On linux /dev/stdin is defined as a symlink to /proc/self/fd/0
This in turn is defined as a "magic" symlink which is allowed to point
to internal kernel objects which really does not have a file
name. As a glaring inconsistency, fopen("/proc/self/fd/0", "r")
works if fd was originally opened using pipe() but not using
socketpair(). As it happens UV_CREATE_PIPE does not create pipes
but creates socket pairs. These two unfortunate conditions
means that using /dev/stdin and similar does not work in
shell commands in nvim on linux. as a work around, override
libuv's descicion and create an actual pipe pair.
This change is not needed on BSD:s but done unconditionally for simplicity,
except for on windows where it is not done for stdout because of windows
fixes#35984
Problem: Installing plugin with submodules doesn't check out their
state (due to `git clone --no-checkout` to not end up with default
branch code in case of invalid `version`).
Updating a plugin with submodules doesn't update their state.
Solution: Update `git_checkout` helper to account for submodules.
Another approach would be `git checkout --recurse-submodules ...`,
but that doesn't seem to allow `--filter=blob:none` for submodules,
which is nice to have.
Also make `git_clone` wrapper simpler since `--no-checkout` makes
`--recurse-submodules` and `--also-filter-submodules` do nothing.
Problem:
vim.glob.to_lpeg() errors when patterns contain numeric literals
(like the '1' in '.ps*1') because LPeg interprets numeric strings
as indexed grammar rule references. For example:
vim.glob.to_lpeg('.ps*1')
E5108: Lua: rule '1' undefined in given grammar
Solution:
Prefix all rule names with '_' in the end_seg() function to prevent
literal numbers from being interpreted as LPeg indexed rules. This
ensures pattern components like '1', '2', etc. are treated as
regular rule names rather than special references.
Problem: Installing plugin is done via `git clone --no-checkout ...`
(to not end up with default branch code in case of invalid `version`).
This leaves cloned repo in a state that `git stash` will actually add
an entry to the stash list. Although not critical, better to not have
that if possible.
Solution: explicitly skip `git stash` step in checkout during install.
Problem: The :help command lacks command termination tests.
Solution: Add tests for command termination at "|", "^M" and "^J".
- Check special handling of "|" in arguments.
- Update the Vim syntax file.
closes: vim/vim#18932c4dc4d8f1e
Co-authored-by: Doug Kearns <dougkearns@gmail.com>
Add .rst to suffixesadd. This allows gf and similar commands to work for
rst documentation such as in the linux kernel Documentation.
closes: vim/vim#19149be6225976c
Co-authored-by: Anakin Childerhose <anakin@childerhose.ca>
Co-authored-by: Christian Brabandt <cb@256bit.org>
Problem:
vim.lsp.tagfunc looks for the presence of 'c' (cursor) flag and issues
sync textDocument/definition requests to all clients, otherwise
workspace/symbol requests. But 'c' flag can also be set during the
insert mode completion, e.g. with an empty tag completion query, the tag
func receives pattern of '\<\k\k' with flags 'cir'.
Solution:
check for 'i' (insert mode completion) flag and don't issue any LSP
requests, return vim.NIL for immediate fallback to tags.
Only set fileformat=dos for new files; preserve existing line endings
when editing. This satisfies RFC 5228 for new files while avoiding
issues with version control and existing workflows.
The previous change (3cb4148) unconditionally set fileformat=dos, which
converts existing files with LF line endings to CRLF on save. This
causes issues with version control (entire file appears changed) and
breaks workflows where sieve files are stored with unix line endings.
Dovecot Pigeonhole (the main sieve implementation) has explicitly
accepted LF line endings since 2008 (commit 97b967b5):
/* Loose LF is allowed (non-standard) and converted to CRLF */
This behavior has remained unchanged for almost 18 years.
closes: vim/vim#19144fc00006777
Co-authored-by: André-Patrick Bubel <code@apb.name>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
vim-patch:536ee91: runtime(doc): add termdebug tag, remove term "floating window"
Problem:
- When I type `:h termdebug`, I will expect to see the introduction of
the termdebug plugin. But instead, it shows me document of
`termdebug_wide`, and I have to scroll up quite much to find the
introduction.
- `:h popup` says `floating-window`? Why? As I have tried both features
(of Vim and Neovim), I think they are _very different_ things, even
more different than job features in Vim and Neovim.
Solution:
- In `:h terminal.txt`, add tag `*termdebug*` to the introduction of
termdebug plugin.
- In `:h popup.txt`, "floating window" -> "popup window".
closes: vim/vim#19135536ee91189
Change the title ":Termdebug plugin" to simply ":Termdebug", otherwise
the line is too long.
Co-authored-by: Phạm Bình An <phambinhanctb2004@gmail.com>
Problem:
The `"Running healthchecks..."` message doesn't inform the user much and
is a hack from before we got a way to emit actual progress messages.
Solution:
Use `nvim_echo` to emit progress messages showing the name of the report
that is currently running.
Problem: Setting pumborder=shadow doesn't blend proerly.
Solution: Check fconfig.shadow when setting pum_grid.blending, like what
is done in check_blending().
Problem:
ICCF Holland is dissolved by the end of 2025 ^1 and sponsorships are
transferred to Kuwasha (https://kuwasha.net). Their SSL certificate is
already expired, so the https URL mentioned in the intro text doesn't
work anymore. Reported by https://github.com/neovim/neovim/issues/36597#issue-3635391949.
Solution:
URL is removed from the text. We'll keep the vim.org URL for now as it
points to the same information.
[^1]: See June 2025 news on https://iccf-holland.org/index.html
This brings us to upstream commit f251a43 (Merge syntax based folding
for block comments, 2025-09-10). Note that not all upstream files are
included.
closes: vim/vim#19126a71e52a0b1
Co-authored-by: D. Ben Knoble <ben.knoble+github@gmail.com>
When a list item contains a mapping key (e.g., '- element1:'), the
content under that key was incorrectly indented. The indent function
was not accounting for the '- ' prefix when calculating indentation
for nested content.
Example that now works correctly:
list:
- element1:
foo: bar # Now correctly at indent 6, not 4
The fix adds special handling in two places:
1. When previous line ends with ':' and starts with '- '
2. When looking up previous mapping key that is a list item
Fixes indentation to account for the 2-character '- ' prefix.
fixes: vim/vim#18943closes: vim/vim#191339179ddc060
Co-authored-by: Cezar Dimoiu <cezar.dimoiu@keysight.com>
Problem:
Can't use `:source` to run a Lua codeblock (treesitter injection) in
a help (vimdoc) file.
Solution:
Use treesitter to parse the range and treat it as Lua if detected as
such.
This commit updates `syntax/cangjie.vim` to match the latest `std.core`
library:
New Features:
* Documentation: Added highlighting for documentation keywords (e.g., `@param`, `@return`).
* Standard Library: Added highlighting for `std.core` functions, interfaces, and classes.
* Exceptions: Added highlighting for standard exception types.
* FFI Support: Added highlighting for C interoperability types (e.g., `CPointer`).
Improvements:
* Configuration: Added a unified switch (`s:enabled('builtin')`) for standard library highlighting.
* Type System: Updated `Int` and `UInt` aliases.
* Interpolation: Enabled standard library highlighting inside string interpolation.
* Cleanup: Removed non-core types like `ArrayList` and `HashMap`.
closes: vim/vim#1908577a59be594
Co-authored-by: Neila <wu.junkai@qq.com>
Problem
LSP Related Information line and column numbers are 0-based. Displaying
them this way can confuse the user, since vim line/col numbers are
typically displayed 1-based.
Solution
Display the line and column numbers as 1-based.
* fix(vim-patch): test_codestyle.vim is NA
Neovim has make lint, specifically clint.lua.
Unless vim9jit, https://github.com/tjdevries/vim9jit, or similar tool
compiles vim9 "test" code into lua, it's futile to port it.
* fix(vim-patch): NA files from dev_vimpatch.txt
- channel
- json
- job
- term
- terminal
Nvim has its own incompatible implementation or uses 3rd-party
libraries (ie. libuv, libtermkey).
* fix(vim-patch): term.c is applicable for replace_termcodes()
* fix(vim-patch/na): blowfish,dlldata,iid_ole,nbdebug,os_w32
blowfish is for encryption.
dlldata,iid_ole,os_w32 are for Windows (GUI, registry, entry point).
Nvim doesn't need them to be installed on Windows.
nbdebug.[ch] is for Netbeans Debugging Tools.
They're N/A.
* fix(vim-patch): runtime/
Nvim has its own documentation for development
- runtime/doc/debug.txt
- runtime/doc/develop.txt
N/A since the beginning:
- runtime/doc/os_haiku.txt
- runtime/doc/pi_logipat.txt
- runtime/doc/quotes.txt
- runtime/plugin/logiPat.vim
- runtime/plugin/manpager.vim
- runtime/tools/demoserver.py
- runtime/spell/.*\.latin1
Vim9 is N/A.
- runtime/doc/os_haiku.txt
* fix(vim-patch): src/
win32yank replaces "src/winclip.c".
Vim's desktop files are N/A.
Nvim removed 'termencoding'.
Nvim doesn't use Vim's lsan-suppress.txt to configures LSAN.
Add documentation notes explaining that visual marks '< and '> have
different behaviors in getpos() and setpos().
Also fix a small typo.
closes: vim/vim#1907093eb081eee
Co-authored-by: Larson, Eric <numeric.larson@gmail.com>
