portaudio-py-client.py

# Silvius microphone client based on Tanel's client.py
__author__ = 'dwk'

import argparse
from ws4py.client.threadedclient import WebSocketClient
import threading
import sys
import urllib
import json

reconnect_mode = False
fatal_error = False

class MyClient(WebSocketClient):

    def __init__(self, url, mic=1, protocols=None, extensions=None, heartbeat_freq=None, byterate=16000,
                 show_hypotheses=True,
                 save_adaptation_state_filename=None, send_adaptation_state_filename=None, audio_gate=0):
        super(MyClient, self).__init__(url, protocols, extensions, heartbeat_freq)
        self.mic = mic
        self.show_hypotheses = show_hypotheses
        self.byterate = byterate
        self.save_adaptation_state_filename = save_adaptation_state_filename
        self.send_adaptation_state_filename = send_adaptation_state_filename
        self.chunk = 0
        self.audio_gate = audio_gate

    def send_data(self, data):
        self.send(data, binary=True)

    def opened(self):
        import pyaudio
        import audioop
        pa = pyaudio.PyAudio()
        sample_rate = self.byterate
        stream = None 
        
        while stream is None:
            try:
                # try adjusting this if you want fewer network packets
                self.chunk = 2048 * 2 * sample_rate / self.byterate

                mic = self.mic
                if mic == -1:
                    mic = pa.get_default_input_device_info()['index']
                    print >> sys.stderr, "Selecting default mic"
                print >> sys.stderr, "Using mic #", mic
                stream = pa.open(
                    rate = sample_rate,
                    format = pyaudio.paInt16,
                    channels = 1,
                    input = True,
                    input_device_index = mic,
                    frames_per_buffer = self.chunk)
            except IOError, e:
                if(e.errno == -9997 or e.errno == 'Invalid sample rate'):
                    new_sample_rate = int(pa.get_device_info_by_index(mic)['defaultSampleRate'])
                    if(sample_rate != new_sample_rate):
                        sample_rate = new_sample_rate
                        continue
                print >> sys.stderr, "\n", e
                print >> sys.stderr, "\nCould not open microphone. Please try a different device."
                global fatal_error
                fatal_error = True
                sys.exit(0)
     
        def mic_to_ws():  # uses stream
            try:
                print >> sys.stderr, "\nLISTENING TO MICROPHONE"
                last_state = None
                while True:
                    data = stream.read(self.chunk)
                    if self.audio_gate > 0:
                        rms = audioop.rms(data, 2)
                        if rms < self.audio_gate:
                            data = '\00' * len(data)
                    #if sample_chan == 2:
                    #    data = audioop.tomono(data, 2, 1, 1)
                    if sample_rate != self.byterate:
                        (data, last_state) = audioop.ratecv(data, 2, 1, sample_rate, self.byterate, last_state)

                    self.send_data(data)
            except IOError, e:
                # usually a broken pipe
                print e
            except AttributeError:
                # currently raised when the socket gets closed by main thread
                pass

            # to voluntarily close the connection, we would use
            #self.send_data("")
            #self.send("EOS")

            try:
                self.close()
            except IOError:
                pass

        threading.Thread(target=mic_to_ws).start()


    def received_message(self, m):
        response = json.loads(str(m))
        #print >> sys.stderr, "RESPONSE:", response
        #print >> sys.stderr, "JSON was:", m
        if response['status'] == 0:
            if 'result' in response:
                trans = response['result']['hypotheses'][0]['transcript']
                if response['result']['final']:
                    if self.show_hypotheses:
                        print >> sys.stderr, '\r%s' % trans.replace("\n", "\\n")
                    print '%s' % trans.replace("\n", "\\n")  # final result!
                    sys.stdout.flush()
                elif self.show_hypotheses:
                    print_trans = trans.replace("\n", "\\n")
                    if len(print_trans) > 80:
                        print_trans = "... %s" % print_trans[-76:]
                    print >> sys.stderr, '\r%s' % print_trans,
            if 'adaptation_state' in response:
                if self.save_adaptation_state_filename:
                    print >> sys.stderr, "Saving adaptation state to %s" % self.save_adaptation_state_filename
                    with open(self.save_adaptation_state_filename, "w") as f:
                        f.write(json.dumps(response['adaptation_state']))
        else:
            print >> sys.stderr, "Received error from server (status %d)" % response['status']
            if 'message' in response:
                print >> sys.stderr, "Error message:",  response['message']
            
            global reconnect_mode
            if reconnect_mode:
                import time
                print >> sys.stderr, "Sleeping for five seconds before reconnecting"
                time.sleep(5)


    def closed(self, code, reason=None):
        #print "Websocket closed() called"
        #print >> sys.stderr
        pass


def setup():
    content_type = "audio/x-raw, layout=(string)interleaved, rate=(int)16000, format=(string)S16LE, channels=(int)1"
    path = 'client/ws/speech'

    parser = argparse.ArgumentParser(description='Microphone client for silvius')
    parser.add_argument('-s', '--server', default="localhost", dest="server", help="Speech-recognition server")
    parser.add_argument('-p', '--port', default="8019", dest="port", help="Server port")
    #parser.add_argument('-r', '--rate', default=16000, dest="rate", type=int, help="Rate in bytes/sec at which audio should be sent to the server.")
    parser.add_argument('-d', '--device', default="-1", dest="device", type=int, help="Select a different microphone (give device ID)")
    parser.add_argument('-k', '--keep-going', action="store_true", help="Keep reconnecting to the server after periods of silence")
    parser.add_argument('--save-adaptation-state', help="Save adaptation state to file")
    parser.add_argument('--send-adaptation-state', help="Send adaptation state from file")
    parser.add_argument('--content-type', default=content_type, help="Use the specified content type (default is " + content_type + ")")
    parser.add_argument('--hypotheses', default=True, type=int, help="Show partial recognition hypotheses (default: 1)")
    parser.add_argument('-g', '--audio-gate', default=0, type=int, help="Audio-gate level to reduce detections when not talking")
    args = parser.parse_args()

    content_type = args.content_type
    print >> sys.stderr, "Content-Type:", content_type

    if(args.keep_going):
        global reconnect_mode
        global fatal_error
        reconnect_mode = True
        while(fatal_error == False):
            print >> sys.stderr, "Reconnecting..."
            run(args, content_type, path)
    else:
        run(args, content_type, path)

def run(args, content_type, path):
    uri = "ws://%s:%s/%s?%s" % (args.server, args.port, path, urllib.urlencode([("content-type", content_type)]))
    print >> sys.stderr, "Connecting to", uri

    ws = MyClient(uri, byterate=16000, mic=args.device, show_hypotheses=args.hypotheses,
                  save_adaptation_state_filename=args.save_adaptation_state, send_adaptation_state_filename=args.send_adaptation_state, audio_gate=args.audio_gate)
    ws.connect()
    #result = ws.get_full_hyp()
    #print result.encode('utf-8')
    ws.run_forever()

def main():
    try:
        setup()
    except KeyboardInterrupt:
        print >> sys.stderr, "\nexiting..."

if __name__ == "__main__":
    main()