Windows下 使用Python 3 调用讯飞 TTS 引擎实现文本转语音

Windows下使用Python 调用科大讯飞TTS引擎离线转换文本到语音

使用前将讯飞SDK bin下的文件及文件夹复制到当前文件夹下。

py -3 ttsdemo.py

 

#! /usr/bin/env python
# -*- coding: utf-8 -*-
# python version >3.5

__author__ = 'onestab'

import sys, os, time
from datetime import datetime
from ctypes import *

dll = windll.LoadLibrary("msc.dll")
login_params = b"appid = 12345678, work_dir = ."
session_begin_params = b"engine_type = local, voice_name = xiaoyan, text_encoding = UTF8, tts_res_path = fo|res\\tts\\xiaoyan.jet;fo|res\\tts\\common.jet, sample_rate = 16000, speed = 80, volume = 70, pitch = 50, rdn = 0"

wav_header = b'RIFF6n\x06\x00WAVEfmt \x10\x00\x00\x00\x01\x00\x01\x00\x80>\x00\x00\x00}\x00\x00\x02\x00\x10\x00data\x00\x00\x00\x00'
MSP_SUCCESS = 0
MSP_TTS_FLAG_STILL_HAVE_DATA = 1
MSP_TTS_FLAG_DATA_END = 2
MSP_TTS_FLAG_CMD_CANCELED = 4

filename = "tts_sample.wav"

class Msp:
    def __init__(self):
        pass

    def login(self):
        ret = dll.MSPLogin(None, None, login_params)
        print(('MSPLogin =>'), ret)

    def tts(self, text, filename, session_begin_params):
        ret, audio_len, synth_status, getret, wav_datasize = c_int(), c_int(), c_int(), c_int(), c_int() 
        sessionID = dll.QTTSSessionBegin(session_begin_params, byref(ret))
        ret = dll.QTTSTextPut(sessionID, text, len(text), None)

        wavFile = open(filename, 'wb')
        wavFile.write(wav_header)
        while True:
            data = dll.QTTSAudioGet(sessionID, byref(audio_len), byref(synth_status), byref(getret))
            if audio_len.value>0:
                print('datasize:            %d\r' % wav_datasize.value, end='\r')
            if getret.value != MSP_SUCCESS:
                break
            if data:
                wavFile.write(string_at(data, audio_len))
                wav_datasize.value += audio_len.value
            if synth_status.value == MSP_TTS_FLAG_DATA_END:
                break
            #time.sleep(0.01)
        # fix wav header
        WAVE_HEADER_SIZE = 44
        if wav_datasize.value>0:
            wav_size8 = c_int()
            wav_size8.value = WAVE_HEADER_SIZE + wav_datasize.value - 8
            wavFile.seek(4)
            wavFile.write(wav_size8)
            wavFile.seek(40)
            wavFile.write(wav_datasize)
        wavFile.close()
        dll.QTTSSessionEnd(sessionID, "Normal")

if __name__ == "__main__":
    text = "欢迎来到方工的CSDN博客。CSDN创立于1999年，是中国专业的IT技术社区和开发者服务平台。拥有5000万注册用户以及60万注册企业及合作伙伴。"
    print(text)
    msp = Msp()
    msp.login()
    msp.tts(text.encode('U8'), filename, session_begin_params)

 ===================================

关于Python3 以上遇到access violation reading错误的问题，是64位和32位指针兼容的bug。暂时修复如下

    def tts(self, text, filename, session_begin_params):
        ret, audio_len, synth_status, getret, wav_datasize = c_int(), c_int(), c_int(), c_int(), c_int() 
        # fix return type
        dll.QTTSSessionBegin.restype = ctypes.c_char_p
        dll.QTTSAudioGet.restype = ctypes.c_uint64

 代码其他部分不变，除了python3.7 64位，装入的是msc_64.dll）

人生苦短，我用Python.