Rust iterators are data structures that contain a sequence of objects and allow the programmer to iterate them efficiently. However, like most things in Rust, iterators have a steep learning curve. The primary functions and concepts can be daunting, but don’t worry, we are here to do our absolute best to explain them.

All three methods iter(), into_iter() and iter_mut() ultimately iterate the iterator. They differ by the handle of each iterated object T:

  • iter() yields &T immutable references.
  • iter_mut() yields, as the name suggests, &mut T – mutable references.
  • into_iter() yields any of T (moved value), &T or &mut T depending on the usage.

If the differences between iter(), into_iter(), and iter_mut() are not still clear, keep reading this post. You will learn more about each of these methods, helping you understand them by looking at easy-to-follow examples.

Understanding iter() and Basic Iterator Functionality

iter() is the most common method out of the three, and most examples online use it. It allows the programmer to iterate each value as an immutable reference. Let us look at a small example:

let fruits = vec!["Banana", "Apple", "Grapes", "Pineapple"];
            .map(|fruit_name: &&str| fruit_name.len())
            .fold(0, |accumulated, fruit_name_length| {
                accumulated + fruit_name_length
    ); // Should print 26

This example demonstrates a couple of things; let us go over it one line at a time.

The first line defines a new variable named fruits and assigns to it a vector (Vec<&str>) of four fruit names using the vec! Macro.

The second and third lines order Rust to print the output of the statement that starts in the fourth line.

Here (line 5) we see the call to the iter() method, which allows us to call the first actual iterator method – map. map is one of many methods that is called on an iterator and returns a new iterator.

The name “map” is not specific to Rust and is used in many other programming languages for a function that maps each element in a sequence of elements to another element (most commonly creates a modified version of the original element) in a new sequence.

In this simple example, I use the map method to map each fruit_name element in the fruits vector to its length. Notice the parameter that map receives – that is called a closure. A closure is an anonymous function (a function with no name) that can be stored in a variable or passed as an argument to another function.

I passed a closure to map that receives as a parameter (between the | characters) a fruit_name of type &&str. The type &&str is expected; we are iterating &str elements using iter() which yields immutable references to the elements in the iterator. The closure returns the length of fruit_name.

The next call is to the fold method, which iterates the entire iterator and calculates a cumulating single value. The fold method receives two parameters:

  1. The initial accumulated value (the starting value; the value built upon).
  2. A closure that executes for each element in the iterator and receives two parameters:
    1. The accumulated value so far.
    2. The current element.

In this example, the sum of all fruit lengths is calculated, with an initial value of 0. In this case, it would also make sense to use the reduce method that uses the first element of the iterator instead of taking the initial accumulated value as a parameter.

Lazy Evaluation and Consuming Iterators

The call to the fold method is special since it consumes the entire iterator. To understand what consuming means, we first must understand that Rust iterators are lazily evaluated, meaning no actual calculation is performed until the result is requested. A classic example of this mechanism is the next method. Let’s look at how it is used:

    ); // Should print “Banana”

The next method consumes the first element of the iterator, meaning it triggers its calculation and removes it from the iterator. It is important to note that next returns an Option (more about the Option enum here), which might contain None if next is called on an empty iterator.

Understanding iter_mut()

iter_mut() is the least used method out of the three. It allows the programmer to iterate each element as a mutable reference. Let us look at a small example:

let mut numbers = vec![1, 2, 3, 4];
numbers.iter_mut().for_each(|n| *n += 1);
println!("{:?}", numbers); // Should print [2, 3, 4, 5]

In this example, we define a mutable vector of numbers and iterate it using iter_mut(). The second line also introduces a new method: for_each(). for_each() is the one-line version of iterating using the for keyword like so:

for n in numbers.iter_mut() {
        *n += 1;

for_each() receives a parameter which is a closure to execute the for each element in the iterator. Since we are using iter_mut, we are iterating mutable references, which means the closure can modify every number n in the original vector using dereferencing.

It is important to note that since iterators are lazy, using methods like map, fold, and reduce without consuming the iterators won’t change the elements of the vector.

Understanding into_iter()

into_iter() is a method whose name comes from the trait IntoIterator which we must understand before we continue. IntoIterator is a trait that should be implemented when the developer wants to specify how a certain type should be converted into an iterator. It is useful when the type represents some sort of collection of data and most notably allows for an iteration over objects of the type in a for loop syntax.

Let’s look at an example before we continue to into_iter().

The type Vec<T> implements IntoIterator for three cases:

  • impl<T> IntoIterator for Vec<T>
  • impl<'a, T> IntoIterator for &'a Vec<T>
  • impl<'a, T> IntoIterator for &'a mut Vec<T>

The first implementation allows for iterating by value (yields Ts), it implements IntoIterator for a Vec that holds elements of some generic type T.

The second implementation allows for iterating by reference (yields &Ts), it implements IntoIterator for a reference of generic lifetime ‘a to a Vec that holds elements of some generic type T.

The third implementation is similar to the second implementation, but with a generic mutable reference instead.

Those three implementations allow the following code snippet to compile:

let by_value = vec![1, 2, 3];
    for i in by_value {
        println!("{}", i);
    // println!("{:?}", by_value);

    let by_reference = vec![1, 2, 3];

    for i in &by_reference {
        println!("{}", i);

    println!("{:?}", by_reference);

    let mut by_mutable_reference = vec![1, 2, 3];
    for i in &mut by_mutable_reference {
        *i += 1;
    println!("{:?}", by_mutable_reference);

This snippet is long so I’ll explain it in multiple parts.

In the first part (lines 1 to 5), we iterate the by_value vector by value and print its elements. It is important to note that since we are iterating the elements by value, the elements are moved and so it is impossible to print its elements after the for loop.

In the second part (lines 7 to 13), we iterate the by_reference vector by reference and print its elements. Now, since we are iterating the elements by reference, the elements are not moved and so printing the elements again after the for loop is indeed possible.

In the third and last part (lines 15 to 19), we iterate the by_mutable_reference vector by mutable reference, add 1 to every one of its elements and then we can print the vector as expected.

In all three cases, into_iter is called implicitly and in all three cases, the elements are handled differently – i32, &i32, and &mut i32 (in order).


In conclusion, iter(), iter_mut() and into_iter() are ultimately methods that allow the programmer to iterate over a collection of data.

  • iter() provides iteration over immutable references (&T).
  • iter_mut() provides iteration over mutable references (&mut T).
  • into_iter() allows for iteration over any of the moved values (T), immutable references, or mutable references. In addition, it is also implicitly called (when possible) when using a for loop to iterate the collection.

Was this article helpful?

Share your thoughts by replying on Twitter of Become A Better Programmer or to my personal Twitter account.