def main(): args = StreamingRecognitionParser().parse_args() with audio_open_read(args.audio_file, args.encoding, args.rate, args.num_channels, args.chunk_size, args.pyaudio_max_seconds) as reader: stub = stt_pb2_grpc.SpeechToTextStub(make_channel(args)) metadata = authorization_metadata(args.api_key, args.secret_key, "tinkoff.cloud.stt") responses = stub.StreamingRecognize(stt_generate_requests(args, reader), metadata=metadata) print_streaming_recognition_responses(responses)
def main(): args = RecognitionParser().parse_args() if args.encoding == "RAW_OPUS": raise ValueError("RAW_OPUS encoding is not supported by this script") stub = stt_pb2_grpc.SpeechToTextStub(make_channel(args.host, args.port)) metadata = authorization_metadata(args.api_key, args.secret_key, "tinkoff.cloud.stt") response = stub.Recognize(build_recognition_request(args), metadata=metadata) print_recognition_response(response)
def main(): args = BaseRecognitionParser().parse_args() if args.encoding == stt_pb2.RAW_OPUS: raise ValueError("RAW_OPUS encoding is not supported by this script") with audio_open_read(args.audio_file, args.encoding, args.rate, args.num_channels, args.chunk_size, args.pyaudio_max_seconds) as reader: stub = stt_pb2_grpc.SpeechToTextStub(make_channel(args)) metadata = authorization_metadata(args.api_key, args.secret_key, "tinkoff.cloud.stt") response = stub.Recognize(build_recognition_request(args, reader), metadata=metadata) print_recognition_response(response)
def main(): args = BaseSynthesisParser().parse_args() if args.encoding == tts_pb2.LINEAR16 and args.rate != 48000: raise ValueError("LINEAR16 supports only 48kHz for now, use RAW_OPUS") with audio_open_write(args.output_file, args.encoding, args.rate) as audio_writer: stub = tts_pb2_grpc.TextToSpeechStub(make_channel(args)) request = build_synthesis_request(args, args.input_text) metadata = authorization_metadata(args.api_key, args.secret_key, "tinkoff.cloud.tts") responses = stub.StreamingSynthesize(request, metadata=metadata) for stream_response in responses: audio_writer.write(stream_response.audio_chunk)
def main(): args = StreamingRecognitionParser().parse_args() stub = stt_pb2_grpc.SpeechToTextStub(make_channel(args.host, args.port)) metadata = authorization_metadata(args.api_key, args.secret_key, "tinkoff.cloud.stt") responses = stub.StreamingRecognize(generate_requests(args), metadata=metadata) for response in responses: for result in response.results: print("Channel", result.recognition_result.channel) print("Phrase start", result.recognition_result.start_time.ToTimedelta()) print("Phrase end", result.recognition_result.end_time.ToTimedelta()) print("Is final", result.is_final) for alternative in result.recognition_result.alternatives: print("Transcription", alternative.transcript) print("Confidence", alternative.confidence) print("------------------")
def main(): agi = AGI() #agi = None if agi != None: agi.verbose("EAGI script started...") ani = agi.env['agi_callerid'] uid = agi.env['agi_uniqueid'] agi.verbose("Call answered from: %s with id %s" % (ani, uid)) else: ani = "" uid = str(uuid.uuid4()) try: with dbcon.cursor() as cursor: sql = "INSERT INTO calls SET uniqueid=%s,callerid=%s,calldate=NOW()" cursor.execute(sql, (uid, ani)) call_id = cursor.lastrowid finally: dbcon.commit() data = { "type": "call", "unqueid": uid, "callerid": ani[-4:], "calldate": time.strftime('%Y-%m-%d %H:%M:%S'), "call_id": call_id } client.publish(cent_channel, data) if agi == None: ic(data) args = StreamingRecognitionParser().parse_args() stub = stt_pb2_grpc.SpeechToTextStub(make_channel(args)) metadata = authorization_metadata(cfg.api_key, cfg.secret_key, "tinkoff.cloud.stt") responses = stub.StreamingRecognize(generate_requests(args, agi), metadata=metadata) save_streaming_recognition_responses(responses, agi, ani, uid, call_id)
def main(): args = BaseRecognitionParser().parse_args() total = '' if args.encoding == stt_pb2.RAW_OPUS: raise ValueError("RAW_OPUS encoding is not supported by this script") with audio_open_read(args.audio_file, args.encoding, args.rate, args.num_channels, args.chunk_size, args.pyaudio_max_seconds) as reader: stub = stt_pb2_grpc.SpeechToTextStub(make_channel(args)) metadata = authorization_metadata(args.api_key, args.secret_key, "tinkoff.cloud.stt") response = stub.Recognize(build_recognition_request(args, reader), metadata=metadata) if not isinstance(response, dict): # https://developers.google.com/protocol-buffers/docs/proto3#json response = MessageToDict(response, including_default_value_fields=True, preserving_proto_field_name=True) for result in response["results"]: for alternative in result["alternatives"]: total = total + alternative["transcript"] print(total)