CaubleStone Ink

.Net development and other geeky stuff

Application Helpers: Protect your method calls

Posted on January 26th, 2009


First thing, refactor, refactor, refactor. Most ideas come from refactoring of your code. It’s amazing how much you do when you start refactoring things. The big benefit of course is less code. If you have a couple hundred thousand lines of code or even ten thousand lines of code, refactoring even a small thing can cut down your code base and improve readability. This article is geared towards refactoring.

In the course of working on multiple applications over the years you find you tend to do the same things over and over again. One of those things is constantly checking objects for null, data type compatibility, etc. How many times have you written the following lines of code.

if (object == null)
{
   throw new ArgumentNullException();
}

or

if (object.GetType() == typeof(sometype))

The amount of code there justs adds up. Not only that but you probably have lots of null checks maybe even in the same method. So what if we wrote something like this instead:

Protect.IsNotNull(obj, "The object you are checking is null");

So if you had multiple objects going into an object you would have something like this:

Protect.IsNotNull(obj1, "Object 1 is null");
Protect.IsNotNull(obj2, "Object 2 is null");

Now since we are talking about refactoring what if we even refactor that code to make use of the newer features of the framework. Since the Protect methods throw an exception if the object is null in this case what if we wanted to expand it to allow for multiple checks. We could do something like this:

Protect.Against(obj1 == null, "Object 1 is null");
Protect.Against(String.IsNullOrEmpty(stringData), "The string is emtpy.");

As you can see we can now handle multiple things at the same time. So let’s look at what is under the Protect method.

public static class Protect
{
  public static void Against(bool condition, string message) where T: Exception
  {
    if (condition)
      throw new (T)Activator.CreateInstance(typeof(T), message);
  }
}

As you can see not much but it makes your code easier to read and it adds a nice bit of conformity while reducing the total lines you need to maintain and worry about. The T parameter is the type of exception you want thrown based on your condition. As long as you have a boolean condition your good. Anything that evaluates to true will throw the exception. The message will be on the exception that is thrown. Based on this pattern we can also add things like Inheritance checks, Type checks, Enum, checks, etc. It’s really easy for it to morph into what you need as you need it. You can even extend it with delegates, Func<> methods, setup for linq style syntax, etc. the sky is the limit.

Here is what it would look like in use inside of a method:

public void DoSomething(List data, Dictionary other)
{
    Protect.Against(data == null, "Data is null");
    Protect.Against(other==null, "Other is null");

    foreach (int i in data)
    {
        Console.WriteLine(other[i]);
     }
}

Introduction to the Circuit Breaker pattern

Posted on March 6th, 2008


Download the code (14kb)

I was reading the book Release It! about a month ago and ran across this pattern for creating code that has the ability to essentially shut itself off.  It just so happens that I had a use for this type of process and it helped me to create my final design.  What I was working on was a component that used a third party webservice for batch processing.  In this batch we would send over anywhere from 50-1500 web request one right after the other.  As such if we had any problems on the service end we could have thrown 1500 or even more error messages depending on how it was coded.  Of course we did not want that as it could potentially cause even worse problems by letting our app continue to try and process our data even if the service was down.  So instead of needlessly killing each of our apps I took a queue from the book and implemented the circuit breaker pattern. 

Now you might ask what is the circuit breaker pattern?  Well lets take a look at a real world scenario that everybody is familiar with.  Every home today has a main power box that is the central point at which power enters the home.  Now depending on how old the home is you could either have circuit breakers, fuses, or even worse a piece of metal jammed between the connections (I’d hope not).  Now fuses and circuit breakers all work the same way and for the same reason, to regulate the amount of load that is allowed over the electrical wire.  This is important because if there were no limit then it would cause an electrical short somewhere and possibly a fire.  It happened a lot before the fuse was created.  So, the use of a fuse or circuit breaker is meant to make sure we don’t put too much of a load on the circuit or wire.  Once a threshold is met then the circuit is tripped or the fuse is blown.  Thus stopping the flow of power.  Now if we look at our code base we want the same thing to occur.  So if I were to start sending records to the service in question and they start constantly failing as long as we have setup a threshold for error counts or failures we can then trip our circuit so to say and stop the flow of data to this service. 

Well you might ask how can we do this?  In our code bases is really not all that difficult to implement once you get the concept down.  So let’s look at some code to create a small example application.  Now you don’t have to create an interface but I wanted to really show what you need to do to add this functionality to your existing code.

public interface ICircuitBreaker
{
    void BreakCircuit();
    void ResetCircuit();
    bool CircuitBroken { get; set; }
    int Threshold { get; set; }
}

Next I’m going to embed an entire class with the interface implemented.  To show how this works I have created a class that writes to a file.  If there is a problem writing to the file for a specified number of tries then it will fail gracefully vs throwing an exception.

public class FileWriterWithCB : ICircuitBreaker
  {
    private bool _circuitBroken;
    private int _threshold;
    private string _fileName;
    private int _batchSize;

    public FileWriterWithCB()
    {
      this._circuitBroken = false;
      this._threshold = 5;
      this._fileName = "CBTest.txt";
      this._batchSize = 50;
    }

    public string FileName
    {
      get { return this._fileName; }
    }

    public int BatchSize
    {
      get { return this._batchSize; }
      set { this._batchSize = value; }
    }

    public void WriteToFile()
    {
      FileStream fs = new FileStream(this._fileName, FileMode.Open, FileAccess.ReadWrite);

      StreamWriter sw = new StreamWriter(fs);

      string data = "This is a test file write";

      sw.Write(data);

      sw.Flush();
      sw.Close();
      fs.Flush();
      fs.Close();

      sw = null;
      fs = null;
    }

    public void RunBatchWithCircuitBreaker()
    {
      int failureCount = 1;

      for (int x = 0; x &lt;= this._batchSize; x++)
      {
        if (failureCount &gt; this._threshold)
        {
          Console.WriteLine("Threshold reached breaking circuit.");
          this.BreakCircuit();
        }
        if (this.CircuitBroken)
        {
          Console.WriteLine("Circuit is Broken exiting code loop.");
          return;
        }

        try
        {
          // This should throw a FileNotFound Exception
          this.WriteToFile();
        }
        catch (Exception ex)
        {
          failureCount++;
          // Just write to the console to show that it is throwing the errors.
          Console.WriteLine(ex.Message);
        }

      }

    }

    #region ICircuitBreaker Members

    public void BreakCircuit()
    {
      Console.WriteLine("Circuit is being broken.");
      this._circuitBroken = true;
    }

    public void ResetCircuit()
    {
      Console.WriteLine("Circuit is being reset.");
      this._circuitBroken = false;
    }

    public bool CircuitBroken
    {
      get
      {
        return this._circuitBroken;
      }
      set
      {
        this._circuitBroken = value;
      }
    }

    public int Threshold
    {
      get
      {
        return this._threshold;
      }
      set
      {
        this._threshold = value;
      }
    }

    #endregion
  }

Ok, so let’s look at what this is doing.  First the class is used to write to a file.  There are two methods to write to the file.  One is a single write only the other is a batch writer for us to show the use of the circuit breaker pattern.  By default we set the system to perform 50 writes and the threshold is 5 errors, at which point we would expect the circuit to be proverbial broken.  In the attached code you will find a unit test that you can run to see the results for yourself. Needless to say when you run the codebase you should see that there are five errors thrown at which point the codebase effectively shuts down.  Try playing with the threshold.

Overall the pattern shown here can be very effective in integration scenarios but it’s especially effective in batch processes.  In the integration scenario you could have your app try to connect to say an external service and if it fails you want it to return a specified value vs crashing or throwing an exception.  This can be easily accomplished by adjusting how you use the circuit breaker.  All you would need to do is set your boolean variable if your external service is down.

Join the Darkside…

Posted on January 19th, 2008


Not that long ago I saw a post on one of the blogs out there about joining the DarkSide for Visual Studio.  I’m proud to say that I’ve been a Dark Sider now for over 10 years…  Yeah I know that’s before Visual Studio.net.  Prior to that I did a lot of VB and VBA work and I used a dark color scheme there as well.  Listed below is a screenshot of my dark scheme.  Which you can also get here for VS2005.  The font I am using here is the BitStream Vera Sans Mono which you can get here.

So come on over to the darkside and give us a try. 🙂

Book Review: Release It!: Design and Deploy Production-Ready Software

Posted on January 17th, 2008


After reading on several blogs that this was a book worth getting I decided to get myself a copy. I will say that this is a great book and that it should be on every programmer’s bookshelf. This is right up there with Code Complete. If you have not already gotten a copy you should do so.

When it comes to technical books I like to only purchase books that I will continue to use over the years vs ones that are very targeted and will go out of date. Thus architectural books are a prime candidate for purchase. However knowing which to buy can be daunting. After all who wants to plop down $50 or more for a book that might only have 1 or 2 good chapters in it or is a rehash of books you already have. That’s where these reviews come in. I picked up this book based on reviews on other blogs and I was not disappointed. While the information in the book is presented by an obvious Java programmer it in no ways is tied to Java programming. It provides insight into several projects from the real world and things that they learned from them. As such it is more of a best practices book vs a code book. From Anti-Patterns to Patterns it walks you through various scenarios and things to watch out for when developing release ready software.

After reading the book myself I was able to make changes to the way I was coding my components to take advantage of these patterns. By making use of this pattern I was able to take a service interface layer and change it such that the application using it would not crash from this component as it had done in the past. I was able to accomplish this by using the circuit breaker pattern that he describes in his book. Basically what it does is very much what you would expect. Just like in a home if you plug too many items into your outlet and it overloads the circuit it throws the breaker so you don’t burn your house down. So in code what this does it allows you to implement your code such that if you run into problems it can shut down the interface or process without taking down your application with it. After all who wants to crash their application or website because an integration point is down or broken.

There are quite a few other patterns some of which I have already taken advantage of like the Fail Fast pattern. I can honestly say that after reading this book almost any developer should be able to walk away a better programmer and will most likely start to use some of the patterns presented in this book. I know I have.

All in all this was an excellent book and it was written very well and not all that expensive. If you would like to see a more in-depth review see this blog post, Ayende’s Blog.
This book should be on every developer’s bookshelf. Get it now. You can get it here: Release It

Tip: Open a .Net Command Prompt Here

Posted on December 21st, 2006


Download (1k)

If you find you want to open a command prompt on a certain folder and have access to all the underlying Visual Studio.Net <your version here> options this is for you.

This is based on similar updates to your windows system in which you can add a new option to your right-click context menu inside of your explorer windows. What this particular tip does is shows you how to add a item to your menu that will open a VS.Net 2003 command prompt on the specified folder, however you follow the same procedure for all versions of visual studio.

Note: This involves updating the registry. You should always backup your registry before making manual changes as your changes could make your system unstable. If you are not comfortable with doing these types of changes you should pass on this tip.

Following the steps below:

1. Open regedit.exe
  Start->Run->regedit.exe
2. Navigate to the following location:
HKEY_CLASSES_ROOT/Directory/shell
3. Add a new key
  VS2003CMD
4. In this key create another key called:
  command
5. Click on the key created in step 3. On the right double-click the Default value and type in the following: (this will be displayed in the context menu).
  Open &VS2003 Command Prompt
6. Click OK
7. Click on the command subkey and then double-click on the default value on the right. Enter the following: (this is the command that will be run).

cmd.exe /k “cd %L && “c:/Program Files/Microsoft Visual Studio .NET 2003/Common7/Tools/vsvars32.bat” && ver”

Make sure you entered the quotes. This also assumes that you have installed VS.Net 2003 in the default folder. If you did not then adjust accordingly.

That should do it. Now if you go into windows explorer you should see a new menu item when you right-click.

To make this easier we have also included a download of the reg file that you can just apply to your system. As with anything to do with the registry you should always double check files you download that update the registry to make sure nothing malicious exists.

Enjoy.

Tutorial – Google Desktop Sidebar – Hello World Part 4, Loading and Saving our properties

Posted on December 10th, 2006


Requirements

.Net Framework SDK 1.1
Google Desktop SDK

Downloads

The zip file listed here is all inclusive of all four tutorial parts. We broke up the article into four parts for ease of reading. Included in the zip is a Word doc version of all four articles.
Download (231KB)

Purpose

In this article I intend to show you how to extend our simple Google Desktop Sidebar (GDS) component. All the source is in C# 1.1. We assume that all developers looking at this article are familiar with writing code in C# and the .Net framework. For this component we will be doing a simple Hello World type sample. We will show screenshots where appropriate. Otherwise we assume you know your way around.

What it Shows

In this part of our tutorial we will be showing you how to save and load the changes you have made. We do not take it so far as to keep up with the content but it gives you the general idea.

Storing our changes

Ok, as we all know if you make a change to something and you want it to stick we must save it. Well the GDK provides that functionality as well. In order to provide this save / load functionality we must implement the interface IPersistStreamInit. To start with we need to add some basic data to our class. We are going to add the following member variables.

const int dataVersion = 1;
private bool isDirty;

The dataVersion property can probably be replaced to read from your assemblyInfo.cs file but we will leave that to you. We need this property so we can support versioning of our component. It will also help if you happen to have multiple versions running that you get the correct data loaded for your component. You don’t really need this but it does provide a bit of added support for later. The isDirty flag will be set and used so we know when anything has changed so that we know when to save our data.

So, for us to know that something has changed we need to add it to our layout property. Let’s change our code:

public GoogleDesktopDisplayContentItemLayout Layout
{
  get { return this.layout; }
  set { this.layout = value; this.isDirty = true; }
}

All we have done here is add the isDirty flag to our property. Now we also need to initialize it so let’s add some code to our constructor.

this.isDirty = false;

Now we can start looking into the IPersistStreamInit interface. The first thing you will probably notice is that it has an IsDirty property. This is used by the GDK to determine if the Save operation needs to occur. So lets modify the code, note that we need to change a Boolean value into an unsigned integer.

public UInt32 IsDirty()
{
  return (UInt32)(isDirty ? 1 : 0);
}

Now that we can tell if our component needs to be saved letss add the code to save our changes. We will do this by adding code to the Save method.

public void Save(UCOMIStream stm, bool clearDirty)
{
  MemoryStream memStream = new MemoryStream();
  BinaryWriter binWriter = new BinaryWriter(memStream);
  // first our version
  binWriter.Write(dataVersion);
  // then the data.
  binWriter.Write(this.Layout.ToString());
  // cleanup binWriter.Flush();
  // write to actual output stream
  byte[] data = memStream.ToArray();
  stm.Write(data, data.Length, IntPtr.Zero);
  if (clearDirty) isDirty = false;
}

Now again, we assume you know how to use the framework and that includes the IO process. So, what we are doing is setting up a memory stream since the GDK actually handles the file IO. To Write to the stream you should be able to use any type of writer and pass it the memory stream to write to. The first thing we write out is our dataVersion and then we write out the value of our Layout property. You will notice that we are writing the string value representation, this will make it easier to pull back in later. Once we are done we convert the memory stream into an array of bytes so we can pass it to the underlying COM stream that the GDK has provided to us. Last but not least we set the dirty flag only if the stream write was successful.

Once we have saved our data we can now add code to load our data. This works the same way just in reverse. In a more advanced scenario there is no reason why you could not serialize / de-serialize a class into this stream. Let’s add our Load code.

public void Load(UCOMIStream stm)
{
  const int maxDataLen = 1000;
  // read a large enough number of bytes for us.
  byte[] data = new byte[maxDataLen];
  stm.Read(data, data.Length, IntPtr.Zero);
  MemoryStream memStream = new MemoryStream(data);
  BinaryReader binReader = new BinaryReader(memStream);

  int version = binReader.ReadInt32();
  if (version != dataVersion)
  {
    MessageBox.Show(&quot;Invalid data version, using default values&quot;);
  }
  else
  {
    string value = binReader.ReadString();
    switch (value)
    {
      case &quot;GDD_CONTENT_ITEM_LAYOUT_NEWS&quot;:
        this.Layout = GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_NEWS;
        break;
      case &quot;GDD_CONTENT_ITEM_LAYOUT_EMAIL&quot;:
        this.Layout = GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_EMAIL;
        break;
      case &quot;GDD_CONTENT_ITEM_LAYOUT_NOWRAP_ITEMS&quot;:
        this.Layout = GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_NOWRAP_ITEMS;
        break;
    }
  }
}

In this section of code we are again creating a Memory stream and this time a Reader object. You should create the same type of Reader as you did writer so that it will be compatible. Seeing as how we have only written a small amount we have set an arbitrary size of 1000 bytes to read which is way more than enough to cover getting our info. If you had a large set of data you would need to get the actual size of your data and initialize accordingly. Since we wrote out the first value of our dataVersion we can check it first to make sure we are working with the same file. If we are then we can read the rest of the data and set our Layout property to what we saved the component with last.

Well that should do it so compile your app and you should get the following prompt.

install

If you don’t fix any problems you have and then try again. You are ready to try it out. So open your GDS and then go into the properties and change the value and then exit the GDS and re-open your data should be displayed the same as when you exited the app.

The first time you run your app you will get an error saying that you have an invalid data version. We could wrap this but we will leave this to you the reader as an exercise.

What you have learned

In this last part of our tutorial set you have learned how to load and save your data changes that you make to your plug-in. If you want to make it more extensive we will leave that to you the reader as an exercise.

Conclusion

We hope that you have found these tutorials helpful. In looking through the developer forum we saw that a lot of people had asked for one and hope this meets the needs of what you were looking for.

Enjoy.

Tutorial – Google Desktop Sidebar – Hello World Part 3, Adding a Property Page

Posted on December 10th, 2006


Requirements

.Net Framework SDK 1.1
Google Desktop SDK

Downloads

The zip file listed here is all inclusive of all four tutorial parts. We broke up the article into four parts for ease of reading. Included in the zip is a Word doc version of all four articles.
Download (231KB)

Purpose

In this article I intend to show you how to extend our simple Google Desktop Sidebar (GDS) component. All the source is in C# 1.1. We assume that all developers looking at this article are familiar with writing code in C# and the .Net framework. For this component we will be doing a simple Hello World type sample. We will show screenshots where appropriate. Otherwise we assume you know your way around.

What it shows

In this part of our tutorial we will be adding a property page to our component that will allow us to change the content format among the three built-in types.

Adding Property Pages

We are assuming you have kept up so far if not pull down the code from above and you can follow along or you can work through the other two parts.

Since we are going to add a property page lets start with the obvious. Add a new Windows Form file to your project. Once complete open up the code window and add the following Using statements.

using System;
using System.Drawing;
using System.Collections;
using System.ComponentModel;
using System.Windows.Forms;
using System.Runtime.InteropServices;
using Google.Base;
using Google.Base.Utils;
using GoogleDesktopAPILib; // Google Type libraries
using GoogleDesktopDisplayLib; // Google Type libraries

As you can see it is mostly the same as a windows form but we have added the google items to the mix. If you look in the GoogleBase folder you will see a base property page object. This is taken pretty much straight from the samples that google provided. What you want to do is change your form class to inherit from the base property page class. That will hookup the form so it can be used by our plug-in. Your class should look something like this:

[GuidAttribute("DFAEFC1E-2466-40d5-80D6-3FF5136627DA")] // CHANGE THIS in your plug-in
[ClassInterface(ClassInterfaceType.None)]
public class frmHelloProperties : Google.Base.PropertyPage

Notice that again just like our main plug-in class we have a GUID Attribute. Since the plug-in / GDS needs a handle to the property page we need the GUID so that COM can get a hook into it for our use. Next by marking our class with the ClassInterface attribute we can control whether or not the class is exposed via a typelib. Since this is used internally we can mark it as none.

Now once you get that done you can setup your form however you want. Just like a normal form. For this tutorial we are going to add a label and a combo box. So add them to the form and change the Label to say “Change Display Layout”. Then change your combobox to be a drop-down list. Next add the following values to the Items collection, NoWrap, Email, and News.

Next let’s get back to our code. We are going to override two methods on our property page, OnPageActivate and OnPageApply. As there names imply we will be handling code when you open the property page and when you hit the Apply or OK button.

Let’s do the Activate method first.

/// <summary>
/// Called just before the property page is displayed
/// </summary>
protected override void OnPageActivate(System.IntPtr wndParent, Rectangle Rect, bool modal)
{
  base.OnPageActivate(wndParent, Rect, modal);
  HelloWorldPlugin ctl = dataObject as HelloWorldPlugin;
  if (ctl != null)
  {
    // Get the parameter values from the plug-in
    switch (ctl.Layout)
    {
      case GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_EMAIL:
         this.cboLayout.SelectedItem = "Email";
         break;
      case GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_NEWS:
         this.cboLayout.SelectedItem = "News";
         break;
      case GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_NOWRAP_ITEMS:
         this.cboLayout.SelectedItem = "NoWrap";
         break;
    }
  }
}

The first thing we do is let the base class do what it needs to do, then we apply our content information / changes. Each property page has a dataObject that equates to the plug-in you are using. So we need to set and cast our object to the type of our plug-in. If we have a plug-in handle then we can continue with our code. The property we have here, which we will create in just a few, is called Layout. It is of the type GoogleDesktopDisplayContentItemLayout which defines the three currently supported display formats out of the box. So what we do here is figure out what we are currently using and then set our combobox to the appropriate value.

Next lets look at the Apply process which is very similar.

/// <summary>
/// Called when user hits OK in the property page
/// </summary>
protected override void OnPageApply()
{
  HelloWorldPlugin ctl = dataObject as HelloWorldPlugin;
  if (ctl != null)
  {
    // Set the new parameter values back to the plug-in
    switch (this.cboLayout.SelectedItem.ToString())
    {
      case "NoWrap":
        ctl.Layout = GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_NOWRAP_ITEMS;
        break;
      case "Email":
        ctl.Layout = GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_EMAIL;
        break;
      case "News":
        ctl.Layout = GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_NEWS;
        break;
    }
    ctl.GetOurStuff();
  }
  base.OnPageApply();
}

Again we get a handle to our plug-in object. However this time we do our setting of our data. Now you will also notice that we are making a direct call to the GetOurStuff method. This way we can go through and re-apply our formatting to our content items. Last but not least we call our base method so that the GDK can finish anything else it needs to.

Since we defined the layout variable in part two you should be able to just add this class. Next lets go through and hookup the property pages. To do this we need to implement the ISpecifyPropertyPages interface. It’s pretty simple and straight forward so here is the code:

public void GetPages(ref CAUUID pages)
{
  // get our guid
  Guid[] g = new Guid[1];
  g[0] = typeof(frmHelloProperties).GUID;
  pages.SetPages(g);
}

What we are doing here is getting the GUID from our property page and then setting it into the pages object provided by the GDK. It will then get the hook to the property page and display it.

Let’s go ahead and compile it and see what we have so far. When you compile your code you should hopefully get the following dialog.

If you did not fix any errors you have and then continue. Next if you do not have the GDS running start it up.

To start you should see something like the following:

Now let’s open the property page by selecting the down arrow and choosing Options. As you will see it will show the NoWrap as the current display style. This is what we set in the constructor of our main plug-in code. Try changing it around and see how it looks.

Email View

News View

What you have learned

In this part of our tutorial you have learned how to add a property page and how to apply those changes to your plug-in. In our next part of this tutorial we will show you how to save your view layout information and restore it at run-time.

Tutorial – Google Desktop Sidebar – Hello World Part 2, A simple plug-in with Content

Posted on December 10th, 2006


Requirements

.Net Framework SDK 1.1
Google Desktop SDK

Downloads

The zip file listed here is all inclusive of all four tutorial parts. We broke up the article into four parts for ease of reading. Included in the zip is a Word doc version of all four articles.
Download (231KB)

Purpose

In this article I intend to show you how to extend our simple Google Desktop Sidebar (GDS) component. All the source is in C# 1.1. We assume that all developers looking at this article are familiar with writing code in C# and the .Net framework. For this component we will be doing a simple Hello World type sample. We will show screenshots where appropriate. Otherwise we assume you know your way around.

What it Shows

In this part of our tutorial we will be showing you how to add content items to your basic plug-in. It is assumed that you have the code from the previous article as we will be building from that base. So have it ready.

Adding Content

As stated we are now going to look at what it takes to add content items to our GDS plug-in. We are going to take the simple approach and just build a for loop that will add 10 items to the plug-in every time it’s loaded. That’s it.

So let’s get started. First add a new class file to your project from earlier and add the following using statements:

using System;
using System.Text;
using System.Windows.Forms;
using Google.Base;
// Base module to ease plug-in development
using Google.Base.Utils;
// Base module to ease plug-in development
using GoogleDesktopAPILib;
// Google Type libraries
using GoogleDesktopDisplayLib;
// Google Type libraries

As you will see this is very similar to what we had in our earlier tutorial. Now let’s add our class definition. If you want to change the class name go ahead.

public class HellowWorldCI :  GoogleDesktopDisplayContentItemHelperClass, IGoogleDesktopDisplayContentItemHandler 

Since this class is not created directly you do not need a GUID attribute on it. As you can see we are inheriting directly from the Helper class provided by the GDK and an Interface. Make sure you use the Class viewer to implement the interface for it.

Most everything you can pretty much leave alone. We will only show you the actual methods that we will be changing from the default implementation. As per the documentation you should throw the NotImplementedException when ever you do not implement a particular method. By doing this the underlying api will know to use the default method for handling this as per the application framework.

A lot of this block of code is taken from the samples that come with the GDK only we have cleaned it up a little bit and by breaking it into a separate file we feel that it makes it easier to understand and conceptualize what is going on.

Add the following code to the ProcessDetailsViewFeedback method:

if ((flags &amp; GoogleDesktopDisplayDetailsViewFlags.GDD_DETAILS_VIEW_FLAG_TOOLBAR_OPEN) != 0)
{
   OpenItem();
   return;
}
if ((flags &amp; GoogleDesktopDisplayDetailsViewFlags.GDD_DETAILS_VIEW_FLAG_NEGATIVE_FEEDBACK) != 0)
{
  MessageBox.Show(&quot;Not interesting: &quot; + heading + &quot;Doing default processing&quot;);
}
// throw this exception to let the default processing happen for all other
// details view commands
throw new NotImplementedException(); 

The OpenItem method will be defined in just a few minutes. All it will do is popup a message box. What this code does is when you are in the detail view for an item there are several processing options. You can click on the header, the Remove item, or the X in the right hand corner. By clicking on each different element you can have it do several things. In the case of an RSS feed if you click on the title it will normally take you to the web article in your current browser. If you click on the Remove Item it normally tells the interface to delete it, etc. By handling the information in this section you can have your code execute various actions depending on how you interact with it.

Next for the OnDetailsView method add the following code:

// Here we can create any ActiveX control for displaying the details, and
// return it via the detailsControl parameter. We choose the
// 'GoogleDesktopDisplayDetailsViewHelper' control that will give us a
// UI that's inline with how the sidebar looks.
GoogleDesktopDisplayDetailsViewHelperClass details = new GoogleDesktopDisplayDetailsViewHelperClass();
// set the details control's content from our own data
string text = &quot;Hello World... How are you/r/n/r/nSome sample text/r/n/r/nwithout formatting/r/n&quot;;  details.SetContent(source, time_created, text, false, layout);
detailsControl = details;
title = heading;
flags =  GoogleDesktopDisplayDetailsViewFlags.GDD_DETAILS_VIEW_FLAG_TOOLBAR_OPEN | GoogleDesktopDisplayDetailsViewFlags.GDD_DETAILS_VIEW_FLAG_NEGATIVE_FEEDBACK;
return false;
// return false to show details view, true to cancel it.

This block of code is executed when you click on an item and it goes to open the expanded details view for an article. The base class used here as seen is the GoogleDesktopDisplayDetailsViewHelper class. This is the base class used by the API to display the standard details view. If you want to implement your own you would need to write the appropriate ActiveX control to handle this. What we are doing I this class in our case is setting the internal details for each item when opened into details view. This is not good for dynamic content but for this example it works perfect. The SetContent method will take the values of the object in question and use it for rendering the object.

Add this code for the OpenItem method:

MessageBox.Show(&quot;Open: &quot; + this.heading);

All are doing here is that when you double click on the item it will display the heading for the item that you double-clicked on.

That pretty much does it for this class. Now we need to update the main plug-in class from Part 1. First we need to add a property to our class to handle the Layout of our content. This will play a bigger part in the later tutorials. For now let’s add this code.

public GoogleDesktopDisplayContentItemLayout Layout
{
  get { return this.layout; }
  set { this.layout = value; }
}

On our main plug-in class lets add the following method:

public void GetOurStuff()
{
  // clear our stuff out first this.
  RemoveAllContentItems();
  // setup the basic display options needed to add to object.
  GoogleDesktopContentItemDisplayOptions options =        GoogleDesktopContentItemDisplayOptions.GDD_ITEM_DISPLAY_IN_SIDEBAR |        GoogleDesktopContentItemDisplayOptions.GDD_ITEM_DISPLAY_AS_NOTIFICATION_IF_SIDEBAR_HIDDEN;
  // We are going to add 10 hello world items.

  for (int x=0; x&lt;=9; x++)
  {
    GDP_HelloWorld.HellowWorldCI content = new HellowWorldCI();
    content.heading = &quot;Hello World - &quot; + x.ToString();
    content.source = &quot;CaubleStone Ink&quot;;
    content.snippet = &quot;Sample Content Item from CaubleStone Ink.&quot;;
    content.time_created = DateTime.UtcNow;
    content.flags = GoogleDesktopDisplayContentItemFlags.GDD_CONTENT_ITEM_FLAG_NONE;
    content.layout = this.Layout;
    this.AddContentItem(content, options);
  }
}

The first thing for this method is that we will remove all the existing items from the plug-in. Essentially clear all. In the real-world this should be dynamic but for this tutorial it allows us to recreate all the items easily. Next we set our base options in this case we tell the item that it will be displayed in the Sidebar and that if the Sidebar is set to auto-hide you will get a toast notification of the change. After that we will be adding our content. We setup the loop to add 10 items to our content page. To do that create an instance of our content class that we created earlier. Then set the heading, source, snippet, etc. For the time use the UtcNow otherwise it will not calculate the added time correctly. We have also setup a property on our main class called Layout that will retrieve our base layout. If you look at Part 1 you will see that by default it is set to the No Wrap option. This will also come into play for our Part 3 tutorial. The last and final step is to add our content item.

Next we need to modify the SetSite method so that we can call our setup method and add our content to the screen.

displaySite = (IGoogleDesktopDisplaySite)site;
// Display our content here...
if (displaySite != null)
{
    this.SetFlags(GoogleDesktopDisplayPluginFlags.GDD_PLUGIN_FLAG_NONE,
                GoogleDesktopDisplayContentFlags.GDD_CONTENT_FLAG_HAVE_DETAILS);
  this.GetOurStuff();
}

What we have added is the IF condition and it’s contents. The first thing we do is set the flags on the plug-in to tell it that "Hey we have content here that needs to be displayed". After that we call our GetOurStuff method. If you do not add the flags the content will not be displayed.

Ok it’s now time to compile and test. If you get the following screen that’s good, if not adjust your errors until it compiles. This is what you should see.

Once you click yes open your GDS unless it is already open. You should see something like the screen below.

What we learned

In this part of our tutorial we have walked you through the process of adding content items to your component. You have seen how you can add individual items, remove those items, and the different changes needed in order to add it to your plug-in.

Tutorial – Google Desktop Sidebar – Hello World Part 1, A simple plug-in

Posted on December 10th, 2006


Requirements

.Net Framework SDK 1.1
Google Desktop SDK

Downloads

The zip file listed here is all inclusive of all four tutorial parts. We broke up the article into four parts for ease of reading. Included in the zip is a Word doc version of all four articles.

Download (231KB)

Purpose

In this article I intend to show you how to build a simple Google Desktop Sidebar (GDS) component. The entire source is in C# 1.1. We assume that all developers looking at this article are familiar with writing code in C# and the .Net framework. For this component we will be doing a simple Hello World type sample. We will show screenshots where appropriate. Otherwise we assume you know your way around.

What it shows

We intend to show you how to hook up a basic no frills .Net C# GDS component. What we have done is setup some basic modules under the folder GoogleBase that you can re-use in any component that you develop. These files wrap up all the API and com interaction that Google had setup in the samples sdk. Now it’s easy for you to use. In the article we will show you how to setup the component. Work with a title changing. Adjusting Layout formats (the canned ones). Adding content items. Using the default reader pane, and setting up Property pages.

This tutorial will walk you through the steps taken to get where the sample code is.

Project Setup

To start with create a new solution with and add a Class Library project. If you have not downloaded the code for this article please do so now as it will have some base files you will need to get going. Also, you need to make sure you have the Google Desktop SDK installed for this to work.

Copy and include the GoogleBase folder and all of it’s contents from the zip file into your project. These modules setup the following, ImageConverter object for getting your resource streams down to COM layer, the COM API, structures, etc needed by the GDK, and the base property page object that you can inherit from in order to create your property pages.

Once you have the files include you need to add some references to your project. The first two references are needed for the plug-in to work. In the references window click on the COM Tab and find the following two objects (figure 1) and add them to your project.

Once you have selected those objects it will create the necessary Interop dll’s needed since we are interacting with COM objects instead of managed code.

You also need to include the following references:

  • System.Drawing
  • System.Windows.Forms

Once that is done you need to set your project properties. The Google Sidebar is a COM object so we need to build a COM library. To do so we need to enable that in our project properties. So open your project properties and set the basics like Assembly name, root namespace, etc. Once you have that done goto the Configuration Properties -> Build section and at the bottom of your options change the Register for COM Interop to TRUE. This will enable us to build a com library.

That’s the basic setup for the project. Now we need to start putting together the basic plug-in.

Compilation Notes

When building a COM library you will need some registration functions in your code to allow it to hook in properly. We will show that after we get the basic plug-in up and running with a title.

However, once you compile and register your component with the GD Sidebar you cannot modify it unless you exit the Sidebar application. This will remove the hooks so you can re-compile and re-register. Tedious process if you are doing a lot of testing but it works.

Your first plug-in

Ok, lets setup our first plug-in. Create a new Class file in your project and the following includes for the file.

using System;
using System.Text;
using System.Runtime.InteropServices;
using System.Windows.Forms;
using System.IO;

// used for Registry / Com interop registration
using Microsoft.Win32;
using Google.Base;
// Base module to ease plug-in development
using Google.Base.Utils;
// Base module to ease plug-in development
using GoogleDesktopAPILib;
// Google Type libraries
using GoogleDesktopDisplayLib;
// Google Type libraries

Once you have these in let’s create our first plug-in class. For your class, add the following inheritance chain (or cut / paste and change your class name):

public class HelloWorldPlugin : GoogleDesktopDisplayPluginHelperClass, IGoogleDesktopDisplayPluginHandler, ISpecifyPropertyPages, IObjectWithSite, IPersistStreamInit

Once you have the class inheritance model in place use the Class view to go in and implement all the interfaces.

Now let’s setup some of our base internal values that we will use with our com object. These values will be used with the registration and read by the GDS when you view available plug-ins.

// since this is a com object we can store some info as static members.
static string baseTitle = "Hello World";
static string pluginName = "Hello World";
static string pluginDesc = "Hello World sample written in C#";

There is one other item you must add and it is extremely important. All COM objects need a GUID. So let’s add ours. Using the GUIDGEN utility from the framework choose the registry format to generate and take the value and add it to our class in the following ways:

[GuidAttribute("A674FDFF-D816-4eaf-9D5E-97317158733E")]
public class HelloWorldPlugin : GoogleDesktopDisplayPluginHelperClass, IGoogleDesktopDisplayPluginHandler, ISpecifyPropertyPages, IObjectWithSite, IPersistStreamInit
{
  // You must set a control GUID and it must match the guid attribute in order to work static string     controlGuid = "{A674FDFF-D816-4eaf-9D5E-97317158733E}";
}

The GUID’s in both places need to match. Without this you will be unable to register and use your plug-in.

You will also need to add the following member variables to the class:

// Member variables needed to hook into the Sidebar api
// this retrieves the site info for this plug-in.
private IGoogleDesktopDisplaySite displaySite;

// setup the layout options so the sidebar api knows how to render our information.
private GoogleDesktopDisplayContentItemLayout layout;

What this does is allows your app to hook into the panel that is showing your information. You will notice that we set up above a baseTitle object as static. You do not need to do this if you plan to have your title bar be dynamic. However it aids in this demonstration. The layout property will be used in a later tutorial.

Ok, now that we have some member variables, our interfaces implemented (roughly) and a class to use let’s build the constructor. Here we will setup an initial layout format as well as hook in the About box information. Real quick here about the about box. There are three distinct areas that text is put. The first line will use your logo if any and display your main info like a company name. If you use two new line commands it will kick your text down to the main content area.

Here is our constructor:

public HelloWorldPlugin()
{
  // setup some basic info.
  StringBuilder sb = new StringBuilder();
  sb.Append("Hello World Plug-in by CaubleStone Ink");
  sb.Append("//n//n");
  sb.Append("Shows how to setup basic functionality");
  sb.Append("for a sidebar plug-in.//n");
  this.layout = GoogleDesktopDisplayContentItemLayout.GDD_CONTENT_ITEM_LAYOUT_NOWRAP_ITEMS;   this.about_text = sb.ToString();
}

What we have done here is setup our basic info using a StringBuilder and setting our basic layout to just the plain jane no wrap value. Similar to the TODO list from Google.

Now lets walk through setting up the interfaces. By default you can leave all the interfaces alone however you may need to add a few new keywords on some to aid in the code generation and to help avoid memory leaks.

We are going to start with the IObjectWithSite interface class. All we need to do here is add the following so our app can get a handle to the display surface that the GDS will be using.

public void GetPageContainer(out Object site)
{
  // TODO: Add HelloWorldPlugin.GetPageContainer implementation
  site = null;

}

public void SetSite(Object site)
{
  displaySite = (IGoogleDesktopDisplaySite)site;
}

That’s it. We are slowly on our way to a plug-in. Now on to the next interface IGoogleDesktopDisplayPluginHandler. In this interface we will hook up our Title for our plug-in. What we plan on having is the plug-in adjust the title from Hello World to Hello World – Min when it’s minimized or maximized.

public new void OnCommand(GoogleDesktopDisplayPluginCommand command)
{
  if (command == GoogleDesktopDisplayPluginCommand.GDD_CMD_ABOUT_DLG)
  {
    // always throw this exception when you dont intend to override the default behavior
    // and you want the plug-in helper to do it[ES][SQ]s usual stuff.
    throw new NotImplementedException();
  }
}

public new void OnDisplayStateChange(GoogleDesktopDisplayTileDisplayState display_state)
{
  if (display_state == GoogleDesktopDisplayTileDisplayState.GDD_TILE_DISPLAY_STATE_MINIMIZED ||        display_state == GoogleDesktopDisplayTileDisplayState.GDD_TILE_DISPLAY_STATE_RESTORED)
  {
    // switching between minimized and normal mode, so update the title
    this.GetTitle();
  }
}

Notice that we have the new keyword on both of these methods. That is so it hides the underlying implementation (which there is none since it’s an interface) and helps us to prevent memory leaks. What we are doing is checking the value of display_state which comes in from the interface call. If the component has been minimized or restored it will change our title.

Let’s get our GetTitle function in here.

public void GetTitle()
{
  this.title = baseTitle;
  if (displaySite != null && displaySite.display_state ==  GoogleDesktopDisplayTileDisplayState.GDD_TILE_DISPLAY_STATE_MINIMIZED)
  {
    this.title = baseTitle + " - Min";
  }
}

What we are doing here is setting the underlying classes title property. We are starting with our baseTitle static variable and then appending the “- Min” to it if we are minimized.

Now we are finished with our basic code. Let’s get our registration functions in place so we can see the final results.

COM Registration functions

#region Registration functions
/// <summary>
/// Called when the plug-in is registered with the system. We add a few registry
/// keys and register with GoogleDesktop as a plug-in.
/// </summary>
/// <param name="t" /></param>
[ComRegisterFunctionAttribute]
static void RegisterFunction(Type t)
{
  try
  {
    // Set the "MiscStatus" value in registry to a valid value string keyName = @"CLSID\" + t.GUID.ToString("B");

    using (RegistryKey key = Registry.ClassesRoot.OpenSubKey(keyName, true))
    {
      key.CreateSubKey("Control").Close();
      using (RegistryKey subkey = key.CreateSubKey("MiscStatus"))
      {
        subkey.SetValue("", "whateveryouwant");
      }
      key.CreateSubKey("Control").Close();
      key.SetValue("",pluginName);
      using (RegistryKey subkey = key.CreateSubKey("Description"))
      {
        subkey.SetValue("", pluginDesc);
      }
    }

    // Create the registrar object
    GoogleDesktopRegistrarClass registrar = new GoogleDesktopRegistrarClass();

    // Start component registration by specifying our attributes
    object[] descriptions = { "Title", pluginName, "Description", pluginName, "Icon", "" };

    registrar.StartComponentRegistration(controlGuid, descriptions);
    // A single component can register for multiple plugins with Google Desktop.
    // Here we register a single display plug-in.
    IGoogleDesktopRegisterDisplayPlugin displayRegistration =
        (IGoogleDesktopRegisterDisplayPlugin)         registrar.GetRegistrationInterface("GoogleDesktop.DisplayPluginRegistration");

    displayRegistration.RegisterPlugin(controlGuid, false);
    // Done with component registration.
    registrar.FinishComponentRegistration();
  }
  catch (Exception e)
  {
    MessageBox.Show("Exception thrown during registration. Description=" + e.Message);
  }
}

/// <summary>
/// Called when the plug-in is unregistered. We unregister our plug-in with
/// GoogleDesktop.
/// </summary>
/// <param name=""t"" /></param>
[ComUnregisterFunctionAttribute]
static void UnregisterFunction(Type t)
{
  try
  {
    // Create the registrar object
    GoogleDesktopRegistrarClass registrar = new GoogleDesktopRegistrarClass();
    // Unregister ourselves
    registrar.UnregisterComponent(controlGuid);
  }
  catch (Exception e)
  {
    MessageBox.Show("Exception thrown during registration. Description=" + e.Message);
  }
}
#endregion

What this block of code does is writes to the registry values needed to register your component for COM interop. By using the GoogleDesktopRegistarClass it also registers your COM object for use and visibility by the GDK. For the most part you can just copy / paste whenever you build a plug-in.

Compiling and Running

Ok, now that you have all the code in place it’s time to compile and run your application. If your build was successful you should see the following:

Once you get this message hit yes. If you have the Google Sidebar running you should see it show up immediately. If not startup the Sidebar and if not there add the panel.

Based on what we have done you should see something like the following:

Normal

Minimized

What you are seeing is the Hello World plug-in with only a title. The title will change when you Minimize or Restore the panel.

What you have learned

In this part of our tutorial you have learned the basics of creating a very simple plug-in for the Google Desktop Sidebar using C#. As you can see it only took minimal setup steps thanks to some of the items that Google provided and we broke out for you.

In our next article we will cover adding content items, basic format to the plug-in.

Errata

04-12-2006

Thanks to M. Loeffen for catching this. If you were working through the examples you would indeed recieve a compile error because we did not define the layout property. We have adjusted the tutorial to add this variable to our main plugin class. If you want to see it here this is what you need to add:

  // setup the layout options so the sidebar api knows how to render our information.
  private GoogleDesktopDisplayContentItemLayout layout;

Understanding Value and Reference Types

Posted on March 10th, 2004


Supporting Downloads
valuereferencetypes.doc (52kb)
valuereferencetypes.zip (11kb)

Purpose

The purpose of this document is to explain the usage of Value and Reference types within the dot net framework. Since each type behaves differently when used as a plain value and when used as a parameter to a method, it is important to understand what is actually happening. We will be covering how each type actually works within the framework when used by themselves and when used as parameters to a method.

Meet the Types

When dealing with the dot net framework you will come across many objects. Each object will belong to either the Value type or Reference type category. To understand how these two objects differ we need to look at how memory is allocated for our application.

There are two areas in memory that are used to store this data, the stack and the heap. If you are a C++ programmer you are probably familiar with these memory units. In the dot net framework all Value types are pushed to the stack where as Reference types are pushed to the heap while a pointer to this memory usage is placed on the stack. Let’s break it down. When you create a Value type in your application it asks the framework to allocate memory based on the size of the type to hold whatever data you need. This data is then placed on the stack for usage. Once data goes onto the stack it stays on the stack until it is removed by our application. This can happen when a variable goes out of scope. Also, if you assign a Value type to another Value type a copy is made of this data and placed on the stack so, even though they are both the same thing the variables are accessing the data from two very different areas of memory. Let’s look at an example:

VB.Net

Public Sub TestValueType()
  Dim x As Integer
  Dim y As Integer

  x = 5
  y = x + 5

  Console.WriteLine(x)
  Console.WriteLine(y)
End Sub

C#

public void TestValueType()
{
  int x, y;

  x = 5;
  y = x + 5;

  Console.WriteLine(x);
  Console.WriteLine(y);
}

In our example we created two variables x and y, which we told the framework were Integers (which is a value type, see the chart at the end of this document for a list of common Value types.). Next we assign the variable x to the value of 5, this tells the framework to place on the stack a value of 5 and assign it to the variable x. Next, we assign the value of y to the value of x. This has just told the framework to create a value of 5 on the stack and assign it to the variable y. Notice, that it did not set the value on the stack to be x but rather the contents of the variable x. This is important because we are dealing with value types and the stack, we have created an exact duplicate of the value stored in the variable x and it did not just point the variable y to the value of x, but instead copied it’s data. Next, we assign the value of y to be equal to its current value plus 5. This will of course make the y variable equal to the value of y plus 5. So, our variable y is now 10. Now, how did this effect our value of x, it did not. Since the variables we are using are value types the data was copied rather than referenced. Later I will show how this can be very beneficial. The key to understand is that any object that is placed on the stack is always copied or modified directly it is never referenced.

Reference types however are placed on the heap. The heap is used in a very different manner than the stack. The heap is a dynamic memory location where objects that may exist at any point in time are created, destroyed, used, etc. in any order, there is no logical flow to the creating or removing of elements on the heap. When working in the dot net framework it is important to understand the heap is managed by the Garbage Collection services in the dot net framework. This means that it is possible and highly likely that an object that goes out of scope in your application is still in memory and can be accessed. This is where you get memory leaks and the nice blue screen of death (when they are not handled properly). When you create a reference type the framework tells the computer to create an area of memory on the heap (wherever it can) and then assigns our objects data to this memory location. Now for our object to be able to have access to this data the framework also creates a location on the stack in which it creates a pointer to the memory address at which our object resides on the heap. Now if you remember earlier we stated that when you create a copy of a value type it actually copies the data from the stack to another area of the stack, almost like a photocopier makes exact copies yet they are in themselves separate objects in that you can highlight data on one page and throw another away and whatever you do only effects the page you are dealing with. However with a reference type when you copy the object you are coping the memory pointer not the actual data itself. What does this mean, it means that whatever you do to one you do to the other. Let’s look at an example:

VB.Net

Public Sub TestReferenceType()
  Dim f1 As New System.Windows.Forms.Form()
  Dim f2 As New System.Windows.Forms.Form()

  f1.Text = "This is form1"

  Console.WriteLine(f1.Text)
  f2 = f1
  f2.Text = "This is form2"

  Console.WriteLine(f1.Text)
End Sub

C#

public void TestReferenceType()
{
  System.Windows.Forms.Form f1, f2;

  f1 = new System.Windows.Forms.Form();
  f2 = new System.Windows.Forms.Form();

  f1.Text = "This is Form1";

  Console.WriteLine(f1.Text);
  f2 = f1;
  f2.Text = "This is Form2";

  Console.WriteLine(f2.Text);
}

What will the second writeline statement print? If you said This is Form2 then you are correct. Why did this happen? Earlier we stated that a reference types data exists on the heap and that a pointer to the memory reference exits on the stack, now when we copied our object into the f2 variable it copied the data that exits on the stack. In the reference type world this is also known as a shallow copy. The reason for this is that only the data that exists on the stack is copied such that you now have two objects in code that are pointing to and using the same information. Why would you want this?, you might ask. Well let’s say that you have an application that needs to modify a value by passing it to a method and return the result. If you can pass a reference to this object so that it will point to the original, you can modify it without having to use a lot of extra memory to do so. This will also help the performance of your application since you will not need to perform a complete copy and use twice as much data space in memory.

Using Types with Methods

We are now going to look at how types are used when passed to a method. As you may know there are two ways to pass a parameter to a method, by value, and by reference. If you remember from earlier how Value and Reference types are different when assigning them to different values then you can probably guess how the data is passed to the methods. For those of you who are coming from prior versions of VB you probably remember that by default all parameters were passed by reference. In the dot net framework by default everything is passed by value. Which in most cases is what we intend to begin with.

Now the difficulty is now remembering how the types interact with the application when they are passed by value or by reference. Let’s first look at how to pass a Value type by value and by reference.

VB.Net

Public Sub PassingValues()
  Dim x As Integer
  Dim y As Integer

  x = 5
  y = x + 5

  PassValueTypes(x, y)

  Console.WriteLine(x)
  Console.WriteLine(y)
End Sub

Public Sub PassValueTypes(ByVal xval As Integer, _ ByRef yval As Integer)
  xval += 10
  yval += 20
End Sub

C#

public void PassingValues()
{
  int x,y;

  x = 5;
  y = x + 5;

  PassValueTypes(x,ref y);

  Console.WriteLine(x);
  Console.WriteLine(y);
}

public void PassValueTypes(int xval, ref int yval)
{
  xval += 10;
  yval += 20;
}

Now lets find out what happened. Can you guess what is going to be written to the console? If you guessed 5 for x and 30 for y then you are correct. However you might be asking how since you added 10 to the x value as well. When you pass a Value type by value to a method the framework creates a copy of the data and assigns it to the variable that is defined in the function and places that data into a new location on the stack. Now, when you pass a Value type by reference something different takes place. While the data is still located on the stack a pointer is created, again on the stack, that tells the framework where the data is for the value in question. In this way you have created a way for the method to access directly the data contained in the passed Value type variable. When this happened you now have the ability to alter the data that exists in that variable and since you are accessing it directly the calling variable will be able to see those changes. Once the calling method is finished it clears the two variables off of the stack, the by value variable is destroyed while the by reference variable only has the pointer destroyed.

So, when you pass a Value type to a method by value it is just like creating another Value type and assigning it the data, which creates a copy of the other variable on the stack. However when you pass a Value type by reference it behaves more like a Reference type in that the framework creates a pointer instead of a copy of the data so that the method will know where the data is on the stack, by doing this the method can modify the value of the variable directly.

Reference types however behave much differently. Lets look at some code.

VB.Net

Public Sub PassingReferences()
  Dim f1 As New System.Windows.Forms.Form()
  Dim f2 As New System.Windows.Forms.Form()

  f1.Text = "This is Form1"
  f2.Text = "This is Form2"

  PassRefTypes(f1, f2)

  Console.WriteLine(f1.Text)
  Console.WriteLine(f2.Text)
End Sub

Public Sub PassRefTypes(ByVal frm1 As System.Windows.Forms.Form, ByRef frm2 As System.Windows.Forms.Form)
  frm1.Text = "We have modified form1"
  frm2.Text = "We have modified form2"
End Sub

C#

public void PassingReferences()
{
  System.Windows.Forms.Form f1, f2;

  f1 = new System.Windows.Forms.Form();
  f2 = new System.Windows.Forms.Form();

  f1.Text = "This is Form1";
  f2.Text = "This is Form2";

  PassReferenceTypes(f1,ref f2);

  Console.WriteLine(f1.Text);
  Console.WriteLine(f2.Text);
}

public void PassReferenceTypes(System.Windows.Forms.Form frm1, ref System.Windows.Forms.Form frm2)
{
  frm1.Text = "We have modified Form1";
  frm2.Text = "We have modified Form2";
}

Can you guess what is going to happen here? If you guessed that the information written to the console will be what is in the PassReferenceTypes function then you are correct. Let’s see why.

As you know when you create a Reference type the data for that type is created on the heap with a pointer placed on the stack. When you copy a Reference type, its pointer is copied into the new variable and placed on the stack. The data it contains on the heap is not replicated, so in essence they both point to the same data. Now here is the gotcha, when you pass a Reference type by value it works the same as when you pass it by reference. Why is this, you might ask. Well when you pass a variable by value what happens? The framework makes a copy of what is on the stack and sends it to the method. Now what does a Reference type have on the stack, a pointer to the heap. So, if we pass a Reference type by value it is going to copy its stack value which in this case is the pointer. This is very important to understand because when you pass a reference type to any method whether it is by value or by reference it will always create a new variable on the stack that contains a pointer to data that is managed on the heap. Thus, you need to be careful when passing a reference type and be sure of what your intending when doing it.

Value Type(s)

Below is a list of the common Value types that are defined in the dot net framework. This list was pulled from the MSDN Library.

C#
enum, struct, bool, byte, sbyte, char, decimal, double, float, int, uint, long, ulong, short, ushort, and string.

VB.Net
Enum, Struct, Boolean, Byte, Char, Date, Decimal, Double, Integer, Long, Short, Single, and String.

Conclusion

I hope that this document will help alleviate some of the confusion around Value and Reference types, and how they are used and managed within the dot net framework. It is also my hope that this document will help eliminate more questions than it answers. So, when venturing into the world of dot net be careful of the types that you use.

Page 2 of 212