Question

I’m using the following Coffeescript code to validate that initialization of one backbone.js view constructs another:

describe 'Avia.AviaView', ->

  beforeEach ->
    @aviaView = new Avia.AviaView(addFixtureDiv('avia'))
    @matricesView = new Backbone.View()

    spyOn(Avia, 'MatricesView').andCallFake(
      (element) =>
        if !element
          throw "Expected MatricesView to be constructed with a parent element"
        else if element.attr('id') != 'tabs-3'
          throw "Expected MatricesView to be constructed with the parent element #tabs-3"
        else
          @matricesView
    )

  describe 'initialize', ->

    beforeEach ->
      @aviaView.initialize()

    it 'creates a new MatricesView ', ->
      expect(Avia.MatricesView).toHaveBeenCalledOnce()

This works nicely, but I can’t help but think it ought to be possible to improve it. I’m imagining a syntax like:

it 'creates a new MatricesView ', ->
  expect(Avia.MatricesView).toHaveBeenCalledMatching((args...) => args[0].attr('id') == 'tabs-3')

… where toHaveBeenCalledMatching takes a function that takes a splat of the arguments, and returns truthy to indicate success, and falsy otherwise.

Has anyone come across something like this, or do I need to roll my own here? Or, has anyone a better suggestion of how to improve this code?

Answer 1

Does toHaveBeenCalledWith() not work for you? In most cases you either know the values ahead of time, or can calculate them.

If you really need to use a function to evaluate the args for the call, you could use separate expectations and argsForCall to test each arg individually.

If worst comes to worst, you can always write your own matcher. But that sounds a lot harder than the first two 🙂