NAME

SQLite::Abstract - Object-oriented wrapper for SQLite

SYNOPSIS

 use SQLite::Abstract;

 
 my $sql = SQLite::Abstract->new($dbname);
 
 my $table_name = "smt";
 
 $SQLite::Abstract::glob->{'default_table'} = $table_name;
 
 my $table = {
 	'struct', [
		'phone',  [qw(INTEGER(32) NOT NULL)],
		'name',   [qw(VARCHAR(512) NOT NULL)],
		'address',[qw(VARCHAR(1024) NOT NULL)]
	]
 };

 $sql->create($table);

 my ($phone, $name, $address, $data, %unique);

 ...

 while (<FH>){
 	($name, $phone, $address) = split ';',$_ and
	push @$data, [$phone, $name, $address];
 }

 $sql->insert($data);

 
 #then select, update, insert again and so on ...
 
 my $what = "where phone like %6% and name like %reni%";
 my $search = { 'where' => $what, 'col' => 'name, phone' };
 my $result = $sql->search($search);

 #or just SELECT * FROM sm_table_name
 
 $sql->search({});

 
 my $update = { 'name set', 'RENI where name like %reni%' };
 my $result = $sql->update($update);

 #insert after delete - SQLite is fast enough
 
 $sql->delete_insert($data);
 
 
 

DESCRIPTION

SQLite::Abstract is just another try to wrap sql and to be more concrete - SQLite. Primary goals are ease and speed in development of sql front-end with the excellent DBD::SQLite.

METHODS

Each method works into a single transaction.

$sql = SQLite::Abstract->new( $dbname );

Object creation expects database name in order to init DB connection.

$sql->search( $search )

Where $search must be e hash reference, containing 'where' and 'col' keys which are optional. Each key's name may be changed through the global %{$glob}:

'where'

speciafies which rows you want

'col'

specifies the columns you want

In brief

$SQLite::Abstract::glob->{'where'} = 'what';
$what   = "where phone like %6% and name like %reni%";
$search = { 'what' => $what, 'col' => 'name, phone' };
$result = $sql->search($search);

where $result is print $_->[0], "\t", $_->[1], "\n" for @$result;

$sql->select( { } )

Synonym for search

$sql->update( { } )

This methos expects hash ref where the only one key is a column name and the value is WHERE clause. The key may be in comapy with 'set':

$update = { 'name set', 'RENI where name like %reni%' }; #or
$update = { 'set name', 'RENI where name like %reni%' }; #or
$update = { 'name', 'RENI where name like %reni%' };

$result = $sql->update($update);

print "$result rows updates\n";

$sql->delete( { } )

The same mthod arguments as insert method except that the key does not have any special meaning - WHERE clause and value which contains the actual sql code:

#for more comfort
$SQLite::Abstract::glob->{'where'} = 'remove';

$delete = { 'remove', 'where name like %myself%' };
$result = $sql->delete($delete);

print "$result rows deleted\n";

$sql->insert( @$data )

Where the array must contain the same number of columns as the table

$sql->delete_insert( @$data )

The same methos as 'insert' except that DELETE the table before INSERT And because it's SQLite - speed is amazing.

$sql->create( $table )

This methos needs table structure and eventually table name unless another global var is not set (by default):

$table_name = 'somewhere';
$SQLite::Abstract::glob->{'default_table'} = $table_name;

Than the table structure where the key 'struct' is also modules' $glob value

   $SQLite::Abstract::glob->{'struct'} = 'structure';
   $table = {
   	'tablename', $table_name,
   	'structure', [
		'column_1', [qw( column_1 properties )],
		'column_2', [qw( column_2 properties )]
		#...
	]
   };

$sql->drop( $table_name )

Pretty self explanatory

$sql->create($table);

NOTES

The database is expected to be -e $file

SEE ALSO

" DBI ", " DBD::SQLite "

AUTHOR

Vidul Petrov, <vidul@cpan.org>

COPYRIGHT AND LICENSE

Copyright 2005 by Vidul Petrov

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.

cpanm SQLite::Abstract

perl -MCPAN -e shell
install SQLite::Abstract