SharePoint Best Practices in display rows visual webpart and Visual Web Part SharePoint 2010 Tree View

1. Don't cache SharePoint objects that are not Thread Safe

Let's take real time scenario, we have to bind dropdown list with the SharePoint list. We know SharePoint list which we are binding with dropdown list is not going to change frequently, so caching will be better option to improve the performance of the application. Sample code is mentioned below:

                      SPListItemCollection colItems;

            if(null != HttpRuntime.Cache[key])

            {

                ddlItems = (SPListItemCollection)HttpRuntime.Cache[key];

            }

            else

            {

                // Do query on sharepoint listand cach SPListItemCollection object

            }

But we are unaware that SPListItemCollection object contains an embedded SPWeb object that is not thread safe and should not be cached.

Good Coding Practice is to cache DataTable in place of SPListItemCollection object.

Sample Code:

         private static object _lock =  new object();

          private DataTable GetDropDownItems()

         {

         DataTable dtDropDownItems;

   SPListItemCollection objListItems;

   lock(_lock)

   {

             dtDropDownItems= (DataTable)Cache["CachedDropDownItems"];

           if(dtDropDownItems== null)

           {

              objListItems = //Do query onSharePoint List

              dtDropDownItems= objListItems.GetDataTable();

              Cache.Add("CachedDropDownItems ", dtDropDownItems,..); // provide rest of the parameter of this method

           }

   }

   }

Note: There are various methods to cache object like HttpRuntime, System.Web.Caching.Cache,  and HttpContext.Current.Cache. Choose one to the method to cache depending upon your application requirement.

    2. Retrieve or fetch SPList instance from SPWeb object: There might be the multiple approach to get the instance of SPList in SharePoint. But while developing an application we should keep in mind about performance of the application. Mentioned below are two approach to achieve the same.

        Approach 1: Not Recommended
        SPList objSpList = <SPWeb>.Lists["<List Title>"];  

        Approach 2: Recommended
        SPList objSpList = <SPWeb>.GetList("<List URL>");
   
        In the first approach, SharePoint loads "metadata" of all the SharePoint lists of current SPWeb object and then create SPList instance by Title comparison and also if Title of SharePoint list gets changed, There will be a code change. But in the second approach, SharePoint fetches GUID of the SharePoint list by list url and then loads the "metadata" of that particular list. User can not update the list URL after SharePoint list creation. So there is no possibility of any code change in this approach.

    3. JavaScript methods call on page load in SharePoint: Sometime we need to call JavaScript methods on page to change UI or to set some controls value. we can use window.onload event to achieve this functionality. But I personally recommend that to use "_spBodyOnLoadFunctionNames" method provided by SharePoint because sometimes window.onload don't work properly.
        Approach 1: Not Recommended
        window.onload = funcation(){//JavaScript code};
        window.onload = <JavaScript method name>;

        Approach 2: Recommended
        _spBodyOnLoadFunctionNames.push('<JavaScript method name>');

     4. Set RowLimit, ViewFields, and ViewFiledsOnly properties while fetching data from SharePoint list:
        Let take me take an example to make understand importance of these properties.

      Scenario: Suppose there isa list to store employee details. Following fields are present in this list:

·    Name

·    Designation

·    Domain

·    Address

·    Email

·    Contact Number

·    Rating

      We have to create a webpart to show top 5 employees name, designation and email in the home page of site based on employee rating.

        To achieve above mention scenario, we will end up with mentioned below code sample:

         Bad coding:

using (SPSite objSite = new SPSite("<Site URL>")) {     using (SPWeb objWeb = objSite.OpenWeb())     {         SPList objList = objWeb.GetList("<List URL>");         SPQuery objQuery = new SPQuery();         objQuery.Query = @"<OrderBy>                                 <FieldRef Name='ID' Ascending='False' />                             </OrderBy>";         SPListItemCollection employeeCollection = objList.GetItems(objQuery);     } }

In the above mentioned codeblock, we will fetch all the fields data but we have to display only 3 fieldsdata in the webpart. So there will be performance hit. As a good practice weshould pass ViewFields also. so that our query will return required fields data only.

            Add following line of code to improve the performation:
         Good coding

//We have specified view fields, now our query will fetch value of these three fields data. objQuery.ViewFields = @"<FieldRef Name='Name'/>                         <FieldRef Name='Designation'/>                         <FieldRef Name='Email'/>";

Even though we have specified view fields, still SharePoint fetches following fields data:

·     ID

·    Created

·   Modified

As per problem statement, we do not want these fields data. SPQuery object has property called ViewFieldsOnly. If we set this property to true, these fields data will not come with the query result. It means, to improve our query performance, we will have to add one more line in our code:

Good coding:

objQuery.ViewFieldsOnly = true;

Note: If you have to do write operation on fetched records by our query, do not set ViewFieldsOnly property to true otherwise while updating any item you will get following exception: "Value does not fall within the expected range."

Still we can improve our code. Here we need to show only 5 employees details and our query will fetch all employee details. So we can improve our code by providing RowLimit to 5.

Good coding:

objQuery.RowLimit = 5;

Note:  Most of the time we do not know how many records will fetch after query execution. But as best practice we should set RowLimit if we know the threshold of our query.


Tree View Code:
============