added starttime option for ffmpeg decoders, fixed ffmpeg ogg encoding…

…, stepping stone for #390 and #275

audiotranscode/__init__.py

@@ -36,7 +36,7 @@ def __init__(self,filetype,command):
     def encode(self, decoder_process, bitrate):
         cmd = self.command[:]
         if 'BITRATE' in cmd:
-            cmd[cmd.index('BITRATE')] = str(bitrate)
+            cmd[cmd.index('BITRATE')] = '%dk' % bitrate
         return subprocess.Popen(cmd,
                                 stdin=decoder_process.stdout,
                                 stdout=subprocess.PIPE,
@@ -52,10 +52,13 @@ def __init__(self,filetype,command):
         self.mimetype = MimeTypes[filetype]
         self.command = command        
 
-    def decode(self, filepath):
+    def decode(self, filepath, starttime=0):
         cmd = self.command[:]
         if 'INPUT' in cmd:
             cmd[cmd.index('INPUT')] = filepath
+        if 'STARTTIME' in cmd:
+            hours, minutes, seconds = starttime//3600, starttime//60%60, starttime%60
+            cmd[cmd.index('STARTTIME')] = '%d:%d:%d' % (hours, minutes, seconds)
         return subprocess.Popen(cmd,
                                 stdout=subprocess.PIPE,
                                 stderr=Transcoder.devnull
@@ -88,7 +91,7 @@ class AudioTranscode:
     Encoders = [
         #encoders take input from stdin and write output to stout
         Encoder('ogg', ['oggenc', '-b','BITRATE','-']),
-        #Encoder('ogg', ['ffmpeg', '-i', '-', '-f', 'ogg', '-acodec', 'vorbis', '-']), #doesn't work yet
+        Encoder('ogg', ['ffmpeg', '-i', '-', '-f', 'ogg', '-acodec', 'libvorbis', '-ab', 'BITRATE', '-']),
         Encoder('mp3', ['lame','-b','BITRATE','-','-']),
         Encoder('mp3', ['ffmpeg', '-i', '-', '-f', 'mp3', '-acodec', 'libmp3lame', '-ab', 'BITRATE', '-']),
         Encoder('aac', ['faac','-b','BITRATE','-P','-X','-o','-','-']),
@@ -100,10 +103,10 @@ class AudioTranscode:
     Decoders = [
         #filepath must be appendable!
         Decoder('mp3'  , ['mpg123', '-w', '-', 'INPUT']),
-        Decoder('mp3'  , ['ffmpeg', '-i', 'INPUT', '-f', 'wav', '-acodec', 'pcm_s16le', '-']),
-        Decoder('wma'  , ['ffmpeg', '-i', 'INPUT', '-f', 'wav', '-acodec', 'pcm_s16le', '-']),
+        Decoder('mp3'  , ['ffmpeg', '-ss', 'STARTTIME', '-i', 'INPUT', '-f', 'wav', '-acodec', 'pcm_s16le', '-']),
+        Decoder('wma'  , ['ffmpeg', '-ss', 'STARTTIME', '-i', 'INPUT', '-f', 'wav', '-acodec', 'pcm_s16le', '-']),
         Decoder('ogg'  , ['oggdec', '-Q','-b', '16', '-o', '-', 'INPUT']),
-        Decoder('ogg'  , ['ffmpeg', '-i', 'INPUT', '-f', 'wav', '-acodec', 'pcm_s16le', '-']),
+        Decoder('ogg'  , ['ffmpeg', '-ss', 'STARTTIME', '-i', 'INPUT', '-f', 'wav', '-acodec', 'pcm_s16le', '-']),
         Decoder('flac' , ['flac', '-F','-d', '-c', 'INPUT']),
         Decoder('aac'  , ['faad', '-w', 'INPUT']), 
         Decoder('m4a'  , ['faad', '-w', 'INPUT']), 
@@ -126,7 +129,7 @@ def _filetype(self, filepath):
         if '.' in filepath:
             return filepath.lower()[filepath.rindex('.')+1:]
 
-    def _decode(self, filepath, decoder=None):
+    def _decode(self, filepath, decoder=None, starttime=0):
         if not os.path.exists(filepath):
             filepath = os.path.abspath(filepath)
             raise DecodeError('File not Found! Cannot decode "file" %s'%filepath)
@@ -140,7 +143,7 @@ def _decode(self, filepath, decoder=None):
                     break
             if self.debug:
                 print(decoder)
-        return decoder.decode(filepath)
+        return decoder.decode(filepath, starttime=starttime)
 
     def _encode(self, audio_format, decoder_process, bitrate=None,encoder=None):
         if not audio_format in self.availableEncoderFormats():
@@ -166,11 +169,12 @@ def transcode(self, in_file, out_file, bitrate=None):
                 fh.write(data)
             fh.close()
 
-    def transcodeStream(self, filepath, newformat,bitrate=None,encoder=None,decoder=None):
+    def transcodeStream(self, filepath, newformat, bitrate=None,
+            encoder=None, decoder=None, starttime=0):
         decoder_process = None
         encoder_process = None
         try:
-            decoder_process = self._decode(filepath, decoder)
+            decoder_process = self._decode(filepath, decoder, starttime=starttime)
             encoder_process = self._encode(newformat, decoder_process,bitrate=bitrate,encoder=encoder)
             while encoder_process.poll() == None:
                 data = encoder_process.stdout.read(AudioTranscode.READ_BUFFER)

cherrymusicserver/httphandler.py

@@ -251,11 +251,14 @@ def trans(self, newformat, *path, **params):
                 path = codecs.decode(codecs.encode(path, 'latin1'), 'utf-8')
             fullpath = os.path.join(cherry.config['media.basedir'], path)
 
+            starttime = int(params.pop('starttime', 0))
+
             transcoder = audiotranscode.AudioTranscode()
             mimetype = transcoder.mimeType(newformat)
             cherrypy.response.headers["Content-Type"] = mimetype
             try:
-                return transcoder.transcodeStream(fullpath, newformat, bitrate=bitrate)
+                return transcoder.transcodeStream(fullpath, newformat,
+                            bitrate=bitrate, starttime=starttime)
             except audiotranscode.TranscodeError as e:
                 raise cherrypy.HTTPError(404, e.value)
     trans.exposed = True

cherrymusicserver/test/test_httphandler.py

@@ -313,7 +313,7 @@ def test_trans(self):
 
                         httphandler.HTTPHandler(config).trans('newformat', 'path', bitrate=111)
 
-                        transcoder.transcodeStream.assert_called_with(expectPath, 'newformat', bitrate=111)
+                        transcoder.transcodeStream.assert_called_with(expectPath, 'newformat', bitrate=111, starttime=0)
 
 
 if __name__ == "__main__":