Chae-Tree

Create multi-level navigation in elm easily.

This package provides essential abstractions for manipulating and creating tree data structure directly from your collection data.

Imagine you have collection of some items which has defined parent <-> children relationship. This package provides easy and universal way to transform this list to tree like data structure and comes with easy to use functions you can use to work with it. For example you can easily browse this tree by levels, getting ancestors of node etc.

This package makes no decision about neither update nor model nor view. You can build your own application level logic as you wish and just use provided api to manipulate your data. Every abstraction which you might need is build in so you're saved from thinking about implementation details and rather focus on actual UI and business logic.

The name comes from Chaenomeles and is reference to similarities with RoseTree

The good place to start is by looking on example code.

Installation

elm-package install turboMaCk/chae-tree

For more informations please follow documentation or see examples of usage.

More About Data Structure

You might be familiar with Rose Tree data structure. The data structure this plugin uses is quite similar but not really the same. This paragraph(s) is to explain similarities and differences between the two and more importantly explain why they are there. First things first please do not use this implementation as an alternative to plain Rose Trees! Even they share some similarities you can find it unnecessary hard to use this in every place you might want to use Rose Tree. Think about this as about more domain specific tree like structure which might be quite handy for one thing but not that good for some other.

Chae.Tree is Forest

In fact Tree in the context of this plugin is not node containing item and children. Tree is collection of multiple nodes (List of Node a to be more precise). Thanks to this Chae.Tree allows you to have multiple nodes in root. For example think about collection of categories where every category can have multiple sub-categories. In that case Tree is actual collection and every category is one Node in that tree.

Chae.Node is not Rose Tree Either

Now you might be thinking that if Tree is just an alias for List (Node a) than Node a is actually a Rose Tree. You're partially right! In fact Node is pretty close to Rose Tree definition which might look like:

type RoseTree a = Node a (List (RoseTree a))

in fact node is defined as follows:

type Node a = Node Id a (List (Node a))

So what is that magic Id? Id is just alias for String. And this Id/String is used as identifier for that node. The string is used since you can easily represent any value using string and since it do not makes much sense to allow any arithmetics over identifier it's perfect fit for ids.

So every node in Tree has its id. What is this good for? Well first of all you can lookup any node by knowing its id. Also it is super easy to build tree from plain List if you just know what is a parent of each item in that list.

All these functions are implemented for you so you can simple transform your flat collection into tree and start quering it by ids.

Functors

Beside this both Node and Tree are functors and you can find essential functions like map, reduce, zip and similar in particular module.