So I'm looking at the impl for is_some()
for Option
and I noticed that it uses match *self {}
under the hood...so it moves it internally.
My question is, if it gets moved, how am I able to do something like this? https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=f094da12290b77bad526674467e51043
fn main() {
let x = Option::Some(3);
x.is_some();
x.is_some();
}
My expectation is that I should be able to call is_some()
only once, and the next time I call it I should get some sort of error saying it's been moved...but no, it all compiles fine.
What am I misunderstanding?
*self
in match *self { ... }
does not move (or copy) what self
points to. From " The Rust Reference " (emphasis mine),
A
match
behaves differently depending on whether or not the scrutinee expression is a place expression or value expression. If the scrutinee expression is a value expression, it is first evaluated into a temporary location, ...When the scrutinee expression is a place expression, the match does not allocate a temporary location; however, a by-value binding may copy or move from the memory location. ...
*self
is a place expression. From " The Rust Reference " (emphasis mine),
Expressions are divided into two main categories: place expressions and value expressions. ...
A place expression is an expression that represents a memory location. These expressions are paths which refer to local variables, static variables, dereferences (
*expr
) , array indexing expressions (expr[expr]
), field references (expr.f
) and parenthesized place expressions. All other expressions are value expressions.A value expression is an expression that represents an actual value.
You might also be interested to know that the Some(_) => true
arm in the match
body does not bind anything. From " The Rust Reference ",
Unlike identifier patterns, it does not copy, move or borrow the value it matches.
where "it" means the wildcard pattern ( _
).
(See the @dkim's answer for a more official and cited answer)
If the signature of the function accepts by reference, it does not take ownership of the value. Option.is_some() actually takes &self not self.
The interesting part is how *self
is allowed to be used in a function that receives &self
when Self: Copy
is not bounded.
To test this, let's create a minimal example containing something similar: https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=d7b137b74b5cd8f8bb57398ae01bf4e3
#[derive(Debug)]
pub enum A {
X(String),
Y(i32),
}
pub fn f(a: &A) {
match *a {
A::X(_) => {
// dbg!(s);
}
A::Y(_i) => {
// dbg!(i);
}
};
}
This compiles fine. But let's change the A::X(_)
pattern to A::X(_s)
: https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=93030e5c3f84532532c5db966c798bd6
#[derive(Debug)]
pub enum A {
X(String),
Y(i32),
}
pub fn f(a: &A) {
match *a {
A::X(_s) => {
// dbg!(s);
}
A::Y(_i) => {
// dbg!(i);
}
};
}
This fails to compile:
error[E0507]: cannot move out of `a.0` which is behind a shared reference
--> src/lib.rs:7:11
|
7 | match *a {
| ^^ help: consider borrowing here: `&*a`
8 | A::X(_s) => {
| --
| |
| data moved here
| move occurs because `_s` has type `std::string::String`, which does not implement the `Copy` trait
So it appears that dereferencing a non-Copy enum is perfectly fine, as long as it wouldn't be used for moving inner non-Copy values. _
is fine because it guarantees that the underlying value would never be used, while _s
does not compile because it is just a normal allow(unused) variable.
This also makes sense because it allows the same match arm to work on both Copy and non-Copy types, as long as no usages violate the ownership rules
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.