Mendix Tree Table

Mendix Tree Table using Ant Design Table (MIT License).

Show a tree-like structure in Mendix.

See test-project https://reacttreeview-sandbox.mxapps.io/ for a live demo! This project is also used for all E2E tests. See at the bottom of this README a preview of the domain model. Test-project can be downloaded here

Features

• Display a tree structure in a table
• DataSource:
  • Root:
  • Xpath
  • Microflow
  • Nanoflow
  • Children:
  • Reference
  • Microflow
  • Nanoflow
• Row classes
• Row icons
• Select row/rows, execute on change/action button
• Dynamic columns
• Much more...

This widget is only ~500Kb uncompressed, so in your cloud environment this would be about ~140Kb. This is light-weight for any project.

Tested:

• IE11
• Chrome,Firefox,Safari
• Should work on Mobile Web, but might not be very usable

Basic configuration

1. Data

• Select an Entity that will serve as your Node.
• You can choose whether to load the whole tree or partial (top level)
• Nodes can be retrieved through a few methods: Xpath, Microflow, Nanoflow
• Please note that this only loads the top-level data (first nodes)
• When you set this to XPath, define your constraint (if needed)
• Microflow & Nanoflow should return a list of Nodes

2. Children

• You do not have to load children if you just want to load a flat list
• If you load children, it can be done in two ways:
  • Child reference: Your Node should have a self-reference, as a reference set (so a many-to-many relation). You define this reference, the widget will take care of checking if it has children or not and load them when opening the parent
  • Microflow/Nanoflow: Your Node should have an attribute that tells whether or not it has children. This can be accomplished during the creation of your nodes, or a calculated attribute. If the widget sees a node has a child it will load the children.

3. Helper

• For doing clicks and selections, we need a Helper object. This will be created by the widget for an action and passed down to a microflow/nanoflow. This helper object has a reference to the context object (view) and a reference SET to the Node objects. When you execute a microflow/nanoflow, you will need to retrieve the node object itself over a reference.

4. Columns

• Easiest way to configure columns is a list. This is a sorted list of all the attributes you want to show in the table. See 3.1 Columns List
• If you want to do this dynamically (because your client wants to configure it), please look at 4. Columns

4.1. Column List

• A column should have a caption (shown at the top) and an attribute (to display the value)
• If you need to transform your value client-side, you can use a Transform Nanoflow. This should always return a string
• In the column UI tab you can set the width of the column (as '100', '25%' etc)
• You can also set a class name

5. Dynamic Columns

• Dynamic columns can be used to let the user define their column. This is a bit tricky though:
  • The attribute of the Node entity should be saved in the Attribute attribute of the column Entity (this is confusing, right 😜?).
• All other attributes should be straightforward.

Note:

In order to get the attributes of your Node entity in your project, I encourage you to use the Model Reflection module in the AppStore. This has the ability to load all entities and read their attributes. Then use this to create your Column entities.

6. Events

• Events are pretty straightforward. The widget uses a debounce, which means it will not single click when you double-click.
• For the microflow/nanoflow, you will need to configure the helper object!

7. UI

• Header row (which contains the captions of the column) can be disabled
• You can set a class on the row through an attribute in your Node. This way it is possible to change colors (needs custom styling)
• Icons can be shown on the first column of the row. See test-project

8. Selection

• Just like a normal table, you can use selection to do things with the Nodes.
• Configure buttons for a selection.
• The microflow/nanoflows use the Helper object
• Please note that when we use the Selection onChange, chances are that you will make a change to your context object (for example, set a selection reference/set). This would mean that the table re-renders, losing the selection. We circumvent this by temporarily lift the subscriptions (that will trigger when the context changes), execute the action, then reapply the subscriptions.

9. Misc

• You can now save the state of your table. This only works for the whole tree, no partial trees!
• Experimentally expose window.__TreeTable_{guid}_select. This will accept a guid as string, or array of guids for multi selection

Test-project

The test-project uses the following domain model. This is purely an example, but can be used as an inspiration to use in your own project:

Test-project can be downloaded here

Issues, suggestions and feature requests

Please report your issues here

License

Apache 2