Video Transcript

Hello and welcome to another SQL Server Video. My name is Michael Campbell, and in this short SQL Server Video, we're going to take a look at some very simple new changes and improvements to SQL Server 2008 that provide some great improvements to T-SQL coding.

Specifically, in this video we'll be looking at Row Constructors, Inline Variable Assignment, Compound Assignment operators, and Grouping Sets. We'll also briefly talk about some of the great new capabilities added to the OUTPUT clause that makes it even more powerful and useful for very complex operations.

And, in this video, we'll also take a very brief look at Table-Valued Parameters and the new T-SQL MERGE statement. But, since these two changes are significant enough in their own right, subsequent SQL Server Videos will look at these two improvements in more detail.

To get started, let's take a look at T-SQL's new Row Constructors, which are best demonstrated by way of an example. And, in this example, I'm just creating a User Types lookup table for a sample application, and populating it with some stock User Types.

The standard way of accomplishing that is to just fire off multiple, distinct, SQL INSERT statements for each row that you need to INSERT. Though, it is common to see people cheating and using INSERTs based off of SELECT statements and UNIONs as well. I'm lazy and have always liked this UNION approach - even if it does cause a decrease in readability.

But with SQL Server 2008's Row Constructors, it's now much easier to INSERT multiple rows into a table while maintaining readability.

It's important to understand though, that under the covers, when using these new Row Constructors, that the kinds of performance you're going to see are really on par with what you'd get using the UNION SELECT INSERT approach - because, really, all this new feature represents is some syntactical sugar that makes life a bit easier.

So, if you need to insert LARGE numbers of rows, then you'll want to look into the Bulk Copy Program or BCP as it provides better throughput and performance.

Another new T-SQL improvement that you'll probably use a bit more than Row Constructors is T-SQL's new ability to do inline variable assignment. And, if you're familiar with C# or VB.NET development, you'll take to this new feature like a fish to water.

Without this feature, to create and assign a variable, you need to first create, and then, well, assign the variable. Assignment of static values to variables is correctly done by using the SET statement.

Though, in previous versions of SQL Server you could also 'SELECT' a static value into a variable, but database snobs will, correctly, look down on you for doing so. And that's because the use of the SELECT statement in assigning variables is used when you are assigning a dynamically assigned, or SELECT-ED, value to your variable - like so.

With SQL Server 2008 though, it's now possible to do static variable assignment with a single line of code - or inline as it were. So, in cases where we want to see everything greater than a given date, we can do that in a single statement now, instead of in two different lines.

And, while this isn't a huge or sweeping change, once you've used it for a while you'll be hard pressed to work without it, and you'll find that it does help save a bit of time when writing sprocs and so on.

And just as SQL Server 2008 provides some simple improvements that make variable assignment much easier and more terse, it also introduces the notion of compound assignment operators that can make working with your variables that much easier.

So for example, prior to SQL Server 2008 if I wanted to change or modify a variable, I basically had to set it equal to itself and any other value that I wanted added - as in the case of these varchar and int variable assignments.
But with SQL Server 2008, you can now do the same thing in a much more terse fashion.

And, again, while this isn't an earth-shattering improvement, it can help speed up your code - and if you're familiar with C# and VB.NET, you'll likely find this new syntax much more comfortable as well.

And, what's really great about these compound assignment operators is that they apply to a wide variety of data types and operations as well - as you can see here in Books Online.

SQL Server 2008 also introduces a number of T-SQL changes and improvements to the GROUP BY clause - including a new ISO compatible syntax, a new GROUPING SETS operator, and a new GROUPING_ID() function.

The new ISO compliant syntax isn't terribly different from previous versions - which still work, though those previous formats are only supported in SQL Server 2008 for backwards compatibility reasons.

Otherwise, in terms of new capabilities, SQL Server 2005 introduced the CUBE and ROLLUP operators - which provide some great capabilities, as you can see from this very simple example.

With SQL Server 2008 though, we now have much more control over how we'd like data aggregated, or grouped, and can explicitly define grouping configurations, or sets, that really let us get close to approximating some of the functionality that you'd get when working with SQL Server Analysis Services Cubes.

The end result though, is that simple reporting applications can be drastically improved to provide increased business intelligence - by just using T-SQL.

And if these capabilities are something that interest you, you'll want to take a look at the section of Books Online that introduces GROUPING SETS, as it's really well put together and will walk you through a series of great examples that will help you understand how to take advantage of these new features.

Otherwise, another improvement that SQL Server 2008 offers is an additional improvement on something introduced, again, in SQL Server 2005: the OUTPUT clause.

Interestingly enough, this change or addition isn't mentioned in the "What's New" section of SQL Server 2008 Books Online, but if you dig around a bit in the documentation for the OUTPUT clause you'll notice some new improvements that allow the OUTPUT clause to be used within nested DML operations. This, in turn, allows for some pretty powerful options and possibilities that just weren't available before.

For example, notice how, in this example, I'm actually INSERTing 10 rows into my table variable instead of just 5 - because I'm actually redirecting output from my OUTPUT clause back in upon the INSERT statement - in a sort of nested 'feedback loop'. And the cool thing about this is that it's a single, atomic, operation - which you can get an inkling for by looking at the way the results are ordered.

And what's great about these new improvements is that you can now use the OUTPUT clause in any nested DML statement including INSERTs, UPDATEs, DELETEs, as well as the new MERGE statement.

Which, of course, leads us to the MERGE statement - which is another great new T-SQL improvement introduced by SQL Server 2008. And the best way to think of this statement is to think of it as a sort of hybrid between the UPDATE and INSERT statements already found in T-SQL.

In fact, this statement very nicely fits the need that many database developers jokingly call the UPSERT clause.

And in terms of operations, this statement kind of acts like a hybrid between a way to JOIN tables, and then define CASE statements that, in turn, define what to do when data being evaluated is found, or matched, and when it isn't - as you can see from this very simplistic example.

And, while this statement is frequently associated with Extract, Transform, and Load operations - or part of Data Warehouse Management, it can also be useful in OLTP applications as well. So, we'll take a look at this statement in more depth in an upcoming SQL Server Video - where we can examine some of its benefits, and work through a number of examples in more detail.

Finally, one of the coolest new things that SQL Server 2008 introduces is the notion of table valued parameters. And, if you've worked with table variables before, then you'll love table valued parameters as they effectively allow you to use table variables as parameter that can now be passed into sprocs and UDFs - which makes it much easier to process entire collections of data instead of single sets of scalar variables.

Prior to SQL Server 2008, if you wanted to achieve this kind of functionality, you were likely using some sort of hack that passed in serialized values that you could then re-hydrate into arrays and then process as needed.

But with SQL Server 2008, you no longer have to jump through those hoops to take advantage of the ability to process multiple rows of data within a single procedure.

Therefore, in a subsequent video we'll take a look at both how to use Table Valued Parameters from within SQL Server as well as take a high-level look at how to pass arrays of data in from .NET applications to take advantage of this functionality from within applications as well.

So, make sure to check out the other SQL Server Videos in this series to learn more, and to stay abreast of new releases, make sure to sign up for SSV's RSS feed.
Otherwise, thanks for joining and we look forward to seeing you again.