I have a set of functions I want to be available to my web pages and user controls in my c# .net 3.5 web project. My standard approach for pages is to create a "base page" that extends the System.Web.UI.Page class and then have my pages inherit from this, rather than directly from the Page class.
I now want to expose some of these functions to my web controls (ascx) and web services. I can think of a number of ways to do this, but they seem a little clumsy and I think I'm missing a trick.
Is there an easy way to provide some common functions to both my pages, web services and controls using inheritance, or do I need to wrap these functions in a class that all of them can access?
An example to clarify:
I have a singleton that handles most functionality for my web application.
At the start of each request I want to check that the class exists in the web cache and initialise it if not.
Initially this was handled in a page base that the pages all used. Now I need to be able to access my singleton safely from services and controls, with the same checks. I have therefore extracted the checking and initialisation logic into another class, that then each of my base page, control and web service, all instantiate. Even with this model I have the same code repeated in 3 places (each of my base classes for controls, ws and pages), albeit not much code, this seems wrong too!
It works, but it seems clumsy...I look forward to you guys humbling me with your wisdom!
Sounds to mee like a case of aspect-oriented programming. .NET is ill equipped for this. I'm afraid that your solution is one of the best.
Alternatively perhaps you can move all or some of those functions to a static class/singleton and then use that class from your aspx/ascx/asmx? Not much in the way of inheritance, but at least less code duplication.
My solution to this is to put all the methods and functions I want to share in my base master page class. I then put an equivalent for each method and function in the user control base class as follows:
//Property in masterpage base
public string QsSearchTerm
{
get
{
if (!String.IsNullOrEmpty(Request.QueryString["q"]))
{
return Helpers.SanitiseString(Server.UrlDecode(Request.QueryString["q"]));
}
return String.Empty;
}
}
//Property in usercontrol base
public string QsSearchTerm
{
get
{
if (Page.Master is BaseMasterPage)
{
return ((BaseMasterPage)Page.Master).QsSearchTerm;
}
return string.Empty;
}
}
What this doesn't help with, is your code repetition with web service base classes. I would think that refactoring the above into a class with a constructor that accepts an HttpContext instance would be the way forward. You can then expose a singleton instance of this class in your base web service, master page, user control, page etc.
Hope this helps, but I too would be interested in hearing if there's a better way.
In your Singleton you could provide a Strategy interface to allow variations of the code depending on the configured environment. This would allow you to switch between web/windows/wcf...and so on.
I think using a BasePage is the right approach.
I have multiple base pages and custom user controls that load differently depending on which basepage is used by the current page.
In your custom user control you can use something like:
if (this.Page is BasePageName)
{
BasePageName bp = (BasePageName)this.Page;
bp.BasePageFunction();
}
No you can get ride of the repetitive code in the custom user control and just call it from the base page.
You can also have a hierarchy of inherited base pages depending on page functionality and needs. ie.) BasePageName2 : BasePageName
Related
There is an abstract base page
public abstract class BasePage : PageData
There are some pages that are inherited from this page
public class ChildPage : BasePage
public class ChildPage2 : BasePage
I need to retrieve all the pages that are inherited from the BasePage.
In PropertyCriteria I could use only ChildPage and ChildPage2 for the search by PageTypeID. So this approach would not solve the issue in my case
Your most viable options are to 1) retrieve all pages (with GetChildren or GetDescendants, depending on content structure) of the base type, or 2) use Find.
Episerver Find makes this kind of thing trivial, provided your site actually uses Find. :/
A third (not recommended) option would be to identify all of the sub-types and use multiple FindPagesWithCriteria calls, but that would be slow. FPWC is entirely uncached, and results in DB calls.
Personally, I never use FindPagesWithCriteria for anything other than certain administrative features not dealing with rendering the actual site.
Let's suppose I have a layer of abstract controllers, which delegates the request to its child controller class, until it reaches the implementation.
Think of it like a pipeline of controllers, that the request must go through, and includes caching the responses, authorizing and authenticating the user, validating the input and output data, handling repository access, etc.
My leaf class (the last child of the hierarchy), may have the following signature:
public class SeasonsController : DefaultPersistenceRestController
<int, Season, SeasonPutDTO, SeasonPostDTO, SeasonQueryData> {
/** Controller implementation here **/
}
The base classes have a lot of reusable code located in one module, this is good and has helped me a lot when changing the logic of my controllers at a global level.
Now, suppose SeasonsController need to call EpisodesController, for irrelevant reasons.
The call would be like this:
EpisodesController episodeController = new EpisodesController();
//Do something with EpisodesController
The problem is that I don't want EpisodesController to be accessed from the outside, such as client's request. ASP.NET automatically identifies controllers and creates a public endpoint for them, such as http://localhost:80/episodes.
I created EpisodesController because it uses a lot of logic from the controller's base classes, but I intend to use it internally.
I can desactivate authentication, authorization, cache and all other stuff that will be useless if a controller is used in this way, so that's not a problem.
However, I cannot manage to prevent ASP.NET to ignore my EpisodesController class, and to not consider it like a controller.
Is there an attribute or annotation maybe that will tell the compiler to do this? Maybe some modification in Web.config?.
Also note that I don't want to change EpisodesController's class name to another name, as it is really a controller, but an internal one.
You could try to use the IgnoreRoute extension method. Or you could try the internal as suggested by beautifulcoder. If it's in another assembly (and you can modify it) you could also make it visible to other assemblies with InternalsVisibleToAttribute.
Although to be honest, using one controller within another controller doesn't seem right to me. I would try and refactor you common functionality to services/helpers, then you could probably also make your EpisodesController into a simple service. Composition over inheritance and all that :)
If you make a controller public it will be accessible. From what I understand, you can change it to protected or internal.
I've read the questions/answers explaining that there is no multiple inheritance in C#, that we don't even need it, and that it causes too much problems.
Now, I'm working on a project where I don't really understand how can I do things without multiple inheritance, without duplicating code.
Here's the situation. There is a website with a home page and other pages inheriting from a masterpage (the home page does not inherit from). Both the page and the masterpage are performing some stuff: custom login, statistics, loading of users settings for customization, etc. For the moment, the solution is crappy, since the source code for those tasks is just copied twice.
The home page class inherits from Page. The masterpage, on the other hand, inherits from Masterpage. Logically, it would be great to inherit from a common class too, but it's multiple inheritance, so it's impossible.
So what to do instead?
I thought about several ways, but dislike them:
Create a standalone class which will be called from the page/masterpage class. So for example instead of writing bool isDisplayingTips = this.CurrentUser.IsDisplayingTips, I would write bool isDisplayingTips = this.SharedObjects.CurrentUser.IsDisplayingTips. I don't like it, since it's longer to write.
Create a "real", empty, common masterpage, and inherit both the home page and the masterpage from it. Not only it will require to write more code to access masterpage parameters, but it will also slow the things down, requiring an additional masterpage on each request.
Any idea?
MasterPage is a just control (that get embedded into the actual page) so you can not have the later approach. However, first approach of creating another helper class is quite feasible.
Yet another approach that we typically use is to have
Common base page class - all pages will inherit from the common base page.
Put common functionality in base page class
From master page, the base page can be referred by casting - for example, myBasePage = (BasePage)this.Page;. This way master page may access common functionality from base page class.
I don't find your 2nd option that dislikable.
I presume you mean creating a base class, e.g. MasterPageBase, derived from System.Web.UI.MasterPage, and creating an empty MasterPage for your homepage, that will inherit from this MasterPageBase.
If done right, it shouldn't slow things down...
I suggest you to use the first of your option. If you (understandably)
don't feel comfortable with increased level of indirection, you could just create new methods on your standalone classe, e.g:
public bool IsDisplayingTips(){
return CurrentUser.IsDisplayingTips;
}
and the from your pages just call
bool isDisplayingTips = this.SharedObjects.IsDisplayingTips()
Use:
standalone class which will be called
from the page/masterpage class
but instead of stopping there, add a base page and a base master page. Both use the shared class, and keep the specific pages/master pages code from the indirection.
I strongly believe that, reading code and reading good code is key to great programming. If not one of the many.
I had been facing some problems in visualizing and having a "feel" of using inheritance to better my code architecture.
Can somebody give me some link to good code to emulate, where folks have used inheritance in an absolute "kung-fooey ruthless" manner [in a good way]
I strongly believe that, reading code and reading good code is key to great programming
Hard to disagree.
Actually the qestion is pretty hard - becouse there is some alternatives to inheritance, such as composite reuse principle, so sometimes it's very hard to diside if inheritance is used in "kung-fooey ruthless" manner or there ware some better way to implement the same wich will make code esier to understand/test/make it lossely coupled and so on.
In my humble opinion Enterprise Library Application validation block whith it's Microsoft.Practices.EnterpriseLibrary.Validation.Validator class with all it's descendants
is a very good example of inheritance, becouse
concept of validation is easy to understand
there is good example how to find common in objects of with pretty different nature (i.e. OrCompositeValidator/DateTimeRangeValidator/ObjectCollectionValidator)
many of us tried to implement something more or less like this, so this background will give more quality for understanding
this is clear(for me, but I can be wrong:) that inheritance has no alternatives there
You can download source code from codeplex.
An example of good usage of inheritance would be the .NET framework classes. You can download the MONO project to gain access to some source code. If you want to better your code architecture, invest some time in studying architectural design patterns.
Here is a personal example where I've made use of inheritance to greatly benefit a development situation:
I needed to develop a asp.net (c#) form control set, which would allow both standard web forms (bunch of form fields with submit button), as well as a secure version which talks to a web service to submit the information over SSL.
Because of all of the similarities between the controls and concepts, I developed a number of classes:
BaseWebControl
custom base control class that inherits from System.Web.UI.WebControl class (part of .NET framework
has custom properties and methods that are used in our application by all custom controls (control state info, etc.)
BaseFormControl
inherits from BaseWebControl, gaining all of its underlying functionality
handles base form functionality, such as dynamically adding fieldsets, marking required fields, adding submit button, etc. etc.
contains a label and associated control index for easy lookups
marked as an abstract class, with abstract method called SubmitForm. This method is not defined on this class, however it is called by the submit button click event. This means that any specific form control class that inherits from this base class can implement the abstract SubmitForm functionality as needed.
EmailFormControl
Inherits from BaseFormControl, so it gains all underlying functionality above without any duplication
contains very little, except overrides the abstract method SubmitForm, and generates an email based on the form fields.
all other control functionality and event handling is dealt with by the base class. In the base class when the submit button is clicked and handled, it calls this specific implementation of SubmitForm
SecureFormControl
Again inherits from BaseFormControl, so it gains all underlying functionality above without any duplication
In its implementation of SubmitForm, it connects to a WCF web service and passes the information in over SSL.
no other functionality is required because base class handles the rest.
In stripped down code form, the general outline is as such:
public class BaseWebControl : System.Web.UI.WebControl
{
//base web control with application wide functionality built in
}
public abstract class BaseFormControl : BaseWebControl
{
//handles all 'common' form functionality
//...
//...
//event handler for submit button calls abstract method submit form,
//which must be implemented by each inheriting class
protected void btnSubmit_Click(object sender, EventArgs e)
{
SubmitForm();
}
protected abstract SubmitForm();
}
public class EmailFormControl : BaseFormControl
{
protected override SubmitForm()
{
//implement specific functionality to email form contents
}
}
public class SecureFormControl : BaseFormControl
{
protected override SubmitForm()
{
//connect to WCF web service and submit contents
}
}
As a result of the above, BaseFormControl has about 1000 lines of code in a whole bunch of methods, properties, etc. SecureFormControl and EmailFormControl each have about 40 lines. All other functionality is shared and controlled by the base class. This promotes:
maintainability
efficiency
flexibility
consistency
Now I can create any type of web form, such as DataBaseFormControl, etc. etc. very easily. I can add great new functionality to all forms by adding methods and properties to the base classes, etc.
And the list goes on.
Phew that was a lot of typing. Hope this helps give you a good example. This was one instance where I found inheritance to be a key success point in a project.
I agree with the recommendation to look at the .NET base class library, as it has excellent examples of abstraction via both inheritance and interfaces. The goal is to insulate consumer code from having to care about the details of a particular implementation. The WinForms designer works with Controls, but it has no idea what specific kinds of Controls will be implemented. It doesn't care, because inheritance abstracts away the unnecessary details. LINQ works similarly with IEnumerable; it doesn't really matter what's being enumerated, as there are algorithms you can write that work with anything enumerable. Both are excellent examples of abstraction used well.
is there a way to reference the page control in a webservice? something like this:
[WebService(Namespace = "http://test.org/")]
public class Search : System.Web.Services.WebService
{
public Search()
{
Page.Controls.Add(new Control()); // can I get a reference to Page?
}
}
This seems like a very odd design approach. In general a called method should have no knowledge about, or dependencies on the caller. In this case the web method would need knowledge about the page calling it. I don't think this is possible, and even if it was, consider the possibility that it may not even be a page calling the web service. It could be any kind of application.
What you are trying to do (at least the way you are trying to do it) is impossible. The web service cannot, at server side, modify the page by modifying the control tree, server side. The page object that was rendered to the user does not exists any more.
I think that what you should be doing instead is using an update panel. That will allow you to do exactly what you want to.