Re: [BUGS] Bug in create operator and/or initdb

#1

"John Hansen" <john (AT) geeknet (DOT) com.au> writes:

Quote:

CREATE FUNCTION my_func (inet, inet) as '$libdir/my_func.so' LANGUAGE 'C' IMMUTABLE STRICT;
CREATE OPERATOR <<< (
PROCEDURE = my_func,
LEFTARG = cidr,
RIGHTARG = cidr,
RESTRICT = contsel,
JOIN = contjoinsel
);

ERROR: function my_func(cidr, cidr) does not exist

Right ...

Quote:

Now, if you look at the catalog, and the < (less than operator) as an example you will see that:

Two operators are defined for < - one for inet,inet and another for cidr,cidr.
Only one function exists named network_lt, and is declared as taking (inet,inet) as arguments.

My opinion is that this is a very bogus shortcut in the network datatype
code. There are no cases outside the inet/cidr group where an operator
doesn't exactly match its underlying function. (The whole business of
inet and cidr being almost but not quite the same type is maldesigned
anyway...)

The right solution for you is to declare two SQL functions. Whether you
make them point at the same underlying C code is up to you.

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?

http://archives.postgresql.org

#2

Quote:

My opinion is that this is a very bogus shortcut in the=20
network datatype code. There are no cases outside the=20
inet/cidr group where an operator doesn't exactly match its=20
underlying function. (The whole business of inet and cidr=20
being almost but not quite the same type is maldesigned
anyway...)
=20
The right solution for you is to declare two SQL functions.=20=20
Whether you make them point at the same underlying C code is=20
up to you.

Right,...

In that case may I suggest fixing the catalog so network_* functions exists=
for both datatypes!
Anything less I'd consider inconsistent...

Kind regards,

John

---------------------------(end of broadcast)---------------------------
TIP 3: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to majordomo (AT) postgresql (DOT) org so that your
message can get through to the mailing list cleanly

#3

"John Hansen" <john (AT) geeknet (DOT) com.au> writes:

Quote:

In that case may I suggest fixing the catalog so network_* functions exists for both datatypes!

Redesigning the inet/cidr distinction is on the to-do list (though I'm
afraid not very high on the list). ISTM it should either be one type
with a distinguishing bit in the runtime representation, or two types
with no such bit needed. Having both is a schizophrenic design. It's
led directly to bugs in the past, and I think there are still some
corner cases that act oddly (see the archives).

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 9: the planner will ignore your desire to choose an index scan if your
joining column's datatypes do not match

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Re: [BUGS] Bug in create operator and/or initdb

mailing.database.pgsql-bugs mailing.database.pgsql-bugs

Re: [BUGS] Bug in create operator and/or initdb - 01-29-2005 , 08:56 PM

Re: [BUGS] Bug in create operator and/or initdb - 01-29-2005 , 08:57 PM

Re: [BUGS] Bug in create operator and/or initdb - 01-29-2005 , 09:18 PM