Problem: Clicking on window separator in statusline crashes Nvim due
to out of bound memory access
Solution: Check if the click location is within clicking range before
applying it.
Problem: Diff mode's inline highlighting is lackluster. It only
performs a line-by-line comparison, and calculates a single
shortest range within a line that could encompass all the
changes. In lines with multiple changes, or those that span
multiple lines, this approach tends to end up highlighting
much more than necessary.
Solution: Implement new inline highlighting modes by doing per-character
or per-word diff within the diff block, and highlight only the
relevant parts, add "inline:simple" to the defaults (which is
the old behaviour)
This change introduces a new diffopt option "inline:<type>". Setting to
"none" will disable all inline highlighting, "simple" (the default) will
use the old behavior, "char" / "word" will perform a character/word-wise
diff of the texts within each diff block and only highlight the
differences.
The new char/word inline diff only use the internal xdiff, and will
respect diff options such as algorithm choice, icase, and misc iwhite
options. indent-heuristics is always on to perform better sliding.
For character highlight, a post-process of the diff results is first
applied before we show the highlight. This is because a naive diff will
create a result with a lot of small diff chunks and gaps, due to the
repetitive nature of individual characters. The post-process is a
heuristic-based refinement that attempts to merge adjacent diff blocks
if they are separated by a short gap (1-3 characters), and can be
further tuned in the future for better results. This process results in
more characters than necessary being highlighted but overall less visual
noise.
For word highlight, always use first buffer's iskeyword definition.
Otherwise if each buffer has different iskeyword settings we would not
be able to group words properly.
The char/word diffing is always per-diff block, not per line, meaning
that changes that span multiple lines will show up correctly.
Added/removed newlines are not shown by default, but if the user has
'list' set (with "eol" listchar defined), the eol character will be be
highlighted correctly for the specific newline characters.
Also, add a new "DiffTextAdd" highlight group linked to "DiffText" by
default. It allows color schemes to use different colors for texts that
have been added within a line versus modified.
This doesn't interact with linematch perfectly currently. The linematch
feature splits up diff blocks into multiple smaller blocks for better
visual matching, which makes inline highlight less useful especially for
multi-line change (e.g. a line is broken into two lines). This could be
addressed in the future.
As a side change, this also removes the bounds checking introduced to
diff_read() as they were added to mask existing logic bugs that were
properly fixed in vim/vim#16768.
closes: vim/vim#168819943d4790e
Co-authored-by: Yee Cheng Chin <ychin.git@gmail.com>
Problem: fragile setup to get (preferred) keys from key_name_entry
(after v9.1.1179)
Solution: refactor the code further, fix a bug with "pref_name" key
entry introduced in v9.1.1180 (Yee Cheng Chin)
The optimization introduced for using bsearch() with key_name_entry
in vim/vim#16788 was fragile as it required synchronizing a non-obvious index
(e.g. IDX_KEYNAME_SWU) with the array that could be accidentally changed
by any one adding a key to it. Furthermore, the "pref_name" that was
introduced in that change was unnecessary, and in fact introduced a bug,
as we don't always want to use the canonical name.
The bug is triggered when the user triggers auto-complete using a
keycode, such as `:set <Scroll<Tab>`. The bug would end up showing two
copies of `<ScrollWheelUp>` because both entries end up using the
canonical name.
In this change, remove `pref_name`, and simply use a boolean to track
whether an entry is an alt name or not and modify logic to respect that.
Add test to make sure auto-complete works with alt names
closes: vim/vim#169877d8e7df551
In Nvim there is no `enabled` field, so put `is_alt` before `name` to
reduce the size of the struct.
Co-authored-by: Yee Cheng Chin <ychin.git@gmail.com>
Problem: Strange error with type for matchfuzzy() "camelcase".
Solution: Show the error "Invalid value for argument camelcase" instead
of "Invalid argument: camelcase" (zeertzjq).
Note that using tv_get_string() will lead to confusion, as when the
value cannot be converted to a string tv_get_string() will also give an
error about that, but "camelcase" takes a boolean, not a string. Also
don't use tv_get_string() for the "limit" argument above.
closes: vim/vim#16926c4815c157b
Problem: When searching for "Cur", CamelCase matches like "lCursor" score
higher than exact prefix matches like Cursor, which is
counter-intuitive (Maxim Kim).
Solution: Add a 'camelcase' option to matchfuzzy() that lets users disable
CamelCase bonuses when needed, making prefix matches rank higher.
(glepnir)
fixes: vim/vim#16504closes: vim/vim#1679728e40a7b55
Co-authored-by: glepnir <glephunter@gmail.com>
Problem: The current_pos.col was incorrectly updated to the length of
the matching text. This will cause the next search to start
from the wrong position.
Solution: current_pos has already been updated in search_str_in_line and
does not need to be changed (glepnir)
closes: vim/vim#169415753084042
Co-authored-by: glepnir <glephunter@gmail.com>
Problem: 'completefuzzycollect' does not handle dictionary correctly
Solution: check for ctrl_x_mode_dictionary (glepnir)
closes: vim/vim#16867587601671c
Cherry-pick a documentation fix from later.
Co-authored-by: glepnir <glephunter@gmail.com>
- Fix grammar
- Use "matches" instead of "items" ("completion candidates" is used in
some other places, but it's a bit verbose)
- "When set" is a bit vague, instead use "For specified modes"
closes: vim/vim#168711dc731a49f
Problem: endless loop with completefuzzycollect and no match found
Solution: move pointer to line end and break loop
closes: vim/vim#16820dd42b05f8a
Co-authored-by: glepnir <glephunter@gmail.com>
Problem: No cmdline completion for the 'completefuzzycollect' option
(after v9.1.1178)
Solution: Add cmdline completion for the 'completefuzzycollect' option,
improve its description in optwin.vim (zeertzjq).
closes: vim/vim#1681353d59ecc1d
No code change is needed in Nvim as Nvim uses expand_set_str_generic()
by default.
Problem: Unnecessary STRLEN() calls in insexpand.c (after 9.1.1178).
Solution: Use the already available length (zeertzjq).
closes: vim/vim#168144422de6316
Problem: potential out-of-memory issue in search.c
Solution: improve situation and refactor search.c slightly
(John Marriott)
- In function update_search_stat():
add a check for a theoretical null pointer reference, set and remember
the length of lastpat, remove the three calls to STRLEN() and use the
various string's associated lengths instead, add a check for an
out-of-memory condition.
- In function search_for_fuzz_match():
remove a call to strnsave() and thus avoid having to add a check for
an out-of-memory condition, also replace the call to STRLEN() by
ml_get_buf_len().
closes: vim/vim#16689b79fa3d9c8
Co-authored-by: John Marriott <basilisk@internode.on.net>
Originally, `:set completeopt+=fuzzy` did not affect how the candidate
list is collected in default keyword completion. A comment was added to
documentation as part of vim/vim#14912 to clarify it. vim/vim#15193 later changed the
fuzzy behavior to now change the candidate collection behavior as well
so the clarification in docs is now wrong. Remove them here.
closes: vim/vim#1565626e4b00002
Co-authored-by: Yee Cheng Chin <ychin.git@gmail.com>
Problem: Sorting of fuzzy filename completion is not stable
Solution: Compare indexes when scores are equal. Fix some typos.
(zeertzjq)
closes: vim/vim#1559358d705238c
Problem: completion does not respect completeslash with fuzzy
(egesip)
Solution: Change path separator on Windows, depending on 'completeslash'
option value (glepnir)
fixes: vim/vim#15392closes: vim/vim#15418b9de1a057f
Co-authored-by: glepnir <glephunter@gmail.com>
Problem: Ctrl-P not working by default
(Jesse Pavel, after v9.1.0598)
Solution: Revert part of v9.1.0598 and set cur_match_pos
correctly according to compl_dir_forward()
fixes: vim/vim#15370closes: vim/vim#1537913032a49b7
Co-authored-by: Christian Brabandt <cb@256bit.org>
Problem: wrong completion list displayed with non-existing dir + fuzzy
completion (kawarimidoll)
Solution: clear list of matches, if leader did not use fuzzy match
(glepnir)
fixes: vim/vim#15357closes: vim/vim#153656b6280c4a2
Co-authored-by: glepnir <glephunter@gmail.com>
Problem: fuzzy completion does not work with default completion
Solution: Make it work (glepnir)
closes: vim/vim#151938159fb18a9
Cherry-pick insexpand.c changes from patch 9.1.0608.
N/A patch:
vim-patch:9.1.0632: MS-Windows: Compiler Warnings
Co-authored-by: glepnir <glephunter@gmail.com>
Problem: With ':set splitkeep=screen', cursor did't restore column
correctly when splitting a window on a line longer than the
last line on the screen (after v9.1.0707)
Solution: Restore cursor column in `win_fix_scroll()` since it may be
changed in `getvcol()` after 396fd1ec2956 (phanium).
Example:
```
echo longlonglongling\nshort | vim - -u NONE --cmd 'set
splitkeep=screen' +'norm $' +new +q
```
fixes: vim/vim#16968closes: vim/vim#169717746348c5d
Co-authored-by: phanium <91544758+phanen@users.noreply.github.com>
Problem:
On Windows, since b360c06085, UI may hang on exit. #33019
Solution:
Restore the hack in on_proc_exit, until we can figure out why rpc_close
is not called in the UI client when the server closes the channel.
Problem:
chan_close_on_err() writes to the log file before peforming its actual
work. This could be slightly misleading in terms of log timestamps, or
could delay important work on Windows where file-open/write can be slow.
Solution:
Log after closing the channel, instead of before.
Problem:
`receive_msgpack()` calls `channel_close()` just before calling
`chan_close_on_err()`, even though `chan_close_on_err()` always calls
`channel_close()`.
Solution:
Remove the redunant `channel_close()` in `receive_msgpack()`.
Problem: Error messages that cause a vim.ui_attach() namespace to
detach are not visible in the message history. Decoration
provider and vim.ui_attach error messages are dissimilar.
Solution: Emit vim.ui_attach() errors as an actual message in addition
to logging it. Adjust error message format.
Problem: "shellcmdline" completion doesn't work with input().
Solution: Use set_context_for_wildcard_arg(). Fix indent in nextwild()
(zeertzjq).
There are some other inconsistencies for input() completion (ref vim/vim#948),
but since "shellcmdline" currently doesn't work at all, it makse sense
to at least make it work.
fixes: vim/vim#16932closes: vim/vim#169347a5115ce50
Problem: Message kind logic for emitting an error message is convoluted
and still results in emitting an unfinished message earlier than
wanted.
Solution: Ensure emsg_multiline() always sets the kind wanted by the caller
and doesn't isn't unset to logic for emitting the source message.
Caller is responsible for making sure multiple message chunks are
not emitted as multiple events by setting `msg_ext_skip_flush`...
Problem: Crash with nvim_get_runtime_file during wildcard expansion.
Solution: Ensure recursive gen_expand_wildcards() is allowed when
calling os_breakcheck()
Problem: Reconfiguring a float window applies the global 'winborder'.
Solution:
- Ignore 'winborder' when reconfiguring a float window.
- Still apply 'winborder' when converting a split to a float window.
Problem: Paired mark whose end is in front of its start should not have
its decor removed (as fixed by 72f630f9), but may still need to
have its range redrawn.
Solution: Still call `buf_decor_remove()` but ensure it is not called with
an inverse range when `extmark_del()` is called on an end mark.
Problem: 'winborder' accepting multiple string values.
Solution: Use the fallback did_set_str_generic() callback instead of
did_set_winborder() which calls opt_strings_flags() with
incorrect last argument.
Problem: Wrong cursor position and '^' mark when leaving Insert mode
just after 'autoindent' and cursor on last char of line.
Solution: Don't move cursor to NUL when it wasn't moved to the left
(zeertzjq).
fixes: vim/vim#15581
related: neovim/neovim#30165neovim/neovim#32943closes: vim/vim#16922a3a7d10bfb
Problem: using wrong length for last inserted string
(Christ van Willegen, after v9.1.1212)
Solution: use the correct length in get_last_insert_save(), make
get_last_insert() return a string_T (John Marriott)
closes: vim/vim#169218ac0f73eb1
N/A patches:
vim-patch:9.1.1129: missing out-of-memory test in buf_write()
vim-patch:9.1.1218: missing out-of-memory check in filepath.c
Co-authored-by: John Marriott <basilisk@internode.on.net>
Problem:
There is currently no global option to define the default border style for floating windows. This leads to repetitive code when developers need consistent styling across multiple floating windows.
Solution:
Introduce a global option winborder to specify the default border style for floating windows. When a floating window is created without explicitly specifying a border style, the value of the winborder option will be used. This simplifies configuration and ensures consistency in floating window appearance.
Co-authored-by: Gregory Anders <greg@gpanders.com>
Problem: Pasting the '.' register multiple times may work incorrectly
when the last insert starts with Ctrl-D and ends with '0'.
(after 9.1.1212)
Solution: Restore the missing assignment (zeertzjq).
closes: vim/vim#1690861b3544424
Problem: too many strlen() calls in edit.c
Solution: refactor edit.c and remove strlen() calls
(John Marriott)
This commit attempts to make edit.c more efficient by:
- in truncate_spaces() pass in the length of the string.
- return a string_T from get_last_insert(), so that the length of the
string is available to the caller.
- refactor stuff_insert():
- replace calls to stuffReadbuff() (which calls STRLEN() on it's
string argument) with stuffReadbuffLen() (which gets the length of
it's string argument passed in).
- replace call to vim_strrchr() which searches from the start of the
string with a loop which searches from end of the string to find the
last ESC character.
- change get_last_insert_save() to call get_last_insert() to get the
last_insert string (the logic is in one place).
closes: vim/vim#1686334954972c2
Co-authored-by: John Marriott <basilisk@internode.on.net>
Problem: Since 3cb1e825, all windows with 'statuscolumn' set, and a
resized 'signcolumn' for a particular buffer are marked
to be fully redrawn when the first window is encountered.
The "resized" variable is only unset after all windows have
been drawn, so this results in windows that have just been
draw to be marked for redraw again, even though the
signcolumn did not change size again.
Solution: Replace the `resized` variable with a `last_max` variable that
is changed when the first window into buf is encountered.
