fix audio file bug: add the ability to read multiple audio file format.

Dockerfile

@@ -23,6 +23,7 @@ RUN apt-get update &&\
  libtool-bin \
  make \
  sox \
+ libsox-fmt-all \
  libav-tools \
  subversion \
  vorbis-tools \

document/swagger.yml

@@ -21,7 +21,7 @@ paths:
  produces:
  - "application/json"
  parameters: 
- - name: "wavFile"
+ - name: "file"
  in: "formData"
  description: "Wave File"
  required: true

run.py

@@ -45,14 +45,14 @@ def run_shell_command(command_line):
  return False, ''
 
 def decode(wav_file,wav_name):
- b,o=run_shell_command("sox "+wav_file+".wav -t wav -b 16 -r 16000 -c 1 "+wav_file+"_tmp.wav")
- if not b:
- return False, ''
+ b,o=run_shell_command("sox "+wav_file+" -t wav -b 16 -r 16000 -c 1 "+wav_file+"_tmp.wav")
+ if not b or 'error' in o or 'FAIL' in o:
+ return False, 'Audio file format is not supported!!! Supported formats are wav, mp3, aiff, flac, and ogg.'
  b,o=run_shell_command("mv "+wav_file+"_tmp.wav "+wav_file+".wav")
  if not b:
  return False, ''
 
-
+ wav_file = wav_file+".wav"
  decode_conf = TEMP_FILE_PATH1+"/online.conf"
  decode_mdl = AM_PATH+"/"+AM_FILE_PATH+"/final.mdl"
  decode_graph = LM_PATH+"/HCLG.fst"
@@ -64,12 +64,9 @@ def decode(wav_file,wav_name):
  b,o=run_shell_command("kaldi-nnet2-latgen-faster --do-endpointing=false --online=false --config="+decode_conf+" --min-active="+DECODER_MINACT+" --max-active="+DECODER_MAXACT+" --beam="+DECODER_BEAM+" --lattice-beam="+DECODER_LATBEAM+" --acoustic-scale="+DECODER_ACWT+" --word-symbol-table="+decode_words+" "+decode_mdl+" "+decode_graph+" \"ark:echo "+wav_name+" "+wav_name+"|\" \"scp:echo "+wav_name+" "+wav_file+"|\" ark:"+TEMP_FILE_PATH+"/"+wav_name+".lat")
  else:
  b=False
- o='KaldiFatalError decode param is not recognized'
 
  if not b or 'KaldiFatalError' in o:
- print(o)
- return False, ''
-
+ return False, 'The decoder params of the acoustic model in "decode.cfg" file are not correct!!'
  hypothesis = re.findall('\n'+wav_name+'.*',o)
  #app.logger.info(hypothesis)
  o=re.sub(wav_name,'',hypothesis[0]).strip()
@@ -83,17 +80,23 @@ def transcribe():
  global busy
  busy=1
  fileid = str(uuid.uuid4())
- if 'wavFile' in request.files.keys():
- file = request.files['wavFile']
- filename = TEMP_FILE_PATH+'/'+fileid+'.wav'
- file.save(filename)
- b, out = decode(filename,fileid)
- if not b:
+ if 'file' in request.files.keys():
+ file = request.files['file']
+ file_ext = file.filename.rsplit('.', 1)[-1].lower()
+ file_type = file.content_type.rsplit('/', 1)[0]
+ if file_type == "audio":
+ filename = TEMP_FILE_PATH+'/'+fileid+'.'+file_ext
+ file.save(filename)
+ b, out = decode(filename,fileid)
+ if not b:
+ busy=0
+ return 'Error while file transcription: '+out, 400
+ else:
  busy=0
- abort(403)
+ return 'Error while file transcription: The uploaded file format is not supported!!! Supported formats are wav, mp3, aiff, flac, and ogg.', 400
  else:
  busy=0
- return 'No wave file was uploaded', 404
+ return 'No audio file was uploaded', 400
 
  # Delete temporary files
  for file in os.listdir(TEMP_FILE_PATH):