Re: "fix" for plpgsql polymorphism

From: Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us>
To: Joe Conway <mail(at)joeconway(dot)com>
Cc: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, "Patches (PostgreSQL)" <pgsql-patches(at)postgresql(dot)org>
Subject: Re: "fix" for plpgsql polymorphism
Date: 2003-07-20 17:52:32
Message-ID: 200307201752.h6KHqWt03180@candle.pha.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-patches


Your patch has been added to the PostgreSQL unapplied patches list at:

http://momjian.postgresql.org/cgi-bin/pgpatches

I will try to apply it within the next 48 hours.

---------------------------------------------------------------------------

Joe Conway wrote:
> Tom Lane wrote:
> >>You can alias $0, similar to the argument variables. And, I confirmed
> >>that you cannot change the value, similar to the argument variables:
> >
> > Perhaps you shouldn't mark it isconst; then it would actually have some
> > usefulness (you could use it directly as a temporary variable to hold
> > the intended result). I can't see much value in aliasing it if it's
> > const, either.
>
> OK; the only change in this version is "isconst = false;". Now you can
> use $0 as a result placeholder if desired. E.g.:
>
> create or replace function tmp(anyelement, anyelement) returns anyarray as '
> declare
> v_ret alias for $0;
> v_el1 alias for $1;
> v_el2 alias for $2;
> begin
> v_ret := ARRAY[v_el1, v_el2];
> return v_ret;
> end;
> ' language 'plpgsql';
>
> create table f(f1 text, f2 text, f3 int, f4 int);
> insert into f values ('a','b',1,2);
> insert into f values ('z','x',3,4);
>
> select tmp(f1,f2) from f;
> select tmp(f3,f4) from f;
>
> Joe

> Index: src/pl/plpgsql/src/pl_comp.c
> ===================================================================
> RCS file: /opt/src/cvs/pgsql-server/src/pl/plpgsql/src/pl_comp.c,v
> retrieving revision 1.59
> diff -c -r1.59 pl_comp.c
> *** src/pl/plpgsql/src/pl_comp.c 1 Jul 2003 21:47:09 -0000 1.59
> --- src/pl/plpgsql/src/pl_comp.c 3 Jul 2003 17:59:48 -0000
> ***************
> *** 354,359 ****
> --- 354,395 ----
> function->fn_rettyplen = typeStruct->typlen;
> function->fn_rettypelem = typeStruct->typelem;
> perm_fmgr_info(typeStruct->typinput, &(function->fn_retinput));
> +
> + /*
> + * install $0 reference, but only for polymorphic
> + * return types
> + */
> + if (procStruct->prorettype == ANYARRAYOID ||
> + procStruct->prorettype == ANYELEMENTOID)
> + {
> + char buf[32];
> +
> + /* name for variable */
> + snprintf(buf, sizeof(buf), "$%d", 0);
> +
> + /*
> + * Normal return values get a var node
> + */
> + var = malloc(sizeof(PLpgSQL_var));
> + memset(var, 0, sizeof(PLpgSQL_var));
> +
> + var->dtype = PLPGSQL_DTYPE_VAR;
> + var->refname = strdup(buf);
> + var->lineno = 0;
> + var->datatype = build_datatype(typeTup, -1);
> + var->isconst = false;
> + var->notnull = false;
> + var->default_val = NULL;
> +
> + /* preset to NULL */
> + var->value = 0;
> + var->isnull = true;
> + var->freeval = false;
> +
> + plpgsql_adddatum((PLpgSQL_datum *) var);
> + plpgsql_ns_additem(PLPGSQL_NSTYPE_VAR, var->varno,
> + var->refname);
> + }
> }
> ReleaseSysCache(typeTup);
>

>
> ---------------------------(end of broadcast)---------------------------
> TIP 4: Don't 'kill -9' the postmaster

--
Bruce Momjian | http://candle.pha.pa.us
pgman(at)candle(dot)pha(dot)pa(dot)us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073

In response to

Browse pgsql-patches by date

  From Date Subject
Next Message Bruce Momjian 2003-07-21 03:39:04 Re: An additional foreign key test for regression
Previous Message Bruce Momjian 2003-07-20 17:16:16 Re: timestamp date_trunc('quarter',...)