遍历 JObject：https://www.newtonsoft.com/json/help/html/JObjectProperties.htm
遍历 JArray：
1
2
3
4
5
6
7
8
9
10
11
12
string json = @"[
'Small',
'Medium',
'Large'
]";
JArray jarray = JArray.Parse(json);
for (int i = 0; i < jarray.Count; i++)
{
Console.WriteLine(jarray[i]);
}
查找并遍历：https://www.newtonsoft.com/json/help/html/QueryJsonSelectTokenJsonPath.htm
例子：
打开 Feature 和 Addition 这两个 JSON 字符串，并去除全部空字符（我这回的数据只有三层，不知道数据层数时要递归遍历）
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
private JToken Preprocessing(DataTable tbl)
{
JArray jr = JArray.Parse(SerializeToJson(tbl));
for (int i = 0; i < jr.Count; i++)
{
foreach (JProperty p in jr[i].ToObject<JObject>().Properties())
{
if (p.Name.Equals("Feature") || p.Name.Equals("Addition"))
{
p.Value = JObject.Parse(p.Value.ToString());
jr[i][p.Name] = p.Value;
foreach (JProperty subp in p.Value.ToObject<JObject>().Properties())
{
jr[i][p.Name][subp.Name] = Regex.Replace(subp.Value.ToString(), @"\s", "");
}
}
else
{
jr[i][p.Name] = Regex.Replace(p.Value.ToString(), @"\s", "");
}
}
}
return jr;
}