Commit graph

9 commits

Author SHA1 Message Date
Devyn Cairns
21ebdfe8d7
Bump version to 0.93.1 (#12710)
# Description

Next patch/dev release, `0.93.1`
2024-05-01 17:19:20 -05:00
Devyn Cairns
3b220e07e3
Bump version to 0.93.0 (#12709)
# Description

Bump version to `0.93.0`
2024-04-30 15:51:13 -07:00
Devyn Cairns
0c4d5330ee
Split the plugin crate (#12563)
# Description

This breaks `nu-plugin` up into four crates:

- `nu-plugin-protocol`: just the type definitions for the protocol, no
I/O. If someone wanted to wire up something more bare metal, maybe for
async I/O, they could use this.
- `nu-plugin-core`: the shared stuff between engine/plugin. Less stable
interface.
- `nu-plugin-engine`: everything required for the engine to talk to
plugins. Less stable interface.
- `nu-plugin`: everything required for the plugin to talk to the engine,
what plugin developers use. Should be the most stable interface.

No changes are made to the interface exposed by `nu-plugin` - it should
all still be there. Re-exports from `nu-plugin-protocol` or
`nu-plugin-core` are used as required. Plugins shouldn't ever have to
use those crates directly.

This should be somewhat faster to compile as `nu-plugin-engine` and
`nu-plugin` can compile in parallel, and the engine doesn't need
`nu-plugin` and plugins don't need `nu-plugin-engine` (except for test
support), so that should reduce what needs to be compiled too.

The only significant change here other than splitting stuff up was to
break the `source` out of `PluginCustomValue` and create a new
`PluginCustomValueWithSource` type that contains that instead. One bonus
of that is we get rid of the option and it's now more type-safe, but it
also means that the logic for that stuff (actually running the plugin
for custom value ops) can live entirely within the `nu-plugin-engine`
crate.

# User-Facing Changes
- New crates.
- Added `local-socket` feature for `nu` to try to make it possible to
compile without that support if needed.

# Tests + Formatting
- 🟢 `toolkit fmt`
- 🟢 `toolkit clippy`
- 🟢 `toolkit test`
- 🟢 `toolkit test stdlib`
2024-04-27 12:08:12 -05:00
Devyn Cairns
25cbcb511d
Rename plugin cache file ⇒ plugin registry file (#12634)
# Description
So far this seems like the winner of my poll on what the name should be.
I'll take this off draft once the poll expires, if this is indeed the
winner.
2024-04-24 17:40:39 -05:00
Ian Manske
9996e4a1f8
Shrink the size of Expr (#12610)
# Description
Continuing from #12568, this PR further reduces the size of `Expr` from
64 to 40 bytes. It also reduces `Expression` from 128 to 96 bytes and
`Type` from 32 to 24 bytes.

This was accomplished by:
- for `Expr` with multiple fields (e.g., `Expr::Thing(A, B, C)`),
merging the fields into new AST struct types and then boxing this struct
(e.g. `Expr::Thing(Box<ABC>)`).
- replacing `Vec<T>` with `Box<[T]>` in multiple places. `Expr`s and
`Expression`s should rarely be mutated, if at all, so this optimization
makes sense.

By reducing the size of these types, I didn't notice a large performance
improvement (at least compared to #12568). But this PR does reduce the
memory usage of nushell. My config is somewhat light so I only noticed a
difference of 1.4MiB (38.9MiB vs 37.5MiB).

---------

Co-authored-by: Stefan Holderbach <sholderbach@users.noreply.github.com>
2024-04-24 15:46:35 +00:00
Devyn Cairns
b576123b0a
Accept filenames in other plugin management commands (#12639)
# Description

This allows the following commands to all accept a filename instead of a
plugin name:

- `plugin use`
- `plugin rm`
- `plugin stop`

Slightly complicated because of the need to also check against
`NU_PLUGIN_DIRS`, but I also fixed some issues with that at the same
time

Requested by @fdncred

# User-Facing Changes

The new commands are updated as described.

# Tests + Formatting

Tests for `NU_PLUGIN_DIRS` handling also made more robust.

- 🟢 `toolkit fmt`
- 🟢 `toolkit clippy`
- 🟢 `toolkit test`
- 🟢 `toolkit test stdlib`

# After Submitting

- [ ] Double check new docs to make sure they describe this capability
2024-04-24 06:28:45 -05:00
Devyn Cairns
1f4131532d
Deprecate register and add plugin use (#12607)
# Description

Adds a new keyword, `plugin use`. Unlike `register`, this merely loads
the signatures from the plugin cache file. The file is configurable with
the `--plugin-config` option either to `nu` or to `plugin use` itself,
just like the other `plugin` family of commands. At the REPL, one might
do this to replace `register`:

```nushell
> plugin add ~/.cargo/bin/nu_plugin_foo
> plugin use foo
```

This will not work in a script, because `plugin use` is a keyword and
`plugin add` does not evaluate at parse time (intentionally). This means
we no longer run random binaries during parse.

The `--plugins` option has been added to allow running `nu` with certain
plugins in one step. This is used especially for the `nu_with_plugins!`
test macro, but I'd imagine is generally useful. The only weird quirk is
that it has to be a list, and we don't really do this for any of our
other CLI args at the moment.

`register` now prints a deprecation parse warning.

This should fix #11923, as we now have a complete alternative to
`register`.

# User-Facing Changes

- Add `plugin use` command
- Deprecate `register`
- Add `--plugins` option to `nu` to replace a common use of `register`

# Tests + Formatting

I think I've tested it thoroughly enough and every existing test passes.
Testing nu CLI options and alternate config files is a little hairy and
I wish there were some more generic helpers for this, so this will go on
my TODO list for refactoring.

- 🟢 `toolkit fmt`
- 🟢 `toolkit clippy`
- 🟢 `toolkit test`
- 🟢 `toolkit test stdlib`

# After Submitting

- [ ] Update plugins sections of book
- [ ] Release notes
2024-04-23 06:37:50 -05:00
Devyn Cairns
5c7f7883c8
Add ErrSpan extension trait for Result (#12626)
# Description
This adds an extension trait to `Result` that wraps errors in `Spanned`,
saving the effort of calling `.map_err(|err| err.into_spanned(span))`
every time. This will hopefully make it even more likely that someone
will want to use a spanned `io::Error` and make it easier to remove the
impl for `From<io::Error> for ShellError` because that doesn't have span
information.
2024-04-23 10:39:55 +02:00
Devyn Cairns
2595f31541
Overhaul the plugin cache file with a new msgpack+brotli format (#12579)
# Description

- Plugin signatures are now saved to `plugin.msgpackz`, which is
brotli-compressed MessagePack.
- The file is updated incrementally, rather than writing all plugin
commands in the engine every time.
- The file always contains the result of the `Signature` call to the
plugin, even if commands were removed.
- Invalid data for a particular plugin just causes an error to be
reported, but the rest of the plugins can still be parsed

# User-Facing Changes

- The plugin file has a different filename, and it's not a nushell
script.
- The default `plugin.nu` file will be automatically migrated the first
time, but not other plugin config files.
- We don't currently provide any utilities that could help edit this
file, beyond `plugin add` and `plugin rm`
  - `from msgpackz`, `to msgpackz` could also help
- New commands: `plugin add`, `plugin rm`

# Tests + Formatting

Tests added for the format and for the invalid handling.

- 🟢 `toolkit fmt`
- 🟢 `toolkit clippy`
- 🟢 `toolkit test`
- 🟢 `toolkit test stdlib`

# After Submitting

- [ ] Check for documentation changes
- [ ] Definitely needs release notes
2024-04-21 07:36:26 -05:00