package SPOPS::Import::DBI::TableTransform::Pg;
# $Id: Pg.pm,v 3.4 2004/06/02 00:48:23 lachoy Exp $
use strict;
use base qw( SPOPS::Import::DBI::TableTransform );
$SPOPS::Import::DBI::TableTransform::Pg::VERSION = sprintf("%d.%02d", q$Revision: 3.4 $ =~ /(\d+)\.(\d+)/);
sub increment {
my ( $self, $sql ) = @_;
$$sql =~ s/%%INCREMENT%%/INT/g;
}
sub increment_type {
my ( $self, $sql ) = @_;
$$sql =~ s/%%INCREMENT_TYPE%%/INT/g;
}
sub datetime {
my ( $self, $sql ) = @_;
$$sql =~ s/%%DATETIME%%/TIMESTAMP/g;
}
1;
__END__
=head1 NAME
SPOPS::Import::DBI::TableTransform::Pg - Table transformations for PostgreSQL
=head1 SYNOPSIS
my $table = qq/
CREATE TABLE blah ( id %%INCREMENT%% primary key,
name varchar(50) )
/;
my $transformer = SPOPS::Import::DBI::TableTransform->new( 'pg' );
$transformer->increment( \$table );
print $table;
# Output:
# CREATE TABLE blah ( id INT primary key,
# name varchar(50) )
=head1 DESCRIPTION
PostgreSQL-specific type conversions for the auto-increment and other
field types.
=head1 METHODS
B<increment>
Returns 'INT NOT NULL' -- relying on the sequence autocreated by
'SERIAL' can get you into trouble since long table names get
truncated. Just create your own sequence and specify it in the
'sequence_name' key of your object config (see
L<SPOPS::DBI::Pg|SPOPS::DBI::Pg>).
B<increment_type>
Returns 'INT'
B<datetime>
Returns 'TIMESTAMP'
=head1 BUGS
None known.
=head1 TO DO
B<Add hook for extra statement>
Since PostgreSQL supports a sequence-based increment type, think about
adding a hook for an extra statement to be registered and modifying
'%%INCREMENT%%' to be 'INT NOT NULL' and the extra statement to create
a sequence of a given name.
=head1 SEE ALSO
L<SPOPS::Import::DBI::TableTransform|SPOPS::Import::DBI::TableTransform>
=head1 COPYRIGHT
Copyright (c) 2002-2004 intes.net, inc.. All rights reserved.
This library is free software; you can redistribute it and/or modify
it under the same terms as Perl itself.
=head1 AUTHORS
Chris Winters E<lt>chris@cwinters.comE<gt>