Keyboard events with keypad had wrong keycodes on ozone.

Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.

Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the
xkb_keyboard_layout_engine.cc part of ozone but for
simplicity sake and that this needs to be backported to M48; restore some of
https://codereview.chromium.org/1308063007 CL with respect to rewrite numpads based events.

BUG=570388
Committed: https://crrev.com/f3c39ae68b3c462a7918b8ae092e6d322c418047
Cr-Commit-Position: refs/heads/master@{#367418}

[dtapuska, 2015-12-29 17:44:56]

Description was changed from

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code caused this issue. Adding back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.

BUG=570388
==========

to

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.
Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the xkb_keyboard_layout_engine.cc part of
ozone but for simplicity sake and that this needs to be backported to M48;
restore some of https://codereview.chromium.org/1308063007 CL with respect to
rewrite numpads based events.

BUG=570388
==========

[dtapuska, 2015-12-29 17:45:20]

Description was changed from

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.
Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the xkb_keyboard_layout_engine.cc part of
ozone but for simplicity sake and that this needs to be backported to M48;
restore some of https://codereview.chromium.org/1308063007 CL with respect to
rewrite numpads based events.

BUG=570388
==========

to

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.

Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the xkb_keyboard_layout_engine.cc part of
ozone but for 
simplicity sake and that this needs to be backported to M48; restore some of
https://codereview.chromium.org/1308063007 CL with respect to rewrite numpads
based events.

BUG=570388
==========

[dtapuska, 2015-12-29 17:45:41]

Description was changed from

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.

Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the xkb_keyboard_layout_engine.cc part of
ozone but for 
simplicity sake and that this needs to be backported to M48; restore some of
https://codereview.chromium.org/1308063007 CL with respect to rewrite numpads
based events.

BUG=570388
==========

to

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.

Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the 
xkb_keyboard_layout_engine.cc part of ozone but for 
simplicity sake and that this needs to be backported to M48; restore some of 
https://codereview.chromium.org/1308063007 CL with respect to rewrite numpads
based events.

BUG=570388
==========

[dtapuska, 2015-12-29 17:54:19]

dtapuska@chromium.org changed reviewers:
+ kpschoedel@chromium.org, sadrul@chromium.org, wez@chromium.org

[dtapuska, 2015-12-29 17:56:11]

I've created issue https://code.google.com/p/chromium/issues/detail?id=572942

to cover the actual bug with ozone.

But the ask was to fix this for citrix users and this I felt was the least
possible impact.

[Wez, 2016-01-04 19:52:27]

On 2015/12/29 at 17:56:11, dtapuska wrote:
> I've created issue https://code.google.com/p/chromium/issues/detail?id=572942
> 
> to cover the actual bug with ozone.
> 
> But the ask was to fix this for citrix users and this I felt was the least
possible impact.

This looks fine, but do we actually need the
NonLocatedToLocatedKeypadKeyboardCode() function to be split-out - do we have
reason to believe that calling NonLocatedToLocatedKeyboardCode() and keeping it
as one function would cause problems with Shift, Control, etc?

[Wez, 2016-01-04 19:53:01]

On 2016/01/04 at 19:52:27, Wez wrote:
> On 2015/12/29 at 17:56:11, dtapuska wrote:
> > I've created issue
https://code.google.com/p/chromium/issues/detail?id=572942
> > 
> > to cover the actual bug with ozone.
> > 
> > But the ask was to fix this for citrix users and this I felt was the least
possible impact.
> 
> This looks fine, but do we actually need the
NonLocatedToLocatedKeypadKeyboardCode() function to be split-out - do we have
reason to believe that calling NonLocatedToLocatedKeyboardCode() and keeping it
as one function would cause problems with Shift, Control, etc?

LGTM assuming that the function split is necessary.

[sadrul, 2016-01-04 20:02:29]

stamp lgtm

[dtapuska, 2016-01-04 20:09:04]

On 2016/01/04 19:52:27, Wez wrote:
> On 2015/12/29 at 17:56:11, dtapuska wrote:
> > I've created issue
https://code.google.com/p/chromium/issues/detail?id=572942
> > 
> > to cover the actual bug with ozone.
> > 
> > But the ask was to fix this for citrix users and this I felt was the least
> possible impact.
> 
> This looks fine, but do we actually need the
> NonLocatedToLocatedKeypadKeyboardCode() function to be split-out - do we have
> reason to believe that calling NonLocatedToLocatedKeyboardCode() and keeping
it
> as one function would cause problems with Shift, Control, etc?

Yes; issue 362551 would be caused again if that occurred.

[dtapuska, 2016-01-04 20:53:25]

The CQ bit was checked by dtapuska@chromium.org

[commit-bot: I haz the power, 2016-01-04 20:54:42]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1550953002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1550953002/1

[commit-bot: I haz the power, 2016-01-04 23:57:02]

Description was changed from

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.

Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the 
xkb_keyboard_layout_engine.cc part of ozone but for 
simplicity sake and that this needs to be backported to M48; restore some of 
https://codereview.chromium.org/1308063007 CL with respect to rewrite numpads
based events.

BUG=570388
==========

to

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.

Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the 
xkb_keyboard_layout_engine.cc part of ozone but for 
simplicity sake and that this needs to be backported to M48; restore some of 
https://codereview.chromium.org/1308063007 CL with respect to rewrite numpads
based events.

BUG=570388
==========

[commit-bot: I haz the power, 2016-01-04 23:57:03]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-01-04 23:58:38]

Description was changed from

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.

Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the 
xkb_keyboard_layout_engine.cc part of ozone but for 
simplicity sake and that this needs to be backported to M48; restore some of 
https://codereview.chromium.org/1308063007 CL with respect to rewrite numpads
based events.

BUG=570388
==========

to

==========
Keyboard events with keypad had wrong keycodes on ozone.

Ozone is generating virtual keys not like Windows and X11 are generating
them. The removal of some code undercovered the issue. Add back some
of the partial conversion code (numpad conversions only) to address
the issue for the users. Whether numpad keys are "located" or not
needs to be re-evaluated and all platforms made consistent.
Documentation claims that numpad events are located but the
platforms generate them to be consistent with windows.

Ozone seemed to be the only platform that didn't generate
them. Ideally the fix should be in the
xkb_keyboard_layout_engine.cc part of ozone but for
simplicity sake and that this needs to be backported to M48; restore some of
https://codereview.chromium.org/1308063007 CL with respect to rewrite numpads
based events.

BUG=570388

Committed: https://crrev.com/f3c39ae68b3c462a7918b8ae092e6d322c418047
Cr-Commit-Position: refs/heads/master@{#367418}
==========

[commit-bot: I haz the power, 2016-01-04 23:58:38]

Patchset 1 (id:??) landed as
https://crrev.com/f3c39ae68b3c462a7918b8ae092e6d322c418047
Cr-Commit-Position: refs/heads/master@{#367418}