Possible improvements

[cibernox]

I'm on mobile so I haven't tried any of this, but I've seen a couple improvements:

• this.$('.foo').attr('id') => find('.foo').id. I think that it's more idiomatic.
• this.$('.foo').val('bar').trigger('input') => await fillIn('.foo', 'bar'). There is no need to trigger the event as fillIn does that by default.
• this.$('.foo').val('bar').trigger('change') => await fillIn('.foo', 'bar'); await blur('.foo'). On input fields the change event is not fired until the input is blurred, so I guess that we should blur it.

This is all I've seen on the readme. Once I'm back from holidays I'll try it myself.

[simonihmig]

this.$('.foo').attr('id') => find('.foo').id. I think that it's more idiomatic.

@cibernox That's a source of constant confusion I think, in what is a DOM node attribute and what a property, at least for me. And certain attributes are so called 'reflected properties', like id, that can be used as properties as well (see http://stackoverflow.com/a/6004028). So I don't know how to best handle this. id still is primary an attribute, so using getAttribute seems perfectly valid, but as a reflected property, using .id would work as well. But that applies to many attributes, so to be consistent, should we use the property syntax for all reflected properties?

This would require maintaining a list of all reflected properties, and using that to convert all instances of .attr('reflectedProp') to .reflectedProp, instead of .getAttribute('reflectedProp'). Right now the conversion is pretty straightforward and simple: .attr('foo') -> .getAttribute('foo') and .prop('bar') -> .bar, without any knowledge of what foo and bar really are (attr or prop).

I would think the current approach is "good enough", but open for other thoughts...

Regarding the two other transforms you mentioned: good to know, will try to change them!

If you've got any other suggestions, would be happy to fix them! :)

[cibernox]

Yeah, I'm not very concerned, but I thought that id and class are so prevalent that it might worth treating them differently and create less verbose code.

As a general rule I think that your approach of .attr() -> getAttribute() and .prop('foo') -> .foo is correct..

But ids and classes easily account for for 90% of the times I check for attributes of an element, that's why I think that we could make an exception for those two.

In particular, I'm thinking in accessing the id as a property instead of an attribute, because they match, and also being smart enough to replace $el.hasClass('foobar') by classList.contains('foobar').

Again, I wouldn't push this very hard if you think it doesn't worth the effort, I just thought that it would be closer to the code I'd write if I did the transformation manually.

[simonihmig]

Ok, I will see, shouldn't be that hard. I was a bit hesitant to make this work for all possible cases of reflected properties, as this would require maintaining and coupling this very strongly to the HTML spec. But treating id as a special case here probably makes sense...

[simonihmig]

Your proposed improvements are implemented now! :)