Use Cases

can-key-tree can be used for a wide variety of purposes. In CanJS, it is used extensively for storing event handlers organized by key and event queue. Its use of can-reflect means it can simplify complex patterns such as implementing event delegation as shown in the Advanced Use section below.

When you are adding, removing, and retrieving items from a nested structure, can-key-tree can likely help.

Use

Import the KeyTree constructor from can-key-tree:

import KeyTree from  "can-key-tree";

Create an instance of KeyTree with an array of types. An instance of each type will be used as the nodes of the tree. The following creates a tree structure 3 levels deep:

const keyTree = new KeyTree( [ Object, Object, Array ], { onFirst, onEmpty } );

Once you've created a keyTree, you can .add, .delete and .get values from it.

.add(keys)

The following adds three handlers:

function handler1() {}
function handler2() {}
function handler3() {}

keyTree.add( [ "click", "li", handler1 ] );
keyTree.add( [ "click", "li", handler2 ] );
keyTree.add( [ "click", "span", handler3 ] );

The keyTree data structure will look like:

{
    "click": {
        "li": [ handler1, handler2 ],
        "span": [ handler3 ]
    }
}

.get(keys)

To get all the li click handlers, use .get:

keyTree.get( [ "click", "li" ] ); //-> [handler1, handler2]

To get all click handlers, you can also use .get:

keyTree.get( [ "click" ] ); //-> [handler1, handler2, handler3]

.delete(keys)

To delete a handler, use .delete:

keyTree.delete( [ "click", "li", handler1 ] );

The keyTree data structure will look like:

{
    "click": {
        "li": [ handler2 ],
        "span": [ handler3 ]
    }
}

To delete the remaining click handlers:

keyTree.delete( [ "click" ] );

The keyTree data structure will look like:

{}

Advanced Use

Often, when a node is created, there needs to be some initial setup, and when a node is empty, some teardown.

This can be achieved by creating custom types. For example, perhaps we want to build an event delegation system where we can delegate from an element like:

eventTree.add( [ document.body, "click", "li", handler ] );

And remove that handler like:

eventTree.delete( [ document.body, "click", "li", handler ] );

We can do that as follows:

// Create an event handler type.
const Delegator = function( parentKey ) {

    // Custom constructors get called with their parentKey.
    // In this case, the `parentKey` is the element we will
    // delegate from.
    this.element = parentKey;

    // the nested data `{click: [handlers...], dblclick: [handlers...]}`
    this.events = {};

    // the callbacks added for each handler.
    this.delegated = {};
};
canReflect.assignSymbols( Delegator.prototype, {

    // when a new event happens, setup event delegation.
    "can.setKeyValue": function( eventName, handlersBySelector ) {

        this.delegated[ eventName ] = function( ev ) {
            canReflect.each( handlersBySelector, function( handlers, selector ) {
                let cur = ev.target;
                do {
                    if ( cur.matches( selector ) ) {
                        handlers.forEach( function( handler ) {
                            handler.call( cur, ev );
                        } );
                    }
                    cur = cur.parentNode;
                } while ( cur && cur !== ev.currentTarget );
            } );
        };
        this.events[ eventName ] = handlersBySelector;
        this.element.addEventListener( eventName, this.delegated[ eventName ] );
    },
    "can.getKeyValue": function( eventName ) {
        return this.events[ eventName ];
    },

    // when an event gets removed, teardown event delegation and clean up.
    "can.deleteKeyValue": function( eventName ) {
        this.element.removeEventListener( eventName, this.delegated[ eventName ] );
        delete this.delegated[ eventName ];
        delete this.events[ eventName ];
    },

    // we need to know how many items at this node
    "can.getOwnEnumerableKeys": function() {
        return Object.keys( this.events );
    }
} );

// create an event tree that stores:
// - "element being delegated" ->
//   - A "delegator" instance for an event ->
//     - The "selectors" we are delegating ->
//       - The handlers to call
const eventTree = new KeyTree( [ Map, Delegator, Object, Array ] );


// to listen to an event:
function handler() {
    console.log( "an li clicked" );
}

eventTree.add( [ document.body, "click", "li", handler ] );

// to stop listening:
eventTree.delete( [ document.body, "click", "li", handler ] );

// to stop listening to all clicks on the body:
eventTree.delete( [ document.body, "click" ] );

// to stop listening to all events on the body:
eventTree.delete( [ document.body ] );

How it works

can-key-tree's' commented source can be found here.

On a high level, KeyTree instances maintain a this.root map-type whose keys point to instances of other map-types. Those keys eventually point to some list-type instance that contains a list of the leaf-values added to the tree. KeyTree's methods walk down the tree structure and use can-reflect to perform operations on it.

Watch this overview video on can-key-tree.