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
* refactor(fold): update comment formatting
* refactor(fold): use post-increment/decrement
* refactor(fold): reduce scope of local variables
* refactor(fold): use booleans
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.
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.
Consolidate commit author details
- Prefer user/email with either most commits or most recent commit
- Generally less preference for @users.noreply.github.com emails
- Avoided consolidating common names
Don't add sign extmarks to state->active. Instead when drawing signs,
perform a full line scan for sign marks. This allows decor_redraw_line
to be moved back inside the `!number_only` block in screen.c, which
prevents decor scans when redrawing the number column when
'relativenumber' is set.
Fixes: #17638
Previously b_signcols was invalidated whenever a sign was added/removed
or when a buffer line was added/removed.
This change introduces a sentinel linenr_T into the buffer state which
is a line number used to determine the signcolumn. With this
information, we can invalidate the signcolumn less often. Now the
signcolumn is only invalidated when a sign or line at the sentinel line
number is removed.
Add the following options to extmarks:
- sign_text
- sign_hl_group
- number_hl_group
- line_hl_group
- cursorline_hl_group
Note: ranges are unsupported and decorations are only applied to
start_row
50250542 failed to consider that the maximum passed to buf_signcols
is window scoped whereas the signcols value is buffer scoped. This can
lead to a bug where the signcolumn becomes incorrect if:
- global signcolumn is set to auto:N
- signcolumn in a window is changed locally to auto:M where M > N
- the buffer has a line with M or greater signs.
Neovim currently places its own loader for searching runtime files at
the front of `package.loaders`. This prevents any preloaders in
`package.preload` from being used. This change fixes that by moving the
default package preloader to run before Neovim's loader. For example,
LuaJIT provides preloaders for the built-in modules `ffi` and `bit`, so
this optimisation will improve the loading of those.
When trying to load a language parser, escape the value of
the language.
With language injection, the language might be picked up from the
buffer. If this value is erroneous it can cause `nvim_get_runtime_file`
to hard error.
E.g., the markdown expression `~~~{` will extract '{' as a language and
then try to get the parser using `parser/{*` as the pattern.
vim.bo can target a specific buffer by indexing with a number, e.g:
`vim.bo[2].filetype` can get/set the filetype for buffer 2. This change
replicates that behaviour for the variable namespace.
* feat(api): add lua C bindings for xdiff
* chore: opt.hunk_lines -> opt.result_type
opt.on_hunk now takes precedence over opt.result_type
* chore: fix indents
Fix indents
* chore: change how priv is managed
Assign priv NULL and unconditionally apply XFREE_CLEAR to it when
finished.
This bumps the bundled LibLUV to 1.34.1 in order to make the newly introduced `uv.fs_mkstemp` available in the official builds. (The minimum required version remains at 1.30.1.)
The bump of lua-compat-53 to 0.9 is needed to build with the updated LibLUV.
...signs is slow
Problem: Placing a larger number of signs is slow.
Solution: Add functions for dealing with a list of signs. (Yegappan
Lakshmanan, closes#4636)
+ support for neovim's dynamic width signcolumn
Problem: Displaying signs is inefficient.
Solution: Avoid making multiple calls to get information about a placed
sign. (Yegappan Lakshmanan, closes#4586)