Re: regproc's lack of certainty is dangerous

Tom Lane wrote:
> I think that we can actually get away (from an implementation point of
> view) with a column containing arrays of different base types; array_out
> will still work AFAIR. It's an interesting question though how such a
> column could reasonably be declared. This ties into your recent
> investigations into polymorphic array functions, perhaps.
>
> Maybe "anyarray" shouldn't be quite so pseudo a pseudotype?
>

More on this idea; here is a simple experiment:

regression=# update pg_type set typtype = 'b', typinput = 'array_in',
typoutput = 'array_out' where oid = 2277;
UPDATE 1
regression=# create table bar(f1 int, f2 anyarray);
CREATE TABLE
regression=# insert into bar values (1,'{1,2}'::integer[]);
INSERT 744428 1
regression=# insert into bar values (2,'{a,b}'::text[]);
INSERT 744429 1
regression=# select * from bar;
f1 | f2
----+-------
1 | {1,2}
2 | {a,b}
(2 rows)

Interesting ... but then there is:
regression=# select f1, f2[2] from bar;
ERROR: transformArraySubscripts: type anyarray is not an array

A bit more to do I guess.

Joe