Hi Justin,
So is your goal to restore Flex 2 behavior? Honestly, I don't recall if
refresh was required back then or not, but I think the code paths that
require a call to refresh() after sort=null have been around as several
releases.
I'd rather just make sure that there is no undesirable behavior if you do
call refresh() after setting sort=null.
Thoughts?
On 5/15/13 9:59 PM, "Justin Mclean" <jus...@classsoftware.com> wrote:
>Hi,
>
>> But first, let's nail this sort=null thing. I can look into it if you
>> want.
>
>Just looking at it now.
>
>I've written a couple of dozen unit tests and have 3 failing in both
>ApacheFlex 4.10 that I think should pass.
>
>Leave it with me but so you get an idea this fails with an RTE:
>
> addNumbers(); // adds 1 and 2
> var sort:Sort = new Sort();
> sort.fields = [new SortField(null, false, true, true)];
> ac.sort = sort;
> ac.refresh();
> ac.sort = null;
> addNumbers(); // adds 1 and 2
>
>I'd expect this result:
> assertEquals("Length is not four", ac.length, 4);
> assertEquals("First element not correct", ac[0], 2);
> assertEquals("Second element not correct", ac[1], 1);
> assertEquals("Third element not correct", ac[2], 1);
> assertEquals("Fourth element not correct", ac[3], 2);
>
>This also fails:
> addNumbers(); // adds 1 and 2
> var sort:Sort = new Sort();
> sort.fields = [new SortField(null, false, true, true)];
> ac.sort = sort;
> ac.refresh();
> ac.sort = null;
> ac.removeItemAt(0); // still sorted so 2 is removed leaving 1
>
>I'd expect this result:
> assertEquals("First element not correct", ac[0], 1);
>
>
>Justin
