Re: general purpose array_sort

From: jian he <jian(dot)universality(at)gmail(dot)com>
To: Junwang Zhao <zhjwpku(at)gmail(dot)com>
Cc: "David G(dot) Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>, "andreas(at)proxel(dot)se" <andreas(at)proxel(dot)se>, Robert Haas <robertmhaas(at)gmail(dot)com>, Amit Langote <amitlangote09(at)gmail(dot)com>
Subject: Re: general purpose array_sort
Date: 2024-10-02 01:50:47
Message-ID: CACJufxGAQz5UoY_5jZqBMPBjcx-oau49yGAa_JMhdQRMaEKn5Q@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

> >
> > + typentry = (TypeCacheEntry *) fcinfo->flinfo->fn_extra;
> > + if (typentry == NULL || typentry->type_id != elmtyp)
> > + {
> > + typentry = lookup_type_cache(elmtyp, sort_asc ? TYPECACHE_LT_OPR :
> > TYPECACHE_GT_OPR);
> > + fcinfo->flinfo->fn_extra = (void *) typentry;
> > + }
> > you need to one-time check typentry->lt_opr or typentry->gt_opr exists?
> > see CreateStatistics.
> > /* Disallow data types without a less-than operator */
> > type = lookup_type_cache(attForm->atttypid, TYPECACHE_LT_OPR);
> > if (type->lt_opr == InvalidOid)
> > ereport(ERROR,
> > (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
> > errmsg("column \"%s\" cannot be used in
> > statistics because its type %s has no default btree operator class",
> > attname, format_type_be(attForm->atttypid))));
>
> I added an Assert for this part, not sure if that is enough.
>

i think it really should be:

if (typentry == NULL || typentry->type_id != elmtyp)
{
typentry = lookup_type_cache(elmtyp, sort_asc ? TYPECACHE_LT_OPR :
TYPECACHE_GT_OPR);
fcinfo->flinfo->fn_extra = (void *) typentry;
if ((sort_asc && !OidIsValid(typentry->lt_opr) || (!sort_as &&
OidIsValid(typentry->gt_opr));
ereport(ERROR,....)
}

Imagine a type that doesn't have TYPECACHE_LT_OPR or TYPECACHE_GT_OPR
then we cannot do the sort, we should just error out.

I just tried this colour type [1] with (CREATE TYPE colour (INPUT =
colour_in, OUTPUT = colour_out, LIKE = pg_catalog.int4);

select array_sort('{#FF0000, #FF0000}'::colour[]);
of course it will segfault with your new Assert.

[1] https://github.com/hlinnaka/colour-datatype/blob/master/colour.c

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Fujii Masao 2024-10-02 02:21:50 Re: Inconsistency in reporting checkpointer stats
Previous Message Michael Paquier 2024-10-02 00:52:59 Re: query_id, pg_stat_activity, extended query protocol