File Coverage

=head1 NAME

Database::Abstraction - Read-only Database Abstraction Layer (ORM)

=head1 VERSION

Version 0.32

=cut

=head1 DESCRIPTION

C<Database::Abstraction> is a read-only database abstraction layer (ORM) for Perl,
designed to provide a simple interface for accessing and querying various types of databases such as CSV, XML, and SQLite without the need to write SQL queries.
It promotes code maintainability by abstracting database access logic into a single interface,
allowing users to switch between different storage formats seamlessly.
The module supports caching for performance optimization,
flexible logging for debugging and monitoring,
and includes features like the AUTOLOAD method for convenient access to database columns.
By handling numerous database and file formats,
C<Database::Abstraction> adds versatility and simplifies the management of read-intensive applications.

=head1 SYNOPSIS

Abstract class giving read-only access to CSV,
XML,
BerkeleyDB and SQLite databases via Perl without writing any SQL,
using caching for performance optimization.

The module promotes code maintainability by abstracting database access logic into a single interface.
Users can switch between different storage formats without changing application logic.
The ability to handle numerous database and file formats adds versatility and makes it useful for a variety of applications.

It's a simple ORM like interface which,
for all of its simplicity,
allows you to do a lot of the heavy lifting of simple database operations without any SQL.
It offers functionalities like opening the database and fetching data based on various criteria.

Built-in support for flexible and configurable caching improves performance for read-intensive applications.

Supports logging to debug and monitor database operations.

Look for databases in $directory in this order:

=over 4

=item 1 C<SQLite>

File ends with .sql

=item 2 C<PSV>

Pipe separated file, file ends with .psv

=item 3 C<CSV>

File ends with .csv or .db, can be gzipped. Note the default sep_char is '!' not ','

=item 4 C<XML>

File ends with .xml

=item 5 C<BerkeleyDB>

File ends with .db

=back

The AUTOLOAD feature allows for convenient access to database columns using method calls.
It hides the complexity of querying the underlying data storage.

If the table has a key column,
entries are keyed on that and sorts are based on it.
To turn that off, pass 'no_entry' to the constructor, for legacy
reasons it's enabled by default.
The key column's default name is 'entry', but it can be overridden by the 'id' parameter.

Arrays are made read-only before being returned.
To disable that, pass C<no_fixate> to the constructor.

CSV files that are not no_entry can have empty lines or comment lines starting with '#',
to make them more readable.

=head1 EXAMPLE

If the file /var/dat/foo.csv contains something like:

    "customer_id","name"
    "plugh","John"
    "xyzzy","Jane"

Create a driver for the file in .../Database/foo.pm:

    package Database::foo;

    use Database::Abstraction;

    our @ISA = ('Database::Abstraction');

    # Regular CSV: There is no entry column and the separators are commas
    sub new
    {
        my $class = shift;
        my %args = (ref($_[0]) eq 'HASH') ? %{$_[0]} : @_;

        return $class->SUPER::new(no_entry => 1, sep_char => ',', %args);
    }

You can then use this code to access the data via the driver:

    # Opens the file, e.g. /var/dat/foo.csv
    my $foo = Database::foo->new(directory => '/var/dat');

    # Prints "John"
    print 'Customer name ', $foo->name(customer_id => 'plugh'), "\n";

    # Prints:
    #  $VAR1 = {
    #     'customer_id' => 'xyzzy',
    #     'name' => 'Jane'
    #  };
    my $row = $foo->fetchrow_hashref(customer_id => 'xyzzy');
    print Data::Dumper->new([$row])->Dump();

=head1 SUBROUTINES/METHODS

=head2 init

Initializes the abstraction class and its subclasses with optional arguments for configuration.

    Database::Abstraction::init(directory => '../data');

See the documentation for new to see what variables can be set.

Returns a reference to a hash of the current values.
Therefore when given with no arguments you can get the current default values:

    my $defaults = Database::Abstraction::init();
    print $defaults->{'directory'}, "\n";

=cut

=head2 import

The module can be initialised by the C<use> directive.

    use Database::Abstraction 'directory' => '/etc/data';

or

    use Database::Abstraction { 'directory' => '/etc/data' };

=cut

=head2 new

Create an object to point to a read-only database.

Arguments:

Takes different argument formats (hash or positional)

=over 4

=item * C<auto_load>

Enable/disable the AUTOLOAD feature.
The default is to have it enabled.

=item * C<cache>

Place to store results

=item * C<cache_duration>

How long to store results in the cache (default is 1 hour).

=item * C<config_file>

Points to a configuration file which contains the parameters to C<new()>.
The file can be in any common format including C<YAML>, C<XML>, and C<INI>.
This allows the parameters to be set at run time.

=item * C<expires_in>

Synonym of C<cache_duration>, for compatibility with C<CHI>.

=item * C<dbname>

The prefix of name of the database file (default is name of the table).
The database will be held in a file such as $dbname.csv.

=item * C<directory>

Where the database file is held.
If only one argument is given to C<new()>, it is taken to be C<directory>.

=item * C<filename>

Filename containing the data.
When not given,
the filename is derived from the tablename
which in turn comes from the class name.

=item * C<logger>

Takes an optional parameter logger, which is used for warnings and traces.
Can be an object that understands warn() and trace() messages,
such as a L<Log::Log4perl> or L<Log::Any> object,
a reference to code,
or a filename.

=item * C<max_slurp_size>

CSV/PSV/XML files smaller than this are held in a HASH in RAM (default is 16K),
falling back to SQL on larger data sets.
Setting this value to 0 will turn this feature off,
thus forcing SQL to be used to access the database

=back

If the arguments are not set, tries to take from class level defaults.

Checks for abstract class usage.

Slurp mode assumes that the key column (entry) is unique.
If it isn't, searches will be incomplete.
Turn off slurp mode on those databases,
by setting a low value for max_slurp_size.

Clones existing objects with or without modifications.
Uses Carp::carp to log warnings for incorrect usage or potential mistakes.

=cut

=head2  set_logger

Sets the class, code reference, or file that will be used for logging.

=cut

=head2  selectall_arrayref

Returns a reference to an array of hash references of all the data meeting
the given criteria.

Note that since this returns an array ref,
optimisations such as "LIMIT 1" will not be used.

Use caching if that is available.

Returns undef if there are no matches.

=cut

=head2  selectall_hashref

Deprecated misleading legacy name for selectall_arrayref.

=cut

=head2  selectall_array

Similar to selectall_array but returns an array of hash references.

Con:    Copies more data around than selectall_arrayref
Pro:    Better determination of list vs scalar mode than selectall_arrayref by setting "LIMIT 1"

TODO:   Remove duplicated code

=cut

=head2  selectall_hash

Deprecated misleading legacy name for selectall_array.

=cut

=head2  count

Return the number items/rows matching the given criteria

=cut

=head2  fetchrow_hashref

Returns a hash reference for a single row in a table.

Special argument: table: determines the table to read from if not the default,
which is worked out from the class name

When no_entry is not set allow just one argument to be given: the entry value.

=cut