Basic CRUD with rust using tide - refactoring

In the last post I started a basic crud using tide and we end up with a simple api that allow us to store dinosaurs information.
Starting from there, let's clean the code a little bit to be more organized. First, we had a closure in every route (let's call it endpoint from here) and will be more clear if we extract that to functions.
async fn dinos_create(mut req: Request<State>) -> tide::Result {
    let dino: Dino = req.body_json().await?;
    // let get a mut ref of our store ( hashMap )
    let mut dinos = req.state().dinos.write().await;
    dinos.insert(String::from(&dino.name), dino.clone());
    let mut res = Response::new(201);
    res.set_body(Body::from_json(&dino)?);
    Ok(res)
}

async fn dinos_list(req: tide::Request<State>) -> tide::Result {
    let dinos = req.state().dinos.read().await;
    // get all the dinos as a vector
    let dinos_vec: Vec<Dino> = dinos.values().cloned().collect();
    let mut res = Response::new(200);
    res.set_body(Body::from_json(&dinos_vec)?);
    Ok(res)
}


( ... )

    app.at("/dinos")
        .post(dinos_create)
        .get(dinos_list);
We moved the closures for this two endpoints to its own functions, let's run the tests to make sure we didn't break anything.
$ cargo test

    Finished test [unoptimized + debuginfo] target(s) in 12.48s
     Running target/debug/deps/tide_basic_crud-3d6db2bae3cd08a5

running 5 tests
test delete_dino ... ok
test index_page ... ok
test list_dinos ... ok
test create_dino ... ok
test update_dino ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out
Great! works as expected. We can move now the rest of the endpoints
app.at("/dinos/:name")
        .get( dinos_get )
        .put( dinos_update )
        .delete( dinos_delete );
Awesome, but now we have five dinos_ functions in the main file. Let's refactor this to be more organized
First, let's create a new struct to represent a rest entity with a base_path field.
struct RestEntity {
    base_path: String,
}
And implement the same methods we had earlier
impl RestEntity {
    async fn create(mut req: Request<State>) -> tide::Result {
        let dino: Dino = req.body_json().await?;
        // let get a mut ref of our store ( hashMap )
        let mut dinos = req.state().dinos.write().await;
        dinos.insert(String::from(&dino.name), dino.clone());
        let mut res = Response::new(201);
        res.set_body(Body::from_json(&dino)?);
        Ok(res)
    }

    async fn list(req: tide::Request<State>) -> tide::Result {
        let dinos = req.state().dinos.read().await;
        // get all the dinos as a vector
        let dinos_vec: Vec<Dino> = dinos.values().cloned().collect();
        let mut res = Response::new(200);
        res.set_body(Body::from_json(&dinos_vec)?);
        Ok(res)
    }

    async fn get(req: tide::Request<State>) -> tide::Result {
        let mut dinos = req.state().dinos.write().await;
        let key: String = req.param("id")?;
        let res = match dinos.entry(key) {
            Entry::Vacant(_entry) => Response::new(404),
            Entry::Occupied(entry) => {
                let mut res = Response::new(200);
                res.set_body(Body::from_json(&entry.get())?);
                res
            }
        };
        Ok(res)
    }

    async fn update(mut req: tide::Request<State>) -> tide::Result {
        let dino_update: Dino = req.body_json().await?;
        let mut dinos = req.state().dinos.write().await;
        let key: String = req.param("id")?;
        let res = match dinos.entry(key) {
            Entry::Vacant(_entry) => Response::new(404),
            Entry::Occupied(mut entry) => {
                *entry.get_mut() = dino_update;
                let mut res = Response::new(200);
                res.set_body(Body::from_json(&entry.get())?);
                res
            }
        };
        Ok(res)
    }

    async fn delete(req: tide::Request<State>) -> tide::Result {
        let mut dinos = req.state().dinos.write().await;
        let key: String = req.param("id")?;
        let deleted = dinos.remove(&key);
        let res = match deleted {
            None => Response::new(404),
            Some(_) => Response::new(204),
        };
        Ok(res)
    }
}
Now we can create a helper function that allow us to register rest like entities to our server, registering five different endpoints to handle the list/create/read/update/delete operations.
fn register_rest_entity(app: &mut Server<State>, entity: RestEntity) {
    app.at(&entity.base_path)
        .get(RestEntity::list)
        .post(RestEntity::create);

    app.at(&format!("{}/:id", entity.base_path))
        .get(RestEntity::get)
        .put(RestEntity::update)
        .delete(RestEntity::delete);
}
And in the server we just need to create a new instance of the struct with the desired base_path and call the helper fn
let dinos_endpoint = RestEntity {
        base_path: String::from("/dinos"),
    };

    register_rest_entity(&mut app, dinos_endpoint);
Great, let's just run the test to ensure that all the operations are still working...
cargo test
   Compiling tide-basic-crud v0.1.0
     Running target/debug/deps/tide_basic_crud-3d6db2bae3cd08a5

running 5 tests
test delete_dino ... ok
test list_dinos ... ok
test create_dino ... ok
test index_page ... ok
test update_dino ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out
Awesome, we just create a nice abstraction that allow us to create easily more rest like entities and implement the basic operations.
That's all for today, in the next iteration I will try to move away from the HashMap and persist the entities information in a db.
As always, I write this as a learning journal and there could be another more elegant and correct way to do it and any feedback is welcome.
I leave here the repo of this example and the pr of the refactor.
Thanks!
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
blog/content/post on  blog [$!?] on ☁️ (us-east-1)
❯ cat basic-crud-with-rust-using-tide-refactoring.md
layout: blog
title: "Basic CRUD with rust using tide - refactoring"
tags:
  • rust
  • notes
  • tide
  • http_rs
  • tide-basic-crud date: 2020-10-03T19:20:36.901Z --- In the last post I started a basic crud using tide and we end up with a simple api that allow us to store dinosaurs information.
  • Starting from there, let's clean the code a little bit to be more organized. First, we had a closure in every route (let's call it endpoint from here) and will be more clear if we extract that to functions.
    async fn dinos_create(mut req: Request<State>) -> tide::Result {
        let dino: Dino = req.body_json().await?;
        // let get a mut ref of our store ( hashMap )
        let mut dinos = req.state().dinos.write().await;
        dinos.insert(String::from(&dino.name), dino.clone());
        let mut res = Response::new(201);
        res.set_body(Body::from_json(&dino)?);
        Ok(res)
    }
    
    async fn dinos_list(req: tide::Request<State>) -> tide::Result {
        let dinos = req.state().dinos.read().await;
        // get all the dinos as a vector
        let dinos_vec: Vec<Dino> = dinos.values().cloned().collect();
        let mut res = Response::new(200);
        res.set_body(Body::from_json(&dinos_vec)?);
        Ok(res)
    }
    
    
    ( ... )
    
        app.at("/dinos")
            .post(dinos_create)
            .get(dinos_list);
    We moved the closures for this two endpoints to its own functions, let's run the tests to make sure we didn't break anything.
    $ cargo test
    
        Finished test [unoptimized + debuginfo] target(s) in 12.48s
         Running target/debug/deps/tide_basic_crud-3d6db2bae3cd08a5
    
    running 5 tests
    test delete_dino ... ok
    test index_page ... ok
    test list_dinos ... ok
    test create_dino ... ok
    test update_dino ... ok
    
    test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out
    Great! works as expected. We can move now the rest of the endpoints
    app.at("/dinos/:name")
            .get( dinos_get )
            .put( dinos_update )
            .delete( dinos_delete );
    Awesome, but now we have five dinos_ functions in the main file. Let's refactor this to be more organized
    First, let's create a new struct to represent a rest entity with a base_path field.
    struct RestEntity {
        base_path: String,
    }
    And implement the same methods we had earlier
    impl RestEntity {
        async fn create(mut req: Request<State>) -> tide::Result {
            let dino: Dino = req.body_json().await?;
            // let get a mut ref of our store ( hashMap )
            let mut dinos = req.state().dinos.write().await;
            dinos.insert(String::from(&dino.name), dino.clone());
            let mut res = Response::new(201);
            res.set_body(Body::from_json(&dino)?);
            Ok(res)
        }
    
        async fn list(req: tide::Request<State>) -> tide::Result {
            let dinos = req.state().dinos.read().await;
            // get all the dinos as a vector
            let dinos_vec: Vec<Dino> = dinos.values().cloned().collect();
            let mut res = Response::new(200);
            res.set_body(Body::from_json(&dinos_vec)?);
            Ok(res)
        }
    
        async fn get(req: tide::Request<State>) -> tide::Result {
            let mut dinos = req.state().dinos.write().await;
            let key: String = req.param("id")?;
            let res = match dinos.entry(key) {
                Entry::Vacant(_entry) => Response::new(404),
                Entry::Occupied(entry) => {
                    let mut res = Response::new(200);
                    res.set_body(Body::from_json(&entry.get())?);
                    res
                }
            };
            Ok(res)
        }
    
        async fn update(mut req: tide::Request<State>) -> tide::Result {
            let dino_update: Dino = req.body_json().await?;
            let mut dinos = req.state().dinos.write().await;
            let key: String = req.param("id")?;
            let res = match dinos.entry(key) {
                Entry::Vacant(_entry) => Response::new(404),
                Entry::Occupied(mut entry) => {
                    *entry.get_mut() = dino_update;
                    let mut res = Response::new(200);
                    res.set_body(Body::from_json(&entry.get())?);
                    res
                }
            };
            Ok(res)
        }
    
        async fn delete(req: tide::Request<State>) -> tide::Result {
            let mut dinos = req.state().dinos.write().await;
            let key: String = req.param("id")?;
            let deleted = dinos.remove(&key);
            let res = match deleted {
                None => Response::new(404),
                Some(_) => Response::new(204),
            };
            Ok(res)
        }
    }
    Now we can create a helper function that allow us to register rest like entities to our server, registering five different endpoints to handle the list/create/read/update/delete operations.
    fn register_rest_entity(app: &mut Server<State>, entity: RestEntity) {
        app.at(&entity.base_path)
            .get(RestEntity::list)
            .post(RestEntity::create);
    
        app.at(&format!("{}/:id", entity.base_path))
            .get(RestEntity::get)
            .put(RestEntity::update)
            .delete(RestEntity::delete);
    }
    And in the server we just need to create a new instance of the struct with the desired base_path and call the helper fn
    let dinos_endpoint = RestEntity {
            base_path: String::from("/dinos"),
        };
    
        register_rest_entity(&mut app, dinos_endpoint);
    Great, let's just run the test to ensure that all the operations are still working...
    cargo test
       Compiling tide-basic-crud v0.1.0
         Running target/debug/deps/tide_basic_crud-3d6db2bae3cd08a5
    
    running 5 tests
    test delete_dino ... ok
    test list_dinos ... ok
    test create_dino ... ok
    test index_page ... ok
    test update_dino ... ok
    
    test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out
    Awesome, we just create a nice abstraction that allow us to create easily more rest like entities and implement the basic operations.
    That's all for today, in the next iteration I will try to move away from the HashMap and persist the entities information in a db.
    As always, I write this as a learning journal and there could be another more elegant and correct way to do it and any feedback is welcome.
    I leave here the repo of this example and the pr of the refactor.
    Thanks!

    29

    This website collects cookies to deliver better user experience

    Basic CRUD with rust using tide - refactoring