Category Archives: Open Source

Unit testing Entity Framework Core–Mocking DBSet Async methods

While working on a new project I decided to use Entity Framework core as the project had a very simple setup and didn’t needed to complex Database settings or queries. A perfect use case to try out EF core.

Implementation was just a breeze while making use of the Repository pattern and the Unit Of Work pattern. Although the application hasn’t have a load of business logic. still I wanted to add Unit tests to validate the implemented logic.

In previous projects we used Moq to create mocks to support the NUnit test framework. For this project I decided to have a look at the Fake It Easy framework. The website promises:

  • Easier semantics: all fake objects are just that — fakes. Usage determines whether they’re mocks or stubs.
  • Context-aware fluent interface guides the developer
  • Easy to use and compatible with both C# and VB.Net.

And I must say, I had to dive in the documentation a few times but for the rest, it was quite simple to set up the mocks and implement the Unit tests.

Mocking DBSet

With the experience of previous projects I knew that mocking a DBSet needed some extra setup. As long as you don’t have to mock queries on the DBSet you’re fine with the default implementation but not if you want to test async queries asFirstOrDefaultAsync(), SingleOrDefaultAsync(). Those are only supported when running against an EF query. For EF 6 there is a very well worked out example (Testing with a mocking framework) on Microsoft Docs.

For EF core there isn’t the same documentation available on the docs pages. Instead there is a page about Testing with InMemory. As already stated by Thomas Galliker in the comments ( In Memory testing doesn’t cover the whole test process.

My first attempt was to use the demo code available for EF core to create the mock and try out if it would work for EF core. But instead of implementing Moq, I’ve used Fake It Easy. First you have to create a TestDbAsyncQueryProvider. I’ve copied the code from the demo on the docs page and noticed the IDbAsyncQueryProvider doesn’t exist any more. After some searching I found that it was renamed to IAsyncQueryProvider. I had to update a few implementations but got it building. The full implementation of the class:

But you also have to make some changes to setup of your mock of the DBSet:

If you now run your tests you will see they run successfully. If you are getting an error like below, make sure your set the mock of DBSet to implement IAsyncEnumerable<TEntity>

The source IQueryable doesn't implement IAsyncEnumerable{0}. Only sources that implement IAsyncEnumerable can be used for Entity Framework asynchronous operations.

Generic solution

Off course you don’t want to repeat all this setup every time you have to write tests for a repository. The solution is to create a helper class where you can define the Entity (and the context if you like) like I did in the class below:

Create Nuget package to reuse in other projects

As we are going to use EF Core in future projects, it would be a timesaver to add the generic implementation in a nuget package to reuse in the Unit tests. As I was looking into creating a package to publish on Nuget I found already a generic solution created by Roman Titov called MockQueryable that was published on Nuget.

Roman already added implementations for Moq and NSubstitute but didn’t had an implementation yet for Fake It Easy. I could off course just used his core package and create my own but instead I cloned his repo and added the implementation for Fake It Easy.

So know you can just use: (

Install-Package MockQueryable.FakeItEasy -Version 1.1.0

and get started right away on your Unit tests.

All source code from this post can  be find on Github:

Personal DEVIntersections review

It’s a fast moving world for software developers. New frameworks, functions, possibilities, are rising up almost every day. Although you can find many resources online in different formats (blogs, videos, live coding, …) it’s always an advantage to attend a conference where you can learn the new stuff and meet up with some of the leaders of the industry.

Until 2014 Microsoft organised an European Teched conference but decided to stop with the Teched at all (also the North America edition) and focus on their BUILD conference. Due to a lack of an European alternative we tried to get tickets for the BUILD conference in San Francisco this year but you have better luck trying to buy U2 tickets for a small venue. Luckily we noticed the DEVIntersections conference in Orlando that takes place only a 3 weeks after the BUILD conference and  with an impressing line up of speakers.

Now the last day of the conference is arrived, is it time for a little wrap up. Please note, these are my personal findings not only about the technical items from the conference but also travel, stay, hotel, …

Long way from home

Not really a downside, but it’s reality that the travel from Belgium to Orlando takes a while. It’s not my first visit to the States so I did know what to expect. Due to the terrorist attacks on march 22 in Brussels our flight was diverted to Schiphol what led to some extra travelling. From the moment I’ve parked my car in Berchem till we entered the hotel a 20 hour trip was finished and has taken us from a train ride to Schiphol, an 8 hour flight to Atlanta, a 2 hour flight to Orlando and an 45 min taxi ride to the hotel.

Disney World

The conference takes place in one of the Disney World hotels, the Walt Disney World Swann and Dolphin. And the hotel is huge! You can’t compare with any hotel you find in Belgium (or the parts of Europe where I travelled before) and due to the Disney park, it houses a mix of different conference attendees (and yeah, you can pick the IT dudes out of the crowd Glimlach) and families ready for some days of Disney fun.

The room was fairly standard but clean and comfortable. A bit a pity of the view as we’re just above a roof of one of the adjusting buildings of the hotel. But once you step outside the room and start walking around you are directly in vacation atmosphere. You have the 5 different swimming pools, different hot tubs, 7 or 8 different restaurants in the hotel complex, the Disney Broadwalk around an artificial lake, boat trips from one side to the other, …

Fun for the first days, but after a while you have seen it all. Plans to go to downtown Orlando, we didn’t travel that far for just the Disney magic, weren’t realistic by the lack of fast public transportation and the cost for a taxi fair in combination with the time we could spend downtown after a day of conference.

DEVIntersections – the workshops

Next to the conference itself, they also organised 4 days of workshops and hands on labs the days before and after, where we registered for the 2 days pre work shops that are titled: ‘Making the Jump to ES6 and TypeScript’ and ‘Building Single Page Applications with Angular 2’.

I’ve only been looking into Angular (1.4) and Typescript shortly so the workshop was very interesting. Learned a lot about Typescript and how it can construct and organize your codebase on a much better way with a lot of advantages. Certain something to look further into.

After the first day of Typescript introduction, you notice on the second day how it perfects integrates with Angular 2! A lot of new stuff was thrown at us and it will take me some time to digest all that information but it’s a very good base to start building Angular 2 apps in the future.

Both work shops were presented by John Papa (John_Papa) and Dan Wahlin (DanWahlin), both experts in these matters. You could notice they worked together before as they seamlessly took over from each other and were joking around without falling into a comedy show. 

Although it was announced as a “hands on” workshop, the “hands on” moments  where few and maybe there should have been more time for the attendees for trail and error on their own machines, especially for an all day workshop.

Note to the organisation itself, if you ask attendees to bring there own device, please make sure there are power sockets for those people. More then half of the attendees ran out of battery before lunch and were unable to use their machines for the rest of the workshop. We were so wise to come early the second day and choose one of the few places where there was a power socket in reach.

DEVIntersections – the sessions

The conference content was of a very good level! Most of the speakers are the experts in their field and many Microsoft employees stood on stage. You could notice there were a lot more attendees then the first days during the workshops but still it wasn’t too busy. Not too much queues (except at the men’s rooms during breaks Glimlach, the ladies still have a huge advantage on these IT events).

The first keynote was by Scott Guthrie (scottgu) and was a general overview about Azure and could be in my opinion a lot wider and a bit more developer focussed instead of the commercial tone.  The videos shown were the same as shown on the build conference but that was to be expected.

The second keynote on the first day was from Scott Hanselman (shanselman). You can’t always predict what he’s going to talk about, but you can at least be sure it’s a good mix of technical innovations an a lot of humour. I’ve also went to the other session that was presented by him and on both occasions it was top notch entertainment with a lot of new exiting things that are coming our way. He’s not afraid to point out where there are still some improvement opportunities for Microsoft but is straight on about where the the different teams of Microsoft are focussing on at the moment. He managed to install the latest development version of Visual Studio on one of the attendees his laptop, who had surprisingly no Visual Studio instance installed??? While trying to connect the laptop, bringing in the technical fellows, he managed to bring that without boring for 1 second.

The last day of the conference was lacking good interesting sessions for me what gave the feeling all was said during the first days. Except some 3rd party frameworks like NodeJS and ReactJS there were not many full developer sessions that were based on Microsoft technology except then the keynote about Sharepoint. (what gave me the time to write this post Glimlach).

But overall good content, excellent speakers and we picked up a lot of new things. If I find the time, I make a follow up post.

DEVIntersections – social

Maybe I’m a bit spoiled by the Tech Ed conferences and the Xamarin Evolve conference I attended in November 2014 but I missed the social parts that makes a good conference a wow conference.

For one, the sponsors expo was limited. With 10 company stands there was not so much to see or to speak about. The boots were small as the expo took place in the hallway between the different rooms.

Ok, a bit childish but there were not so many goodies to find. On previous occasions I’ve been able to make my son very happy with a bunch of goodies that I’ve taken with me from the exhibitors expo. 

There were 2 evening gatherings (if I didn’t miss one, but saw no other in the schedule or announcements). On Tuesday the opening of the partner expo with drinks and snacks. It’s a pity they just gave one coupon for a drink and not so many snacks. Because there was only one drink included, the reception was over quite early as everybody went to somewhere else.

I didn’t went to the after dark sessions on Wednesday but the general feedback from the people I spoke, wasn’t that overwhelming and many bailed out during the session.

As other conferences were held in the same hotel complex, you could see the difference. We’ve saw at least 3 beach parties with open bar, music, a lot of people and seemed to be a good atmosphere. It would have been great if DEVIntersections had organised one of those.

DEVIntersections – General

Overall it is a good conference with some grow opportunities on their way. Off course Rome wasn’t build in one day, and DEVIntersections is a very young conference. I’ve enjoyed my stay in Florida and learned a lot of new interesting stuff. I’m a big fan on how Microsoft is proceeding and it will be challenging times for aswell Microsoft as the developers using their stack to create excellent apps on different platfoms, mobile, web, desktop, cloud, IoT, …

Searching with a Lucene.NET wrapper

For a project we’re currently working on we needed to be able to search on different fields that are shown to the end user. We didn’t want to rely fully on the SQL server Full Text Search capabilities. Luckily I knew the Lucene.NET engine from some previous work with the Umbraco CMS.

We just wanted to receive from our search engine the ID’s of the objects where the search term in was found rather then complete objects. When working with a database it just doesn’t feel right to store all your data on 2 different places (the database and the search index store). We only wanted to add the fields where we would search on to the search index and let Lucene decide the match, afterwards we’ll pull the objects from the database using Entity Framework.

Setting up Lucene.NET

Setting up Lucene.Net is not that difficult and you can find a lot of blog posts on how to use Lucene (like Introducing Lucene.NET on Code Project) and with a Nuget package available it’s easy to add the engine to your project.

I’ve followed a few of those blog posts I could find. Off course these posts give you a starting point and for the simplicity all code is written in one or two classes. After the first implementation I started with some refactoring’s and decided it would be easier for future implementations if I created my own search DLL or package.

At the bottom you’ll find the link to the Github repository where you can browse and download the package.


If you want to be able to search in Lucene you’ll have to add the parameter to a Lucene document. To simplify the creation of such documents I’ve created an abstract class ADocument where you have to inherit from.

You’ll see that I’ve added the public property “Id” so that every class that will inherit should have that Id. In the setter of the ID you’ll see I add the field to the Lucene.Net Document with the AddParameterToDocument method.

Next to the private AddParameterToDocument method you’ll find two other methods, one with the possibility to store the parameter in the Lucene index and one to just analyze the parameter but not to store it in the Lucene index.


You’ll see in the ADocument class that I’ve decorated the “Id” with the “SearchField” attribute. This attribute was created to simplify the search on multiple fields that you’ll see when implementing the BaseSearcher. You can find the implementation of this attribute below.


Both the writer as the searcher have to have access to the index that is written to the local file system. To avoid multiple implementations (check if the directory exists, load a FSDirectory object, …) I’ve created a BaseSearch class. Not much to see in this class, just some basic settings of the folder where the index is stored.


before we can search we’ll have to add some items into the search index off course. Following the DRY principle we’ve created a new abstract base class, the “BaseWriter”. In this class we added the methods to add and update new or existing item(s) of the type of ADocument. Next to adding and updating we added the corresponding delete methods. Because we delete everything based on the Id property of the ADocument implementation it’s important its value is always set in the deriving class.

NOTE: the log messages are added to be used with Log4Net.


Adding items to the Lucene index wasn’t that hard. When I first started to implement the search methods I wanted to be able to search on a specific property or on all properties available on that object.

To be able to search on all properties you’ll have to user a MultifieldQueryParser instead of the default QueryParser. But with the MultifieldQueryParser you have to enter all the fields where Lucene have to search on. I really didn’t wanted to have to implement the same search code for every type we have to add to the index.

We’ll no choice then to turn to reflection to fetch all the parameters. But the class could have more properties then the one we are searching on. To avoid these extra properties (and the errors because we’ll be searching on properties that are not indexed) I’ve added the SearchField attribute.

So the first thing we’ll do when searching is to fetch all properties from the class and add them in a list. By using the T parameter we can easily reuse the same method for different classes as long they inherit from ADocument.

With this list we can now implement the actual searching. If we search on a specific field we’ll use the default QueryParser and if we are searching on more then one field, the MultifieldQueryParser.

When you look into the code, you’ll see that even when searching on a specific field, there’s still a possibility that the MultiFieldQueryParser is used. I’ve added the option to search on multiple parameters in the Lucene index if they are related to each other.

For example: you have a registration number for each person that always starts with the year of registration, a number (sequence) and a suffix: 2014-0023-aaa. When you add this complete registration number to the index and you search on 2014 (without wildcards) the Lucene engine will return no results. To avoid that the end user have to use wildcards you can store the registration number in three different parts. But when you want to search on the field RegistrationNumber you’ll have to indicate that multiple fields have to be used.

Therefor can the SearchField parameter contain a array of strings that contain the other searchfields that have to be taken into account. (confused, see the TestApp project on github)


The search term that the end user enters has to be translated to a Lucene.NET Query. There a re build in methods to convert a string to a Query object. Although the method can throw a ParseException if invalid characters are used. Therefor I added a private ParseQuery method to catch those exceptions and to filter out the invalid characters.


Because we only want to return the Ids of the objects we’re searching for we can use a generic SearchResult class to return the results. The amount of hits and the search term are added to the result to be shown in the UI.

Test application

The above classes (except a custom exception) are all the parts we need to start testing our search wrapper. On the Github repository you’ll find a TestApp project in the solution where the classes below are implemented.


The example used is of a Person that will register for a service. For simplicity sake local classes are used as repository instead of a database but you’ll get the point.

We”’ll start with the Person class that is used in some application. All default stuff and an override for the ToString method to print out Person class.


It are these Person classes we want to add to our search index. To add these we have to create a PersonDocument class that implements from our ADocument abstract class. We”’ll add all properties with private backing fields so we can call the ‘AddParameterToDocument’ methods. Next to adding the property to the index we’ll have to decorate the properties we want to search on. (See the RegistrationString property where we added multiple SearchFields)

At the bottom I added a operator method to cast the person object to a PersonDocument object so I don’t have to repeat the cast in the business logic.


The PersonWriter class will be not that difficult to implement now our PersonDocument class is defined. We just have to add the add and update methods and the delete methods that will call the base class methods.


Also the PersonSearcher class will not be that difficult with the Search method implemented in the BaseSeacher class.


In the program.cs file you’ll find the creation of the multiple Person objects an how they are added to the index. Underneath you’ll find the different search methods and their results.


Although Lucene.NET has far more options then showed here in this blog post, will the created wrapper at least give you the basic search possibilities. Off course can you extend the base classes to add more search options, index options etc.

With the basic settings in the Document class that inherits from ADocument you’ll avoid to create numerous searchers or indexers.

All source code can be found on GitHub and feel free to fork or download.

Create an installer for website with WIX, part 3

In part 1 we saw how we can use a MS Build script to harvest all the files that needs to be installed and to create our msi installer. In part 2 we saw how we can add our own custom UI.

With only the files installed you don’t have a website running. At least not an ASP.NET MVC website. You have to register the website in the Internet Information Services (IIS), create an app pool with the correct framework selected and activate (start) the website.

This post will handle that part of the creation of the installer where we extend the project we used in part 1 & part 2.

Add the IISConfiguration WiX file

We start by adding a new item in our setup project, choose for an installer file and name it IISConfiguration.wxs. The template will create an empty WiX page with just an empty “Fragment” element.

We’ll start with adding a component for the application pool.

We add a “DirectoryRef” element that we point to our installation folder so that this will be run when handling the installation directory in the installer.

In the “Component” element we add the “iis:WebAppPool” element. Give it an id and name, the name will be displayed in IIS. Choose the Identity that will start the application pool.

You can choose “networkService”, “localService”, “localSystem”,”applicationPoolIdentity” or “other”. With “other’” you’ll have to create a user with a username and password that you have to assign to the “iis:WebAppPool” element “users” parameter (not in this example).

NOTE: I had to install the 3.8 version of the WiX toolset to be able to assign the “applicationPoolIdentity”. In previous versions this was apparently not possible.

If you work in Visual Studio you’ll notice the editor adds an red line under the “iis:WebAppPool” element. We have to add the IISExtension namespace in the Wix declaration.

Add the install website component

Next we’ll add a second component to our file to create the website in IIS.

Here we’ll have to an unique id, a description of the website ‘(will be shown in IIS), the directory where the website is installed (our “INSTALLFOLDER”) and the auto start variables.

In the “iis:Website” element you can add one or more “iis:WebAddress” elements to assign DNS names and TCP ports to the website.

Last but not least, we’ll have to connect the correct application pool to the website by adding a “iis:WebApplication” with a unique id and name (I took the website id) and a reference to the “iis:WebAppPool” element we created before.

Create and reference component group

It’s best practice to group these two components in a component group so you only have to add one reference to iis configuration elements.

Add this “ComponentGroup” in the “Target” element.

In the product.wxs file we have to place a reference to this component group in the “Feature” tag.

Update the MS Build file

Because we created an extra file to include in the build process we must be sure that the file is added in the MS Build file.

We have to update the item groups that contain the list of files to include as shown above.

Now open up the Developer Command Prompt for Visual Studio 2012 again and change the prompt to the setup project folder and type the next statement:

msbuild /t: Build;PublishWebsite;Harvest;WIX;DeleteTmpFiles

and hit enter.

You’ll see the build process will return an error. We added a reference to the IISExtensions library in our IISConfiguration.wxs file but we didn’t told the candle.exe and light.exe tools to take that into account.

Add in both commands the –ext WixIISExtension flag and run the MS build script again. This time it shouldn’t be returning any errors.

Test the installer

Let’s run the installer and see if it really adds our website.

During the installation you’ll see a message passing by that it’s configuring IIS. Complete the installation and open up the IIS manager (run inetmgr).


We’ll see a new Demo Website website on the left hand side and a new application pool DemoWebsite on the right hand side running .NET 4.0 under the correct application pool identity.


If we open up the advanced settings of the website we’ll see the correct physical path is used and the binding is correct.

You may have noticed that our website isn’t started (the stop sign on the website icon). This isn’t a fault of the installer, it tried to start the website but because we choose port 80 in our installer IIS noticed there is another site running on that port (Default Web Site) and refuses to start our installed website.

If you change the port in the installer (or remove/stop the default webs site before installing) you’ll notice it starts upon install.

Next parts

This concludes the installation in IIS. Next part we’ll handle the creation of the database using the entered credentials and running SQL scripts from the installer.

  • Install the .NET 4.5 framework if that isn’t installed already
  • Install the MVC 4 framework if that isn’t installed already.
  • Create a folder and copy all needed files to run the application (done)
  • Create a new database on an existing SQL server and prefill the database with the correct tables and values. (the connection details and database name should be entered by the end user running the installer)
  • Create a new website in IIS 7.5 (create website and application pool running under .NET 4.5) (done)
  • Alter the config file so the correct connection settings are used (entered by the end user)

Complete source code

Can be found on Github.

Other posts in this series:

Create an installer for website with WIX, part 2

In part 1 we saw how we can use a MS Build script to harvest all the files that needs to be installed and to create our msi installer.

When you run the installer we created you saw it had a very minimalistic UI, better none UI at all. In part 2 we’re going to add a user interface for the installer. We’ll make use of the build in UI elements of WiX but also add a new dialog so the end user can enter the database connection values and credentials.

The WiX library has already a lot of build in UI dialogs and flows that you can use out of the box. If you download the source code from codeplex you can take a look at the standard UI. (unzip and then go to \src\ext\UIExtension\wixlib\)

The dialogs we wanted to show our end customers where:

  • Welcome screen
  • Let the user select the install directory (default c:\inetpub\)
  • Let the user fill out database connection values (database server IP or name, database name, database user end password)
  • Confirmation to install
  • Show some progress of the installation
  • Show a install succeeded screen

The WixUI_InstallDir default UI comes closest to what we want to implement. So we start from that default UI theme.

Create our own UI

Add a new file in our setup project and name it DemoUI.wxs.and copy the default WixUI_InstallDir file content into this file.

You can see the flow the UI will follow. First a welcome screen, then a license screen, then the installation directory screen, a verify ready dialog and a maintenance dialog. Except for the license screen this is exactly what we needed.

Our first job will be to remove the license screen from our UI flow. You can delete the lines 60 and 61 but we’ll have to take care of the references to the license screen.

On line 58 you’ll see that the welcome dialog next button will send you to the license screen, alter that so it will send you to the installation directory screen.

Same for line 63 where the back button will send you to the license screen. Update the Value tag so it references the welcome dialog instead.

Reference our own UI

We still have to tell our installer that he has to use our own UI instead of the default one. In our DemoUI.wxs change the Id of the UI element (line 33) from “WixUI_InstallDir” to “MyUi”.

In our Product.wxs we’ll have to reference our own UI by adding a “UIRef” element in the “Product” element with Id “MyUI”. (line 11)

And we have to tell the UI the default folder he has to use in the Install directory dialog by adding the “WIXUI_INSTALLDIR” property that we reference to our “INETPUB” property.

Update our MS build file

Now that we have a new source file (DemoUI.wxs) we have to update our MS Build file to include our user interface.

Open up the file and update the Itemgroups containing the WiX source files and WiX compiled files.

In our WIX target we’ll have to tell the light.exe tool that we are using parts of the WixUIExtension library.

And last but not least I added a new target to clean up all previous created and build files to avoid conflicts.

Now build the installer by opening a Developer Command Prompt for Visual Studio 2012, change the directory to the setup project and enter the next command:

msbuild /t: Build;PublishWebsite;Harvest;WIX;DeleteTmpFiles

and hit enter.

In the \bin\release folder of our setup project you’ll find our freshly build DemoWebsite_Setup.msi. Double-click the installer and you’ll see the welcome screen as intended.


If we click next the destination folder is prefilled with the c:\Inetpub folder but can be changed by the end user. Clicking next will show the “ready for installation” and the “installation succeeded” screens.

During installation we’ll see a progress bar showing the status of the installer. Running the installer will, just like in the last post, install the necessary files in the C:\Inetpub\DemoWebsite folder (if no other location is selected by you).

If you run the installer again you’ll get the default repair/remove interface.

Add the SQL connection screen

Now that we have a default UI for our installer we can add the dialog to let the end user enter the database connection values.

Add properties

The database connection values that will be entered by the end user will need to be reused in a later process of installation of the database. Therefor we have to store them in properties.

In our Product.wxs file we add four properties (DB_USER, DB_PASSWORD, DB_SERVER, DB_DATABASE) and give them some default values (line 7 to 10).

Add a new dialog

Now we can add the new dialog, add a new Installer file in our setup project and name it UIDialogs.wxs. I added the controls we need.

We’ll see a label and a text box for our 4 properties and a next and cancel button that we’ll link in our UI file.

Insert the dialog in the UI flow

Open up our DemoUI.wxs file again and add our new dialog between the “Install directory” and “verify ready” dialogs. Add 2 references, one for the next button and one for the back button.

The next button we reference to the “VerifyReadyDlg” and the back button to the “InstallDirDlg”.

We’ve got to alter the flow to insert our dialog by changing line 5 (reference the next button of “InstallDirDlg” to our one) and change line 12 (reference the back button of “VerifyReadyDlg” to our one).

Update our MS build file

We have to tell our script there is a new source file to handle, the dialog file. Open up the MS build script again and alter the ItemGroup settings by adding the new file.

Open your command prompt again and run the next statement:

msbuild /t: Build;PublishWebsite;Harvest;WIX;DeleteTmpFiles

and hit enter.

When our installer is build, run it and you’ll see the new dialog appear after the installation directory dialog.


Check the installer

You may see the properties now in the UI you want off course to be sure they are set properly. You can check that by letting the install process create a log file.

Browse to the /bin/release/ folder in the setup project and open a command prompt. On the prompt type the next command:

msiexec /I Demowebsite_Setup.msi /l*v demo.log

You’ll see that during the installation process a “demo.log” file will appear in the /bin/release/ folder. In that log file you will see the parameters are set to the entered values (second line in the screenshot)


Next parts

This concludes the custom UI we wanted to create. In the next chapters we’ll use the entered information to create a database.

Off course there is (a lot) of room to enhance the UI with custom bitmaps, text, … More information you can find in the WiX documentation page.

The items in yellow are handled, the next post will be probably on how to add a website in IIS.

  • Install the .NET 4.5 framework if that isn’t installed already
  • Install the MVC 4 framework if that isn’t installed already.
  • Create a folder and copy all needed files to run the application (done)
  • Create a new database on an existing SQL server and prefill the database with the correct tables and values. (the connection details and database name should be entered by the end user running the installer)
  • Create a new website in IIS 7.5 (create website and application pool running under .NET 4.5)
  • Alter the config file so the correct connection settings are used (entered by the end user)

Complete source code

Can be found on Github.

Other posts in this series

Create an installer for website with WIX, part 1

For a customer we created a new web application in MVC4 with an underlying SQL database. One of the requirements was to provide an installer to install this website at their customers local installations. The installer had to do a few tasks:

  • Install the .NET 4.5 framework if that isn’t installed already
  • Install the MVC 4 framework if that isn’t installed already.
  • Create a folder and copy all needed files to run the application
  • Create a new database on an existing SQL server and prefill the database with the correct tables and values. (the connection details and database name should be entered by the end user running the installer)
  • Create a new website in IIS 7.5 (create website and application pool running under .NET 4.5)
  • Alter the config file so the correct connection settings are used (entered by the end user)

From Visual Studio 2012 on there is no Windows Installer project available any more. You can use the InstallShield Express edition with limited capabilities or the Windows Installer XML (WiX) open source package created by Rob Mensching when he was working for Microsoft. (It’s actually the oldest open source project from Microsoft and now under the OuterCurve foundation)

The Installshield express version doesn’t support IIS installations and falls out of the boat. WiX does support all actions that we have to do but has a steep learning curve. I used WIX in a previous project and still had some hassle to put all of it together. This series of posts will walk us through the creation of the installer.

Start project

Start by downloading the WiX toolset from For this demo I used the 3.8.826.0 version. This is not the latest stable published release but I haven’t got any problems with this version.

Next step is creating a MVC 4 web project and choose for an internet application so we have some default files (javascript, css, views, controllers, …).

Right click on the solution in the solution explorer and choose to add another project. In the ‘Add New Project’ dialog select Windows Installer XML on the left hand side. Choose for ‘Setup Project’ and click the OK button.


You should get a new project with only one file (Product.wxs).

WiX flow

WiX source files are written in XML and have the wxs or wxi (for variables) extensions. Those files have to be compiled to wixobj files. This can be done in Visual Studio or by command line by using the candle.exe tool in the WiX toolset. After compiling the wixobj files another tool is needed to create the msi (installer) file, the light.exe tool.

The most simple installer can be created by just using one wxs file. You will notice that it will make your project more clear to use different wxs files. One for every part of the installation.

You can use the default UI that is available in the WiX toolset but with bigger projects (and in this demo) you can create your own UI and flow. Even the UI is defined in XML and has the same wxs extension. Another reason to split up your installer code in different files to keep the overview.

Step 1: install all needed files

Open up the Product.wxs file in Visual Studio. You’ll see already a few standard values filled out.

At line 3 you get the product attributes that have to be set. Leave the asterix ( * ) at the Id tag. WiX will replace this with an unique Guid when compiling the source. Set the version, language and the name to the desired values.

IIS default location for websites is the c:\inetpub directory. We’ll alter the installer so this default location is used. In one of the next chapters we’ll be able to change this folder. Navigate to line 15 and alter the Directory tag.

You’ll see I’ve changed the default InstallFolder to Inetpub. (from c:\Program Files to c:\Inetpub). This is all we have to change for the install location.

On line 26 you’ll see the ComponentGroup where we’ll have to define all files that have to be installed in our installation folder. Let’s start with adding some files. In the example below I added 3 files in the root of out application (favicon.ico, web.config and global.asax). To add the bin folder I had to add a new ComponentGroup, a new Component, and a new Directory element before I could add the files (2 dll’s).

As you can see this is a tedious job to add every file you want to be installed. Luckily there is a faster way to create this.

Use the heat component from the WiX toolset

The WiX toolset has another tool heat.exe that can help us to harvest all files that we need to install. Although heat was incorporated in Votive (the Visual Studio environment for WiX) in earlier versions, in the 3.7 – 3.8 version this is not available in Visual Studio.

MSBuild to the rescue

If we want to make use of the heat component we’ll have to script it. We can create a bat file we can run every time before we build the installer or we can create an MS build script that we can run. The MS build script has the advantage that we can reuse this script for out build server (continuous integration).

Create a new text file in the Setup project and rename it to First we’ll add some properties in a ‘PropertyGroup’: the source of our website and the name of the WiX file we want to build. We also include the path where we should publish all files. In the ‘itemgroups’ we define the temporary files witch is the content of the web site and the list of WiX input files.

Add build target

We first have to build our website so we are sure we have the latest build we are deploying. Therefor we add a target in the MS build file

Add Publish website target

We’ll use the build in publish feature of MS build to deploy the website to a new folder so we have only the files we need. (and not the .cs files etc)

Harvest the files in WiX

Now that we have all the files we need under a temporary folder we can use the heat.exe tool in the WiX tool belt to harvest the files and create a wxs file.

The parameters used in this command:

  • dir $(Publish) tells to harvest a directory (our published website)
  • -dr:  The directory where the files have to be installed to
  • -ke:  Keep the Empty directories
  • -srd:  Suppress harvesting the root directory as an element.
  • -cg: The ComponentGroup name that have to be used
  • -var var.publishDir: Will substitute the source directory with a wix variable so we can use $(var.publishDir)\myfile.txt in the wxs files
  • -out $(WebsiteContentCode)  the file we want to be created (see PropertyGroup settings)

Test the script

With the heat command inserted we can test our script. Open up a Developer Command Prompt for VS2012. Change the prompt to the DemoWebsite.Setup project folder and type the following command:

msbuild /t: Build;PublishWebsite;Harvest

and hit enter. If everything goes well you’ll see a lot of command coming by. The script will create a folder Setup\publish under the root and publish the website. At last a WebsiteContent.wxs file will be created in the setup project folder.

If you open up the WebsiteContent.wxs file you’ll see all files and folders are added with their own Id under a ComponentGroup MyWebComponents.

If you looked closely you’ll have seen a few WiX commands passing by when executing the build file. Because we are going to handle the WiX build process in our build file we can exclude the setup project from the build configuration. Right click on the solution in the Solution Explorer in Visual Studio and choose for Configuration manager.


Change the active solution configuration to ‘Release’ and uncheck the build flag next to the setup project.


Update the Product.wxs file

Now we have all our files we have to install we have to reference to the created MyWebComponents CompenentGroup and delete the entries we made before to add the files.

Build the installer

Now that we have all the components for the first fase (installing the files) we can use the candle.exe and light.exe tools from the WiX tool belt to built our installer.

Add properties in the build file

First we need some more properties in our build file. Add the WebSiteContentObject parameter that will hold the compiled WiX code (WebSiteContent.wixobj). And also add the MsiOut parameter that will hold the path and name of the installer (.msi) file.

Add candle.exe in the build file

Add a new target tag in the build file and add the candle.exe tool with the parameters where to find the publish directory and witch files he has to compile.

Add light.exe in the build file

In the same target (WIX) add the light.exe command with parameters where to put the generated msi and witch source files to include.

Final run

Open up your Developer Command Prompt and type the next command where we added the WIX target:

msbuild /t: Build;PublishWebsite;Harvest;WIX

Hit enter and keep your fingers crossed. If you had no error messages you should find a msi file in the bin/release folder of the setup project. Run that installer and you’ll see that under the C:\inetpub folder a DemoWebsite folder is added with all the published files from our webapplication.

If you had any errors you can find the complete files here:

Next parts

Enough for one blog post I would say. the next posts in this series will handle the next actions:

  • Install the .NET 4.5 framework if that isn’t installed already
  • Install the MVC 4 framework if that isn’t installed already.
  • Create a folder and copy all needed files to run the application (done)
  • Create a new database on an existing SQL server and prefill the database with the correct tables and values. (the connection details and database name should be entered by the end user running the installer)
  • Create a new website in IIS 7.5 (create website and application pool running under .NET 4.5)
  • Alter the config file so the correct connection settings are used (entered by the end user)

Complete source code

You can find the complete source code for this project on GitHub. Keep in mind that this project will be altered when the next parts are implemented. I will try to keep the commits together with the series.

Other posts in this series

Securing Backend Pages in MVC for Umbraco 4.11.1

Beginning of the year I posted an article on how you can create backend pages in the open source ASP.NET CMS Umbraco ( At the end of the post I published a little TODO list:

  • Block access to the backend controller for non authenticated users
  • Adapt the actions that you get in the Umbraco back end by right clicking
  • Due to a very busy work schedule I didn’t had the time to look into this list. In the mean while version 6 was released by the Umbraco core team.

    Last week I received the question if I had already found a solution for the first item on the todo list: blocking access to the backend controller for non authenticated users.

    Between the search for Easter eggs together with my son, backing pancakes and nothing interesting on the TV this evening I found the time to implement the feature.

    Reminder: the problem

    A small reminder of the problem we are facing. We’ve created a MVC controller to access from the Umbraco backend to perform our own business logic like shown in the image below.


    We used the implementation if the Umbraco’s Surface  Controller that will create a MVC area. If we know the structure of the URL to this area – controller – action we can access the page without logging in the Umbraco backend.


    This is off course a security risk we have to solve. We don’t want non authenticated users to be able to access our business logic.

    Using the Authorization attribute

    In ASP.NET MVC access to certain controllers or individual actions is managed by using the Authorization annotation attribute. You can use the default attribute where you are able to filter request based on specific users or specific user roles.

    But you can off course inherit from the default attribute and implement your own authorization attribute.

    Create the UmbracoAuthorizeAttribute

    In this post I’ll extend the project we’ve created during the previous post and that you can find on Github.

    Add a folder ‘Helpers’ in the UmbracoMVCDemo.BackEnd project. In that folder create a new class ‘UmbracoAuthorizeAttribute’. Add the public keyword to the class implementation and we shall inherit from the System.Web.MVC.AuthorizeAttribute.

    Implementing authorization

    Implementing your own authorization login can be done by overriding two methods from the base class: AuthorizeCore(HttpContextBase httpContext) and OnAuthorization(AuthorizationContext filterContext).

    In AuthorizeCore(HttpContextBase httpContext) we will connect to the Umbraco backend to check if the user is authenticated or not.

    I’ll first check if the httpContext contains any value, if not we’ll throw an exception. After that I’ll use a method from the Umbraco BasePage: ValidateUserContextID. This method will return true if the user is authenticated and false if not. With that result we’ll set a class variable we can use in the implementation of the OnAuthorization method.

    In this implementation we’ll first execute the base class implementation before we check the failedAuthentication class variable. If the authentication in the failedAuthentication failed we’ll send the user to the login page using the FormsAuthentication class.

    Adding authorization to our controller

    It’s not enough to create the attribute, we still have to tell our controller to use the authentications we’ve implemented. Open up the DemoAdminSurfaceController class we created in the previous post and add the UmbracoAuthorize attribute (line 4).


    I’ve opened up another browser and accessed the same action again. (keep in mind that we make use of the default FormsAuthentication that will use a session cookie. Therefor will a new tab in the same browser not give the desired effect). Because we’re not authenticated we get redirected to the login page.


    You’ll see the url we entered in the ReturnUrl parameter. If we now enter our credentials we’ll get redirected to the Umbraco back end. Although the ReturnUrl is filled out we’ll not be redirected to the correct page. We still have to click on our demo Section and demo node.


    I’ve had to alter the default web.config file to be correctly redirected to the login page. I’m not sure this needs to be done if you have a fresh installation, I’ll add it here to avoid bad redirections.


    You can find the full source code from this post and the earlier posts on Github. Feel free to download, fork, comment, …

    Creating a JIRA interface in 1-2-3

    For one of our projects where close to the first test release. While preparing the necessary documents (test overview, legal documents, bug report, …) I felt a bit ashamed to give the client a Word document to report bugs to us. For a software development company it’s a bit shameful to let the client fill out a paper, scan it, mail it and then have someone copy the document in our issue tracker.

    While driving home I thought it couldn’t be that difficult to set up a small site where the client can report the bugs. But I didn’t want to create a new bug tracker nor give the client direct access to our issue tracker (JIRA from Confluence).

    JIRA has a API that you can address to create and fetch issues from their tracker. But I didn’t want to  spend a couple of hours implementing the REST service. (jeah, I know, lazy as hell)

    Nuget to the rescue

    When I got home, the first thing I checked was the nuget site to see if anyone had already created a package to communicate with the JIRA API. First thing that showed up in the search was Atlassian.SDK a complete package, even including a Linq Query implementation to search for issues created by Federico Silva Armas.


    Great, let’s get started.

    New MVC 4 web application

    I fired up Visual Studio 2012 and choose a new MVC 4 Default application with Razor support. Opened up the Package Manager Console and ran the following command:

    Add controller

    I added a new controller in the controllers folder and named it HomeController. In the Index method I’ve added the call to JIRA to fetch all items from a certain filter.

    The implementation is really simple, create a new Jira object with as parameters the URL to your JIRA instance, a username and a password. (in the constructor)
    In the Index method you then can call GetIssuesFromFilter and add the filtername as parameter. The advantage of using a filter is that you can always change this filter settings if you want to remove certain issues from the result without to change anything in your application.

    You’ll need off course the user rights correctly set in JIRA to access the issues that are returned from the filter.

    Add View

    Next I created a new view under the Views – Home folder using the add view action after a right click on the folder. Create a strongly-typed view by selecting the Atlassian.Jira class as Model class.


    Click Add and the view is created.

    If you hit F5 now, your browser should open up and you get already the overview screen for all issues returned from the filter you have selected.

    In 10 minutes I’ve created a working (but ugly) web application that can fetch all the tickets I wanted to show to the customer. Of course your view will need some tweaking to show the values of the issues you want to be shown.

    Adding a new issue

    Adding a new issue to JIRA is as simply. (You can check the complete code on Github). I give you a (very) short overview.

    Create 2 new methods Create in the HomeController. The first one just returned the view, the second one add the HttpPost attribute and this will receive an Atlassion.Issue object with all the values. Pushing this to JIRA is as simple to create a new issue object from the jira object and fill up the values. Hit SaveChanges on the jira object and your issue is created.

    On thing I had a problem with is the projetc name – key difference in the jira.CreateIssue() method. I first tried with the project name but I got some HTML 500 error returning. Although for fetching issues this worked fine but for creating issues you apparently have to gibe the project key instead of the project name.

    Look and feel

    Ok, we have a working application but the look and feel isn’t that, certainly not enough to send to an end user. I didn’t wanted to spend to much time on the look and feel and decided to use the Twitter Bootstrap.

    Don’t start downloading the package from the Twitter Bootstrap website. Instead use the nuget package manager again. There is a Bootstrap for MVC 4 package that will add all necessary items to your solution.

    Open up the Package Manager Console again and run the following command:

    Add bootstrap to the bundles

    To add the bootstrap files to your application we’ve to alter the style and script bundles that are in use. Open up the BundleConfig.cs class in the App_Start folder. There we can add the necessary files to the bundles.

    Online 28 I changed the the Content/css bundle to include the bootstrap css instead of the default Site.Css. On line 21 I added a script bundle to include the bootstrap.js file.

    Because I added a script bundle we’ll have to add this bundle in the _ Layout.cshtml view.

    Just underneath the rendering of the jQuery bundle add the rendering of the bootstrap bundle. (at the bottom of the view.


    I was able to create an new interface for our client to report bugs in just an hour of two. (the above only took 30 min but I’ve added some extra logic to send emails and stuff). Off course it’s not correct to add all business logic in your controllers and to hard code some strings and stuff. For this small application it’s more then enough. Underneath the screenshots of the 2 pages (in Dutch).



    Source code

    You can find this project on Github, feel free to fork, comment, …

    One note, to avoid adding my personal credentials to the project I’ve created an xml file that holds the credentials in the App_Data folder. When creating the Jira object I fetch the values from this file. For obvious reasons I added this file to the .gitignore so it wouldn’t be pushed to Github. The structure of the file:

    The hard search for missing Extension methods in Umbraco 4.7.0

    Implementing SignalR

    For one of our company websites we got the question from HR to implement a chatbox on the website so interested candidates could contact them in an easy way. After comparing a few open source alternatives we decided to build our own using the just released SignalR framework.

    Everything worked fine on our development server where we’re running a backup of the production Umbraco website. Our dev team could easily create the chatbox and sooner as expected we had a working chatbox to present to HR.

    Moving changes to staging

    Like all our development projects we work in different steps. Development on the local machine or development server, client tests on the staging machine. This way we remove the risk that some developers test come in the way of our clients testing.

    For this implementation the client is our own HR department, still we follow the same development guidelines to avoid test data on the acceptance (staging) environment.

    Extension methods missing

    After copying all DLL’s and usercontrols to the staging environment we noticed that all of our Macros in Razor where failing. Suddenly the ‘umbraco.MacroEngines.DynamicNodeList‘ was missing the extension methods like ‘First’ and ‘FirstOrDefault’.


    Start debugging

    First reaction was of course to delete all the DLL’s we’ve altered for this last release and to reinstate the old DLL’s. Still we faced the same problem.

    Back to start, we took a new backup from the production server and started the deploy again. There we noticed we overwritten a third party DLL ‘Newtonsoft.Json.dll’, a well known library for Json conversions created by James Newton-King. More information on his website.

    We skipped the replacement of the DLL and all of our Razor scripts we’re working like before. Problem solved (we thought).

    SignalR failed to work

    After the re-deploy of our chatbox DLL’s and usercontrols we couldn’t get the chatbox to work. It seemed the SignalR hub was missing to contact the server to register a chat user. A check with Firebug reveled the hub couldn’t be loaded because a DLL reference couldn’t find the correct version of the DLL. Of course the ‘Newtonsoft.Json.dll’ was playing up again.


    Some more debugging

    No problem, we find out where Umbraco references the DLL and update it to the latest version and rebuild the Umbraco DLL and we’re good to go. Wrong, we checked the source of the 4.7.0 version that was available on codeplex but no sign of ‘Newtonsoft.Json.dll’.

    Then it had to be one of our own projects referencing the DLL? No, none of our own project had a reference to the ‘Newtonsoft.Json.dll’.

    I loaded the project on disk, configured it in IIS and attached the Visual Studio debugger. In the Debug menu –> windows –> Modules component I could find the ‘Newtonsoft.Json.dll’ reference but no way to find out witch component was responsible for loading the DLL.

    Some more (manual) debugging

    As I couldn’t find a way to find out who was responsible for loading the DLL I had only one option left.

    I took a look at the download at codeplex and removed all DLL’s that were not in the original download. Of course some other problems showed up as all custom created usercontrols failed to load but the Razor Extension methods worked and I could start a tedious process of adding DLL’s to the bin folder with my Visual Studio debugger attached and to reload the site every time.

    Found the problem

    When I copied the ‘Twitterizer2.dll’ back into the bin folder the Razor scripts failed again. I finally found the source of the problem. After a quick look at the Github repository of the Twitterizer project I found the Newtonsoft references and the different versions in the packages folder.

    Resolving the problem

    I did found the problem but we wanted to keep the Twitter possibilities that we had implemented before so we needed the Twitterizer2 DLL.I forked the project on Github and cloned the project locally.

    After updating the package to the latest version (the same our SignalR implementation was using) I’ve rebuilded the DLL and copied it to the Umbraco bin folder. Finally we had our Razor scripts working and our chatbox was responding!


    It took me half a day of debugging before I did found the problem. If anyone has a good way to find out witch component loads a certain DLL, let me know in the comments or by a tweet, …

    Using MVC backend pages in Umbraco 4.11.1

    Umbraco backend

    In a previous post we saw the use of your own MVC controllers in the Umbraco front end. The Umbraco community has made effort to document these possibilities on The use of MVC in the back end of the Umbraco interface is not very well documented and it took me some analysis and debugging of the source code to find out how this could be implemented.

    Default you have 7 sections in the backend:

    • Content
    • Media
    • Settings
    • Developer
    • Users
    • Members
    • Translation (hidden by default for the admin user)

    These sections will do in a plain website where you don’t want to use custom code.

    One of my personal projects is to create a new website for the sport club where I’m already 15 years affiliated with. The current website is already 7 or 8 years old and was written in PHP. Needless to say that the look and feel is outdated, the PHP code I’ve written 8 years ago wasn’t the most beautiful piece of code I’ve created.
    I wanted to create a new website in ASP.NET MVC but I don’t want to put the effort in creating controllers and views for the administration part where a good CMS can fill up this gap.

    Because I’m already familiar with Umbraco it looked a good choice to combine the strength of Umbraco with custom development. But therefor I had to get custom MVC controllers working in the backend.

    New backend project

    This blog post goes further on the project created in the previous post. You can download the source at Github.

    In the existing solution we’ll add a new project by right clicking the solution and choose Add – New project.


    Choose for a ASP.NET MVC 3 Web Application and give it a name (in this demo: UmbracoMVCDemo.BackEnd)


    In the next window choose for Internet Application and the Razor view engine. We’ll need the core Umbraco libraries so open up the Package Manager Console and type ‘”Install-Package UmbracoCms.Core” and hit enter. Wait for the “Successfully added ‘UmbracoCms.Core 4.11.1’ to UmbracoMVCDemo.BackEnd.” message and we’re ready to go.

    Adding a new section

    First we’ll have to create a new section in the Umbraco backend. Open op your windows explorer and go to the root of the Umbraco installation. Open up the config folder and open up the applications.config file.

    <?xml version="1.0" encoding="utf-8"?>
      <add alias="content" name="Content" icon=".traycontent" sortOrder="0" />
      <add alias="media" name="Media" icon=".traymedia" sortOrder="1" />
      <add alias="settings" name="Settings" icon=".traysettings" sortOrder="2" />
      <add alias="developer" name="Developer" icon=".traydeveloper" sortOrder="3" />
      <add alias="users" name="Users" icon=".trayusers" sortOrder="4" />
      <add alias="member" name="Members" icon=".traymember" sortOrder="5" />
      <add alias="translation" name="Translation" icon=".traytranslation" sortOrder="6" />

    You’ll see the existing 7 sections configured in the config file. We’ll add a new one:

      <add alias="demo" name="Demo" icon="demo.gif" sortOrder="7" />

    We give the new application or section the alias ‘demo’ and the name ‘Demo’. For the icon we’ll add a new picture called ‘demo.gif‘ and set the sort order to 7. (The default images are showed using sprites where the icon name is actually the css class to define the background image. You can choose to alter the existing sprite and add a css class or select new picture). Create a new picture with your favorite image editor, name it ‘demo.gif’ and place it in the folder: [Umbraco root]/umbraco/images/tray/ folder.

    We’ll have to force the application pool to restart before Umbraco loads the new section. Open up the web.config file in the umbraco installation root and add a line break, save and refresh the page.

    Still the new section is not available in the backend because none of the users have access to this new section. Go to the users section, open up your user and add a check in the sections area next to the demo section. Refresh your page and you’ll see the new section appear.


    When you click the Demo section at the bottom of the page we’ll see a new section opening up with an empty tree and the section name in square brackets.

    Adding a tree

    Next step will be to create the tree for our Demo section. This tree you can populate with data you choose. For this demo we’ll use some dummy data.

    In our backed project add a folder ‘Trees’ and add a new class named DemoTree. This class will have to inhered from the umbraco.cms.presentation.Trees.BaseTree class. This class has 3 abstract methods that we’ll need to implement.

    • RenderJS: the javascript that will run when a tree node is clicked
    • Render: the creation of tree nodes
    • CreateRootNode: to define the root node.

    We’ll also have to create a constructor that will receive a string that will forward to the base class.

    Last but not least we’ll have to decorate the class with the umbraco.businesslogic.TreeArrtibute where we define the application alias and title.

    using System;
    using System.Collections.Generic;
    using System.Linq;
    using System.Text;
    using System.Web;
    using umbraco.cms.presentation.Trees;
    namespace UmbracoMVCDemo.BackEnd.Trees
        public class DemoTree : umbraco.cms.presentation.Trees.BaseTree
            public DemoTree(string application) : base(application)
            public override void RenderJS(ref StringBuilder Javascript)
                throw new NotImplementedException();
            public override void Render(ref XmlTree tree)
                throw new NotImplementedException();
            protected override void CreateRootNode(ref XmlTreeNode rootNode)
                throw new NotImplementedException();

    Let’s start with defining the root node what is as simple as setting the NodeType and NodeId parameter.

    protected override void CreateRootNode(ref XmlTreeNode rootNode)
    	rootNode.NodeType = "init" + TreeAlias;
    	rootNode.NodeID = "init";

    Next we’ll implement the Render method and attach 2 nodes to the tree.

    public override void Render(ref XmlTree tree)
    	XmlTreeNode xNode = XmlTreeNode.Create(this);
    	xNode.NodeID = "1";
    	xNode.Text = "Demo Node 1";
    	xNode.Action = "";
    	xNode.Icon = "folder.gif";
    	xNode.OpenIcon = "folder_o.gif";
    	OnBeforeNodeRender(ref tree, ref xNode, EventArgs.Empty);
    	if (xNode != null)
    		OnAfterNodeRender(ref tree, ref xNode, EventArgs.Empty);
    	XmlTreeNode xNode2 = XmlTreeNode.Create(this);
    	xNode2.NodeID = "2";
    	xNode2.Text = "Demo node 2";
    	xNode2.Action = "";
    	xNode2.Icon = "folder.gif";
    	xNode2.OpenIcon = "folder_o.gif";
    	OnBeforeNodeRender(ref tree, ref xNode2, EventArgs.Empty);
    	if (xNode2 != null)
    		OnAfterNodeRender(ref tree, ref xNode2, EventArgs.Empty);

    In the RenderJS method remove the ‘throw new NotImplementedException” line and build the project. We’ll have to copy the UmbracoMVCDemo.BackEnd.dll from the bin folder to the Umbraco installation bin folder.

    Refresh the page and you’ll see two new node appear. Notice that the Demo root node isn’t surrounded by square brackets any more.


    If you open up the trees.config file in the umbraco root – config folder you’ll see Umbraco automatically added the tree.

    For now the nodes don’t do much if you click on it because we left the Action attribute empty. Before we can add an action we need the editor for the right hand side.

    Adding an editor – controller and views

    Like I said before we want to create the editor in a MVC pattern instead of the default Webforms like was possible before. Therefor will create a new controller by right clicking the controllers folder in the backend project.


    Name the controller DemoAdminController and choose for the empty controller template. Because we’ll already have a DemoController in the frontend application I named this one DemoAdmin to avoid confusion.

    The Controller will have to inhered from the Umbraco.Web.Mvc.SurfaceController before we can use it in the backend.

    IMPORTANT: Although the documentation of version 5 stated the controller didn’t need to be suffixed by ‘Surface’ I noticed there are constraints set on the PluginControllerResolver in the Umbraco source code. We can remove this constraint in the source code but the easiest way is to rename our controller to DemoAdminSurfaceController. This will change our URL’s but they won’t be visible in the Umbraco backend anyway.

    We have to decorate the class with the Umbraco.Web.Mvc.PluginController attribute. (yes, the plugincontroller attribute although we’re inheriting a SurfaceController). In this attribute we’ll have to enter the area name. Every plugin will be placed in his own MVC area to avoid duplicate views and controllers. We’ll set the area name to ‘demo’.

    Last but not least we’ll have to create two constructors. One without parameters, one that will take a Umbraco.Web.UmbracoContext instance and forward this to the base class.

    using System;
    using System.Collections.Generic;
    using System.Linq;
    using System.Web;
    using System.Web.Mvc;
    using Umbraco.Web;
    using Umbraco.Web.Mvc;
    namespace UmbracoMVCDemo.BackEnd.Controllers
        public class DemoAdminSurfaceController : SurfaceController
            public DemoAdminSurfaceController()
            public DemoAdminSurfaceController(UmbracoContext umbracoContext)
                : base(umbracoContext)
            public ActionResult Index()
                return View();

    Last but not least we’ll have to create a view to back up the Index action on our controller that will for now will be one line:

    “<h1>Backend Demo Controller – view</h1>”

    Build the backend project and copy the dll to the Umbraco installation bin folder. Default will the view be searched in the following folder:

    [Umbraco ROOT]/App_Plugins/Demo/Views/DemoAdminSurface/. That is App_Plugins/[AreaName]/Views/[ControllerName]/. Create the folders and add the index view.

    I we go to the Umbraco backend we”’ll notice the 2 demo nodes still don’t do anything if we click them. We’ll first have to assign an action to these nodes.

    Assign an action to the nodes

    To assign an action to the nodes we’ll have to do two things. We’ll have to implement the RenderJS method in out Tree class and fill out the Action parameter from our nodes in the Render method.

    In the RenderJS method will render the necessary javascript that will open our editor in the right hand pane by using the ‘UmbClientMgr’ javascript object.

    public override void RenderJS(ref StringBuilder Javascript)
    	@"function openDemoController() {

    In the Render method will add a call to the javascript action we just created (for now only on the first node).

    public override void Render(ref XmlTree tree)
    	XmlTreeNode xNode = XmlTreeNode.Create(this);
    	xNode.NodeID = "1";
    	xNode.Text = "Demo Node 1";
    	xNode.Action = "javascript:openDemoController();";
    	xNode.Icon = "folder.gif";
    	xNode.OpenIcon = "folder_o.gif";
    	OnBeforeNodeRender(ref tree, ref xNode, EventArgs.Empty);
    	if (xNode != null)
    		OnAfterNodeRender(ref tree, ref xNode, EventArgs.Empty);
    	XmlTreeNode xNode2 = XmlTreeNode.Create(this);
    	xNode2.NodeID = "2";
    	xNode2.Text = "Demo node 2";
    	xNode2.Action = "";
    	xNode2.Icon = "folder.gif";
    	xNode2.OpenIcon = "folder_o.gif";
    	OnBeforeNodeRender(ref tree, ref xNode2, EventArgs.Empty);
    	if (xNode2 != null)
    		OnAfterNodeRender(ref tree, ref xNode2, EventArgs.Empty);

    Build the project and copy the dll to the Umbraco installation bin folder and refresh the backend in the browser. If we now click the first test node we’ll see the test view appear in the right hand side frame.

    Automating the copy process

    Just like in the previous blog post you can use the build events to automate the copy process of DLL’s and views. Place the 2 xcopy commands in the post build event.

    xcopy $(TargetPath) $(SolutionDir)UmbracoMVCDemo\bin\ /C /Y
    xcopy $(ProjectDir)Views $(SolutionDir)UmbracoMVCDemo\App_Plugins\Demo\Views\ /E /Y

    Source code

    You can find the solution above on Github.

    To do’s

    There are still some things I need to resolve:

    • Block access to the backend controller for non authenticated users
    • Adapt the actions that you get in the Umbraco back end by right clicking
    • ….