Re: Dynamically update NEW columns in plpgsql trigger - Mailing list pgsql-general

From Pavel Stehule
Subject Re: Dynamically update NEW columns in plpgsql trigger
Date
Msg-id AANLkTikOmX0pZm-zBZNCjcFR5cCy+SP44k=NopqN_Xh5@mail.gmail.com
Whole thread Raw
In response to Dynamically update NEW columns in plpgsql trigger  (Nick <nboutelier@gmail.com>)
List pgsql-general
Hello

plpgsql isn't good tool for this. use a plperl or plpython instead.
your solution is extremly slow.

Regards

Pavel Stehule

2010/9/9 Nick <nboutelier@gmail.com>:
> I need to dynamically update NEW columns. Ive been inserting the NEW
> values into a temp table, updating them, then passing the temp table
> values back to NEW (is there a better way?). Ive had success with this
> method unless there is a null value...
>
> EXECUTE 'CREATE TEMP TABLE new AS SELECT $1.*' USING NEW;
> EXECUTE 'UPDATE new SET ' || dyn_col_name || ' = 1';
> EXECUTE 'SELECT * FROM new' INTO NEW;
> EXECUTE 'DROP TABLE new';
>
> This last line...
> EXECUTE 'SELECT $1.' || dyn_col_name INTO val USING NEW;
>
> gives the ERROR: could not identify column "col_one" in record data
> type.
>
> However RAISE EXCEPTION '%',NEW.col_one;
> returns "1" correctly.
>
> If col_one does does not start out as a null value, then everything
> works. Why does the passing from temp table back to NEW lose the USING
> functionality?
>
> --
> Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-general
>

pgsql-general by date:

Previous
From: Nick
Date:
Subject: Dynamically update NEW columns in plpgsql trigger
Next
From: Dmitriy Igrishin
Date:
Subject: Re: Dynamically update NEW columns in plpgsql trigger