Almost Sorta Real Dynamic in .NET

This post discusses the basics of writing your own dynamic type – If you’re interested in the finished code make sure to download the source.

The new dynamic keyword is a great way to create and modify a dynamic object in your .NET code – For example, check out this bit of code.

dynamic counter = new {
    sent = 0,
    received = 0
};
counter.sent++;

… Oh… wait… That doesn’t actually work…

That’s because even though you have an object declared as dynamic, it is still just the same old object it always was under the hood. Maybe not the dynamic utopia we were expecting. Seems like plain ol’ Reflection with a slightly cleaner syntax.

Bummer…

However, Microsoft was awesome enough to actually allow us to create our own dynamic types by simply inheriting from the DynamicObject class. This post looks at how we can create a semi-dynamic object by handling the binding methods ourselves.

[Source Code: Dynamic.cs]

Our goal is to create an object that we can add, remove and change values on the fly. It ends up being a little easier than we expect. Let’s look at the minimum code we are going to need to get started.

public class SimpleDynamic : DynamicObject {

    //contains the values to use
    private Dictionary<string, object> _Values = new Dictionary<string, object>();

    //handles getting values from our object
    public override void TryGetMember(GetMemberBinder binder, out object result) {
        //...
    }

    //handles assigning values to our object
    public override bool TrySetMember(SetMemberBinder binder, object value) {
        //...
    }

}

Note: If you downloaded the source code above you might be wondering why these examples don’t look the same. The source code is the complete Dynamic class while this post only goes over the basics of using the DynamicObject class.

In order to create our dynamic object we must handle getting and setting values. We can use a simple Dictionary<string,object> to hold our values until we need them.

//handles getting values from our object
public override void TryGetMember(GetMemberBinder binder, out object result) {
    return this._Values.TryGetValue(binder.name, out result);
}

//handles assigning values to our object
public override bool TrySetMember(SetMemberBinder binder, object value) {
    this._Values.Remove(binder.name);
    if (value is object) { this._Values.Add(binder.name, value); }
    return true;
}

Not much code, but now our dynamic object is a little more dynamic than it was before. We can now work with properties a lot easier.

//create the new object
dynamic simple = new SimpleDynamic();

//add and update some values
simple.red = "#f00";
simple.green = "#0f0";
simple.count = 10;
simple.add = new Action<int>(value => simple.count += value);

//and check the values
Console.WriteLine(simple.red); // #f00
Console.WriteLine(simple.green); // #0f0
Console.WriteLine(simple.count); // 10

//invoke methods
simple.add(5);
Console.WriteLine(simple.count); // 15

Of course this class isn’t that helpful since it’s only regular Dictionary without string indexes but it is interesting to see how you can make a ‘dynamic’ object with only a few lines of code.

As for the included source code, it has a few extra features that this class doesn’t have.

• Use Anonymous Types directly to build out entire objects.
• Support for the += operator.
• Enumeration through member names.
• String indexes (for both single and nested properties).

Here are a few examples the Dynamic.cs class can do.

//create using anonymous types (and nested anonymous types)
dynamic info = new Dynamic(new {
    name = "Hugo",
    age = 30,
    settings = new {
        color = "orange",
        homepage = "website.com"
    }});

//add values using +=
info.create = new { value = false; }
info.nested.value.allowed = 55;

//enumeration of properties
foreach(string prop in info) {
    Console.WriteLine("{0}: {1}", prop, info[prop]);
}

I’d caution against using this a lot in your projects, but it is really interesting to see the kind of control Microsoft has built into the new DynamicObject class.

[Source Code: Dynamic.cs]

Using MongoDB With Visual Studio

My new project CSMongo, a Mongo driver for .NET, is now online! Check it out!

I recently found out about MongoDB and decided to check it out and so far I’ve been really impressed. The most interesting thing about MongoDB that it doesn’t work like a normal database — there aren’t really any schemas for any of the “tables”. In fact, you can make changes to the structure of any record at any time without affecting the rest of the “table”.

Getting Started

Here are a few steps I used to get MongoDB running and testable from my Visual Studio. To keep my main computer clean I used a virtual instance of Ubuntu to host the “database server”.

This part might take a little bit of time but start by downloading VirtualBox and Ubuntu 9.10 (You don’t need to download Mongo just yet). Once everything is downloaded install and configure Ubuntu but don’t start it up right away (you need to configure some stuff).

You’re going to want to make sure that your PC can connect to the virtual instance on the standard MongoDB port (unless you change it of course). If not, open a command prompt (on the host system) and then run the following commands (from the VirtualBox directory).

VBoxManage setextradata UbuntuDev "VBoxInternal/Devices/pcnet/0/LUN#0/Config/MongoDB/HostPort" 27017
VBoxManage setextradata UbuntuDev "VBoxInternal/Devices/pcnet/0/LUN#0/Config/MongoDB/GuestPort" 27017
VBoxManage setextradata UbuntuDev "VBoxInternal/Devices/pcnet/0/LUN#0/Config/MongoDB/Protocol" TCP

This example uses the same name (UbuntuDev) that I used for the screenshot example above. Make sure you use the correct name when setting yours up.

It is also worth mentioning that I had to use a Bridged Connection for my Network connection so I’d get an IP address from my wireless. But, if you aren’t on a wireless network… say, like your in-laws house for several hours… you can use Host Only Adapter so you can keep working… just sayin’…

Inside Ubuntu

Once you’ve successfully setup your Ubuntu system go on and download the MongoDB binaries. I don’t know the correct location to install these files so I placed them inside /etc/mongodb.

If you’ve never used Linux before then warm up your typing fingers and open a terminal window (Applications > Accessories > Terminal). Start typing in the following commands…

sudo bash
mkdir /data
mkdir /data/db
mkdir /etc/mongodb
chmod a=rwx /etc/mongodb

Note: This is certainly not the recommended security setup for this folder but for our testing purposes it is sufficient.

At this point we can revert back to our lazy Windows ways and drag the contents of the .tgz file into the /etc/mongodb directory. Once we have the contents copied over switch back to the terminal window and then type…

sudo /etc/mongodb/bin/mongod

And you should see something like the screenshot below…

Once you see this message you should be ready to test from Visual Studio but you can always test it from Ubuntu by opening a new Terminal window and typing…

sudo /etc/mongodb/bin/mongo

Which allows you to enter commands and make changes to the database similar to the online demo on their website.

Connecting Via Visual Studio

I haven’t found much for C# code to connect to Mongo but there is currently a project hosted on GitHub that allows you to perform queries and edit documents. It is a mess of code – but it *does* at least work… mostly… (nothing personal guys) 😉 Here is a simple example of how to use the code…

//connect to the source (IP address of virtual)
Mongo mongo = new Mongo("100.0.0.1");
mongo.Connect();

//get the database and 'table'
Database website = mongo.getDB("website");
IMongoCollection users = website.GetCollection("users");

//make a new document of information
Document user = new Document();
user["name"] = "Hugoware";
user["age"] = 29;
user["isAdmin"] = false;

//then save the document
users.Insert(user);

If you still have your Terminal window up then you might have noticed messages listed in response to your update. If you still have the MongoDB command line up you can view your changes by entering in a few commands. For example, to see the database I just created in the sample above I would enter…

use website
db.users.find()

And I would get a response similar to this…

What Is Next?

Personally, I think MongoDB is going to end up being huge. The main problem I see for C# developers is that MongoDB really favors Dynamic Languages which isn’t really a strong suit of the language.

Right now I’m working on my own driver to talk to Mongo that heavily relies on my AnonymousType code (anybody remember that old stuff?). It is still early on in the project so if you’re interested in helping feel free to contact me.

Simple External Templates With jQuery

New code available: I’ve written a follow up post about this topic with new code and new features.

If you’ve ever tried building a jQuery object from scratch then you’ve probably found that it is a little time consuming and probably a little ugly by the time you’re finished. You could put additional HTML templates as hidden elements on the page but maybe you’d rather keep that extra mark up out of your page.

One option would be load a template from a website address and then make your changes once it arrives. The code would probably look something like this.

//create the container
var element = null;

//make the request for the HTML
$.ajax({
    url:"template.html",
    success:function(html) {
        element = $(html);
        //make changes here
    }
});

There isn’t anything wrong with this code but it does put a bit of separation of your declaration of an element and the actual usage of the element. The problem is that you can’t do the whole process at once since the AJAX call takes a little time and you have to rely on a call back before you have access to the object.

It would be nice if we could have immediate access to our jQuery object on an AJAX call, but not block anything else on the page.

Time For Some Javascript Magic ™

I’m not really sure what you would call the example below, but the general idea is to capture (and queue up) any actions that would be invoked on our jQuery object and then release them once the AJAX call has been finished.

Let’s look at some code and see what it would look like (extra comments to try and explain the process)

/*
 * Arguments
 * $.template(url) - Pass in a string to a url to load 
 * $.template(params) - An object with parameters found below
 *   url: Path to the template resource (required)
 *   data: Same as the 'data' argument in the $.ajax call
 *   error: Same as the 'error' argument in the $.ajax call
 *   complete: Same as the 'complete' argument in the $.ajax call
 *   beforeUpdate: delegate(html) called just before the actual object is created
 *   afterUpdate: delegate(html, actual) called just after the actual object is created
 */

//creates a function to access a web template
jQuery.template = function(params) {
	
	//format the passed in parameters
	//check if this was only the resource url
	if (typeof(params) === "string") {
		params = { url:params };
	}
	
	//prepare the arguments passed into the class
	if (!$.isFunction(params.beforeUpdate)) { params.beforeUpdate = function() {}; }
	if (!$.isFunction(params.afterUpdate)) { params.afterUpdate = function() {}; }
	

	//create the object that handles the work
	var self = {
	
	//Properties
	//---------------------------------------------------------
	
		//handles forwarding methods onto the actual object
		container:null,
		
		//the actual jQuery object being created
		actual:null,
		
		//method calls that are waiting to be called
		queue:[],
		
	//Methods
	//---------------------------------------------------------
		
		//prepares the container for use
		setup:function() {
			
			//apply each of the methods
			self.container = $("<div/>");
			for(var item in self.container) {
				if (!$.isFunction(self.container[item])) { continue; }
				self.register(item);
			}
			
		},
		
		//handles creating method forwarding on the returned object
		register:function(method) {
		
			//create a method that handles routing the original method calls
			self.container[method] = 
				function(p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13, p14) {
				
				//if the actual object has been called, just invoke
				if (self.container.actual) {
					return self.container.actual[method](
						p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13, p14
						);
				}
				//otherwise, queue the request
				else {
					self.queue.push({
						action:method,
						params:[p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13, p14]
					});
					
					//then return the temporary object
					return self.container;
				}
			
			};
		
		},
		
		//executes any queued commands and updates the jQuery object
		update:function(html) {
			
			//create the jQuery object
			self.container.actual = $(html);
			
			//then execute all of the waiting commands
			$.each(self.queue, function(i, arg) {
				self.container.actual[arg.action](
					arg.params[0], arg.params[1], arg.params[2], arg.params[3], arg.params[4],
					arg.params[5], arg.params[6], arg.params[7], arg.params[8], arg.params[9],
					arg.params[10], arg.params[11], arg.params[12], arg.params[13], arg.params[14]
					);
			});
		
		},
		
		//starts the ajax request to download the template HTML
		download:function() {
			
			//starts downloading content
			$.ajax({
			
				//parameters for the template request
				url:params.url,
				data:params.data,
				dataType:"html",
				
				//performs the catch up work for the ajax
				success:function(html) {
					//** Optional: Uncomment 'setTimeout' to simulate a delay 
					//setTimeout(function() {
					params.beforeUpdate(html);
					self.update(html);
					params.afterUpdate(html, self.container.actual);
					//}, 2000);
				},
				
				//additional handling of the request
				error:params.error,
				complete:params.complete
			});
		
		},
		
	//Initalization
	//---------------------------------------------------------
		
		//setup the object to work
		init:function() {
		
			//prepare the temporary container
			self.setup();
		
			//start downloading the content
			self.download();
		}
	
	};
	
	//prepare the the template code
	self.init();
	
	//return the custom container to forward method calls
	return self.container;
	
};

This code allows you to have direct access to a jQuery object even before it has finished loading the content from the server. So, instead of using a callback we can write our jQuery like we normally would.

//note: the first command is all one long chain with a comments
//between functions to explain a bit more

//create the object and immediately apply changes
var template = $.template("template.txt")
	.css({"color":"#f00", "width":"200", "background":"#333"})
	.animate({width:800, height:900}, 3000)

	//you can even append and appendTo the object in advance
	.appendTo($("#container"))

	//or search for and update child elements
	.find(".title").text("howdy")

	//and parent elements
	.parent().css({"background":"#00f"});
	
//you can also still access the template from the 
//assigned variable
template.find("div").click(function() {
	alert('clicked');
});

//even if if you call it much later, it still works
setTimeout(function() {
	template.css({"color":"#0f0"});
}, 10000);

What Is Going On Here?

As I mentioned before the real problem is that the jQuery object we create isn’t ready as soon as we want to assign to it. Because of that we have to use a callback to resume the work. However, the cool thing about dynamic languages is that we can override anything we want.

In this example we start by creating a container that has all the same functions as a typical jQuery object but has one slight modification — all the methods are overridden and placed into a queue (man, I love enclosures). Once our AJAX call has completed we run an update command that executes everything we have saved against our new jQuery object instead of the one that received the calls to begin with — madness!

After we’ve caught up and our actual object is created we can stop saving actions to the queue and instead just invoke them immediately. I nicknamed this method forwarding for the sake of having a cool, buzzword sounding sort of name but if anyone knows what this is really called, please tell me 🙂

It is worth noting that this is only going to work with functions that return the jQuery object (at least until the ‘real’ object is created). The reason is that since we’re just capturing methods and saving them for later then we don’t know what the actual return type is. Needless to say, using a property probably won’t be accurate either (since we can’t intercept the request for the property like we would with a method)

In any case, this might simplify the next time you need to download content in jQuery but you don’t want to break up your functionality.

Using Delegates As Parameters

I’m not sure the community stance on passing around delegates as arguments but personally I love it. I think there is something magical about the way you can provide instructions on how how to do something and then pass it into another method and let it do its job without needing to think about it. It is almost like writing a code template and then filling in the blanks with the specifics of what you need.

Using delegates as arguments can simplify code and in some cases reduce it. This post goes over a sample scenario where you might use delegates as arguments instead of writing multiple functions to perform different tasks.

Simple Resource Loader

Let’s pretend we have a class that is responsible for loading and saving files from an external resource. We could write a class like the one below.

//handles loading resources for the application
public static class ResourceLoader {

    //returns the bytes for a resource
    public static byte[] LoadBytes(string resource) {
        string path = ResourceLoader._GetResourcePath(resource);
        return File.ReadAllBytes(path);
    }

    //returns a xml document resource
    public static XDocument LoadXmlDocument(string resource) {
        string path = ResourceLoader._GetResourcePath(resource);
        return XDocument.Load(path);
    }

    //returns a bitmap resource
    public static Bitmap LoadBitmap(string resource) {
        string path = ResourceLoader._GetResourcePath(resource);
        return Bitmap.FromFile(path) as Bitmap;
    }

    //returns the string text for a resource
    public static string LoadText(string resource) {
        string path = ResourceLoader._GetResourcePath(resource);
        return File.ReadAllText(path);
    }

    //generates a path to a resource
    private static string _GetResourcePath(string file) {
        return Path.Combine(@"c:\path\to\files\", file);
    }

}

Nothing is really wrong with this code. If we need to add a new type to this class then we simply create a new method and plug it in. Additionally, each of the methods could use some sort of exception handling in case the conversion doesn’t go so well. As you can imagine, the more try catch blocks we add, the larger this class starts to get.

However, if you think about it, all of these resources could be handled roughly the same way. They all need a way to convert bytes to whatever type you’re wanting.

Using A Delegate To Fill In The Blank

So instead, lets address this problem using a delegate to handle the conversion of bytes to the type that we need.

//handles loading resources for the application
public static class ResourceLoader {

    //returns the bytes for a resource
    public static T Load<T>(string resource, Func<byte[], T> convert) {

        //find the correct path
        string path = Path.Combine(@"c:\path\to\files\", resource);
        byte[] bytes = File.ReadAllBytes(path);

        //attempt convert the file
        try {
            return convert(bytes);
        }
        //if it fails, forward the error to the caller
        catch (Exception ex) {
            throw new FormatException(
                string.Format(
                    "Could not load resource '{0}' as a type {1}", 
                    resource, typeof(T).Name
                    ), ex);
        }

    }

}

Great — Now we can provide any method we want to format the bytes and then return the type we’re looking for. So for example, instead of calling ResourceHandler.LoadBitmap we can use our new method as shown below.

Bitmap bitmap = ResourceLoader.Load("test.jpg", (bytes) => {
    using (MemoryStream stream = new MemoryStream(bytes)) {
        return Bitmap.FromStream(stream) as Bitmap;
    }
});

Pretty slick, right? …oh, wait… this code is longer and more difficult to use… This can’t be right!

Bringing The Concept Together

Clearly, the example above isn’t improving anything for us. The code is longer and requires that we write the same functionality for reading a resource in multiple places. Even though this code is more flexible, it also requires more work. So instead, lets write some definitions of common conversions as part of our class.

//handles loading resources for the application
public static class ResourceLoader {

    //Load<T>(resource, convert)
    //snip...

    //converts bytes to a bitmap
    public static readonly Func<byte[], Bitmap> AsBitmap = (bytes) => {
        using (MemoryStream stream = new MemoryStream(bytes)) {
            return Bitmap.FromStream(stream) as Bitmap;
        }
    };

    //converts bytes to a xml document
    public static readonly Func<byte[], XDocument> AsXml = (bytes) => {
        using (MemoryStream stream = new MemoryStream(bytes)) {
            using (StreamReader reader = new StreamReader(stream)) {
                return XDocument.Parse(reader.ReadToEnd());
            }
        }
    };

    //converts bytes to a string
    public static readonly Func<byte[], string> AsText = (bytes) => {
        return Encoding.UTF8.GetString(bytes);
    };

    //simply returns the byte array
    public static readonly Func<byte[], byte[]> AsBytes = (bytes) => bytes;

}

Now, instead of needing to manually create a delegate to handle the conversion process we can write code like the example below.

 
//Loads a resource by passing the static delegate that is part of the class we created
Bitmap bitmap = ResourceLoader.Load("test.jpg", ResourceLoader.AsBitmap);
 

By using delegates as an argument we allow our loading function to be flexible enough to accept other methods for performing the conversion (that we can create as we need them) or the common conversion methods that we added as part of the class.

You’ll also notice that we don’t need to declare the generic type we’re wanting to have returned since it can be inferred from the return type of the delegate we pass in! (so cool… to me at least)

Of course, there is about a hundred different ways to can already load resources in .NET but you could apply this same concept many other areas of applications you develop.

How could you use delegates to create “templates” for code?

Undo, Redo and Whatever… From A Web Application

Long ago a brilliant developer invented the Undo command — probably one of the most important features to be added in computing history. It seems funny to me that today, in 2009, it is still a very rare feature to find on a website. This isn’t to say no one is doing it but more that it isn’t a common feature to find.

Using Lambdas For Undo Actions

In .NET you can use Lambdas to create a function and pass it around like an argument and then that argument can then be invoked at a later time in a different place. That said, with a little careful planning and a few correctly stored Lambdas, you could have a working ‘Undo’ model in your ASP.NET web sites For this example I’m going to use MVC but this would really work anywhere.

Warning: This is very crude code, more of a proof of concept, so you probably don’t want to use it in your projects as is.

Let’s start with an abstract controller that contains our Undo handling code.

[UndoController.cs]

using System;
using System.Web;
using System.Web.Mvc;

namespace UndoAction {

	//class that supports an undo action
	public class UndoController : Controller {

		#region Constants

		private const string SESSION_UNDO_ACTION = "Session:UndoAction";
		private const string SESSION_UNDO_MESSAGE = "Session:UndoMessage";
		private const string TEMP_DATA_UNDO_MESSAGE = "UndoResult";
		private const string DEFAULT_UNDO_MESSAGE = "Previous action was undone!";
		private const string DEFAULT_MISSING_UNDO_MESSAGE = "No actions to undo!";

		#endregion

		#region Undo Action Container (Session)

		//contains the current undo action
		private static Action _UndoAction {
			get {
				return System.Web.HttpContext.Current
					.Session[SESSION_UNDO_ACTION] as Action ;
			}
			set {
				System.Web.HttpContext.Current
					.Session[SESSION_UNDO_ACTION] = value;
			}
		}

		//the message to return in the view data for this action
		private static string _UndoMessage {
			get {
				return System.Web.HttpContext.Current
					.Session[SESSION_UNDO_MESSAGE] as string ;
			}
			set {
				System.Web.HttpContext.Current
					.Session[SESSION_UNDO_MESSAGE] = value;
			}
		}

		#endregion

		#region Setting Undo Actions

		/// <summary>
		/// Applies an undo action
		/// </summary>
		protected void SetUndo(Action action) {
			this.SetUndo(DEFAULT_UNDO_MESSAGE, action);
		}

		/// <summary>
		/// Applies an undo action with a return message for the user
		/// </summary>
		protected void SetUndo(string message, Action action) {
			UndoController._UndoAction = action;
			UndoController._UndoMessage = message;
		}

		/// <summary>
		/// Performs the undo action (if any) and saves the message
		/// </summary>
		protected void PerformUndo() {

			//check if there is an action
			if (UndoController._UndoAction is Action) {

				//perform the action and save the message
				Action action = UndoController._UndoAction;
				action();
				this.TempData[TEMP_DATA_UNDO_MESSAGE] =
					UndoController._UndoMessage;

				//and clear out the previous information
				UndoController._UndoAction = null;
				UndoController._UndoMessage = null;

			}
			//just save a generic message
			else {
				this.TempData[TEMP_DATA_UNDO_MESSAGE] =
					DEFAULT_MISSING_UNDO_MESSAGE;
			}

		}

		#endregion

	}

}

Basically, we create an abstract Controller that allows us to set an Undo action using a Lambda. You can also set a message for to return to the user that sumarizes what the action had done. This example uses a couple Session variables to hold the parts of the undo action, except I recommend that you create an actual class to house all of the information.

Next, let’s look at how we would actually use this controller.

[HomeController.cs]

using System;
using System.Collections.Generic;
using System.Web.Mvc;
using System.Linq;

namespace UndoAction {

	//example of using undo actions
	public class HomeController : UndoController {

		#region Constants

		private const string SESSION_LIST_CONTAINER = "Session:ListContainer";

		#endregion

		#region Properties

		//a simple list of items for the list
		public static List<string> ListContainer {
			get {
				List<string> container = System.Web.HttpContext.Current.Session[SESSION_LIST_CONTAINER] as List<string>;
				if (container == null) {
					container = new List<string>();
					System.Web.HttpContext.Current.Session[SESSION_LIST_CONTAINER] = container;
				}
				return container;
			}
		}

		#endregion

		#region Actions

		//shows the list of items
		public ActionResult Index() {
			return this.View(HomeController.ListContainer.OrderBy(item =&gt; item.ToLower()));
		}

		//adds an item to the list
		public ActionResult Add(string phrase) {

			//format the value
			phrase = (phrase ?? string.Empty).Trim();

			//add the item if it isn't there yet
			if (!HomeController.ListContainer.Any(item => item.Equals(phrase, StringComparison.OrdinalIgnoreCase)) &amp;&amp;
			    !string.IsNullOrEmpty(phrase)) {
				HomeController.ListContainer.Add(phrase);
			}

			//return to the list view
			return this.RedirectToAction("Index");

		}

		//removes an item from the list
		public ActionResult Delete(string phrase) {

			//make sure the item even exists first
			if (HomeController.ListContainer.Any(item => item.Equals(phrase, StringComparison.OrdinalIgnoreCase))) {

				//since it exists, save the logging message
				this.SetUndo(
					string.Format("Restored '{0}' to the list!", phrase),
					() => {
						HomeController.ListContainer.Add(phrase);
					});

				//and then actually remove it
				HomeController.ListContainer.Remove(phrase);

			}

			//return to the main page
			return this.RedirectToAction("Index");

		}

		//tries to undo the previous action
		public ActionResult Undo() {

			//attempts to undo the previous action (if any)
			this.PerformUndo();

			//return to the main page
			return this.RedirectToAction("Index");
		}

		#endregion

	}

}

This controller allows the user to manage a simple shopping list with an Undo action to allow them to restore a deleted item. It is some ugly code, but you get the idea on how it works.

The important thing you should note is that the Lambda doesn’t refer to the instance of the class (this) but instead that the list is a static property. This is important to keep in mind as you develop something like this. When you create the Lambda action, you can’t refer to the instance that it was created in. Instead, you need to work with static properties.

Finally, let’s look at the ViewPage that is used in this example.

[Index.aspx]

<%@ Page Language="C#" 
	Inherits="System.Web.Mvc.ViewPage" %>
<%@ Import Namespace="System.Collections.Generic" %>

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
	<head>
		<title>Shopping List</title>
	</head>
	<body>
		<% IEnumerable<string> list = this.Model as IEnumerable<string>; %>
		
		<h2>Shopping List</h2>
		
		<p><% =this.Html.ActionLink("Undo Previous Action", "Undo") %></p>
		
		<% if (this.TempData["UndoResult"] is string) { %>
		<p><em><% =this.TempData["UndoResult"] as string %></em></p>
		<% } %>
		
		<hr />
		
		<table>
		<% foreach(string item in list) { %>
			<tr>
				<td><% =this.Html.Encode(item) %></td>
				<td><% =this.Html.ActionLink("Remove", "Delete", new { phrase = item }) %></td>
			</tr>
		<% } %>
		</table>
		
		<hr />
		
		<form action="<% =this.Url.Action("Add") %>" method="post" >
			<strong>Add an item:</strong>
			<input type="text" name="phrase" />
			<input type="submit" value="Add" />
		</form>
		
	</body>
</html>

The ViewPage itself is nothing fancy — If you’re working in a project you should create a strongly-typed view but in this example we just read the information we need and go on.

Using The Page

Below are a few screen shots of what happens as the user makes changes to the page. This example starts with a list of a few different items.

After removing an item from the list (the ‘Orange’) pressing the ‘Undo Previous Action’ button will restore the item. The custom message that was added is displayed (from the TempData field).

If you press the ‘Undo’ button again there aren’t any actions waiting anymore and the default message is displayed instead (no actions waiting).

Things To Consider

This code is just a sample of how you could implement and Undo function into your web applications. Here are a few things that you might want to keep in mind before starting something like this.

• Make sure that your Undo command doesn’t cause you to duplicate code. In this example, the undo command uses a different ‘Add’ approach than the actual ‘Add’ action on the controller. This could result in duplicated code or incorrect results.
• Determine if you can queue up ‘Undo’ commands. This example can only do one action at a time, but you could potentially create a Stack of actions that are executed in the reverse order that they were added.
• Could you use the same concept to ‘Redo’ a command after it has been undone?
• Be careful with your context – Don’t use ‘this’ inside of your Undo Lambdas.

Anyways, just an interesting concept to share. This approach requires a lot more thought but could pay off in the end.

Do you have a web application that could benefit from an ‘Undo’ command?

Combining WebServices with MVC

I always liked the concept behind WebServices. Having a single place to store a bunch of complex but commonly used functions is a great way to decrease complexity of other programs that all sit on the same network. If you’re like me and tend to do a lot of intranet applications, a web service can prevent a lot of duplicate code.

My only real problem with WebServices was having to use SOAP. Adding a Web Reference to a project wasn’t that big of a deal but if you ever wanted to just call a function real quick, say from a script file (yes, I do VBScript occasionally… ick) then it isn’t quite as I’d prefer. You end up spending more time making sure your XML is well formed and less time on the logic inside your quick script.

MVC helps get around that problem by allowing you to perform normal HTTP calls and plug all your arguments into the query string or the body of the request – something much easier to do. The problem, however, is that you end up losing the convenience of using a WebService with other applications.

A Simple Solution

The idea here is to create a controller that acts as a wrapper to a WebService. By doing this we can override a few methods on our Controller that use reflection to invoke the matching method on the WebService. Also, because the Controller is still a unique class, we can still attach any number of Actions to it as we normally would. Below is some code that I wrote the other day. It isn’t battle tested so if you use it be sure to verify it does everything that you need it to do.

using System;
using System.Collections.Generic;
using System.Linq;
using System.Web;
using System.Web.Services;
using System.Web.Mvc;
using System.Reflection;
using System.Xml.Serialization;
using System.Xml.Linq;
using System.IO;

namespace Interface.Controllers {

    /// <summary>
    /// Abstract wrapper that handles calling WebService methods on behalf of a controller
    /// </summary
    public abstract class WebServiceControllerWrapper<T> : Controller where T : WebService {

        #region Constructors

        /// <summary>
        /// Creates a new Controller Wrapper for a WebService
        /// </summary>
        public WebServiceControllerWrapper() {
            this.Service = Activator.CreateInstance<T>();
        }

        #endregion

        #region Properties

        //The service that is being used for this call
        private T Service { get; set; }

        #endregion

        #region Overriding Methods

        //finds the correct method for the WebService method
        protected override void HandleUnknownAction(string actionName) {

            //find if the method exists
            MethodInfo method = this.Service
                .GetType()
                .GetMethods()
                .Where(found =>
                    found.IsPublic &&
                    found.Name.Equals(actionName, StringComparison.OrdinalIgnoreCase) &&
                    found.GetCustomAttributes(typeof(WebMethodAttribute), true).Count() > 0
                    )
                    .FirstOrDefault();

            //if no method was found, just give up
            if (method == null) { return; }

            //check if all the arguments were found
            List<object> arguments = new List<object>();
            ParameterInfo[] parameters = method.GetParameters();
            foreach (ParameterInfo param in parameters) {

                //check if this argument was found
                object arg = this.ValueProvider[param.Name].ConvertTo(param.ParameterType);
                arguments.Add(arg);
            }

            //with the arguments try and call the result
            object result = method.Invoke(this.Service, arguments.ToArray());

            //if there is a return value, serialize it and write it
            this.Response.ContentType = "text/xml";
            if (method.ReturnType != null) {

                //if this is an XElement of some kind, just use it as is
                if (result is XObject) {

                    //write the string
                    using (StreamWriter writer = new StreamWriter(this.Response.OutputStream)) {
                        writer.Write(result.ToString());
                    }

                }
                else {

                    //try and serialize it
                    XmlSerializer serialize = new XmlSerializer(result.GetType());
                    serialize.Serialize(this.Response.OutputStream, result);
                }

            }

        }

        #endregion

    }

}

The idea here is to inherit this class instead of the standard Controller class and provide the name of the WebService we want to wrap around as our Generic argument. For example…

//that's about it - actions are mapped automatically to the correct method on the webservice
public class AccountController : WebServiceControllerWrapper<AccountWebService> { }

By doing this, when our controller receives an Action, is checks the WebService instance for the same method and then tries to call the method with the arguments it finds as part of the request!

And that’s it! A quick and simple way to map incoming actions to a matching method on a Web Service!

The Return Type

I’m not sure the best way to handle the return type at this time. It seems to me that the XmlSerializer should be sufficient for object with the exception of Xml which should probably just be written out as a string. If you have a suggestion on a better way to respond to incoming requests, please let me know. 🙂

Remember: This is just some quick and dirty code – This needs some more polish and exception handling before I’d use it in a real project, but at least it might help you get going in the right direction.

Build A Smarter Loop With C#

One of the first things every programmer learns in how to loop through an array of information. We learn how to do for, foreach, while loops but then we never really improve on them. For the most part the standard loop is as far as a programmer goes.

It seems to me that loops should be more intelligent than that. In fact we often write code that uses the loop for additional information. It isn’t uncommon to see a loop check if it is an even or odd index or the first or last value.

Modestly Smarter Loops

I really like lambdas a lot – in fact probably, too much. Being able to pass an delegate around like an argument definitely can be abused but in some cases it can make for some really elegant code. Let’s look at some code that can improve working with arrays.

namespace LoopExtensions {
//Some comments removed for brevity...

    #region Extension Methods

    /// <summary>
    /// Handles looping through collections with additional details
    /// </summary>
    public static class LoopExtensionMethods {

        public static IEnumerable<T> Each<T>(this IEnumerable<T> collection, 
            Action<ElementDetail<T>> each) {
            return LoopExtensionMethods.Each<T>(collection, 0, collection.Count(), each);            
        }

        public static IEnumerable<T> Each<T>(this IEnumerable<T> collection, 
            int start, 
            Action<ElementDetail<T>> each) {
            return LoopExtensionMethods.Each<T>(collection, start, collection.Count(), each);
        }

        public static IEnumerable<T> Each<T>(this IEnumerable<T> collection,
            int start,
            int end,
            Action<ElementDetail<T>> each) {
            Action<ElementDetail<T>, T> handle = (detail, item) => { each(detail); };
            return LoopExtensionMethods.Each<T>(collection, start, end, handle);
        }

        public static IEnumerable<T> Each<T>(this IEnumerable<T> collection, 
            Action<ElementDetail<T>, T> each) {
            return LoopExtensionMethods.Each<T>(collection, 0, collection.Count(), each);            
        }

        public static IEnumerable<T> Each<T>(this IEnumerable<T> collection, 
            int start, 
            Action<ElementDetail<T>, T> each) {
            return LoopExtensionMethods.Each<T>(collection, start, collection.Count(), each);
        }

        public static IEnumerable<T> Each<T>(this IEnumerable<T> collection, 
            int start, 
            int end, 
            Action<ElementDetail<T>, T> each) {

            //verify the ranges
            if (start < 0 || end > collection.Count()) {
                throw new ArgumentOutOfRangeException();
            }

            //perform the work
            foreach (T value in collection) {
                each(new ElementDetail<T>(value, start++, end, collection), value);
                if (start == end) { break; }
            }
            return collection;
        }

    }

    #endregion

    #region Detail Information Class

    /// <summary>  
    /// Contains a summary of information about the item current in the loop  
    /// </summary>  
    public class ElementDetail<T> {

        #region Constructors

        internal ElementDetail(T value, int index, int total, IEnumerable<T> collection) {
            this.Value = value;
            this.Index = index;
            this.Total = total;
            this.Collection = collection;
        }

        #endregion

        #region Loop Information

        public int Index { get; private set; }

        public int Total { get; private set; }

        public T Value { get; private set; }

        public IEnumerable<T> Collection { get; private set; }

        #endregion

        #region Value Properties

        public T Previous {
            get {
                return !this.First
                    ? this.Collection.ElementAt(this.Index - 1)
                    : default(T);
            }
        }

        public T Next {
            get {
                return !this.Last
                    ? this.Collection.ElementAt(this.Index + 1)
                    : default(T);
            }
        }

        public bool Last {
            get { return this.Index == (this.Total - 1); }
        }

        public bool First {
            get { return this.Index == 0; }
        }

        public bool Outer {
            get { return this.First || this.Last; }
        }

        public bool Inner {
            get { return !this.Outer; }
        }

        public bool Even {
            get { return this.Index % 2 == 0; }
        }

        public bool Odd {
            get { return !this.Even; }
        }

        #endregion

        #region Collection Properties

        public int StepNumber {
            get { return this.Index + 1; }
        }

        public float PercentCompleted {
            get { return (((float)this.Index / (float)this.Total) * 100); }
        }

        public float PercentRemaining {
            get { return 100 - this.PercentCompleted; }
        }

        public int StepsCompleted {
            get { return this.Index; }
        }

        public int StepsRemaining {
            get { return this.Total - this.Index; }
        }

        #endregion

    } 

    #endregion

}

Now that is a lot of code but if you review over it you’ll find that it is a set of extension methods that can be used with IEnumerable. The basic idea is that we can perform our loop within a delegate that accepts additional information about the elements in our loop.

So here is a quick example…

string[] items = {
    "Apple",
    "Orange",
    "Grape",
    "Watermellon",
    "Kiwi"
};

items.Each((item) => {
    Console.Write("{0} > ", item.Inner ? "Inner" : "Outer");
    if (!item.First) { Console.Write("Previous: {0}, ", item.Previous); }
    Console.Write("Current: {0} ({1})", item.Value, item.StepNumber);
    if (!item.Last) { Console.Write(", Next: {0}", item.Next); }
    Console.WriteLine(" -- {0}% remaining", item.PercentRemaining);
});

// Output
// Outer > Current: Apple (1), Next: Orange -- 100% remaining
// Inner > Previous: Apple, Current: Orange (2), Next: Grape -- 80% remaining
// Inner > Previous: Orange, Current: Grape (3), Next: Watermellon -- 60% remaining
// Inner > Previous: Grape, Current: Watermellon (4), Next: Kiwi -- 40% remaining
// Outer > Previous: Watermellon, Current: Kiwi (5) -- 20% remaining

Normally you would simply write all the comparisons within your loop and then use them as needed. Instead, in this code, we pass in an additional parameter that contains shortcuts to many common comparisons you might find inside of loops. By doing this we improve readability and focus on what the loop is doing.

Whats Wrong With The Old Way

Nothing! Writing your comparisons inline can be advantageous since you aren’t doing any more work than you need to do. However, in some cases, the improved readability would make a big difference in the quality of the code. Consider the two snippets of MVC code and decide with one is easier to read.

<ul>
<% foreach(SiteMapNode node in breadcrumb) { %>
    <li class="item <% =(breadcrumb.IndexOf(node) % 2 == 0 ? "item-even" : "item-odd") %>" >
    <% if (breadcrumb.First().Equals(node) || breadcrumb.Last().Equals(node)) { %>
        <strong>
    <% } %>
        <a href="<% =node.Url %>" >
            (<% =(breadcrumb.IndexOf(node) + 1) %>) : <% =node.Text %>
        </a>
    <% if (breadcrumb.First().Equals(node) || breadcrumb.Last().Equals(node)) { %>
        </strong>
    <% } %>
    </li>
<% } %>
</ul>

Or the same code using a loop helper…

<ul>
<% breadcrumb.Each((node) => { %>
    <li class="item <% =(node.Even ? "item-even" : "item-odd") %>" >
    <% if (node.Outer) { %><strong><% } %>
        <a href="<% =node.Value.Url %>" >
            (<% = node.StepNumber %>) : <% =node.Value.Text %>
        </a>
    <% if (node.Outer) { %></strong><% } %>
    </li>
<% }); %>
</ul>

This example shows using only the ElementDetail class but of course you could use the other extension method that also supplies the value by itself.

Write Code For Humans Not For Computers

Everyone has heard the advice before – more or less, readability is probably the most important part of your code. The computer can read both of those examples easily whereas a human might need to think for a moment before they are positive what the code does.

In any case, this simple but effective approach can transform previously unreadable code into works of programming art… or something like that…

Keep Your Website Alive (Don’t Let IIS Recycle Your Website)!

Have you ever opened a page for one of your websites and it lags for awhile before it finally shows a page but then all of your following requests are quick? If you were to look up the problem you’d find that often it ends up having to do with IIS meeting an idle time limit and shuts down your site. There is even some software you can purchase to fix the problem for you.

But who wants to spend money on something like that? Especially when we can solve this ourselves — even for Hosted Environments!

Stayin’ Alive — (ack! bad pun again!)

If you happened to check out that software above then you can probably glean what it does just from the title. I’d rather not devote my personal machine to something like that so lets see if we can’t approach this from an alternative route.

private static void _SetupRefreshJob() {

    //remove a previous job
    Action remove = HttpContext.Current.Cache["Refresh"] as Action;
    if (remove is Action) {
        HttpContext.Current.Cache.Remove("Refresh");
        remove.EndInvoke(null);
    }

    //get the worker
    Action work = () => {
        while (true) {
            Thread.Sleep(60000);
            //TODO: Refresh Code (Explained in a moment)
        }
    };
    work.BeginInvoke(null, null);

    //add this job to the cache
    HttpContext.Current.Cache.Add(
        "Refresh",
        work,
        null,
        Cache.NoAbsoluteExpiration,
        Cache.NoSlidingExpiration,
        CacheItemPriority.Normal,
        (s, o, r) => { _SetupRefreshJob(); }
        );
}

If we place this bit of code in the Global.asax and call it when Application_Start() is raised, we can basically start a job that keeps our website alive. You could just as easily use a Thread to host the refresh method but for this example we simply used an Action delegate (but if you are using an earlier version of .NET then you might HAVE to use a Thread to do this).

Once our application starts the refresh job is also started and is saved to the cache. In this example we’re using 60 seconds, but you can change this to be as often as you like.

So How Can We Keep It Fresh?

So how about an example of some code we can use? Here is a simple example that could keep our website alive. Replace the //TODO: in the example above with something like the following.

WebClient refresh = new WebClient();
try {
    refresh.UploadString("http://www.website.com/", string.Empty);
}
catch (Exception ex) {
    //snip...
}
finally {
    refresh.Dispose();
}

This snippet uses a WebClient to actually make an HTTP call to our website, thus keeping the site alive! We could do any number of things from this code like updating local data or get information from external resource. This can be used to keep our site alive and our content refreshed, even if we’re using a Hosted Environment!

It is worth nothing that might not actually need to do an HTTP call back to your website. It is possible that using any method will keep your website from being killed off (but I haven’t tested it yet so let me know what happens if you try it). This example, however, has been tested and works quite well with my provider.

Using WebControls In ASP.NET MVC Views – Part 3

Check out my newest blog post about using WebControls inside of MVC (source code included)🙂

WebControls In MVC Series

• Part 1
• Part 2
• Part 3 (Working source code)
• Part 4

I mean’t to have this post out this morning, but I was busy working on getting my personal website online — Please check it out and tell me what you think!

Up until now all of the post in this series have been tests to see if we could make the process work. Now, finally, now we get into some real code that we can actually use. I’ll take just a little bit of space to explain how it works and how to set it up, but then the rest of the post to explain how to use it.

Getting The Code Setup

Before we discuss how to use the code lets go over briefly how to setup the code.

To start, the code you need to download is found at the end of the post. When you download it, put both the .cs and the .aspx file into your project. By default the code expects the .aspx to be in your Views directory, but you can move it — but if you do then you need to update MVCWebFormExtensionMethods._WebFormRenderControl.TEMPLATE_PAGE with the new location (yeah, it’s a long name :)).

Optionally, you can make a couple changes to your Web.config to make sure that the extension method and MvcControlGroup are available to the rest of your application without needing to add anything special onto each of your pages.

<configuration>
    <!-- snip... -->
    <system.web>
        <pages>
            <controls>
                <!-- snip... -->
                <add tagPrefix="mvc" 
					namespace="MvcWebControls" 
					assembly="YourProjectAssemblyName" />
            </controls>
            <namespaces>
                <!-- snip... -->
                <add namespace="MvcWebControls"/>
            </namespaces>
        </pages>
    </system.web>
    <!-- snip... -->
</configuration>

I’ve hidden most of the Web.config from this example, so make sure to add the areas you see above, not simply replace your entire config.

The rest of this post goes over some demos found out on my website, you may want to follow along to better understand what is going on.

Example 1 — Simple Postbacks

<html>
    <head runat="server">
        <title>Simple Postback Example</title>
    </head>
    <body>    
        <h2>Just A Simple Form</h2>
        <% this.Html.WebForm((form) => { %>        
            <% form.RenderControl(new TextBox()); %>        
            <% form.RenderControl(new Button() { Text = "Save" }); %>
            <hr />
            <% form.RenderControl(new Calendar()); %>
        <% }); %>    
    </body>
</html>

For the first example, we want to see if we can use a few simple WebControls and see if our values are posted back the way we would expect. In this example we add a TextBox, Button and Calendar. Pay attention to how this HtmlHelper method works.

If you notice, the method excepts a delegate to render the control onto the page. We do it like this because it allows us to provide markup to our page along side with the WebControls. The method accepts a single parameter the MvcWebForm. Think of this as a very, very simple version of a WebForm.

The MvcWebForm gives you access to the page that is being used to render the controls. This is important to remember and I’ll go over it in the next section.

Example 2 — Using Events

This is handy so far but unless we are using some events for our controls, nothing much has changed. Let’s look at another example.

<html>
    <head runat="server">
        <title>Using Events</title>
    </head>
    <body>    
        <h2>Simple Button Click Event</h2>
        <% this.Html.WebForm((form) => { %>        
            <strong>Clicked : </strong>
            <% Label text = form.RenderControl(new Label() { Text = "0" }); %>
            
            <br /><br />            
            <% Button submit = form.RenderControl(new Button() { Text = "Add" });
               submit.Click += (s, e) => {
                   int value = 0;
                   int.TryParse(text.Text, out value);
                   text.Text = (++value).ToString();
                   
                   //post back this information
                   form.Page.ClientScript.RegisterStartupScript(
                       typeof(Page),
                       "confirm",
                       string.Format("alert('Updated to {0}');", text.Text),
                       true
                       );
               }; %>
        <% }); %>    
    </body>
</html>

In this example we are assigning a Click event to the button we added to the page. You’ll notice when we run this example the value is incremented up by one on each press.

Notice that we use form.Page instead of this.Page. When you use this code you must remember that you are working with two separate Page instances. One for the View and the other for rendering the control. In this example we use RegisterStartupScript to display an alert box. If you were to have used this.Page.ClientScript, nothing would happen.

It’s a subtle difference that you are going to want to keep in mind while you use this class.

Example 3 — MvcControlGroup

So far all of our examples have only used a couple controls that were all rendered inline with the rest of the page. While this is going to work in most situations, some places won’t work as well.

The example below shows how you can use the MvcControlGroup to group controls together, for example, the Wizard control.

<html>
    <head runat="server">
        <title>Using Events</title>
    </head>
    <body>    
        <h2>Using Wizard Control with MvcControlGroup</h2>
        <div>
        <% this.Html.WebForm((form) => { %>
        <mvc:MvcControlGroup runat="server" >
        <asp:Wizard runat="server" >
            <WizardSteps>
                <asp:WizardStep runat="server" Title="Personal Information" >
                    <p>Leaving fields blank will catch the validators.</p>
                    <strong>Name</strong>
                    <asp:TextBox ID="name" runat="server" />
                    <asp:RequiredFieldValidator runat="server" 
                        ControlToValidate="name" ErrorMessage="Must provide a name" />
                    <br /><br />
                
                    <strong>City</strong>
                    <asp:TextBox ID="city" runat="server" />
                    <asp:RequiredFieldValidator runat="server" 
                        ControlToValidate="city" ErrorMessage="Must provide a city" />                    
                </asp:WizardStep>
                
                <asp:WizardStep runat="server" Title="Reservation Date" >
                    <strong>Date Requested</strong>
                    <asp:Calendar ID="date" runat="server" />                 
                
                </asp:WizardStep>
                
                <asp:WizardStep runat="server" Title="Confirm" >
                    
                    <h2>Confirm This Order?</h2>
                    <p>This step maps to a Controller Action before submitting</p>
                    
                    <% MvcWebForm.Current.Action = "/Projects/MvcWebForms/Submit"; %>
                    <% MvcWebForm.Map(() => new {
                        name = name.Text,
                        city = city.Text,
                        date = date.SelectedDate.ToShortDateString()
                       }); %>
                       
                </asp:WizardStep>
            </WizardSteps>
        </asp:Wizard>
        </mvc:MvcControlGroup>
        <% }); %>
        </div>
        
    </body>
</html>

Just like with other controls, the MvcControlGroup must be rendered in a separate page instance. This control moves it to the correct context before it is processed.

If you remember, the idea behind this project was to allow people to use WebControls with Mvc — not just use WebControls in Mvc. On the last step we point our form to a Controller Action and use a method called Map to format our information for the postback.

When the user finished the Wizard they can post the finished form back to the correct Controller Action!

**whew** — that’s a lot to take in!

Limitations

There are a couple limitations that still need to be worked through…

• HTML output must be well formed – If your control outputs bad HTML, this thing will croak. The reason is that instead of using Regular Expressions to try and match content the output is parsed using the XDocument.
• PageMethods, AJAX calls will probably not work – Controls like the UpdatePanel or using PageMethods will most likely not work anymore. Some additional work is going to be required to make those work.

I’m sure there is more that needs to be looked at but at this point it seems as if it can be used — at least for testing.

Please let me know if you have any ideas, feedback or suggestions! I’m going to continue to work on this and see what kinds of improvements can be made.

Source Code

Download MvcWebForms.zip

.

Include Stylesheets and Scripts From A WebControl In MVC

As of MVC2 this code may not work anymore – Check here for updated code.

Have you ever created a WebControl in MVC and though “Gee, it sure would be nice if I could add a stylesheet/script to the header of my page.” It’s not as easy as it used to be. Lets just pretend you ran this code right here.

<%@ Page Language="C#" Inherits="System.Web.Mvc.ViewPage" %>
<html>
    <head runat="server">
        <title>Just A Test</title>
    </head>
    <body>
        <% Page.Header.Controls.Add(
            new LiteralControl(
                @"<script src=""/script.js"" type=""text/javascript"" ></script>"
                )); %>
    </body>
</html>

It would compile and run just fine but you wouldn’t see your script! That’s because the page life-cycle has come and gone and you’re in the middle of the page render! We’re too late!

Inline code doesn’t make it easy to communicate with other parts of the page. As far as I can tell, inline code is rendered in the same order that it appears on the page (or at least, parsed), so once you finally get to the end of the page, the code at the start has already been executed.

A Workable Solution

I’ve included some code at the end of this post that simplifies the whole process and for the remaining of this post I explain how it works. In an effort to solve this problem I ended up coming up with two extension methods that are attached to the HtmlHelper.

InsertMarker(id) : Creates a point on the page that will render any content that is added to it. Uses the id provided (either an enum or a string)

AppendToMarker(id, content) : Appends the string content to the matching marker id (either an enum or a string)

These methods allow marker points to be added to the page inline and then content appended to each marker, regardless of where they are used. Let’s look at an example of the code. If this code doesn’t make sense just yet, don’t worry — I’ll explain 🙂

A Simple Example

Lets use this code to add a script to the header of our page — but from within a WebControl.

[Index.aspx (the View)]

<%@ Page Language="C#" Inherits="System.Web.Mvc.ViewPage" %>
<html>
    <head runat="server">
        <title>Just A Test</title>
        <% this.Html.InsertMarker(Document.Head); %>
    </head>
    <body>
        <% this.Html.RenderPartial("SomeControl"); %>
    </body>
</html>

[SomeControl.ascx (the WebControl)]

<%@ Control Language="C#" Inherits="System.Web.Mvc.ViewUserControl" %>
<% this.Html.AppendToMarker(
    Document.Head,
    @"<script src=""/script.js"" type=""text/javascript"" ></script>"
    ); %>

As you can see, we can pick the name of the marker that we want to append content and it is added to the page — all from within our WebControl! You can define your own custom markers and append content to really anywhere you want on the page. If the marker isn’t found, then the content is never rendered.

How It Works (The Short Version)

Once we’re in the middle of our render event for our page it makes it quite difficult for us to make many changes to the rest of the document. This is where the HttpContext.Response.Filter stream comes in handy. With this stream we’re able to intercept all of the content before it is really sent out.

Each time we set a marker onto the page we place a bit of text to mark it as a point we need to replace before we push the content out. I had preferred the idea of remembering the position of the output stream, but apparently the output stream is written to all at once which took that out of the picture.

Once we’re done its as simple as using a Regular Expression and replacing all of the markers with the correct content.

MVC-ish – Kinda Sorta…

This may not have been what the guys who designed MVC were thinking of when they created it, but the ability to work with multiple areas of the page was definitely a feature that I felt was missing. How could something like this help you with your projects?

Source Code

ContentMarkerHtmlHelper.cs (Source)

.