676195 - IME doesn't work with hardware keyboard on Android

Status: RESOLVED FIXED

(Core Graveyard :: Widget: Android, defect)

Description

[Makoto Kato [:m_kato]]

IME works well on software keyboard, but it doesn't on hardware keyboard.  It should work.

- Env
Milestone + OpenWnn
Softbank Mobile AQUOS PHONE THE HYBRID 007SH

Comment 1

[Makoto Kato [:m_kato]]

Attachment: WIP

Comment 2

[Makoto Kato [:m_kato]]

Attachment: fix v1

Comment 3

[(inactive) Jim Chen [:jchen] [:darchons]]

Comment on attachment 555709 [details] [diff] [review]
fix v1

>Bug 676195 - IME doesn't work with hardware keyboard on Android
>
>@@ -560,6 +560,10 @@
> 
>     @Override
>     public boolean onKeyDown(int keyCode, KeyEvent event) {
>+        return processKeyDown(keyCode, event, false);
>+    }
>+
>+    public boolean processKeyDown(int keyCode, KeyEvent event, boolean isPreIme) {

Can this be private?

>+
>+        if (isPreIme && mIMEState != IME_STATE_DISABLED) {
>+            // To support IME on hardware keyboard, onKeyPreIme has to return false.

How about "Let active IME process pre-IME key events"?

And the style is to not use braces on single-line statements

>@@ -608,6 +618,10 @@
> 
>     @Override
>     public boolean onKeyUp(int keyCode, KeyEvent event) {
>+        return processKeyUp(keyCode, event, false);
>+    }
>+
>+    public boolean processKeyUp(int keyCode, KeyEvent event, boolean isPreIme) {

Private?

>+
>+        if (isPreIme && mIMEState != IME_STATE_DISABLED) {
>+            // To support IME on hardware keyboard, onKeyPreIme has to return false.

See above about comment and style.

Thanks!

Comment 4

[Makoto Kato [:m_kato]]

Attachment: fix v2

Comment 5

[(inactive) Jim Chen [:jchen] [:darchons]]

Comment on attachment 557457 [details] [diff] [review]
fix v2

Thank you!

Comment 6

[Makoto Kato [:m_kato]]

http://hg.mozilla.org/integration/mozilla-inbound/rev/b3fa7b442186

Comment 7

[Ed Morley [:emorley]]

http://hg.mozilla.org/mozilla-central/rev/b3fa7b442186

Comment 8

[Makoto Kato [:m_kato]]

Comment on attachment 557457 [details] [diff] [review]
fix v2

requsting aurora.

In Japan mobile market, some devices with hardware keyboard are being released now.  But due to this bug, fennec cannot input any non-ascii string on these devices.

Comment 9

[Aaron Train [:aaronmt]]

This bug might have caused a regression in bug 685537 .

Comment 10

[JP Rosevear [:jpr]]

Need to know for sure if its causing the regression before we can approve this (or if we can live with the regression).  Mark is looking at this.

Comment 11

[Asa Dotzler [:asa]]

Comment on attachment 557457 [details] [diff] [review]
fix v2

drivers agree we'd like this (and the regression fix) both landed in Aurora.

Comment 12

[Makoto Kato [:m_kato]]

Comment on attachment 557457 [details] [diff] [review]
fix v2

cancel approval-aurora since related bug isn't in time until aurora.