Skip to content

Commit

Permalink
Rework iterator section (#2523)
Browse files Browse the repository at this point in the history
  • Loading branch information
randomPoison authored Dec 17, 2024
1 parent e902b1e commit 7f0c591
Show file tree
Hide file tree
Showing 5 changed files with 152 additions and 26 deletions.
6 changes: 4 additions & 2 deletions src/SUMMARY.md
Original file line number Diff line number Diff line change
Expand Up @@ -171,9 +171,11 @@

- [Welcome](welcome-day-4.md)
- [Iterators](iterators.md)
- [`Iterator`](iterators/iterator.md)
- [`IntoIterator`](iterators/intoiterator.md)
- [Motivation](iterators/motivation.md)
- [`Iterator` Trait](iterators/iterator.md)
- [`Iterator` Helper Methods](iterators/helpers.md)
- [`collect`](iterators/collect.md)
- [`IntoIterator`](iterators/intoiterator.md)
- [Exercise: Iterator Method Chaining](iterators/exercise.md)
- [Solution](iterators/solution.md)
- [Modules](modules.md)
Expand Down
42 changes: 42 additions & 0 deletions src/iterators/helpers.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,42 @@
---
minutes: 5
---

# `Iterator` Helper Methods

In addition to the `next` method that defines how an iterator behaves, the
`Iterator` trait provides 70+ helper methods that can be used to build
customized iterators.

```rust,editable
let result: i32 = (1..=10) // Create a range from 1 to 10
.filter(|&x| x % 2 == 0) // Keep only even numbers
.map(|x| x * x) // Square each number
.sum(); // Sum up all the squared numbers
println!("The sum of squares of even numbers from 1 to 10 is: {}", result);
```

<details>

- The `Iterator` trait implements many common functional programming operations
over collections (e.g. `map`, `filter`, `reduce`, etc). This is the trait
where you can find all the documentation about them.

- Many of these helper methods take the original iterator and produce a new
iterator with different behavior. These are know as "iterator adapter
methods".

- Some methods, like `sum` and `count`, consume the iterator and pull all of the
elements out of it.

- These methods are designed to be chained together so that it's easy to build a
custom iterator that does exactly what you need.

## More to Explore

- Rust's iterators are extremely efficient and highly optimizable. Even complex
iterators made by combining many adapter methods will still result in code as
efficient as equivalent imperative implementations.

</details>
5 changes: 5 additions & 0 deletions src/iterators/intoiterator.md
Original file line number Diff line number Diff line change
Expand Up @@ -57,6 +57,11 @@ fn main() {

<details>

- `IntoIterator` is the trait that makes for loops work. It is implemented by
collection types such as `Vec<T>` and references to them such as `&Vec<T>` and
`&[T]`. Ranges also implement it. This is why you can iterate over a vector
with `for i in some_vec { .. }` but `some_vec.next()` doesn't exist.

Click through to the docs for `IntoIterator`. Every implementation of
`IntoIterator` must declare two types:

Expand Down
66 changes: 42 additions & 24 deletions src/iterators/iterator.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,49 +2,67 @@
minutes: 5
---

# `Iterator`
# `Iterator` Trait

The [`Iterator`][1] trait supports iterating over values in a collection. It
requires a `next` method and provides lots of methods. Many standard library
types implement `Iterator`, and you can implement it yourself, too:
The [`Iterator`][1] trait defines how an object can be used to produce a
sequence of values. For example, if we wanted to create an iterator that can
produce the elements of a slice it might look something like this:

```rust,editable
struct Fibonacci {
curr: u32,
next: u32,
struct SliceIter<'s> {
slice: &'s [i32],
i: usize,
}
impl Iterator for Fibonacci {
type Item = u32;
impl<'s> Iterator for SliceIter<'s> {
type Item = &'s i32;
fn next(&mut self) -> Option<Self::Item> {
let new_next = self.curr + self.next;
self.curr = self.next;
self.next = new_next;
Some(self.curr)
if self.i == self.slice.len() {
None
} else {
let next = &self.slice[self.i];
self.i += 1;
Some(next)
}
}
}
fn main() {
let fib = Fibonacci { curr: 0, next: 1 };
for (i, n) in fib.enumerate().take(5) {
println!("fib({i}): {n}");
let slice = [2, 4, 6, 8].as_slice();
let iter = SliceIter { slice, i: 0 };
for elem in iter {
println!("elem: {elem}");
}
}
```

<details>

- The `Iterator` trait implements many common functional programming operations
over collections (e.g. `map`, `filter`, `reduce`, etc). This is the trait
where you can find all the documentation about them. In Rust these functions
should produce the code as efficient as equivalent imperative implementations.
- The `SliceIter` example implements the same logic as the C-style `for` loop
demonstrated on the last slide.

- `IntoIterator` is the trait that makes for loops work. It is implemented by
collection types such as `Vec<T>` and references to them such as `&Vec<T>` and
`&[T]`. Ranges also implement it. This is why you can iterate over a vector
with `for i in some_vec { .. }` but `some_vec.next()` doesn't exist.
- Point out to the students that iterators are lazy: Creating the iterator just
initializes the struct but does not otherwise do any work. No work happens
until the `next` method is called.

- Iterators don't need to be finite! It's entirely valid to have an iterator
that will produce values forever. For example, a half open range like `0..`
will keep going until integer overflow occurs.

## More to Explore

- The "real" version of `SliceIter` is the [`slice::Iter`][2] type in the
standard library, however the real version uses pointers under the hood
instead of an index in order to eliminate bounds checks.

- The `SliceIter` example is a good example of a struct that contains a
reference and therefore uses lifetime annotations.

- You can also demonstrate adding a generic parameter to `SliceIter` to allow it
to work with any kind of slice (not just `&[i32]`).

</details>

[1]: https://doc.rust-lang.org/std/iter/trait.Iterator.html
[2]: https://doc.rust-lang.org/stable/std/slice/struct.Iter.html
59 changes: 59 additions & 0 deletions src/iterators/motivation.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,59 @@
---
minutes: 3
---

# Motivating Iterators

If you want to iterate over the contents of an array, you'll need to define:

- Some state to keep track of where you are in the iteration process, e.g. an
index.
- A condition to determine when iteration is done.
- Logic for updating the state of iteration each loop.
- Logic for fetching each element using that iteration state.

In a C-style for loop you declare these things directly:

```c,editable
for (int i = 0; i < array_len; i += 1) {
int elem = array[i];
}
```

In Rust we bundle this state and logic together into an object known as an
"iterator".

<details>

- This slide provides context for what Rust iterators do under the hood. We use
the (hopefully) familiar construct of a C-style `for` loop to show how
iteration requires some state and some logic, that way on the next slide we
can show how an iterator bundles these together.

- Rust doesn't have a C-style `for` loop, but we can express the same thing with
`while`:
```rust,editable
let array = [2, 4, 6, 8];
let mut i = 0;
while i < array.len() {
let elem = array[i];
i += 1;
}
```

## More to Explore

There's another way to express array iteration using `for` in C and C++: You can
use a pointer to the front and a pointer to the end of the array and then
compare those pointers to determine when the loop should end.

```c,editable
for (int *ptr = array; ptr < array + len; ptr += 1) {
int elem = *ptr;
}
```

If students ask, you can point out that this is how Rust's slice and array
iterators work under the hood (though implemented as a Rust iterator).

</details>

0 comments on commit 7f0c591

Please sign in to comment.