做物联网涉及到实际使用环境语音识别准确率的测试，大量的语料如果是人来朗读的话太费功夫，编一个自动测试程序即可无人通宵测试，大大提高了效率；朗读上可以通过调用powershell来朗文档中的语料，通然后过串口抓取设备设备对语料响应的数据进行分析把识别结果写入到文档当中;
需要安装serial（串口支持模块）和openxl（Exel文档支持模块）;编译调试完成打包为.exe文件，可在没有配置python环境的电脑上运行；
使用到python多线程（通过Queue线程间通信），还有文档读写 *** 作；很有参考价值，
import serialimport threadingimport timeimport osimport reimport fileinputfrom glob import globimport subprocess as spfrom queue import Queue# import xlsxwriter# import xlrdfrom openpyxl import load_workbookfrom openpyxl import Workbook
print('Hellow World:',time.time()) #时间戳：秒
TheTimeIs=time.ctime()event = threading.Event()q=Queue(1)q1=Queue(1)# q2=Queue(1)#//////////////////////////////////////////////////////////////////////////////调用powershell的类class PowerShell: # from scapy def __init__(self, Coding, ): cmd = [self._where('PowerShell.exe'), "-Nologo", "-NonInteractive", # Do not print headers "-Command", "-"] # Listen commands from stdin startupinfo = sp.STARTUPINFO() startupinfo.DWFlags |= sp.STARTF_USESHOWWINDOW self.popen = sp.Popen(cmd, stdout=sp.PIPE, stdin=sp.PIPE, stderr=sp.STDOUT, startupinfo=startupinfo) self.Coding = Coding
def __enter__(self): return self
def __exit__(self, a, b, c): self.popen.kill()
def run(self, cmd, timeout=6): #timeout=0 延时 b_cmd = cmd.encode(enCoding=self.Coding) try: b_outs, errs = self.popen.communicate(b_cmd, timeout=timeout) except sp.TimeoutExpired: self.popen.kill() b_outs, errs = self.popen.communicate() outs = b_outs.decode(enCoding=self.Coding) return outs, errs
@staticmethod def _where(filename, dirs=None, env="PATH"): """Find file in current dir, in deep_lookup cache or in system path""" if dirs is None: dirs = [] if not isinstance(dirs, List): dirs = [dirs] if glob(filename): return filename paths = [os.curdir] + os.environ[env].split(os.path.pathsep) + dirs try: return next(os.path.normpath(match) for path in paths for match in glob(os.path.join(path, filename)) if match) except (stopiteration, RuntimeError): raise IOError("file not found: %s" % filename)#//////////////////////////////////////////////////////////////////////////
#/////////////////////////////////////////////////////////////////////////文本处理pscmd='add-type -Assemblyname System.speech;\ $speak = New-Object System.Speech.Synthesis.SpeechSynthesizer;\ $speak.Rate = 0;\ $speak.Speak("天猫精灵")' #powershell文本朗读指令
cmd="天猫精灵"StandardCmd_n=''# Number_n=0SuccessNum_n=0Faildispaly_n=[]uart_error=0 #控制，朗读10次天猫精灵都没有回应，退出程序cmd_error=0
# input()data = load_workbook('test.xlsx')#打开表格mydata=data["Sheet1"] #sheet1
log='' #提取日志
def SpeakTm(): global uart_error #读天猫 with PowerShell('GBK') as ps: ps.run(pscmd) print("天猫精灵") uart_error+=1 if uart_error>10: #朗读10次没有回应，保存信息，退出程序 data.save('TestResult.xlsx') os._exit(0) with open('Show.txt', mode='a',enCoding='utf-8') as LogUart: LogUart.writelines(["天猫精灵\n"]) LogUart.close()
def changepscmd(fcmd): #读设定朗读语句 cmd=str(fcmd) pscmd0=pscmd p = re.compile("天猫精灵") pscmd1 = p.sub(cmd,pscmd0)
with PowerShell('GBK') as ps: #time.sleep(1) outs, errs = ps.run(pscmd1) #在run函数中输入powershell指令 print('error:', os.linesep, errs) print('output:', os.linesep, outs) # with open('LogSpeech.txt', mode='a',enCoding='utf-8') as LogSpeech: # LogSpeech.writelines([time.ctime(), ' say :',cmd,'\n']) # LogSpeech.close() with open('Show.txt', mode='a',enCoding='utf-8') as LogUart: LogUart.writelines([time.ctime(), ' say :',cmd,'\n']) LogUart.close() print(cmd) #打印朗读的语料
def checkcmd(StandardCmd_n): #比较提取的指令和标准指令 global SuccessNum_n global Faildispaly_n # q2.get() p = re.compile(", 0x") pscmd0 = p.sub('',log) pscmd1=pscmd0.upper() scmd0=str(StandardCmd_n) scmd=scmd0.upper() print('yu--->',cmd,'pscmd1--->',pscmd1,'standard--->',StandardCmd_n) with open('Show.txt', mode='a',enCoding='utf-8') as LogUart: LogUart.writelines(['yu--->',cmd,'pscmd1--->',pscmd1,'standard--->',StandardCmd_n]) LogUart.close() if pscmd1.find(scmd)!=-1: return True else: return False
def Rouse_Tm(log): if log.find('tian mao jing ling')!=-1: return True else: return False
#//////////////////////////////////////////////////////////////////////////////////朗读线程函数入口def Speech(): global StandardCmd_n global Corpus_n global order order=1 for i in mydata['A']: if order > 1: #使得order读取的数据和i的下标同步，好自然结束for循环 Corpus_n=str(i.value) #取语料 StandardCmd_n=str(mydata['B%d'%order].value) #取语料对应的标准指令
Number_n=mydata['C%d'%order].value #取语料对应的朗读次数 if Number_n==None: Number_n=0 print(Number_n) SuccessNum_n=0 #初始成功次数置0 Faildispaly_n="" #初始失败次数置空 FailSum=0 error_check=0 while Number_n>0: #朗读次数大于0朗读，并递减 SpeakTm() time.sleep(1) #使线程等待收到相应天猫响应指令后继续执行 print(time.ctime()) event.wait(7) #线程等待触发 print(time.ctime())
if q.empty(): print('cant Listen') continue #天猫精灵没有唤醒，重新进行唤醒 else: q.get() print('good') event.clear() #set会重置所有wait为不再等待，clear使所有wait重新生效 changepscmd(Corpus_n) #朗读 Number_n-=1 # time.sleep(5) #必须有个时间等待接收指令 print(time.ctime()) event.wait(10) #线程等待触发 print(time.ctime()) event.clear()
if q1.empty(): print('cant kNow') #天猫精灵唤醒成功，朗读语料，没有回应相关的指令 error_check+=1 with open('Show.txt', mode='a',enCoding='utf-8') as LogUart: LogUart.writelines([time.ctime(), 'cant kNow\n']) LogUart.close() if cmd_error==2: if FailSum<10: FailSum+=1 Faildispaly_n+="识别T 发送F- \n" mydata['E%d'%order].value=Faildispaly_n #直接写入没有识别进行下一次朗读 else: if FailSum<10: FailSum+=1 Faildispaly_n+="识别错误- \n" mydata['E%d'%order].value=Faildispaly_n #直接写入没有识别进行下一次朗读 else: q1.get() print('good') with open('Show.txt', mode='a',enCoding='utf-8') as LogUart: LogUart.writelines([time.ctime(), 'good\n']) LogUart.close() if checkcmd(StandardCmd_n): #判断指令是否正确，正确添加成功数，不正确打印错误案例 error_check=0 SuccessNum_n+=1 else: error_check+=1 if FailSum<10: FailSum+=1 Faildispaly_n+=log+'\n' mydata['E%d'%order].value=Faildispaly_n if error_check>=10: break mydata['D%d'%order].value=SuccessNum_n #直接将所获得得数据输入对应单元格 try: data.save('TestResult.xlsx')#保存到一个新的excel except(): os._exit(0) order+=1
time.sleep(5) os._exit(0)
#/////////////////////////////////////////////////////////////////////////////////////串口调试ser = serial.Serial()
print("请输入正确的串口号按回车健运行：")MycomX='COM'+str(input())
ser.port =MycomX # 串口print(ser.name)
ser.baudrate = 1500000 #波特率ser.bytesize = 8ser.parity = serial.PARITY_NONEser.stopbits = 1ser.open() # 开启串口
#///////////////////////////////////////////////////////////////////////////////////主线程串口接收def UartData():
global uart_error global cmd_error global log while True: #ch = ser.read() # 收一个bytes ch=ser.readline()#读取以“\n”结尾的行 log_can=ch.decode(enCoding='utf-8') itlog=log_can.find('send')
error_1=log_can.find('fresult') #接收到指令的特征 with open('LogUart.txt', mode='a',enCoding='utf-8') as LogUart: LogUart.writelines([time.ctime(), ' ',log_can]) LogUart.close() # if Rouse_Tm(log_can): print("Tm已唤醒：") uart_error=0 with open('Show.txt', mode='a',enCoding='utf-8') as LogUart: LogUart.writelines([time.ctime(), 'Tm已唤醒：\n']) LogUart.close() if q.empty(): q.put(1) #使用queue队列来完成线程之间的通信，queue是线程安全的 event.set() # event.clear()
if itlog==0: cmdl=len(log_can) if cmdl==104: log=log_can cmd_error=1 print('cmdlen is:\n',len(log),'\n') print('log is-----> %s' % log) with open('Show.txt', mode='a',enCoding='utf-8') as LogUart: LogUart.writelines([time.ctime(), 'log is----->',log,'\n']) LogUart.close()
if q.empty(): q1.put(1) #使用queue队列来完成线程之间的通信，queue是线程安全的 event.set()
if cmd_error==0 and error_1!=-1: print('error_1log-->',log_can) cmd_error=2
#/////////////////////////////////////////////////////////////////////////////////////
#多线程SpeechThread=threading.Thread(target=Speech)#SpeechThread.daemon=True #设置线程为守护线程UartDataThread=threading.Thread(target=UartData)#UartDataThread.daemon=True
def main(): UartDataThread.start() SpeechThread.start() time.sleep(60*60*24) #主函数结束，会导致线程结束，阻塞main,让线程继续运行，朗读线程运行完毕通过 os._exit（）结束程序
if __name__ == '__main__': main()
总结
以上是内存溢出为你收集整理的用python写一个自动朗读的自动化测试程序全部内容，希望文章能够帮你解决用python写一个自动朗读的自动化测试程序所遇到的程序开发问题。
如果觉得内存溢出网站内容还不错，欢迎将内存溢出网站推荐给程序员好友。
欢迎分享，转载请注明来源：内存溢出
微信扫一扫
支付宝扫一扫
评论列表（0条）