Problem: terminal: redraws are slow (Mao-Yining)
Solution: Disable redrawing in handle_movecursor()
(Yasuhiro Matsumoto)
handle_movecursor callback was calling update_cursor() with redraw=TRUE
on every cursor move inside vterm_input_write(). This triggered
gui_mch_flush() (GdiFlush + DWriteContext_Flush) and TextChangedT
autocmd for each cursor move. ConPTY output contains ~17 cursor moves
per 4KB chunk, each flush taking ~5ms, resulting in 80-110ms per chunk.
Fix by passing FALSE to update_cursor() in handle_movecursor since
write_to_term() already calls update_cursor() with proper redraw after
vterm_input_write() finishes.
Also set vterm_screen_set_damage_merge() to VTERM_DAMAGE_SCROLL so that
damage callbacks are buffered until vterm_screen_flush_damage() instead
of being emitted per cell.
fixes: #19845closes: #19846
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: When a popup window overlapping a status line is closed or
hidden, the status line is not redrawn, leaving ghost
artifacts from the popup.
Solution: popup_free() and popup_hide() call
redraw_all_later(UPD_NOT_VALID) which marks window contents
for redraw but does not set w_redr_status. The diff-based path
in may_update_popup_mask() that normally sets w_redr_status
is skipped when redrawing_all_win is TRUE. Add status_redraw_all
calls to ensure status lines are properly redrawn
(Yasuhiro Matsumoto).
closes: #19830
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: tabpanel: corrupted tabpanel during scrolling causing flicker
Solution: When the tabpanel is visible, force a line-by-line redraw in
win_do_lines() similarly to popup handling (Yasuhiro Matsumoto).
When a vertical tabpanel is visible, terminal scroll operations in
win_do_lines() affect the full screen width, corrupting the tabpanel
area. The tabpanel is then redrawn via redraw_tabpanel, causing visible
flicker. Return FAIL to force line-by-line redraw instead, analogous to
the existing popup_visible check.
closes: #19832
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: memory leak in add_mark()
Solution: Free lpos in the error case when it hasn't been added to the
dict yet (Huihui Huang)
closes: #19827
Signed-off-by: Huihui Huang <625173@qq.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: Unnecessary memory allocation in set_callback(); after
set_callback(), callers must manually free the source
callback's name if cb_free_name is set.
Solution: Refactor set_callback() to re-use the callback name when
possible to avoid extra memory allocations and clean up so the
callers do not have to take care themselves (Shane Harper).
closes: #19831
Signed-off-by: Shane Harper <shane@shaneharper.net>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: The visual selection size is not displayed in the showcmd area
when entering visual mode from a script or mapping, because
char_avail() incorrectly reports input as pending. This causes
test failure on CI with the ASAN CI runner.
Solution: Replace char_avail() with explicit checks for an empty stuff
buffer, empty typeahead buffer, and not running a script
(zeertzjq).
related: #19801
closes: #19824
Co-authored-by: zeertzjq <zeertzjq@outlook.com>
Signed-off-by: zeertzjq <zeertzjq@outlook.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: tests: Test_popup_opacity_vsplit() fails in a wide terminal
(after v9.2.0230)
Solution: Reduce terminal window size to 60, force termguicolors to make
the opacity visible; enable termguicolors so that the opacity
is visually apparent in the screen dump.
related: #19824
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: w_locked can be bypassed when recursively set if not restored
to its prior value.
Solution: Rather than save/restore everywhere, just make it a count,
like other locks (Sean Dewar)
Requires the previous commit, otherwise b_nwindows will be wrong in
tests, which causes a bunch of weird failures.
closes: #19728
Signed-off-by: Sean Dewar <6256228+seandewar@users.noreply.github.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: close_buffer() callers incorrectly handle b_nwindows,
especially after nasty autocmds, allowing it to go
out-of-sync. May lead to buffers that can't be unloaded, or
buffers that are prematurely freed whilst displayed.
Solution: Modify close_buffer() and review its callers; let them
decrement b_nwindows if it didn't unload the buffer. Remove
some now unneeded workarounds like 8.2.2354, 9.1.0143,
9.1.0764, which didn't always work (Sean Dewar)
close_buffer() now doesn't decrement b_nwindows when not unloading buf, or when
buf isn't w_buffer after autocmds (they would've already decremented it).
Callers are now expected to decrement b_nwindows if w_buffer is not NULL after
close_buffer(), and when still intending to switch buffers or close win, for two
reasons:
- close_buffer() autocmds may have switched buffers. The new w_buffer's
b_nwindows would also need decrementing.
- After close_buffer(), callers may opt to not switch w_buffer or close win.
b_nwindows would need to be incremented again. (unless w_buffer is NULL from
being unloaded; callers are already forced to find a new buffer then)
These were the main causes of b_nwindows bugs, as these cases could not be
reliably detected, and went largely unhandled.
NOTE: if close_buffer() autocmds switch buffers, close_buffer() is not called
for that new buffer before decrementing b_nwindows. This may skip side-effects
like from 'bufhidden', but I think it's mostly harmless, and was already
happening in other places.
Let's see how this goes... Other details: (I have lots to say!)
It's OK to pass a win to close_buffer() that isn't showing buf (used by
set_curbuf()). In that case, we skip some side-effects and don't decrement
b_nwindows, but may still unload buf if hidden.
buf_freeall() now returns whether it freed anything. Removes some repeated
checks in close_buffer().
Preserve close_buffer()'s behaviour when called by win_free_popup() after its
popup was already removed from the window list. This made win_valid_any_tab()
return FALSE, so we skip things that originally checked it in that case.
Add "set_context" to close_buffer() to preserve do_ecmd()'s behaviour of only
setting b_last_cursor and/or calling buflist_setfpos() when not splitting
(see 7.2.041: https://groups.google.com/g/vim_dev/c/ZGgNvaylNzI/m/WHxjhnuxqB0J)
Without this, Test_marks_cmd() fails from its ' mark differing. Don't use
oldwin though; it's not always the window with the closed buf, especially
after BufLeave autocmds in do_ecmd(). Also, only set context if win is really
displaying buf.
Don't bail in do_ecmd() if buf was deleted but curwin->w_buffer is NULL; that
leaves curwin open to a NULL buffer! Use lastbuf instead, like set_curbuf().
I don't think it's possible for buf to be deleted by close_buffer() anyway, as
b_locked was set (which I can't see a way to bypass, unlike b_locked_split).
Maybe such checks can be removed, but I'd rather not risk that here.
Don't set curwin to previouswin in set_curbuf(); shouldn't be needed, otherwise
may lead to curbuf != curwin->w_buffer if autocmds switch to a window showing
buf, as that skips enter_buffer()? Was introduced back in 7.3.557 to avoid
cases where autocmds switch windows, possibly leaving previouswin with a NULL
buffer. Since 7.4.2312 and 7.4.2328, close_buffer() and buf_freeall() already
handles this. I've added an assert() as a sanity check anyway.
In free_all_mem(), set b_nwindows to 0 before close_buffer() so buffers can be
wiped if still in a window before win_free_all(). Needed as close_buffer() now
skips unloading buffers that aren't hidden if win is NULL. If it's possible for
free_all_mem()'s :tabonly! and :only! to not close all windows before freeing,
then this issue was also previously possible if b_nwindows > 1.
related: #19728
Signed-off-by: Sean Dewar <6256228+seandewar@users.noreply.github.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: if close_buffer() in set_curbuf() unloads curbuf, NULL pointer
accesses may occur from enter_buffer() calling
end_visual_mode(), as curbuf is already abandoned and possibly
unloaded. Also, selection registers may not contain the
selection with clipboard+=autoselect(plus).
Solution: Move close_buffer()'s end_visual_mode() call to buf_freeall(), after
any autocmds that may restart it, but just before freeing anything
(Sean Dewar)
related: #19728
Signed-off-by: Sean Dewar <6256228+seandewar@users.noreply.github.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: Compile error when building without channel feature
(John Marriott, after v9.2.0250)
Solution: Update ifdefs and move implementation out of FEAT_JOB_CHANNEL
(Yasuhiro Matsumoto)
Move build_argv_from_list() and mch_get_cmd_output_direct() out of
FEAT_JOB_CHANNEL guards so that system() with a List argument works
in builds that have FEAT_EVAL but not FEAT_JOB_CHANNEL (e.g.
FEAT_NORMAL without GUI).
related: #19791
closes: #19826
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: system() and systemlist() only accept a String, requiring
manual shell escaping for arguments with special characters.
Solution: Accept a List as the first argument and execute the command
bypassing the shell (Yasuhiro Matsumoto).
fixes: #19789closes: #19791
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: json_decode() accepted keywords case-insensitively, violating
RFC 7159. Both json_decode() and js_decode() silently accepted
lone surrogates, which are invalid Unicode.
Solution: Only allow lowercase keyword in json_decode(), reject lone
surrogates, improve encoding performance in write_string() and
blob byte serialization.
1. Fix surrogate pair range check (0xDFFF -> 0xDBFF) so only high
surrogates trigger pair decoding. Reject lone surrogates that do
not form a valid pair instead of producing invalid UTF-8.
2. Use case-sensitive matching for JSON keywords (true, false, null,
NaN, Infinity) in json_decode() per RFC 7159. js_decode() retains
case-insensitive behavior.
3. Replace double ga_append() calls for escape sequences with single
GA_CONCAT_LITERAL() calls, halving function call and buffer growth
check overhead.
4. Replace vim_snprintf_safelen() for blob byte encoding (0-255) with
direct digit conversion.
closes: #19807
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: popup: popups may not wrap as expected
(Enrico Maria De Angelis, after v9.1.0949)
Solution: don't shift popupwin left when 'wrap' is on and maxwidth is
set (Hirohito Higashi)
When a non-fixed popup with 'wrap' enabled and an explicit maxwidth was
placed near the right edge of the screen, the shift-left logic increased
maxwidth beyond the user-specified value, preventing text from wrapping.
Instead cap the shift amount so that maxwidth does not exceed w_maxwidth
when wrapping is enabled, letting text wrap as expected.
fixes: #19767closes: #19809
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Hirohito Higashi <h.east.727@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: xxd: color output detection is broken
(Juergen Weigert)
Solution: Fix the issues (Hirohito Higashi)
- Disable auto color when output goes to a file (two-argument form)
- Check TERM variable: disable color when unset, empty or "dumb"
- Add color_forced flag to preserve -R always behavior
- Add tests for the new behavior
fixes: #19790closes: #19813
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Hirohito Higashi <h.east.727@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: memory leak in eval8()
Solution: Free type_list() even when evaluate is FALSE
(Huihui Huang)
closes: #19819
Signed-off-by: Huihui Huang <625173@qq.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: memory leak in change_indent()
Solution: Free orig_line in the error case
(Huihui Huang)
closes: #19820
Signed-off-by: Huihui Huang <625173@qq.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: memory leak in check_for_cryptkey()
Solution: Free cryptkey if it had been allocated
(Huihui Huang)
closes: #19821
Signed-off-by: Huihui Huang <625173@qq.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: tests: Test_visual_block_hl_with_autosel() is flaky on ASAN CI
(after v9.2.0221)
Solution: Add a TermWait() call
related: #19784
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: Looking up highlight group names uses a linear scan of the
highlight table, which is slow at startup when many groups
are defined.
Solution: Use a hashtable for O(1) highlight group name lookup.
(Yasuhiro Matsumoto).
Benchmark (523 highlight groups, :highlight + highlight link + syntax
keyword + hlID() + redraw, 20 iterations):
master: 0.057 sec (~295,000 ops/sec)
this branch: 0.036 sec (~463,000 ops/sec)
speedup: ~1.57x
closes: #19788
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
- Highlight keywords "function" and "namespace" with
the "Keyword" group ("shFunctionKey").
- Highlight function body delimiters "{" and "}" with the
"Delimiter" group ("shFunctionExprRegion").
- Highlight function body delimiters "(" and ")" with the
"Operator" group ("shFunctionSubShRegion").
- Also, follow one style in folding all supported variants
of function bodies for grouping commands too by enclosing
a delimited function body, e.g. "{" and "}", in a fold and
leaving its function header, e.g. "function f()", out of
it when the header is written on a separate line.
To restore previous colouring, add to "after/syntax/sh.vim":
------------------------------------------------------------
hi link shFunctionKey Function
hi link shFunctionExprRegion Function
hi link shFunctionSubShRegion Function
------------------------------------------------------------
fixes: https://github.com/vim/vim/pull/19638#issuecomment-4052635546closes: #19638
Signed-off-by: Aliaksei Budavei <0x000c70@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: Changing the 'signcolumn' may cause flicker, because it uses
the P_RCLR flag for redrawing (Yggdroot)
Solution: Change 'signcolumn' to use P_RWIN instead of P_RCLR, matching
the behavior of 'number' and 'foldcolumn'
(Yasuhiro Matsumoto).
fixes: #19663closes: #19713
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: showmode message may not be displayed (Yee Cheng Chin)
Solution: Don't call char_avail() in skip_showmode(), but check
for anything in the stuff and typeahead buffer
(Hirohito Higashi).
Fix "-- VISUAL --" not shown when terminal responses are pending
When starting Vim with a script that enters Visual mode (e.g.
"vim -S script.vim"), the "-- VISUAL --" mode message sometimes
doesn't appear. This happens because skip_showmode() calls
char_avail(), which reads raw terminal input and picks up terminal
response escape sequences (e.g. t_RV response). Combined with
!KeyTyped (which is TRUE after script execution), this causes
skip_showmode() to return TRUE, preventing the mode from being
displayed.
Fix this by checking only the stuff buffer and typeahead buffer for
pending characters, instead of char_avail() which also reads raw
terminal input. This way, terminal response sequences no longer
interfere with mode display.
fixes: #16620closes: #19801
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: Christian Brabandt <cb@256bit.org>
Signed-off-by: Hirohito Higashi <h.east.727@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
\a in Vim's regex matches [A-Za-z], not the BEL character (0x07).
This caused the OSC 8 hyperlink stripping pattern to incorrectly
match alphabetic characters, breaking man page display.
Use \x07 (inside []) and %x07 (outside []) to correctly match BEL.
closes: #19806
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: stack-overflow with deeply nested data in json_encode/decode()
(ZyX-I)
Solution: Add depth limit check using 'maxfuncdepth' to
json_encode_item() and json_decode_item() to avoid crash when
encoding/decoding deeply nested lists, dicts, or JSON arrays/objects,
fix typo in error name, add tests (Yasuhiro Matsumoto).
fixes: #588closes: #19808
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: test: Test_close_handle() is flaky
Solution: Use WaitForAssert() to wait for the channel to be closed
(Yasuhiro Matsumoto)
Ch_close_handle() did not wait for the channel to be fully closed,
which could cause Ch_CloseHandler to fire during the next test's
GetPort() sleep loop, resulting in E906.
Wait for ch_status() to become 'closed' before returning.
closes: #19797
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: Compiler warning in strings.c
(Timothy Rice, after v9.2.0031)
Solution: Return early when str_m is zero
(Hirohito Higashi)
fixes: #19795closes: #19800
Signed-off-by: Hirohito Higashi <h.east.727@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: fileinfo not shown after :bd of last listed buffer
(memeplex)
Solution: Set need_fileinfo to true in empty_curbuf()
(Hirohito Higashi)
When deleting the last listed buffer with :bd, the new empty buffer's
file info (e.g. "[No Name]" --No lines in buffer--) was not displayed.
do_ecmd() only calls fileinfo() for existing buffers (oldbuf), not for
newly created empty buffers.
Set need_fileinfo in empty_curbuf() so the file info is displayed after
redraw.
fixes: #548closes: #19802
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Hirohito Higashi <h.east.727@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: AROS/MorphOS/AmigaOS tiny/small link error
Solution: Include get_cmd_output by defining HAVE_LOCALE_H
(Ola Söder)
closes: #19792
Signed-off-by: Ola Söder <rolfkopman@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: popup: opacity not working across vert splits
Solution: Iterate over the full popup width to find all underlying
windows (Yasuhiro Matsumoto)
redraw_win_under_opacity_popup() only checked the popup's left column
to find underlying windows. When a popup spans a vertical split, the
right-side window was never forced to redraw, causing blended
ScreenAttrs values to accumulate across redraw cycles.
Iterate across the full popup width to find and redraw all underlying
windows.
closes: #19737
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: In XTerm, with 'xtermcodes' enabled (default), vim will
request keypad keys after editing pad keys, and will remove
the latter when they're duplicates of the former.
Solution: When a termcode reply is a keypad key and it would replace a
different key, skip it.
fixes: #19182 (case 2)
related: #19643
closes: #19644
Signed-off-by: AstroSnail <astrosnail@protonmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: still possible flicker
Solution: Use redraw_as_cleared() instead of screenclear() to avoid
flicker (Yasuhiro Matsumoto).
When the window width matches the frame width, screenclear() causes
visible flicker. Use redraw_as_cleared() instead which marks the
screen for redraw without actually clearing it.
closes: #19732
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: MS-Windows: CSI sequences may be written to screen
(after v9.2.0200).
Solution: Move the CSI parsing and skipping logic outside of the USE_VTP
guard in mch_write() so that sequences are always consumed
even if VTP is inactive (Yasuhiro Matsumoto).
After v9.2.0200 added send_decrqm_modes(), running vim -u NONE -U NONE
-N on Windows displays raw ESC[?2026$p on screen. This is because
USE_VTP is false when termguicolors is off and t_colors < 256, and the
CSI skip logic in mch_write() was inside the if (USE_VTP) block, so
s/len were never advanced and the escape sequence leaked out as plain
text.
Move the CSI parsing/skipping outside the USE_VTP guard so sequences are
always consumed. Only the vtp_printf() pass-through for DECSCUSR (q)
remains conditional on USE_VTP.
related: #19755
closes: #19762
Signed-off-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: No 'incsearch' highlighting support for :uniq
Solution: Add :uniq support (Hirohito Higashi)
closes: #19780
Signed-off-by: Hirohito Higashi <h.east.727@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: runtime(compiler): No compiler plugin for just
Solution: Add a compiler plugin for the just command runner, add a test
(Aditya Malik)
Sets makeprg and a custom errorformat to parse just's multi-line
error output into quickfix entries with file, line, column, and
message. Includes a test.
Reference:
- https://github.com/casey/justcloses: #19773
Signed-off-by: Aditya Malik <adityamalik2833@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: channel: 2 issues with out/err callbacks
Solution: fix indentation and output order with term_start()
(Hirohito Higashi).
When term_start() is called with err_cb (and optionally out_cb),
two issues occur:
1. Unexpected indentation in the terminal buffer display: stderr
arrives via a pipe which lacks the PTY ONLCR line discipline,
so bare LF moves the cursor down without a CR, causing each
subsequent line to be indented one column further.
2. stdout appears in the terminal and callbacks before stderr,
even when the child process writes to stderr first. This is
because channel_parse_messages() iterates parts in enum order
(PART_OUT before PART_ERR), so when both have buffered data
they are always processed in the wrong order.
Solution:
- In may_invoke_callback(), before writing PART_ERR data to the
terminal buffer, convert bare LF to CR+LF so that vterm renders
each line at column 0.
- In channel_parse_messages(), when about to process PART_OUT of a
terminal PTY job, if PART_ERR already has readahead data, skip
PART_OUT and process PART_ERR first. This works without any
artificial delay because channel_select_check() reads all ready
file descriptors into their readahead buffers in a single
select() pass before any callbacks are invoked; by the time
channel_parse_messages() runs, both buffers are populated and
the skip logic can enforce the correct order.
- Also add a per-line split for out_cb/err_cb on terminal PTY
jobs: instead of passing a potentially multi-line raw chunk to
the callback, split on NL and strip trailing CR so each callback
receives exactly one clean line.
Add Test_term_start_cb_per_line() to verify that err_cb and out_cb
each receive one line per call, with correct stderr-before-stdout
ordering, without any sleep between the writes.
fixes: #16354closes: #19776
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Hirohito Higashi <h.east.727@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
I don't know whether it is necessary to copy desktop files to runtime/
manually, but these two files seem not to have been updated for some
time.
The change in Makefile is because the previous pattern will throw an error:
C:\msys64\ucrt64\bin\sed.exe: -e expression #1, char 3: unknown command: `.'
which I am not sure why.
closes: #19777
Signed-off-by: Mao-Yining <mao.yining@outlook.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: Option handling for key:value suboptions is limited
Solution: Improve :set+=, :set-= and :set^= for options that use
"key:value" pairs (Hirohito Higashi)
For comma-separated options with P_COLON (e.g., diffopt, listchars,
fillchars), :set += -= ^= now processes each comma-separated item
individually instead of treating the whole value as a single string.
For :set += and :set ^=:
- A "key:value" item where the key already exists with a different value:
the old item is replaced.
- An exact duplicate item is left unchanged.
- A new item is appended (+=) or prepended (^=).
For :set -=:
- A "key:value" or "key:" item removes by key match regardless of value.
- A non-colon item removes by exact match.
This also handles multiple non-colon items (e.g., :set
diffopt-=filler,internal) by processing each item individually, making
the behavior order-independent.
Previously, :set += simply appended the value, causing duplicate keys to
accumulate.
fixes: #18495closes: #19783
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Hirohito Higashi <h.east.727@gmail.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Problem: "zb" scrolls incorrectly with cursor on fold.
Solution: Set w_botline to the line below the fold (zeertzjq).
related: neovim/neovim#38413
closes: #19785
Signed-off-by: zeertzjq <zeertzjq@outlook.com>
Signed-off-by: Christian Brabandt <cb@256bit.org>
Yasuhiro Matsumoto
Huihui Huang
Shane Harper
zeertzjq
Christian Brabandt
Sean Dewar
Ola Söder
Antonio Giovanni Colombo
Foxe Chen
Hirohito Higashi
Aliaksei Budavei
Hannah Kimura
Anakin Childerhose
AstroSnail
Aditya Malik
Mao-Yining