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 DBIx::Class::Carp;
11 use DBIx::Class::Exception;
12 use Scalar::Util qw/refaddr weaken reftype blessed/;
13 use List::Util qw/first/;
14 use Sub::Name 'subname';
15 use Context::Preserve 'preserve_context';
18 use Data::Compare (); # no imports!!! guard against insane architecture
19 use DBI::Const::GetInfoType (); # no import of retarded global hash
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/
26 sql_limit_dialect sql_quote_char sql_name_sep
29 __PACKAGE__->mk_group_accessors('component_class' => qw/sql_maker_class datetime_parser_type/);
31 __PACKAGE__->sql_maker_class('DBIx::Class::SQLMaker');
32 __PACKAGE__->datetime_parser_type('DateTime::Format::MySQL'); # historic default
34 __PACKAGE__->sql_name_sep('.');
36 __PACKAGE__->mk_group_accessors('simple' => qw/
37 _connect_info _dbi_connect_info _dbic_connect_attributes _driver_determined
38 _dbh _dbh_details _conn_pid _sql_maker _sql_maker_opts _dbh_autocommit
39 _perform_autoinc_retrieval _autoinc_supplied_for_op
42 # the values for these accessors are picked out (and deleted) from
43 # the attribute hashref passed to connect_info
44 my @storage_options = qw/
45 on_connect_call on_disconnect_call on_connect_do on_disconnect_do
46 disable_sth_caching unsafe auto_savepoint
48 __PACKAGE__->mk_group_accessors('simple' => @storage_options);
51 # capability definitions, using a 2-tiered accessor system
54 # A driver/user may define _use_X, which blindly without any checks says:
55 # "(do not) use this capability", (use_dbms_capability is an "inherited"
58 # If _use_X is undef, _supports_X is then queried. This is a "simple" style
59 # accessor, which in turn calls _determine_supports_X, and stores the return
60 # in a special slot on the storage object, which is wiped every time a $dbh
61 # reconnection takes place (it is not guaranteed that upon reconnection we
62 # will get the same rdbms version). _determine_supports_X does not need to
63 # exist on a driver, as we ->can for it before calling.
65 my @capabilities = (qw/
67 insert_returning_bound
76 __PACKAGE__->mk_group_accessors( dbms_capability => map { "_supports_$_" } @capabilities );
77 __PACKAGE__->mk_group_accessors( use_dbms_capability => map { "_use_$_" } (@capabilities ) );
79 # on by default, not strictly a capability (pending rewrite)
80 __PACKAGE__->_use_join_optimizer (1);
81 sub _determine_supports_join_optimizer { 1 };
83 # Each of these methods need _determine_driver called before itself
84 # in order to function reliably. This is a purely DRY optimization
86 # get_(use)_dbms_capability need to be called on the correct Storage
87 # class, as _use_X may be hardcoded class-wide, and _supports_X calls
88 # _determine_supports_X which obv. needs a correct driver as well
89 my @rdbms_specific_methods = qw/
103 with_deferred_fk_checks
105 get_use_dbms_capability
112 for my $meth (@rdbms_specific_methods) {
114 my $orig = __PACKAGE__->can ($meth)
115 or die "$meth is not a ::Storage::DBI method!";
118 no warnings qw/redefine/;
119 *{__PACKAGE__ ."::$meth"} = subname $meth => sub {
121 # only fire when invoked on an instance, a valid class-based invocation
122 # would e.g. be setting a default for an inherited accessor
125 ! $_[0]->_driver_determined
127 ! $_[0]->{_in_determine_driver}
129 $_[0]->_determine_driver;
131 # This for some reason crashes and burns on perl 5.8.1
132 # IFF the method ends up throwing an exception
133 #goto $_[0]->can ($meth);
135 my $cref = $_[0]->can ($meth);
145 DBIx::Class::Storage::DBI - DBI storage handler
149 my $schema = MySchema->connect('dbi:SQLite:my.db');
151 $schema->storage->debug(1);
153 my @stuff = $schema->storage->dbh_do(
155 my ($storage, $dbh, @args) = @_;
156 $dbh->do("DROP TABLE authors");
161 $schema->resultset('Book')->search({
162 written_on => $schema->storage->datetime_parser->format_datetime(DateTime->now)
167 This class represents the connection to an RDBMS via L<DBI>. See
168 L<DBIx::Class::Storage> for general information. This pod only
169 documents DBI-specific methods and behaviors.
176 my $new = shift->next::method(@_);
178 $new->_sql_maker_opts({});
179 $new->_dbh_details({});
180 $new->{_in_do_block} = 0;
181 $new->{_dbh_gen} = 0;
183 # read below to see what this does
184 $new->_arm_global_destructor;
189 # This is hack to work around perl shooting stuff in random
190 # order on exit(). If we do not walk the remaining storage
191 # objects in an END block, there is a *small but real* chance
192 # of a fork()ed child to kill the parent's shared DBI handle,
193 # *before perl reaches the DESTROY in this package*
194 # Yes, it is ugly and effective.
195 # Additionally this registry is used by the CLONE method to
196 # make sure no handles are shared between threads
198 my %seek_and_destroy;
200 sub _arm_global_destructor {
202 $seek_and_destroy{ refaddr($_[0]) } = $_[0]
207 local $?; # just in case the DBI destructor changes it somehow
209 # destroy just the object if not native to this process
210 $_->_verify_pid for (grep
212 values %seek_and_destroy
217 # As per DBI's recommendation, DBIC disconnects all handles as
218 # soon as possible (DBIC will reconnect only on demand from within
220 my @instances = grep { defined $_ } values %seek_and_destroy;
222 $_->{_dbh_gen}++; # so that existing cursors will drop as well
225 $_->transaction_depth(0);
229 # properly renumber all existing refs
230 %seek_and_destroy = ();
231 $_->_arm_global_destructor for @instances;
238 # some databases spew warnings on implicit disconnect
239 $self->_verify_pid unless DBIx::Class::_ENV_::BROKEN_FORK;
240 local $SIG{__WARN__} = sub {};
243 # this op is necessary, since the very last perl runtime statement
244 # triggers a global destruction shootout, and the $SIG localization
245 # may very well be destroyed before perl actually gets to do the
250 # handle pid changes correctly - do not destroy parent's connection
254 my $pid = $self->_conn_pid;
255 if( defined $pid and $pid != $$ and my $dbh = $self->_dbh ) {
256 $dbh->{InactiveDestroy} = 1;
259 $self->transaction_depth(0);
260 $self->savepoints([]);
268 This method is normally called by L<DBIx::Class::Schema/connection>, which
269 encapsulates its argument list in an arrayref before passing them here.
271 The argument list may contain:
277 The same 4-element argument set one would normally pass to
278 L<DBI/connect>, optionally followed by
279 L<extra attributes|/DBIx::Class specific connection attributes>
280 recognized by DBIx::Class:
282 $connect_info_args = [ $dsn, $user, $password, \%dbi_attributes?, \%extra_attributes? ];
286 A single code reference which returns a connected
287 L<DBI database handle|DBI/connect> optionally followed by
288 L<extra attributes|/DBIx::Class specific connection attributes> recognized
291 $connect_info_args = [ sub { DBI->connect (...) }, \%extra_attributes? ];
295 A single hashref with all the attributes and the dsn/user/password
298 $connect_info_args = [{
306 $connect_info_args = [{
307 dbh_maker => sub { DBI->connect (...) },
312 This is particularly useful for L<Catalyst> based applications, allowing the
313 following config (L<Config::General> style):
318 dsn dbi:mysql:database=test
325 The C<dsn>/C<user>/C<password> combination can be substituted by the
326 C<dbh_maker> key whose value is a coderef that returns a connected
327 L<DBI database handle|DBI/connect>
331 Please note that the L<DBI> docs recommend that you always explicitly
332 set C<AutoCommit> to either I<0> or I<1>. L<DBIx::Class> further
333 recommends that it be set to I<1>, and that you perform transactions
334 via our L<DBIx::Class::Schema/txn_do> method. L<DBIx::Class> will set it
335 to I<1> if you do not do explicitly set it to zero. This is the default
336 for most DBDs. See L</DBIx::Class and AutoCommit> for details.
338 =head3 DBIx::Class specific connection attributes
340 In addition to the standard L<DBI|DBI/ATTRIBUTES COMMON TO ALL HANDLES>
341 L<connection|DBI/Database Handle Attributes> attributes, DBIx::Class recognizes
342 the following connection options. These options can be mixed in with your other
343 L<DBI> connection attributes, or placed in a separate hashref
344 (C<\%extra_attributes>) as shown above.
346 Every time C<connect_info> is invoked, any previous settings for
347 these options will be cleared before setting the new ones, regardless of
348 whether any options are specified in the new C<connect_info>.
355 Specifies things to do immediately after connecting or re-connecting to
356 the database. Its value may contain:
362 This contains one SQL statement to execute.
364 =item an array reference
366 This contains SQL statements to execute in order. Each element contains
367 a string or a code reference that returns a string.
369 =item a code reference
371 This contains some code to execute. Unlike code references within an
372 array reference, its return value is ignored.
376 =item on_disconnect_do
378 Takes arguments in the same form as L</on_connect_do> and executes them
379 immediately before disconnecting from the database.
381 Note, this only runs if you explicitly call L</disconnect> on the
384 =item on_connect_call
386 A more generalized form of L</on_connect_do> that calls the specified
387 C<connect_call_METHOD> methods in your storage driver.
389 on_connect_do => 'select 1'
393 on_connect_call => [ [ do_sql => 'select 1' ] ]
395 Its values may contain:
401 Will call the C<connect_call_METHOD> method.
403 =item a code reference
405 Will execute C<< $code->($storage) >>
407 =item an array reference
409 Each value can be a method name or code reference.
411 =item an array of arrays
413 For each array, the first item is taken to be the C<connect_call_> method name
414 or code reference, and the rest are parameters to it.
418 Some predefined storage methods you may use:
424 Executes a SQL string or a code reference that returns a SQL string. This is
425 what L</on_connect_do> and L</on_disconnect_do> use.
433 Will execute the scalar as SQL.
437 Taken to be arguments to L<DBI/do>, the SQL string optionally followed by the
438 attributes hashref and bind values.
440 =item a code reference
442 Will execute C<< $code->($storage) >> and execute the return array refs as
449 Execute any statements necessary to initialize the database session to return
450 and accept datetime/timestamp values used with
451 L<DBIx::Class::InflateColumn::DateTime>.
453 Only necessary for some databases, see your specific storage driver for
454 implementation details.
458 =item on_disconnect_call
460 Takes arguments in the same form as L</on_connect_call> and executes them
461 immediately before disconnecting from the database.
463 Calls the C<disconnect_call_METHOD> methods as opposed to the
464 C<connect_call_METHOD> methods called by L</on_connect_call>.
466 Note, this only runs if you explicitly call L</disconnect> on the
469 =item disable_sth_caching
471 If set to a true value, this option will disable the caching of
472 statement handles via L<DBI/prepare_cached>.
476 Sets a specific SQL::Abstract::Limit-style limit dialect, overriding the
477 default L</sql_limit_dialect> setting of the storage (if any). For a list
478 of available limit dialects see L<DBIx::Class::SQLMaker::LimitDialects>.
482 When true automatically sets L</quote_char> and L</name_sep> to the characters
483 appropriate for your particular RDBMS. This option is preferred over specifying
484 L</quote_char> directly.
488 Specifies what characters to use to quote table and column names.
490 C<quote_char> expects either a single character, in which case is it
491 is placed on either side of the table/column name, or an arrayref of length
492 2 in which case the table/column name is placed between the elements.
494 For example under MySQL you should use C<< quote_char => '`' >>, and for
495 SQL Server you should use C<< quote_char => [qw/[ ]/] >>.
499 This parameter is only useful in conjunction with C<quote_char>, and is used to
500 specify the character that separates elements (schemas, tables, columns) from
501 each other. If unspecified it defaults to the most commonly used C<.>.
505 This Storage driver normally installs its own C<HandleError>, sets
506 C<RaiseError> and C<ShowErrorStatement> on, and sets C<PrintError> off on
507 all database handles, including those supplied by a coderef. It does this
508 so that it can have consistent and useful error behavior.
510 If you set this option to a true value, Storage will not do its usual
511 modifications to the database handle's attributes, and instead relies on
512 the settings in your connect_info DBI options (or the values you set in
513 your connection coderef, in the case that you are connecting via coderef).
515 Note that your custom settings can cause Storage to malfunction,
516 especially if you set a C<HandleError> handler that suppresses exceptions
517 and/or disable C<RaiseError>.
521 If this option is true, L<DBIx::Class> will use savepoints when nesting
522 transactions, making it possible to recover from failure in the inner
523 transaction without having to abort all outer transactions.
527 Use this argument to supply a cursor class other than the default
528 L<DBIx::Class::Storage::DBI::Cursor>.
532 Some real-life examples of arguments to L</connect_info> and
533 L<DBIx::Class::Schema/connect>
535 # Simple SQLite connection
536 ->connect_info([ 'dbi:SQLite:./foo.db' ]);
539 ->connect_info([ sub { DBI->connect(...) } ]);
541 # Connect via subref in hashref
543 dbh_maker => sub { DBI->connect(...) },
544 on_connect_do => 'alter session ...',
547 # A bit more complicated
554 { quote_char => q{"} },
558 # Equivalent to the previous example
564 { AutoCommit => 1, quote_char => q{"}, name_sep => q{.} },
568 # Same, but with hashref as argument
569 # See parse_connect_info for explanation
572 dsn => 'dbi:Pg:dbname=foo',
574 password => 'my_pg_password',
581 # Subref + DBIx::Class-specific connection options
584 sub { DBI->connect(...) },
588 on_connect_do => ['SET search_path TO myschema,otherschema,public'],
589 disable_sth_caching => 1,
599 my ($self, $info) = @_;
601 return $self->_connect_info if !$info;
603 $self->_connect_info($info); # copy for _connect_info
605 $info = $self->_normalize_connect_info($info)
606 if ref $info eq 'ARRAY';
608 for my $storage_opt (keys %{ $info->{storage_options} }) {
609 my $value = $info->{storage_options}{$storage_opt};
611 $self->$storage_opt($value);
614 # Kill sql_maker/_sql_maker_opts, so we get a fresh one with only
615 # the new set of options
616 $self->_sql_maker(undef);
617 $self->_sql_maker_opts({});
619 for my $sql_maker_opt (keys %{ $info->{sql_maker_options} }) {
620 my $value = $info->{sql_maker_options}{$sql_maker_opt};
622 $self->_sql_maker_opts->{$sql_maker_opt} = $value;
626 %{ $self->_default_dbi_connect_attributes || {} },
627 %{ $info->{attributes} || {} },
630 my @args = @{ $info->{arguments} };
632 if (keys %attrs and ref $args[0] ne 'CODE') {
634 'You provided explicit AutoCommit => 0 in your connection_info. '
635 . 'This is almost universally a bad idea (see the footnotes of '
636 . 'DBIx::Class::Storage::DBI for more info). If you still want to '
637 . 'do this you can set $ENV{DBIC_UNSAFE_AUTOCOMMIT_OK} to disable '
639 if ! $attrs{AutoCommit} and ! $ENV{DBIC_UNSAFE_AUTOCOMMIT_OK};
641 push @args, \%attrs if keys %attrs;
643 $self->_dbi_connect_info(\@args);
646 # save attributes them in a separate accessor so they are always
647 # introspectable, even in case of a CODE $dbhmaker
648 $self->_dbic_connect_attributes (\%attrs);
650 return $self->_connect_info;
653 sub _normalize_connect_info {
654 my ($self, $info_arg) = @_;
657 my @args = @$info_arg; # take a shallow copy for further mutilation
659 # combine/pre-parse arguments depending on invocation style
662 if (ref $args[0] eq 'CODE') { # coderef with optional \%extra_attributes
663 %attrs = %{ $args[1] || {} };
666 elsif (ref $args[0] eq 'HASH') { # single hashref (i.e. Catalyst config)
667 %attrs = %{$args[0]};
669 if (my $code = delete $attrs{dbh_maker}) {
672 my @ignored = grep { delete $attrs{$_} } (qw/dsn user password/);
675 'Attribute(s) %s in connect_info were ignored, as they can not be applied '
676 . "to the result of 'dbh_maker'",
678 join (', ', map { "'$_'" } (@ignored) ),
683 @args = delete @attrs{qw/dsn user password/};
686 else { # otherwise assume dsn/user/password + \%attrs + \%extra_attrs
688 % { $args[3] || {} },
689 % { $args[4] || {} },
691 @args = @args[0,1,2];
694 $info{arguments} = \@args;
696 my @storage_opts = grep exists $attrs{$_},
697 @storage_options, 'cursor_class';
699 @{ $info{storage_options} }{@storage_opts} =
700 delete @attrs{@storage_opts} if @storage_opts;
702 my @sql_maker_opts = grep exists $attrs{$_},
703 qw/limit_dialect quote_char name_sep quote_names/;
705 @{ $info{sql_maker_options} }{@sql_maker_opts} =
706 delete @attrs{@sql_maker_opts} if @sql_maker_opts;
708 $info{attributes} = \%attrs if %attrs;
713 sub _default_dbi_connect_attributes () {
718 ShowErrorStatement => 1,
724 This method is deprecated in favour of setting via L</connect_info>.
728 =head2 on_disconnect_do
730 This method is deprecated in favour of setting via L</connect_info>.
734 sub _parse_connect_do {
735 my ($self, $type) = @_;
737 my $val = $self->$type;
738 return () if not defined $val;
743 push @res, [ 'do_sql', $val ];
744 } elsif (ref($val) eq 'CODE') {
746 } elsif (ref($val) eq 'ARRAY') {
747 push @res, map { [ 'do_sql', $_ ] } @$val;
749 $self->throw_exception("Invalid type for $type: ".ref($val));
757 Arguments: ($subref | $method_name), @extra_coderef_args?
759 Execute the given $subref or $method_name using the new exception-based
760 connection management.
762 The first two arguments will be the storage object that C<dbh_do> was called
763 on and a database handle to use. Any additional arguments will be passed
764 verbatim to the called subref as arguments 2 and onwards.
766 Using this (instead of $self->_dbh or $self->dbh) ensures correct
767 exception handling and reconnection (or failover in future subclasses).
769 Your subref should have no side-effects outside of the database, as
770 there is the potential for your subref to be partially double-executed
771 if the database connection was stale/dysfunctional.
775 my @stuff = $schema->storage->dbh_do(
777 my ($storage, $dbh, @cols) = @_;
778 my $cols = join(q{, }, @cols);
779 $dbh->selectrow_array("SELECT $cols FROM foo");
788 my $run_target = shift;
790 # short circuit when we know there is no need for a runner
792 # FIXME - asumption may be wrong
793 # the rationale for the txn_depth check is that if this block is a part
794 # of a larger transaction, everything up to that point is screwed anyway
795 return $self->$run_target($self->_get_dbh, @_)
796 if $self->{_in_do_block} or $self->transaction_depth;
800 DBIx::Class::Storage::BlockRunner->new(
802 run_code => sub { $self->$run_target ($self->_get_dbh, @$args ) },
804 retry_handler => sub { ! ( $_[0]->retried_count or $_[0]->storage->connected ) },
809 $_[0]->_get_dbh; # connects or reconnects on pid change, necessary to grab correct txn_depth
810 shift->next::method(@_);
815 Our C<disconnect> method also performs a rollback first if the
816 database is not in C<AutoCommit> mode.
826 push @actions, ( $self->on_disconnect_call || () );
827 push @actions, $self->_parse_connect_do ('on_disconnect_do');
829 $self->_do_connection_actions(disconnect_call_ => $_) for @actions;
831 # stops the "implicit rollback on disconnect" warning
832 $self->_exec_txn_rollback unless $self->_dbh_autocommit;
834 %{ $self->_dbh->{CachedKids} } = ();
835 $self->_dbh->disconnect;
841 =head2 with_deferred_fk_checks
845 =item Arguments: C<$coderef>
847 =item Return Value: The return value of $coderef
851 Storage specific method to run the code ref with FK checks deferred or
852 in MySQL's case disabled entirely.
856 # Storage subclasses should override this
857 sub with_deferred_fk_checks {
858 my ($self, $sub) = @_;
866 =item Arguments: none
868 =item Return Value: 1|0
872 Verifies that the current database handle is active and ready to execute
873 an SQL statement (e.g. the connection did not get stale, server is still
874 answering, etc.) This method is used internally by L</dbh>.
880 return 0 unless $self->_seems_connected;
883 local $self->_dbh->{RaiseError} = 1;
888 sub _seems_connected {
891 $self->_verify_pid unless DBIx::Class::_ENV_::BROKEN_FORK;
893 my $dbh = $self->_dbh
896 return $dbh->FETCH('Active');
902 my $dbh = $self->_dbh or return 0;
907 sub ensure_connected {
910 unless ($self->connected) {
911 $self->_populate_dbh;
917 Returns a C<$dbh> - a data base handle of class L<DBI>. The returned handle
918 is guaranteed to be healthy by implicitly calling L</connected>, and if
919 necessary performing a reconnection before returning. Keep in mind that this
920 is very B<expensive> on some database engines. Consider using L</dbh_do>
928 if (not $self->_dbh) {
929 $self->_populate_dbh;
931 $self->ensure_connected;
936 # this is the internal "get dbh or connect (don't check)" method
939 $self->_verify_pid unless DBIx::Class::_ENV_::BROKEN_FORK;
940 $self->_populate_dbh unless $self->_dbh;
946 unless ($self->_sql_maker) {
947 my $sql_maker_class = $self->sql_maker_class;
949 my %opts = %{$self->_sql_maker_opts||{}};
953 $self->sql_limit_dialect
956 my $s_class = (ref $self) || $self;
958 "Your storage class ($s_class) does not set sql_limit_dialect and you "
959 . 'have not supplied an explicit limit_dialect in your connection_info. '
960 . 'DBIC will attempt to use the GenericSubQ dialect, which works on most '
961 . 'databases but can be (and often is) painfully slow. '
962 . "Please file an RT ticket against '$s_class' ."
969 my ($quote_char, $name_sep);
971 if ($opts{quote_names}) {
972 $quote_char = (delete $opts{quote_char}) || $self->sql_quote_char || do {
973 my $s_class = (ref $self) || $self;
975 "You requested 'quote_names' but your storage class ($s_class) does "
976 . 'not explicitly define a default sql_quote_char and you have not '
977 . 'supplied a quote_char as part of your connection_info. DBIC will '
978 .q{default to the ANSI SQL standard quote '"', which works most of }
979 . "the time. Please file an RT ticket against '$s_class'."
985 $name_sep = (delete $opts{name_sep}) || $self->sql_name_sep;
988 $self->_sql_maker($sql_maker_class->new(
990 array_datatypes => 1,
991 limit_dialect => $dialect,
992 ($quote_char ? (quote_char => $quote_char) : ()),
993 name_sep => ($name_sep || '.'),
997 return $self->_sql_maker;
1000 # nothing to do by default
1007 my @info = @{$self->_dbi_connect_info || []};
1008 $self->_dbh(undef); # in case ->connected failed we might get sent here
1009 $self->_dbh_details({}); # reset everything we know
1011 $self->_dbh($self->_connect(@info));
1013 $self->_conn_pid($$) unless DBIx::Class::_ENV_::BROKEN_FORK; # on win32 these are in fact threads
1015 $self->_determine_driver;
1017 # Always set the transaction depth on connect, since
1018 # there is no transaction in progress by definition
1019 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
1021 $self->_run_connection_actions unless $self->{_in_determine_driver};
1024 sub _run_connection_actions {
1028 push @actions, ( $self->on_connect_call || () );
1029 push @actions, $self->_parse_connect_do ('on_connect_do');
1031 $self->_do_connection_actions(connect_call_ => $_) for @actions;
1036 sub set_use_dbms_capability {
1037 $_[0]->set_inherited ($_[1], $_[2]);
1040 sub get_use_dbms_capability {
1041 my ($self, $capname) = @_;
1043 my $use = $self->get_inherited ($capname);
1046 : do { $capname =~ s/^_use_/_supports_/; $self->get_dbms_capability ($capname) }
1050 sub set_dbms_capability {
1051 $_[0]->_dbh_details->{capability}{$_[1]} = $_[2];
1054 sub get_dbms_capability {
1055 my ($self, $capname) = @_;
1057 my $cap = $self->_dbh_details->{capability}{$capname};
1059 unless (defined $cap) {
1060 if (my $meth = $self->can ("_determine$capname")) {
1061 $cap = $self->$meth ? 1 : 0;
1067 $self->set_dbms_capability ($capname, $cap);
1077 unless ($info = $self->_dbh_details->{info}) {
1083 $server_version = $self->_get_server_version;
1086 if ($self->{_in_determine_driver}) {
1087 $self->throw_exception($_);
1089 $server_version = undef;
1092 if (defined $server_version) {
1093 $info->{dbms_version} = $server_version;
1095 my ($numeric_version) = $server_version =~ /^([\d\.]+)/;
1096 my @verparts = split (/\./, $numeric_version);
1102 # consider only up to 3 version parts, iff not more than 3 digits
1104 while (@verparts && @use_parts < 3) {
1105 my $p = shift @verparts;
1107 push @use_parts, $p;
1109 push @use_parts, 0 while @use_parts < 3;
1111 $info->{normalized_dbms_version} = sprintf "%d.%03d%03d", @use_parts;
1115 $self->_dbh_details->{info} = $info;
1121 sub _get_server_version {
1122 shift->_dbh_get_info('SQL_DBMS_VER');
1126 my ($self, $info) = @_;
1128 if ($info =~ /[^0-9]/) {
1129 $info = $DBI::Const::GetInfoType::GetInfoType{$info};
1130 $self->throw_exception("Info type '$_[1]' not provided by DBI::Const::GetInfoType")
1131 unless defined $info;
1137 $res = $self->_get_dbh->get_info($info);
1140 if ($self->{_in_determine_driver}) {
1141 $self->throw_exception($_);
1149 sub _describe_connection {
1150 require DBI::Const::GetInfoReturn;
1153 $self->ensure_connected;
1156 DBIC_DSN => $self->_dbi_connect_info->[0],
1157 DBI_VER => DBI->VERSION,
1158 DBIC_VER => DBIx::Class->VERSION,
1159 DBIC_DRIVER => ref $self,
1163 #keys %DBI::Const::GetInfoType::GetInfoType,
1165 SQL_CURSOR_COMMIT_BEHAVIOR
1166 SQL_CURSOR_ROLLBACK_BEHAVIOR
1167 SQL_CURSOR_SENSITIVITY
1168 SQL_DATA_SOURCE_NAME
1171 SQL_DEFAULT_TXN_ISOLATION
1176 SQL_EXPRESSIONS_IN_ORDERBY
1179 SQL_IDENTIFIER_QUOTE_CHAR
1180 SQL_MAX_CATALOG_NAME_LEN
1181 SQL_MAX_COLUMN_NAME_LEN
1182 SQL_MAX_IDENTIFIER_LEN
1183 SQL_MAX_TABLE_NAME_LEN
1184 SQL_MULTIPLE_ACTIVE_TXN
1185 SQL_MULT_RESULT_SETS
1186 SQL_NEED_LONG_DATA_LEN
1187 SQL_NON_NULLABLE_COLUMNS
1189 SQL_QUALIFIER_NAME_SEPARATOR
1190 SQL_QUOTED_IDENTIFIER_CASE
1192 SQL_TXN_ISOLATION_OPTION
1195 my $v = $self->_dbh_get_info($inf);
1196 next unless defined $v;
1198 #my $key = sprintf( '%s(%s)', $inf, $DBI::Const::GetInfoType::GetInfoType{$inf} );
1199 my $expl = DBI::Const::GetInfoReturn::Explain($inf, $v);
1200 $res->{$inf} = DBI::Const::GetInfoReturn::Format($inf, $v) . ( $expl ? " ($expl)" : '' );
1206 sub _determine_driver {
1209 if ((not $self->_driver_determined) && (not $self->{_in_determine_driver})) {
1210 my $started_connected = 0;
1211 local $self->{_in_determine_driver} = 1;
1213 if (ref($self) eq __PACKAGE__) {
1215 if ($self->_dbh) { # we are connected
1216 $driver = $self->_dbh->{Driver}{Name};
1217 $started_connected = 1;
1220 # if connect_info is a CODEREF, we have no choice but to connect
1221 if (ref $self->_dbi_connect_info->[0] &&
1222 reftype $self->_dbi_connect_info->[0] eq 'CODE') {
1223 $self->_populate_dbh;
1224 $driver = $self->_dbh->{Driver}{Name};
1227 # try to use dsn to not require being connected, the driver may still
1228 # force a connection in _rebless to determine version
1229 # (dsn may not be supplied at all if all we do is make a mock-schema)
1230 my $dsn = $self->_dbi_connect_info->[0] || $ENV{DBI_DSN} || '';
1231 ($driver) = $dsn =~ /dbi:([^:]+):/i;
1232 $driver ||= $ENV{DBI_DRIVER};
1237 my $storage_class = "DBIx::Class::Storage::DBI::${driver}";
1238 if ($self->load_optional_class($storage_class)) {
1239 mro::set_mro($storage_class, 'c3');
1240 bless $self, $storage_class;
1244 $self->_warn_undetermined_driver(
1245 'This version of DBIC does not yet seem to supply a driver for '
1246 . "your particular RDBMS and/or connection method ('$driver')."
1251 $self->_warn_undetermined_driver(
1252 'Unable to extract a driver name from connect info - this '
1253 . 'should not have happened.'
1258 $self->_driver_determined(1);
1260 Class::C3->reinitialize() if DBIx::Class::_ENV_::OLD_MRO;
1262 $self->_init; # run driver-specific initializations
1264 $self->_run_connection_actions
1265 if !$started_connected && defined $self->_dbh;
1269 sub _determine_connector_driver {
1270 my ($self, $conn) = @_;
1272 my $dbtype = $self->_dbh_get_info('SQL_DBMS_NAME');
1275 $self->_warn_undetermined_driver(
1276 'Unable to retrieve RDBMS type (SQL_DBMS_NAME) of the engine behind your '
1277 . "$conn connector - this should not have happened."
1282 $dbtype =~ s/\W/_/gi;
1284 my $subclass = "DBIx::Class::Storage::DBI::${conn}::${dbtype}";
1285 return if $self->isa($subclass);
1287 if ($self->load_optional_class($subclass)) {
1288 bless $self, $subclass;
1292 $self->_warn_undetermined_driver(
1293 'This version of DBIC does not yet seem to supply a driver for '
1294 . "your particular RDBMS and/or connection method ('$conn/$dbtype')."
1299 sub _warn_undetermined_driver {
1300 my ($self, $msg) = @_;
1302 require Data::Dumper::Concise;
1304 carp_once ($msg . ' While we will attempt to continue anyway, the results '
1305 . 'are likely to be underwhelming. Please upgrade DBIC, and if this message '
1306 . "does not go away, file a bugreport including the following info:\n"
1307 . Data::Dumper::Concise::Dumper($self->_describe_connection)
1311 sub _do_connection_actions {
1313 my $method_prefix = shift;
1316 if (not ref($call)) {
1317 my $method = $method_prefix . $call;
1319 } elsif (ref($call) eq 'CODE') {
1321 } elsif (ref($call) eq 'ARRAY') {
1322 if (ref($call->[0]) ne 'ARRAY') {
1323 $self->_do_connection_actions($method_prefix, $_) for @$call;
1325 $self->_do_connection_actions($method_prefix, @$_) for @$call;
1328 $self->throw_exception (sprintf ("Don't know how to process conection actions of type '%s'", ref($call)) );
1334 sub connect_call_do_sql {
1336 $self->_do_query(@_);
1339 sub disconnect_call_do_sql {
1341 $self->_do_query(@_);
1344 # override in db-specific backend when necessary
1345 sub connect_call_datetime_setup { 1 }
1348 my ($self, $action) = @_;
1350 if (ref $action eq 'CODE') {
1351 $action = $action->($self);
1352 $self->_do_query($_) foreach @$action;
1355 # Most debuggers expect ($sql, @bind), so we need to exclude
1356 # the attribute hash which is the second argument to $dbh->do
1357 # furthermore the bind values are usually to be presented
1358 # as named arrayref pairs, so wrap those here too
1359 my @do_args = (ref $action eq 'ARRAY') ? (@$action) : ($action);
1360 my $sql = shift @do_args;
1361 my $attrs = shift @do_args;
1362 my @bind = map { [ undef, $_ ] } @do_args;
1365 $_[0]->_query_start($sql, \@bind);
1366 $_[1]->do($sql, $attrs, @do_args);
1367 $_[0]->_query_end($sql, \@bind);
1375 my ($self, @info) = @_;
1377 $self->throw_exception("You failed to provide any connection info")
1380 my ($old_connect_via, $dbh);
1382 local $DBI::connect_via = 'connect' if $INC{'Apache/DBI.pm'} && $ENV{MOD_PERL};
1385 if(ref $info[0] eq 'CODE') {
1386 $dbh = $info[0]->();
1390 $dbh = DBI->connect(@info);
1397 unless ($self->unsafe) {
1399 $self->throw_exception(
1400 'Refusing clobbering of {HandleError} installed on externally supplied '
1401 ."DBI handle $dbh. Either remove the handler or use the 'unsafe' attribute."
1402 ) if $dbh->{HandleError} and ref $dbh->{HandleError} ne '__DBIC__DBH__ERROR__HANDLER__';
1404 # Default via _default_dbi_connect_attributes is 1, hence it was an explicit
1405 # request, or an external handle. Complain and set anyway
1406 unless ($dbh->{RaiseError}) {
1407 carp( ref $info[0] eq 'CODE'
1409 ? "The 'RaiseError' of the externally supplied DBI handle is set to false. "
1410 ."DBIx::Class will toggle it back to true, unless the 'unsafe' connect "
1411 .'attribute has been supplied'
1413 : 'RaiseError => 0 supplied in your connection_info, without an explicit '
1414 .'unsafe => 1. Toggling RaiseError back to true'
1417 $dbh->{RaiseError} = 1;
1420 # this odd anonymous coderef dereference is in fact really
1421 # necessary to avoid the unwanted effect described in perl5
1424 my $weak_self = $_[0];
1427 # the coderef is blessed so we can distinguish it from externally
1428 # supplied handles (which must be preserved)
1429 $_[1]->{HandleError} = bless sub {
1431 $weak_self->throw_exception("DBI Exception: $_[0]");
1434 # the handler may be invoked by something totally out of
1436 DBIx::Class::Exception->throw("DBI Exception (unhandled by DBIC, ::Schema GCed): $_[0]");
1438 }, '__DBIC__DBH__ERROR__HANDLER__';
1443 $self->throw_exception("DBI Connection failed: $_")
1446 $self->_dbh_autocommit($dbh->{AutoCommit});
1453 # this means we have not yet connected and do not know the AC status
1454 # (e.g. coderef $dbh), need a full-fledged connection check
1455 if (! defined $self->_dbh_autocommit) {
1456 $self->ensure_connected;
1458 # Otherwise simply connect or re-connect on pid changes
1463 $self->next::method(@_);
1466 sub _exec_txn_begin {
1469 # if the user is utilizing txn_do - good for him, otherwise we need to
1470 # ensure that the $dbh is healthy on BEGIN.
1471 # We do this via ->dbh_do instead of ->dbh, so that the ->dbh "ping"
1472 # will be replaced by a failure of begin_work itself (which will be
1473 # then retried on reconnect)
1474 if ($self->{_in_do_block}) {
1475 $self->_dbh->begin_work;
1477 $self->dbh_do(sub { $_[1]->begin_work });
1484 $self->_verify_pid unless DBIx::Class::_ENV_::BROKEN_FORK;
1485 $self->throw_exception("Unable to txn_commit() on a disconnected storage")
1488 # esoteric case for folks using external $dbh handles
1489 if (! $self->transaction_depth and ! $self->_dbh->FETCH('AutoCommit') ) {
1490 carp "Storage transaction_depth 0 does not match "
1491 ."false AutoCommit of $self->{_dbh}, attempting COMMIT anyway";
1492 $self->transaction_depth(1);
1495 $self->next::method(@_);
1497 # if AutoCommit is disabled txn_depth never goes to 0
1498 # as a new txn is started immediately on commit
1499 $self->transaction_depth(1) if (
1500 !$self->transaction_depth
1502 defined $self->_dbh_autocommit
1504 ! $self->_dbh_autocommit
1508 sub _exec_txn_commit {
1509 shift->_dbh->commit;
1515 $self->_verify_pid unless DBIx::Class::_ENV_::BROKEN_FORK;
1516 $self->throw_exception("Unable to txn_rollback() on a disconnected storage")
1519 # esoteric case for folks using external $dbh handles
1520 if (! $self->transaction_depth and ! $self->_dbh->FETCH('AutoCommit') ) {
1521 carp "Storage transaction_depth 0 does not match "
1522 ."false AutoCommit of $self->{_dbh}, attempting ROLLBACK anyway";
1523 $self->transaction_depth(1);
1526 $self->next::method(@_);
1528 # if AutoCommit is disabled txn_depth never goes to 0
1529 # as a new txn is started immediately on commit
1530 $self->transaction_depth(1) if (
1531 !$self->transaction_depth
1533 defined $self->_dbh_autocommit
1535 ! $self->_dbh_autocommit
1539 sub _exec_txn_rollback {
1540 shift->_dbh->rollback;
1543 # generate some identical methods
1544 for my $meth (qw/svp_begin svp_release svp_rollback/) {
1546 *{__PACKAGE__ ."::$meth"} = subname $meth => sub {
1548 $self->_verify_pid unless DBIx::Class::_ENV_::BROKEN_FORK;
1549 $self->throw_exception("Unable to $meth() on a disconnected storage")
1551 $self->next::method(@_);
1555 # This used to be the top-half of _execute. It was split out to make it
1556 # easier to override in NoBindVars without duping the rest. It takes up
1557 # all of _execute's args, and emits $sql, @bind.
1558 sub _prep_for_execute {
1559 #my ($self, $op, $ident, $args) = @_;
1560 return shift->_gen_sql_bind(@_)
1564 my ($self, $op, $ident, $args) = @_;
1566 my ($sql, @bind) = $self->sql_maker->$op(
1567 blessed($ident) ? $ident->from : $ident,
1572 ! $ENV{DBIC_DT_SEARCH_OK}
1576 first { blessed($_->[1]) && $_->[1]->isa('DateTime') } @bind
1578 carp_unique 'DateTime objects passed to search() are not supported '
1579 . 'properly (InflateColumn::DateTime formats and settings are not '
1580 . 'respected.) See "Formatting DateTime objects in queries" in '
1581 . 'DBIx::Class::Manual::Cookbook. To disable this warning for good '
1582 . 'set $ENV{DBIC_DT_SEARCH_OK} to true'
1585 return( $sql, $self->_resolve_bindattrs(
1586 $ident, [ @{$args->[2]{bind}||[]}, @bind ]
1590 sub _resolve_bindattrs {
1591 my ($self, $ident, $bind, $colinfos) = @_;
1595 my $resolve_bindinfo = sub {
1596 #my $infohash = shift;
1598 %$colinfos = %{ $self->_resolve_column_info($ident) }
1599 unless keys %$colinfos;
1602 if (my $col = $_[0]->{dbic_colname}) {
1603 $ret = { %{$_[0]} };
1605 $ret->{sqlt_datatype} ||= $colinfos->{$col}{data_type}
1606 if $colinfos->{$col}{data_type};
1608 $ret->{sqlt_size} ||= $colinfos->{$col}{size}
1609 if $colinfos->{$col}{size};
1616 if (ref $_ ne 'ARRAY') {
1619 elsif (! defined $_->[0]) {
1622 elsif (ref $_->[0] eq 'HASH') {
1624 ($_->[0]{dbd_attrs} or $_->[0]{sqlt_datatype}) ? $_->[0] : $resolve_bindinfo->($_->[0]),
1628 elsif (ref $_->[0] eq 'SCALAR') {
1629 [ { sqlt_datatype => ${$_->[0]} }, $_->[1] ]
1632 [ $resolve_bindinfo->({ dbic_colname => $_->[0] }), $_->[1] ]
1637 sub _format_for_trace {
1638 #my ($self, $bind) = @_;
1640 ### Turn @bind from something like this:
1641 ### ( [ "artist", 1 ], [ \%attrs, 3 ] )
1643 ### ( "'1'", "'3'" )
1646 defined( $_ && $_->[1] )
1653 my ( $self, $sql, $bind ) = @_;
1655 $self->debugobj->query_start( $sql, $self->_format_for_trace($bind) )
1660 my ( $self, $sql, $bind ) = @_;
1662 $self->debugobj->query_end( $sql, $self->_format_for_trace($bind) )
1667 sub _dbi_attrs_for_bind {
1668 my ($self, $ident, $bind) = @_;
1670 if (! defined $sba_compat) {
1671 $self->_determine_driver;
1672 $sba_compat = $self->can('source_bind_attributes') == \&source_bind_attributes
1680 my $class = ref $self;
1682 "The source_bind_attributes() override in $class relies on a deprecated codepath. "
1683 .'You are strongly advised to switch your code to override bind_attribute_by_datatype() '
1684 .'instead. This legacy compat shim will also disappear some time before DBIC 0.09'
1687 my $sba_attrs = $self->source_bind_attributes
1692 for (map { $_->[0] } @$bind) {
1694 if (exists $_->{dbd_attrs}) {
1697 elsif($_->{sqlt_datatype}) {
1698 # cache the result in the dbh_details hash, as it can not change unless
1699 # we connect to something else
1700 my $cache = $self->_dbh_details->{_datatype_map_cache} ||= {};
1701 if (not exists $cache->{$_->{sqlt_datatype}}) {
1702 $cache->{$_->{sqlt_datatype}} = $self->bind_attribute_by_data_type($_->{sqlt_datatype}) || undef;
1704 $cache->{$_->{sqlt_datatype}};
1706 elsif ($sba_attrs and $_->{dbic_colname}) {
1707 $sba_attrs->{$_->{dbic_colname}} || undef;
1710 undef; # always push something at this position
1719 my ($self, $op, $ident, @args) = @_;
1721 my ($sql, $bind) = $self->_prep_for_execute($op, $ident, \@args);
1723 shift->dbh_do( # retry over disconnects
1727 $self->_dbi_attrs_for_bind($ident, $bind)
1732 my ($self, undef, $sql, $bind, $bind_attrs) = @_;
1734 $self->_query_start( $sql, $bind );
1735 my $sth = $self->_sth($sql);
1737 for my $i (0 .. $#$bind) {
1738 if (ref $bind->[$i][1] eq 'SCALAR') { # any scalarrefs are assumed to be bind_inouts
1739 $sth->bind_param_inout(
1740 $i + 1, # bind params counts are 1-based
1742 $bind->[$i][0]{dbd_size} || $self->_max_column_bytesize($bind->[$i][0]), # size
1749 (ref $bind->[$i][1] and overload::Method($bind->[$i][1], '""'))
1758 # Can this fail without throwing an exception anyways???
1759 my $rv = $sth->execute();
1760 $self->throw_exception(
1761 $sth->errstr || $sth->err || 'Unknown error: execute() returned false, but error flags were not set...'
1764 $self->_query_end( $sql, $bind );
1766 return (wantarray ? ($rv, $sth, @$bind) : $rv);
1769 sub _prefetch_autovalues {
1770 my ($self, $source, $to_insert) = @_;
1772 my $colinfo = $source->columns_info;
1775 for my $col (keys %$colinfo) {
1777 $colinfo->{$col}{auto_nextval}
1780 ! exists $to_insert->{$col}
1782 ref $to_insert->{$col} eq 'SCALAR'
1784 (ref $to_insert->{$col} eq 'REF' and ref ${$to_insert->{$col}} eq 'ARRAY')
1787 $values{$col} = $self->_sequence_fetch(
1789 ( $colinfo->{$col}{sequence} ||=
1790 $self->_dbh_get_autoinc_seq($self->_get_dbh, $source, $col)
1800 my ($self, $source, $to_insert) = @_;
1802 my $prefetched_values = $self->_prefetch_autovalues($source, $to_insert);
1804 # fuse the values, but keep a separate list of prefetched_values so that
1805 # they can be fused once again with the final return
1806 $to_insert = { %$to_insert, %$prefetched_values };
1808 # FIXME - we seem to assume undef values as non-supplied. This is wrong.
1809 # Investigate what does it take to s/defined/exists/
1810 my $col_infos = $source->columns_info;
1811 my %pcols = map { $_ => 1 } $source->primary_columns;
1812 my (%retrieve_cols, $autoinc_supplied, $retrieve_autoinc_col);
1813 for my $col ($source->columns) {
1814 if ($col_infos->{$col}{is_auto_increment}) {
1815 $autoinc_supplied ||= 1 if defined $to_insert->{$col};
1816 $retrieve_autoinc_col ||= $col unless $autoinc_supplied;
1819 # nothing to retrieve when explicit values are supplied
1820 next if (defined $to_insert->{$col} and ! (
1821 ref $to_insert->{$col} eq 'SCALAR'
1823 (ref $to_insert->{$col} eq 'REF' and ref ${$to_insert->{$col}} eq 'ARRAY')
1826 # the 'scalar keys' is a trick to preserve the ->columns declaration order
1827 $retrieve_cols{$col} = scalar keys %retrieve_cols if (
1830 $col_infos->{$col}{retrieve_on_insert}
1834 local $self->{_autoinc_supplied_for_op} = $autoinc_supplied;
1835 local $self->{_perform_autoinc_retrieval} = $retrieve_autoinc_col;
1837 my ($sqla_opts, @ir_container);
1838 if (%retrieve_cols and $self->_use_insert_returning) {
1839 $sqla_opts->{returning_container} = \@ir_container
1840 if $self->_use_insert_returning_bound;
1842 $sqla_opts->{returning} = [
1843 sort { $retrieve_cols{$a} <=> $retrieve_cols{$b} } keys %retrieve_cols
1847 my ($rv, $sth) = $self->_execute('insert', $source, $to_insert, $sqla_opts);
1849 my %returned_cols = %$to_insert;
1850 if (my $retlist = $sqla_opts->{returning}) { # if IR is supported - we will get everything in one set
1851 @ir_container = try {
1852 local $SIG{__WARN__} = sub {};
1853 my @r = $sth->fetchrow_array;
1856 } unless @ir_container;
1858 @returned_cols{@$retlist} = @ir_container if @ir_container;
1861 # pull in PK if needed and then everything else
1862 if (my @missing_pri = grep { $pcols{$_} } keys %retrieve_cols) {
1864 $self->throw_exception( "Missing primary key but Storage doesn't support last_insert_id" )
1865 unless $self->can('last_insert_id');
1867 my @pri_values = $self->last_insert_id($source, @missing_pri);
1869 $self->throw_exception( "Can't get last insert id" )
1870 unless (@pri_values == @missing_pri);
1872 @returned_cols{@missing_pri} = @pri_values;
1873 delete $retrieve_cols{$_} for @missing_pri;
1876 # if there is more left to pull
1877 if (%retrieve_cols) {
1878 $self->throw_exception(
1879 'Unable to retrieve additional columns without a Primary Key on ' . $source->source_name
1882 my @left_to_fetch = sort { $retrieve_cols{$a} <=> $retrieve_cols{$b} } keys %retrieve_cols;
1884 my $cur = DBIx::Class::ResultSet->new($source, {
1885 where => { map { $_ => $returned_cols{$_} } (keys %pcols) },
1886 select => \@left_to_fetch,
1889 @returned_cols{@left_to_fetch} = $cur->next;
1891 $self->throw_exception('Duplicate row returned for PK-search after fresh insert')
1892 if scalar $cur->next;
1896 return { %$prefetched_values, %returned_cols };
1900 my ($self, $source, $cols, $data) = @_;
1902 my @col_range = (0..$#$cols);
1904 # FIXME - perhaps this is not even needed? does DBI stringify?
1906 # forcibly stringify whatever is stringifiable
1907 # ResultSet::populate() hands us a copy - safe to mangle
1908 for my $r (0 .. $#$data) {
1909 for my $c (0 .. $#{$data->[$r]}) {
1910 $data->[$r][$c] = "$data->[$r][$c]"
1911 if ( ref $data->[$r][$c] and overload::Method($data->[$r][$c], '""') );
1915 my $colinfos = $source->columns_info($cols);
1917 local $self->{_autoinc_supplied_for_op} =
1918 (first { $_->{is_auto_increment} } values %$colinfos)
1923 # get a slice type index based on first row of data
1924 # a "column" in this context may refer to more than one bind value
1925 # e.g. \[ '?, ?', [...], [...] ]
1927 # construct the value type index - a description of values types for every
1928 # per-column slice of $data:
1930 # nonexistent - nonbind literal
1932 # [] of bindattrs - resolved attribute(s) of bind(s) passed via literal+bind \[] combo
1934 # also construct the column hash to pass to the SQL generator. For plain
1935 # (non literal) values - convert the members of the first row into a
1936 # literal+bind combo, with extra positional info in the bind attr hashref.
1937 # This will allow us to match the order properly, and is so contrived
1938 # because a user-supplied literal/bind (or something else specific to a
1939 # resultsource and/or storage driver) can inject extra binds along the
1940 # way, so one can't rely on "shift positions" ordering at all. Also we
1941 # can't just hand SQLA a set of some known "values" (e.g. hashrefs that
1942 # can be later matched up by address), because we want to supply a real
1943 # value on which perhaps e.g. datatype checks will be performed
1944 my ($proto_data, $value_type_idx);
1945 for my $i (@col_range) {
1946 my $colname = $cols->[$i];
1947 if (ref $data->[0][$i] eq 'SCALAR') {
1948 # no bind value at all - no type
1950 $proto_data->{$colname} = $data->[0][$i];
1952 elsif (ref $data->[0][$i] eq 'REF' and ref ${$data->[0][$i]} eq 'ARRAY' ) {
1953 # repack, so we don't end up mangling the original \[]
1954 my ($sql, @bind) = @${$data->[0][$i]};
1956 # normalization of user supplied stuff
1957 my $resolved_bind = $self->_resolve_bindattrs(
1958 $source, \@bind, $colinfos,
1961 # store value-less (attrs only) bind info - we will be comparing all
1962 # supplied binds against this for sanity
1963 $value_type_idx->{$i} = [ map { $_->[0] } @$resolved_bind ];
1965 $proto_data->{$colname} = \[ $sql, map { [
1966 # inject slice order to use for $proto_bind construction
1967 { %{$resolved_bind->[$_][0]}, _bind_data_slice_idx => $i }
1969 $resolved_bind->[$_][1]
1974 $value_type_idx->{$i} = 0;
1976 $proto_data->{$colname} = \[ '?', [
1977 { dbic_colname => $colname, _bind_data_slice_idx => $i }
1984 my ($sql, $proto_bind) = $self->_prep_for_execute (
1990 if (! @$proto_bind and keys %$value_type_idx) {
1991 # if the bindlist is empty and we had some dynamic binds, this means the
1992 # storage ate them away (e.g. the NoBindVars component) and interpolated
1993 # them directly into the SQL. This obviously can't be good for multi-inserts
1994 $self->throw_exception('Cannot insert_bulk without support for placeholders');
1998 # FIXME - devise a flag "no babysitting" or somesuch to shut this off
2000 # use an error reporting closure for convenience (less to pass)
2001 my $bad_slice_report_cref = sub {
2002 my ($msg, $r_idx, $c_idx) = @_;
2003 $self->throw_exception(sprintf "%s for column '%s' in populate slice:\n%s",
2007 require Data::Dumper::Concise;
2008 local $Data::Dumper::Maxdepth = 5;
2009 Data::Dumper::Concise::Dumper ({
2010 map { $cols->[$_] =>
2018 for my $col_idx (@col_range) {
2019 my $reference_val = $data->[0][$col_idx];
2021 for my $row_idx (1..$#$data) { # we are comparing against what we got from [0] above, hence start from 1
2022 my $val = $data->[$row_idx][$col_idx];
2024 if (! exists $value_type_idx->{$col_idx}) { # literal no binds
2025 if (ref $val ne 'SCALAR') {
2026 $bad_slice_report_cref->(
2027 "Incorrect value (expecting SCALAR-ref \\'$$reference_val')",
2032 elsif ($$val ne $$reference_val) {
2033 $bad_slice_report_cref->(
2034 "Inconsistent literal SQL value (expecting \\'$$reference_val')",
2040 elsif (! $value_type_idx->{$col_idx} ) { # regular non-literal value
2041 if (ref $val eq 'SCALAR' or (ref $val eq 'REF' and ref $$val eq 'ARRAY') ) {
2042 $bad_slice_report_cref->("Literal SQL found where a plain bind value is expected", $row_idx, $col_idx);
2045 else { # binds from a \[], compare type and attrs
2046 if (ref $val ne 'REF' or ref $$val ne 'ARRAY') {
2047 $bad_slice_report_cref->(
2048 "Incorrect value (expecting ARRAYREF-ref \\['${$reference_val}->[0]', ... ])",
2053 # start drilling down and bail out early on identical refs
2055 $reference_val != $val
2057 $$reference_val != $$val
2059 if (${$val}->[0] ne ${$reference_val}->[0]) {
2060 $bad_slice_report_cref->(
2061 "Inconsistent literal/bind SQL (expecting \\['${$reference_val}->[0]', ... ])",
2066 # need to check the bind attrs - a bind will happen only once for
2067 # the entire dataset, so any changes further down will be ignored.
2068 elsif (! Data::Compare::Compare(
2069 $value_type_idx->{$col_idx},
2073 @{$self->_resolve_bindattrs(
2074 $source, [ @{$$val}[1 .. $#$$val] ], $colinfos,
2078 $bad_slice_report_cref->(
2079 'Differing bind attributes on literal/bind values not supported',
2089 # neither _dbh_execute_for_fetch, nor _dbh_execute_inserts_with_no_binds
2090 # are atomic (even if execute_for_fetch is a single call). Thus a safety
2092 my $guard = $self->txn_scope_guard;
2094 $self->_query_start( $sql, @$proto_bind ? [[undef => '__BULK_INSERT__' ]] : () );
2095 my $sth = $self->_sth($sql);
2098 # proto bind contains the information on which pieces of $data to pull
2099 # $cols is passed in only for prettier error-reporting
2100 $self->_dbh_execute_for_fetch( $source, $sth, $proto_bind, $cols, $data );
2103 # bind_param_array doesn't work if there are no binds
2104 $self->_dbh_execute_inserts_with_no_binds( $sth, scalar @$data );
2108 $self->_query_end( $sql, @$proto_bind ? [[ undef => '__BULK_INSERT__' ]] : () );
2112 return wantarray ? ($rv, $sth, @$proto_bind) : $rv;
2115 # execute_for_fetch is capable of returning data just fine (it means it
2116 # can be used for INSERT...RETURNING and UPDATE...RETURNING. Since this
2117 # is the void-populate fast-path we will just ignore this altogether
2118 # for the time being.
2119 sub _dbh_execute_for_fetch {
2120 my ($self, $source, $sth, $proto_bind, $cols, $data) = @_;
2122 my @idx_range = ( 0 .. $#$proto_bind );
2124 # If we have any bind attributes to take care of, we will bind the
2125 # proto-bind data (which will never be used by execute_for_fetch)
2126 # However since column bindtypes are "sticky", this is sufficient
2127 # to get the DBD to apply the bindtype to all values later on
2129 my $bind_attrs = $self->_dbi_attrs_for_bind($source, $proto_bind);
2131 for my $i (@idx_range) {
2133 $i+1, # DBI bind indexes are 1-based
2134 $proto_bind->[$i][1],
2136 ) if defined $bind_attrs->[$i];
2139 # At this point $data slots named in the _bind_data_slice_idx of
2140 # each piece of $proto_bind are either \[]s or plain values to be
2141 # passed in. Construct the dispensing coderef. *NOTE* the order
2142 # of $data will differ from this of the ?s in the SQL (due to
2143 # alphabetical ordering by colname). We actually do want to
2144 # preserve this behavior so that prepare_cached has a better
2145 # chance of matching on unrelated calls
2146 my %data_reorder = map { $proto_bind->[$_][0]{_bind_data_slice_idx} => $_ } @idx_range;
2148 my $fetch_row_idx = -1; # saner loop this way
2149 my $fetch_tuple = sub {
2150 return undef if ++$fetch_row_idx > $#$data;
2153 { (ref $_ eq 'REF' and ref $$_ eq 'ARRAY')
2154 ? map { $_->[-1] } @{$$_}[1 .. $#$$_]
2158 { $data->[$fetch_row_idx][$_]}
2160 { $data_reorder{$a} <=> $data_reorder{$b} }
2165 my $tuple_status = [];
2168 $rv = $sth->execute_for_fetch(
2177 # Not all DBDs are create equal. Some throw on error, some return
2178 # an undef $rv, and some set $sth->err - try whatever we can
2179 $err = ($sth->errstr || 'UNKNOWN ERROR ($sth->errstr is unset)') if (
2182 ( !defined $rv or $sth->err )
2185 # Statement must finish even if there was an exception.
2190 $err = shift unless defined $err
2195 ++$i while $i <= $#$tuple_status && !ref $tuple_status->[$i];
2197 $self->throw_exception("Unexpected populate error: $err")
2198 if ($i > $#$tuple_status);
2200 require Data::Dumper::Concise;
2201 $self->throw_exception(sprintf "execute_for_fetch() aborted with '%s' at populate slice:\n%s",
2202 ($tuple_status->[$i][1] || $err),
2203 Data::Dumper::Concise::Dumper( { map { $cols->[$_] => $data->[$i][$_] } (0 .. $#$cols) } ),
2210 sub _dbh_execute_inserts_with_no_binds {
2211 my ($self, $sth, $count) = @_;
2215 my $dbh = $self->_get_dbh;
2216 local $dbh->{RaiseError} = 1;
2217 local $dbh->{PrintError} = 0;
2219 $sth->execute foreach 1..$count;
2225 # Make sure statement is finished even if there was an exception.
2230 $err = shift unless defined $err;
2233 $self->throw_exception($err) if defined $err;
2239 #my ($self, $source, @args) = @_;
2240 shift->_execute('update', @_);
2245 #my ($self, $source, @args) = @_;
2246 shift->_execute('delete', @_);
2251 $self->_execute($self->_select_args(@_));
2254 sub _select_args_to_query {
2257 $self->throw_exception(
2258 "Unable to generate limited query representation with 'software_limit' enabled"
2259 ) if ($_[3]->{software_limit} and ($_[3]->{offset} or $_[3]->{rows}) );
2261 # my ($op, $ident, $select, $cond, $rs_attrs, $rows, $offset)
2262 # = $self->_select_args($ident, $select, $cond, $attrs);
2263 my ($op, $ident, @args) =
2264 $self->_select_args(@_);
2266 # my ($sql, $prepared_bind) = $self->_gen_sql_bind($op, $ident, [ $select, $cond, $rs_attrs, $rows, $offset ]);
2267 my ($sql, $prepared_bind) = $self->_gen_sql_bind($op, $ident, \@args);
2268 $prepared_bind ||= [];
2271 ? ($sql, $prepared_bind)
2272 : \[ "($sql)", @$prepared_bind ]
2277 my ($self, $ident, $select, $where, $attrs) = @_;
2279 my $sql_maker = $self->sql_maker;
2280 my ($alias2source, $rs_alias) = $self->_resolve_ident_sources ($ident);
2287 $rs_alias && $alias2source->{$rs_alias}
2288 ? ( _rsroot_rsrc => $alias2source->{$rs_alias} )
2293 # Sanity check the attributes (SQLMaker does it too, but
2294 # in case of a software_limit we'll never reach there)
2295 if (defined $attrs->{offset}) {
2296 $self->throw_exception('A supplied offset attribute must be a non-negative integer')
2297 if ( $attrs->{offset} =~ /\D/ or $attrs->{offset} < 0 );
2300 if (defined $attrs->{rows}) {
2301 $self->throw_exception("The rows attribute must be a positive integer if present")
2302 if ( $attrs->{rows} =~ /\D/ or $attrs->{rows} <= 0 );
2304 elsif ($attrs->{offset}) {
2305 # MySQL actually recommends this approach. I cringe.
2306 $attrs->{rows} = $sql_maker->__max_int;
2311 # see if we need to tear the prefetch apart otherwise delegate the limiting to the
2312 # storage, unless software limit was requested
2315 ( $attrs->{rows} && keys %{$attrs->{collapse}} )
2317 # grouped prefetch (to satisfy group_by == select)
2318 ( $attrs->{group_by}
2320 @{$attrs->{group_by}}
2322 $attrs->{_prefetch_selector_range}
2325 ($ident, $select, $where, $attrs)
2326 = $self->_adjust_select_args_for_complex_prefetch ($ident, $select, $where, $attrs);
2328 elsif (! $attrs->{software_limit} ) {
2330 $attrs->{rows} || (),
2331 $attrs->{offset} || (),
2335 # try to simplify the joinmap further (prune unreferenced type-single joins)
2339 reftype $ident eq 'ARRAY'
2343 $ident = $self->_prune_unused_joins ($ident, $select, $where, $attrs);
2347 # This would be the point to deflate anything found in $where
2348 # (and leave $attrs->{bind} intact). Problem is - inflators historically
2349 # expect a row object. And all we have is a resultsource (it is trivial
2350 # to extract deflator coderefs via $alias2source above).
2352 # I don't see a way forward other than changing the way deflators are
2353 # invoked, and that's just bad...
2356 return ('select', $ident, $select, $where, $attrs, @limit);
2359 # Returns a counting SELECT for a simple count
2360 # query. Abstracted so that a storage could override
2361 # this to { count => 'firstcol' } or whatever makes
2362 # sense as a performance optimization
2364 #my ($self, $source, $rs_attrs) = @_;
2365 return { count => '*' };
2368 sub source_bind_attributes {
2369 shift->throw_exception(
2370 'source_bind_attributes() was never meant to be a callable public method - '
2371 .'please contact the DBIC dev-team and describe your use case so that a reasonable '
2372 .'solution can be provided'
2373 ."\nhttp://search.cpan.org/dist/DBIx-Class/lib/DBIx/Class.pm#GETTING_HELP/SUPPORT"
2381 =item Arguments: $ident, $select, $condition, $attrs
2385 Handle a SQL select statement.
2391 my ($ident, $select, $condition, $attrs) = @_;
2392 return $self->cursor_class->new($self, \@_, $attrs);
2397 my ($rv, $sth, @bind) = $self->_select(@_);
2398 my @row = $sth->fetchrow_array;
2399 my @nextrow = $sth->fetchrow_array if @row;
2400 if(@row && @nextrow) {
2401 carp "Query returned more than one row. SQL that returns multiple rows is DEPRECATED for ->find and ->single";
2403 # Need to call finish() to work round broken DBDs
2408 =head2 sql_limit_dialect
2410 This is an accessor for the default SQL limit dialect used by a particular
2411 storage driver. Can be overridden by supplying an explicit L</limit_dialect>
2412 to L<DBIx::Class::Schema/connect>. For a list of available limit dialects
2413 see L<DBIx::Class::SQLMaker::LimitDialects>.
2418 my ($self, $dbh, $sql) = @_;
2420 # 3 is the if_active parameter which avoids active sth re-use
2421 my $sth = $self->disable_sth_caching
2422 ? $dbh->prepare($sql)
2423 : $dbh->prepare_cached($sql, {}, 3);
2425 # XXX You would think RaiseError would make this impossible,
2426 # but apparently that's not true :(
2427 $self->throw_exception(
2430 sprintf( "\$dbh->prepare() of '%s' through %s failed *silently* without "
2431 .'an exception and/or setting $dbh->errstr',
2433 ? substr($sql, 0, 20) . '...'
2436 'DBD::' . $dbh->{Driver}{Name},
2444 carp_unique 'sth was mistakenly marked/documented as public, stop calling it (will be removed before DBIC v0.09)';
2449 my ($self, $sql) = @_;
2450 $self->dbh_do('_dbh_sth', $sql); # retry over disconnects
2453 sub _dbh_columns_info_for {
2454 my ($self, $dbh, $table) = @_;
2456 if ($dbh->can('column_info')) {
2460 my ($schema,$tab) = $table =~ /^(.+?)\.(.+)$/ ? ($1,$2) : (undef,$table);
2461 my $sth = $dbh->column_info( undef,$schema, $tab, '%' );
2463 while ( my $info = $sth->fetchrow_hashref() ){
2465 $column_info{data_type} = $info->{TYPE_NAME};
2466 $column_info{size} = $info->{COLUMN_SIZE};
2467 $column_info{is_nullable} = $info->{NULLABLE} ? 1 : 0;
2468 $column_info{default_value} = $info->{COLUMN_DEF};
2469 my $col_name = $info->{COLUMN_NAME};
2470 $col_name =~ s/^\"(.*)\"$/$1/;
2472 $result{$col_name} = \%column_info;
2477 return \%result if !$caught && scalar keys %result;
2481 my $sth = $dbh->prepare($self->sql_maker->select($table, undef, \'1 = 0'));
2483 my @columns = @{$sth->{NAME_lc}};
2484 for my $i ( 0 .. $#columns ){
2486 $column_info{data_type} = $sth->{TYPE}->[$i];
2487 $column_info{size} = $sth->{PRECISION}->[$i];
2488 $column_info{is_nullable} = $sth->{NULLABLE}->[$i] ? 1 : 0;
2490 if ($column_info{data_type} =~ m/^(.*?)\((.*?)\)$/) {
2491 $column_info{data_type} = $1;
2492 $column_info{size} = $2;
2495 $result{$columns[$i]} = \%column_info;
2499 foreach my $col (keys %result) {
2500 my $colinfo = $result{$col};
2501 my $type_num = $colinfo->{data_type};
2503 if(defined $type_num && $dbh->can('type_info')) {
2504 my $type_info = $dbh->type_info($type_num);
2505 $type_name = $type_info->{TYPE_NAME} if $type_info;
2506 $colinfo->{data_type} = $type_name if $type_name;
2513 sub columns_info_for {
2514 my ($self, $table) = @_;
2515 $self->_dbh_columns_info_for ($self->_get_dbh, $table);
2518 =head2 last_insert_id
2520 Return the row id of the last insert.
2524 sub _dbh_last_insert_id {
2525 my ($self, $dbh, $source, $col) = @_;
2527 my $id = try { $dbh->last_insert_id (undef, undef, $source->name, $col) };
2529 return $id if defined $id;
2531 my $class = ref $self;
2532 $self->throw_exception ("No storage specific _dbh_last_insert_id() method implemented in $class, and the generic DBI::last_insert_id() failed");
2535 sub last_insert_id {
2537 $self->_dbh_last_insert_id ($self->_dbh, @_);
2540 =head2 _native_data_type
2544 =item Arguments: $type_name
2548 This API is B<EXPERIMENTAL>, will almost definitely change in the future, and
2549 currently only used by L<::AutoCast|DBIx::Class::Storage::DBI::AutoCast> and
2550 L<::Sybase::ASE|DBIx::Class::Storage::DBI::Sybase::ASE>.
2552 The default implementation returns C<undef>, implement in your Storage driver if
2553 you need this functionality.
2555 Should map types from other databases to the native RDBMS type, for example
2556 C<VARCHAR2> to C<VARCHAR>.
2558 Types with modifiers should map to the underlying data type. For example,
2559 C<INTEGER AUTO_INCREMENT> should become C<INTEGER>.
2561 Composite types should map to the container type, for example
2562 C<ENUM(foo,bar,baz)> becomes C<ENUM>.
2566 sub _native_data_type {
2567 #my ($self, $data_type) = @_;
2571 # Check if placeholders are supported at all
2572 sub _determine_supports_placeholders {
2574 my $dbh = $self->_get_dbh;
2576 # some drivers provide a $dbh attribute (e.g. Sybase and $dbh->{syb_dynamic_supported})
2577 # but it is inaccurate more often than not
2579 local $dbh->{PrintError} = 0;
2580 local $dbh->{RaiseError} = 1;
2581 $dbh->do('select ?', {}, 1);
2589 # Check if placeholders bound to non-string types throw exceptions
2591 sub _determine_supports_typeless_placeholders {
2593 my $dbh = $self->_get_dbh;
2596 local $dbh->{PrintError} = 0;
2597 local $dbh->{RaiseError} = 1;
2598 # this specifically tests a bind that is NOT a string
2599 $dbh->do('select 1 where 1 = ?', {}, 1);
2609 Returns the database driver name.
2614 shift->_get_dbh->{Driver}->{Name};
2617 =head2 bind_attribute_by_data_type
2619 Given a datatype from column info, returns a database specific bind
2620 attribute for C<< $dbh->bind_param($val,$attribute) >> or nothing if we will
2621 let the database planner just handle it.
2623 Generally only needed for special case column types, like bytea in postgres.
2627 sub bind_attribute_by_data_type {
2631 =head2 is_datatype_numeric
2633 Given a datatype from column_info, returns a boolean value indicating if
2634 the current RDBMS considers it a numeric value. This controls how
2635 L<DBIx::Class::Row/set_column> decides whether to mark the column as
2636 dirty - when the datatype is deemed numeric a C<< != >> comparison will
2637 be performed instead of the usual C<eq>.
2641 sub is_datatype_numeric {
2642 #my ($self, $dt) = @_;
2644 return 0 unless $_[1];
2647 numeric | int(?:eger)? | (?:tiny|small|medium|big)int | dec(?:imal)? | real | float | double (?: \s+ precision)? | (?:big)?serial
2652 =head2 create_ddl_dir
2656 =item Arguments: $schema \@databases, $version, $directory, $preversion, \%sqlt_args
2660 Creates a SQL file based on the Schema, for each of the specified
2661 database engines in C<\@databases> in the given directory.
2662 (note: specify L<SQL::Translator> names, not L<DBI> driver names).
2664 Given a previous version number, this will also create a file containing
2665 the ALTER TABLE statements to transform the previous schema into the
2666 current one. Note that these statements may contain C<DROP TABLE> or
2667 C<DROP COLUMN> statements that can potentially destroy data.
2669 The file names are created using the C<ddl_filename> method below, please
2670 override this method in your schema if you would like a different file
2671 name format. For the ALTER file, the same format is used, replacing
2672 $version in the name with "$preversion-$version".
2674 See L<SQL::Translator/METHODS> for a list of values for C<\%sqlt_args>.
2675 The most common value for this would be C<< { add_drop_table => 1 } >>
2676 to have the SQL produced include a C<DROP TABLE> statement for each table
2677 created. For quoting purposes supply C<quote_table_names> and
2678 C<quote_field_names>.
2680 If no arguments are passed, then the following default values are assumed:
2684 =item databases - ['MySQL', 'SQLite', 'PostgreSQL']
2686 =item version - $schema->schema_version
2688 =item directory - './'
2690 =item preversion - <none>
2694 By default, C<\%sqlt_args> will have
2696 { add_drop_table => 1, ignore_constraint_names => 1, ignore_index_names => 1 }
2698 merged with the hash passed in. To disable any of those features, pass in a
2699 hashref like the following
2701 { ignore_constraint_names => 0, # ... other options }
2704 WARNING: You are strongly advised to check all SQL files created, before applying
2709 sub create_ddl_dir {
2710 my ($self, $schema, $databases, $version, $dir, $preversion, $sqltargs) = @_;
2713 carp "No directory given, using ./\n";
2718 (require File::Path and File::Path::make_path ("$dir")) # make_path does not like objects (i.e. Path::Class::Dir)
2720 $self->throw_exception(
2721 "Failed to create '$dir': " . ($! || $@ || 'error unknown')
2725 $self->throw_exception ("Directory '$dir' does not exist\n") unless(-d $dir);
2727 $databases ||= ['MySQL', 'SQLite', 'PostgreSQL'];
2728 $databases = [ $databases ] if(ref($databases) ne 'ARRAY');
2730 my $schema_version = $schema->schema_version || '1.x';
2731 $version ||= $schema_version;
2734 add_drop_table => 1,
2735 ignore_constraint_names => 1,
2736 ignore_index_names => 1,
2740 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy')) {
2741 $self->throw_exception("Can't create a ddl file without " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2744 my $sqlt = SQL::Translator->new( $sqltargs );
2746 $sqlt->parser('SQL::Translator::Parser::DBIx::Class');
2747 my $sqlt_schema = $sqlt->translate({ data => $schema })
2748 or $self->throw_exception ($sqlt->error);
2750 foreach my $db (@$databases) {
2752 $sqlt->{schema} = $sqlt_schema;
2753 $sqlt->producer($db);
2756 my $filename = $schema->ddl_filename($db, $version, $dir);
2757 if (-e $filename && ($version eq $schema_version )) {
2758 # if we are dumping the current version, overwrite the DDL
2759 carp "Overwriting existing DDL file - $filename";
2763 my $output = $sqlt->translate;
2765 carp("Failed to translate to $db, skipping. (" . $sqlt->error . ")");
2768 if(!open($file, ">$filename")) {
2769 $self->throw_exception("Can't open $filename for writing ($!)");
2772 print $file $output;
2775 next unless ($preversion);
2777 require SQL::Translator::Diff;
2779 my $prefilename = $schema->ddl_filename($db, $preversion, $dir);
2780 if(!-e $prefilename) {
2781 carp("No previous schema file found ($prefilename)");
2785 my $difffile = $schema->ddl_filename($db, $version, $dir, $preversion);
2787 carp("Overwriting existing diff file - $difffile");
2793 my $t = SQL::Translator->new($sqltargs);
2798 or $self->throw_exception ($t->error);
2800 my $out = $t->translate( $prefilename )
2801 or $self->throw_exception ($t->error);
2803 $source_schema = $t->schema;
2805 $source_schema->name( $prefilename )
2806 unless ( $source_schema->name );
2809 # The "new" style of producers have sane normalization and can support
2810 # diffing a SQL file against a DBIC->SQLT schema. Old style ones don't
2811 # And we have to diff parsed SQL against parsed SQL.
2812 my $dest_schema = $sqlt_schema;
2814 unless ( "SQL::Translator::Producer::$db"->can('preprocess_schema') ) {
2815 my $t = SQL::Translator->new($sqltargs);
2820 or $self->throw_exception ($t->error);
2822 my $out = $t->translate( $filename )
2823 or $self->throw_exception ($t->error);
2825 $dest_schema = $t->schema;
2827 $dest_schema->name( $filename )
2828 unless $dest_schema->name;
2831 my $diff = SQL::Translator::Diff::schema_diff($source_schema, $db,
2835 if(!open $file, ">$difffile") {
2836 $self->throw_exception("Can't write to $difffile ($!)");
2844 =head2 deployment_statements
2848 =item Arguments: $schema, $type, $version, $directory, $sqlt_args
2852 Returns the statements used by L</deploy> and L<DBIx::Class::Schema/deploy>.
2854 The L<SQL::Translator> (not L<DBI>) database driver name can be explicitly
2855 provided in C<$type>, otherwise the result of L</sqlt_type> is used as default.
2857 C<$directory> is used to return statements from files in a previously created
2858 L</create_ddl_dir> directory and is optional. The filenames are constructed
2859 from L<DBIx::Class::Schema/ddl_filename>, the schema name and the C<$version>.
2861 If no C<$directory> is specified then the statements are constructed on the
2862 fly using L<SQL::Translator> and C<$version> is ignored.
2864 See L<SQL::Translator/METHODS> for a list of values for C<$sqlt_args>.
2868 sub deployment_statements {
2869 my ($self, $schema, $type, $version, $dir, $sqltargs) = @_;
2870 $type ||= $self->sqlt_type;
2871 $version ||= $schema->schema_version || '1.x';
2873 my $filename = $schema->ddl_filename($type, $version, $dir);
2876 # FIXME replace this block when a proper sane sql parser is available
2878 open($file, "<$filename")
2879 or $self->throw_exception("Can't open $filename ($!)");
2882 return join('', @rows);
2885 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy') ) {
2886 $self->throw_exception("Can't deploy without a ddl_dir or " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2889 # sources needs to be a parser arg, but for simplicty allow at top level
2891 $sqltargs->{parser_args}{sources} = delete $sqltargs->{sources}
2892 if exists $sqltargs->{sources};
2894 my $tr = SQL::Translator->new(
2895 producer => "SQL::Translator::Producer::${type}",
2897 parser => 'SQL::Translator::Parser::DBIx::Class',
2901 return preserve_context {
2904 $self->throw_exception( 'Unable to produce deployment statements: ' . $tr->error)
2905 unless defined $_[0];
2909 # FIXME deploy() currently does not accurately report sql errors
2910 # Will always return true while errors are warned
2912 my ($self, $schema, $type, $sqltargs, $dir) = @_;
2916 return if($line =~ /^--/);
2917 # next if($line =~ /^DROP/m);
2918 return if($line =~ /^BEGIN TRANSACTION/m);
2919 return if($line =~ /^COMMIT/m);
2920 return if $line =~ /^\s+$/; # skip whitespace only
2921 $self->_query_start($line);
2923 # do a dbh_do cycle here, as we need some error checking in
2924 # place (even though we will ignore errors)
2925 $self->dbh_do (sub { $_[1]->do($line) });
2927 carp qq{$_ (running "${line}")};
2929 $self->_query_end($line);
2931 my @statements = $schema->deployment_statements($type, undef, $dir, { %{ $sqltargs || {} }, no_comments => 1 } );
2932 if (@statements > 1) {
2933 foreach my $statement (@statements) {
2934 $deploy->( $statement );
2937 elsif (@statements == 1) {
2938 # split on single line comments and end of statements
2939 foreach my $line ( split(/\s*--.*\n|;\n/, $statements[0])) {
2945 =head2 datetime_parser
2947 Returns the datetime parser class
2951 sub datetime_parser {
2953 return $self->{datetime_parser} ||= do {
2954 $self->build_datetime_parser(@_);
2958 =head2 datetime_parser_type
2960 Defines the datetime parser class - currently defaults to L<DateTime::Format::MySQL>
2962 =head2 build_datetime_parser
2964 See L</datetime_parser>
2968 sub build_datetime_parser {
2970 my $type = $self->datetime_parser_type(@_);
2975 =head2 is_replicating
2977 A boolean that reports if a particular L<DBIx::Class::Storage::DBI> is set to
2978 replicate from a master database. Default is undef, which is the result
2979 returned by databases that don't support replication.
2983 sub is_replicating {
2988 =head2 lag_behind_master
2990 Returns a number that represents a certain amount of lag behind a master db
2991 when a given storage is replicating. The number is database dependent, but
2992 starts at zero and increases with the amount of lag. Default in undef
2996 sub lag_behind_master {
3000 =head2 relname_to_table_alias
3004 =item Arguments: $relname, $join_count
3008 L<DBIx::Class> uses L<DBIx::Class::Relationship> names as table aliases in
3011 This hook is to allow specific L<DBIx::Class::Storage> drivers to change the
3012 way these aliases are named.
3014 The default behavior is C<< "$relname_$join_count" if $join_count > 1 >>,
3015 otherwise C<"$relname">.
3019 sub relname_to_table_alias {
3020 my ($self, $relname, $join_count) = @_;
3022 my $alias = ($join_count && $join_count > 1 ?
3023 join('_', $relname, $join_count) : $relname);
3028 # The size in bytes to use for DBI's ->bind_param_inout, this is the generic
3029 # version and it may be necessary to amend or override it for a specific storage
3030 # if such binds are necessary.
3031 sub _max_column_bytesize {
3032 my ($self, $attr) = @_;
3036 if ($attr->{sqlt_datatype}) {
3037 my $data_type = lc($attr->{sqlt_datatype});
3039 if ($attr->{sqlt_size}) {
3041 # String/sized-binary types
3042 if ($data_type =~ /^(?:
3043 l? (?:var)? char(?:acter)? (?:\s*varying)?
3045 (?:var)? binary (?:\s*varying)?
3050 $max_size = $attr->{sqlt_size};
3052 # Other charset/unicode types, assume scale of 4
3053 elsif ($data_type =~ /^(?:
3054 national \s* character (?:\s*varying)?
3063 $max_size = $attr->{sqlt_size} * 4;
3067 if (!$max_size and !$self->_is_lob_type($data_type)) {
3068 $max_size = 100 # for all other (numeric?) datatypes
3072 $max_size || $self->_dbic_connect_attributes->{LongReadLen} || $self->_get_dbh->{LongReadLen} || 8000;
3075 # Determine if a data_type is some type of BLOB
3077 my ($self, $data_type) = @_;
3078 $data_type && ($data_type =~ /lob|bfile|text|image|bytea|memo/i
3079 || $data_type =~ /^long(?:\s+(?:raw|bit\s*varying|varbit|binary
3080 |varchar|character\s*varying|nvarchar
3081 |national\s*character\s*varying))?\z/xi);
3084 sub _is_binary_lob_type {
3085 my ($self, $data_type) = @_;
3086 $data_type && ($data_type =~ /blob|bfile|image|bytea/i
3087 || $data_type =~ /^long(?:\s+(?:raw|bit\s*varying|varbit|binary))?\z/xi);
3090 sub _is_text_lob_type {
3091 my ($self, $data_type) = @_;
3092 $data_type && ($data_type =~ /^(?:clob|memo)\z/i
3093 || $data_type =~ /^long(?:\s+(?:varchar|character\s*varying|nvarchar
3094 |national\s*character\s*varying))\z/xi);
3097 # Determine if a data_type is some type of a binary type
3098 sub _is_binary_type {
3099 my ($self, $data_type) = @_;
3100 $data_type && ($self->_is_binary_lob_type($data_type)
3101 || $data_type =~ /(?:var)?(?:binary|bit|graphic)(?:\s*varying)?/i);
3108 =head2 DBIx::Class and AutoCommit
3110 DBIx::Class can do some wonderful magic with handling exceptions,
3111 disconnections, and transactions when you use C<< AutoCommit => 1 >>
3112 (the default) combined with L<txn_do|DBIx::Class::Storage/txn_do> for
3113 transaction support.
3115 If you set C<< AutoCommit => 0 >> in your connect info, then you are always
3116 in an assumed transaction between commits, and you're telling us you'd
3117 like to manage that manually. A lot of the magic protections offered by
3118 this module will go away. We can't protect you from exceptions due to database
3119 disconnects because we don't know anything about how to restart your
3120 transactions. You're on your own for handling all sorts of exceptional
3121 cases if you choose the C<< AutoCommit => 0 >> path, just as you would
3127 Matt S. Trout <mst@shadowcatsystems.co.uk>
3129 Andy Grundman <andy@hybridized.org>
3133 You may distribute this code under the same terms as Perl itself.