Re: lead() with arrays - strange behaviour - Mailing list pgsql-general

From Thomas Kellerer
Subject Re: lead() with arrays - strange behaviour
Date
Msg-id d4f4eb9d-2e00-899f-d9ba-e386d098021d@gmx.net
Whole thread Raw
In response to Re: lead() with arrays - strange behaviour  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-general
Tom Lane schrieb am 08.08.2019 um 16:10:
> David's point is that the two occurrences of lead() don't mean the
> same thing.  A window function is directly tied to the SELECT that
> it is in the select-list of, and its notion of next and previous
> rows is concerned with the set of rows that that SELECT's FROM-clause
> generates.  In this example, the inner SELECT has an empty FROM that
> returns one row, so the lead() in that SELECT doesn't do anything
> useful.

Ah! Now I get it ;) 

Thanks for clearing that up.

> You could probably get where you want to go with something along
> the lines of

Yes, that's what I did in the end (see my initial post)




pgsql-general by date:

Previous
From: Tom Lane
Date:
Subject: Re: lead() with arrays - strange behaviour
Next
From: Tony Shelver
Date:
Subject: Re: Recomended front ends?