Unity基于BestHTTP插件实现讯飞语音评测(英文文章题型)WebAPI接入方案
本方案基于 Unity+BestHTTP 插件，采用 WebSocket 协议对接科大讯飞流式口语评测（ISE）API。流程为：用户在 UI 下拉选择麦克风并录音，录音经 TrimSilence 和 PCM 转码后分帧 Base64 编码，通过协程依次发送首帧文本与多帧音频数据；服务端返回 XML，解析 `<read_chapter>` 节点并通过 UnityEvent 回调将准确度、标准度、流利
·
一、概述
本方案采用Unity引擎配合BestHTTP插件，通过WebSocket协议对接科大讯飞语音评测流式版API（ISE），实现跨平台语音评测功能。
二、环境准备
1. 开发依赖
-
Unity 2019.3.6f1或更高版本(本文基于Unity 2022.3 LTS(2022.3.42f1c1)编写)
-
BestHTTP插件（Project Setting -> Player -> Other Settings -> Allow downloads over HTTP 设置为Always allowed）
-
科大讯飞开发者账号（申请地址：https://www.xfyun.cn）
三、核心实现解析
1. 申请账号（https://www.xfyun.cn/）
(1)注册账号
(2)进入控制台
(3)创建应用并提交
(4)找到语音评测,记录服务接口认证信息
2.语音评测逻辑
(1)参数
//signature
private string signature_origin = null;
private string signature_sha = null;
private string signature = null;
//author
private string authorization_origin = null;
private string authorization = null;
//url
private string url = "wss://ise-api.xfyun.cn/v2/open-ise";
//WebSocket实例
public WebSocket webSocket;
//Websocket服务接口认证信息
public string APPID = "91b4059e";
public string APISecret = "MGNjOWY4YzgyNjAxNGFkNWIyY2IxMjUw";
public string APIKey = "13f5b50899e2adc74fb939502616bc02";
//接收
public UnityEvent<Dictionary<string, string>> messageSend = new();(2)编码
#region 编码
/// <summary>
///空格替换为 %20
/// </summary>
/// <param name="input"></param>
/// <returns></returns>
private static string ReplaceSpacesWithPercent20(string input)
{
return string.IsNullOrEmpty(input) ? input : input.Replace(" ", "%20");
}
/// <summary>
/// HmacSHA256
/// </summary>
/// <param name="secret">A</param>
/// <param name="signKey">B</param>
/// <returns></returns>
private static string HmacSHA256(string secret, string signKey)
{
var signRet = string.Empty;
using HMACSHA256 mac = new HMACSHA256(Encoding.UTF8.GetBytes(signKey));
var hash = mac.ComputeHash(Encoding.UTF8.GetBytes(secret));
signRet = Convert.ToBase64String(hash);
return signRet;
}
/// <summary>
/// Base64编码
/// </summary>
/// <remarks></remarks>
/// <param name="value"></param>
/// <returns></returns>
private static string ToBase64String(string value)
{
if (string.IsNullOrEmpty(value))
{
return "";
}
var bytes = Encoding.UTF8.GetBytes(value);
return Convert.ToBase64String(bytes);
}
#endregion(3)鉴权
#region 鉴权/URL
/// <summary>
/// 获取URL
/// </summary>
/// <param name="url"></param>
/// <returns></returns>
private string GetUrl(string url)
{
var uri = new Uri(url);
var
date = DateTime.UtcNow.ToString("r"); //官方文档要求时间必须是UTC+0或GMT时区，RFC1123格式(Thu, 01 Aug 2019 01:53:21 GMT)。
ComposeAuthUrl(uri, date);
var uriStr = string.Format("{0}?authorization={1}&date={2}&host={3}", uri, authorization,
ReplaceSpacesWithPercent20(date), uri.Host); //生成最终鉴权
Debug.Log(uriStr);
Debug.Log(date);
return uriStr;
}
/// <summary>
/// 组装生成鉴权
/// </summary>
private void ComposeAuthUrl(Uri uri, string date)
{
signature_origin =
string.Format("host: " + uri.Host + "\ndate: " + date + "\nGET " + uri.AbsolutePath + " HTTP/1.1");
//Debug.Log("signature_origin: " + signature_origin);
signature_sha = HmacSHA256(signature_origin, APISecret); //使用hmac - sha256算法结合apiSecret对signature_origin签名
//Debug.Log("signature_sha: " + signature_sha);
signature = signature_sha;
//Debug.Log("signature: " + signature);
string auth = "api_key=\"{0}\", algorithm=\"{1}\", headers=\"{2}\", signature=\"{3}\"";
authorization_origin = string.Format(auth, APIKey, "hmac-sha256", "host date request-line", signature);
//Debug.Log("authorization_origin: " + authorization_origin);
authorization = ToBase64String(authorization_origin);
//Debug.Log("authorization: " + authorization);
}
#endregion(4)WebSocket(BestHTTP)
#region Web
/// <summary>
/// 开启连接
/// </summary>
public void Content()
{
webSocket = new WebSocket(new Uri(GetUrl(url)));
webSocket.OnOpen += OnOpen;
webSocket.OnBinary += OnBinaryReceived;
webSocket.OnMessage += OnMessage;
webSocket.OnError += OnError;
webSocket.OnClosed += OnClosed;
webSocket.Open();
Debug.Log("讯飞WebSocket打开成功！");
}
private void OnOpen(WebSocket ws)
{
Debug.Log("WebSocket 连接成功！");
}
private void OnBinaryReceived(WebSocket ws, byte[] data)
{
}
private void OnMessage(WebSocket ws, string message)
{
try
{
JsonData js = JsonMapper.ToObject(message);
// 检查响应的状态
if (js["code"].ToString() != "0")
{
Debug.LogError("联网错误：" + js["message"].ToString());
return;
}
// 检查返回结果
if (js["message"].ToString() == "success" && js["data"] != null)
{
// 获取数据状态
var status = Convert.ToString(js["data"]["status"]);
if (status == "2" && js["data"]["data"] != null)
{
// 解码 Base64 数据
byte[] decodedBytes = Convert.FromBase64String(js["data"]["data"].ToString());
string xmlString = System.Text.Encoding.UTF8.GetString(decodedBytes);
ProcessXml(xmlString);
webSocket.Close();
}
else
{
Debug.LogWarning("返回的数据状态不符合预期，status: " + status);
}
}
else
{
Debug.LogError("错误：未能获取到有效的数据");
webSocket.Close();
}
}
catch (Exception ex)
{
Debug.LogError("处理消息时发生错误: " + ex.Message);
webSocket.Close();
}
}
private void ProcessXml(string xmlString)
{
try
{
// 加载 XML 字符串
XmlDocument xmlDoc = new XmlDocument();
xmlDoc.LoadXml(xmlString);
// 获取所有 <read_chapter> 标签的节点
XmlNodeList nodes = xmlDoc.GetElementsByTagName("read_chapter");
if (nodes.Count == 0)
{
Debug.Log("没有找到 'read_chapter' 标签节点。");
return;
}
var tempMessage = new Dictionary<string, string>();
foreach (XmlNode node in nodes)
{
if (node == null) continue;
// 处理节点内容
if (!string.IsNullOrEmpty(node.InnerText))
{
Debug.Log("Node InnerText: " + node.InnerText);
}
// 处理节点属性
if (node.Attributes != null)
{
foreach (XmlAttribute attribute in node.Attributes)
{
if (attribute != null)
{
Debug.Log("Attribute: " + attribute.Name + " = " + attribute.Value);
tempMessage[attribute.Name] = attribute.Value;
}
}
}
// 处理子节点
foreach (XmlNode childNode in node.ChildNodes)
{
if (childNode != null && !string.IsNullOrEmpty(childNode.InnerText))
{
Debug.Log("Child Node: " + childNode.Name + " = " + childNode.InnerText);
}
}
}
// 调用委托返回处理后的信息
messageSend.Invoke(tempMessage);
}
catch (Exception ex)
{
Debug.LogError("处理 XML 数据时发生错误: " + ex.Message);
}
}
private void OnError(WebSocket ws, string ex)
{
var sf = $"编码{ws.InternalRequest.Response.StatusCode},详细信息{ws.InternalRequest.Response.Message}";
Debug.Log("错误编码： " + sf);
}
private void OnClosed(WebSocket ws, ushort code, string message)
{
Debug.Log("WebSocket 连接关闭，关闭代码: " + code + ", 原因: " + message);
}
#endregion(5)分次发送
#region 发送
/// <summary>
/// 发送
/// </summary>
/// <param name="text"></param>
/// <param name="audioData"></param>
/// <returns></returns>
private IEnumerator XinghuoISE(string text, byte[] audioData)
{
yield return new WaitUntil(()=>webSocket.IsOpen);
// 剩余要发送的字节数
var remainLength = audioData.Length;
//第一次发送
var jsonData = new JsonData();
jsonData = CreateJson_1(text);
webSocket.Send(JsonMapper.ToJson(jsonData));
// 分段发送消息，直到所有数据都发送完成
while (remainLength > 0)
{
// 当前要发送的字节数组
byte[] currBytes;
var status = -1;
var aus = -1;
//非结尾帧
if (remainLength > 1280)
{
//是否为第一帧
status = remainLength == audioData.Length ? 0 : 1;
aus = remainLength == audioData.Length ? 1 : 2;
currBytes = new byte[1280];
Array.Copy(audioData, audioData.Length - remainLength, currBytes, 0, 1280);
remainLength -= 1280;
}
//结尾帧
else
{
status = 2;
aus = 4;
currBytes = new byte[remainLength];
Array.Copy(audioData, audioData.Length - remainLength, currBytes, 0, remainLength);
remainLength = 0;
}
//发送
jsonData = CreateJson_2(aus,status,currBytes);
webSocket.Send(JsonMapper.ToJson(jsonData));
yield return null;
}
Debug.Log("数据发送完成");
yield break;
}
/// <summary>
/// 第一次发送
/// </summary>
/// <returns></returns>
private JsonData CreateJson_1(string text)
{
JsonData requestObj = new JsonData();
// common 部分
JsonData commonJson = new JsonData
{
["app_id"] = APPID
};
requestObj["common"] = commonJson;
JsonData businessJson = new JsonData()
{
["sub"] = "ise",
["ent"] = "en_vip",
["category"] = "read_chapter",
["cmd"] = "ssb",
["text"] = '\uFEFF' + "[content]" + text,
["tte"] = "utf-8",
["ttp_skip"] = true,
["aue"] = "raw"
};
requestObj["business"] = businessJson;
JsonData dataJson = new JsonData()
{
["status"] = 0
};
requestObj["data"] = dataJson;
return requestObj;
}
/// <summary>
/// 第二次发送(音频发送阶段)
/// </summary>
/// <param name="aus"></param>
/// <param name="status"></param>
/// <param name="value"></param>
/// <returns></returns>
private JsonData CreateJson_2(int aus, int status,byte[] value)
{
JsonData requestObj = new JsonData();
JsonData businessJson = new JsonData()
{
["cmd"] = "auw",
["aus"] = aus,
};
requestObj["business"] = businessJson;
JsonData dataJson = new JsonData()
{
["status"]=status,
["data"] = Convert.ToBase64String(value)
};
requestObj["data"] = dataJson;
return requestObj;
}
#endregion(6)音频处理(AudioClip -> byte[])
#region 音频处理
/// <summary>
/// 处理音频
/// </summary>
/// <param name="audio">音频</param>
/// <returns></returns>
private static byte[] DealAudio(AudioClip audio)
{
var recordedAudioClip = audio;
//剔除沉默音域
recordedAudioClip = TrimSilence(recordedAudioClip, 0.01f);
//bytes
var bytes = AudioClipToBytes(recordedAudioClip);
return bytes;
}
/// <summary>
/// 将AudioClip转换成byte[]数据
/// </summary>
/// <param name="audioClip">Unity中的音频数据</param>
/// <returns>byte[]数据</returns>
private static byte[] AudioClipToBytes(AudioClip audioClip)
{
var data = new float[audioClip.samples];
audioClip.GetData(data, 0);
var rescaleFactor = 32767; //to convert float to Int16
var outData = new byte[data.Length * 2];
for (var i = 0; i < data.Length; i++)
{
var temshort = (short)(data[i] * rescaleFactor);
var temdata = BitConverter.GetBytes(temshort);
outData[i * 2] = temdata[0];
outData[i * 2 + 1] = temdata[1];
}
return outData;
}
/// <summary>
/// 剔除沉默音域
/// </summary>
/// <param name="clip"></param>
/// <param name="min"></param>
/// <returns></returns>
private static AudioClip TrimSilence(AudioClip clip, float min)
{
var samples = new float[clip.samples];
clip.GetData(samples, 0);
return TrimSilence(new List<float>(samples), min, clip.channels, clip.frequency);
}
private static AudioClip TrimSilence(List<float> samples, float min, int channels, int hz, bool _3D = false)
{
var origSamples = samples.Count;
int i;
for (i = 0; i < samples.Count; i++)
{
if (Mathf.Abs(samples[i]) > min)
{
break;
}
}
i -= (int)(hz * .1f);
i = Mathf.Max(i, 0);
// Remove start silence
samples.RemoveRange(0, i);
for (i = samples.Count - 1; i > 0; i--)
{
if (Mathf.Abs(samples[i]) > min)
{
break;
}
}
// Add some tail onto it
i += (int)(hz * .1f);
i = Mathf.Min(i, samples.Count - 1);
samples.RemoveRange(i, samples.Count - i);
if (samples.Count == 0)
{
Debug.Log("剔除后的AudioClip长度为0");
return null;
}
var clip = AudioClip.Create("TempClip", samples.Count, channels, hz, _3D);
clip.SetData(samples.ToArray(), 0);
return clip;
}
#endregion(7)使用
/// <summary>
/// 使用ISE
/// </summary>
/// <param name="text"></param>
/// <param name="audio"></param>
public void XingHuo_ISE_SA(string text, AudioClip audio)
{
//打开
Content();
//转二进制
var _bytes = DealAudio(audio);
//发送
StartCoroutine(XinghuoISE(text, _bytes));
}(8)XingHuo_ISE 完整脚本
using System;
using System.Collections;
using System.Collections.Generic;
using System.Security.Cryptography;
using System.Text;
using UnityEngine;
using System.Xml;
using WebSocket = BestHTTP.WebSocket.WebSocket;
using LitJson;
using UnityEngine.Events;
namespace ISE
{
public class XingHuo_ISE : MonoBehaviour
{
//signature
private string signature_origin = null;
private string signature_sha = null;
private string signature = null;
//author
private string authorization_origin = null;
private string authorization = null;
//url
private string url = "wss://ise-api.xfyun.cn/v2/open-ise";
//WebSocket实例
public WebSocket webSocket;
//Websocket服务接口认证信息
public string APPID = "91b4059e";
public string APISecret = "MGNjOWY4YzgyNjAxNGFkNWIyY2IxMjUw";
public string APIKey = "13f5b50899e2adc74fb939502616bc02";
//接收
public UnityEvent<Dictionary<string, string>> messageSend = new();
/// <summary>
/// 使用ISE
/// </summary>
/// <param name="text"></param>
/// <param name="audio"></param>
public void XingHuo_ISE_SA(string text, AudioClip audio)
{
//打开
Content();
//转二进制
var _bytes = DealAudio(audio);
//发送
StartCoroutine(XinghuoISE(text, _bytes));
}
#region 发送
/// <summary>
/// 发送
/// </summary>
/// <param name="text"></param>
/// <param name="audioData"></param>
/// <returns></returns>
private IEnumerator XinghuoISE(string text, byte[] audioData)
{
yield return new WaitUntil(()=>webSocket.IsOpen);
// 剩余要发送的字节数
var remainLength = audioData.Length;
//第一次发送
var jsonData = new JsonData();
jsonData = CreateJson_1(text);
webSocket.Send(JsonMapper.ToJson(jsonData));
// 分段发送消息，直到所有数据都发送完成
while (remainLength > 0)
{
// 当前要发送的字节数组
byte[] currBytes;
var status = -1;
var aus = -1;
//非结尾帧
if (remainLength > 1280)
{
//是否为第一帧
status = remainLength == audioData.Length ? 0 : 1;
aus = remainLength == audioData.Length ? 1 : 2;
currBytes = new byte[1280];
Array.Copy(audioData, audioData.Length - remainLength, currBytes, 0, 1280);
remainLength -= 1280;
}
//结尾帧
else
{
status = 2;
aus = 4;
currBytes = new byte[remainLength];
Array.Copy(audioData, audioData.Length - remainLength, currBytes, 0, remainLength);
remainLength = 0;
}
//发送
jsonData = CreateJson_2(aus,status,currBytes);
webSocket.Send(JsonMapper.ToJson(jsonData));
yield return null;
}
Debug.Log("数据发送完成");
yield break;
}
/// <summary>
/// 第一次发送
/// </summary>
/// <returns></returns>
private JsonData CreateJson_1(string text)
{
JsonData requestObj = new JsonData();
// common 部分
JsonData commonJson = new JsonData
{
["app_id"] = APPID
};
requestObj["common"] = commonJson;
JsonData businessJson = new JsonData()
{
["sub"] = "ise",
["ent"] = "en_vip",
["category"] = "read_chapter",
["cmd"] = "ssb",
["text"] = '\uFEFF' + "[content]" + text,
["tte"] = "utf-8",
["ttp_skip"] = true,
["aue"] = "raw"
};
requestObj["business"] = businessJson;
JsonData dataJson = new JsonData()
{
["status"] = 0
};
requestObj["data"] = dataJson;
return requestObj;
}
/// <summary>
/// 第二次发送(音频发送阶段)
/// </summary>
/// <param name="aus"></param>
/// <param name="status"></param>
/// <param name="value"></param>
/// <returns></returns>
private JsonData CreateJson_2(int aus, int status,byte[] value)
{
JsonData requestObj = new JsonData();
JsonData businessJson = new JsonData()
{
["cmd"] = "auw",
["aus"] = aus,
};
requestObj["business"] = businessJson;
JsonData dataJson = new JsonData()
{
["status"]=status,
["data"] = Convert.ToBase64String(value)
};
requestObj["data"] = dataJson;
return requestObj;
}
#endregion
#region 音频处理
/// <summary>
/// 处理音频
/// </summary>
/// <param name="audio">音频</param>
/// <returns></returns>
private static byte[] DealAudio(AudioClip audio)
{
var recordedAudioClip = audio;
//剔除沉默音域
recordedAudioClip = TrimSilence(recordedAudioClip, 0.01f);
//bytes
var bytes = AudioClipToBytes(recordedAudioClip);
return bytes;
}
/// <summary>
/// 将AudioClip转换成byte[]数据
/// </summary>
/// <param name="audioClip">Unity中的音频数据</param>
/// <returns>byte[]数据</returns>
private static byte[] AudioClipToBytes(AudioClip audioClip)
{
var data = new float[audioClip.samples];
audioClip.GetData(data, 0);
var rescaleFactor = 32767; //to convert float to Int16
var outData = new byte[data.Length * 2];
for (var i = 0; i < data.Length; i++)
{
var temshort = (short)(data[i] * rescaleFactor);
var temdata = BitConverter.GetBytes(temshort);
outData[i * 2] = temdata[0];
outData[i * 2 + 1] = temdata[1];
}
return outData;
}
/// <summary>
/// 剔除沉默音域
/// </summary>
/// <param name="clip"></param>
/// <param name="min"></param>
/// <returns></returns>
private static AudioClip TrimSilence(AudioClip clip, float min)
{
var samples = new float[clip.samples];
clip.GetData(samples, 0);
return TrimSilence(new List<float>(samples), min, clip.channels, clip.frequency);
}
private static AudioClip TrimSilence(List<float> samples, float min, int channels, int hz, bool _3D = false)
{
var origSamples = samples.Count;
int i;
for (i = 0; i < samples.Count; i++)
{
if (Mathf.Abs(samples[i]) > min)
{
break;
}
}
i -= (int)(hz * .1f);
i = Mathf.Max(i, 0);
// Remove start silence
samples.RemoveRange(0, i);
for (i = samples.Count - 1; i > 0; i--)
{
if (Mathf.Abs(samples[i]) > min)
{
break;
}
}
// Add some tail onto it
i += (int)(hz * .1f);
i = Mathf.Min(i, samples.Count - 1);
samples.RemoveRange(i, samples.Count - i);
if (samples.Count == 0)
{
Debug.Log("剔除后的AudioClip长度为0");
return null;
}
var clip = AudioClip.Create("TempClip", samples.Count, channels, hz, _3D);
clip.SetData(samples.ToArray(), 0);
return clip;
}
#endregion
#region Web
/// <summary>
/// 开启连接
/// </summary>
public void Content()
{
webSocket = new WebSocket(new Uri(GetUrl(url)));
webSocket.OnOpen += OnOpen;
webSocket.OnBinary += OnBinaryReceived;
webSocket.OnMessage += OnMessage;
webSocket.OnError += OnError;
webSocket.OnClosed += OnClosed;
webSocket.Open();
Debug.Log("讯飞WebSocket打开成功！");
}
private void OnOpen(WebSocket ws)
{
Debug.Log("WebSocket 连接成功！");
}
private void OnBinaryReceived(WebSocket ws, byte[] data)
{
}
private void OnMessage(WebSocket ws, string message)
{
try
{
JsonData js = JsonMapper.ToObject(message);
// 检查响应的状态
if (js["code"].ToString() != "0")
{
Debug.LogError("联网错误：" + js["message"].ToString());
return;
}
// 检查返回结果
if (js["message"].ToString() == "success" && js["data"] != null)
{
// 获取数据状态
var status = Convert.ToString(js["data"]["status"]);
if (status == "2" && js["data"]["data"] != null)
{
// 解码 Base64 数据
byte[] decodedBytes = Convert.FromBase64String(js["data"]["data"].ToString());
string xmlString = System.Text.Encoding.UTF8.GetString(decodedBytes);
ProcessXml(xmlString);
webSocket.Close();
}
else
{
Debug.LogWarning("返回的数据状态不符合预期，status: " + status);
}
}
else
{
Debug.LogError("错误：未能获取到有效的数据");
webSocket.Close();
}
}
catch (Exception ex)
{
Debug.LogError("处理消息时发生错误: " + ex.Message);
webSocket.Close();
}
}
private void ProcessXml(string xmlString)
{
try
{
// 加载 XML 字符串
XmlDocument xmlDoc = new XmlDocument();
xmlDoc.LoadXml(xmlString);
// 获取所有 <read_chapter> 标签的节点
XmlNodeList nodes = xmlDoc.GetElementsByTagName("read_chapter");
if (nodes.Count == 0)
{
Debug.Log("没有找到 'read_chapter' 标签节点。");
return;
}
var tempMessage = new Dictionary<string, string>();
foreach (XmlNode node in nodes)
{
if (node == null) continue;
// 处理节点内容
if (!string.IsNullOrEmpty(node.InnerText))
{
Debug.Log("Node InnerText: " + node.InnerText);
}
// 处理节点属性
if (node.Attributes != null)
{
foreach (XmlAttribute attribute in node.Attributes)
{
if (attribute != null)
{
Debug.Log("Attribute: " + attribute.Name + " = " + attribute.Value);
tempMessage[attribute.Name] = attribute.Value;
}
}
}
// 处理子节点
foreach (XmlNode childNode in node.ChildNodes)
{
if (childNode != null && !string.IsNullOrEmpty(childNode.InnerText))
{
Debug.Log("Child Node: " + childNode.Name + " = " + childNode.InnerText);
}
}
}
// 调用委托返回处理后的信息
messageSend.Invoke(tempMessage);
}
catch (Exception ex)
{
Debug.LogError("处理 XML 数据时发生错误: " + ex.Message);
}
}
private void OnError(WebSocket ws, string ex)
{
var sf = $"编码{ws.InternalRequest.Response.StatusCode},详细信息{ws.InternalRequest.Response.Message}";
Debug.Log("错误编码： " + sf);
}
private void OnClosed(WebSocket ws, ushort code, string message)
{
Debug.Log("WebSocket 连接关闭，关闭代码: " + code + ", 原因: " + message);
}
#endregion
#region 鉴权/URL
/// <summary>
/// 获取URL
/// </summary>
/// <param name="url"></param>
/// <returns></returns>
private string GetUrl(string url)
{
var uri = new Uri(url);
var
date = DateTime.UtcNow.ToString("r"); //官方文档要求时间必须是UTC+0或GMT时区，RFC1123格式(Thu, 01 Aug 2019 01:53:21 GMT)。
ComposeAuthUrl(uri, date);
var uriStr = string.Format("{0}?authorization={1}&date={2}&host={3}", uri, authorization,
ReplaceSpacesWithPercent20(date), uri.Host); //生成最终鉴权
Debug.Log(uriStr);
Debug.Log(date);
return uriStr;
}
/// <summary>
/// 组装生成鉴权
/// </summary>
private void ComposeAuthUrl(Uri uri, string date)
{
signature_origin =
string.Format("host: " + uri.Host + "\ndate: " + date + "\nGET " + uri.AbsolutePath + " HTTP/1.1");
//Debug.Log("signature_origin: " + signature_origin);
signature_sha = HmacSHA256(signature_origin, APISecret); //使用hmac - sha256算法结合apiSecret对signature_origin签名
//Debug.Log("signature_sha: " + signature_sha);
signature = signature_sha;
//Debug.Log("signature: " + signature);
string auth = "api_key=\"{0}\", algorithm=\"{1}\", headers=\"{2}\", signature=\"{3}\"";
authorization_origin = string.Format(auth, APIKey, "hmac-sha256", "host date request-line", signature);
//Debug.Log("authorization_origin: " + authorization_origin);
authorization = ToBase64String(authorization_origin);
//Debug.Log("authorization: " + authorization);
}
#endregion
#region 编码
/// <summary>
///空格替换为 %20
/// </summary>
/// <param name="input"></param>
/// <returns></returns>
private static string ReplaceSpacesWithPercent20(string input)
{
return string.IsNullOrEmpty(input) ? input : input.Replace(" ", "%20");
}
/// <summary>
/// HmacSHA256
/// </summary>
/// <param name="secret">A</param>
/// <param name="signKey">B</param>
/// <returns></returns>
private static string HmacSHA256(string secret, string signKey)
{
var signRet = string.Empty;
using HMACSHA256 mac = new HMACSHA256(Encoding.UTF8.GetBytes(signKey));
var hash = mac.ComputeHash(Encoding.UTF8.GetBytes(secret));
signRet = Convert.ToBase64String(hash);
return signRet;
}
/// <summary>
/// Base64编码
/// </summary>
/// <remarks></remarks>
/// <param name="value"></param>
/// <returns></returns>
private static string ToBase64String(string value)
{
if (string.IsNullOrEmpty(value))
{
return "";
}
var bytes = Encoding.UTF8.GetBytes(value);
return Convert.ToBase64String(bytes);
}
#endregion
}
}(9)代码详解
该脚本主要封装了科大讯飞在线口语评测的完整接入流程：
XingHuo_ISE_SA 方法接收文本与 AudioClip，并将音频交由 DealAudio 处理；
DealAudio 依次调用 TrimSilence（及其重载）剔除前后静默段，再通过 AudioClipToBytes 将浮点 PCM 转为 16 位小端字节；
XinghuoISE 协程先用 CreateJson_1 构建首帧 JSON（包含 common、business 与 data.status=0），通过 WebSocket 发送文本及评测参数，然后循环分 1280 字节调用 CreateJson_2 拼装音频分片 JSON（status 与 aus 表示帧类型）并依次推送；
Content 方法负责调用 GetUrl（内部用 ComposeAuthUrl、ReplaceSpacesWithPercent20、HmacSHA256 和 ToBase64String 构造带签名的鉴权 URL）打开 BestHTTP 的 WebSocket，并注册 OnOpen、OnBinaryReceived、OnMessage、OnError 与 OnClosed 回调；
其中 OnMessage 检查 code 与 status、解 Base64 得到 XML 字符串并交给 ProcessXml 解析 <read_chapter> 节点属性和子节点内容；
最终通过 UnityEvent<Dictionary<string,string>> messageSend 将评测结果发回给上层。
3.外部调用逻辑示例
(1)参数
//讯飞ISE
public XingHuo_ISE XingHuoIse;
//评分显示
public Text textAsset_1;
public Text textAsset_2;
public Text textAsset_3;
public Text textAsset_4;
public Text textAsset_5;
//麦克风选择
public Dropdown microphoneDropdown;
//选择的麦克风设备
private string defaultMicrophone;
//待评测文本
public InputField inputField;
//启动录音
public Button Button_AMic;
//停止录音
public Button Button_SMic;
//启动ISE
public Button Button_ISE;
//缓存录音音频
public AudioClip audioClip_1;(2)通过下拉菜单选择麦克风(UI-Dropdown)
/// <summary>
/// 初始化下拉菜单
/// </summary>
public void UpdateMicrophoneDropdown()
{
microphoneDropdown.ClearOptions();
string[] devices = Microphone.devices;
List<string> options = new List<string>();
foreach (string device in devices)
{
options.Add(device);
}
microphoneDropdown.AddOptions(options);
if (devices.Length > 0)
{
microphoneDropdown.value = 0;
defaultMicrophone = devices[0]; // 初始默认值
}
else
{
defaultMicrophone = "No Microphone Found";
}
microphoneDropdown.RefreshShownValue();
}
/// <summary>
/// 当下拉菜单选项变化时调用
/// </summary>
/// <param name="selectedIndex"></param>
private void OnMicrophoneDropdownChanged(int selectedIndex)
{
string[] devices = Microphone.devices;
if (devices.Length > selectedIndex)
{
defaultMicrophone = devices[selectedIndex]; // 更新当前选中的麦克风
Debug.Log("麦克风已切换至: " + defaultMicrophone);
}
}(3)按钮监听
/// <summary>
/// 开始录音
/// </summary>
public void OnClick_UseMic()
{
audioClip_1 = Microphone.Start(defaultMicrophone, true, 40, 16000);
}
/// <summary>
/// 停止录音
/// </summary>
public void OnClick_StopMic()
{
Microphone.End(defaultMicrophone);
}
/// <summary>
/// 开始评分
/// </summary>
public void OnClick_UseISE()
{
XingHuoIse.XingHuo_ISE_SA(inputField.text, audioClip_1);
}(4)返回评分监听
/// <summary>
/// 评分回调
/// </summary>
/// <param name="parameters"></param>
public void InterpretationParameters(Dictionary<string,string> parameters)
{
textAsset_1.text ="准确度："+ parameters["accuracy_score"];
textAsset_2.text ="标准度："+ parameters["standard_score"];
textAsset_3.text ="流利度："+ parameters["fluency_score"];
textAsset_4.text ="完整度："+ parameters["integrity_score"];
textAsset_5.text ="总分: "+ parameters["total_score"];
}(5)下拉菜单初始化
private void Start()
{
// 初始化下拉菜单
UpdateMicrophoneDropdown();
// 监听下拉菜单的变化
microphoneDropdown.onValueChanged.AddListener(OnMicrophoneDropdownChanged);
}(6)ISE_Test 完整脚本
using System;
using System.Collections;
using System.Collections.Generic;
using ISE;
using UnityEngine;
using UnityEngine.UI;
public class ISE_Test : MonoBehaviour
{
//讯飞ISE
public XingHuo_ISE XingHuoIse;
//评分显示
public Text textAsset_1;
public Text textAsset_2;
public Text textAsset_3;
public Text textAsset_4;
public Text textAsset_5;
//麦克风选择
public Dropdown microphoneDropdown;
//选择的麦克风设备
private string defaultMicrophone;
//待评测文本
public InputField inputField;
//启动录音
public Button Button_AMic;
//停止录音
public Button Button_SMic;
//启动ISE
public Button Button_ISE;
//缓存录音音频
public AudioClip audioClip_1;
private void Start()
{
// 初始化下拉菜单
UpdateMicrophoneDropdown();
// 监听下拉菜单的变化
microphoneDropdown.onValueChanged.AddListener(OnMicrophoneDropdownChanged);
}
/// <summary>
/// 初始化下拉菜单
/// </summary>
public void UpdateMicrophoneDropdown()
{
microphoneDropdown.ClearOptions();
string[] devices = Microphone.devices;
List<string> options = new List<string>();
foreach (string device in devices)
{
options.Add(device);
}
microphoneDropdown.AddOptions(options);
if (devices.Length > 0)
{
microphoneDropdown.value = 0;
defaultMicrophone = devices[0]; // 初始默认值
}
else
{
defaultMicrophone = "No Microphone Found";
}
microphoneDropdown.RefreshShownValue();
}
/// <summary>
/// 当下拉菜单选项变化时调用
/// </summary>
/// <param name="selectedIndex"></param>
private void OnMicrophoneDropdownChanged(int selectedIndex)
{
string[] devices = Microphone.devices;
if (devices.Length > selectedIndex)
{
defaultMicrophone = devices[selectedIndex]; // 更新当前选中的麦克风
Debug.Log("麦克风已切换至: " + defaultMicrophone);
}
}
/// <summary>
/// 开始录音
/// </summary>
public void OnClick_UseMic()
{
audioClip_1 = Microphone.Start(defaultMicrophone, true, 40, 16000);
}
/// <summary>
/// 停止录音
/// </summary>
public void OnClick_StopMic()
{
Microphone.End(defaultMicrophone);
}
/// <summary>
/// 开始评分
/// </summary>
public void OnClick_UseISE()
{
XingHuoIse.XingHuo_ISE_SA(inputField.text, audioClip_1);
}
/// <summary>
/// 评分回调
/// </summary>
/// <param name="parameters"></param>
public void InterpretationParameters(Dictionary<string,string> parameters)
{
textAsset_1.text ="准确度："+ parameters["accuracy_score"];
textAsset_2.text ="标准度："+ parameters["standard_score"];
textAsset_3.text ="流利度："+ parameters["fluency_score"];
textAsset_4.text ="完整度："+ parameters["integrity_score"];
textAsset_5.text ="总分: "+ parameters["total_score"];
}
}
(7)编辑器中使用
1)Project Setting中设置HTTP请求
2)添加脚本
创建空物体，命名为ISEManager。
在ISEManager上添加XingHuo_ISE脚本与ISE_Test脚本，并在XingHuo_ISE脚本中填写对应的服务接口认证信息。
3)创建并挂载对应的UI元素
创建
挂载
4)返回评分监听
5)为三个按钮分别注册监听
6)运行测试
(8)代码详解
该脚本首先通过 Microphone.devices 获取系统中所有可用的麦克风设备并将其填充到 Dropdown 列表中以供用户选择；
然后通过调用 Microphone.Start(defaultMicrophone, true, 40, 16000) 开始录制 16 kHz 的音频并在需要时用 Microphone.End(defaultMicrophone) 停止录音；
接着，用户在 InputField 中输入待测文本，点击评测按钮时将文本和录音得到的 AudioClip 一并传给 XingHuo_ISE_SA 方法发起在线口语评测；
最后在 InterpretationParameters 回调中分别将准确度、标准度、流利度、完整度和总分显示到对应的 Text 组件上。
四、后记&下载地址
在本篇 Unity 技术博客即将付梓之际，作为一名刚入行的新人，我既怀揣着与大家分享所学所感的激动，也难免对自己文字与技术深度的不足心存忐忑。然而技法尚嫌浅薄，阐述不周或示例不够完善之处在所难免，代码编写之中有失偏颇之处，还望大佬们不吝赐教和批评。
BestHTTP2_v2.0.5下载地址(转载)：GitCode - 全球开发者的开源社区,开源代码托管平台
技术共进，成长同行——讯飞AI开发者社区
更多推荐
27
0- 0
已为社区贡献1条内容
所有评论(0)