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|^Try::Tiny/;
12 use DBIx::Class::Storage::DBI::Cursor;
13 use DBIx::Class::Storage::Statistics;
14 use Scalar::Util qw/refaddr weaken reftype blessed/;
15 use Data::Dumper::Concise 'Dumper';
16 use Sub::Name 'subname';
18 use File::Path 'make_path';
22 # default cursor class, overridable in connect_info attributes
23 __PACKAGE__->cursor_class('DBIx::Class::Storage::DBI::Cursor');
25 __PACKAGE__->mk_group_accessors('inherited' => qw/sql_maker_class sql_limit_dialect/);
26 __PACKAGE__->sql_maker_class('DBIx::Class::SQLMaker');
28 __PACKAGE__->mk_group_accessors('simple' => qw/
29 _connect_info _dbi_connect_info _dbic_connect_attributes _driver_determined
30 _dbh _dbh_details _conn_pid _conn_tid _sql_maker _sql_maker_opts
31 transaction_depth _dbh_autocommit savepoints
34 # the values for these accessors are picked out (and deleted) from
35 # the attribute hashref passed to connect_info
36 my @storage_options = qw/
37 on_connect_call on_disconnect_call on_connect_do on_disconnect_do
38 disable_sth_caching unsafe auto_savepoint
40 __PACKAGE__->mk_group_accessors('simple' => @storage_options);
43 # capability definitions, using a 2-tiered accessor system
46 # A driver/user may define _use_X, which blindly without any checks says:
47 # "(do not) use this capability", (use_dbms_capability is an "inherited"
50 # If _use_X is undef, _supports_X is then queried. This is a "simple" style
51 # accessor, which in turn calls _determine_supports_X, and stores the return
52 # in a special slot on the storage object, which is wiped every time a $dbh
53 # reconnection takes place (it is not guaranteed that upon reconnection we
54 # will get the same rdbms version). _determine_supports_X does not need to
55 # exist on a driver, as we ->can for it before calling.
57 my @capabilities = (qw/insert_returning placeholders typeless_placeholders/);
58 __PACKAGE__->mk_group_accessors( dbms_capability => map { "_supports_$_" } @capabilities );
59 __PACKAGE__->mk_group_accessors( use_dbms_capability => map { "_use_$_" } @capabilities );
62 # Each of these methods need _determine_driver called before itself
63 # in order to function reliably. This is a purely DRY optimization
65 # get_(use)_dbms_capability need to be called on the correct Storage
66 # class, as _use_X may be hardcoded class-wide, and _supports_X calls
67 # _determine_supports_X which obv. needs a correct driver as well
68 my @rdbms_specific_methods = qw/
82 get_use_dbms_capability
86 for my $meth (@rdbms_specific_methods) {
88 my $orig = __PACKAGE__->can ($meth)
89 or die "$meth is not a ::Storage::DBI method!";
92 no warnings qw/redefine/;
93 *{__PACKAGE__ ."::$meth"} = subname $meth => sub {
94 if (not $_[0]->_driver_determined and not $_[0]->{_in_determine_driver}) {
95 $_[0]->_determine_driver;
96 goto $_[0]->can($meth);
105 DBIx::Class::Storage::DBI - DBI storage handler
109 my $schema = MySchema->connect('dbi:SQLite:my.db');
111 $schema->storage->debug(1);
113 my @stuff = $schema->storage->dbh_do(
115 my ($storage, $dbh, @args) = @_;
116 $dbh->do("DROP TABLE authors");
121 $schema->resultset('Book')->search({
122 written_on => $schema->storage->datetime_parser->format_datetime(DateTime->now)
127 This class represents the connection to an RDBMS via L<DBI>. See
128 L<DBIx::Class::Storage> for general information. This pod only
129 documents DBI-specific methods and behaviors.
136 my $new = shift->next::method(@_);
138 $new->transaction_depth(0);
139 $new->_sql_maker_opts({});
140 $new->_dbh_details({});
141 $new->{savepoints} = [];
142 $new->{_in_dbh_do} = 0;
143 $new->{_dbh_gen} = 0;
145 # read below to see what this does
146 $new->_arm_global_destructor;
151 # This is hack to work around perl shooting stuff in random
152 # order on exit(). If we do not walk the remaining storage
153 # objects in an END block, there is a *small but real* chance
154 # of a fork()ed child to kill the parent's shared DBI handle,
155 # *before perl reaches the DESTROY in this package*
156 # Yes, it is ugly and effective.
158 my %seek_and_destroy;
160 sub _arm_global_destructor {
162 my $key = Scalar::Util::refaddr ($self);
163 $seek_and_destroy{$key} = $self;
164 Scalar::Util::weaken ($seek_and_destroy{$key});
168 local $?; # just in case the DBI destructor changes it somehow
170 # destroy just the object if not native to this process/thread
171 $_->_preserve_foreign_dbh for (grep
173 values %seek_and_destroy
181 # destroy just the object if not native to this process/thread
182 $self->_preserve_foreign_dbh;
184 # some databases need this to stop spewing warnings
185 if (my $dbh = $self->_dbh) {
187 %{ $dbh->{CachedKids} } = ();
195 sub _preserve_foreign_dbh {
198 return unless $self->_dbh;
202 return unless $self->_dbh;
208 # handle pid changes correctly - do not destroy parent's connection
212 return if ( defined $self->_conn_pid and $self->_conn_pid == $$ );
214 $self->_dbh->{InactiveDestroy} = 1;
221 # very similar to above, but seems to FAIL if I set InactiveDestroy
225 if ( ! defined $self->_conn_tid ) {
228 elsif ( $self->_conn_tid == threads->tid ) {
229 return; # same thread
232 #$self->_dbh->{InactiveDestroy} = 1; # why does t/51threads.t fail...?
242 This method is normally called by L<DBIx::Class::Schema/connection>, which
243 encapsulates its argument list in an arrayref before passing them here.
245 The argument list may contain:
251 The same 4-element argument set one would normally pass to
252 L<DBI/connect>, optionally followed by
253 L<extra attributes|/DBIx::Class specific connection attributes>
254 recognized by DBIx::Class:
256 $connect_info_args = [ $dsn, $user, $password, \%dbi_attributes?, \%extra_attributes? ];
260 A single code reference which returns a connected
261 L<DBI database handle|DBI/connect> optionally followed by
262 L<extra attributes|/DBIx::Class specific connection attributes> recognized
265 $connect_info_args = [ sub { DBI->connect (...) }, \%extra_attributes? ];
269 A single hashref with all the attributes and the dsn/user/password
272 $connect_info_args = [{
280 $connect_info_args = [{
281 dbh_maker => sub { DBI->connect (...) },
286 This is particularly useful for L<Catalyst> based applications, allowing the
287 following config (L<Config::General> style):
292 dsn dbi:mysql:database=test
299 The C<dsn>/C<user>/C<password> combination can be substituted by the
300 C<dbh_maker> key whose value is a coderef that returns a connected
301 L<DBI database handle|DBI/connect>
305 Please note that the L<DBI> docs recommend that you always explicitly
306 set C<AutoCommit> to either I<0> or I<1>. L<DBIx::Class> further
307 recommends that it be set to I<1>, and that you perform transactions
308 via our L<DBIx::Class::Schema/txn_do> method. L<DBIx::Class> will set it
309 to I<1> if you do not do explicitly set it to zero. This is the default
310 for most DBDs. See L</DBIx::Class and AutoCommit> for details.
312 =head3 DBIx::Class specific connection attributes
314 In addition to the standard L<DBI|DBI/ATTRIBUTES_COMMON_TO_ALL_HANDLES>
315 L<connection|DBI/Database_Handle_Attributes> attributes, DBIx::Class recognizes
316 the following connection options. These options can be mixed in with your other
317 L<DBI> connection attributes, or placed in a separate hashref
318 (C<\%extra_attributes>) as shown above.
320 Every time C<connect_info> is invoked, any previous settings for
321 these options will be cleared before setting the new ones, regardless of
322 whether any options are specified in the new C<connect_info>.
329 Specifies things to do immediately after connecting or re-connecting to
330 the database. Its value may contain:
336 This contains one SQL statement to execute.
338 =item an array reference
340 This contains SQL statements to execute in order. Each element contains
341 a string or a code reference that returns a string.
343 =item a code reference
345 This contains some code to execute. Unlike code references within an
346 array reference, its return value is ignored.
350 =item on_disconnect_do
352 Takes arguments in the same form as L</on_connect_do> and executes them
353 immediately before disconnecting from the database.
355 Note, this only runs if you explicitly call L</disconnect> on the
358 =item on_connect_call
360 A more generalized form of L</on_connect_do> that calls the specified
361 C<connect_call_METHOD> methods in your storage driver.
363 on_connect_do => 'select 1'
367 on_connect_call => [ [ do_sql => 'select 1' ] ]
369 Its values may contain:
375 Will call the C<connect_call_METHOD> method.
377 =item a code reference
379 Will execute C<< $code->($storage) >>
381 =item an array reference
383 Each value can be a method name or code reference.
385 =item an array of arrays
387 For each array, the first item is taken to be the C<connect_call_> method name
388 or code reference, and the rest are parameters to it.
392 Some predefined storage methods you may use:
398 Executes a SQL string or a code reference that returns a SQL string. This is
399 what L</on_connect_do> and L</on_disconnect_do> use.
407 Will execute the scalar as SQL.
411 Taken to be arguments to L<DBI/do>, the SQL string optionally followed by the
412 attributes hashref and bind values.
414 =item a code reference
416 Will execute C<< $code->($storage) >> and execute the return array refs as
423 Execute any statements necessary to initialize the database session to return
424 and accept datetime/timestamp values used with
425 L<DBIx::Class::InflateColumn::DateTime>.
427 Only necessary for some databases, see your specific storage driver for
428 implementation details.
432 =item on_disconnect_call
434 Takes arguments in the same form as L</on_connect_call> and executes them
435 immediately before disconnecting from the database.
437 Calls the C<disconnect_call_METHOD> methods as opposed to the
438 C<connect_call_METHOD> methods called by L</on_connect_call>.
440 Note, this only runs if you explicitly call L</disconnect> on the
443 =item disable_sth_caching
445 If set to a true value, this option will disable the caching of
446 statement handles via L<DBI/prepare_cached>.
450 Sets a specific SQL::Abstract::Limit-style limit dialect, overriding the
451 default L</sql_limit_dialect> setting of the storage (if any). For a list
452 of available limit dialects see L<DBIx::Class::SQLMaker::LimitDialects>.
456 Specifies what characters to use to quote table and column names. If
457 you use this you will want to specify L</name_sep> as well.
459 C<quote_char> expects either a single character, in which case is it
460 is placed on either side of the table/column name, or an arrayref of length
461 2 in which case the table/column name is placed between the elements.
463 For example under MySQL you should use C<< quote_char => '`' >>, and for
464 SQL Server you should use C<< quote_char => [qw/[ ]/] >>.
468 This only needs to be used in conjunction with C<quote_char>, and is used to
469 specify the character that separates elements (schemas, tables, columns) from
470 each other. In most cases this is simply a C<.>.
472 The consequences of not supplying this value is that L<SQL::Abstract>
473 will assume DBIx::Class' uses of aliases to be complete column
474 names. The output will look like I<"me.name"> when it should actually
479 This Storage driver normally installs its own C<HandleError>, sets
480 C<RaiseError> and C<ShowErrorStatement> on, and sets C<PrintError> off on
481 all database handles, including those supplied by a coderef. It does this
482 so that it can have consistent and useful error behavior.
484 If you set this option to a true value, Storage will not do its usual
485 modifications to the database handle's attributes, and instead relies on
486 the settings in your connect_info DBI options (or the values you set in
487 your connection coderef, in the case that you are connecting via coderef).
489 Note that your custom settings can cause Storage to malfunction,
490 especially if you set a C<HandleError> handler that suppresses exceptions
491 and/or disable C<RaiseError>.
495 If this option is true, L<DBIx::Class> will use savepoints when nesting
496 transactions, making it possible to recover from failure in the inner
497 transaction without having to abort all outer transactions.
501 Use this argument to supply a cursor class other than the default
502 L<DBIx::Class::Storage::DBI::Cursor>.
506 Some real-life examples of arguments to L</connect_info> and
507 L<DBIx::Class::Schema/connect>
509 # Simple SQLite connection
510 ->connect_info([ 'dbi:SQLite:./foo.db' ]);
513 ->connect_info([ sub { DBI->connect(...) } ]);
515 # Connect via subref in hashref
517 dbh_maker => sub { DBI->connect(...) },
518 on_connect_do => 'alter session ...',
521 # A bit more complicated
528 { quote_char => q{"}, name_sep => q{.} },
532 # Equivalent to the previous example
538 { AutoCommit => 1, quote_char => q{"}, name_sep => q{.} },
542 # Same, but with hashref as argument
543 # See parse_connect_info for explanation
546 dsn => 'dbi:Pg:dbname=foo',
548 password => 'my_pg_password',
555 # Subref + DBIx::Class-specific connection options
558 sub { DBI->connect(...) },
562 on_connect_do => ['SET search_path TO myschema,otherschema,public'],
563 disable_sth_caching => 1,
573 my ($self, $info) = @_;
575 return $self->_connect_info if !$info;
577 $self->_connect_info($info); # copy for _connect_info
579 $info = $self->_normalize_connect_info($info)
580 if ref $info eq 'ARRAY';
582 for my $storage_opt (keys %{ $info->{storage_options} }) {
583 my $value = $info->{storage_options}{$storage_opt};
585 $self->$storage_opt($value);
588 # Kill sql_maker/_sql_maker_opts, so we get a fresh one with only
589 # the new set of options
590 $self->_sql_maker(undef);
591 $self->_sql_maker_opts({});
593 for my $sql_maker_opt (keys %{ $info->{sql_maker_options} }) {
594 my $value = $info->{sql_maker_options}{$sql_maker_opt};
596 $self->_sql_maker_opts->{$sql_maker_opt} = $value;
600 %{ $self->_default_dbi_connect_attributes || {} },
601 %{ $info->{attributes} || {} },
604 my @args = @{ $info->{arguments} };
606 $self->_dbi_connect_info([@args,
607 %attrs && !(ref $args[0] eq 'CODE') ? \%attrs : ()]);
610 # save attributes them in a separate accessor so they are always
611 # introspectable, even in case of a CODE $dbhmaker
612 $self->_dbic_connect_attributes (\%attrs);
614 return $self->_connect_info;
617 sub _normalize_connect_info {
618 my ($self, $info_arg) = @_;
621 my @args = @$info_arg; # take a shallow copy for further mutilation
623 # combine/pre-parse arguments depending on invocation style
626 if (ref $args[0] eq 'CODE') { # coderef with optional \%extra_attributes
627 %attrs = %{ $args[1] || {} };
630 elsif (ref $args[0] eq 'HASH') { # single hashref (i.e. Catalyst config)
631 %attrs = %{$args[0]};
633 if (my $code = delete $attrs{dbh_maker}) {
636 my @ignored = grep { delete $attrs{$_} } (qw/dsn user password/);
639 'Attribute(s) %s in connect_info were ignored, as they can not be applied '
640 . "to the result of 'dbh_maker'",
642 join (', ', map { "'$_'" } (@ignored) ),
647 @args = delete @attrs{qw/dsn user password/};
650 else { # otherwise assume dsn/user/password + \%attrs + \%extra_attrs
652 % { $args[3] || {} },
653 % { $args[4] || {} },
655 @args = @args[0,1,2];
658 $info{arguments} = \@args;
660 my @storage_opts = grep exists $attrs{$_},
661 @storage_options, 'cursor_class';
663 @{ $info{storage_options} }{@storage_opts} =
664 delete @attrs{@storage_opts} if @storage_opts;
666 my @sql_maker_opts = grep exists $attrs{$_},
667 qw/limit_dialect quote_char name_sep/;
669 @{ $info{sql_maker_options} }{@sql_maker_opts} =
670 delete @attrs{@sql_maker_opts} if @sql_maker_opts;
672 $info{attributes} = \%attrs if %attrs;
677 sub _default_dbi_connect_attributes {
687 This method is deprecated in favour of setting via L</connect_info>.
691 =head2 on_disconnect_do
693 This method is deprecated in favour of setting via L</connect_info>.
697 sub _parse_connect_do {
698 my ($self, $type) = @_;
700 my $val = $self->$type;
701 return () if not defined $val;
706 push @res, [ 'do_sql', $val ];
707 } elsif (ref($val) eq 'CODE') {
709 } elsif (ref($val) eq 'ARRAY') {
710 push @res, map { [ 'do_sql', $_ ] } @$val;
712 $self->throw_exception("Invalid type for $type: ".ref($val));
720 Arguments: ($subref | $method_name), @extra_coderef_args?
722 Execute the given $subref or $method_name using the new exception-based
723 connection management.
725 The first two arguments will be the storage object that C<dbh_do> was called
726 on and a database handle to use. Any additional arguments will be passed
727 verbatim to the called subref as arguments 2 and onwards.
729 Using this (instead of $self->_dbh or $self->dbh) ensures correct
730 exception handling and reconnection (or failover in future subclasses).
732 Your subref should have no side-effects outside of the database, as
733 there is the potential for your subref to be partially double-executed
734 if the database connection was stale/dysfunctional.
738 my @stuff = $schema->storage->dbh_do(
740 my ($storage, $dbh, @cols) = @_;
741 my $cols = join(q{, }, @cols);
742 $dbh->selectrow_array("SELECT $cols FROM foo");
753 my $dbh = $self->_get_dbh;
755 return $self->$code($dbh, @_)
756 if ( $self->{_in_dbh_do} || $self->{transaction_depth} );
758 local $self->{_in_dbh_do} = 1;
760 # take a ref instead of a copy, to preserve coderef @_ aliasing semantics
763 $self->$code ($dbh, @$args);
765 $self->throw_exception($_) if $self->connected;
767 # We were not connected - reconnect and retry, but let any
768 # exception fall right through this time
769 carp "Retrying $code after catching disconnected exception: $_"
770 if $ENV{DBIC_DBIRETRY_DEBUG};
772 $self->_populate_dbh;
773 $self->$code($self->_dbh, @$args);
777 # This is basically a blend of dbh_do above and DBIx::Class::Storage::txn_do.
778 # It also informs dbh_do to bypass itself while under the direction of txn_do,
779 # via $self->{_in_dbh_do} (this saves some redundant eval and errorcheck, etc)
784 ref $coderef eq 'CODE' or $self->throw_exception
785 ('$coderef must be a CODE reference');
787 return $coderef->(@_) if $self->{transaction_depth} && ! $self->auto_savepoint;
789 local $self->{_in_dbh_do} = 1;
792 my $want_array = wantarray;
798 # take a ref instead of a copy, to preserve coderef @_ aliasing semantics
806 @result = $coderef->(@$args);
808 elsif(defined $want_array) {
809 $result[0] = $coderef->(@$args);
819 if(! defined $exception) { return $want_array ? @result : $result[0] }
821 if($tried++ || $self->connected) {
822 my $rollback_exception;
823 try { $self->txn_rollback } catch { $rollback_exception = shift };
824 if(defined $rollback_exception) {
825 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
826 $self->throw_exception($exception) # propagate nested rollback
827 if $rollback_exception =~ /$exception_class/;
829 $self->throw_exception(
830 "Transaction aborted: ${exception}. "
831 . "Rollback failed: ${rollback_exception}"
834 $self->throw_exception($exception)
837 # We were not connected, and was first try - reconnect and retry
839 carp "Retrying $coderef after catching disconnected exception: $exception"
840 if $ENV{DBIC_DBIRETRY_DEBUG};
841 $self->_populate_dbh;
847 Our C<disconnect> method also performs a rollback first if the
848 database is not in C<AutoCommit> mode.
858 push @actions, ( $self->on_disconnect_call || () );
859 push @actions, $self->_parse_connect_do ('on_disconnect_do');
861 $self->_do_connection_actions(disconnect_call_ => $_) for @actions;
863 $self->_dbh_rollback unless $self->_dbh_autocommit;
865 %{ $self->_dbh->{CachedKids} } = ();
866 $self->_dbh->disconnect;
872 =head2 with_deferred_fk_checks
876 =item Arguments: C<$coderef>
878 =item Return Value: The return value of $coderef
882 Storage specific method to run the code ref with FK checks deferred or
883 in MySQL's case disabled entirely.
887 # Storage subclasses should override this
888 sub with_deferred_fk_checks {
889 my ($self, $sub) = @_;
897 =item Arguments: none
899 =item Return Value: 1|0
903 Verifies that the current database handle is active and ready to execute
904 an SQL statement (e.g. the connection did not get stale, server is still
905 answering, etc.) This method is used internally by L</dbh>.
911 return 0 unless $self->_seems_connected;
914 local $self->_dbh->{RaiseError} = 1;
919 sub _seems_connected {
922 $self->_preserve_foreign_dbh;
924 my $dbh = $self->_dbh
927 return $dbh->FETCH('Active');
933 my $dbh = $self->_dbh or return 0;
938 sub ensure_connected {
941 unless ($self->connected) {
942 $self->_populate_dbh;
948 Returns a C<$dbh> - a data base handle of class L<DBI>. The returned handle
949 is guaranteed to be healthy by implicitly calling L</connected>, and if
950 necessary performing a reconnection before returning. Keep in mind that this
951 is very B<expensive> on some database engines. Consider using L</dbh_do>
959 if (not $self->_dbh) {
960 $self->_populate_dbh;
962 $self->ensure_connected;
967 # this is the internal "get dbh or connect (don't check)" method
970 $self->_preserve_foreign_dbh;
971 $self->_populate_dbh unless $self->_dbh;
977 unless ($self->_sql_maker) {
978 my $sql_maker_class = $self->sql_maker_class;
979 $self->ensure_class_loaded ($sql_maker_class);
981 my %opts = %{$self->_sql_maker_opts||{}};
985 $self->sql_limit_dialect
988 my $s_class = (ref $self) || $self;
990 "Your storage class ($s_class) does not set sql_limit_dialect and you "
991 . 'have not supplied an explicit limit_dialect in your connection_info. '
992 . 'DBIC will attempt to use the GenericSubQ dialect, which works on most '
993 . 'databases but can be (and often is) painfully slow.'
1000 $self->_sql_maker($sql_maker_class->new(
1001 bindtype=>'columns',
1002 array_datatypes => 1,
1003 limit_dialect => $dialect,
1007 return $self->_sql_maker;
1010 # nothing to do by default
1017 my @info = @{$self->_dbi_connect_info || []};
1018 $self->_dbh(undef); # in case ->connected failed we might get sent here
1019 $self->_dbh_details({}); # reset everything we know
1021 $self->_dbh($self->_connect(@info));
1023 $self->_conn_pid($$);
1024 $self->_conn_tid(threads->tid) if $INC{'threads.pm'};
1026 $self->_determine_driver;
1028 # Always set the transaction depth on connect, since
1029 # there is no transaction in progress by definition
1030 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
1032 $self->_run_connection_actions unless $self->{_in_determine_driver};
1035 sub _run_connection_actions {
1039 push @actions, ( $self->on_connect_call || () );
1040 push @actions, $self->_parse_connect_do ('on_connect_do');
1042 $self->_do_connection_actions(connect_call_ => $_) for @actions;
1047 sub set_use_dbms_capability {
1048 $_[0]->set_inherited ($_[1], $_[2]);
1051 sub get_use_dbms_capability {
1052 my ($self, $capname) = @_;
1054 my $use = $self->get_inherited ($capname);
1057 : do { $capname =~ s/^_use_/_supports_/; $self->get_dbms_capability ($capname) }
1061 sub set_dbms_capability {
1062 $_[0]->_dbh_details->{capability}{$_[1]} = $_[2];
1065 sub get_dbms_capability {
1066 my ($self, $capname) = @_;
1068 my $cap = $self->_dbh_details->{capability}{$capname};
1070 unless (defined $cap) {
1071 if (my $meth = $self->can ("_determine$capname")) {
1072 $cap = $self->$meth ? 1 : 0;
1078 $self->set_dbms_capability ($capname, $cap);
1088 unless ($info = $self->_dbh_details->{info}) {
1092 my $server_version = try { $self->_get_server_version };
1094 if (defined $server_version) {
1095 $info->{dbms_version} = $server_version;
1097 my ($numeric_version) = $server_version =~ /^([\d\.]+)/;
1098 my @verparts = split (/\./, $numeric_version);
1104 # consider only up to 3 version parts, iff not more than 3 digits
1106 while (@verparts && @use_parts < 3) {
1107 my $p = shift @verparts;
1109 push @use_parts, $p;
1111 push @use_parts, 0 while @use_parts < 3;
1113 $info->{normalized_dbms_version} = sprintf "%d.%03d%03d", @use_parts;
1117 $self->_dbh_details->{info} = $info;
1123 sub _get_server_version {
1124 shift->_get_dbh->get_info(18);
1127 sub _determine_driver {
1130 if ((not $self->_driver_determined) && (not $self->{_in_determine_driver})) {
1131 my $started_connected = 0;
1132 local $self->{_in_determine_driver} = 1;
1134 if (ref($self) eq __PACKAGE__) {
1136 if ($self->_dbh) { # we are connected
1137 $driver = $self->_dbh->{Driver}{Name};
1138 $started_connected = 1;
1140 # if connect_info is a CODEREF, we have no choice but to connect
1141 if (ref $self->_dbi_connect_info->[0] &&
1142 reftype $self->_dbi_connect_info->[0] eq 'CODE') {
1143 $self->_populate_dbh;
1144 $driver = $self->_dbh->{Driver}{Name};
1147 # try to use dsn to not require being connected, the driver may still
1148 # force a connection in _rebless to determine version
1149 # (dsn may not be supplied at all if all we do is make a mock-schema)
1150 my $dsn = $self->_dbi_connect_info->[0] || $ENV{DBI_DSN} || '';
1151 ($driver) = $dsn =~ /dbi:([^:]+):/i;
1152 $driver ||= $ENV{DBI_DRIVER};
1157 my $storage_class = "DBIx::Class::Storage::DBI::${driver}";
1158 if ($self->load_optional_class($storage_class)) {
1159 mro::set_mro($storage_class, 'c3');
1160 bless $self, $storage_class;
1166 $self->_driver_determined(1);
1168 $self->_init; # run driver-specific initializations
1170 $self->_run_connection_actions
1171 if !$started_connected && defined $self->_dbh;
1175 sub _do_connection_actions {
1177 my $method_prefix = shift;
1180 if (not ref($call)) {
1181 my $method = $method_prefix . $call;
1183 } elsif (ref($call) eq 'CODE') {
1185 } elsif (ref($call) eq 'ARRAY') {
1186 if (ref($call->[0]) ne 'ARRAY') {
1187 $self->_do_connection_actions($method_prefix, $_) for @$call;
1189 $self->_do_connection_actions($method_prefix, @$_) for @$call;
1192 $self->throw_exception (sprintf ("Don't know how to process conection actions of type '%s'", ref($call)) );
1198 sub connect_call_do_sql {
1200 $self->_do_query(@_);
1203 sub disconnect_call_do_sql {
1205 $self->_do_query(@_);
1208 # override in db-specific backend when necessary
1209 sub connect_call_datetime_setup { 1 }
1212 my ($self, $action) = @_;
1214 if (ref $action eq 'CODE') {
1215 $action = $action->($self);
1216 $self->_do_query($_) foreach @$action;
1219 # Most debuggers expect ($sql, @bind), so we need to exclude
1220 # the attribute hash which is the second argument to $dbh->do
1221 # furthermore the bind values are usually to be presented
1222 # as named arrayref pairs, so wrap those here too
1223 my @do_args = (ref $action eq 'ARRAY') ? (@$action) : ($action);
1224 my $sql = shift @do_args;
1225 my $attrs = shift @do_args;
1226 my @bind = map { [ undef, $_ ] } @do_args;
1228 $self->_query_start($sql, @bind);
1229 $self->_get_dbh->do($sql, $attrs, @do_args);
1230 $self->_query_end($sql, @bind);
1237 my ($self, @info) = @_;
1239 $self->throw_exception("You failed to provide any connection info")
1242 my ($old_connect_via, $dbh);
1244 if ($INC{'Apache/DBI.pm'} && $ENV{MOD_PERL}) {
1245 $old_connect_via = $DBI::connect_via;
1246 $DBI::connect_via = 'connect';
1250 if(ref $info[0] eq 'CODE') {
1251 $dbh = $info[0]->();
1254 $dbh = DBI->connect(@info);
1261 unless ($self->unsafe) {
1263 # this odd anonymous coderef dereference is in fact really
1264 # necessary to avoid the unwanted effect described in perl5
1267 my $weak_self = $_[0];
1270 $_[1]->{HandleError} = sub {
1272 $weak_self->throw_exception("DBI Exception: $_[0]");
1275 # the handler may be invoked by something totally out of
1277 croak ("DBI Exception (unhandled by DBIC, ::Schema GCed): $_[0]");
1282 $dbh->{ShowErrorStatement} = 1;
1283 $dbh->{RaiseError} = 1;
1284 $dbh->{PrintError} = 0;
1288 $self->throw_exception("DBI Connection failed: $_")
1291 $DBI::connect_via = $old_connect_via if $old_connect_via;
1294 $self->_dbh_autocommit($dbh->{AutoCommit});
1299 my ($self, $name) = @_;
1301 $name = $self->_svp_generate_name
1302 unless defined $name;
1304 $self->throw_exception ("You can't use savepoints outside a transaction")
1305 if $self->{transaction_depth} == 0;
1307 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1308 unless $self->can('_svp_begin');
1310 push @{ $self->{savepoints} }, $name;
1312 $self->debugobj->svp_begin($name) if $self->debug;
1314 return $self->_svp_begin($name);
1318 my ($self, $name) = @_;
1320 $self->throw_exception ("You can't use savepoints outside a transaction")
1321 if $self->{transaction_depth} == 0;
1323 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1324 unless $self->can('_svp_release');
1326 if (defined $name) {
1327 $self->throw_exception ("Savepoint '$name' does not exist")
1328 unless grep { $_ eq $name } @{ $self->{savepoints} };
1330 # Dig through the stack until we find the one we are releasing. This keeps
1331 # the stack up to date.
1334 do { $svp = pop @{ $self->{savepoints} } } while $svp ne $name;
1336 $name = pop @{ $self->{savepoints} };
1339 $self->debugobj->svp_release($name) if $self->debug;
1341 return $self->_svp_release($name);
1345 my ($self, $name) = @_;
1347 $self->throw_exception ("You can't use savepoints outside a transaction")
1348 if $self->{transaction_depth} == 0;
1350 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1351 unless $self->can('_svp_rollback');
1353 if (defined $name) {
1354 # If they passed us a name, verify that it exists in the stack
1355 unless(grep({ $_ eq $name } @{ $self->{savepoints} })) {
1356 $self->throw_exception("Savepoint '$name' does not exist!");
1359 # Dig through the stack until we find the one we are releasing. This keeps
1360 # the stack up to date.
1361 while(my $s = pop(@{ $self->{savepoints} })) {
1362 last if($s eq $name);
1364 # Add the savepoint back to the stack, as a rollback doesn't remove the
1365 # named savepoint, only everything after it.
1366 push(@{ $self->{savepoints} }, $name);
1368 # We'll assume they want to rollback to the last savepoint
1369 $name = $self->{savepoints}->[-1];
1372 $self->debugobj->svp_rollback($name) if $self->debug;
1374 return $self->_svp_rollback($name);
1377 sub _svp_generate_name {
1380 return 'savepoint_'.scalar(@{ $self->{'savepoints'} });
1386 # this means we have not yet connected and do not know the AC status
1387 # (e.g. coderef $dbh)
1388 $self->ensure_connected if (! defined $self->_dbh_autocommit);
1390 if($self->{transaction_depth} == 0) {
1391 $self->debugobj->txn_begin()
1393 $self->_dbh_begin_work;
1395 elsif ($self->auto_savepoint) {
1398 $self->{transaction_depth}++;
1401 sub _dbh_begin_work {
1404 # if the user is utilizing txn_do - good for him, otherwise we need to
1405 # ensure that the $dbh is healthy on BEGIN.
1406 # We do this via ->dbh_do instead of ->dbh, so that the ->dbh "ping"
1407 # will be replaced by a failure of begin_work itself (which will be
1408 # then retried on reconnect)
1409 if ($self->{_in_dbh_do}) {
1410 $self->_dbh->begin_work;
1412 $self->dbh_do(sub { $_[1]->begin_work });
1418 if ($self->{transaction_depth} == 1) {
1419 $self->debugobj->txn_commit()
1422 $self->{transaction_depth} = 0
1423 if $self->_dbh_autocommit;
1425 elsif($self->{transaction_depth} > 1) {
1426 $self->{transaction_depth}--;
1428 if $self->auto_savepoint;
1434 my $dbh = $self->_dbh
1435 or $self->throw_exception('cannot COMMIT on a disconnected handle');
1441 my $dbh = $self->_dbh;
1443 if ($self->{transaction_depth} == 1) {
1444 $self->debugobj->txn_rollback()
1446 $self->{transaction_depth} = 0
1447 if $self->_dbh_autocommit;
1448 $self->_dbh_rollback;
1450 elsif($self->{transaction_depth} > 1) {
1451 $self->{transaction_depth}--;
1452 if ($self->auto_savepoint) {
1453 $self->svp_rollback;
1458 die DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION->new;
1462 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
1464 if ($_ !~ /$exception_class/) {
1465 # ensure that a failed rollback resets the transaction depth
1466 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
1469 $self->throw_exception($_)
1475 my $dbh = $self->_dbh
1476 or $self->throw_exception('cannot ROLLBACK on a disconnected handle');
1480 # This used to be the top-half of _execute. It was split out to make it
1481 # easier to override in NoBindVars without duping the rest. It takes up
1482 # all of _execute's args, and emits $sql, @bind.
1483 sub _prep_for_execute {
1484 my ($self, $op, $extra_bind, $ident, $args) = @_;
1486 if( blessed $ident && $ident->isa("DBIx::Class::ResultSource") ) {
1487 $ident = $ident->from();
1490 my ($sql, @bind) = $self->sql_maker->$op($ident, @$args);
1493 map { ref $_ eq 'ARRAY' ? $_ : [ '!!dummy', $_ ] } @$extra_bind)
1495 return ($sql, \@bind);
1499 sub _fix_bind_params {
1500 my ($self, @bind) = @_;
1502 ### Turn @bind from something like this:
1503 ### ( [ "artist", 1 ], [ "cdid", 1, 3 ] )
1505 ### ( "'1'", "'1'", "'3'" )
1508 if ( defined( $_ && $_->[1] ) ) {
1509 map { qq{'$_'}; } @{$_}[ 1 .. $#$_ ];
1516 my ( $self, $sql, @bind ) = @_;
1518 if ( $self->debug ) {
1519 @bind = $self->_fix_bind_params(@bind);
1521 $self->debugobj->query_start( $sql, @bind );
1526 my ( $self, $sql, @bind ) = @_;
1528 if ( $self->debug ) {
1529 @bind = $self->_fix_bind_params(@bind);
1530 $self->debugobj->query_end( $sql, @bind );
1535 my ($self, $dbh, $op, $extra_bind, $ident, $bind_attributes, @args) = @_;
1537 my ($sql, $bind) = $self->_prep_for_execute($op, $extra_bind, $ident, \@args);
1539 $self->_query_start( $sql, @$bind );
1541 my $sth = $self->sth($sql,$op);
1543 my $placeholder_index = 1;
1545 foreach my $bound (@$bind) {
1546 my $attributes = {};
1547 my($column_name, @data) = @$bound;
1549 if ($bind_attributes) {
1550 $attributes = $bind_attributes->{$column_name}
1551 if defined $bind_attributes->{$column_name};
1554 foreach my $data (@data) {
1555 my $ref = ref $data;
1556 $data = $ref && $ref ne 'ARRAY' ? ''.$data : $data; # stringify args (except arrayrefs)
1558 $sth->bind_param($placeholder_index, $data, $attributes);
1559 $placeholder_index++;
1563 # Can this fail without throwing an exception anyways???
1564 my $rv = $sth->execute();
1565 $self->throw_exception(
1566 $sth->errstr || $sth->err || 'Unknown error: execute() returned false, but error flags were not set...'
1569 $self->_query_end( $sql, @$bind );
1571 return (wantarray ? ($rv, $sth, @$bind) : $rv);
1576 $self->dbh_do('_dbh_execute', @_); # retry over disconnects
1579 sub _prefetch_insert_auto_nextvals {
1580 my ($self, $source, $to_insert) = @_;
1584 foreach my $col ( $source->columns ) {
1585 if ( !defined $to_insert->{$col} ) {
1586 my $col_info = $source->column_info($col);
1588 if ( $col_info->{auto_nextval} ) {
1589 $upd->{$col} = $to_insert->{$col} = $self->_sequence_fetch(
1591 $col_info->{sequence} ||=
1592 $self->_dbh_get_autoinc_seq($self->_get_dbh, $source, $col)
1603 my ($source, $to_insert, $opts) = @_;
1605 my $updated_cols = $self->_prefetch_insert_auto_nextvals (@_);
1607 my $bind_attributes = $self->source_bind_attributes($source);
1609 my ($rv, $sth) = $self->_execute('insert' => [], $source, $bind_attributes, $to_insert, $opts);
1611 if ($opts->{returning}) {
1612 my @ret_cols = @{$opts->{returning}};
1614 my @ret_vals = try {
1615 local $SIG{__WARN__} = sub {};
1616 my @r = $sth->fetchrow_array;
1622 @ret{@ret_cols} = @ret_vals if (@ret_vals);
1630 return $updated_cols;
1633 ## Currently it is assumed that all values passed will be "normal", i.e. not
1634 ## scalar refs, or at least, all the same type as the first set, the statement is
1635 ## only prepped once.
1637 my ($self, $source, $cols, $data) = @_;
1640 @colvalues{@$cols} = (0..$#$cols);
1642 for my $i (0..$#$cols) {
1643 my $first_val = $data->[0][$i];
1644 next unless ref $first_val eq 'SCALAR';
1646 $colvalues{ $cols->[$i] } = $first_val;
1649 # check for bad data and stringify stringifiable objects
1650 my $bad_slice = sub {
1651 my ($msg, $col_idx, $slice_idx) = @_;
1652 $self->throw_exception(sprintf "%s for column '%s' in populate slice:\n%s",
1656 local $Data::Dumper::Maxdepth = 1; # don't dump objects, if any
1658 map { $cols->[$_] => $data->[$slice_idx][$_] } (0 .. $#$cols)
1664 for my $datum_idx (0..$#$data) {
1665 my $datum = $data->[$datum_idx];
1667 for my $col_idx (0..$#$cols) {
1668 my $val = $datum->[$col_idx];
1669 my $sqla_bind = $colvalues{ $cols->[$col_idx] };
1670 my $is_literal_sql = (ref $sqla_bind) eq 'SCALAR';
1672 if ($is_literal_sql) {
1674 $bad_slice->('bind found where literal SQL expected', $col_idx, $datum_idx);
1676 elsif ((my $reftype = ref $val) ne 'SCALAR') {
1677 $bad_slice->("$reftype reference found where literal SQL expected",
1678 $col_idx, $datum_idx);
1680 elsif ($$val ne $$sqla_bind){
1681 $bad_slice->("inconsistent literal SQL value, expecting: '$$sqla_bind'",
1682 $col_idx, $datum_idx);
1685 elsif (my $reftype = ref $val) {
1687 if (overload::Method($val, '""')) {
1688 $datum->[$col_idx] = "".$val;
1691 $bad_slice->("$reftype reference found where bind expected",
1692 $col_idx, $datum_idx);
1698 my ($sql, $bind) = $self->_prep_for_execute (
1699 'insert', undef, $source, [\%colvalues]
1703 my $empty_bind = 1 if (not @bind) &&
1704 (grep { ref $_ eq 'SCALAR' } values %colvalues) == @$cols;
1706 if ((not @bind) && (not $empty_bind)) {
1707 $self->throw_exception(
1708 'Cannot insert_bulk without support for placeholders'
1712 # neither _execute_array, nor _execute_inserts_with_no_binds are
1713 # atomic (even if _execute _array is a single call). Thus a safety
1715 my $guard = $self->txn_scope_guard;
1717 $self->_query_start( $sql, [ dummy => '__BULK_INSERT__' ] );
1718 my $sth = $self->sth($sql);
1721 # bind_param_array doesn't work if there are no binds
1722 $self->_dbh_execute_inserts_with_no_binds( $sth, scalar @$data );
1725 # @bind = map { ref $_ ? ''.$_ : $_ } @bind; # stringify args
1726 $self->_execute_array( $source, $sth, \@bind, $cols, $data );
1730 $self->_query_end( $sql, [ dummy => '__BULK_INSERT__' ] );
1734 return (wantarray ? ($rv, $sth, @bind) : $rv);
1737 sub _execute_array {
1738 my ($self, $source, $sth, $bind, $cols, $data, @extra) = @_;
1740 ## This must be an arrayref, else nothing works!
1741 my $tuple_status = [];
1743 ## Get the bind_attributes, if any exist
1744 my $bind_attributes = $self->source_bind_attributes($source);
1746 ## Bind the values and execute
1747 my $placeholder_index = 1;
1749 foreach my $bound (@$bind) {
1751 my $attributes = {};
1752 my ($column_name, $data_index) = @$bound;
1754 if( $bind_attributes ) {
1755 $attributes = $bind_attributes->{$column_name}
1756 if defined $bind_attributes->{$column_name};
1759 my @data = map { $_->[$data_index] } @$data;
1761 $sth->bind_param_array(
1764 (%$attributes ? $attributes : ()),
1766 $placeholder_index++;
1771 $rv = $self->_dbh_execute_array($sth, $tuple_status, @extra);
1777 # Statement must finish even if there was an exception.
1782 $err = shift unless defined $err
1787 if (! defined $err and $sth->err);
1791 ++$i while $i <= $#$tuple_status && !ref $tuple_status->[$i];
1793 $self->throw_exception("Unexpected populate error: $err")
1794 if ($i > $#$tuple_status);
1796 $self->throw_exception(sprintf "%s for populate slice:\n%s",
1797 ($tuple_status->[$i][1] || $err),
1798 Dumper { map { $cols->[$_] => $data->[$i][$_] } (0 .. $#$cols) },
1805 sub _dbh_execute_array {
1806 my ($self, $sth, $tuple_status, @extra) = @_;
1808 return $sth->execute_array({ArrayTupleStatus => $tuple_status});
1811 sub _dbh_execute_inserts_with_no_binds {
1812 my ($self, $sth, $count) = @_;
1816 my $dbh = $self->_get_dbh;
1817 local $dbh->{RaiseError} = 1;
1818 local $dbh->{PrintError} = 0;
1820 $sth->execute foreach 1..$count;
1826 # Make sure statement is finished even if there was an exception.
1831 $err = shift unless defined $err;
1835 $self->throw_exception($err) if defined $err;
1841 my ($self, $source, @args) = @_;
1843 my $bind_attrs = $self->source_bind_attributes($source);
1845 return $self->_execute('update' => [], $source, $bind_attrs, @args);
1850 my ($self, $source, @args) = @_;
1852 my $bind_attrs = $self->source_bind_attributes($source);
1854 return $self->_execute('delete' => [], $source, $bind_attrs, @args);
1857 # We were sent here because the $rs contains a complex search
1858 # which will require a subquery to select the correct rows
1859 # (i.e. joined or limited resultsets, or non-introspectable conditions)
1861 # Generating a single PK column subquery is trivial and supported
1862 # by all RDBMS. However if we have a multicolumn PK, things get ugly.
1863 # Look at _multipk_update_delete()
1864 sub _subq_update_delete {
1866 my ($rs, $op, $values) = @_;
1868 my $rsrc = $rs->result_source;
1870 # quick check if we got a sane rs on our hands
1871 my @pcols = $rsrc->_pri_cols;
1873 my $sel = $rs->_resolved_attrs->{select};
1874 $sel = [ $sel ] unless ref $sel eq 'ARRAY';
1877 join ("\x00", map { join '.', $rs->{attrs}{alias}, $_ } sort @pcols)
1879 join ("\x00", sort @$sel )
1881 $self->throw_exception (
1882 '_subq_update_delete can not be called on resultsets selecting columns other than the primary keys'
1889 $op eq 'update' ? $values : (),
1890 { $pcols[0] => { -in => $rs->as_query } },
1895 return $self->_multipk_update_delete (@_);
1899 # ANSI SQL does not provide a reliable way to perform a multicol-PK
1900 # resultset update/delete involving subqueries. So by default resort
1901 # to simple (and inefficient) delete_all style per-row opearations,
1902 # while allowing specific storages to override this with a faster
1905 sub _multipk_update_delete {
1906 return shift->_per_row_update_delete (@_);
1909 # This is the default loop used to delete/update rows for multi PK
1910 # resultsets, and used by mysql exclusively (because it can't do anything
1913 # We do not use $row->$op style queries, because resultset update/delete
1914 # is not expected to cascade (this is what delete_all/update_all is for).
1916 # There should be no race conditions as the entire operation is rolled
1919 sub _per_row_update_delete {
1921 my ($rs, $op, $values) = @_;
1923 my $rsrc = $rs->result_source;
1924 my @pcols = $rsrc->_pri_cols;
1926 my $guard = $self->txn_scope_guard;
1928 # emulate the return value of $sth->execute for non-selects
1929 my $row_cnt = '0E0';
1931 my $subrs_cur = $rs->cursor;
1932 my @all_pk = $subrs_cur->all;
1933 for my $pks ( @all_pk) {
1936 for my $i (0.. $#pcols) {
1937 $cond->{$pcols[$i]} = $pks->[$i];
1942 $op eq 'update' ? $values : (),
1956 $self->_execute($self->_select_args(@_));
1959 sub _select_args_to_query {
1962 # my ($op, $bind, $ident, $bind_attrs, $select, $cond, $rs_attrs, $rows, $offset)
1963 # = $self->_select_args($ident, $select, $cond, $attrs);
1964 my ($op, $bind, $ident, $bind_attrs, @args) =
1965 $self->_select_args(@_);
1967 # my ($sql, $prepared_bind) = $self->_prep_for_execute($op, $bind, $ident, [ $select, $cond, $rs_attrs, $rows, $offset ]);
1968 my ($sql, $prepared_bind) = $self->_prep_for_execute($op, $bind, $ident, \@args);
1969 $prepared_bind ||= [];
1972 ? ($sql, $prepared_bind, $bind_attrs)
1973 : \[ "($sql)", @$prepared_bind ]
1978 my ($self, $ident, $select, $where, $attrs) = @_;
1980 my $sql_maker = $self->sql_maker;
1981 my ($alias2source, $rs_alias) = $self->_resolve_ident_sources ($ident);
1988 $rs_alias && $alias2source->{$rs_alias}
1989 ? ( _rsroot_source_handle => $alias2source->{$rs_alias}->handle )
1994 # calculate bind_attrs before possible $ident mangling
1995 my $bind_attrs = {};
1996 for my $alias (keys %$alias2source) {
1997 my $bindtypes = $self->source_bind_attributes ($alias2source->{$alias}) || {};
1998 for my $col (keys %$bindtypes) {
2000 my $fqcn = join ('.', $alias, $col);
2001 $bind_attrs->{$fqcn} = $bindtypes->{$col} if $bindtypes->{$col};
2003 # Unqialified column names are nice, but at the same time can be
2004 # rather ambiguous. What we do here is basically go along with
2005 # the loop, adding an unqualified column slot to $bind_attrs,
2006 # alongside the fully qualified name. As soon as we encounter
2007 # another column by that name (which would imply another table)
2008 # we unset the unqualified slot and never add any info to it
2009 # to avoid erroneous type binding. If this happens the users
2010 # only choice will be to fully qualify his column name
2012 if (exists $bind_attrs->{$col}) {
2013 $bind_attrs->{$col} = {};
2016 $bind_attrs->{$col} = $bind_attrs->{$fqcn};
2021 # Sanity check the attributes (SQLMaker does it too, but
2022 # in case of a software_limit we'll never reach there)
2023 if (defined $attrs->{offset}) {
2024 $self->throw_exception('A supplied offset attribute must be a non-negative integer')
2025 if ( $attrs->{offset} =~ /\D/ or $attrs->{offset} < 0 );
2027 $attrs->{offset} ||= 0;
2029 if (defined $attrs->{rows}) {
2030 $self->throw_exception("The rows attribute must be a positive integer if present")
2031 if ( $attrs->{rows} =~ /\D/ or $attrs->{rows} <= 0 );
2033 elsif ($attrs->{offset}) {
2034 # MySQL actually recommends this approach. I cringe.
2035 $attrs->{rows} = $sql_maker->__max_int;
2040 # see if we need to tear the prefetch apart otherwise delegate the limiting to the
2041 # storage, unless software limit was requested
2044 ( $attrs->{rows} && keys %{$attrs->{collapse}} )
2046 # grouped prefetch (to satisfy group_by == select)
2047 ( $attrs->{group_by}
2049 @{$attrs->{group_by}}
2051 $attrs->{_prefetch_select}
2053 @{$attrs->{_prefetch_select}}
2056 ($ident, $select, $where, $attrs)
2057 = $self->_adjust_select_args_for_complex_prefetch ($ident, $select, $where, $attrs);
2059 elsif (! $attrs->{software_limit} ) {
2060 push @limit, $attrs->{rows}, $attrs->{offset};
2063 # try to simplify the joinmap further (prune unreferenced type-single joins)
2064 $ident = $self->_prune_unused_joins ($ident, $select, $where, $attrs);
2067 # This would be the point to deflate anything found in $where
2068 # (and leave $attrs->{bind} intact). Problem is - inflators historically
2069 # expect a row object. And all we have is a resultsource (it is trivial
2070 # to extract deflator coderefs via $alias2source above).
2072 # I don't see a way forward other than changing the way deflators are
2073 # invoked, and that's just bad...
2076 return ('select', $attrs->{bind}, $ident, $bind_attrs, $select, $where, $attrs, @limit);
2079 # Returns a counting SELECT for a simple count
2080 # query. Abstracted so that a storage could override
2081 # this to { count => 'firstcol' } or whatever makes
2082 # sense as a performance optimization
2084 #my ($self, $source, $rs_attrs) = @_;
2085 return { count => '*' };
2089 sub source_bind_attributes {
2090 my ($self, $source) = @_;
2092 my $bind_attributes;
2093 foreach my $column ($source->columns) {
2095 my $data_type = $source->column_info($column)->{data_type} || '';
2096 $bind_attributes->{$column} = $self->bind_attribute_by_data_type($data_type)
2100 return $bind_attributes;
2107 =item Arguments: $ident, $select, $condition, $attrs
2111 Handle a SQL select statement.
2117 my ($ident, $select, $condition, $attrs) = @_;
2118 return $self->cursor_class->new($self, \@_, $attrs);
2123 my ($rv, $sth, @bind) = $self->_select(@_);
2124 my @row = $sth->fetchrow_array;
2125 my @nextrow = $sth->fetchrow_array if @row;
2126 if(@row && @nextrow) {
2127 carp "Query returned more than one row. SQL that returns multiple rows is DEPRECATED for ->find and ->single";
2129 # Need to call finish() to work round broken DBDs
2134 =head2 sql_limit_dialect
2136 This is an accessor for the default SQL limit dialect used by a particular
2137 storage driver. Can be overriden by supplying an explicit L</limit_dialect>
2138 to L<DBIx::Class::Schema/connect>. For a list of available limit dialects
2139 see L<DBIx::Class::SQLMaker::LimitDialects>.
2145 =item Arguments: $sql
2149 Returns a L<DBI> sth (statement handle) for the supplied SQL.
2154 my ($self, $dbh, $sql) = @_;
2156 # 3 is the if_active parameter which avoids active sth re-use
2157 my $sth = $self->disable_sth_caching
2158 ? $dbh->prepare($sql)
2159 : $dbh->prepare_cached($sql, {}, 3);
2161 # XXX You would think RaiseError would make this impossible,
2162 # but apparently that's not true :(
2163 $self->throw_exception($dbh->errstr) if !$sth;
2169 my ($self, $sql) = @_;
2170 $self->dbh_do('_dbh_sth', $sql); # retry over disconnects
2173 sub _dbh_columns_info_for {
2174 my ($self, $dbh, $table) = @_;
2176 if ($dbh->can('column_info')) {
2180 my ($schema,$tab) = $table =~ /^(.+?)\.(.+)$/ ? ($1,$2) : (undef,$table);
2181 my $sth = $dbh->column_info( undef,$schema, $tab, '%' );
2183 while ( my $info = $sth->fetchrow_hashref() ){
2185 $column_info{data_type} = $info->{TYPE_NAME};
2186 $column_info{size} = $info->{COLUMN_SIZE};
2187 $column_info{is_nullable} = $info->{NULLABLE} ? 1 : 0;
2188 $column_info{default_value} = $info->{COLUMN_DEF};
2189 my $col_name = $info->{COLUMN_NAME};
2190 $col_name =~ s/^\"(.*)\"$/$1/;
2192 $result{$col_name} = \%column_info;
2197 return \%result if !$caught && scalar keys %result;
2201 my $sth = $dbh->prepare($self->sql_maker->select($table, undef, \'1 = 0'));
2203 my @columns = @{$sth->{NAME_lc}};
2204 for my $i ( 0 .. $#columns ){
2206 $column_info{data_type} = $sth->{TYPE}->[$i];
2207 $column_info{size} = $sth->{PRECISION}->[$i];
2208 $column_info{is_nullable} = $sth->{NULLABLE}->[$i] ? 1 : 0;
2210 if ($column_info{data_type} =~ m/^(.*?)\((.*?)\)$/) {
2211 $column_info{data_type} = $1;
2212 $column_info{size} = $2;
2215 $result{$columns[$i]} = \%column_info;
2219 foreach my $col (keys %result) {
2220 my $colinfo = $result{$col};
2221 my $type_num = $colinfo->{data_type};
2223 if(defined $type_num && $dbh->can('type_info')) {
2224 my $type_info = $dbh->type_info($type_num);
2225 $type_name = $type_info->{TYPE_NAME} if $type_info;
2226 $colinfo->{data_type} = $type_name if $type_name;
2233 sub columns_info_for {
2234 my ($self, $table) = @_;
2235 $self->_dbh_columns_info_for ($self->_get_dbh, $table);
2238 =head2 last_insert_id
2240 Return the row id of the last insert.
2244 sub _dbh_last_insert_id {
2245 my ($self, $dbh, $source, $col) = @_;
2247 my $id = try { $dbh->last_insert_id (undef, undef, $source->name, $col) };
2249 return $id if defined $id;
2251 my $class = ref $self;
2252 $self->throw_exception ("No storage specific _dbh_last_insert_id() method implemented in $class, and the generic DBI::last_insert_id() failed");
2255 sub last_insert_id {
2257 $self->_dbh_last_insert_id ($self->_dbh, @_);
2260 =head2 _native_data_type
2264 =item Arguments: $type_name
2268 This API is B<EXPERIMENTAL>, will almost definitely change in the future, and
2269 currently only used by L<::AutoCast|DBIx::Class::Storage::DBI::AutoCast> and
2270 L<::Sybase::ASE|DBIx::Class::Storage::DBI::Sybase::ASE>.
2272 The default implementation returns C<undef>, implement in your Storage driver if
2273 you need this functionality.
2275 Should map types from other databases to the native RDBMS type, for example
2276 C<VARCHAR2> to C<VARCHAR>.
2278 Types with modifiers should map to the underlying data type. For example,
2279 C<INTEGER AUTO_INCREMENT> should become C<INTEGER>.
2281 Composite types should map to the container type, for example
2282 C<ENUM(foo,bar,baz)> becomes C<ENUM>.
2286 sub _native_data_type {
2287 #my ($self, $data_type) = @_;
2291 # Check if placeholders are supported at all
2292 sub _determine_supports_placeholders {
2294 my $dbh = $self->_get_dbh;
2296 # some drivers provide a $dbh attribute (e.g. Sybase and $dbh->{syb_dynamic_supported})
2297 # but it is inaccurate more often than not
2299 local $dbh->{PrintError} = 0;
2300 local $dbh->{RaiseError} = 1;
2301 $dbh->do('select ?', {}, 1);
2309 # Check if placeholders bound to non-string types throw exceptions
2311 sub _determine_supports_typeless_placeholders {
2313 my $dbh = $self->_get_dbh;
2316 local $dbh->{PrintError} = 0;
2317 local $dbh->{RaiseError} = 1;
2318 # this specifically tests a bind that is NOT a string
2319 $dbh->do('select 1 where 1 = ?', {}, 1);
2329 Returns the database driver name.
2334 shift->_get_dbh->{Driver}->{Name};
2337 =head2 bind_attribute_by_data_type
2339 Given a datatype from column info, returns a database specific bind
2340 attribute for C<< $dbh->bind_param($val,$attribute) >> or nothing if we will
2341 let the database planner just handle it.
2343 Generally only needed for special case column types, like bytea in postgres.
2347 sub bind_attribute_by_data_type {
2351 =head2 is_datatype_numeric
2353 Given a datatype from column_info, returns a boolean value indicating if
2354 the current RDBMS considers it a numeric value. This controls how
2355 L<DBIx::Class::Row/set_column> decides whether to mark the column as
2356 dirty - when the datatype is deemed numeric a C<< != >> comparison will
2357 be performed instead of the usual C<eq>.
2361 sub is_datatype_numeric {
2362 my ($self, $dt) = @_;
2364 return 0 unless $dt;
2366 return $dt =~ /^ (?:
2367 numeric | int(?:eger)? | (?:tiny|small|medium|big)int | dec(?:imal)? | real | float | double (?: \s+ precision)? | (?:big)?serial
2372 =head2 create_ddl_dir
2376 =item Arguments: $schema \@databases, $version, $directory, $preversion, \%sqlt_args
2380 Creates a SQL file based on the Schema, for each of the specified
2381 database engines in C<\@databases> in the given directory.
2382 (note: specify L<SQL::Translator> names, not L<DBI> driver names).
2384 Given a previous version number, this will also create a file containing
2385 the ALTER TABLE statements to transform the previous schema into the
2386 current one. Note that these statements may contain C<DROP TABLE> or
2387 C<DROP COLUMN> statements that can potentially destroy data.
2389 The file names are created using the C<ddl_filename> method below, please
2390 override this method in your schema if you would like a different file
2391 name format. For the ALTER file, the same format is used, replacing
2392 $version in the name with "$preversion-$version".
2394 See L<SQL::Translator/METHODS> for a list of values for C<\%sqlt_args>.
2395 The most common value for this would be C<< { add_drop_table => 1 } >>
2396 to have the SQL produced include a C<DROP TABLE> statement for each table
2397 created. For quoting purposes supply C<quote_table_names> and
2398 C<quote_field_names>.
2400 If no arguments are passed, then the following default values are assumed:
2404 =item databases - ['MySQL', 'SQLite', 'PostgreSQL']
2406 =item version - $schema->schema_version
2408 =item directory - './'
2410 =item preversion - <none>
2414 By default, C<\%sqlt_args> will have
2416 { add_drop_table => 1, ignore_constraint_names => 1, ignore_index_names => 1 }
2418 merged with the hash passed in. To disable any of those features, pass in a
2419 hashref like the following
2421 { ignore_constraint_names => 0, # ... other options }
2424 WARNING: You are strongly advised to check all SQL files created, before applying
2429 sub create_ddl_dir {
2430 my ($self, $schema, $databases, $version, $dir, $preversion, $sqltargs) = @_;
2433 carp "No directory given, using ./\n";
2438 make_path ("$dir") # make_path does not like objects (i.e. Path::Class::Dir)
2440 $self->throw_exception(
2441 "Failed to create '$dir': " . ($! || $@ || 'error unknow')
2445 $self->throw_exception ("Directory '$dir' does not exist\n") unless(-d $dir);
2447 $databases ||= ['MySQL', 'SQLite', 'PostgreSQL'];
2448 $databases = [ $databases ] if(ref($databases) ne 'ARRAY');
2450 my $schema_version = $schema->schema_version || '1.x';
2451 $version ||= $schema_version;
2454 add_drop_table => 1,
2455 ignore_constraint_names => 1,
2456 ignore_index_names => 1,
2460 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy')) {
2461 $self->throw_exception("Can't create a ddl file without " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2464 my $sqlt = SQL::Translator->new( $sqltargs );
2466 $sqlt->parser('SQL::Translator::Parser::DBIx::Class');
2467 my $sqlt_schema = $sqlt->translate({ data => $schema })
2468 or $self->throw_exception ($sqlt->error);
2470 foreach my $db (@$databases) {
2472 $sqlt->{schema} = $sqlt_schema;
2473 $sqlt->producer($db);
2476 my $filename = $schema->ddl_filename($db, $version, $dir);
2477 if (-e $filename && ($version eq $schema_version )) {
2478 # if we are dumping the current version, overwrite the DDL
2479 carp "Overwriting existing DDL file - $filename";
2483 my $output = $sqlt->translate;
2485 carp("Failed to translate to $db, skipping. (" . $sqlt->error . ")");
2488 if(!open($file, ">$filename")) {
2489 $self->throw_exception("Can't open $filename for writing ($!)");
2492 print $file $output;
2495 next unless ($preversion);
2497 require SQL::Translator::Diff;
2499 my $prefilename = $schema->ddl_filename($db, $preversion, $dir);
2500 if(!-e $prefilename) {
2501 carp("No previous schema file found ($prefilename)");
2505 my $difffile = $schema->ddl_filename($db, $version, $dir, $preversion);
2507 carp("Overwriting existing diff file - $difffile");
2513 my $t = SQL::Translator->new($sqltargs);
2518 or $self->throw_exception ($t->error);
2520 my $out = $t->translate( $prefilename )
2521 or $self->throw_exception ($t->error);
2523 $source_schema = $t->schema;
2525 $source_schema->name( $prefilename )
2526 unless ( $source_schema->name );
2529 # The "new" style of producers have sane normalization and can support
2530 # diffing a SQL file against a DBIC->SQLT schema. Old style ones don't
2531 # And we have to diff parsed SQL against parsed SQL.
2532 my $dest_schema = $sqlt_schema;
2534 unless ( "SQL::Translator::Producer::$db"->can('preprocess_schema') ) {
2535 my $t = SQL::Translator->new($sqltargs);
2540 or $self->throw_exception ($t->error);
2542 my $out = $t->translate( $filename )
2543 or $self->throw_exception ($t->error);
2545 $dest_schema = $t->schema;
2547 $dest_schema->name( $filename )
2548 unless $dest_schema->name;
2551 my $diff = SQL::Translator::Diff::schema_diff($source_schema, $db,
2555 if(!open $file, ">$difffile") {
2556 $self->throw_exception("Can't write to $difffile ($!)");
2564 =head2 deployment_statements
2568 =item Arguments: $schema, $type, $version, $directory, $sqlt_args
2572 Returns the statements used by L</deploy> and L<DBIx::Class::Schema/deploy>.
2574 The L<SQL::Translator> (not L<DBI>) database driver name can be explicitly
2575 provided in C<$type>, otherwise the result of L</sqlt_type> is used as default.
2577 C<$directory> is used to return statements from files in a previously created
2578 L</create_ddl_dir> directory and is optional. The filenames are constructed
2579 from L<DBIx::Class::Schema/ddl_filename>, the schema name and the C<$version>.
2581 If no C<$directory> is specified then the statements are constructed on the
2582 fly using L<SQL::Translator> and C<$version> is ignored.
2584 See L<SQL::Translator/METHODS> for a list of values for C<$sqlt_args>.
2588 sub deployment_statements {
2589 my ($self, $schema, $type, $version, $dir, $sqltargs) = @_;
2590 $type ||= $self->sqlt_type;
2591 $version ||= $schema->schema_version || '1.x';
2593 my $filename = $schema->ddl_filename($type, $version, $dir);
2597 open($file, "<$filename")
2598 or $self->throw_exception("Can't open $filename ($!)");
2601 return join('', @rows);
2604 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy') ) {
2605 $self->throw_exception("Can't deploy without a ddl_dir or " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2608 # sources needs to be a parser arg, but for simplicty allow at top level
2610 $sqltargs->{parser_args}{sources} = delete $sqltargs->{sources}
2611 if exists $sqltargs->{sources};
2613 my $tr = SQL::Translator->new(
2614 producer => "SQL::Translator::Producer::${type}",
2616 parser => 'SQL::Translator::Parser::DBIx::Class',
2623 @ret = $tr->translate;
2626 $ret[0] = $tr->translate;
2629 $self->throw_exception( 'Unable to produce deployment statements: ' . $tr->error)
2630 unless (@ret && defined $ret[0]);
2632 return $wa ? @ret : $ret[0];
2636 my ($self, $schema, $type, $sqltargs, $dir) = @_;
2639 return if($line =~ /^--/);
2641 # next if($line =~ /^DROP/m);
2642 return if($line =~ /^BEGIN TRANSACTION/m);
2643 return if($line =~ /^COMMIT/m);
2644 return if $line =~ /^\s+$/; # skip whitespace only
2645 $self->_query_start($line);
2647 # do a dbh_do cycle here, as we need some error checking in
2648 # place (even though we will ignore errors)
2649 $self->dbh_do (sub { $_[1]->do($line) });
2651 carp qq{$_ (running "${line}")};
2653 $self->_query_end($line);
2655 my @statements = $schema->deployment_statements($type, undef, $dir, { %{ $sqltargs || {} }, no_comments => 1 } );
2656 if (@statements > 1) {
2657 foreach my $statement (@statements) {
2658 $deploy->( $statement );
2661 elsif (@statements == 1) {
2662 foreach my $line ( split(";\n", $statements[0])) {
2668 =head2 datetime_parser
2670 Returns the datetime parser class
2674 sub datetime_parser {
2676 return $self->{datetime_parser} ||= do {
2677 $self->build_datetime_parser(@_);
2681 =head2 datetime_parser_type
2683 Defines (returns) the datetime parser class - currently hardwired to
2684 L<DateTime::Format::MySQL>
2688 sub datetime_parser_type { "DateTime::Format::MySQL"; }
2690 =head2 build_datetime_parser
2692 See L</datetime_parser>
2696 sub build_datetime_parser {
2698 my $type = $self->datetime_parser_type(@_);
2699 $self->ensure_class_loaded ($type);
2704 =head2 is_replicating
2706 A boolean that reports if a particular L<DBIx::Class::Storage::DBI> is set to
2707 replicate from a master database. Default is undef, which is the result
2708 returned by databases that don't support replication.
2712 sub is_replicating {
2717 =head2 lag_behind_master
2719 Returns a number that represents a certain amount of lag behind a master db
2720 when a given storage is replicating. The number is database dependent, but
2721 starts at zero and increases with the amount of lag. Default in undef
2725 sub lag_behind_master {
2729 =head2 relname_to_table_alias
2733 =item Arguments: $relname, $join_count
2737 L<DBIx::Class> uses L<DBIx::Class::Relationship> names as table aliases in
2740 This hook is to allow specific L<DBIx::Class::Storage> drivers to change the
2741 way these aliases are named.
2743 The default behavior is C<< "$relname_$join_count" if $join_count > 1 >>,
2744 otherwise C<"$relname">.
2748 sub relname_to_table_alias {
2749 my ($self, $relname, $join_count) = @_;
2751 my $alias = ($join_count && $join_count > 1 ?
2752 join('_', $relname, $join_count) : $relname);
2761 =head2 DBIx::Class and AutoCommit
2763 DBIx::Class can do some wonderful magic with handling exceptions,
2764 disconnections, and transactions when you use C<< AutoCommit => 1 >>
2765 (the default) combined with C<txn_do> for transaction support.
2767 If you set C<< AutoCommit => 0 >> in your connect info, then you are always
2768 in an assumed transaction between commits, and you're telling us you'd
2769 like to manage that manually. A lot of the magic protections offered by
2770 this module will go away. We can't protect you from exceptions due to database
2771 disconnects because we don't know anything about how to restart your
2772 transactions. You're on your own for handling all sorts of exceptional
2773 cases if you choose the C<< AutoCommit => 0 >> path, just as you would
2779 Matt S. Trout <mst@shadowcatsystems.co.uk>
2781 Andy Grundman <andy@hybridized.org>
2785 You may distribute this code under the same terms as Perl itself.