Crates.io | flat_enum |
lib.rs | flat_enum |
version | 0.1.0 |
source | src |
created_at | 2024-01-22 12:42:10.383527 |
updated_at | 2024-01-22 12:42:10.383527 |
description | Expand nested enum into flattened enum |
homepage | |
repository | https://github.com/yasuo-ozu/flat_enum |
max_upload_size | |
id | 1108740 |
size | 5,648 |
This crate expands nesting enums. See the example:
# use flat_enum::{flat, into_flat, FlatTarget};
#[derive(FlatTarget)]
pub enum Enum1<A> {
E1(A),
E2(),
E3(String),
}
#[into_flat(Enum2Flat<A>)]
pub enum Enum2<A> {
#[flatten]
Enum1(Enum1<A>),
E4,
}
#[flat(Enum2<A>)]
pub enum Enum2Flat<A> {}
In macro invocation, the Enum2Flat
expands into something like:
pub enum Enum2Flat<A> {
E1(A),
E2(),
E3(String),
E4,
}
In this example, Enum1
and Enum2
are not required to be defined in the same crate. But Enum2
and Enum2Flat
should be defined in the same context (module).
In Rust's enum representation on memory, we have std::mem::Disctiminant
value in addition to the field values of each variants. If two enums are nesting, it should have two discriminants on memory. The compiler's optimization algorithm does not do such work.
This crate gives a way to generate flattened enum automatically to deal with the problem.
When using a value of nested enum types in match-like expression, the matchers are easily to become complex. The flattened enum solves that.