XML 特定于版本的反序列化

XML version-specific deserialization

所以,我有一个基础项目和它的几个修改版本。 BaseProject 包含 class BaseClass,

namespace BaseProject.SomeClasses{
    public abstract class BaseClass{
    //...
    }
}

而每个版本都包含 BaseClass - f.e 的几个继承者。 ProjectOne:

namespace BaseProject.VersionOne.SomeClasses{
    public class InheritorClass : BaseClass{
    //Some logic here...
    }
}

ProjectTwo:

namespace BaseProject.VersionTwo.SomeClasses{
    public class InheritorClass : BaseClass{
    //Some different logic here...
    }
}

唯一不同的是命名空间的名称。

基础项目在运行时加载每个程序集并获取所有继承类型。我需要创建 xml 文件,该文件应包含继承者的实例和一些指向 class 的指针,实例应反序列化为:

...
<BaseClass xsi:type="InheritorClass"> <!-- From VersionOne -->
    <PropOne></PropOne>
    <PropTwo></PropTwo>
    <PropThree></PropThree>
    <!-- ... -->
</BaseClass>
<BaseClass xsi:type="InheritorClass"> <!-- From VersionTwo -->
    <PropFour></PropFour>
    <PropFive></PropFive>
    <PropSix></PropSix>
    <!-- ... -->
</BaseClass>
...

有什么方法可以将这个 xml(包含两个版本的继承者的实例)反序列化为 IEnumerable<BaseClass>

class 的命名空间应该没有什么区别。请参阅此代码生成的 XML 文件。

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Xml;
using System.Xml.Serialization;
using System.IO;

namespace ConsoleApplication1
{
    class Program
    {
        const string FILENAME = @"c:\temp\test.xml";
        static void Main(string[] args)
        {
            Root root = new Root()
            {
                baseClass = new List<BaseClass>(){
                    new InheritorClass1(){ name = "class1"},
                    new InheritorClass2(){ name = "class2"}
                }
            };
            XmlSerializer serializer = new XmlSerializer(typeof(Root));

            StreamWriter writer = new StreamWriter(FILENAME);
            serializer.Serialize(writer, root);
            writer.Flush();
            writer.Close();
            writer.Dispose();

            XmlSerializer xs = new XmlSerializer(typeof(Root));
            XmlTextReader reader = new XmlTextReader(FILENAME);
            Root newRoot = (Root)xs.Deserialize(reader);
        }
    }
    [XmlRoot("Root")]
    public class Root
    {
        [XmlElement("BaseClass")]
        public List<BaseClass> baseClass { get; set; }
    }
    [XmlRoot("BaseClass")]
    [XmlInclude(typeof(InheritorClass1))]
    [XmlInclude(typeof(InheritorClass2))]
    public class BaseClass
    {
        [XmlElement("name")]
        public string name { get; set; }    

    }
    [XmlRoot("InheritorClass1")]
    public class InheritorClass1 : BaseClass
    {
    }

    [XmlRoot("InheritorClass2")]
    public class InheritorClass2 : BaseClass
    {
    }

}
​

您需要将 [XmlTypeAttribute(name)] 属性应用于派生的 类 以消除 xst:type 名称的歧义:

namespace BaseProject.VersionOne.SomeClasses
{
    [XmlType("VersionOneInheritorClass")]
    public class InheritorClass : BaseClass
    {
        public string VersionOneProperty { get; set; } // For instance
    }
}

namespace BaseProject.VersionTwo.SomeClasses
{
    [XmlType("VersionTwoInheritorClass")]
    public class InheritorClass : BaseClass
    {
        public string VersionTwoProperty { get; set; } // For instance
    }
}

那么你的XML会出现如下,可以序列化反序列化成功,不丢失信息:

<BaseClass xsi:type="VersionOneInheritorClass">
    <VersionOneProperty>one</VersionOneProperty>
</BaseClass>
<BaseClass xsi:type="VersionTwoInheritorClass">
    <VersionTwoProperty>two</VersionTwoProperty>
</BaseClass>

顺便说一下,如果您使用 XmlSerializer (Type, Type[]) constructor to add in your discovered derived types, you must cache the serializer in a static cache somewhere 构建 XmlSerializer,否则您将遇到可怕的资源泄漏。