1 package DBIx::Class::Storage::DBI;
2 # -*- mode: cperl; cperl-indent-level: 2 -*-
7 use base qw/DBIx::Class::Storage::DBIHacks DBIx::Class::Storage/;
10 use Carp::Clan qw/^DBIx::Class/;
12 use DBIx::Class::Storage::DBI::Cursor;
13 use DBIx::Class::Storage::Statistics;
16 use Data::Dumper::Concise();
19 __PACKAGE__->mk_group_accessors('simple' =>
20 qw/_connect_info _dbi_connect_info _dbh _sql_maker _sql_maker_opts _conn_pid
21 _conn_tid transaction_depth _dbh_autocommit _driver_determined savepoints
25 # the values for these accessors are picked out (and deleted) from
26 # the attribute hashref passed to connect_info
27 my @storage_options = qw/
28 on_connect_call on_disconnect_call on_connect_do on_disconnect_do
29 disable_sth_caching unsafe auto_savepoint
31 __PACKAGE__->mk_group_accessors('simple' => @storage_options);
34 # default cursor class, overridable in connect_info attributes
35 __PACKAGE__->cursor_class('DBIx::Class::Storage::DBI::Cursor');
37 __PACKAGE__->mk_group_accessors('inherited' => qw/
41 __PACKAGE__->sql_maker_class('DBIx::Class::SQLAHacks');
44 # Each of these methods need _determine_driver called before itself
45 # in order to function reliably. This is a purely DRY optimization
46 my @rdbms_specific_methods = qw/
60 for my $meth (@rdbms_specific_methods) {
62 my $orig = __PACKAGE__->can ($meth)
66 no warnings qw/redefine/;
67 *{__PACKAGE__ ."::$meth"} = Sub::Name::subname $meth => sub {
68 if (not $_[0]->_driver_determined) {
69 $_[0]->_determine_driver;
70 goto $_[0]->can($meth);
79 DBIx::Class::Storage::DBI - DBI storage handler
83 my $schema = MySchema->connect('dbi:SQLite:my.db');
85 $schema->storage->debug(1);
87 my @stuff = $schema->storage->dbh_do(
89 my ($storage, $dbh, @args) = @_;
90 $dbh->do("DROP TABLE authors");
95 $schema->resultset('Book')->search({
96 written_on => $schema->storage->datetime_parser(DateTime->now)
101 This class represents the connection to an RDBMS via L<DBI>. See
102 L<DBIx::Class::Storage> for general information. This pod only
103 documents DBI-specific methods and behaviors.
110 my $new = shift->next::method(@_);
112 $new->transaction_depth(0);
113 $new->_sql_maker_opts({});
114 $new->{savepoints} = [];
115 $new->{_in_dbh_do} = 0;
116 $new->{_dbh_gen} = 0;
123 This method is normally called by L<DBIx::Class::Schema/connection>, which
124 encapsulates its argument list in an arrayref before passing them here.
126 The argument list may contain:
132 The same 4-element argument set one would normally pass to
133 L<DBI/connect>, optionally followed by
134 L<extra attributes|/DBIx::Class specific connection attributes>
135 recognized by DBIx::Class:
137 $connect_info_args = [ $dsn, $user, $password, \%dbi_attributes?, \%extra_attributes? ];
141 A single code reference which returns a connected
142 L<DBI database handle|DBI/connect> optionally followed by
143 L<extra attributes|/DBIx::Class specific connection attributes> recognized
146 $connect_info_args = [ sub { DBI->connect (...) }, \%extra_attributes? ];
150 A single hashref with all the attributes and the dsn/user/password
153 $connect_info_args = [{
161 $connect_info_args = [{
162 dbh_maker => sub { DBI->connect (...) },
167 This is particularly useful for L<Catalyst> based applications, allowing the
168 following config (L<Config::General> style):
173 dsn dbi:mysql:database=test
180 The C<dsn>/C<user>/C<password> combination can be substituted by the
181 C<dbh_maker> key whose value is a coderef that returns a connected
182 L<DBI database handle|DBI/connect>
186 Please note that the L<DBI> docs recommend that you always explicitly
187 set C<AutoCommit> to either I<0> or I<1>. L<DBIx::Class> further
188 recommends that it be set to I<1>, and that you perform transactions
189 via our L<DBIx::Class::Schema/txn_do> method. L<DBIx::Class> will set it
190 to I<1> if you do not do explicitly set it to zero. This is the default
191 for most DBDs. See L</DBIx::Class and AutoCommit> for details.
193 =head3 DBIx::Class specific connection attributes
195 In addition to the standard L<DBI|DBI/ATTRIBUTES_COMMON_TO_ALL_HANDLES>
196 L<connection|DBI/Database_Handle_Attributes> attributes, DBIx::Class recognizes
197 the following connection options. These options can be mixed in with your other
198 L<DBI> connection attributes, or placed in a separate hashref
199 (C<\%extra_attributes>) as shown above.
201 Every time C<connect_info> is invoked, any previous settings for
202 these options will be cleared before setting the new ones, regardless of
203 whether any options are specified in the new C<connect_info>.
210 Specifies things to do immediately after connecting or re-connecting to
211 the database. Its value may contain:
217 This contains one SQL statement to execute.
219 =item an array reference
221 This contains SQL statements to execute in order. Each element contains
222 a string or a code reference that returns a string.
224 =item a code reference
226 This contains some code to execute. Unlike code references within an
227 array reference, its return value is ignored.
231 =item on_disconnect_do
233 Takes arguments in the same form as L</on_connect_do> and executes them
234 immediately before disconnecting from the database.
236 Note, this only runs if you explicitly call L</disconnect> on the
239 =item on_connect_call
241 A more generalized form of L</on_connect_do> that calls the specified
242 C<connect_call_METHOD> methods in your storage driver.
244 on_connect_do => 'select 1'
248 on_connect_call => [ [ do_sql => 'select 1' ] ]
250 Its values may contain:
256 Will call the C<connect_call_METHOD> method.
258 =item a code reference
260 Will execute C<< $code->($storage) >>
262 =item an array reference
264 Each value can be a method name or code reference.
266 =item an array of arrays
268 For each array, the first item is taken to be the C<connect_call_> method name
269 or code reference, and the rest are parameters to it.
273 Some predefined storage methods you may use:
279 Executes a SQL string or a code reference that returns a SQL string. This is
280 what L</on_connect_do> and L</on_disconnect_do> use.
288 Will execute the scalar as SQL.
292 Taken to be arguments to L<DBI/do>, the SQL string optionally followed by the
293 attributes hashref and bind values.
295 =item a code reference
297 Will execute C<< $code->($storage) >> and execute the return array refs as
304 Execute any statements necessary to initialize the database session to return
305 and accept datetime/timestamp values used with
306 L<DBIx::Class::InflateColumn::DateTime>.
308 Only necessary for some databases, see your specific storage driver for
309 implementation details.
313 =item on_disconnect_call
315 Takes arguments in the same form as L</on_connect_call> and executes them
316 immediately before disconnecting from the database.
318 Calls the C<disconnect_call_METHOD> methods as opposed to the
319 C<connect_call_METHOD> methods called by L</on_connect_call>.
321 Note, this only runs if you explicitly call L</disconnect> on the
324 =item disable_sth_caching
326 If set to a true value, this option will disable the caching of
327 statement handles via L<DBI/prepare_cached>.
331 Sets the limit dialect. This is useful for JDBC-bridge among others
332 where the remote SQL-dialect cannot be determined by the name of the
333 driver alone. See also L<SQL::Abstract::Limit>.
337 Specifies what characters to use to quote table and column names. If
338 you use this you will want to specify L</name_sep> as well.
340 C<quote_char> expects either a single character, in which case is it
341 is placed on either side of the table/column name, or an arrayref of length
342 2 in which case the table/column name is placed between the elements.
344 For example under MySQL you should use C<< quote_char => '`' >>, and for
345 SQL Server you should use C<< quote_char => [qw/[ ]/] >>.
349 This only needs to be used in conjunction with C<quote_char>, and is used to
350 specify the character that separates elements (schemas, tables, columns) from
351 each other. In most cases this is simply a C<.>.
353 The consequences of not supplying this value is that L<SQL::Abstract>
354 will assume DBIx::Class' uses of aliases to be complete column
355 names. The output will look like I<"me.name"> when it should actually
360 This Storage driver normally installs its own C<HandleError>, sets
361 C<RaiseError> and C<ShowErrorStatement> on, and sets C<PrintError> off on
362 all database handles, including those supplied by a coderef. It does this
363 so that it can have consistent and useful error behavior.
365 If you set this option to a true value, Storage will not do its usual
366 modifications to the database handle's attributes, and instead relies on
367 the settings in your connect_info DBI options (or the values you set in
368 your connection coderef, in the case that you are connecting via coderef).
370 Note that your custom settings can cause Storage to malfunction,
371 especially if you set a C<HandleError> handler that suppresses exceptions
372 and/or disable C<RaiseError>.
376 If this option is true, L<DBIx::Class> will use savepoints when nesting
377 transactions, making it possible to recover from failure in the inner
378 transaction without having to abort all outer transactions.
382 Use this argument to supply a cursor class other than the default
383 L<DBIx::Class::Storage::DBI::Cursor>.
387 Some real-life examples of arguments to L</connect_info> and
388 L<DBIx::Class::Schema/connect>
390 # Simple SQLite connection
391 ->connect_info([ 'dbi:SQLite:./foo.db' ]);
394 ->connect_info([ sub { DBI->connect(...) } ]);
396 # Connect via subref in hashref
398 dbh_maker => sub { DBI->connect(...) },
399 on_connect_do => 'alter session ...',
402 # A bit more complicated
409 { quote_char => q{"}, name_sep => q{.} },
413 # Equivalent to the previous example
419 { AutoCommit => 1, quote_char => q{"}, name_sep => q{.} },
423 # Same, but with hashref as argument
424 # See parse_connect_info for explanation
427 dsn => 'dbi:Pg:dbname=foo',
429 password => 'my_pg_password',
436 # Subref + DBIx::Class-specific connection options
439 sub { DBI->connect(...) },
443 on_connect_do => ['SET search_path TO myschema,otherschema,public'],
444 disable_sth_caching => 1,
454 my ($self, $info) = @_;
456 return $self->_connect_info if !$info;
458 $self->_connect_info($info); # copy for _connect_info
460 $info = $self->_normalize_connect_info($info)
461 if ref $info eq 'ARRAY';
463 for my $storage_opt (keys %{ $info->{storage_options} }) {
464 my $value = $info->{storage_options}{$storage_opt};
466 $self->$storage_opt($value);
469 # Kill sql_maker/_sql_maker_opts, so we get a fresh one with only
470 # the new set of options
471 $self->_sql_maker(undef);
472 $self->_sql_maker_opts({});
474 for my $sql_maker_opt (keys %{ $info->{sql_maker_options} }) {
475 my $value = $info->{sql_maker_options}{$sql_maker_opt};
477 $self->_sql_maker_opts->{$sql_maker_opt} = $value;
481 %{ $self->_default_dbi_connect_attributes || {} },
482 %{ $info->{attributes} || {} },
485 my @args = @{ $info->{arguments} };
487 $self->_dbi_connect_info([@args,
488 %attrs && !(ref $args[0] eq 'CODE') ? \%attrs : ()]);
490 return $self->_connect_info;
493 sub _normalize_connect_info {
494 my ($self, $info_arg) = @_;
497 my @args = @$info_arg; # take a shallow copy for further mutilation
499 # combine/pre-parse arguments depending on invocation style
502 if (ref $args[0] eq 'CODE') { # coderef with optional \%extra_attributes
503 %attrs = %{ $args[1] || {} };
506 elsif (ref $args[0] eq 'HASH') { # single hashref (i.e. Catalyst config)
507 %attrs = %{$args[0]};
509 if (my $code = delete $attrs{dbh_maker}) {
512 my @ignored = grep { delete $attrs{$_} } (qw/dsn user password/);
515 'Attribute(s) %s in connect_info were ignored, as they can not be applied '
516 . "to the result of 'dbh_maker'",
518 join (', ', map { "'$_'" } (@ignored) ),
523 @args = delete @attrs{qw/dsn user password/};
526 else { # otherwise assume dsn/user/password + \%attrs + \%extra_attrs
528 % { $args[3] || {} },
529 % { $args[4] || {} },
531 @args = @args[0,1,2];
534 $info{arguments} = \@args;
536 my @storage_opts = grep exists $attrs{$_},
537 @storage_options, 'cursor_class';
539 @{ $info{storage_options} }{@storage_opts} =
540 delete @attrs{@storage_opts} if @storage_opts;
542 my @sql_maker_opts = grep exists $attrs{$_},
543 qw/limit_dialect quote_char name_sep/;
545 @{ $info{sql_maker_options} }{@sql_maker_opts} =
546 delete @attrs{@sql_maker_opts} if @sql_maker_opts;
548 $info{attributes} = \%attrs if %attrs;
553 sub _default_dbi_connect_attributes {
563 This method is deprecated in favour of setting via L</connect_info>.
567 =head2 on_disconnect_do
569 This method is deprecated in favour of setting via L</connect_info>.
573 sub _parse_connect_do {
574 my ($self, $type) = @_;
576 my $val = $self->$type;
577 return () if not defined $val;
582 push @res, [ 'do_sql', $val ];
583 } elsif (ref($val) eq 'CODE') {
585 } elsif (ref($val) eq 'ARRAY') {
586 push @res, map { [ 'do_sql', $_ ] } @$val;
588 $self->throw_exception("Invalid type for $type: ".ref($val));
596 Arguments: ($subref | $method_name), @extra_coderef_args?
598 Execute the given $subref or $method_name using the new exception-based
599 connection management.
601 The first two arguments will be the storage object that C<dbh_do> was called
602 on and a database handle to use. Any additional arguments will be passed
603 verbatim to the called subref as arguments 2 and onwards.
605 Using this (instead of $self->_dbh or $self->dbh) ensures correct
606 exception handling and reconnection (or failover in future subclasses).
608 Your subref should have no side-effects outside of the database, as
609 there is the potential for your subref to be partially double-executed
610 if the database connection was stale/dysfunctional.
614 my @stuff = $schema->storage->dbh_do(
616 my ($storage, $dbh, @cols) = @_;
617 my $cols = join(q{, }, @cols);
618 $dbh->selectrow_array("SELECT $cols FROM foo");
629 my $dbh = $self->_get_dbh;
631 return $self->$code($dbh, @_) if $self->{_in_dbh_do}
632 || $self->{transaction_depth};
634 local $self->{_in_dbh_do} = 1;
637 my $want_array = wantarray;
642 @result = $self->$code($dbh, @_);
644 elsif(defined $want_array) {
645 $result[0] = $self->$code($dbh, @_);
648 $self->$code($dbh, @_);
652 # ->connected might unset $@ - copy
654 if(!$exception) { return $want_array ? @result : $result[0] }
656 $self->throw_exception($exception) if $self->connected;
658 # We were not connected - reconnect and retry, but let any
659 # exception fall right through this time
660 carp "Retrying $code after catching disconnected exception: $exception"
661 if $ENV{DBIC_DBIRETRY_DEBUG};
662 $self->_populate_dbh;
663 $self->$code($self->_dbh, @_);
666 # This is basically a blend of dbh_do above and DBIx::Class::Storage::txn_do.
667 # It also informs dbh_do to bypass itself while under the direction of txn_do,
668 # via $self->{_in_dbh_do} (this saves some redundant eval and errorcheck, etc)
673 ref $coderef eq 'CODE' or $self->throw_exception
674 ('$coderef must be a CODE reference');
676 return $coderef->(@_) if $self->{transaction_depth} && ! $self->auto_savepoint;
678 local $self->{_in_dbh_do} = 1;
681 my $want_array = wantarray;
690 @result = $coderef->(@_);
692 elsif(defined $want_array) {
693 $result[0] = $coderef->(@_);
701 # ->connected might unset $@ - copy
703 if(!$exception) { return $want_array ? @result : $result[0] }
705 if($tried++ || $self->connected) {
706 eval { $self->txn_rollback };
707 my $rollback_exception = $@;
708 if($rollback_exception) {
709 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
710 $self->throw_exception($exception) # propagate nested rollback
711 if $rollback_exception =~ /$exception_class/;
713 $self->throw_exception(
714 "Transaction aborted: ${exception}. "
715 . "Rollback failed: ${rollback_exception}"
718 $self->throw_exception($exception)
721 # We were not connected, and was first try - reconnect and retry
723 carp "Retrying $coderef after catching disconnected exception: $exception"
724 if $ENV{DBIC_DBIRETRY_DEBUG};
725 $self->_populate_dbh;
731 Our C<disconnect> method also performs a rollback first if the
732 database is not in C<AutoCommit> mode.
742 push @actions, ( $self->on_disconnect_call || () );
743 push @actions, $self->_parse_connect_do ('on_disconnect_do');
745 $self->_do_connection_actions(disconnect_call_ => $_) for @actions;
747 $self->_dbh_rollback unless $self->_dbh_autocommit;
749 %{ $self->_dbh->{CachedKids} } = ();
750 $self->_dbh->disconnect;
756 =head2 with_deferred_fk_checks
760 =item Arguments: C<$coderef>
762 =item Return Value: The return value of $coderef
766 Storage specific method to run the code ref with FK checks deferred or
767 in MySQL's case disabled entirely.
771 # Storage subclasses should override this
772 sub with_deferred_fk_checks {
773 my ($self, $sub) = @_;
781 =item Arguments: none
783 =item Return Value: 1|0
787 Verifies that the current database handle is active and ready to execute
788 an SQL statement (e.g. the connection did not get stale, server is still
789 answering, etc.) This method is used internally by L</dbh>.
795 return 0 unless $self->_seems_connected;
798 local $self->_dbh->{RaiseError} = 1;
803 sub _seems_connected {
806 my $dbh = $self->_dbh
809 if(defined $self->_conn_tid && $self->_conn_tid != threads->tid) {
816 return 0 if !$self->_dbh;
819 return $dbh->FETCH('Active');
825 my $dbh = $self->_dbh or return 0;
830 # handle pid changes correctly
831 # NOTE: assumes $self->_dbh is a valid $dbh
835 return if defined $self->_conn_pid && $self->_conn_pid == $$;
837 $self->_dbh->{InactiveDestroy} = 1;
844 sub ensure_connected {
847 unless ($self->connected) {
848 $self->_populate_dbh;
854 Returns a C<$dbh> - a data base handle of class L<DBI>. The returned handle
855 is guaranteed to be healthy by implicitly calling L</connected>, and if
856 necessary performing a reconnection before returning. Keep in mind that this
857 is very B<expensive> on some database engines. Consider using L</dbh_do>
865 if (not $self->_dbh) {
866 $self->_populate_dbh;
868 $self->ensure_connected;
873 # this is the internal "get dbh or connect (don't check)" method
876 $self->_verify_pid if $self->_dbh;
877 $self->_populate_dbh unless $self->_dbh;
881 sub _sql_maker_args {
886 array_datatypes => 1,
887 limit_dialect => $self->_get_dbh,
888 %{$self->_sql_maker_opts}
894 unless ($self->_sql_maker) {
895 my $sql_maker_class = $self->sql_maker_class;
896 $self->ensure_class_loaded ($sql_maker_class);
897 $self->_sql_maker($sql_maker_class->new( $self->_sql_maker_args ));
899 return $self->_sql_maker;
902 # nothing to do by default
909 my @info = @{$self->_dbi_connect_info || []};
910 $self->_dbh(undef); # in case ->connected failed we might get sent here
911 $self->_dbh($self->_connect(@info));
913 $self->_conn_pid($$);
914 $self->_conn_tid(threads->tid) if $INC{'threads.pm'};
916 $self->_determine_driver;
918 # Always set the transaction depth on connect, since
919 # there is no transaction in progress by definition
920 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
922 $self->_run_connection_actions unless $self->{_in_determine_driver};
924 $self->_get_server_info;
927 sub _run_connection_actions {
931 push @actions, ( $self->on_connect_call || () );
932 push @actions, $self->_parse_connect_do ('on_connect_do');
934 $self->_do_connection_actions(connect_call_ => $_) for @actions;
937 sub _get_server_info {
941 $info{dbms_ver} = $self->_get_dbh->get_info(18);
943 $self->_server_info(\%info);
948 sub _determine_driver {
951 if ((not $self->_driver_determined) && (not $self->{_in_determine_driver})) {
952 my $started_connected = 0;
953 local $self->{_in_determine_driver} = 1;
955 if (ref($self) eq __PACKAGE__) {
957 if ($self->_dbh) { # we are connected
958 $driver = $self->_dbh->{Driver}{Name};
959 $started_connected = 1;
961 # if connect_info is a CODEREF, we have no choice but to connect
962 if (ref $self->_dbi_connect_info->[0] &&
963 Scalar::Util::reftype($self->_dbi_connect_info->[0]) eq 'CODE') {
964 $self->_populate_dbh;
965 $driver = $self->_dbh->{Driver}{Name};
968 # try to use dsn to not require being connected, the driver may still
969 # force a connection in _rebless to determine version
970 # (dsn may not be supplied at all if all we do is make a mock-schema)
971 my $dsn = $self->_dbi_connect_info->[0] || '';
972 ($driver) = $dsn =~ /dbi:([^:]+):/i;
977 my $storage_class = "DBIx::Class::Storage::DBI::${driver}";
978 if ($self->load_optional_class($storage_class)) {
979 mro::set_mro($storage_class, 'c3');
980 bless $self, $storage_class;
986 $self->_driver_determined(1);
988 $self->_init; # run driver-specific initializations
990 $self->_run_connection_actions
991 if !$started_connected && defined $self->_dbh;
995 sub _do_connection_actions {
997 my $method_prefix = shift;
1000 if (not ref($call)) {
1001 my $method = $method_prefix . $call;
1003 } elsif (ref($call) eq 'CODE') {
1005 } elsif (ref($call) eq 'ARRAY') {
1006 if (ref($call->[0]) ne 'ARRAY') {
1007 $self->_do_connection_actions($method_prefix, $_) for @$call;
1009 $self->_do_connection_actions($method_prefix, @$_) for @$call;
1012 $self->throw_exception (sprintf ("Don't know how to process conection actions of type '%s'", ref($call)) );
1018 sub connect_call_do_sql {
1020 $self->_do_query(@_);
1023 sub disconnect_call_do_sql {
1025 $self->_do_query(@_);
1028 # override in db-specific backend when necessary
1029 sub connect_call_datetime_setup { 1 }
1032 my ($self, $action) = @_;
1034 if (ref $action eq 'CODE') {
1035 $action = $action->($self);
1036 $self->_do_query($_) foreach @$action;
1039 # Most debuggers expect ($sql, @bind), so we need to exclude
1040 # the attribute hash which is the second argument to $dbh->do
1041 # furthermore the bind values are usually to be presented
1042 # as named arrayref pairs, so wrap those here too
1043 my @do_args = (ref $action eq 'ARRAY') ? (@$action) : ($action);
1044 my $sql = shift @do_args;
1045 my $attrs = shift @do_args;
1046 my @bind = map { [ undef, $_ ] } @do_args;
1048 $self->_query_start($sql, @bind);
1049 $self->_get_dbh->do($sql, $attrs, @do_args);
1050 $self->_query_end($sql, @bind);
1057 my ($self, @info) = @_;
1059 $self->throw_exception("You failed to provide any connection info")
1062 my ($old_connect_via, $dbh);
1064 if ($INC{'Apache/DBI.pm'} && $ENV{MOD_PERL}) {
1065 $old_connect_via = $DBI::connect_via;
1066 $DBI::connect_via = 'connect';
1070 if(ref $info[0] eq 'CODE') {
1071 $dbh = $info[0]->();
1074 $dbh = DBI->connect(@info);
1077 if($dbh && !$self->unsafe) {
1078 my $weak_self = $self;
1079 Scalar::Util::weaken($weak_self);
1080 $dbh->{HandleError} = sub {
1082 $weak_self->throw_exception("DBI Exception: $_[0]");
1085 # the handler may be invoked by something totally out of
1087 croak ("DBI Exception: $_[0]");
1090 $dbh->{ShowErrorStatement} = 1;
1091 $dbh->{RaiseError} = 1;
1092 $dbh->{PrintError} = 0;
1096 $DBI::connect_via = $old_connect_via if $old_connect_via;
1098 $self->throw_exception("DBI Connection failed: " . ($@||$DBI::errstr))
1101 $self->_dbh_autocommit($dbh->{AutoCommit});
1107 my ($self, $name) = @_;
1109 $name = $self->_svp_generate_name
1110 unless defined $name;
1112 $self->throw_exception ("You can't use savepoints outside a transaction")
1113 if $self->{transaction_depth} == 0;
1115 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1116 unless $self->can('_svp_begin');
1118 push @{ $self->{savepoints} }, $name;
1120 $self->debugobj->svp_begin($name) if $self->debug;
1122 return $self->_svp_begin($name);
1126 my ($self, $name) = @_;
1128 $self->throw_exception ("You can't use savepoints outside a transaction")
1129 if $self->{transaction_depth} == 0;
1131 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1132 unless $self->can('_svp_release');
1134 if (defined $name) {
1135 $self->throw_exception ("Savepoint '$name' does not exist")
1136 unless grep { $_ eq $name } @{ $self->{savepoints} };
1138 # Dig through the stack until we find the one we are releasing. This keeps
1139 # the stack up to date.
1142 do { $svp = pop @{ $self->{savepoints} } } while $svp ne $name;
1144 $name = pop @{ $self->{savepoints} };
1147 $self->debugobj->svp_release($name) if $self->debug;
1149 return $self->_svp_release($name);
1153 my ($self, $name) = @_;
1155 $self->throw_exception ("You can't use savepoints outside a transaction")
1156 if $self->{transaction_depth} == 0;
1158 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1159 unless $self->can('_svp_rollback');
1161 if (defined $name) {
1162 # If they passed us a name, verify that it exists in the stack
1163 unless(grep({ $_ eq $name } @{ $self->{savepoints} })) {
1164 $self->throw_exception("Savepoint '$name' does not exist!");
1167 # Dig through the stack until we find the one we are releasing. This keeps
1168 # the stack up to date.
1169 while(my $s = pop(@{ $self->{savepoints} })) {
1170 last if($s eq $name);
1172 # Add the savepoint back to the stack, as a rollback doesn't remove the
1173 # named savepoint, only everything after it.
1174 push(@{ $self->{savepoints} }, $name);
1176 # We'll assume they want to rollback to the last savepoint
1177 $name = $self->{savepoints}->[-1];
1180 $self->debugobj->svp_rollback($name) if $self->debug;
1182 return $self->_svp_rollback($name);
1185 sub _svp_generate_name {
1188 return 'savepoint_'.scalar(@{ $self->{'savepoints'} });
1194 # this means we have not yet connected and do not know the AC status
1195 # (e.g. coderef $dbh)
1196 $self->ensure_connected if (! defined $self->_dbh_autocommit);
1198 if($self->{transaction_depth} == 0) {
1199 $self->debugobj->txn_begin()
1201 $self->_dbh_begin_work;
1203 elsif ($self->auto_savepoint) {
1206 $self->{transaction_depth}++;
1209 sub _dbh_begin_work {
1212 # if the user is utilizing txn_do - good for him, otherwise we need to
1213 # ensure that the $dbh is healthy on BEGIN.
1214 # We do this via ->dbh_do instead of ->dbh, so that the ->dbh "ping"
1215 # will be replaced by a failure of begin_work itself (which will be
1216 # then retried on reconnect)
1217 if ($self->{_in_dbh_do}) {
1218 $self->_dbh->begin_work;
1220 $self->dbh_do(sub { $_[1]->begin_work });
1226 if ($self->{transaction_depth} == 1) {
1227 $self->debugobj->txn_commit()
1230 $self->{transaction_depth} = 0
1231 if $self->_dbh_autocommit;
1233 elsif($self->{transaction_depth} > 1) {
1234 $self->{transaction_depth}--;
1236 if $self->auto_savepoint;
1242 my $dbh = $self->_dbh
1243 or $self->throw_exception('cannot COMMIT on a disconnected handle');
1249 my $dbh = $self->_dbh;
1251 if ($self->{transaction_depth} == 1) {
1252 $self->debugobj->txn_rollback()
1254 $self->{transaction_depth} = 0
1255 if $self->_dbh_autocommit;
1256 $self->_dbh_rollback;
1258 elsif($self->{transaction_depth} > 1) {
1259 $self->{transaction_depth}--;
1260 if ($self->auto_savepoint) {
1261 $self->svp_rollback;
1266 die DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION->new;
1271 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
1272 $error =~ /$exception_class/ and $self->throw_exception($error);
1273 # ensure that a failed rollback resets the transaction depth
1274 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
1275 $self->throw_exception($error);
1281 my $dbh = $self->_dbh
1282 or $self->throw_exception('cannot ROLLBACK on a disconnected handle');
1286 # This used to be the top-half of _execute. It was split out to make it
1287 # easier to override in NoBindVars without duping the rest. It takes up
1288 # all of _execute's args, and emits $sql, @bind.
1289 sub _prep_for_execute {
1290 my ($self, $op, $extra_bind, $ident, $args) = @_;
1292 if( Scalar::Util::blessed($ident) && $ident->isa("DBIx::Class::ResultSource") ) {
1293 $ident = $ident->from();
1296 my ($sql, @bind) = $self->sql_maker->$op($ident, @$args);
1299 map { ref $_ eq 'ARRAY' ? $_ : [ '!!dummy', $_ ] } @$extra_bind)
1301 return ($sql, \@bind);
1305 sub _fix_bind_params {
1306 my ($self, @bind) = @_;
1308 ### Turn @bind from something like this:
1309 ### ( [ "artist", 1 ], [ "cdid", 1, 3 ] )
1311 ### ( "'1'", "'1'", "'3'" )
1314 if ( defined( $_ && $_->[1] ) ) {
1315 map { qq{'$_'}; } @{$_}[ 1 .. $#$_ ];
1322 my ( $self, $sql, @bind ) = @_;
1324 if ( $self->debug ) {
1325 @bind = $self->_fix_bind_params(@bind);
1327 $self->debugobj->query_start( $sql, @bind );
1332 my ( $self, $sql, @bind ) = @_;
1334 if ( $self->debug ) {
1335 @bind = $self->_fix_bind_params(@bind);
1336 $self->debugobj->query_end( $sql, @bind );
1341 my ($self, $dbh, $op, $extra_bind, $ident, $bind_attributes, @args) = @_;
1343 my ($sql, $bind) = $self->_prep_for_execute($op, $extra_bind, $ident, \@args);
1345 $self->_query_start( $sql, @$bind );
1347 my $sth = $self->sth($sql,$op);
1349 my $placeholder_index = 1;
1351 foreach my $bound (@$bind) {
1352 my $attributes = {};
1353 my($column_name, @data) = @$bound;
1355 if ($bind_attributes) {
1356 $attributes = $bind_attributes->{$column_name}
1357 if defined $bind_attributes->{$column_name};
1360 foreach my $data (@data) {
1361 my $ref = ref $data;
1362 $data = $ref && $ref ne 'ARRAY' ? ''.$data : $data; # stringify args (except arrayrefs)
1364 $sth->bind_param($placeholder_index, $data, $attributes);
1365 $placeholder_index++;
1369 # Can this fail without throwing an exception anyways???
1370 my $rv = $sth->execute();
1371 $self->throw_exception($sth->errstr) if !$rv;
1373 $self->_query_end( $sql, @$bind );
1375 return (wantarray ? ($rv, $sth, @$bind) : $rv);
1380 $self->dbh_do('_dbh_execute', @_); # retry over disconnects
1383 sub _prefetch_insert_auto_nextvals {
1384 my ($self, $source, $to_insert) = @_;
1388 foreach my $col ( $source->columns ) {
1389 if ( !defined $to_insert->{$col} ) {
1390 my $col_info = $source->column_info($col);
1392 if ( $col_info->{auto_nextval} ) {
1393 $upd->{$col} = $to_insert->{$col} = $self->_sequence_fetch(
1395 $col_info->{sequence} ||=
1396 $self->_dbh_get_autoinc_seq($self->_get_dbh, $source, $col)
1407 my ($source, $to_insert, $opts) = @_;
1409 my $updated_cols = $self->_prefetch_insert_auto_nextvals (@_);
1411 my $bind_attributes = $self->source_bind_attributes($source);
1413 my ($rv, $sth) = $self->_execute('insert' => [], $source, $bind_attributes, $to_insert, $opts);
1415 if ($opts->{returning}) {
1416 my @ret_cols = @{$opts->{returning}};
1418 my @ret_vals = eval {
1419 local $SIG{__WARN__} = sub {};
1420 my @r = $sth->fetchrow_array;
1426 @ret{@ret_cols} = @ret_vals if (@ret_vals);
1434 return $updated_cols;
1437 ## Currently it is assumed that all values passed will be "normal", i.e. not
1438 ## scalar refs, or at least, all the same type as the first set, the statement is
1439 ## only prepped once.
1441 my ($self, $source, $cols, $data) = @_;
1444 @colvalues{@$cols} = (0..$#$cols);
1446 for my $i (0..$#$cols) {
1447 my $first_val = $data->[0][$i];
1448 next unless ref $first_val eq 'SCALAR';
1450 $colvalues{ $cols->[$i] } = $first_val;
1453 # check for bad data and stringify stringifiable objects
1454 my $bad_slice = sub {
1455 my ($msg, $col_idx, $slice_idx) = @_;
1456 $self->throw_exception(sprintf "%s for column '%s' in populate slice:\n%s",
1460 local $Data::Dumper::Maxdepth = 1; # don't dump objects, if any
1461 Data::Dumper::Concise::Dumper({
1462 map { $cols->[$_] => $data->[$slice_idx][$_] } (0 .. $#$cols)
1468 for my $datum_idx (0..$#$data) {
1469 my $datum = $data->[$datum_idx];
1471 for my $col_idx (0..$#$cols) {
1472 my $val = $datum->[$col_idx];
1473 my $sqla_bind = $colvalues{ $cols->[$col_idx] };
1474 my $is_literal_sql = (ref $sqla_bind) eq 'SCALAR';
1476 if ($is_literal_sql) {
1478 $bad_slice->('bind found where literal SQL expected', $col_idx, $datum_idx);
1480 elsif ((my $reftype = ref $val) ne 'SCALAR') {
1481 $bad_slice->("$reftype reference found where literal SQL expected",
1482 $col_idx, $datum_idx);
1484 elsif ($$val ne $$sqla_bind){
1485 $bad_slice->("inconsistent literal SQL value, expecting: '$$sqla_bind'",
1486 $col_idx, $datum_idx);
1489 elsif (my $reftype = ref $val) {
1491 if (overload::Method($val, '""')) {
1492 $datum->[$col_idx] = "".$val;
1495 $bad_slice->("$reftype reference found where bind expected",
1496 $col_idx, $datum_idx);
1502 my ($sql, $bind) = $self->_prep_for_execute (
1503 'insert', undef, $source, [\%colvalues]
1507 my $empty_bind = 1 if (not @bind) &&
1508 (grep { ref $_ eq 'SCALAR' } values %colvalues) == @$cols;
1510 if ((not @bind) && (not $empty_bind)) {
1511 $self->throw_exception(
1512 'Cannot insert_bulk without support for placeholders'
1516 # neither _execute_array, nor _execute_inserts_with_no_binds are
1517 # atomic (even if _execute _array is a single call). Thus a safety
1519 my $guard = $self->txn_scope_guard;
1521 $self->_query_start( $sql, ['__BULK__'] );
1522 my $sth = $self->sth($sql);
1525 # bind_param_array doesn't work if there are no binds
1526 $self->_dbh_execute_inserts_with_no_binds( $sth, scalar @$data );
1529 # @bind = map { ref $_ ? ''.$_ : $_ } @bind; # stringify args
1530 $self->_execute_array( $source, $sth, \@bind, $cols, $data );
1534 $self->_query_end( $sql, ['__BULK__'] );
1538 return (wantarray ? ($rv, $sth, @bind) : $rv);
1541 sub _execute_array {
1542 my ($self, $source, $sth, $bind, $cols, $data, @extra) = @_;
1544 ## This must be an arrayref, else nothing works!
1545 my $tuple_status = [];
1547 ## Get the bind_attributes, if any exist
1548 my $bind_attributes = $self->source_bind_attributes($source);
1550 ## Bind the values and execute
1551 my $placeholder_index = 1;
1553 foreach my $bound (@$bind) {
1555 my $attributes = {};
1556 my ($column_name, $data_index) = @$bound;
1558 if( $bind_attributes ) {
1559 $attributes = $bind_attributes->{$column_name}
1560 if defined $bind_attributes->{$column_name};
1563 my @data = map { $_->[$data_index] } @$data;
1565 $sth->bind_param_array(
1568 (%$attributes ? $attributes : ()),
1570 $placeholder_index++;
1574 $self->_dbh_execute_array($sth, $tuple_status, @extra);
1576 my $err = $@ || $sth->errstr;
1578 # Statement must finish even if there was an exception.
1579 eval { $sth->finish };
1580 $err = $@ unless $err;
1584 ++$i while $i <= $#$tuple_status && !ref $tuple_status->[$i];
1586 $self->throw_exception("Unexpected populate error: $err")
1587 if ($i > $#$tuple_status);
1589 $self->throw_exception(sprintf "%s for populate slice:\n%s",
1590 ($tuple_status->[$i][1] || $err),
1591 Data::Dumper::Concise::Dumper({
1592 map { $cols->[$_] => $data->[$i][$_] } (0 .. $#$cols)
1599 sub _dbh_execute_array {
1600 my ($self, $sth, $tuple_status, @extra) = @_;
1602 return $sth->execute_array({ArrayTupleStatus => $tuple_status});
1605 sub _dbh_execute_inserts_with_no_binds {
1606 my ($self, $sth, $count) = @_;
1609 my $dbh = $self->_get_dbh;
1610 local $dbh->{RaiseError} = 1;
1611 local $dbh->{PrintError} = 0;
1613 $sth->execute foreach 1..$count;
1617 # Make sure statement is finished even if there was an exception.
1618 eval { $sth->finish };
1619 $exception = $@ unless $exception;
1621 $self->throw_exception($exception) if $exception;
1627 my ($self, $source, @args) = @_;
1629 my $bind_attrs = $self->source_bind_attributes($source);
1631 return $self->_execute('update' => [], $source, $bind_attrs, @args);
1636 my ($self, $source, @args) = @_;
1638 my $bind_attrs = $self->source_bind_attributes($source);
1640 return $self->_execute('delete' => [], $source, $bind_attrs, @args);
1643 # We were sent here because the $rs contains a complex search
1644 # which will require a subquery to select the correct rows
1645 # (i.e. joined or limited resultsets, or non-introspectable conditions)
1647 # Generating a single PK column subquery is trivial and supported
1648 # by all RDBMS. However if we have a multicolumn PK, things get ugly.
1649 # Look at _multipk_update_delete()
1650 sub _subq_update_delete {
1652 my ($rs, $op, $values) = @_;
1654 my $rsrc = $rs->result_source;
1656 # quick check if we got a sane rs on our hands
1657 my @pcols = $rsrc->_pri_cols;
1659 my $sel = $rs->_resolved_attrs->{select};
1660 $sel = [ $sel ] unless ref $sel eq 'ARRAY';
1663 join ("\x00", map { join '.', $rs->{attrs}{alias}, $_ } sort @pcols)
1665 join ("\x00", sort @$sel )
1667 $self->throw_exception (
1668 '_subq_update_delete can not be called on resultsets selecting columns other than the primary keys'
1675 $op eq 'update' ? $values : (),
1676 { $pcols[0] => { -in => $rs->as_query } },
1681 return $self->_multipk_update_delete (@_);
1685 # ANSI SQL does not provide a reliable way to perform a multicol-PK
1686 # resultset update/delete involving subqueries. So by default resort
1687 # to simple (and inefficient) delete_all style per-row opearations,
1688 # while allowing specific storages to override this with a faster
1691 sub _multipk_update_delete {
1692 return shift->_per_row_update_delete (@_);
1695 # This is the default loop used to delete/update rows for multi PK
1696 # resultsets, and used by mysql exclusively (because it can't do anything
1699 # We do not use $row->$op style queries, because resultset update/delete
1700 # is not expected to cascade (this is what delete_all/update_all is for).
1702 # There should be no race conditions as the entire operation is rolled
1705 sub _per_row_update_delete {
1707 my ($rs, $op, $values) = @_;
1709 my $rsrc = $rs->result_source;
1710 my @pcols = $rsrc->_pri_cols;
1712 my $guard = $self->txn_scope_guard;
1714 # emulate the return value of $sth->execute for non-selects
1715 my $row_cnt = '0E0';
1717 my $subrs_cur = $rs->cursor;
1718 my @all_pk = $subrs_cur->all;
1719 for my $pks ( @all_pk) {
1722 for my $i (0.. $#pcols) {
1723 $cond->{$pcols[$i]} = $pks->[$i];
1728 $op eq 'update' ? $values : (),
1743 # localization is neccessary as
1744 # 1) there is no infrastructure to pass this around before SQLA2
1745 # 2) _select_args sets it and _prep_for_execute consumes it
1746 my $sql_maker = $self->sql_maker;
1747 local $sql_maker->{_dbic_rs_attrs};
1749 return $self->_execute($self->_select_args(@_));
1752 sub _select_args_to_query {
1755 # localization is neccessary as
1756 # 1) there is no infrastructure to pass this around before SQLA2
1757 # 2) _select_args sets it and _prep_for_execute consumes it
1758 my $sql_maker = $self->sql_maker;
1759 local $sql_maker->{_dbic_rs_attrs};
1761 # my ($op, $bind, $ident, $bind_attrs, $select, $cond, $order, $rows, $offset)
1762 # = $self->_select_args($ident, $select, $cond, $attrs);
1763 my ($op, $bind, $ident, $bind_attrs, @args) =
1764 $self->_select_args(@_);
1766 # my ($sql, $prepared_bind) = $self->_prep_for_execute($op, $bind, $ident, [ $select, $cond, $order, $rows, $offset ]);
1767 my ($sql, $prepared_bind) = $self->_prep_for_execute($op, $bind, $ident, \@args);
1768 $prepared_bind ||= [];
1771 ? ($sql, $prepared_bind, $bind_attrs)
1772 : \[ "($sql)", @$prepared_bind ]
1777 my ($self, $ident, $select, $where, $attrs) = @_;
1779 my ($alias2source, $rs_alias) = $self->_resolve_ident_sources ($ident);
1781 my $sql_maker = $self->sql_maker;
1782 $sql_maker->{_dbic_rs_attrs} = {
1787 $rs_alias && $alias2source->{$rs_alias}
1788 ? ( _source_handle => $alias2source->{$rs_alias}->handle )
1793 # calculate bind_attrs before possible $ident mangling
1794 my $bind_attrs = {};
1795 for my $alias (keys %$alias2source) {
1796 my $bindtypes = $self->source_bind_attributes ($alias2source->{$alias}) || {};
1797 for my $col (keys %$bindtypes) {
1799 my $fqcn = join ('.', $alias, $col);
1800 $bind_attrs->{$fqcn} = $bindtypes->{$col} if $bindtypes->{$col};
1802 # Unqialified column names are nice, but at the same time can be
1803 # rather ambiguous. What we do here is basically go along with
1804 # the loop, adding an unqualified column slot to $bind_attrs,
1805 # alongside the fully qualified name. As soon as we encounter
1806 # another column by that name (which would imply another table)
1807 # we unset the unqualified slot and never add any info to it
1808 # to avoid erroneous type binding. If this happens the users
1809 # only choice will be to fully qualify his column name
1811 if (exists $bind_attrs->{$col}) {
1812 $bind_attrs->{$col} = {};
1815 $bind_attrs->{$col} = $bind_attrs->{$fqcn};
1822 $attrs->{software_limit}
1824 $sql_maker->_default_limit_syntax eq "GenericSubQ"
1826 $attrs->{software_limit} = 1;
1829 $self->throw_exception("rows attribute must be positive if present")
1830 if (defined($attrs->{rows}) && !($attrs->{rows} > 0));
1832 # MySQL actually recommends this approach. I cringe.
1833 $attrs->{rows} = 2**48 if not defined $attrs->{rows} and defined $attrs->{offset};
1838 # see if we need to tear the prefetch apart otherwise delegate the limiting to the
1839 # storage, unless software limit was requested
1842 ( $attrs->{rows} && keys %{$attrs->{collapse}} )
1844 # limited prefetch with RNO subqueries
1848 $sql_maker->limit_dialect eq 'RowNumberOver'
1850 $attrs->{_prefetch_select}
1852 @{$attrs->{_prefetch_select}}
1856 ( $attrs->{group_by}
1858 @{$attrs->{group_by}}
1860 $attrs->{_prefetch_select}
1862 @{$attrs->{_prefetch_select}}
1865 ($ident, $select, $where, $attrs)
1866 = $self->_adjust_select_args_for_complex_prefetch ($ident, $select, $where, $attrs);
1870 ($attrs->{rows} || $attrs->{offset})
1872 $sql_maker->limit_dialect eq 'RowNumberOver'
1874 (ref $ident eq 'ARRAY' && @$ident > 1) # indicates a join
1876 scalar $self->_parse_order_by ($attrs->{order_by})
1878 # the RNO limit dialect above mangles the SQL such that the join gets lost
1879 # wrap a subquery here
1881 push @limit, delete @{$attrs}{qw/rows offset/};
1883 my $subq = $self->_select_args_to_query (
1891 -alias => $attrs->{alias},
1892 -source_handle => $ident->[0]{-source_handle},
1893 $attrs->{alias} => $subq,
1896 # all part of the subquery now
1897 delete @{$attrs}{qw/order_by group_by having/};
1901 elsif (! $attrs->{software_limit} ) {
1902 push @limit, $attrs->{rows}, $attrs->{offset};
1905 # try to simplify the joinmap further (prune unreferenced type-single joins)
1906 $ident = $self->_prune_unused_joins ($ident, $select, $where, $attrs);
1909 # This would be the point to deflate anything found in $where
1910 # (and leave $attrs->{bind} intact). Problem is - inflators historically
1911 # expect a row object. And all we have is a resultsource (it is trivial
1912 # to extract deflator coderefs via $alias2source above).
1914 # I don't see a way forward other than changing the way deflators are
1915 # invoked, and that's just bad...
1919 { $attrs->{$_} ? ( $_ => $attrs->{$_} ) : () }
1920 (qw/order_by group_by having/ )
1923 return ('select', $attrs->{bind}, $ident, $bind_attrs, $select, $where, $order, @limit);
1926 # Returns a counting SELECT for a simple count
1927 # query. Abstracted so that a storage could override
1928 # this to { count => 'firstcol' } or whatever makes
1929 # sense as a performance optimization
1931 #my ($self, $source, $rs_attrs) = @_;
1932 return { count => '*' };
1935 # Returns a SELECT which will end up in the subselect
1936 # There may or may not be a group_by, as the subquery
1937 # might have been called to accomodate a limit
1939 # Most databases would be happy with whatever ends up
1940 # here, but some choke in various ways.
1942 sub _subq_count_select {
1943 my ($self, $source, $rs_attrs) = @_;
1945 if (my $groupby = $rs_attrs->{group_by}) {
1947 my $avail_columns = $self->_resolve_column_info ($rs_attrs->{from});
1950 for my $sel (@{$rs_attrs->{select}}) {
1951 if (ref $sel eq 'HASH' and $sel->{-as}) {
1952 $sel_index->{$sel->{-as}} = $sel;
1957 for my $g_part (@$groupby) {
1958 if (ref $g_part or $avail_columns->{$g_part}) {
1959 push @selection, $g_part;
1961 elsif ($sel_index->{$g_part}) {
1962 push @selection, $sel_index->{$g_part};
1965 $self->throw_exception ("group_by criteria '$g_part' not contained within current resultset source(s)");
1972 my @pcols = map { join '.', $rs_attrs->{alias}, $_ } ($source->primary_columns);
1973 return @pcols ? \@pcols : [ 1 ];
1976 sub source_bind_attributes {
1977 my ($self, $source) = @_;
1979 my $bind_attributes;
1980 foreach my $column ($source->columns) {
1982 my $data_type = $source->column_info($column)->{data_type} || '';
1983 $bind_attributes->{$column} = $self->bind_attribute_by_data_type($data_type)
1987 return $bind_attributes;
1994 =item Arguments: $ident, $select, $condition, $attrs
1998 Handle a SQL select statement.
2004 my ($ident, $select, $condition, $attrs) = @_;
2005 return $self->cursor_class->new($self, \@_, $attrs);
2010 my ($rv, $sth, @bind) = $self->_select(@_);
2011 my @row = $sth->fetchrow_array;
2012 my @nextrow = $sth->fetchrow_array if @row;
2013 if(@row && @nextrow) {
2014 carp "Query returned more than one row. SQL that returns multiple rows is DEPRECATED for ->find and ->single";
2016 # Need to call finish() to work round broken DBDs
2025 =item Arguments: $sql
2029 Returns a L<DBI> sth (statement handle) for the supplied SQL.
2034 my ($self, $dbh, $sql) = @_;
2036 # 3 is the if_active parameter which avoids active sth re-use
2037 my $sth = $self->disable_sth_caching
2038 ? $dbh->prepare($sql)
2039 : $dbh->prepare_cached($sql, {}, 3);
2041 # XXX You would think RaiseError would make this impossible,
2042 # but apparently that's not true :(
2043 $self->throw_exception($dbh->errstr) if !$sth;
2049 my ($self, $sql) = @_;
2050 $self->dbh_do('_dbh_sth', $sql); # retry over disconnects
2053 sub _dbh_columns_info_for {
2054 my ($self, $dbh, $table) = @_;
2056 if ($dbh->can('column_info')) {
2059 my ($schema,$tab) = $table =~ /^(.+?)\.(.+)$/ ? ($1,$2) : (undef,$table);
2060 my $sth = $dbh->column_info( undef,$schema, $tab, '%' );
2062 while ( my $info = $sth->fetchrow_hashref() ){
2064 $column_info{data_type} = $info->{TYPE_NAME};
2065 $column_info{size} = $info->{COLUMN_SIZE};
2066 $column_info{is_nullable} = $info->{NULLABLE} ? 1 : 0;
2067 $column_info{default_value} = $info->{COLUMN_DEF};
2068 my $col_name = $info->{COLUMN_NAME};
2069 $col_name =~ s/^\"(.*)\"$/$1/;
2071 $result{$col_name} = \%column_info;
2074 return \%result if !$@ && scalar keys %result;
2078 my $sth = $dbh->prepare($self->sql_maker->select($table, undef, \'1 = 0'));
2080 my @columns = @{$sth->{NAME_lc}};
2081 for my $i ( 0 .. $#columns ){
2083 $column_info{data_type} = $sth->{TYPE}->[$i];
2084 $column_info{size} = $sth->{PRECISION}->[$i];
2085 $column_info{is_nullable} = $sth->{NULLABLE}->[$i] ? 1 : 0;
2087 if ($column_info{data_type} =~ m/^(.*?)\((.*?)\)$/) {
2088 $column_info{data_type} = $1;
2089 $column_info{size} = $2;
2092 $result{$columns[$i]} = \%column_info;
2096 foreach my $col (keys %result) {
2097 my $colinfo = $result{$col};
2098 my $type_num = $colinfo->{data_type};
2100 if(defined $type_num && $dbh->can('type_info')) {
2101 my $type_info = $dbh->type_info($type_num);
2102 $type_name = $type_info->{TYPE_NAME} if $type_info;
2103 $colinfo->{data_type} = $type_name if $type_name;
2110 sub columns_info_for {
2111 my ($self, $table) = @_;
2112 $self->_dbh_columns_info_for ($self->_get_dbh, $table);
2115 =head2 last_insert_id
2117 Return the row id of the last insert.
2121 sub _dbh_last_insert_id {
2122 my ($self, $dbh, $source, $col) = @_;
2124 my $id = eval { $dbh->last_insert_id (undef, undef, $source->name, $col) };
2126 return $id if defined $id;
2128 my $class = ref $self;
2129 $self->throw_exception ("No storage specific _dbh_last_insert_id() method implemented in $class, and the generic DBI::last_insert_id() failed");
2132 sub last_insert_id {
2134 $self->_dbh_last_insert_id ($self->_dbh, @_);
2137 =head2 _native_data_type
2141 =item Arguments: $type_name
2145 This API is B<EXPERIMENTAL>, will almost definitely change in the future, and
2146 currently only used by L<::AutoCast|DBIx::Class::Storage::DBI::AutoCast> and
2147 L<::Sybase::ASE|DBIx::Class::Storage::DBI::Sybase::ASE>.
2149 The default implementation returns C<undef>, implement in your Storage driver if
2150 you need this functionality.
2152 Should map types from other databases to the native RDBMS type, for example
2153 C<VARCHAR2> to C<VARCHAR>.
2155 Types with modifiers should map to the underlying data type. For example,
2156 C<INTEGER AUTO_INCREMENT> should become C<INTEGER>.
2158 Composite types should map to the container type, for example
2159 C<ENUM(foo,bar,baz)> becomes C<ENUM>.
2163 sub _native_data_type {
2164 #my ($self, $data_type) = @_;
2168 # Check if placeholders are supported at all
2169 sub _placeholders_supported {
2171 my $dbh = $self->_get_dbh;
2173 # some drivers provide a $dbh attribute (e.g. Sybase and $dbh->{syb_dynamic_supported})
2174 # but it is inaccurate more often than not
2176 local $dbh->{PrintError} = 0;
2177 local $dbh->{RaiseError} = 1;
2178 $dbh->do('select ?', {}, 1);
2183 # Check if placeholders bound to non-string types throw exceptions
2185 sub _typeless_placeholders_supported {
2187 my $dbh = $self->_get_dbh;
2190 local $dbh->{PrintError} = 0;
2191 local $dbh->{RaiseError} = 1;
2192 # this specifically tests a bind that is NOT a string
2193 $dbh->do('select 1 where 1 = ?', {}, 1);
2200 Returns the database driver name.
2205 shift->_get_dbh->{Driver}->{Name};
2208 =head2 bind_attribute_by_data_type
2210 Given a datatype from column info, returns a database specific bind
2211 attribute for C<< $dbh->bind_param($val,$attribute) >> or nothing if we will
2212 let the database planner just handle it.
2214 Generally only needed for special case column types, like bytea in postgres.
2218 sub bind_attribute_by_data_type {
2222 =head2 is_datatype_numeric
2224 Given a datatype from column_info, returns a boolean value indicating if
2225 the current RDBMS considers it a numeric value. This controls how
2226 L<DBIx::Class::Row/set_column> decides whether to mark the column as
2227 dirty - when the datatype is deemed numeric a C<< != >> comparison will
2228 be performed instead of the usual C<eq>.
2232 sub is_datatype_numeric {
2233 my ($self, $dt) = @_;
2235 return 0 unless $dt;
2237 return $dt =~ /^ (?:
2238 numeric | int(?:eger)? | (?:tiny|small|medium|big)int | dec(?:imal)? | real | float | double (?: \s+ precision)? | (?:big)?serial
2243 =head2 create_ddl_dir
2247 =item Arguments: $schema \@databases, $version, $directory, $preversion, \%sqlt_args
2251 Creates a SQL file based on the Schema, for each of the specified
2252 database engines in C<\@databases> in the given directory.
2253 (note: specify L<SQL::Translator> names, not L<DBI> driver names).
2255 Given a previous version number, this will also create a file containing
2256 the ALTER TABLE statements to transform the previous schema into the
2257 current one. Note that these statements may contain C<DROP TABLE> or
2258 C<DROP COLUMN> statements that can potentially destroy data.
2260 The file names are created using the C<ddl_filename> method below, please
2261 override this method in your schema if you would like a different file
2262 name format. For the ALTER file, the same format is used, replacing
2263 $version in the name with "$preversion-$version".
2265 See L<SQL::Translator/METHODS> for a list of values for C<\%sqlt_args>.
2266 The most common value for this would be C<< { add_drop_table => 1 } >>
2267 to have the SQL produced include a C<DROP TABLE> statement for each table
2268 created. For quoting purposes supply C<quote_table_names> and
2269 C<quote_field_names>.
2271 If no arguments are passed, then the following default values are assumed:
2275 =item databases - ['MySQL', 'SQLite', 'PostgreSQL']
2277 =item version - $schema->schema_version
2279 =item directory - './'
2281 =item preversion - <none>
2285 By default, C<\%sqlt_args> will have
2287 { add_drop_table => 1, ignore_constraint_names => 1, ignore_index_names => 1 }
2289 merged with the hash passed in. To disable any of those features, pass in a
2290 hashref like the following
2292 { ignore_constraint_names => 0, # ... other options }
2295 WARNING: You are strongly advised to check all SQL files created, before applying
2300 sub create_ddl_dir {
2301 my ($self, $schema, $databases, $version, $dir, $preversion, $sqltargs) = @_;
2304 carp "No directory given, using ./\n";
2308 $self->throw_exception ("Directory '$dir' does not exist\n") unless(-d $dir);
2310 $databases ||= ['MySQL', 'SQLite', 'PostgreSQL'];
2311 $databases = [ $databases ] if(ref($databases) ne 'ARRAY');
2313 my $schema_version = $schema->schema_version || '1.x';
2314 $version ||= $schema_version;
2317 add_drop_table => 1,
2318 ignore_constraint_names => 1,
2319 ignore_index_names => 1,
2323 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy')) {
2324 $self->throw_exception("Can't create a ddl file without " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2327 my $sqlt = SQL::Translator->new( $sqltargs );
2329 $sqlt->parser('SQL::Translator::Parser::DBIx::Class');
2330 my $sqlt_schema = $sqlt->translate({ data => $schema })
2331 or $self->throw_exception ($sqlt->error);
2333 foreach my $db (@$databases) {
2335 $sqlt->{schema} = $sqlt_schema;
2336 $sqlt->producer($db);
2339 my $filename = $schema->ddl_filename($db, $version, $dir);
2340 if (-e $filename && ($version eq $schema_version )) {
2341 # if we are dumping the current version, overwrite the DDL
2342 carp "Overwriting existing DDL file - $filename";
2346 my $output = $sqlt->translate;
2348 carp("Failed to translate to $db, skipping. (" . $sqlt->error . ")");
2351 if(!open($file, ">$filename")) {
2352 $self->throw_exception("Can't open $filename for writing ($!)");
2355 print $file $output;
2358 next unless ($preversion);
2360 require SQL::Translator::Diff;
2362 my $prefilename = $schema->ddl_filename($db, $preversion, $dir);
2363 if(!-e $prefilename) {
2364 carp("No previous schema file found ($prefilename)");
2368 my $difffile = $schema->ddl_filename($db, $version, $dir, $preversion);
2370 carp("Overwriting existing diff file - $difffile");
2376 my $t = SQL::Translator->new($sqltargs);
2381 or $self->throw_exception ($t->error);
2383 my $out = $t->translate( $prefilename )
2384 or $self->throw_exception ($t->error);
2386 $source_schema = $t->schema;
2388 $source_schema->name( $prefilename )
2389 unless ( $source_schema->name );
2392 # The "new" style of producers have sane normalization and can support
2393 # diffing a SQL file against a DBIC->SQLT schema. Old style ones don't
2394 # And we have to diff parsed SQL against parsed SQL.
2395 my $dest_schema = $sqlt_schema;
2397 unless ( "SQL::Translator::Producer::$db"->can('preprocess_schema') ) {
2398 my $t = SQL::Translator->new($sqltargs);
2403 or $self->throw_exception ($t->error);
2405 my $out = $t->translate( $filename )
2406 or $self->throw_exception ($t->error);
2408 $dest_schema = $t->schema;
2410 $dest_schema->name( $filename )
2411 unless $dest_schema->name;
2414 my $diff = SQL::Translator::Diff::schema_diff($source_schema, $db,
2418 if(!open $file, ">$difffile") {
2419 $self->throw_exception("Can't write to $difffile ($!)");
2427 =head2 deployment_statements
2431 =item Arguments: $schema, $type, $version, $directory, $sqlt_args
2435 Returns the statements used by L</deploy> and L<DBIx::Class::Schema/deploy>.
2437 The L<SQL::Translator> (not L<DBI>) database driver name can be explicitly
2438 provided in C<$type>, otherwise the result of L</sqlt_type> is used as default.
2440 C<$directory> is used to return statements from files in a previously created
2441 L</create_ddl_dir> directory and is optional. The filenames are constructed
2442 from L<DBIx::Class::Schema/ddl_filename>, the schema name and the C<$version>.
2444 If no C<$directory> is specified then the statements are constructed on the
2445 fly using L<SQL::Translator> and C<$version> is ignored.
2447 See L<SQL::Translator/METHODS> for a list of values for C<$sqlt_args>.
2451 sub deployment_statements {
2452 my ($self, $schema, $type, $version, $dir, $sqltargs) = @_;
2453 $type ||= $self->sqlt_type;
2454 $version ||= $schema->schema_version || '1.x';
2456 my $filename = $schema->ddl_filename($type, $version, $dir);
2460 open($file, "<$filename")
2461 or $self->throw_exception("Can't open $filename ($!)");
2464 return join('', @rows);
2467 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy') ) {
2468 $self->throw_exception("Can't deploy without a ddl_dir or " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2471 # sources needs to be a parser arg, but for simplicty allow at top level
2473 $sqltargs->{parser_args}{sources} = delete $sqltargs->{sources}
2474 if exists $sqltargs->{sources};
2476 my $tr = SQL::Translator->new(
2477 producer => "SQL::Translator::Producer::${type}",
2479 parser => 'SQL::Translator::Parser::DBIx::Class',
2486 @ret = $tr->translate;
2489 $ret[0] = $tr->translate;
2492 $self->throw_exception( 'Unable to produce deployment statements: ' . $tr->error)
2493 unless (@ret && defined $ret[0]);
2495 return $wa ? @ret : $ret[0];
2499 my ($self, $schema, $type, $sqltargs, $dir) = @_;
2502 return if($line =~ /^--/);
2504 # next if($line =~ /^DROP/m);
2505 return if($line =~ /^BEGIN TRANSACTION/m);
2506 return if($line =~ /^COMMIT/m);
2507 return if $line =~ /^\s+$/; # skip whitespace only
2508 $self->_query_start($line);
2510 # do a dbh_do cycle here, as we need some error checking in
2511 # place (even though we will ignore errors)
2512 $self->dbh_do (sub { $_[1]->do($line) });
2515 carp qq{$@ (running "${line}")};
2517 $self->_query_end($line);
2519 my @statements = $schema->deployment_statements($type, undef, $dir, { %{ $sqltargs || {} }, no_comments => 1 } );
2520 if (@statements > 1) {
2521 foreach my $statement (@statements) {
2522 $deploy->( $statement );
2525 elsif (@statements == 1) {
2526 foreach my $line ( split(";\n", $statements[0])) {
2532 =head2 datetime_parser
2534 Returns the datetime parser class
2538 sub datetime_parser {
2540 return $self->{datetime_parser} ||= do {
2541 $self->build_datetime_parser(@_);
2545 =head2 datetime_parser_type
2547 Defines (returns) the datetime parser class - currently hardwired to
2548 L<DateTime::Format::MySQL>
2552 sub datetime_parser_type { "DateTime::Format::MySQL"; }
2554 =head2 build_datetime_parser
2556 See L</datetime_parser>
2560 sub build_datetime_parser {
2562 my $type = $self->datetime_parser_type(@_);
2563 $self->ensure_class_loaded ($type);
2568 =head2 is_replicating
2570 A boolean that reports if a particular L<DBIx::Class::Storage::DBI> is set to
2571 replicate from a master database. Default is undef, which is the result
2572 returned by databases that don't support replication.
2576 sub is_replicating {
2581 =head2 lag_behind_master
2583 Returns a number that represents a certain amount of lag behind a master db
2584 when a given storage is replicating. The number is database dependent, but
2585 starts at zero and increases with the amount of lag. Default in undef
2589 sub lag_behind_master {
2593 =head2 relname_to_table_alias
2597 =item Arguments: $relname, $join_count
2601 L<DBIx::Class> uses L<DBIx::Class::Relationship> names as table aliases in
2604 This hook is to allow specific L<DBIx::Class::Storage> drivers to change the
2605 way these aliases are named.
2607 The default behavior is C<< "$relname_$join_count" if $join_count > 1 >>,
2608 otherwise C<"$relname">.
2612 sub relname_to_table_alias {
2613 my ($self, $relname, $join_count) = @_;
2615 my $alias = ($join_count && $join_count > 1 ?
2616 join('_', $relname, $join_count) : $relname);
2624 $self->_verify_pid if $self->_dbh;
2626 # some databases need this to stop spewing warnings
2627 if (my $dbh = $self->_dbh) {
2630 %{ $dbh->{CachedKids} } = ();
2642 =head2 DBIx::Class and AutoCommit
2644 DBIx::Class can do some wonderful magic with handling exceptions,
2645 disconnections, and transactions when you use C<< AutoCommit => 1 >>
2646 (the default) combined with C<txn_do> for transaction support.
2648 If you set C<< AutoCommit => 0 >> in your connect info, then you are always
2649 in an assumed transaction between commits, and you're telling us you'd
2650 like to manage that manually. A lot of the magic protections offered by
2651 this module will go away. We can't protect you from exceptions due to database
2652 disconnects because we don't know anything about how to restart your
2653 transactions. You're on your own for handling all sorts of exceptional
2654 cases if you choose the C<< AutoCommit => 0 >> path, just as you would
2660 Matt S. Trout <mst@shadowcatsystems.co.uk>
2662 Andy Grundman <andy@hybridized.org>
2666 You may distribute this code under the same terms as Perl itself.