NAME

DBIx::DR - easy DBI helper (perl inside SQL and blessed results)

SYNOPSIS

my $dbh = DBIx::DR->connect($dsn, $login, $passed);

$dbh->perform(
    'UPDATE tbl SET a = 1 WHERE id = <%= $id %>',
    id => 123
);

my $rowset = $dbh->select(
    'SELECT * FROM tbl WHERE id IN (<% list @$ids %>)',
    ids => [ 123, 456 ]
);
my $rowset = $dbh->select(-f => 'sqlfile.sql.ep', ids => [ 123, 456 ]);

while(my $row = $rowset->next) {
    print "id: %d, value: %s\n", $row->id, $row->value;
}

DESCRIPTION

The package extends DBI and allows You:

  • to use perl inside Your SQL requests;

  • to bless resultsets into Your package;

  • to place Your SQL's into dedicated directory;

  • to use usual DBI methods.

Additional 'connect' options.

dr_iterator

A string describes iterator class. Default value is 'dbix-dr-iterator#new' (decamelized string).

dr_item

A string describes item (one row) class. Default value is 'dbix-dr-iterator-item#new' (decamelized string).

dr_sql_dir

Directory path to seek sql files (If You use dedicated SQLs).

dr_decode_errors

Decode database errors into utf-8

METHODS

All methods can receive the following arguments:

-f => $sql_file_name

It will load SQL-request from file. It will seek file in directory that was defined in dr_sql_dir param of connect.

You needn't to use suffixes (.sql.ep) here, but You can.

-item => 'decamelized_obj_define'

It will bless (or construct) row into specified class. See below. Default value defined by dr_item argument of DBI::connect.

-noitem

Do not bless row into any class.

-iterator => 'decamelized_obj_define'

It will bless (or construct) rowset into specified class. Default value defined by dr_iterator argument of DBI::connect.

-noiterator

Do not bless rowset into any class.

-dbi => HASHREF

Additional DBI arguments.

-hash => FIELDNAME

Selects into HASH. Iterator will operate by names (not numbers).

-die => 0|1

If true the method will die with SQL-request.

Decamelized strings

Are strings that represent class [ and method ].

foo_bar                => FooBar
foo_bar#subroutine     => FooBar->subroutine
foo_bar-baz            => FooBar::Baz

perform

Does SQL-request like 'UPDATE', 'INSERT', etc.

$dbh->perform($sql, value => 1, other_value => 'abc');
$dbh->perform(-f => $sql_file_name, value => 1, other_value => 'abc');

select

Does SQL-request, pack results into iterator class. By default it uses DBIx::DR::Iterator class.

my $res = $dbh->select(-f => $sql_file_name, value => 1);
while(my $row = $res->next) {
    printf "RowId: %d, RowValue: %s\n", $row->id, $row->value;
}

my $row = $row->get(15);  # row 15

my $res = $dbh->select(-f => $sql_file_name,
        value => 1, -hash => 'name');
while(my $row = $res->next) {
    printf "RowId: %d, RowName: %s\n", $row->id, $row->name;
}

my $row = $row->get('Vasya');  # row with name eq 'Vasya'

single

Does SQL-request that returns one row. Pack results into item class. Does SQL-request, pack results (one row) into item class. By default it uses DBIx::DR::Iterator::Item class.

Template language

You can use perl inside Your SQL requests:

% my $foo = 1;
% my $bar = 2;
<% my $foo_bar = $foo + $bar %>

..

% use POSIX;
% my $gid = POSIX::getgid;

There are two functions available inside perl:

quote

Replaces argument to '?', add argument value into bindlist. You can also use shortcut '=' instead of the function.

Example 1

SELECT
    *
FROM
    tbl
WHERE
    id = <% quote $id %>

Result

SELECT
    *
FROM
    tbl
WHERE
    id = ?

and bindlist will contain id value.

If You use DBIx::DR::ByteStream in place of string the function will recall immediate function.

Example 2

SELECT
    *
FROM
    tbl
WHERE
    id = <%= $id %>

immediate

Replaces argument to its value. You can also use shortcut '==' instead of the function.

Example 1

SELECT
    *
FROM
    tbl
WHERE
    id = <% immediate $id %>

Result

SELECT
    *
FROM
    tbl
WHERE
    id = 123

Where 123 is id value.

Be carful! Using the operator You can produce code that will be amenable to SQL-injection.

Example 2

SELECT
    *
FROM
    tbl
WHERE
    id = <%== $id %>

Helpers

There are a few default helpers.

list

Expands array into Your SQL request.

Example

SELECT
    *
FROM
    tbl
WHERE
    status IN (<% list @$ids %>)

Result

SELECT
    *
FROM
    tbl
WHERE
    status IN (?,?,? ...)

and bindlist will contain ids values.

hlist

Expands array of hash into Your SQL request. The first argument can be a list of required keys. Places each group into brackets.

Example

INSERT INTO
    tbl
        ('a', 'b')
VALUES
    <% hlist ['a', 'b'] => @$inserts

Result

INSERT INTO
    tbl
        ('a', 'b')
VALUES
    (?, ?), (?, ?) ...

and bindlist will contain all inserts values.

include

Includes the other SQL-part.

Example

% include 'other_sql', argument1 => 1, argument2 => 2;

User's helpers

You can add Your helpers using method set_helper.

set_helper

Sets (or replaces) helpers.

$dbh->set_helper(foo => sub { ... }, bar => sub { ... });

Each helper receives template object as the first argument.

Examples:

$dbh->set_helper(foo_AxB => sub {
    my ($tpl, $a, $b) = @_;
    $tpl->quote($a * $b);
});

You can use quote and immediate functions inside Your helpers.

If You want use the other helper inside Your helper You have to do that by Yourself. To call the other helper You can also use $tpl->call_helper function.

call_helper

$dbh->set_helper(
    foo => sub {
        my ($tpl, $a, $b) = @_;
        $tpl->quote('foo' . $a . $b);
    },
    bar => sub {
        my $tpl = shift;
        $tpl->call_helper(foo => 'b', 'c');
    }
);

COPYRIGHT

Copyright (C) 2011 Dmitry E. Oboukhov <unera@debian.org>
Copyright (C) 2011 Roman V. Nikolaev <rshadow@rambler.ru>

This program is free software, you can redistribute it and/or
modify it under the terms of the Artistic License.

VCS

The project is placed git repo on github: https://github.com/unera/dbix-dr/