I want to make an API first application in Laravel. I don't know what is the best approach to do this, I will explain what I am trying to do, but please feel free to give answers how to do this in a different way.
I don't want all my frontend to be written in javascript and parse the JSON output of the API with angular.js or something similar. I want my Laravel application to produce the HTML views. I am trying to go down the road of having two controllers one on for the API and one for the web. For the show User action my routes.php looks like this:
# the web controller
Route::controller('user', 'WebUserController');
# the api controller
Route::group(array('prefix' => 'api'), function() {
Route::resource('user', 'UserController');
});
So /user will take me to WebUserController and /api/user will take me to the UserController. Now I want to put all my logic in the API UserController, and call its actions from the WebUserController. Here is the code for both of them:
class UserController extends BaseController
{
public function show($id)
{
$user = User::find($id);
return Response::json(array('success'=>true,'user'=>$user->toArray()));
}
}
class WebUserController extends UserController
{
public function getView($id)
{
# call the show method of the API's User Controller
$response = $this->show($id);
return View::make('user.view')->with('data', $response->getData());
}
}
In the WebUserController I am able to get the json content of the response with getData(), but I am not able to get the headers and status code (they are protected properties of Illuminate\Http\JsonResponse).
I think that my approach might not be the best, so I am open to suggestions how to make this app.
EDIT: The question how to get the headers and status of the response has been answered by Drew Lewis, but I still think that there might be a better way how to design this
You should utilize the Repository / Gateway design pattern: please see the answers here.
For example, when dealing with the User model, first create a User Repository. The only responsibility of the user repository is to communicate with the database (performing CRUD operations). This User Repository extends a common base repository and implements an interface containing all methods you require:
class EloquentUserRepository extends BaseRepository implements UserRepository
{
public function __construct(User $user) {
$this->user = $user;
}
public function all() {
return $this->user->all();
}
public function get($id){}
public function create(array $data){}
public function update(array $data){}
public function delete($id){}
// Any other methods you need go here (getRecent, deleteWhere, etc)
}
Then, create a service provider, which binds your user repository interface to your eloquent user repository. Whenever you require the user repository (by resolving it through the IoC container or injecting the dependency in the constructor), Laravel automatically gives you an instance of the Eloquent user repository you just created. This is so that, if you change ORMs to something other than eloquent, you can simply change this service provider and no other changes to your codebase are required:
use Illuminate\Support\ServiceProvider;
class RepositoryServiceProvider extends ServiceProvider {
public function register() {
$this->app->bind(
'lib\Repositories\UserRepository', // Assuming you used these
'lib\Repositories\EloquentUserRepository' // namespaces
);
}
}
Next, create a User Gateway, who's purpose is to talk to any number of repositories and perform any business logic of your application:
use lib\Repositories\UserRepository;
class UserGateway {
protected $userRepository;
public function __construct(UserRepository $userRepository) {
$this->userRepository = $userRepository;
}
public function createUser(array $input)
{
// perform any sort of validation first
return $this->userRepository->create($input);
}
}
Finally, create your User web controller. This controller talks to your User Gateway:
class UserController extends BaseController
{
public function __construct(UserGatway $userGateway)
{
$this->userGateway = $userGateway;
}
public function create()
{
$user = $this->userGateway->createUser(Input::all());
}
}
By structuring the design of your application in this way, you get several benefits: you achieve a very clear separation of concerns, since your application will be adhering to the Single Responsibility Principle (by separating your business logic from your database logic) . This enables you to perform unit and integration testing in a much easier manner, makes your controllers as slim as possible, as well as allowing you to easily swap out Eloquent for any other database if you desire in the future.
For example, if changing from Eloquent to Mongo, the only things you need to change are the service provider binding as well as creating a MongoUserRepository which implements the UserRepository interface. This is because the repository is the only thing talking to your database - it has no knowledge of anything else. Therefore, the new MongoUserRepository might look something like:
class MongoUserRepository extends BaseRepository implements UserRepository
{
public function __construct(MongoUser $user) {
$this->user = $user;
}
public function all() {
// Retrieve all users from the mongo db
}
...
}
And the service provider will now bind the UserRepository interface to the new MongoUserRepository:
$this->app->bind(
'lib\Repositories\UserRepository',
'lib\Repositories\MongoUserRepository'
);
Throughout all your gateways you have been referencing the UserRepository, so by making this change you're essentially telling Laravel to use the new MongoUserRepository instead of the older Eloquent one. No other changes are required.
You should be use Repository for this design.
Example -
//UserRepository Class
class UserRepository {
public function getById($id)
{
return User::find($id);
}
}
// WebUser Controller
class WebUserController extends BaseController {
protected $user;
public function __construct(UserRepository $user)
{
$this->user = $user;
}
public function show($id)
{
return View::make('user.view')->with('data', $this->user->getById($id));
}
}
// APIUser Controller
class UserController extends BaseController {
protected $user;
public function __construct(UserRepository $user)
{
$this->user = $user;
}
public function show($id)
{
$data =>$this->user->getById($id);
return Response::json(array('success'=>true,'user'= $data->toArray()));
}
}
Checkout Laravel's RESTful controllers:
http://laravel.com/docs/controllers#restful-controllers
Their docs do a pretty good job.
But even better is this tutorial:
http://code.tutsplus.com/tutorials/laravel-4-a-start-at-a-restful-api-updated--net-29785
This is a video by Jeffrey Way he is one of the better Laravel developers. In this tutorial he is connecting a BackboneJS application to a RESTful service that he sets up in Laravel. It doesn't get any better then this. I can write you a lot of boilerplate, but just learn it by watching a nice video and having a coffee. ;)
https://www.youtube.com/watch?v=uykzCfu1RiQ
I have a response to the problem you are having with the Response.
You can get the headers, status code and data from the Response.
// your data
$response->getData();
// the status code of the Response
$response->getStatusCode();
// array of headers
$response->headers->all();
// array of headers with preserved case
$response->headers->allPreserveCase();
$response->headers is a Symfony\Component\HttpFoundation\ResponseHeaderBag which inherits from Symfony\Component\HttpFoundation\HeaderBag
I would also recommend using a repository.
Attempting to call one controller from another would be falling into a pattern called HMVC (Hierarchical model–view–controller).
This means that your entire application relies on lower modules.
In this case, your API would serve as a repository for your data (which isn't the worst thing in the world at first).
However, when you then modify the structure of how data is returned in your API, everything else relying on it would have to know how to respond.
Say you wanted to have authorization checks to see if a logged in user should be able to see the details of a returned user and there was an error.
In the API, you would return a Response object with a 403 forbidden code and some meta data.
Your HTML controller would have to know how to handle this.
Contrast this to a repository which could throw an exception.
public function findById ($id)
{
$user = User::findOrFail($id);
if (Auth::user->hasAccessTo($user)) {
return $user;
} else {
throw new UnauthorizedAccessException('you do not have sufficient access to this resource');
}
}
And your API controller would look more like this:
public function show($id)
{
try {
return $this->user->findById($id);
} catch (UnauthorizedAccessException $e) {
$message = $e->getMessage();
return Response::json('403', ['meta' => ['message' => $message]]));
}
}
Your HTML controller would then look like this:
public function show($id)
{
try {
$user = $this->user->findById($id);
} catch (UnauthorizedAccessException $e) {
Session::flash('error', $e->getMessage());
// Redirect wherever you would like
return Response::redirect('/');
}
}
This gives you very reusable code and let's you change your controller implementations independently without worry of changing the other's behavior.
I wrote more on how to implement the repository pattern in this post: you can ignore the interface and skip right to the implementations if you would like.
Related
I have to call the method of another controller.
I use following code to make a call.
app('App\Http\Controllers\ApiUserController')->getList();
This is working fine.
But I want to try using use function so that I dont have to repeat all line
use App\Http\Controllers\ApiUserController;
class MyMethods
{
public function index()
{
app('ApiUserController')->getList()
Did I made some mistake here?
Instead of using app function, you will need to go through OOP way like so:
use App\Http\Controllers\ApiUserController;
class MyMethods
{
public function index()
{
$apiUserController = new ApiUserController();
$apiUserController->getList();
However, as many people have mentioned here, it is not really the best practice to call a method of one controller from the another.
So if I were at your place, I would create a helper, register its alias in config and use that helper to get the list in both places.
I hope it helps
Calling controller from other controller or other objects is not a good practice. Here is a good article explaining why. Also "fat" controllers is less preferable than "thin" controllers.
You should define a service layer object with common logic and use it. Create a service object and register it with one of service providers.
namespace App\Providers;
use Illuminate\Support\ServiceProvider;
use App\Services\YourUserService;
class AppServiceProvider extends ServiceProvider
{
public function register()
{
$this->app->singleton(YourUserService::class);
}
}
After that you can use your service in DI style.
use App\Services\YourUserService;
class MyMethods
{
protected $userService;
public function __construct(YourUserService $userService)
{
$this->userService = $userService;
}
public function index()
{
$this->userService->foo();
}
}
Why should I use dependency injection?
I agree with the answer Learner has given above, however, I wouldn't recommend it in terms of code organisation and testability.
Looking at the code, I can see that you need to get list of users and thats why you have to call the api user controller from another controller. However, you can easily extract the logic out to a service or even a trait.
if you were to use a trait then you could do some thing like following,
trait ApiUser {
public function getList()
{
// get the list for users from api
}
}
//Then you can simply use this trait any where you want,
class SomeController
{
// correct namespace for ApiUser trait
use ApiUser;
}
Another way of doing it, which i love to use again and again depending on the scenario; is to stick with the principle of coding to interface not to implementation. That would be some thing like follow.
interface ApiUserInterface
{
public function getList();
}
class ApiUser implements ApiUserInterface
{
public function getList()
{
// logic to get users from api
}
}
Make sure that when application requires the interface, it knows where to find its implementation. If you using Laravel, then you could register your interface to class in AppServiceProvider
Once that's done, you can use this service any where you want as a contract.
class OneController
{
protected $apiUserContract;
public function __construct(ApiUserInterface $apiUserContract)
{
$this->apiUserContract = $apiUserContract;
}
public function index()
{
// You can retrieve the list of the contract
$this->apiUserContract->getList();
}
}
// you could also just typehint the contact in method without requiring
// it in constructor and it will get resolved out of IOC i.e. container
class AnotherController
{
public function index(ApiUserInterface $apiUserContract)
{
// You can retrieve the list of the contract
$apiUserContract->getList();
}
}
Let me know if you need further explanation and hope it helps
In my project I have decided to use the Service Pattern (Possibly with the Repository Pattern) to deal with the business logic in my application. I have for example a Client model which represents a customer and a corresponding ClientService that is responsible for client-specific business logic.
class ClientService extends Service implements ClientServiceContract
{
public function create(array $attributes)
{
// Create a new client...
}
public function doSomethingElse(Client $client)
{
// Do something else
}
}
Say for example I have another service UserService, which is similar to the ClientService above in that it has methods to create and do other things to User models.
Now on my site, imagine that I have a form that someone can fill in to register their interest in becoming a client. In my back end system I would like to create a button that takes a client's interest record ClientInterest and creates a Client, a User, associates the two and finally sends an e-mail to the new user with the details.
Where, when using the service pattern would it be best to put this logic?
I have considered:
Create a service and method ClientInterestService::createClientAndUser(...) which would use the ClientService and UserService classes to create the Client and User instances and then carry out the association before triggering an event which sends the email. This approach means that I'm not duplicating code, however I'm coupling classes together and I'm breaking some SOLID principles. I'm not sure but I have a feeling this wouldn't be great for testing either.
As described above, create a service class and method to carry out the logic, but instead of using the other two services I would write the logic to create the Client and User instances, carry out the association and trigger the event to send the email. This approach feels nicer, my code is more loosely coupled and I'm not breaking any SOLID principles, however, I'm potentially duplicating code.
Simply put the logic that I would have had in ClientInterestService::createClientAndUser(...) in my controller. Doing this would mean that I have business logic in my controller which kind of defeats the point of having services.
I think if you break this down into smaller steps you can achieve DRY architecture. The steps I'm seeing are:
Create Client
Create User
Associate (via pivot table, junction table etc)
Email
To avoid having the dreaded duplicate code you'd create a method around each of these in your service class or classes. You'd then create an action encapsulating all of the steps involved based around these methods.
Don't be scared to implement things outside of your service class - this doesn't mean it is outside of your service layer.
I see registering client interest as an action. You follow synchronous steps to achieve your desired action. So based on methods like creating a user, client etc we can build an action to register client interest, like so:
<?php
class ClientService {
public function addAction(IAction $action)
{
return $action->process();
}
public function createUser() {} // business logic for creating a user.
public function createClient() {} // business logic for creating a client.
public function createAssociation() {} // business logic for creating an association.
}
interface IAction {
public function process();
}
class RegisterClientInterestAction implements IAction {
protected $client;
public function __construct(ClientService $client)
{
$this->client = $client;
}
public function process()
{
$this->createUser()->createClient()->createAssociation();
}
private function createUser() {} // interact with your client service to call the method $client->createUser()
private function createClient() {} // interact with your client service to call the method $client->createClient()
private function createAssociation() {} // interact with your client service to call the method $client->createAssociation()
}
//USAGE
$service = new ClientService;
$results = $service->addAction(new RegisterClientInterestAction($service));
?>
By doing it this way you are able to utilise the createUser etc methods in a new action but without duplicating the code. By having the addAction on the service class you are still executing the business logic inside of your service layer.
If two or more services are required, I'd take a slightly different approach by moving where I would execute the action.
In terms of handling more than one service you can use DI within the constructor of your action.
Like this:
<?php
class Service {
public function addAction(IAction $action)
{
return $action->process();
}
// Other stuff for a base service...
}
class UserService extends Service {
public function createUser() {} // business logic for creating a user.
}
class ClientService extends Service {
public function createClient() {} // business logic for creating a client.
public function createAssociation() {} // business logic for creating an association.
}
interface IAction {
public function process();
}
class RegisterClientInterestAction implements IAction {
protected $client;
protected $service;
public function __construct(ClientService $client, UserService $user)
{
$this->user = $user;
$this->client = $client;
}
public function process()
{
$this->createUser()->createClient()->createAssociation();
}
private function createUser() {} // interact with your user service to call the method $client->createUser()
private function createClient() {} // interact with your client service to call the method $client->createClient()
private function createAssociation() {} // interact with your client service to call the method $client->createAssociation()
}
//USAGE
$service = new Service;
$results = $service->addAction(new RegisterClientInterestAction(new ClientService, new UserService));
?>
What feels best for me is your proposed solution of #2.
What I like to do is build out the two service classes and see what the duplication there is, then refactor/extract any duplication to another class. This way all classes are very testable and you have the least chance of breaking any SOLID principles.
I am currently facing a very interesting dilemma with my architecture and implementation.
I have an interface called ServiceInterface which have a method called execute()
Then I have two different implementations for this interface: Service1 and Service2, which implements the execute method properly.
I have a controller called MainController and this controller has a "type-hint" for the ServiceInterface (dependency injection), it means that both, Service1 and Service2, can be called as resolution for that dependency injection.
Now the fun part:
I do not know which of those implementations to use (Service1 or Service2) because I just know if I can use one or other based on a user input from a previous step.
It means the user choose a service and based on that value I know if a can use Service1 or Service2.
I am currently solving the dependency injection using a session value, so depending of the value I return an instance or other, BUT I really think that it is not a good way to do it.
Please, let me know if you faced something similar and, how do you solve it, or what can I do to achieve this in the right way.
Thanks in advance. Please let me know if further information is required.
Finally, after some days of researching and thinking a lot about the best approach for this, using Laravel, I finally solved it.
I have to say that this was especially difficult in Laravel 5.2 because, in this version, the Session middleware only is executed in the controllers used in a route, it means that if for some reason I used a controller (not linked for a rote) and try to get access to the session it is not going to be possible.
So, because I cannot use the session, I decided to use URL parameters. Here you have the solution approach; I hope some of you found it useful.
so, you have an interface:
interface Service
{
public function execute();
}
Then a couple of implementations for the interface:
Service one:
class ServiceOne implements Service
{
public function execute()
{
.......
}
}
Service two.
class ServiceTwo implements Service
{
public function execute()
{
.......
}
}
The interesting part is that I have a controller with a function with a dependency with the Service interface. Still, I need to resolve it dynamically to ServiceOne or ServiceTwo based on user input. So:
The controller
class MyController extends Controller
{
public function index(Service $service, ServiceRequest $request)
{
$service->execute();
.......
}
}
Please note that ServiceRequest, validated that the request already have the parameter that we need to resolve the dependency (call it 'service_name')
Now, in the AppServiceProvider we can resolve the dependency in this way:
class AppServiceProvider extends ServiceProvider
{
public function boot()
{
}
public function register()
{
//This specific dependency is going to be resolved only if
//the request has the service_name field stablished
if(Request::has('service_name'))
{
//Obtaining the name of the service to be used (class name)
$className = $this->resolveClassName(Request::get('service_name')));
$this->app->bind('Including\The\Namespace\For\Service', $className);
}
}
protected function resolveClassName($className)
{
$resolver = new Resolver($className);
$className = $resolver->resolveDependencyName();
return $className;
}
}
So now all the responsibility is for the Resolver class. This class basically use the parameter passed to the constructor to return the full name (with namespace) of the class that is going to be used as an implementation of the Service interface:
class Resolver
{
protected $name;
public function __construct($className)
{
$this->name = $className;
}
public function resolveDependencyName()
{
//This is just an example, you can use whatever as 'service_one'
if($this->name === 'service_one')
{
return Full\Namespace\For\Class\Implementation\ServiceOne::class;
}
if($this->name === 'service_two')
{
return Full\Namespace\For\Class\Implementation\ServiceTwo::class;
}
//If none, so throw an exception because the dependency can not be resolved
throw new ResolverException;
}
}
Well, I really hope it helps some of you.
Best wishes!
---------- EDIT -----------
I just realize that it is not a good idea to use the request data directly inside the container of Laravel. It really is going to cause some trouble in the long term.
The best way is to directly register all the possible instances supported (serviceone and servicetwo) and then resolve one of them directly from a controller or a middleware, so then is the controller "who decides" what service to use (from all the available) based on the input from the request.
In the end, it works at the same, but it is going to allow you to work more naturally.
I have to say thanks to rizqi, a user from the questions channel of the slack chat of Laravel.
He personally created a golden article about this. Please read it because it solves this issue completely and in a very right way.
laravel registry pattern
The fact that you define that your controller works with ServiceInterface is ok
If you have to choose the concrete implementation of the service basing on a previous step (that, as i've understood, happens in a previous request) storing the value in session or in database is right too, as you have no alternative: to choose the implementation you have to know the value of the input
The important point is to 'isolate' the resolution of the concrete implementation from the input value in one place: for example create a method that takes this value as a parameter and returns the concrete implementation of the service from the value:
public function getServiceImplementation($input_val)
{
switch($input_val)
{
case 1 : return new Service1();
case 2 : return new Service2();
}
}
and in your controller:
public function controllerMethod()
{
//create and assign the service implementation
$this->service = ( new ServiceChooser() )->getServiceImplementation( Session::get('input_val') );
}
In this example i've used a different class to store the method, but you can place the method in the controller or use a Simple Factory pattern, depending on where the service should be resolved in your application
It's an interesting problem. I'm currently using Laravel 5.5 and have been mulling it over. I also want my service provider to return a specific class (implementing an interface) based upon user input. I think it's better to manually pass the input from the controller so it's easier to see what's going on. I would also store the possible values of the class names in the config.
So based upon the Service classes and interface you've defined above i came up with this:
/config/services.php
return [
'classes': [
'service1' => 'Service1',
'service2' => 'Service2',
]
]
/app/Http/Controllers/MainController.php
public function index(ServiceRequest $request)
{
$service = app()->makeWith(ServiceInterface::class, ['service'=>$request->get('service)]);
// ... do something with your service
}
/app/Http/Requests/ServiceRequest.php
public function rules(): array
$availableServices = array_keys(config('services.classes'));
return [
'service' => [
'required',
Rule::in($availableServices)
]
];
}
/app/Providers/CustomServiceProvider.php
class CustomServiceProvider extends ServiceProvider
{
public function boot() {}
public function register()
{
// Parameters are passed from the controller action
$this->app->bind(
ServiceInterface::class,
function($app, $parameters) {
$serviceConfigKey = $parameters['service'];
$className = '\\App\\Services\\' . config('services.classes.' . $serviceConfigKey);
return new $className;
}
);
}
}
This way we can validate the input to ensure we are passing a valid service, then the controller handles passing the input from the Request object into the ServiceProvider. I just think when it comes to maintaining this code it will be clear what is going on as opposed to using the request object directly in the ServiceProvider.
PS Remember to register the CustomServiceProvider!
I find the best way to deal with this is using a factory pattern. You can create a class say ServiceFactory and it has a single method create() it can accept an argument which is used to dynamically choose which concrete class to instantiate.
It has a case statement based on the argument.
It will use App::make(ServiceOne::class) or App::make(ServiceTwo::class).depending on which one is required.
You are then able to inject this into your controller (or service which depends on the factory).
You can then mock it in a service unit test.
Recently, I had to implement a similar logic where I was to implement a method to perform mobile top-ups for multiple networks in our application. So, I decided to implement the logic using Factory and Bridge pattern. Factory to create an instance of the concrete Service class based on the user input, and then, the Bridge pattern to set closely related classes into separate hierarchies and route the request to the respective class.
In the controller's method, both Factory and Service classes are injected. The TopUpServiceFactory's create method creates an object of the concrete class. The TopUpService class then routes the request to that concrete class method.
class TopUpController extends Controller
{
public function topUp(Request $request, TopUpServiceFactoryInterface $serviceFactory, TopUpServiceInterface $topUpService)
{
$serviceFactory->create($request->networkCode);
$topUpService->TopUp($request->all());
}
}
The TopUpServiceFactoryInterface and TopUpServiceInterface are bound to TopUpServiceFactory and TopUpService concrete Classes respectively in Service Container.
class AppServiceProvider extends ServiceProvider
{
public function register()
{
$this->app->bind(TopUpServiceFactoryInterface::class, TopUpServiceFactory::class);
$this->app->bind(TopUpServiceInterface::class, TopUpService::class);
}
}
The create method accepts user input and creates an object of the respective class based on the user input.
class TopUpServiceFactory implements TopUpServiceFactoryInterface
{
public function create(string $networkCode)
{
switch ($networkCode) {
case 'network1':
app()->bind(NetworkServiceInterface::class, Network1Service::class);
break;
case 'network2':
app()->bind(NetworkServiceInterface::class, Network2Service::class);
break;
default:
app()->bind(NetworkServiceInterface::class, DefaultNetworkService::class);
break;
}
}
}
The Service Class then picks the object of NetworkService Class and forwards the request.
class TopUpService implements TopUpServiceInterface
{
public function topUp(array $requestParams)
{
$networkService = app()->get(NetworkServiceInterface::class);
$networkService->topUp($requestParams);
}
}
All network's concrete classes implement a common interface NetworkServiceInterface, which is used to inject dependency dynamically, implementing Liskov Substitution Principle
class Network1Service implements NetworkServiceInterface
{
public function topUp(array $requestParam)
{
Process Topup ......
}
}
class Network2Service implements NetworkServiceInterface
{
public function topUp(array $requestParam)
{
Process Topup ......
}
}
...
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 8 years ago.
Improve this question
I've been picking other developers' brains on the concept of "fat models, skinny controllers" after reading:
http://culttt.com/2013/07/01/setting-up-your-first-laravel-4-controller/
http://culttt.com/2013/05/13/setting-up-your-first-laravel-4-model/
Most respondents are using what I'd consider fat controllers.
While the topic has come up on Stack Overflow I haven't found thorough description of the method in practice.
I just found an old related question here.
Skinny Controllers
What you are going to see in PHP (vanilla or Laravel or Symfony) more and more are the skinniest controllers ever. This is something you already see in Rails, and people are also starting to call it (with some other practices) hexagonal. One line of code is all you need in your controller, actually they say this should be a goal for all your methods. This is an example with, yes, a little bit more than that, but still skinny:
<?php
class PostController extends Controller {
private $repository;
public function __construct(PostRepositoryInterface $repository)
{
$this->repository = $repository;
}
public function store()
{
try
{
$this->repository->create(Input::all());
}
catch (ValidationException $e)
{
return Redirect::back()->withInput()->withErrors($e->all());
}
return Redirect::route('posts');
}
}
A controller is a bridge between the HTTP requests, your business logic and your presentation layer. So it should receive one request, send it to an injected object which will process it and redirect to the route (or render a view) responsible for giving feedback to a client (or user). Everything else, including validation, should happen in your repositories, services, models (MVC, yay!), etc.
But we could refactor this controller, in the hexagonal way, to reach the one-line-per-method goal:
<?php
class PostController extends Controller {
private $repository;
public function __construct(PostRepositoryInterface $repository)
{
$this->repository = $repository;
}
public function store()
{
return $this->repository->create(Input::all(), $this);
}
public function createSucceeded()
{
return Redirect::route('posts');
}
public function createFailed()
{
return Redirect::back()->withInput()->withErrors($e->all());
}
}
Basically your repository classes will use the own caller ($this) to fire the succeeded and failed methods.
Fat Repositories / Services / Models
Models are too related to your data, sometimes they are your ORM and talk directly to your database server, so, these days you'll see people use repositories and services as layers between them.
Repositories
A repository is a class that, by talking directly to your models, processes and gather the information your application needs. Your application should not be aware of what is necessary to select some information in your database, select, where, order, group by, those are things sometimes only your models should be aware of, so this is a repository:
class PostRepository implements PostRepositoryInterface {
private $model;
public function __construct(PostInterface $model)
{
$this->model = $model;
}
public function create($input)
{
return $this->model->create($input);
}
public findBySlug($slug)
{
return $this->model->where('slug', $slug)->first();
}
}
Services
Everything that doesn't belongs directly to your business logic, mostly external services, the farthest from your application code, the more decoupled you build them, the better. Creating external packages (Composer packages) for those services are a good way of decoupling them from everything else, and you if you make them framework agnostic you're entitled to receive 10 Sturgeon points. In Laravel you can create services by integrating three kind of classes:
1) Service Class(es): responsible for doing what your service must do, all your service logic goes here.
2) Service Provider: responsible for booting up your service and adding it to Laravel's IoC container so it can be ready to use at any time, but note that Laravel will only instantiate your service classes when your application really use them.
3) Facade: lets you access your service from anywhere in your application using the static (::) syntax:
Mailer::send($user->id, 'Thanks for registering', 'emails.registered');
This the Mailer service:
Service Class
<?php namespace ACR\Services\Mailer;
use Illuminate\Mail\Mailer as IlluminateMailer;
use Sentry;
class Service {
public function __construct(IlluminateMailer $mailer)
{
$this->mailer = $mailer;
}
public function send($userId, $subject, $view, $data = [])
{
return $this->mailer->queue($view, $data, function($message) use ($userId, $subject)
{
$user = Sentry::findUserById($userId);
$message->to($user->email, $user->name);
$message->subject($subject);
});
}
}
Service Provider
<?php namespace ACR\Services\Mailer;
use Illuminate\Support\ServiceProvider as IlluminateServiceProvider;
use ACR\Services\Mailer\Service as Mailer;
class ServiceProvider extends IlluminateServiceProvider {
/**
* Indicates if loading of the provider is deferred.
*
* #var bool
*/
protected $defer = true;
/**
* Register the service provider.
*
* #return void
*/
public function register()
{
$this->app->bind('acr.mailer', function($app) {
return new Mailer($app->make('mailer'));
});
}
/**
* Get the services provided by the provider.
*
* #return array
*/
public function provides()
{
return array('acr.mailer');
}
}
Facade
<?php namespace ACR\Services\Mailer;
use Illuminate\Support\Facades\Facade as IlluminateFacade;
class Facade extends IlluminateFacade {
protected static function getFacadeAccessor() { return 'acr.mailer'; }
}
Models / ORM
Those guys should be highly swappable, today you may be using a Eloquent as your ORM, storing data in a database, but you might need to change it to something else, some foks are storing 100% of their data in Redis, so you better be prepared for a change like this by using an Interface (contract) layer between your ORM and your domain loginc and really develop for your interfaces, not your concrete classes. Taylor Otwell in his book even say that you should completely delete your models folder.
interface PostInterface {
public function all();
public function find($id);
}
class DbPost extends Eloquent implements PostInterface {
}
class RedisPost extends Eloquent implements PostInterface {
}
The idea behind this is to swap implementations easily, so in Laravel you can use the IoC container to tell Laravel which implementation you are using:
App::bind('PostInterface', 'DbPost');
So, if you have a Repository is using your PostInterface:
class PostRepository implements PostRepositoryInterface {
private $model;
public function __construct(PostInterface $model)
{
$this->model = $model;
}
}
Laravel IoC container will automatically instantiate this repository with an instance of DbPost. And if you ever need to change it to Redis, you just need to change one line:
App::bind('PostInterface', 'RedisPost');
Views / Presenters
The dumbest the awesomer.
Views
Views should be responsible only for displaying information. Views should not be aware of your models, services, repositories, or anything else in your system. Views should be editable by webesigners so, the more code you have on them, the more bugs your non-php-programmer-designer will add to them. Your controller should gather the information from your repositories and pass them to your views:
<?php
class PostController extends Controller {
private $repository;
public function __construct(PostRepositoryInterface $repository)
{
$this->repository = $repository;
}
public function index()
{
return View::make('posts.index')->with('posts', $this->repository->getPaginated());
}
}
And the only responsibility of your view should be show that data:
#extends('layout')
#section('contents')
<ul>
#foreach($posts as $post)
<li>
{{ $post->title }} - {{ $post->author }} - {{ $post->published_at }}
</li>
#endforeach
</ul>
{{ $users->links() }}
#stop
Presenters
How do you format your data? You write raw properties in your views, but you should, behind the scenes, be using presenters to, yeah, present your data. Presenters usually use the Decorator Design Pattern to format your data to be presented in your pages. This is an example using Shawn McCool's LaravelAutoPresenter:
<?php namespace App\Presenters;
use McCool\LaravelAutoPresenter\BasePresenter;
class Post extends BasePresenter {
public function __construct(UserModel $user)
{
$this->resource = $user;
}
public function author()
{
return $this->resource->author->name;
}
public function published_at()
{
return $this->date($this->resource->created_at);
}
public function dateTime($date)
{
return \Carbon\Carbon::createFromFormat('d-m-Y', $date, 'Sao_Paulo/Brazil')
->toFormattedDateString();
}
}
Related Books
Taylor Otwell's Laravel: From Apprentice To Artisan
Chris Fidao's Implementing Laravel
Read books From Apprentice To Artisan and Implementing Laravel by Chris Fidao and now i don't know how to correctly work with Models in Repositories.
In Implementing laravel book author is working with models in this way:
Example #1
<?php
use MyApp\Interfaces\UserInterface;
use Illuminate\Database\Eloquent\Model;
class UserRepository implements UserInterface
{
protected $user;
public function __construct(Model $user)
{
$this->user = $user;
}
public function find($userId)
{
return $this->user->find($userId);
}
}
But that can by done in other way, not injecting Model as a dependency, like this:
Example #2
Built example using tutorial http://culttt.com/2013/07/08/creating-flexible-controllers-in-laravel-4-using-repositories/
<?php
use MyApp\Interfaces\UserInterface;
use MyApp\Models\User\User;
class UserRepository implements UserInterface
{
public function find($userId)
{
return User::with('profile')->find($userId);
}
}
Why in first example Model is injected, why not use directly Model like in example two?
Which way is correct and why ?
Also which way will be more testable with integrated to laravel UnitTest package ?
The example 2 is bad because it's coupling your repository to a particular implementation of the User model.
Every time you use your repository it'll need to instantiate Univemba\Models\User\User. The whole idea of Dependency Injection is to inject (send) in to your object whatever dependencies it has. If your object needs a Model to work with you can send it a Laravel Eloquent Model, but any of your co-workers could also need to send to it a Doctrine Model. But if you couple your class to Eloquent, this isn't possible.
So in the first example, there is no instantiation happening on your code and it's not using a concrete class directly as in the second:
return User::with('profile')->find($userId);
It is receiving an implementation in the process of its instantiation:
public function __construct(Model $user)
{
$this->user = $user;
}
There are better ways to do that, because it is still expecting a concrete class while it should be expecting an implementation of an interface
public function __construct(ModelInterface $user)
{
$this->user = $user;
}
In this case you just need to pass to your object something that implements ModelInterface, which could be
Univemba\Models\EloquentModel
Or
Univemba\Models\DoctrineModel
Because both would be implementing
Univemba\Models\ModelInterface
I think if a Repository is intended to be the Eloquent implementation of a RepositoryInterface it isn't a bad idea to use the EloquentModel directly.