Paulo Morgado

Web Site Globalization With ASP.NET Routing

For those who don’t know, I have this web site http://PauloMorgado.NET/ that I use both as a web presence besides my blogs and a playfield.

Because I write both in English and Portuguese, I wanted the web site to have both English and Portuguese versions. This is easily accomplished by using ASP.NET Globalization and Localization.

But I wanted to do more than guessing the user’s language form her/his web browser languages. I wanted something like the MSDN and TechNet web sites have where the culture is embedded in the URL which makes it easy for the user to choose in which language she/he wants to see the web site.

With the release of the ASP.NET Routing, this is as easy as writing a custom route handler that sets the culture for the request and returns the requested page handler.

Something like this:

public class GlobalizationRouteHandler : global::System.Web.Routing.IRouteHandler
{
    System.Globalization.CultureInfo culture;
    System.Globalization.CultureInfo uiCulture;

    public GlobalizationRouteHandler(System.Globalization.CultureInfo culture)
        : this(culture, culture)
    {
    }

    public GlobalizationRouteHandler(CultureInfo culture, CultureInfo uiCulture)
    {
        if (culture == null)
        {
            throw new ArgumentNullException("cultureInfo", "cultureInfo is null.");
        }

        if (uiCulture == null)
        {
            throw new ArgumentNullException("uiCulture", "uiCulture is null.");
        }

        this.culture = culture;
        this.uiCulture = uiCulture;
    }

    private GlobalizationRouteHandler()
    {
    }

    #region IRouteHandler Members

    public IHttpHandler GetHttpHandler(RequestContext requestContext)
    {
        Thread.CurrentThread.CurrentCulture = this.culture;
        Thread.CurrentThread.CurrentUICulture = this.uiCulture;

        string path = "~/" + (requestContext.RouteData.Values["path"] as string);

        var physicalPath = requestContext.HttpContext.Server.MapPath(path);
        if (System.IO.Directory.Exists(physicalPath))
        {
            path = VirtualPathUtility.Combine(path, "Default.aspx");
        }

        var httpHandler = BuildManager.CreateInstanceFromVirtualPath(path, typeof(IHttpHandler)) as IHttpHandler;

        return httpHandler;
    }

    #endregion
}

And now it’s only a matter of registering the handled cultures:

routes.Add("en", new Route("en/{*path}", new GlobalizationRouteHandler(CultureInfo.GetCultureInfo("en-US"))));
routes.Add("pt", new Route("pt/{*path}", new GlobalizationRouteHandler(CultureInfo.GetCultureInfo("pt-PT"))));

Coupling ASP.NET Session State With Forms Authentication

Today I was talking with João about a way to couple the lifetime of the ASP.NET session state with the lifetime of Forms Authentication ticket.

My idea was to store the session ID in the UserData property of the forms authentication ticket upon logon and retrieve it with a custom session ID manager.

The login code would be something like this:

protected void Login1_Authenticate(object sender, AuthenticateEventArgs e)
{
    bool isPersistent = this.Login1.RememberMeSet;
    string username = this.Login1.UserName;
    var ticket = new FormsAuthenticationTicket(
        0,
        username,
        DateTime.Now,
        DateTime.Now.AddMinutes(2),
        isPersistent,
        Guid.NewGuid().ToString("N"));

    // Encrypt the ticket.
    var encryptedTicket = FormsAuthentication.Encrypt(ticket);

    // Create the cookie.
    this.Response.Cookies.Add(new HttpCookie(FormsAuthentication.FormsCookieName, encryptedTicket));

    // Redirect back to original URL.
    this.Response.Redirect(FormsAuthentication.GetRedirectUrl(username, isPersistent));
}

For the purpose of this test I am using a Guid as the session ID.

The session ID manager will return this session ID when queried by the session state HTTP module:

public class SessionIdManager : global::System.Web.SessionState.ISessionIDManager
{
    #region ISessionIDManager Members

    public string CreateSessionID(HttpContext context)
    {
        return GetDummySessionIdOrRedirectToLoginPage(context);
    }

    public string GetSessionID(HttpContext context)
    {
        return GetSessionIdFromFormsIdentity(context);
    }

    public void Initialize()
    {
    }

    public bool InitializeRequest(HttpContext context, bool suppressAutoDetectRedirect, out bool supportSessionIDReissue)
    {
        supportSessionIDReissue = false;
        return GetSessionIdFromFormsIdentity(context) == null;
    }

    public void RemoveSessionID(HttpContext context)
    {
    }

    public void SaveSessionID(HttpContext context, string id, out bool redirected, out bool cookieAdded)
    {
        redirected = false;
        cookieAdded = false;
    }

    public bool Validate(string id)
    {
        return true;
    }

    #endregion

    private static string GetSessionIdFromFormsIdentity(HttpContext context)
    {
        var identity = context.User != null ? context.User.Identity as FormsIdentity : null;

        if ((identity == null) || (identity.Ticket == null) || string.IsNullOrEmpty(identity.Ticket.UserData))
        {
            return GetDummySessionIdOrRedirectToLoginPage(context);
        }
        else
        {
            return identity.Ticket.UserData;
        }
    }

    private static string GetDummySessionIdOrRedirectToLoginPage(HttpContext context)
    {
        if (context.Request.CurrentExecutionFilePath.Equals(FormsAuthentication.DefaultUrl, StringComparison.OrdinalIgnoreCase)
                        || context.Request.CurrentExecutionFilePath.Equals(FormsAuthentication.LoginUrl, StringComparison.OrdinalIgnoreCase))
        {
            return Guid.NewGuid().ToString("N");
        }
        else
        {
            FormsAuthentication.RedirectToLoginPage();
            return null;
        }
    }
}

NOTE: Although this might work, it’s just an intellectual exercise and wasn’t fully tested.

Playing With LINQ: Getting Interface Property Implementations

Today, my friend Nuno was writing some code to get the PropertyInfos of a class implementation of an interface.

Given this interface:

public interface ISomeInterface
{
    int IntProperty { get; set; }
    string StringProperty { get; }
    void Method();
}

and this class:

public class SomeClass : ISomeInterface
{
    int ISomeInterface.IntProperty { get; set; }
    public int IntProperty { get; private set; }
    public string StringProperty { get; private set; }
    public void Method() { }
}

Nuno wanted to retrieve:

• Int32 ISomeInterface.IntProperty
• System.String StringProperty

The code is fairly simple. First we need to get the interface mappings for the class:

typeof(SomeClass).GetInterfaceMap(typeof(ISomeInterface)).TargetMethods

and get all PropertyInfos for which the MethodInfo in the list is part of the implementation of the property (is either the set method or the get method).

Something like his:

public static bool Implements(this MethodInfo methodInfo, PropertyInfo propertyInfo)
{
    return (propertyInfo.GetGetMethod(true) == methodInfo) || (propertyInfo.GetSetMethod(true) == methodInfo);
}

But what caught my eye was that, with the above extension methods, I can use LINQ to retrieve a list of the desired PropertyInfos.

Something like this:

public static IEnumerable<PropertyInfo> GetInterfacePropertyImplementation(Type implementer, Type implemented)
{
    return (from propertyInfo in implementer.GetProperties(BindingFlags.Instance | BindingFlags.Public | BindingFlags.NonPublic).AsEnumerable()
            from methodInfo in implementer.GetInterfaceMap(implemented).TargetMethods.AsEnumerable()
            where methodInfo.Implements(propertyInfo)
            select propertyInfo).Distinct();
}

For the sample class and interface, the usage would be something like this:

var q = GetInterfacePropertyImplementation(typeof(SomeClass), typeof(ISomeInterface));

foreach (var p in q)
{
    Console.WriteLine(p);
}

Which would produce the following output:

Int32 ISomeInterface.IntProperty
System.String StringProperty

UPDATED: The previous implementation was overcomplicated and had some string based logic. Kudos to Nuno.

A TraceListener For Tests

In my code, I make extensive use of debug assertions (see System.Diagnostics.Debug.Assert). These assertions are very helpful while debugging because you don’t need to step into every line of code to see if all pre-conditions are met. As soon as a pre-condition fails, an assertion failed window will pop up and will allow us to either abort, ignore or go to the assertion instruction (retry).

Imagine you have this code:

private void IKnowForSureThatANullStringWillNeverBePassed(string text)
{
    System.Diagnostics.Debug.Assert(string != null, "text is null.");

    // ...
}

Because this method is private, I have full control of what is passed to the text parameter, therefore I’m asserting that it will never be null. Because it might not be obvious that text will never be null, the assertion also acts as documentation.

I usually run my unit tests and integration test on debug compilations and these assertions would be helpful by making the test fail on an assertion fail instead of running through the method and failing with an NullReferenceException. That’s why I (and more people out there) wrote this simple TraceListener:

public class TraceListener : global::System.Diagnostics.TraceListener
{
    public static readonly TraceListener Default = new TraceListener();

    protected TraceListener()
    {
        this.Name = "Testing Trace Listener";    
    }

    protected TraceListener(string name)
        : base(name)
    {
    }

    public override void Write(string message)
    {
    }

    public override void WriteLine(string message)
    {
    }

    public override void Fail(string message, string detailMessage)
    {
        var builder = new global::System.Text.StringBuilder();

        builder.Append(message);

        if (detailMessage != null)
        {
            builder.Append(" ");
            builder.Append(detailMessage);
        }

        throw new global::Microsoft.VisualStudio.TestTools.UnitTesting.AssertFailedException(builder.ToString());

    }
}

This trace listener won’t write anything to anywhere. It will just throw an AssertFailedException if the Fail method is called, which is what happens when an assertion fails.

Because an assertion window is not desired when running the tests (specially if they are automated tests ran as part of a build process) it’s better to disable the assert UI on the configuration file of the test project.

<?xml version="1.0" encoding="utf-8" ?>
<configuration>
    <system.diagnostics>
        <assert assertuienabled="false"/>
        <trace>
            <listeners>
                <add name="TestTraceListener"
                     type="PauloMorgado.TestTools.VisualStudio.UnitTesting.Diagnostics.TraceListener, PauloMorgado.TestTools.VisualStudio" />
            </listeners>
        </trace>
    </system.diagnostics>
</configuration>

How To Set Elements Of An Array Of A Private Type Using Visual Studio Shadows

Visual Studio uses Publicize to create accessors public for private members and types of a type.

But when you try to set elements of a private array of elements of a private type, things get complicated.

Imagine this hypothetic class to test:

public static class MyClass
{
    private static readonly MyInnerClass[] myArray = new MyInnerClass[10];

    public static bool IsEmpty()
    {
        foreach (var item in myArray)
        {
            if ((item != null) && (!string.IsNullOrEmpty(item.Field)))
            {
                return false;
            }
        }

        return true;
    }

    private class MyInnerClass
    {
        public string Field;
    }
}

If I want to write a test for the case when the array has “non empty” entries, I need to setup the array first.

Using the accessors generated by Visual Studio, I would write something like this:

[TestClass()]
public class MyClassTest
{
    [TestMethod()]
    public void IsEmpty_NotEmpty_ReturnsFalse()
    {
        for (int i = 0; i < 10; i++)
        {
            MyClass_Accessor.myArray[i] = new MyClass_Accessor.MyInnerClass { Field = i.ToString() };
        }

        bool expected = false;
        bool actual;

        actual = MyClass.IsEmpty();

        Assert.AreEqual(expected, actual);
    }
}

But the test will fail because, although the elements of the private array myArray can be read as MyClass_Accessor.MyInnerClass instances, they can’t be written as such.

To do so, the test would have to be written like this:

[TestClass()]
public class MyClassTest
{
    [TestMethod()]
    public void IsEmpty_NotEmpty_ReturnsFalse()
    {
        for (int i = 0; i < 10; i++)
        {
            MyClass_Accessor.ShadowedType.SetStaticArrayElement("myArray", new MyClass_Accessor.MyInnerClass { Field = i.ToString() }.Target, i);
        }

        bool expected = false;
        bool actual;

        actual = MyClass.IsEmpty();

        Assert.AreEqual(expected, actual);
    }
}

But, this way, we loose all the strong typing of the accessors because we need to write the name of the array field.

Because the accessor for the field is a property, we could write a set of extension methods that take care of getting the field name for us. Something like this:

public static class PrivateypeExtensions
{
    public static void SetStaticArrayElement<T>(this PrivateType self, Expression<Func<T[]>> expression, T value, params int[] indices)
    {
        object elementValue = (value is BaseShadow) ? (value as BaseShadow).Target : value;

        self.SetStaticArrayElement(
            ((PropertyInfo)((MemberExpression)(expression.Body)).Member).Name,
            elementValue,
            indices);
    }

    public static void SetStaticArrayElement<T>(this PrivateType self, Expression<Func<T[]>> expression, BindingFlags invokeAttr, T value, params int[] indices)
    {
        object elementValue = (value is BaseShadow) ? (value as BaseShadow).Target : value;

        self.SetStaticArrayElement(
            ((PropertyInfo)((MemberExpression)(expression.Body)).Member).Name,
            invokeAttr,
            elementValue,
            indices);
    }
}

Now, we can write the test like this:

[TestClass()]
public class MyClassTest
{
    [TestMethod()]
    public void IsEmpty_NotEmpty_ReturnsFalse()
    {
        for (int i = 0; i < 10; i++)
        {
            MyClass_Accessor.ShadowedType.SetStaticArrayElement(() => MyClass_Accessor.myArray, new MyClass_Accessor.MyInnerClass { Field = i.ToString() }, i);
        }

        bool expected = false;
        bool actual;

        actual = MyClass.IsEmpty();

        Assert.AreEqual(expected, actual);
    }
}

It’s not the same as the first form, but it’s strongly typed and we’ll get a compiler error instead of a test run error if we change the name of the myArray field.

You can find this and other tools on the PauloMorgado.TestTools on CodePlex.

Extended WebBrowser Control - Version 0.0.0.0 Uploaded

After a long time, I finally managed to upload a version of the Extended WebBrowser Control to CodePlex.

It's still a work in progress, but it's usable. Feel free to download, comment and file issues. A nice tabbed browser demo is included.

Compiling .NET 1.1 Projects In Visual Studio 2008

After having put my .NET 1.1 application running on the .NET 2.0 runtime (^), I’m planning on migrating it to .NET 2.0, but not all at once.

Because I don’t want to have 2 solutions (one on Visual Studio 2003 for the .NET 1.1 assemblies and another on Visual Studio 2008 for the .NET 2.0 assemblies) I decide to try using MSBee and have only one Visual Studio 2008 solution.

MSBee has a CodePlex project. You can download it from there or from Microsoft Downloads. Because the build on Microsoft Downloads seemed to be the most stable one, that was the one I downloaded and installed. The package comes with a Word document that explains all that needs to be done.

Before you can install and use MSBee you’ll need to install the .NET 1.1 SDK.

Having everything installed, I just opened the Visual Studio 2003 solution in Visual Studio 2008 and let it convert the solution and projects (near 30).

After the conversion, for building the projects with the .NET 1.1 C# compiler, the project files need to be edited to add the override the default targets with the MSBee ones by adding the MSBee imports after the default imports for the language:

<Import Project="$(MSBuildBinPath)\Microsoft.CSharp.targets" />
<Import Project="$(MSBuildExtensionsPath)\MSBee\MSBuildExtras.FX1_1.CSharp.targets" />

Another change needed (for Visual Studio 2008 - I don't know if it was needed for Visual Studio 2005) is the tools version. MSBee needs version 2.0. To change that you'll have to change the ToolsVersion attribute of the project’s root element:

<Project DefaultTargets="Build" ToolsVersion="2.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">

MSBee likes has own idea about output paths and I had set up custom output paths on my project. There’s information about this on the documentation but I decided to simply comment that out of the $(MSBuildExtensionsPath)\MSBee\MSBuildExtras.FX1_1.Common.targets file:

<!-- Paulo
<Choose>
  <When Condition=" '$(BaseFX1_1OutputPath)' == '' ">
    <PropertyGroup>
      <OutputPath>bin\FX1_1\</OutputPath>
    </PropertyGroup>
  </When>
  <Otherwise>
    <PropertyGroup>
      <OutputPath>$(BaseFX1_1OutputPath)</OutputPath>
      <OutputPath Condition=" !HasTrailingSlash('$(OutputPath)') ">$(OutputPath)\</OutputPath>
    </PropertyGroup>
  </Otherwise>
</Choose>
-->

<!-- Paulo
<PropertyGroup>
  <BaseIntermediateOutputPath>obj\FX1_1\</BaseIntermediateOutputPath>
  <IntermediateOutputPath Condition=" '$(PlatformName)' == 'AnyCPU' ">$(BaseIntermediateOutputPath)$(Configuration)\</IntermediateOutputPath>
  <IntermediateOutputPath Condition=" '$(PlatformName)' != 'AnyCPU' ">$(BaseIntermediateOutputPath)$(PlatformName)\$(Configuration)\</IntermediateOutputPath>

  <OutputPath Condition=" '$(PlatformName)' == 'AnyCPU' ">$(OutputPath)$(Configuration)\</OutputPath>
  <OutputPath Condition=" '$(PlatformName)' != 'AnyCPU' ">$(OutputPath)$(PlatformName)\$(Configuration)\</OutputPath>
  
  <- Once OutputPath is determined, set OutDir to its value. ->
  <OutDir>$(OutputPath)</OutDir>
</PropertyGroup>
-->

This all seemed to work fine on my old Windows XP machine without any third party Visual Studio plug-ins, but when I tried it on my Windows Vista X64 machine, I came across some problems:

• License Compiler

  Because I'm using Infragistics' controls, there's a licences.licx file and the build will compile it. And that proved to be a problem.

  MSBee copies all the files it needs to the build process to a temporary folder, builds it in there and then copies the result to the output path.

  LC.exe seemed to never be able to find all the assemblies it needed. Searching seemed to me to be an old issue (even from the .NET 1.1 times) and the solution always pointed to not compile the license file. So, I commented that part out of the $(MSBuildExtensionsPath)\MSBee\MSBuildExtras.FX1_1.Common.targets file:

  <Target
      Name="CompileLicxFiles"  Condition="'@(_LicxFile)'!=''"
      DependsOnTargets="$(CompileLicxFilesDependsOn)"
      Inputs="$(MSBuildAllProjects);@(_LicxFile);@(ReferencePath);@(ReferenceDependencyPaths)"
      Outputs="$(IntermediateOutputPath)$(TargetFileName).licenses">
  
    <!--
    <LC
        Sources="@(_LicxFile)"
        LicenseTarget="$(TargetFileName)"
        OutputDirectory="$(IntermediateOutputPath)"
        OutputLicense="$(IntermediateOutputPath)$(TargetFileName).licenses"
        ReferencedAssemblies="@(ReferencePath);@(ReferenceDependencyPaths)"
        ToolPath="$(TargetFrameworkSDKDirectory)bin\">
  
      <Output TaskParameter="OutputLicense" ItemName="CompiledLicenseFile"/>
      <Output TaskParameter="OutputLicense" ItemName="FileWrites"/>
  
    </LC>
    -->
  </Target>

• Resource Generator

  Although this worked fine on the command line, inside Visual Studio ResGen.exe would throw some error and needed to be closed.

  Looking at the Windows Application Log I found out this:

  Faulting application Resgen.exe, version 1.1.4322.573, time stamp 0x3e559b5f, faulting module MockWeaver.dll, version 0.0.0.0, time stamp 0x4adb072e, exception code 0xc0000005, fault offset 0x00018fac, process id 0x4a50, application start time 0x01ca53c14488a2fb.

  MockWeaver.dll belongs to Isolator and I just disable it when building inside Visual Studio. I was hoping to start using Isolator on this project, but, for now, I can't.

I hope this can be of some help and, if you need more, you’ll probably find it at the MSBee’s CodePlex forum.

The bottom line is: You don’t need Visual Studio 2003!

Running .NET 1.1 Applications On .NET 2.0

One of the applications I develop is a .NET 1.1 Windows Forms application used by more than 5000 users and critical for the business.

Being a complex and critical application, porting it to the 2.0 runtime just because it was not an option because it would mean installing the new runtime and framework on the stable environment of the workstations (Windows XP) and test fully the application. That was not an option.

As time went by, a developer received a brand new laptop with Windows Vista. Since he only needed ,NET 2.0 for his developments, he never installed .NET 1.1.

Another developer on my team had already tried to port the application to .NET 2.0 and run into some issues:

• The main component of this application is the Web Browser Control. This control derives from AxHost, which changed going from 1.1 to 2.0 and needed major changes to compile for the 2.0 framework.
• Another change was that mixing synchronous and asynchronous calls is not allowed in the 2.0 framework and we had, at least, one of those in our use of HttpWebRequest/HttpWebResponse.

The .NET 2.0 runtime and frameworks were developed to be highly compatible with applications written and compiled to the 1.1 runtime and frameworks. In fact, some of the changes were just applying the ObsoleteAttribute set to throw a compiler error when used, which doesn’t prevent its use by already compiled assemblies. This was the case of the WebBrowserControl/AxHost and just using the assembly compiled for .NET 1.1 would probably run fine. And it did.

The synchronous/asynchronous was also very easy to fix. All it took was changing this:

request.GetRequestStream()

into this:

request.EndGetRequestStream(response.BeginGetRequestStream(null, null))

And it all worked as if it was running on .NET 1.1.

But that’s not the end of it. Latter came a requirement to use, in one of the web pages that ran in the web browser control, an ActiveX component developed in .NET 2.0.

But that time, the 2.0 runtime and framework were already installed on the workstations.

But how would we force the application to run in the 2.0 runtime if the 1.1 runtime was still there?

As simple as adding this to the configuration file (App.config):

<configuration>
  <startup>
    <requiredRuntime version="v2.0.50727" safemode="true"/>
  </startup>
</configuration>

LINQ To SQL Tips & Tricks: String Operations

LINQ brought developers a very user friendly and domain independent style of writing queries.

The fact that the way queries are written is domain independent doesn’t mean that any query will compile the same way or even run the same way. You’ll always need to know how the provider will behave.

LINQ To Objects, for example, will compile queries as a Func<> delegate and the query methods will return IEnumerable(T) implementations.

On the other hand, LINQ To SQL will compile queries as an Expression<Func<>> (which is, in fact, an expression tree) instance and the query methods will return IQueryable(T) implementations.

Because LINQ To SQL queries are compiled to an expression tree, that allows the provider to treat the query elements as it sees fit.

In this case, this means that all operations that can be done on the database will be done on the database and the developer must be aware of this when she/he is writing the queries.

Lets take an example using the AdventureWorks database (if you don’t have it, you can download it from here).

I want to build a list of salutation for every employee that has the SalariedFlag set, in the form of:

[Mr.|Mrs.|Miss] <first name> <middle name> <last name>

But there’s also one detail about the data in the database: FirstName, MiddleName and LastName may have trailing spaces and I don’t want them.

This is a simple query like this:

var q1 = from e in context.Employees
         where e.SalariedFlag
         select
            ((e.Gender == 'F') ? ((e.MaritalStatus == 'S') ? "Miss" : "Mrs.") : "Mr.") + " " +
            e.Person.FirstName.Trim() +
            (e.Person.MiddleName == null || e.Person.MiddleName.Trim().Length == 0 ? " " : " " + e.Person.MiddleName.Trim() + " ") +
            e.Person.LastName.Trim();

and it will be executed as:

SELECT ((((
    (CASE
        WHEN UNICODE([t0].[Gender]) = @p0 THEN
            (CASE
                WHEN UNICODE([t0].[MaritalStatus]) = @p1 THEN @p2
                ELSE @p3
             END)
        ELSE CONVERT(NVarChar(4),@p4)
     END)) + @p5) + LTRIM(RTRIM([t1].[FirstName]))) + (
    (CASE
        WHEN ([t1].[MiddleName] IS NULL) OR (LEN(LTRIM(RTRIM([t1].[MiddleName]))) = @p6) THEN CONVERT(NVarChar(MAX),@p7)
        ELSE (@p8 + LTRIM(RTRIM([t1].[MiddleName]))) + @p9
     END))) + LTRIM(RTRIM([t1].[LastName])) AS [value]
FROM [HumanResources].[Employee] AS [t0]
INNER JOIN [Person].[Person] AS [t1] ON [t1].[BusinessEntityID] = [t0].[BusinessEntityID]
WHERE [t0].[SalariedFlag] = 1
-- @p0: Input Int (Size = 0; Prec = 0; Scale = 0) [70]
-- @p1: Input Int (Size = 0; Prec = 0; Scale = 0) [83]
-- @p2: Input NVarChar (Size = 4; Prec = 0; Scale = 0) [Miss]
-- @p3: Input NVarChar (Size = 4; Prec = 0; Scale = 0) [Mrs.]
-- @p4: Input NVarChar (Size = 3; Prec = 0; Scale = 0) [Mr.]
-- @p5: Input NVarChar (Size = 1; Prec = 0; Scale = 0) [ ]
-- @p6: Input Int (Size = 0; Prec = 0; Scale = 0) [0]
-- @p7: Input NVarChar (Size = 1; Prec = 0; Scale = 0) [ ]
-- @p8: Input NVarChar (Size = 1; Prec = 0; Scale = 0) [ ]
-- @p9: Input NVarChar (Size = 1; Prec = 0; Scale = 0) [ ]
-- Context: SqlProvider(Sql2008) Model: AttributedMetaModel Build: 3.5.30729.4926

If you notice the query, there are a lot of text operations going on for each row.

Depending on the number of rows or database load this can prove to be very bad. The result might even be just a timeout.

So, how do we force the string operations to occur on the client instead of the database?

Only IQueryable<T> will be translated to T-SQL. So, all we need to do is change the type of the enumerator being iterated.

One way to do this is using the the AsEnumerable method of the Enumerable class.

The query would now be written as:

var q2 = from e in context.Employees.Where(e => e.SalariedFlag).AsEnumerable()
         select
            ((e.Gender == 'F') ? ((e.MaritalStatus == 'S') ? "Miss" : "Mrs.") : "Mr.") + " " + e.Person.FirstName.Trim() +
            (e.Person.MiddleName == null || e.Person.MiddleName.Trim().Length == 0 ? " " : " " + e.Person.MiddleName.Trim() + " ") +
            e.Person.LastName.Trim();

and it will be executed as:

SELECT
    [t0].[BusinessEntityID],
    [t0].[LoginID],
    [t0].[NationalIDNumber],
    [t0].[JobTitle],
    [t0].[MaritalStatus],
    [t0].[BirthDate],
    [t0].[Gender],
    [t0].[HireDate],
    [t0].[SalariedFlag],
    [t0].[VacationHours],
    [t0].[SickLeaveHours],
    [t0].[CurrentFlag],
    [t0].[rowguid],
    [t0].[ModifiedDate],
    [t1].[BusinessEntityID] AS [BusinessEntityID2],
    [t1].[PersonType],
    [t1].[NameStyle],
    [t1].[Title],
    [t1].[FirstName],
    [t1].[MiddleName],
    [t1].[LastName],
    [t1].[Suffix],
    [t1].[EmailPromotion],
    [t1].[AdditionalContactInfo],
    [t1].[Demographics],
    [t1].[rowguid] AS [rowguid2],
    [t1].[ModifiedDate] AS [ModifiedDate2]
FROM [HumanResources].[Employee] AS [t0]
INNER JOIN [Person].[Person] AS [t1] ON [t1].[BusinessEntityID] = [t0].[BusinessEntityID]
WHERE [t0].[SalariedFlag] = 1
-- Context: SqlProvider(Sql2008) Model: AttributedMetaModel Build: 3.5.30729.4926

As you can notice, text operations are no longer done on the database, but all the columns of both tables are being returned. And this is still a bad thing because we are using network bandwidth with data that won’t be used.

The way to choose the columns that will be retrieved from the database is by selecting only the ones wanted in the select statement. But because we still want string operations the be done on the client, we’ll need to project the desired columns into an intermediary object. Since we won’t need this object outside the query, we’ll use an anonymous type.

The query would now be written as:

var q3 = from n in
             (
                 from e in context.Employees
                 where e.SalariedFlag
                 select new
                 {
                     Gender = e.Gender,
                     MaritalStatus = e.MaritalStatus,
                     FirstName = e.Person.FirstName,
                     MiddleName = e.Person.MiddleName,
                     LastName = e.Person.LastName
                 }
             ).AsEnumerable()
         select ((n.Gender == 'F') ? ((n.MaritalStatus == 'S') ? "Miss" : "Mrs.") : "Mr.") + " " + n.FirstName.Trim()
         + (n.MiddleName == null || n.MiddleName.Trim().Length == 0 ? " " : " " + n.MiddleName.Trim() + " ")
         + n.LastName.Trim();

and it will be executed as:

SELECT
    [t0].[Gender],
    [t0].[MaritalStatus],
    [t1].[FirstName],
    [t1].[MiddleName],
    [t1].[LastName]
FROM [HumanResources].[Employee] AS [t0]
INNER JOIN [Person].[Person] AS [t1] ON [t1].[BusinessEntityID] = [t0].[BusinessEntityID]
WHERE [t0].[SalariedFlag] = 1
-- Context: SqlProvider(Sql2008) Model: AttributedMetaModel Build: 3.5.30729.4926

Notice the call to Enumerable.As Enumerable to translate the LINQ To SQL query into a LINQ To Objects query.

And, to end this long blog post, if you don’t use any string operations on the query, they, obviously, won’t be translated to T-SQL:

var q4 = from e in context.Employees
         where e.SalariedFlag
         select BuildSalutation(e.Gender, e.MaritalStatus, e.Person.FirstName, e.Person.MiddleName, e.Person.LastName);

where BuildSalutation is implemented as:

private static object BuildSalutation(char gender, char maritalStatus, string firstName, string middleName, string lastName)
{
    return ((gender == 'F') ? ((maritalStatus == 'S') ? "Miss" : "Mrs.") : "Mr.") + " "
        + firstName.Trim()
        + (middleName == null || middleName.Trim().Length == 0 ? " " : " " + middleName.Trim() + " ")
        + lastName.Trim();
}

and it will be executed as:

SELECT
    [t0].[Gender] AS [gender],
    [t0].[MaritalStatus] AS [maritalStatus],
    [t1].[FirstName] AS [firstName],
    [t1].[MiddleName] AS [middleName],
    [t1].[LastName] AS [lastName]
FROM [HumanResources].[Employee] AS [t0]
INNER JOIN [Person].[Person] AS [t1] ON [t1].[BusinessEntityID] = [t0].[BusinessEntityID]
WHERE [t0].[SalariedFlag] = 1
-- Context: SqlProvider(Sql2008) Model: AttributedMetaModel Build: 3.5.30729.4926

Have you noticed that this T-SQL query is pretty much the same in the previous example?

If you are still reading this, I hope you now aware of how you write your LINQ To SQL queries affect the generated T-SQL.

PowerShell For The .NET Developer

Some time ago I needed to have the validationKey of the machineKey element of an ASP.NET application changed and found out that ASP.NET doesn’t provide a command-line tool (or any other) to do this.

Looking around I found several applications and code samples to do it, but to have a system administrator do this I needed to test and document the application and it was to much work for such task.

I’ve always been a supporter of the idea of PowerShell but I never used it my self. Just because I almost always have Visual Studio open and writing a simple console application is quicker and easier than learning PowerShell.

This time I decide that I would do a PowerShell script instead.

In C# I would have done something like this:

class Program
{
    private static string GenerateKey()
    {
        var buff = new byte[64];
        (new System.Security.Cryptography.RNGCryptoServiceProvider()).GetBytes(buff);
        var sb = new System.Text.StringBuilder();
        foreach (var b in buff)
        {
            sb.AppendFormat("{0:X2}", b);
        }
        return sb.ToString();
    }

    private static void Main(string[] args)
    {
        var path = args[0];
        var config = System.Web.Configuration.WebConfigurationManager.OpenMachineConfiguration(path);
        var systemWeb = config.GetSectionGroup("system.web") as System.Web.Configuration.SystemWebSectionGroup;
        var machineKey = systemWeb.MachineKey;
        machineKey.ValidationKey = GenerateKey();
        config.Save(System.Configuration.ConfigurationSaveMode.Modified);
    }
}

How would it be in PowerShell? As simple as this:

function GenerateKey
{
    [System.Byte[]]$buff = 0..63
    (new-object System.Security.Cryptography.RNGCryptoServiceProvider).GetBytes($buff)
    $sb = new-object System.Text.StringBuilder(128)
    for($i = 0; ($i -lt $buff.Length); $i++)
    {
        $sb = $sb.AppendFormat("{0:X2}", $buff[$i])
    }
    return $sb.ToString()
}

[System.Reflection.Assembly]::LoadWithPartialName("System.Web")
$config = [System.Web.Configuration.WebConfigurationManager]::OpenWebConfiguration("<path>")
$systemWeb = $config.GetSectionGroup("system.web");
$machineKey = $systemWeb.MachineKey
$machineKey.ValidationKey=GenerateKey
$config.save("Modified")

Wonder how I got from no knowledge of PowerShell to this? Simple. Something that every real .NET developer has and loves: .NET Reflector (with a PowerShell add-in, of course).