/ 
SPOPS-0.75
River stage one • 2 direct dependents • 2 total dependents
/ SPOPS::DBI::Pg

NAME

SPOPS::DBI::Pg -- PostgreSQL-specific routines for the SPOPS::DBI

SYNOPSIS

# In your configuration:

'myspops' => {
    'isa' => [ qw/ SPOPS::DBI::Pg SPOPS::DBI / ],

    # If you have a SERIAL field, just set increment_field to a true
    # value

    'increment_field' => 1,

    # If you want to specify the name of your sequence (whether using
    # a SERIAL field or not):

    'sequence_name'   => 'myseq',
    ...
},

DESCRIPTION

This just implements some Postgres-specific routines so we can abstract them out.

One of them optionally returns the sequence value of the just-inserted id field. Of course, this only works if you have a the field marked as 'SERIAL' or using a sequence value in your table:

CREATE TABLE my_table (
  id  SERIAL,
  ...
)

or

CREATE TABLE my_table (
  id int not null primary key,
  ...
);

CREATE SEQUENCE myobject_sequence;

You must to let this module know if you are using this option by setting in your class configuration the key 'increment_field' to a true value:

$spops = {
   myobj => {
      class => 'My::Object',
      isa   => [ qw/ SPOPS::DBI::Pg  SPOPS::DBI / ],
      increment_field => 1,
      ...
   },
};

If you use the 'SERIAL' datatype then you do not have to specify a sequence name. Otherwise you need to tell SPOPS what sequence to use in the class configuration:

$spops = {
   myobj => {
      class           => 'My::Object',
      isa             => [ qw/ SPOPS::DBI::Pg  SPOPS::DBI / ],
      increment_field => 1,
      sequence_name   => 'myobject_sequence',
   },
};

NOTE: The name automatically created by PostgreSQL when you use the 'SERIAL' datatype follows a certain convention ($table-$idfield-seq). But if the table or ID field are too long, PostgreSQL will truncate the name so it will fit in the 32-character limit for symbols. In this case you will either need to recompile PostgreSQL (yuck) or list the sequence name in the class configuration. See a message from the openinteract-help mailing list at:

http://www.geocrawler.com/archives/3/8429/2002/1/0/7551783/

for more information on recompiling if you are so inclined.

METHODS

sql_current_date()

Returns 'CURRENT_TIMESTAMP()', used in PostgreSQL to return the value for right now.

sql_quote( $value, $data_type, [ $db_handle ] )

DBD::Pg depends on the type of a field if you are quoting values to put into a statement, so we override the default 'sql_quote' from SPOPS::SQLInterface to ensure the type of the field is used in the DBI->quote call.

The $data_type should correspond to one of the DBI datatypes (see the file 'dbi_sql.h' in your Perl library tree for more info). If the DBI database handle $db_handle is not passed in, we try to find it with the class method global_datasource_handle().

pre_fetch_id( \%params )

If 'increment_field' is not set we do not fetch an ID. If 'sequence_name' is not also set we do not fetch an ID, assuming that you have defined the ID field using the 'SERIAL' datatype.

post_fetch_id( \%params )

Retrieve the value just put into the database for the ID field. To use this you must in the configuration for your object set 'increment_field' to a true value and either specify a 'sequence_name' or use the SERIAL-default name of:

<table_name>_<id_field_name>_seq

This is the sequence created by default when you use the 'SERIAL' datatype.

BUGS

None known.

TO DO

Nothing known.

SEE ALSO

DBD::Pg

DBI

COPYRIGHT

Copyright (c) 2001-2002 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.

AUTHORS

Chris Winters <chris@cwinters.com>