refactor(tests): move canvas events tests from qunit to vitest #10563
Conversation
Review or Edit in CodeSandboxOpen the branch in Web Editor • VS Code • Insiders |
|
|
||
| canvas.on('mouse:down', function (options) { | ||
| // TODO: fix this, for some reason first target on mouse down is active selection and then second target is the green rectangle | ||
| if (options.target instanceof ActiveSelection) { |
There was a problem hiding this comment.
@asturur I don't know how this test worked in qunit but in both jsdom and in the browser the first event that comes here is Active selection and then the green box
I guess you will have to have a look here since you have far more knowledge than I do when it comes to fabric source code
for now I workarounded this with a return if it is now a green box but clearly this is not a solution
|
Build Stats
|
| "mousemove: subTargetCheck: setCursorFromEvent considers subTargets in reverse order, so the top-most subTarget's .hoverCursor takes precedence", | ||
| ); | ||
|
|
||
| // TODO: ported from qunit, check if is still valid for vitest |
There was a problem hiding this comment.
I ported this thing as well but I don't know if it is relevant
| expect(getTarget(), 'should unregister b').toBeFalsy(); | ||
| }); | ||
|
|
||
| it.todo('mousemove: subTargetCheck: setCursorFromEvent considers subTargets'); |
There was a problem hiding this comment.
I also ported these two as test todos so it doesn't get lost in case it is something that should be actually done
|
this concludes the porting of tests from qunit to vitest after that is done I will work on making tests run on firefox and webkit as well @asturur maybe you could do the cleanup of the qunit setup for unit tests? |
|
i ll try to clean up yes |
| // expect(!!canvas.actionIsDisabled('br', target, e), 'br action is not disabled lockSkewingX').toBe(false); | ||
| // expect(!!canvas.actionIsDisabled('mtr', target, e), 'mtr action is not disabled lockSkewingX').toBe(false); | ||
| // }); | ||
| }); |
There was a problem hiding this comment.
right there was this disabled action logic that was never ported to the action themselves i guess. I don't remember very well, controls could be refactored twice and still be sub optimal.
trying to have both a configuration approach and a code approach to default actions and custom actions has proved to be both confusing and hard.
|
i may have merged too quick. i ll check and fix in case |
this is the last unit test in qunit that was left
there is a todo in the test in one test which I will comment inline