Comments on: SQL Single Column Table

By: SAffer

SAffer — Tue, 28 Aug 2012 07:29:53 +0000

It looks like the PIVOT command is exactly what I am looking for.

My view structure is based on what the Help for the PIVOT command suggests – a simpler way of creating summarised data rather than using a complex CASE command.
I’m going to try this first - if I don’t come right then I’ll be back for futher assistance and explain in much more detail what I am trying to do – thanks for the feedback, I really appreciate it.

By: carlosdl

carlosdl — Mon, 27 Aug 2012 23:31:33 +0000

"You couldn’t even give the columns names." That's why I mentioned the possibility to alias the column names with numbers when creating the view (I know, you still wouldn't know the number of columns, but if you knew, you would know how to call them), but as I mentioned before, I don't think I would want to have something like that. I think the main problem is that we don't know te reasons why this table exists in the first place, and why SAffer wants to create such a view from it.

By: TomLiotta

TomLiotta — Mon, 27 Aug 2012 22:35:44 +0000

@carlosdl:

Couldn’t the view be created over a recursive CTE? The major problem would be that it would result in a single column with all of the values in a single string. But that would be the only way to handle such a view anyway.

You couldn’t do anything useful with a view that had an unknown number of columns. You couldn’t do any WHERE or ORDER BY if you never knew that columns would be there. You couldn’t even give the columns names.

Tom

By: carlosdl

carlosdl — Mon, 27 Aug 2012 18:32:00 +0000

Sorry, the editor created a mess with the formatting.

By: carlosdl

carlosdl — Mon, 27 Aug 2012 18:27:30 +0000

The PIVOT operator is usually what you would use to create cross tab queries, but I don’t think doing it on single-column tables is its intended use. Comes to mind something like this, but you would still need to know the possible values before hand:SELECT *
FROM (SELECT col1,col1 col2 FROM test1)
pivot
(
MAX(col1)
FOR col2 IN (‘AL1′,’AL2′,’AL3′,’AK01′)
); You could alias the column names if necessary:
SELECT *
FROM (SELECT col1,col1 col2 FROM test1)
pivot
(
MAX(col1)
FOR col2 IN (‘AL1′ AS “1″,’AL2′ AS “2″,’AL3′ AS “3″,’AK01′ AS “4″)
);

But since you are asking for a way to create a view, I suppose you expect it to automatically change as you insert new rows into the table, which is not possible with this approach

You could write some code to generate this query dinamically based on the current data on the table, and put it in a trigger so the view is recreated every time a new row is inserted, but I’m not sure I would like to do something like that (I’m not sure I would have a table like that either…)

By: SAffer

SAffer — Mon, 27 Aug 2012 16:34:11 +0000

Hi Tom
It would most definately look like this e.g.
AL1
AL2
AL3
AK01
AK02 etc. - the column being 4 alpha characters long and either 3 or 4 characters used
to give AL1 AL2 AL3 AK01 AK02 AL1 AL2 BX01 BX01 etc.
I hope you can help – obviously the easiest would be a single command if there is one, otherwise a recurring structure without having to ‘hard code’ each column with the CASE command

By: TomLiotta

TomLiotta — Mon, 27 Aug 2012 07:16:51 +0000

Just to be sure, you have a table with single-column data:

Abc
Defghi
Jklmn

And you want a view that gives this result:

Abc Defghi Jklmn

Is that right?

Tom

By: SAffer

SAffer — Mon, 27 Aug 2012 07:05:23 +0000

Hi Carlos
The tag got missmatched somehow – SQL Server 2005.

By: carlosdl

carlosdl — Fri, 24 Aug 2012 19:00:32 +0000

What database platform (and version) are you using ?