named_params

Crates.ionamed_params
lib.rsnamed_params
version1.0.1
sourcesrc
created_at2024-09-26 21:38:45.833929
updated_at2024-09-26 21:38:45.833929
descriptionFast, simple named parameters for Rust functions
homepage
repositoryhttps://github.com/michaeleisel/named_params
max_upload_size
id1388047
size7,572
Michael Eisel (michaeleisel)

documentation

README

named_params

Fast, simple named parameters for Rust functions

Named function parameters have been a much-requested feature for Rust, with various approaches for it. This crate is a proc macro crate that streamlines one particular approach, that of replacing the parameters of a function with a struct that has each parameter as a member (note that optional parameters are not currently supported, although they could be if people want it).

Motivation

Although there are many potential reasons for wanting to use named parameters, this section will focus on one specific example: avoiding misorderings of variables. For some Rust functions, it can be tricky to get the order of parameters right, e.g. if a function takes a large number of parameters that are all of the same type. For example, for the function fn register_user(username: String, password: String, email: String, bio: String, address: String) { ... }, where all 5 parameters have the same type, the type system won't catch misorderings of the parameters. It may be safer to change it to something like the following:

struct RegisterUserArgs {
    username: String,
    password: String,
    email: String,
    bio: String,
    address: String
}

// Call this function with, e.g., register_user(RegisterUserArgs { username, password, email, bio, address });
fn register_user(RegisterUserArgs { username, password, email, bio, address }: RegisterUserArgs) {
    // ...
}

Now whenever someone calls it, they can see very clearly which arguments they pass in correspond to which parameters. The downside of the approach is that it requires a lot of boilerplate. This crate, on the other hand, can do the above with far less boilerplate:

// Gets expanded to match the previous code block
#[named_params]
fn register_user(username: String, password: String, email: String, bio: String, address: String) { ... }

We see that the struct definition can be completely omitted.

The crate also supports references, by tying the lifetime of the reference to that of the args struct:

#[named_params]
fn some_fn(some_str: &str, some_slice: &[u8]) { ... }

gets expanded to:

struct SomeFnArgs<'a> {
    some_str: &'a str,
    some_slice: &'a [u8]
}

fn some_fn(SomeFnArgs { some_str, some_slice }: SomeFnArgs) { ... }

Note that although this crate can be used for low-overhead named parameters, it's important to still follow best practices. If a function has a huge and unwieldy number of parameters, that can be a sign of being a "god function" that should be split up into smaller functions if possible.

License

This crate is licensed under the MIT license.

Commit count: 8

cargo fmt