C# develops WeChat portals and applications using voice processing

高洛峰
Release: 2017-03-02 10:04:02
Original
1521 people have browsed it

We know that WeChat was originally designed for voice chat, which made it more popular. Therefore, voice recognition processing has naturally become an important way for WeChat communication. WeChat’s development interface also provides processing of voice message requests. . This article mainly introduces how to use speech recognition to process the entire event chain of the WeChat portal application developed in C#, making it more convenient and diversified to process user input in our WeChat account.

1. Definition of WeChat voice interface 0

WeChat’s API defines voice recognition in this way: When the voice recognition function is enabled, every time the user sends a voice to the official account, WeChat will In the pushed voice message XML packet, add a Recongnition field .

The voice message format is as follows.

<xml>
<ToUserName><![CDATA[toUser]]></ToUserName>
<FromUserName><![CDATA[fromUser]]></FromUserName>
<CreateTime>1357290913</CreateTime>
<MsgType><![CDATA[voice]]></MsgType>
<MediaId><![CDATA[media_id]]></MediaId>
<Format><![CDATA[Format]]></Format>
<MsgId>1234567890123456</MsgId>
</xml>
Copy after login

ParametersDescription
ToUserNameDevelopment人微信
FromUserNameSender account (an OpenID)
CreateTimeMessage creation time (integer)
MsgTypeVoice is voice
MediaIdVoice Message media id, you can call the multimedia file download interface to pull data.
FormatVoice format, such as amr, speex, etc.
MsgIDMessage id, 64 Bit integer

根据以上微信接口的定义,我们可以定义一个实体类来对消息的传递进行处理,如下所示。

/// <summary>
    /// 接收的语音消息
    /// </summary>
    [System.Xml.Serialization.XmlRoot(ElementName = "xml")]
    public class RequestVoice : BaseMessage
    {
        public RequestVoice()
        {
            this.MsgType = RequestMsgType.Voice.ToString().ToLower();
        }

        /// <summary>
        /// 语音格式,如amr,speex等
        /// </summary>
        public string Format { get; set; }

        /// <summary>
        /// 语音消息媒体id,可以调用多媒体文件下载接口拉取数据。
        /// </summary>
        public string MediaId { get; set; }
  
        /// <summary>
        /// 消息ID
        /// </summary>
        public Int64 MsgId { get; set; }

        /// <summary>
        /// 语音识别结果,UTF8编码
        /// </summary>
        public string Recognition  { get; set; }

    }
Copy after login

我们看到,这里我们最感兴趣的是语音的识别结果,也就是Recognition的字段,这个就是微信服务器自动根据用户的语音转换过来的内容,我测试过,识别率还是非常高的。

这个实体类,在整个微信应用的消息传递中的关系如下所示:

C# develops WeChat portals and applications using voice processing

2、语音的处理操作

明确了上面的语音对象实体,我们就可以看看它们之间是如何处理的。

微信消息的处理逻辑如下图所示。

C# develops WeChat portals and applications using voice processing

其中我们来看看语音的处理操作,我的代码处理逻辑如下所示。

/// <summary>
        /// 对语音请求信息进行处理
        /// </summary>
        /// <param name="info">语音请求信息实体</param>
        /// <returns></returns>
        public string HandleVoice(Entity.RequestVoice info)
        {
            string xml = "";
            // 开通语音识别功能,用户每次发送语音给公众号时,
            // 微信会在推送的语音消息XML数据包中,增加一个Recongnition字段。
            if (!string.IsNullOrEmpty(info.Recognition))
            {
                TextDispatch dispatch = new TextDispatch();
                xml = dispatch.HandleVoiceText(info, info.Recognition);
            }
            else
            {
                xml = "";
            }

            return xml;
        }
Copy after login

在这里,我先看看,是否获得了微信的语音识别结果,如果获得,那么这个时候,就是和处理用户文本输入的操作差不多了,因此把它转给TextDispatch的处理类进行处理。

其中这里面的处理逻辑如下所示。

C# develops WeChat portals and applications using voice processing

首先我根据识别结果,寻找是否用户读出了微信门户的菜单名称,如果根据语音结果找到对应的菜单记录,那么我们执行菜单事件(如果是URL的View类型菜单,我们没办法重定向到指定的链接,因此给出一个链接文本提示,给用户单击进入;如果没有找到菜单记录,那么我们就把语音识别结果作为一般的事件进行处理,如果事件逻辑没有处理,那么我们最后给出一个默认的语音应答提示结果就可以了。

具体的处理代码如下所示。

/// <summary>
        /// 如果用户用语音读出菜单的内容,那么我们应该先根据菜单对应的事件触发,最后再交给普通事件处理
        /// </summary>
        /// <param name="info"></param>
        /// <returns></returns>
        public string HandleVoiceText(BaseMessage info, string voiceText)
        {
            string xml = "";
            MenuInfo menuInfo = BLLFactory<Menu>.Instance.FindByName(voiceText);
            if (menuInfo != null)
            {
                #region 如果找到菜单对象的处理
                if (menuInfo.Type == "click")
                {
                    //模拟单击事件
                    RequestEventClick eventInfo = new RequestEventClick();
                    eventInfo.CreateTime = info.CreateTime;
                    eventInfo.EventKey = menuInfo.Key;
                    eventInfo.FromUserName = info.FromUserName;
                    eventInfo.ToUserName = info.ToUserName;

                    xml = base.DealEvent(eventInfo, eventInfo.EventKey);
                }
                else
                {
                    //由于无法自动切换到连接,
                    //转换为连接文本供用户进入
                    string content = string.Format("请单击链接进入<a href=\"{0}\">{1}</a> ", menuInfo.Url, menuInfo.Name);

                    ResponseText textInfo = new ResponseText(info);
                    textInfo.Content = content;

                    xml = textInfo.ToXml();
                } 
                #endregion
            }
            else
            {
                //交给事件机制处理
                if (string.IsNullOrEmpty(xml))
                {
                    xml = HandleText(info, voiceText);
                }
            }

            //最后如果没有处理到,那么提示用户的语音内容
            if (string.IsNullOrEmpty(xml))
            {
                ResponseText textInfo = new ResponseText(info);
                textInfo.Content = string.Format("非常抱歉,您输入的语音内容没有找到对应的处理方式。您的语音内容为:{0}", voiceText);
                xml = textInfo.ToXml();
            }

            return xml;
        }
Copy after login

微信门户测试界面效果如下所示。

C# develops WeChat portals and applications using voice processing                 C# develops WeChat portals and applications using voice processing   

 C# develops WeChat portals and applications using voice processing

为了方便对客户会话的记录,我的微信门户后台,会记录用户的语音输入内容,如下所示。

C# develops WeChat portals and applications using voice processing

 当然,微信后台的管理界面,也能够查到相应的语音记录,界面如下所示。

C# develops WeChat portals and applications using voice processing

以上就是我对微信语音的消息定义和事件处理的逻辑,其实语音是一个重要的输入,如果正确的识别内容,比手工输入的效果更好,给用户提供另外一种高效的输入和事件处理操作。

这样的处理模式,能够使得我们整个微信门户框架,不管是对于用户的语音输入,还是文本输入,还是菜单事件的处理,都可以融为一体,实现更加完美的衔接。

更多C# develops WeChat portals and applications using voice processing相关文章请关注PHP中文网!

Related labels:
source:php.cn
Statement of this Website
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn
Popular Tutorials
More>
Latest Downloads
More>
Web Effects
Website Source Code
Website Materials
Front End Template