Ejemplo n.º 1
0
 def setUp(self):
     stream = BytesIO()
     w = wave.open(stream, 'wb')
     sample_rate = 16000
     bytes_per_sample = 2
     w.setframerate(sample_rate)
     w.setsampwidth(bytes_per_sample)
     w.setnchannels(1)
     w.writeframes(b'audiodata')
     self.stream = BytesIO(stream.getvalue())
     self.source = audio_helpers.WaveSource(self.stream, sample_rate,
                                            bytes_per_sample)
     self.sleep_time_1024 = self.source._sleep_time(1024)
     self.sleep_time_512 = self.source._sleep_time(512)
Ejemplo n.º 2
0
def main(keyword_file_paths, library_path, model_file_path, asensitivities,
         input_audio_device_index, output_path, show_audio_devices_info,
         api_endpoint, credentials, project_id, device_model_id, device_id,
         device_config, lang, display, verbose, input_audio_file,
         output_audio_file, audio_sample_rate, audio_sample_width,
         audio_iter_size, audio_block_size, audio_flush_size, grpc_deadline,
         once, *args, **kwargs):

    # Load OAuth 2.0 credentials.
    try:
        with open(credentials, 'r') as f:
            credentials = google.oauth2.credentials.Credentials(token=None,
                                                                **json.load(f))
            http_request = google.auth.transport.requests.Request()
            #print http_request
            credentials.refresh(http_request)
    except Exception as e:
        logging.error('Error loading credentials: %s', e)
        logging.error('Run google-oauthlib-tool to initialize '
                      'new OAuth 2.0 credentials.')
        sys.exit(-1)

    # Create an authorized gRPC channel.
    grpc_channel = google.auth.transport.grpc.secure_authorized_channel(
        credentials, http_request, api_endpoint)
    logging.info('Connecting to %s', api_endpoint)

    # Configure audio source and sink.
    audio_device = None
    if input_audio_file:
        audio_source = audio_helpers.WaveSource(
            open(input_audio_file, 'rb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width)
    else:
        audio_source = audio_device = (audio_device
                                       or audio_helpers.SoundDeviceStream(
                                           sample_rate=audio_sample_rate,
                                           sample_width=audio_sample_width,
                                           block_size=audio_block_size,
                                           flush_size=audio_flush_size))
    if output_audio_file:
        audio_sink = audio_helpers.WaveSink(open(output_audio_file, 'wb'),
                                            sample_rate=audio_sample_rate,
                                            sample_width=audio_sample_width)
    else:
        audio_sink = audio_device = (audio_device
                                     or audio_helpers.SoundDeviceStream(
                                         sample_rate=audio_sample_rate,
                                         sample_width=audio_sample_width,
                                         block_size=audio_block_size,
                                         flush_size=audio_flush_size))
    # Create conversation stream with the given audio source and sink.
    conversation_stream = audio_helpers.ConversationStream(
        source=audio_source,
        sink=audio_sink,
        iter_size=audio_iter_size,
        sample_width=audio_sample_width,
    )

    if not device_id or not device_model_id:
        try:
            with open(device_config) as f:
                device = json.load(f)
                device_id = device['id']
                device_model_id = device['model_id']
                logging.info("Using device model %s and device id %s",
                             device_model_id, device_id)
        except Exception as e:
            logging.warning('Device config not found: %s' % e)
            logging.info('Registering device')
            if not device_model_id:
                logging.error('Option --device-model-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            if not project_id:
                logging.error('Option --project-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            device_base_url = ('https://%s/v1alpha2/projects/%s/devices' %
                               (api_endpoint, project_id))
            device_id = str(uuid.uuid1())
            payload = {
                'id': device_id,
                'model_id': device_model_id,
                'client_type': 'SDK_SERVICE'
            }
            session = google.auth.transport.requests.AuthorizedSession(
                credentials)
            r = session.post(device_base_url, data=json.dumps(payload))
            if r.status_code != 200:
                logging.error('Failed to register device: %s', r.text)
                sys.exit(-1)
            logging.info('Device registered: %s', device_id)
            pathlib.Path(os.path.dirname(device_config)).mkdir(exist_ok=True)
            with open(device_config, 'w') as f:
                json.dump(payload, f)

    device_handler = device_helpers.DeviceRequestHandler(device_id)

    @device_handler.command('action.devices.commands.OnOff')
    def onoff(on):
        if on:
            logging.info('Turning device on')
        else:
            logging.info('Turning device off')

    @device_handler.command('com.example.commands.BlinkLight')
    def blink(speed, number):
        logging.info('Blinking device %s times.' % number)
        delay = 1
        if speed == "slowly":
            delay = 2
        elif speed == "quickly":
            delay = 0.5
        for i in range(int(number)):
            logging.info('Device is blinking.')
            time.sleep(delay)

    if show_audio_devices_info:
        PorcupineDemo.show_audio_devices_info()
    else:
        if not keyword_file_paths:
            raise ValueError('keyword file paths are missing')

        keyword_file_paths = [x.strip() for x in keyword_file_paths.split(',')]

        if isinstance(asensitivities, float):
            sensitivities = [asensitivities] * len(keyword_file_paths)
        else:
            sensitivities = [float(x) for x in asensitivities.split(',')]

        PorcupineDemo(lang,
                      device_model_id,
                      device_id,
                      conversation_stream,
                      display,
                      grpc_channel,
                      grpc_deadline,
                      device_handler,
                      library_path,
                      model_file_path,
                      keyword_file_paths,
                      sensitivities,
                      input_device_index=input_audio_device_index,
                      output_path=output_path).run()
        print input_audio_device_index
Ejemplo n.º 3
0
def main(api_endpoint, credentials, project_id,
         device_model_id, device_id, device_config,
         lang, display, verbose,
         input_audio_file, output_audio_file,
         audio_sample_rate, audio_sample_width,
         audio_iter_size, audio_block_size, audio_flush_size,
         grpc_deadline, once, *args, **kwargs):
    """Samples for the Google Assistant API.

    Examples:
      Run the sample with microphone input and speaker output:

        $ python -m googlesamples.assistant

      Run the sample with file input and speaker output:

        $ python -m googlesamples.assistant -i <input file>

      Run the sample with file input and output:

        $ python -m googlesamples.assistant -i <input file> -o <output file>
    """
    if gender=='Male':
        subprocess.Popen(["aplay", "{}/sample-audio-files/Startup-Male.wav".format(ROOT_PATH)], stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    else:
        subprocess.Popen(["aplay", "{}/sample-audio-files/Startup-Female.wav".format(ROOT_PATH)], stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    # Setup logging.
    logging.basicConfig(level=logging.DEBUG if verbose else logging.INFO)

    # Load OAuth 2.0 credentials.
    try:
        with open(credentials, 'r') as f:
            credentials = google.oauth2.credentials.Credentials(token=None,
                                                                **json.load(f))
            http_request = google.auth.transport.requests.Request()
            credentials.refresh(http_request)
    except Exception as e:
        logging.error('Error loading credentials: %s', e)
        logging.error('Run google-oauthlib-tool to initialize '
                      'new OAuth 2.0 credentials.')
        sys.exit(-1)

    # Create an authorized gRPC channel.
    grpc_channel = google.auth.transport.grpc.secure_authorized_channel(
        credentials, http_request, api_endpoint)
    logging.info('Connecting to %s', api_endpoint)

    # Configure audio source and sink.
    audio_device = None
    if input_audio_file:
        audio_source = audio_helpers.WaveSource(
            open(input_audio_file, 'rb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width
        )
    else:
        audio_source = audio_device = (
            audio_device or audio_helpers.SoundDeviceStream(
                sample_rate=audio_sample_rate,
                sample_width=audio_sample_width,
                block_size=audio_block_size,
                flush_size=audio_flush_size
            )
        )
    if output_audio_file:
        audio_sink = audio_helpers.WaveSink(
            open(output_audio_file, 'wb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width
        )
    else:
        audio_sink = audio_device = (
            audio_device or audio_helpers.SoundDeviceStream(
                sample_rate=audio_sample_rate,
                sample_width=audio_sample_width,
                block_size=audio_block_size,
                flush_size=audio_flush_size
            )
        )
    # Create conversation stream with the given audio source and sink.
    conversation_stream = audio_helpers.ConversationStream(
        source=audio_source,
        sink=audio_sink,
        iter_size=audio_iter_size,
        sample_width=audio_sample_width,
    )

    if not device_id or not device_model_id:
        try:
            with open(device_config) as f:
                device = json.load(f)
                device_id = device['id']
                device_model_id = device['model_id']
                logging.info("Using device model %s and device id %s",
                             device_model_id,
                             device_id)
        except Exception as e:
            logging.warning('Device config not found: %s' % e)
            logging.info('Registering device')
            if not device_model_id:
                logging.error('Option --device-model-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            if not project_id:
                logging.error('Option --project-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            device_base_url = (
                'https://%s/v1alpha2/projects/%s/devices' % (api_endpoint,
                                                             project_id)
            )
            device_id = str(uuid.uuid1())
            payload = {
                'id': device_id,
                'model_id': device_model_id,
                'client_type': 'SDK_SERVICE'
            }
            session = google.auth.transport.requests.AuthorizedSession(
                credentials
            )
            r = session.post(device_base_url, data=json.dumps(payload))
            if r.status_code != 200:
                logging.error('Failed to register device: %s', r.text)
                sys.exit(-1)
            logging.info('Device registered: %s', device_id)
            pathlib.Path(os.path.dirname(device_config)).mkdir(exist_ok=True)
            with open(device_config, 'w') as f:
                json.dump(payload, f)

    device_handler = device_helpers.DeviceRequestHandler(device_id)

    @device_handler.command('action.devices.commands.OnOff')
    def onoff(on):
        if on:
            logging.info('Turning device on')
        else:
            logging.info('Turning device off')



    @device_handler.command('com.example.commands.BlinkLight')
    def blink(speed, number):
        logging.info('Blinking device %s times.' % number)
        delay = 1
        if speed == "slowly":
            delay = 2
        elif speed == "quickly":
            delay = 0.5
        for i in range(int(number)):
            logging.info('Device is blinking.')
            time.sleep(delay)

    with SampleAssistant(lang, device_model_id, device_id,
                         conversation_stream, display,
                         grpc_channel, grpc_deadline,
                         device_handler) as assistant:
        # If file arguments are supplied:
        # exit after the first turn of the conversation.
        if input_audio_file or output_audio_file:
            assistant.assist()
            return

        def detected():
            continue_conversation=assistant.assist()
            if continue_conversation:
                print('Continuing conversation')
                assistant.assist()

        signal.signal(signal.SIGINT, signal_handler)
        sensitivity = [0.5]*len(models)
        callbacks = [detected]*len(models)
        detector = snowboydecoder.HotwordDetector(models, sensitivity=sensitivity)
        def start_detector():
            detector.start(detected_callback=callbacks,
                                   interrupt_check=interrupt_callback,
                                   sleep_time=0.03)

        # If no file arguments supplied:
        # keep recording voice requests using the microphone
        # and playing back assistant response using the speaker.
        # When the once flag is set, don't wait for a trigger. Otherwise, wait.

        wait_for_user_trigger = not once
        while True:
            if wait_for_user_trigger:
                if custom_wakeword:
                    start_detector()
                else:
                    button_state=GPIO.input(pushbuttontrigger)
                    if button_state==True:
                       continue
                    else:
                       pass
            continue_conversation = assistant.assist()
            # wait for user trigger if there is no follow-up turn in
            # the conversation.
            wait_for_user_trigger = not continue_conversation

            # If we only want one conversation, break.
            if once and (not continue_conversation):
                break



    detector.terminate()
Ejemplo n.º 4
0
def main(api_endpoint, credentials, project_id,
         device_model_id, device_id, device_config, lang, verbose,
         input_audio_file, output_audio_file,
         audio_sample_rate, audio_sample_width,
         audio_iter_size, audio_block_size, audio_flush_size,
         grpc_deadline, once, *args, **kwargs):
    """Samples for the Google Assistant API.

    Examples:
      Run the sample with microphone input and speaker output:

        $ python -m googlesamples.assistant

      Run the sample with file input and speaker output:

        $ python -m googlesamples.assistant -i <input file>

      Run the sample with file input and output:

        $ python -m googlesamples.assistant -i <input file> -o <output file>
    """
    # Setup logging.
    logging.basicConfig(level=logging.DEBUG if verbose else logging.INFO)

    # Load OAuth 2.0 credentials.
    try:
        with open(credentials, 'r') as f:
            credentials = google.oauth2.credentials.Credentials(token=None,
                                                                **json.load(f))
            http_request = google.auth.transport.requests.Request()
            credentials.refresh(http_request)
    except Exception as e:
        logging.error('Error loading credentials: %s', e)
        logging.error('Run google-oauthlib-tool to initialize '
                      'new OAuth 2.0 credentials.')
        sys.exit(-1)

    # Create an authorized gRPC channel.
    grpc_channel = google.auth.transport.grpc.secure_authorized_channel(
        credentials, http_request, api_endpoint)
    logging.info('Connecting to %s', api_endpoint)

    # Configure audio source and sink.
    audio_device = None
    if input_audio_file:
        audio_source = audio_helpers.WaveSource(
            open(input_audio_file, 'rb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width
        )
    else:
        audio_source = audio_device = (
            audio_device or audio_helpers.SoundDeviceStream(
                sample_rate=audio_sample_rate,
                sample_width=audio_sample_width,
                block_size=audio_block_size,
                flush_size=audio_flush_size
            )
        )
    if output_audio_file:
        audio_sink = audio_helpers.WaveSink(
            open(output_audio_file, 'wb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width
        )
    else:
        audio_sink = audio_device = (
            audio_device or audio_helpers.SoundDeviceStream(
                sample_rate=audio_sample_rate,
                sample_width=audio_sample_width,
                block_size=audio_block_size,
                flush_size=audio_flush_size
            )
        )
    # Create conversation stream with the given audio source and sink.
    conversation_stream = audio_helpers.ConversationStream(
        source=audio_source,
        sink=audio_sink,
        iter_size=audio_iter_size,
        sample_width=audio_sample_width,
    )

    device_handler = device_helpers.DeviceRequestHandler(device_id)

    @device_handler.command('action.devices.commands.OnOff')
    def onoff(on):
        if on:
            logging.info('Turning device on')
        else:
            logging.info('Turning device off')

    if not device_id or not device_model_id:
        try:
            with open(device_config) as f:
                device = json.load(f)
                device_id = device['id']
                device_model_id = device['model_id']
        except Exception as e:
            logging.warning('Device config not found: %s' % e)
            logging.info('Registering device')
            if not device_model_id:
                logging.error('Option --device-model-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            if not project_id:
                logging.error('Option --project-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            device_base_url = (
                'https://%s/v1alpha2/projects/%s/devices' % (api_endpoint,
                                                             project_id)
            )
            device_id = str(uuid.uuid1())
            payload = {
                'id': device_id,
                'model_id': device_model_id
            }
            session = google.auth.transport.requests.AuthorizedSession(
                credentials
            )
            r = session.post(device_base_url, data=json.dumps(payload))
            if r.status_code != 200:
                logging.error('Failed to register device: %s', r.text)
                sys.exit(-1)
            logging.info('Device registered: %s', device_id)
            os.makedirs(os.path.dirname(device_config), exist_ok=True)
            with open(device_config, 'w') as f:
                json.dump(payload, f)

    with SampleAssistant(lang, device_model_id, device_id,
                         conversation_stream,
                         grpc_channel, grpc_deadline,
                         device_handler) as assistant:
        # If file arguments are supplied:
        # exit after the first turn of the conversation.
        if input_audio_file or output_audio_file:
            assistant.assist()
            return

        # If no file arguments supplied:
        # keep recording voice requests using the microphone
        # and playing back assistant response using the speaker.
        # When the once flag is set, don't wait for a trigger. Otherwise, wait.
        wait_for_user_trigger = not once
        while True:
            if wait_for_user_trigger:
                button_state=GPIO.input(22)
                if button_state==True:
                    continue
                else:
                    pass
            continue_conversation = assistant.assist()
            # wait for user trigger if there is no follow-up turn in
            # the conversation.
            wait_for_user_trigger = not continue_conversation


            # If we only want one conversation, break.
            if once and (not continue_conversation):
                break
Ejemplo n.º 5
0
 def test_raw(self):
     self.stream = BytesIO(b'audiodata')
     self.source = audio_helpers.WaveSource(self.stream, 16000, 2)
     self.assertEqual(b'audiodata', self.source.read(9))
Ejemplo n.º 6
0
def return_ga():

    api_endpoint = 'embeddedassistant.googleapis.com'
    project_id = "assistantdevice-1f1e8"
    device_model_id = "assistantdevice-1f1e8-ga1-v3slut"
    device_id = None
    credentials = os.path.join(click.get_app_dir('google-oauthlib-tool'),
                               'credentials.json')
    device_config = os.path.join(click.get_app_dir('googlesamples-assistant'),
                                 'device_config.json')
    lang = "en-US"
    display = False
    verbose = False
    input_audio_file = None
    output_audio_file = None
    audio_sample_rate = audio_helpers.DEFAULT_AUDIO_SAMPLE_RATE
    audio_sample_width = audio_helpers.DEFAULT_AUDIO_SAMPLE_WIDTH
    audio_iter_size = audio_helpers.DEFAULT_AUDIO_ITER_SIZE
    audio_block_size = audio_helpers.DEFAULT_AUDIO_DEVICE_BLOCK_SIZE
    audio_flush_size = audio_helpers.DEFAULT_AUDIO_DEVICE_FLUSH_SIZE
    grpc_deadline = 60 * 3 + 5
    once = False

    # Load OAuth 2.0 credentials.
    try:
        with open(credentials, 'r') as f:
            credentials = google.oauth2.credentials.Credentials(token=None,
                                                                **json.load(f))
            http_request = google.auth.transport.requests.Request()
            credentials.refresh(http_request)
    except Exception as e:
        logging.error('Error loading credentials: %s', e)
        logging.error('Run google-oauthlib-tool to initialize '
                      'new OAuth 2.0 credentials.')
        sys.exit(-1)

    # Create an authorized gRPC channel.
    grpc_channel = google.auth.transport.grpc.secure_authorized_channel(
        credentials, http_request, api_endpoint)
    logger.info('Connecting to %s', api_endpoint)

    # Configure audio source and sink.
    audio_device = None
    if input_audio_file:
        audio_source = audio_helpers.WaveSource(
            open(input_audio_file, 'rb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width)
    else:
        audio_source = audio_device = (audio_device
                                       or audio_helpers.SoundDeviceStream(
                                           sample_rate=audio_sample_rate,
                                           sample_width=audio_sample_width,
                                           block_size=audio_block_size,
                                           flush_size=audio_flush_size))
    if output_audio_file:
        audio_sink = audio_helpers.WaveSink(open(output_audio_file, 'wb'),
                                            sample_rate=audio_sample_rate,
                                            sample_width=audio_sample_width)
    else:
        audio_sink = audio_device = (audio_device
                                     or audio_helpers.SoundDeviceStream(
                                         sample_rate=audio_sample_rate,
                                         sample_width=audio_sample_width,
                                         block_size=audio_block_size,
                                         flush_size=audio_flush_size))
    # Create conversation stream with the given audio source and sink.
    conversation_stream = audio_helpers.ConversationStream(
        source=audio_source,
        sink=audio_sink,
        iter_size=audio_iter_size,
        sample_width=audio_sample_width,
    )

    if not device_id or not device_model_id:
        try:
            with open(device_config) as f:
                device = json.load(f)
                device_id = device['id']
                device_model_id = device['model_id']
                logger.info("Using device model %s and device id %s",
                            device_model_id, device_id)
        except Exception as e:
            logging.warning('Device config not found: %s' % e)
            logger.info('Registering device')
            if not device_model_id:
                logging.error('Option --device-model-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            if not project_id:
                logging.error('Option --project-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            device_base_url = ('https://%s/v1alpha2/projects/%s/devices' %
                               (api_endpoint, project_id))
            device_id = str(uuid.uuid1())
            payload = {
                'id': device_id,
                'model_id': device_model_id,
                'client_type': 'SDK_SERVICE'
            }
            session = google.auth.transport.requests.AuthorizedSession(
                credentials)
            r = session.post(device_base_url, data=json.dumps(payload))
            if r.status_code != 200:
                logging.error('Failed to register device: %s', r.text)
                sys.exit(-1)
            logger.info('Device registered: %s', device_id)
            pathlib.Path(os.path.dirname(device_config)).mkdir(exist_ok=True)
            with open(device_config, 'w') as f:
                json.dump(payload, f)

    device_handler = device_helpers.DeviceRequestHandler(device_id)

    @device_handler.command('action.devices.commands.OnOff')
    def onoff(on):
        if on:
            logger.info('Turning device on')
        else:
            logger.info('Turning device off')

    @device_handler.command('com.example.commands.BlinkLight')
    def blink(speed, number):
        logger.info('Blinking device %s times.' % number)
        delay = 1
        if speed == "SLOWLY":
            delay = 2
        elif speed == "QUICKLY":
            delay = 0.5
        for i in range(int(number)):
            logger.info('Device is blinking.')
            time.sleep(delay)

    logger.info("Start Google Assistant")
    assistant = SampleAssistant(lang, device_model_id, device_id,
                                conversation_stream, display, grpc_channel,
                                grpc_deadline, device_handler)

    return assistant
Ejemplo n.º 7
0
def main(api_endpoint, credentials, project_id, device_model_id, device_id,
         device_config, lang, verbose, input_audio_file, output_audio_file,
         audio_sample_rate, audio_sample_width, audio_iter_size,
         audio_block_size, audio_flush_size, grpc_deadline, once, *args,
         **kwargs):
    """Samples for the Google Assistant API.

    Examples:
      Run the sample with microphone input and speaker output:

        $ python -m googlesamples.assistant

      Run the sample with file input and speaker output:

        $ python -m googlesamples.assistant -i <input file>

      Run the sample with file input and output:

        $ python -m googlesamples.assistant -i <input file> -o <output file>
    """
    # Setup logging.
    logging.basicConfig(level=logging.DEBUG if verbose else logging.INFO)

    # Load OAuth 2.0 credentials.
    try:
        with open(credentials, 'r') as f:
            credentials = google.oauth2.credentials.Credentials(token=None,
                                                                **json.load(f))
            http_request = google.auth.transport.requests.Request()
            credentials.refresh(http_request)
    except Exception as e:
        logging.error('Error loading credentials: %s', e)
        logging.error('Run google-oauthlib-tool to initialize '
                      'new OAuth 2.0 credentials.')
        sys.exit(-1)

# D-Bus preparation
    cv_trigger = Condition(Lock())
    eventq = Queue()
    dbus_obj = DBusSignals()
    loop = GLib.MainLoop()
    system_bus = SystemBus()

    def dbus_handler(sender, object, iface, signal, params):
        # logging.info(sender)
        # logging.info(object)
        # logging.info(signal)
        # logging.info(params)

        logging.debug('Received D-Bus signal: {}'.format(signal))

        if signal == 'trigger':
            cv_trigger.acquire()
            cv_trigger.notify()
            cv_trigger.release()

    pub = system_bus.publish("io.respeaker.respeakerd", dbus_obj)
    sub = system_bus.subscribe(iface='respeakerd.signal',
                               signal_fired=dbus_handler)

    def exit_dbus():
        sub.unsubscribe()
        pub.unpublish()

    # Create an authorized gRPC channel.
    grpc_channel = google.auth.transport.grpc.secure_authorized_channel(
        credentials, http_request, api_endpoint)
    logging.info('Connecting to %s', api_endpoint)

    # Configure audio source and sink.
    audio_device = None
    if input_audio_file:
        audio_source = audio_helpers.WaveSource(
            open(input_audio_file, 'rb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width)
    else:
        audio_source = audio_device = (audio_device
                                       or audio_helpers.SoundDeviceStream(
                                           sample_rate=audio_sample_rate,
                                           sample_width=audio_sample_width,
                                           block_size=audio_block_size,
                                           flush_size=audio_flush_size))
    if output_audio_file:
        audio_sink = audio_helpers.WaveSink(open(output_audio_file, 'wb'),
                                            sample_rate=audio_sample_rate,
                                            sample_width=audio_sample_width)
    else:
        audio_sink = audio_device = (audio_device
                                     or audio_helpers.SoundDeviceStream(
                                         sample_rate=audio_sample_rate,
                                         sample_width=audio_sample_width,
                                         block_size=audio_block_size,
                                         flush_size=audio_flush_size))
    # Create conversation stream with the given audio source and sink.
    conversation_stream = audio_helpers.ConversationStream(
        source=audio_source,
        sink=audio_sink,
        iter_size=audio_iter_size,
        sample_width=audio_sample_width,
    )

    if not device_id or not device_model_id:
        try:
            with open(device_config) as f:
                device = json.load(f)
                device_id = device['id']
                device_model_id = device['model_id']
                logging.info("Using device model %s and device id %s",
                             device_model_id, device_id)
        except Exception as e:
            logging.warning('Device config not found: %s' % e)
            logging.info('Registering device')
            if not device_model_id:
                logging.error('Option --device-model-id required '
                              'when registering a device instance.')
                exit_dbus()
                sys.exit(-1)
            if not project_id:
                logging.error('Option --project-id required '
                              'when registering a device instance.')
                exit_dbus()
                sys.exit(-1)
            device_base_url = ('https://%s/v1alpha2/projects/%s/devices' %
                               (api_endpoint, project_id))
            device_id = str(uuid.uuid1())
            payload = {
                'id': device_id,
                'model_id': device_model_id,
                'client_type': 'SDK_SERVICE'
            }
            session = google.auth.transport.requests.AuthorizedSession(
                credentials)
            r = session.post(device_base_url, data=json.dumps(payload))
            if r.status_code != 200:
                logging.error('Failed to register device: %s', r.text)
                exit_dbus()
                sys.exit(-1)
            logging.info('Device registered: %s', device_id)
            pathlib.Path(os.path.dirname(device_config)).mkdir(exist_ok=True)
            with open(device_config, 'w') as f:
                json.dump(payload, f)

    device_handler = device_helpers.DeviceRequestHandler(device_id)

    @device_handler.command('action.devices.commands.OnOff')
    def onoff(on):
        if on:
            logging.info('Turning device on')
        else:
            logging.info('Turning device off')

    # If file arguments are supplied:
    # exit after the first turn of the conversation.
    if input_audio_file or output_audio_file:
        with SampleAssistant(lang, device_model_id, device_id,
                             conversation_stream, grpc_channel, grpc_deadline,
                             device_handler, eventq) as assistant:
            dbus_obj.ready()
            assistant.assist()
            exit_dbus()
            return

    # else it's a long term run

    def assistant_thread(conversation_stream_):

        with SampleAssistant(lang, device_model_id, device_id,
                             conversation_stream_, grpc_channel, grpc_deadline,
                             device_handler, eventq) as assistant:

            dbus_obj.ready()

            # If no file arguments supplied:
            # keep recording voice requests using the microphone
            # and playing back assistant response using the speaker.
            # When the once flag is set, don't wait for a trigger. Otherwise, wait.
            wait_for_user_trigger = not once
            # capture device should always be opened when not playing
            conversation_stream_.start_recording()

            while True:
                if wait_for_user_trigger:
                    logging.info("speak hotword to wake up")
                    cv_trigger.acquire()
                    cv_trigger.wait()
                    cv_trigger.release()
                    logging.info("wake up!")

                continue_conversation = False
                try:
                    continue_conversation = assistant.assist()
                except PortAudioError as e:
                    logging.warn('PortAudio Error: {}'.format(str(e)))
                    eventq.put('on_idle')
                # wait for user trigger if there is no follow-up turn in
                # the conversation.
                wait_for_user_trigger = not continue_conversation

                # If we only want one conversation, break.
                if once and (not continue_conversation):
                    break
        exit_dbus()
        logging.debug('Exit from the assistant thread...')
        loop.quit()

    def event_process_thread():
        while True:
            event = eventq.get()
            if event == 'on_listen':
                dbus_obj.on_listen()
            elif event == 'on_think':
                dbus_obj.on_think()
            elif event == 'on_speak':
                dbus_obj.on_speak()
            elif event == 'on_idle':
                dbus_obj.on_idle()
            time.sleep(0.5)

    def on_exit(sig):
        exit_dbus()
        loop.quit()
        logging.info("Quit...")

    setup_signals(
        signals=[sys_signal.SIGINT, sys_signal.SIGTERM, sys_signal.SIGHUP],
        handler=on_exit)

    # make conversation_stream writable inside thread
    thrd1 = Thread(target=assistant_thread, args=(conversation_stream, ))
    thrd2 = Thread(target=event_process_thread)
    thrd1.daemon = True
    thrd2.daemon = True
    thrd1.start()
    thrd2.start()

    logging.info("Glib mainloop start running...")
    loop.run()

    dbus_obj.on_idle()
Ejemplo n.º 8
0
def main(api_endpoint=ASSISTANT_API_ENDPOINT,
         credentials=os.path.join(click.get_app_dir('google-oauthlib-tool'),
                                  'credentials.json'),
         project_id=None,
         device_model_id=None,
         device_id=None,
         device_config=os.path.join(
             click.get_app_dir('googlesamples-assistant'),
             'device_config.json'),
         lang='en-US',
         display=False,
         verbose=False,
         input_audio_file=None,
         output_audio_file=None,
         audio_sample_rate=audio_helpers.DEFAULT_AUDIO_SAMPLE_RATE,
         audio_sample_width=audio_helpers.DEFAULT_AUDIO_SAMPLE_WIDTH,
         audio_iter_size=audio_helpers.DEFAULT_AUDIO_ITER_SIZE,
         audio_block_size=audio_helpers.DEFAULT_AUDIO_DEVICE_BLOCK_SIZE,
         audio_flush_size=audio_helpers.DEFAULT_AUDIO_DEVICE_FLUSH_SIZE,
         grpc_deadline=DEFAULT_GRPC_DEADLINE,
         once=False,
         *args,
         **kwargs):
    # Setup logging.
    #logging.basicConfig(level=logging.DEBUG if verbose else logging.INFO)
    logging.basicConfig(level=logging.INFO)

    # Load OAuth 2.0 credentials.
    try:
        with open(credentials, 'r') as f:
            credentials = google.oauth2.credentials.Credentials(token=None,
                                                                **json.load(f))
            http_request = google.auth.transport.requests.Request()
            credentials.refresh(http_request)
    except Exception as e:
        logging.error('Error loading credentials: %s', e)
        logging.error('Run google-oauthlib-tool to initialize '
                      'new OAuth 2.0 credentials.')
        sys.exit(-1)

    # Create an authorized gRPC channel.
    grpc_channel = google.auth.transport.grpc.secure_authorized_channel(
        credentials, http_request, api_endpoint)
    logging.info('Connecting to %s', api_endpoint)

    # Configure audio source and sink.
    audio_device = None
    if input_audio_file:
        audio_source = audio_helpers.WaveSource(
            open(input_audio_file, 'rb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width)
    else:
        audio_source = audio_device = (audio_device
                                       or audio_helpers.SoundDeviceStream(
                                           sample_rate=audio_sample_rate,
                                           sample_width=audio_sample_width,
                                           block_size=audio_block_size,
                                           flush_size=audio_flush_size))
    if output_audio_file:
        audio_sink = audio_helpers.WaveSink(open(output_audio_file, 'wb'),
                                            sample_rate=audio_sample_rate,
                                            sample_width=audio_sample_width)
    else:
        audio_sink = audio_device = (audio_device
                                     or audio_helpers.SoundDeviceStream(
                                         sample_rate=audio_sample_rate,
                                         sample_width=audio_sample_width,
                                         block_size=audio_block_size,
                                         flush_size=audio_flush_size))
    # Create conversation stream with the given audio source and sink.
    conversation_stream = audio_helpers.ConversationStream(
        source=audio_source,
        sink=audio_sink,
        iter_size=audio_iter_size,
        sample_width=audio_sample_width,
    )

    if not device_id or not device_model_id:
        try:
            with open(device_config) as f:
                device = json.load(f)
                device_id = device['id']
                device_model_id = device['model_id']
                logging.info("Using device model %s and device id %s",
                             device_model_id, device_id)
        except Exception as e:
            logging.warning('Device config not found: %s' % e)
            logging.info('Registering device')
            if not device_model_id:
                logging.error('Option --device-model-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            if not project_id:
                logging.error('Option --project-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            device_base_url = ('https://%s/v1alpha2/projects/%s/devices' %
                               (api_endpoint, project_id))
            device_id = str(uuid.uuid1())
            payload = {
                'id': device_id,
                'model_id': device_model_id,
                'client_type': 'SDK_SERVICE'
            }
            session = google.auth.transport.requests.AuthorizedSession(
                credentials)
            r = session.post(device_base_url, data=json.dumps(payload))
            if r.status_code != 200:
                logging.error('Failed to register device: %s', r.text)
                sys.exit(-1)
            logging.info('Device registered: %s', device_id)
            pathlib.Path(os.path.dirname(device_config)).mkdir(exist_ok=True)
            with open(device_config, 'w') as f:
                json.dump(payload, f)

    device_handler = device_helpers.DeviceRequestHandler(device_id)

    @device_handler.command('action.devices.commands.OnOff')
    def onoff(on):
        if on:
            logging.info('Turning device on')
        else:
            logging.info('Turning device off')

    @device_handler.command('com.example.commands.BlinkLight')
    def blink(speed, number):
        logging.info('Blinking device %s times.' % number)
        delay = 1
        if speed == "SLOWLY":
            delay = 2
        elif speed == "QUICKLY":
            delay = 0.5
        for i in range(int(number)):
            logging.info('Device is blinking.')
            time.sleep(delay)

    audio_stream = None
    handle = None
    pa = None
    library_path = '../../lib/windows/amd64/libpv_porcupine.dll'
    model_file_path = '../../lib/common/porcupine_params.pv'
    keyword_file_paths = ['bumblebee_windows.ppn']
    num_keywords = len(keyword_file_paths)
    sensitivities = [0.7]  #0.2
    handle = porcupine.Porcupine(library_path,
                                 model_file_path,
                                 keyword_file_paths=keyword_file_paths,
                                 sensitivities=sensitivities)

    pa = pyaudio.PyAudio()
    audio_stream = pa.open(rate=handle.sample_rate,
                           channels=1,
                           format=pyaudio.paInt16,
                           input=True,
                           frames_per_buffer=handle.frame_length)

    conversation_stream.volume_percentage = 100
    print('Listening for keyword bumblebee...')
    with SampleAssistant(lang, device_model_id, device_id, conversation_stream,
                         display, grpc_channel, grpc_deadline,
                         device_handler) as assistant:
        while True:

            pcm = audio_stream.read(handle.frame_length)
            pcm = struct.unpack_from("h" * handle.frame_length, pcm)
            result = handle.process(pcm)
            if num_keywords > 0 and result:
                play_audio_file('ding.wav')
                print('[%s] detected keyword!!!!' % str(datetime.now()))
                assistant.assist()
                play_audio_file('dong.wav')
                print('Listening for keyword bumblebee...')

        print('stopping ...')
        if handle is not None:
            handle.delete()

        if audio_stream is not None:
            audio_stream.close()

        if pa is not None:
            pa.terminate()
def main(api_endpoint, credentials, verbose,
         input_audio_file, output_audio_file,
         audio_sample_rate, audio_sample_width,
         audio_iter_size, audio_block_size, audio_flush_size,
         grpc_deadline, once, *args, **kwargs):
    """Samples for the Google Assistant API.
    Examples:
      Run the sample with microphone input and speaker output:
        $ python -m googlesamples.assistant
      Run the sample with file input and speaker output:
        $ python -m googlesamples.assistant -i <input file>
      Run the sample with file input and output:
        $ python -m googlesamples.assistant -i <input file> -o <output file>
    """
    # Setup logging.
    logging.basicConfig(level=logging.DEBUG if verbose else logging.INFO)

    # Load OAuth 2.0 credentials.
    try:
        with open(credentials, 'r') as f:
            credentials = google.oauth2.credentials.Credentials(token=None,
                                                                **json.load(f))
            http_request = google.auth.transport.requests.Request()
            credentials.refresh(http_request)
    except Exception as e:
        logging.error('Error loading credentials: %s', e)
        logging.error('Run google-oauthlib-tool to initialize '
                      'new OAuth 2.0 credentials.')
        return

    # Create an authorized gRPC channel.
    grpc_channel = google.auth.transport.grpc.secure_authorized_channel(
        credentials, http_request, api_endpoint)
    logging.info('Connecting to %s', api_endpoint)

    # Configure audio source and sink.
    audio_device = None
    if input_audio_file:
        audio_source = audio_helpers.WaveSource(
            open(input_audio_file, 'rb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width
        )
    else:
        audio_source = audio_device = (
            audio_device or audio_helpers.SoundDeviceStream(
                sample_rate=audio_sample_rate,
                sample_width=audio_sample_width,
                block_size=audio_block_size,
                flush_size=audio_flush_size
            )
        )
    if output_audio_file:
        audio_sink = audio_helpers.WaveSink(
            open(output_audio_file, 'wb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width
        )
    else:
        audio_sink = audio_device = (
            audio_device or audio_helpers.SoundDeviceStream(
                sample_rate=audio_sample_rate,
                sample_width=audio_sample_width,
                block_size=audio_block_size,
                flush_size=audio_flush_size
            )
        )
    # Create conversation stream with the given audio source and sink.
    conversation_stream = audio_helpers.ConversationStream(
        source=audio_source,
        sink=audio_sink,
        iter_size=audio_iter_size,
        sample_width=audio_sample_width,
    )

    with SampleAssistant(conversation_stream,
                         grpc_channel, grpc_deadline) as assistant:
        # If file arguments are supplied:
        # exit after the first turn of the conversation.
        if input_audio_file or output_audio_file:
            assistant.converse()
            return

        # If no file arguments supplied:
        # keep recording voice requests using the microphone
        # and playing back assistant response using the speaker.
        # When the once flag is set, don't wait for a trigger. Otherwise, wait.
		
        continue_conversation = assistant.converse()
Ejemplo n.º 10
0
def main(api_endpoint, credentials, project_id, device_model_id, device_id,
         device_config, lang, display, verbose, input_audio_file,
         output_audio_file, audio_sample_rate, audio_sample_width,
         audio_iter_size, audio_block_size, audio_flush_size, grpc_deadline,
         once, *args, **kwargs):
    """Samples for the Google Assistant API.

    Examples:
      Run the sample with microphone input and speaker output:

        $ python -m googlesamples.assistant

      Run the sample with file input and speaker output:

        $ python -m googlesamples.assistant -i <input file>

      Run the sample with file input and output:

        $ python -m googlesamples.assistant -i <input file> -o <output file>
    """
    # Setup logging.
    logging.basicConfig(level=logging.DEBUG if verbose else logging.INFO)

    # Load OAuth 2.0 credentials.
    try:
        with open(credentials, 'r') as f:
            credentials = google.oauth2.credentials.Credentials(token=None,
                                                                **json.load(f))
            http_request = google.auth.transport.requests.Request()
            credentials.refresh(http_request)
    except Exception as e:
        logging.error('Error loading credentials: %s', e)
        logging.error('Run google-oauthlib-tool to initialize '
                      'new OAuth 2.0 credentials.')
        sys.exit(-1)

    # Create an authorized gRPC channel.
    grpc_channel = google.auth.transport.grpc.secure_authorized_channel(
        credentials, http_request, api_endpoint)
    logging.info('Connecting to %s', api_endpoint)

    # Configure audio source and sink.
    audio_device = None
    if input_audio_file:
        audio_source = audio_helpers.WaveSource(
            open(input_audio_file, 'rb'),
            sample_rate=audio_sample_rate,
            sample_width=audio_sample_width)
    else:
        audio_source = audio_device = (audio_device
                                       or audio_helpers.SoundDeviceStream(
                                           sample_rate=audio_sample_rate,
                                           sample_width=audio_sample_width,
                                           block_size=audio_block_size,
                                           flush_size=audio_flush_size))
    if output_audio_file:
        audio_sink = audio_helpers.WaveSink(open(output_audio_file, 'wb'),
                                            sample_rate=audio_sample_rate,
                                            sample_width=audio_sample_width)
    else:
        audio_sink = audio_device = (audio_device
                                     or audio_helpers.SoundDeviceStream(
                                         sample_rate=audio_sample_rate,
                                         sample_width=audio_sample_width,
                                         block_size=audio_block_size,
                                         flush_size=audio_flush_size))
    # Create conversation stream with the given audio source and sink.
    conversation_stream = audio_helpers.ConversationStream(
        source=audio_source,
        sink=audio_sink,
        iter_size=audio_iter_size,
        sample_width=audio_sample_width,
    )

    if not device_id or not device_model_id:
        try:
            with open(device_config) as f:
                device = json.load(f)
                device_id = device['id']
                device_model_id = device['model_id']
                logging.info("Using device model %s and device id %s",
                             device_model_id, device_id)
        except Exception as e:
            logging.warning('Device config not found: %s' % e)
            logging.info('Registering device')
            if not device_model_id:
                logging.error('Option --device-model-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            if not project_id:
                logging.error('Option --project-id required '
                              'when registering a device instance.')
                sys.exit(-1)
            device_base_url = ('https://%s/v1alpha2/projects/%s/devices' %
                               (api_endpoint, project_id))
            device_id = str(uuid.uuid1())
            payload = {
                'id': device_id,
                'model_id': device_model_id,
                'client_type': 'SDK_SERVICE'
            }
            session = google.auth.transport.requests.AuthorizedSession(
                credentials)
            r = session.post(device_base_url, data=json.dumps(payload))
            if r.status_code != 200:
                logging.error('Failed to register device: %s', r.text)
                sys.exit(-1)
            logging.info('Device registered: %s', device_id)
            pathlib.Path(os.path.dirname(device_config)).mkdir(exist_ok=True)
            with open(device_config, 'w') as f:
                json.dump(payload, f)

    device_handler = device_helpers.DeviceRequestHandler(device_id)

    @device_handler.command('action.devices.commands.OnOff')
    def onoff(on):
        if on:
            logging.info('Turning device on')
            logging.warning('U CAN DO SOMETHING IN THERE FOR ON!!!')
        else:
            logging.info('Turning device off')
            logging.warning('U CAN DO SOMETHING IN THERE FOR OFF!!!')

    @device_handler.command('action.devices.commands.BrightnessAbsolute')
    def brightnessCheck(brightness):
        if brightness > 50:
            logging.info('brightness > 50')
        else:
            logging.info('brightness <= 50')

    @device_handler.command('action.devices.commands.ColorAbsolute')
    def color(color):
        if color.get('name') == "blue":
            logging.info('color is blue')
        else:
            logging.info('color is not blue')

    @device_handler.command('com.example.commands.BlinkLight')
    def blink(speed, number):
        logging.info('Blinking device %s times.' % number)
        delay = 1
        if speed == "SLOWLY":
            delay = 2
        elif speed == "QUICKLY":
            delay = 0.5
        for i in range(int(number)):
            logging.info('Device is blinking.')
            time.sleep(delay)

    with SampleAssistant(lang, device_model_id, device_id, conversation_stream,
                         display, grpc_channel, grpc_deadline,
                         device_handler) as assistant:
        # If file arguments are supplied:
        # exit after the first turn of the conversation.
        if input_audio_file or output_audio_file:
            assistant.assist()
            return

        # If no file arguments supplied:
        # keep recording voice requests using the microphone
        # and playing back assistant response using the speaker.
        # When the once flag is set, don't wait for a trigger. Otherwise, wait.
        wait_for_user_trigger = not once
        while True:
            if wait_for_user_trigger:
                click.pause(info='Press Enter to send a new request...')
            continue_conversation = assistant.assist()
            # wait for user trigger if there is no follow-up turn in
            # the conversation.
            wait_for_user_trigger = not continue_conversation

            # If we only want one conversation, break.
            if once and (not continue_conversation):
                break