nushell/crates/nu-plugin-protocol/Cargo.toml

[package]
authors = ["The Nushell Project Developers"]
description = "Protocol type definitions for Nushell plugins"
repository = "https://github.com/nushell/nushell/tree/main/crates/nu-plugin-protocol"
edition = "2021"
license = "MIT"
name = "nu-plugin-protocol"
version = "0.98.0"

[lib]
bench = false

[lints]
workspace = true

[dependencies]
nu-protocol = { path = "../nu-protocol", version = "0.98.0", features = ["plugin"] }
nu-utils = { path = "../nu-utils", version = "0.98.0" }

rmp-serde = { workspace = true }
serde = { workspace = true, features = ["derive"] }
semver = "1.0"
typetag = "0.2"

[features]
default = ["local-socket"]
local-socket = []
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 - :green_circle: `toolkit fmt` - :green_circle: `toolkit clippy` - :green_circle: `toolkit test` - :green_circle: `toolkit test stdlib` 2024-04-27 17:08:12 +00:00			`[package]`
			`authors = ["The Nushell Project Developers"]`
			`description = "Protocol type definitions for Nushell plugins"`
			`repository = "https://github.com/nushell/nushell/tree/main/crates/nu-plugin-protocol"`
			`edition = "2021"`
			`license = "MIT"`
			`name = "nu-plugin-protocol"`
Bump version to `0.98.0` (#13865) 2024-09-18 07:48:46 +00:00			`version = "0.98.0"`
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 - :green_circle: `toolkit fmt` - :green_circle: `toolkit clippy` - :green_circle: `toolkit test` - :green_circle: `toolkit test stdlib` 2024-04-27 17:08:12 +00:00
			`[lib]`
			`bench = false`

Setup global `cargo` lint configuration (#13691) # Description `cargo` somewhat recently gained the capability to store `lints` settings for the crate and workspace, that can override the defaults from `rustc` and `clippy` lints. This means we can enforce some lints without having to actively pass them to clippy via `cargo clippy -- -W ...`. So users just forking the repo have an easier time to follow similar requirements like our CI. ## Limitation An exception that remains is that those lints apply to both the primary code base and the tests. Thus we can't include e.g. `unwrap_used` without generating noise in the tests. Here the setup in the CI remains the most helpful. ## Included lints - Add `clippy::unchecked_duration_subtraction` (added by #12549) # User-Facing Changes Running `cargo clippy --workspace` should be closer to the CI. This has benefits for editor configured runs of clippy and saves you from having to use `toolkit` to be close to CI in more cases. 2024-08-28 21:37:17 +00:00			`[lints]`
			`workspace = true`

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 - :green_circle: `toolkit fmt` - :green_circle: `toolkit clippy` - :green_circle: `toolkit test` - :green_circle: `toolkit test stdlib` 2024-04-27 17:08:12 +00:00			`[dependencies]`
Bump version to `0.98.0` (#13865) 2024-09-18 07:48:46 +00:00			`nu-protocol = { path = "../nu-protocol", version = "0.98.0", features = ["plugin"] }`
			`nu-utils = { path = "../nu-utils", version = "0.98.0" }`
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 - :green_circle: `toolkit fmt` - :green_circle: `toolkit clippy` - :green_circle: `toolkit test` - :green_circle: `toolkit test stdlib` 2024-04-27 17:08:12 +00:00
Remove bincode and use MessagePack instead for plugin custom values (#13745) # Description This changes the serialization of custom values within the plugin protocol to use MessagePack instead of bincode, removing the dependency on bincode entirely. Bincode does not seem to be very maintained anymore, and the externally tagged enum representation doesn't seem to always work now even though it should. Since we use MessagePack already anyway for the plugin protocol, this seems like an obvious choice. This uses the unnamed variant of the serialization rather than the named variant, which is what the plugin protocol in general uses. The unnamed variant does not include field names, which aren't really required here, so this should give us something that's more or less as efficient as bincode is. Should fix #13743. # User-Facing Changes - Will need to recompile plugins (but always do anyway) - Doesn't technically break the plugin protocol (custom value data is a black box / plugin implementation specific), but breaks compatibility between `nu-plugin-engine` and `nu-plugin` so they do need to both be updated to match. # Tests + Formatting All tests pass. # After Submitting - [ ] release notes 2024-09-01 15:33:10 +00:00			`rmp-serde = { workspace = true }`
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 - :green_circle: `toolkit fmt` - :green_circle: `toolkit clippy` - :green_circle: `toolkit test` - :green_circle: `toolkit test stdlib` 2024-04-27 17:08:12 +00:00			`serde = { workspace = true, features = ["derive"] }`
			`semver = "1.0"`
			`typetag = "0.2"`
Fix missing local socket feature (#12698) # Description So sorry to do this during the pre-release freeze, but my plugin crate split PR broke local socket mode, because `nu-plugin-protocol` didn't have the compile feature to advertise the `LocalSocket` protocol feature. This is a very simple, configuration-only bugfix that I think really needs to be merged before the release, or else local socket mode won't work at all. # Tests + Formatting There's an oversight in my testing that caused this to not be caught: the engine really did have the feature, but it just wasn't advertising it, so for `stress_internals` it was still able to use it successfully. Post-release I'll try to make sure this is properly handled somehow. 2024-04-29 07:02:56 +00:00
			`[features]`
			`default = ["local-socket"]`
Bump version to `0.98.0` (#13865) 2024-09-18 07:48:46 +00:00			`local-socket = []`