RESOLVED FIXED62350
[Chromium] Forward shift + up/down key events to text field when autofill popup window is shown.
https://bugs.webkit.org/show_bug.cgi?id=62350
Summary [Chromium] Forward shift + up/down key events to text field when autofill pop...
Naoki Takano
Reported 2011-06-08 20:31:09 PDT
[Chromium] Forward shit + up/down key events to text field when autofill popup window is shown.
Attachments
Patch (3.34 KB, patch)
2011-06-08 21:08 PDT, Naoki Takano
no flags
Patch (3.43 KB, patch)
2011-06-13 22:56 PDT, Naoki Takano
no flags
Patch (3.46 KB, patch)
2011-06-14 18:22 PDT, Naoki Takano
no flags
Patch (3.48 KB, patch)
2011-06-14 18:54 PDT, Naoki Takano
no flags
Naoki Takano
Comment 1 2011-06-08 21:08:14 PDT
Naoki Takano
Comment 2 2011-06-08 21:13:40 PDT
Please review my patch.
Kent Tamura
Comment 3 2011-06-08 21:31:42 PDT
Comment on attachment 96540 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=96540&action=review > Source/WebCore/ChangeLog:5 > + [Chromium] Forward shit + up/down key events to text field when autofill popup window is shown. Why do we forward only these specific key combinations? No need to forward other keys? What about Command+cursor on Mac?
Naoki Takano
Comment 4 2011-06-08 21:47:37 PDT
Kent-san, Thank you for your review. Original bug report only refers to "shift + up". (I typo "shit" though, orz) Is there any other combination I have to consider? If shift/control/alt/command keys are appended, do I have to forward all up/down key events? (In reply to comment #3) > (From update of attachment 96540 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=96540&action=review > > > Source/WebCore/ChangeLog:5 > > + [Chromium] Forward shit + up/down key events to text field when autofill popup window is shown. > > Why do we forward only these specific key combinations? No need to forward other keys? > What about Command+cursor on Mac?
Kent Tamura
Comment 5 2011-06-08 22:45:42 PDT
(In reply to comment #4) > Kent-san, > > Thank you for your review. > > Original bug report only refers to "shift + up". (I typo "shit" though, orz) > > Is there any other combination I have to consider? > If shift/control/alt/command keys are appended, do I have to forward all up/down key events? I don't know. You should discuss it with Autofill team and Chrome UX team.
Naoki Takano
Comment 6 2011-06-08 22:47:47 PDT
I know Autofill team, David and Ilya. Actually I already added Ilya here. But I don't know who are in Chrome UX team. Do you know them? Thanks, (In reply to comment #5) > (In reply to comment #4) > > Kent-san, > > > > Thank you for your review. > > > > Original bug report only refers to "shift + up". (I typo "shit" though, orz) > > > > Is there any other combination I have to consider? > > If shift/control/alt/command keys are appended, do I have to forward all up/down key events? > > I don't know. You should discuss it with Autofill team and Chrome UX team.
Ilya Sherman
Comment 7 2011-06-08 23:07:56 PDT
(In reply to comment #4) > Kent-san, > > Thank you for your review. > > Original bug report only refers to "shift + up". (I typo "shit" though, orz) > > Is there any other combination I have to consider? > If shift/control/alt/command keys are appended, do I have to forward all up/down key events? > > (In reply to comment #3) > > (From update of attachment 96540 [details] [details]) > > View in context: https://bugs.webkit.org/attachment.cgi?id=96540&action=review > > > > > Source/WebCore/ChangeLog:5 > > > + [Chromium] Forward shit + up/down key events to text field when autofill popup window is shown. > > > > Why do we forward only these specific key combinations? No need to forward other keys? > > What about Command+cursor on Mac? If we want to be absolutely perfect, we should probably forward more key combinations. For example, on Mac, option-up/down should move the cursor without selecting text. On the other hand, the ideal behavior for ctrl-up/down is to jump to the top/bottom of the popup list. That's on Mac -- I'm not as sure what the behavior should be on the other platforms. That said, I am fine with forwarding either only key combinations involving shift, as shift-up/down is by far the most common cursor movement of this sort. I am also fine with forwarding all key presses with modifiers, and only intercepting literal up/down arrow presses. David, Glen, any thoughts to add into the mix?
Naoki Takano
Comment 8 2011-06-09 18:12:12 PDT
David and Glen, Could you give us your thoughts? Thanks, (In reply to comment #7) > (In reply to comment #4) > > Kent-san, > > > > Thank you for your review. > > > > Original bug report only refers to "shift + up". (I typo "shit" though, orz) > > > > Is there any other combination I have to consider? > > If shift/control/alt/command keys are appended, do I have to forward all up/down key events? > > > > (In reply to comment #3) > > > (From update of attachment 96540 [details] [details] [details]) > > > View in context: https://bugs.webkit.org/attachment.cgi?id=96540&action=review > > > > > > > Source/WebCore/ChangeLog:5 > > > > + [Chromium] Forward shit + up/down key events to text field when autofill popup window is shown. > > > > > > Why do we forward only these specific key combinations? No need to forward other keys? > > > What about Command+cursor on Mac? > > If we want to be absolutely perfect, we should probably forward more key combinations. For example, on Mac, option-up/down should move the cursor without selecting text. On the other hand, the ideal behavior for ctrl-up/down is to jump to the top/bottom of the popup list. That's on Mac -- I'm not as sure what the behavior should be on the other platforms. > > That said, I am fine with forwarding either only key combinations involving shift, as shift-up/down is by far the most common cursor movement of this sort. I am also fine with forwarding all key presses with modifiers, and only intercepting literal up/down arrow presses. > > David, Glen, any thoughts to add into the mix?
Glen Murphy
Comment 9 2011-06-09 19:58:30 PDT
Ilya's proposal of forwarding all keypresses with modifiers sounds good to me.
David Holloway
Comment 10 2011-06-10 08:47:22 PDT
(In reply to comment #9) > Ilya's proposal of forwarding all keypresses with modifiers sounds good to me. Yes, sounds good to me.
Naoki Takano
Comment 11 2011-06-10 09:53:24 PDT
Thanks, guys. I'll do that. (In reply to comment #10) > (In reply to comment #9) > > Ilya's proposal of forwarding all keypresses with modifiers sounds good to me. > > Yes, sounds good to me.
Naoki Takano
Comment 12 2011-06-13 22:56:36 PDT
Kent Tamura
Comment 13 2011-06-13 23:12:07 PDT
Comment on attachment 97071 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=97071&action=review > Source/WebCore/ChangeLog:11 > + * manual-tests/autofill-popup-shiftupdown.hml: Added to check modifier + up/down key is working correctly for text area in autofill popup is shown. This test is Chromium-specific. It should be put to WebCore/manual-tests/chromium/.
Naoki Takano
Comment 14 2011-06-14 18:22:41 PDT
Kent Tamura
Comment 15 2011-06-14 18:23:58 PDT
Comment on attachment 97211 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=97211&action=review > Source/WebCore/ChangeLog:11 > + Manual test: autofill-popup-shiftupdown.html > + > + * manual-tests/autofill-popup-shiftupdown.hml: Added to check modifier + up/down key is working correctly for text area in autofill popup is shown. Need to update ChangeLog
Naoki Takano
Comment 16 2011-06-14 18:54:04 PDT
Kent Tamura
Comment 17 2011-06-14 19:01:31 PDT
Comment on attachment 97219 [details] Patch Looks ok.
Naoki Takano
Comment 18 2011-06-14 19:09:39 PDT
Thank, kent-san!! (In reply to comment #17) > (From update of attachment 97219 [details]) > Looks ok.
WebKit Review Bot
Comment 19 2011-06-14 19:10:45 PDT
Comment on attachment 97219 [details] Patch Clearing flags on attachment: 97219 Committed r88895: <http://trac.webkit.org/changeset/88895>
WebKit Review Bot
Comment 20 2011-06-14 19:10:50 PDT
All reviewed patches have been landed. Closing bug.
Note You need to log in before you can comment on or make changes to this bug.