Symfony 3 - Should I create two different controllers for REST and normal HTML?
I already have a web application where I am using Symfony controller and twig html templates. So I already have the business logic written to fetch the content. Now I want to expose REST API to share this content with third party. Should I write a separate controller using fosrestbundle ? Or can I use the same controller for both json and HTML? If yes, how?
This might be a highly subjective question and each programmer will have another opinion to this.
I'll try to give you an answer ...
Is the REST Data requested by the page which gets already handled by the controller it might be a better choice to add just a simple json returning function for simplicity.
Are the requirements bigger and for example a third party application may have access to this, you should go with a dedicated REST API, since there may be complete different requirements for security, response times and so on.
At the one hand, a simple Controller can't give you that features and on the other hand it's much cleaner to follow the single responsibility principle with an extra API.
But consider also creating an extra API takes also more time.
So now you should decide yourself for now and for future whether you need an dedicated API or not. The need of third party access sounds to me to go for an extra API.
Related
As mentioned in another answer, our use case for Webflow is a bit unusual. Instead of a traditional MVC approach driving UI, with a little effort we exposed SWF as a RESTful HATEOAS API (with some caveats). Our Views are Collection+JSON format with links modelling SWF transitions.
The problem is that SWF tightly binds to the MVC pattern of returning ModelandView, with the View states resolving a view and calling render(). As such till now our choices have largely been to expose the API via a JSP (yuck). We're looking to use Spring Boot (no JSP) and moreover use RestController and Spring Hateoas where instead we want to return ResponseBody<> or ResponseEntity<T> from a controller method, with that being an object formed from the View(state) - i.e. get the Model but no view and generate C-J or HAL+Forms or similar from it with transitions as Links.
We use XML flow definitions (essential to the solution) so I figure either I'd have to
Define a new Flow xsd with custom attributes and monkey with the ModelBuilder code and View implementation. Seems a lot of hackery.
Somehow define a do-nothing MVC view and defer to the customer controller for the HttpResponse when entering View, but how?
Write a complete custom FlowHandlerAdapter (this might help alter the POST-redirect-GET behaviour but won't tackle the ModelandView bit)
"Insert some more elegant answer here"
What is the best route to pursue?
Before this whole thing is considered lunacy, think what SWF brings to proper REST API exposure of business logic - not just another tedious boilerplate example of a CRUD repository API, but loosely coupled declarative workflow. I looked at Spring State Machine also but frankly it seems miles behind SWF on key aspects and doesn't help with state exposure as REST.
My question might seem strange for pros but please take to account that I am coming from ruby on rails world =)
So, I am learning ASP.NET Core. And I like what I am seeing in it compared to rails. But there is always that but... Let me describe the theoretical problem.
Let's say I have a Product model. And there are over 9000 records in the database. It is obvious that I have to paginate them. I've read this article, but it seems to me that something is wrong here since the controller shouldn't use context directly. It has to use some repository (but that example might be provided in such a way only for simplicity).
So my question is: who should be responsible for pagination? Should it be the controller which will receive some queryable object from the repository and take only those records it needs? Or should it be my own business service which does the same? Or should the repository has a method like public IEnumerable<Product> ListProducts(int offset, int page)?
One Domain-Driven-Design solution to this problem is to use a Specification. The Specification design pattern describes a query in an object. So you might create a PagedProduct specification which would take in any necessary parameters (pageSize, pageNumber, filter). Then one of your repository methods (usually a List() overload) would accept an ISpecification and would be able to produce the expected result given the specification. There are several benefits to this approach. The specification has a name (as opposed to just a bunch of LINQ) that you can reason about and discuss. It can be unit tested in isolation to ensure correctness. And it can easily be reused if you need the same behavior (say on an MVC View action and a Web API action).
I cover the Specification pattern in the Pluralsight Design Patterns Library.
For first, I would like to remind you that all such examples you linked are overly simplified, so it shouldn't drive you to believe that that is the correct way. Simple things, with fewer abstraction layers are easier to oversee and understand (at least in the case of simple examples for beginners when the reader may not know where to look for what) and that's why they are presented like that.
Regarding the question: I would say none of the above. If I had to decide between them then I would say the service and/or the repository, but that depends on how you define your storage layer, etc.
"None of the above", then what? My preference is to implement an intermediary layer between the service layer and the Web UI layer. The service layer exposes manipulation functionality but for read operations, exposes the whole collection as an IQueryable, and not as an IEnumerable, so that you can utilize LINQ-to-whatever-storage.
Why am I doing this, many may ask. Because almost all the time you will use specialized viewmodels. To display the list of products on an admin page, for example, you would need to display values of columns in the products table, but you are very likely to need to display its category as well. Very rarely is it the case that you need data only from one table and by exposing the items as an IQueryable<T> you get the benefit of being able to do Selects like this:
public IEnumerable<ProductAdminTableViewModel> GetProducts(int page, int pageSize)
{
backingQueryable.Select(prod => new ProductAdminTableViewModel
{
Id = prod.Id,
Category = prod.Category.Name, // your provider will likely resolve this to a Join
Name = prod.Name
}).Skip((page - 1) * pageSize).Take(pageSize).ToList();
}
As commented, by using the backing store as an IQueryable you will be able to do projections before your query hits the DB and thus you can avoid any nasty Select N+1s.
The reason that this sits in an intermediary layer is simply you do not want to add references to your web project neither in your repo nor in your service layer (project) but because of this you cannot implement the viewmodel-specific queries in your service layer simply because the viewmodels cannot be resolved there. This implies that the viewmodels reside in this same project as well, and to this end, the MVC project only contains views, controllers and the ASP.NET MVC-related guttings of your app. I usually call this intermediate layer as 'SolutionName.Web.Core' and it references the service layer to be able to access the IQueryable<T>-returning method.
I am designing a system that uses asp.net webapi to serve data that is used by a number of jquery grid controls. The grids call back for the data after the page has loaded. I have a User table and a Project table. In between these is a Membership table that stores the many to many relationships.
User
userID
Username
Email
Project
projectID
name
code
Membership
membershipID
projectID
userID
My question is what is best way to describe this data and relationships as a webapi?
I have the following routes
GET: user // gets all users
GET: user/{id} // gets a single user
GET: project
GET: project/{id}
I think one way to do it would be to have:
GET: user/{id}/projects // gets all the projects for a given user
GET: project/{id}/users // gets all the users for a given project
I'm not sure what the configuration of the routes and the controllers should look like for this, or even if this is the correct way to do it.
Modern standard for that is a very simple approach called REST Just read carefully and implement it.
Like Ph0en1x said, REST is the new trend for web services. It looks like you're on the right track already with some of your proposed routes. I've been doing some REST design at my job and here are some things to think about:
Be consistent with your routes. You're already doing that, but watch out for when/if another developer starts writing routes. A user wants consistent routes for using your API.
Keep it simple. A major goal should be discoverability. What I mean is that if I'm a regular user of your system, and I know there are users and projects and maybe another entity called "goal" ... I want to guess at /goal and get a list of goals. That makes a user very happy. The less they have to reference the documentation, the better.
Avoid appending a ton of junk to the query string. We suffer from this currently at my job. Once the API gets some traction, users might want more fine grained control. Be careful not to turn the URL into something messy. Something like /user?sort=asc&limit=5&filter=...&projectid=...
Keep the URL nice and simple. Again I love this in a well design API. I can easily remember something like http://api.twitter.com. Something like http://www.mylongdomainnamethatishardtospell.com/api/v1/api/user_entity/user ... is much harder to remember and is frustrating.
Just because a REST API is on the web doesn't mean it's all that different than a normal method in client side only code. I've read arguments that any method should have no more than 3 parameters. This idea is similar to (3). If you find yourself wanting to expand, consider adding more methods/routes, not more parameters.
I know what I want in a REST API these days and that is intuition, discoverability, simplicity and to avoid having to constantly dig through complex documentation.
I am trying to determine the best way to implement the retrieval of detail records based upon the master's ID.
Obviously you would set up API controllers for both master and detail.
Solutions I've considered:
Have API consumers use OData to get all details filtered by a master ID. While I don't have an issue with this solution, I kinda feel bad putting that onto the API consumer and feel it is something that should be handled internally by the API
Go against the convention of just having the Get/Put/Post/Delete methods and create an action of "GetMastersDetails" on the detail controller and make it accessible via routing. While this would certainly work, I feel this gets away from the whole point of Web API (to an extent).
Create a 3rd controller named "MastersDetailsController" which would have a Get based upon a master ID with different possible return types:
Which would return a list of detail IDs which would then be used to call a Get on the details controller to get those actual details
Which would return a list of actual detail objects. What I don't like about that is having a controller returning a different type than what it is based upon.
Option 2 will be fine. Option 1 opens up a lot more risk depending on your scenario, and what you want to allow the user to get at.
It's not really "against convention" to add custom methods to an ApiController. You can do that however you like. It would only be "against convention" if you did so and used the wrong HTTP methods (i.e. a GET when you're deleting something in your custom method).
I'd go with either #1 or #2.
For #1, OData support enables not only the scenario you describe but offers a lot of additional functionality which might be desired in the future.
For #2, I don't think it gets away from the point of Web API's. Maybe a bit from a true RESTful service, but it's easy to implement and easy to understand.
I've spent a lot of time working in Django, and have grokked the framework well enough that I have started replacing the original components (view engine, etc.) with my own custom components and the sky hasn't fallen down.
I've been looking at ASP.NET MVC, and been quite interested (I really like C#/F#) but so far have learned... just about nothing. I've been digging through http://www.asp.net/mvc/mvc4 without much success. I suppose my main questions would be:
What are the main moving parts in a typical workflow? Let's say a request comes in. Who takes it, does stuff, and passes it on to who else? In Django, for example, a request goes through the URL Mapper, Middleware, goes to a controller, which may dig through some models (via explicit function calls) and get some data, pass it into a template (also via an explicit function call) to be rendered and pass it back.
What kind of client-server coupling is there? For example, in many frameworks there is a explicit coupling of each HTML-form with a serverside-validator, with a model, with a database table, such that client side validation code is automatically generated. Another example is Quora's Livenode, which explicitly links client-side HTML components with their dependencies in the model, allowing changes in the database to propagate and automagically update client-side code.
I think there is no better answer to your first question than ASP.NET MVC Pipeline :
http://www.simple-talk.com/content/file.ashx?file=6068
explained in more detail here :
http://www.simple-talk.com/dotnet/.net-framework/an-introduction-to-asp.net-mvc-extensibility/
To your second question : answer is none. ASP.NET application dont even have to render HTML output, you can write your own viewengine to give any representation of the data, not consumed by browser, but any http (REST) capable device. The only things you can consider as coupling "conventions" (for model binding for example), but they can be replaced and extended in any way you like.
What kind of client-server coupling is there?
As rouen said, none.
I am not familiar with Django, but unlike other MVC frameworks (including Rails) ASP.NET MVC is very skinny in that it only implements Views and Controllers of the traditional MVC pattern. You are pretty much on your own for the model part. That means there is no built-in support for database creation, ORM, et cetera.
ASP.NET MVC does implement a bunch of plumbing to route requests to the appropriate controllers and even some binding of parameters (e.g. query string parameters, form values) when instantiating controllers but this binding is not to a full blown model. The binding in this context is usually either single values or "viewModels"
ASP.NET MVC also implements the plumbing to select the right view to render.