Accordion: correct height calculated when closed

[apushak]

Fixes #11938

[apushak]

I have no idea why that unit test is failing. All the tests pass when I run them on my machine, so I don't know how I can fix it.

[scottgonzalez]

Can we get a test for this? heightStyle: "fill" needs to be updated as well.

[scottgonzalez]

Just use $( this ).show().

[scottgonzalez]

Well, more explicitly, follow this logic:

jquery-ui/ui/dialog.js

Lines 634 to 642 in adcc8ef

	// Need to show the dialog to get the actual offset in the position plugin
	var isVisible = this.uiDialog.is( ":visible" );
	if ( !isVisible ) {
	this.uiDialog.show();
	}
	this.uiDialog.position( this.options.position );
	if ( !isVisible ) {
	this.uiDialog.hide();
	}

[apushak]

I've changed it to use is( ":visible" ) and show() / hide(), but that unit test is still failing while it passes if I run it on my machine. It seems like calling show() and hide() causes some change in the element that assert.domEqual detects, but not when I run the test locally.

[scottgonzalez]

I'm not sure what's going on with that failure either. Can you go ahead with the rest of the changes? I'll try to dig into the Travis failure.

[apushak]

I've added a unit test.

heightStyle "fill" uses height() - innerHeight() which just measures padding, and is not effected by the visibility of the element, so no changes should be necessary. I tested this in Chrome, Firefox, and IE.

[scottgonzalez]

You're testing a set of options, not a method, so this is in the wrong file and the test title doesn't really make sense.

[scottgonzalez]

Add // http://bugs.jqueryui.com/ticket/11938 above the test.

[scottgonzalez]

The changes and test look good. Just a minor note about the placement of the test. If you'd prefer, I can make that change myself when I land this. Otherwise, I'll just wait to hear back from you. Thanks.

[scottgonzalez]

Also, we still have no idea why that test is failing in Travis, but we'll keep digging.

[scottgonzalez]

Can you rebase on master and see if that helps at all with Travis? It's a clean rebase, I tested locally.

[apushak]

I did a rebase on master which did not fix the Travis issue. I also moved and renamed the unit test.

[scottgonzalez]

I'm still trying to figure out what's causing the test to fail. So far I've tracked it down to this difference:

< "WebkitTransformOrigin": "500px 22px",
> "WebkitTransformOrigin": "492px 22px",
< "WebkitPerspectiveOrigin": "500px 22px",
> "WebkitPerspectiveOrigin": "492px 22px",
< "width": "1000px",
> "width": "984px",

[scottgonzalez]

It's been almost a year, so you've probably forgotten about this by now. But I have good news! We just upgraded to Phantom 2 and the tests now pass on Travis.