- Upload and apply display templates to the Content Search Web Part
- Understanding how search results are displayed

If you have created custom display temples, you have to upload them to the SharePoint sites where you want to apply them. If you only have a few sites in your SharePoint farm, you can easily do this by mapping your network drive, and dragging and dropping your custom display templates into the correct file directory. However, if you have a large number of sites in your farm, this process can be quite time consuming.

In this blog post, I’ll explain how you can upload display templates to all sites in your farm by using PowerShell scripts.

About the scripts

All display templates are located in Master Page Gallery --> Display Templates.

You can use one script to upload display templates to the Content Web Parts folder, and another to upload to the Search folder.

Upload display templates to the Content Web Parts folder

The display templates that are located in the Content Web Parts folder are used by the Content Search Web Part.

To upload display templates to the Content Web Parts folder to all sites in your farm, all you have to do is modify the UploadToContentWebParts script so that it contains the file path to where you have stored your custom display templates.

Then, run this script on the machine that hosts your SharePoint farm. Complete script is attached at bottom of the post.

Upload display templates to the Search folder

The display templates that are located in the Search folder are used by the Search Results Web Part. These display templates are connected to a result type.

The UploadToSearch script will upload your custom display templates to the Search folder to all sites in your farm, create new result types on all sites in your farm, and connect these result types to your custom display templates.

To do this, you have to modify the UploadToSearch with the following:

-          FileType match: FileType that the new result type should match.

-          Result type name: name of the new result type.

-          Full file path: full file path to where you have stored your custom display template.

Then, run this script on the machine that hosts your SharePoint farm.

IMPORTANT: You must include minimum two custom lines for the UploadToSearch script to run correct.

In the following example, I have modified the UploadToSearch script with two custom lines.

By running this script, my two custom display templates are uploaded to all sites in my farm.

Two new result sources are created on all sites in my farm. These new result sources are connected to my custom display templates.

ScriptSample.zip

The reason is missing some required configuration like services and settings.

The error can be seen in both cases when you try to deploy app using PowerShell or Application Center.

Here is the powershell sample

Add-PSSnapin Microsoft.SharePoint.PowerShell;

$spapp = Import-SPAppPackage -Site http://SPHVM-2906 -Path c:\temp\My_App.app -Source ([microsoft.sharepoint.administration.spappsource]::DeveloperSite);

Install-SPApp -Web 'http://SPHVM-2906' -Identity $spapp; 

If you will try to deploy from Visual Studio you will see an error like this

The main reason is that SharePoint Apps require two services: App Management Service and SharePoint Foundation Subscription Settings Service. Ensure that both are installed and running.

Add these services if they are not in list and do not forget to start the services

Also check that web application is started.

Restart IIS form UI or running IISRESET from comand line.

And here is small code sample how to uninstall application 

Add-PSSnapin Microsoft.SharePoint.PowerShell;

$installedapps = Get-SPAppInstance -Web http://SPHVM-2906

$yourInstalledapp = $installedapps | where Title -eq 'My App'

$yourInstalledapp

Uninstall-SPAppInstance -Identity $yourInstalledapp 

So what will happen when during fix/upgrade process (I will call it upgrade) a value which was declared as public const was changed. For clear understanding we should see how build process is done. Actually during build process all variables which consume const will be hardcoded. It is not a problem when it is references from the same project, component will be rebuilt anyway. But what if it is referenced from another dll? The value will remain hardcoded and will contain old value.

Tools.dll MyConstants.cs
public const int DEFAULT_PORT = 80;

Service.dll Host.cs
StartService(DEFAULT_PORT);

So after upgrade of Tools.dll and change of the DEFAULT_PORT to 8080, Service.dll still will try to start service on 80 port! That is easy to test, just build a solution, copy components to another folder, change const value, rebuild and copy only updated DLL, run an application. You will see that application will show the old value.

To avoid troubles with component upgrade and constants it is strongly recommended to mark all constants as internal, or use them inside private or internal classes, classes which are not visible from the outside.

Another solution if you still want to use the same style of constant definition const should be changed to static readonly, which will create reference and value will not be hardcoded.

public static readonly int DEFAULT_PORT = 80;

Which will work correctly after component upgrade.

Happy coding.

This feature is very useful when new managed property is configured or site column is modified so as a result search schema needs to be updated and list owner who initially has not access to central admin to fire full crawl which will be very expensive step for large deployment and will affect a lot of content.

So this cool feature can be found in list (I will not mention library anymore) settings under Advanced Settings link.

Scroll a little Advanced Settings page and pay attention to Search block where you can enable the list to appear in search results. This option should be enabled otherwise Reindex Document feature will not have any sense.

Only then you can press Reindex Document Library or Reindex List button

In popup dialog just confirm that you are sure and want to re-index list.

Now your list will be picked up by next continuous or incremental crawl. Make a note that you do not need to scroll down on Advanced Settings page and press OK, list already marked as dirty.

I am not sure if someone will need to make the same using PowerShell or OM, but anyway drop a line in comment then we can discuss it. 

and registration will start from Feb 12 2013

So TechEd North America will be in New Orleans, LA, June 3 - 6 and

TechEd Europe in Madrid Spain, June 25 - 28

For more info visit http://www.msteched.com/ 

Stay connected https://www.facebook.com/TechEdEurope and https://www.facebook.com/teched

It took some time for me to investigate how to force development process to increase the

quality of HTML markup and minimize cross browser compatibility issues in web UI development.

HTML validation process always left by developer for the last point and has lowest priority.

It is almost not done during development process and only when page or probably in many cases

whole web site is ready for production. 

Compatibility and accessibility testing is another duty of testers but looks like they are in general

more busy with other things and markup left as is with low priority and extremely high number of errors

and warnings in production code. So looks like it is rendered, then no one cares.

For manual testing there are some tools available for devs and tests but I think that

W3C's online validator is out of competition. Just open http://validator.w3.org/ online validator

and try to validate some popular resources.

I will be surprised if any one of them will have low number of validation errors.

I will not publish some numbers here and will try to avoid validating software/hardware and IT service providers.

Try some popular news portals like cnn.com and nytimes.com .

Number of errors is extremely high I think. To be true I do not see any reason to keep that so high

and the only reason it is so just web UI was not designed and developed with compatibility in mind.

Some will tell but it is rendered and looks nice. I will agree that it is working, but we should understand the price of it.

Making bad html to look good leads to extra coding and development pain.

That means more code we need, more testing we need and more UI bugs we have.

Just some words about automation testing. There is manual tests which is well known when tester

tests unit like feature or page manually executing or opening it to check functionality.

Automation test is done using some tools which checks functionality problematically  like browser simulating,

clicking, creating resources, managing them and disposing. Automation test just a step after small development

milestone which should check if feature works as expected and none of dependencies are broken.

Spend some time using your favorite search engine to find more about "Automation testing"

like http://en.wikipedia.org/wiki/Test_automation

Solution is not new and it is web ui automation test. Automation tests will force devs to keep HTML

clear and valid from the first steps of development. It will minimize tester's efforts and save a lot of time for big portal testing.

Code can be as simple as

[TestMethod]
..............
var result = Validator.Client.Check(new Uri("http://www.cnn.com", UriKind.Absolute));

if(!result.Validity)
{
   .... Show errors.... result.Errors
} 

 W3C provides API (web service) for testing. More information can be found here http://validator.w3.org/docs/api.html

API will check for markup validity and will return a list of errors with HTML fragments and comments with suggestions.

So you can consume that API from your test code for HTML validation or use sample library which I am using for my projects,

it is published with sources under http://w3cvalidatorclient.codeplex.com/

So the only thing you need now is to create Test project and implement some lines of code.

<!DOCTYPE html>
<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
<title>Scope test</title>
</head>
<body>
<input type="button" onclick="TestScope()" value="Test" />
<script type="text/javascript">
var MyValue = 10;
function TestScope()
{
	if(true)
	{
		var MyValue = 20;
	}
	
	alert('Final value: ' + MyValue);
}
</script>
</body>
</html>
 
So scopes in JavaScript sometime can be more confusing if 
developer does not clearly understand the usage of var keyword.
I saw previously that developers just skip declaration of the variable 
and use it. Sure may be that worked but what is the price of omitting var?  
So variable which was used without var keyword will become global variable! 
Example will show that variable defined in Scope3 method will not 
affect global MyValue 

<!DOCTYPE html>
<html>

<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
<title>Scope test</title>
</head>
<body>
<script type="text/javascript">
function TestScope1()
{
	MyValue = 'Scope 1,';
}

function TestScope2()
{
	MyValue += 'Scope 2';
}

function TestScope3()
{
	var MyValue = 'Scope 3';
}
TestScope1();
TestScope2();
TestScope3();
alert(MyValue);
</script>
</body>
</html>

For demo I will use Visual Studio 2012 and MVC4. Client call will be hosted in MVC4 page and service will be provided by WebAPI which is also based on MVC4.

Lets first see how it works in one server environment, more complex scenario will be in next posts. 

Start Visual Studio 2012 and create a web site using ASP.NET MVC4 Web Application template.

Then choose Basic template without creating test project.

Create HomeController and corresponding View which will response to root address

I want to skip details about creating controllers and views, you can find a lot of samples and blog posts about it.

On next step I will create another controller which will response to JavaScript calls. To create WebAPI page right click on Controllers folder and click on Add / Web API Controller class.

In new window just edit ValuesController1 by removing trailing 1.

Controller with default methods will be created and that is enough for demo. To confirm and check what is created just build project and run it. Add /api/values which is Web API address by default and confirm response from the server.

The only thing we need is to create a script which will call this web service and show results.
Modify Index view and add to it script block, button and div which will contain the response.

 <div>
 <div id="response">
 
 </div>
 <input type="button" value="Call Web API" onclick="javascript:CallWebAPI();" />
 </div>
 <script type="text/javascript">
 function CallWebAPI()
 {
 var responseContainer = document.getElementById('response');
 var request = new XMLHttpRequest();
 request.open("get", "/api/values", false);
 request.send();
 
 response.innerHTML = request.responseText;
 }
 </script>

 So as a result you will see injected JSON response. You can manipulate and request response to be in XMLor JSON format.

In the next post I will show how to make cross site requests and what kind of difficulties can occur.

Happy coding!