C#中如何反序列化含MTOM边界的SOAP XML响应为POCO

实现方案 #

问题1解答 #

不需要自行编写边界和元数据处理逻辑，.NET 内置了MIME多部分内容解析能力，符合C#开发规范，可以直接从MTOM响应中提取目标SOAP XML：

using System.IO;
using System.Net.Http.Headers;
using System.Net.Http;

public static async Task<string> ExtractSoapXmlFromMtomResponse(string rawResponse, string contentTypeHeaderValue)
{
    // 解析响应头的Content-Type获取边界值
    var contentType = MediaTypeHeaderValue.Parse(contentTypeHeaderValue);
    var boundary = contentType.Boundary.Trim('"');
    
    // 将原始响应转为流供MultipartReader读取
    using var stream = new MemoryStream(System.Text.Encoding.UTF8.GetBytes(rawResponse));
    var reader = new MultipartReader(boundary, stream);
    
    // 遍历多部分内容，找到SOAP XML部分
    while (await reader.ReadNextSectionAsync())
    {
        var section = reader.CurrentSection;
        if (section.ContentType.MediaType == "application/xop+xml")
        {
            using var readerStream = new StreamReader(section.Body);
            return await readerStream.ReadToEndAsync();
        }
    }
    return string.Empty;
}

如果调用接口时直接用HttpClient拿到HttpResponseMessage，可以直接读响应内容的流，不需要先转成字符串，性能更优。

问题2解答 #

命名空间不需要关注别名，只需要匹配实际的命名空间URI即可，return等特殊命名节点可以通过XML序列化属性显式指定映射关系。调整后的POCO定义如下，直接适配响应结构：

using System.Xml.Serialization;

// 对应SOAP Body内的listAttachmentsResponse节点
[XmlRoot("listAttachmentsResponse", Namespace = "http://ws.praxedo.com/v6/businessEvent")]
public class ListAttachmentsResponseDto
{
    [XmlElement("return")]
    public ResponseContent Content { get; set; }
}

public class ResponseContent
{
    [XmlElement("resultCode")]
    public int ResultCode { get; set; }
    
    [XmlElement("entities")]
    public Attachment[] Attachments { get; set; }
}

public class Attachment
{
    [XmlElement("entityId")]
    public string BusinessEventId { get; set; }
    
    [XmlElement("name")]
    public string FileName { get; set; }
    
    [XmlElement("addedOnDevice")]
    public bool IsAddedOnDevice { get; set; }
    
    [XmlElement("creationDate")]
    public DateTime CreationDate { get; set; }
    
    [XmlElement("external")]
    public bool IsExternal { get; set; }
    
    [XmlElement("id")]
    public string Id { get; set; }
    
    [XmlElement("lastModificationDate")]
    public DateTime LastModificationDate { get; set; }
    
    [XmlElement("size")]
    public int Size { get; set; }
    
    [XmlElement("unmodifiable")]
    public bool IsUnmodifiable { get; set; }
}

XML反序列化代码 #

提取到SOAP XML之后，直接用XmlSerializer反序列化即可，会自动处理命名空间和节点映射：

using System.Xml;
using System.Xml.Serialization;

public static ListAttachmentsResponse DeserializeSoapResponse(string soapXml)
{
    var serializer = new XmlSerializer(typeof(ListAttachmentsResponseDto));
    using var stringReader = new StringReader(soapXml);
    using var xmlReader = XmlReader.Create(stringReader);
    
    var dto = (ListAttachmentsResponseDto)serializer.Deserialize(xmlReader);
    
    // 转换为你需要的最终POCO结构
    return new ListAttachmentsResponse
    {
        ResultCode = dto.Content.ResultCode,
        Attachments = dto.Content.Attachments
    };
}

如果不需要分层的DTO结构，也可以直接调整POCO的序列化属性适配层级，减少转换步骤。

内容的提问来源于stack exchange，提问作者Brian Kessler