最新消息:20190717 VPS服务器:Vultr新加坡,WordPress主题:大前端D8,统一介绍入口:关于

【已解决】用和适在线的语言合成接口把文字转语音

接口 crifan 77浏览 0评论
之前已经做好了产品demo:
【已解决】把本地前端页面部署到在线Flask环境中
然后需要去使用在线语音接口,把文字转换为语音
然后集成到产品demo中。
所以不仅要能用语音接口,且要整合进来。
【调研】国内可用的效果好的英语的在线Web语音合成API接口
先去用目前发现的,相对来说最好用的:
百度的语音合成api
刚发现:
在线文字转语音|免费生成语音-百度广播开放平台
还可以生成临时的语音文件,以url的形式输出
但是有个缺点:
好像必须是:
标题也要有,内容也要有,才能生成
然后先去注册百度开发者账号:
【记录】注册百度开发者账号并创建应用
接着:
百度语音-文档中心
详细看官网文档:
百度语音合成-开发文档
“浏览器跨域
目前合成接口支持浏览器跨域。
跨域demo示例: https://github.com/Baidu-AIP/SPEECH-TTS-CORS
由于获取token的接口不支持浏览器跨域。因此需要您从服务端获取或者每个30天手动输入更新。”
Baidu-AIP/SPEECH-TTS-CORS: 百度语音 语音合成 跨域demo以及支持库
SPEECH-TTS-CORS/baidu_tts_cors.js at master · Baidu-AIP/SPEECH-TTS-CORS
调用流程示例
去获取token
然后去用浏览器打开:
https://openapi.baidu.com/oauth/2.0/token?grant_type=client_credentials&client_id=SNjsggdYDNWtnlbKhxsPLcaz&client_secret=47d7c02dxxxxxxxxxxxxxxe7ba
获得了token:
{
    "access_token": "24.569b3b5b470938a522ce60d2e2ea2506.2592000.1528015602.282335-11192483",
    "session_key": "9mzdDoR4p/oexxx0Yp9VoSgFCFOSGEIA==",
    "scope": "public audio_voice_assistant_get audio_tts_post wise_adapt lebo_resource_base lightservice_public hetu_basic lightcms_map_poi kaidian_kaidian ApsMisTest_Test权限 vis-classify_flower lpq_开放 cop_helloScope ApsMis_fangdi_permission smartapp_snsapi_base",
    "refresh_token": "25.5axxxx5-xxx3",
    "session_secret": "12xxxa",
    "expires_in": 2592000
}
再去调用url:
http://tsn.baidu.com/text2audio?lan=zh&ctp=1&cuid=xxx_robot&tok=24.56xxx3&vol=9&per=0&spd=5&pit=5&tex=as+a+book-collector%2c+i+have+the+story+you+just+want+to+listen!
获取合成后的mp3:
很明显,此处是直接返回mp3的内容的
而不是希望的临时的mp3的临时的url
而之前的:
http://developer.baidu.com/vcast
是可以返回临时mp3的url的
比如:
http://boscdn.bpc.baidu.com/v1/developer/df29f25f-6003-4e9f-9f80-15be4babf831.mp3
搜了下:
boscdn bpc baidu.com
boscdn.bpc.baidu.com domain information – Reason Core Security Labs
ecomfe/edpx-bos: edp的bos扩展
感觉是百度的一个cdn的服务器
然后还有js接口上传内容上去,生成临时url的
但是貌似是百度内部自己用的?
接着需要:
1.最好把百度的token,弄成那个永久的,或者至少是1年的,而不是现在的1个月的
2.最好把生成的mp3的文件,弄成一个url可以返回给用户的
感觉需要是用自己的flask的rest的api中,封装百度的接口,给外界一个统一的接口,返回mp3的url,然后是有临时时限的  -》 那内部可以考虑把mp3保存到 /tmp 或者是redis然后设置一个expire时限?
先去弄永久的token的事情:
【无法解决】获取百度的永久的或长期比如1年的有效的access token
而关于文档,从:
SPEECH-TTS-CORS/demo.html at master · Baidu-AIP/SPEECH-TTS-CORS
发现了:
百度语音合成 REST API
也找到了Python文档:
Python SDK文档
不过,对于,想要去模拟,当access_token失效时,百度接口会返回什么
突然想到,可以用刚才已经被refresh_token刷新后,而失效的之前的access_token,去调用看看,返回什么
结果之前的token,竟然还能用:
那算了,把token值随便改一下,去尝试模拟一个无效的token
结果返回:
{
    "err_detail": "Access token invalid or no longer valid",
    "err_msg": "authentication failed.",
    "err_no": 502,
    "err_subcode": 50004,
    "tts_logid": 1007366076
}
REST API文档
错误码解释
错误码
含义
500
不支持输入
501
输入参数不正确
502
token验证失败
503
合成后端错误
百度 err_subcode 50004
百度 authentication failed 50004
产品介绍_百度云推送_免费专业最精准的移动推送服务平台
50004
Passport Not Login
未登录百度账号passport
400
那目前就可以暂定为如下思路了:
用Flask去封装百度的语音合成的api
然后内部使用Python的SDK(用pip去安装)
如果返回dict,且发现是err_no是502的话,则确定是token无效或过期
则使用refresh_token去重新刷新获得有效的token
重新再去尝试一次
然后正常的话,返回得到mp3的数据
再考虑如何处理,放到哪里,生成一个外部可以直接访问的url
此处,参考:
SPEECH-TTS-CORS/demo.html at master · Baidu-AIP/SPEECH-TTS-CORS
发现是:
        // 参数含义请参考 https://ai.baidu.com/docs#/TTS-API/41ac79a6
        audio = btts({
。。。
            onSuccess: function(htmlAudioElement) {

                audio = htmlAudioElement;
                playBtn.innerText = '播放';
            },
发现是btts直接返回了audio这个html的element?
去看:
SPEECH-TTS-CORS/baidu_tts_cors.js at master · Baidu-AIP/SPEECH-TTS-CORS
发现是:
document.body.append(audio);
audio.setAttribute('src', URL.createObjectURL(xhr.response));
好像是创建了本地的文件了?
去搜:
URL.createObjectURL
URL.createObjectURL() – Web API 接口 | MDN
Blob – Web API 接口 | MDN
URL.createObjectURL和URL.revokeObjectURL – 流浪猫の窝 – 博客园
“File对象,就是一个文件,比如我用input type=”file”标签来上传文件,那么里面的每个文件都是一个File对象.
Blob对象,就是二进制数据,比如通过new Blob()创建的对象就是Blob对象.又比如,在XMLHttpRequest里,如果指定responseType为blob,那么得到的返回值也是一个blob对象.”
所以此处就是返回了mp3的二进制数据,是blob格式,传递给createObjectURL,生成了临时的文件,可以用来播放了
-》那么我后续封装出来的接口,倒是也可以考虑支持两种:
  • 直接返回mp3的url
  • 返回mp3的二进制数据文件
而返回的类型,可以通过输入参数指定
然后就是去:
【已解决】后台用Flask封装百度的语音合成功能对外提供REST API接口
接着就可以继续去:
【已解决】Flask中如何保存临时文件且可以指定有效期
接着就是去:
前端web页面中把相关的之前输出text的接口,更新为,解析返回的mp3的(临时文件)的url,以及调用播放器播放出来:
var curResponseDict = respJsonObj["data"]["response"];
console.log("curResponseDict=%s", curResponseDict);

var curResponseText = curResponseDict["text"];
console.log("curResponseText=%s", curResponseText);
$('#response_text p').text(curResponseText);

var curResponseAudioUrl = curResponseDict["audioUrl"];
console.log("curResponseAudioUrl=%s", curResponseAudioUrl);
if (curResponseAudioUrl) {
    console.log("now play the response text's audio %s", curResponseAudioUrl);

    var respTextAudioObj = $(".response_text_audio_player audio")[0];
    console.log("respTextAudioObj=%o", respTextAudioObj);

    $(".response_text_audio_player .col-sm-offset-1").text(curResponseText);

    $(".response_text_audio_player audio source").attr("src", curResponseAudioUrl);
    respTextAudioObj.load();
    console.log("has load respTextAudioObj=%o", respTextAudioObj);

    respTextAudioPromise = respTextAudioObj.play();
    // console.log("respTextAudioPromise=%o", respTextAudioPromise);
    if (respTextAudioPromise !== undefined) {
        respTextAudioPromise.then(() => {
            // Auto-play started
            console.log("Auto paly audio started, respTextAudioPromise=%o", respTextAudioPromise);
        }).catch(error => {
            // Auto-play was prevented
            // Show a UI element to let the user manually start playback

            console.error("play response text's audio promise error=%o", error);
            //NotAllowedError: The request is not allowed by the user agent or the platform in the current context, possibly because the user denied permission.
        });
    }
}
已经可以去播放返回的text的audio了:
然后等个1秒左右,再播放被点播的文件
所以先要去解决:
【已解决】html网页中如何用js或jquery在音频文件播放后去触发其他动作
然后故意再去优化,当出错时显示错误信息,期间:
【已解决】js中如何实现字符串拼接或格式化
以及:
【已解决】js的jquery的ajax的get返回的error错误的详细信息
【总结】
最后实现了想要的效果
后端:
Flask的REST API和百度接口的初始化和调用:
app.py
from flask import Flask
from flask import jsonify
from flask_restful import Resource, Api, reqparse
import logging
from logging.handlers import RotatingFileHandler
from bson.objectid import ObjectId
from flask import send_file
import os
import io
import re
from urllib.parse import quote
import json
import uuid
from flask_cors import CORS
import requests

from celery import Celery

################################################################################
# Global Definitions
################################################################################
"""
http://ai.baidu.com/docs#/TTS-API/top
    500    不支持输入
    501    输入参数不正确
    502    token验证失败
    503    合成后端错误
"""
BAIDU_ERR_NOT_SUPPORT_PARAM = 500
BAIDU_ERR_PARAM_INVALID = 501
BAIDU_ERR_TOKEN_INVALID = 502
BAIDU_ERR_BACKEND_SYNTHESIS_FAILED = 503

################################################################################
# Global Variables
################################################################################
log = None
app = None
"""
{
    "access_token": "24.569bcccccccc11192484",
    "session_key": "9mxxxxxxEIB==",
    "scope": "public audio_voice_assistant_get audio_tts_post wise_adapt lebo_resource_base lightservice_public hetu_basic lightcms_map_poi kaidian_kaidian ApsMisTest_Test权限 vis-classify_flower lpq_开放 cop_helloScope ApsMis_fangdi_permission smartapp_snsapi_base",
    "refresh_token": "25.6acfxxxx2483",
    "session_secret": "121xxxxxfa",
    "expires_in": 2592000
}
"""
gCurBaiduRespDict = {} # get baidu token resp dict
gTempAudioFolder = ""

################################################################################
# Global Function
################################################################################

def generateUUID(prefix = ""):
    generatedUuid4 = uuid.uuid4()
    generatedUuid4Str = str(generatedUuid4)
    newUuid = prefix + generatedUuid4Str
    return newUuid

#----------------------------------------
# Audio Synthesis / TTS
#----------------------------------------

def createAudioTempFolder():
    """create foler to save later temp audio files"""
    global log, gTempAudioFolder

    # init audio temp folder for later store temp audio file
    audioTmpFolder = app.config["AUDIO_TEMP_FOLDER"]
    log.info("audioTmpFolder=%s", audioTmpFolder)
    curFolderAbsPath = os.getcwd() #'/Users/crifan/dev/dev_root/company/xxx/projects/robotDemo/server'
    log.info("curFolderAbsPath=%s", curFolderAbsPath)
    audioTmpFolderFullPath = os.path.join(curFolderAbsPath, audioTmpFolder)
    log.info("audioTmpFolderFullPath=%s", audioTmpFolderFullPath)
    if not os.path.exists(audioTmpFolderFullPath):
        os.makedirs(audioTmpFolderFullPath)
        log.info("++++++ Created tmp audio folder: %s", audioTmpFolderFullPath)

    gTempAudioFolder = audioTmpFolderFullPath
    log.info("gTempAudioFolder=%s", gTempAudioFolder)

def initAudioSynthesis():
    """
    init audio synthesis related:
        init token
    :return:
    """

    getBaiduToken()
    createAudioTempFolder()


def getBaiduToken():
    """get baidu token"""
    global app, log, gCurBaiduRespDict

    getBaiduTokenUrlTemplate = "
https://openapi.baidu.com/oauth/2.0/token?grant_type=client_credentials&client_id=%s&client_secret=%s
"
    getBaiduTokenUrl = getBaiduTokenUrlTemplate % (app.config["BAIDU_API_KEY"], app.config["BAIDU_SECRET_KEY"])
    log.info("getBaiduTokenUrl=%s", getBaiduTokenUrl) #
https://openapi.baidu.com/oauth/2.0/token?grant_type=client_credentials&client_id=xxxz&client_secret=xxxx
    resp = requests.get(getBaiduTokenUrl)
    log.info("resp=%s", resp)
    respJson = resp.json()
    log.info("respJson=%s", respJson) #{'access_token': '24.xxx.2592000.1528609320.282335-11192484', 'session_key': 'xx+I/xx+6KwgZmw==', 'scope': 'public audio_voice_assistant_get audio_tts_post wise_adapt lebo_resource_base lightservice_public hetu_basic lightcms_map_poi kaidian_kaidian ApsMisTest_Test权限 vis-classify_flower lpq_开放 cop_helloScope ApsMis_fangdi_permission smartapp_snsapi_base', 'refresh_token': '25.xxx', 'session_secret': 'cxxx6e', 'expires_in': 2592000}
    if resp.status_code == 200:
        gCurBaiduRespDict = respJson
        log.info("get baidu token resp: %s", gCurBaiduRespDict)
    else:
        log.error("error while get baidu token: %s", respJson)
        #{'error': 'invalid_client', 'error_description': 'Client authentication failed'}
        #{'error': 'invalid_client', 'error_description': 'unknown client id'}
        #{'error': 'unsupported_grant_type', 'error_description': 'The authorization grant type is not supported'}

def refreshBaiduToken():
    """refresh baidu token when current token invalid"""
    global app, log, gCurBaiduRespDict
    if gCurBaiduRespDict:
        refreshBaiduTokenUrlTemplate = "
https://openapi.baidu.com/oauth/2.0/token?grant_type=refresh_token&refresh_token=%s&client_id=%s&client_secret=%s
"
        refreshBaiduTokenUrl = refreshBaiduTokenUrlTemplate % (gCurBaiduRespDict["refresh_token"], app.config["BAIDU_API_KEY"], app.config["BAIDU_SECRET_KEY"])
        log.info("refreshBaiduTokenUrl=%s", refreshBaiduTokenUrl) #
https://openapi.baidu.com/oauth/2.0/token?grant_type=refresh_token&refresh_token=25.1xxxx.xx.1841379583.282335-11192483&client_id=Sxxxxz&client_secret=47dxxxxa
        resp = requests.get(refreshBaiduTokenUrl)
        log.info("resp=%s", resp)
        respJson = resp.json()
        log.info("respJson=%s", respJson)
        if resp.status_code == 200:
            gCurBaiduRespDict = respJson
            log.info("Ok to refresh baidu token response: %s", gCurBaiduRespDict)
        else:
            log.error("error while refresh baidu token: %s", respJson)
    else:
        log.error("Can't refresh baidu token for previous not get token")


def baiduText2Audio(unicodeText):
    """call baidu text2audio to generate mp3 audio from text"""
    global app, log, gCurBaiduRespDict
    log.info("baiduText2Audio: unicodeText=%s", unicodeText)

    isOk = False
    mp3BinData = None
    errNo = 0
    errMsg = "Unknown error"

    if not gCurBaiduRespDict:
        errMsg = "Need get baidu token before call text2audio"
        return isOk, mp3BinData, errNo, errMsg

    utf8Text = unicodeText.encode("utf-8")
    log.info("utf8Text=%s", utf8Text)
    encodedUtf8Text = quote(unicodeText)
    log.info("encodedUtf8Text=%s", encodedUtf8Text)

    #
 
http://ai.baidu.com/docs#/TTS-API/top
    tex = encodedUtf8Text #合成的文本,使用UTF-8编码。小于512个中文字或者英文数字。(文本在百度服务器内转换为GBK后,长度必须小于1024字节)
    tok = gCurBaiduRespDict["access_token"] #开放平台获取到的开发者access_token(见上面的“鉴权认证机制”段落)
    cuid = app.config["FLASK_APP_NAME"] #用户唯一标识,用来区分用户,计算UV值。建议填写能区分用户的机器 MAC 地址或 IMEI 码,长度为60字符以内
    ctp = 1 #客户端类型选择,web端填写固定值1
    lan = "zh" #固定值zh。语言选择,目前只有中英文混合模式,填写固定值zh
    spd = 5 #语速,取值0-9,默认为5中语速
    pit = 5 #音调,取值0-9,默认为5中语调
    # vol = 5 #音量,取值0-9,默认为5中音量
    vol = 9
    per = 0 #发音人选择, 0为普通女声,1为普通男生,3为情感合成-度逍遥,4为情感合成-度丫丫,默认为普通女声
    getBaiduSynthesizedAudioTemplate = "
http://tsn.baidu.com/text2audio?lan=%s&ctp=%s&cuid=%s&tok=%s&vol=%s&per=%s&spd=%s&pit=%s&tex=%s
"
    getBaiduSynthesizedAudioUrl = getBaiduSynthesizedAudioTemplate % (lan, ctp, cuid, tok, vol, per, spd, pit, tex)
    log.info("getBaiduSynthesizedAudioUrl=%s", getBaiduSynthesizedAudioUrl) #
http://tsn.baidu.com/text2audio?lan=zh&ctp=1&cuid=RobotQA&tok=24.5f056b15e9d5da63256bac89f64f61b5.2592000.1528609737.282335-11192483&vol=5&per=0&spd=5&pit=5&tex=as%20a%20book-collector%2C%20i%20have%20the%20story%20you%20just%20want%20to%20listen%21
    resp = requests.get(getBaiduSynthesizedAudioUrl)
    log.info("resp=%s", resp)
    respContentType = resp.headers["Content-Type"]
    respContentTypeLowercase = respContentType.lower() #'audio/mp3'
    log.info("respContentTypeLowercase=%s", respContentTypeLowercase)
    if respContentTypeLowercase == "audio/mp3":
        mp3BinData = resp.content
        log.info("resp content is binary data of mp3, length=%d", len(mp3BinData))
        isOk = True
        errMsg = ""
    elif respContentTypeLowercase == "application/json":
        """                        
            {
              'err_detail': 'Invalid params per or lan!',
              'err_msg': 'parameter error.',
              'err_no': 501,
              'err_subcode': 50000,
              'tts_logid': 642798357
            }

            {
              'err_detail': 'Invalid params per&pdt!',
              'err_msg': 'parameter error.',
              'err_no': 501,
              'err_subcode': 50000,
              'tts_logid': 1675521246
            }

            {
              'err_detail': 'Access token invalid or no longer valid',
              'err_msg': 'authentication failed.',
              'err_no': 502,
              'err_subcode': 50004,
              'tts_logid': 4221215043
            }
        """
        log.info("resp content is json -> occur error")

        isOk = False
        respDict = resp.json()
        log.info("respDict=%s", respDict)
        errNo = respDict["err_no"]
        errMsg = respDict["err_msg"] + " " + respDict["err_detail"]
    else:
        isOk = False
        errMsg = "Unexpected response content-type: %s" % respContentTypeLowercase

    return isOk, mp3BinData, errNo, errMsg

def doAudioSynthesis(unicodeText):
    """
        do audio synthesis from unicode text
        if failed for token invalid/expired, will refresh token to do one more retry
    """
    global app, log, gCurBaiduRespDict
    isOk = False
    audioBinData = None
    errMsg = ""

    # # for debug
    # gCurBaiduRespDict["access_token"] = "99.569b3b5b470938a522ce60d2e2ea2506.2592000.1528015602.282335-11192483"

    log.info("doAudioSynthesis: unicodeText=%s", unicodeText)
    isOk, audioBinData, errNo, errMsg = baiduText2Audio(unicodeText)
    log.info("isOk=%s, errNo=%d, errMsg=%s", isOk, errNo, errMsg)

    if isOk:
        errMsg = ""
        log.info("got synthesized audio binary data length=%d", len(audioBinData))
    else:
        if errNo == BAIDU_ERR_TOKEN_INVALID:
            log.warning("Token invalid -> refresh token")
            refreshBaiduToken()

            isOk, audioBinData, errNo, errMsg = baiduText2Audio(unicodeText)
            log.info("after refresh token: isOk=%ss, errNo=%s, errMsg=%s", isOk, errNo, errMsg)
        else:
            log.warning("try synthesized audio occur error: errNo=%d, errMsg=%s", errNo, errMsg)
            audioBinData = None

    log.info("return isOk=%s, errMsg=%s", isOk, errMsg)
    if audioBinData:
        log.info("audio binary bytes=%d", len(audioBinData))
    return isOk, audioBinData, errMsg


def testAudioSynthesis():
    global app, log, gTempAudioFolder

    testInputUnicodeText = u"as a book-collector, i have the story you just want to listen!"
    isOk, audioBinData, errMsg = doAudioSynthesis(testInputUnicodeText)
    if isOk:
        audioBinDataLen = len(audioBinData)
        log.info("Now will save audio binary data %d bytes to file", audioBinDataLen)

        # 1. save mp3 binary data into tmp file
        newUuid = generateUUID()
        log.info("newUuid=%s", newUuid)
        tempFilename = newUuid + ".mp3"
        log.info("tempFilename=%s", tempFilename)
        if not gTempAudioFolder:
            createAudioTempFolder()
        tempAudioFullname = os.path.join(gTempAudioFolder, tempFilename) #'/Users/crifan/dev/dev_root/company/xxx/projects/robotDemo/server/tmp/audio/2aba73d1-f8d0-4302-9dd3-d1dbfad44458.mp3'
        log.info("tempAudioFullname=%s", tempAudioFullname)

        with open(tempAudioFullname, 'wb') as tmpAudioFp:
            log.info("tmpAudioFp=%s", tmpAudioFp)
            tmpAudioFp.write(audioBinData)
            tmpAudioFp.close()
            log.info("Done to write audio data into file of %d bytes", audioBinDataLen)

        # 2. use celery to delay delete tmp file
    else:
        log.warning("Fail to get synthesis audio for errMsg=%s", errMsg)


#----------------------------------------
# Flask API
#----------------------------------------
def sendFile(fileBytes, contentType, outputFilename):
    """Flask API use this to send out file (to browser, browser can directly download file)"""
    return send_file(
        io.BytesIO(fileBytes),
        # io.BytesIO(fileObj.read()),
        mimetype=contentType,
        as_attachment=True,
        attachment_filename=outputFilename
    )

################################################################################
# Global Init App
################################################################################
app = Flask(__name__)
CORS(app)
# app.config.from_object('config.DevelopmentConfig')
app.config.from_object('config.ProductionConfig')

logFormatterStr = app.config["LOG_FORMAT"]
logFormatter = logging.Formatter(logFormatterStr)

fileHandler = RotatingFileHandler(
    app.config['LOG_FILE_FILENAME'],
    maxBytes=app.config["LOF_FILE_MAX_BYTES"],
    backupCount=app.config["LOF_FILE_BACKUP_COUNT"],
    encoding="UTF-8")
fileHandler.setLevel(logging.DEBUG)
fileHandler.setFormatter(logFormatter)
app.logger.addHandler(fileHandler)


app.logger.setLevel(logging.DEBUG) # set root log level

log = app.logger
log.info("app=%s", app)
# log.debug("app.config=%s", app.config)

api = Api(app)
log.info("api=%s", api)

celeryApp = Celery(app.name, broker=app.config['CELERY_BROKER_URL'])
celeryApp.conf.update(app.config)
log.info("celeryApp=%s", celeryApp)

aiContext = Context()
log.info("aiContext=%s", aiContext)

initAudioSynthesis()
# testAudioSynthesis()

...

#----------------------------------------
# Celery tasks
#----------------------------------------

# @celeryApp.task()
@celeryApp.task
# @celeryApp.task(name=app.config["CELERY_TASK_NAME"] + ".deleteTmpAudioFile")
def deleteTmpAudioFile(filename):
    """
        delete tmp audio file from filename
            eg: 98fc7c46-7aa0-4dd7-aa9d-89fdf516abd6.mp3
    """
    global log

    log.info("deleteTmpAudioFile: filename=%s", filename)

    audioTmpFolder = app.config["AUDIO_TEMP_FOLDER"]
    # audioTmpFolder = "tmp/audio"
    log.info("audioTmpFolder=%s", audioTmpFolder)
    curFolderAbsPath = os.getcwd() #'/Users/crifan/dev/dev_root/company/xxx/projects/robotDemo/server'
    log.info("curFolderAbsPath=%s", curFolderAbsPath)
    audioTmpFolderFullPath = os.path.join(curFolderAbsPath, audioTmpFolder)
    log.info("audioTmpFolderFullPath=%s", audioTmpFolderFullPath)
    tempAudioFullname = os.path.join(audioTmpFolderFullPath, filename)
    #'/Users/crifan/dev/dev_root/company/xxx/projects/robotDemo/server/tmp/audio/2aba73d1-f8d0-4302-9dd3-d1dbfad44458.mp3'
    if os.path.isfile(tempAudioFullname):
        os.remove(tempAudioFullname)
        log.info("Ok to delete file %s", tempAudioFullname)
    else:
        log.warning("No need to remove for not exist file %s", tempAudioFullname)

# log.info("deleteTmpAudioFile=%s", deleteTmpAudioFile)
# log.info("deleteTmpAudioFile.name=%s", deleteTmpAudioFile.name)
# log.info("celeryApp.tasks=%s", celeryApp.tasks)

#----------------------------------------
# Rest API
#----------------------------------------

class RobotQaAPI(Resource):

    def processResponse(self, respDict):
        """
            process response dict before return
                generate audio for response text part
        """
        global log, gTempAudioFolder

        tmpAudioUrl = ""

        unicodeText = respDict["data"]["response"]["text"]
        log.info("unicodeText=%s")

        if not unicodeText:
            log.info("No response text to do audio synthesis")
            return jsonify(respDict)

        isOk, audioBinData, errMsg = doAudioSynthesis(unicodeText)
        if isOk:
            audioBinDataLen = len(audioBinData)
            log.info("audioBinDataLen=%s", audioBinDataLen)

            # 1. save mp3 binary data into tmp file
            newUuid = generateUUID()
            log.info("newUuid=%s", newUuid)
            tempFilename = newUuid + ".mp3"
            log.info("tempFilename=%s", tempFilename)
            if not gTempAudioFolder:
                createAudioTempFolder()
            tempAudioFullname = os.path.join(gTempAudioFolder, tempFilename)
            log.info("tempAudioFullname=%s", tempAudioFullname) # 'xxx/tmp/audio/2aba73d1-f8d0-4302-9dd3-d1dbfad44458.mp3'

            with open(tempAudioFullname, 'wb') as tmpAudioFp:
                log.info("tmpAudioFp=%s", tmpAudioFp)
                tmpAudioFp.write(audioBinData)
                tmpAudioFp.close()
                log.info("Saved %d bytes data into temp audio file %s", audioBinDataLen, tempAudioFullname)

            # 2. use celery to delay delete tmp file
            delayTimeToDelete = app.config["CELERY_DELETE_TMP_AUDIO_FILE_DELAY"]
            deleteTmpAudioFile.apply_async([tempFilename], countdown=delayTimeToDelete)
            log.info("Delay %s seconds to delete %s", delayTimeToDelete, tempFilename)

            # generate temp audio file url
            # /tmp/audio
            tmpAudioUrl = "http://%s:%d/tmp/audio/%s" % (
                app.config["FILE_URL_HOST"],
                app.config["FLASK_PORT"],
                tempFilename)
            log.info("tmpAudioUrl=%s", tmpAudioUrl)
            respDict["data"]["response"]["audioUrl"] = tmpAudioUrl
        else:
            log.warning("Fail to get synthesis audio for errMsg=%s", errMsg)

        log.info("respDict=%s", respDict)
        return jsonify(respDict)

    def get(self):
        respDict = {
            "code": 200,
            "message": "generate response ok",
            "data": {
                "input": "",
                "response": {
                    "text": "",
                    "audioUrl": ""
                },
                "control": "",
                "audio": {}
            }
        }

        parser = reqparse.RequestParser()
        # i want to hear the story of Baby Sister Says No
        parser.add_argument('input', type=str, help="input words")
        log.info("parser=%s", parser)

        parsedArgs = parser.parse_args()  #
        log.info("parsedArgs=%s", parsedArgs)
        if not parsedArgs:
            respDict["data"]["response"]["text"] = "Can not recognize input"
            return self.processResponse(respDict)

        inputStr = parsedArgs["input"]
        log.info("inputStr=%s", inputStr)

        if not inputStr:
            respDict["data"]["response"]["text"] = "Can not recognize parameter input"
            return self.processResponse(respDict)

        respDict["data"]["input"] = inputStr

        aiResult = QueryAnalyse(inputStr, aiContext)
        log.info("aiResult=%s", aiResult)

        if aiResult["response"]:
            respDict["data"]["response"]["text"] = aiResult["response"]
        if aiResult["control"]:
            respDict["data"]["control"] = aiResult["control"]
        log.info('respDict["data"]=%s', respDict["data"])

        audioFileIdStr = aiResult["mediaId"]
        log.info("audioFileIdStr=%s", audioFileIdStr)

        if audioFileIdStr:
            audioFileObjectId = ObjectId(audioFileIdStr)
            log.info("audioFileObjectId=%s", audioFileObjectId)

            if fsCollection.exists(audioFileObjectId):
                audioFileObj = fsCollection.get(audioFileObjectId)
                log.info("audioFileObj=%s", audioFileObj)
                encodedFilename = quote(audioFileObj.filename)
                log.info("encodedFilename=%s", encodedFilename)

                respDict["data"]["audio"] = {
                    "contentType": audioFileObj.contentType,
                    "name": audioFileObj.filename,
                    "size": audioFileObj.length,
                    "url": "http://%s:%d/files/%s/%s" %
                           (app.config["FILE_URL_HOST"],
                            app.config["FLASK_PORT"],
                            audioFileObj._id,
                            encodedFilename)
                }
                log.info("respDict=%s", respDict)
                return self.processResponse(respDict)
            else:
                log.info("Can not find file from id %s", audioFileIdStr)
                respDict["data"]["audio"] = {}
                return self.processResponse(respDict)
        else:
            log.info("Not response file id")
            respDict["data"]["audio"] = {}
            return self.processResponse(respDict)


class GridfsAPI(Resource):

    def get(self, fileId, fileName=None):
        log.info("fileId=%s, file_name=%s", fileId, fileName)

        fileIdObj = ObjectId(fileId)
        log.info("fileIdObj=%s", fileIdObj)
        if not fsCollection.exists({"_id": fileIdObj}):
            respDict = {
                "code": 404,
                "message": "Can not find file from object id %s" % (fileId),
                "data": {}
            }
            return jsonify(respDict)

        fileObj = fsCollection.get(fileIdObj)
        log.info("fileObj=%s, filename=%s, chunkSize=%s, length=%s, contentType=%s",
                 fileObj, fileObj.filename, fileObj.chunk_size, fileObj.length, fileObj.content_type)
        log.info("lengthInMB=%.2f MB", float(fileObj.length / (1024 * 1024)))

        fileBytes = fileObj.read()
        log.info("len(fileBytes)=%s", len(fileBytes))

        outputFilename = fileObj.filename
        if fileName:
            outputFilename = fileName
        log.info("outputFilename=%s", outputFilename)

        return sendFile(fileBytes, fileObj.content_type, outputFilename)


class TmpAudioAPI(Resource):

    def get(self, filename=None):
        global gTempAudioFolder

        log.info("TmpAudioAPI: filename=%s", filename)

        tmpAudioFullPath = os.path.join(gTempAudioFolder, filename)
        log.info("tmpAudioFullPath=%s", tmpAudioFullPath)

        if not os.path.isfile(tmpAudioFullPath):
            log.warning("Not exists file %s", tmpAudioFullPath)
            respDict = {
                "code": 404,
                "message": "Can not find temp audio file %s" % filename,
                "data": {}
            }
            return jsonify(respDict)

        fileSize = os.path.getsize(tmpAudioFullPath)
        log.info("fileSize=%s", fileSize)

        with open(tmpAudioFullPath, "rb") as tmpAudioFp:
            fileBytes = tmpAudioFp.read()
            log.info("read out fileBytes length=%s", len(fileBytes))

            outputFilename = filename
            # contentType = "audio/mp3" # chrome use this
            contentType = "audio/mpeg" # most common and compatible

            return sendFile(fileBytes, contentType, outputFilename)


api.add_resource(PlaySongAPI, '/playsong', endpoint='playsong')
api.add_resource(RobotQaAPI, '/qa', endpoint='qa')
api.add_resource(GridfsAPI, '/files/<fileId>', '/files/<fileId>/<fileName>', endpoint='gridfs')
api.add_resource(TmpAudioAPI, '/tmp/audio/<filename>', endpoint='TmpAudio')

if __name__ == "__main__":
    app.run(
        host=app.config["FLASK_HOST"],
        port=app.config["FLASK_PORT"],
        debug=app.config["DEBUG"]
    )
config.py
class BaseConfig(object):
    DEBUG = False

    FLASK_PORT = 3xxxx
    # FLASK_HOST = "127.0.0.1"
    # FLASK_HOST = "localhost"
    # Note:
    # 1. to allow external access this server
    # 2. make sure here gunicorn parameter "bind" is same with here !!!
    FLASK_HOST = "0.0.0.0"

    # Flask app name
    FLASK_APP_NAME = "RobotQA"

    # Log File
    LOG_FILE_FILENAME = "logs/" + FLASK_APP_NAME + ".log"
    LOG_FORMAT = "[%(asctime)s %(levelname)s %(filename)s:%(lineno)d %(funcName)s] %(message)s"
    LOF_FILE_MAX_BYTES = 2*1024*1024
    LOF_FILE_BACKUP_COUNT = 10

    # reuturn file url's host
    # FILE_URL_HOST = FLASK_HOST
    FILE_URL_HOST = "127.0.0.1"

    # Audio Synthesis / TTS
    # BAIDU_APP_ID = "1xxx3"
    BAIDU_API_KEY = "Sxxxxz"
    BAIDU_SECRET_KEY = "4xxxxxa"

    AUDIO_TEMP_FOLDER = "tmp/audio"

    # CELERY_TASK_NAME = "Celery_" + FLASK_APP_NAME
    # CELERY_BROKER_URL = "
redis://localhost
"
    CELERY_BROKER_URL = "
redis://localhost:6379/0
"
    # CELERY_RESULT_BACKEND = "
redis://localhost:6379/0
" # current not use result
    CELERY_DELETE_TMP_AUDIO_FILE_DELAY = 60 * 2 # two minutes

class DevelopmentConfig(BaseConfig):
    # DEBUG = True
    # for local dev, need access remote mongodb
    MONGODB_HOST = "47.xx.xx.xx"
    FILE_URL_HOST = "127.0.0.1"


class ProductionConfig(BaseConfig):
    FILE_URL_HOST = "47.xx.xx.xx"
前端:
main.html
<!doctype html>
<html lang="en">
  <head>
    <!-- Required meta tags -->
    <meta charset="utf-8">
    <meta http-equiv="X-UA-Compatible" content="IE=edge">
    <meta name="viewport" content="width=device-width, initial-scale=1">
    <!-- <meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no"> -->

    <!-- Bootstrap CSS -->
    <link rel="stylesheet" href="css/bootstrap-3.3.1/bootstrap.css">
    <!-- <link rel="stylesheet" href="css/highlightjs_default.css"> -->
    <link rel="stylesheet" href="css/highlight_atom-one-dark.css">
    <!-- <link rel="stylesheet" href="css/highlight_monokai-sublime.css"> -->
    <link rel="stylesheet" href="css/bootstrap3_player.css">
    <link rel="stylesheet" href="css/main.css">

    <title>xxx英语智能机器人演示</title>

    <!-- HTML5 shim and Respond.js for IE8 support of HTML5 elements and media queries -->
    <!-- WARNING: Respond.js doesn't work if you view the page via file:// -->
    <!--[if lt IE 9]>
      <script src="
https://oss.maxcdn.com/html5shiv/3.7.2/html5shiv.min.js
"></script>
      <script src="
https://oss.maxcdn.com/respond/1.4.2/respond.min.js
"></script>
    <![endif]-->
  </head>
    <div class="logo text-center">
        <img class="mb-4" src="img/logo_transparent_183x160.png" alt="xxx Logo" width="72" height="72">
    </div>

    <h2>xxx英语智能机器人</h2>
    <h4>xxx Bot for Kids</h4>

    <div class="panel panel-primary">
      <div class="panel-heading">
        <h3 class="panel-title">Input</h3>
      </div>
      <div class="panel-body">
          <ul class="list-group">
            <li class="list-group-item">
                <h3 class="panel-title">Input Example</h3>
                <ul>
                    <li>i want to hear the story of apple</li>
                    <li>say story apple</li>
                    <li>say apple</li>
                    <li>next episode</li>
                    <li>next</li>
                    <li>i want you stop reading</li>
                    <li>stop reading</li>
                    <li>please go on</li>
                    <li>go on</li>
                </ul>
            </li>

            <li class="list-group-item">
                <!--
                <form>
                  <div class="form-group input_request">
                    <input id="inputRequest" type="text" class="form-control" placeholder="请输入您要说的话" value="i want to hear the story of apple">
                  </div>
                  <div class="form-group">
                    <button id="submitInput" type="submit" class="btn btn-primary btn-lg col-sm-3 btn-block">提交</button>
                    <button id="clearInput" class="btn btn-secondary btn-lg col-sm-3" type="button">清除</button>
                    <button id="clearInput" class="btn btn-info btn-lg col-sm-3 btn-block" type="button">清除</button>
                  </div>
                </form>
                 -->

                <div class="row">
                  <div class="col-lg-12">
                    <div class="input-group">
                      <input id="inputRequest" type="text" class="form-control" placeholder="请输入您要说的话" value="say apple">
                      <span class="input-group-btn">
                        <button id="submitInput" type="submit" class="btn btn-primary" type="button">提交</button>
                      </span>
                    </div><!-- /input-group -->
                  </div><!-- /.col-lg-6 -->
                </div>
            </li>
          </ul>

      </div>
    </div>

<!--
  <div class="input_example bg-light box-shadow">
    <h5>Input Example:</h5>
    <ul>
        <li>i want to hear the story of apple</li>
        <li>next episode</li>
        <li>i want you stop reading</li>
        <li>please go on</li>
    </ul>
  </div>
-->

    <div class="panel panel-success">
      <div class="panel-heading">
        <h3 class="panel-title">Output</h3>
      </div>
      <div class="panel-body">

        <div id="response_text" class="alert alert-success" role="alert">
          <p>here will output response text</p>

          <div class="response_text_audio_player">
            <audio
                controls
                data-info-att="response text's audio">
                <source src="" type="audio/mpeg" />
            </audio>
          </div>
        </div>

        <div class="audio_player col-md-12 col-xs-12">
            <audio
                controls
                data-info-att="">
                <source src="" type="" />
            </audio>
        </div>

        <!--
        <div id="audio_play_prevented" class="alert alert-warning alert-dismissible col-md-12 col-xs-12">
          <button type = "button" class="close" data-dismiss = "alert">x</button>
          <strong>Notice:</strong> Auto play prevented, please mannually click above play button to play
        </div>
        -->

    <!--
        <div id="response_json" class="bg-light box-shadow">
            <pre><code class="json">here will output response</code></pre>
        </div>
    -->

    <!--
        <pre id="response_json">
            <code class="json">here will output response</code>
        </pre>
    -->

        <div id="response_json">
          <code class="json">here will output response</code>
        </div>

      </div>
    </div>


    <!-- Optional JavaScript -->
    <!-- jQuery first, then Popper.js, then Bootstrap JS -->
    <!-- <script src="js/jquery-3.3.1.js"></script> -->
    <!-- <script src="
https://ajax.googleapis.com/ajax/libs/jquery/1.11.1/jquery.min.js
"></script> -->
    <script src="js/jquery/1.11.1/jquery-1.11.1.js"></script>
    <!-- <script src="
https://code.jquery.com/jquery-3.3.1.slim.min.js
" integrity="sha384-q8i/X+965DzO0rT7abK41JStQIAqVgRVzpbzo5smXKp4YfRvH+8abtTE1Pi6jizo" crossorigin="anonymous"></script> -->
    <!-- <script src="
https://cdnjs.cloudflare.com/ajax/libs/popper.js/1.14.0/umd/popper.min.js
" integrity="sha384-cs/chFZiN24E4KMATLdqdvsezGxaGsi4hLGOzlXwp5UZB1LY//20VyM2taTB4QvJ" crossorigin="anonymous"></script> -->
    <script src="js/popper-1.14.0/popper.min.js"></script>
    <!-- <script src="js/bootstrap.js"></script> -->
    <script src="js/bootstrap-3.3.1/bootstrap.min.js"></script>
    <script src="js/highlight.js"></script>
    <script src="js/bootstrap3_player.js"></script>
    <script src="js/main.js"></script>
  </body>
</html>
main.css
.logo{
    padding: 10px 2%;
}

h2{
    text-align: center;
    margin-top: 10px;
    margin-bottom: 10px;
}

h4{
    text-align: center;
    margin-top: 0px;
    margin-bottom: 20px;
}

form {
    text-align: center;
}

.form-group {
    /*padding-left: 1%;*/
    /*padding-right: 1%;*/
}

.input_example {
    /*padding: 1px 1%;*/
}

#response_json {
    /*width: 96%;*/
    height: 380px;
    border-radius: 10px;
    padding-top: 20px;

    /*padding-left: 1%;*/
    /*padding-right: 1%;*/
}

#response_text {
    text-align: center !important;
    font-size: 14px;

    /* padding-left: 4%;
    padding-right: 4%; */
}

/*pre {*/
    /*padding-left: 2%;*/
    /*padding-right: 2%;*/
/*}*/

.audio_player {
    margin-top: 10px;
    margin-bottom: 5px;
    text-align: center;

    padding-left: 0 !important;
    padding-right: 0 !important;
}

.response_text_audio_player{
    /* visibility: hidden; */

    width: 100%;
    /* height: 1px !important; */
    height: 100px;
}

/* #audio_play_prevented {
    display: none;
} */
main.js
if (!String.format) {
    String.format = function(format) {
      var args = Array.prototype.slice.call(arguments, 1);
      return format.replace(/{(\d+)}/g, function(match, number) {
        return typeof args[number] != 'undefined'
          ? args[number]
          : match
        ;
      });
    };
}


$(document).ready(function(){

    $('[data-toggle="tooltip"]').tooltip();

    // when got response json, update to highlight it
    function updateHighlight() {
        console.log("updateHighlight");

        $('pre code').each(function(i, block) {
            hljs.highlightBlock(block);
        });
    }

    updateHighlight();

    $("#submitInput").click(function(event){
        event.preventDefault();

        ajaxSubmitInput();
    });
    
    function ajaxSubmitInput() {
        console.log("ajaxSubmitInput");

        var inputRequest = $("#inputRequest").val();
        console.log("inputRequest=%s", inputRequest);
        var encodedInputRequest = encodeURIComponent(inputRequest)
        console.log("encodedInputRequest=%s", encodedInputRequest);

        // var qaUrl = "http://127.0.0.1:32851/qa";
        var qaUrl = "http://xxx:32851/qa";
        console.log("qaUrl=%s", qaUrl);
        var fullQaUrl = qaUrl + "?input=" + encodedInputRequest
        console.log("fullQaUrl=%s", fullQaUrl);

        $.ajax({
            type : "GET",
            url : fullQaUrl,
            success: function(respJsonObj){
                console.log("respJsonObj=%o", respJsonObj);

                // var respnJsonStr = JSON.stringify(respJsonObj);
                //var beautifiedJespnJsonStr = JSON.stringify(respJsonObj, null, '\t');
                var beautifiedJespnJsonStr = JSON.stringify(respJsonObj, null, 2);
                console.log("beautifiedJespnJsonStr=%s", beautifiedJespnJsonStr);
                var prevOutputValue = $('#response_json').text();
                console.log("prevOutputValue=%o", prevOutputValue);
                var afterOutputValue = $('#response_json').html('<pre><code class="json">' + beautifiedJespnJsonStr + "</code></pre>");
                console.log("afterOutputValue=%o", afterOutputValue);
                updateHighlight();

                var curResponseDict = respJsonObj["data"]["response"];
                console.log("curResponseDict=%s", curResponseDict);

                var curResponseText = curResponseDict["text"];
                console.log("curResponseText=%s", curResponseText);
                $('#response_text p').text(curResponseText);

                var curResponseAudioUrl = curResponseDict["audioUrl"];
                console.log("curResponseAudioUrl=%s", curResponseAudioUrl);
                if (curResponseAudioUrl) {
                    console.log("now play the response text's audio %s", curResponseAudioUrl);

                    var respTextAudioObj = $(".response_text_audio_player audio")[0];
                    console.log("respTextAudioObj=%o", respTextAudioObj);

                    $(".response_text_audio_player .col-sm-offset-1").text(curResponseText);

                    $(".response_text_audio_player audio source").attr("src", curResponseAudioUrl);
                    respTextAudioObj.load();
                    console.log("has load respTextAudioObj=%o", respTextAudioObj);

                    respTextAudioObj.onended = function() {
                        console.log("play response text's audio ended");

                        var dataControl = respJsonObj["data"]["control"];
                        console.log("dataControl=%o", dataControl);

                        var audioElt = $(".audio_player audio");
                        console.log("audioElt=%o", audioElt);
                        var audioObject = audioElt[0];
                        console.log("audioObject=%o", audioObject);

                        var playAudioPromise = undefined;

                        if (dataControl === "stop") {
                            //audioObject.stop();
                            audioObject.pause();
                            console.log("has pause audioObject=%o", audioObject);
                        } else if (dataControl === "continue") {
                            // // audioObject.load();
                            // audioObject.play();
                            // // audioObject.continue();
                            // console.log("has load and play audioObject=%o", audioObject);

                            playAudioPromise = audioObject.play();
                        }

                        if (respJsonObj["data"]["audio"]) {
                            var audioDict = respJsonObj["data"]["audio"];
                            console.log("audioDict=%o", audioDict);

                            var audioName = audioDict["name"];
                            console.log("audioName=%o", audioName);
                            var audioSize = audioDict["size"];
                            console.log("audioSize=%o", audioSize);
                            var audioType = audioDict["contentType"];
                            console.log("audioType=%o", audioType);
                            var audioUrl = audioDict["url"];
                            console.log("audioUrl=%o", audioUrl);

                            var isAudioEmpty = (!audioName && !audioSize && !audioType && !audioUrl)
                            console.log("isAudioEmpty=%o", isAudioEmpty);

                            if (isAudioEmpty) {
                                // var pauseAudioResult = audioObject.pause();
                                // console.log("pauseAudioResult=%o", pauseAudioResult);

                                // audioElt.attr("data-info-att", "");
                                // $(".col-sm-offset-1").text("");
                            } else {
                                if (audioName) {
                                    audioElt.attr("data-info-att", audioName);
                                    $(".audio_player .col-sm-offset-1").text(audioName);
                                }

                                if (audioType) {
                                    $(".audio_player audio source").attr("type", audioType);
                                }

                                if (audioUrl) {
                                    $(".audio_player audio source").attr("src", audioUrl);

                                    audioObject.load();
                                    console.log("has load audioObject=%o", audioObject);
                                }

                                console.log("dataControl=%s,audioUrl=%s", dataControl, audioUrl);

                                if ((dataControl === "") && audioUrl) {
                                    playAudioPromise = audioObject.play();
                                } else if ((dataControl === "next") && (audioUrl)) {
                                    playAudioPromise = audioObject.play();
                                }
                            }
                        } else {
                            console.log("empty respJsonObj['data']['audio']=%o", respJsonObj["data"]["audio"]);
                        }

                        if (playAudioPromise !== undefined) {
                            playAudioPromise.then(() => {
                                // Auto-play started
                                console.log("Auto paly audio started, playAudioPromise=%o", playAudioPromise);

                                //for debug
                                // showAudioPlayPreventedNotice();
                            }).catch(error => {
                                // Auto-play was prevented
                                // Show a UI element to let the user manually start playback
                                showAudioPlayPreventedNotice();

                                console.error("play audio promise error=%o", error);
                                //NotAllowedError: The request is not allowed by the user agent or the platform in the current context, possibly because the user denied permission.
                            });
                        }
                    }

                    respTextAudioPromise = respTextAudioObj.play();
                    // console.log("respTextAudioPromise=%o", respTextAudioPromise);
                    if (respTextAudioPromise !== undefined) {
                        respTextAudioPromise.then(() => {
                            // Auto-play started
                            console.log("Auto paly audio started, respTextAudioPromise=%o", respTextAudioPromise);
                        }).catch(error => {
                            // Auto-play was prevented
                            // Show a UI element to let the user manually start playback

                            console.error("play response text's audio promise error=%o", error);
                            //NotAllowedError: The request is not allowed by the user agent or the platform in the current context, possibly because the user denied permission.
                        });
                    }
                }

            },
            error : function(jqXHR, textStatus, errorThrown) {
                console.error("jqXHR=%o, textStatus=%s, errorThrown=%s", jqXHR, textStatus, errorThrown);
                var errDetail = String.format("status={0}\n\tstatusText={1}\n\tresponseText={2}", jqXHR.status, jqXHR.statusText, jqXHR.responseText);
                var errStr = String.format("GET: {0}\nERROR:\t{1}", fullQaUrl,  errDetail);
                // $('#response_text p').text(errStr);
                var responseError = $('#response_json').html('<pre><code class="html">' + errStr + "</code></pre>");
                console.log("responseError=%o", responseError);
                updateHighlight();
            }
        });
    }

    function showAudioPlayPreventedNotice(){
        console.log("showAudioPlayPreventedNotice");
        // var prevDisplayValue = $("#audio_play_prevented").css("display");
        // console.log("prevDisplayValue=%o", prevDisplayValue);
        // $("#audio_play_prevented").css({"display":"block"});

        var curAudioPlayPreventedNoticeEltHtml = $("#audio_play_prevented").html();
        console.log("curAudioPlayPreventedNoticeEltHtml=%o", curAudioPlayPreventedNoticeEltHtml);
        if (curAudioPlayPreventedNoticeEltHtml !== undefined) {
            console.log("already exist audio play prevented notice, so not insert again");
        } else {
            var audioPlayPreventedNoticeHtml = '<div id="audio_play_prevented" class="alert alert-warning alert-dismissible col-md-12 col-xs-12"><button type = "button" class="close" data-dismiss = "alert">x</button><strong>Notice:</strong> Auto play prevented, please mannually click above play button to play</div>';
            console.log("audioPlayPreventedNoticeHtml=%o", audioPlayPreventedNoticeHtml);
            $(".audio_player").append(audioPlayPreventedNoticeHtml);    
        }
    }

    $("#clearInput").click(function(event){
        // event.preventDefault();
        console.log("event=%o", event);
        $('#inputRequest').val("");

        $('#response_json').html('<pre><code class="json">here will output response</code></pre>');
        updateHighlight();
    });
});
效果:
点击提交后,后端生成临时的mp3的文件,返回到前端,前端可以正常加载并播放:

转载请注明:在路上 » 【已解决】用和适在线的语言合成接口把文字转语音

发表我的评论
取消评论

表情

Hi,您需要填写昵称和邮箱!

  • 昵称 (必填)
  • 邮箱 (必填)
  • 网址
29 queries in 0.083 seconds, using 9.64MB memory