Chapman Flack <chap@anastigmatix.net> writes:
> Where should I look to learn more about the capabilities of the type
> inference done in planning?
It's not terribly well documented outside the source code, I fear.
> Not everything-there-is-to-know, but basics
> like, can some simple query constructs reliably cause an intended type to
> be reported? I assume if the query is "SELECT ?" then the reported argtype
> will be unknown. If changed to, say, "SELECT ?::foo", will that report
> a type of foo for the parameter, or continue to report the parameter
> as untyped because it isn't needed to type the cast expression?
That will result in reporting the parameter as having type foo, cf
variable_coerce_param_hook(). Experimenting with this sort of stuff
isn't hard, eg in psql:
regression=# prepare p1 as select $1;
PREPARE
regression=# prepare p2 as select $1::bigint;
PREPARE
regression=# select name, parameter_types from pg_prepared_statements ;
name | parameter_types
------+-----------------
p1 | {text}
p2 | {bigint}
(2 rows)
(Before v10, p1 would have failed PREPARE for lack of a determinate
type for the parameter, which I think corresponds to sending back
UNKNOWN in the wire-protocol-prepare case.)
regards, tom lane