1 package DBIx::DBSchema;
4 use vars qw(@ISA $VERSION $DEBUG);
7 use DBIx::DBSchema::_util qw(_load_driver _dbh);
8 use DBIx::DBSchema::Table;
9 use DBIx::DBSchema::Column;
10 use DBIx::DBSchema::ColGroup::Unique;
11 use DBIx::DBSchema::ColGroup::Index;
21 DBIx::DBSchema - Database-independent schema objects
27 $schema = new DBIx::DBSchema @dbix_dbschema_table_objects;
28 $schema = new_odbc DBIx::DBSchema $dbh;
29 $schema = new_odbc DBIx::DBSchema $dsn, $user, $pass;
30 $schema = new_native DBIx::DBSchema $dbh;
31 $schema = new_native DBIx::DBSchema $dsn, $user, $pass;
33 $schema->save("filename");
34 $schema = load DBIx::DBSchema "filename";
36 $schema->addtable($dbix_dbschema_table_object);
38 @table_names = $schema->tables;
40 $DBIx_DBSchema_table_object = $schema->table("table_name");
42 @sql = $schema->sql($dbh);
43 @sql = $schema->sql($dsn, $username, $password);
44 @sql = $schema->sql($dsn); #doesn't connect to database - less reliable
46 $perl_code = $schema->pretty_print;
47 %hash = eval $perl_code;
48 use DBI qw(:sql_types); $schema = pretty_read DBIx::DBSchema \%hash;
52 DBIx::DBSchema objects are collections of DBIx::DBSchema::Table objects and
53 represent a database schema.
55 This module implements an OO-interface to database schemas. Using this module,
56 you can create a database schema with an OO Perl interface. You can read the
57 schema from an existing database. You can save the schema to disk and restore
58 it a different process. Most importantly, DBIx::DBSchema can write SQL
59 CREATE statements statements for different databases from a single source.
61 Currently supported databases are MySQL and PostgreSQL. Sybase support is
62 partially implemented. DBIx::DBSchema will attempt to use generic SQL syntax
63 for other databases. Assistance adding support for other databases is
64 welcomed. See L<DBIx::DBSchema::DBD>, "Driver Writer's Guide and Base Class".
70 =item new TABLE_OBJECT, TABLE_OBJECT, ...
72 Creates a new DBIx::DBSchema object.
77 my($proto, @tables) = @_;
78 my %tables = map { $_->name, $_ } @tables; #check for duplicates?
80 my $class = ref($proto) || $proto;
85 bless ($self, $class);
89 =item new_odbc DATABASE_HANDLE | DATA_SOURCE USERNAME PASSWORD [ ATTR ]
91 Creates a new DBIx::DBSchema object from an existing data source, which can be
92 specified by passing an open DBI database handle, or by passing the DBI data
93 source name, username, and password. This uses the experimental DBI type_info
94 method to create a schema with standard (ODBC) SQL column types that most
95 closely correspond to any non-portable column types. Use this to import a
96 schema that you wish to use with many different database engines. Although
97 primary key and (unique) index information will only be read from databases
98 with DBIx::DBSchema::DBD drivers (currently MySQL and PostgreSQL), import of
99 column names and attributes *should* work for any database. Note that this
100 method only uses "ODBC" column types; it does not require or use an ODBC
106 my($proto, $dbh) = ( shift, _dbh(@_) );
108 map { new_odbc DBIx::DBSchema::Table $dbh, $_ } _tables_from_dbh($dbh)
112 =item new_native DATABASE_HANDLE | DATA_SOURCE USERNAME PASSWORD [ ATTR ]
114 Creates a new DBIx::DBSchema object from an existing data source, which can be
115 specified by passing an open DBI database handle, or by passing the DBI data
116 source name, username and password. This uses database-native methods to read
117 the schema, and will preserve any non-portable column types. The method is
118 only available if there is a DBIx::DBSchema::DBD for the corresponding database engine (currently, MySQL and PostgreSQL).
123 my($proto, $dbh) = (shift, _dbh(@_) );
125 map { new_native DBIx::DBSchema::Table ( $dbh, $_ ) } _tables_from_dbh($dbh)
131 Loads a DBIx::DBSchema object from a file.
136 my($proto,$file)=@_; #use $proto ?
141 eval { $self = Storable::retrieve($file); };
143 if ( $@ && $@ =~ /not.*storable/i ) { #then try FreezeThaw
144 eval "use FreezeThaw;";
146 open(FILE,"<$file") or die "Can't open $file: $!";
147 my $string = join('',<FILE>);
148 close FILE or die "Can't close $file: $!";
149 ($self) = FreezeThaw::thaw($string);
158 Saves a DBIx::DBSchema object to a file.
163 #my($self, $file) = @_;
164 Storable::nstore(@_);
167 =item addtable TABLE_OBJECT
169 Adds the given DBIx::DBSchema::Table object to this DBIx::DBSchema.
175 $self->{'tables'}->{$table->name} = $table; #check for dupliates?
180 Returns a list of the names of all tables.
186 keys %{$self->{'tables'}};
189 =item table TABLENAME
191 Returns the specified DBIx::DBSchema::Table object.
197 $self->{'tables'}->{$table};
200 =item sql [ DATABASE_HANDLE | DATA_SOURCE [ USERNAME PASSWORD [ ATTR ] ] ]
202 Returns a list of SQL `CREATE' statements for this schema.
204 The data source can be specified by passing an open DBI database handle, or by
205 passing the DBI data source name, username and password.
207 Although the username and password are optional, it is best to call this method
208 with a database handle or data source including a valid username and password -
209 a DBI connection will be opened and the quoting and type mapping will be more
212 If passed a DBI data source (or handle) such as `DBI:mysql:database' or
213 `DBI:Pg:dbname=database', will use syntax specific to that database engine.
214 Currently supported databases are MySQL and PostgreSQL.
216 If not passed a data source (or handle), or if there is no driver for the
217 specified database, will attempt to use generic SQL syntax.
222 my($self, $dbh) = ( shift, _dbh(@_) );
223 map { $self->table($_)->sql_create_table($dbh); } $self->tables;
226 =item sql_update_schema PROTOTYPE_SCHEMA [ DATABASE_HANDLE | DATA_SOURCE [ USERNAME PASSWORD [ ATTR ] ] ]
228 Returns a list of SQL statements to update this schema so that it is idential
229 to the provided prototype schema, also a DBIx::DBSchema object.
231 #Optionally, the data source can be specified by passing an open DBI database
232 #handle, or by passing the DBI data source name, username and password.
234 #If passed a DBI data source (or handle) such as `DBI:mysql:database' or
235 #`DBI:Pg:dbname=database', will use syntax specific to that database engine.
236 #Currently supported databases are MySQL and PostgreSQL.
238 #If not passed a data source (or handle), or if there is no driver for the
239 #specified database, will attempt to use generic SQL syntax.
241 Right now this method knows how to add new tables and alter existing tables.
242 It doesn't know how to drop tables yet.
244 See L<DBIx::DBSchema::Table/sql_alter_table>,
245 L<DBIx::DBSchema::Column/sql_add_coumn> and
246 L<DBIx::DBSchema::Column/sql_alter_column> for additional specifics and
251 #gosh, false laziness w/DBSchema::Table::sql_alter_schema
253 sub sql_update_schema {
254 my($self, $new, $dbh) = ( shift, shift, _dbh(@_) );
258 foreach my $table ( $new->tables ) {
260 if ( $self->table($table) ) {
262 warn "$table exists\n" if $DEBUG > 1;
265 $self->table($table)->sql_alter_table( $new->table($table), $dbh );
269 warn "table $table does not exist.\n" if $DEBUG;
272 $new->table($table)->sql_create_table( $dbh );
278 # should eventually drop tables not in $new
280 warn join("\n", @r). "\n"
287 =item update_schema PROTOTYPE_SCHEMA, DATABASE_HANDLE | DATA_SOURCE [ USERNAME PASSWORD [ ATTR ] ]
289 Same as sql_update_schema, except actually runs the SQL commands to update
290 the schema. Throws a fatal error if any statement fails.
295 my($self, $new, $dbh) = ( shift, shift, _dbh(@_) );
297 foreach my $statement ( $self->sql_update_schema( $new, $dbh ) ) {
298 $dbh->do( $statement )
299 or die "Error: ". $dbh->errstr. "\n executing: $statement";
306 Returns the data in this schema as Perl source, suitable for assigning to a
319 #cant because -w complains about , in qw()
320 # (also biiiig problems with empty lengths)
322 #$self->table($table)->column($_)->type. " ".
323 #( $self->table($table)->column($_)->null ? 'NULL' : 0 ). " ".
324 #$self->table($table)->column($_)->length. " ),\n"
326 "'". $self->table($table)->column($_)->type. "', ".
327 "'". $self->table($table)->column($_)->null. "', ".
328 "'". $self->table($table)->column($_)->length. "', ".
329 "'". $self->table($table)->column($_)->default. "', ".
330 "'". $self->table($table)->column($_)->local. "',\n"
331 } $self->table($table)->columns
334 " 'primary_key' => '". $self->table($table)->primary_key. "',\n".
335 " 'unique' => [ ". join(', ',
336 map { "[ '". join("', '", @{$_}). "' ]" }
337 @{$self->table($table)->unique->lol_ref}
339 " 'index' => [ ". join(', ',
340 map { "[ '". join("', '", @{$_}). "' ]" }
341 @{$self->table($table)->index->lol_ref}
343 #" 'index' => [ ". " ],\n"
350 =item pretty_read HASHREF
352 Creates a schema as specified by a data structure such as that created by
353 B<pretty_print> method.
358 my($proto, $href) = @_;
359 my $schema = $proto->new( map {
361 while ( @{$href->{$_}{'columns'}} ) {
362 push @columns, DBIx::DBSchema::Column->new(
363 splice @{$href->{$_}{'columns'}}, 0, 6
366 DBIx::DBSchema::Table->new(
368 $href->{$_}{'primary_key'},
369 DBIx::DBSchema::ColGroup::Unique->new($href->{$_}{'unique'}),
370 DBIx::DBSchema::ColGroup::Index->new($href->{$_}{'index'}),
376 # private subroutines
378 sub _tables_from_dbh {
380 my $driver = _load_driver($dbh);
382 scalar(eval "DBIx::DBSchema::DBD::$driver->default_db_catalog");
384 scalar(eval "DBIx::DBSchema::DBD::$driver->default_db_schema");
385 my $sth = $dbh->table_info($db_catalog, $db_schema, '', 'TABLE')
387 #map { $_->{TABLE_NAME} } grep { $_->{TABLE_TYPE} eq 'TABLE' }
388 # @{ $sth->fetchall_arrayref({ TABLE_NAME=>1, TABLE_TYPE=>1}) };
389 map { $_->[0] } grep { $_->[1] =~ /^TABLE$/i }
390 @{ $sth->fetchall_arrayref([2,3]) };
397 Ivan Kohler <ivan-dbix-dbschema@420.am>
399 Charles Shapiro <charles.shapiro@numethods.com> and Mitchell Friedman
400 <mitchell.friedman@numethods.com> contributed the start of a Sybase driver.
402 Daniel Hanks <hanksdc@about-inc.com> contributed the Oracle driver.
404 Jesse Vincent contributed the SQLite driver.
408 Contributions are welcome! I'm especially keen on any interest in the first
409 three items/projects below under BUGS.
413 Copyright (c) 2000-2006 Ivan Kohler
414 Copyright (c) 2000 Mail Abuse Prevention System LLC
416 This program is free software; you can redistribute it and/or modify it under
417 the same terms as Perl itself.
421 Indices are not stored by name. Index representation could use an overhaul.
423 Multiple primary keys are not yet supported.
425 Foreign keys and other constraints are not yet supported.
427 Eventually it would be nice to have additional transformations (deleted,
428 modified columns, added/modified/indices (probably need em named first),
431 Need to port and test with additional databases
433 Each DBIx::DBSchema object should have a name which corresponds to its name
434 within the SQL database engine (DBI data source).
436 pretty_print is actually pretty ugly.
438 Perhaps pretty_read should eval column types so that we can use DBI
439 qw(:sql_types) here instead of externally.
441 sql CREATE TABLE output should convert integers
442 (i.e. use DBI qw(:sql_types);) to local types using DBI->type_info plus a hash
445 sql_update_schema doesn't drop tables yet.
449 L<DBIx::DBSchema::Table>, L<DBIx::DBSchema::ColGroup>,
450 L<DBIx::DBSchema::ColGroup::Unique>, L<DBIx::DBSchema::ColGroup::Index>,
451 L<DBIx::DBSchema::Column>, L<DBIx::DBSchema::DBD>,
452 L<DBIx::DBSchema::DBD::mysql>, L<DBIx::DBSchema::DBD::Pg>, L<FS::Record>,