[RFC] vim-patch:7.4.{569,573} #2621

glts · 2015-05-09T12:52:18Z

Problem:    Having CTRL-C interrupt or not does not check the mode of the
        mapping. (Ingo Karkat)
Solution:   Use a bitmask with the map mode. (Christian Brabandt)

vim/vim@v7-4-569

Original patch:

diff --git a/src/nvim/getchar.c b/src/nvim/getchar.c
--- a/src/nvim/getchar.c
+++ b/src/nvim/getchar.c
@@ -3708,8 +3708,13 @@ do_map(maptype, arg, mode, abbrev)
    if (!did_it)
        retval = 2;             /* no match */
    else if (*keys == Ctrl_C)
+   {
        /* If CTRL-C has been unmapped, reuse it for Interrupting. */
-       mapped_ctrl_c = FALSE;
+       if (map_table == curbuf->b_maphash)
+       curbuf->b_mapped_ctrl_c &= ~mode;
+       else
+       mapped_ctrl_c &= ~mode;
+   }
    goto theend;
     }

@@ -3744,7 +3749,12 @@ do_map(maptype, arg, mode, abbrev)

     /* If CTRL-C has been mapped, don't always use it for Interrupting. */
     if (*keys == Ctrl_C)
-   mapped_ctrl_c = TRUE;
+    {
+   if (map_table == curbuf->b_maphash)
+       curbuf->b_mapped_ctrl_c |= mode;
+   else
+       mapped_ctrl_c |= mode;
+    }

     mp->m_keys = vim_strsave(keys);
     mp->m_str = vim_strsave(rhs);
diff --git a/src/nvim/globals.h b/src/nvim/globals.h
--- a/src/nvim/globals.h
+++ b/src/nvim/globals.h
@@ -958,7 +958,7 @@ EXTERN char_u   *exe_name;      /* the name of
 #ifdef USE_ON_FLY_SCROLL
 EXTERN int dont_scroll INIT(= FALSE);/* don't use scrollbars when TRUE */
 #endif
-EXTERN int mapped_ctrl_c INIT(= FALSE); /* CTRL-C is mapped */
+EXTERN int mapped_ctrl_c INIT(= FALSE); /* modes where CTRL-C is mapped */
 EXTERN int ctrl_c_interrupts INIT(= TRUE); /* CTRL-C sets got_int */

 EXTERN cmdmod_T    cmdmod;         /* Ex command modifiers */
diff --git a/src/nvim/structs.h b/src/nvim/structs.h
--- a/src/nvim/structs.h
+++ b/src/nvim/structs.h
@@ -1802,6 +1802,7 @@ struct file_buffer
     cryptstate_T *b_cryptstate;    /* Encryption state while reading or writing
                 * the file. NULL when not using encryption. */
 #endif
+    int        b_mapped_ctrl_c; /* modes where CTRL-C is mapped */

 }; /* file_buffer */

diff --git a/src/nvim/testdir/test_mapping.in b/src/nvim/testdir/test_mapping.in
--- a/src/nvim/testdir/test_mapping.in
+++ b/src/nvim/testdir/test_mapping.in
@@ -8,6 +8,15 @@ STARTTEST
 :inoreab чкпр   vim
 GAчкпр 
 �
+:" mapping of ctrl-c in insert mode
+:set cpo-=< cpo-=k
+:inoremap <c-c> <ctrl-c>
+:cnoremap <c-c> dummy
+:cunmap <c-c>
+GA
+TEST2: CTRL-C |�A|
+�
+:nunmap <c-c>

 : " langmap should not get remapped in insert mode
 :inoremap { FAIL_ilangmap
diff --git a/src/nvim/testdir/test_mapping.ok b/src/nvim/testdir/test_mapping.ok
--- a/src/nvim/testdir/test_mapping.ok
+++ b/src/nvim/testdir/test_mapping.ok
@@ -1,4 +1,6 @@
 test starts here:
 vim
+TEST2: CTRL-C |<ctrl-c>A|
+
 +
 +
diff --git a/src/nvim/ui.c b/src/nvim/ui.c
--- a/src/nvim/ui.c
+++ b/src/nvim/ui.c
@@ -180,7 +180,7 @@ ui_inchar(buf, maxlen, wtime, tb_change_

    /* ... there is no need for CTRL-C to interrupt something, don't let
     * it set got_int when it was mapped. */
-   if (mapped_ctrl_c)
+   if ((mapped_ctrl_c | curbuf->b_mapped_ctrl_c) & State)
        ctrl_c_interrupts = FALSE;
     }

diff --git a/src/nvim/version.c b/src/nvim/version.c
--- a/src/nvim/version.c
+++ b/src/nvim/version.c
@@ -742,6 +742,8 @@ static char *(features[]) =
 static int included_patches[] =
 {   /* Add new patch number below this line */
 /**/
+    569,
+/**/
     568,
 /**/
     567,

Problem:    Mapping CTRL-C in Visual mode doesn't work. (Ingo Karkat)
Solution:   Call get_real_state() instead of using State directly.

vim/vim@v7-4-573

Original patch:

diff --git a/src/nvim/testdir/test_mapping.in b/src/nvim/testdir/test_mapping.in
--- a/src/nvim/testdir/test_mapping.in
+++ b/src/nvim/testdir/test_mapping.in
@@ -8,7 +8,7 @@ STARTTEST
 :inoreab чкпр   vim
 GAчкпр 
 �
-:" mapping of ctrl-c in insert mode
+:" mapping of ctrl-c in Insert mode
 :set cpo-=< cpo-=k
 :inoremap <c-c> <ctrl-c>
 :cnoremap <c-c> dummy
@@ -16,9 +16,15 @@ GAчкпр
 GA
 TEST2: CTRL-C |�A|
 �
-:nunmap <c-c>
-
-: " langmap should not get remapped in insert mode
+:unmap <c-c>
+:unmap! <c-c>
+:"
+:" mapping of ctrl-c in Visual mode
+:vnoremap <c-c> :<C-u>$put ='vmap works'
+GV�
+:vunmap <c-c>
+:"
+:" langmap should not get remapped in insert mode
 :inoremap { FAIL_ilangmap
 :set langmap=+{ langnoremap
 o+�
diff --git a/src/nvim/testdir/test_mapping.ok b/src/nvim/testdir/test_mapping.ok
--- a/src/nvim/testdir/test_mapping.ok
+++ b/src/nvim/testdir/test_mapping.ok
@@ -2,5 +2,6 @@ test starts here:
 vim
 TEST2: CTRL-C |<ctrl-c>A|

+vmap works
 +
 +
diff --git a/src/nvim/ui.c b/src/nvim/ui.c
--- a/src/nvim/ui.c
+++ b/src/nvim/ui.c
@@ -180,7 +180,7 @@ ui_inchar(buf, maxlen, wtime, tb_change_

    /* ... there is no need for CTRL-C to interrupt something, don't let
     * it set got_int when it was mapped. */
-   if ((mapped_ctrl_c | curbuf->b_mapped_ctrl_c) & State)
+   if ((mapped_ctrl_c | curbuf->b_mapped_ctrl_c) & get_real_state())
        ctrl_c_interrupts = FALSE;
     }

diff --git a/src/nvim/version.c b/src/nvim/version.c
--- a/src/nvim/version.c
+++ b/src/nvim/version.c
@@ -742,6 +742,8 @@ static char *(features[]) =
 static int included_patches[] =
 {   /* Add new patch number below this line */
 /**/
+    573,
+/**/
     572,
 /**/
     571,

glts · 2015-05-09T13:53:32Z

src/nvim/terminal.c

Before, mapped_ctrl_c was a Boolean value, where true indicated that Ctrl-C was mapped and false that it wasn't mapped. Now that mapped_ctrl_c is an int with mode flags, the true case becomes 'all flags turned on', which I think can be expressed as MAP_ALL_MODES. Correct?

(Why is Ctrl-C always 'mapped' when entering the terminal though? I don't understand this.)

glts · 2015-05-09T18:12:49Z

Test failure.

It passed locally but now I see there is some indeterminism in this test. Look at the failing test results, they're not all the same. Hmm.

glts · 2015-05-10T12:02:51Z

Still failing! Frustrating. Does anybody know why these tests fail and how to make them pass in the CI? It's quite bad with these brittle functional tests at the moment. Is this a fundamental problem in the test infrastructure?

jszakmeister · 2015-05-10T13:08:31Z

It looks like the quickbuild-related issue (with the functional tests timing out) could be fixed if you rebased on master. The tests were hanging on QB because a necessary executable was not being built, and the tests didn't really do well picking up on that.

I wish I knew about the Travis ones. I know we've got a couple that are a bit racy, but I'm not sure what's happening here. :-(

glts · 2015-05-10T16:41:39Z

blimey the test passed on Travis 🍸 No idea why.

Now on to clean up the mess I made.

Problem: Having CTRL-C interrupt or not does not check the mode of the mapping. (Ingo Karkat) Solution: Use a bitmask with the map mode. (Christian Brabandt) vim/vim@v7-4-569

Problem: Mapping CTRL-C in Visual mode doesn't work. (Ingo Karkat) Solution: Call get_real_state() instead of using State directly. vim/vim@v7-4-573 The code change in src/ui.c was already applied in src/nvim/os/input.c in patch 7.4.569. Also split monolithic test into four "it" blocks because of persistent non-deterministic test failures on CI.

glts · 2015-05-26T17:55:20Z

Rebased, no changes.

felipecrv · 2015-05-29T05:25:56Z

LGTM.

justinmk · 2015-06-14T01:43:20Z

restarted build (in case anyone missed it: travis infrastructure problems appear to be the cause of the many build failures in May; they have been running smoothly since then)

Shougo · 2015-12-19T03:05:44Z

@glts Can you fix the conflicts?

Problem: Having CTRL-C interrupt or not does not check the mode of the mapping. (Ingo Karkat) Solution: Use a bitmask with the map mode. (Christian Brabandt) vim/vim@651863c Problem: Mapping CTRL-C in Visual mode doesn't work. (Ingo Karkat) Solution: Call get_real_state() instead of using State directly. vim/vim@5000869

Shougo · 2015-12-19T03:55:17Z

@glts I fixed the conflicts in #3866.
Thanks.

Problem: Having CTRL-C interrupt or not does not check the mode of the mapping. (Ingo Karkat) Solution: Use a bitmask with the map mode. (Christian Brabandt) vim/vim@651863c Problem: Mapping CTRL-C in Visual mode doesn't work. (Ingo Karkat) Solution: Call get_real_state() instead of using State directly. vim/vim@5000869

@glts

vim-patch:7.4.569 vim-patch:7.4.573 Helped-by: @glts neovim#2621 Problem: Having CTRL-C interrupt or not does not check the mode of the mapping. (Ingo Karkat) Solution: Use a bitmask with the map mode. (Christian Brabandt) vim/vim@651863c Problem: Mapping CTRL-C in Visual mode doesn't work. (Ingo Karkat) Solution: Call get_real_state() instead of using State directly. vim/vim@5000869

marvim added the WIP label May 9, 2015

glts reviewed May 9, 2015
View reviewed changes

glts changed the title ~~[WIP] vim-patch:7.4.{569,573}~~ [RFC] vim-patch:7.4.{569,573} May 9, 2015

marvim added RFC and removed WIP labels May 9, 2015

glts changed the title ~~[RFC] vim-patch:7.4.{569,573}~~ [WIP] vim-patch:7.4.{569,573} May 9, 2015

marvim added WIP and removed RFC labels May 9, 2015

glts changed the title ~~[WIP] vim-patch:7.4.{569,573}~~ [RFC] vim-patch:7.4.{569,573} May 10, 2015

marvim added RFC and removed WIP labels May 10, 2015

glts added 2 commits May 26, 2015 19:51

vim-patch:7.4.569

77f3eb9

Problem: Having CTRL-C interrupt or not does not check the mode of the mapping. (Ingo Karkat) Solution: Use a bitmask with the map mode. (Christian Brabandt) vim/vim@v7-4-569

Shougo mentioned this pull request Dec 19, 2015

[RFC] vim-patch:7.4.569, 7.4.573 #3866

Closed

justinmk closed this Dec 26, 2015

justinmk removed the RFC label Dec 26, 2015

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

[RFC] vim-patch:7.4.{569,573} #2621

[RFC] vim-patch:7.4.{569,573} #2621

Uh oh!

glts commented May 9, 2015

Uh oh!

glts May 9, 2015

Uh oh!

glts commented May 9, 2015

Uh oh!

glts commented May 10, 2015

Uh oh!

jszakmeister commented May 10, 2015

Uh oh!

glts commented May 10, 2015

Uh oh!

glts commented May 26, 2015

Uh oh!

felipecrv commented May 29, 2015

Uh oh!

justinmk commented Jun 14, 2015

Uh oh!

Shougo commented Dec 19, 2015

Uh oh!

Shougo commented Dec 19, 2015

Uh oh!

Uh oh!

Uh oh!

[RFC] vim-patch:7.4.{569,573} #2621

[RFC] vim-patch:7.4.{569,573} #2621

Uh oh!

Conversation

glts commented May 9, 2015

Uh oh!

glts May 9, 2015

Choose a reason for hiding this comment

Uh oh!

glts commented May 9, 2015

Uh oh!

glts commented May 10, 2015

Uh oh!

jszakmeister commented May 10, 2015

Uh oh!

glts commented May 10, 2015

Uh oh!

glts commented May 26, 2015

Uh oh!

felipecrv commented May 29, 2015

Uh oh!

justinmk commented Jun 14, 2015

Uh oh!

Shougo commented Dec 19, 2015

Uh oh!

Shougo commented Dec 19, 2015

Uh oh!

Uh oh!