Problem: v:register is wrong in v_: command (after 9.1.1858).
Solution: Don't reset v:register for OP_COLON (zeertzjq)
related: https://github.com/vim/vim/pull/18583#issuecomment-3418030021closes: vim/vim#185970124320c97
While at it, also fix using stale set_prevcount value. That only matters
when readbuf1 ends with an operator or a register, which never happens,
but it's still good to avoid using a stale value.
The processing of terminfo can be separated into two steps:
1. The initialization of terminfo, which includes trying to find $TERM
in a terminfo database file. As a fallback, common terminfo
definitions are compiled in. After this, we apply a lot of ad-hoc
patching to cover over limitations of terminfo.
2. While processing updates from nvim, actually using terminfo strings
and formatting them with runtime values. for this part, terminfo
essentially is a hyper-enhanced version of snprintf(), including
a sm0l stack based virtual machine which can manipulate the runtime
parameters.
This PR completely replaces libuniblium for step 2, with code
vendored from NetBSD's libtermkey which has been adapted to use typesafe
input parameters and to write into an output buffer in place.
The most immedatiate effects is a performance enhancement of
update_attrs() which is a very hot function when profiling the
TUI-process part of screen updates. In a stupid microbenchmark
(essentially calling nvim__screenshot over and over in a loop) this
leads to a speedup of ca 1.5x for redrawing the screen on the TUI-side.
What this means in practise when using nvim as a text editor is probably
no noticible effect at all, and when reabusing nvim as idk a full screen
RGB ASCII art rendrer maybe an increase from 72 to 75 FPS LMAO.
As nice side-effect, reduce the usage of unibilium to initialization only..
which will make it easier to remove, replace or make unibilium optional,
adressing #31989. Specifically, the builtin fallback doesn't use
unibilium at all, so a unibilium-free build is in principle possible
if the builtin definitions are good enough.
As a caveat, this PR doesn't touch libtermkey at all, which still has a
conditional dependency on unibilium. This will be investigated in a
follow-up PR
Note: the check of $TERMCOLOR was moved from tui/tui.c to
_defaults.lua in d7651b27d5 as we want to
skip the logic in _defaults.lua if the env var was set, but there
is no harm in TUI getting the right value when the TUI is trying to
initialize its terminfo shenanigans. Also this check is needed when
a TUI connects to a `--headless` server later, which will observe
a different $TERMCOLOR value than the nvim core process itself.
* x64: Add support for CET IBT.
* Gracefully handle broken custom allocator.
* Add GNU/Hurd build support.
* Fix io.write() of newly created buffer.
* Fix reporting of an error during error handling.
Problem: completion: whitespace not cleared with 'ai'
Solution: Remove spaces added by 'autoindent' when autocomplete is set
and restore did_ai in ins_compl_start() (Maxim Kim)
fixesvim/vim#18560closes: vim/vim#185826180d65751
Co-authored-by: Maxim Kim <habamax@gmail.com>
docs: fix issue query in github templates
Problem:
Error "Filter contains 1 issue: Invalid value enhancement for label" when
clicking on the "existing feature requests" links in the github feature
template or github lsp bug template.
Solution:
Update the queries to use the new type:Enhancement categorization instead of the
old label:enhancement system.
Similar to #36107.
docs(lsp): specify exactly how `vim.lsp.config` merges configs…
Problem:
It already says that the behaviour is defined by `vim.tbl_deep_extend`,
but that can mean very different things depending on the `behavior`
parameter.
Solution:
Specify that it uses "force".
Problem: v:register not reset after Visual mode command.
(laktak)
Solution: Reset v:register if Visual mode was active before
do_pending_operator() (zeertzjq)
fixes: vim/vim#18579
related: vim/vim#5305closes: vim/vim#18583b3b47e540d
vim-patch: 8.1.1985: code for dealing with paths is spread out
vim-patch:8.2.0375: Coverity warning for not using return value
vim-patch:8.2.0471: missing change to compile_list()
vim-patch:8.2.0489: Vim9: memory leaks
vim-patch:8.2.0494: Vim9: asan error
vim-patch:8.2.0685: Build failure
vim-patch:8.2.0759: Vim9: missing changes for performance improvements
vim-patch:8.2.1740: test fails without the terminal feature
vim-patch:8.2.1742: test still fails without the terminal feature
vim-patch:8.2.1880: Vim9: asan complains about adding zero to NULL
vim-patch:8.2.1958: build failure with timers
vim-patch:8.2.2248: ASAN error on exit with GUI
vim-patch:8.2.2273: build failure
vim-patch:8.2.2382: build failure
vim-patch:8.2.2401: build fails without +profiling feature
vim-patch:8.2.2536: Coverity complains about unchecked return value
vim-patch:8.2.2599: build failure
vim-patch:8.2.2715: Vim9: tests fail without the channel feature
vim-patch:8.2.2987: build failure with normal features
vim-patch:8.2.3092: Vim9: builtin function test fails without +channel feature
vim-patch:8.2.3291: Coverity warns for not checking return value
vim-patch:8.2.3505: Vim9: build failure without the +eval feature
vim-patch:8.2.3541: compiler warning for unused variable in tiny version
vim-patch:8.2.3546: build failure without the +eval feature
vim-patch:8.2.3563: build failure with +eval but without GUI or +termguicolors
vim-patch:8.2.3566: build failure on old systems when using nano timestamp
vim-patch:8.2.3621: build failure
vim-patch:8.2.3636: Coverity warns for unreachable code
vim-patch:8.2.3672: build failure with unsigned char
vim-patch:8.2.3691: build failure with small features
vim-patch:8.2.3693: Coverity warns for possibly using a NULL pointer
vim-patch:8.2.3718: compiler warns for unused variable without +textprop
vim-patch:8.2.3724: build error for missing error message in small build
vim-patch:8.2.3736: test fails without the channel feature
vim-patch:8.2.3740: memory left allocated on exit when using Tcl
vim-patch:8.2.3752: build error when using Photon GUI
vim-patch:8.2.3791: build error with +cindent but without +smartindent
vim-patch:8.2.3868: Vim9: function test fails
vim-patch:8.2.3926: build failure without the 'autochdir' option
vim-patch:8.2.3931: Coverity reports a memory leak
vim-patch:8.2.3958: build failure compiling xxd with "-std=c2x"
vim-patch:8.2.3963: build failure with tiny and small features
vim-patch:8.2.3988: tiny build fails
vim-patch:8.2.3998: asan error for adding zero to NULL
vim-patch:8.2.4015: build failure with tiny features
vim-patch:8.2.4034: Coverity warns for possibly using a NULL pointer
vim-patch:8.2.4042: Vim9: build error
vim-patch:8.2.4054: Vim9 script test fails
vim-patch:8.2.4079: MS-Windows: "gvim --version" didn't work with VIMDLL
vim-patch:8.2.4081: CodeQL reports problem in if_cscope causing it to fail
vim-patch:8.2.4119: build failure when disabling the channel feature
vim-patch:8.2.4127: build failure without the +eval feature
vim-patch:8.2.4150: Coverity warns for using pointer after free
vim-patch:8.2.4239: build fails with unusual configuration
vim-patch:8.2.4272: Vim9 expr test fails without the channel feature
vim-patch:8.2.4278: build with Athena GUI fails
vim-patch:8.2.4300: cannot build tiny version
vim-patch:8.2.4368: Amiga: a few compiler warnings
vim-patch:8.2.4369: redundant #ifdef argument
vim-patch:8.2.4385: cannot build tiny version
vim-patch:8.2.4386: still cannot build tiny version
vim-patch:8.2.4405: compiler warning for unused variable without +folding
vim-patch:8.2.4413: Vim9: Coverity warns for using NULL pointer
vim-patch:8.2.4470: Coverity warns for uninitialized variable
vim-patch:8.2.4488: build error with +eval but without +channel or +job
vim-patch:8.2.4516: build failure without the +eval feature
vim-patch:8.2.4521: build failure without the +diff feature
vim-patch:8.2.4532: suspending with CTRL-Z does not work on OpenBSD
vim-patch:8.2.4549: cannot build with Motif and editres
vim-patch:8.2.4550: Motif: cannot set the color of the scrollbar thumb
vim-patch:8.2.4556: test fails without the +job or +channel feature
vim-patch:8.2.4560: suspending with CTRL-Z does not work on DragonFlyBSD
vim-patch:8.2.4654: missing changes for import check
vim-patch:8.2.4680: build failure without +postscript
vim-patch:8.2.4681: build fails with a combination of features
vim-patch:8.2.4725: unused variable in tiny build
vim-patch:8.2.4756: build error without the +eval feature
vim-patch:8.2.4769: build warning with UCRT
vim-patch:8.2.4771: Coverity warns for not checking return value
vim-patch:8.2.4815: cannot build with older GTK version
vim-patch:8.2.4863: accessing freed memory in test without the +channel feature
vim-patch:8.2.4878: valgrind warning for using uninitialized variable
vim-patch:8.2.4884: test fails without the job/channel feature
vim-patch:8.2.4898: Coverity complains about pointer usage
vim-patch:8.2.4900: Vim9 expression test fails without the job feature
vim-patch:8.2.4976: Coverity complains about not restoring a saved value
vim-patch:8.2.5021: build fails with normal features and +terminal
vim-patch:8.2.5067: timer_create is not available on every Mac system
vim-patch:8.2.5071: with some Mac OS version clockid_t is redefined
vim-patch:8.2.5125: MS-Windows: warnings from MinGW compiler
vim-patch:8.2.5133: MacOS: build fails
vim-patch:8.2.5137: cannot build without the +channel feature
vim-patch:8.2.5168: cannot build with Python 3.11
vim-patch:9.0.0123: cannot build with small features
vim-patch:9.0.0159: cannot build with small features
vim-patch:9.0.0161: warning for uninitialized variable
vim-patch:9.0.0237: Mac: cannot build if dispatch.h is not available
vim-patch:9.0.0256: compiler warning for uninitialized variables
vim-patch:9.0.0378: compiler warning for uninitialized variable
vim-patch:9.0.0413: ASAN reports a memory leak
vim-patch:9.0.0462: ASAN warning for integer overflow
vim-patch:9.0.0970: Coverity warns for uninitialized variable
vim-patch:9.0.1042: ASAN gives false alarm about array access.
vim-patch:9.0.1076: ASAN complains about NULL argument
vim-patch:9.0.1085: compiler warns for uninitialized variable
vim-patch:9.0.1666: compiler may warn for uninitialized variable
vim-patch:9.1.0816: tests: not clear what tests cause asan failures
vim-patch:9.1.1345: tests: Test_xxd_color2() test failure dump diff is misleading
Problem: when create a hidden terminal job with `nvim_buf_call`+
`jobstart(…,{term=true})`, program like `lazygit` cannot figure out the
correct width and height.
Solution: `jobstart(…,{term=true})` accepts `width`/`height`
Problem: if pumblend >= 50, non-whitespace menu character gets italic,
bold, underline (or similar) attribute from the cell underneath, which
is not really useful and can't be "blended".
Solution: drop highlight combination for that special case (@zeertzjq
on #36133).
Co-authored-by: zeertzjq <zeertzjq@outlook.com>
Problem:
If a client doesn't have a config then an error may be thrown.
Probably caused by: 2f78ff816b
Lua callback: …/lsp.lua:442: attempt to index local 'config' (a nil value)
stack traceback:
…/lsp.lua:442: in function 'can_start'
…/lsp.lua:479: in function 'lsp_enable_callback'
…/lsp.lua:566: in function <…/lsp.lua:565>
Solution:
Not all clients necessarily have configs.
- Handle `config=nil` in `can_start`.
- If user "enables" an invalid name that happens to match a *client*
name, don't auto-detach the client.
This patch preserves indentation in comments. It work by removing the
explicit 3-space indentation and replaces with with an expression which
uses the current value with a minimum of three spaces.
Discussed in the mailing list:
https://groups.google.com/g/vim_dev/c/rn8ZLDrCbYU
Thanks to Friedrich Romstedt for reporting and Christian Brabandt for
investigating the issue.
closes: vim/vim#1856664c8105867
Co-authored-by: Marshall Ward <marshall.ward@gmail.com>
The value of this variable is evaluated when the 'filetype' option is
changed and should be a command string that will undo any configuration
changes that the plugin has made.
See :help undo_indent for details
related: vim/vim#185665ffb23c967
Co-authored-by: Doug Kearns <dougkearns@gmail.com>
Minor style and whitespace changes to the reST syntax file, to sync with
the development branch.
related: vim/vim#1856646d86979a9
Co-authored-by: Marshall Ward <marshall.ward@gmail.com>
Escape character support was disabled for inline literals, which are
handled separately from standard inline highlights, in that escape
characters are unsupported.
related: vim/vim#18566d7fb4cd2f8
Co-authored-by: Marshall Ward <marshall.ward@gmail.com>
Previously, a comment in a directive block would incorrectly mark
all subsequent lines in the directive block as comment, because the
syn-region did not check the leading indent.
related: vim/vim#185664e9f16dd1e
Co-authored-by: Antony Lee <anntzer.lee@gmail.com>
The leading two dots of a RST directive gets matched by
rstExplicitMarkup group first, and then the directive name and directive
body will be matched by the groups contained in rstDirectives cluster in
order.
The rstExDirective group in rstDiretives matches any RST directives
other than footnote, citation and hyperlink target, therefore admonition
and code block will be matched by rstExDirective. This fix has
rstExDirective contain rstExplicitMarkup so as to enable recursive RST
directives match.
The interpreted texts enclosed in quotes are not highlighted within a
RST directive body, because the rstCruft cluster contains a non-existing
rstInterpretedText group.It should be renamed to
rstInterpretedTextOrHyperlinkReference which is defined in a
DefineInlineMarkup function call.
related: vim/vim#1856647949b4b46
Co-authored-by: Minjie Xu <madjxatw@gmail.com>
Following the approach used in other syntax definitions, add support for
defining the "syntax sync minlines=..." values for rst files in the
users' ~/.vimrc files, to allow the users to adjust that value in case
syntax highlighting stops working for some of the files they edit.
related: vim/vim#18566310082f3cf
Co-authored-by: Dragan Simic <dsimic@manjaro.org>
A doctest block usually spans multiple lines, e.g.
>>> print('this is a Doctest block')
this is a Doctest block
Remove ``oneline`` argument to syntax region as this requirement is not
met. Consequently, also remove ``display`` as the prerequisite (the
syntax region is on a single line) is no longer met.
----
Recognise '>>>' inside doctest blocks
Recognise subsequent '>>>' prompts in doctest blocks, e.g.
>>> print('this is a Doctest block')
this is a Doctest block
>>> print('this is a second Doctest block')
this is a second Doctest block
A doctest block usually spans multiple lines, e.g.
>>> print('this is a Doctest block')
this is a Doctest block
related: vim/vim#185660940465866
Co-authored-by: Kirk Roemer <91125534+kirk-roemer@users.noreply.github.com>
- Column align tags
- Move tags to the same line as the function signature
- Move descriptions to the line below the function signature
- Add missing hyperlinks to builtins in the description text
closes: vim/vim#1847883eb1da19e
Co-authored-by: Doug Kearns <dougkearns@gmail.com>
Fix syntax highlighting for def-style Python functions, with their
parameters spanning multiple lines. E.g. the following should match as
valid Python code in Bitbake recipes:
def myFunction(one, two, \
three, four):
pass
For this to work, use the prefix modifier "\_" before the wildcard ".",
to also match newline characters.
closes: vim/vim#185655fe4faa711
Co-authored-by: Martin Schwan <m.schwan@phytec.de>
Problem: Ctrl-F and Ctrl-B at more prompt not working with kitty
keyboard protocol or modifyOtherKeys (after v9.1.1849).
Solution: Call merge_modifyOtherKeys() in get_keystroke() (zeertzjq).
closes: vim/vim#185586622dc2ded
Problem: CTRL-F and CTRL-B don't work in more prompt
Solution: Make CTRL-F and CTRL-B scroll by a screen down/up
(Bjoern Foersterling)
closes: vim/vim#18545fcf4c435af
Co-authored-by: bfoersterling <bjoern.foersterling@gmail.com>
Suppresses output from the :only command by passing the { silent = true }
modifier to vim.cmd.only(). This prevents unnecessary messages when
setting up the diff layout.
Closes#36167
- Move autocmd cleanup logic back to setup_layout as WinClosed autocmds
were triggering for every closed window even when buf scoped, and when
buf scoped, pattern filter did not work.
- Cleanup when quickfix window is closed as well to prevent state where
on closing quickfix setup_layout will reset it after with empty
content because of WinClosed autocmds triggering.
Signed-off-by: Tomas Slusny <slusnucky@gmail.com>
Problem:
Checking if the tag is nvim-owned (defined in a $VIMRUNTIME/doc/*.txt
file) is too expensive for now without non-obvious workarounds
Solution:
Only set url extmarks for $VIMRUNTIME/doc/*.txt files.
Closes: #36163
Use double sentence spacing and wrap lines at 'textwidth'. Code
examples and tables were not wrapped unless this had already been done
locally.
closes: vim/vim#18453c58f91c035
Fix incorrect docs in :h ModeChanged.
Cherry-pick :h bufnr() changes from patch 8.1.2080.
Co-authored-by: Doug Kearns <dougkearns@gmail.com>
Problem: No cmdline completion for :echoconsole, :echowindow and second
expression after :echoerr.
Solution: Set EXPAND_EXPRESSION for :echoconsole and :echowindow, and
check for multiple expressions after :echoerr (zeertzjq).
closes: vim/vim#185527aaca97fc5
