I have just started an MVC project for the first time, and I have encountered a problem that I really wasn't expecting: Too many models.
Not too many different models, but too many which are subtly different, but almost exactly the same.
I have a fairly complex page, with several combo boxes, a list of detail objects and some other extraneous unrelated things that I need to keep track of. More information goes to the view than I get back (which is fine).
I have a domain model, which has remained constant throughout. (1 model)
I have a single use view model, which is the domain model and some extra information wrapped round it. (1 model)
I have a single use form model, which is a copy of the view model with an empty shell copy of the domain model inside it with validation stuff on it. (2 models)
I needed to change the type of a property this afternoon, and I had three separate places to change it in. It seems that I have too many models which are single use. It seems so much work for something that was supposed to be much easier than classic ASP.NET.
My question is: Am I doing it right? Are there supposed to be a multitude of models, or am I missing something obvious?
Not entirely sure what the form model in your example is for - are you using that to limit the post-able fields from the client or using that to map in to your domain?
In my own MVC websites I will have a domain model (which may just be a local domain or come from a remote WCF service) which I map one to one in to my individual ViewModels in order to render on screen and receive the post back. These also contain my validation (be it annotations or fluent etc).
Any property name changes would then be restricted to just the relevant page's ViewModel and the domain.
If you're in a situation where you're replicating properties between domain models then some sort of base view model structure would serve you well for common entities but on the whole it's not something I find to be a massive overhead personally.
Related
I've worked with OOP for a while now, and have gotten into the habit of creating classes for 'things', such as person, account, etc. I was coding in Java for this.
Recently I begun to work with MVC (in PHP), and I've come to realise that, contrary to what I originally thought, the model is not like an OOP class - feel free to correct me if I'm wrong, but my understanding is that the model is simply an interface between the controller and the database (maybe with data processing - more on this below). My reason for thinking this stems from my recent messing around with the CodeIgniter framework for PHP. CI doesn't allow for instances of models. Rather, it's a singleton pattern, and in most tutorials I've seen, it's used only for database queries and sometimes some static method.
Now, coming from OOP, I've gotten into the habit of having classes where the data is stored and able to be manipulated (an OOP class). My issue is that in MVC, I'm not sure where this takes place, if it does (I initially thought 'class' was synonymous with 'model'). So, I guess what I'm looking for is someone to explain to me:
Where does the manipulation of data (business logic) take place? I've read many articles and posts, and it seems some prefer to do it in the controller, and others in the model. Is one of these more correct than the other in terms of MVC?
Where/how do I store data for use within my application, such as that from a database or JSON/XML files returned from an API call? I'm talking things that would usually be attributes in an OOP class. Does this even still take place, or is it straight from the database to the view without being stored in variables in a class?
If you could link me to any guides/sites that may help me to understand this all better, I would appreciate it.
Representing MVC in terms of classes, you can say that:
Model - A class which stores/provides data.
View - A class which provides functionality to display the provided data in a particular fashion.
Controller - A class which controls that how the data is manipulated and passed around.
MVC design is adopted to provide facility to easily replace any of the module (Model/View/Controller) without affecting the other. In most cases, it is the view which gets changed (like same data represented in form of graphs, chats) but other modules must also be made independent of others.
This is an interesting question, and you often hear contrary things about how clean and stripped down models should be vs them actually doing everything you'd think they should do. If I were you I'd take a look at data mappers. Tons of people much smarter than I have written on the subject, so please do some research, but I'll see if I can summarize.
The idea is that you separate out your concept of a model into two things. The first is the standard model. This simple becomes a property bag. All it does is have a state that often reflects your database data, or what your database data will be once saved, or however you use it.
The second is a data mapper. This is where you put the heavy-lifting stuff. It's job becomes providing a layer between your pure model, and the database, that's it. It talks to the database, retrieves data specific to your model, and maps it to the model. Or it reads the model, and writes that data to the database. Or it compares the model's data with that of the DB.
What this does is it allows each layer to concern itself with only one thing. The model only has to keep a state, and be aware of its state. Or have some functionality related to the model that does not involve the database or storage. Now the model no longer cares about talking to the database which is hugely freeing! Now if you ever switch to a different database, or you switch to cookies or storing things in files or caching or any other form of persistence, you never need to change your models. All you have to change is the mapping layer that communicates between your models and the DB.
Wouldn't creating ViewModels lead to redundancy? In the sense I have my domain model and I need to display the data from it on a view. So we create ViewModels, add DataAnnotations to it and display it on the View. At this point I have 2 object with almost identical data.
As others have already said, only the most trivial application can get away with passing their domain models directly to the view. Even then, it's still not a good idea for a lot of reasons.
The requirements of your view are different from the requirements for your data model. For instance, you may have a field that is required in your view, but is nullable in your vie model. If you add a `[Required]' attribute, then your model will now consider this field non-nullable.
However, my single biggest reason for never using domain models in views is for security. MVC allows you to post any value to it, and the default model binder will happily plug values you post into the model, which means if you had an IsAdmin flag, and someone posted a true value for IsAdmin, then when you saved the changes to the model, someone is now an admin.
The first rule of web security is never trust input from the user, and passing your view models directly to the view basically gives up sanitizing your data.
Yes, it is a form of redundancy. But redundancy is often required to achieve other goals. In case of Models, having this separation of view models and domain models helps in achieving a decoupled setup between your view and data-store. And it is not often that ViewModels are exact copies of Domain.
Which means, either can change without having an impact on other. I can see cases where this would be valuable - data-type changes in table need not call for deployment of the web application.
So, in summary, while there is redundancy, it is a design choice on whether the system is complex enough to benefit from this redundancy.
In 99% cases ViewModels don't lead to redundency.
The only 1% which comes to my mind is simple application with anemic domain models and pages, which shows nothing but a single model on a page. This is peculiar to content management pages.
In any other case:
1) your ViewModels will join information from multiple domain models
2) you'll have a logic specific to your domain in domain models
3) it's not a good idea to mix view-specific metainformation like DataAnnotations into your domain
Nope, using ViewModel has its own purpose. Let's think about a situation when your view has two or more Entities from the Domain Model, without a ViewModel, how are you gonna organize data? The data needed for a view sometime is not exactly like domain model, it can has
less or more information and sometime we have to pre-process data from domain before rendering view(e.g. format date time depends on client's culture).
Furthermore, ViewModel help de-couple the Web UI from the domain layer. Entities in Domain Model is not just about data representation(which is the only purpose of view model), they also have operations that mimics the business rule, you don't want expose too much domain knowledge to the UI layer who doesn't need to know.
I'm building a multi-lingual web application using the MVC pattern as the starting position. The application has a number of forms that users will be interacting with and many of these forms will have fields that do a lookup from a database table, 'Province' for example.
If I need the options in these lists to be displayed in the users' language on the screen, I can see a couple of ways to do this:
In the model. When querying the model, I can provide the language that I desire the results to be returned in. This would allow translations to be used everywhere that data from the model is displayed without changes. However, this also means that the Province model in my example (plus all other application models) now need to know how to do language translations.
In the controller. I can query the model in the controller action as usual and then create a 'Translator' object that I can pass the results into before completing the action. This would imply that each controller action would potentially be duplicating the same translation code, violating the DRY principle.
In the view. Since the presentation of the application is generally expected to exist in the views, and the language of the user doesn't impact the business logic of the system, an argument could be made that language translations belong here. Especially considering that a page could also contain static content that will need to be translated. The downside to this is that it would complicate the views somewhat, especially for the front-end designers who will have to work around the new translation code.
Is there an accepted best-practice for where text translations belong in the MCV pattern for web applications? Does this change at all if I were to be loading the select list options via an AJAX call instead of at page load time?
Thanks for the help!
The best place to handle it is in the view. Your question only references dynamic data from the database, but you also have to handle the static information in your views as well. Best to handle both of those in the same place. A common practice in MVC for handling multiple language is resource strings, separate views for each language, or a combination of both. For information from the database resource strings would work well. You would store a token in the database for the options in the list (that token could be the English translation) and the view would get the appropriate translation from a resource for the specified country/locale. There is a good detailed explanation on this approach in this blog post.
If you need to translate part of your UI, then I would create a helper method that would read a resource file and output a translated string for that resources. E.g.
#Translate("NewUserHeading")
So regarding the UI, it makes sense to handle this in the UI.
If the data that you are going to translate at some point might be shown in a Flash client, or a mobile app, then it should be translated by a server and it should have nothing to do with your MVC application.
Honestly, any interaction with a database should be handled in the model, and that's the only thing handled in the model. Interpreting/organizing that data should be handled in the controller. I guess more information would be need as to where this translation is coming from and how that works to really give a solid answer.
The view will just display strings from a resource file. Including the correct resource file for the locale should do it. In Web applications, it's often a single language JS file defining the UI strings per locale, e.g, strings.en-us.js, strings.pt-br.js and so on.
Some strings do come from the server dynamically, the controller doesn't need to know about it, the model should just grab the localized strings and return is as part of the data request.
So it's either in the view (if it's static) or in the model, (if it's dynamic). Controllers should just translate data from the view to the model and from the model to the view
Another very important thing to consider is WHEN to Translate. Using a once-per-each-translation method like #Translate("NewUserHeading") is fine if you only have 100 translations on your page, but what if you have more?
Because think about it, translation is based upon language choice and that only happens once-per-page or even once-per-session. The way that the Kendo UI people do it in their demo is that way, when the User clicks a new language a new set of server files is loaded. That would mean having something like:
/src/html/en-us/
/src/js/en-us/
/src/html/es/
/src/js/es/
etc..
Check it out at https://demos.telerik.com/kendo-ui/grid/localization
It's a trade off of more work for better performance. The ultimate use-case of translation right now is https://www.jw.org/ with over 900 languages!
Before rendering into a view model should be formatted:
multilingual data localized;
date, time values formatted;
numbers formatted.
Who performs all this formatting - Controller or View?
Am I right that all the formatting is performed by the Controller which creates so called ViewModel containing only formatted values and sends this ViewModel to the View?
Thanks in advance!
Eric Petroelje is right, but I would build helper a class(es) to get localised content/dates etc, because localisation isn't always in the views, e.g. sending emails with localised content. I would have something like LocalisationHelper.GetString("MyKey"), or LocalisationHelper.GetDate(Date.Now), where the LocalisationHelper knows the users current locale (maybe from Session).
Then use this directly in the views where possible:
<%= Html.Encode(LocalisationHelper.GetDate(Date.Now)) %>
Design Patterns 101.
Model is for storing data (usually database-backed).
View is for presenting the data (not manipulating it).
Controller is for manipulating the model and passing that to the view (choosing the right locale for instance would go here).
MVC doesn't necessarily mean you have 3 distinct classes, but rather 3 components or layers. These are abstract and don't necessarily have to be tied to one physical class. Inside your Controller layer, that can consist of any number of helper classes or whatever.
I agree with part of what cartoonfox is saying, everything is intertwined. For instance, if you develop a view for a shopping cart, but the model is containing birthday information, then it isn't going to work. It is simply a design pattern to help remove duplication of effort. When you have fewer variables and less noise, it is much easier to focus on what needs to be done and understand it very well.
I had a discussion with our team about using annotations to render forms on a web page. These annotations were placed in the model or entity class. You will often work directly with entity classes, so it eliminates quite a bit of overhead and duplication of effort if you put your annotations here. Because your annotations are placed directly on the model class, you cannot end up with a view for birthday information, it just isn't possible. In addition, by following patterns, you strip out the junk that doesn't add value to the end result. You simply write the business logic and close to nothing else.
Although the annotations were in the same class as the model layer, the presentation or view layer consisted of the annotations and the helper classes. It doesn't need to necessarily be distinct classes or boundaries.
Another example would be. I've worked on some PHP web "applications" in the past. I use applications because they were a monolithic block of code, more or less a single main method with all the logic in there (which is hardly a functional application).
If you don't abstract code into functions and simply use a single method, you will end up with a lot of duplication of effort. If you tried to understand that monolithic code block, you would be in for a heap of trouble (as was I, it was hard to learn what was going on, then I would find a similar block of code somewhere else and be dumbfounded why a few things were tweaked the way they were).
It can be done any way you want, but design patterns such as MVC help you simplify what you write in order to get it to work as well as more easily wrap your head around the solution. Another popular design pattern or approach is divide and conquer.
To answer your original questions:
In Java, localized data would be done transparently by the application. For instance, if you wanted US English locale support, you would create a properties file:
messages_en-US.properties
and then place all your US English content in there. Depending on how much content this may be, you might want to use a different approach.
For my application I have whole pages in a different language so what I do is this. My controller determines the client's locale or a best match. Then it chooses which view to display and passes the model (data) to the view.
If you need dynamic formatting for your date/times, then your controller is responsible for determining which one will be used. Then, your view is responsible for using that converter to format the value.
I'm using JBoss Seam so the MVC design pattern is still used, but more abstractly. I don't have actual 'controllers', but an interceptor which is responsible for handling a specific piece of functionality (determining the client's locale, then another for their date/time preference). My controllers that would be the equivalent of a Spring Controller are action components responsible for handling page actions.
Walter
The view is responsible for this not the controller.
To explain I need to elaborate a bit about what the "view" is:
In both web-based and traditional GUI MVC patterns, the View is responsible for the external representation of whatever parts of the model that are shown. If that means "formatting" data from the model then so be it.
The controller shouldn't format the data - because the controller is responsible for doing things to the model based on events/commands coming in from the outside world. The controller doesn't have anything to do with rendering the output.
Example:
I want to display a shopping cart with several order lines:
Adding things to the cart causes the controller to change what's in the cart model.
Changing my locale causes the controller to change a setting in the user model to indicate preferred locale.
Whenever the cart is displayed, the view has to ask the model for the order lines and decide how to display that - maybe doing locale-sensitive work.
The same customer could ask to see the shopping cart in multiple different currencies. That just means changing what it looks like in the view. It's still the same shopping cart for the same customer with the same things in it.
If this is a web-app made from web template pages, you might embed code to pull in localized messages e.g. <%= message_renderer.text(:insufficient_funds) %>. In this case the "message_renderer" and the template file are both part of the view.
Views aren't necessarily just web-based templates. In some Java-based frameworks for example, we put "view objects" into a velocity template. Those view objects are linked back to objects in the model, but they also perform on-demand rendering and formatting behaviour. Those are part of the view although they're not just a template.
It's a bit confusing with frameworks like ruby on rails and groovy on grails where they call the template the "view" - when a view is really more than just a template. Traditionally (in Smalltalk MVC and in Java's Swing) views are just code that can perform formatting, rendering and any other display-related behaviour.
All that stuff seems like presentation layer logic, so personally I think it should go in the view.
ETA:
I wouldn't advocate for having the view call service layer objects in the general sense, but for localization I think it makes sense.
For example, lets say you have lots of static text on your pages that you have tokenized and stored in a database in various localized forms. I don't think it would make sense for the controller to have to lookup and put all those text tokens in the model. Even though a text token lookup is technically a "service layer" operation, I still think it makes sense for the view to call that service directly via some utility class rather than having the controller do it.
I've read some MVC advice in the past regarding models stating that you should not reuse the same model objects for the domain and the view; but I haven't been able to find anyone willing to discuss why this is bad.
It is my opinion that creating two separate models - one for the domain, one for the view - and then mapping between them creates a lot of duplication, plus tedious mapping code (some of which might be alleviated by things like AutoMapper) that will likely be error prone.
What makes having a separate model for the two concerns worth the trouble of duplication and mapping code?
At its heart, two models is about Separation of Concerns. I want my View to work off of a single Model. I want my Domain Model to represent the conceptual model I build with the domain experts. ViewModel often has technical constraints. Domain Model is about POCO, and not being bound by technical constraints of either data shown (View) or persisted (in a DB or otherwise).
Suppose I have three entities shown on a screen. Does that mean I need to force a relationship between the three? Or just create a ViewModel component object that contains all three items. With a separate ViewModel, View concerns are separated from my domain.
why? 'cause the view shouldn't have the ability to use the model object!
Imagine you pass the project to a web designer to do the view layer. Suddenly he/she has the ability to mess around with your application's data through the model layer. That's not good.
So always only pass the data the view needs, instead of the object with methods.
J.P. Boodhoo's article Screen Bound DTOs will help you understand the design benefit.
There is also a security benefit that I have written about.
Having a presentation model simplifies your views. This is especially important because views are typically very hard to test. By having a presentation model you move a lot of work out of the view and into the domain->presentation model. Things like formatting, handling null values and flattening object graphs.
I agree that the extra mapping is a pain but I think you probably need to try both approaches in your specific context to see what works best for you.
There are even plainer concerns including the view model's ability to be specially formatted and certainly null-safe.
I guess the idea is that your domain models might extend to other implementations, not just your MVC application and that would break the seperations of concerns principle. If your View Model IS your domain model then your domain model has two reasons to change: a domain change requirement AND a view requirement change.
Seems I have duplication of rules as well.
ie. client object validation on the UI, then mapping to domain object that has to be validated.
What I tend to do however is map my set of domain objects to create a model - ie. a webpage that shows customer info, stock info, etc... my model becomes a structure that holds a Customer object and a Stock object.
CompanyPageModel
public Customer Customer{get;}
public Stock Stock {get;}
then in my mvc project ViewData.Model.Customer.Name ViewData.Model.Stock.CurrentStocks
Separation 'seems' like more work, but later, it's good to have this division of UI/Domain model...sorta like writing tests :)
I have drunk the cool-aide finally, I do like being able to mark my viewmodel with display instructions and have that all auto wired up.
What I demand now is some kind of auto generator of viewmodels from poco entities. I always set some string as an int and it takes me forever to find it. Don't even think of doing this without automapper unless you like pain.