Problem: It is possible (and documented in `:h vim.pack`) that plugin's
`src` uses "insteadOf" Git config. In that case comparing it directly
to repo's `origin` will error.
Solution: Add extra check that lockfile's `src` is not equal to repo's
`origin` when taking Git's "insteadOf" into account.
However, still report the original lockfile's `src` in the
`:checkhealth` output, as it seems to be a clearer indication of what
actually is wrong.
If a user accepts completion and immediately exits insert mode, it
could happen that `Context.cursor` was nil by the time the
`completion/resolve` response arrives, leading to an error.
Problem:
Default statusline doesn't show progress status.
Solution:
- Provide `vim.ui.progress_status()`.
- Include it in the default 'statusline'.
How it works:
Status text summarizes "running" progress messages.
- If none: returns empty string
- If one running item: "title: percent%"
- If multiple running items: "Progress: N items avg-percent%"
Problem:
Checking the extension of a file is done often, e.g. in Nvim's codebase
for differentiating Lua and Vimscript files in the runtime. The current
way to do this in Lua is (1) a Lua pattern match, which has pitfalls
such as not considering filenames starting with a dot, or (2)
fnamemodify() which is both hard to discover and hard to use / read if
not very familiar with the possible modifiers.
vim.fs.ext() returns the file extension including the leading dot of
the extension. Similar to the "file extension" implementation of many
other stdlibs (including fnamemodify(file, ":e")), a leading dot
doesn't indicate the start of the extension. E.g.: the .git folder in a
repository doesn't have the extension .git, but it simply has no
extension, similar to a folder named git or any other filename without
dot(s).
Problem
The format of LSP log messages is inconsistent; some include underscores, while others are not logged at all.
Solution
Standardize log recording and unify the log message prefixes with the module names.
Problem:
The capability attach path for client/registerCapability can initialize a capability even when the capability was only registered in specific buffers
Solution:
Check supports_method() before attaching capabilities from the dynamic registration handler so unsupported registrations are ignored.
AI-assisted: OpenCode
Problem:
Terminal refresh may be missed if buffer update callbacks poll for uv
events. Example test failure on FreeBSD:
FAILED test/functional/terminal/buffer_spec.lua @ 1049: :terminal buffer scrollback is correct if buffer update callbacks poll for uv events
test/functional/terminal/buffer_spec.lua:1004: Row 1 did not match.
Expected:
|*19995: TEST{MATCH: +}|
|*19996: TEST{MATCH: +}|
|*19997: TEST{MATCH: +}|
|*19998: TEST{MATCH: +}|
|*19999: TEST{MATCH: +}|
|^[Process exited 0] |
|{5:-- TERMINAL --} |
Actual:
|*19696: TEST |
|*19697: TEST |
|*19698: TEST |
|*19699: TEST |
|*19700: TEST |
|^[Process exited 0] |
|{5:-- TERMINAL --} |
Solution:
Call changed_lines() after resetting invalid region in refresh_screen().
Handle terminal being invalidated in the middle of refresh_timer_cb().
Problem: Vim may freeze if setcmdline() is called while the wildmenu or
cmdline popup menu is active (rendcrx)
Solution: Cleanup completion state if cmdbuff_replaced flag has been set
(Yasuhiro Matsumoto)
fixes: vim/vim#19742closes: vim/vim#19744332dd22ed4
Co-authored-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Problem: The glob() function on Unix-like systems does not escape
newline characters when expanding wildcards. A maliciously
crafted string containing '\n' can be used as a command
separator to execute arbitrary shell commands via
mch_expand_wildcards(). This depends on the user's 'shell'
setting.
Solution: Add the newline character ('\n') to the SHELL_SPECIAL
definition to ensure it is properly escaped before being
passed to the shell (pyllyukko).
closes: vim/vim#19746
Github Advisory:
https://github.com/vim/vim/security/advisories/GHSA-w5jw-f54h-x46c645ed6597d
Co-authored-by: pyllyukko <pyllyukko@maimed.org>
- Add TODO comments for aggregating diagnostics from all pull namespaces
and for clearing diagnostics when an empty array is received, referencing
the LSP specification.
- Update diagnostics refresh logic to safely access previousResultId,
preventing potential nil errors.
Previously, resultId for diagnostics was keyed only by client_id, which
could cause issues when multiple identifiers are used by the same client.
This change introduces a composite key of client_id and identifier for
client_result_id, ensuring correct tracking of diagnostic results per
identifier. Updates all relevant logic to use the new keying scheme.
Previously, the LSP client assumed all providers had subcapabilities,
which could cause issues when a provider did not. This change adds a
check for the presence of subcapabilities before attempting to access
them, ensuring correct handling of both cases. This improves
compatibility with servers that register providers without additional
capabilities.
Update diagnostic refresh to request diagnostics from all provider
registrations using _provider_foreach. This ensures diagnostics are
fetched from every registered provider during a refresh.
Co-authored-by: ZieMcd <ziemcd@gmail.com>
Introduce _provider_foreach to iterate over all matching provider
capabilities for a given LSP method, handling both static and dynamic
registrations. Update diagnostic logic and tests to use the new
iteration approach, simplifying capability access and improving
consistency across features.
Problem
Unlike inlay hints, code lenses are closely related to running commands;
a significant number of code lenses are used to execute a command (such
as running tests). Therefore, it is necessary to provide a default
mapping for them.
Solution
Add a new default mapping "grx" (mnemonic: "eXecute", like "gx").
* docs(example_init): consistently use `:h`
* docs(example_init): align PLUGINS section with the overall style
Since this entire file is example, all sections use declarative voice,
avoiding "For example".
Replaces summary list with related help tags.
* docs(example_init): adjust mini.completion URI
See https://github.com/nvim-mini/mini.nvim/discussions/1970
* docs(example_init): misc grammar fix
Problem:
Mouse popup menus (right-click context menus) do not respect the
'pumborder' option and could overflow screen boundaries when borders
were enabled near the edge.
Solution:
- Remove the mouse menu exclusion from border rendering.
- Add boundary check to shift menu left when border would exceed screen
width, ensuring complete visibility of menu content and borders.
Problem:
`vim.lsp.document_color.color_presentation()` throws, due to incorrect use of the `vim.Range` API.
Solution:
Change `vim.Range.has(a, b)` call to `a:has(b)`.
Problem: Entering the pager is scheduled to avoid errors while in the
cmdwin. Meanwhile the current window is used as a condition to
detect an active pager.
Solution: Keep track of when the user is in the pager (or entered the
cmdwin while the pager was open).
Problem: Windows are closed and re-opened after changing tabpage (and
might be removed from the current tabpage after 094b297a).
Solution: Ensure windows are on the current tabpage by moving them.
Problem:
No way to iterate configs. Users need to reach
for `vim.lsp.config._configs`, an internal interface.
Solution:
Provide vim.lsp.get_configs().
Also indirectly improves :lsp enable/disable completion
by discarding invalid configs from completion.
Problem: using copy_option_part() can be improved
Solution: Refactor and use the return value of copy_option_part() to
avoid strlen() calls (John Marriott).
In addition, this commit includes the following changes:
memline.c:
- In recover_names():
- Replace calls to vim_strsave() with vim_strnsave() for the literal
strings
- Use a string_T to store local variable dir_name.
bufwrite.c:
- In buf_write()
- move variable wp to where it is used.
help.c:
- In fix_help_buffer():
- replace call to add_pathsep() with after_pathsep()
optionstr.c:
- In export_myvimdir():
- use a string_T to store local variable buf
- replace call to add_pathsep() with after_pathsep()
scriptfile.c:
- In do_in_path():
- use a string_T to store local variable buf
- measure the lengths of prefix and name once before the while loop
- replace call to add_pathsep() with after_pathsep()
- move some variables closer to where they are used
spellfile.c:
- In init_spellfile():
- use a string_T to store local variable buf
closes: vim/vim#19725a74e5fc5b9
Co-authored-by: John Marriott <basilisk@internode.on.net>
Problem:
If buffer update callbacks poll for uv events during terminal scrollback
refresh, new output from PTY process may lead to incorrect scrollback.
Solution:
Don't poll for output to the same terminal as the one being refreshed.
Add a minimal ftplugin `runtime/ftplugin/yara.vim` that sets:
- `commentstring` for YARA line comments (`//`)
- `comments` for YARA block comment (`/* */`)
- `formatoptions` to wrap comment lines and continue comment after newlines
This was heavily inspired from `runtime/ftplugin/c.vim`
closes: vim/vim#1973639ee7d17b9
Co-authored-by: Thomas Dupuy <thom4s.d@gmail.com>
Problem:
`nvim_echo(…, {id=…})` accepts user-defined id as a string or integer.
Generated ids are always higher than last highest msg-id used. Thus
plugins may accidentally advance the integer id "address space", which,
at minimum, could lead to confusion when troubleshooting, or in the
worst case, could overflow or "exhaust" the id address space.
There's no use-case for it, and it could be the mildly confusing, so we
should just disallow it.
Solution:
Disallow *integer* user-defined message-id.
Only allow *string* user-defined message-id.
Problem:
nvim_get_option_value with "filetype" set silently returns incorrect
defaults if autocommands are blocked, like when they're already running.
Solution:
Allow its FileType autocommands to nest: `do_filetype_autocmd(force=true)`.
Also error if executing them fails, rather than silently return wrong defaults.
Endless nesting from misbehaving scripts should be prevented by the recursion
limit in apply_autocmds_group, which is 10.
Problem:
On windows, if a drive-relative path doesn't contain a slash,
`path_to_absolute` can't split out the relative component, causing
expansion to fails. e.g., `c:` `c:.` `c:..` `c:foo.md`
Solution:
For these cases, we can pass letter and colon to `path_full_dir_name`.
Notably, `..` is included as well.
if that relative path exists, it can be expanded correctly.
Problem:
When `to_vim_value[info.metatype](info, value)` is called, a list value
such as `{'eob:a'}` is treated like a map, which generates `1:eob:a`.
Note: commands like `:lua vim.opt.wildmode={'longest:full'}` are not an
issue because only cases harcoded in `key_value_options` have metatype `map`.
Solution:
Check for array type and use the same logic as in array metatypes.
Problem:
Right-click menu fails with E335 when using V in Insert mode (after
i_ctrl-o). The mode detection checks restart_edit before VIsual_active,
incorrectly selecting Insert mode binding even when Visual mode is
active.
Solution:
Check Visual mode before Insert mode, to match get_menu_mode() priority
order.
Problem:
When code lens is enabled, `on_attach` is executed, but it does not trigger a redraw. Another event, eg, moving the cursor, is required to trigger a redraw and execute the decoration provider's `on_win`.
Solution:
Trigger a `redraw` after each request is completed.
Problem:
"[Process exited]" is implemented in C with anonymous namespace
and users have no way to hide it.
Solution:
- Handle "TermClose" event in Lua.
- User can delete the "nvim.terminal" augroup to avoid "[Process exited]".
Problem:
Building with GCC emits -Wfree-nonheap-object warnings in lua_cjson code
when calling strbuf_free(), despite the free being conditionally valid.
This is an unresolved GCC issue: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98753
Solution:
Supress warnings with the -Wno-free-nonheap-object flag for now.
Problem:
Applications running inside :terminal that use DEC private mode 2026
(synchronized output) to batch screen updates get garbled rendering.
Neovim's embedded libvterm does not handle mode 2026, so the
synchronization sequences are ignored and intermediate screen states
leak through as visual corruption.
Solution:
Add mode 2026 support to libvterm's state machine and wire it through
to terminal.c. When an application enables mode 2026, invalidation of
the terminal buffer is deferred until the application disables it,
causing all accumulated screen updates to flush as a single
atomic refresh.
* fix(terminal): harden sync output redraw gating
Problem:
The initial mode 2026 implementation gated invalidate_terminal()
but missed three other redraw paths: term_sb_push/term_sb_pop
bypassed the gate by directly adding to invalidated_terminals,
refresh_timer_cb could fire mid-sync flushing partial state, and
the 10ms timer delay after sync-end left a window for stale
repaints.
Solution:
- Gate term_sb_push/term_sb_pop during synchronized output
- Skip syncing terminals in refresh_timer_cb
- On sync end, schedule a zero-delay full-screen refresh via
sync_flush_pending flag in terminal_receive()
- Add news.txt entry for mode 2026 support
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* test(terminal): add vterm unit tests for mode 2026
Add unit-level tests for synchronized output (mode 2026) to
vterm_spec.lua, covering settermprop callbacks and DECRQM
query/response.
Suggested-by: justinmk
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix(terminal): address review feedback for mode 2026
- Use multiqueue_put(main_loop.events) instead of restarting the
global refresh timer on sync end, to avoid affecting other
invalidated terminals.
- Add screen:expect_unchanged() to verify screen doesn't update
during sync mode.
- Merge buffer-lines test into existing test.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
