Problem:
On "arm clang unittest" CI job, `make unittest` sometimes fails with
lots of these messages:
FAILED test/unit/testutil.lua @ 773: ...
test/unit/testutil.lua:773:
test/unit/testutil.lua:297: declaration specifier expected near '_Static_assert' at line 429
exit code: 256
stack traceback:
test/unit/testutil.lua:773: in function 'itp_parent'
test/unit/testutil.lua:811: in function <test/unit/testutil.lua:801>
Solution:
Update filter_complex_blocks.
Problem:
Many color schemes assume the Comment hl group is dim text and use it
for secondary text, decorations, or parts of UI. This is true for many
color schemes but not all.
Solution:
Introducing a new highlight group with a more specific meaning, similar
to Underlined or Ignore.
The new group links to Comment by default so the behavior is unchanged
for color schemes that don't define it.
Problem:
The Lua test harness still ran through standalone -ll mode, so tests
depended on the low-level Lua path instead of the regular Nvim Lua
environment. That also meant os.exit() coverage had to carry an ASAN
workaround because Lua's raw process exit skipped Nvim teardown and let
LeakSanitizer interfere with the observed exit code.
Solution:
Run the harness and related fixtures with nvim -l. Patch os.exit() in
the main Lua state to exit through getout(), so scripts observe normal
Nvim shutdown while standalone -ll remains available for generator-style
scripts. As a consequence, the startup test can assert os.exit() without
disabling leak detection.
AI-assisted: Codex
Problem: Pasting ". register without TextPut* autocommands breaks
subsequent TextPut* autocommands (after 9.2.0470).
Solution: Only decrement add_last_insert if it has been incremented
(zeertzjq).
closes: vim/vim#20192a70b7a85af
Problem: No way to hook into put commands
(yochem)
Solution: Introduce TextPutPre and TextPutPost autocommands
(Foxe Chen).
fixes: vim/vim#18701closes: vim/vim#20144e0781bd5bf
Co-authored-by: Foxe Chen <chen.foxe@gmail.com>
Problem: vim.ui_attach() callback for nvim_echo() call that spoofs an
internal message kind is executed in fast context.
Solution: Set msg_show callback |api-fast| context dynamically at
external message callsites, and for internal list_cmd",
"progress" and "shell*" messages.
Problem:
When using prompt_appendbuf with multi-element list,
the first item is concated and rest replace the prompt instead of
inserting the lines before the prompt.
Solution:
Concat first element with replace_buf and insert the rest of the list
with set_buffer_lines.
Signed-off-by: Tomas Slusny <slusnucky@gmail.com>
Problem: the bookmarks list can have duplicate entries, more often
in win32 (due to mixed slashes and capitalization) and when
g:netrw_keepdir=0 (which could introduce relative paths).
Duplicate entries could be: C:\foo\BAR\baz.file
c:\foo\bar\baz.file
c:/foo\BAR/baz.file
BAR/baz.file
Solution: Normalize the paths and make sure they are always absolute
(J. Paulo Seibt).
closes: vim/vim#2019474019bea8c
Co-authored-by: J. Paulo Seibt <jpseibt@gmail.com>
Problem: - "bufwrite" message identifier is encoded in the message ID
of a "progress" kind message (since ff68fd6b); UI2 does not
allow routing by message ID.
- No documented way to set a default message target for all
but a few kinds (without copying all of |ui-messages| kinds
to cfg.msg.targets).
- A user adding a message route for the documented empty ""
kind can result in unexpected behavior.
- Showing duplicate message (x) indicator in msg and cmd
targets simultaneously is unsupported.
- Manually triggering CursorMoved autocommand to add matchparen
highlighting in the cmdline.
Solution: - Match cfg.msg.targets keys as Lua pattern to a message ID.
- Recognize "default" as key in cfg.msg.targets, drop the
undocumented cfg.msg.target field.
- Don't try to get configured target for "" message kind/trigger.
- Maintain msg indicator virtual text for the cmd and msg target.
- Add matchparen highlighting by directly calling the Lua module
(possible since b813c7e0).
Problem: 'title' is updated when changing the name of a non-current
buffer with nvim_buf_set_name().
Solution: Set RedrawingDisabled when renaming the buffer.
Problem: Message redirection column for captured output is not reset
after :echon since (4260f73, e63346df).
Solution: Ensure msg_ext_append is set before the kind with :echon.
Problem: some functions can be run from the sandbox
Solution: Block them, so they are not accessible from a modeline
(q1uf3ng)
closes: vim/vim#19975fcc4276db3
Co-authored-by: q1uf3ng <q1uf3ng@protone.me>
Problem: runtime(netrw): bookmarking directory uses current dir
Solution: Correctly handle netrw actual directory (J. Paulo Seibt)
fixes: vim/vim#10481closes: vim/vim#20169ec76ac620b
Co-authored-by: J. Paulo Seibt <jpseibt@gmail.com>
Problem:
Regression from c822a2657c: `vim.wait(0)` does not call `loop_poll`,
so `vim.wait(1)` is needed to "yield" from Lua.
Solution:
- Ensure that `vim._core.loop_poll()` is always called, even when `time=0`.
- Document how to interrupt Lua code (ctrl-c).
ref https://github.com/neovim/neovim/issues/6800
Problem: <Esc> in a Select-mode tabstop leaves the session and highlight active.
CursorMoved isn’t triggered since the cursor doesn’t move.
Solution: use ModeChanged (s:n) instead. Defer with vim.schedule() to avoid transient
s:n from jump().
Problem: The four pointer-resolution loops in u_read_undo() lack
an i != j guard, so a header whose uh_next.seq equals
its own uh_seq resolves uh_next.ptr to itself. On
buffer close, u_freeheader() sees uhp->uh_next.ptr !=
NULL and skips updating b_u_oldhead, so u_blockfree()
dereferences the freed header on the next iteration.
The same pattern applies to uh_prev, uh_alt_next and
uh_alt_prev. A crafted .un~ file in the same directory
as a text file can trigger the use-after-free and
subsequent double-free when the buffer is closed.
(Daniel Cervera)
Solution: Add an i != j guard to each of the four resolution
loops, matching the guard already present in the
duplicate-detection loop above.
closes: vim/vim#20168
Supported by AI
4f610f07b7
Co-authored-by: Christian Brabandt <cb@256bit.org>
Problem: #39625 retained `matchparen.vim` as a shim that sources the new
`matchparen.lua` entrypoint, which is redundant since Lua runtime/plugin
files are sourced automatically and incurs startuptime cost solely for
the sake of not touching the tests.
Solution: Remove the shim and `source` the Lua plugin in directly in
tests.
Problem: Crash with invalid shellredir/shellpipe value
(bfredl)
Solution: Validate the option and allow only a single "%s".
fixes: vim/vim#20157closes: vim/vim#2015984ae09dd79
Co-authored-by: Christian Brabandt <cb@256bit.org>
Problem: 'findfunc' only allows extra info for cmdline completion, not
for actually finding files (Maxim Kim, after 9.2.0451).
Solution: Handle returning a list of dicts when actually finding files.
Also fix crash on NULL string (zeertzjq).
fixes: vim/vim#20163closes: vim/vim#201649694ff58fe
Problem: No test that "abbr" in customlist completion is shown in pum.
Solution: Add some "abbr" fields to the existing test (zeertzjq).
closes: vim/vim#20165b207b5a2a3
Problem: 'findfunc' can't return extra info for cmdline completion
(Maxim Kim).
Solution: Handle 'findfunc' return value in cmdline completion like that
of "customlist" functions (zeertzjq).
fixes: vim/vim#20155closes: vim/vim#2015858124789aa
Problem: Trying to execute code action on an active plugin without
updates leads to nothing. It is more useful if code actions "do
something" on a bigger portion of the confirm buffer.
Solution: Suggest "delete" code action even for active plugins. Trying
to execute it will first show a confirmation buffer with relevant
warning of why this might be not a good idea. Confirming will delete
a plugin.
Problem:
- Empty ranges have different `<`, `<=`, `has` and `intersect` semantics compared to regular ranges.
- `to_inclusive_pos` assumes that the end position of a range is exclusive, which is not true for empty ranges
Solution:
Special case empty ranges in these operations.
Problem:
When `:!` writes shell output to a buffer, write_output() splits on `\r`, `\n`,
and `\r\n`, replacing the terminator byte with NUL. For a binary-mode buffer
this is wrong: `\r` should be preserved verbatim, not treated as a line
terminator. This wrong behavior causes a file like `\r\n` round-trips through
`:%!cat` to `\n`.
This was masked when 'shelltemp' was enabled, because output went through a temp
file and the regular file I/O path handled binary-mode correctly. Switching the
default to 'noshelltemp' exposed the bug, since output is now piped directly
into write_output().
Solution:
In `write_output()`, skip the `\r` and `\r\n` splits for a binary-mode buffer;
only split on `\n`.
Problem: read_compound() in spellfile.c computes the size of the regex
pattern buffer using signed-int arithmetic on the attacker
controlled SN_COMPOUND sectionlen. With sectionlen=0x40000008
and UTF-8 encoding active the multiplication wraps to 27 while
the per-byte loop writes up to ~1B bytes, overflowing the heap.
Reachable when loading a crafted .spl file (e.g. via 'set spell'
after a modeline sets 'spelllang'). The cp/ap/crp allocations
have the same int + 1 overflow class (Daniel Cervera)
Solution: Use type size_t as buffer size and reject values larger than
COMPOUND_MAX_LEN (100000). Apply the same size_t treatment to
the cp/ap/crp allocations.
Github Advisory:
https://github.com/vim/vim/security/advisories/GHSA-q4jv-r9gj-6cwv9299332917
Co-authored-by: Christian Brabandt <cb@256bit.org>
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Problem:
When showing the :connect menu, it is useful to know which servers
are most-recently active. But we don't have a good way to detect that.
Solution:
- Introduce `v:useractive`.
- Include this timestamp in `serverlist({info=true})`.
Co-authored-by: Justin M. Keyes <justinkz@gmail.com>
Signed-off-by: Szymon Wilczek <swilczek.lx@gmail.com>
Problem:
parser_gc() calls ts_parser_delete() but leaves the userdata pointer
pointing to freed memory. If the GC finalizer runs at an unexpected time
(e.g. inside nvim_buf_get_lines #39411), a stale pointer could cause a crash.
Solution:
- NULL out `*ud` after ts_parser_delete() in parser_gc()
- Update parser_check() to handle NULL with a clear error message,
guarding all parser methods against UAF
Co-authored-by: Lewis Russell <lewis6991@gmail.com>
Signed-off-by: Szymon Wilczek <swilczek.lx@gmail.com>
Problem:
For a given position, it is not easy to compare which of several other positions is closest to it.
Solution:
Add support for converting `vim.Pos` to a buffer byte offset.
This allows for sorting, e.g:
```lua
table.sort(positions, function(pos1, pos2)
return pos1:to_offset() < pos2:to_offset()
end
```
Or a binary search, e.g:
```lua
vim.list.bisect(positions, pos, { key = function(pos) return pos:to_offset() end })
```
Problem: Internal progress messages use the "nvim" source (since
ff68fd6b), plugins shouldn't be allowed to set the progress
message source to "nvim". The message ID used for internal
progress messages is not identifiable as such.
Solution: Disallow setting opts->source to "nvim" with nvim_echo().
Refactor msg_progress() and callees to bypass nvim_echo().
Prepend message id for internal progress messages with "nvim.".
Improve the vim.iter annotations with richer generics that track element and
tuple types through iterator pipelines, including multi-value stages and
list-specific methods.
Extend the LuaCATS parser and vimdoc generator so those richer generic classes
and overloads round-trip into the generated help. These annotations are only
supported by EmmyLua, so LuaLS still uses a broader fallback in _meta.lua.
AI-assisted: Codex
Problem: Entering the pager fails if <ESC> is remapped to :fclose by user.
Solution: Avoid executing mappings with nvim_feedkeys() that closes expanded cmdline.
Problem:
UI tools and orchestration engines need more context than just raw
socket addresses from serverlist(). Without knowing if a server belongs
to the current instance or knowing its PID, UIs cannot display
meaningful options to users.
Solution:
- Added the `info=v:true` option to `serverlist()`.
- When `info` is requested, it implies `peer=true` and returns a list of
dictionaries (defined as `vim.ServerInfo`) with `addr`, `pid` and
`own`.
- Uses an RPC request to `getpid()` across the socket to fetch the
peer's actual process ID.
Signed-off-by: Szymon Wilczek <swilczek.lx@gmail.com>
Problem:
Can't get a command's description from nvim_get_commands when
cmd is string.
Solution:
Returns "desc" field in nvim_get_commands.
`definition` is now empty when cmd is function type.
Problem: When find_start_brace() scans backwards for the enclosing
block, '{' and '}' inside // and /* */ comments are counted,
producing wrong indent for code following such comments
(rendcrx).
Solution: Implement FM_SKIPCOMM in findmatchlimit() to track block-
comment state and skip matches inside comments. Pass
FM_SKIPCOMM from cindent's call sites
(find_start_brace, find_match_char, cin_iswhileofdo,
get_c_indent).
fixes: vim/vim#4
fixes: vim/vim#648
fixes: vim/vim#19578closes: vim/vim#19581closes: vim/vim#20111c06002f3cb
Co-authored-by: magnus-rattlehead <magnus-rattlehead@users.noreply.github.com>
Problem: off-by-one bug in s:NetrwUnMarkFile()
Solution: Correctly loop through all buffers to unlet all variables
(J. Paulo Seibt)
When the function loops through buffers to clear s:netrwmarkfilelist_#
and s:netrwmarkfilemtch_#, it skips the last one at bufnr('$'), messing
up mark highlights and causing other functions that operate on those
arrays (like delete or rename) to target stale marked files.
The bufnr() help page says that bufnr("$") returns the highest buffer
number of existing buffers, so while ibuf < bufnr("$") does not clear
the last buffer-local arrays.
To reproduce:
Just opening a fresh Vim and running :Ex opens a netrw buffer at the
highest number. Then, typing mu after marking some files triggers the
mark highlight bug, and finally typing D would act like calling the
delete function against the previous marked files, as the buffer-local
arrays where not touched by s:NetrwUnMarkFile.
closes: vim/vim#201297ccc273a4c
Co-authored-by: J. Paulo Seibt <jpseibt@gmail.com>
Problem: When closing gvim with an unsaved unnamed buffer, choosing
"Yes" in the "Save changes?" dialog and then "Cancel" in the
file selection dialog either silently writes the buffer to a
file named "Untitled" (overwriting any existing file with
that name) or discards the buffer altogether
(vibs29, after v9.1.0265).
Solution: In dialog_changed(), if browse_save_fname() leaves the buffer
without a file name, treat it as a cancel and return without
saving. Also stop clearing the modified flag in the restore
path on write failure, so the unsaved changes are kept and
the caller (e.g. gui_shell_closed()) can also cancel the
close. Pre-fill the file dialog with "Untitled" to match
the preceding "Save changes to ..." prompt. Add a test for
the write-failure path (Hirohito Higashi).
fixes: vim/vim#20132closes: vim/vim#20143cf947e7ef0
Co-authored-by: Hirohito Higashi <h.east.727@gmail.com>
