Are some of the Haskell idioms of tacit programming translatable to Rust?
You can try to build a macro for that:
#[feature(macro_rules)];
macro_rules! compose_inner(
($var:ident, $f:ident) => (
$f($var)
);
($var:ident, $f:ident $($rest:ident )+) => (
$f(compose_inner!($var, $($rest )+))
);
)
macro_rules! compose(
($($f:ident )+) => (
|x| compose_inner!(x, $($f )+)
)
)
fn foo(x: int) -> f64 {
(x*x) as f64
}
fn bar(y: f64) -> ~str {
(y+2.0).to_str()
}
fn baz(z: ~str) -> ~[u8] {
z.into_bytes()
}
fn main() {
let f = compose!(baz bar foo);
println!("{:?}", f(10));
}
The macros probably could be simpler, but that's what I came up with.
But it certainly is not supported in the language itself. Rust is not a functional nor concatenative language after all.
Very similar idiom is method chaining, which is absolutely supported by Rust. The most prominent example, I think, would be iterator transformations:
let v: ~[int] = ...;
let x: int = v.iter().map(|x| x + 1).filter(|x| x > 0).fold(0, |acc, x| acc + x/2);
True, it is not as flexible as arbitrary functions composition, but it looks much more natural and feels much more convenient.