Problem: Using `buf=0`/`win=0` context in `vim._with` should be
equivalent to using explicit buffer/window identifier respectively.
Solution: Explicitly adjust context in case of `buf=0` or `win=0`.
Problem:
Too much boilerplate needed to use Lua to impl an excmd or f_xx
function.
Solution:
- Add `nlua_call_vimfn` which takes the args typval, executes
Lua, and returns a typval.
- refactor(excmd): lua impl for :log, :lsp
Problem:
- Builtin "Vimscript" functions (f_xx) are mostly implemented in C.
Partly that's because there is some boilerplate required to call out
to Lua.
- Calls to `vim.fn.foo()` always marshall over the Lua <=> Vimscript
("typval") bridge, even if `fn.foo()` is implemented entirely in Lua:
```
Lua => typval => Object => Lua => Object => typval => Lua.
```
Solution:
Functions declared in eval.lua with `func_lua` are implemented in
entirely in Lua (`_core/vimfn.lua`).
- `gen_eval.lua` wires `func_lua` entries to `lua_wrapper`, which handles
the typval conversion for Vimscript callers (slow path).
- `nlua_call()` detects `func_lua` functions and calls the Lua
implementation directly. This eliminates all conversion overhead for
Lua callers (fast path).
- Validate at build-time that `func`, `func_float`, and `func_lua` are
mutually exclusive.
- Migrate `hostname()` as a toy example, to show the idea.
Problem: - Paging keys in the dialog window consume input when the user
may not expect it. The dismissable title hint intended to
mitigate that results in having to press Escape twice to
abandon the prompt.
- Mimicked "msgsep" float border is taking up unnecessary
space when window takes up the entire screen.
Solution: - Use (conventional, albeit less convenient) keys intended
for scrolling to page the dialog window:
<(Mousewheel/Page)Up/Down>, <Home/End>.
- Only set the float top border when separation is actually
necessary, i.e. window does not reach the first row.
Problem:
Naming conventions are not automatically checked.
Solution:
Add a check to the doc generator. Eventually we should extract this
somehow, but that will require refactoring the doc generator...
Note: this also checks non-public functions, basically anything that
passes through `gen_eval_files.lua` and `gen_vimdoc.lua`. And that's
a good thing.
Replace the busted-based Lua test runner with a repo-local harness.
The new harness runs spec files directly under `nvim -ll`, ships its own
reporter and lightweight `luassert` shim, and keeps the helper/preload
flow used by the functional and unit test suites.
Keep the file boundary model shallow and busted-like by restoring `_G`,
`package.loaded`, `package.preload`, `arg`, and the process environment
between files, without carrying extra reset APIs or custom assertion
machinery.
Update the build and test entrypoints to use the new runner, add
black-box coverage for the harness itself, and drop the bundled
busted/luacheck dependency path.
AI-assisted: Codex
Problem:
Running `:Open` on an open buffer does not run `vim.ui.open()` on that file, requiring the user to type `:Open %`. This is inconsistent with most other vim commands which accept files, which default to the current buffer's file.
Solution:
Default to the current file when `:Open` is used without arguments.
Problem: When messages are appended to an already expanded cmdline,
the spilled lines indicator is not updated.
Solution: Remove early return for updating virtual text while cmdline is
expanded, guard updating "msg" virt_text at callsite instead.
Problem: 'showcmd' causes flickering when pressing "g<" to enter the
pager when the cmdline is expanded for messages.
Initial keypress for an incomplete mapping is not giving 'showcmd'
feedback while cmdline is expanded for messages (which is only
dismissed upon the vim.on_key callback after 'timeoutlen').
Solution: Delay dismissing expanded cmdline when vim.on_key() callback
receives "g".
Place 'showcmd' "last" virtual text during expanded cmdline.
Problem: Cmdline is not redrawn after an empty message clears it.
Remembered last drawn cursor position may be outdated but
equal to the current cmdline content with UI2.
Solution: Ensure cmdline is redrawn after an empty message clears it.
Compare wanted cursor position with actual cursor position.
Problem: `vim._core.ui2.enable` requires passing an empty table for
default options, unlike other (native) Lua API.
Solution: Initialize `opts` with empty table if `nil` is passed.
Problem: Mixing "buf" and "M.bufs.msg" in M.msg:start_timer().
Cannot run `require("vim._core.ui2").enable(nil)`.
Solution: Replace "M.bufs.msg" with "buf". Allow `opts == nil`.
Co-authored-by: Luuk van Baal <luukvbaal@gmail.com>
fix(ui2): prevent <CR> from focusing pager in insert/terminal mode
Problem: <CR> in insert/terminal mode can focus pager unexpectedly.
Solution: Don't enter the pager when <CR> is pressed during expanded
cmdline in insert/terminal mode.
Problem: When entering the cmdline below expanded messages, those
messages are moved to the dialog window. The dialog window
supports paging but that is unexpected in this situation where
it just serves to keep (some of, exactly those that were
visible before the cmdline was entered) the messages visible.
Wrong highlight group for dialog "more" message.
Solution: Don't create the `vim.on_key()` dialog pager callback after
entering the cmdline below expanded messages.
Use the MsgMore highlight group for the paging hint title.
Problem: - With expanded messages exceeding cfg.msg.cmd.height, entering
the cmdline scrolls to the bottom and expands to the full "cmd"
buffer text height.
- Cursor in the pager is not always at the last message and at
the bottom of the window when appending to the pager.
- unreliable test: messages2_spec: "closed msg window timer removes
empty lines".
Solution: - Achieve separation of the cmdline and message text by moving
messages to the dialog window when entering the cmdline below
expanded messages.
- Set cursor to start of the first message only when first
entering the pager. Use `norm! zb` to position last message
at the bottom of the window (which shouldn't crash anymore
since 911337eb).
- Increase cfg.msg.msg.timeout used in the test file.
The `buffer` option remains functional but is now undocumented.
Providing both will raise an error. Since providing `buf` was disallowed
before, there is no code that will break due to using `buffer` alongside
`buf`.
Problem:
- Window height is set dynamically to match the text height,
making it difficult for the user to use a different height.
- Cmdwin is closed to enter the pager but still taken into
account for the pager position, and not restored when
the pager is closed.
- Dialog pager handler may unnecessarily consume <Esc>.
Solution:
- Add maximum height config fields for each of the UI2 windows,
where a number smaller than one is a fraction of 'lines',
absolute height otherwise (i.e. `cfg.msg.pager.height = 0.5`).
- If the cmdwin will be closed to enter the pager, don't try
to position the pager above it. Re-enter the cmdwin when the
pager is closed.
- Only add vim.on_key() handler for the dialog paging is actually
possible.
Problem
Unlike inlay hints, code lenses are closely related to running commands;
a significant number of code lenses are used to execute a command (such
as running tests). Therefore, it is necessary to provide a default
mapping for them.
Solution
Add a new default mapping "grx" (mnemonic: "eXecute", like "gx").
Problem: Entering the pager is scheduled to avoid errors while in the
cmdwin. Meanwhile the current window is used as a condition to
detect an active pager.
Solution: Keep track of when the user is in the pager (or entered the
cmdwin while the pager was open).
Problem: Windows are closed and re-opened after changing tabpage (and
might be removed from the current tabpage after 094b297a).
Solution: Ensure windows are on the current tabpage by moving them.
Problem:
No way to iterate configs. Users need to reach
for `vim.lsp.config._configs`, an internal interface.
Solution:
Provide vim.lsp.get_configs().
Also indirectly improves :lsp enable/disable completion
by discarding invalid configs from completion.
Problem:
When `to_vim_value[info.metatype](info, value)` is called, a list value
such as `{'eob:a'}` is treated like a map, which generates `1:eob:a`.
Note: commands like `:lua vim.opt.wildmode={'longest:full'}` are not an
issue because only cases harcoded in `key_value_options` have metatype `map`.
Solution:
Check for array type and use the same logic as in array metatypes.
Problem:
"[Process exited]" is implemented in C with anonymous namespace
and users have no way to hide it.
Solution:
- Handle "TermClose" event in Lua.
- User can delete the "nvim.terminal" augroup to avoid "[Process exited]".
Problem:
`K` in help files may fail in some noisy text. Example:
(`fun(config: vim.lsp.ClientConfig): boolean`)
^cursor
Solution:
- `:help!` (bang, no args) activates DWIM behavior: tries `<cWORD>`,
then trims punctuation until a valid tag is found.
- Set `keywordprg=:help!` by default.
- Does not affect `CTRL-]`, that is still fully "tags" based.
Problem: Mimicked block mode for cmdline entered while expanded
does not work intuitively for repeated commands yielding
messages exceeding the screen height. The expanded cmdline
resizes and scrolls to bottom/top when appending a message
and entering the cmdline. Also includes the entered command,
as opposed to the UI1 behavior.
Crash when scrolling to bottom of pager due to recursive
uv_run after shell message callback executes `nvim_command()`
with 'showcmd'.
Solution: Still mimic block mode when entering the expanded cmdline,
but when the entered command emits a message open the pager
with the current message content in the expanded cmdline.
Always route typed commands to the pager when it is open.
Use `nvim_buf_set_cursor()` instead of `nvim_command()`.
refactor(lua): add integer coercion helpers
Add vim._tointeger() and vim._ensure_integer(), including optional base
support, and switch integer-only tonumber()/assert call sites in the Lua
runtime to use them.
This also cleans up related integer parsing in LSP, health, loader, URI,
tohtml, and Treesitter code.
supported by AI
Problem:
In autocmd examples, using "args" as the event-object name is vague and
may be confused with a user-command.
Solution:
Use "ev" as the conventional event-object name.
Problem:
If NVIM_LOG_FILE, or the default fallback, is inaccessible (e.g.
directory is owned by root), users get confused.
Solution:
Show a warning when $NVIM_LOG_FILE or $XDG_STATE_HOME are inaccessible.
Also fix a latent memory leak: `os_mkdir_recurse` returns a uv error
code (int), but it was stored as `bool`, causing `os_strerror` to
receive an invalid error code and leak memory.
See: https://docs.libuv.org/en/v1.x/errors.html#c.uv_strerror
Co-authored-by: Sean Dewar <6256228+seandewar@users.noreply.github.com>
Co-authored-by: Justin M. Keyes <justinkz@gmail.com>
