[PATCH 0/4] MR3115: imm32: Fixes for comp string on character deletion and result commit.

[Rémi Bernon (＠rbernon)]

I believe this fixes some bad behavior, and it includes one fix from https://gitlab.winehq.org/wine/wine/-/merge_requests/2698 to avoid duplicated entries (though doesn't supersedes that MR).

[PATCH 1/4] imm32/tests: Test how deleting a character can behave.

[Rémi Bernon]

From: Rémi Bernon

---
 dlls/imm32/tests/imm32.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/dlls/imm32/tests/imm32.c b/dlls/imm32/tests/imm32.c
index 95f832b97e4..07caa054357 100644
--- a/dlls/imm32/tests/imm32.c
+++ b/dlls/imm32/tests/imm32.c
@@ -7213,6 +7213,20 @@ static void test_ga_na_da(void)
     /* These sequences have some additional WM_IME_NOTIFY messages with unknown wparam > IMN_PRIVATE */
     struct ime_call complete_seq[] =
     {
+        /* G */
+        {.hkl = default_hkl, .himc = 0/*himc*/, .func = MSG_TEST_WIN, .message = {.msg = WM_IME_STARTCOMPOSITION, .wparam = 0, .lparam = 0}},
+        {
+            .hkl = default_hkl, .himc = 0/*himc*/, .func = MSG_TEST_WIN, .comp = L"\u3131", .result = L"",
+            .message = {.msg = WM_IME_COMPOSITION, .wparam = 0x3131, .lparam = GCS_COMPSTR|GCS_COMPATTR|CS_INSERTCHAR|CS_NOMOVECARET},
+        },
+
+        {
+            .hkl = default_hkl, .himc = 0/*himc*/, .func = MSG_TEST_WIN, .comp = L"", .result = L"",
+            .message = {.msg = WM_IME_COMPOSITION, .wparam = 0x1b, .lparam = GCS_CURSORPOS|GCS_DELTASTART|GCS_COMPSTR|GCS_COMPATTR|GCS_COMPCLAUSE|
+                                                                             GCS_COMPREADSTR|GCS_COMPREADATTR|GCS_COMPREADCLAUSE},
+        },
+        {.hkl = default_hkl, .himc = 0/*himc*/, .func = MSG_TEST_WIN, .message = {.msg = WM_IME_ENDCOMPOSITION, .wparam = 0, .lparam = 0}},
+
         /* G */
         {.hkl = default_hkl, .himc = 0/*himc*/, .func = MSG_TEST_WIN, .message = {.msg = WM_IME_STARTCOMPOSITION, .wparam = 0, .lparam = 0}},
         {
@@ -7377,6 +7391,14 @@ static void test_ga_na_da(void)
     flush_events();
     keybd_event( 'R', 0x13, KEYEVENTF_KEYUP, 0 );

+    keybd_event( VK_BACK, 0x0e, 0, 0 );
+    flush_events();
+    keybd_event( VK_BACK, 0x0e, KEYEVENTF_KEYUP, 0 );
+
+    keybd_event( 'R', 0x13, 0, 0 );
+    flush_events();
+    keybd_event( 'R', 0x13, KEYEVENTF_KEYUP, 0 );
+
     keybd_event( 'K', 0x25, 0, 0 );
     flush_events();
     keybd_event( 'K', 0x25, KEYEVENTF_KEYUP, 0 );

[PATCH 2/4] imm32: Use offsets instead of lengths to decide on sending updates.

[Rémi Bernon]

From: Rémi Bernon

The composition or result strings may be present, but with zero length.
In which case we still want to send the messages, to indicate any change
for instance whenever a character is deleted and strings become empty.

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=55055
---
 dlls/imm32/ime.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/dlls/imm32/ime.c b/dlls/imm32/ime.c
index b5c4fc5b2e1..0470f3b6d6c 100644
--- a/dlls/imm32/ime.c
+++ b/dlls/imm32/ime.c
@@ -559,7 +559,7 @@ UINT WINAPI ImeToAsciiEx( UINT vkey, UINT vsc, BYTE *state, TRANSMSGLIST *msgs,
         TRANSMSG status_msg = {.message = ime_set_composition_status( himc, !!compstr->dwCompStrOffset )};
         if (status_msg.message) msgs->TransMsg[count++] = status_msg;

-        if (compstr->dwResultStrLen)
+        if (compstr->dwResultStrOffset)
         {
             const WCHAR *result = (WCHAR *)((BYTE *)compstr + compstr->dwResultStrOffset);
             TRANSMSG msg = {.message = WM_IME_COMPOSITION, .wParam = result[0], .lParam = GCS_RESULTSTR};
@@ -567,7 +567,7 @@ UINT WINAPI ImeToAsciiEx( UINT vkey, UINT vsc, BYTE *state, TRANSMSGLIST *msgs,
             msgs->TransMsg[count++] = msg;
         }

-        if (compstr->dwCompStrLen)
+        if (compstr->dwCompStrOffset)
         {
             const WCHAR *comp = (WCHAR *)((BYTE *)compstr + compstr->dwCompStrOffset);
             TRANSMSG msg = {.message = WM_IME_COMPOSITION, .wParam = comp[0], .lParam = GCS_COMPSTR | GCS_CURSORPOS | GCS_DELTASTART};

[PATCH 3/4] winex11: Report empty preedit string when result string is committed.

[Rémi Bernon]

From: Rémi Bernon

Based on a patch from Byeong-Sik Jeon .

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=55027
---
 dlls/winex11.drv/xim.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/dlls/winex11.drv/xim.c b/dlls/winex11.drv/xim.c
index 1c3d2dd9875..1b063cfe886 100644
--- a/dlls/winex11.drv/xim.c
+++ b/dlls/winex11.drv/xim.c
@@ -140,7 +140,7 @@ void xim_set_result_string( HWND hwnd, const char *str, UINT count )
     len = ntdll_umbstowcs( str, count, output, count );
     output[len] = 0;

-    post_ime_update( hwnd, 0, ime_comp_buf, output );
+    post_ime_update( hwnd, 0, NULL, output );

free( output );
 }

[PATCH 4/4] winex11: Make sure HIMC is opened before sending IME updates.

[Rémi Bernon]

From: Rémi Bernon

---
 dlls/winex11.drv/xim.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/dlls/winex11.drv/xim.c b/dlls/winex11.drv/xim.c
index 1b063cfe886..209d63f0402 100644
--- a/dlls/winex11.drv/xim.c
+++ b/dlls/winex11.drv/xim.c
@@ -175,6 +175,7 @@ static int xic_preedit_start( XIC xic, XPointer user, XPointer arg )
     if ((ime_comp_buf = realloc( ime_comp_buf, sizeof(WCHAR) ))) *ime_comp_buf = 0;
     else ERR( "Failed to allocate preedit buffer\n" );

+    NtUserPostMessage( hwnd, WM_IME_NOTIFY, IMN_WINE_SET_OPEN_STATUS, TRUE );
     post_ime_update( hwnd, 0, ime_comp_buf, NULL );

return -1;
@@ -190,6 +191,7 @@ static int xic_preedit_done( XIC xic, XPointer user, XPointer arg )
     ime_comp_buf = NULL;

post_ime_update( hwnd, 0, NULL, NULL );
+    NtUserPostMessage( hwnd, WM_IME_NOTIFY, IMN_WINE_SET_OPEN_STATUS, FALSE );

return 0;
 }

[PATCH 4/4] winex11: Make sure HIMC is opened before sending IME updates.

[Marvin]

Hi,

It looks like your patch introduced the new failures shown below.
Please investigate and fix them before resubmitting your patch.
If they are not new, fixing them anyway would help a lot. Otherwise
please ask for the known failures list to be updated.

The tests also ran into some preexisting test failures. If you know how
to fix them that would be helpful. See the TestBot job for the details:

The full results can be found at:
https://testbot.winehq.org/JobDetails.pl?Key=133963

Your paranoid android.

=== debian11 (32 bit report) ===

Report validation errors:
quartz:vmr7 has no test summary line (early exit of the main process?)
quartz:vmr7 has unaccounted for todo messages
quartz:vmr7 has unaccounted for skip messages

[Masahito Suzuki (＠firelzrd)]

Great job, and thank you, @rbernon for your contribution.  
I did quick tests on the built version, and here's the result.

### Problem Type-1
Input: a{Enter}  
Expected: あ  
Actual: ああ

Status (Wordpad): Fixed by these commits.  
Status (Excel): Fixed by these commits.  
Status (VBA): Remains to always happen.

### Problem Type-2
Input: nihongo n  
Expected: 日本語ｎ  
Actual: 日本語

Status: Remains to happen at some probability.

### Problem Type-3
Input: aa{BackSpace}{BackSpace}  
Expected:  
Actual: あ

Status: Fixed by these commits.

[Masahito Suzuki (＠firelzrd)]

The above result shows that these patches beautifully addressed at least Type-1(partially) and Type-3 input string issues.  
And I think that fact alone is already enough to merge them to the upcoming release.