Hi,
I've been a CPAN consumer for so long, I thought it's about time to start contributing.
I wrote a little module that does exactly the same as File::CounterFile, but with a database as storage engine.
It came into this world after a thread on www.experts-exchange.com; so there's at least one interested user for it.
Granted, it's nothing earth-shattering, but it's simple and generic enough to be useful outside my usual work, and it presents me with a manageable package to start my CPAN career.
Before I go all the way by applying for a PAUSE account and uploading it, I would like to put it to public scrutiny first. I'd like to know I'm doing things properly from the start.
Here's my list of questions:
- what about the module name? is it good?
- code review would be nice
- documentation: clear enough?
- tests: for a module this simple?
- exception handling: should I do this myself, or let DBI throw them?
The archive is here: http://www.rhesa.com/DBIx-Counter-0.01.tar.gz
Thanks for your attention!
Rhesa
Here's the pod:
NAME
DBIx::Counter - Manipulate named counters stored in a database
SYNOPSIS
use DBIx::Counter;
$c = new DBIx::Counter('my counter',
dsn => 'dbi:mysql:mydb',
login => 'username',
password => 'secret'
);
print $c->value;
$c->inc;
DESCRIPTION
This module creates and maintains named counters in a database. It has a
simple interface, with methods to increment and decrement the counter by
one, and a method for retrieving the value. It supports operator
overloading for increment (++), decrement (--) and stringification ("").
It should perform well in persistent environments, since it uses the
connect_cached and prepare_cached methods of DBI. The biggest advantage
over its main inspiration - File::CounterFile - is that it allows
distributed, concurrent access to the counters and isn't tied to a
single file system.
Connection settings can be set in the constructor, or by using the
package variables $DSN, $LOGIN and $PASSWORD and $TABLENAME. The table
name is configurable, but the column names are currently hard-coded to
counter_id and value. The following SQL statement can be used to create
the table:
CREATE TABLE counters (
counter_id varchar(64) primary key,
value int not null default 0
);
This module attempts to mimick the File::CounterFile interface, except
currently it only supports integer counters. The locking functions in
File::CounterFile are present for compatibility only: they always return
0.
METHODS
new Creates a new counter instance. First parameter is the required
counter name. Second, optional, argument is an initial value for the
counter on its very first use. It also accepts named parameters for
the dbi connection string, dbi login and dbi password, and the table
name:
dsn - overrides $DBIx::Counter::DSN
login - overrides $DBIx::Counter::LOGIN
password - overrides $DBIx::Counter::PASSWORD
tablename - overrides $DBIx::Counter::TABLENAME
Examples:
$c = new DBIx::Counter('my counter');
$c = new DBIx::Counter('my counter',
dsn => 'dbi:mysql:mydb',
login => 'username',
password => 'secret'
);
$c = new DBIx::Counter('my counter',
42,
dsn => 'dbi:mysql:mydb',
tablename => 'gauges'
);
inc increases the counter by one.
$c->inc;
# or using overload:
$c++;
dec decreases the counter by one.
$c->dec;
# or using overload:
$c--;
value
returns the current value of the counter.
print $c->value;
# or using overload:
print $c;
SEE ALSO
File::CounterFile
AUTHOR
Rhesa Rozendaal,
COPYRIGHT AND LICENSE
Copyright (C) 2005 by Rhesa Rozendaal
This library is free software; you can redistribute it and/or modify it
under the same terms as Perl itself, either Perl version 5.8.2 or, at
your option, any later version of Perl 5 you may have available.
Re:SQL sequence?
rhesa on 2005-04-14T09:50:34
I suppose it could be used for sequences in databases that do not have those (such as Mysql, SQLLite and SQL Server), but the main difference is that this module doesn't care if you decrement a counter.
The fact that it uses a database for storage, doesn't mean that it can only be used in a database context. You can count things, that's it.
It's meant as a (drop in) replacement of File::CounterFile. My module gives you an (well, almost) identical API, but the storage is centralised. That means you can access the same counter from multiple hosts (provided your database is configured to allow that, of course).
