Problem:
`vim.diagnostic.fromqflist` ignores lines that are `item.valid == 0` (see
`getqflist`). Many qflists have messages that span multiple lines, which look
like this:
collection/src/Modelling/CdOd/Central.hs|496 col 80| error: [GHC-83865]
|| • Couldn't match expected type: InstanceWithForm
|| (FilePath
|| -> SelectValidCdInstWithForm
...
calling `vim.diagnostic.fromqflist(vim.fn.getqflist)` gets a diagnostic message
like this:
error: [GHC-83865]
only the first line is kept, but often, the remaing lines are useful as well.
Solution:
Introduce `merge_lines` option, which "squashes" lines from invalid qflist items
into the error message of the previous valid item, so that we get this
diagnostic message instead:
error: [GHC-83865]
• Couldn't match expected type: InstanceWithForm
(FilePath
-> SelectValidCdInstWithForm
Problem:
- `:restart <cmd>` prepends `-c <cmd>` before the original `-c` args (if
any). So the original `-c` args may "override" it, which is
surprising.
- Confusing logic: `v:argv` is partially prepared in `ex_docmd.c`, and
then later `ui.c` skips other parts of it.
Current behavior is nonsense, for example this sequence:
:restart echo "Hello"
:restart +qall echo "Hello" | echo "World"
results in this v:argv:
[
'nvim'
'-c'
'echo "Hello" | echo "World"'
'--embed'
'-c'
'echo "Hello"'
...
]
Whereas after this commit, v:argv is:
[
'nvim'
'--embed'
...
'-c'
'echo "Hello" | echo "World"'
]
Solution:
- Append `-c <cmd>` at the _end_ of `v:argv`, not the start.
- Use a dummy placeholder `+:::` to mark where the "restart command"
appears in `v:argv`.
- Do all `v:argv` preparation in `ex_docmd.c`. This simplifies `ui.c`.
- Drop `-- [files…]` from `v:argv` since it is probably more annoying
than useful. (Users can use sessions to restore files on restart.)
Problem:
- ~200 line function of hard-to-maintain C code.
- Local Addition section looks messy because of the varying description
formats.
Solution:
- Move code to Lua.
- Have a best-effort approach where short descriptions are right
aligned, giving a cleaner look. Long descriptions are untouched.
Effective use of 'formatprg' requires both an understanding of the
specific capabilities of the formatting tool and Vim's formatting
commands. This is overly burdensome for some users.
Rather than address each complaint on a filetype by filetype basis,
remove 'formatprg' settings from all ftplugins.
It is expected that formatter plugins will be available in the near
future as a better solution. See vim/vim#17145 (Add "formatter" feature using
"compiler" as a template).
Note: 'formatprg' will be removed from older ftplugins after the release
of Vim 9.2. The setting was added to the go and gleam ftplugins during
the current development cycle and have not been included in a Vim
release.
See: vim/vim#18650 (rust.vim: stop setting formatprg to rustfmt)
closes: vim/vim#19108dcc4175284
Co-authored-by: Doug Kearns <dougkearns@gmail.com>
The username/group/netgroup patterns used \l\+ which only matched
lowercase letters. Linux usernames commonly contain hyphens, digits,
and underscores (e.g. www-data, deploy01, test_user).
Update the pattern to \l[-a-z0-9_]* to allow matching the additional
characters "-_" and numbers.
fixes: vim/vim#18963closes: vim/vim#19396a39d7c2617
Co-authored-by: Bozhidar Batsov <bozhidar@batsov.dev>
Problem: Terminal doesn't detect if the PTY process is suspended or
offer a convenient way for the user to resume the process.
Solution: Detect suspended PTY process on SIGCHLD and show virtual text
"[Process suspended]" at the bottom-left. Resume the process
when the user presses a key.
- Amend syntax highlighting to allow for ksh93 discipline function names
(e.g. 'foo.get()') and mksh's odd function naming idiosyncrasies
(shNamespaceOne was introduced to enforce stricter naming rules for
ksh93 namespaces).
- Remove 'bind' from ksh93 syntax (such a builtin has never been
implemented in ksh93).
- 'xgrep' is only available in ksh93v- as an alternative way to
invoke the builtin 'grep -X', so reflect that in the syntax
highlighting.
- Forbid bash-style 'function name() {' syntax when highlighting
ksh88 and ksh93 scripts.
- Fix bug causing ' ()' to be incorrectly validated in mksh scripts.
- Add the many ksh93/ksh2020 .sh.* variables to the list of special
variables.
- Amend iskeyword to allow '.' so that '.sh.tilde.get' and such are
valid function names/variable names. (For mksh functions starting
with odd characters like '%' and '@' this would probably have too
many bad side effects, so I've omitted such a change for that shell.)
- Add new syntax tests and regenerate syntax dump files
closes: vim/vim#1938356033b9df3
Co-authored-by: Johnothan King <johnothanking@protonmail.com>
We add new key exchange algorithms and new enums for PubkeyAuthOptions.
We also add new keywords from sshd_config.5 not present here and remove
keywords present here that are not present in the official
documentation, with the exception of those patched in by Debian and
Fedora, as well as ChallengeResponseAuthentication which is deprecated
but still functional.
closes: vim/vim#1934704c3c6871e
Co-authored-by: Fionn Fitzmaurice <fionn@github.com>
Problem:
Iter:peek() only works if the iterator is a |list-iterator| (internally, an `ArrayIter`).
However, it is possible to implement :peek() support for any iterator.
Solution:
- add `_peeked` buffer for lookahead without actually consuming values
- `peek()` now works for function, pairs(), and array iterators
- `skip(predicate)` stops at the first non matching element without consuming it
- keep existing optimized behavior for `ArrayIter` to maintain backward compatibility
- use `pack`/`unpack` to support iterators that return multiple values
fix(treesitter): more distinctive highlight for EditQuery captures
Problem: EditQuery shows captures in the source buffer using the Title
highlight group, which could be too similar to Normal.
Solution: Use a virtual text diagnostic highlight group: they are
displayed in a similar manner to the query captures so we can assume
that the color scheme should have appropriate styling applied to make
them visible.
Problem:
`vim.json.decode()` could not parse JSONC (JSON with Comments)
extension, which is commonly used in configuration files.
Solution:
Introduce an `skip_comments` option, which is disabled by default. When
enabled, allows JavaScript-style comments within JSON data.
Problem:
Escaping logic for {subject} in ex cmd `:help {subject}` is done in a
messy 200+ lines C function which is hard to maintain and improve.
Solution:
Rewrite in Lua. Use `string.gsub()` instead of looping over characters
to improve clarity and add many more tests to be able to confidently
improve current code later on.
Problem: cterm field in Dict(highlight) is declared as Union(Integer, String)
but it actually expects a Dict(highlight_cterm).
Solution: change cterm type to DictAs(highlight__cterm) and simplify the
handling in dict2hlattrs since type validation and empty array compat are
already handled by api_dict_to_keydict.
In particular, also mention the difference between the regex atom \k and
what Vim considers for a word character.
closes: vim/vim#186889e456e52df
Co-authored-by: Christian Brabandt <cb@256bit.org>
The set_by_lua_block directive of the Lua module takes an additional
variable as an argument which currently breaks the detection of inline
Lua blocks. For example:
set_by_lua_block $myvar {
return tonumber(ngx.var.myothervar)-1
}
closes: vim/vim#19362e92998ea4d
Co-authored-by: Josef Schönberger <josef.schoenberger@tum.de>
Problem:
After eaacdc9, complete with emmylua_ls error with:
runtime/lua/vim/lsp/completion.lua:586: attempt to get length of field
'items' (a nil value)
Solution:
Result can be CompletionItem[] according the spec:
> If a `CompletionItem[]` is provided, it is interpreted to be complete,
> so it is the same as `{ isIncomplete: false, items }`
Problem:
Mappings with a control modifier follow the (tag) format of:
{key_before}_CTRL-{key}_{key_after}
Where `{key_before}` and `{key_after}` can be any other key combination.
However, for the Nvim default mappings `[_CTRL-Q`, `]_CTRL-Q`,
`[_CTRL-L`, `]_CTRL-L`, `[_CTRL-T` and `]_CTRL-T`, the underscore
between the bracket and CTRL was absent. This lead to problems finding
the relevant docs with `:help [_CTRL-Q` and made parsing the {subject}
argument for `:help` harder.
Solution:
Use the right tag format.
Problem: Terminal doesn't handle ED 3 (clear scrollback) properly.
Solution: Add vterm callback for sb_clear().
Also fix another problem that scrollback lines may be duplicated when
pushing to scrollback immediately after reducing window height, as can
be seen in the changes to test/functional/terminal/window_spec.lua.
- Change syntax file maintainer.
- Add Guile and Python command highlighting.
- Update command list to version 12.
- Add foldable regions for the commands 'define', 'if' and 'while'
multiline commands.
- Support documented partial command names.
- Add matchit, browsefilter, and comment formatting support.
- Support embedded C in compiler {code|print} commands.
- Add largely complete settings highlighting and folding.
- Add syntax tests (incomplete).
Thanks to Claudio Fleiner for many years of maintenance.
closes: vim/vim#10649b422a33ac2
Co-authored-by: Doug Kearns <dougkearns@gmail.com>
Problem:
Empty response affects server start boundary computed before.
Solution:
Ignore empty responses. This is mostly micro-optimization that avoids
extending existing results with empty responses.
Problem: Code lenses currently display as virtual text on the same line
and after the relevant item. While the spec does not say how lenses
should be rendered, above the line is most typical. For longer lines,
lenses rendered as virtual text can run off the side of the screen.
Solution: Display lenses as virtual lines above the text.
Closes https://github.com/neovim/neovim/issues/33923
Co-authored-by: Yi Ming <ofseed@foxmail.com>
Problem: The following strchar functions have incorrect types:
strcharlen() - Currently any. Always returns an integer, including on
error
strcharpart() - The skipcc annotation does not specify that 0 and 1 are
valid. These inputs are required for vimscript usage. The current
return type is any, even though the function returns an empty string
on error
strchars() - The skipcc annotation does not specify that 0 and 1 are
valid
Solution: Update the problem types.
Problem: In eval.lua, setcharsearch() has an incorrect param type,
causing Lua_Ls to display an error when a valid table is passed.
While getcharsearch correctly states that it returns a table, the type
is non-specific about the contents.
Solution: Update eval.lua with the correct types.
