diff --git a/src/ch17-02-trait-objects.md b/src/ch17-02-trait-objects.md index 6f947b2..edc481f 100644 --- a/src/ch17-02-trait-objects.md +++ b/src/ch17-02-trait-objects.md @@ -258,28 +258,16 @@ objects. Clone is an example of one. You'll get errors that will let you know if a trait can't be a trait object, look up object safety if you're interested in the details"? Thanks! /Carol --> -Not all traits can be made into trait objects; only *object safe* traits can. A -trait is object safe as long as both of the following are true: +不是所有的trait都可以被放进trait对象中; 只有*对象安全的*trait才可以这样做. 一个trait只有同时满足如下两点时才被认为是对象安全的: -* The trait does not require `Self` to be `Sized` -* All of the trait's methods are object safe. +* 该trait要求`Self`不是`Sized`; +* 该trait的所有方法都是对象安全的; -`Self` is a keyword that is an alias for the type that we're implementing -traits or methods on. `Sized` is a marker trait like the `Send` and `Sync` -traits that we talked about in Chapter 16. `Sized` is automatically implemented -on types that have a known size at compile time, such as `i32` and references. -Types that do not have a known size include slices (`[T]`) and trait objects. +`Self`是一个类型的别名关键字，它表示当前正被实现的trait类型或者是方法所属的类型. `Sized`是一个像在第16章中介绍的`Send`和`Sync`那样的标记trait, 在编译时它会自动被放进大小确定的类型里，比如`i32`和引用. 大小不确定的类型有切片(`[T]`)和trait对象. -`Sized` is an implicit trait bound on all generic type parameters by default. -Most useful operations in Rust require a type to be `Sized`, so making `Sized` -a default requirement on trait bounds means we don't have to write `T: Sized` -with most every use of generics. If we want to be able to use a trait on -slices, however, we need to opt out of the `Sized` trait bound, and we can do -that by specifying `T: ?Sized` as a trait bound. +`Sized`是一个默认会被绑定到所有常规类型参数的内隐trait. Rust中要求一个类型是`Sized`的最具可用性的用法是让`Sized`成为一个默认的trait绑定，这样我们就可以在大多数的常规的用法中不去写`T: Sized`了. 如果我们想在切片(slice)中使用一个trait, 我们需要取消对`Sized`的trait绑定, 我们只需制定`T: ?Sized`作为trait绑定. -Traits have a default bound of `Self: ?Sized`, which means that they can be -implemented on types that may or may not be `Sized`. If we create a trait `Foo` -that opts out of the `Self: ?Sized` bound, that would look like the following: +默认绑定到`Self: ?Sized`的trait可以被实现到是`Sized`或非`Sized`的类型上. 如果我们创建一个不绑定`Self: ?Sized`的trait`Foo`，它看上去应该像这样: ```rust trait Foo: Sized { @@ -287,40 +275,21 @@ trait Foo: Sized { } ``` -The trait `Sized` is now a *super trait* of trait `Foo`, which means trait -`Foo` requires types that implement `Foo` (that is, `Self`) to be `Sized`. -We're going to talk about super traits in more detail in Chapter 19. - -The reason a trait like `Foo` that requires `Self` to be `Sized` is not allowed -to be a trait object is that it would be impossible to implement the trait -`Foo` for the trait object `Foo`: trait objects aren't sized, but `Foo` -requires `Self` to be `Sized`. A type can't be both sized and unsized at the -same time! - -For the second object safety requirement that says all of a trait's methods -must be object safe, a method is object safe if either: - -* It requires `Self` to be `Sized` or -* It meets all three of the following: - * It must not have any generic type parameters - * Its first argument must be of type `Self` or a type that dereferences to - the Self type (that is, it must be a method rather than an associated - function and have `self`, `&self`, or `&mut self` as the first argument) - * It must not use `Self` anywhere else in the signature except for the - first argument - -Those rules are a bit formal, but think of it this way: if your method requires -the concrete `Self` type somewhere in its signature, but an object forgets the -exact type that it is, there's no way that the method can use the original -concrete type that it's forgotten. Same with generic type parameters that are -filled in with concrete type parameters when the trait is used: the concrete -types become part of the type that implements the trait. When the type is -erased by the use of a trait object, there's no way to know what types to fill -in the generic type parameters with. - -An example of a trait whose methods are not object safe is the standard -library's `Clone` trait. The signature for the `clone` method in the `Clone` -trait looks like this: +Trait`Sized`现在就是trait`Foo`的一个*超级trait*, 也就是说trait`Foo`需要实现了`Foo`的类型(即`Self`)是`Sized`. 我们将在第19章中更详细的介绍超trait(supertrait). + +像`Foo`那样要求`Self`是`Sized`的trait不允许成为trait对象的原因是不可能为trait对象`Foo`实现trait`Foo`: trait对象是无确定大小的，但是`Foo`要求`Self`是`Sized`. 一个类型不可能同时既是有大小的又是无确定大小的. + +第二点说对象安全要求一个trait的所有方法必须是对象安全的. 一个对象安全的方法满足下列条件: + +* 它要求`Self`是`Sized`或者 +* 它符合下面全部三点: + * 它不包含任意类型的常规参数 + * 它的第一个参数必须是类型`Self`或一个引用到`Self`的类型(也就是说它必须是一个方法而非关联函数并且以`self`、`&self`或`&mut self`作为第一个参数) + * 除了第一个参数外它不能在其它地方用`Self`作为方法的参数签名 + +虽然这些规则有一点形式化, 但是换个角度想一下: 如果你的方法在它的参数签名的其它地方也需要具体的`Self`类型参数, 但是一个对象又忘记了它的具体类型是什么, 这时该方法就无法使用被它忘记的原先的具体类型. 当该trait被使用时, 被具体类型参数填充的常规类型参数也是如此: 这个具体的类型就成了实现该trait的类型的某一部分, 如果使用一个trait对象时这个类型被抹掉了, 就没有办法知道该用什么类型来填充这个常规类型参数. + +一个trait的方法不是对象安全的一个例子是标准库中的`Clone`trait. `Clone`trait的`clone`方法的参数签名是这样的: ```rust pub trait Clone { @@ -328,21 +297,11 @@ pub trait Clone { } ``` -`String` implements the `Clone` trait, and when we call the `clone` method on -an instance of `String` we get back an instance of `String`. Similarly, if we -call `clone` on an instance of `Vec`, we get back an instance of `Vec`. The -signature of `clone` needs to know what type will stand in for `Self`, since -that's the return type. +`String`实现了`Clone` trait, 当我们在一个`String实例上调用`clone`方法时, 我们会得到一个`String`实例. 同样地, 如果我们在一个`Vec`实例上调用`clone`方法, 我们会得到一个`Vec`实例. `clone`的参数签名需要知道`Self`是什么类型, 因为它需要返回这个类型. -If we try to implement `Clone` on a trait like the `Draw` trait from Listing -17-3, we wouldn't know whether `Self` would end up being a `Button`, a -`SelectBox`, or some other type that will implement the `Draw` trait in the -future. +如果我们想在像17-3中列出的`Draw`trait那样的trait上实现`Clone`, 我们就不知道`Self`将会是一个`Button`, 一个`SelectBox`, 或者是其它的在将来要实现`Draw`trait的类型. -The compiler will tell you if you're trying to do something that violates the -rules of object safety in regards to trait objects. For example, if we had -tried to implement the `Screen` struct in Listing 17-4 to hold types that -implement the `Clone` trait instead of the `Draw` trait, like this: +如果你做了违反trait对象的对象安全性规则的事情, 编译器将会告诉你. 比如, 如果你实现在17-4中列出的`Screen`结构, 你想让该结构像这样持有实现了`Clone`trait的类型而不是`Draw`trait: ```rust,ignore pub struct Screen { @@ -350,7 +309,7 @@ pub struct Screen { } ``` -We'll get this error: +我们将会得到下面的错误: ```text error[E0038]: the trait `std::clone::Clone` cannot be made into an object diff --git a/src/ch17-03-oo-design-patterns.md b/src/ch17-03-oo-design-patterns.md new file mode 100644 index 0000000..5e65df5 --- /dev/null +++ b/src/ch17-03-oo-design-patterns.md @@ -0,0 +1,693 @@ +## 面向对象设计模式的实现 + +让我们看一下状态设计模式和怎样在Rust中来使用它的例子. *状态模式*就是当一个值有多个内部状态时，值的行为改变基于内部状态. Let's look at an example of the state design pattern and how to use it in Rust. +The *state pattern* is when a value has some internal state, and the value's +behavior changes based on the internal state. The internal state is represented +by a set of objects that inherit shared functionality (we'll use structs and +traits since Rust doesn't have objects and inheritance). Each state object is +responsible for its own behavior and the rules for when it should change into +another state. The value that holds one of these state objects doesn't know +anything about the different behavior of the states or when to transition +between states. In the future when requirements change, we won't need to change +the code of the value holding the state or the code that uses the value. We'll +only need to update the code inside one of the state objects to change its +rules, or perhaps add more state objects. + +In order to explore this idea, we're going to implement a blog post workflow in +an incremental way. The workflow that we want our blog posts to follow, once +we're done with the implementation, is: + +1. A blog post starts as an empty draft. +2. Once the draft is done, we request a review of the post. +3. Once the post is approved, it gets published. +4. Only published blog posts return content to print so that we can't + accidentally print the text of a post that hasn't been approved. + +Any other changes attempted on a post should have no effect. For example, if we +try to approve a draft blog post before we've requested a review, the post +should stay an unpublished draft. + +Listing 17-11 shows this workflow in code form. This is an example usage of the +API we're going to implement in a library crate named `blog`: + +Filename: src/main.rs + +```rust,ignore +extern crate blog; +use blog::Post; + +fn main() { + let mut post = Post::new(); + + post.add_text("I ate a salad for lunch today"); + assert_eq!("", post.content()); + + post.request_review(); + assert_eq!("", post.content()); + + post.approve(); + assert_eq!("I ate a salad for lunch today", post.content()); +} +``` + +Listing 17-11: Code that demonstrates the desired +behavior we want our `blog` crate to have + +We want to be able to create a new draft blog post with `Post::new`. Then, we +want to add some text to the blog post while we're in the draft state. If we +try to print out the post's content immediately, though, we shouldn't get any +text, since the post is still a draft. We've added an `assert_eq!` here for +demonstration purposes. Asserting that a draft blog post returns an empty +string from the `content` method would make an excellent unit test in our +library, but we're not going to write tests for this example. + +Next, we want to be able to request a review of our post, and `content` should +still return an empty string while waiting for a review. Lastly, when we +approve the blog post, it should get published, which means the text we added +will be returned when we call `content`. + +Notice that the only type we're interacting with from the crate is the `Post` +type. The various states a post can be in (draft, waiting for review, +published) are managed internally to the `Post` type. The states change due to +the methods we call on the `Post` instance, but we don't have to manage the +state changes directly. This also means we won't make a mistake with the +states, like forgetting to request a review before publishing. + +### Defining `Post` and Creating a New Instance in the Draft State + +Let's get started on the implementation of the library! We know we want to have +a public `Post` struct that holds some content, so let's start with the +definition of the struct and an associated public `new` function to create an +instance of `Post` as shown in Listing 17-12. We're also going to have a +private trait `State`. `Post` will hold a trait object of `Box` inside +an `Option` in a private field named `state`. We'll see why the `Option` is +necessary in a bit. The `State` trait defines all the behavior different post +states share, and the `Draft`, `PendingReview`, and `Published` states will all +implement the `State` trait. For now, the trait does not have any methods, and +we're going to start by defining just the `Draft` state since that's the state +we want to start in: + +Filename: src/lib.rs + +```rust +pub struct Post { + state: Option>, + content: String, +} + +impl Post { + pub fn new() -> Post { + Post { + state: Some(Box::new(Draft {})), + content: String::new(), + } + } +} + +trait State {} + +struct Draft {} + +impl State for Draft {} +``` + +Listing 17-12: Definition of a `Post` struct and a `new` +function that creates a new `Post` instance, a `State` trait, and a `Draft` +struct that implements `State` + +When we create a new `Post`, we set its `state` field to a `Some` value holding +a `Box` pointing to a new instance of the `Draft` struct. This ensures whenever +we create a new instance of `Post`, it'll start out as a draft. Because the +`state` field of `Post` is private, there's no way to create a `Post` in any +other state! + +### Storing the Text of the Post Content + +In the `Post::new` function, we set the `content` field to a new, empty +`String`. In Listing 17-11, we showed that we want to be able to call a method +named `add_text` and pass a `&str` to it to add that text to the content of the +blog post. We're choosing to implement this as a method rather than exposing +the `content` field as `pub` because we want to be able to control how the +`content` field's data is read by implementing a method later. The `add_text` +method is pretty straightforward though, let's add the implementation in +Listing 17-13 to the `impl Post` block: + +Filename: src/lib.rs + +```rust +# pub struct Post { +# content: String, +# } +# +impl Post { + // ...snip... + pub fn add_text(&mut self, text: &str) { + self.content.push_str(text); + } +} +``` + +Listing 17-13: Implementing the `add_text` method to add +text to a post's `content` + +`add_text` takes a mutable reference to `self`, since we're changing the `Post` +instance that we're calling `add_text` on. We then call `push_str` on the +`String` in `content` and pass the `text` argument to add to the saved +`content`. This isn't part of the state pattern since its behavior doesn't +depend on the state that the post is in. The `add_text` method doesn't interact +with the `state` field at all, but it is part of the behavior we want to +support. + +### Content of a Draft Post is Empty + +After we've called `add_text` and added some content to our post, we still want +the `content` method to return an empty string slice since the post is still in +the draft state, as shown on line 8 of Listing 17-11. For now, let's implement +the `content` method with the simplest thing that will fulfill this requirement: +always returning an empty string slice. We're going to change this later once +we implement the ability to change a post's state to be published. With what we +have so far, though, posts can only be in the draft state, which means the post +content should always be empty. Listing 17-14 shows this placeholder +implementation: + +Filename: src/lib.rs + +```rust +# pub struct Post { +# content: String, +# } +# +impl Post { + // ...snip... + pub fn content(&self) -> &str { + "" + } +} +``` + +Listing 17-14: Adding a placeholder implementation for +the `content` method on `Post` that always returns an empty string slice + +With this added `content` method, everything in Listing 17-11 up to line 8 +works as we intend. + +### Requesting a Review of the Post Changes its State + +Next up is requesting a review of a post, which should change its state from +`Draft` to `PendingReview`. We want `post` to have a public method named +`request_review` that will take a mutable reference to `self`. Then we're going +to call an internal `request_review` method on the state that we're holding, and +this second `request_review` method will consume the current state and return a +new state. In order to be able to consume the old state, the second `request_review` +method needs to take ownership of the state value. This is where the `Option` comes +in: we're going to `take` the `Some` value out of the `state` field and leave a +`None` in its place since Rust doesn't let us have unpopulated fields in +structs. Then we'll set the post's `state` value to the result of this +operation. Listing 17-15 shows this code: + +Filename: src/lib.rs + +```rust +# pub struct Post { +# state: Option>, +# content: String, +# } +# +impl Post { + // ...snip... + pub fn request_review(&mut self) { + if let Some(s) = self.state.take() { + self.state = Some(s.request_review()) + } + } +} + +trait State { + fn request_review(self: Box) -> Box; +} + +struct Draft {} + +impl State for Draft { + fn request_review(self: Box) -> Box { + Box::new(PendingReview {}) + } +} + +struct PendingReview {} + +impl State for PendingReview { + fn request_review(self: Box) -> Box { + self + } +} +``` + +Listing 17-15: Implementing `request_review` methods on +`Post` and the `State` trait + +We've added the `request_review` method to the `State` trait; all types that +implement the trait will now need to implement the `request_review` method. +Note that rather than having `self`, `&self`, or `&mut self` as the first +parameter of the method, we have `self: Box`. This syntax means the +method is only valid when called on a `Box` holding the type. This syntax takes +ownership of `Box`, which is what we want because we're transforming the +old state into a new state, and we want the old state to no longer be valid. + +The implementation for the `request_review` method on `Draft` is to return a +new, boxed instance of the `PendingReview` struct, which is a new type we've +introduced that represents the state when a post is waiting for a review. The +`PendingReview` struct also implements the `request_review` method, but it +doesn't do any transformations. It returns itself since requesting a review on +a post that's already in the `PendingReview` state should stay in the +`PendingReview` state. + +Now we can start seeing the advantages of the state pattern: the +`request_review` method on `Post` is the same no matter what its `state` value +is. Each state is responsible for its own rules. + +We're going to leave the `content` method on `Post` as it is, returning an +empty string slice. We can now have a `Post` in the `PendingReview` state, not +just the `Draft` state, but we want the same behavior in the `PendingReview` +state. Listing 17-11 now works up until line 11! + +### Approving a Post Changes the Behavior of `content` + +The `approve` method on `Post` will be similar to that of the `request_review` +method: it will set the `state` to the value that the current state says it +should have when that state is approved. We'll need to add the `approve` method +to the `State` trait, and we'll add a new struct that implements `State`, the +`Published` state. Listing 17-16 shows the new code: + +Filename: src/lib.rs + +```rust +# pub struct Post { +# state: Option>, +# content: String, +# } +# +impl Post { + // ...snip... + pub fn approve(&mut self) { + if let Some(s) = self.state.take() { + self.state = Some(s.approve()) + } + } +} + +trait State { + fn request_review(self: Box) -> Box; + fn approve(self: Box) -> Box; +} + +struct Draft {} + +impl State for Draft { +# fn request_review(self: Box) -> Box { +# Box::new(PendingReview {}) +# } +# + // ...snip... + fn approve(self: Box) -> Box { + self + } +} + +struct PendingReview {} + +impl State for PendingReview { +# fn request_review(self: Box) -> Box { +# Box::new(PendingReview {}) +# } +# + // ...snip... + fn approve(self: Box) -> Box { + Box::new(Published {}) + } +} + +struct Published {} + +impl State for Published { + fn request_review(self: Box) -> Box { + self + } + + fn approve(self: Box) -> Box { + self + } +} +``` + +Listing 17-16: Implementing the `approve` method on +`Post` and the `State` trait + +Similarly to `request_review`, if we call the `approve` method on a `Draft`, it +will have no effect since it will return `self`. When we call `approve` on +`PendingReview`, it returns a new, boxed instance of the `Published` struct. +The `Published` struct implements the `State` trait, and for both the +`request_review` method and the `approve` method, it returns itself since the +post should stay in the `Published` state in those cases. + +Now for updating the `content` method on `Post`: we want to return the value in +the post's `content` field if its state is `Published`, otherwise we want to +return an empty string slice. Because the goal is to keep all the rules like +this in the structs that implement `State`, we're going to call a `content` +method on the value in `state` and pass the post instance (that is, `self`) as +an argument. Then we'll return the value returned from the `content` method on +the `state` value as shown in Listing 17-17: + +Filename: src/lib.rs + +```rust +# trait State { +# fn content<'a>(&self, post: &'a Post) -> &'a str; +# } +# pub struct Post { +# state: Option>, +# content: String, +# } +# +impl Post { + // ...snip... + pub fn content(&self) -> &str { + self.state.as_ref().unwrap().content(&self) + } + // ...snip... +} +``` + +Listing 17-17: Updating the `content` method on `Post` to +delegate to a `content` method on `State` + +We're calling the `as_ref` method on the `Option` because we want a reference +to the value inside the `Option`. We're then calling the `unwrap` method, which +we know will never panic because all the methods on `Post` ensure that the +`state` value will have a `Some` value in it when those methods are done. This +is one of the cases we talked about in Chapter 12 where we know that a `None` +value is never possible even though the compiler isn't able to understand that. + +The `content` method on the `State` trait is where the logic for what content +to return will be. We're going to add a default implementation for the +`content` method that returns an empty string slice. That lets us not need to +implement `content` on the `Draft` and `PendingReview` structs. The `Published` +struct will override the `content` method and will return the value in +`post.content`, as shown in Listing 17-18: + +Filename: src/lib.rs + +```rust +# pub struct Post { +# content: String +# } +trait State { + // ...snip... + fn content<'a>(&self, post: &'a Post) -> &'a str { + "" + } +} + +// ...snip... +struct Published {} + +impl State for Published { + // ...snip... + fn content<'a>(&self, post: &'a Post) -> &'a str { + &post.content + } +} +``` + +Listing 17-18: Adding the `content` method to the `State` +trait + +Note that we need lifetime annotations on this method, like we discussed in +Chapter 10. We're taking a reference to a `post` as an argument, and we're +returning a reference to a part of that `post`, so the lifetime of the returned +reference is related to the lifetime of the `post` argument. + +### Tradeoffs of the State Pattern + +We've shown that Rust is capable of implementing the object-oriented state +pattern in order to encapsulate the different kinds of behavior that a post +should have that depends on the state that the post is in. The methods on +`Post` don't know anything about the different kinds of behavior. The way this +code is organized, we have one place to look in order to find out all the +different ways that a published post behaves: the implementation of the `State` +trait on the `Published` struct. + +An alternative implementation that didn't use the state pattern might have +`match` statements in the methods on `Post` or even in the code that uses +`Post` (`main` in our case) that checks what the state of the post is and +changes behavior in those places instead. That would mean we'd have a lot of +places to look in order to understand all the implications of a post being in +the published state! This would get worse the more states we added: each of +those `match` statements would need another arm. With the state pattern, the +`Post` methods and the places we use `Post` don't need `match` statements and +adding a new state only involves adding a new `struct` and implementing the +trait methods on that one struct. + +This implementation is easy to extend to add more functionality. Here are some +changes you can try making to the code in this section to see for yourself what +it's like to maintain code using this pattern over time: + +- Only allow adding text content when a post is in the `Draft` state +- Add a `reject` method that changes the post's state from `PendingReview` back + to `Draft` +- Require two calls to `approve` before changing the state to `Published` + +A downside of the state pattern is that since the states implement the +transitions between the states, some of the states are coupled to each other. +If we add another state between `PendingReview` and `Published`, such as +`Scheduled`, we would have to change the code in `PendingReview` to transition +to `Scheduled` instead. It would be nicer if `PendingReview` wouldn't need to +change because of the addition of a new state, but that would mean switching to +another design pattern. + +There are a few bits of duplicated logic that are a downside of this +implementation in Rust. It would be nice if we could make default +implementations for the `request_review` and `approve` methods on the `State` +trait that return `self`, but this would violate object safety since the trait +doesn't know what the concrete `self` will be exactly. We want to be able to +use `State` as a trait object, so we need its methods to be object safe. + +The other duplication that would be nice to get rid of is the similar +implementations of the `request_review` and `approve` methods on `Post`. They +both delegate to the implementation of the same method on the value in the +`Option` in the `state` field, and set the new value of the `state` field to +the result. If we had a lot of methods on `Post` that followed this pattern, we +might consider defining a macro to eliminate the repetition (see Appendix E on +macros). + +A downside of implementing this object-oriented pattern exactly as it's defined +for object-oriented languages is that we're not taking advantage of Rust's +strengths as much as we could be. Let's take a look at some changes we can make +to this code that can make invalid states and transitions into compile time +errors. + +#### Encoding States and Behavior as Types + +We're going to show how to rethink the state pattern a bit in order to get a +different set of tradeoffs. Rather than encapsulating the states and +transitions completely so that outside code has no knowledge of them, we're +going to encode the states into different types. When the states are types, +Rust's type checking will make any attempt to use a draft post where we should +only use published posts into a compiler error. + +Let's consider the first part of `main` from Listing 17-11: + +Filename: src/main.rs + +```rust,ignore +fn main() { + let mut post = Post::new(); + + post.add_text("I ate a salad for lunch today"); + assert_eq!("", post.content()); +} +``` + +We still want to create a new post in the draft state using `Post::new`, and we +still want to be able to add text to the post's content. But instead of having +a `content` method on a draft post that returns an empty string, we're going to +make it so that draft posts don't have the `content` method at all. That way, +if we try to get a draft post's content, we'll get a compiler error that the +method doesn't exist. This will make it impossible for us to accidentally +display draft post content in production, since that code won't even compile. +Listing 17-19 shows the definition of a `Post` struct, a `DraftPost` struct, +and methods on each: + +Filename: src/lib.rs + +```rust +pub struct Post { + content: String, +} + +pub struct DraftPost { + content: String, +} + +impl Post { + pub fn new() -> DraftPost { + DraftPost { + content: String::new(), + } + } + + pub fn content(&self) -> &str { + &self.content + } +} + +impl DraftPost { + pub fn add_text(&mut self, text: &str) { + self.content.push_str(text); + } +} +``` + +Listing 17-19: A `Post` with a `content` method and a +`DraftPost` without a `content` method + +Both the `Post` and `DraftPost` structs have a private `content` field that stores the +blog post text. The structs no longer have the `state` field since we're moving +the encoding of the state to the types of the structs. `Post` will represent a +published post, and it has a `content` method that returns the `content`. + +We still have a `Post::new` function, but instead of returning an instance of +`Post`, it returns an instance of `DraftPost`. It's not possible to create an +instance of `Post` right now since `content` is private and there aren't any +functions that return `Post`. `DraftPost` has an `add_text` method defined on +it so that we can add text to `content` as before, but note that `DraftPost` +does not have a `content` method defined! So we've enforced that all posts +start as draft posts, and draft posts don't have their content available for +display. Any attempt to get around these constraints will be a compiler error. + +#### Implementing Transitions as Transformations into Different Types + +So how do we get a published post then? The rule we want to enforce is that a +draft post has to be reviewed and approved before it can be published. A post +in the pending review state should still not display any content. Let's +implement these constraints by adding another struct, `PendingReviewPost`, +defining the `request_review` method on `DraftPost` to return a +`PendingReviewPost`, and defining an `approve` method on `PendingReviewPost` to +return a `Post` as shown in Listing 17-20: + +Filename: src/lib.rs + +```rust +# pub struct Post { +# content: String, +# } +# +# pub struct DraftPost { +# content: String, +# } +# +impl DraftPost { + // ...snip... + + pub fn request_review(self) -> PendingReviewPost { + PendingReviewPost { + content: self.content, + } + } +} + +pub struct PendingReviewPost { + content: String, +} + +impl PendingReviewPost { + pub fn approve(self) -> Post { + Post { + content: self.content, + } + } +} +``` + +Listing 17-20: A `PendingReviewPost` that gets created by +calling `request_review` on `DraftPost`, and an `approve` method that turns a +`PendingReviewPost` into a published `Post` + +The `request_review` and `approve` methods take ownership of `self`, thus +consuming the `DraftPost` and `PendingReviewPost` instances and transforming +them into a `PendingReviewPost` and a published `Post`, respectively. This way, +we won't have any `DraftPost` instances lingering around after we've called +`request_review` on them, and so forth. `PendingReviewPost` doesn't have a +`content` method defined on it, so attempting to read its content is a compiler +error like it is with `DraftPost`. Because the only way to get a published +`Post` instance that does have a `content` method defined is to call the +`approve` method on a `PendingReviewPost`, and the only way to get a +`PendingReviewPost` is to call the `request_review` method on a `DraftPost`, +we've now encoded the blog post workflow into the type system. + +This does mean we have to make some small changes to `main`. Because +`request_review` and `approve` return new instances rather than modifying the +struct they're called on, we need to add more `let post = ` shadowing +assignments to save the returned instances. We also can't have the assertions +about the draft and pending review post's contents being empty string anymore, +nor do we need them: we can't compile code that tries to use the content of +posts in those states any longer. The updated code in `main` is shown in +Listing 17-21: + +Filename: src/main.rs + +```rust,ignore +extern crate blog; +use blog::Post; + +fn main() { + let mut post = Post::new(); + + post.add_text("I ate a salad for lunch today"); + + let post = post.request_review(); + + let post = post.approve(); + + assert_eq!("I ate a salad for lunch today", post.content()); +} +``` + +Listing 17-21: Modifications to `main` to use the new +implementation of the blog post workflow + +Having to change `main` to reassign `post` is what makes this implementation +not quite following the object-oriented state pattern anymore: the +transformations between the states are no longer encapsulated entirely within +the `Post` implementation. However, we've gained the property of having invalid +states be impossible because of the type system and type checking that happens +at compile time! This ensures that certain bugs, such as displaying the content +of an unpublished post, will be discovered before they make it to production. + +Try the tasks suggested that add additional requirements that we mentioned at +the start of this section to see how working with this version of the code +feels. + +Even though Rust is capable of implementing object-oriented design patterns, +there are other patterns like encoding state into the type system that are +available in Rust. These patterns have different tradeoffs than the +object-oriented patterns do. While you may be very familiar with +object-oriented patterns, rethinking the problem in order to take advantage of +Rust's features can give benefits like preventing some bugs at compile-time. +Object-oriented patterns won't always be the best solution in Rust, since Rust +has features like ownership that object-oriented languages don't have. + +## Summary + +No matter whether you think Rust is an object-oriented language or not after +reading this chapter, you've now seen that trait objects are a way to get some +object-oriented features in Rust. Dynamic dispatch can give your code some +flexibility in exchange for a bit of runtime performance. This flexibility can +be used to implement object-oriented patterns that can help with the +maintainability of your code. Rust also has different features, like ownership, +than object-oriented languages. An object-oriented pattern won't always be the +best way to take advantage of Rust's strengths. + +Next, let's look at another feature of Rust that enables lots of flexibility: +patterns. We've looked at them briefly throughout the book, but haven't seen +everything they're capable of yet. Let's go!