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();
22 __PACKAGE__->mk_group_accessors('simple' =>
23 qw/_connect_info _dbi_connect_info _dbh _sql_maker _sql_maker_opts _conn_pid
24 _conn_tid transaction_depth _dbh_autocommit _driver_determined savepoints
28 # the values for these accessors are picked out (and deleted) from
29 # the attribute hashref passed to connect_info
30 my @storage_options = qw/
31 on_connect_call on_disconnect_call on_connect_do on_disconnect_do
32 disable_sth_caching unsafe auto_savepoint
34 __PACKAGE__->mk_group_accessors('simple' => @storage_options);
37 # default cursor class, overridable in connect_info attributes
38 __PACKAGE__->cursor_class('DBIx::Class::Storage::DBI::Cursor');
40 __PACKAGE__->mk_group_accessors('inherited' => qw/
42 _supports_insert_returning
44 __PACKAGE__->sql_maker_class('DBIx::Class::SQLAHacks');
46 # Each of these methods need _determine_driver called before itself
47 # in order to function reliably. This is a purely DRY optimization
48 my @rdbms_specific_methods = qw/
63 for my $meth (@rdbms_specific_methods) {
65 my $orig = __PACKAGE__->can ($meth)
69 no warnings qw/redefine/;
70 *{__PACKAGE__ ."::$meth"} = Sub::Name::subname $meth => sub {
71 if (not $_[0]->_driver_determined) {
72 $_[0]->_determine_driver;
73 goto $_[0]->can($meth);
82 DBIx::Class::Storage::DBI - DBI storage handler
86 my $schema = MySchema->connect('dbi:SQLite:my.db');
88 $schema->storage->debug(1);
90 my @stuff = $schema->storage->dbh_do(
92 my ($storage, $dbh, @args) = @_;
93 $dbh->do("DROP TABLE authors");
98 $schema->resultset('Book')->search({
99 written_on => $schema->storage->datetime_parser->format_datetime(DateTime->now)
104 This class represents the connection to an RDBMS via L<DBI>. See
105 L<DBIx::Class::Storage> for general information. This pod only
106 documents DBI-specific methods and behaviors.
113 my $new = shift->next::method(@_);
115 $new->transaction_depth(0);
116 $new->_sql_maker_opts({});
117 $new->{savepoints} = [];
118 $new->{_in_dbh_do} = 0;
119 $new->{_dbh_gen} = 0;
121 # read below to see what this does
122 $new->_arm_global_destructor;
127 # This is hack to work around perl shooting stuff in random
128 # order on exit(). If we do not walk the remaining storage
129 # objects in an END block, there is a *small but real* chance
130 # of a fork()ed child to kill the parent's shared DBI handle,
131 # *before perl reaches the DESTROY in this package*
132 # Yes, it is ugly and effective.
134 my %seek_and_destroy;
136 sub _arm_global_destructor {
138 my $key = Scalar::Util::refaddr ($self);
139 $seek_and_destroy{$key} = $self;
140 Scalar::Util::weaken ($seek_and_destroy{$key});
144 local $?; # just in case the DBI destructor changes it somehow
146 # destroy just the object if not native to this process/thread
147 $_->_preserve_foreign_dbh for (grep
149 values %seek_and_destroy
157 # destroy just the object if not native to this process/thread
158 $self->_preserve_foreign_dbh;
160 # some databases need this to stop spewing warnings
161 if (my $dbh = $self->_dbh) {
163 %{ $dbh->{CachedKids} } = ();
171 sub _preserve_foreign_dbh {
174 return unless $self->_dbh;
178 return unless $self->_dbh;
184 # handle pid changes correctly - do not destroy parent's connection
188 return if ( defined $self->_conn_pid and $self->_conn_pid == $$ );
190 $self->_dbh->{InactiveDestroy} = 1;
197 # very similar to above, but seems to FAIL if I set InactiveDestroy
201 if ( ! defined $self->_conn_tid ) {
204 elsif ( $self->_conn_tid == threads->tid ) {
205 return; # same thread
208 #$self->_dbh->{InactiveDestroy} = 1; # why does t/51threads.t fail...?
218 This method is normally called by L<DBIx::Class::Schema/connection>, which
219 encapsulates its argument list in an arrayref before passing them here.
221 The argument list may contain:
227 The same 4-element argument set one would normally pass to
228 L<DBI/connect>, optionally followed by
229 L<extra attributes|/DBIx::Class specific connection attributes>
230 recognized by DBIx::Class:
232 $connect_info_args = [ $dsn, $user, $password, \%dbi_attributes?, \%extra_attributes? ];
236 A single code reference which returns a connected
237 L<DBI database handle|DBI/connect> optionally followed by
238 L<extra attributes|/DBIx::Class specific connection attributes> recognized
241 $connect_info_args = [ sub { DBI->connect (...) }, \%extra_attributes? ];
245 A single hashref with all the attributes and the dsn/user/password
248 $connect_info_args = [{
256 $connect_info_args = [{
257 dbh_maker => sub { DBI->connect (...) },
262 This is particularly useful for L<Catalyst> based applications, allowing the
263 following config (L<Config::General> style):
268 dsn dbi:mysql:database=test
275 The C<dsn>/C<user>/C<password> combination can be substituted by the
276 C<dbh_maker> key whose value is a coderef that returns a connected
277 L<DBI database handle|DBI/connect>
281 Please note that the L<DBI> docs recommend that you always explicitly
282 set C<AutoCommit> to either I<0> or I<1>. L<DBIx::Class> further
283 recommends that it be set to I<1>, and that you perform transactions
284 via our L<DBIx::Class::Schema/txn_do> method. L<DBIx::Class> will set it
285 to I<1> if you do not do explicitly set it to zero. This is the default
286 for most DBDs. See L</DBIx::Class and AutoCommit> for details.
288 =head3 DBIx::Class specific connection attributes
290 In addition to the standard L<DBI|DBI/ATTRIBUTES_COMMON_TO_ALL_HANDLES>
291 L<connection|DBI/Database_Handle_Attributes> attributes, DBIx::Class recognizes
292 the following connection options. These options can be mixed in with your other
293 L<DBI> connection attributes, or placed in a separate hashref
294 (C<\%extra_attributes>) as shown above.
296 Every time C<connect_info> is invoked, any previous settings for
297 these options will be cleared before setting the new ones, regardless of
298 whether any options are specified in the new C<connect_info>.
305 Specifies things to do immediately after connecting or re-connecting to
306 the database. Its value may contain:
312 This contains one SQL statement to execute.
314 =item an array reference
316 This contains SQL statements to execute in order. Each element contains
317 a string or a code reference that returns a string.
319 =item a code reference
321 This contains some code to execute. Unlike code references within an
322 array reference, its return value is ignored.
326 =item on_disconnect_do
328 Takes arguments in the same form as L</on_connect_do> and executes them
329 immediately before disconnecting from the database.
331 Note, this only runs if you explicitly call L</disconnect> on the
334 =item on_connect_call
336 A more generalized form of L</on_connect_do> that calls the specified
337 C<connect_call_METHOD> methods in your storage driver.
339 on_connect_do => 'select 1'
343 on_connect_call => [ [ do_sql => 'select 1' ] ]
345 Its values may contain:
351 Will call the C<connect_call_METHOD> method.
353 =item a code reference
355 Will execute C<< $code->($storage) >>
357 =item an array reference
359 Each value can be a method name or code reference.
361 =item an array of arrays
363 For each array, the first item is taken to be the C<connect_call_> method name
364 or code reference, and the rest are parameters to it.
368 Some predefined storage methods you may use:
374 Executes a SQL string or a code reference that returns a SQL string. This is
375 what L</on_connect_do> and L</on_disconnect_do> use.
383 Will execute the scalar as SQL.
387 Taken to be arguments to L<DBI/do>, the SQL string optionally followed by the
388 attributes hashref and bind values.
390 =item a code reference
392 Will execute C<< $code->($storage) >> and execute the return array refs as
399 Execute any statements necessary to initialize the database session to return
400 and accept datetime/timestamp values used with
401 L<DBIx::Class::InflateColumn::DateTime>.
403 Only necessary for some databases, see your specific storage driver for
404 implementation details.
408 =item on_disconnect_call
410 Takes arguments in the same form as L</on_connect_call> and executes them
411 immediately before disconnecting from the database.
413 Calls the C<disconnect_call_METHOD> methods as opposed to the
414 C<connect_call_METHOD> methods called by L</on_connect_call>.
416 Note, this only runs if you explicitly call L</disconnect> on the
419 =item disable_sth_caching
421 If set to a true value, this option will disable the caching of
422 statement handles via L<DBI/prepare_cached>.
426 Sets the limit dialect. This is useful for JDBC-bridge among others
427 where the remote SQL-dialect cannot be determined by the name of the
428 driver alone. See also L<SQL::Abstract::Limit>.
432 Specifies what characters to use to quote table and column names. If
433 you use this you will want to specify L</name_sep> as well.
435 C<quote_char> expects either a single character, in which case is it
436 is placed on either side of the table/column name, or an arrayref of length
437 2 in which case the table/column name is placed between the elements.
439 For example under MySQL you should use C<< quote_char => '`' >>, and for
440 SQL Server you should use C<< quote_char => [qw/[ ]/] >>.
444 This only needs to be used in conjunction with C<quote_char>, and is used to
445 specify the character that separates elements (schemas, tables, columns) from
446 each other. In most cases this is simply a C<.>.
448 The consequences of not supplying this value is that L<SQL::Abstract>
449 will assume DBIx::Class' uses of aliases to be complete column
450 names. The output will look like I<"me.name"> when it should actually
455 This Storage driver normally installs its own C<HandleError>, sets
456 C<RaiseError> and C<ShowErrorStatement> on, and sets C<PrintError> off on
457 all database handles, including those supplied by a coderef. It does this
458 so that it can have consistent and useful error behavior.
460 If you set this option to a true value, Storage will not do its usual
461 modifications to the database handle's attributes, and instead relies on
462 the settings in your connect_info DBI options (or the values you set in
463 your connection coderef, in the case that you are connecting via coderef).
465 Note that your custom settings can cause Storage to malfunction,
466 especially if you set a C<HandleError> handler that suppresses exceptions
467 and/or disable C<RaiseError>.
471 If this option is true, L<DBIx::Class> will use savepoints when nesting
472 transactions, making it possible to recover from failure in the inner
473 transaction without having to abort all outer transactions.
477 Use this argument to supply a cursor class other than the default
478 L<DBIx::Class::Storage::DBI::Cursor>.
482 Some real-life examples of arguments to L</connect_info> and
483 L<DBIx::Class::Schema/connect>
485 # Simple SQLite connection
486 ->connect_info([ 'dbi:SQLite:./foo.db' ]);
489 ->connect_info([ sub { DBI->connect(...) } ]);
491 # Connect via subref in hashref
493 dbh_maker => sub { DBI->connect(...) },
494 on_connect_do => 'alter session ...',
497 # A bit more complicated
504 { quote_char => q{"}, name_sep => q{.} },
508 # Equivalent to the previous example
514 { AutoCommit => 1, quote_char => q{"}, name_sep => q{.} },
518 # Same, but with hashref as argument
519 # See parse_connect_info for explanation
522 dsn => 'dbi:Pg:dbname=foo',
524 password => 'my_pg_password',
531 # Subref + DBIx::Class-specific connection options
534 sub { DBI->connect(...) },
538 on_connect_do => ['SET search_path TO myschema,otherschema,public'],
539 disable_sth_caching => 1,
549 my ($self, $info) = @_;
551 return $self->_connect_info if !$info;
553 $self->_connect_info($info); # copy for _connect_info
555 $info = $self->_normalize_connect_info($info)
556 if ref $info eq 'ARRAY';
558 for my $storage_opt (keys %{ $info->{storage_options} }) {
559 my $value = $info->{storage_options}{$storage_opt};
561 $self->$storage_opt($value);
564 # Kill sql_maker/_sql_maker_opts, so we get a fresh one with only
565 # the new set of options
566 $self->_sql_maker(undef);
567 $self->_sql_maker_opts({});
569 for my $sql_maker_opt (keys %{ $info->{sql_maker_options} }) {
570 my $value = $info->{sql_maker_options}{$sql_maker_opt};
572 $self->_sql_maker_opts->{$sql_maker_opt} = $value;
576 %{ $self->_default_dbi_connect_attributes || {} },
577 %{ $info->{attributes} || {} },
580 my @args = @{ $info->{arguments} };
582 $self->_dbi_connect_info([@args,
583 %attrs && !(ref $args[0] eq 'CODE') ? \%attrs : ()]);
585 return $self->_connect_info;
588 sub _normalize_connect_info {
589 my ($self, $info_arg) = @_;
592 my @args = @$info_arg; # take a shallow copy for further mutilation
594 # combine/pre-parse arguments depending on invocation style
597 if (ref $args[0] eq 'CODE') { # coderef with optional \%extra_attributes
598 %attrs = %{ $args[1] || {} };
601 elsif (ref $args[0] eq 'HASH') { # single hashref (i.e. Catalyst config)
602 %attrs = %{$args[0]};
604 if (my $code = delete $attrs{dbh_maker}) {
607 my @ignored = grep { delete $attrs{$_} } (qw/dsn user password/);
610 'Attribute(s) %s in connect_info were ignored, as they can not be applied '
611 . "to the result of 'dbh_maker'",
613 join (', ', map { "'$_'" } (@ignored) ),
618 @args = delete @attrs{qw/dsn user password/};
621 else { # otherwise assume dsn/user/password + \%attrs + \%extra_attrs
623 % { $args[3] || {} },
624 % { $args[4] || {} },
626 @args = @args[0,1,2];
629 $info{arguments} = \@args;
631 my @storage_opts = grep exists $attrs{$_},
632 @storage_options, 'cursor_class';
634 @{ $info{storage_options} }{@storage_opts} =
635 delete @attrs{@storage_opts} if @storage_opts;
637 my @sql_maker_opts = grep exists $attrs{$_},
638 qw/limit_dialect quote_char name_sep/;
640 @{ $info{sql_maker_options} }{@sql_maker_opts} =
641 delete @attrs{@sql_maker_opts} if @sql_maker_opts;
643 $info{attributes} = \%attrs if %attrs;
648 sub _default_dbi_connect_attributes {
658 This method is deprecated in favour of setting via L</connect_info>.
662 =head2 on_disconnect_do
664 This method is deprecated in favour of setting via L</connect_info>.
668 sub _parse_connect_do {
669 my ($self, $type) = @_;
671 my $val = $self->$type;
672 return () if not defined $val;
677 push @res, [ 'do_sql', $val ];
678 } elsif (ref($val) eq 'CODE') {
680 } elsif (ref($val) eq 'ARRAY') {
681 push @res, map { [ 'do_sql', $_ ] } @$val;
683 $self->throw_exception("Invalid type for $type: ".ref($val));
691 Arguments: ($subref | $method_name), @extra_coderef_args?
693 Execute the given $subref or $method_name using the new exception-based
694 connection management.
696 The first two arguments will be the storage object that C<dbh_do> was called
697 on and a database handle to use. Any additional arguments will be passed
698 verbatim to the called subref as arguments 2 and onwards.
700 Using this (instead of $self->_dbh or $self->dbh) ensures correct
701 exception handling and reconnection (or failover in future subclasses).
703 Your subref should have no side-effects outside of the database, as
704 there is the potential for your subref to be partially double-executed
705 if the database connection was stale/dysfunctional.
709 my @stuff = $schema->storage->dbh_do(
711 my ($storage, $dbh, @cols) = @_;
712 my $cols = join(q{, }, @cols);
713 $dbh->selectrow_array("SELECT $cols FROM foo");
724 my $dbh = $self->_get_dbh;
726 return $self->$code($dbh, @_)
727 if ( $self->{_in_dbh_do} || $self->{transaction_depth} );
729 local $self->{_in_dbh_do} = 1;
733 $self->$code ($dbh, @args);
735 $self->throw_exception($_) if $self->connected;
737 # We were not connected - reconnect and retry, but let any
738 # exception fall right through this time
739 carp "Retrying $code after catching disconnected exception: $_"
740 if $ENV{DBIC_DBIRETRY_DEBUG};
742 $self->_populate_dbh;
743 $self->$code($self->_dbh, @args);
747 # This is basically a blend of dbh_do above and DBIx::Class::Storage::txn_do.
748 # It also informs dbh_do to bypass itself while under the direction of txn_do,
749 # via $self->{_in_dbh_do} (this saves some redundant eval and errorcheck, etc)
754 ref $coderef eq 'CODE' or $self->throw_exception
755 ('$coderef must be a CODE reference');
757 return $coderef->(@_) if $self->{transaction_depth} && ! $self->auto_savepoint;
759 local $self->{_in_dbh_do} = 1;
762 my $want_array = wantarray;
773 @result = $coderef->(@args);
775 elsif(defined $want_array) {
776 $result[0] = $coderef->(@args);
786 if(! defined $exception) { return $want_array ? @result : $result[0] }
788 if($tried++ || $self->connected) {
789 my $rollback_exception;
790 try { $self->txn_rollback } catch { $rollback_exception = shift };
791 if(defined $rollback_exception) {
792 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
793 $self->throw_exception($exception) # propagate nested rollback
794 if $rollback_exception =~ /$exception_class/;
796 $self->throw_exception(
797 "Transaction aborted: ${exception}. "
798 . "Rollback failed: ${rollback_exception}"
801 $self->throw_exception($exception)
804 # We were not connected, and was first try - reconnect and retry
806 carp "Retrying $coderef after catching disconnected exception: $exception"
807 if $ENV{DBIC_DBIRETRY_DEBUG};
808 $self->_populate_dbh;
814 Our C<disconnect> method also performs a rollback first if the
815 database is not in C<AutoCommit> mode.
825 push @actions, ( $self->on_disconnect_call || () );
826 push @actions, $self->_parse_connect_do ('on_disconnect_do');
828 $self->_do_connection_actions(disconnect_call_ => $_) for @actions;
830 $self->_dbh_rollback unless $self->_dbh_autocommit;
832 %{ $self->_dbh->{CachedKids} } = ();
833 $self->_dbh->disconnect;
839 =head2 with_deferred_fk_checks
843 =item Arguments: C<$coderef>
845 =item Return Value: The return value of $coderef
849 Storage specific method to run the code ref with FK checks deferred or
850 in MySQL's case disabled entirely.
854 # Storage subclasses should override this
855 sub with_deferred_fk_checks {
856 my ($self, $sub) = @_;
864 =item Arguments: none
866 =item Return Value: 1|0
870 Verifies that the current database handle is active and ready to execute
871 an SQL statement (e.g. the connection did not get stale, server is still
872 answering, etc.) This method is used internally by L</dbh>.
878 return 0 unless $self->_seems_connected;
881 local $self->_dbh->{RaiseError} = 1;
886 sub _seems_connected {
889 $self->_preserve_foreign_dbh;
891 my $dbh = $self->_dbh
894 return $dbh->FETCH('Active');
900 my $dbh = $self->_dbh or return 0;
905 sub ensure_connected {
908 unless ($self->connected) {
909 $self->_populate_dbh;
915 Returns a C<$dbh> - a data base handle of class L<DBI>. The returned handle
916 is guaranteed to be healthy by implicitly calling L</connected>, and if
917 necessary performing a reconnection before returning. Keep in mind that this
918 is very B<expensive> on some database engines. Consider using L</dbh_do>
926 if (not $self->_dbh) {
927 $self->_populate_dbh;
929 $self->ensure_connected;
934 # this is the internal "get dbh or connect (don't check)" method
937 $self->_preserve_foreign_dbh;
938 $self->_populate_dbh unless $self->_dbh;
942 sub _sql_maker_args {
947 array_datatypes => 1,
948 limit_dialect => $self->_get_dbh,
949 %{$self->_sql_maker_opts}
955 unless ($self->_sql_maker) {
956 my $sql_maker_class = $self->sql_maker_class;
957 $self->ensure_class_loaded ($sql_maker_class);
958 $self->_sql_maker($sql_maker_class->new( $self->_sql_maker_args ));
960 return $self->_sql_maker;
963 # nothing to do by default
970 my @info = @{$self->_dbi_connect_info || []};
971 $self->_dbh(undef); # in case ->connected failed we might get sent here
972 $self->_server_info_hash (undef);
973 $self->_dbh($self->_connect(@info));
975 $self->_conn_pid($$);
976 $self->_conn_tid(threads->tid) if $INC{'threads.pm'};
978 $self->_determine_driver;
980 # Always set the transaction depth on connect, since
981 # there is no transaction in progress by definition
982 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
984 $self->_run_connection_actions unless $self->{_in_determine_driver};
987 sub _run_connection_actions {
991 push @actions, ( $self->on_connect_call || () );
992 push @actions, $self->_parse_connect_do ('on_connect_do');
994 $self->_do_connection_actions(connect_call_ => $_) for @actions;
1000 unless ($self->_server_info_hash) {
1004 my $server_version = try { $self->_get_server_version };
1006 if (defined $server_version) {
1007 $info{dbms_version} = $server_version;
1009 my ($numeric_version) = $server_version =~ /^([\d\.]+)/;
1010 my @verparts = split (/\./, $numeric_version);
1016 # consider only up to 3 version parts, iff not more than 3 digits
1018 while (@verparts && @use_parts < 3) {
1019 my $p = shift @verparts;
1021 push @use_parts, $p;
1023 push @use_parts, 0 while @use_parts < 3;
1025 $info{normalized_dbms_version} = sprintf "%d.%03d%03d", @use_parts;
1029 $self->_server_info_hash(\%info);
1032 return $self->_server_info_hash
1035 sub _get_server_version {
1036 shift->_get_dbh->get_info(18);
1039 sub _determine_driver {
1042 if ((not $self->_driver_determined) && (not $self->{_in_determine_driver})) {
1043 my $started_connected = 0;
1044 local $self->{_in_determine_driver} = 1;
1046 if (ref($self) eq __PACKAGE__) {
1048 if ($self->_dbh) { # we are connected
1049 $driver = $self->_dbh->{Driver}{Name};
1050 $started_connected = 1;
1052 # if connect_info is a CODEREF, we have no choice but to connect
1053 if (ref $self->_dbi_connect_info->[0] &&
1054 Scalar::Util::reftype($self->_dbi_connect_info->[0]) eq 'CODE') {
1055 $self->_populate_dbh;
1056 $driver = $self->_dbh->{Driver}{Name};
1059 # try to use dsn to not require being connected, the driver may still
1060 # force a connection in _rebless to determine version
1061 # (dsn may not be supplied at all if all we do is make a mock-schema)
1062 my $dsn = $self->_dbi_connect_info->[0] || $ENV{DBI_DSN} || '';
1063 ($driver) = $dsn =~ /dbi:([^:]+):/i;
1064 $driver ||= $ENV{DBI_DRIVER};
1069 my $storage_class = "DBIx::Class::Storage::DBI::${driver}";
1070 if ($self->load_optional_class($storage_class)) {
1071 mro::set_mro($storage_class, 'c3');
1072 bless $self, $storage_class;
1078 $self->_driver_determined(1);
1080 $self->_init; # run driver-specific initializations
1082 $self->_run_connection_actions
1083 if !$started_connected && defined $self->_dbh;
1087 sub _do_connection_actions {
1089 my $method_prefix = shift;
1092 if (not ref($call)) {
1093 my $method = $method_prefix . $call;
1095 } elsif (ref($call) eq 'CODE') {
1097 } elsif (ref($call) eq 'ARRAY') {
1098 if (ref($call->[0]) ne 'ARRAY') {
1099 $self->_do_connection_actions($method_prefix, $_) for @$call;
1101 $self->_do_connection_actions($method_prefix, @$_) for @$call;
1104 $self->throw_exception (sprintf ("Don't know how to process conection actions of type '%s'", ref($call)) );
1110 sub connect_call_do_sql {
1112 $self->_do_query(@_);
1115 sub disconnect_call_do_sql {
1117 $self->_do_query(@_);
1120 # override in db-specific backend when necessary
1121 sub connect_call_datetime_setup { 1 }
1124 my ($self, $action) = @_;
1126 if (ref $action eq 'CODE') {
1127 $action = $action->($self);
1128 $self->_do_query($_) foreach @$action;
1131 # Most debuggers expect ($sql, @bind), so we need to exclude
1132 # the attribute hash which is the second argument to $dbh->do
1133 # furthermore the bind values are usually to be presented
1134 # as named arrayref pairs, so wrap those here too
1135 my @do_args = (ref $action eq 'ARRAY') ? (@$action) : ($action);
1136 my $sql = shift @do_args;
1137 my $attrs = shift @do_args;
1138 my @bind = map { [ undef, $_ ] } @do_args;
1140 $self->_query_start($sql, @bind);
1141 $self->_get_dbh->do($sql, $attrs, @do_args);
1142 $self->_query_end($sql, @bind);
1149 my ($self, @info) = @_;
1151 $self->throw_exception("You failed to provide any connection info")
1154 my ($old_connect_via, $dbh);
1156 if ($INC{'Apache/DBI.pm'} && $ENV{MOD_PERL}) {
1157 $old_connect_via = $DBI::connect_via;
1158 $DBI::connect_via = 'connect';
1162 if(ref $info[0] eq 'CODE') {
1163 $dbh = $info[0]->();
1166 $dbh = DBI->connect(@info);
1173 unless ($self->unsafe) {
1174 my $weak_self = $self;
1175 Scalar::Util::weaken($weak_self);
1176 $dbh->{HandleError} = sub {
1178 $weak_self->throw_exception("DBI Exception: $_[0]");
1181 # the handler may be invoked by something totally out of
1183 croak ("DBI Exception: $_[0]");
1186 $dbh->{ShowErrorStatement} = 1;
1187 $dbh->{RaiseError} = 1;
1188 $dbh->{PrintError} = 0;
1192 $self->throw_exception("DBI Connection failed: $_")
1195 $DBI::connect_via = $old_connect_via if $old_connect_via;
1198 $self->_dbh_autocommit($dbh->{AutoCommit});
1203 my ($self, $name) = @_;
1205 $name = $self->_svp_generate_name
1206 unless defined $name;
1208 $self->throw_exception ("You can't use savepoints outside a transaction")
1209 if $self->{transaction_depth} == 0;
1211 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1212 unless $self->can('_svp_begin');
1214 push @{ $self->{savepoints} }, $name;
1216 $self->debugobj->svp_begin($name) if $self->debug;
1218 return $self->_svp_begin($name);
1222 my ($self, $name) = @_;
1224 $self->throw_exception ("You can't use savepoints outside a transaction")
1225 if $self->{transaction_depth} == 0;
1227 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1228 unless $self->can('_svp_release');
1230 if (defined $name) {
1231 $self->throw_exception ("Savepoint '$name' does not exist")
1232 unless grep { $_ eq $name } @{ $self->{savepoints} };
1234 # Dig through the stack until we find the one we are releasing. This keeps
1235 # the stack up to date.
1238 do { $svp = pop @{ $self->{savepoints} } } while $svp ne $name;
1240 $name = pop @{ $self->{savepoints} };
1243 $self->debugobj->svp_release($name) if $self->debug;
1245 return $self->_svp_release($name);
1249 my ($self, $name) = @_;
1251 $self->throw_exception ("You can't use savepoints outside a transaction")
1252 if $self->{transaction_depth} == 0;
1254 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1255 unless $self->can('_svp_rollback');
1257 if (defined $name) {
1258 # If they passed us a name, verify that it exists in the stack
1259 unless(grep({ $_ eq $name } @{ $self->{savepoints} })) {
1260 $self->throw_exception("Savepoint '$name' does not exist!");
1263 # Dig through the stack until we find the one we are releasing. This keeps
1264 # the stack up to date.
1265 while(my $s = pop(@{ $self->{savepoints} })) {
1266 last if($s eq $name);
1268 # Add the savepoint back to the stack, as a rollback doesn't remove the
1269 # named savepoint, only everything after it.
1270 push(@{ $self->{savepoints} }, $name);
1272 # We'll assume they want to rollback to the last savepoint
1273 $name = $self->{savepoints}->[-1];
1276 $self->debugobj->svp_rollback($name) if $self->debug;
1278 return $self->_svp_rollback($name);
1281 sub _svp_generate_name {
1284 return 'savepoint_'.scalar(@{ $self->{'savepoints'} });
1290 # this means we have not yet connected and do not know the AC status
1291 # (e.g. coderef $dbh)
1292 $self->ensure_connected if (! defined $self->_dbh_autocommit);
1294 if($self->{transaction_depth} == 0) {
1295 $self->debugobj->txn_begin()
1297 $self->_dbh_begin_work;
1299 elsif ($self->auto_savepoint) {
1302 $self->{transaction_depth}++;
1305 sub _dbh_begin_work {
1308 # if the user is utilizing txn_do - good for him, otherwise we need to
1309 # ensure that the $dbh is healthy on BEGIN.
1310 # We do this via ->dbh_do instead of ->dbh, so that the ->dbh "ping"
1311 # will be replaced by a failure of begin_work itself (which will be
1312 # then retried on reconnect)
1313 if ($self->{_in_dbh_do}) {
1314 $self->_dbh->begin_work;
1316 $self->dbh_do(sub { $_[1]->begin_work });
1322 if ($self->{transaction_depth} == 1) {
1323 $self->debugobj->txn_commit()
1326 $self->{transaction_depth} = 0
1327 if $self->_dbh_autocommit;
1329 elsif($self->{transaction_depth} > 1) {
1330 $self->{transaction_depth}--;
1332 if $self->auto_savepoint;
1338 my $dbh = $self->_dbh
1339 or $self->throw_exception('cannot COMMIT on a disconnected handle');
1345 my $dbh = $self->_dbh;
1347 if ($self->{transaction_depth} == 1) {
1348 $self->debugobj->txn_rollback()
1350 $self->{transaction_depth} = 0
1351 if $self->_dbh_autocommit;
1352 $self->_dbh_rollback;
1354 elsif($self->{transaction_depth} > 1) {
1355 $self->{transaction_depth}--;
1356 if ($self->auto_savepoint) {
1357 $self->svp_rollback;
1362 die DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION->new;
1366 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
1368 if ($_ !~ /$exception_class/) {
1369 # ensure that a failed rollback resets the transaction depth
1370 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
1373 $self->throw_exception($_)
1379 my $dbh = $self->_dbh
1380 or $self->throw_exception('cannot ROLLBACK on a disconnected handle');
1384 # This used to be the top-half of _execute. It was split out to make it
1385 # easier to override in NoBindVars without duping the rest. It takes up
1386 # all of _execute's args, and emits $sql, @bind.
1387 sub _prep_for_execute {
1388 my ($self, $op, $extra_bind, $ident, $args) = @_;
1390 if( Scalar::Util::blessed($ident) && $ident->isa("DBIx::Class::ResultSource") ) {
1391 $ident = $ident->from();
1394 my ($sql, @bind) = $self->sql_maker->$op($ident, @$args);
1397 map { ref $_ eq 'ARRAY' ? $_ : [ '!!dummy', $_ ] } @$extra_bind)
1399 return ($sql, \@bind);
1403 sub _fix_bind_params {
1404 my ($self, @bind) = @_;
1406 ### Turn @bind from something like this:
1407 ### ( [ "artist", 1 ], [ "cdid", 1, 3 ] )
1409 ### ( "'1'", "'1'", "'3'" )
1412 if ( defined( $_ && $_->[1] ) ) {
1413 map { qq{'$_'}; } @{$_}[ 1 .. $#$_ ];
1420 my ( $self, $sql, @bind ) = @_;
1422 if ( $self->debug ) {
1423 @bind = $self->_fix_bind_params(@bind);
1425 $self->debugobj->query_start( $sql, @bind );
1430 my ( $self, $sql, @bind ) = @_;
1432 if ( $self->debug ) {
1433 @bind = $self->_fix_bind_params(@bind);
1434 $self->debugobj->query_end( $sql, @bind );
1439 my ($self, $dbh, $op, $extra_bind, $ident, $bind_attributes, @args) = @_;
1441 my ($sql, $bind) = $self->_prep_for_execute($op, $extra_bind, $ident, \@args);
1443 $self->_query_start( $sql, @$bind );
1445 my $sth = $self->sth($sql,$op);
1447 my $placeholder_index = 1;
1449 foreach my $bound (@$bind) {
1450 my $attributes = {};
1451 my($column_name, @data) = @$bound;
1453 if ($bind_attributes) {
1454 $attributes = $bind_attributes->{$column_name}
1455 if defined $bind_attributes->{$column_name};
1458 foreach my $data (@data) {
1459 my $ref = ref $data;
1460 $data = $ref && $ref ne 'ARRAY' ? ''.$data : $data; # stringify args (except arrayrefs)
1462 $sth->bind_param($placeholder_index, $data, $attributes);
1463 $placeholder_index++;
1467 # Can this fail without throwing an exception anyways???
1468 my $rv = $sth->execute();
1469 $self->throw_exception(
1470 $sth->errstr || $sth->err || 'Unknown error: execute() returned false, but error flags were not set...'
1473 $self->_query_end( $sql, @$bind );
1475 return (wantarray ? ($rv, $sth, @$bind) : $rv);
1480 $self->dbh_do('_dbh_execute', @_); # retry over disconnects
1483 sub _prefetch_insert_auto_nextvals {
1484 my ($self, $source, $to_insert) = @_;
1488 foreach my $col ( $source->columns ) {
1489 if ( !defined $to_insert->{$col} ) {
1490 my $col_info = $source->column_info($col);
1492 if ( $col_info->{auto_nextval} ) {
1493 $upd->{$col} = $to_insert->{$col} = $self->_sequence_fetch(
1495 $col_info->{sequence} ||=
1496 $self->_dbh_get_autoinc_seq($self->_get_dbh, $source, $col)
1507 my ($source, $to_insert, $opts) = @_;
1509 my $updated_cols = $self->_prefetch_insert_auto_nextvals (@_);
1511 my $bind_attributes = $self->source_bind_attributes($source);
1513 my ($rv, $sth) = $self->_execute('insert' => [], $source, $bind_attributes, $to_insert, $opts);
1515 if ($opts->{returning}) {
1516 my @ret_cols = @{$opts->{returning}};
1518 my @ret_vals = try {
1519 local $SIG{__WARN__} = sub {};
1520 my @r = $sth->fetchrow_array;
1526 @ret{@ret_cols} = @ret_vals if (@ret_vals);
1534 return $updated_cols;
1537 ## Currently it is assumed that all values passed will be "normal", i.e. not
1538 ## scalar refs, or at least, all the same type as the first set, the statement is
1539 ## only prepped once.
1541 my ($self, $source, $cols, $data) = @_;
1544 @colvalues{@$cols} = (0..$#$cols);
1546 for my $i (0..$#$cols) {
1547 my $first_val = $data->[0][$i];
1548 next unless ref $first_val eq 'SCALAR';
1550 $colvalues{ $cols->[$i] } = $first_val;
1553 # check for bad data and stringify stringifiable objects
1554 my $bad_slice = sub {
1555 my ($msg, $col_idx, $slice_idx) = @_;
1556 $self->throw_exception(sprintf "%s for column '%s' in populate slice:\n%s",
1560 local $Data::Dumper::Maxdepth = 1; # don't dump objects, if any
1561 Data::Dumper::Concise::Dumper({
1562 map { $cols->[$_] => $data->[$slice_idx][$_] } (0 .. $#$cols)
1568 for my $datum_idx (0..$#$data) {
1569 my $datum = $data->[$datum_idx];
1571 for my $col_idx (0..$#$cols) {
1572 my $val = $datum->[$col_idx];
1573 my $sqla_bind = $colvalues{ $cols->[$col_idx] };
1574 my $is_literal_sql = (ref $sqla_bind) eq 'SCALAR';
1576 if ($is_literal_sql) {
1578 $bad_slice->('bind found where literal SQL expected', $col_idx, $datum_idx);
1580 elsif ((my $reftype = ref $val) ne 'SCALAR') {
1581 $bad_slice->("$reftype reference found where literal SQL expected",
1582 $col_idx, $datum_idx);
1584 elsif ($$val ne $$sqla_bind){
1585 $bad_slice->("inconsistent literal SQL value, expecting: '$$sqla_bind'",
1586 $col_idx, $datum_idx);
1589 elsif (my $reftype = ref $val) {
1591 if (overload::Method($val, '""')) {
1592 $datum->[$col_idx] = "".$val;
1595 $bad_slice->("$reftype reference found where bind expected",
1596 $col_idx, $datum_idx);
1602 my ($sql, $bind) = $self->_prep_for_execute (
1603 'insert', undef, $source, [\%colvalues]
1607 my $empty_bind = 1 if (not @bind) &&
1608 (grep { ref $_ eq 'SCALAR' } values %colvalues) == @$cols;
1610 if ((not @bind) && (not $empty_bind)) {
1611 $self->throw_exception(
1612 'Cannot insert_bulk without support for placeholders'
1616 # neither _execute_array, nor _execute_inserts_with_no_binds are
1617 # atomic (even if _execute _array is a single call). Thus a safety
1619 my $guard = $self->txn_scope_guard;
1621 $self->_query_start( $sql, ['__BULK__'] );
1622 my $sth = $self->sth($sql);
1625 # bind_param_array doesn't work if there are no binds
1626 $self->_dbh_execute_inserts_with_no_binds( $sth, scalar @$data );
1629 # @bind = map { ref $_ ? ''.$_ : $_ } @bind; # stringify args
1630 $self->_execute_array( $source, $sth, \@bind, $cols, $data );
1634 $self->_query_end( $sql, ['__BULK__'] );
1638 return (wantarray ? ($rv, $sth, @bind) : $rv);
1641 sub _execute_array {
1642 my ($self, $source, $sth, $bind, $cols, $data, @extra) = @_;
1644 ## This must be an arrayref, else nothing works!
1645 my $tuple_status = [];
1647 ## Get the bind_attributes, if any exist
1648 my $bind_attributes = $self->source_bind_attributes($source);
1650 ## Bind the values and execute
1651 my $placeholder_index = 1;
1653 foreach my $bound (@$bind) {
1655 my $attributes = {};
1656 my ($column_name, $data_index) = @$bound;
1658 if( $bind_attributes ) {
1659 $attributes = $bind_attributes->{$column_name}
1660 if defined $bind_attributes->{$column_name};
1663 my @data = map { $_->[$data_index] } @$data;
1665 $sth->bind_param_array(
1668 (%$attributes ? $attributes : ()),
1670 $placeholder_index++;
1675 $rv = $self->_dbh_execute_array($sth, $tuple_status, @extra);
1681 # Statement must finish even if there was an exception.
1686 $err = shift unless defined $err
1691 if (! defined $err and $sth->err);
1695 ++$i while $i <= $#$tuple_status && !ref $tuple_status->[$i];
1697 $self->throw_exception("Unexpected populate error: $err")
1698 if ($i > $#$tuple_status);
1700 $self->throw_exception(sprintf "%s for populate slice:\n%s",
1701 ($tuple_status->[$i][1] || $err),
1702 Data::Dumper::Concise::Dumper({
1703 map { $cols->[$_] => $data->[$i][$_] } (0 .. $#$cols)
1711 sub _dbh_execute_array {
1712 my ($self, $sth, $tuple_status, @extra) = @_;
1714 return $sth->execute_array({ArrayTupleStatus => $tuple_status});
1717 sub _dbh_execute_inserts_with_no_binds {
1718 my ($self, $sth, $count) = @_;
1722 my $dbh = $self->_get_dbh;
1723 local $dbh->{RaiseError} = 1;
1724 local $dbh->{PrintError} = 0;
1726 $sth->execute foreach 1..$count;
1732 # Make sure statement is finished even if there was an exception.
1737 $err = shift unless defined $err;
1741 $self->throw_exception($err) if defined $err;
1747 my ($self, $source, @args) = @_;
1749 my $bind_attrs = $self->source_bind_attributes($source);
1751 return $self->_execute('update' => [], $source, $bind_attrs, @args);
1756 my ($self, $source, @args) = @_;
1758 my $bind_attrs = $self->source_bind_attributes($source);
1760 return $self->_execute('delete' => [], $source, $bind_attrs, @args);
1763 # We were sent here because the $rs contains a complex search
1764 # which will require a subquery to select the correct rows
1765 # (i.e. joined or limited resultsets, or non-introspectable conditions)
1767 # Generating a single PK column subquery is trivial and supported
1768 # by all RDBMS. However if we have a multicolumn PK, things get ugly.
1769 # Look at _multipk_update_delete()
1770 sub _subq_update_delete {
1772 my ($rs, $op, $values) = @_;
1774 my $rsrc = $rs->result_source;
1776 # quick check if we got a sane rs on our hands
1777 my @pcols = $rsrc->_pri_cols;
1779 my $sel = $rs->_resolved_attrs->{select};
1780 $sel = [ $sel ] unless ref $sel eq 'ARRAY';
1783 join ("\x00", map { join '.', $rs->{attrs}{alias}, $_ } sort @pcols)
1785 join ("\x00", sort @$sel )
1787 $self->throw_exception (
1788 '_subq_update_delete can not be called on resultsets selecting columns other than the primary keys'
1795 $op eq 'update' ? $values : (),
1796 { $pcols[0] => { -in => $rs->as_query } },
1801 return $self->_multipk_update_delete (@_);
1805 # ANSI SQL does not provide a reliable way to perform a multicol-PK
1806 # resultset update/delete involving subqueries. So by default resort
1807 # to simple (and inefficient) delete_all style per-row opearations,
1808 # while allowing specific storages to override this with a faster
1811 sub _multipk_update_delete {
1812 return shift->_per_row_update_delete (@_);
1815 # This is the default loop used to delete/update rows for multi PK
1816 # resultsets, and used by mysql exclusively (because it can't do anything
1819 # We do not use $row->$op style queries, because resultset update/delete
1820 # is not expected to cascade (this is what delete_all/update_all is for).
1822 # There should be no race conditions as the entire operation is rolled
1825 sub _per_row_update_delete {
1827 my ($rs, $op, $values) = @_;
1829 my $rsrc = $rs->result_source;
1830 my @pcols = $rsrc->_pri_cols;
1832 my $guard = $self->txn_scope_guard;
1834 # emulate the return value of $sth->execute for non-selects
1835 my $row_cnt = '0E0';
1837 my $subrs_cur = $rs->cursor;
1838 my @all_pk = $subrs_cur->all;
1839 for my $pks ( @all_pk) {
1842 for my $i (0.. $#pcols) {
1843 $cond->{$pcols[$i]} = $pks->[$i];
1848 $op eq 'update' ? $values : (),
1862 $self->_execute($self->_select_args(@_));
1865 sub _select_args_to_query {
1868 # my ($op, $bind, $ident, $bind_attrs, $select, $cond, $rs_attrs, $rows, $offset)
1869 # = $self->_select_args($ident, $select, $cond, $attrs);
1870 my ($op, $bind, $ident, $bind_attrs, @args) =
1871 $self->_select_args(@_);
1873 # my ($sql, $prepared_bind) = $self->_prep_for_execute($op, $bind, $ident, [ $select, $cond, $rs_attrs, $rows, $offset ]);
1874 my ($sql, $prepared_bind) = $self->_prep_for_execute($op, $bind, $ident, \@args);
1875 $prepared_bind ||= [];
1878 ? ($sql, $prepared_bind, $bind_attrs)
1879 : \[ "($sql)", @$prepared_bind ]
1884 my ($self, $ident, $select, $where, $attrs) = @_;
1886 my $sql_maker = $self->sql_maker;
1887 my ($alias2source, $rs_alias) = $self->_resolve_ident_sources ($ident);
1894 $rs_alias && $alias2source->{$rs_alias}
1895 ? ( _rsroot_source_handle => $alias2source->{$rs_alias}->handle )
1900 # calculate bind_attrs before possible $ident mangling
1901 my $bind_attrs = {};
1902 for my $alias (keys %$alias2source) {
1903 my $bindtypes = $self->source_bind_attributes ($alias2source->{$alias}) || {};
1904 for my $col (keys %$bindtypes) {
1906 my $fqcn = join ('.', $alias, $col);
1907 $bind_attrs->{$fqcn} = $bindtypes->{$col} if $bindtypes->{$col};
1909 # Unqialified column names are nice, but at the same time can be
1910 # rather ambiguous. What we do here is basically go along with
1911 # the loop, adding an unqualified column slot to $bind_attrs,
1912 # alongside the fully qualified name. As soon as we encounter
1913 # another column by that name (which would imply another table)
1914 # we unset the unqualified slot and never add any info to it
1915 # to avoid erroneous type binding. If this happens the users
1916 # only choice will be to fully qualify his column name
1918 if (exists $bind_attrs->{$col}) {
1919 $bind_attrs->{$col} = {};
1922 $bind_attrs->{$col} = $bind_attrs->{$fqcn};
1928 if (defined $attrs->{rows}) {
1929 $self->throw_exception("rows attribute must be positive if present")
1930 unless $attrs->{rows} > 0;
1932 elsif (defined $attrs->{offset}) {
1933 # MySQL actually recommends this approach. I cringe.
1934 $attrs->{rows} = 2**32;
1939 # see if we need to tear the prefetch apart otherwise delegate the limiting to the
1940 # storage, unless software limit was requested
1943 ( $attrs->{rows} && keys %{$attrs->{collapse}} )
1945 # grouped prefetch (to satisfy group_by == select)
1946 ( $attrs->{group_by}
1948 @{$attrs->{group_by}}
1950 $attrs->{_prefetch_select}
1952 @{$attrs->{_prefetch_select}}
1955 ($ident, $select, $where, $attrs)
1956 = $self->_adjust_select_args_for_complex_prefetch ($ident, $select, $where, $attrs);
1958 elsif (! $attrs->{software_limit} ) {
1959 push @limit, $attrs->{rows}, $attrs->{offset};
1962 # try to simplify the joinmap further (prune unreferenced type-single joins)
1963 $ident = $self->_prune_unused_joins ($ident, $select, $where, $attrs);
1966 # This would be the point to deflate anything found in $where
1967 # (and leave $attrs->{bind} intact). Problem is - inflators historically
1968 # expect a row object. And all we have is a resultsource (it is trivial
1969 # to extract deflator coderefs via $alias2source above).
1971 # I don't see a way forward other than changing the way deflators are
1972 # invoked, and that's just bad...
1975 return ('select', $attrs->{bind}, $ident, $bind_attrs, $select, $where, $attrs, @limit);
1978 # Returns a counting SELECT for a simple count
1979 # query. Abstracted so that a storage could override
1980 # this to { count => 'firstcol' } or whatever makes
1981 # sense as a performance optimization
1983 #my ($self, $source, $rs_attrs) = @_;
1984 return { count => '*' };
1988 sub source_bind_attributes {
1989 my ($self, $source) = @_;
1991 my $bind_attributes;
1992 foreach my $column ($source->columns) {
1994 my $data_type = $source->column_info($column)->{data_type} || '';
1995 $bind_attributes->{$column} = $self->bind_attribute_by_data_type($data_type)
1999 return $bind_attributes;
2006 =item Arguments: $ident, $select, $condition, $attrs
2010 Handle a SQL select statement.
2016 my ($ident, $select, $condition, $attrs) = @_;
2017 return $self->cursor_class->new($self, \@_, $attrs);
2022 my ($rv, $sth, @bind) = $self->_select(@_);
2023 my @row = $sth->fetchrow_array;
2024 my @nextrow = $sth->fetchrow_array if @row;
2025 if(@row && @nextrow) {
2026 carp "Query returned more than one row. SQL that returns multiple rows is DEPRECATED for ->find and ->single";
2028 # Need to call finish() to work round broken DBDs
2037 =item Arguments: $sql
2041 Returns a L<DBI> sth (statement handle) for the supplied SQL.
2046 my ($self, $dbh, $sql) = @_;
2048 # 3 is the if_active parameter which avoids active sth re-use
2049 my $sth = $self->disable_sth_caching
2050 ? $dbh->prepare($sql)
2051 : $dbh->prepare_cached($sql, {}, 3);
2053 # XXX You would think RaiseError would make this impossible,
2054 # but apparently that's not true :(
2055 $self->throw_exception($dbh->errstr) if !$sth;
2061 my ($self, $sql) = @_;
2062 $self->dbh_do('_dbh_sth', $sql); # retry over disconnects
2065 sub _dbh_columns_info_for {
2066 my ($self, $dbh, $table) = @_;
2068 if ($dbh->can('column_info')) {
2072 my ($schema,$tab) = $table =~ /^(.+?)\.(.+)$/ ? ($1,$2) : (undef,$table);
2073 my $sth = $dbh->column_info( undef,$schema, $tab, '%' );
2075 while ( my $info = $sth->fetchrow_hashref() ){
2077 $column_info{data_type} = $info->{TYPE_NAME};
2078 $column_info{size} = $info->{COLUMN_SIZE};
2079 $column_info{is_nullable} = $info->{NULLABLE} ? 1 : 0;
2080 $column_info{default_value} = $info->{COLUMN_DEF};
2081 my $col_name = $info->{COLUMN_NAME};
2082 $col_name =~ s/^\"(.*)\"$/$1/;
2084 $result{$col_name} = \%column_info;
2089 return \%result if !$caught && scalar keys %result;
2093 my $sth = $dbh->prepare($self->sql_maker->select($table, undef, \'1 = 0'));
2095 my @columns = @{$sth->{NAME_lc}};
2096 for my $i ( 0 .. $#columns ){
2098 $column_info{data_type} = $sth->{TYPE}->[$i];
2099 $column_info{size} = $sth->{PRECISION}->[$i];
2100 $column_info{is_nullable} = $sth->{NULLABLE}->[$i] ? 1 : 0;
2102 if ($column_info{data_type} =~ m/^(.*?)\((.*?)\)$/) {
2103 $column_info{data_type} = $1;
2104 $column_info{size} = $2;
2107 $result{$columns[$i]} = \%column_info;
2111 foreach my $col (keys %result) {
2112 my $colinfo = $result{$col};
2113 my $type_num = $colinfo->{data_type};
2115 if(defined $type_num && $dbh->can('type_info')) {
2116 my $type_info = $dbh->type_info($type_num);
2117 $type_name = $type_info->{TYPE_NAME} if $type_info;
2118 $colinfo->{data_type} = $type_name if $type_name;
2125 sub columns_info_for {
2126 my ($self, $table) = @_;
2127 $self->_dbh_columns_info_for ($self->_get_dbh, $table);
2130 =head2 last_insert_id
2132 Return the row id of the last insert.
2136 sub _dbh_last_insert_id {
2137 my ($self, $dbh, $source, $col) = @_;
2139 my $id = try { $dbh->last_insert_id (undef, undef, $source->name, $col) };
2141 return $id if defined $id;
2143 my $class = ref $self;
2144 $self->throw_exception ("No storage specific _dbh_last_insert_id() method implemented in $class, and the generic DBI::last_insert_id() failed");
2147 sub last_insert_id {
2149 $self->_dbh_last_insert_id ($self->_dbh, @_);
2152 =head2 _native_data_type
2156 =item Arguments: $type_name
2160 This API is B<EXPERIMENTAL>, will almost definitely change in the future, and
2161 currently only used by L<::AutoCast|DBIx::Class::Storage::DBI::AutoCast> and
2162 L<::Sybase::ASE|DBIx::Class::Storage::DBI::Sybase::ASE>.
2164 The default implementation returns C<undef>, implement in your Storage driver if
2165 you need this functionality.
2167 Should map types from other databases to the native RDBMS type, for example
2168 C<VARCHAR2> to C<VARCHAR>.
2170 Types with modifiers should map to the underlying data type. For example,
2171 C<INTEGER AUTO_INCREMENT> should become C<INTEGER>.
2173 Composite types should map to the container type, for example
2174 C<ENUM(foo,bar,baz)> becomes C<ENUM>.
2178 sub _native_data_type {
2179 #my ($self, $data_type) = @_;
2183 # Check if placeholders are supported at all
2184 sub _placeholders_supported {
2186 my $dbh = $self->_get_dbh;
2188 # some drivers provide a $dbh attribute (e.g. Sybase and $dbh->{syb_dynamic_supported})
2189 # but it is inaccurate more often than not
2191 local $dbh->{PrintError} = 0;
2192 local $dbh->{RaiseError} = 1;
2193 $dbh->do('select ?', {}, 1);
2201 # Check if placeholders bound to non-string types throw exceptions
2203 sub _typeless_placeholders_supported {
2205 my $dbh = $self->_get_dbh;
2208 local $dbh->{PrintError} = 0;
2209 local $dbh->{RaiseError} = 1;
2210 # this specifically tests a bind that is NOT a string
2211 $dbh->do('select 1 where 1 = ?', {}, 1);
2221 Returns the database driver name.
2226 shift->_get_dbh->{Driver}->{Name};
2229 =head2 bind_attribute_by_data_type
2231 Given a datatype from column info, returns a database specific bind
2232 attribute for C<< $dbh->bind_param($val,$attribute) >> or nothing if we will
2233 let the database planner just handle it.
2235 Generally only needed for special case column types, like bytea in postgres.
2239 sub bind_attribute_by_data_type {
2243 =head2 is_datatype_numeric
2245 Given a datatype from column_info, returns a boolean value indicating if
2246 the current RDBMS considers it a numeric value. This controls how
2247 L<DBIx::Class::Row/set_column> decides whether to mark the column as
2248 dirty - when the datatype is deemed numeric a C<< != >> comparison will
2249 be performed instead of the usual C<eq>.
2253 sub is_datatype_numeric {
2254 my ($self, $dt) = @_;
2256 return 0 unless $dt;
2258 return $dt =~ /^ (?:
2259 numeric | int(?:eger)? | (?:tiny|small|medium|big)int | dec(?:imal)? | real | float | double (?: \s+ precision)? | (?:big)?serial
2264 =head2 create_ddl_dir
2268 =item Arguments: $schema \@databases, $version, $directory, $preversion, \%sqlt_args
2272 Creates a SQL file based on the Schema, for each of the specified
2273 database engines in C<\@databases> in the given directory.
2274 (note: specify L<SQL::Translator> names, not L<DBI> driver names).
2276 Given a previous version number, this will also create a file containing
2277 the ALTER TABLE statements to transform the previous schema into the
2278 current one. Note that these statements may contain C<DROP TABLE> or
2279 C<DROP COLUMN> statements that can potentially destroy data.
2281 The file names are created using the C<ddl_filename> method below, please
2282 override this method in your schema if you would like a different file
2283 name format. For the ALTER file, the same format is used, replacing
2284 $version in the name with "$preversion-$version".
2286 See L<SQL::Translator/METHODS> for a list of values for C<\%sqlt_args>.
2287 The most common value for this would be C<< { add_drop_table => 1 } >>
2288 to have the SQL produced include a C<DROP TABLE> statement for each table
2289 created. For quoting purposes supply C<quote_table_names> and
2290 C<quote_field_names>.
2292 If no arguments are passed, then the following default values are assumed:
2296 =item databases - ['MySQL', 'SQLite', 'PostgreSQL']
2298 =item version - $schema->schema_version
2300 =item directory - './'
2302 =item preversion - <none>
2306 By default, C<\%sqlt_args> will have
2308 { add_drop_table => 1, ignore_constraint_names => 1, ignore_index_names => 1 }
2310 merged with the hash passed in. To disable any of those features, pass in a
2311 hashref like the following
2313 { ignore_constraint_names => 0, # ... other options }
2316 WARNING: You are strongly advised to check all SQL files created, before applying
2321 sub create_ddl_dir {
2322 my ($self, $schema, $databases, $version, $dir, $preversion, $sqltargs) = @_;
2325 carp "No directory given, using ./\n";
2328 -d $dir or File::Path::mkpath($dir)
2329 or $self->throw_exception("create_ddl_dir: $! creating dir '$dir'");
2332 $self->throw_exception ("Directory '$dir' does not exist\n") unless(-d $dir);
2334 $databases ||= ['MySQL', 'SQLite', 'PostgreSQL'];
2335 $databases = [ $databases ] if(ref($databases) ne 'ARRAY');
2337 my $schema_version = $schema->schema_version || '1.x';
2338 $version ||= $schema_version;
2341 add_drop_table => 1,
2342 ignore_constraint_names => 1,
2343 ignore_index_names => 1,
2347 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy')) {
2348 $self->throw_exception("Can't create a ddl file without " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2351 my $sqlt = SQL::Translator->new( $sqltargs );
2353 $sqlt->parser('SQL::Translator::Parser::DBIx::Class');
2354 my $sqlt_schema = $sqlt->translate({ data => $schema })
2355 or $self->throw_exception ($sqlt->error);
2357 foreach my $db (@$databases) {
2359 $sqlt->{schema} = $sqlt_schema;
2360 $sqlt->producer($db);
2363 my $filename = $schema->ddl_filename($db, $version, $dir);
2364 if (-e $filename && ($version eq $schema_version )) {
2365 # if we are dumping the current version, overwrite the DDL
2366 carp "Overwriting existing DDL file - $filename";
2370 my $output = $sqlt->translate;
2372 carp("Failed to translate to $db, skipping. (" . $sqlt->error . ")");
2375 if(!open($file, ">$filename")) {
2376 $self->throw_exception("Can't open $filename for writing ($!)");
2379 print $file $output;
2382 next unless ($preversion);
2384 require SQL::Translator::Diff;
2386 my $prefilename = $schema->ddl_filename($db, $preversion, $dir);
2387 if(!-e $prefilename) {
2388 carp("No previous schema file found ($prefilename)");
2392 my $difffile = $schema->ddl_filename($db, $version, $dir, $preversion);
2394 carp("Overwriting existing diff file - $difffile");
2400 my $t = SQL::Translator->new($sqltargs);
2405 or $self->throw_exception ($t->error);
2407 my $out = $t->translate( $prefilename )
2408 or $self->throw_exception ($t->error);
2410 $source_schema = $t->schema;
2412 $source_schema->name( $prefilename )
2413 unless ( $source_schema->name );
2416 # The "new" style of producers have sane normalization and can support
2417 # diffing a SQL file against a DBIC->SQLT schema. Old style ones don't
2418 # And we have to diff parsed SQL against parsed SQL.
2419 my $dest_schema = $sqlt_schema;
2421 unless ( "SQL::Translator::Producer::$db"->can('preprocess_schema') ) {
2422 my $t = SQL::Translator->new($sqltargs);
2427 or $self->throw_exception ($t->error);
2429 my $out = $t->translate( $filename )
2430 or $self->throw_exception ($t->error);
2432 $dest_schema = $t->schema;
2434 $dest_schema->name( $filename )
2435 unless $dest_schema->name;
2438 my $diff = SQL::Translator::Diff::schema_diff($source_schema, $db,
2442 if(!open $file, ">$difffile") {
2443 $self->throw_exception("Can't write to $difffile ($!)");
2451 =head2 deployment_statements
2455 =item Arguments: $schema, $type, $version, $directory, $sqlt_args
2459 Returns the statements used by L</deploy> and L<DBIx::Class::Schema/deploy>.
2461 The L<SQL::Translator> (not L<DBI>) database driver name can be explicitly
2462 provided in C<$type>, otherwise the result of L</sqlt_type> is used as default.
2464 C<$directory> is used to return statements from files in a previously created
2465 L</create_ddl_dir> directory and is optional. The filenames are constructed
2466 from L<DBIx::Class::Schema/ddl_filename>, the schema name and the C<$version>.
2468 If no C<$directory> is specified then the statements are constructed on the
2469 fly using L<SQL::Translator> and C<$version> is ignored.
2471 See L<SQL::Translator/METHODS> for a list of values for C<$sqlt_args>.
2475 sub deployment_statements {
2476 my ($self, $schema, $type, $version, $dir, $sqltargs) = @_;
2477 $type ||= $self->sqlt_type;
2478 $version ||= $schema->schema_version || '1.x';
2480 my $filename = $schema->ddl_filename($type, $version, $dir);
2484 open($file, "<$filename")
2485 or $self->throw_exception("Can't open $filename ($!)");
2488 return join('', @rows);
2491 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy') ) {
2492 $self->throw_exception("Can't deploy without a ddl_dir or " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2495 # sources needs to be a parser arg, but for simplicty allow at top level
2497 $sqltargs->{parser_args}{sources} = delete $sqltargs->{sources}
2498 if exists $sqltargs->{sources};
2500 my $tr = SQL::Translator->new(
2501 producer => "SQL::Translator::Producer::${type}",
2503 parser => 'SQL::Translator::Parser::DBIx::Class',
2510 @ret = $tr->translate;
2513 $ret[0] = $tr->translate;
2516 $self->throw_exception( 'Unable to produce deployment statements: ' . $tr->error)
2517 unless (@ret && defined $ret[0]);
2519 return $wa ? @ret : $ret[0];
2523 my ($self, $schema, $type, $sqltargs, $dir) = @_;
2526 return if($line =~ /^--/);
2528 # next if($line =~ /^DROP/m);
2529 return if($line =~ /^BEGIN TRANSACTION/m);
2530 return if($line =~ /^COMMIT/m);
2531 return if $line =~ /^\s+$/; # skip whitespace only
2532 $self->_query_start($line);
2534 # do a dbh_do cycle here, as we need some error checking in
2535 # place (even though we will ignore errors)
2536 $self->dbh_do (sub { $_[1]->do($line) });
2538 carp qq{$_ (running "${line}")};
2540 $self->_query_end($line);
2542 my @statements = $schema->deployment_statements($type, undef, $dir, { %{ $sqltargs || {} }, no_comments => 1 } );
2543 if (@statements > 1) {
2544 foreach my $statement (@statements) {
2545 $deploy->( $statement );
2548 elsif (@statements == 1) {
2549 foreach my $line ( split(";\n", $statements[0])) {
2555 =head2 datetime_parser
2557 Returns the datetime parser class
2561 sub datetime_parser {
2563 return $self->{datetime_parser} ||= do {
2564 $self->build_datetime_parser(@_);
2568 =head2 datetime_parser_type
2570 Defines (returns) the datetime parser class - currently hardwired to
2571 L<DateTime::Format::MySQL>
2575 sub datetime_parser_type { "DateTime::Format::MySQL"; }
2577 =head2 build_datetime_parser
2579 See L</datetime_parser>
2583 sub build_datetime_parser {
2585 my $type = $self->datetime_parser_type(@_);
2586 $self->ensure_class_loaded ($type);
2591 =head2 is_replicating
2593 A boolean that reports if a particular L<DBIx::Class::Storage::DBI> is set to
2594 replicate from a master database. Default is undef, which is the result
2595 returned by databases that don't support replication.
2599 sub is_replicating {
2604 =head2 lag_behind_master
2606 Returns a number that represents a certain amount of lag behind a master db
2607 when a given storage is replicating. The number is database dependent, but
2608 starts at zero and increases with the amount of lag. Default in undef
2612 sub lag_behind_master {
2616 =head2 relname_to_table_alias
2620 =item Arguments: $relname, $join_count
2624 L<DBIx::Class> uses L<DBIx::Class::Relationship> names as table aliases in
2627 This hook is to allow specific L<DBIx::Class::Storage> drivers to change the
2628 way these aliases are named.
2630 The default behavior is C<< "$relname_$join_count" if $join_count > 1 >>,
2631 otherwise C<"$relname">.
2635 sub relname_to_table_alias {
2636 my ($self, $relname, $join_count) = @_;
2638 my $alias = ($join_count && $join_count > 1 ?
2639 join('_', $relname, $join_count) : $relname);
2648 =head2 DBIx::Class and AutoCommit
2650 DBIx::Class can do some wonderful magic with handling exceptions,
2651 disconnections, and transactions when you use C<< AutoCommit => 1 >>
2652 (the default) combined with C<txn_do> for transaction support.
2654 If you set C<< AutoCommit => 0 >> in your connect info, then you are always
2655 in an assumed transaction between commits, and you're telling us you'd
2656 like to manage that manually. A lot of the magic protections offered by
2657 this module will go away. We can't protect you from exceptions due to database
2658 disconnects because we don't know anything about how to restart your
2659 transactions. You're on your own for handling all sorts of exceptional
2660 cases if you choose the C<< AutoCommit => 0 >> path, just as you would
2666 Matt S. Trout <mst@shadowcatsystems.co.uk>
2668 Andy Grundman <andy@hybridized.org>
2672 You may distribute this code under the same terms as Perl itself.