binary operators on integers
I want to do simple and/or/xor/test operations on integers
but PostgreSQL seems not to support this. I found in conrib
area the 'bit' package, which supports 'SQL-compliant bitstrings'
but this is too weird for me and anyway it crashes too.
So, my question 'How can I do this?' Is there any plans to implement
those operators for integer types, or good reasons why they cannot
be implemented? Should I try to implement those myself or start
debugging the 'bit' package?
My situation: I want to implement a 'capabilities' field for
application server, so different bits are for different
capabilities.
TIA.
--
marko
Marko Kreen writes:
I want to do simple and/or/xor/test operations on integers
but PostgreSQL seems not to support this. I found in conrib
area the 'bit' package, which supports 'SQL-compliant bitstrings'
but this is too weird for me and anyway it crashes too.
These were the remainders of a slightly too late implementation attempt
for 7.0. In 7.1 they should be fully functional.
My situation: I want to implement a 'capabilities' field for
application server, so different bits are for different
capabilities.
Personally, I'd say using bit fields for that is pushing it a little too
hard. You could use shortint fields, or char(1)'s if you are not that
concerned about space.
--
Peter Eisentraut peter_e@gmx.net http://yi.org/peter-e/
On Mon, Sep 18, 2000 at 10:17:55PM +0200, Peter Eisentraut wrote:
Marko Kreen writes:
My situation: I want to implement a 'capabilities' field for
application server, so different bits are for different
capabilities.Personally, I'd say using bit fields for that is pushing it a little too
hard. You could use shortint fields, or char(1)'s if you are not that
concerned about space.
Hmm, probably you are right.
But now I am only curious: Will PostgreSQL support binary
arithmetics on ordinary integers someday or is the 'bit-string'
only way to go?
--
marko
Marko Kreen writes:
But now I am only curious: Will PostgreSQL support binary
arithmetics on ordinary integers someday or is the 'bit-string'
only way to go?
AFAIK, there's no one working on the former. Feel free to contribute. :-)
--
Peter Eisentraut peter_e@gmx.net http://yi.org/peter-e/