| From: | "David G(dot) Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com> |
|---|---|
| To: | Jim Nasby <Jim(dot)Nasby(at)bluetreble(dot)com> |
| Cc: | pgsql-general <pgsql-general(at)postgresql(dot)org> |
| Subject: | Re: "Cast" SRF returning record to a table type? |
| Date: | 2015-04-18 00:39:27 |
| Message-ID: | CAKFQuwZTEBBoK_AgeD3oCUbAU11RH4kUiWcTdtt7pa1OogQ-HA@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
On Friday, April 17, 2015, Jim Nasby <Jim(dot)Nasby(at)bluetreble(dot)com> wrote:
> I'm working on a function that will return a set of test data, for unit
> testing database stuff. It does a few things, but ultimately returns SETOF
> record that's essentially:
>
> RETURN QUERY EXECUTE 'SELECT * FROM ' || table_name;
>
> Because it's always going to return a real relation, I'd like to be able
> to the equivalent of:
>
> SELECT ... FROM my_function( 'some_table' )::some_table;
Unfortunately this means "cast the existing type to some_table" and
"record" is not a valid type in this context.
>
> Is there any trick that would allow that to work? I know that instead of
> 'SELECT * ...' I can do 'SELECT row(t.*) FROM ' || table_name || ' AS t'
> and then do
>
> SELECT ... FROM my_function( 'some_table' ) AS data( d some_table )
>
> but I'm hoping to avoid the extra level of indirection.
>
>
Haven't explored this specific code in depth...but which part - the
function alias or the select row(t.*)? They seem to be independent
concerns.
David J.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Jim Nasby | 2015-04-18 00:41:54 | Re: Waiting on ExclusiveLock on extension |
| Previous Message | Jim Nasby | 2015-04-18 00:20:26 | "Cast" SRF returning record to a table type? |