Problem:
The Lua test harness still ran through standalone -ll mode, so tests
depended on the low-level Lua path instead of the regular Nvim Lua
environment. That also meant os.exit() coverage had to carry an ASAN
workaround because Lua's raw process exit skipped Nvim teardown and let
LeakSanitizer interfere with the observed exit code.
Solution:
Run the harness and related fixtures with nvim -l. Patch os.exit() in
the main Lua state to exit through getout(), so scripts observe normal
Nvim shutdown while standalone -ll remains available for generator-style
scripts. As a consequence, the startup test can assert os.exit() without
disabling leak detection.
AI-assisted: Codex
Both for tests and for system wide install, $PREFIX/lib/nvim/parser
is a valid path for tree-sitter parsers. This also brings the build.zig
behavior in line with how we set up the paths in CMakeLists.txt
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
If the Nvim session has exited, the nvim_command will fail too.
ERROR test/functional/core/fileio_spec.lua @ 342: tmpdir failure modes
test\functional\testnvim.lua:133: sending request after EOF from Nvim
stack traceback:
test\functional\testnvim.lua:133: in function 'command'
test\functional\testnvim.lua:847: in function 'rmdir'
test/functional/core\fileio_spec.lua:353: in function <test/functional/core\fileio_spec.lua:342>
Problem: Fake LSP server does not timeout or respond to SIGTERM as it
does not run the event loop.
Solution: Instead of io.read(), use stdioopen()'s on_stdin callback to
accumulate input and use vim.wait() to wait for input.
Also, in the test suite, don't stop a session when it's not running, as
calling uv.stop() outside uv.run() will instead cause the next uv.run()
to stop immediately, which cancels the next RPC request.
Problem:
Temporary files from /tmp/ and /private/ paths clutter :oldfiles list.
Additionally, the documented Windows default (rA:,rB:) was never applied
due to a missing platform condition.
Solution:
Drop platform-specific shada differences and default to excluding
/tmp/ and /private/ paths.
Problem:
Using vim.defer_fn() just before Nvim exit leaks luv handles.
Solution:
Make vim.schedule() return an error message if scheduling failed.
Make vim.defer_fn() close timer if vim.schedule() failed.
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:
Every CI log has a lot of noise at the end, which makes it harder to
find relevant test failures:
Running tests from test/functional/terminal/tui_spec.lua
...
T5831 TUI bg color queries the terminal for background color:
T5832 TUI bg color triggers OptionSet from automatic background processing:
T5833 TUI bg color sends theme update notifications when background changes #31652:
...
Running tests from test/functional/ui/output_spec.lua
...
WRN 2025-12-02T03:36:47.304 ui/c/T5831.28003.0 tui_handle_term_mode:223: TUI: terminal mode 2026 unavailable, state 0
WRN 2025-12-02T03:36:47.359 ui/c/T5832.28006.0 tui_handle_term_mode:223: TUI: terminal mode 2048 unavailable, state 0
WRN 2025-12-02T03:36:47.414 ui/c/T5833.28009.0 tui_handle_term_mode:223: TUI: terminal mode 2048 unavailable, state 0
Solution:
- Skip logging in test-mode.
- This can be reverted later, when these logs are changed to "INFO"
level, per this TODO comment:
```
// TODO(bfredl): This is really ILOG but we want it in all builds.
// add to show_verbose_terminfo() without being too racy ????
WLOG("TUI: terminal mode %d unavailable, state %d", mode, state);
```
NEW BUILD SYSTEM!
This is a MVP implementation which supports building the "nvim" binary,
including cross-compilation for some targets.
As an example, you can build a aarch64-macos binary from
an x86-64-linux-gnu host, or vice versa
Add CI target for build.zig currently for functionaltests on linux
x86_64 only
Follow up items:
- praxis for version and dependency bumping
- windows 💀
- full integration of libintl and gettext (or a desicion not to)
- update help and API metadata files
- installation into a $PREFIX
- more tests and linters
Problem:
Default 'statusline' is implemented in C and not representable as
a statusline expression. This makes it hard for user configs/plugins to
extend it.
Solution:
- Change the default 'statusline' slightly to a statusline expression.
- Remove the C implementation.
Problem:
`feed_command()` was added as a "bridge" for old test code. 99% of those
cases should be using `n.command()`, which raises errors instead of
silently continuing the test.
Solution:
Deprecate `feed_command()`. It should not be used in new tests.
All usages of `feed_command()` should be converted to `command()` or
`feed()`.
Problem:
Indenting text is a common task in plugins/scripts for
presentation/formatting, yet vim has no way of doing it (especially
"dedent", and especially non-buffer text).
Solution:
Introduce `vim.text.indent()`. It sets the *exact* indentation because
that's a more difficult (and thus more useful) task than merely
"increasing the current indent" (which is somewhat easy with a `gsub()`
one-liner).
Problem:
- `n.spawn()` is misleading because it also connects RPC, it's not just
"spawning" a process.
- It's confusing that `n.spawn()` and `n.spawn_argv()` are separate.
Solution:
- Replace `n.spawn()`/`n.spawn_argv()` with a single function `n.new_session()`.
This name aligns with the existing functions `n.set_session`/`n.get_session`.
- Note: removes direct handling of `prepend_argv`, but I doubt that was
important or intentional. If callers want to control use of `prepend_argv`
then we should add a new flag to `test.session.Opts`.
- Move `keep` to first parameter of `n.new_session()`.
- Add a `merge` flag to `test.session.Opts`
- Mark `_new_argv()` as private. Test should use clear/new_session/spawn_wait
instead.
Problem:
Tests that need to check `nvim` CLI behavior (no RPC session) create
their own ad-hoc `system()` wrappers.
Solution:
- Use `n.spawn_wait` instead of `system()`.
- Bonus: this also improves the tests by explicitly checking for
`stdout` or `stderr`. And if a signal is raised, `ProcStream.status`
will reflect it.
Problem:
Can't use `n.clear()` to test non-RPC `nvim` invocations. So tests end
up creating ad-hoc wrappers around `system()` or `jobstart()`.
Solution:
- Introduce `n.spawn_wait()`
- TODO (followup PR): Rename `n.spawn()` and `n.spawn_wait()`.
It's misleading that `n.spawn()` returns a RPC session...
Similar to how there is a `GDB` environment variable to let the nvim
test instances to be run under `gdbserver` this adds a `OSV_PORT`
variable to start nvim test instances with `osv` in blocking mode to let
a debug client attach to it for debugging of `exec_lua` code blocks.
By making it a separate module, the embedded Nvim session can require
this module directly instead of setup code sending over the module via
RPC.
Also make exec_lua wrap _G.print so messages can be seen
in the test output immediately as the exec_lua returns.
Problem:
When tmpdir is local. The returned values from tmpname may already
exist. This can cause problems for tests which pass `create=false` as
they may require the file to not exist yet.
Solution:
When creating tmp names, always remove it to ensure it doesn't exist,
and optionally open it if `create~=false`
Additionally refactor the tmpname code and flattrn some functions into
constants.
Also while debugging this issue. It was discovered that `exec_lua()`
doesn't report error messages properly. This has been fixed.
Problem:
The default builtin UI client does not declare its client info. This
reduces discoverability and makes it difficult for plugins to identify
the UI.
Solution:
- Call nvim_set_client_info after attaching, as recommended by `:help dev-ui`.
- Also set the "pid" field.
- Also change `ui_active()` to return a count. Not directly relevant to
this commit, but will be useful later.
This module is generally used by any tests that need the full Nvim TUI
instead of `screen.lua`. Thus it should live in `functional/` instead of
in `functional/terminal/`.
- `alter_slashes` belongs in `testutil.lua`, not `testnvim.lua`.
- `alter_slashes` is an unusual name. Rename it to `fix_slashes`.
- invert its behavior, to emphasize that `/` slashes are the preferred,
pervasive convention, not `\` slashes.
Problem:
If $NVIM_APPNAME is a relative dir path, Nvim fails to start its
primary/default server, and `v:servername` is empty.
Root cause is d34c64e342, but this wasn't
noticed until 96128a5076 started reporting the error more loudly.
Solution:
- `server_address_new`: replace slashes "/" in the appname before using
it as a servername.
- `vim_mktempdir`: always prefer the system-wide top-level "nvim.user/"
directory. That isn't intended to be specific to NVIM_APPNAME; rather,
each *subdirectory* ("nvim.user/xxx") is owned by each Nvim instance.
Nvim "apps" can be identified by the server socket(s) stored in those
per-Nvim subdirs.
fix#30256
Problem:
`'scrollbind'` does not work properly if the window being scrolled
automatically contains any filler/virtual lines (except for diff filler
lines).
This is because when the scrollbind check is done, the logic only
considers changes to topline which are represented as line numbers.
Solution:
Write the logic for determine the scroll amount to take into account
filler/virtual lines.
Fixes#29751
Problem:
- The test for vim.deprecate() has a "mock" which is outdated because
vim.deprecate() no longer uses that.
- The tests get confused after a version bump.
Solution:
Make the tests adapt to the current version.
Specifically, functions that are run in the context of the test runner
are put in module `test/testutil.lua` while the functions that are run
in the context of the test session are put in
`test/functional/testnvim.lua`.
Closes https://github.com/neovim/neovim/issues/27004.
