                    x.write(srt)
+                print('Done.')
+
+            if self.danmaku is not None and not dry_run:
+                filename = '{}.cmt.xml'.format(get_filename(self.title))
+                print('Downloading {} ...\n'.format(filename))
+                with open(os.path.join(kwargs['output_dir'], filename), 'w', encoding='utf8') as fp:
+                    fp.write(self.danmaku)
+
+            if self.lyrics is not None and not dry_run:
+                filename = '{}.lrc'.format(get_filename(self.title))
+                print('Downloading {} ...\n'.format(filename))
+                with open(os.path.join(kwargs['output_dir'], filename), 'w', encoding='utf8') as fp:
+                    fp.write(self.lyrics)
+
+            # For main_dev()
+            #download_urls(urls, self.title, self.streams[stream_id]['container'], self.streams[stream_id]['size'])
+        keep_obj = kwargs.get('keep_obj', False)
+        if not keep_obj:
+            self.__init__()
+
+
+    def acfun_download(self, url, output_dir='.', merge=True, info_only=False, **kwargs):
         assert re.match(r'https?://[^\.]*\.*acfun\.[^\.]+/(\D|bangumi)/\D\D(\d+)', url)
 
         def getM3u8UrlFromCurrentVideoInfo(currentVideoInfo):
@@ -162,7 +207,7 @@ def acfun_download(url, output_dir='.', merge=True, info_only=False, **kwargs):
         if not info_only:
             download_url_ffmpeg(m3u8_url, title, 'mp4', output_dir=output_dir, merge=merge)
 
-
+site = AcFun()
 site_info = "AcFun.cn"
-download = acfun_download
+download = site.download_by_url
 download_playlist = playlist_not_supported('acfun')