Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

Javascript: Building a hierarchical tree

Tags:

javascript

My data has these properties:

  1. Each entry has a unique id (Id)
  2. Each has a Parent field, which points to the Id of the parent.
  3. A node can have multiple children, but only one parent.

My first attempt to build a tree is below. It is buggy as the recursion causes an infinite loop. Even if I solve it, I am not sure if there is a better approach to do this. Currently, I am doing it in 2 passes.

I would like it to be as efficient as possible as I have a decent amount of data. It also needs to rebuild the tree dynamically (the root can be any node)

There is sample data in the program below:

 arry = [{"Id":"1", "Name":"abc", "Parent":""}, {"Id":"2", "Name":"abc", "Parent":"1"},
    {"Id":"3", "Name":"abc", "Parent":"2"},{"Id":"4", "Name":"abc", "Parent":"2"}]//for testing

I was hoping the output to be (it might be wrong nested structure, as I manually wrote it. but, what I am hoping is a valid JSON structure with node as a field 'value' and children as an array.)

{
 "value": {"Id":"1", "Name":"abc", "Parent":""},
 "children": [
  {
   "value": {"Id":"2", "Name":"abc", "Parent":"1"},
   "children": [
    {
     "value": {"Id":"3", "Name":"abc", "Parent":"2"},
     "children": []
     },
     {
     "value": {"Id":"4", "Name":"abc", "Parent":"2"},
     "children": []
     }
   ]
..
}

Sample program:

function convertToHierarchy(arry, root) 
{
//root can be treated a special case, as the id is known
    arry = [{"Id":"1", "Name":"abc", "Parent":""}, {"Id":"2", "Name":"abc", "Parent":"1"},
    {"Id":"3", "Name":"abc", "Parent":"2"},{"Id":"4", "Name":"abc", "Parent":"2"}]//for testing


    var mapping = {}; // parent : [children]
    for (var i = 0; i < array.length; i++) 
    {
        var node = arry[i];

    if (!mapping[node.Id]) { 
          mapping[node.Id] = {value: node, children:[] } ;
        }else{
      mapping[node.Id] = {value: node} //children is already set    
    }

    if (!mapping[node.Parent]) { //TODO what if parent doesn't exist.
                mapping[node.Parent] =  {value: undefined, children:[ {value: node,children:[]} ]};
        }else {//parent is already in the list
        mapping[node.Parent].children.push({value: node,children:[]} )
    }

    }
    //by now we will have an index with all nodes and their children.

    //Now, recursively add children for root element.

    var root = mapping[1]  //hardcoded for testing, but a function argument
    recurse(root, root, mapping)
    console.log(root)

    //json dump
}

function recurse(root, node, mapping)
{
    var nodeChildren = mapping[node.value.Id].children;
    root.children.push({value:node.value, children:nodeChildren})
   for (var i = 0; i < nodeChildren.length; i++) {
        recurse(root, nodeChildren[i], mapping);
    }
    return root;
}

I have 3 good solutions so far, and hope the upvotes suggest more idiomatic, efficient implementation. I am not sure, utilizing the property of my data that, there will be only one root element in the set of input array, and also the root is always given, any of these implementation could be better. I should also be learning how to benchmark, as my requirement is how efficiently (fast/without much memory) the tree can be rebuild. For example, the input is already cached (array) and rebuild the tree like

convertToHierarchy(parentid)
....
convertToHierarchy(parentid2)
...
like image 884
bsr Avatar asked Oct 11 '12 03:10

bsr


People also ask

What is tree structure in JavaScript?

Definition. A tree is a data structure consisting of a set of linked nodes that represent a hierarchical tree structure. Each node is linked to others via parent-children relationship. The first node in the tree is the root, whereas nodes without any children are the leaves.

Why tree is hierarchical data structure?

A tree is non-linear and a hierarchical data structure consisting of a collection of nodes such that each node of the tree stores a value and a list of references to other nodes (the “children”). This data structure is a specialized method to organize and store data in the computer to be used more effectively.


1 Answers

Here's one solution:

var items = [
    {"Id": "1", "Name": "abc", "Parent": "2"},
    {"Id": "2", "Name": "abc", "Parent": ""},
    {"Id": "3", "Name": "abc", "Parent": "5"},
    {"Id": "4", "Name": "abc", "Parent": "2"},
    {"Id": "5", "Name": "abc", "Parent": ""},
    {"Id": "6", "Name": "abc", "Parent": "2"},
    {"Id": "7", "Name": "abc", "Parent": "6"},
    {"Id": "8", "Name": "abc", "Parent": "6"}
];

function buildHierarchy(arry) {

    var roots = [], children = {};

    // find the top level nodes and hash the children based on parent
    for (var i = 0, len = arry.length; i < len; ++i) {
        var item = arry[i],
            p = item.Parent,
            target = !p ? roots : (children[p] || (children[p] = []));

        target.push({ value: item });
    }

    // function to recursively build the tree
    var findChildren = function(parent) {
        if (children[parent.value.Id]) {
            parent.children = children[parent.value.Id];
            for (var i = 0, len = parent.children.length; i < len; ++i) {
                findChildren(parent.children[i]);
            }
        }
    };

    // enumerate through to handle the case where there are multiple roots
    for (var i = 0, len = roots.length; i < len; ++i) {
        findChildren(roots[i]);
    }

    return roots;
}

console.log(buildHierarchy(items));​
like image 162
Bill Avatar answered Sep 19 '22 16:09

Bill