Re: [vim/vim] The nv_g_cmd() function is too long. Refactor it using a table of key… (PR #9561)

Hi Bram, On Wed, Jan 19, 2022 at 8:58 AM Bram Moolenaar wrote: > > Yegappan wrote: > > > On Wed, Jan 19, 2022 at 4:34 AM Bram Moolenaar ***@***.***> > > wrote: > > > > > I'm not sure this is an improvement. Using a switch statement the compiler > > > can find the optimal way to select the code to

Patch 8.2.4148

Patch 8.2.4148 Problem:Deleting any mapping may cause to not set the script context. Solution: Only reset last_used_map if it is the deleted mapping. (closes #9568) Files: src/map.c, src/getchar.c, src/proto/getchar.pro, src/testdir/test_mapping.vim

Patch 8.2.4147

Patch 8.2.4147 Problem:E464 does not always include the offending command. Solution: Add another error message with "%s". (closes #9564) Files: src/errors.h, src/vim9compile.c, src/ex_docmd.c, src/testdir/test_vim9_script.vim *** ../vim-8.2.4146/src/errors.h2022-0

Patch 8.2.4146

Patch 8.2.4146 Problem:Vim9: shadowed function can be used in compiled function but not at script level. Solution: Also give an error in a compiled function. (closes #9563) Files: src/vim9expr.c *** ../vim-8.2.4145/src/vim9expr.c 2022-01-18 17:43:01.061598437 + -

Patch 8.2.4145

Patch 8.2.4145 Problem:Confusing error when using name of import for a function. Solution: Pass a flag to trans_function_name(). Files: src/vim.h, src/userfunc.c, src/proto/userfunc.pro, src/eval.c, src/testdir/test_vim9_import.vim *** ../vim-8.2.4144/src/vim.h 2022-01-

Re: [vim/vim] The nv_g_cmd() function is too long. Refactor it using a table of key… (PR #9561)

Yegappan wrote: > On Wed, Jan 19, 2022 at 4:34 AM Bram Moolenaar ***@***.***> > wrote: > > > I'm not sure this is an improvement. Using a switch statement the compiler > > can find the optimal way to select the code to execute. A linear search > > through an array is likely to always be slower.

Re: [vim/vim] The nv_g_cmd() function is too long. Refactor it using a table of key… (PR #9561)

Hi Bram, On Wed, Jan 19, 2022 at 4:34 AM Bram Moolenaar wrote: > I'm not sure this is an improvement. Using a switch statement the compiler > can find the optimal way to select the code to execute. A linear search > through an array is likely to always be slower. > A binary search would be much

Patch 8.2.4144

Patch 8.2.4144 Problem:Cannot load libsodium dynamically. Solution: Support dynamic loading on MS-Windows. (Ken Takata, closes #9554) Files: src/Make_cyg_ming.mak, src/Make_mvc.mak, src/buffer.c, src/crypt.c, src/memline.c, src/proto/crypt.pro *** ../vim-8.2.4143/src/Make

Patch 8.2.4143

Patch 8.2.4143 Problem:MS-Windows: IME support for Win9x is obsolete. Solution: Remove the Win9x code. (Ken Takata, closes #9559) Files: src/gui_w32.c *** ../vim-8.2.4142/src/gui_w32.c 2022-01-16 14:15:45.517045001 + --- src/gui_w32.c 2022-01-19 12:58:34.751910173 +00

Re: [vim/vim] Dotted/dashed underline support (Issue #9553)

On wsltty (windows) everything is rendered: [image: 2022-01-19_15-14-58.png] среда, 19 января 2022 г. в 15:10:42 UTC+3, Dominique Pellé: > @brammool wrote: > > The double underline looks like a thin underline, is that a display error? > > Do all the terminals use th

Patch 8.2.4142

Patch 8.2.4142 Problem:Build failure with normal features without persistent undo. Solution: Adjust #ifdef. (closes #9557) Files: src/fileio.c *** ../vim-8.2.4141/src/fileio.c2022-01-08 16:19:18.501639918 + --- src/fileio.c2022-01-19 11:19:57.570065914 + *