Support "Set input mode to X" commands on Linux

[GoogleCodeExporter]

Currently "InputModeX" commands ("InputModeSwitchKanaType", 
"InputModeHiragana", etc...) have been enabled only on Windows as follows.
http://code.google.com/p/mozc/source/browse/trunk/src/session/internal/keymap.cc
?r=93#581
http://code.google.com/p/mozc/source/browse/trunk/src/session/internal/keymap.cc
?r=93#691
http://code.google.com/p/mozc/source/browse/trunk/src/session/internal/keymap.cc
?r=93#780

Although we have intentionally disabled these features on Mac OS X due to some 
UX reason, current ibus-mozc seems to be ready to support these commands.

FYI, this issue was originally reported in Issue 114 and 
https://bugzilla.redhat.com/show_bug.cgi?id=771799.
I made this thread because this limitation on Linux has been implemented in the 
converter engine itself and not related to the language-bar nor ibus-mozc.

What steps will reproduce the problem?
1. Build Mozc on Linux.
2. Create a custom keymap with any combination of following commands and input 
states.
- On Precomposition (入力文字なし)
-- "InputModeSwitchKanaType" (次のかな文字種に入力切替)
- On Precomposition (入力文字なし) or Composition (変換前入力中) or 
Conversion (変換中)
-- "InputModeHiragana" (ひらがなに入力切替)
-- "InputModeFullKatakana" (全角カタカナに入力切替)
-- "InputModeHalfKatakana" (半角カタカナに入力切替)
-- "InputModeFullAlphanumeric" (全角英数に入力切替)
-- "InputModeHalfAlphanumeric" (半角英数に入力切替)
3. Type the shortcut key on such input state.

What is the expected output? What do you see instead?
Functional.

Please use labels and text to provide additional information.
Nothing happens.

Original issue reported on code.google.com by yukawa@google.com on 28 Feb 2012 at 7:10

[GoogleCodeExporter]

Original comment by yukawa@google.com on 28 Feb 2012 at 4:22

• Changed title: Support "Set input mode to X" commands on Linux

[GoogleCodeExporter]

Updated the repro steps by using English UI messages instead of internal 
command names just for readability and searchability.

What steps will reproduce the problem?
1. Build Mozc on Linux.
2. Create a custom keymap with any combination of following commands and input 
states.
- On Precomposition (入力文字なし)
-- "Set input mode to next kana type" (次のかな文字種に入力切替)
- On Precomposition (入力文字なし) or Composition (変換前入力中) or 
Conversion (変換中)
-- "Set input mode to Hiragana" (ひらがなに入力切替)
-- "Set input mode to full-width katakana" (全角カタカナに入力切替)
-- "Set input mode to half-width katakana" (半角カタカナに入力切替)
-- "Set input mode to full-width alphanumeric" (全角英数に入力切替)
-- "Set input mode to half-width alphanumeric" (半角英数に入力切替)
3. Type the shortcut key on such input state.

What is the expected output? What do you see instead?
Functional.

Please use labels and text to provide additional information.

Original comment by yukawa@google.com on 28 Feb 2012 at 4:28

[GoogleCodeExporter]

The converter-side issue has been fixed in r96 (1.4.1003.102).
I believe that now you can use the following key mappings you reported in Issue 
114.
- Composition/Precomposition | Muhenkan | Switch kana type > Not working.
- Precomposition | Ctl+. | Switch kana type > Not working

Unfortunately, you may find that the following key mapping you reported in 
Issue 114 is still not functional.
- Composition/Precomposition | Hiragana | Set input mode to Hiragana

This is because our client-side code (mozc.el, ibus-mozc, scim-mozc and 
key-binding editor on the config dialog) still cannot handle 
"Hiragana/Katakana" key event in some cases, as filed in Issue 136 and Issue 
137. We are planning to fix Issue 136 and Issue 137 in the next push.

Anyway, I'd like to close this issue because server-side implementation has 
been finished. Please feel free to reopen this issue if you find any "Set input 
mode to X" command is still not functional in r96 (1.4.1003.102) or later.

Thank you for thank you for taking your time to report this.

Original comment by yukawa@google.com on 13 Mar 2012 at 2:20

• Changed state: Fixed

[GoogleCodeExporter]

Thank you in return.
I am looking forward to try the update, as soon as it is available in the 
update testing of my distribution.
I will keep updated with issue 136, 137, and I am already looking for the next 
push.
I definitely like using mozec over anthy for Japanese input.

Original comment by nodata...@gmail.com on 17 Mar 2012 at 2:07

[GoogleCodeExporter]

Issue 114 has been merged into this issue.

Original comment by yukawa@google.com on 18 Mar 2012 at 11:21

[GoogleCodeExporter]

Issue 114 has been merged into this issue.

Original comment by yukawa@google.com on 21 Jan 2015 at 8:52