Parameters can be passed to functions and modified:
fn set_42(int: &mut i32) {
*int += 42;
}
fn main() {
let mut int = 0;
set_42(&mut int);
println!("{:?}", int);
}
Output:
42
Changing the code to use a slice fails with a whole bunch of errors:
fn pop_front(slice: &mut [i32]) {
*slice = &{slice}[1..];
}
fn main() {
let mut slice = &[0, 1, 2, 3][..];
pop_front(&mut slice);
println!("{:?}", slice);
}
Output:
error[E0308]: mismatched types
--> src/main.rs:2:14
|
2 | *slice = &{ slice }[1..];
| ^^^^^^^^^^^^^^^
| |
| expected slice `[i32]`, found `&[i32]`
| help: consider removing the borrow: `{ slice }[1..]`
error[E0277]: the size for values of type `[i32]` cannot be known at compilation time
--> src/main.rs:2:5
|
2 | *slice = &{ slice }[1..];
| ^^^^^^ doesn't have a size known at compile-time
|
= help: the trait `std::marker::Sized` is not implemented for `[i32]`
= note: to learn more, visit <https://doc.rust-lang.org/book/ch19-04-advanced-types.html#dynamically-sized-types-and-the-sized-trait>
= note: the left-hand-side of an assignment must have a statically known size
If we try using a mutable slice (which isn't what I really want; I don't want to modify the values within the slice, I just want to modify the slice itself so it covers a smaller range of elements) and a mutable parameter, it has no effect on the original slice:
fn pop_front(mut slice: &mut [i32]) {
slice = &mut {slice}[1..];
}
fn main() {
let mut slice = &mut [0, 1, 2, 3][..];
pop_front(&mut slice);
println!("{:?}", slice);
}
Output:
[0, 1, 2, 3]
Is there a way to modify a slice that's a function parameter? I don't want to modify the elements within the slice; I just want to modify the range of the slice itself so it becomes a smaller "sub-slice".
As others have said, the core idea here is to take a &mut &... [T] (where ... is mut or empty) and read/write to the internal slice. The other answers demonstrate it is possible for &mut &[T] in safe code, and possible for &mut &mut [T] with unsafe code, but they don't explain why there's the difference... and &mut &mut [T] is possible with safe code too.
In explicit-lifetime terms, the nested reference is something like &'a mut &'b ... [T] for some lifetimes 'a and 'b, and the goal here is to get a &'b ... [T], slice it and write that into the &'a mut.
For &'a mut &'b [T], this is easy: &[T] is copy, so writing *slice = &slice[1..] will effectively copy the &'b [T] out of the &mut and then, later, overwrite the existing value with the shorter one. The copy means that one literally gets a &'b [T] to operate with, and so there's no direct connection between that and the &'a mut, and hence it is legal to mutate. It is effectively something like
fn pop_front<'a, 'b>(slice: &'a mut &'b[i32]) {
// *slice = &slice[1..] behaves like
let value: &'b [i32] = *slice;
*slice = &value[1..]
}
(I've labelled the lifetimes and annotated the type to tie into my explanation, but this is not required for the code to work.)
For &'a mut &'b mut [T] things are a little trickier: &mut [T] cannot be copied: dereferencing won't copy, it will reborrow to give a &'a mut [T] i.e. the slice has a lifetime that is connected to the outer &'a mut, not the inner &'b mut [T]. This means the sliced reference has a shorter lifetime than the type it is trying to overwrite, so it's invalid to store the slice into that position. In other words:
fn pop_front<'a, 'b>(slice: &'a mut &'b mut [i32]) {
let value: &'a mut [i32] = &mut **slice;
*slice = &mut value[1..] // error
}
The way to do this safely for &'a mut &'b mut [T] is to get the internal slice out of the reference with that 'b lifetime. This requires keeping track of the "one owner" rule, doing no borrowing, and the best function for this sort of ownership manipulation is mem::replace. It allows us to extract the inner &'b mut [T] by swapping it with some placeholder, which we can then overwrite with the short version. The best/only placeholder is an empty array: writing &mut [] can be a &'c mut [X] for any type X and any lifetime 'c, since there's no data to store and so nothing needs initialisation, and no data will ever become invalid. In particular, it can be a &'b mut [T]:
fn pop_front<'a, 'b>(slice: &'a mut &'b mut [i32]) {
let value: &'b mut [i32] = mem::replace(slice, &mut []);
*slice = &mut value[1..]
}
Since &mut[T] implements Default, we can also use mem::take:
fn pop_front<'a, 'b>(slice: &'a mut &'b mut [i32]) {
let value: &'b mut [i32] = mem::take(slice);
*slice = &mut value[1..]
}
(As above, I've made things more explicit than necessary.)
See also:
Why can't I assign one dereference of a reference of a reference to another when the outer lifetimes differ?
Why does linking lifetimes matter only with mutable references?
How can I swap in a new value for a field in a mutable reference to a structure?
If you need to modify an immutable slice, see Cornstalks's answer.
You cannot modify a mutable slice in safe Rust. When you take a subslice of a mutable slice, you effectively borrow from the original slice. This means that the subslice must not outlive the original slice.
You want something that looks like this:
fn pop_front(slice: &mut &mut [i32]) {
*slice = &mut slice[1..];
}
but the subslice slice[1..] is only valid until the end of the function, and which point the borrow will end and the original slice (the slice parameter) will be usable again.
We can use some unsafe code to construct manually the slice we want:
use std::slice;
fn pop_front(slice: &mut &mut [i32]) {
let ptr = slice.as_mut_ptr();
let len = slice.len();
*slice = unsafe { slice::from_raw_parts_mut(ptr.offset(1), len - 1) };
}
fn main() {
let mut slice = &mut [0, 1, 2, 3][..];
pop_front(&mut slice);
println!("{:?}", slice);
}
playground
This program outputs:
[1, 2, 3]
Using part of Francis Gagné's answer (I didn't think of trying &mut &), I was able to get it working without using unsafe code:
fn pop_front(mut slice: &mut &[i32]) {
*slice = &slice[1..];
}
fn main() {
let mut slice = &[0, 1, 2, 3][..];
pop_front(&mut slice);
println!("{:?}", slice);
}
Output:
[1, 2, 3]
Related
fn change(a: &mut i32, b: &mut i32) {
let c = *a;
*a = *b;
*b = c;
}
fn main() {
let mut v = vec![1, 2, 3];
change(&mut v[0], &mut v[1]);
}
When I compile the code above, it has the error:
error[E0499]: cannot borrow `v` as mutable more than once at a time
--> src/main.rs:9:32
|
9 | change(&mut v[0], &mut v[1]);
| - ^ - first borrow ends here
| | |
| | second mutable borrow occurs here
| first mutable borrow occurs here
Why does the compiler prohibit it? v[0] and v[1] occupy different memory positions, so it's not dangerous to use these together. And what should I do if I come across this problem?
You can solve this with split_at_mut():
let mut v = vec![1, 2, 3];
let (a, b) = v.split_at_mut(1); // Returns (&mut [1], &mut [2, 3])
change(&mut a[0], &mut b[0]);
There are uncountably many safe things to do that the compiler unfortunately does not recognize yet. split_at_mut() is just like that, a safe abstraction implemented with an unsafe block internally.
We can do that too, for this problem. The following is something I use in code where I need to separate all three cases anyway (I: Index out of bounds, II: Indices equal, III: Separate indices).
enum Pair<T> {
Both(T, T),
One(T),
None,
}
fn index_twice<T>(slc: &mut [T], a: usize, b: usize) -> Pair<&mut T> {
if a == b {
slc.get_mut(a).map_or(Pair::None, Pair::One)
} else {
if a >= slc.len() || b >= slc.len() {
Pair::None
} else {
// safe because a, b are in bounds and distinct
unsafe {
let ar = &mut *(slc.get_unchecked_mut(a) as *mut _);
let br = &mut *(slc.get_unchecked_mut(b) as *mut _);
Pair::Both(ar, br)
}
}
}
}
Since Rust 1.26, pattern matching can be done on slices. You can use that as long as you don't have huge indices and your indices are known at compile-time.
fn change(a: &mut i32, b: &mut i32) {
let c = *a;
*a = *b;
*b = c;
}
fn main() {
let mut arr = [5, 6, 7, 8];
{
let [ref mut a, _, ref mut b, ..] = arr;
change(a, b);
}
assert_eq!(arr, [7, 6, 5, 8]);
}
The borrow rules of Rust need to be checked at compilation time, that is why something like mutably borrowing a part of a Vec is a very hard problem to solve (if not impossible), and why it is not possible with Rust.
Thus, when you do something like &mut v[i], it will mutably borrow the entire vector.
Imagine I did something like
let guard = something(&mut v[i]);
do_something_else(&mut v[j]);
guard.do_job();
Here, I create an object guard that internally stores a mutable reference to v[i], and will do something with it when I call do_job().
In the meantime, I did something that changed v[j]. guard holds a mutable reference that is supposed to guarantee nothing else can modify v[i]. In this case, all is good, as long as i is different from j; if the two values are equal it is a huge violation of the borrow rules.
As the compiler cannot guarantee that i != j, it is thus forbidden.
This was a simple example, but similar cases are legions, and are why such access mutably borrows the whole container. Plus the fact that the compiler actually does not know enough about the internals of Vec to ensure that this operation is safe even if i != j.
In your precise case, you can have a look at the swap(..) method available on Vec that does the swap you are manually implementing.
On a more generic case, you'll probably need an other container. Possibilities are wrapping all the values of your Vec into a type with interior mutability, such as Cell or RefCell, or even using a completely different container, as #llogiq suggested in his answer with par-vec.
The method [T]::iter_mut() returns an iterator that can yield a mutable reference for each element in the slice. Other collections have an iter_mut method too. These methods often encapsulate unsafe code, but their interface is totally safe.
Here's a general purpose extension trait that adds a method on slices that returns mutable references to two distinct items by index:
pub trait SliceExt {
type Item;
fn get_two_mut(&mut self, index0: usize, index1: usize) -> (&mut Self::Item, &mut Self::Item);
}
impl<T> SliceExt for [T] {
type Item = T;
fn get_two_mut(&mut self, index0: usize, index1: usize) -> (&mut Self::Item, &mut Self::Item) {
match index0.cmp(&index1) {
Ordering::Less => {
let mut iter = self.iter_mut();
let item0 = iter.nth(index0).unwrap();
let item1 = iter.nth(index1 - index0 - 1).unwrap();
(item0, item1)
}
Ordering::Equal => panic!("[T]::get_two_mut(): received same index twice ({})", index0),
Ordering::Greater => {
let mut iter = self.iter_mut();
let item1 = iter.nth(index1).unwrap();
let item0 = iter.nth(index0 - index1 - 1).unwrap();
(item0, item1)
}
}
}
}
On recent nightlies, there is get_many_mut():
#![feature(get_many_mut)]
fn main() {
let mut v = vec![1, 2, 3];
let [a, b] = v
.get_many_mut([0, 1])
.expect("out of bounds or overlapping indices");
change(a, b);
}
Building up on the answer by #bluss you can use split_at_mut() to create a function that can turn mutable borrow of a vector into a vector of mutable borrows of vector elements:
fn borrow_mut_elementwise<'a, T>(v:&'a mut Vec<T>) -> Vec<&'a mut T> {
let mut result:Vec<&mut T> = Vec::new();
let mut current: &mut [T];
let mut rest = &mut v[..];
while rest.len() > 0 {
(current, rest) = rest.split_at_mut(1);
result.push(&mut current[0]);
}
result
}
Then you can use it to get a binding that lets you mutate many items of original Vec at once, even while you are iterating over them (if you access them by index in your loop, not through any iterator):
let mut items = vec![1,2,3];
let mut items_mut = borrow_mut_elementwise(&mut items);
for i in 1..items_mut.len() {
*items_mut[i-1] = *items_mut[i];
}
println!("{:?}", items); // [2, 3, 3]
The problem is that &mut v[…] first mutably borrows v and then gives the mutable reference to the element to the change-function.
This reddit comment has a solution to your problem.
Edit: Thanks for the heads-up, Shepmaster. par-vec is a library that allows to mutably borrow disjunct partitions of a vec.
I publish my daily utils for this to crate.io. Link to the doc.
You may use it like
use arref::array_mut_ref;
let mut arr = vec![1, 2, 3, 4];
let (a, b) = array_mut_ref!(&mut arr, [1, 2]);
assert_eq!(*a, 2);
assert_eq!(*b, 3);
let (a, b, c) = array_mut_ref!(&mut arr, [1, 2, 0]);
assert_eq!(*c, 1);
// ⚠️ The following code will panic. Because we borrow the same element twice.
// let (a, b) = array_mut_ref!(&mut arr, [1, 1]);
It's a simple wrapper around the following code, which is sound. But it requires that the two indexes are different at runtime.
pub fn array_mut_ref<T>(arr: &mut [T], a0: usize, a1: usize) -> (&mut T, &mut T) {
assert!(a0 != a1);
// SAFETY: this is safe because we know a0 != a1
unsafe {
(
&mut *(&mut arr[a0] as *mut _),
&mut *(&mut arr[a1] as *mut _),
)
}
}
Alternatively, you may use a method that won't panic with mut_twice
#[inline]
pub fn mut_twice<T>(arr: &mut [T], a0: usize, a1: usize) -> Result<(&mut T, &mut T), &mut T> {
if a0 == a1 {
Err(&mut arr[a0])
} else {
unsafe {
Ok((
&mut *(&mut arr[a0] as *mut _),
&mut *(&mut arr[a1] as *mut _),
))
}
}
}
fn change(a: &mut i32, b: &mut i32) {
let c = *a;
*a = *b;
*b = c;
}
fn main() {
let mut v = vec![1, 2, 3];
change(&mut v[0], &mut v[1]);
}
When I compile the code above, it has the error:
error[E0499]: cannot borrow `v` as mutable more than once at a time
--> src/main.rs:9:32
|
9 | change(&mut v[0], &mut v[1]);
| - ^ - first borrow ends here
| | |
| | second mutable borrow occurs here
| first mutable borrow occurs here
Why does the compiler prohibit it? v[0] and v[1] occupy different memory positions, so it's not dangerous to use these together. And what should I do if I come across this problem?
You can solve this with split_at_mut():
let mut v = vec![1, 2, 3];
let (a, b) = v.split_at_mut(1); // Returns (&mut [1], &mut [2, 3])
change(&mut a[0], &mut b[0]);
There are uncountably many safe things to do that the compiler unfortunately does not recognize yet. split_at_mut() is just like that, a safe abstraction implemented with an unsafe block internally.
We can do that too, for this problem. The following is something I use in code where I need to separate all three cases anyway (I: Index out of bounds, II: Indices equal, III: Separate indices).
enum Pair<T> {
Both(T, T),
One(T),
None,
}
fn index_twice<T>(slc: &mut [T], a: usize, b: usize) -> Pair<&mut T> {
if a == b {
slc.get_mut(a).map_or(Pair::None, Pair::One)
} else {
if a >= slc.len() || b >= slc.len() {
Pair::None
} else {
// safe because a, b are in bounds and distinct
unsafe {
let ar = &mut *(slc.get_unchecked_mut(a) as *mut _);
let br = &mut *(slc.get_unchecked_mut(b) as *mut _);
Pair::Both(ar, br)
}
}
}
}
Since Rust 1.26, pattern matching can be done on slices. You can use that as long as you don't have huge indices and your indices are known at compile-time.
fn change(a: &mut i32, b: &mut i32) {
let c = *a;
*a = *b;
*b = c;
}
fn main() {
let mut arr = [5, 6, 7, 8];
{
let [ref mut a, _, ref mut b, ..] = arr;
change(a, b);
}
assert_eq!(arr, [7, 6, 5, 8]);
}
The borrow rules of Rust need to be checked at compilation time, that is why something like mutably borrowing a part of a Vec is a very hard problem to solve (if not impossible), and why it is not possible with Rust.
Thus, when you do something like &mut v[i], it will mutably borrow the entire vector.
Imagine I did something like
let guard = something(&mut v[i]);
do_something_else(&mut v[j]);
guard.do_job();
Here, I create an object guard that internally stores a mutable reference to v[i], and will do something with it when I call do_job().
In the meantime, I did something that changed v[j]. guard holds a mutable reference that is supposed to guarantee nothing else can modify v[i]. In this case, all is good, as long as i is different from j; if the two values are equal it is a huge violation of the borrow rules.
As the compiler cannot guarantee that i != j, it is thus forbidden.
This was a simple example, but similar cases are legions, and are why such access mutably borrows the whole container. Plus the fact that the compiler actually does not know enough about the internals of Vec to ensure that this operation is safe even if i != j.
In your precise case, you can have a look at the swap(..) method available on Vec that does the swap you are manually implementing.
On a more generic case, you'll probably need an other container. Possibilities are wrapping all the values of your Vec into a type with interior mutability, such as Cell or RefCell, or even using a completely different container, as #llogiq suggested in his answer with par-vec.
The method [T]::iter_mut() returns an iterator that can yield a mutable reference for each element in the slice. Other collections have an iter_mut method too. These methods often encapsulate unsafe code, but their interface is totally safe.
Here's a general purpose extension trait that adds a method on slices that returns mutable references to two distinct items by index:
pub trait SliceExt {
type Item;
fn get_two_mut(&mut self, index0: usize, index1: usize) -> (&mut Self::Item, &mut Self::Item);
}
impl<T> SliceExt for [T] {
type Item = T;
fn get_two_mut(&mut self, index0: usize, index1: usize) -> (&mut Self::Item, &mut Self::Item) {
match index0.cmp(&index1) {
Ordering::Less => {
let mut iter = self.iter_mut();
let item0 = iter.nth(index0).unwrap();
let item1 = iter.nth(index1 - index0 - 1).unwrap();
(item0, item1)
}
Ordering::Equal => panic!("[T]::get_two_mut(): received same index twice ({})", index0),
Ordering::Greater => {
let mut iter = self.iter_mut();
let item1 = iter.nth(index1).unwrap();
let item0 = iter.nth(index0 - index1 - 1).unwrap();
(item0, item1)
}
}
}
}
On recent nightlies, there is get_many_mut():
#![feature(get_many_mut)]
fn main() {
let mut v = vec![1, 2, 3];
let [a, b] = v
.get_many_mut([0, 1])
.expect("out of bounds or overlapping indices");
change(a, b);
}
Building up on the answer by #bluss you can use split_at_mut() to create a function that can turn mutable borrow of a vector into a vector of mutable borrows of vector elements:
fn borrow_mut_elementwise<'a, T>(v:&'a mut Vec<T>) -> Vec<&'a mut T> {
let mut result:Vec<&mut T> = Vec::new();
let mut current: &mut [T];
let mut rest = &mut v[..];
while rest.len() > 0 {
(current, rest) = rest.split_at_mut(1);
result.push(&mut current[0]);
}
result
}
Then you can use it to get a binding that lets you mutate many items of original Vec at once, even while you are iterating over them (if you access them by index in your loop, not through any iterator):
let mut items = vec![1,2,3];
let mut items_mut = borrow_mut_elementwise(&mut items);
for i in 1..items_mut.len() {
*items_mut[i-1] = *items_mut[i];
}
println!("{:?}", items); // [2, 3, 3]
The problem is that &mut v[…] first mutably borrows v and then gives the mutable reference to the element to the change-function.
This reddit comment has a solution to your problem.
Edit: Thanks for the heads-up, Shepmaster. par-vec is a library that allows to mutably borrow disjunct partitions of a vec.
I publish my daily utils for this to crate.io. Link to the doc.
You may use it like
use arref::array_mut_ref;
let mut arr = vec![1, 2, 3, 4];
let (a, b) = array_mut_ref!(&mut arr, [1, 2]);
assert_eq!(*a, 2);
assert_eq!(*b, 3);
let (a, b, c) = array_mut_ref!(&mut arr, [1, 2, 0]);
assert_eq!(*c, 1);
// ⚠️ The following code will panic. Because we borrow the same element twice.
// let (a, b) = array_mut_ref!(&mut arr, [1, 1]);
It's a simple wrapper around the following code, which is sound. But it requires that the two indexes are different at runtime.
pub fn array_mut_ref<T>(arr: &mut [T], a0: usize, a1: usize) -> (&mut T, &mut T) {
assert!(a0 != a1);
// SAFETY: this is safe because we know a0 != a1
unsafe {
(
&mut *(&mut arr[a0] as *mut _),
&mut *(&mut arr[a1] as *mut _),
)
}
}
Alternatively, you may use a method that won't panic with mut_twice
#[inline]
pub fn mut_twice<T>(arr: &mut [T], a0: usize, a1: usize) -> Result<(&mut T, &mut T), &mut T> {
if a0 == a1 {
Err(&mut arr[a0])
} else {
unsafe {
Ok((
&mut *(&mut arr[a0] as *mut _),
&mut *(&mut arr[a1] as *mut _),
))
}
}
}
This question already has an answer here:
How can I create my own data structure with an iterator that returns mutable references?
(1 answer)
Closed 6 months ago.
I'm having a lifetime issue when implementing an iterator on custom struct containing borrows of vecs.
I've been trying different solutions but can't fix it by myself (I'm still a beginner) and I want to understand what is going on.
here is a playground example of my issue, that should be simple enough.
This is the example :
struct SomeData;
struct CustomIterator<'a> {
pub vec: &'a mut Vec<SomeData>,
pub index: usize,
}
struct MultipleIterator<'a> {
iter1: CustomIterator<'a>,
iter2: CustomIterator<'a>,
}
impl<'a> Iterator for MultipleIterator<'a> {
type Item = (&'a mut SomeData, &'a mut SomeData);
fn next(&mut self) -> Option<Self::Item> {
Some((
match self.iter1.vec.get_mut(self.iter1.index) {
Some(mut data) => &mut data,
None => return None,
},
match self.iter2.vec.get_mut(self.iter2.index) {
Some(mut data) => &mut data,
None => return None,
}
))
}
}
I don't unerstand why I can't borrow out of the next function, since I am borrowing the struct anyway when calling next()
This is actually quite tricky to implement safely and requires a lot of care to do correctly.
First things first though:
match self.iter1.vec.get_mut(self.iter1.index) {
Some(mut data) => &mut data,
None => return None,
},
This is a problem because Vec::get_mut already returns a Option<&mut T>. So in the Some(mut data) arm, data already is a mutable reference. When you try to return &mut data, you're trying to return a &mut &mut T which doesn't work. Instead, just do this:
match self.iter1.vec.get_mut(self.iter1.index) {
Some(data) => data,
None => return None,
},
We can tidy this up even more with the ? operator which does the same thing. I'm gonna substitute SomeData for i32 from now on to demonstrate something later.
impl<'a> Iterator for MultipleIterator<'a> {
type Item = (&'a mut i32, &'a mut i32);
fn next(&mut self) -> Option<Self::Item> {
Some((
self.iter1.vec.get_mut(self.iter1.index)?,
self.iter2.vec.get_mut(self.iter2.index)?,
))
}
}
This still doesn't work and now we're getting to the core of the problem. The signature of next is
fn next(&mut self) -> Option<(&'a mut i32, &'a mut i32)>
which can be desugared to
fn next<'b>(&'b mut self) -> Option<(&'a mut i32, &'a mut i32)>
This means that the lifetime of &mut self ('b) is completely decoupled from the lifetime of the references we return ('a). Which makes total sense. If that wasn't the case, we couldn't do
let mut v = vec![1,2,3];
let mut iter = v.iter_mut();
let next1: &mut i32 = iter.next().unwrap();
let next2: &mut i32 = iter.next().unwrap();
because the lifetime of next1 would have to be the same lifetime of iter, i.e. that of v. But you can't have multiple mutable references to v at the same time, so next2 would be illegal.
So you're getting an error because Rust only knows that self is borrowed for 'b but you're telling it that you're returning a reference with lifetime 'a which it can't verify to be true.
And for good reason, because as it stands right now, your implementation isn't safe! Let's just throw caution to the wind and tell Rust that this is okay with unsafe:
impl<'a> Iterator for MultipleIterator<'a> {
type Item = (&'a mut i32, &'a mut i32);
fn next(&mut self) -> Option<Self::Item> {
unsafe {
Some((
&mut *(self.iter1.vec.get_mut(self.iter1.index)? as *mut _),
&mut *(self.iter2.vec.get_mut(self.iter2.index)? as *mut _),
))
}
}
}
It's not important what exactly this does, it basically just tells the compiler to shut up and trust me.
But now, we can do this:
let mut v1 = vec![1, 2, 3];
let mut v2 = vec![4, 5, 6];
let mut mi = MultipleIterator {
iter1: CustomIterator {
vec: &mut v1,
index: 0,
},
iter2: CustomIterator {
vec: &mut v2,
index: 0,
},
};
let next1 = mi.next().unwrap();
let next2 = mi.next().unwrap();
assert_eq!(next1, (&mut 1, &mut 4));
assert_eq!(next2, (&mut 1, &mut 4));
*next1.0 += 1;
assert_eq!(next1, (&mut 2, &mut 4));
assert_eq!(next2, (&mut 2, &mut 4));
We have broken Rust's most important rule: never have two mutable references to the same thing at once.
This can only be safe if your Iterator implementation can never return a mutable reference to something more than once. You could increment index each time, for example (although this still requires unsafe):
impl<'a> Iterator for MultipleIterator<'a> {
type Item = (&'a mut i32, &'a mut i32);
fn next(&mut self) -> Option<Self::Item> {
let next1 = self.iter1.vec.get_mut(self.iter1.index)?;
let next2 = self.iter2.vec.get_mut(self.iter2.index)?;
self.iter1.index += 1;
self.iter2.index += 1;
// SAFETY: this is safe because we will never return a reference
// to the same index more than once
unsafe { Some((&mut *(next1 as *mut _), &mut *(next2 as *mut _))) }
}
}
Here is an interesting related read from the nomicon; the "mutable slice" example being particularly relevant to your problem.
Parameters can be passed to functions and modified:
fn set_42(int: &mut i32) {
*int += 42;
}
fn main() {
let mut int = 0;
set_42(&mut int);
println!("{:?}", int);
}
Output:
42
Changing the code to use a slice fails with a whole bunch of errors:
fn pop_front(slice: &mut [i32]) {
*slice = &{slice}[1..];
}
fn main() {
let mut slice = &[0, 1, 2, 3][..];
pop_front(&mut slice);
println!("{:?}", slice);
}
Output:
error[E0308]: mismatched types
--> src/main.rs:2:14
|
2 | *slice = &{ slice }[1..];
| ^^^^^^^^^^^^^^^
| |
| expected slice `[i32]`, found `&[i32]`
| help: consider removing the borrow: `{ slice }[1..]`
error[E0277]: the size for values of type `[i32]` cannot be known at compilation time
--> src/main.rs:2:5
|
2 | *slice = &{ slice }[1..];
| ^^^^^^ doesn't have a size known at compile-time
|
= help: the trait `std::marker::Sized` is not implemented for `[i32]`
= note: to learn more, visit <https://doc.rust-lang.org/book/ch19-04-advanced-types.html#dynamically-sized-types-and-the-sized-trait>
= note: the left-hand-side of an assignment must have a statically known size
If we try using a mutable slice (which isn't what I really want; I don't want to modify the values within the slice, I just want to modify the slice itself so it covers a smaller range of elements) and a mutable parameter, it has no effect on the original slice:
fn pop_front(mut slice: &mut [i32]) {
slice = &mut {slice}[1..];
}
fn main() {
let mut slice = &mut [0, 1, 2, 3][..];
pop_front(&mut slice);
println!("{:?}", slice);
}
Output:
[0, 1, 2, 3]
Is there a way to modify a slice that's a function parameter? I don't want to modify the elements within the slice; I just want to modify the range of the slice itself so it becomes a smaller "sub-slice".
As others have said, the core idea here is to take a &mut &... [T] (where ... is mut or empty) and read/write to the internal slice. The other answers demonstrate it is possible for &mut &[T] in safe code, and possible for &mut &mut [T] with unsafe code, but they don't explain why there's the difference... and &mut &mut [T] is possible with safe code too.
In explicit-lifetime terms, the nested reference is something like &'a mut &'b ... [T] for some lifetimes 'a and 'b, and the goal here is to get a &'b ... [T], slice it and write that into the &'a mut.
For &'a mut &'b [T], this is easy: &[T] is copy, so writing *slice = &slice[1..] will effectively copy the &'b [T] out of the &mut and then, later, overwrite the existing value with the shorter one. The copy means that one literally gets a &'b [T] to operate with, and so there's no direct connection between that and the &'a mut, and hence it is legal to mutate. It is effectively something like
fn pop_front<'a, 'b>(slice: &'a mut &'b[i32]) {
// *slice = &slice[1..] behaves like
let value: &'b [i32] = *slice;
*slice = &value[1..]
}
(I've labelled the lifetimes and annotated the type to tie into my explanation, but this is not required for the code to work.)
For &'a mut &'b mut [T] things are a little trickier: &mut [T] cannot be copied: dereferencing won't copy, it will reborrow to give a &'a mut [T] i.e. the slice has a lifetime that is connected to the outer &'a mut, not the inner &'b mut [T]. This means the sliced reference has a shorter lifetime than the type it is trying to overwrite, so it's invalid to store the slice into that position. In other words:
fn pop_front<'a, 'b>(slice: &'a mut &'b mut [i32]) {
let value: &'a mut [i32] = &mut **slice;
*slice = &mut value[1..] // error
}
The way to do this safely for &'a mut &'b mut [T] is to get the internal slice out of the reference with that 'b lifetime. This requires keeping track of the "one owner" rule, doing no borrowing, and the best function for this sort of ownership manipulation is mem::replace. It allows us to extract the inner &'b mut [T] by swapping it with some placeholder, which we can then overwrite with the short version. The best/only placeholder is an empty array: writing &mut [] can be a &'c mut [X] for any type X and any lifetime 'c, since there's no data to store and so nothing needs initialisation, and no data will ever become invalid. In particular, it can be a &'b mut [T]:
fn pop_front<'a, 'b>(slice: &'a mut &'b mut [i32]) {
let value: &'b mut [i32] = mem::replace(slice, &mut []);
*slice = &mut value[1..]
}
Since &mut[T] implements Default, we can also use mem::take:
fn pop_front<'a, 'b>(slice: &'a mut &'b mut [i32]) {
let value: &'b mut [i32] = mem::take(slice);
*slice = &mut value[1..]
}
(As above, I've made things more explicit than necessary.)
See also:
Why can't I assign one dereference of a reference of a reference to another when the outer lifetimes differ?
Why does linking lifetimes matter only with mutable references?
How can I swap in a new value for a field in a mutable reference to a structure?
If you need to modify an immutable slice, see Cornstalks's answer.
You cannot modify a mutable slice in safe Rust. When you take a subslice of a mutable slice, you effectively borrow from the original slice. This means that the subslice must not outlive the original slice.
You want something that looks like this:
fn pop_front(slice: &mut &mut [i32]) {
*slice = &mut slice[1..];
}
but the subslice slice[1..] is only valid until the end of the function, and which point the borrow will end and the original slice (the slice parameter) will be usable again.
We can use some unsafe code to construct manually the slice we want:
use std::slice;
fn pop_front(slice: &mut &mut [i32]) {
let ptr = slice.as_mut_ptr();
let len = slice.len();
*slice = unsafe { slice::from_raw_parts_mut(ptr.offset(1), len - 1) };
}
fn main() {
let mut slice = &mut [0, 1, 2, 3][..];
pop_front(&mut slice);
println!("{:?}", slice);
}
playground
This program outputs:
[1, 2, 3]
Using part of Francis Gagné's answer (I didn't think of trying &mut &), I was able to get it working without using unsafe code:
fn pop_front(mut slice: &mut &[i32]) {
*slice = &slice[1..];
}
fn main() {
let mut slice = &[0, 1, 2, 3][..];
pop_front(&mut slice);
println!("{:?}", slice);
}
Output:
[1, 2, 3]
fn change(a: &mut i32, b: &mut i32) {
let c = *a;
*a = *b;
*b = c;
}
fn main() {
let mut v = vec![1, 2, 3];
change(&mut v[0], &mut v[1]);
}
When I compile the code above, it has the error:
error[E0499]: cannot borrow `v` as mutable more than once at a time
--> src/main.rs:9:32
|
9 | change(&mut v[0], &mut v[1]);
| - ^ - first borrow ends here
| | |
| | second mutable borrow occurs here
| first mutable borrow occurs here
Why does the compiler prohibit it? v[0] and v[1] occupy different memory positions, so it's not dangerous to use these together. And what should I do if I come across this problem?
You can solve this with split_at_mut():
let mut v = vec![1, 2, 3];
let (a, b) = v.split_at_mut(1); // Returns (&mut [1], &mut [2, 3])
change(&mut a[0], &mut b[0]);
There are uncountably many safe things to do that the compiler unfortunately does not recognize yet. split_at_mut() is just like that, a safe abstraction implemented with an unsafe block internally.
We can do that too, for this problem. The following is something I use in code where I need to separate all three cases anyway (I: Index out of bounds, II: Indices equal, III: Separate indices).
enum Pair<T> {
Both(T, T),
One(T),
None,
}
fn index_twice<T>(slc: &mut [T], a: usize, b: usize) -> Pair<&mut T> {
if a == b {
slc.get_mut(a).map_or(Pair::None, Pair::One)
} else {
if a >= slc.len() || b >= slc.len() {
Pair::None
} else {
// safe because a, b are in bounds and distinct
unsafe {
let ar = &mut *(slc.get_unchecked_mut(a) as *mut _);
let br = &mut *(slc.get_unchecked_mut(b) as *mut _);
Pair::Both(ar, br)
}
}
}
}
Since Rust 1.26, pattern matching can be done on slices. You can use that as long as you don't have huge indices and your indices are known at compile-time.
fn change(a: &mut i32, b: &mut i32) {
let c = *a;
*a = *b;
*b = c;
}
fn main() {
let mut arr = [5, 6, 7, 8];
{
let [ref mut a, _, ref mut b, ..] = arr;
change(a, b);
}
assert_eq!(arr, [7, 6, 5, 8]);
}
The borrow rules of Rust need to be checked at compilation time, that is why something like mutably borrowing a part of a Vec is a very hard problem to solve (if not impossible), and why it is not possible with Rust.
Thus, when you do something like &mut v[i], it will mutably borrow the entire vector.
Imagine I did something like
let guard = something(&mut v[i]);
do_something_else(&mut v[j]);
guard.do_job();
Here, I create an object guard that internally stores a mutable reference to v[i], and will do something with it when I call do_job().
In the meantime, I did something that changed v[j]. guard holds a mutable reference that is supposed to guarantee nothing else can modify v[i]. In this case, all is good, as long as i is different from j; if the two values are equal it is a huge violation of the borrow rules.
As the compiler cannot guarantee that i != j, it is thus forbidden.
This was a simple example, but similar cases are legions, and are why such access mutably borrows the whole container. Plus the fact that the compiler actually does not know enough about the internals of Vec to ensure that this operation is safe even if i != j.
In your precise case, you can have a look at the swap(..) method available on Vec that does the swap you are manually implementing.
On a more generic case, you'll probably need an other container. Possibilities are wrapping all the values of your Vec into a type with interior mutability, such as Cell or RefCell, or even using a completely different container, as #llogiq suggested in his answer with par-vec.
The method [T]::iter_mut() returns an iterator that can yield a mutable reference for each element in the slice. Other collections have an iter_mut method too. These methods often encapsulate unsafe code, but their interface is totally safe.
Here's a general purpose extension trait that adds a method on slices that returns mutable references to two distinct items by index:
pub trait SliceExt {
type Item;
fn get_two_mut(&mut self, index0: usize, index1: usize) -> (&mut Self::Item, &mut Self::Item);
}
impl<T> SliceExt for [T] {
type Item = T;
fn get_two_mut(&mut self, index0: usize, index1: usize) -> (&mut Self::Item, &mut Self::Item) {
match index0.cmp(&index1) {
Ordering::Less => {
let mut iter = self.iter_mut();
let item0 = iter.nth(index0).unwrap();
let item1 = iter.nth(index1 - index0 - 1).unwrap();
(item0, item1)
}
Ordering::Equal => panic!("[T]::get_two_mut(): received same index twice ({})", index0),
Ordering::Greater => {
let mut iter = self.iter_mut();
let item1 = iter.nth(index1).unwrap();
let item0 = iter.nth(index0 - index1 - 1).unwrap();
(item0, item1)
}
}
}
}
On recent nightlies, there is get_many_mut():
#![feature(get_many_mut)]
fn main() {
let mut v = vec![1, 2, 3];
let [a, b] = v
.get_many_mut([0, 1])
.expect("out of bounds or overlapping indices");
change(a, b);
}
Building up on the answer by #bluss you can use split_at_mut() to create a function that can turn mutable borrow of a vector into a vector of mutable borrows of vector elements:
fn borrow_mut_elementwise<'a, T>(v:&'a mut Vec<T>) -> Vec<&'a mut T> {
let mut result:Vec<&mut T> = Vec::new();
let mut current: &mut [T];
let mut rest = &mut v[..];
while rest.len() > 0 {
(current, rest) = rest.split_at_mut(1);
result.push(&mut current[0]);
}
result
}
Then you can use it to get a binding that lets you mutate many items of original Vec at once, even while you are iterating over them (if you access them by index in your loop, not through any iterator):
let mut items = vec![1,2,3];
let mut items_mut = borrow_mut_elementwise(&mut items);
for i in 1..items_mut.len() {
*items_mut[i-1] = *items_mut[i];
}
println!("{:?}", items); // [2, 3, 3]
The problem is that &mut v[…] first mutably borrows v and then gives the mutable reference to the element to the change-function.
This reddit comment has a solution to your problem.
Edit: Thanks for the heads-up, Shepmaster. par-vec is a library that allows to mutably borrow disjunct partitions of a vec.
I publish my daily utils for this to crate.io. Link to the doc.
You may use it like
use arref::array_mut_ref;
let mut arr = vec![1, 2, 3, 4];
let (a, b) = array_mut_ref!(&mut arr, [1, 2]);
assert_eq!(*a, 2);
assert_eq!(*b, 3);
let (a, b, c) = array_mut_ref!(&mut arr, [1, 2, 0]);
assert_eq!(*c, 1);
// ⚠️ The following code will panic. Because we borrow the same element twice.
// let (a, b) = array_mut_ref!(&mut arr, [1, 1]);
It's a simple wrapper around the following code, which is sound. But it requires that the two indexes are different at runtime.
pub fn array_mut_ref<T>(arr: &mut [T], a0: usize, a1: usize) -> (&mut T, &mut T) {
assert!(a0 != a1);
// SAFETY: this is safe because we know a0 != a1
unsafe {
(
&mut *(&mut arr[a0] as *mut _),
&mut *(&mut arr[a1] as *mut _),
)
}
}
Alternatively, you may use a method that won't panic with mut_twice
#[inline]
pub fn mut_twice<T>(arr: &mut [T], a0: usize, a1: usize) -> Result<(&mut T, &mut T), &mut T> {
if a0 == a1 {
Err(&mut arr[a0])
} else {
unsafe {
Ok((
&mut *(&mut arr[a0] as *mut _),
&mut *(&mut arr[a1] as *mut _),
))
}
}
}