使用DataSet(c#)和JavaScript。一个创建带有DataSet输入的JSON流的简单函数。创建JSON内容，如(多表数据集):

[[{a:1,b:2,c:3},{a:3,b:5,c:6}],[{a:23,b:45,c:35},{a:58,b:59,c:45}]]

只是客户端，使用eval。例如,

var d = eval('[[{a:1,b:2,c:3},{a:3,b:5,c:6}],[{a:23,b:45,c:35},{a:58,b:59,c:45}]]')

然后使用:

d[0][0].a // out 1 from table 0 row 0

d[1][1].b // out 59 from table 1 row 1

// Created by Behnam Mohammadi And Saeed Ahmadian
public string jsonMini(DataSet ds)
{
    int t = 0, r = 0, c = 0;
    string stream = "[";

    for (t = 0; t < ds.Tables.Count; t++)
    {
        stream += "[";
        for (r = 0; r < ds.Tables[t].Rows.Count; r++)
        {
            stream += "{";
            for (c = 0; c < ds.Tables[t].Columns.Count; c++)
            {
                stream += ds.Tables[t].Columns[c].ToString() + ":'" +
                          ds.Tables[t].Rows[r][c].ToString() + "',";
            }
            if (c>0)
                stream = stream.Substring(0, stream.Length - 1);
            stream += "},";
        }
        if (r>0)
            stream = stream.Substring(0, stream.Length - 1);
        stream += "],";
    }
    if (t>0)
        stream = stream.Substring(0, stream.Length - 1);
    stream += "];";
    return stream;
}

获取一个ExpandoObject:

using Newtonsoft.Json;
using Newtonsoft.Json.Converters;

Container container = JsonConvert.Deserialize<Container>(jsonAsString, new ExpandoObjectConverter());

使用DataSet(c#)和JavaScript。一个创建带有DataSet输入的JSON流的简单函数。创建JSON内容，如(多表数据集):

[[{a:1,b:2,c:3},{a:3,b:5,c:6}],[{a:23,b:45,c:35},{a:58,b:59,c:45}]]

只是客户端，使用eval。例如,

var d = eval('[[{a:1,b:2,c:3},{a:3,b:5,c:6}],[{a:23,b:45,c:35},{a:58,b:59,c:45}]]')

然后使用:

d[0][0].a // out 1 from table 0 row 0

d[1][1].b // out 59 from table 1 row 1

// Created by Behnam Mohammadi And Saeed Ahmadian
public string jsonMini(DataSet ds)
{
    int t = 0, r = 0, c = 0;
    string stream = "[";

    for (t = 0; t < ds.Tables.Count; t++)
    {
        stream += "[";
        for (r = 0; r < ds.Tables[t].Rows.Count; r++)
        {
            stream += "{";
            for (c = 0; c < ds.Tables[t].Columns.Count; c++)
            {
                stream += ds.Tables[t].Columns[c].ToString() + ":'" +
                          ds.Tables[t].Rows[r][c].ToString() + "',";
            }
            if (c>0)
                stream = stream.Substring(0, stream.Length - 1);
            stream += "},";
        }
        if (r>0)
            stream = stream.Substring(0, stream.Length - 1);
        stream += "],";
    }
    if (t>0)
        stream = stream.Substring(0, stream.Length - 1);
    stream += "];";
    return stream;
}

在JSON中反序列化。NET可以使用包含在该库中的JObject类来实现动态。我的JSON字符串表示这些类:

public class Foo {
   public int Age {get;set;}
   public Bar Bar {get;set;}
}

public class Bar {
   public DateTime BDay {get;set;}
}

现在我们在不引用上述类的情况下反序列化字符串:

var dyn = JsonConvert.DeserializeObject<JObject>(jsonAsFooString);

JProperty propAge = dyn.Properties().FirstOrDefault(i=>i.Name == "Age");
if(propAge != null) {
    int age = int.Parse(propAge.Value.ToString());
    Console.WriteLine("age=" + age);
}

//or as a one-liner:
int myage = int.Parse(dyn.Properties().First(i=>i.Name == "Age").Value.ToString());

或者如果你想深入一点:

var propBar = dyn.Properties().FirstOrDefault(i=>i.Name == "Bar");
if(propBar != null) {
    JObject o = (JObject)propBar.First();
    var propBDay = o.Properties().FirstOrDefault (i => i.Name=="BDay");
    if(propBDay != null) {
        DateTime bday = DateTime.Parse(propBDay.Value.ToString());
        Console.WriteLine("birthday=" + bday.ToString("MM/dd/yyyy"));
    }
}

//or as a one-liner:
DateTime mybday = DateTime.Parse(((JObject)dyn.Properties().First(i=>i.Name == "Bar").First()).Properties().First(i=>i.Name == "BDay").Value.ToString());

完整的示例请参见文章。

为此，我将使用JSON。NET对JSON流进行低级解析，然后从ExpandoObject类的实例构建对象层次结构。

.NET 4.0有一个内置库可以做到这一点:

using System.Web.Script.Serialization;
JavaScriptSerializer jss = new JavaScriptSerializer();
var d = jss.Deserialize<dynamic>(str);

这是最简单的方法。