Problem:
Tests call `clear()` even though `clear_notrace()` is already called in
a `before_each()` handler, wasting precious milliseconds!
Solution:
Remove redundant `clear()` calls.
Problem:
Users of the Roslyn (C#) LSP have encountered significant delays when
retrieving pull diagnostics in large documents while using Neovim. For
instance, diagnostics in a 2000-line .cs file can take over 20 seconds
to display after edits in Neovim, whereas in VS Code, diagnostics for
the same file are displayed almost instantly.
As [mparq noted](https://github.com/seblj/roslyn.nvim/issues/93#issuecomment-2508940330)
in https://github.com/seblj/roslyn.nvim/issues/93, VS Code leverages
additional parameters specified in the [LSP documentation for
textDocument/diagnostic](https://microsoft.github.io/language-server-protocol/specifications/lsp/3.17/specification/#documentDiagnosticParams),
specifically:
- previousResultId
- identifier
Solution:
When requesting diagnostics, Neovim should include the
`previousResultId` and `identifier` parameters as part of the request.
These parameters enable the server to utilize caching and return
incremental results.
Support for maintaining state is already present in the
[textDocument/semanticTokens implementation](8f84167c30/runtime/lua/vim/lsp/semantic_tokens.lua (L289)).
A similar mechanism can be implemented in `textDocument/diagnostic` handler.
Problem:
In cases when the (in-process) LSP server responds to the request
immediately and calls `notify_reply_callback` the request will still be
marked as pending, because the code assumes that the response will occur
asynchronously. Then the request will be pending forever, because it was
already set as "completed" before we even set it as "pending".
A workaround is to wrap `notify_replay_callback` in `vim.shedule` ([like
so](https://github.com/neovim/neovim/pull/24338#issuecomment-2809568617)]
but that seems counterintuitive.
Solution:
Handle this case in Client:request().
Problem: missing out-of-memory check in textformat.c
Solution: add out-of-memory check, add small optimizations to
internal_format() and same_leader() (John Marriott)
closes: vim/vim#17200c25368ba14
Co-authored-by: John Marriott <basilisk@internode.on.net>
Problem: As in f85bc41, assume unfocusable windows to be UI windows
whose buffer content is unexpectedly included in 'complete'
"w" completion.
Solution: Exclude unfocusable windows when looping over windows.
Problem: 'showmode' ext_messages state is not cleared after insert_expand mode.
Solution: Replace empty message with more idiomatic way to clear the showmode.
Problem: Undo corrupted with 'completeopt' "preinsert" when switching
buffer or window.
Solution: Do not delete preinsert text when switching buffer or window.
(zeertzjq)
related: neovim/neovim#33581closes: vim/vim#171931343681aba
Problem: cannot get information about command line completion
Solution: add CmdlineLeavePre autocommand and cmdcomplete_info() Vim
script function (Girish Palya)
This commit introduces two features to improve introspection and control
over command-line completion in Vim:
- Add CmdlineLeavePre autocmd event:
A new event triggered just before leaving the command line and before
CmdlineLeave. It allows capturing completion-related state that is
otherwise cleared by the time CmdlineLeave fires.
- Add cmdcomplete_info() Vim script function:
Returns a Dictionary with details about the current command-line
completion state.
These are similar in spirit to InsertLeavePre and complete_info(),
but focused on command-line mode.
**Use case:**
In [[PR vim/vim#16759](https://github.com/vim/vim/pull/16759)], two examples
demonstrate command-line completion: one for live grep, and another for
fuzzy file finding. However, both examples share two key limitations:
1. **Broken history recall (`<Up>`)**
When selecting a completion item via `<Tab>` or `<C-n>`, the original
pattern used for searching (e.g., a regex or fuzzy string) is
overwritten in the command-line history. This makes it impossible to
recall the original query later.
This is especially problematic for interactive grep workflows, where
it’s useful to recall a previous search and simply select a different
match from the menu.
2. **Lack of default selection on `<CR>`**
Often, it’s helpful to allow `<CR>` (Enter) to accept the first match
in the completion list, even when no item is explicitly selected. This
behavior is particularly useful in fuzzy file finding.
----
Below are the updated examples incorporating these improvements:
**Live grep, fuzzy find file, fuzzy find buffer:**
```vim
command! -nargs=+ -complete=customlist,GrepComplete Grep VisitFile()
def GrepComplete(arglead: string, cmdline: string, cursorpos: number):
list<any>
return arglead->len() > 1 ? systemlist($'grep -REIHns "{arglead}"' ..
' --exclude-dir=.git --exclude=".*" --exclude="tags" --exclude="*.swp"') : []
enddef
def VisitFile()
if (selected_match != null_string)
var qfitem = getqflist({lines: [selected_match]}).items[0]
if qfitem->has_key('bufnr') && qfitem.lnum > 0
var pos = qfitem.vcol > 0 ? 'setcharpos' : 'setpos'
exec $':b +call\ {pos}(".",\ [0,\ {qfitem.lnum},\ {qfitem.col},\ 0]) {qfitem.bufnr}'
setbufvar(qfitem.bufnr, '&buflisted', 1)
endif
endif
enddef
nnoremap <leader>g :Grep<space>
nnoremap <leader>G :Grep <c-r>=expand("<cword>")<cr>
command! -nargs=* -complete=customlist,FuzzyFind Find
execute(selected_match != '' ? $'edit {selected_match}' : '')
var allfiles: list<string>
autocmd CmdlineEnter : allfiles = null_list
def FuzzyFind(arglead: string, _: string, _: number): list<string>
if allfiles == null_list
allfiles = systemlist($'find {get(g:, "fzfind_root", ".")} \! \(
-path "*/.git" -prune -o -name "*.swp" \) -type f -follow')
endif
return arglead == '' ? allfiles : allfiles->matchfuzzy(arglead)
enddef
nnoremap <leader><space> :<c-r>=execute('let
fzfind_root="."')\|''<cr>Find<space><c-@>
nnoremap <leader>fv :<c-r>=execute('let
fzfind_root="$HOME/.vim"')\|''<cr>Find<space><c-@>
nnoremap <leader>fV :<c-r>=execute('let
fzfind_root="$VIMRUNTIME"')\|''<cr>Find<space><c-@>
command! -nargs=* -complete=customlist,FuzzyBuffer Buffer execute('b '
.. selected_match->matchstr('\d\+'))
def FuzzyBuffer(arglead: string, _: string, _: number): list<string>
var bufs = execute('buffers', 'silent!')->split("\n")
var altbuf = bufs->indexof((_, v) => v =~ '^\s*\d\+\s\+#')
if altbuf != -1
[bufs[0], bufs[altbuf]] = [bufs[altbuf], bufs[0]]
endif
return arglead == '' ? bufs : bufs->matchfuzzy(arglead)
enddef
nnoremap <leader><bs> :Buffer <c-@>
var selected_match = null_string
autocmd CmdlineLeavePre : SelectItem()
def SelectItem()
selected_match = ''
if getcmdline() =~ '^\s*\%(Grep\|Find\|Buffer\)\s'
var info = cmdcomplete_info()
if info != {} && info.pum_visible && !info.matches->empty()
selected_match = info.selected != -1 ? info.matches[info.selected] : info.matches[0]
setcmdline(info.cmdline_orig). # Preserve search pattern in history
endif
endif
enddef
```
**Auto-completion snippet:**
```vim
set wim=noselect:lastused,full wop=pum wcm=<C-@> wmnu
autocmd CmdlineChanged : CmdComplete()
def CmdComplete()
var [cmdline, curpos] = [getcmdline(), getcmdpos()]
if getchar(1, {number: true}) == 0 # Typehead is empty (no more pasted input)
&& !pumvisible() && curpos == cmdline->len() + 1
&& cmdline =~ '\%(\w\|[*/:.-]\)$' && cmdline !~ '^\d\+$' # Reduce noise
feedkeys("\<C-@>", "ti")
SkipCmdlineChanged() # Suppress redundant completion attempts
# Remove <C-@> that get inserted when no items are available
timer_start(0, (_) => getcmdline()->substitute('\%x00', '', 'g')->setcmdline())
endif
enddef
cnoremap <expr> <up> SkipCmdlineChanged("\<up>")
cnoremap <expr> <down> SkipCmdlineChanged("\<down>")
autocmd CmdlineEnter : set bo+=error
autocmd CmdlineLeave : set bo-=error
def SkipCmdlineChanged(key = ''): string
set ei+=CmdlineChanged
timer_start(0, (_) => execute('set ei-=CmdlineChanged'))
return key != '' ? ((pumvisible() ? "\<c-e>" : '') .. key) : ''
enddef
```
These customizable snippets can serve as *lightweight* and *native*
alternatives to picker plugins like **FZF** or **Telescope** for common,
everyday workflows. Also, live grep snippet can replace **cscope**
without the overhead of building its database.
closes: vim/vim#1711592f68e26ec
Co-authored-by: Girish Palya <girishji@gmail.com>
Problem:
nvim_parse_cmd returns invalid 'range' field for cmd like `:bdelete`.
Solution:
Add the condtion `ea.add_count > 0` as required to put 'range'
into result.
Co-authored-by: zeertzjq <zeertzjq@outlook.com>
Problem:
'winblend' does not display text behind blank unicode characters other than 0x20 ascii space.
Solution:
In ui_compositor.c -> compose_line() check for either 0x20 or 0x80A0E2 (utf8 \u2800).
Problem:
Cursor is visible in "hidden" floating window.
Solution:
Hide cursor when curwin is a hidden floating window.
Show cursor after returning to a normal (non-hidden) window.
Problem: Change applied in d3e495ce uses a byte-offset where a virtual
column is expected.
Solution: Set the cursor directly through a <Cmd> mapping, while making
sure the commands are ordered correctly by adding them to the
type-ahead buffer.
This commit fixes the following error message:
```
Compiler not supported: make inc< sw< sts<
```
1. orginal value: `setl com< cms< et< fo<| compiler make inc< sw< sts<`
2. correct value: `setl com< cms< et< fo< inc< sw< sts< | compiler make`
While at it, let's also document the g:yaml_recommended_style variable.
closes: vim/vim#17179229f79c168
Co-authored-by: Vincent Law <vlaw@users.noreply.github.com>
Co-authored-by: Christian Brabandt <cb@256bit.org>
Problem:
Neovim disables a number of terminal modes when it exits, some of which
cause the terminal to send asynchronous events to Neovim. It's possible
that Neovim exits before the terminal has received and processed all of
the sequences to disable these modes, causing the terminal to emit one
of these asynchronous sequences after Neovim has already exited. If this
happens, then the sequence is received by the user's shell (or some
other program that is not Neovim).
Solution:
When Neovim exits, it now emits a Device Attributes request (DA1)
after disabling all of the different modes. When the terminal responds
to this request we know that it has already received all of our other
sequences disabling the other modes. At that point, it should not be
emitting any further asynchronous sequences. This means the process of
exiting Neovim is now asynchronous as well since it depends on receiving
the DA1 response from the terminal.
Problem: Transpiled Lua code from vim9script is not amenable to static
analysis, requiring manual cleanup or ignoring parts of our codebase.
Solution: Revert to pre-rewrite version of (low-impact) legacy plugins
and remove the vim9jit shim.
Problem: When iterating in reverse with {start} > {end} in
`nvim_buf_get_extmarks()`, marks that overlap {start} and are
greater than {end} are included in the return value twice.
Marks that overlap {end} and do not overlap {start} are not
not included in the return value at all. Marks are not
actually returned in a meaningful "traversal order".
Solution: Rather than actually iterating in reverse, (also possible but
requires convoluted conditions and would require fetching
overlapping marks for both the {start} and {end} position,
while still ending up with non-traversal ordered marks),
iterate normally and reverse the return value.
Problem:
Default 'statusline' is implemented in C and not representable as
a statusline expression. This makes it hard for user configs/plugins to
extend it.
Solution:
- Change the default 'statusline' slightly to a statusline expression.
- Remove the C implementation.
Problem: We allow setting 'cmdheight' to 0 with ext_messages enabled
since b72931e7. Enabling ext_messages with vim.ui_attach()
implicitly sets 'cmdheight' to 0 for BWC. When non-zero
'cmdheight' is wanted, this behavior make it unnecessarily
hard to keep track of the user configured value.
Solution: Add set_cmdheight to vim.ui_attach() opts table that can be
set to false to avoid setting 'cmdheight' to 0.
Useful to e.g. limit the height to the window height, avoiding unnecessary
work. Or to find out how many buffer lines beyond "start_row" take up a
certain number of logical lines (returned in "end_row" and "end_vcol").
