X-Git-Url: https://git.ralfj.de/rust-101.git/blobdiff_plain/e2374eed1c3ae8d0063138ea011e86bbd42473ab..d9fc9a4d89d8e4950f43308318962d9179bd1177:/src/part06.rs diff --git a/src/part06.rs b/src/part06.rs index 26fa124..e357d0e 100644 --- a/src/part06.rs +++ b/src/part06.rs @@ -1,100 +1,101 @@ -// Rust-101, Part 06: Copy -// ======================= +// Rust-101, Part 06: Copy, Lifetimes +// ================================== +// We continue to work on our `BigInt`, so we start by importing what we already established. use part05::BigInt; // With `BigInt` being about numbers, we should be able to write a version of `vec_min` -// that computes the minimum of a list of `BigInt`. We start by writing `min` for -// `BigInt`. Now our assumption of having no trailing zeros comes in handy! +// that computes the minimum of a list of `BigInt`. First, we have to write `min` for `BigInt`. impl BigInt { - fn min(self, other: Self) -> Self { - // Just to be sure, we first check that both operands actually satisfy our invariant. - // `debug_assert!` is a macro that checks that its argument (must be of type `bool`) - // is `true`, and panics otherwise. It gets removed in release builds, which you do with - // `cargo build --release`. + fn min_try1(self, other: Self) -> Self { + // Just to be sure, we first check that both operands actually satisfy our invariant. `debug_assert!` is a + // macro that checks that its argument (must be of type `bool`) is `true`, and panics otherwise. It gets + // removed in release builds, which you do with `cargo build --release`. debug_assert!(self.test_invariant() && other.test_invariant()); + // Now our assumption of having no trailing zeros comes in handy: // If the lengths of the two numbers differ, we already know which is larger. if self.data.len() < other.data.len() { self } else if self.data.len() > other.data.len() { other } else { - // **Exercise 05.1**: Fill in this code. - panic!("Not yet implemented."); + // **Exercise 06.1**: Fill in this code. + unimplemented!() } } } -// Now we can write `vec_min`. In order to make it type-check, we have to write it as follows. +// Now we can write `vec_min`. However, in order to make it type-check, we have to make a full (deep) copy of e +// by calling `clone()`. fn vec_min(v: &Vec) -> Option { let mut min: Option = None; for e in v { + let e = e.clone(); min = Some(match min { - None => e.clone(), - Some(n) => e.clone().min(n) + None => e, + Some(n) => e.min_try1(n) }); } min } -// Now, what's happening here? Why do we have to write `clone()`, and why did we not -// have to write that in our previous version? +// Now, what's happening here? Why do we have to clone `e`, and why did we not +// have to do that in our previous version? // // The answer is already hidden in the type of `vec_min`: `v` is just borrowed, but // the Option that it returns is *owned*. We can't just return one of the elements of `v`, // as that would mean that it is no longer in the vector! In our code, this comes up when we update -// the intermediate variable `min`, which also has type `Option`. If you replace `e.clone()` -// in the `None` arm with `*e`, Rust will complain "Cannot move out of borrowed content". That's because +// the intermediate variable `min`, which also has type `Option`. If you replace get rid of the +// `e.clone()`, Rust will complain "Cannot move out of borrowed content". That's because // `e` is a `&BigInt`. Assigning `min = Some(*e)` works just like a function call: Ownership of the // underlying data is transferred from where `e` borrows from to `min`. But that's not allowed, since // we just borrowed `e`, so we cannot empty it! We can, however, call `clone()` on it. Then we own // the copy that was created, and hence we can store it in `min`.
-// Of course, making such a full copy is expensive, so we'd like to avoid it. We'll some to that soon. +// Of course, making such a full copy is expensive, so we'd like to avoid it. We'll some to that in the next part. // ## `Copy` types // But before we go there, I should answer the second question I brought up above: Why did our old `vec_min` work? // We stored the minimal `i32` locally without cloning, and Rust did not complain. That's because there isn't // really much of an "ownership" when it comes to types like `i32` or `bool`: If you move the value from one -// place to another, then both instance are "complete". We also say the value has been *duplicated*. This is in +// place to another, then both instances are "complete". We also say the value has been *duplicated*. This is in // stark contrast to types like `Vec`, where moving the value results in both the old and the new vector to -// point to the same underlying buffer. We don't have two vectors, there's no duplication. +// point to the same underlying buffer. We don't have two vectors, there's no proper duplication. // -// Rust calls types that can be freely duplicated `Copy` types. `Copy` is another trait, and it -// is implemented for types like `i32` and `bool`. Remember how we defined the trait `Minimum` by writing -// `trait Minimum : Copy { ...`? This tells Rust that every type that implements `Minimum` must also -// implement `Copy`, and that's why the compiler accepted our generic `vec_min` in part 02. -// `Copy` is the first *marker trait* that we encounter: It does not provide any methods, but -// makes a promise about the behavior of the type - in this case, being duplicable. +// Rust calls types that can be easily duplicated `Copy` types. `Copy` is another trait, and it is implemented for +// types like `i32` and `bool`. Remember how we defined the trait `Minimum` by writing `trait Minimum : Copy { ...`? +// This tells Rust that every type that implements `Minimum` must also implement `Copy`, and that's why the compiler +// accepted our generic `vec_min` in part 02. `Copy` is the first *marker trait* that we encounter: It does not provide +// any methods, but makes a promise about the behavior of the type - in this case, being duplicable. // If you try to implement `Copy` for `BigInt`, you will notice that Rust // does not let you do that. A type can only be `Copy` if all its elements // are `Copy`, and that's not the case for `BigInt`. However, we can make // `SomethingOrNothing` copy if `T` is `Copy`. use part02::{SomethingOrNothing,Something,Nothing}; -impl Copy for SomethingOrNothing{} +impl Copy for SomethingOrNothing {} // Again, Rust can generate implementations of `Copy` automatically. If // you add `#[derive(Copy,Clone)]` right before the definition of `SomethingOrNothing`, // both `Copy` and `Clone` will automatically be implemented. // ## An operational perspective // Instead of looking at what happens "at the surface" (i.e., visible in Rust), one can also explain -// ownership passing and how `Copy` and `Clone` fit by looking at what happens on the machine.
+// ownership passing and how `Copy` and `Clone` fit in by looking at what happens on the machine.
// When Rust code is executed, passing a value (like `i32` or `Vec`) to a function will always // result in a shallow copy being performed: Rust just copies the bytes representing that value, and // considers itself done. That's just like the default copy constructor in C++. Rust, however, will // consider this a destructive operation: After copying the bytes elsewhere, the original value must -// no longer be used. After all, the two could not share a pointer! If, however, you mark a type `Copy`, +// no longer be used. After all, the two could now share a pointer! If, however, you mark a type `Copy`, // then Rust will *not* consider a move destructive, and just like in C++, the old and new value -// can happily coexist. Now, Rust does not allow to to overload the copy constructor. This means that +// can happily coexist. Now, Rust does not allow you to overload the copy constructor. This means that // passing a value around will always be a fast operation, no allocation or any other kind of heap access // will happen. In the situations where you would write a copy constructor in C++ (and hence // incur a hidden cost on every copy of this type), you'd have the type *not* implement `Copy`, but only // `Clone`. This makes the cost explicit. // ## Lifetimes -// To fix the performance problems of `vec_min`, we need ti avoid using `clone()`. We'd like +// To fix the performance problems of `vec_min`, we need to avoid using `clone()`. We'd like // the return value to not be owned (remember that this was the source of our need for cloning), but *borrowed*. -// This is demonstrated by the function `head` that borrows the first element of a vector if it is non-empty. +// The function `head` demonstrates how that could work: It borrows the first element of a vector if it is non-empty. // The type of the function says that it will either return nothing, or it will return a borrowed `T`. // We can then borrow the first element of `v` and use it to construct the return value. fn head(v: &Vec) -> Option<&T> { @@ -104,10 +105,8 @@ fn head(v: &Vec) -> Option<&T> { None } } - -// Now, coming back to `head` - here, we are returning a pointer to the first element. But doesn't -// that mean that callers have to be careful? Imagine `head` would be a C++ function, and we would -// write the following code. +// Technically, we are returning a pointer to the first element. But doesn't that mean that callers have to be +// careful? Imagine `head` would be a C++ function, and we would write the following code. /* int foo(std::vector v) { int *first = head(v); @@ -115,8 +114,9 @@ fn head(v: &Vec) -> Option<&T> { return *first; } */ -// This is very much like our very first motivating example for ownership, at the beginning of part 04. -// But this time, the bug is hidden behind the call to `head`. How does Rust solve this? If we translate +// This is very much like our very first motivating example for ownership, at the beginning of part 04: +// `push_back` could reallocate the buffer, making `first` an invalid pointer. Again, we have aliasing (of `first` +// and `v`) and mutation. But this time, the bug is hidden behind the call to `head`. How does Rust solve this? If we translate // the code above to Rust, it doesn't compile, so clearly we are good - but how and why? // (Notice that have to explicitly assert using `unwrap` that `first` is not `None`, whereas the C++ code // above would silently dereference a `NULL`-pointer. But that's another point.) @@ -128,11 +128,11 @@ fn rust_foo(mut v: Vec) -> i32 { // To give the answer to this question, we have to talk about the *lifetime* of a borrow. The point is, saying that // you borrowed your friend a `Vec`, or a book, is not good enough, unless you also agree on *how long* -// your friend can borrow. After all, you need to know when you can rely on owning your data (or book) again. +// your friend can borrow it. After all, you need to know when you can rely on owning your data (or book) again. // -// Every borrow in Rust has an associated lifetime. The full type of `head` reads as follows: -// `fn<'a, T>(&'a Vec) -> Option<&'a T>`. Here, `'a` is a *lifetime variable*, which represents how long the vector has -// been borrowed. The function type expresses that argument and return value have *the same lifetime*. +// Every borrow in Rust has an associated lifetime, written `&'a T` for a borrow of type `T` with lifetime `'a`. The full +// type of `head` reads as follows: `fn<'a, T>(&'a Vec) -> Option<&'a T>`. Here, `'a` is a *lifetime variable*, which +// represents how long the vector has been borrowed. The function type expresses that argument and return value have *the same lifetime*. // // When analyzing the code of `rust_foo`, Rust has to assign a lifetime to `first`. It will choose the scope // where `first` is valid, which is the entire rest of the function. Because `head` ties the lifetime of its @@ -147,4 +147,4 @@ fn rust_foo(mut v: Vec) -> i32 { // Most of the time, we don't have to explicitly add lifetimes to function types. This is thanks to *lifetimes elision*, // where Rust will automatically insert lifetimes we did not specify, following some [simple, well-documented rules](http://doc.rust-lang.org/stable/book/lifetimes.html#lifetime-elision). -// [index](main.html) | [previous](part05.html) | [next](main.html) +// [index](main.html) | [previous](part05.html) | [next](part07.html)