How to use web api to develop WeChat public account to call Turing robot interface

Passive response message (returning XML)

WeChat requires us to return XML data, and the format is specified. For details, please see

WeChat public platform developer documentation.

We have already written the response entity class before, because the requirement is in XML format.
We use the System.Xml.Serialization.XmlSerializer provided by Microsoft here to serialize our data into XML.
So we marked the XmlRoot attribute on the class, the XmlEnum attribute on the enumerated field, and the XmlArray and XmlArrayItem attributes on the article list in NewsMsg. What is then deserialized is the format required by WeChat.

The serialization method is as follows:

public string ResponseXML(object value, Type type){
    StringWriter sw = new StringWriter();
    XmlSerializerNamespaces ns = new XmlSerializerNamespaces();

    ns.Add("", "");　　//去除命名空间
    XmlSerializer serializer = new XmlSerializer(type);

    serializer.Serialize(sw, value, ns);    return sw.ToString();
}

Note: The XML namespace must be removed here, otherwise WeChat will not recognize it.

The complete method is provided:

public HttpResponseMessage Post(){    var requestContent = Request.Content.ReadAsStreamAsync().Result;    //从正文参数中加载微信的请求参数
    XmlDocument xmlDoc = new XmlDocument();
    xmlDoc.Load(requestContent);

    logger.DebugFormat("WX请求XML内容:{0}", xmlDoc.InnerText);    string msgTypeStr = xmlDoc.SelectSingleNode("xml/MsgType").InnerText;    string userName = xmlDoc.SelectSingleNode("xml/FromUserName").InnerText;    string efhName = xmlDoc.SelectSingleNode("xml/ToUserName").InnerText;    string responseContent;
    MsgType msgType;
    //获取消息类型，若未定义，则返回。
    if (!Enum.TryParse(msgTypeStr, true, out msgType))
    {
        responseContent = MsgService.Instance.ResponseXML(new TextMsg
            {
                FromUserName = efhName,
                MsgType = MsgType.Text,
                Content = "俺还小，不知道你在说啥子(⊙_⊙)?",
                CreateTime = UnixTimestamp.Now.ToNumeric(),
                ToUserName = userName
            }, typeof(TextMsg));        return new HttpResponseMessage(HttpStatusCode.OK)
        {
            Content = new StringContent(responseContent, Encoding.UTF8, "application/xml"),
        };
    }    if (msgType == MsgType.Event)
    {        return ProcessEvent(xmlDoc, userName, efhName);
    }
    //图灵消息转换为微信响应消息，下一节奉上
    string content = xmlDoc.SelectSingleNode("xml/Content").InnerText;    var requestResult = TuLingService.Instance.GetMsgFromResponse(content, userName, efhName);

    responseContent = MsgService.Instance.ResponseXML(requestResult.Data, requestResult.DataType);    return new HttpResponseMessage(HttpStatusCode.OK)
    {
        Content = new StringContent(responseContent, Encoding.UTF8, "application/xml"),
    };
}private HttpResponseMessage ProcessEvent(XmlDocument xmlDoc, string userName, string efhName){    string eventValue = xmlDoc.SelectSingleNode("xml/Event").InnerText;    var responseContent = MsgService.Instance.ResponseXML(new TextMsg
    {
        FromUserName = efhName,
        MsgType = MsgType.Text,
        Content = eventValue.ToLower().Equals("subscribe") ? "lei好哇~" : "大爷，奴家会想你的",//其实取消订阅是不会发送消息的
        CreateTime = UnixTimestamp.Now.ToNumeric(),
        ToUserName = userName
    }, typeof(TextMsg));    return new HttpResponseMessage(HttpStatusCode.OK)
    {
        Content = new StringContent(responseContent, Encoding.UTF8, "application/xml"),
    };
}

So far, we have completed the response to WeChat passive reply messages.

Mapping Turing messages and WeChat messages

We have already implemented the function of passively replying to messages. Next, we need to associate the Turing robot interface with our public platform.

Analyzing the parameters returned by the Turing robot, we found that all types of content have code and text parameters. And because we need to directly correspond Turing's message to WeChat's response message, we define an interface and provide a conversion method

public class TuLingResult{    //消息类型(我们在序列化为XML的时候需要提供类型)
    public Type DataType { get; set; }    public object Data { get; set; }
}public interface IResponse{    TuLingResult ToTuLingResult(string fromUserName, string toUserName);
}

Create an entity of text data as the base class of Turing message (corresponding to WeChat's Text message)

public class TextResult : IResponse{    public int Code { get; set; }    public string Text { get; set; }    public virtual TuLingResult ToTuLingResult(string fromUserName, string toUserName)    {        return new TuLingResult
        {
            DataType = typeof(TextMsg),
            Data = new TextMsg
             {
                 FromUserName = fromUserName,
                 ToUserName = toUserName,
                 Content = Text,
                 CreateTime = UnixTimestamp.Now.ToNumeric(),
                 MsgType = MsgType.Text
             }
        };
    }
}

Then create entity classes for various data in sequence.
For example: News (corresponding to WeChat graphic messages)

public class NewsResult : TextResult{    public List<NewsInfo> List { get; set; }    public override TuLingResult ToTuLingResult(string fromUserName, string toUserName)    {        if (List.Count > 10)
        {
            List = List.Take(10).ToList();
        }        return new TuLingResult
        {
            DataType = typeof(NewsMsg),
            Data = new NewsMsg
            {
                FromUserName = fromUserName,
                ToUserName = toUserName,
                ArticleCount = List.Count,
                Articles = List.Select(m => new MsgNewsInfo
                {
                    Title = m.Article,
                    Description = m.Source,
                    Url = m.DetailUrl,
                    PicUrl = m.Icon
                }).ToList(),
                CreateTime = UnixTimestamp.Now.ToNumeric(),
                MsgType=MsgType.News
            }
        };
    }
}public class NewsInfo{    /// <summary>
    /// 标题
    /// </summary>
    public string Article { get; set; }    /// <summary>
    /// 来源
    /// </summary>
    public string Source { get; set; }    /// <summary>
    /// 详情地址
    /// </summary>
    public string DetailUrl { get; set; }    /// <summary>
    /// 图标地址
    /// </summary>
    public string Icon { get; set; }
}

Similarly create various data entity classes provided by Turing robot

The data entities we want to support have been defined After that, we can start to request the Turing interface and obtain the real message. Here we use HttpClient to implement it.

private const string TULING_API_URL = "http://www.tuling123.com/openapi/api";private const string TULING_API_KEY = "XXXXX";//图灵的APIKEYpublic TuLingResult GetMsgFromResponse(string keyword, string userFlag, string efhName){    string linkString = string.Format("{0}?key={1}&info={2}&userid={3}"
        , TULING_API_URL, TULING_API_KEY, keyword, userFlag);    string content = string.Empty;    using (HttpClient client = new HttpClient())
    {
        HttpResponseMessage response = client.GetAsync(linkString).Result;

        content = response.Content.ReadAsStringAsync().Result;
        logger.DebugFormat("图灵机器人响应:{0}", content);
    }    return ConvertToMsg(content, userFlag, efhName);
}

Turing returns the code identification message type and error information, so we first parse the response message into TextResult and get the Turing type.

First define the Turing type enumeration

public enum ResultType
{
    TL_FORMAT_DATA = 50000,
    TL_TEXT_DATA = 100000,
    TL_LINK_DATA = 200000,
    TL_NOVEL_DATA = 301000,
    TL_NEWS_DATA = 302000,
    TL_APP_DATA = 304000,
    TL_TRAIN_DATA = 305000,
    TL_AIRPORT_DATA = 306000,
    TL_TUAN_DATA = 307000,
    TL_TUWEN_DATA = 308000,
    TL_HOTEL_DATA = 309000,
    TL_LOTTERY_DATA = 310000,
    TL_PRICE_DATA = 311000,
    TL_RESTAURANT_DATA = 312000,

    TL_ERROR_LENGTH = 40001,
    TL_ERROR_EMPTY = 40002,
    TL_ERROR_INVALID = 40003,
    TL_ERROR_OUTLIMIT = 40004,
    TL_ERROR_NOTSUPPORT = 40005,
    TL_ERROR_SERVERUPDATE = 40006,
    TL_ERROR_SERVERERROR = 40007
}

The return code corresponding to Turing

100000  文本类数据
200000  网址类数据
301000  小说
302000  新闻
304000  应用、软件、下载
305000  列车
306000  航班
307000  团购
308000  优惠
309000  酒店
310000  彩票
311000  价格
312000  餐厅
40001   key的长度错误（32位）
40002   请求内容为空
40003   key错误或帐号未激活
40004   当天请求次数已用完
40005   暂不支持该功能
40006   服务器升级中
40007   服务器数据格式异常
50000   机器人设定的“学用户说话”或者“默认回答”

And then get the message type

private ResultType GetResultType(string response)
{
    var result = JsonConvert.DeserializeObject<TextResult>(response);

    return (ResultType)result.Code;
}

, we will Corresponding TuLingResult can be returned according to different types.

public TuLingResult ConvertToMsg(string response, string userFlag, string efhName)
{
    IResponse result = null;

    var resultType = GetResultType(response);    switch (resultType)
    {        case ResultType.TL_TEXT_DATA:
            result = JsonConvert.DeserializeObject<TextResult>(response);            break;        case ResultType.TL_LINK_DATA:
            result = JsonConvert.DeserializeObject<LinkResult>(response);            break;        case ResultType.TL_NEWS_DATA:
            result = JsonConvert.DeserializeObject<NewsResult>(response);            break;        case ResultType.TL_TUWEN_DATA:
            result = JsonConvert.DeserializeObject<TuWenResult>(response);            break;        case ResultType.TL_TRAIN_DATA:
            result = JsonConvert.DeserializeObject<TrainResult>(response);            break;        case ResultType.TL_AIRPORT_DATA:
            result = JsonConvert.DeserializeObject<AirportResult>(response);            break;        case ResultType.TL_APP_DATA:
            result = JsonConvert.DeserializeObject<AppResult>(response);            break;        case ResultType.TL_HOTEL_DATA:
            result = JsonConvert.DeserializeObject<HotelResult>(response);            break;        case ResultType.TL_PRICE_DATA:
            result = JsonConvert.DeserializeObject<PriceResult>(response);            break;        case ResultType.TL_ERROR_LENGTH:
        case ResultType.TL_ERROR_INVALID:
        case ResultType.TL_ERROR_EMPTY:
        case ResultType.TL_ERROR_OUTLIMIT:
            result = new TextResult { Text = "您的输入有误" };            break;        case ResultType.TL_ERROR_SERVERERROR:
        case ResultType.TL_ERROR_SERVERUPDATE:
            result = new TextResult { Text = "服务器忙，暂时无法为您提供服务" };            break;        case ResultType.TL_ERROR_NOTSUPPORT:
            result = new TextResult { Text = "俺还小，您说的这个还得慢慢学习，以后再来试吧" };            break;
        default:
            result = new TextResult { Text = "俺还小，不知道你在说啥子(⊙_⊙)?" };            break;
    }    return result.ToTuLingResult(efhName, userFlag);
}

Then, we can serialize the Data in the TuLingResult we got into the XML required by WeChat

var requestResult = TuLingService.Instance.GetMsgFromResponse(content, userName, efhName);

responseContent = MsgService.Instance.ResponseXML(requestResult.Data, requestResult.DataType);return new HttpResponseMessage(HttpStatusCode.OK)
{
    Content = new StringContent(responseContent, Encoding.UTF8, "application/xml"),
};

The above is the detailed content of How to use web api to develop WeChat public account to call Turing robot interface. For more information, please follow other related articles on the PHP Chinese website!