select stuff((

select N', ' + cols.COLUMN_NAME as [text()]

from INFORMATION_SCHEMA.COLUMNS cols

Where cols.TABLE_NAME=...

I know, since they were created over 2 years ago, that I would probably implement parts of the scripts differently now (f.e. converting the password binaries into hex can probably...

I know, this was the only one I knew for sure where it was online. 🙂

In the attachment is the complete set of scripts.

- logins,

- server permissions

- server role members

-...

I had overlooked the contest comment (or better: I incorrectly interpreted the implications). Luckily I wasn't too complete in my answer :-D. Then again, all information you may need is...

Back in 2009 I created a set of scripts to generate T-SQL statements for all server and database principals and their rights from an existing server/database. The generated statements can...

You've got a cross join in your cte.

Change the recursive part like this and you're done:

SELECT m.ID,

DATEADD(dd,1,c.TheDATE),

...

Ah, then that's a different problem than I assumed it was: I was thinking you wanted the "greater than" symbol in your output, but found that it got filtered out...

The > is most likely exported by SQL server. It's probably one of the layers above that which filters the text from your output. You haven't however given us any...

Ok, a cross-tab implementation. It is faster and enables more possibilities than the Pivot implementation. Enough reasons to choose the cross-tab version. Even though I personally don't share the ideas...

You're looking for the "pivot" operator:

SELECT t.*

,x.*

FROM PS_TestForOnline t

CROSS APPLY (

select pvt.[1]

,pvt.[2]

...

Hi SwePeso,

Unless something drastic happens, this will be my last posting in this topic. I found and fixed a bug in the precedence detection and to my surprise your...

Thanks a lot Peso, I'll have a look at your "pattern-approach". My initial impression though is that it will not work when we encounter other expressions like for example a.col1...

I am not completely sure yet that I understand you correctly. Are you saying that there is another algorithm, producing slightly different RPN expressions -but still valid RPN-, which evaluate...

Hi SwePeso,

Thanks for the response, but I don't yet see how that helps me?

I think it doesn't help me because your RPN representation is not exactly the same expression any...

Searching for rows for which more than one with the same value exists by using a "having count() > 1" is going to be slower the bigger your data set...