| Summary: | Pause is needed while simulating mouse actions | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | cathiechen <cathiechen> | ||||||||||
| Component: | Tools / Tests | Assignee: | cathiechen <cathiechen> | ||||||||||
| Status: | RESOLVED FIXED | ||||||||||||
| Severity: | Normal | CC: | bburg, clopez, darin, ews-watchlist, fred.wang, gsnedders, simon.fraser, webkit-bug-importer, youennf | ||||||||||
| Priority: | P2 | Keywords: | InRadar | ||||||||||
| Version: | WebKit Nightly Build | ||||||||||||
| Hardware: | Unspecified | ||||||||||||
| OS: | Unspecified | ||||||||||||
| Attachments: |
|
||||||||||||
Created attachment 409589 [details]
Patch
This patch modifies the imported WPT tests. Please ensure that any changes on the tests (not coming from a WPT import) are exported to WPT. Please see https://trac.webkit.org/wiki/WPTExportProcess Comment on attachment 409589 [details]
Patch
This seems to have regressed a few tests according to EWS. Please investigate.
Created attachment 409713 [details]
Patch
(In reply to Brian Burg from comment #3) > Comment on attachment 409589 [details] > Patch > > This seems to have regressed a few tests according to EWS. Please > investigate. Done! Thanks Comment on attachment 409713 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=409713&action=review > LayoutTests/imported/w3c/web-platform-tests/resources/testdriver-vendor.js:10 > +// Provide an imprecise way of pause. This comment is not grammatically correct and it’s also unclear. I suppose it’s saying that there is something wrong with the implementation, but what? I suggest omitting the comment, or we could include it if we can pinpoint *what* we are trying to say and find a good way to word it. Is there something wrong with using setTimeout to pause? If so, what is it? Is there a problem with precision of the duration? Or is the problem that a timeout is queued behind other operations or on the wrong run loop? Maybe there is nothing wrong! > LayoutTests/imported/w3c/web-platform-tests/resources/testdriver-vendor.js:50 > + // FIXME: Provide a precise way of pause. Same thing here. Please just remove this comment or if you like I can try to help write one for whatever we mean here. Comment on attachment 409713 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=409713&action=review >> LayoutTests/imported/w3c/web-platform-tests/resources/testdriver-vendor.js:10 >> +// Provide an imprecise way of pause. > > This comment is not grammatically correct and it’s also unclear. I suppose it’s saying that there is something wrong with the implementation, but what? I suggest omitting the comment, or we could include it if we can pinpoint *what* we are trying to say and find a good way to word it. > > Is there something wrong with using setTimeout to pause? If so, what is it? Is there a problem with precision of the duration? Or is the problem that a timeout is queued behind other operations or on the wrong run loop? Maybe there is nothing wrong! Thanks! I'll remove them. Yeah, looks like nothing is wrong. I added the comments because setTimeout seems not as precise as the pause in dispatchTouchActions. Created attachment 409824 [details]
Patch
Created attachment 409825 [details]
Patch
Committed r267656: <https://trac.webkit.org/changeset/267656> All reviewed patches have been landed. Closing bug and clearing flags on attachment 409825 [details]. |
Test code like below would not work if pause is not supported. test_driver.Actions() .addPointer("pointer1", "mouse") .pointerMove(0, 0, { origin: target }) .pointerDown() .pointerMove(originpos.x, originpos.y, { origin: target }) .pointerMove(originpos.x + delta.x / 2, originpos.y + delta.y / 2, { origin: target }) .pointerMove(originpos.x + delta.x, originpos.y + delta.y, { origin: target }) .pause(duriation) .pointerUp() .send();