Features
- help: Highlighted codeblocks for plugin documentation
Documentation
- lua: Add guide to using Lua in Neovim (#21137)
- Mention how to remove the "How-to disable mouse" menu item
- Fix order of numbers in syntax.txt
- lua: Correct the tags for vim.opt_local and vim.opt_global (#21138)
Bug Fixes
- api: Set correct curbuf when temporarily changing curwin
- api: "emsg_silent" should imply "silent" in nvim_cmd
- decoration: Do not reset must_redraw after calling providers
- diagnostic: Clear stale cache on reset (#21454)
- events: Save v:event for cmdline autocommands separately
- float: Fix ml_get error with bufpos
- float: Fix crash with bufpos and non-existent window
- folds: Use long for number of folded lines
- mappings: Use all buckets in second round of unmap
- memory: Fix memory alignment for dynamic allocation
- options: Fix local 'sidescrolloff' doesn't work for mouse
- options: Restore exists() behavior for options
- terminal: Fix 'mousescroll' not respected in terminal mode
- tui: Set cursor color parameter as string when required
- win_close: Remove float grid after closing buffer
Build System Fixes
- Restrict `git describe` to top level source directory
Problem: ModeChanged autocmd not executed when Visual mode is ended with
CTRL-C.
Solution: Do not trigger the autocmd when got_int is set. (closesvim/vim#11394)
61c4b04799
Cherry-pick removal of cmdwin feature check from patch 9.0.0663.
(cherry picked from commit 0b3328d15a)
It is not safe to remove the float grid when autocommands can still be
triggered, as autocommands may use the float grid.
(cherry picked from commit 9a5e80af56)
Add introductory guide explaining how to use Lua in Neovim:
where to put Lua files, how to set variables and options, how
to create mappings, autocommands, and user commands.
Adapted with kind permission from
https://github.com/nanotee/nvim-lua-guide
Duplicating get_option_value() logic for an obscure future refactor
isn't really worthwhile, and findoption() isn't used anywhere else
outside the options code.
(cherry picked from commit 09841ccbc3)
The BufWipeout autocmd is not 100% reliable and may leave stale entries
in the cache. This is sort of a hack/workaround to ensure
`vim.diagnostic.reset` calls don't fail if there are stale cache entries
but instead clears them
Fixes errors like
Error executing vim.schedule lua callback: /usr/share/nvim/runtime/lua/vim/diagnostic.lua:1458: Invalid buffer id: 22
stack traceback:
[C]: in function 'nvim_exec_autocmds'
/usr/share/nvim/runtime/lua/vim/diagnostic.lua:1458: in function 'reset'
(cherry picked from commit 1743359235)
Problem: screenpos() column result in fold may be too small.
Solution: Add space of 'number', sign column, etc. (closesvim/vim#11715)
ba2d191932
(cherry picked from commit 56b77dc171)
Problem: Test fails when terminal feature is missing.
Solution: Use CheckRunVimInTerminal.
b9603f6498
Co-authored-by: Bram Moolenaar <Bram@vim.org>
(cherry picked from commit 95c655fedc)
Problem: Using freed memory with the cmdline popup menu.
Solution: Clear the popup menu when clearing the matches. (closesvim/vim#11677)
038e6d20e6
Co-authored-by: Bram Moolenaar <Bram@vim.org>
(cherry picked from commit 4d22424d9e)
* fix(chansend): sending lines to terminal in reverse order on Windows #19315
Problem: `chansend()` on Windows sends lines in reverse order.
Cause: Using \n instead of \r\n for newlines on Windows.
Solution: on Windows, use CRLF newline characters.
Fixes#18501
(cherry picked from commit d5004aab27)
* fixup! fix(chansend): sending lines to terminal in reverse order on Windows #19315
(cherry picked from commit d313491a8b)
Co-authored-by: Enan Ajmain <3nan.ajmain@gmail.com>
Co-authored-by: Justin M. Keyes <justinkz@gmail.com>
Problem: ml_get error when using screenpos().
Solution: Give an error for the line number. (closesvim/vim#11661)
99d19438ca
Co-authored-by: Bram Moolenaar <Bram@vim.org>
(cherry picked from commit 0909d987fe)
Problem: screenpos() does not handle a position in a closed fold.
Solution: Check if the position is inside a closed fold. (closesvim/vim#9778)
4556a2e868
Co-authored-by: Bram Moolenaar <Bram@vim.org>
(cherry picked from commit 10af0549df)
Problem: screenpos() has non-zero row for invisible text.
Solution: Only add the window row when the text is visible. (closesvim/vim#9618)
7924a17791
Co-authored-by: Bram Moolenaar <Bram@vim.org>
(cherry picked from commit 6f9cda0f0a)
Problem: screenpos() is wrong when the last line is partially visible and
'display' is "lastline".
Solution: Also compute the position for a partially visible line.
(closesvim/vim#8599)
189663bdac
Co-authored-by: Bram Moolenaar <Bram@vim.org>
(cherry picked from commit 11d2704274)
vim-patch:86e6717ace4f
Problem: Crash when using win_move_statusline() in another tab page.
Solution: Check for valid window pointer. (issue vim/vim#11427)
86e6717ace
Co-authored-by: Bram Moolenaar <Bram@vim.org>
(cherry picked from commit d2a22242fb)
all pointers returned by arena_alloc residing in arena block should be
properly aligned
to meet neovim's alignment requirements but keeping it simple settle on
ARENA_ALIGN = MAX(sizeof(void *), sizeof(double)).
(cherry picked from commit 0240fd6d0f)
This is mainly a bugfix release, but there a few feature additions.
Features
- extmarks: Allow preventing spellchecking with spell = false
- spell: Support nospell in treesitter queries
- man: Add health check
- docs-html: Try to use tags for ToC headings
Bug Fixes
- ui: Setting 'cmdheight' with global statusline
- ui: Fix highlights not being updated with 'winhl'
- Make_filter_cmd for :! powershell
- :! pwsh redirection for `command not found`
- Find multibyte file name in line
- Avoid unsigned overflow in home_replace()
- docs-html: Update parser
- folds: Fix fold marker multibyte comparison
- health: Correct tmux rgb verification
- man.lua: Set modifiable before writing page
- mouse: Ensure no scrolling with "ver:0" in 'mousescroll'
- paste: Feed keys as typed in cmdline mode
- qflist: Avoid read of uninitialized memory
- spell: Fix wrong cast
- stdpath: Default to /tmp if stdpath('run') cannot be created
- tui: Resume main thread if suspending isn't implemented
Documentation
- lsp: Vim.lsp.range_code_action() is not deprecated
docs(api): pattern is not expanded for autocommands
Problem: Unlike `:autocmd`, `nvim_create_autocommand()` does not expand
environment variables in the `pattern`, which is unexpected.
Solution: Add a note to the documentation explaining this and suggesting
using `expand()` explicitly.
(cherry picked from commit eeaf943ca3)
Co-authored-by: Christian Clason <c.clason@uni-graz.at>
Problem: 'langmap' works differently when there are modifiers.
Solution: Only apply 'langmap' to a character where modifiers have no
effect. (closesvim/vim#11395, closesvim/vim#11404)
49660f5139
(cherry picked from commit e8877c4d7b)
build: don't append git tag on release versions
This prevents duplicate version strings such as v0.8.0-v0.8.0.
Also change the format for git releases from v0.9.0-dev-67-g625ba79be
to v0.9.0-dev-67+g625ba79be. The neovim versions are now:
release: v0.9.0
prerelease without git info: v0.9.0-dev
prerelease with git info: v0.9.0-dev-67+g625ba79be
(cherry picked from commit 90785aafe1)
Co-authored-by: dundargoc <gocdundar@gmail.com>
fix(highlight): link more treesitter groups by default
Problem: Captures used by bundled parsers are not highlighted by default
Solution: Add links to default groups
A link is added for a capture if
* there is a default group of the same name (e.g., `@tag` -> `Tag`)
* it's used in a bundled query and doesn't have a reasonable fallback
(e.g., `@text.literal`)
Also add all linked groups to the treesitter docs.
(cherry picked from commit 0ee8b88b10)
Co-authored-by: Christian Clason <c.clason@uni-graz.at>
Problem: Mouse column not correctly used for popup_setpos.
Solution: Adjust off-by-one error and handle Visual line selection properly.
(Yee Cheng Chin, closesvim/vim#11356)
17822c507c
The test_termcodes.vim test cannot be used. Use a Lua test instead.
(cherry picked from commit 72d2f5aef4)
Problem: User command does not get number from :tab modifier.
Solution: Include the number. (closesvim/vim#11393, closesvim/vim#6901)
208567e9d7
(cherry picked from commit a43a46f1e6)
If the call to `qf_setup_state()` in `qf_init_ext()` fails, control flow jumps
to label `qf_init_end` where a call to `qf_update_buffer()` is made with
`old_last` as a function call argument.
Prior to this patch, `old_last` would not yet have been initialized to its
default value of `NULL`, resulting in `qf_update_buffer()` attempting to compare
against its uninitialized value (quickfix.c:3882) then later forwarding it to
`qf_fill_buffer()` where the address is dereferenced and repeatedly read from/
written to for performing core qflist operations.
Depending on what the default value of `old_last` was, the results may range
from a best case scenario of neovim terminating with SIGSEGV from an attempt to
dereference an invalid pointer (quickfix.c:4056) to memory corruption if it
contained some other value that results in `qfp` being initialized from
`old_last->qf_next` (after which it is subsequently written to and read from in
a fairly arbitrary fashion).
Though extremely unlikely, it's possible for there to be security considerations
as a user can ensure that the next call to `qf_setup_state()` fails.
This patch ensures that `old_last` is NULL-assigned before control flow jumps to
`qf_init_end`.
Closes#20639.
(cherry picked from commit 8a762eba60)
Problem: SpellFileMissing autocmd may delete buffer.
Solution: Disallow deleting the current buffer to avoid using freed memory.
ef976323e7
(cherry picked from commit 6bc2d6b66b)
Problem: Crash when a callback deletes a window that is being used.
Solution: Do not unload a buffer that is being displayed while redrawing the
screen. Also avoid invoking callbacks while redrawing.
(closesvim/vim#2107)
94f01956a5
Omit parse_queued_messages(): N/A.
Cherry-pick a break statement from patch 8.1.0425.
(cherry picked from commit 65cbe0cc35)
vim-patch:9.0.0750: crash when popup closed in callback
Problem: Crash when popup closed in callback. (Maxim Kim)
Solution: In syntax_end_parsing() check that syn_block is valid.
0abd6cf62d
(cherry picked from commit 2a2ae32ceb)
Problem:
The generated ToC (table of contents) uses anchors derived from the
heading title, e.g. the "Global Plugins" heading yields:
https://neovim.io/doc/user/usr_05.html#_global-plugins-
so if the heading title changes, then the old URL (anchor) is broken.
Solution:
:help tags change less often than heading titles, so if a heading
contains a *tag*, use that as its anchor name instead. Example:
https://neovim.io/doc/user/usr_05.html#standard-plugin
(cherry picked from commit 6b01e9bf87)
- Improve generated HTML by updating parser which includes fixes for
single "'" and single "|":
https://github.com/neovim/tree-sitter-vimdoc/pull/31
- Updated parser also fixes the conceal issue for "help" highlight
queries https://github.com/neovim/tree-sitter-vimdoc/issues/23 by
NOT including whitespace in nodes.
- But this means we need to restore the getws() function which scrapes
leading whitespace from the original input (buffer).
(cherry picked from commit a7a83bc4c2)
fix: incorrect :help tag
vim.lsp.format() doesn't exist, which causes functionaltest to fail.
Change to vim.lsp.buf.format().
(cherry picked from commit 29a7f92e2e)
Co-authored-by: dundargoc <gocdundar@gmail.com>
docs: add vim APIs to deprecated.txt
Add vim.lsp.buf.formatting() to deprecated.txt.
Add vim.lsp.buf.range_formatting() to deprecated.txt.
(cherry picked from commit a99669df1c)
Co-authored-by: steventhorne <steven.thorne91@gmail.com>
fix(runtime): properly rely on t_Co for colorschemes
Problem: check for available colors failed
Solution: simply trust t_Co, which is always available
(cherry picked from commit 5dbc713de7)
Co-authored-by: Christian Clason <c.clason@uni-graz.at>
fix(docs-html): keycodes, taglinks, column_heading
Problem:
- Docs HTML: "foo ~" headings (column_heading) are not aligned with
their table columns/contents because the leading whitespace is not
emitted.
- taglinks starting with hyphen like |-x| were not recognized.
- keycodes like `<foo>` and `CTRL-x` were not recognized.
- ToC is not scrollable.
Solution:
- Add ws() to the column_heading case.
- Update help parser to latest version
- supports `keycode`
- fixes for taglink, argument
- Update .toc CSS. https://github.com/neovim/neovim.github.io/issues/297
fix https://github.com/neovim/neovim.github.io/issues/297
(cherry picked from commit 6d74676848)
Co-authored-by: Justin M. Keyes <justinkz@gmail.com>
feat(docs): format parameter list as a list
Problem:
The {foo} parameters listed in `:help api` and similar generated docs,
are intended to be a "list" but they aren't prefixed with a list symbol.
This prevents parsers from understanding the list, which forces
generators like `gen_help_html.lua` to use hard-wrapped/preformatted
layout instead of a soft-wrapped "flow" layout.
Solution:
Modify gen_vimdoc.py to prefix {foo} parameters with a "•" symbol.
(cherry picked from commit 24e5b5c8c7)
Co-authored-by: Justin M. Keyes <justinkz@gmail.com>
Problem:
The first parameter of "_open_osfhandle()" is an intptr_t; however, a HANDLE is passed.
The official documentation of [_open_osfhandle](https://learn.microsoft.com/en-us/cpp/c-runtime-library/reference/open-osfhandle) suggests to cast the HANDLE to intptr_t.
MSVC/cl is able to perform an implicit type cast.
However, LLVM/clang-cl will generate an compilation error.
Solution:
Explicitly convert HANDLE to intptr_t for _open_osfhandle().
(cherry picked from commit 63ab3337e0)
Problem: <amatch> of MenuPopup event is expanded like a file name.
Solution: Do not expand <amatch> for MenuPopup. (closesvim/vim#11328)
c601d988b6
(cherry picked from commit 4b2c5be6e6)
Switch back to Ubuntu 18.04 for buliding the appimage. This allows for
using the appimage on older systems that do not provide GLIBC_2.29.
Fixes#19711.
Fixes#20113.
Problem: No space for command line when there is a tabline.
Solution: Correct computation of where the command line should be.
(closesvim/vim#11295)
c9f5f73206
(cherry picked from commit 31996f7748)
Problem: Setting 'cmdheight' has no effect if last window was resized.
Solution: Do apply 'cmdheight' when told to. Use the frame height instead
of the cmdline_row. (closesvim/vim#11286)
0816f473ab
Problem:
If the shell command passed to the filtered bang command isn't found,
the error isn't redirected to the temp.out file when shell is set to
powershell.
Solution: Use anonymous function with Invoke-Command operator (&).
(cherry picked from commit f2482b3b53)
Problem:
`Start-Process` requires the command to be split into the shell
command and its arguments. Previously it was done by parsing, which
didn't handle cases such as
- commands with escaped space in their filepath
- quoted commands with space in their filepath
Solution:
Use
- `pwsh -Command` instead of `Start-Process`
- `Get-Content` instead of `-RedirectStandardInput`
- `Out-File` instead of `-RedirectStandardOutput`
(cherry picked from commit b4d42bb905)
BREAKING CHANGES
- Remove 'insertmode' 'remap' and 'terse' options
- highlight: Rename attributes to match Vim (#19159)
- highlight: Error on invalid names and allow '.' and '@'
- terminal: Drop winpty, require Windows 10 #18253
- treesitter: Use @foo.bar style highlight groups
- treesitter: Do not merge queries by default (#20105)
FEATURES
- runtime: Enable filetype.lua by default (#19216)
- Add `undo!`
- Add "prerelease" to version dict
- Click support for 'statusline', 'winbar' #18650
- Add preview functionality to user commands
- allow cmdheight=0 (EXPERIMENTAL) #16251
- Stdpath('run'), /tmp/nvim.user/ #18993
- Add 'mousescroll' option (#12355)
- Allow :wincmd to accept a count (#19815)
- Multibuffer preview support for inccommand
- Man: Port to Lua (#19912)
- api: Ui options relevant for remote TUI
- api: Add `nvim_parse_cmd` And `nvim_cmd`
- api: Support handling stdin stream in remote ui
- api: Add `group_name` to `nvim_get_autocmds`
- api: Enable nvim_exec_autocmds to pass arbitrary data (#18613)
- api: Support pattern array for exec_autocmds
- api: Add `unsilent` to command APIs
- api: Add replace_keycodes to nvim_set_keymap (#19598)
- api: Add support for :horizontal modifier
- api: Add "move" to nvim_input_mouse
- api/ui: "ui_watched" option for ui-side extmarks
- build: Add_glob_target runs only on changed files #19070
- checkhealth: Check for slow shell #17829
- defaults: Session data in $XDG_STATE_HOME #15583
- defaults: Search selection by * and # in visual mode (#18538)
- defaults: Nnoremap & :&&<CR> #19365
- defaults: enable mouse by default (set mouse=nvi) #19290
- diagnostic: Pass diagnostics as data to DiagnosticChanged autocmd (#20173)
- docs: Gen_help_html.lua
- edit: Insert an unsimplified key using CTRL-SHIFT-V
- treesitter: Full support for spelling
- filetype: Convert more patterns to filetype.lua
- filetype: Remove side effects from vim.filetype.match (#18894)
- filetype: Expand environment variables in filetype patterns (#20145)
- fs: Add vim.fs functions: parents() dirname() basename() dir() find() normalize()
- highlight: Implement CurSearch highlight
- highlight: Support scoped @spam.eggs.baked_beans groups
- input: Delay all simplifications
- l10n: Turkish translations #19441
- l10n: Improve zh_CN translations (#19483)
- lsp: Remove capabilities sanitization (#17814)
- lsp: Show feedback on empty hover response (#18308)
- lsp: Options to filter and auto-apply code actions (#18221)
- lsp: Add vim.lsp.buf.format (#18193)
- lsp: Add logging level "OFF" (#18379)
- lsp: Add LspAttach and LspDetach autocommands
- lsp: Add filter to vim.lsp.get_active_clients()
- lsp: Option to reuse_win for jump actions (#18577)
- lsp: Add a start function (#18631)
- lsp: Send didChangeConfiguration after init (#18847)
- lsp: Defaults: tagfunc, omnifunc, formatexpr (#19003, #19677)
- lsp: Allow passing custom list handler to LSP functions that return lists (#19213)
- lsp: Provide feedback if server can't compute rename result (#19546)
- lsp: Add range option to code_action; deprecate range_code_action (#19551)
- lsp: Disable exit_timeout by default (#19672)
- lsp: Add tcp support
- lua: Vim.deprecate() #18320
- lua: Vim.cmd() with kwargs acts like nvim_cmd() #18523
- lua: Allow some vim script functions to run in fast callbacks
- lua: Measure require in --startuptime
- lua: Allow vim.cmd to be indexed (#19238)
- lua: Print source locations of lua callbacks (#19597)
- lua: Add vim.iconv (#18286)
- lua: Vim.ui_attach to get ui events from lua (EXPERIMENTAL)
- man.vim: List command flags in "gO" outline #17558
- mappings: Do not replace existing mapping for simplified form
- mappings: Allow special keys and modifiers in <Cmd> mapping
- mapset: Support restoring "replace_keycodes" and "desc"
- mapset: Support restoring Lua callback (#20024)
- marks: Restore viewport on jump #15831
- nvim_create_user_command: Pass structured modifiers to commands
- pum: Pretend 'mousemoveevent' is set when showing right-click menu
- server: Set $NVIM, unset $NVIM_LISTEN_ADDRESS #11009
- server: Instance "name", store pipes in stdpath(state)
- term: Add support for libvterm >= 0.2
- terminal: Implement <c-\><c-o> for terminal mode
- terminal: Recognize underdouble and undercurl
- terminfo: Bump built-in terminfo entries (#18570)
- treesitter: Allow customizing language symbol name
- treesitter: Add ability to retreive a tree/node given a range
- treesitter: Upstream node methods from nvim-treesitter
- treesitter: Include language in invalid query error (#14053)
- treesitter: Bundle Lua parser and queries
- treesitter: Add viml parser and queries
- treesitter: Add injections
- treesitter: Add vim.treesitter.start(), enable for Lua
- treesitter: Bundle :help parser and queries
- tui: Query terminal for CSI u support (#18181)
- tui: Recognize keypad keys when using kitty keyboard protocol
- tui: Try terminfo for [re]set_cursor_color OSC #19255
- tui: Allow grid and host to disagree on ambiguous-width chars (#19686)
- tui: Recognize sidescroll events (#19992)
- tui: Support 'mousemoveevent'
- ui: Add `'winbar'`
- ui: Clear message history explicitly with msg_history_clear event
- ui: Make right-click menu work properly with ext_multigrid
- ui: Allow to set the highlight namespace per window
- ui: Use msg_grid based implementation for cmdheight=0
- ui-ext: Make 'mousemoveevent' a ui_option
- eval: Make Lua Funcref work as method and in substitute() (#20217)
- eval: Input() support any type for "cancelreturn" in a dict (#19357)
BUG FIXES
- Show autocmd output when F is in shortmess (#18251)
- Has() should preserve v:shell_error #18280
- Suppress "is a directory" messages with shortmess 'F' (#18296)
- Display global statusline correctly with ext_messages
- Correct nlua_wait error message #18867
- Right-click in clickable statusline #19252
- Remote UI may get invalid 'pumblend' value #19379
- Assertion failure when requiring missing module in autocmd
- api: Nvim_eval_statusline should validate input #18347
- api: Check error after getting win/buf handle (#19052)
- api: Check for inclusive buffer line index out of bounds correctly (#19056)
- api: Change default value of 'pattern' in nvim_exec_autocmds (#19115)
- api: Do not switch win/buf if getting option in current win/buf (#19383)
- api: Make nvim_set_hl(ns=0, ...) redraw screen properly
- api: Nvim_set_hl bail out on invalid group name (#20021)
- api: Notify dict watchers on nvim_set_var and vim.g setter
- api/command: Fargs behavior when no arguments are passed (#19862)
- autocmds: Separate command from desc (#18617)
- buffer: Disable buffer-updates before removing from window #18933
- build: Missing definitions for sizeof macros #16393
- build: Only pass -municode if MINGW #19049
- build: Strip trailing newline from variable (#19084)
- build: Don't disable byte precompilation on debug builds
- build: Fails if git is missing #19366
- charclass: Make behavior with empty str match latest Vim (#19749)
- checkhealth: Skip vim.health #18816
- ci: Remove 2000ms blocking wait in many plugin/lsp_spec.lua tests
- cmd: Make :-tabmove work with modifiers (#18447)
- cmdheight=0: Various issues part3 #19816
- cmdline: Fix passing -1 as char
- cmdline: Trigger CmdlineEnter and ModeChanged earlier (#19474)
- cmdline: Do not trigger completion at wrong time (#19920)
- cmdline: Don't send invalid cursor with incsearch and cmdheight=0
- column: Move sign sentinel after inserting/deleting lines (#20400)
- completion: Remove wrong FUNC_ATTR_NONNULL_ALL (#19627)
- diagnostic: Use nvim_exec_autocmds to trigger DiagnosticChanged (#18188)
- diagnostic: Check for negative column value (#18868)
- diagnostic: Remove buf from cache on `BufWipeout` (#20099)
- diagnostic: Populate data key in DiagnosticChanged autocmd in reset (#20207)
- docs: Correct obsolete note about 'writedelay' in dev docs
- docs: Remove internal function from docs
- edit.c: Indentkeys double indent after "!" #12894
- eval: Check for v:lua when calling callback (#19855)
- eval/f_getmatches: Return empty list for invalid win argument (#18893)
- events: Make CursorHold behave as documented
- ex_cmds: Correct flags for :const (#19387)
- exceptions: Restore `did_throw` (#20000)
- exmode: Do not throttle messages when pressing enter to print line
- extmarks: Make virt_lines always start at 0 virtcol
- filetype: Fix and improve filetype patterns
- fillchars: Change fallback after setcellwidths()
- float: Make `screen*()` functions respect floating windows
- float: Fix glitch when making float window with border a split
- float: Fix mouse drag position if float window turned to a split
- folds: Fix fold regression with :move (#18685)
- folds: Fix fold remains when :delete makes buffer empty (#19673)
- ftdetect: Source plugins in autogroup (#18237)
- gen_vimdoc.py: Handle missing luajit
- getchar: Flush screen before doing a blocking wait
- handlers: More specific error messages (#16772)
- health: Correct shada file path #18603
- health: Handle non-existent log file #18610
- highlight: Use ctermbg/fg instead of bg/fg when use_rgb=false #18982
- highlight: Add missing 'nocombine' to nvim_get_hl apis (#19586)
- highlight: Set the window namespace when redrawing statusline
- hl: Set Normal hl group sg_attr value #18820
- hl: Return cterm fg/bg even if they match Normal #18981
- inccommand: Do not try to preview an ambiguous command (#18827)
- inccommand: Avoid crash if callback changes inccommand option (#18830)
- inccommand: Clear cmdpreview state if preview is not shown (#18923)
- inccommand: Skip split window if not enough room #18937
- inccommand: Never preview if parsing command failed (#18944)
- inccommand: Parse the command to check if it is previewable
- inccommand: Deal with unsynced undo (#20041)
- input: Allow Ctrl-C to interrupt a recursive mapping even if mapped (#18885)
- input: Fix macro recording with ALT and special key (#18917)
- input: Use correct grid when restoring cursor for <expr> mapping (#19047)
- input: Do no reinterpret mouse keys with ALT modifiers
- input: Use click number of last click for mouse drag (#20300)
- inspect: Escape identifiers that are lua keywords (#19898)
- keywordprg: Default to :help if set to empty string (#19983)
- l10n: Improve zh_CN and zh_TW translations (#19969)
- log: Back even again
- logging: Skip recursion, fix crash #18764
- logging: Make logmsg() thread-safe again
- logging: Try harder to resolve Nvim "name" #19016
- lsp: Unify progress message handling (#18040)
- lsp: Fix unnecessary buffers being added on empty diagnostics (#18275)
- lsp: Fix infinite loop in resolved_capabilities deprecation message (#18333)
- lsp: Add missing bufnr argument (#18382)
- lsp: Fix rename capability checks and multi client support (#18441)
- lsp: Detach spawned LSP server processes (#18477)
- lsp: Perform client side filtering of code actions (#18392)
- lsp: Only send diagnostics from current buffer in code_action() (#18639)
- lsp: Respect global syntax setting in open float preview (#15225)
- lsp: Include cancellable in progress message table (#18809)
- lsp: Adjust offset encoding in lsp.buf.rename() (#18829)
- lsp: Set buflisted before switching to buffer (#18854)
- lsp: Fix multi client handling in code action (#18869)
- lsp: Small bugs in snippet-parser #18998
- lsp: Pcall nvim_del_augroup_by_name (#19302)
- lsp: Abort pending changes after flush when debouncing (#19314)
- lsp: Don't attach a client in lsp.start() if there is none (#19328)
- lsp: Account for initializing servers in vim.lsp.start (#19329)
- lsp: Move augroup define to if statement (#19406)
- lsp: Set workspace.configuration capability (#19548)
- lsp: Send didOpen if name changes on write (#19583)
- lsp: Prevent unexpected position jumps (#19370)
- lsp: Avoid ^M character in hover window on Windows (#19640)
- lsp: Set end_col in formatexpr (#19676)
- lsp: Handle multiple clients with incremental sync (#19658)
- lsp: Fix some type annotations in lsp.rpc (#19714)
- lsp: Avoid pipe leaks if lsp cmd isn't executable (#19717)
- lsp: Handle nil client in onexit callback (#19722)
- lsp: Fix nil value error in get_group (#19735)
- lsp: Clean the diagnostic cache when buffer delete (#19449)
- lsp: When buffer detach remove buffer from client attached buffers (#20081)
- lsp: Schedule removal of client object (#20148)
- lsp: Support `false` result in handlers (#20252)
- lsp: Out of bounds error in lsp.util.apply_text_edits (#20137)
- lsp: Use correct function name in deprecated message (#20308)
- lsp: Create missing directory before creating file (#19835)
- lua: Don't mutate opts parameter of vim.keymap.del (#18227)
- lua: Stop pending highlight.on_yank timer, if any (#18824)
- lua: Highlight.on_yank can close timer in twice #18976
- lua: Clear got_int when calling vim.on_key() callback (#18979)
- lua: Don't leak memory on error
- lua: Double entries in :lua completion #19410
- lua: Make it possible to cancel vim.wait() with Ctrl-C (#19217)
- lua: Make ui_attach()/ui_detach() take effect immediately (#20037)
- lua: Make vim.str_utfindex and vim.str_byteindex handle NUL bytes
- lua: Free vim.ui_attach callback before lua close (#20205)
- lua: Fix architecture-dependent behavior in usercmd "reg" (#20384)
- mac: Use same $LANG fallback mechanism as Vim
- mac: Add CoreServices to flake.nix #18358
- man.vim: Q in "$MANPAGER mode" does not quit #18443
- maparg: Remove double allocation (#20033)
- mappings: Fix double-free when unmapping simplifiable Lua mapping
- mapset: Remove existing abbreviation of same lhs (#20320)
- mark: Set mark fnum from buffer (#19195)
- mark: Mark without a view restores at topline #19224
- mark: Fix unexpected cursor movements (#19253)
- mark: Give correct error message when mark is in another buffer (#19454)
- menu: Make :menu still print header when there are no menus
- messages: Add color when showing nvim_echo in :messages history
- messages: Do not crash on cmdheight=0 and g< redisplay
- messages: Validate msg_grid before silent! message with cmdheight=0
- mksession: Don't store floats in session #18635
- mouse: Click on global statusline with splits (#19390)
- mouse: Fix using uninitialized memory with K_MOUSEMOVE (#19480)
- mpack: Make sure a `bool` always is a `bool`
- normal: Fix segfault with bracket command jumping to a mark
- options: Properly free string options (#19510)
- options: Mark `winhighlight` as list style (#19477)
- packaging: Remove excess forward slash in Wix Patch (#18121)
- paste: Ignore mappings in Cmdline mode (#18114)
- path: Path_is_url returns false for "foo:/" #19797
- powershell: Filter ":!" commands with args #19268
- pum: Make right drag in anchor grid to select work in multigrid UI (#19382)
- query: Fix unnatural order for inherits in treesitter queries (#20298)
- redraw: Make sure :redraw! redraws command line
- redraw: Handle switching to a tabpage with larger p_ch value
- redraw: Avoid unnecessary redraws and glitches with floats+messages
- redraw: Make redrawdebug=nodelta handle all the cases
- rpc: Break nvim_error_event feedback loop between two nvim instances
- runtime/genvimvim: Omit s[ubstitute] from vimCommand #18480
- screen: Restart win_update() if a decor provider changes signcols (#18768)
- screen: Check for col instead of vcol when drawing fold (#19572)
- session: Respect sessionoptions=terminal #19497
- shared: Avoid indexing unindexable values in vim.tbl_get() (#18337)
- signs: Priority of extmark signs (#19718)
- source: Make changing 'shellslash' change expand() result
- source: Fix expand('<sfile>') no longer works for Lua
- spell: Make setting 'encoding' clear word list
- spell: Correct spell move behavior without "noplainbuffer" (#20386)
- startup: Nvim with --clean should not load user rplugins
- substitute: Subtract number of backslashes later
- tabpage: Check if ROWS_AVAIL changed for resize (#19620)
- terminal: Invalid pointer comparison #18453
- terminal: Do not trim whitespace that is actually in the terminal (#16423)
- terminal: Scrollback delete lines immediately #18832
- terminal: Coverity USE_AFTER_FREE #18978
- terminal: Crash if TermClose deletes own buffer #19222
- terminal: Avoid reading over the end of cell.chars (#19580)
- terminal: Skip aucmd_win when checking terminal size (#19668)
- terminal: Adopt altscreen test for libvterm 0.2 changes
- terminfo: Disable smglr for vtpcon and conemu (#18855)
- termopen: Avoid ambiguity in URI when CWD is root dir (#16988)
- tests: Fix some screen.lua warnings
- tests: Fix some issues with ui/inccommand_spec.lua causing slowness
- tests: Unreliable parser_spec #18911
- tests: Check for EOF on exit of nvim properly
- tests: Missing clear() #18927
- tests: Remove misleading $TEST_PATH segment #19050
- tests: Remove irrelevant usage of display-=msgsep
- tests: Use pending_c_parser when needed
- tests: Indicate in test logs when nvim exit times out
- tmpdir: Invalid tempname() if username has slashes #19323
- treesitter: Create new parser if language is not the same as cached parser (#18149)
- treesitter: Bump match limit up
- treesitter: Offset directive associates range with capture (#18276)
- treesitter: Correct region for string parser (#18794)
- treesitter: New iter if folded
- treesitter: Don't error when node argument of predicate is nil (#19355)
- treesitter: Free memory on removing parser (#19933)
- treesitter: More efficient node:root()
- treesitter: Make it get_captures_at_position
- treesitter: Do not link @error by default
- treesitter: Don't support legacy syntax in start()
- treesitter: Use the right loading order for base queries (#20117)
- treesitter: Prevent endless loop on self-inheritence
- treesitter: Return full metadata for get_captures_at_position (#20203)
- ts: Do not clobber spelloptions (#20095)
- tui: Update modifyOtherKeys reporting (#18158)
- tui: Disable extended keys before exiting alternate screen (#18318)
- tui: Piping nodejs to nvim breaks input handling #18932
- tui: Resize at startup #17795
- tui: Add fixups for hterm family #19078
- tui: Handle padding requirements for visual bell (#20238)
- ui: Require window-local value to show winbar on floating windows (#18773)
- ui: Do not call showmode() when setting window height (#18969)
- ui: Set redraw_cmdline when setting window height (#19630)
- ui: Don't allow decor provider with ns_id==0
- ui: Ui compositor does not correctly free event callbacks
- ui: Allow redrawing statusline when msgsep is used (#20337)
- ui: Redraw end of buffer if last line is modified (#20354)
- unittests: Coredump when running unit tests #18663
- usercmd: Also check for whitespace after escaped character (#19942)
- version.c: Mark N/A vim patches #18587
- vim.ui.input: Accept nil or empty "opts" #19109
- window: Close floats first when closing buffer in other tab (#20284)
- window: Fix equalization with cmdheight=0 (#20369)
- windows: Stdpath("state") => "nvim-data" #18546
- windows: Exepath, stdpath return wrong slashes #19111
- winhl: Do not crash when unsetting winhl in just opened window
- Make :undo! notify buffer update callbacks (#20344)
- eval: Make Vim functions return inner window width and height (#19743)
BUILD SYSTEM
- Bump Doxyfile to minimum required version 1.9.0 #18118
- Bump msgpack to 4.0.0
- Enable EXITFREE on Debug builds (#17783)
- Add formatting targets for c and lua files (#19488)
- clang-format: Align with project style #18192
- clint: Remove all python2-isms with pyupgrade
- clint: Remove "function size is too large" warning
- clint: Remove rules for includes, whitespace, tabs #18611
- cmake: Simplify and speed up the uninstall target
- cmake: Simplify def_cmd_target function
- cmake: Use glob_wrapper instead of file(GLOB in main CMakeLists
- cmake: Fix static `libintl` test on macOS
- deps: Bump LuaJIT, Luv and libuv
- deps: Support universal builds on macOS
- deps: Bump tree-sitter to v0.20.7 (#20067)
- deps: Bump tree-sitter parsers
- deps: Bump required libvterm to v0.3 (#20222)
- deps: Require libtermkey version 0.22
- deps: update neovim-qt, win32tools.zip
PERFORMANCE
- Only redraw for CurSearch when it is currently in use
- highlight: Allocate permanent names in an arena for fun and cache locality
- highlight: Use binary search to lookup RGB color names
- map: Visit only one hash bucket instead of all, like an actual hash table
- memory: Get rid of extraneous heap allocations
- memory: implement arena memory allocation with a shared freelist
- memory: Use an arena for RPC decoding and some API return values
- messages: Don't call ui_flush() per message line in various places
- treesitter: Use a reuse list for query cursors
- ui: Reduce allocations when encoding and decoding "redraw" events
- ui: Avoid ui_flush() work in headless mode
REFACTOR
- checkhealth: Rename to vim.health, move logic to Lua #18720
- Add pure attribute to pure functions
- Replace char_u variables and functions with char
- Enable -Wconversion warning for all Nvim source files
- Add warnings for deprecated functions (#18662)
- Change type of linenr_T from long to int32_t
- Use nvim_get/set_option_value for vim.{b,w}o
- Remove functions marked for deprecation in 0.8 (#19299)
- Rename function prefix mb to the more accurate utf_cp (#19590)
- Remove some unused includes
- Change remaining sourcing_name/sourcing_lnum to exestack
- Change FALSE/TRUE to false/true
- api: Use a hashy hash for looking up api method and event names
- api: Use a unpacker based on libmpack instead of msgpack-c
- api: restructure api/vim.c and api/private/helpers.c code in separate files
- api: Remove redundant fields of CmdParseInfo
- aucmd: Call define_autocmd() directly for default autocmds
- ci: Cleanup release.yml #19132
- cmd: Format do_one_cmd()
- cmd: Hoist out some code into functions
- cmd: Unify execute_cmd with do_one_cmd
- decor: Use decor levels properly
- drawline.c: Factor out utf8 multibyte check
- eval: Use Hashy McHashFace instead of gperf
- eval.c: Resolve all clint issues (#19774)
- eval/funcs.c: Resolve all clint errors
- events: Remove unnecessary fudging of updating_screen
- ex_cd: Add an early return to fix clint warning
- ex_docmd.c: Resolve most clint errors (#19775)
- filetype: Allow vim.filetype.match to accept buf and filename (#19114)
- highlight: Make hlattrs2dict always use pre-allocated dict
- log: Simplify log_path_init #18898
- log: Use msg_schedule_semsg #18950
- lsp: Remove redundant client cleanup (#18744)
- lsp: Make the use of local aliases more consistent
- lsp: Use autocmd api (#19407)
- lsp: Factor out read_loop function
- lsp: Encapsulate rpc uv handle
- lsp: Extract rpc client from rpc.start
- lua: Replace hard-coded gsub with vim.pesc() (#18407)
- lua: Reformat with stylua 0.14.0 (#19264)
- lua: Git-blame-ignore stylua update PR (#19273)
- lua: Replace vim.cmd use with API calls (#19283)
- map: Simplify free_all_mem handling
- map: Statically initialize maphash array
- map: Simplify add_map params
- normal: Convert function comments to doxygen format
- object: Get rid of redundant FIXED_TEMP_ARRAY
- ops: Doxygen docstrings #17743
- option: DRY get/set option value #19038
- plines: Use a struct for chartabsize state
- provider: Use list comprehension #19027
- regexp_nfa.c: Match where Vim calls fopen() (#18778)
- runtime: Convert dist#ft functions to lua (#18247)
- runtime: Convert more dist#ft functions to lua (#18430)
- runtime: Convert the remaining dist#ft functions to lua (#18623)
- runtime: Port remaining patterns from filetype.vim to filetype.lua (#18814)
- runtime: Refactor filetype.lua (#18813)
- runtime: Port scripts.vim to lua (#18710)
- setcellwidths: Use TV_LIST_ITEM_NEXT properly
- signs: Handle non-sign attrs separately (#19784)
- tests: Introduce testprg()
- treesitter: Get_{nodes,captures}_at_{position,cursor}
- typval: Change FC_CFUNC abstraction into FC_LUAREF
- ui: Simplify stdin handling
- uncrustify: Format all c code under /src/nvim/
- vim.opt: Remove del arg
- vim.opt: Unify vim.bo/wo building
- vim.opt: Optimize append/prepend/remove
- Format runtime with stylua
- fix regression by #20411
- `diff.exe` is required for non-default 'diffopt' (diffopt=filler, diffopt=context, …)
- the names of some required nvim-qt DLLs changed
I don't think using an integer as a NUL-terminated string can work on
big-endian systems, at least.
This is also not tested. Add a test.
Also fix a mistake in the docs of nvim_parse_cmd.
Replaces unnecessary helper functions in `optionstr.c` such as
`get_option_flags()`, `get_option_fullname()`, `set_option_flag()`,
`is_global_option()`, etc. with a single `get_option()` helper function
that allows direct access to the `options` array.
Also refactors `f_exists()` to use `get_varp_scope` instead of using
`get_option_tv`. This opens up the path for removing `getoptions_T`
altogether later down the line since the hidden option logic is no
longer needed.
vim-patch:9.0.0602: new TypeScript extensions are not recognized
Problem: New TypeScript extensions are not recognized.
Solution: Recognize .mts and .cts files. (closesvim/vim#11237)
7fc6c0e4da
vim-patch:9.0.0600: GYP files are not recognized
Problem: GYP files are not recognized.
Solution: Recognize GYP files. (closesvim/vim#11242)
d324742292
vim-patch:9.0.0599: latexmkrc files are not recognized
Problem: Latexmkrc files are not recognized.
Solution: Use Perl filetype for latexmkrc files. (closesvim/vim#11241)
cde0319385
Problem: Extra newline in messages after a verbose shell message.
Solution: Output the newline with msg_putchar_attr(). (closesvim/vim#11233)
Make it possible to filter a screendump before comparing it.
1190139ed0
Cherry-pick Test_message_more_scrolledback() from patch 9.0.0592 because
Nvim already behaves as intended.
Problem: Missing change in test.
Solution: Add the test change.
124af71a28
vim-patch:9.0.0585: when long message test fails the error message is not visible
Problem: When long message test fails the error message is not visible.
Solution: Dump more lines.
6a879878f4
When :undo! was introduced to Nvim the implementation of 'inccommand'
preview callback hasn't been fully decided yet, so not notifying buffer
update callbacks made sense for 'inccommand' preview callback in case it
needs to undo the changes itself.
Now it turns out that the undo-and-forget is done automatically for
'inccommand', so it doesn't make sense for :undo! to avoid notifying
buffer update callbacks anymore.
Problem:
Cannot opt-out of "WARNING: The file has been changed since reading
it!!!", even with ":write!".
Solution:
Change ":write!" to skip the warning.
closes#7270
Problem: Running filetype test leaves file behind.
Solution: Delete the file.
0e71b7d4ce
vim-patch:8.2.4466: MS-Windows: illegal memory access in installer
Problem: MS-Windows: illegal memory access in installer when using
"create-directories" as the final argument.
Solution: Check the argument count. (Cam Sinclair, closesvim/vim#9844)
5c6edf41f9
Problem: 'completeopt' "longest" is not used for complete().
Solution: Also use "longest" for complete(). (Bjorn Linse, closesvim/vim#11206)
87af60c915
Problem: Autocmd code is indented more than needed.
Solution: Break out sooner. (Yegappan Lakshmanan, closesvim/vim#11208)
Also in user function code.
e9dcf13a30
Problem: HSL playlist files are not recognized.
Solution: Add a pattern to recognize HSL palylist files. (Benoît Ryder,
closesvim/vim#11204)
35fdd9a67d
- YouCompleteMe is unnecessary since Nvim LSP works well.
- vim-addon-local-vimrc is not needed since we added `.editorconfig`.
- Inline flake8 arguments. Eventually we will remove all python code,
don't need a top-level `.flake8` file meanwhile.
Problem:
- The layout of the doxygen HTML "dev docs" has been broken for 1+ years.
- There is no evidence that anyone uses the doxygen HTML.
- In the meantime since we introduced the doxygen HTML pages, the Nvim
C docstrings have been surfaced in other ways:
- Nvim LSP client
- Generated :help docs like ":help api" and ":help ui".
Solution:
- Remove `contrib/doxygen/` from neovim repo.
- Note: keep `src/Doxyfile`, it is used by `scripts/gen_vimdoc.py`. ☠️
- Remove `doc/dev/` from website. https://github.com/neovim/neovim.github.io/pull/291
- Remove doxygen job from the neovim/doc repo. https://github.com/neovim/doc/pull/28
- Future/ongoing: ":help dev" and other :help docs will continue to be
improved as the "source of truth" for developer-related docs.
ref https://github.com/neovim/neovim/pull/824
This partially reverts commit 42aeb5c5b1.
Setting cmake policies is normally not required as
cmake_minimum_required automatically sets these. One exception is cmake
script mode (-P) since it automatically resets all policy changes.
Closes: https://github.com/neovim/neovim/issues/20286
Problem: Crash when closing a tabpage and buffer is NULL.
Solution: Adjust how autocommands are triggered when closing a window.
(closesvim/vim#11198, closesvim/vim#11197)
62de54b48d
Problem: Long message test can be flaky.
Solution: Wait for more prompt instead of ruler.
21d393a12b
Comment N/A lines out instead of deleting in buftype tests.
When it happens it wastes 2 seconds which is NOT included in the
normal busted timing info. It is hard to correct this, but we
can at least print a warning when this happens.
Before only win_line lines were considered. this applies nodelta
to all screen elements. Causes some failures, which might indeed
indicate excessive redraws.
Problem: Ctrl-hat test fails with Athena and Motif. (Elimar Riesebieter)
Solution: Run the test only with GTK. (Dominique Pellé, closesvim/vim#9069)
8753c1dd2c
Problem: Using freed memory when replacing. (Dhiraj Mishra)
Solution: Get the line pointer after calling ins_copychar().
35a9a00afc
This patch is N/A as it only applies to non-UTF-8 encoding.
Problem: Various code is not fully tested.
Solution: Add a few more tests. (Yegappan Lakshmanan, closesvim/vim#8378)
2d6d718dde
Nvim does not support encoding=latin1 or compatible mode.
The two paste tests are applicable.
Problem: Various code not fully tested.
Solution: Add a few more tests. (Yegappan Lakshmanan, closesvim/vim#6049)
0ff5dedf0f
Nvim does not support encoding=latin1 or setting keycodes/termcaps.
Problem: The do_set() function is much too long.
Solution: Move setting of a string option to a separate function.
4740394f23
Cherry-pick some tests from Vim patch 8.2.0540.
Problem: When at the command line :redrawstatus does not work well.
Solution: Only update the statuslines instead of the screen. (closesvim/vim#11180)
320d910064
Problem: Cannot redraw the status lines when editing a command.
Solution: Only postpone the redraw when messages have scrolled.
(closesvim/vim#11170)
c14bfc31d9
Problem: Command line cleared when using :redrawstatus in CmdlineChanged
autocommand event.
Solution: Postpone the redraw. (closesvim/vim#11162)
bcd6924245
Cherry-pick Test_redraw_in_autocmd() from Vim patch 8.2.4789.
This allows a more declarative way of defining targets which may be easier to those not familiar with cmake quirks.
Example usage:
# get a list of pre-configured presets
cmake --list-presets
# configure a preset
cmake . --preset=default
# build it
cmake --build --preset=default
Problem: Macros for MS-Windows are inconsistent, using "32", "3264 and
others.
Solution: Use MSWIN for all MS-Windows builds. Use FEAT_GUI_MSWIN for the
GUI build. (Hirohito Higashi, closesvim/vim#3932)
4f97475d32
Problem: Illegal memory access when replacing in virtualedit mode.
Solution: Check for replacing NUL after Tab.
c249913edc
Cherry-pick Test_virtualedit_mouse() from patch 9.0.0177.
vim-patch:9.0.0479: in :def function all closures in loop get the same variables
Problem: In a :def function all closures in a loop get the same variables.
Solution: Use a separate list of variables for LOADOUTER and SAVEOUTER.
1aea184a0d
(note: patch description is wrong)
BREAKING CHANGE: When using a Funcref converted from a Lua function as
a method in Vim script, the result of the base expression is now passed
as the first argument instead of being ignored.
vim-patch:8.2.5117: crash when calling a Lua callback from a :def function
Problem: Crash when calling a Lua callback from a :def function. (Bohdan
Makohin)
Solution: Handle FC_CFUNC in call_user_func_check(). (closesvim/vim#10587)
7d149f899d
Problem: Varargs does not work for replacement function of substitute().
Solution: Check the varargs flag of the function. (closesvim/vim#11142)
48db5dafec
Problem: The funcexe_T struct members are not named consistently.
Solution: Prefix "fe_" to all the members.
851f86b951
Omit fe_check_type: always NULL in legacy Vim script.
Problem: ERROR_UNKNOWN clashes on some systems.
Solution: Rename ERROR_ to FCERR_. (Ola Söder, closesvim/vim#5415)
ef140544f6
Remove ERROR_BOTH which was removed from Vim in patch 7.4.1582.
fix(treesitter): get_captures_at_position returns metadata
Return the full `metadata` table for the capture instead of just the
priority.
Further cleanup of related docs.
Makes it possible to use `vim.fs.find` to find files where only a
substring is known.
This is useful for `vim.lsp.start` to get the `root_dir` for languages
where the project-file is only known by its extension, not by the full
name.
For example in .NET projects there is usually a `<projectname>.csproj`
file in the project root.
Example:
vim.fs.find(function(x) return vim.endswith(x, '.csproj') end, { upward = true })
Problem: Function called at debug prompt is also debugged.
Solution: Reset the debug level while entering the debug command.
(closesvim/vim#11118)
b1842de5ca
Problem: There is no easy way to translate a string with a key code into a
readable string.
Solution: Add the keytrans() function. (closesvim/vim#11114)
cdc839353f
vim-patch:7b2d87220c6c
Add missing part of patch
7b2d87220c
Problem: Error messages for setcmdline() could be better.
Solution: Use more specific error messages. (Yegappan Lakshmanan,
closesvim/vim#10995)
25f1e55562
Cherry-pick tv_check_for_opt_number_arg() from Vim.
Problem: 'wildignorecase' is sometimes not used for glob().
Solution: Also use 'wildignorecase' when there are no wildcards.
(closesvim/vim#10066, closesvim/vim#8350)
a3157a476b
Problem: There is no way to get the byte index from a virtual column.
Solution: Add virtcol2col(). (Yegappan Lakshmanan, closesvim/vim#10477,
closesvim/vim#10098)
5a6ec10cc8
Cherry-pick tv_check_for_number_arg() from Vim.
Cherry-pick pathshorten() doc change.
Problem: Vim9: not enough function arguments checked for string.
Solution: Check in balloon functions. Refactor function arguments.
32105ae88f
Cherry-pick removal of useless check from patch 8.2.3840.
vim-patch:8.2.3083: crash when passing null string to charclass()
Problem: Crash when passing null string to charclass().
Solution: Bail out when string pointer is NULL. (Christian Brabandt,
closesvim/vim#8498, closesvim/vim#8260)
72463f883c
The execution of the LspDetach autocommands in the LSP client's on_exit
function are scheduled on the event loop to avoid making API calls in a
fast context; however, this means that by the time the LspDetach
autocommands finally run the client object has already been deleted.
To address this, we also schedule the deletion of the client on the
event loop so that it is guaranteed to occur after all of the LspDetach
autocommands have fired.
Add --always flag to `git describe` so version generation succeeds if
current directory is in a git repo. If not in git repo, fall back to a
default version in the format vx.y.z-dev
dispatch.sr.ht is being deprecated, meaning that using sourcehut CI
won't be possible (see https://github.com/neovim/neovim/issues/19609).
Since Github Actions doesn't provide any BSD runners an external service
is required and Cirrus CI seems like a good replacement for sourcehut.
Initially experimented with using FreeBSD and OpenBSD virtual machines
in GitHub Actions, but Cirrus has been a much better fit with better
performance, logs and overall experience.
Failing tests are automatically skipped on FreeBSD regardless if it's on
CI or not. Ideally these tests should only be skipped in CI with the
help of `isCI` helper function. Unfortunately, the tests don't recognize
the environment variable CIRRUS_CI even if it's set manually. This
workaround is good enough for the time being, but we might want to only
skip tests when using the CI (or even better, fix the failing tests).
Closes: https://github.com/neovim/neovim/issues/19609
Problem: Jsonnet files are not recognized.
Solution: Add a pattern for Jsonnet files. (Cezary Drożak, closesvim/vim#11073,
closesvim/vim#11081)
2a4c885d54
build: remove ARGS from add_custom_command
It's a command that doesn't do anything, kept only for compatibility
reasons.
build: remove unnecessary policy related code
Having cmake version 3.10 as the required minimum version ensures these
are set to new by default.
build: replace deprecated CMAKE_COMPILER_IS_GNUCC variable
Instead use the recommended form CMAKE_C_COMPILER_ID MATCHES "GNU"
build: remove unnecessary build functions
These functions serve no purpose if they're only used as intermediary
functions that passes on arguments to ExternalProject_Add.
Use the first, not last, query for a language on runtimepath. Typically,
this implies that a user query will override a site plugin query, which
will override a bundled runtime query.
Problem: Treesitter queries for a given language in runtime were merged together,
leading to errors if they targeted different parser versions (e.g., bundled viml queries
and those shipped by nvim-treesitter).
Solution: Runtime queries now work as follows:
* The last query in the rtp without `; extends` in the header will be used as the base query
* All queries (without a specific order) with `; extends` are concatenated with the base query
BREAKING CHANGE: queries need to be updated if they are meant to extend other queries
Doing so on `BufDelete` has issues:
- `BufDelete` is also fired for listed buffers that are made unlisted.
- `BufDelete` is not fired for unlisted buffers that are deleted.
This means that diagnostics will be lost for a buffer that becomes unlisted.
It also means that if an entry exists for an unlisted buffer, deleting that
buffer later will not remove its entry from the cache (and you may see "Invalid
buffer id" errors when using diagnostic functions if it was wiped).
Instead, remove a buffer from the cache if it is wiped out.
This means simply `:bd`ing a buffer will not clear its diagnostics now.
"cfuncs" was only ever used to wrap luarefs. As vim8script is
finished and will not be developed further, support for "cfuncs"
for other usecases are not planned. This abstraction was immediately
broken anyway in order to get luarefs out of userfuncs again.
Even if a new kind of userfunc needs to be invented in the future,
likely just extending the FC_... flag union directy, instead of
invoking unnecessary heap object and c function pointer indirection,
will be a more straightforward design pattern.
Change the default build type to always be Debug, and allow only four
predefined build types: Debug, Release, RelWithDebInfo and MinRelSize.
Furthermore, flags meant for single-configuration generator (make,
ninja) will not be used for multi-configuration generators (visual
studio, Xcode), and flags meant for multi-configuration generators will
not be used for single-configuration generators.
This will allow Debug builds to be built with MSVC which requires that
all dependencies are also built with the Debug build type to avoid
runtime library mismatch.
The correct way to specify build type (for example Release) for
single-configuration generators (Make and Ninja) is to run
cmake -B build -D CMAKE_BUILD_TYPE=Release
cmake --build build
while for multi-configuration generators (Visual Studio, Xcode and Ninja
Multi-Config) is to run
cmake -B build
cmake --build build --config Release
Passing CMAKE_BUILD_TYPE for multi-config generators will now not only
not be used, but also generate a warning for the user.
Co-authored-by: dundargoc <gocundar@gmail.com>
- Added 'spell' option to extmarks:
Extmarks with this set will have the region spellchecked.
- Added 'noplainbuffer' option to 'spelloptions':
This is used to tell Neovim not to spellcheck the buffer. The old
behaviour was to spell check the whole buffer unless :syntax was set.
- Added spelling support to the treesitter highlighter:
@spell captures in highlights.scm are used to define regions which
should be spell checked.
- Added support for navigating spell errors for extmarks:
Works for both ephemeral and static extmarks
- Added '_on_spell_nav' callback for decoration providers:
Since ephemeral callbacks are only drawn for the visible screen,
providers must implement this callback to instruct Neovim which
regions in the buffer need can be spell checked.
The callback takes a start position and an end position.
Note: this callback is subject to change hence the _ prefix.
- Added spell captures for built-in support languages
Co-authored-by: Lewis Russell <lewis6991@gmail.com>
Co-authored-by: Björn Linse <bjorn.linse@gmail.com>
* Add vim.treesitter.start() for starting treesitter highlighting via
ftplugin or autocommand (can be extended later for fold, indent,
matchpairs, ...)
* Add vim.treesitter.stop() for manually stopping treesitter
highlighting
* Enable treesitter highlighting for Lua if
`vim.g.ts_highlight_lua = true` is set in `init.lua`
The @error capture is used for tree-sitter's ERROR node, which indicates
a parsing error -- which can be quite frequent (and jarring) while typing.
Users can still manually `hi link @error Error` in their config.
vim-patch:9.0.0386: some code blocks are nested too deep
N/A patches for version.c:
vim-patch:9.0.0385: GUI: when CTRL-D is mapped in Insert mode it gets inserted
Problem: Cannot force getting MouseMove events.
Solution: Add the 'mousemoveevent' option with implementaiton for the GUI.
(Ernie Rael, closesvim/vim#10044)
c4cb544cd5
This only ports the docs and the option variable.
The following commits will actually implement it as a UI option.
Problem: filetype detection does not run on piped input
Solution: add `StdinReadPost` to main filetype.lua autocommand
Rationale: legacy filetype detection checked contents by sourcing
`scripts.vim` in separate autocommands, including on `StdinReadPost`.
For Lua filetype detection, this was moved into the main autocommand,
with bundled `scripts.vim` gated behind `g:do_legacy_filetype` (i.e.,
only user `scripts.vim` are sourced for compatibility by default). Adding
`StdinReadPost` to the main autocommand again runs content checks on
piped input without requiring code duplication and low-payoff
refactoring.
Problem: Crash when invalid line number on :for is ignored.
Solution: Do not check breakpoint for non-existing line.
35d21c6830
Test does not fail without the fix in Nvim as Nvim uses 0 when line
number overflows. If it is changed to MAXLNUM then the test does fail
without the fix, but using 0 seems better as E481 is still given.
Problem: Not all file read and writecode is tested.
Solution: Add a few tests. (Dominique Pellé, closesvim/vim#6764)
1b04ce2d40
Cherry-pick Test_glob() from patch 8.2.0634.
Now nvim_parse_cmd and nvim_create_user_command use a "tab" value which
is the same as the number passed before :tab modifier instead of the
number plus 1, and "tab" value is -1 if :tab modifier is not used.
Prior to this PR, when freeing event callbacks, UI compositor did not
free the luarefs which could cause potential memory leaks. This PR fixes
that by freeing the luarefs properly.
Problem: ":wincmd =" equalizes in two directions.
Solution: Make ":vertical wincmd =" equalize vertically only and
":horizontal wincmd =" equalize horizontally only.
21c3a80a7f
Problem: ColorScheme autocommand triggered when colorscheme is not found.
(Romain Lafourcade)
Solution: Only trigger ColorScheme when loading the colorscheme succeeds.
(closesvim/vim#11024)
5d09a401ec
Most of Test_colorscheme() is applicable to Nvim.
vim-patch:9.0.0341: mapset() does not restore <Nop> mapping properly
Problem: mapset() does not restore <Nop> mapping properly.
Solution: Use an empty string for <Nop>. (closesvim/vim#11022)
92a3d20682
This introduces the following breaking changes:
- nvim_get_keymap now always returns a LuaRef object as "callback" for a
Lua mapping regardless of how it is called. The LuaRef object can be
called from Lua and Vim script, but is lost over RPC.
- maparg() now returns a Funcref instead of a ref number as "callback"
for a Lua mapping. The Funcref can be called from Lua and Vim script,
but is lost over RPC.
This may also make nvim_get_keymap faster, but make maparg() slower.
Problem: Test for crash fix does not fail without the fix.
Solution: Adjust the test sequence. (closesvim/vim#8506)
3777d6e32b
Cherry-pick CheckUnix from patch 8.2.1432.
Problem: Insufficient testing for exception handling and the "attention"
prompt.
Solution: Add test cases. (Yegappan Lakshmanan, closesvim/vim#5681)
b654103ad1
Fix memory leak from last char_u refactor.
`!did_throw` doesn't exactly imply `!current_exception`, as `did_throw = false`
is sometimes used to defer exception handling for later (without forgetting the
exception). E.g: uncaught exception handling in `do_cmdline()` may be deferred
to a different call (e.g: when `try_level > 0`).
In #7881, `current_exception = NULL` in `do_cmdline()` is used as an analogue of
`did_throw = false`, but also causes the pending exception to be lost, which
also leaks as `discard_exception()` wasn't used.
It may be possible to fix this by saving/restoring `current_exception`, but
handling all of `did_throw`'s edge cases seems messier. Maybe not worth
diverging over.
This fix also uncovers a `man_spec.lua` bug on Windows: exceptions are thrown
due to Windows missing `man`, but they're lost; skip these tests if `man` isn't
executable.
* vim-patch:9.0.0314: VDM files are not recognized
Problem: VDM files are not recognized.
Solution: Add patterns for VDM files. (Alessandro Pezzoni, closesvim/vim#11004)
bf26941f40
* vim-patch:9.0.0319: Godot shader files are not recognized
Problem: Godot shader files are not recognized.
Solution: Add patterns for "gdshader". (Maxim Kim, closesvim/vim#11006)
d5c8f11905
Problem: People are confused about `vim.o` and `vim.opt`
Solution: Clarify that `vim.o` is the default interface, with `vim.opt`
specifically meant for interacting with list-style options.
Problem: Clearing screen causes flicker.
Solution: Do not clear but redraw in more cases. Add () to "wait_return".
13608d851a
Only 2 lines of actual code change.
This is a refactor extracted from vim-patch 9.0.0067: cannot show virtual text
The logic for inline virtual text is going to be different in nvim than
text property based text in vim, but this refactor is still useful,
as calculation of displayed linesize is going to be stateful in a
similar way.
vim-patch:9.0.0285: it is not easy to change the command line from a plugin
Problem: It is not easy to change the command line from a plugin.
Solution: Add setcmdline(). (Shougo Matsushita, closesvim/vim#10869)
07ea5f1509
fix(events): trigger WinScrolled when only skipcol changed
vim-patch:9.0.0304: WinScrolled is not triggered when only skipcol changes
Problem: WinScrolled is not triggered when only skipcol changes.
Solution: Add w_last_skipcol and use it. (closesvim/vim#10998)
670ab0334b
Problem: confusing error when using :cc without error list. (Gary Johnson)
Solution: Give the "no errors" error.
a5d78d1f11
Cherry-pick e_quickfix -> e_no_errors rename from patch 8.2.3190.
Problem: The search.c file is a bit big.
Solution: Split off the text object code to a separate file. (Yegappan
Lakshmanan, closesvim/vim#6007)
ed8ce057b7
Problem: The tiny version has the popup menu but not 'wildmenu'.
Solution: Graduate the wildmenu feature.
5416232707
N/A patches for version.c:
vim-patch:9.0.0281: build failure without the +eval feature
Problem: Build failure without the +eval feature.
Solution: Add #ifdef.
58dcbf1c65
Problem: 'buftype' values not sufficiently tested.
Solution: Add and extend tests with 'buftype' values. (closesvim/vim#10988)
93f72cc119
"terminal" and "popup" buffer types cannot be tested, and commenting
them out causes an error, so just remove them.
Problem: BufReadCmd not triggered when loading a "nofile" buffer. (Maxim
Kim)
Solution: Call readfile() but bail out before reading a file.
(closesvim/vim#10983)
b1d2c8116c
Problem: Some source files are too big.
Solution: Move text formatting functions to a new file. (Yegappan
Lakshmanan, closesvim/vim#6021)
11abd09521
Cherry-pick set_can_cindent() from patch 8.1.2062.
Cherry-pick global old_indent from patch 8.2.2127.
This removes the support for defining links via
vim.treesitter.highlighter.hl_map (never documented, but plugins did
anyway), or the uppercase-only `@FooGroup.Bar` to `FooGroup` rule.
The fallback is now strictly `@foo.bar.lang` to `@foo.bar` to `@foo`,
and casing is irrelevant (as it already was outside of treesitter)
For compatibility, define default links to builting syntax groups
as defined by pre-existing color schemes
Problem: The option.c file is still very big.
Solution: Move a few functions to where they fit better. (Yegappan
Lakshmanan, closesvim/vim#4895)
7bae0b1bc8
vim-patch:9.0.0271: using INIT() in non-header files
Problem: Using INIT() in non-header files.
Solution: Remove INIT(). (closesvim/vim#10981)
9b7d2a9596
Problem: Cannot change the characters displayed in the foldcolumn.
Solution: Add fields to 'fillchars'. (Yegappan Lakshmanan, Matthieu Coudron,
closesvim/vim#7860)
3aca5a6fbc
Problem: Too many #ifdefs.
Solution: Make some functions always available.
6d4b2f54df
N/A patches for version.c:
vim-patch:9.0.0262: build failure without the +quickfix feature
Problem: Build failure without the +quickfix feature.
Solution: Add #ifdef.
2e6dcbc445
vim-patch:9.0.0266: compiler warning for unused argument
Problem: Compiler warning for unused argument.
Solution: Add UNUSED.
340dafd155
vim-patch:9.0.0268: build error without the +eval feature
Problem: Build error without the +eval feature.
Solution: Remove #ifdef.
0166e398d1
The private 'get_node_range' function from the languagetree module has
been renamed and remains private as it serve a purpose that is only
relevant inside the languagetree module.
The 'get_node_range' upstreamed from nvim-treesitter in the treesitter
module has been made public as it is in itself a utlity function.
As part of the upstream of utility functions from nvim-treesitter, this
option when set to false allows to return a table (downstream behavior).
Effectively making the switch from the downstream to the upstream
function much easier.
Problem: Crash when opening a buffer for the cmdline window fails. (Chris
Barber)
Solution: Check do_ecmd() succeeds. Reset got_int if "q" was used at the
more prompt. (closesvim/vim#6211)
9b7cce28d5
Make code match latest Vim instead.
Problem: bufload() reads a file even if the name is not a file name. (Cyker
Way)
Solution: Do not read the file when the buffer name is not a file name.
(closesvim/vim#10975)
2eddbacd6d
Problem: Command line completion functions are very long.
Solution: Refactor into multiple functions. (Yegappan Lakshmanan,
closesvim/vim#9753)
620d8edba0
Change fifth argument of set_context_by_cmdname() to "context".
Problem: Warnings reported by MSVC.
Solution: Rename variables and other fixes. (Ken Takata, closesvim/vim#9689)
5411910c77
N/A patches for version.c:
vim-patch:8.2.0091: compiler warnings for size_t / int types
Problem: Compiler warnings for size_t / int types.
Solution: Change type to size_t. (Mike Williams)
4d7a248b64
vim-patch:8.2.1299: compiler warning for using size_t for int and void pointer
Problem: Compiler warning for using size_t for int and void pointer.
Solution: Add type casts.
d3bb6a82a5
vim-patch:8.2.1906: warning for signed/unsigned
Problem: Warning for signed/unsigned.
Solution: Use size_t instead of int. (Mike Williams)
a360dbe3b6
vim-patch:8.2.4531: LGTM warnings for condition and buffer size
Problem: LGTM warnings for condition always true and buffer size too small.
Solution: Remove the useless condition. Make the buffer larger. (Goc
Dundar, closesvim/vim#9914)
f01a653ac5
vim-patch:8.2.4624: old Coverity warning for resource leak
Problem: Old Coverity warning for resource leak.
Solution: Close the file if memory allocation fails.
5d46dcfeed
vim-patch:9.0.0129: compiler warning for int/size_t usage
Problem: Compiler warning for int/size_t usage.
Solution: Add a type cast. (Mike Williams, closesvim/vim#10830)
ab146dac6b
Problem: Compiler warnings for int to size_t conversion. (Randall W.
Morris)
Solution: Add type casts.
551c1aed65
Still keep it size_t, but avoid calculating multiple times.
This marks the remaining part of Vim patch 8.2.3953 as ported, because
ins_compl_upd_pum() has been removed.
vim-patch:8.2.3953: insert completion code is too complicated
Problem: Insert completion code is too complicated.
Solution: More refactoring. Move function arguments into a struct.
(Yegappan Lakshmanan, closesvim/vim#9437)
6ad84ab3e4
Problem: Insert complete code uses global variables.
Solution: Make variables local to the file and use accessor functions.
(Yegappan Lakshmanan, closesvim/vim#9470)
d94fbfc74a
Skip changes in comments for callback-related functions (not ported).
Also make compl_busy static again.
Problem: Some insert completion code is not tested.
Solution: Add a few tests. Refactor thesaurus completion. (Yegappan
Lakshmanan, closesvim/vim#9460)
e982586f8e
vim-patch:9.0.0254: typo in function name
Problem: Typo in function name.
Solution: Rename the function. (closesvim/vim#10971)
5fb3aabc2b
Problem: Missing test coverage.
Solution: Improve test coverage for completion with different encodings,
mapset(), and term function failures. (Dominique Pellé,
closesvim/vim#7877)
a1070eae77
Cherry-pick E716 -> E715 change from patch 8.2.4861.
Problem: Reading past end of completion with a long line and 'infercase'
set.
Solution: Allocate the string if needed.
caea66442d
Cherry-pick the deletion of a blank line from patch 9.0.0027.
N/A patches for version.c:
vim-patch:9.0.0054: compiler warning for size_t to int conversion
Problem: Compiler warning for size_t to int conversion.
Solution: Add type cast. (Mike Williams, closesvim/vim#10741)
c7bd2f08e5
Previously if a highlight group with a name outside the regexp
[a-zA-Z0-9_] was defined, Nvim would emit an "invalid character"
warning message. This was annoying for Lua scripts, as it was very hard
to debug what line of code was triggering this message since it didn't
produce a stack trace.
This has now been promoted to an error with the code E5248.
Additionally the ASCII character period ('.') and at-sign ('@') have
been added to the allowed list of characters of a highlight group name
to support the application of defining hierarchical highlight groups,
e.g. 'TS.keyword'.
Co-authored-by: Christian Clason <christian.clason@uni-due.de>
In non-multigrid UI the only change is that the returned height now
excludes winbar, and this is compatible with Vim.
In multigrid UI this means the return value of these functions now
reflect the space available for buffer lines in a window.
No change in nvim_win_get_height() and nvim_win_get_width().
Problem: Insert completion code is too complicated.
Solution: More refactoring. Move function arguments into a struct.
(Yegappan Lakshmanan, closesvim/vim#9437)
6ad84ab3e4
Skip most pum-related refactoring.
Cherry-pick rename to match_at_original_text() from patch 8.2.4001.
Problem: Insert mode completion functions are too long.
Solution: Split up into multiple functions. (Yegappan Lakshmanan,
closesvim/vim#9431)
5d2e007ccb
Cherry-pick can_cindent_get() -> get_can_cindent() from patch 8.1.2062.
Problem: Insert mode completion function is too long.
Solution: Refactor into multiple functions. (Yegappan Lakshmanan,
closesvim/vim#9423)
edc6f10390
Cherry-pick a typo fix from patch 8.2.3637.
Problem: Empty string considered an error for expand() when 'verbose' is
set. (Christian Brabandt)
Solution: Do not give an error for an empty result. (closesvim/vim#10307)
a96edb736d
Problem: <script> is not expanded in autocmd context.
Solution: Add the context to the pattern struct. (closesvim/vim#10144)
Rename AutoPatCmd to AutoPatCmd_T.
eca7c60d68
Omit AutoPatCmd -> AutoPatCmd_T rename as it is inconsistent.
Use `.sn_name` instead of `->sn_name` as v8.2.0154 hasn't been ported.
Omit acp_script_stx(), use member directly.
Problem: Startup test fails.
Solution: Avoid an error for verbose expansion. Fix that the "0verbose"
command modifier doesn't work.
60895f3e36
Most code changes has already been ported.
Problem: Cannot use expand() to get the script name.
Solution: Support expand('<script>'). (closesvim/vim#10121)
6013d0045d
Use `.sn_name` instead of `->sn_name` as v8.2.0154 hasn't been ported.
Cherry-pick builtin.txt expand() doc from latest Vim.
Problem: Insufficient tests for src/buffer.c.
Solution: Add more tests. Move comments related tests to a separate file.
(Yegappan Lakshmanan, closesvim/vim#6325)
b7e2483655
Problem: Several errors are not tested for.
Solution: Add tests. (Yegappan Lakshmanan, closesvim/vim#5892)
ee4e0c1e9a
Omit Test_range() change: reverted in patch 8.2.0615.
Cherry-pick Test_z_no_space_before_xxx() from patch 8.2.0195.
Cherry-pick Test_reverse_sort_uniq() change from patch 8.2.0183.
Make uniq() error behavior consistent with sort().
Cherry-pick Test_set_ttytype() change from patch 8.1.1826.
Cherry-pick quickfix checks from patch 8.1.2373 to test_preview.vim.
Test_viminfo_error() is applicable.
Cherry-pick E1058 from patch 8.2.0149 and port Test_deep_nest() to Lua.
vim-patch:8.1.1915: more functions can be used as methods
Problem: More functions can be used as methods.
Solution: Make various functions usable as a method.
1a3a89168d
Move debugbreak() to the right place.
vim-patch:8.1.1921: more functions can be used as methods
Problem: More functions can be used as methods.
Solution: Make various functions usable as a method.
a4208966fb
vim-patch:8.1.1953: more functions can be used as a method
Problem: More functions can be used as a method.
Solution: Allow more functions to be used as a method.
f9f24ce7a0
Omit test_termcodes.vim: cannot be used and superseded by later patches.
Cherry-pick test_bufline.vim change from patch 8.1.1993.
Problem: A bit of code is not covered by tests.
Solution: Add a few more test cases. (Dominique Pellé, closesvim/vim#8857)
d176ca3dde
Cherry-pick Test_trim() change from patch 8.2.0448.
Cherry-pick Test_History() change from patch 8.2.1736.
Cherry-pick charidx() and trim() type checks from patch 8.2.3135.
which is both unexpected and different from the Vim behaviour.
Indent was triggered once by the '!' check in insert_execute(), and
inserting the char was correctly skipped, but then triggered again in
insert_check() (provided that cindent was not being ignored after manual
indentation, i.e. `can_cindent == true`).
While this is the smallest fix, another solution would be to remove
VimState#check and instead move that to *_enter()/-_execute(), since the
control flow is pretty unnecessarily convoluted as is. That would also
have the benefit of differing less from the Vim source code.
Problem: Vim9: 'cpo' can become empty.
Solution: Use empty_option instead of an empty string. Update quickfix
buffer after restoring 'cpo'. (closesvim/vim#7608)
e5a2dc87fd
Omit test as it is Vim9 script.
Problem: Some tests fail on Mac.
Solution: Avoid Mac test failures. Add additional test for wildmenu.
(Yegappan Lakshmanan, closesvim/vim#7341)
4b2ce1297e
Problem: Command line popup menu not positioned correctly.
Solution: Also use vim_strsize() on the existing text. (Naruhiko Nishino,
closesvim/vim#9727)
68cc2b8a37
Problem: Popup menu not removed when 'wildmenu' reset while it is visible.
Solution: Do not check p_wmnu, only pum_visible(). (closesvim/vim#10953)
b82a2ab8ad
Problem: 'wildmenu' only shows few matches.
Solution: Add the "pum" option: use a popup menu to show the matches.
(Yegappan Lakshmanan et al., closesvim/vim#9707)
3908ef5017
Omit p_wmnu check in cmdline_pum_active() as it can cause problems.
Omit vim_strchr() flags as that isn't really better than bitmasks.
Omit key translations and document it in vim_diff.txt.
Problem: A command defined with `nargs="?"` returns `fargs={""}` to
a Lua callback when executed with no arguments, which is inconsistent
with how`nargs="*"` behaves.
Solution: Pass `fargs={}` for no argument with `nargs="?"` as well.
Problem: Cannot make difference between the end of :normal and a character
in its argument.
Solution: Add the "typebuf_was_empty" flag. (closesvim/vim#10950)
8d69637133
Problem: Still using cached values after unsetting some known environment
variables.
Solution: Take care of the side effects. (closesvim/vim#10194)
7714231bb5
Cherry-pick vim_setenv_ext() from patch 8.2.0200.
Problem: Loop for handling keys for the command line is too long.
Solution: Move wild menu handling to separate functions. (Yegappan
Lakshmanan, closesvim/vim#6856)
eadee486c7
Problem: The option.c file is too big.
Solution: Split off the code dealing with strings. (Yegappan Lakshmanan,
closesvim/vim#4937)
dac1347b4d
Cherry-pick set_string_option_direct_in_win() from patch 8.1.1405.
Cherry-pick shift_line() comment change from patch 8.1.2096.
Move 'clipboard' default parsing to didset_string_options().
Reorder option flags to put Nvim-only flags at the end.
Problem: "$*" is expanded to "nonomatch".
Solution: Only add "set nonomatch" when using a csh-like shell. (Christian
Brabandt, closesvim/vim#9159, closesvim/vim#9153)
8b8d829faf
Cherry-pick a line from patch 8.2.0522.
Problem: Cannot use a lambda for 'operatorfunc'.
Solution: Support using a lambda or partial. (Yegappan Lakshmanan,
closesvim/vim#8775)
777175b0df
Omit duplicate docs. It's removed in patch 8.2.3623.
Nvim doesn't seem to need callback_set() as it was omitted when patch 8.1.1437
was first ported.
problem: code for drawing statusline is arbitrarily spreadout between drawscreen.c, screen.c and buffer.c
solution: move it to a new file statusline.c
- rename archaic internal name "status match" to public name "wildmenu"
- showruler() does not show the ruler. it show anything which displays
info about the cursor. Rename it accordingy.
fix(filetype): only pass first 100 and last lines to contents check
sufficient for current content checks and avoids performance issues for
buffers with a large number of lines
fixes#19817
docs(lua): update vim.{go,bo,wo} documentation
* document indexing by buffer/window handle
* correct wrapper information (`nvim_buf_{g,s}et_value` now)
* make clear what is considered "invalid key" (consistently)
Problem: Insufficient testing for reading/writing files.
Solution: Add more tests. (Yegappan Lakshmanan, closesvim/vim#6257)
Add "ui_delay" to test_override() and use it for the CTRL-O test.
b340baed9f
Omit test_override().
Reorder test_writefile.vim to match Vim.
Problem: Code for modeless selection not sufficiently tested.
Solution: Add tests. Move mouse code functionality to a common script file.
(Yegappan Lakshmanan, closesvim/vim#5821)
515545e11f
Skip termcode tests for now.
Problem: Cannot read error message when abort() is called.
Solution: Output a newline before calling abort().
213e70e284
Add emsg_not_now() check to make code equivalent.
Problem: When an internal error makes Vim exit the error is not seen.
Solution: Add the error to the test output.
1c67f3a977
Add emsg_not_now() check to make code equivalent.
Problem: Some code in options.c fits better elsewhere.
Solution: Move functions from options.c to other files. (Yegappan
Lakshmanan, closesvim/vim#4889)
e677df8d93
Problem: The screen.c file is much too big.
Solution: Split it in three parts. (Yegappan Lakshmanan, closesvim/vim#4943)
7528d1f6b5
This is an approximation vim-patch 8.1.2057. Applying the patch directly
isn't feasible since our version of screen.c has diverged too much,
however we still introduce drawscreen.c and drawline.c:
- screen.c is now a much smaller file used for low level screen functions
- drawline.c contains everything needed for win_line()
- drawscreen.c contains everything needed for update_screen()
Co-authored-by: zeertzjq <zeertzjq@outlook.com>
Allows preview callbacks to modify multiple buffers in order to show the
preview. Previously, if multiple buffers were modified, only the current
buffer would have its state restored. After this change, all buffers
have their state restored after preview.
Closes#19103.
vim-patch:8.1.2082: some files have a weird name to fit in 8.3 characters
Problem: Some files have a weird name to fit in 8.3 characters.
Solution: Use a nicer names.
30e8e73506
This starts a soft phase-out of `buf_request`.
`buf_request` is quite error prone:
- Positional `params` depend on the client because of the
`offset_encoding`. Currently if there is one client using UTF-8 offset
encoding and another using UTF-16, the positions in the request are
wrong for one of the clients. To solve this the params would need to
be created per client instead of once for all of them.
- `handler` is called *per* client but many users of it assume it is
only called once.
This can lead to a "select n + 1"
kind of problem, where the handler makes another call to `buf_request`,
multiplying the amount of requests.
(There are in fact still some places where this happens in core)
Or it leads to erratic behavior if called multiple times (E.g. the
quicklist list flickering & being overwritten)
(See hover or references implementation)
`buf_request_all` returns an aggregate of the responses which is more
sensible as it avoids this problem.
For off-spec extensions it also has the problem that it sends requests to
clients which cannot handle a given request.
Given that `buf_request` is in use by a lot of plugins this starts a
soft-phase out. Planned Steps:
- Remove from docs
- Provide an alternative, either `buf_request_all`, maybe with
extensions (params being a function), or an entirely new method.
- Mark as deprecated in 0.9
- Remove in 0.10
To note:
- `buf_request_all` currently isn't ideal either because it suffers from
the `params` problem as well.
- This implies that the `vim.lsp.with` pattern will die, because the
global handlers as they are don't fit a multi-client model, as most of
the time an aggregate is needed.
These were just added to avoid churn when changing the default
of 'display'. To simplify message handling logic, we might want
to remove support for printing messages in default_grid later on.
This would allow things like printing error messages safely in the
middle of redraw, or a future graduation of the 'multigrid' feature.
Let :wincmd command accept a count like what its documentation suggests.
Previously it could only accept a range, which led to some ambiguity on
which attribute should be used when executing :wincmd using nvim_cmd.
Closes#19662.
Also fix a typo in a related Vim test:
vim-patch:9.0.0223: typo in diffmode test
Problem: Typo in diffmode test.
Solution: Fix the typo. (closesvim/vim#10932)
5fd6ab820b
Problem: Messages test fails; window size incorrect when 'cmdheight' is
made smaller.
Solution: Properly cleanup after test with cmdheight zero. Resize windows
correctly when 'cmdheight' gets smaller.
d4cf9fc53e
N/A patches for version.c:
vim-patch:9.0.0192: possible invalid memory access when 'cmdheight' is zero
Problem: Possible invalid memory access when 'cmdheight' is zero. (Martin
Tournoij)
Solution: Avoid going over the end of w_lines[] when w_height is Rows.
(closesvim/vim#10882)
fdc5d17d58
Problem: The way 'cmdheight' can be made zero is inconsistent.
Solution: Only make 'cmdheight' zero when setting it explicitly, not when
resizing windows. (closesvim/vim#10890)
f797e309ca
Problem: The argument list may contain duplicates.
Solution: Add the :argdedeupe command. (Nir Lichtman, closesvim/vim#6235)
73a024209c
Use latest index.txt :argdedupe doc from Vim.
Problem: The evalfunc.c file is too big.
Solution: Move getchar() and parse_queued_messages() to getchar.c.
9c658c9eac
Omit parse_queued_messages(): Nvim does not have MESSAGE_QUEUE.
Problem: A few failures are not tested.
Solution: Test a few failures. (Dominique Pellé, closesvim/vim#7075)
afe8cf6170
Cherry-pick a line in Test_argdelete() from patch 8.2.1736.
In Vim :redraw! uses CLEAR, which calls screenclear(), which sets
redraw_cmdline. In Nvim :redraw! uses NOT_VALID, which does not set
redraw_cmdline automatically, so set it explicitly.
Problem: Some Normal mode commands not fully tested.
Solution: Add a few more tests. (Yegappan Lakshmanan, closesvim/vim#7073)
d1ad99b654
Cherry-pick test_registers.vim change from patch 8.2.0644.
Problem: Various Normal mode commands not fully tested.
Solution: Add more tests. (Yegappan Lakshmanan, closesvim/vim#7059)
8a9bc95eae
Cherry-pick Test_normal_gdollar_cmd() change from patch 8.2.0540.
Problem: Mode is not cleared when leaving Insert mode with mapped Esc.
Solution: Clear the mode when redraw_cmdline is set. (closesvim/vim#4269)
4c25bd785a
Problem: Peeking and flushing output slows down execution.
Solution: Do not update the mode message when global_busy is set. Do not
flush when only peeking for a character. (Ken Takata)
cb574f4154
Omit inchar() change: it breaks too many tests.
N/A patches for version.c:
vim-patch:8.2.5170: tiny issues
Problem: Tiny issues.
Solution: Tiny improvements.
944cc9ceba
Patch 8.2.4594 (36a5b6867b) added support for sourcing a buffer without a name.
Patch 8.2.4325 (3908ef5017) added support for using a popup menu for wildmode completion.
Problem: When a test fails it's often not easy to see what the call stack
is.
Solution: Add more entries from the call stack in the exception message.
a5d0423fa1
Use docs from latest Vim.
Problem: The spell.c file is too big.
Solution: Move the code for spell suggestions to a separate file. (Yegappan
Lakshmanan, closesvim/vim#4988)
46a426c9ac
vim-patch:8.2.3675: using freed memory when vim_strsave() fails
Problem: Using freed memory when vim_strsave() fails.
Solution: Clear "last_sourcing_name". Check for msg_source() called
recursively. (closesvim/vim#8217)
ba8c92687d
Although xstrdup() cannot fail in Nvim, it may still be possible that an
error appears (e.g. in regexp engine) when printing the message.
Problem: Vim9: script reload test is disabled.
Solution: Compile a function in the context of the script where it was
defined. Set execution stack for compiled function. Add a test
that an error is reported for the right file/function.
25e0f5863e
Omit stack_top_is_ufunc(): only used by Vim9 script.
Problem: Expanding <sfile> works differently the second time.
Solution: Keep the expanded name when redefining a function. (closesvim/vim#5425)
b9adef79ec
Problem: Execution stack is incomplete and inefficient.
Solution: Introduce a proper execution stack and use it instead of
sourcing_name/sourcing_lnum. Create a string only when used.
1a47ae32cd
Omit test_debugger.vim: superseded by later patches.
Omit check_map_keycodes(): N/A.
Omit kword_test.c: N/A (converted to a unit test).
vim-patch:8.2.1536: cannot get the class of a character; emoji widths are wrong
Problem: Cannot get the class of a character; emoji widths are wrong in
some environments.
Solution: Add charclass(). Update some emoji widths. Add script to check
emoji widths.
4e4473c927
Use latest charclass() docs from Vim.
Rewrite DoIt() in emoji_list.vim in Lua.
Omit emoji table updates:
- emoji_width update looks wrong as these added ranges are only double-width when followed by 0xFE0F.
- Other updates are too old.
vim-patch:8.2.1540: the user cannot try out emoji character widths
Problem: The user cannot try out emoji character widths.
Solution: Move the emoji script to the runtime/tools directory.
98945560c1
`server_capabilities` can be nil until the server is initialized.
Reproduced with:
vim.lsp.stop_client(vim.lsp.start_client {
cmd = { vim.v.progpath, '-es', '-u', 'NONE', '--headless' };
})
vim-patch:8.1.1927: code for dealing with script files is spread out
Problem: Code for dealing with script files is spread out.
Solution: Move the code to scriptfile.c. (Yegappan Lakshmanan, closesvim/vim#4861)
307c5a5bb7
Problem: Profiling functionality is spread out.
Solution: Put profiling functionality in profiling.c. (Yegappan Lakshmanan,
closesvim/vim#4666)
fa55cfc69d
Move proftime_T to types.h for now to avoid recursive #include.
vim-patch:8.2.0559: clearing a struct is verbose
Problem: Clearing a struct is verbose.
Solution: Define and use CLEAR_FIELD() and CLEAR_POINTER().
a80faa8930
before, RelWithDebInfo linking gave this warning:
src/mpack/conv.h:36:16: warning: type of ‘mpack_unpack_boolean’ does not match original declaration [-Wlto-type-mismatch]
36 | MPACK_API bool mpack_unpack_boolean(mpack_token_t t) FUNUSED FPURE;
| ^
src/mpack/conv.c:196:16: note: return value type mismatch
196 | MPACK_API bool mpack_unpack_boolean(mpack_token_t t)
| ^
The change tracking used a single lines/lines_tmp table to track
changes to a buffer.
If multiple clients using incremental sync are connected to a buffer,
they both made changes to the same lines table. That resulted in an
inconsistent state.
This commit changes the didChange handling to group clients by
synchronization scheme and offset encoding.
This avoids computing the diff multiple times for clients using the
same scheme and resolves the lines/lines_tmp conflicts.
Fixes https://github.com/neovim/neovim/issues/19325
Problem: Command line history code is spread out.
Solution: Put the code in a new file. (Yegappan Lakshmanan, closesvim/vim#4779)
Also graduate the +cmdline_hist feature.
d7663c22c6
Problem: setcellwidths() may make 'listchars' or 'fillchars' invalid.
Solution: Check the value and give an error. (closesvim/vim#9024)
94358a1e6e
Cherry-pick f_setcellwidths() change from patch 9.0.0036.
Cherry-pick 'ambiwidth' docs update from runtime update 079ba76ae7a7.
Problem: It is not possible to specify cell widths of characters.
Solution: Add setcellwidths().
08aac3c619
Co-Authored-By: delphinus <me@delphinus.dev>
The lsp client used to wait up to 500ms for a language server to
shutdown before sending a TERM signal.
The intention behind the 500ms grace period was to ensure the language
server exits to prevent stale processes, but it has the side-effect that
it can interrupt language-servers which are too slow to shutdown within
500ms. Language servers tend to write out index files or project files
on shutdown, and being interrupted during this process can cause
corruption of those files.
This changes the default to not wait at all, at the risk of leaving
stale processes around if the language server isn't well behaved.
An alternative would be to wait indefinitely, but that can cause neovim
to take several seconds to exit.
Problem: Ex command error cannot contain an argument.
Solution: Add ex_errmsg() and translate earlier. Use e_trailing_arg where
possible.
8930caaa1a
Remove duplicate test file 062_tab_pages_spec.lua
Problem: Support for user commands is spread out. No good reason to make
user commands optional.
Solution: Move user command support to usercmd.c. Always enable the
user_commands feature.
ac9fb18020
Problem: Options window still checks for the multi_byte feature.
Solution: Remove the unnecessary check. (Dominique Pelle, closesvim/vim#3990)
76cbe811da
Problem: "delmenu" does not remove autocmmands. Running menu test function
alone fails.
Solution: Delete autocommands Make sure there is at least one menu.
(closesvim/vim#10848)
206fce307b
- Separate preview and callback functions to make the example easier to understand
- Use false instead of 0 for boolean arguments in API function calls
- Remove explicit nil checks for consistency
- Format with stylua
Test ":unlet self-referencing node in a List graph #6070" feeds many characters into typeahead, so a timeout of only 100 milliseconds sometimes fails. Change that timeout to 1000 milliseconds.
Problem: Buffer menu does not handle special buffers properly.
Solution: Keep a dictionary with buffer names to reliably keep track of
entries.
Also trigger BufFilePre and BufFilePost for command-line and
terminal buffers when the name changes.
5e94a29ebb
Problem: Vim9: need more tests for empty string arguments.
Solution: Add more tests. Also use empty argument with menu_info() to get
the top-level menu names. (Yegappan Lakshmanan, closesvim/vim#8925)
51491adfa8
Problem: Menu functionality insufficiently tested.
Solution: Add tests. Add menu_info(). (Yegappan Lakshmanan, closesvim/vim#5760)
0eabd4dc8f
Omit feedkeys() change: even if "L" flag is implemented it will likely
use input_enqueue(), which already checks for interrupts.
Omit Test_mouse_popup_menu(): already tested in Lua.
- Use DIRECTORY instead of PATH in get_filename_component
- Use COMPILE_OPTIONS instead of COMPILE_FLAGS. COMPILE_FLAGS is treated
as a single string while COMPILE_OPTIONS is a list, meaning that cmake
will take care of any escaping and quoting automatically.
N/A patches for version.c:
vim-patch:9.0.0135: comment about tabpage line above the wrong code
Problem: Comment about tabpage line above the wrong code.
Solution: Move the comment. (closesvim/vim#10836)
0b0ccbbfb0
The targets will only format files that have been changed in current
branch compared to the master branch. This includes unstaged, staged and
committed files.
Add following make and cmake targets:
formatc - format changed c files
formatlua - format changed lua files
format - run formatc and formatlua
Remove scripts/uncrustify.sh as this deprecates it.
Problem: Reading past end of line with "gf" in Visual block mode.
Solution: Do not include the NUL in the length.
395bd1f6d3
Omit trailing space: removed in patch 9.0.0126.
Problem: With latin1 encoding CTRL-W might go before the start of the
command line.
Solution: Check already being at the start of the command line.
ef02f16609
vim-patch:9.0.0126: expanding file names fails in dir with more than 255 entries
Problem: Expanding file names fails in directory with more than 255
entries.
Solution: Use an int instead of char_u to count. (John Drouhard,
closesvim/vim#10818)
95fca12b0e
`:saveas newName` changes the name of an existing buffer.
Due to the buffer re-use it skips the lsp attach phase and immediately
sends a `didSave` notification to the server.
Servers get confused about this, because they expect a `didOpen`
notification first.
Closes https://github.com/neovim/neovim/issues/18688
Problem: Default cmdwin mappings are re-mappable.
Solution: Make the default mappings not re-mappable. (closesvim/vim#10580) Use
symbols for the first do_map() argument.
44068e97db
Problem: Crash when using maparg() and unmapping simplified keys.
Solution: Do not keep a mapblock pointer. (closesvim/vim#10294)
2c8a7ebdad
N/A patches for version.c:
vim-patch:8.2.0832: compiler warning for uninitialized variable
Problem: Compiler warning for uninitialized variable. (Tony Mechelynck)
Solution: Add initial value.
a55ba06f69
Problem: maparg() does not provide enough information for mapset().
Solution: Add "lhsraw" and "lhsrawalt" items. Drop "simplified"
9c65253fe7
vim-patch:9.0.0127: unused variable
Problem: Unused variable.
Solution: Remove the variable. (closesvim/vim#10829)
e95f22f63a
Problem: Cannot easily restore a mapping.
Solution: Add mapset().
4c9243f9fb
Use MapArgument to reduce number of arguments of map_add().
N/A patches for version.c:
vim-patch:8.2.0809: build failure with small features
Problem: Build failure with small features. (Tony Mechelynck)
Solution: Move "expr" inside #ifdef.
5a80f8ad5d
Problem: When 'cmdheight' is zero pressing ':' may scroll a window.
Solution: Add the made_cmdheight_nonzero flag and set 'scrolloff' to zero.
6747cf1671
Problem: The command line takes up space even when not used.
Solution: Allow for 'cmdheight' to be set to zero. (Shougo Matsushita,
closesvim/vim#10675, closesvim/vim#940)
f39cfb7262
Omit win_redr_ruler() change: winbar may still need redraw.
Omit win_update() changes: Nvim doesn't use `Rows` there.
Omit redraw_asap(): removed.
Problem: Writing over the end of a buffer on stack when making list of
spell suggestions.
Solution: Make sure suggested word is not too long. (closesvim/vim#10812)
1eead4cf1d
Most casts where PVS warns for V1028 aren't added to prevent overflows
in the first place, but to avoid other warnings, like printf argument or
-Wconversion warnings. PVS/V1028 is more annoying than useful.
`code_action` gained extra functions (`filter` and `apply`) which
`range_code_action` didn't have.
To close this gap, this adds a `range` option to `code_action` and
deprecates `range_code_action`.
The option defaults to the current selection if in visual mode.
This allows users to setup a mapping like `vim.keymap.set({'v', 'n'},
'<a-CR>', vim.lsp.buf.code_action)`
`range_code_action` used to use the `<` and `>` markers to get the
_last_ selection which required using a `<Esc><Cmd>lua
vim.lsp.buf.range_code_action()<CR>` (note the `<ESC>`) mapping.
Problem: Scrollback can be wrong after redrawing the command line.
Solution: Clear unfinished scrollback when redrawing. (closesvim/vim#10807)
46af7bc08d
Problem: Cursor restored unexpected with nested autocommand.
Solution: Do not restore the cursor when it was moved intentionally.
(closesvim/vim#10780)
3d6ee8bda0
Problem: When switching window in autocmd the restored cursor position may
be wrong.
Solution: Do not restore the cursor if it was not set. (closesvim/vim#10775)
b03950fafa
This patch cannot be fully ported because it depends on patch 8.2.3518.
Problem: The ins_complete() function is much too long.
Solution: Split it up into multiple functions. (Yegappan Lakshmanan,
closesvim/vim#9414)
bf7ff61af4
Reduce the amount of string allocations and length calculations.
With the following benchmark:
```lua
total = 0
for _ = 1, loops do
local start = now()
vim.api.nvim_cmd({cmd = 'let', args = {'a', '=', '1'}}, {})
total = total + (now() - start)
end
print('nvim_cmd', total / loops)
```
```
hyperfine 'nvim --clean test.lua +source +q'
```
Before: 234.5ms
After: 173.8ms
Without some form of feedback a user cannot easily tell if the server is
still computing the result (which can take a while in large projects),
or whether the server couldn't compute the rename result.
Problem: Coverity warns for using uninitialized field.
Solution: Initialize he field to zero.
03a297c63f
Also only initialize used fields in f_fullcommand().
Problem: Command line completion of user command may have duplicates.
(Dani Dickstein)
Solution: Skip global user command if an identical buffer-local one is
defined. (closesvim/vim#10797)
c2842adfb2
Problem: ModeChanged event not triggered when leaving the cmdline window.
Solution: Call may_trigger_modechanged(). (closesvim/vim#10791)
c9e8fd6fc7
Code is already present in Nvim. Add some other related missing changes.
It currently falls back to texlua if luajit doesn't exist. However,
the documentation generation does not work with texlua. Instead use lua
as a fall back instead.
Problem: Using "terraform" filetype for .tfvars file is bad.
Solution: use "terraform-vars", so that different completion and other
mechanisms can be used. (Radek Simko, closesvim/vim#10755)
15b87b6610
It's more convenient to not specify the version and let openbsd's
package manager figure it out. This will help us avoid manually bumping
dependency versions when a new version of openbsd is released.
Some packages have multiple versions and not specifying a version in
those cases fails the CI job, so providing a version seems to be
necessary for some key packages.
When using a Makefile generator, older CMake versions are not building
custom targets when depending on their byproducts, making version
generation fail. (works on 3.19.4, fails on 3.10.2) As a workaround,
add a dependency on the custom target explicitly.
Problem: Command overlaps with printed text in scrollback.
Solution: Clear until end-of-line and use correct message chunk.
(closesvim/vim#10765, closesvim/vim#10764)
ecdc82e74e
N/A patches for version.c:
vim-patch:9.0.0070: using utfc_ptr2char_len() when length is negative
Problem: Using utfc_ptr2char_len() when length is negative.
Solution: Check value of length. (closesvim/vim#10760)
4dc513a22c
Currently LSP allows only using loclist or quickfix list window. I
normally prefer to review all quickfix items without opening quickfix
window. This fix allows passing `on_list` option which allows full
control what to do with list.
Here is example how to use it with quick fix list:
```lua
local function on_list(options)
vim.fn.setqflist({}, ' ', options)
vim.api.nvim_command('cfirst')
end
local bufopts = { noremap=true, silent=true, buffer=bufnr }
vim.keymap.set('n', '<leader>ad', function() vim.lsp.buf.declaration{on_list=on_list} end, bufopts)
vim.keymap.set('n', '<leader>d', function() vim.lsp.buf.definition{on_list=on_list} end, bufopts)
vim.keymap.set('n', '<leader>ai', function() vim.lsp.buf.implementation{on_list=on_list} end, bufopts)
vim.keymap.set('n', '<leader>at', function() vim.lsp.buf.type_definition{on_list=on_list} end, bufopts)
vim.keymap.set('n', '<leader>af', function() vim.lsp.buf.references(nil, {on_list=on_list}) end, bufopts)
```
If you prefer loclist do something like this:
```lua
local function on_list(options)
vim.fn.setloclist(0, {}, ' ', options)
vim.api.nvim_command('lopen')
end
```
close#19182
Co-authored-by: Mathias Fußenegger <mfussenegger@users.noreply.github.com>
Problem: Switching window uneccarily when getting buffer options.
Solution: Do not switch window when getting buffer options. (closesvim/vim#10767)
cd6ad6439d
Problem: Duplicate code in "get" functions.
Solution: Use get_var_from() for getwinvar(), gettabvar(), gettabwinvar()
and getbufvar(). (closesvim/vim#10335)
47d4e317f8
f_setbufvar() can use tv_get_buf_from_arg() as it sets emsg_off.
Problem: No tests for clicking in the GUI tabline.
Solution: Add test functions to generate the events. Add tests using the
functions. (Yegappan Lakshmanan, closesvim/vim#9638)
b0ad2d92fd
Problem: Subtracting from number option fails when result is zero. (Ingo
Karkat)
Solution: Reset the string value when using the numeric value.
(closesvim/vim#8351)
a42e6e0082
Cherry-pick Test_compound_assignment_operators() changes from patch 8.2.1593
Problem: Vim9: bool option type is number.
Solution: Have get_option_value() return a different value for bool and
number options. (closesvim/vim#7583)
dd1f426bd6
Problem: When splitting a window the changelist position moves.
Solution: Set the changelist index a bit later. (closesvim/vim#10493)
e6f13b473c
Cherry-pick Test_nv_hat_count() change from patch 8.2.1593.
Problem: The changelist index is not remembered per buffer.
Solution: Keep the changelist index per window and buffer. (closesvim/vim#10135,
closesvim/vim#2173)
db0ea7f2b0
Cherry-pick FOR_ALL_BUF_WININFO from patch 8.2.0500.
Cherry-pick test_changelist.vim change from patch 8.2.3795.
Problem: Vim9: error when setting an option with setbufvar().
Solution: Do not get a number from a string value. (closesvim/vim#6740)
191929b182
Vim9 is N/A, so this just refactors the code without changing behavior.
Problem:
Build fails without git or .git/.
ref #19289
Solution:
Fix the version generation logic.
Test cases:
If `git` is missing:
-- Using NVIM_VERSION_MEDIUM: v0.8.0-dev
If `.git/` is missing:
-- Git tag extraction failed:
fatal: not a git repository (or any of the parent directories): .git
-- Using NVIM_VERSION_MEDIUM: v0.8.0-dev
If `git describe` fails
-- Git tag extraction failed:
fatal: ...
-- Using NVIM_VERSION_MEDIUM: v0.8.0-dev
Co-authored-by: Justin M. Keyes <justinkz@gmail.com>
This type itself is not eval-specific. Moving it to types.h can avoid
including eval/funcs.h in many headers, and types.h is already included
by many headers.
Problem: The eval.c file is too big.
Solution: Move code related to variables to evalvars.c. (Yegappan
Lakshmanan, closesvim/vim#4868)
0522ba0359
Name the new file eval/vars.c instead.
vim-patch:8.2.4502: in the GUI a modifier is not recognized after CTRL-X
Problem: In the GUI a modifier is not recognized for the key typed after
CTRL-X, which may result in a mapping to be used. (Daniel
Steinberg)
Solution: Recognize a modifier starting with CSI. (closesvim/vim#9889)
d979d64fa2
Code is N/A. This just ports the test change.
Test depends on README.txt in testdir. Add that file.
Reorder test_ins_complete.vim to match upstream.
Problem: File for Insert mode is much too big.
Solution: Split off the code for Insert completion. (Yegappan Lakshmanan,
closesvim/vim#4044)
7591bb39d5
Cherry-pick ins_compl_len() -> get_compl_len() from patch 8.2.4001.
Revert a71c5e9eb9: ctrl_x_mode is no
longer a global variable, so l_ctrl_x_mode is no longer needed.
* Problem
Quotes are special in doxygen, and should be escaped. *Sometimes* they
cause doc generation issues. Like in #17785
* Solution
Replace double quotes with single quotes
Use the same logic for both deps (including LuaJIT, for which setting
this variable is mandatory) and Nvim: either the eponymous environment
variable if set, or the current software version if not.
Removes annoying warnings when building locally on macOS.
Problem: Not all command line arguments are tested.
Solution: Add tests for -D and -serverlist. (Dominique Pellé, closesvim/vim#7992)
c5cf369e95
Cherry-pick two deletions from patch 8.2.1799.
Problem: various code is not properly tested.
Solution: Add more tests. (Yegappan Lakshmanan, closesvim/vim#5871)
cde0ff39da
Cherry-pick test_clientserver.vim change from patch 8.1.1826.
Cherry-pick a comment from patch 8.2.0301.
Omit test_viminfo.vim.
In case nvim A sends nvim_error_event to nvim B, it would
respond with another nvim_error_event due to unknown
request name. Fix this by adding dummy request handler for now.
vim-patch:8.2.1147: :confirm may happen in cooked mode
Problem: :confirm may happen in cooked mode. (Jason Franklin)
Solution: Switch to raw mode before prompting. (Brandon Pfeifer)
27321dbeed
Code is N/A, but this improves tests. Port the improved tests to Lua.
Problem:
Since 0b9664f524 powershell filtered
:[range]! commands with args causes error:
"Start-Process: A positional parameter cannot be found that accepts argument ..."
Solution:
Pass args to Start-Process via `-ArgumentList`.
closes#19250
vim-patch:9.0.0055: bitbake files are not detected
Problem: Bitbake files are not detected.
Solution: Add bitbake filetype detection by file name and contents. (Gregory
Anders, closesvim/vim#10697)
fa49eb4827
Problem:
Since right-click can now show a popup menu, we can provide messaging to
guide users who expect 'mouse' to be disabled by default. So 'mouse' can
now be enabled by default.
Solution:
Do it.
Closes#15521
Problem: Can still switch windows for 'completefunc'.
Solution: Also disallow switching windows for other completions.
3eb6bd9c2b
Assert E565 instead of E578.
Need to assert a different string because patch 8.2.1919 hasn't been
ported yet.
Problem: Allowing 'completefunc' to switch windows causes trouble.
Solution: use "textwinlock" instead of "textlock".
28976e2acc
Assert E565 instead of E578.
vim-patch:8.2.0670: cannot change window when evaluating 'completefunc'
Problem: Cannot change window when evaluating 'completefunc'.
Solution: Make a difference between not changing text or buffers and also
not changing window.
6adb9ea0a6
vim-patch:8.2.5029: "textlock" is always zero
Problem: "textlock" is always zero.
Solution: Remove "textlock" and rename "textwinlock" to "textlock".
(closesvim/vim#10489)
cfe456543e
Problem: Some tests are known to cause an error with ASAN.
Solution: Add CheckNotAsan.
97202d9516
Move CheckNotMSWindows to the right place.
Omit test_memory_usage.vim: a Lua test is used for this file.
Problem: Cannot use :vimgrep in omni completion, causing C completion to
fail.
Solution: Add the EX_LOCK_OK flag to :vimgrep. (closesvim/vim#7292)
33aecb1f2c
Cherry-pick Test_resize_from_copen() indent change from patch 8.2.1432.
Problem: Vim9: error when using vim9script in TextYankPost.
Solution: Use EX_LOCKOK instead of the EX_CMDWIN flag for command that can
be used when text is locked. (closesvim/vim#6529)
37394ff752
Problem: Insert mode completion not fully tested.
Solution: Add a few more tests. (Yegappan Lakshmanan, closesvim/vim#6041)
f9ab52e155
Cherry-pick test_ins_complete.vim changes from patches 8.2.{0522,0615}.
Problem: Not enough testing for movement commands.
Solution: Add more tests. (Yegappan Lakshmanan, closesvim/vim#6313)
bdd2c290d3
Cherry-pick test_functions.vim changes from patch 8.2.0183.
Cherry-pick Test_normal18_z_fold() change from patch 8.2.0540.
Problem: Various code not covered by tests.
Solution: Add a few more test. (Yegappan Lakshmanan, closesvim/vim#7995)
1f448d906b
Cherry-pick some test_edit.vim changes from patches 8.2.{1022,1432}.
Reorder test_undo.vim to match upstream.
Problem: Character input not fully tested.
Solution: Add more tests. (Yegappan Lakshmanan, closesvim/vim#7963)
f4fcedc59d
Cherry-pick related changes from patches 8.2.{0433,0866}.
Problem:
1. Right-click does not work in statusline unless you left-click first (to focus
the statusline).
2. Modifier (e.g. shift+rightclick) does not work in statusline.
Solution:
Make clickable statusline sections receive right-clicks regardless of whether
the statusline is focused.
Closes#18994
Problem: No test for E386.
Solution: Add a test. (Dominique Pelle, closesvim/vim#5911)
8832a34578
Cherry-pick Test_search_errors() from patch 8.2.0448.
Problem: When 'buftype' is "nofile" there is no overwrite check.
Solution: Also check for existing file when 'buftype' is set.
(closesvim/vim#5807)
a3a9c8ef69
Nvim doesn't have buftype=popup and doesn't allow buftype=terminal for
non-terminal buffer.
Problem: Some errors were not tested for.
Solution: Add tests. (Dominique Pelle, closesvim/vim#5824)
9b9be007e7
Cherry-pick get_highest_fnum() from patch 8.1.1908 to make tests pass.
Problem: FEAT_WINDOWS adds a lot of #ifdefs while it is nearly always
enabled and only adds 7% to the binary size of the tiny build.
Solution: Graduate FEAT_WINDOWS.
4033c55eca
Problem: "zG" may throw an error if invalid character follows.
Solution: Pass the word length to valid_spell_word(). (Ken Takata,
closesvim/vim#10737)
2ebcc35826
Problem: CTRL-W T in cmdline window causes trouble.
Solution: Disallow CTRL-W T in the cmdline window. Add more tests.
(Naruhiko Nishino, closesvim/vim#6219)
4fdb8bd054
Test already passes in Nvim because of later patches.
Move Test_cmdwin_jump_to_win() to the right place.
Problem: Various Normal mode commands not fully tested.
Solution: Add more tests. (Yegappan Lakshmanan, closesvim/vim#5751)
1671f44881
Cherry-pick a fix from patch 8.2.3162.
Omit test_iminsert.vim as previous patches to that file are N/A, and
Nvim doesn't support iminsert=2 either, so that test isn't useful.
Problem: Various code not covered by tests.
Solution: Add more test coverage. (Yegappan Lakshmanan, closesvim/vim#5720)
91ffc8a5f5
Test_Ex_echo_backslash() is not applicable to Vim enhanced Ex mode.
Problem: Some code in ex_getln.c not covered by tests.
Solution: Add more tests. (Yegappan Lakshmanan, closesvim/vim#5717)
0546d7df13
Nvim doesn't support imsearch=2, commenting out that line is enough.
Test_Ex_append() is not applicable to Vim enhanced Ex mode.
Omit test_iminsert.vim as previous patches to that file are N/A.
- only update git-version if both of these conditions are met:
- `git` command succeeds
- `versiondef_git.h` would change (SHA1-diff)
- else print a status/warning message
also move version generation out of Lua into cmake.
Problem: Using freed memory when searching for pattern in path.
Solution: Make a copy of the line.
409510c588
Cherry-pick Test_def_search() -> Test_macro_search() from patch 8.2.0369
Problem: Various Ex commands not sufficiently tested.
Solution: Add more test cases. (Yegappan Lakshmanan, closesvim/vim#5673)
818fc9ad14
Needs to assert E170 instead of E580 because patch 8.2.3486 has been
ported but patch 8.2.1183 hasn't.
vim.lsp.start_client() may fail (for example if the `cmd` is not
executable). It produces a nice error notification in this case. Passing
the `nil` value returned from an erroneous `vim.lsp.start_client()` call
into `vim.lsp.buf_attach_client()` causes a meaty param validate
exception message. Avoid this.
Problem: Another failing timer test.
Solution: Assert that timers are finished by the end of the test. Rename
test functions to make them easier to find.
9a2fddcf04
Use test_garbagecollect_now() in Test_timer_retain_partial() like Vim.
It's a leftover artifact that currently just acts as an unnecessary
intermediary script that calls the Makefile. It can be replaced by just
calling the Makefile directly.
Problem: When :edit reuses the current buffer the alternate file is set to
the same buffer.
Solution: Only set the alternate file when not reusing the buffer.
(closesvim/vim#8783)
b8bd2e6eba
Cherry-pick Test_cmdline_expand_special() from patches 8.2.{0243,2873}.
Move Test_cmd_backtick() to the right place.
Issuing a server request triggers `changetracking.flush` so as to
make sure we're not operating on a stale state. This immediately
triggers notification of any pending changes (as a result of debouncing)
to the server. However, this happens in addition to the notification
that is waiting on the debounce delay. Because we `nil`
`buf_state.pending_change` when it is called, the fix is to
also check that this is non-`nil` when it is called and exit if it is,
as this being `nil` would mean that it originates from a pending change
that has already been flushed out.
Problem: Arabic support excludes Farsi.
Solution: Add Farsi support to the Arabic support. (Ali Gholami Rudi,
Ameretat Reith)
dc4fa190e7
Omit Test_shape_final_to_medial(): removed in later patches.
Problem: Spellfile functionality not fully tested.
Solution: Add tests for SFX with removal of characters, spelling
suggestions with NOBREAK and others. (Dominique Pellé,
closesvim/vim#8293)
bb162367ac
Problem: Spellfile functionality not fully tested.
Solution: Add tests for CIRCUMFIX, NOBREAK and others. (Dominique Pellé,
closesvim/vim#8283)
5a6cfb3ff2
Reorder test_spellfile.vim to match upstream.
Problem: Cmdexpand.c insufficiently tested.
Solution: Add more tests. (Yegappan Lakshmanan, closesvim/vim#5789)
24ebd83e03
Map Q to gQ before every test since a test uses :mapclear.
Problem: Ex_getln.c code not covered by tests.
Solution: Add a few more tests. (Yegappan Lakshmanan, closesvim/vim#5702)
578fe947e3
Cherry-pick Test_Ex_global() from patch 8.2.0293.
Test_rightleftcmd() fails if incsearch is enabled, so disable it.
Problem
-------
In #19040, I reported two things that started happening somewhen in the
last three months when using neovim in hterm (the Chrome Secure Shell
terminal):
1. Under certain circumstances, the window title (set by nvim
[i0]) would appear over the line I was typing, corrupting the screen.
2. If I changed my $TERM from xterm-256color to the new hterm-256color
(available since ncurses >=20210320), the window title corruption
was gone, but pane scrolling was broken.
Both problems are due to changes in the termcap files, their source of
truth being the ncurses project. See "Timeline of ncurses changes" below
for details.
Cause: title corruption
-----------------------
The title corruption when using hterm + TERM=xterm-256color can be
explained by event #4 (ncurses 2022-03-12) in the ncurses timeline:
The xterm-256color termcap file gained status line termcodes in ncurses
2022-03-12. These termcodes are used by Neovim to set the title when.
hterm does not have a status line. Due to ncurses versions earlier than
2022-03-12 missing the xterm status line capability, Neovim manually
fixed up [t0] the terminfo file if $TERM was xterm-256color. So if
before Neovim manually added fsl/tsl capabilties, and after they were
in the termcap file, why did hterm suddenly start getting corruption?
The answer is that the termcodes for these capabilties are different
when Neovim fixes them up, versus the one in the new termcap database:
fsl=\E[0$} // from xterm-256color
tsl=\E[2$~\E[1$}\E[%i%p1%d` // from xterm-256color
fsl=\x07 // patched by Neovim
tsl=\x1b]0; // patched by Neovim
hterm ignores the latter, but corrupts the screen with the former.
Solution: Make hterm users set hterm-256color, which lacks the new
fsl/tsl codes. Also, to reduce superfluous work, stop patching in this
capability when hterm is detected (even if hterm would ignore the
patched version).
Cause: pane corruption
----------------------
The pane corruption when using hterm + TERM=hterm-256color, but NOT when
using hterm + TERM=xterm-256color can be explained by:
- Neovim uses DECSLRM when available [p1] for performant scrolling.
- Both the hterm-256color and xterm-256color termcap databases
advertise support for DECSLRM (ncurses timeline #1, #2 and #3).
- hterm does not support DESCLRM [p2] (note: it does support DESCTBM for
top/bottom scrolling, but it's broken [p3] and not used by Neovim)
- xterm-alikes that are not real xterm generally don't support DECSLRM
either, so Neovim patches it out [p4].
When using hterm-256color, hterm is no longer considered an xterm-alike
by Neovim. As a result, DECSLRM is not cleared. hterm does not support
it, so corruption ensues.
This is a problem with the hterm-256color termcap file, but we're stuck
with it so the best we can do is patch over it.
Timeline of ncurses changes
---------------------------
1. 2019-05-19: Part of the DECSLRM capability (smglr AKA set_lr_margin)
added to vt420+lrmm, which xterm-256color inherits [n1]
2. 2021-03-20: hterm-256color added, inheriting xterm-256colors. [n2]
3. 2021-09-25: The *parm versions of smglr (AKA set_lr_margin) were
added to vt420+lrmm [n3]. Namely:
1. smglp AKA set_left_margin_parm, and
2. smgrp AKA set_right_margin_parm
4. 2022-03-12: (new) codes for fsl, bsl and tsl added to xterm (add
dec+sl to xterm-new, per patch #371 -TD) [n4]
Fixes#19040.
[i0]: 3a4fa22bad/src/nvim/tui/tui.c (L1377)
[t0]: 3a4fa22bad/src/nvim/tui/tui.c (L1728),L1729
[p1]: 3a4fa22bad/src/nvim/tui/tui.c (L1196)
[p2]: https://bugs.chromium.org/p/chromium/issues/detail?id=1175065&q=component%3APlatform%3EApps%3EDefault%3EHterm
[p3]: https://bugs.chromium.org/p/chromium/issues/detail?id=1298796&q=component%3APlatform%3EApps%3EDefault%3EHterm
[p4]: 3a4fa22bad/src/nvim/tui/tui.c (L1740-L1752)
[n1]: https://github.com/mirror/ncurses/commit/8f6d94b8d6211a2323caef53fa4c96c475ec9a6#diff-01544c577762d3308a1d232aa7afc79acf64b9a5057f88a004df82fda89549b7R2742
[n2]: c2650100f8 (diff-01544c577762d3308a1d232aa7afc79acf64b9a5057f88a004df82fda89549b7R5907)
[n3]: f6b436c4fb (diff-01544c577762d3308a1d232aa7afc79acf64b9a5057f88a004df82fda89549b7R2842)
[n4]: 8bf8c836fe (diff-01544c577762d3308a1d232aa7afc79acf64b9a5057f88a004df82fda89549b7R4828)
Signed-off-by: Nicolas Hillegeer <nicolas@hillegeer.com>
This is a port of these two patches combined:
vim-patch:8.1.1524: tests are silently skipped
Problem: Tests are silently skipped.
Solution: Throw an exception for skipped tests in more places.
b0f94c1ff3
vim-patch:8.1.1544: some balloon tests don't run when they can
Problem: Some balloon tests don't run when they can.
Solution: Split GUI balloon tests off into a separate file. (Ozaki Kiichi,
closesvim/vim#4538) Change the feature check into a command for
consistency.
b46fecd345
Omit test_lua.vim: previous patches are N/A
Omit test_memory_usage.vim: previous patches are N/A
Omit test_textprop.vim: previous patches are N/A
Omit test_winbar.vim: previous patches are N/A
Omit test_termcodes.vim: mostly N/A
Skip Test_mouse_positon() because it uses test_setmouse().
Cannot throw error in a Test_nocatch_ test.
Use latest `CheckFeature clipboard_working` for test_quotestar.vim
* revert to filetype.vim by setting `g:do_legacy_filetype`
* skip either filetype.lua or filetype.vim via `g:did_load_filetypes`
(Running both is no longer required and therefore no longer supported.)
Problem: Command line test fails.
Solution: Also beep when cmdline win can't be opened because of locks.
Make the test not beep. Make the test pass on MS-Windows.
be99042b03
Problem: Can open a cmdline window from a substitute expression.
Solution: Disallow opening a command line window when text or buffer is
locked.
71223e2db8
Problem: Get ml_get error when deleting a line in 'completefunc'. (Yegappan
Lakshmanan)
Solution: Lock the text while evaluating 'completefunc'.
ff06f283e3
Fix a mistake in the porting of patch 8.1.0098.
Cherry-pick Test_run_excmd_with_text_locked() from patch 8.2.0270.
Cherry-pick test_gf.vim changes from patch 8.2.0369.
Cherry-pick message change from later patches.
vim-patch:9.0.0043: insufficient testing for bracket commands
Problem: Insufficient testing for bracket commands.
Solution: Add a few more tests. (closesvim/vim#10668)
cf34434b5e
Cherry-pick a change from patch 8.2.0369.
Add 'mousescroll' option to control how many lines to scroll by when a
mouse wheel keycode is received. The mousescroll option controls both
horizontal and vertical scrolling. The option is a string in the format:
set mousescroll=direction:count,direction:count
Where direction is either "ver" or "hor", and count is a non negative
integer. If a direction is omitted, a default value is used. The default
values remain unchanged, that is 3 for vertical scrolling, and 6 for
horizontal scrolling. As such, the mousescroll default is "ver:3,hor:6".
Add mousescroll documentation
- Add option documentation in options.txt
- Add brief summary in quickref.txt
Update :help scroll-mouse-wheel
- Mention mousescroll option as a means of controlling scrolling.
- Remove obsolete suggestion to map scroll wheel keys to <C-U> to
scroll by a single line -- users should prefer the mousescroll option.
- Add some information about the consequences of remapping scroll wheel
keys (they lose their magic ability to affect inactive windows).
Update :help vim-differences
- Add brief mousescroll summary under Options
Add mousescroll tests
- Test option validation
- Test default mousescroll value and behavior
- Test fallback to default values
- Test mouse vertical and horizontal scrolling in normal mode
- Test mouse vertical and horizontal scrolling in insert mode
Problem: The terminal debugger can't set a breakpoint with the mouse.
Solution: Add popup menu entries.
71137fed4d
This ports missing popup menu code to termdebug plugin.
Despite the commit message, the code is copied from latest Vim.
WinBar code is commented out and WinBar docs is not ported.
Problem: If 'operatorfunc' invokes an operator the remembered Visual mode
may be changed. (Naohiro Ono)
Solution: Save and restore the information for redoing the Visual area.
(closesvim/vim#9455)
b3bd1d39e6
Problem: Count for 'operatorfunc' in Visual mode is not redone.
Solution: Add the count to the redo buffer. (closesvim/vim#9174)
2228cd72cf
Cherry-pick a line from patch 8.2.0522.
Problem: "g$" causes scroll if half a double width char is visible.
Solution: Advance to the last fully visible character. (closesvim/vim#8254)
74ede80aeb
- Partially fixes#10386 except for the case where the alternate buffer
is the default, empty, first buffer created on startup. #vimlife
- TODO: port patches related to `can_unload_buffer`, maybe that fully
fixes#10386?
vim-patch:8.0.1732: crash when terminal API call deletes the buffer
Problem: When re-formatting with an indent expression the first line of a
paragraph may get the wrong indent. (Martin F. Krafft)
Solution: Apply the correct indenting function for the first line.
(Christian Brabandt, closesvim/vim#9150, closesvim/vim#9056)
818ff25cd1
Problem: Some Normal mode commands not tested.
Solution: Add more tests. (Yegappan Lakshmanan, closesvim/vim#5746)
f5f1e10d0d
Omit Test_edit_forbidden(): change reverted in patch 8.2.0369.
Cherry-pick Test_normal_yank_with_excmd() from patch 8.2.0293.
Skip Test_normal_cursorhold_with_count().
Cherry-pick Test_star_register() from patch 8.2.0270.
For a local mark without a view, currently trying to restore its view
will put the cursor at topline, which is not the correct behavior.
Initialize `topline_offset` to `MAXLNUM` instead to fix this.
* fix(test): screen.lua nil index
When actual_rows and expected_rows are different avoid a nil index.
* fix(tui): resize at startup
The deleted code is not needed after 402b4e8.
It caused the condition to false positive when the function was called
more than once before startup (first normal mode). Being itself what
set the dimension and not the user, locking the size of the screen to
an incorrect size.
Make got_winch an int to mitigate: tui_grid_resize changing the size
of the host terminal between the signal handler and the call to
sigwinch_cb. Since the actual signal handler uv__signal_handle doesn't
directly call the callback, the event loop does.
Fixes#17285Fixes#15044Fixes#11330
Problem: Spelldump test sometimes hangs.
Solution: Catch the problem of the spell file not being found to avoid
hanging in the download dialog.
fc9f0fd6d1
libuv does not call callback functions(uv_signal_cb) directly from signal
handlers. Therefore, there is no need to use volatile.
Ref. 1a91b51976/src/unix/signal.c (L183)
Problem: Invalid memory use when using SpellFileMissing autocmd.
Solution: Add test case. (Dominique Pellé, closesvim/vim#7036) Fix using a window
that was closed.
d569a9e746
Skip an assert because of #3027.
Problem: Still not enough test coverage for the spell file handling.
Solution: Add spell file tests. (Yegappan Lakshmanan, closesvim/vim#6790)
b9fc192f92
Problem: Not enough test coverage for the spell file handling.
Solution: Add spell file tests. (Yegappan Lakshmanan, closesvim/vim#6763)
c8ec5fe56f
Add missing whitespace in message.
vim-patch:9.0.0031: <mods> of user command does not have correct verbose value
Problem: <mods> of user command does not have correct verbose value.
Solution: Use the value from the command modifier. (closesvim/vim#10651)
9359e8a6d9
Problem: Matchfuzzy test depends on path of current directory.
Solution: Use fnamemodify() to remove the path. (Robin Becker,
closesvim/vim#10650)
22e7e867e2
Disable -Wsuggest compile flags by default.
Although the flags are useful to see what the compiler thinks we can
improve on, it adds a lot of noise to the build output. That makes it
harder to spot actual issues that may be flagged.
Closes https://github.com/neovim/neovim/issues/19160
In the rare case that exit_event is called from process_close_handles,
it stalls waiting for the process to exit (the routine is currently
underway to do just that). This causes `job_spec.lua` to sometimes
stall.
REJECTED IDEAS:
==============================================================
1. Currently `exit_event` is placed on `main_loop.fast_events`. Would the problem
be solved by using `main_loop.events` instead?
- A: Maybe, but it will cause other problems, such as queuing exit_event()
during "Press Enter..." prompt which may result in the event not being
processed, leading to another stall.
2. Can we avoid the timer?
- A: Using a timer is just the easiest way to queue a delayed event without
causing an infinite loop in the queue currently being processed.
3. Can we avoid the new `exit_need_delay` global...
1. by using `process_is_tearing_down` instead?
- A: Can't use `process_is_tearing_down` because its semantics are different.
2. by checking a similar condition as `process_teardown`? f50135a32e/src/nvim/event/process.c (L141-L142)
```
if (!process_is_tearing_down || (kl_empty(main_loop.children) && multiqueue_empty(main_loop.events))) {
uv_timer_start(&main_loop.exit_delay_timer, exit_delay_cb, 0, 0);
return;
}
```
- A: Tried but it did not work (other stalls occurred). Maybe
exit_event() is called from a source other than
process_close_handles() and is delayed, the delayed exit_event() will
be executed before main_loop.events is processed, resulting in an
infinite loop.
* refactor(fold): update comment formatting
* refactor(fold): use post-increment/decrement
* refactor(fold): reduce scope of local variables
* refactor(fold): use booleans
Problem: Some code not covered by tests.
Solution: Add test cases. (Yegappan Lakshmanan, closesvim/vim#5645)
f0cee1971f
Cherry-pick Test_bunload_with_offset() from patch 8.2.0243
Problem: Invalid memory access when adding word with a control character to
the internal spell word list.
Solution: Disallow adding a word with control characters or a trailing
slash.
5e59ea54c0
Problem: With EXITFREE defined terminal menus are not cleared.
Solution: Also clear terminal menus. Remove condition that is always true.
(closesvim/vim#10641)
79ae152697
Problem: Command line completion not fully tested.
Solution: Add more test cases. Make help sorting stable. (Dominique Pelle,
closesvim/vim#5402)
297610ba4b
Problem: Internal error when using :popup.
Solution: When a menu only exists in Terminal mode give an error. (Naruhiko
Nishino, closesvim/vim#3765)
f42b45d719
Problem: No menus specifically for the terminal window.
Solution: Add :tlmenu. (Yee Cheng Chin, closesvim/vim#3439) Add a menu test.
4c5d815256
ADDR_OHTER comes from patch 8.1.1241, which has already been ported.
Problem: Show cursor in wrong place when using popup menu. (Wei Zhang)
Solution: Force updating the cursor position. Fix skipping over unused
entries.
987723e084
Problem: Can't use :popup for a menu in the terminal. (Wei Zhang)
Solution: Make :popup work in the terminal. Also fix that entries were
included that don't work in the current state.
29a2c08d79
** Refactor
Previously most functions used to "get" a mark returned a position,
changed the line number and sometimes changed even the current buffer.
Now functions return a {x}fmark_T making calling context aware whether
the mark is in another buffer without arcane casting. A new function is
provided for switching to the mark buffer and returning a flag style
Enum to convey what happen in the movement. If the cursor changed, line,
columns, if it changed buffer, etc.
The function to get named mark was split into multiple functions.
- mark_get() -> fmark_T
- mark_get_global() -> xfmark_T
- mark_get_local() -> fmark_T
- mark_get_motion() -> fmark_T
- mark_get_visual() -> fmark_T
Functions that manage the changelist and jumplist were also modified to
return mark types.
- get_jumplist -> fmark_T
- get_changelist -> fmark_T
The refactor is also seen mainly on normal.c, where all the mark
movement has been siphoned through one function nv_gomark, while the
other functions handle getting the mark and setting their movement
flags. To handle whether context marks should be left, etc.
** Mark View
While doing the refactor the concept of a mark view was also
implemented:
The view of a mark currently implemented as the number of lines between
the mark position on creation and the window topline. This allows for
moving not only back to the position of a mark but having the window
look similar to when the mark was defined. This is done by carrying and
extra element in the fmark_T struct, which can be extended later to also
restore horizontal shift.
*** User space features
1. There's a new option, jumpoptions+=view enables the mark view restoring
automatically when using the jumplist, changelist, alternate-file and
mark motions. <C-O> <C-I> g; g, <C-^> '[mark] `[mark]
** Limitations
- The view information is not saved in shada.
- Calls to get_mark should copy the value in the pointer since we are
using pos_to_mark() to wrap and provide a homogeneous interfaces. This
was also a limitation in the previous state of things.
The general idea is that add_glob_targets creates a "touch file", a
dummy file that acts as a dependency in order to check which files are
outdated since the last time the target was run.
Remove RunUncrustify.cmake as it's no longer necessary. It was initially
introduced to silence its noisy output. The per-file targets will
suppress the noisy output from uncrustify, except for the very first
run.
Also remove DefCmdTarget.cmake since add_glob_target already
incorporates its functionality.
Problem:
- Since c57f6b28d7#8519, sockets are created in ~/.local/… but XDG
spec says: "XDG_RUNTIME_DIR: Must be on the local filesystem", which
implies that XDG_STATE_DIR is potentially non-local.
- Not easy to inspect Nvim-created temp files (for debugging etc).
Solution:
- Store sockets in stdpath('run') ($XDG_RUNTIME_DIR).
- Establish "/tmp/nvim.user/" as the tempdir root shared by all Nvims.
- Make ok() actually useful.
- Introduce assert_nolog().
closes#3517closes#17093
Ref: 84f5463630
Rename:
- `underlineline` to `underdouble`
- `underdot` to `underdotted`
- `underdash` to `underdashed`
`underdouble` also now takes higher precedence than `undercurl`.
This fixes the following bugs:
`${1:else_text}` -> format with if_text: "else_text"
`${1:-else_text}` -> format with if_text: "else_text"
`${1:}` in `format` (eg. empty else_text) -> error.
`${1:}` (eg. empty placeholder) -> error.
Thanks hrsh7th :)
Problem: Incsearch highlight broken when calling searchcount() in 'tabLine'
function. (Mirko Palmer)
Solution: Save and restore the incsearch state. (Christian Brabandt,
closesvim/vim#9763, closesvim/vim#9633)
6dd7424c7e
Our previous mangling of gettext broke the `HAVE_WORKING_LIBINTL` test
because it prevented CMake from finding `libintl.h`. Let's fix that by
linking Gettext's `include` directory into `/usr/local` too.
If `libintl` is a static library on macOS, we also need to explicitly
link with `libiconv` and the `CoreFoundation` framework. Otherwise, our
`HAVE_WORKING_LIBINTL` test erroneously fails.
Closes#19127Closes#19138
Problem:
Dirs "config", "packaging", and "third-party" are all closely related
but this is not obvious from the layout. This adds friction for new
contributors.
Solution:
- rename config/ to cmake.config/
- rename test/config/ to test/cmakeconfig/ because it is used in Lua
tests: require('test.cmakeconfig.paths').
- rename packaging/ to cmake.packaging/
- rename third-party/ to cmake.deps/ (parallel with .deps/)
This is necessary in cases where filetype detection acts recursively.
For example, when matching files that end with .bak, the "root" of
the filename is matched again against the same buffer (e.g. a buffer
named "foo.c.bak" will be matched again with the filename "foo.c", using
the same underlying buffer).
Omitting 'pattern' in nvim_exec_autocmds should be equivalent to
omitting the 'fname' argument in :doautoall, which is equivalent to
using an empty string as the pattern. Fixes regression introduced
in #19091.
Problem: Crash when going through spell suggestions.
Solution: Limit the text length for finding suggestions to the original
length. Do not update buffers when exiting. (closesvim/vim#8965)
e275ba4fc9
This enables vim.filetype.match to match based on a buffer (most
accurate) or simply a filename or file contents, which are less accurate
but may still be useful for some scenarios.
When matching based on a buffer, the buffer's name and contents are both
used to do full filetype matching. When using a filename, if the file
exists the file is loaded into a buffer and full filetype detection is
performed. If the file does not exist then filetype matching is only
performed against the filename itself. Content-based matching does the
equivalent of scripts.vim, and matches solely based on file contents
without any information from the name of the file itself (e.g. for
shebangs).
BREAKING CHANGE: use `vim.filetype.match({buf = bufnr})` instead
of `vim.filetype.match(name, bufnr)`
With 4c7462fdb3, when `libintl` is turned off via
`-DENABLE_LIBINTL=OFF`, Neovim does not compile in Debug mode with the following
error message:
```
...
[201/331] Building C object src/nvim/CMakeFiles/nvim.dir/eval.c.o
FAILED: src/nvim/CMakeFiles/nvim.dir/eval.c.o
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/cc -DEXITFREE -DINCLUDE_GENERATED_DECLARATIONS -DMIN_LOG_LEVEL=1 -DNVIM_MSGPACK_HAS_FLOAT32 -DNVIM_UNIBI_HAS_VAR_FROM -Dnvim_EXPORTS -I/Users/hat/Projects/neovim/build/config -I/Users/hat/Projects/neovim/src -I/Users/hat/Projects/neovim/build/src/nvim/auto -I/Users/hat/Projects/neovim/build/include -I/Users/hat/Projects/neovim/.deps/usr/include/luajit-2.1 -isystem /Users/hat/Projects/neovim/.deps/usr/include -U_FORTIFY_SOURCE -D_FORTIFY_SOURCE=1 -DNVIM_TS_HAS_SET_MATCH_LIMIT -DNVIM_TS_HAS_SET_ALLOCATOR -g -arch arm64 -isysroot /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX12.3.sdk -Wall -Wextra -pedantic -Wno-unused-parameter -Wstrict-prototypes -std=gnu99 -Wshadow -Wconversion -Wdouble-promotion -Wmissing-noreturn -Wmissing-format-attribute -Wmissing-prototypes -Wimplicit-fallthrough -Wvla -fstack-protector-strong -fno-common -fdiagnostics-color=always -Wno-conversion -MD -MT src/nvim/CMakeFiles/nvim.dir/eval.c.o -MF src/nvim/CMakeFiles/nvim.dir/eval.c.o.d -o src/nvim/CMakeFiles/nvim.dir/eval.c.o -c /Users/hat/Projects/neovim/src/nvim/eval.c
/Users/hat/Projects/neovim/src/nvim/eval.c:473:3: error: implicit declaration of function 'free_locales' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
free_locales();
^
/Users/hat/Projects/neovim/src/nvim/eval.c:473:3: note: did you mean 'free_titles'?
/Users/hat/Projects/neovim/build/include/buffer.h.generated.h:58:16: note: 'free_titles' declared here
DLLEXPORT void free_titles(void);
^
1 error generated.
[212/331] Building C object src/nvim/CMakeFiles/nvim.dir/eval/funcs.c.o
ninja: build stopped: subcommand failed.
make: *** [nvim] Error 1
```
The previous commit, b1b5802009, to 4c7462f does
not have this issue. As you can see in the following, the issue occurs only in
Debug mode with `libintl` turned off:
```bash
# No issue
git checkout b1b5802
# The following Debug build works
make distclean; make SDKROOT=$(xcrun --sdk macosx --show-sdk-path) CMAKE_EXTRA_FLAGS="-DENABLE_LIBINTL=OFF"
# The following Release build works
make distclean; make CMAKE_BUILD_TYPE=Release SDKROOT=$(xcrun --sdk macosx --show-sdk-path) CMAKE_EXTRA_FLAGS="-DENABLE_LIBINTL=OFF"
# Compile error when in Debug and libintl turned off
git checkout 4c7462f
# The following Debug build does not work
make distclean; make SDKROOT=$(xcrun --sdk macosx --show-sdk-path) CMAKE_EXTRA_FLAGS="-DENABLE_LIBINTL=OFF"
# The following Debug build works
make distclean; make SDKROOT=$(xcrun --sdk macosx --show-sdk-path)
# The following Release build works
make distclean; make CMAKE_BUILD_TYPE=Release SDKROOT=$(xcrun --sdk macosx --show-sdk-path) CMAKE_EXTRA_FLAGS="-DENABLE_LIBINTL=OFF"
```
I noticed this because I turn `libintl` off when compiling `libnvim` for VimR.
Fix#19093
Problem:
Nvim fails to create tempfile "…/nvim6UJx04/7" when 'shell' is set to
pwsh (PowerShell Core). This breaks filtered shell commands ":{range}!".
With shell set to cmd, it works.
Solution:
PowerShell doesn't use "<" for stdin redirection. Instead, use
"-RedirectStandardInput".
Closes#15913
Problem: #19029 added a new fallback that sets
`$MACOSX_DEPLOYMENT_TARGET` to the local macOS version via `sw_vers`.
However, the output included a newline, which broke the generated Ninja
build script.
Solution: use `OUTPUT_STRIP_TRAILING_WHITESPACE` for `execute_process`.
Problem: 'completefunc'/'omnifunc' error does not end completion.
Solution: Check if there was an error or exception. (closesvim/vim#10486,
closesvim/vim#4218)
9bcb9ca9c7
This re-introduces the fix that the filetype.lua refactor inadvertently reverted.
The fix ensures that in the case when end_lnum is omitted, a string is always returned.
After some tweaks to our dep builds, we can now build a universal binary
for macOS by using `CMAKE_OSX_ARCHITECTURES`. So, let's do that. This
requires a number of additional changes:
1. We need to build on macOS 11, since earlier versions do not support
building universal (M1 + Intel) binaries.
2. We need to provision a universal `libintl`. The linker will look for
an ARM64 version of this library when linking the `nvim` binary.
While we're here:
1. Link statically to `libintl`. This allows to to avoid having to do
any install name rewriting or codesigning to package Neovim.
2. Bump the `MACOSX_DEPLOYMENT_TARGET` to `11`. We're already using a
`libintl` built by Homebrew (through the pre-installed version of
`gettext`), and that is built for macOS 11.
In order to ensure we link to `libintl.a` instead of `libintl.dylib`, we
have to make sure that CMake can't find the latter. This ideally should
be a matter of doing `brew unlink gettext`. However, CMake is too adept
at finding things that Homebrew has installed (even when not linked), so
we have to do a bit more than that. This appears in the additional step
ensuring static linkage to `libintl`.
We end up breaking some Homebrew-installed software in the process, and
some of these software is called during our build (e.g. curl, git,
wget). To avoid any adverse effects, let's just uninstall them.
CMake can handle building universal binaries on macOS using the
`CMAKE_OSX_ARCHITECTURES` variable. Let's pass this variable to the
relevant dep builds. We use the `LIST_SEPARATOR` argument to prevent the
shell from interpreting the `;` that CMake uses as a list separator.
For dependencies that don't build using CMake, we only need to make sure
that the compiler is invoked with the correct `-arch` flags. The
compiler does the rest.
The only exception to this is the LuaJIT build, which we handle
separately as a special case.
To build universal binaries on macOS, one typically only needs to pass
multiple `-arch` flags to `clang`.
Unfortunately, this strategy causes LuaJIT builds to fail. To work
around this, we build LuaJIT for each requested architecture
individually first and then use `lipo` to package each architecture
slice into a universal binary.
To be able to do this on an Intel macOS host or an M1 macOS host without
a Rosetta installation, we need some special flags in order to tell
LuaJIT that it is cross-compiling for a different target. See [1] for
details.
[1] https://luajit.org/install.html#cross
- don't immediately vim_strsave and then xfree a heap buffer.
- allow replace_termcodes to take in a buffer instead of allocating it
- grug! memory allocation bad!
Problem: Memory leak when substitute expression nests.
Solution: Use an array of expression results.
44ddf19ec0
Cherry-pick a comment change from patch 8.2.5057.
N/A patches for version.c:
vim-patch:8.2.5154: still mentioning version8, some cosmetic issues
Problem: Still mentioning version8, some cosmetic issues.
Solution: Prefer mentioning version9, cosmetic improvements.
abd56da30b
Problem: Memory access error when substitute expression changes window.
Solution: Disallow changing window in substitute expression.
e2bd8600b8
"textwinlock" was renamed back to "textlock" in patch 8.2.5029.
Problem:
RunTests.cmake adds $TEST_PATH to $TMPDIR with the implication that it
gives more isolation. But this is misleading because $TEST_PATH is only
defined once. Full test runs use the same $TMPDIR for all tests.
This was likely added with the intention of invoking RunTests.cmake
once-per-testfile from a wrapper than does the isolation/orchestration.
But even so, Nvim's vim_maketempdir() / vim_mktempdir() _already_
creates a unique tempdir per session.
Solution:
Don't append $TEST_PATH to $TMPDIR. Avoids confusion and makes the path
shorter.
This marks the following Vim patches as ported:
vim-patch:8.1.1785: map functionality mixed with character input
Problem: Map functionality mixed with character input.
Solution: Move the map functionality to a separate file. (Yegappan
Lakshmanan, closesvim/vim#4740) Graduate the +localmap feature.
b66bab381c
vim-patch:8.2.3643: header for source file is outdated
Problem: Header for source file is outdated.
Solution: Make the header more accurate. (closesvim/vim#9186)
a3f83feb63
Also cherry-pick a change for <unique> mappings from patch 8.2.0807.
Rename map_clear_mode() to do_mapclear().
Problem: Transparent syntax item uses start/end of containing region.
Solution: Do not change the startpos and endpos of a transparent region to
that of its containing region. (Adrian Ghizaru, closesvim/vim#7349,
closesvim/vim#7391)
b46f57e87b
Avoids this warning in the Windows build:
2022-06-22T08:58:13.0542153Z LINK : warning LNK4044: unrecognized option '/municode'; ignored [D:\a\neovim\neovim\build\test\functional\fixtures\printenv-test.vcxproj]
global-local window options need to be handled specially. When `win` is
given but `scope` is not, then we want to set the local version of the
option but not the global one, therefore we need to force
`scope='local'`.
Note this does not apply to window-local only options (e.g. 'number')
Example:
nvim_set_option_value('scrolloff', 10, {}) -- global-local window option; set global value
nvim_set_option_value('scrolloff', 20, {win=0}) -- global-local window option; set local value
nvim_set_option_value('number', true, {}) -- local window option
is now equivalent to:
nvim_set_option_value('scrolloff', 10, {})
nvim_set_option_value('scrolloff', 20, {win=0, scope='local'}) -- changed from before
nvim_set_option_value('number', true, {win=0}) -- unchanged from before
Only the global-local option with a `win` provided gets forced to local
scope.
The main motivation for this is for the buf and win cases which need to
set up and restore context, and it's what specifically makes the
semantics of options nuanced, and thus this should not be repeated more
than once.
- nvim_get/set_option_value now share argument validation.
This will allow cmake to check if any directories needs re-globbing, in
other words, if any new file has been added since last time cmake was
run. This will (allegedly) make the configure stage slower but I have
not noticed any difference so I believe this is well worth it.
This is a sham. if the user does
hi ExistingGroup guifg=AliceBlue invalidkey=foobar
the "guifg" part will still be executed. No need to micro-manage
the same case where ANewGroup is added instead.
`nvim_get_option_value` and `nvim_set_option_value` better handle
unsetting local options. For instance, this is currently not possible:
vim.bo.tagfunc = nil
This does not work because 'tagfunc' is marked as "local to buffer" and
does not have a fallback global option. However, using :setlocal *does*
work as expected
:setlocal tagfunc=
`nvim_set_option_value` behaves more like :set and :setlocal (by
design), so using these as the underlying API functions beneath vim.bo
and vim.wo makes those two tables act more like :setlocal. Note that
vim.o *already* uses `nvim_set_option_value` under the hood, so that
vim.o behaves like :set.
Problem: Various small issues.
Solution: Various small improvments.
8088ae95bb
N/A patches for version.c:
vim-patch:8.2.5107: some callers of rettv_list_alloc() check for not OK
Problem: Some callers of rettv_list_alloc() check for not OK. (Christ van
Willegen)
Solution: Use "==" instead of "!=" when checking the return value.
93a1096fe4
Note for external UIs: Nvim can now emit multiple "redraw" event batches
before a final "flush" event is received. To retain existing behavior,
clients should make sure to update visible state at an explicit "flush"
event, not just the end of a "redraw" batch of event.
* Get rid of copy_object() blizzard in the auto-generated ui_event layer
* Special case "grid_line" by encoding screen state directly to
msgpack events with no intermediate API events.
* Get rid of the arcane notion of referring to the screen as the "shell"
* Array and Dictionary are kvec_t:s, so define them as such.
* Allow kvec_t:s, such as Arrays and Dictionaries, to be allocated with
a predetermined size within an arena.
* Eliminate redundant capacity checking when filling such kvec_t:s
with values.
FindLua.cmake is a copy from the cmake repo:
0419ecbcad/Modules/FindLua.cmake.
It's a cmake module, meaning it's already shipped with cmake by default.
There have been two changes done to our version of FindLua.cmake.
The first change is that
include(${CMAKE_CURRENT_LIST_DIR}/FindPackageHandleStandardArgs.cmake)
was changed to
include(FindPackageHandleStandardArgs.cmake)
This change is required only because we have imported FindLua.cmake
module but not FindPackageHandleStandardArgs module. Had FindLua been
called as a module as intended then this file would not need changing.
The second change is that support for Lua 5.4 is added. However, support
for any version of Lua except for 5.1 is disabled since
e322b5c864.
Because these changes from the upstream FindLua.cmake is unnecessary I
believe we can and should use the builtin FindLua.cmake instead of our
own.
Instead of appending to a command output, append to an existing target
instead. The primary benefit is intermediary ...-cmd targets aren't
needed, we can instead append commands to the relevant target directly.
More specifically, replace exec_program with file(REMOVE ...) so that
the uninstall target is run during the build stage instead of the
configure stage, significantly speeding up the target.
The code snippet that was removed is taken from the cmake FAQ
https://gitlab.kitware.com/cmake/community/-/wikis/FAQ#can-i-do-make-uninstall-with-cmake.
However, this uses undocumented features such as IMMEDIATE when calling
configure_file, which is an artifact from cmake 2.x (it's so old it's
difficult to find information on it). Similarly, this particular code
snippet has been around for a long time and originated from the cmake
mailing lists. Based on this I believe the in-file was a workaround for
the limitations of cmake back then and that it's not required anymore.
Problem:
Piping NodeJS output into Neovim makes the editor unusable.
This happens because NodeJS changes the tty state on exit after
Nvim calls uv_tty_set_mode(). (May not always happen due to race
condition.)
This should have been fixed by 4ba5b4a864#13084. But some
commands and functions (:sleep, system(), …) call ui_flush()
internally, in particular the first tui_mode_change() is called before
the end of startup.
Steps to reproduce:
1. node -e "setTimeout(()=>{console.log('test')}, 1000)" | nvim -u NORC +"sleep 500m" -
2. The cursor key letters just overwrite the editor screen, and CTRL+C exits.
Solution:
Skip pending_mode_update during startup.
Note: Delaying ui_flush() entirely could be a more general solution
(emit a new UI event on VimEnter?). But "remote/coprocess TUI" #18375
could make all of this moot anyway.
Fixes#18470
Problem:
The release script bundles a system library (CoreServices) that was
added in #18294, which leads to errors on M1 since the architecture is
different from the Github runner.
Solution:
Skip CoreServices when bundling the libraries (as was done for the
CoreFoundation library that #18294 replaced with CoreServices).
Problem:
1. CI logs have too many (40+) logs mentioning SIGHUP:
```
WRN 2022-06-18T16:05:47.075 T3568.22499.0/c deadly_signal:177: got signal 1 (SIGHUP)
WRN 2022-06-18T16:05:47.273 T3569.91095.0/c deadly_signal:177: got signal 1 (SIGHUP)
WRN 2022-06-18T16:05:47.651 T3570.59545.0/c deadly_signal:177: got signal 1 (SIGHUP)
```
2. TS parser test still sometimes fails on BSD CI.
3. remote_spec test fails too often.
Solution:
1. Log deadly signals at INFO level. It hasn't been helpful in CI, and
for local troubleshooting it's reasonable to adjust the loglevel as
needed.
2. Adjust the TS parser test again. ref #18911
3. Skip the remote_spec test. The `--remote` feature was merged before
it was fully formed and needs to be revisited.
luacheck and clint are the most important, and they are bundled.
The others are nice-to-have, and not always available (and not bundled),
so make them optional. This allows the "lint" target to function as
a convenient and low-hassle handle for contributors.
We still get the full power of lint in CI.
TODO: bundle uncrustify (and then set it as required for "lint")?
Problem:
Coverity reports use after free:
*** CID 352784: Memory - illegal accesses (USE_AFTER_FREE)
/src/nvim/buffer.c: 1508 in set_curbuf()
1502 if (old_tw != curbuf->b_p_tw) {
1503 check_colorcolumn(curwin);
1504 }
1505 }
1506
1507 if (bufref_valid(&prevbufref) && prevbuf->terminal != NULL) {
>>> CID 352784: Memory - illegal accesses (USE_AFTER_FREE)
>>> Calling "terminal_check_size" dereferences freed pointer "prevbuf->terminal".
1508 terminal_check_size(prevbuf->terminal);
1509 }
1510 }
1511
1512 /// Enter a new current buffer.
1513 /// Old curbuf must have been abandoned already! This also means "curbuf" may
Solution:
Change terminal_destroy and terminal_close to set caller storage to NULL,
similar to XFREE_CLEAR. This aligns with the pattern found already in:
terminal_destroy e897ccad3e
term_delayed_free 3e59c1e20d
Problem:
If startup finishes (starting=false) before the logger ever happens to
see a v:servername, we're stuck with the "?.<PID>" fallback name
forever.
Solution:
Drop the `starting` condition. Discard the "?.<PID>" fallback after
using it for the current log message. So logging will always check
v:servername next time.
If `use_rgb` was false, we would attempt to set the `cterm_bg_color` to the
variable `bg`, which is only retrieved from `bg` and `background` keys, not
`ctermbg`. Same for `fg`. This means the values would be `-1` (the default,
un-got value) and we'd always set the returned values to `0`.
My understanding is `fg/bg` is always "gui" values, so instead we should be
using `ctermbg` when needed.
Nb: when looking around I think this function is currently *always* called with
`use_rgb = true`.
Fixes#18980
- 831fa45ad8 is related but this doesn't regress that
- The `cterm_normal_fg_color != ae.cterm_fg_color` comparison is originally
carried from patch to patch starting all the way back in 29bc6dfabd where it
was avoiding setting a HL attr. But `hlattrs2dict()` now is just
informational.
The current emulation script enters Insert mode much too frequently.
Using only BufWinEnter seems to be a closer simulation.
Also add a few more mappings.
Steps to reproduce:
1. setting `vim.highlight.on_yank`
```
vim.api.nvim_create_autocmd({ "TextYankPost" }, {
pattern = { "*" },
callback = function()
vim.highlight.on_yank({ timeout = 200 })
end,
})
```
2. repeat typing `yeye` ...
3. causes the following error.
```
Error executing vim.schedule lua callback: vim/_editor.lua:0: handle 0x01e96970 is already closing
stack traceback:
[C]: in function 'close'
vim/_editor.lua: in function ''
vim/_editor.lua: in function <vim/_editor.lua:0>
```
📝 Test result before fix:
[----------] Global test environment setup.
[----------] Running tests from test/functional/lua/highlight_spec.lua
[ RUN ] vim.highlight.on_yank does not show errors even if buffer is wiped before timeout: 15.07 ms OK
[ RUN ] vim.highlight.on_yank does not show errors even if executed between timeout and clearing highlight: 15.07 ms ERR
test/helpers.lua:73: Expected objects to be the same.
Passed in:
(string) 'Error executing vim.schedule lua callback: vim/_editor.lua:0: handle 0x02025260 is already closing
stack traceback:
[C]: in function 'close'
vim/_editor.lua: in function ''
vim/_editor.lua: in function <vim/_editor.lua:0>'
Expected:
(string) ''
Problem:
- Unix sockets are created in random /tmp dirs.
- /tmp is messy, unclear when OSes actually clear it.
- The generated paths are very ugly. This adds friction to reasoning
about which paths belong to which Nvim instances.
- No way to provide a human-friendly way to identify Nvim instances in
logs or server addresses.
Solution:
- Store unix sockets in stdpath('state')
- Allow --listen "name" and serverstart("name") to given a name (which
is appended to a generated path).
TODO:
- is stdpath(state) the right place?
Problem:
1. Log messages (especially in CI) are hard to correlate with tests.
2. Since b353a5c05f#11886, dumplog() prints the logs next to test
failures. This is noisy and gets in the way of the test results.
Solution:
1. Associate an incrementing id with each test and include it in log
messages.
- FUTURE: add v:name so Nvim instances can be formally "named"?
2. Mention "child" in log messages if the current Nvim is a child (based
on the presence of $NVIM).
BEFORE:
DBG … 12345 UI: event
DBG … 12345 log_server_msg:722: RPC ->ch 1: …
DBG … 12345 UI: flush
DBG … 12345 inbuf_poll:444: blocking... events_enabled=1 events_pending=0
DBG … 23454 UI: stop
INF … 23454 os_exit:594: Nvim exit: 0
AFTER:
DBG … T57 UI: event
DBG … T57 log_server_msg:722: RPC ->ch 1: …
DBG … T57 UI: flush
DBG … T57 inbuf_poll:444: blocking... events_enabled=1 events_pending=0
DBG … T57/child UI: stop
INF … T57/child os_exit:594: Nvim exit: 0
Problem: Value of cmod_verbose is a bit complicated to use.
Solution: Use zero for not set, value + 1 when set. (closesvim/vim#10564)
cd7496382e
Omit has_cmdmod(): only used for Vim9 script
Problem: Command modifiers are saved and set inconsistently.
Solution: Separate parsing and applying command modifiers. Save values in
cmdmod_T.
5661ed6c83
Cherry-pick: :0verbose fix from patch 8.2.4741
drawback: tracing memory errors with ASAN is less accurate for arena
allocated memory.
Therefore, to start with it is being used for Object types around
serialization/deserialization exclusively. This is going to have
a large impact especially when TUI is refactored as a co-prosess
as all UI events will be serialized and deserialized by nvim itself.
Problem: Not all modifiers supported for :options.
Solution: Use all cmdmod.split flags. (closesvim/vim#4401)
7a1637f4c0
Cherry-pick Test_options_command() change from patch 8.2.0540
Problem: DirChanged autocommand may use freed memory. (Shane-XB Qian)
Solution: Free the memory later. (closesvim/vim#10555)
d8c9d32c89
Code change is N/A as Nvim gets the full current directory again before
applying the autocommand, so this just ports the tests.
Problem:
Since 6d57bb89c1#18543, luacheck is not found on some systems when
running the "lintlua" target.
Solution:
- Move the find_program() to the top-level CMakeLists.txt
- Define a def_cmd_target() function with fewer assumptions than the old
lint() function.
- Move "lintuncrustify" to src/nvim/CMakeLists.txt so it can reuse the
$LINT_NVIM_SOURCES already defined there.
- Make the lint targets _fatal_ by default. There is little reason for
the "lint" umbrella target defined in Makefile to exist if it's going
to ignore the absence of the actual linters.
- For now, keep the uncrustify call in a separate cmake script so that
it can be silenced (too noisy).
There can be other places that access window buffer info (e.g.
`tabpagebuflist()`), so checking `w_closing` in `win_findbuf()` doesn't
solve the crash in all cases, and may also cause Nvim's behavior to
diverge from Vim.
Fix#14998
Command preview now behaves like inccommand=nosplit when there's not
enough room for the preview window to be opened instead of aborting,
which is consistent with old behavior of 'inccommand'.
The size of long varies depending on architecture, in contrast to the
MAXLNUM constant which sets the maximum allowable number of lines to
2^32-1. This discrepancy may lead to hard to detect bugs, for example
https://github.com/neovim/neovim/issues/18454. Setting linenr_T to a
fix maximum size of 2^32-1 will prevent this type of errors in the
future.
Also change the variables `amount` and `amount_after` to be linenr_T
since they're referring to "the line number difference" between two
texts.
Many filetypes from filetype.vim set buffer-local variables, meaning
vim.filetype.match cannot be used without side effects. Instead of
setting these buffer-local variables in the filetype detection functions
themselves, have vim.filetype.match return an optional function value
that, when called, sets these variables. This allows vim.filetype.match
to work without side effects.
* build: move the logic for linters to cmake
Cmake is our source of truth. We should have as much of our build
process there as possible so everyone can make use of it.
* build: remove redundant check for ninja generator
The minimum cmake version as of writing this is 3.10, which has ninja
support.
The "first run" has high variability. Looks like the test failures
correlate with 545dc82c1b
, which makes sense because that improves "first run" performance.
So the `1000*` factor of this test could be adjusted to e.g. `300*` or `500*`.
ref https://github.com/neovim/neovim/pull/16945
fixes#15221
I tried to reproduce with a test, but failed. The below patch is able to
cause the out of bound access (I verified by adding a check to the
code), but it doesn't seg fault or trigger asan/valgrind errors.
```
diff --git a/src/nvim/testdir/test_tagfunc.vim b/src/nvim/testdir/test_tagfunc.vim
index ffc1d63b9..22828a39f 100644
--- a/src/nvim/testdir/test_tagfunc.vim
+++ b/src/nvim/testdir/test_tagfunc.vim
@@ -117,4 +117,26 @@ func Test_tagfunc_settagstack()
delfunc Mytagfunc2
endfunc
+func Test_tagfunc_settagstack_many()
+
+ func Mytagfunc1(pat, flags, info)
+ return [{'name' : 'mytag', 'filename' : 'Xtest', 'cmd' : '1'}]
+ endfunc
+ set tagfunc=Mytagfunc1
+ call writefile([''], 'Xtest')
+
+ for i in range(0,20)
+ let pos = [bufnr()] + getcurpos()[1:]
+ let newtag = [{'tagname' : 'mytag' + i, 'from' : pos}]
+ call settagstack(1, {'items' : newtag}, 'a')
+ call settagstack(1, {'curidx' : 21})
+ endfor
+
+ tag
+
+ call delete('Xtest')
+ set tagfunc&
+ delfunc Mytagfunc1
+endfunc
```
* on_scrollback_option_changed renamed to adjust_scrollback. The
function name did not correspond to what it was doing. It is
called unconditionally in every refresh of the terminal
unrelated if the scrollback option was changed.
* new on_scrollback_option_changed function, which calls
refresh_terminal, which then calls adjust_scrollback
* terminal_check_size is not the appropriate function to call when the
option is changed since it only conditionally adjusts the scrollback.
Use the new on_scrollback_option_changed
fixes#15477fixes#11811
Problem:
Since 22b52dd462#11501, log_path_init is called in log_init, so it is
now called at a deterministic time. So the "just in time" complexity of
log_path_init is no longer needed.
Solution:
Remove logic intended to try to "heal" partial initialization.
Problem: input() does not handle composing characters properly.
Solution: Use mb_cptr2char_adv() instead of mb_ptr2char_adv().
(closesvim/vim#10527)
e3a529bc87
Cherry-pick all of Test_input_func() from patch 8.2.0316.
Most LSP servers require the notification to correctly load the
settings and for those who don't it doesn't cause any harm.
So far this is done in lspconfig, but with the addition of vim.lsp.start
it should be part of core.
> DEPRECATION: Python 2.7 reached the end of its life on January 1st, 2020
Testing `pynvim` compatibility with python2 should not be done in core,
and there's only a provider_spec for python3 either way.
A alternative/subset of https://github.com/neovim/neovim/pull/18506 that should be forward compatible with a potential project system.
Configuration of LSP clients (without lspconfig) now looks like this:
vim.lsp.start({
name = 'my-server-name',
cmd = {'name-of-language-server-executable'},
root_dir = vim.fs.dirname(vim.fs.find({'setup.py', 'pyproject.toml'}, { upward = true })[1]),
})
problem: data race when `recursive` is read outside of mutex by thread A
while thread B has taken the mutex and modifies it.
solution: use a recursive lock.
ref #18764
Currently this is more or less a straight off reimplementation,
but this allow further optimizations down the line, especially
for avoiding memory allocations of rpc objects.
Current score for "make functionaltest; make oldtest" on a -DEXITFREE build:
is 117 055 352 xfree(ptr != NULL) calls (that's NUMBERWANG!).
When yanking another range while previous yank is still highlighted, the
pending timer could clear the highlight almost immediately (especially
when using larger `timeout`, i.e. 2000)
clang: Result of operation is garbage or undefined
clang: Uninitialized argument value
Also check for == 's' instead of != 'n' as it is more straightforward.
Also fix another unrelated PVS warning:
PVS/V1071: Return value of win_comp_pos() is not always used
Fix a bug in lsp.buf.rename() where the range returned by the server in
textDocument/prepareRename was interpreted as a byte range directly,
instead of taking the negotiated offset encoding into account. This
caused the placeholder value in vim.ui.input to be incorrect in some
cases, for example when non-ascii characters are used earlier on the
same line.
Problem:
https://github.com/neovim/neovim/pull/18720#issuecomment-1142614996
The vim.health module is detected as a healthcheck, which produces spurious errors:
vim: require("vim.health").check()
========================================================================
- ERROR: Failed to run healthcheck for "vim" plugin. Exception:
function health#check, line 20
Vim(eval):E5108: Error executing lua [string "luaeval()"]:1: attempt to call field 'check' (a nil value)
stack traceback:
[string "luaeval()"]:1: in main chunk
Solution:
Skip vim.health when discovering healthchecks.
Also avoid referring to mappings as "keymaps" in commands and docs.
*map_empty_rhs* *map-empty-rhs*
You can create an empty {rhs} by typing nothing after a single CTRL-V (you
have to type CTRL-V two times). Unfortunately, you cannot do this in a vimrc
file.
- Complete function:
There was lots of unnecessary C code for the complete function, therefore
moving it to Lua and use all the plumbing we have in place to retrieve the
results.
- Moving the module:
It's important we keep nvim lua modules name spaced, avoids conflict with
plugins, luarocks, etc.
Currently the `title`, `message` and `percentage` is stored for a
progress, but there is also an optional `cancellable` that comes in with
both the `WorkDoneProgressBegin` and also `WorkDoneProgressReport`. This
change also stores that value so that a plugin can access it when they
do a lookup in `client.messages`.
Problem:
1. The main log routine does not protect itself against recursion.
log_lock() doesn't guard against recursion, it would deadlock...
2. 22b52dd462 (#11501) regressed 6f27f5ef91 (#10172), because
set_init_1..process_spawn tries to log (see backtrace below), but the
mutex isn't initialized yet. Even if the mutex were valid, we don't
want early logging to fallback to stderr because that can break
embedders when stdio is used for RPC.
frame 1: 0x00000001001d54f4 nvim`open_log_file at log.c:205:7
frame 2: 0x00000001001d5390 nvim`logmsg(log_level=1, context="UI: ", func_name=0x0000000000000000, line_num=-1, eol=true, fmt="win_viewport") at log.c:150:20
frame : 0x000000010039aea2 nvim`ui_call_win_viewport(grid=2, win=1000, topline=0, botline=1, curline=0, curcol=0, line_count=1) at ui_events_call.generated.h:321:3
frame 4: 0x00000001003dfefc nvim`ui_ext_win_viewport(wp=0x0000000101816400) at window.c:939:5
frame 5: 0x00000001003ec5b4 nvim`win_ui_flush at window.c:7303:7
frame 6: 0x00000001003a04c0 nvim`ui_flush at ui.c:508:3
frame 7: 0x00000001002966ba nvim`do_os_system(argv=0x0000600000c0c000, input=0x0000000000000000, len=0, output=0x0000000000000000, nread=0x00007ff7bfefe830, silent=false, forward_output=false) at shell.c:894:3
frame 8: 0x0000000100295f68 nvim`os_call_shell(cmd="unset nonomatch; vimglob() { while [ $# -ge 1 ]; do echo \"$1\"; shift; done }; vimglob >/var/folders/gk/3tttv_md06987tlwpyp62jrw0000gn/T/nvimwwvwfD/0 ~foo", opts=kShellOptExpand | kShellOptSilent | kShellOptHideMess, extra_args=0x0000000000000000) at shell.c:663:18
frame 9: 0x0000000100295845 nvim`call_shell(cmd="unset nonomatch; vimglob() { while [ $# -ge 1 ]; do echo \"$1\"; shift; done }; vimglob >/var/folders/gk/3tttv_md06987tlwpyp62jrw0000gn/T/nvimwwvwfD/0 ~foo", opts=kShellOptExpand | kShellOptSilent | kShellOptHideMess, extra_shell_arg=0x0000000000000000) at shell.c:712:14
frame 10: 0x0000000100294c6f nvim`os_expand_wildcards(num_pat=1, pat=0x00007ff7bfefeb20, num_file=0x00007ff7bfefee58, file=0x00007ff7bfefee60, flags=43) at shell.c:328:7
...
frame 23: 0x000000010028ccef nvim`expand_env_esc(srcp=",~foo", dst="~foo", dstlen=4094, esc=false, one=false, prefix=0x0000000000000000) at env.c:673:17
frame 24: 0x000000010026fdd5 nvim`option_expand(opt_idx=29, val=",~foo") at option.c:1950:3
frame 25: 0x000000010026f129 nvim`set_init_1(clean_arg=false) at option.c:558:19
frame 26: 0x00000001001ea25e nvim`early_init(paramp=0x00007ff7bfeff5f0) at main.c:198:3
frame 27: 0x00000001001ea6bf nvim`main(argc=1, argv=0x00007ff7bfeff848) at main.c:255:3
Solution:
1. Check for recursion, show "internal error" message.
- FUTURE: when "remote TUI" is merged, can we remove log_lock()?
2. Skip logging if log_init wasn't called yet.
PPIndentWidth
The number of columns to use for indentation of preprocessor statements.
When set to -1 (default) IndentWidth is used also for preprocessor statements.
Adds an `smods` key to `nvim_create_user_command` Lua command callbacks,
which has command modifiers but in a structured format. This removes the
need to manually parse command modifiers. It also reduces friction in
using `nvim_cmd` inside a Lua command callback.
Since we're not running tests or other things that are more sensitive to
changes in the VM environment, use ubuntu-latest to avoid the busy work
of updating the VM image.
The async nature of feed_command caused a race condition on the calls
to delete, leaving the files on the root of the repo. Just use os.remove
and no need to wipeout.
In scripts/vim-patch.sh line 335:
printf '
^-- SC2183 (warning): This format string has 4 variables, but is passed 3 arguments.
In scripts/vim-patch.sh line 597:
list_missing_vimpatches 1 "$@" | while read -r vim_commit; do
^--------^ SC2030 (info): Modification of vim_commit is local (to subshell caused by pipeline)
In scripts/vim-patch.sh line 626:
done < <(git -C "${VIM_SOURCE_DIR}" diff-tree --no-commit-id --name-only -r "${vim_commit}" -- . ':!src/version.c')
^-----------^ SC2031 (info): vim_commit was modified in a subshell. That change might be lost.
For more information:
https://www.shellcheck.net/wiki/SC2183 -- This format string has 4 variable...
https://www.shellcheck.net/wiki/SC2030 -- Modification of vim_commit is loc...
https://www.shellcheck.net/wiki/SC2031 -- vim_commit was modified in a subs...
Previously the `offset!` directive populated the metadata in such a way
that the new range could be attributed to a specific capture. #14046
made it so the directive simply stored just the new range in the
metadata and information about what capture the range is based from is
lost.
This change reverts that whilst also correcting the docs.
Previously, there was a bug where setting the local value of 'winbar' to
itself would cause winbar to appear on a floating window, which is
undesirable. This fix makes it so that it's explicitly required for the
window-local value of 'winbar' for a floating window to be set in order
for winbar to be shown on that window.
Problem: The terminal debugger uses various global variables.
Solution: Add a dictionary to hold the terminal debugger preferences.
c9a431c763
Omit popup menu.
When enabling -municode for MSVC the following warning shows up:
"LINK : warning LNK4044: unrecognized option '/municode'; ignored".
This will ensure cleaner logs for the MSVC job.
The client state is cleaned up both in client.stop() as well as in the
client.on_exit() handler. Technically, the client has not actually
stopped until the on_exit handler is called, so we should just do this
cleanup there and remove it from client.stop().
This makes the common use case easier.
If one really needs access to all clients, they can create a filter
function which manually calls `get_active_clients`.
Problem: Cannot select one character inside ().
Solution: Do not try to extend the area if it is empty. (closesvim/vim#10472,
closesvim/vim#6616)
53737b5eee
Neovim already removed `evim` (or any similar flags). The 'insertmode'
option is a weird remnant, so get rid of it.
The 'insertmode' option is replaced with a script that closely emulates
the option. This script is documented at :help 'insertmode'
Problem: setbufline() may change Visual selection. (Qiming Zhao)
Solution: Disable Visual mode when using another buffer. (closesvim/vim#10466)
0ad00a7fd3
Problem:
If there are floating windows when `:mksession` runs, the session cannot
be properly restored.
Solution:
Change `:mksession` to skip floating windows. This matches Vim's
treatment of popup windows.
An alternative approach could have `:mksession` save floating windows
that can be _properly_ restored (rather than skip them entirely, which
is what this PR does). While that would seemingly be a more complete
fix, that could present additional issues since floating windows are
ordinarily created by plugins, and they may no longer be properly under
a plugin's control when restored.
closes#18432
from ~30 to ~20 secs with ASAN build
- feedkeys test: the same substitute was repeated, and not even correctly spelled
- don't clear() in a tight loop. "bwipe!" is enough to erase undo history
- error in cnoremap mapping causes a check_for_delay(). <c-c> it away.
Problem: With 'foldmethod' "indent" some lines are not included in the
fold. (Oleg Koshovetc)
Solution: Fix it. (Brandon Simmons, closesvim/vim#10399, closesvim/vim#3214)
d98e75e236
Cuts down typical run time for `plugin/lsp_spec.lua`
from 70 secs to 12 secs in ASAN CI build.
This happens in ASAN/EXIT_FREE builds where nvim waits 2000ms due to
unclosed handled. I wasn't able to pin-point the exact cause.
But these tests ran in nested context where two server/client pairs
were setup for no good reason. Moving these tests out so only one client
is being setup fixed the exit hang.
Problem: smart/C/lisp indenting is optional, which makes the code more
complex, while it only reduces the executable size a bit.
Solution: Graduate FEAT_CINDENT, FEAT_SMARTINDENT and FEAT_LISP.
8e145b8246
Problem: Increment/decrement might get interrupted by updating folds.
Solution: Disable fold updating for a moment. (Christian Brabandt,
closesvim/vim#3599)
6b731886ca
This lint job will ensure that the C codebase is properly formatted at
all times. This helps eliminate most of clint.py.
To save CI time, it's faster to manually compile uncrustify and cache
the binary instead of using homebrew (the apt-get package is too old).
fs_init() must be called before early_init() in init/helpers.lua
If I run 'make unittest' on my Mac (macOS 10.14/Mojave or 12/Big Sur, intel
CPU), every test produce a core dump.
Call sequence in the core is:
early_init() main.c:197
set_init_1() option.c:508
runtimepath_default() runtime.c:1205
get_lib_dir() runtime.c:1175
os_isdir() fs.c:137
os_getperm() fs.c:777
os_stat() fs.c:761
fs_loop_lock() fs.c:72
uv_mutex_lock(&fs_loop_mutex) thread.c:352
abort()
.deps/build/src/libuv/src/unix/thread.c:
void uv_mutex_lock(uv_mutex_t* mutex) {
if (pthread_mutex_lock(mutex))
abort(); // line 352
}
So pthread_mutex_lock(&fs_loop_mutex) failed. The reason seems to be simple.
fs_init() was not called and fs_loop_mutex has not been initialized. fs_init()
was moved out from early_init() in main.c by
b87867e69e, but unit/helpers.lua was not updated
accordingly.
The earlier vsnprintf() call checks whether str->items is NULL, sets of
the "possible NULL" spidey sense. kv_ensure_space() guarantees
str->items is non-NULL but since it doesn't use NULL checks to decide
whether to alloc, static analyzers can't tell this code path is safe.
register_cfunc allocates a ufunc_T, but doesn't store the pointer
anywhere before returning. The uf_name member variable is stored in a
hashtable and used to lookup the ufunc_T later, but that's too much for
Coverity to track.
Adding the annotation ensures that any new callers to register_cfunc
don't pop up as new "leaks" in the Coverity scans.
Coverity was reporting false positives, particularly around for non-NUL
terminated strings around uses of xmemdupz(). The updated model
ensures Coverity understands xmemdupz allocates an extra byte and sets it
to NUL as well as the main details of our other allocation related
wrappers.
Problem: When 'shortmess' contains 'A' loading a session may still warn for
an existing swap file. (Melker Österberg)
Solution: Keep the 'A' flag to 'shortmess' in the session file.
(closesvim/vim#10443)
aaadb5b6f7
Use readfile() with "B" flag in test as readblob() needs patch 8.2.2343.
The docs for `nvim_buf_set_extmark` mention that you can create a new
extmark when passing in `id=0`, however if you do this you'll get an
error since the code checks that id is positive.
```
id is not a positive integer
```
This change re-words the sentence to make it clearer.
The current approach of using `on_attach` callbacks for configuring
buffers for LSP is suboptimal:
1. It does not use the standard Nvim interface for driving and hooking
into events (i.e. autocommands)
2. There is no way for "third parties" (e.g. plugins) to hook into the
event. This means that *all* buffer configuration must go into the
user-supplied on_attach callback. This also makes it impossible for
these configurations to be modular, since it all must happen in the
same place.
3. There is currently no way to do something when a client detaches from
a buffer (there is no `on_detach` callback).
The solution is to use the traditional method of event handling in Nvim:
autocommands. When a LSP client is attached to a buffer, fire a
`LspAttach`. Likewise, when a client detaches from a buffer fire a
`LspDetach` event.
This enables plugins to easily add LSP-specific configuration to buffers
as well as enabling users to make their own configurations more modular
(e.g. by creating multiple LspAttach autocommands that each do
something unique).
Remove the command('qall!') from mksession_spec.lua because it prevents
helpers.rmdir() from retrying.
Allow extra trailing spaces when matching terminal lines.
Problem: Recursive command line loop may cause a crash.
Solution: Limit recursion of getcmdline().
51f0bfb88a
Cherry-pick e_command_too_recursive from patch 8.2.3957.
Problem:
vim.lsp: require("vim.lsp.health").check()
========================================================================
- ERROR: Failed to run healthcheck for "vim.lsp" plugin. Exception:
function health#check, line 20
Vim(eval):E5108: Error executing lua ...m/HEAD-6613f58/share/nvim/runtime/lua/vim/lsp/health.lua:20: attempt to index a nil value
stack traceback:
...m/HEAD-6613f58/share/nvim/runtime/lua/vim/lsp/health.lua:20: in function 'check'
[string "luaeval()"]:1: in main chunk
Solution:
Check for nil.
fix#18602
vim-patch:8.0.1119: quitting a split terminal window kills the job
N/A, or tracked in https://github.com/neovim/neovim/issues/5431
vim-patch:8.0.1307: compiler warning for ignoring return value
N/A
vim-patch:8.0.1335: writefile() using fsync() may give an error
N/A
vim-patch:8.0.1339: no test for what 8.0.1335 fixes
already merged in 5972ff0056
vim-patch:8.0.1367: terminal test hangs, executing abcde
N/A
vim-patch:8.0.1562: the terminal debugger can't set breakpoint with mouse
we have all :Termdebug changes
vim-patch:8.0.1609: shell commands in the GUI use a dumb terminal
tracked in https://github.com/neovim/neovim/issues/5431
vim-patch:8.0.1616: Win32: shell commands in the GUI open a new console (guioptions="!")
tracked in https://github.com/neovim/neovim/issues/1496
vim-patch:8.0.1706: cannot sent CTRL-\ to a terminal window
already possible via :call chansend()
Removes NOLINT, which is pointless for the generated terminfo_defs.h.
Adds `uncrustify:off`, so it is not uncrustify which complains about the same
things (too long lines, no space after comma) instead.
previous: https://github.com/neovim/neovim/pull/14123
CI tests were disabled on drafts #18566 to manage the
large number of incoming jobs. While this did help, it had the drawback
of making the purpose of the ready-for-review a bit fuzzier. It went
from a clear "my PR is ready" signal to maintainers to somewhere between
"my PR is ready but I need the tests to confirm" to "please don't merge
yet, I just need to see the test results". Worse is that the specific
case of wanting to see the test results but not wanting it merged is
that this needs to be actively conveyed to the maintainers with a [DNM]
or a comment to not merge the PR yet. All of this causes weird
workarounds and noises which I believe isn't necessary.
The reason why I don't think this workaround is needed anymore is that
our CI now aborts a job if a new job from the same pull requests is
created, which makes the "10 simultaneous jobs per PR" situations that
triggered this not possible.
* fix(PVS/V547): remove ifs that are always true or false
* fix(PVS/V560): remove partial conditions that are always true
* fix(PVS/V1044): suppress warning about loop break conditions
* fix(PVS/V1063): suppress "modulo by 1 operation is meaningless"
* fix(PVS/V568): suppress "operator evaluates the size of a pointer"
Also mark vim-patch:8.2.4958 as ported.
This makes it more convenient to find memory leaks since you don't need
to remember to set the EXITFREE flag every time you use valgrind or a
sanitizer.
The formatting for these files were originally disabled as to signal
that "we don't own these files", meaning we intentionally want to
minimize the amount of work put in these files as the return will be
very little. This unfortunately conflicts with other refactoring efforts
that happen to touch these files, and it's easier to simply enable
formatting.
Uncrustify and clang-format are already both excellent at ordering
includes; this isn't something we need to check for ourselves. Also
remove the section on include order in the dev-style documentation.
Problem:
q in "$MANPAGER mode" does not quit Nvim. This is because
ftplugin/man.vim creates its own mapping:
nnoremap <silent> <buffer> <nowait> q :lclose<CR><C-W>c
which overrides the one set by the autoload file when using :Man!
("$MANPAGER mode")
Solution:
Set b:pager during "$MANPAGER mode" so that ftplugin/man.vim can set the
mapping correctly.
Fixes#18281
Ref #17791
Helped-by: Gregory Anders <8965202+gpanders@users.noreply.github.com>
Problem: Changing 'switchbuf' may have no effect.
Solution: Handle 'switchbuf' in didset_string_options(). (Sean Dewar,
closesvim/vim#10406)
39c46b4378
Guess it doesn't hurt to cherry-pick didset_string_options from v8.1.2045
(but in option.c, for now).
Implement filtering of actions based on the kind when passing the 'only'
parameter to code_action(). Action kinds are hierachical with a '.' as
the separator, and the filter thus allows, for example, both 'quickfix'
and 'quickfix.foo' when requestiong only 'quickfix'.
Fix https://github.com/neovim/neovim/pull/18221#issuecomment-1110179121
Hardcoding a charset causes trouble when porting Vim patches.
I previously tried to unset "charset" for certain file extensions, but
vim-patch.sh can generate more files, and automatically detecting file
encoding is more correct anyway.
`:tabmove` takes either an argument (`:tabmove -`) or an address (`:-tabmove`).
The code assumed that `:tabmove` is the first command on the cmdline, but that
is not the case when using additional modifiers like `:silent`.
Make the addr parsing more robust by searching the command first, then going
back to check for a potential address `-`.
This will check if the string after the variable in a @param is either
"number", "string", "table", "boolean" and "function" and if so add a
parenthesis around it. This will help separate the variable type with
the following text. Had all our functions been annotated with emmylua
then a more robust solution might have been preferable (such as always
assuming the third string is parameter type without making any checks).
I believe however this is a clear improvement over the current situation
and will suffice for now.
At the moment of comparison, the pointer save_curwin can be invalid (as
suggested by the comment) because it has been free'd. Worst, the new
curwin could have been re-allocated to that same pointer, altering the
execution flow unpredictably.
While there are many other potential similar cases to fix in the
codebase, the presented scenario is not hypothetical and does happen in
practice (while spawning new windows from fzf for instance).
There are numerous other instances of curwin comparisons in the
codebase, and they may need further investigation.
closes#16941
Most code in keymap.h is for keycode definitions, while most code in
keymap.c is for the parsing and conversion of keycodes.
The name "keymap" may also make people think these two files are for
mappings, while in fact keycodes are used even when no mappings are
involved, so "keycodes" should be a better file name than "keymap".
It's special cased by the vimSubst syntax group, and isn't present in Vim's
vimCommand group.
For example, this fixes `call s:Foo()` highlighting `:` as Error in Nvim, as the
`s` is parsed as vimCommand rather than as vimUserFunc since
`contains=vimCommand` was added to vimUserFunc (and vimFunc) in a rt update.
Interestingly, `g:`, `l:`, etc. have the same issues due to :global, :list, etc.
Vim also has that problem, so it should ideally be fixed upstream.
We could also omit g[lobal] from vimCommand and rely on vimGlobal instead, but
it doesn't work in some cases, like when there's a `:` before the command. Also,
Vim matches only `g` in vimCommand for some reason, which doesn't produce any
highlight for `:global/foo/bar` (with Nvim you at least get some highlights on
the `global` bit despite the leading `:`).
Also, remove special handling of :py3 in syntax/vim.vim, as the generator seems
to have no problems finding it.
Adds the API function `nvim_cmd` which allows executing an Ex-command through a Dictionary which can have the same values as the return value of `nvim_parse_cmd()`. This makes it much easier to do things like passing arguments with a space to commands that otherwise may not allow it, or to make commands interpret certain characters literally when they otherwise would not.
Problem: The mode #defines are not clearly named.
Solution: Prepend MODE_. Renumber them to put the mapped modes first.
249591057b
A hunk from the patch depends on patch 8.2.4861, which hasn't been
ported yet, but that should be easy to notice.
Problem: Trailing backslash may cause reading past end of line.
Solution: Check for NUL after backslash.
53a70289c2
N/A patches for version.c:
vim-patch:8.2.4926: #ifdef for crypt feature around too many lines
Problem: #ifdef for crypt feature around too many lines.
Solution: Move code outside of #ifdef. (closesvim/vim#10388)
51f0bc31d3
Problem: maparg() may return a string that cannot be reused.
Solution: use msg_outtrans_special() instead of str2special().
(closesvim/vim#10384)
0519ce0039
Problem: Conceal character from matchadd() displayed too many times.
Solution: Check the syntax flag. (closesvim/vim#10381, closesvim/vim#7268)
9830db6305
Problem: Mouse in Insert mode test fails.
Solution: Fix the text and check relevant positions.
8e8dc9b323
Use nvim_input_mouse() to set mouse position, and discard mouse event
using getchar().
vim-patch:8.2.4903: cannot get the current cmdline completion type and position
Problem: Cannot get the current cmdline completion type and position.
Solution: Add getcmdcompltype() and getcmdscreenpos(). (Shougo Matsushita,
closesvim/vim#10344)
79d599b877
vim-patch:8.2.4910: imperfect coding
Problem: Imperfect coding.
Solution: Make code nicer.
9ff7d717aa
LSP servers should be daemonized (detached) so that they run in a
separate process group from Neovim's. Among other things, this ensures
the process does not inherit Neovim's TTY (#18475).
Make this configurable so that clients can explicitly opt-out of
detaching from Nvim.
Problem: Some users do not want a line comment always inserted.
Solution: Add the '/' flag to 'formatoptions' to not repeat the comment
leader after a statement when using "o".
2bf875f881
It seems range and count can be used together in commands. This PR fixes
the behavior of `nvim_parse_cmd` for those cases by removing the mutual
exclusivity of "range" and "count". It also removes range line number
validation for `nvim_parse_cmd` as it's not its job to validate the
command.
This warning is essentially only triggered for ported vim functions.
It's unlikely that we'll refactor vim functions solely based on their
size since it'd mean we'd greatly deviate from vim, which is a high cost
when it comes to importing the vim patches. Thus, this warning only
serves as an annoyance and should be removed.
Problem: "P" in Visual mode still changes some registers.
Solution: Make "P" in Visual mode not change any register. (Shougo
Matsushita, closesvim/vim#10349)
509142ab7a
Uncrustify version 0.75 assumes the macro inside the enum is an enum
element, and thus adds a comma after it. This breaks neovim and makes it
impossible to build it.
Adds filter and id options to filter the client to use for rename.
Similar to the recently added `format` function.
rename will use all matching clients one after another and can handle a
mix of prepareRename/rename support. Also ensures the right
`offset_encoding` is used for the `make_position_params` calls
This fixes issues where subsequent calls to vim.lsp.codelens.refresh()
would have no effect due to the buffer not getting cleared from the
active_refresh table.
Examples of how such scenarios would occur are:
- A textDocument/codeLens result yielded an error.
- The 'textDocument/codeLens' handler was overriden in such a way that
it no longer called vim.lsp.codelens.on_codelens().
Changes the `range` value in `nvim_parse_cmd` into an array to describe
range information more concisely. Also makes `range` and `count` be
mutually exclusive by making count `-1` when command takes a range
instead of a count. Additionally corrects the behavior of `count` for
built-in commands by making the default count `0`.
Problem: When closing help window autocmds triggered for the wrong window.
Solution: Figure out the new current window earlier. (closesvim/vim#10348)
2a2707d033
Problem: Listing of mapping with K_SPECIAL is wrong.
Solution: Adjust escaping of special characters. (closesvim/vim#10351)
ac402f4d64
Avoid overshadowing.
Cherry-pick Test_list_mapping() from Vim patches 8.2.{0148,2994}.
Fix rhs_is_noop.
Problem: :startinsert right after :stopinsert does not work when popup menu
is still visible.
Solution: Use ins_compl_active() instead of pum_visible(). (closesvim/vim#10352)
cd5dbad184
Adds range, count and reg to the return values of nvim_parse_cmd. Also makes
line1 and line2 be -1 if the command does not take a range. Also moves
nvim_parse_cmd to vimscript.c because it fits better there.
This is primarily intended to act as documentation for the developer so
they know exactly when and what to remove. This will help prevent the
situation of deprecated code lingering for far too long as developers
don't have to worry if a function is safe to remove.
PROBLEM
------------------------------------------------------------------------
$NVIM_LISTEN_ADDRESS has conflicting purposes as both a parameter ("the
current process should listen on this address") and a descriptor ("the
current process is a child of this address").
This contradiction means the presence of NVIM_LISTEN_ADDRESS is
ambiguous, so child Nvim always tries to listen on its _parent's_
socket. This is the cause of lots of "Failed to start server" spam in
our test/CI logs:
WARN 2022-04-30… server_start:154: Failed to start server: address already in use: \\.\pipe\nvim-4480-0
WARN 2022-04-30… server_start:154: Failed to start server: address already in use: \\.\pipe\nvim-2168-0
SOLUTION
------------------------------------------------------------------------
1. Set $NVIM to the parent v:servername, *only* in child processes.
- Now the correct way to detect a "parent" Nvim is to check for $NVIM.
2. Do NOT set $NVIM_LISTEN_ADDRESS in child processes.
3. On startup if $NVIM_LISTEN_ADDRESS exists, unset it immediately after
server init.
4. Open a channel to parent automatically, expose it as v:parent.
Fixes#3118Fixes#6764Fixes#9336
Ref https://github.com/neovim/neovim/pull/8247#issuecomment-380275696
Ref #8696
* vim-patch:8.2.1835: ":help ??" finds the "!!" tag
Problem: ":help ??" finds the "!!" tag.
Solution: Do not translate "?" into ".". (Naruhiko Nishino, closesvim/vim#7114,
closesvim/vim#7115)
6eb36ade98
Change test because patch 8.2.1794 hasn't been ported yet.
When checking the version of the doxygen installed from conda the output
has the following format:
1.9.2 (ee54ebd4f0ad83d9c44f19a459146de64d0ffba2*)
This would cause an error in the "Missing API docs" CI job. This fix
will correctly parse the doxygen version for both stable releases
("1.9.2") as well as the version with the git commit hash attached.
Problem: Unexpected "No matching autocommands".
Solution: Do not give the message when aborting. Mention the arguments in
the message. (closesvim/vim#8690)
1b154ea121
There have been a few instances where developers got confused as to why
their generated documentation differs from the one generated by the CI.
More often than not, the reason is that their doxygen version is older
than 1.9.0, which is the current minimum version. Having a simple
version check will help save future developers avoid this problem.
`:syntax enable` is no longer necessary (and may even be harmful). Do not recommend it in `usr_05.txt` and `usr_06.txt`, and mention the new default in `syntax.txt`.
Deprecates the existing `vim.lsp.buf.formatting` function.
With this, `vim.lsp.buf.format` will replace all three:
- vim.lsp.buf.formatting
- vim.lsp.buf.formatting_sync
- vim.lsp.buf.formatting_seq_sync
Problem: Local completion with mappings and simplification not working.
Solution: Fix local completion <C-N>/<C-P> mappings not ignored if keys are
not simplified. (closesvim/vim#10323)
ee44603069
* feat(lsp)!: remove capabilities sanitization
Users must now access client.server_capabilities which matches the same
structure as the protocol.
https://microsoft.github.io/language-server-protocol/specification
client.resolved_capabilities is no longer used to gate capabilities, and
will be removed in a future release.
BREAKING CHANGE
Co-authored-by: Mathias Fussenegger <f.mathias@zignar.net>
Implement two new options to vim.lsp.buf.code_action():
- filter (function): predicate taking an Action as input, and returning
a boolean.
- apply (boolean): when set to true, and there is just one remaining
action (after filtering), the action is applied without user query.
These options can, for example, be used to filter out, and automatically
apply, the action indicated by the server to be preferred:
vim.lsp.buf.code_action({
filter = function(action)
return action.isPreferred
end,
apply = true,
})
Fix#17514.
Problem: Duplicate code.
Solution: Move code below if/else. (closesvim/vim#10314)
590f365f91
N/A patches for version.c:
vim-patch:8.2.4844: <C-S-I> is simplified to <S-Tab>
Problem: <C-S-I> is simplified to <S-Tab>.
Solution: Do not simplify CTRL if there is also SHIFT. (closesvim/vim#10313)
758a8d1999
vim-patch:8.2.4846: termcodes test fails
Problem: Termcodes test fails.
Solution: use CTRL-SHIFT-V to insert an unsimplified key. (closesvim/vim#10316)
bad8a013c2
Add space around arithmetic operators '+' and '-'.
Remove space between back-to-back parentheses, i.e. ')(' vs. ') ('.
Remove space between '((' or '))' of control statements.
Add space between ')' and '{' of control statements.
Remove space between function name and '(' on function declaration.
Collapse empty blocks between '{' and '}'.
Remove newline at the end of the file.
Remove newline between 'enum' and '{'.
Remove newline between '}' and ')' in a function invocation.
Remove newline between '}' and 'while' of 'do' statement.
Problem: Modifiers not simplified when timed out or using feedkeys() with
'n" flag.
Solution: Adjust how mapped flag and timeout are used. (closesvim/vim#10305)
68a573ce2b
Problem: When there is a partially matching map and modifyOtherKeys is
active a full map may not work.
Solution: Only simplify modifiers when there is no matching mapping.
(closesvim/vim#8792)
196c3850db
Omit test as it sends terminal codes. Use a Lua test instead.
Problem: Special key names don't work if 'isident' is cleared.
Solution: Add vim_isNormalIDc() and use it for special key names.
(closesvim/vim#2389)
e3d1f4c982
Code is N/A as Nvim already has ascii_isident(), so just port the test.
Problem: CTRL-R CTRL-R doesn't work with modifyOtherKeys.
Solution: Allow key codes when fetching argument for CTRL-R. (closesvim/vim#5266)
Also fix CTRL-G in Insert mode.
38571a04b4
Omit test as it sends terminal codes. Use a Lua test instead.
Problem: Merging modifier for modifyOtherKeys is done twice.
Solution: Remove the merging done in vgetc().
673fc3e23f
Omit ex_getln.c change as it was removed in patch 8.2.2084, so
no_reduce_keys is still not needed in Nvim.
Problem: Mapping with partly modifyOtherKeys code does not work.
Solution: If there is no mapping with a separate modifier include the
modifier in the key and then try mapping again. (closesvim/vim#6200)
975a880a13
Cherry-pick applicable part of put_string_in_typebuf().
Revert related changes from 10a5825.
Use KEYLEN_PART_KEY for incomplete modifier sequence.
Omit test as it sends terminal codes. Use a Lua test instead.
Problem: Using \{xxx} for encoding a modifier is not nice.
Solution: Use \<*xxx> instead, since it's the same as \<xxx> but producing a
different code.
fccd93f091
Use this notation in langmap_spec.
Problem: GUI tests fail because the test doesn't use a modifier.
Solution: Add "\{xxx}" to be able to encode a modifier.
ebe9d34aa0
Change macros to enums to use them in unit tests.
Problem: Can't distinguish <M-a> from accented "a" in the GUI.
Solution: Use another way to make mapping <C-bslash> work. (closesvim/vim#6163)
f4ae6b245a
Problem: Dropping modifier when putting a character back in typeahead.
Solution: Add modifier to ins_char_typebuf(). (closesvim/vim#6158)
b42c0d5427
Vim's test doesn't seem to work properly as the hit-enter prompt seems
to be delayed. Add a Lua screen test.
Problem: maparg() does not provide enough information for mapset().
Solution: Add "lhsraw" and "lhsrawalt" items. Drop "simplified"
9c65253fe7
This only includes the "lhs" value part.
This marks the following Vim patches as ported:
vim-patch:8.1.2333: with modifyOtherKeys CTRL-^ doesn't work
Problem: With modifyOtherKeys CTRL-^ doesn't work.
Solution: Handle the exception.
828ffd5963
vim-patch:8.1.2350: other text for CTRL-V in Insert mode with modifyOtherKeys
Problem: Other text for CTRL-V in Insert mode with modifyOtherKeys.
Solution: Convert the Escape sequence back to key as if modifyOtherKeys is
not set, and use CTRL-SHIFT-V to get the Escape sequence itself.
(closesvim/vim#5254)
fc4ea2a72d
vim-patch:8.2.2084: CTRL-V U doesn't work to enter a Unicode character
Problem: CTRL-V U doesn't work to enter a Unicode character when
modifyOtherKeys is effective. (Ken Takata)
Solution: Add a flag to get_literal() for the shift key. (closesvim/vim#7413)
0684e36a7e
Omit getcmdkeycmd() change as it depends on Vim patch 8.2.2062, which
may introduce a potential breakage.
Problem: Cannot map <C-H> when modifyOtherKeys is enabled.
Solution: Add the <C-H> mapping twice, both with modifier and as 0x08. Use
only the first one when modifyOtherKeys has been detected.
459fd785e4
Add REPTERM_NO_SPECIAL instead of REPTERM_SPECIAL because the meaning of
"special" is different between Vim and Nvim.
Omit seenModifyOtherKeys as Nvim supports attaching multiple UIs.
Omit tests as they send terminal codes.
Keep the behavior of API functions.
In a locale "en_US", "en" is the language and "US" is the region.
Before this change, we were too clever for our own good and tried to handle the
region as well. But if the macOS primary language is set to "English" and the
region to "Norway", we would end up with "en_NO", which is a locale that does
not exist.
Now we only take the language into account. Taking the example from above would
yield "en_US", which is a sensible fallback.
If the region is important to you, set $LANG and the more specific LC_*
variables in your shell config or alternatively use `:help :language`.
References https://github.com/neovim/neovim/issues/18292
Problem:
winpty is only needed for Windows 8.1. Removing it reduces our build and code
complexity.
Solution:
- Remove winpty.
- Require Windows 10.
closes#18252
Problem:
I had some issues where multiple plugins (vim-fzf and fugitive) was slow
because of my `.zshenv`.
Solution:
Check shell performance in :checkhealth.
Co-authored-by: Justin M. Keyes <justinkz@gmail.com>
Some language servers send empty `textDocument/publishDiagnostics`
messages after indexing the project with URIs corresponding to unopened buffers.
This commit guards against opening buffers corresponding to empty diagnostics.
Problem: Using matchfuzzy() on a long list can take a while.
Solution: Add a limit to the number of matches. (Yasuhiro Matsumoto,
closesvim/vim#10189)
9029a6e993
The default value of including F in 'shortmess' has the unfortunate side
effect of hiding output from autocommands. This is a common source of
confusion and often leads people to think their autocommands are not
working when they are. There is a small snippet in the docs for
'shortmess' indicating that the F flag suppresses autocmd output, but
it's not easy to find if you don't already know to look for it.
This commit removes that behavior of the F flag to make it only suppress
file info when opening a new file.
This reverts commit 526798a941.
This will make man filetype not modifiable by default, as it is the
superior behavior in my opinion. More importantly, also make it possible
for a user to modify man filetypes by adding `set modifiable` in
`~/.config/nvim/ftplugin/man.vim` or its equivalent.
ref #11450closes#17595
Co-authored-by: Javier López <graulopezjavier@gmail.com>
The check in redraw_for_cursorline() is not needed because VALID_VIRTCOL
is always cleared when the cursor moves to another line, so the check in
redraw_for_cursorcolumn() is enough.
* fix(PVS/V002): disable rule completely
V002: "Some diagnostic messages may contain incorrect line number in
this file." This particular check seems unreliable. It says on their
website https://pvs-studio.com/en/docs/warnings/v002/ that this warning
occurs when there are multiline pragmas, but there are none in
extmark.c.
* fix(PVS/V756): ignore "counter is not used inside a nested loop" warning
The nested loop starts with "AutoCmd *ac = ap->cmds" so "ap" is
definitely used.
* fix(PVS/V560): disable "a part of conditional expression is always true"
* fix(PVS/V614): potentially uninitialized variable 'blen' used
Problem: No test for what 8.2.4806 fixes.
Solution: Add a test. (closesvim/vim#10727)
ac92ab7719
Test cannot be used because it must use test_setmouse(). Use a Lua test.
Problem: A mapping using <LeftDrag> does not start Select mode.
Solution: When checking for starting select mode with the mouse also do this
when there is typeahead. (closesvim/vim#10249)
53ef573148
Problem: Pasting text while indent folding may mess up folds.
Solution: Adjust the way folds are split. (Brandon Simmons, closesvim/vim#10254)
2c40707baa
In `filetype.lua`, source runtime `ftdetect` scripts within the `filetypedetect` augroup, same as `filetype.vim` (and only do so if `g:did_load_ftdetect` does not exist).
`vim.keymap.del` takes an `opts` parameter that lets caller refer to and
delete buffer-local mappings. For some reason the implementation of
`vim.keymap.del` mutates the table that is passed in, setting
`opts.buffer` to `nil`. This is wrong and also undocumented.
Problem: CurSearch highlight does not work for multi-line match.
Solution: Check cursor position before adjusting columns. (closesvim/vim#10133)
693ccd1160
Problem: Test is not cleaned up.
Solution: Make test clean up after itself. Avoid NUL. (closesvim/vim#10233)
7851c69a12
Adapt test_autocmd_vimgrep() to Nvim.
Problem: Autocmd events triggered in different order when reusing an empty
buffer.
Solution: Call buff_freeall() earlier. (Charlie Groves, closesvim/vim#10198)
fef4485ef5
Test failure becomes very strange.
Problem: Using freed memory when using an autocommand to split a window
while a buffer is being closed.
Solution: Disallow splitting when the buffer has b_locked_split set.
983d83ff1c
Put the error message in window.c.
Cherry-pick a memory leak fix from Vim patch 8.2.0399.
Test still fails.
Problem: Using freed memory when window is closed by autocommand.
(houyunsong)
Solution: Check the window pointer is still valid.
2c7080bf1c
Add missing comment from Vim patch 8.0.1420.
Test fails.
Problem: Crash when using command line window in an autocommand.
(houyunsong)
Solution: Save and restore au_new_curbuf.
aad5f9d79a
Nvim has removed :open, so use :edit in the test instead.
Problem: Autocommand that splits window messes up window layout.
Solution: Disallow splitting a window while closing one. In ":all" give an
error when moving a window will not work.
1417c766f5
Expected error number was changed to E242 in Vim in patch 8.2.1183, and
patch 8.2.2420 (which has already been ported) made the test no longer
throw E249 in Vim, so just use E242 in the test.
Problem: Insufficient testing for indent.c.
Solution: Add indent tests. (Yegappan Lakshmanan, closesvim/vim#5736)
bd7206e02c
Cherry-pick Test_ex_mode() changes from Vim patches 8.2.{0342,0347}.
Reorder test_expand_func.vim to match upstream.
Problem: getwininfo() may get oudated values.
Solution: Make sure w_botline is up-to-date. (closesvim/vim#10226)
8530b41fd3
Correct test order and add a modeline in test_bufwintabinfo.vim.
Problem: Current instance of last search pattern not easily spotted.
Solution: Add CurSearch highlighting. (closesvim/vim#10133)
a43993897a
This fixes CurSearch highlight for multiline match.
Omit screen redrawing code because Nvim redraws CurSearch differently.
Some people ran into issues trying to use `callback = myluafun` because
of the event data payload.
Co-authored-by: Gregory Anders <8965202+gpanders@users.noreply.github.com>
The LSP progress handler would put non-progress messages (such as from
clangd or pyls; not part of the LSP spec) directly into
`client.messages`, while `vim.lsp.util.get_progress_messages()` would
try to fetch them from `client.messages.messages` instead (and come up
empty everytime). This would result in these messages never being
cleaned up by `get_progress_messages()`.
This commit fixes that by treating those messages like show-once
progress messages (by setting `done=true` immediately).
Problem: KRL files using "deffct" not recognized.
Solution: Adjust the pattern used for matching. (Patrick Meiser-Knosowski,
closesvim/vim#10200)
93c7a45e86
Problem: A second popup window with terminal causes trouble.
Solution: Disallow opening a second terminal-popup window. (closesvim/vim#6101,
closesvim/vim#6103) Avoid defaulting to an invalid line number.
b5383b174b
This is the only applicable hunk.
xstrnsave is a clone of vim_strnsave that uses char* instead of char_u*.
Its purpose short-term is to help reduce the number of casts and for
long-term to replace vim_strnsave as the need to use char_u is
eliminated.
Problem: Wrong 'statusline' value can cause illegal memory access.
Solution: Properly check the value. (closesvim/vim#10192)
5dc294a7b6
Make two error messages static variables.
- Fix bugs found by [Clang](#clang-scan-build), [PVS](#pvs-studio) or
[Coverity](#coverity).
- [Improve documentation][wiki-contribute-help]
- [Merge a Vim patch] (familiarity with Vim is *strongly* recommended)
- [Improve documentation](#documenting)
- [Merge a Vim patch] (requires strong familiarity with Vim)
- NOTE: read the above link before sending improvements to "runtime files" (anything in `runtime/`).
- Vimscript and documentation files are (mostly) maintained by [Vim](https://github.com/vim/vim), not Nvim.
- Lua files are maintained by Nvim.
Reporting problems
------------------
@ -30,9 +33,9 @@ Reporting problems
Developer guidelines
--------------------
- Read `:help dev` if you are working on Nvim core.
- Read `:help dev-ui` if you are developing a UI.
- Read `:help dev-api-client` if you are developing an API client.
- Read [:help dev](https://neovim.io/doc/user/develop.html#dev) if you are working on Nvim core.
- Read [:help dev-ui](https://neovim.io/doc/user/develop.html#dev-ui) if you are developing a UI.
- Read [:help dev-api-client](https://neovim.io/doc/user/develop.html#dev-api-client) if you are developing an API client.
- Install `ninja` for faster builds of Nvim.
```
sudo apt-get install ninja-build
@ -47,21 +50,19 @@ Pull requests (PRs)
- Your PR must include [test coverage][run-tests].
- Avoid cosmetic changes to unrelated files in the same commit.
- Use a [feature branch][git-feature-branch] instead of the master branch.
- Use a **rebase workflow** for small PRs.
- After addressing review comments, it's fine to rebase and force-push.
- Use a **merge workflow** for big, high-risk PRs.
- Use a _rebase workflow_ for small PRs.
- After addressing review comments, it's fine to force-push.
- Use a _merge workflow_ (as opposed to "rebase") for big, high-risk PRs.
- Merge `master` into your PR when there are conflicts or when master
introduces breaking changes.
- Use the `ri` git alias:
```
[alias]
ri = "!sh -c 't=\"${1:-master}\"; s=\"${2:-HEAD}\"; mb=\"$(git merge-base \"$t\" \"$s\")\"; if test \"x$mb\" = x ; then o=\"$t\"; else lm=\"$(git log -n1 --merges \"$t..$s\" --pretty=%H)\"; if test \"x$lm\" = x ; then o=\"$mb\"; else o=\"$lm\"; fi; fi; test $# -gt 0 && shift; test $# -gt 0 && shift; git rebase --interactive \"$o\" \"$@\"'"
```
This avoids unnecessary rebases yet still allows you to combine related
commits, separate monolithic commits, etc.
- Do not edit commits that come before the merge commit.
- During a squash/fixup, use `exec make -C build unittest` between each
pick/edit/reword.
### Merging to master
For maintainers: when a PR is ready to merge to master,
- prefer _Squash Merge_ for "single-commit PRs" (when the PR has only one meaningful commit).
- prefer _Merge_ for "multi-commit PRs" (when the PR has multiple meaningful commits).
### Stages: Draft and Ready for review
@ -92,7 +93,7 @@ the VCS/git logs more valuable. The general structure of a commit message is:
```
- Prefix the commit subject with one of these [_types_](https://github.com/commitizen/conventional-commit-types/blob/master/index.json):
make CMAKE_EXTRA_FLAGS=-DCMAKE_C_INCLUDE_WHAT_YOU_USE=include-what-you-use | tee iwyu.txt
```
See [#549][549] for more details.
Documenting
-----------
Many parts of the `:help` documentation are autogenerated from C or Lua docstrings using the `./scripts/gen_vimdoc.py` script.
You can filter the regeneration based on the target (api, lua, or lsp), or the file you changed, that need a doc refresh using `./scripts/gen_vimdoc.py -t <target>`.
## Lua docstrings
Lua documentation uses a subset of [EmmyLua] annotations. A rough outline of a function documentation is
```lua
--- {Brief}
---
--- {Long explanation}
---
---@param arg1 type {description}
---@param arg2 type {description}
{...}
---
---@return type {description}
```
If possible, always add type information (`table`, `string`, `number`, ...). Multiple valid types are separated by a bar (`string|table`). Indicate optional parameters via `type|nil`.
If a function in your Lua module should not be documented (e.g. internal function or local function), you should set the doc comment to:
```
---@private
```
Mark functions that are deprecated as
```
---@deprecated
```
Reviewing
---------
@ -266,30 +305,36 @@ commits in the feature branch which aren't in the `master` branch; `-p`
shows each commit's diff. To show the whole surrounding function of a change
# Need to manually reset terminal to avoid mangled output, nvim does not
# properly restore the terminal when it crashes.
tput reset
for i in "$log_path"/*;do
cat "$i"
done
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
