不使用递归实现无限层次结构

　　我们从数据库里读取的数据都是无层次结构的二维数据，通过O/RM工具一般都可以直接转换到集合类型，之前一直使用递推将无层次的集合类型转换为有层次的集合类型。现在不用递归实现了无限层次结构。具体代码和分析如下：

　　 定义对象，这里使用经常用到的部门对象

 

 

 1 public class Department
 2     {
 3         /// <summary>
 4         /// 唯一标识
 5         /// </summary>
 6         public int Id { get; set; }
 7 
 8         /// <summary>
 9         /// 部门名称
10         /// </summary>
11         public string Name { get; set; }
12 
13         /// <summary>
14         /// 上级部门唯一标识
15         /// </summary>
16         public int ParentId { get; set; }
17 
18         /// <summary>
19         /// 所有直接下一级部门
20         /// </summary>
21         public IList<Department> ChildDepartments;
22     }

　　准备测试用的数据集合，这里使用List

 1 /// <summary>
 2         /// 取得测试数据
 3         /// </summary>
 4         /// <returns></returns>
 5         public static IList<Department> GetDepartments()
 6         {
 7 
 8             var l = new List<Department>()
 9                         {
10                             //ParentId = 0 表示顶层节点
11                             new Department() {Id = 2, Name = "董事会",ParentId = 1},
12                             new Department() {Id = 4, Name = "行政部",ParentId=1},
13                             new Department() {Id = 5, Name = "研发部",ParentId=1},
14                             new Department() {Id = 1, Name = "浙江**股份有限公司",ParentId = 0},
15                             new Department() {Id = 6, Name = "营销中心",ParentId=1},
16                             new Department() {Id = 7, Name = "人力资源部",ParentId=4},
17                             new Department() {Id = 8, Name = "采购部",ParentId =4},
18                             new Department() {Id = 3, Name = "总经理办公室",ParentId =1},
19                             new Department() {Id = 9, Name = ".net技术研发部",ParentId =5},
20                             new Department() {Id = 10, Name = "销售一部",ParentId =6},
21                             new Department() {Id = 11, Name = "java技术研发部",ParentId =5 },
22                             new Department() {Id = 12, Name = "销售二部",ParentId= 6},
23                             new Department() {Id = 13, Name = "第一项目组",ParentId=9},
24                             new Department() {Id = 14, Name = "华东市场部",ParentId=12},
25                             new Department() {Id = 15, Name = "华南市场部",ParentId=12},
26                             new Department() {Id = 16, Name = "第二项目组" ,ParentId = 9},
27                             new Department() {Id = 17, Name = "上海**集团",ParentId = 0}
28 
29                         };
30 
31 
32             return l;
33         }

上面的List集合是没有层次结构的，实际项目中一般来源于数据库的二维表。

 

下面将List转换成具有层次结构的List

 

 1 public static IList<Department> ToTree(IList<Department> ds)
 2         {
 3             //定义字典类型，将List转换成字典类型，集合中的元素个数是相同的
 4             var dic = new Dictionary<int, Department>(ds.Count);
 5 
 6             foreach (var department in ds)
 7             {
 8                 dic.Add(department.Id,department);
 9             }
10 
11             //通过一次遍历集合，转换成具有层次结构的类型
12             foreach (var department in dic.Values)
13             {
14                 if(dic.ContainsKey(department.ParentId))
15                 {
16                     if (dic[department.ParentId].ChildDepartments ==null)
17                         dic[department.ParentId].ChildDepartments = new List<Department>();
18 
19                     dic[department.ParentId].ChildDepartments.Add(department);
20                 }
21             }
22 
23             //仅仅选择最顶层节点返回
24             return dic.Values.Where(t => t.ParentId == 0).ToList();
25         }

以下是测试代码

 var t = Util.ToTree(Util.GetDepartments());

 通过调试可以清楚看到当前的List已经具有层次结构了，该集合中只有两个顶层节点的元素了。

 

以下是打包后代码

 /Files/szp1118/WebApplication4.rar