=head1 NAME

DBD::File::HowTo - Guide to create DBD::File based driver

=head1 SYNOPSIS

  perldoc DBD::File::HowTo
  perldoc DBI
  perldoc DBI::DBD
  perldoc DBD::File::Developers
  perldoc DBI::DBD::SqlEngine::Developers
  perldoc DBI::DBD::SqlEngine
  perldoc SQL::Eval
  perldoc DBI::DBD::SqlEngine::HowTo
  perldoc SQL::Statement::Embed
  perldoc DBD::File
  perldoc DBD::File::HowTo
  perldoc DBD::File::Developers

=head1 DESCRIPTION

This document provides a step-by-step guide, how to create a new
C<DBD::File> based DBD. It expects that you carefully read the L<DBI>
documentation and that you're familiar with L<DBI::DBD> and had read and
understood L<DBD::ExampleP>.

This document addresses experienced developers who are really sure that
they need to invest time when writing a new DBI Driver. Writing a DBI
Driver is neither a weekend project nor an easy job for hobby coders
after work. Expect one or two man-month of time for the first start.

Those who are still reading, should be able to sing the rules of
L<DBI::DBD/CREATING A NEW DRIVER>.

Of course, DBD::File is a DBI::DBD::SqlEngine and you surely read
L<DBI::DBD::SqlEngine::HowTo> before continuing here.

=head1 CREATING DRIVER CLASSES

Do you have an entry in DBI's DBD registry? For this guide, a prefix of
C<foo_> is assumed.

=head2 Sample Skeleton

    package DBD::Foo;

    use strict;
    use warnings;
    use vars qw(@ISA $VERSION);
    use base qw(DBD::File);

    use DBI ();

    $VERSION = "0.001";

    package DBD::Foo::dr;

    use vars qw(@ISA $imp_data_size);

    @ISA = qw(DBD::File::dr);
    $imp_data_size = 0;

    package DBD::Foo::db;

    use vars qw(@ISA $imp_data_size);

    @ISA = qw(DBD::File::db);
    $imp_data_size = 0;

    package DBD::Foo::st;

    use vars qw(@ISA $imp_data_size);

    @ISA = qw(DBD::File::st);
    $imp_data_size = 0;

    package DBD::Foo::Statement;

    use vars qw(@ISA);

    @ISA = qw(DBD::File::Statement);

    package DBD::Foo::Table;

    use vars qw(@ISA);

    @ISA = qw(DBD::File::Table);

    1;

Tiny, eh? And all you have now is a DBD named foo which will is able to
deal with temporary tables, as long as you use L<SQL::Statement>. In
L<DBI::SQL::Nano> environments, this DBD can do nothing.

=head2 Start over

Based on L<DBI::DBD::SqlEngine::HowTo>, we're now having a driver which
could do basic things. Of course, it should now derive from DBD::File
instead of DBI::DBD::SqlEngine, shouldn't it?

DBD::File extends DBI::DBD::SqlEngine to deal with any kind of files.
In principle, the only extensions required are to the table class:

    package DBD::Foo::Table;

    sub bootstrap_table_meta
    {
	my ( $self, $dbh, $meta, $table ) = @_;

	# initialize all $meta attributes which might be relevant for
	# file2table

	return $self->SUPER::bootstrap_table_meta($dbh, $meta, $table);
    }

    sub init_table_meta
    {
	my ( $self, $dbh, $meta, $table ) = @_;

	# called after $meta contains the results from file2table
	# initialize all missing $meta attributes

	$self->SUPER::init_table_meta( $dbh, $meta, $table );
    }

In case C<DBD::File::Table::open_file> doesn't open the files as the driver
needs that, override it!

    sub open_file
    {
	my ( $self, $meta, $attrs, $flags ) = @_;
	# ensure that $meta->{f_dontopen} is set
	$self->SUPER::open_file( $meta, $attrs, $flags );
	# now do what ever needs to be done
    }

Combined with the methods implemented using the L<SQL::Statement::Embed>
guide, the table is full working and you could try a start over.

=head2 User comfort

C<DBD::File> since C<0.39> consolidates all persistent meta data of a table
into a single structure stored in C<< $dbh->{f_meta} >>. While DBD::File
provides only readonly access to this structure, modifications are still
allowed.

Primarily DBD::File provides access via setters C<get_file_meta>,
C<set_file_meta> and C<clear_file_meta>. Those methods are easily
accessible by the users via the C<< $dbh->func () >> interface provided
by DBI. Well, many users don't feel comfortize when calling

    # don't require extension for tables cars
    $dbh->func ("cars", "f_ext", ".csv", "set_file_meta");

DBD::File will inject a method into your driver to increase the user
comfort to allow:

    # don't require extension for tables cars
    $dbh->foo_set_meta ("cars", "f_ext", ".csv");

Better, but here and there users likes to do:

    # don't require extension for tables cars
    $dbh->{foo_tables}->{cars}->{f_ext} = ".csv";

This interface is provided when derived DBD's define following in
C<init_valid_attributes> (please compare carefully with the example in
DBI::DBD::SqlEngine::HowTo):

    sub init_valid_attributes
    {
	my $dbh = $_[0];

	$dbh->SUPER::init_valid_attributes ();

	$dbh->{foo_valid_attrs} = {
	    foo_version 	=> 1,	# contains version of this driver
	    foo_valid_attrs	=> 1,	# contains the valid attributes of foo drivers
	    foo_readonly_attrs	=> 1,	# contains immutable attributes of foo drivers
	    foo_bar		=> 1,	# contains the bar attribute
	    foo_baz		=> 1,   # contains the baz attribute
	    foo_manager		=> 1,	# contains the manager of the driver instance
	    foo_manager_type	=> 1,	# contains the manager class of the driver instance
	    foo_meta            => 1,   # contains the public interface to modify table meta attributes
	};
	$dbh->{foo_readonly_attrs} = {
	    foo_version 	=> 1,	# ensure no-one modifies the driver version
	    foo_valid_attrs	=> 1,	# do not permit to add more valid attributes ...
	    foo_readonly_attrs	=> 1,	# ... or make the immutable mutable
	    foo_manager		=> 1,	# manager is set internally only
	    foo_meta            => 1,   # ensure public interface to modify table meta attributes are immutable
	};

	$dbh->{foo_meta} = "foo_tables";

	return $dbh;
    }

This provides a tied hash in C<< $dbh->{foo_tables} >> and a tied hash for
each table's meta data in C<< $dbh->{foo_tables}->{$table_name} >>.
Modifications on the table meta attributes are done using the table
methods:

    sub get_table_meta_attr { ... }
    sub set_table_meta_attr { ... }

Both methods can adjust the attribute name for compatibility reasons, e.g.
when former versions of the DBD allowed different names to be used for the
same flag:

    my %compat_map = (
		       abc => 'foo_abc',
		       xyz => 'foo_xyz',
		     );
    __PACKAGE__->register_compat_map( \%compat_map );

If any user modification on a meta attribute needs reinitialization of
the meta structure (in case of C<DBD::File> these are the attributes
C<f_file>, C<f_dir>, C<f_ext> and C<f_lockfile>), inform DBD::File by
doing 

    my %reset_on_modify = (
			    foo_xyz => "foo_bar",
			    foo_abc => "foo_bar",
			  );
    __PACKAGE__->register_reset_on_modify( \%reset_on_modify );

The next access to the table meta data will force DBD::File to re-do the
entire meta initialization process.

Any further action which needs to be taken can handled in
C<table_meta_attr_changed>:

    sub table_meta_attr_changed
    {
	my ($class, $meta, $attrib, $value) = @_;
	...
	$class->SUPER::table_meta_attr_changed ($meta, $attrib, $value);
    }

This is done before the new value is set in C<$meta>, so the attribute
changed handler can act depending on the old value.

=head2 Testing

Now you should have your own DBD::File based driver. Was easy, wasn't it?
But does it work well?  Prove it by writing tests and remember to use
dbd_edit_mm_attribs from L<DBI::DBD> to ensure testing even rare cases.

=head1 AUTHOR

This guide is written by Jens Rehsack. DBD::File is written by Jochen
Wiedmann and Jeff Zucker.

The module DBD::File is currently maintained by

H.Merijn Brand < h.m.brand at xs4all.nl > and
Jens Rehsack  < rehsack at googlemail.com >

=head1 COPYRIGHT AND LICENSE

Copyright (C) 2010 by H.Merijn Brand & Jens Rehsack

All rights reserved.

You may freely distribute and/or modify this module under the terms of
either the GNU General Public License (GPL) or the Artistic License, as
specified in the Perl README file.

=cut