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';
19 # default cursor class, overridable in connect_info attributes
20 __PACKAGE__->cursor_class('DBIx::Class::Storage::DBI::Cursor');
22 __PACKAGE__->mk_group_accessors('inherited' => qw/
23 sql_limit_dialect sql_quote_char sql_name_sep
26 __PACKAGE__->mk_group_accessors('component_class' => qw/sql_maker_class datetime_parser_type/);
28 __PACKAGE__->sql_maker_class('DBIx::Class::SQLMaker');
29 __PACKAGE__->datetime_parser_type('DateTime::Format::MySQL'); # historic default
31 __PACKAGE__->sql_name_sep('.');
33 __PACKAGE__->mk_group_accessors('simple' => qw/
34 _connect_info _dbi_connect_info _dbic_connect_attributes _driver_determined
35 _dbh _dbh_details _conn_pid _sql_maker _sql_maker_opts _dbh_autocommit
38 # the values for these accessors are picked out (and deleted) from
39 # the attribute hashref passed to connect_info
40 my @storage_options = qw/
41 on_connect_call on_disconnect_call on_connect_do on_disconnect_do
42 disable_sth_caching unsafe auto_savepoint
44 __PACKAGE__->mk_group_accessors('simple' => @storage_options);
47 # capability definitions, using a 2-tiered accessor system
50 # A driver/user may define _use_X, which blindly without any checks says:
51 # "(do not) use this capability", (use_dbms_capability is an "inherited"
54 # If _use_X is undef, _supports_X is then queried. This is a "simple" style
55 # accessor, which in turn calls _determine_supports_X, and stores the return
56 # in a special slot on the storage object, which is wiped every time a $dbh
57 # reconnection takes place (it is not guaranteed that upon reconnection we
58 # will get the same rdbms version). _determine_supports_X does not need to
59 # exist on a driver, as we ->can for it before calling.
61 my @capabilities = (qw/
63 insert_returning_bound
68 __PACKAGE__->mk_group_accessors( dbms_capability => map { "_supports_$_" } @capabilities );
69 __PACKAGE__->mk_group_accessors( use_dbms_capability => map { "_use_$_" } (@capabilities ) );
71 # on by default, not strictly a capability (pending rewrite)
72 __PACKAGE__->_use_join_optimizer (1);
73 sub _determine_supports_join_optimizer { 1 };
75 # Each of these methods need _determine_driver called before itself
76 # in order to function reliably. This is a purely DRY optimization
78 # get_(use)_dbms_capability need to be called on the correct Storage
79 # class, as _use_X may be hardcoded class-wide, and _supports_X calls
80 # _determine_supports_X which obv. needs a correct driver as well
81 my @rdbms_specific_methods = qw/
95 with_deferred_fk_checks
97 get_use_dbms_capability
104 for my $meth (@rdbms_specific_methods) {
106 my $orig = __PACKAGE__->can ($meth)
107 or die "$meth is not a ::Storage::DBI method!";
110 no warnings qw/redefine/;
111 *{__PACKAGE__ ."::$meth"} = subname $meth => sub {
113 # only fire when invoked on an instance, a valid class-based invocation
114 # would e.g. be setting a default for an inherited accessor
117 ! $_[0]->_driver_determined
119 ! $_[0]->{_in_determine_driver}
121 $_[0]->_determine_driver;
123 # This for some reason crashes and burns on perl 5.8.1
124 # IFF the method ends up throwing an exception
125 #goto $_[0]->can ($meth);
127 my $cref = $_[0]->can ($meth);
137 DBIx::Class::Storage::DBI - DBI storage handler
141 my $schema = MySchema->connect('dbi:SQLite:my.db');
143 $schema->storage->debug(1);
145 my @stuff = $schema->storage->dbh_do(
147 my ($storage, $dbh, @args) = @_;
148 $dbh->do("DROP TABLE authors");
153 $schema->resultset('Book')->search({
154 written_on => $schema->storage->datetime_parser->format_datetime(DateTime->now)
159 This class represents the connection to an RDBMS via L<DBI>. See
160 L<DBIx::Class::Storage> for general information. This pod only
161 documents DBI-specific methods and behaviors.
168 my $new = shift->next::method(@_);
170 $new->_sql_maker_opts({});
171 $new->_dbh_details({});
172 $new->{_in_do_block} = 0;
173 $new->{_dbh_gen} = 0;
175 # read below to see what this does
176 $new->_arm_global_destructor;
181 # This is hack to work around perl shooting stuff in random
182 # order on exit(). If we do not walk the remaining storage
183 # objects in an END block, there is a *small but real* chance
184 # of a fork()ed child to kill the parent's shared DBI handle,
185 # *before perl reaches the DESTROY in this package*
186 # Yes, it is ugly and effective.
187 # Additionally this registry is used by the CLONE method to
188 # make sure no handles are shared between threads
190 my %seek_and_destroy;
192 sub _arm_global_destructor {
194 my $key = refaddr ($self);
195 $seek_and_destroy{$key} = $self;
196 weaken ($seek_and_destroy{$key});
200 local $?; # just in case the DBI destructor changes it somehow
202 # destroy just the object if not native to this process/thread
203 $_->_verify_pid for (grep
205 values %seek_and_destroy
210 # As per DBI's recommendation, DBIC disconnects all handles as
211 # soon as possible (DBIC will reconnect only on demand from within
213 for (values %seek_and_destroy) {
215 $_->{_dbh_gen}++; # so that existing cursors will drop as well
218 $_->transaction_depth(0);
227 # some databases spew warnings on implicit disconnect
228 local $SIG{__WARN__} = sub {};
231 # this op is necessary, since the very last perl runtime statement
232 # triggers a global destruction shootout, and the $SIG localization
233 # may very well be destroyed before perl actually gets to do the
238 # handle pid changes correctly - do not destroy parent's connection
242 my $pid = $self->_conn_pid;
243 if( defined $pid and $pid != $$ and my $dbh = $self->_dbh ) {
244 $dbh->{InactiveDestroy} = 1;
247 $self->transaction_depth(0);
248 $self->savepoints([]);
256 This method is normally called by L<DBIx::Class::Schema/connection>, which
257 encapsulates its argument list in an arrayref before passing them here.
259 The argument list may contain:
265 The same 4-element argument set one would normally pass to
266 L<DBI/connect>, optionally followed by
267 L<extra attributes|/DBIx::Class specific connection attributes>
268 recognized by DBIx::Class:
270 $connect_info_args = [ $dsn, $user, $password, \%dbi_attributes?, \%extra_attributes? ];
274 A single code reference which returns a connected
275 L<DBI database handle|DBI/connect> optionally followed by
276 L<extra attributes|/DBIx::Class specific connection attributes> recognized
279 $connect_info_args = [ sub { DBI->connect (...) }, \%extra_attributes? ];
283 A single hashref with all the attributes and the dsn/user/password
286 $connect_info_args = [{
294 $connect_info_args = [{
295 dbh_maker => sub { DBI->connect (...) },
300 This is particularly useful for L<Catalyst> based applications, allowing the
301 following config (L<Config::General> style):
306 dsn dbi:mysql:database=test
313 The C<dsn>/C<user>/C<password> combination can be substituted by the
314 C<dbh_maker> key whose value is a coderef that returns a connected
315 L<DBI database handle|DBI/connect>
319 Please note that the L<DBI> docs recommend that you always explicitly
320 set C<AutoCommit> to either I<0> or I<1>. L<DBIx::Class> further
321 recommends that it be set to I<1>, and that you perform transactions
322 via our L<DBIx::Class::Schema/txn_do> method. L<DBIx::Class> will set it
323 to I<1> if you do not do explicitly set it to zero. This is the default
324 for most DBDs. See L</DBIx::Class and AutoCommit> for details.
326 =head3 DBIx::Class specific connection attributes
328 In addition to the standard L<DBI|DBI/ATTRIBUTES_COMMON_TO_ALL_HANDLES>
329 L<connection|DBI/Database_Handle_Attributes> attributes, DBIx::Class recognizes
330 the following connection options. These options can be mixed in with your other
331 L<DBI> connection attributes, or placed in a separate hashref
332 (C<\%extra_attributes>) as shown above.
334 Every time C<connect_info> is invoked, any previous settings for
335 these options will be cleared before setting the new ones, regardless of
336 whether any options are specified in the new C<connect_info>.
343 Specifies things to do immediately after connecting or re-connecting to
344 the database. Its value may contain:
350 This contains one SQL statement to execute.
352 =item an array reference
354 This contains SQL statements to execute in order. Each element contains
355 a string or a code reference that returns a string.
357 =item a code reference
359 This contains some code to execute. Unlike code references within an
360 array reference, its return value is ignored.
364 =item on_disconnect_do
366 Takes arguments in the same form as L</on_connect_do> and executes them
367 immediately before disconnecting from the database.
369 Note, this only runs if you explicitly call L</disconnect> on the
372 =item on_connect_call
374 A more generalized form of L</on_connect_do> that calls the specified
375 C<connect_call_METHOD> methods in your storage driver.
377 on_connect_do => 'select 1'
381 on_connect_call => [ [ do_sql => 'select 1' ] ]
383 Its values may contain:
389 Will call the C<connect_call_METHOD> method.
391 =item a code reference
393 Will execute C<< $code->($storage) >>
395 =item an array reference
397 Each value can be a method name or code reference.
399 =item an array of arrays
401 For each array, the first item is taken to be the C<connect_call_> method name
402 or code reference, and the rest are parameters to it.
406 Some predefined storage methods you may use:
412 Executes a SQL string or a code reference that returns a SQL string. This is
413 what L</on_connect_do> and L</on_disconnect_do> use.
421 Will execute the scalar as SQL.
425 Taken to be arguments to L<DBI/do>, the SQL string optionally followed by the
426 attributes hashref and bind values.
428 =item a code reference
430 Will execute C<< $code->($storage) >> and execute the return array refs as
437 Execute any statements necessary to initialize the database session to return
438 and accept datetime/timestamp values used with
439 L<DBIx::Class::InflateColumn::DateTime>.
441 Only necessary for some databases, see your specific storage driver for
442 implementation details.
446 =item on_disconnect_call
448 Takes arguments in the same form as L</on_connect_call> and executes them
449 immediately before disconnecting from the database.
451 Calls the C<disconnect_call_METHOD> methods as opposed to the
452 C<connect_call_METHOD> methods called by L</on_connect_call>.
454 Note, this only runs if you explicitly call L</disconnect> on the
457 =item disable_sth_caching
459 If set to a true value, this option will disable the caching of
460 statement handles via L<DBI/prepare_cached>.
464 Sets a specific SQL::Abstract::Limit-style limit dialect, overriding the
465 default L</sql_limit_dialect> setting of the storage (if any). For a list
466 of available limit dialects see L<DBIx::Class::SQLMaker::LimitDialects>.
470 When true automatically sets L</quote_char> and L</name_sep> to the characters
471 appropriate for your particular RDBMS. This option is preferred over specifying
472 L</quote_char> directly.
476 Specifies what characters to use to quote table and column names.
478 C<quote_char> expects either a single character, in which case is it
479 is placed on either side of the table/column name, or an arrayref of length
480 2 in which case the table/column name is placed between the elements.
482 For example under MySQL you should use C<< quote_char => '`' >>, and for
483 SQL Server you should use C<< quote_char => [qw/[ ]/] >>.
487 This parameter is only useful in conjunction with C<quote_char>, and is used to
488 specify the character that separates elements (schemas, tables, columns) from
489 each other. If unspecified it defaults to the most commonly used C<.>.
493 This Storage driver normally installs its own C<HandleError>, sets
494 C<RaiseError> and C<ShowErrorStatement> on, and sets C<PrintError> off on
495 all database handles, including those supplied by a coderef. It does this
496 so that it can have consistent and useful error behavior.
498 If you set this option to a true value, Storage will not do its usual
499 modifications to the database handle's attributes, and instead relies on
500 the settings in your connect_info DBI options (or the values you set in
501 your connection coderef, in the case that you are connecting via coderef).
503 Note that your custom settings can cause Storage to malfunction,
504 especially if you set a C<HandleError> handler that suppresses exceptions
505 and/or disable C<RaiseError>.
509 If this option is true, L<DBIx::Class> will use savepoints when nesting
510 transactions, making it possible to recover from failure in the inner
511 transaction without having to abort all outer transactions.
515 Use this argument to supply a cursor class other than the default
516 L<DBIx::Class::Storage::DBI::Cursor>.
520 Some real-life examples of arguments to L</connect_info> and
521 L<DBIx::Class::Schema/connect>
523 # Simple SQLite connection
524 ->connect_info([ 'dbi:SQLite:./foo.db' ]);
527 ->connect_info([ sub { DBI->connect(...) } ]);
529 # Connect via subref in hashref
531 dbh_maker => sub { DBI->connect(...) },
532 on_connect_do => 'alter session ...',
535 # A bit more complicated
542 { quote_char => q{"} },
546 # Equivalent to the previous example
552 { AutoCommit => 1, quote_char => q{"}, name_sep => q{.} },
556 # Same, but with hashref as argument
557 # See parse_connect_info for explanation
560 dsn => 'dbi:Pg:dbname=foo',
562 password => 'my_pg_password',
569 # Subref + DBIx::Class-specific connection options
572 sub { DBI->connect(...) },
576 on_connect_do => ['SET search_path TO myschema,otherschema,public'],
577 disable_sth_caching => 1,
587 my ($self, $info) = @_;
589 return $self->_connect_info if !$info;
591 $self->_connect_info($info); # copy for _connect_info
593 $info = $self->_normalize_connect_info($info)
594 if ref $info eq 'ARRAY';
596 for my $storage_opt (keys %{ $info->{storage_options} }) {
597 my $value = $info->{storage_options}{$storage_opt};
599 $self->$storage_opt($value);
602 # Kill sql_maker/_sql_maker_opts, so we get a fresh one with only
603 # the new set of options
604 $self->_sql_maker(undef);
605 $self->_sql_maker_opts({});
607 for my $sql_maker_opt (keys %{ $info->{sql_maker_options} }) {
608 my $value = $info->{sql_maker_options}{$sql_maker_opt};
610 $self->_sql_maker_opts->{$sql_maker_opt} = $value;
614 %{ $self->_default_dbi_connect_attributes || {} },
615 %{ $info->{attributes} || {} },
618 my @args = @{ $info->{arguments} };
620 if (keys %attrs and ref $args[0] ne 'CODE') {
622 'You provided explicit AutoCommit => 0 in your connection_info. '
623 . 'This is almost universally a bad idea (see the footnotes of '
624 . 'DBIx::Class::Storage::DBI for more info). If you still want to '
625 . 'do this you can set $ENV{DBIC_UNSAFE_AUTOCOMMIT_OK} to disable '
627 if ! $attrs{AutoCommit} and ! $ENV{DBIC_UNSAFE_AUTOCOMMIT_OK};
629 push @args, \%attrs if keys %attrs;
631 $self->_dbi_connect_info(\@args);
634 # save attributes them in a separate accessor so they are always
635 # introspectable, even in case of a CODE $dbhmaker
636 $self->_dbic_connect_attributes (\%attrs);
638 return $self->_connect_info;
641 sub _normalize_connect_info {
642 my ($self, $info_arg) = @_;
645 my @args = @$info_arg; # take a shallow copy for further mutilation
647 # combine/pre-parse arguments depending on invocation style
650 if (ref $args[0] eq 'CODE') { # coderef with optional \%extra_attributes
651 %attrs = %{ $args[1] || {} };
654 elsif (ref $args[0] eq 'HASH') { # single hashref (i.e. Catalyst config)
655 %attrs = %{$args[0]};
657 if (my $code = delete $attrs{dbh_maker}) {
660 my @ignored = grep { delete $attrs{$_} } (qw/dsn user password/);
663 'Attribute(s) %s in connect_info were ignored, as they can not be applied '
664 . "to the result of 'dbh_maker'",
666 join (', ', map { "'$_'" } (@ignored) ),
671 @args = delete @attrs{qw/dsn user password/};
674 else { # otherwise assume dsn/user/password + \%attrs + \%extra_attrs
676 % { $args[3] || {} },
677 % { $args[4] || {} },
679 @args = @args[0,1,2];
682 $info{arguments} = \@args;
684 my @storage_opts = grep exists $attrs{$_},
685 @storage_options, 'cursor_class';
687 @{ $info{storage_options} }{@storage_opts} =
688 delete @attrs{@storage_opts} if @storage_opts;
690 my @sql_maker_opts = grep exists $attrs{$_},
691 qw/limit_dialect quote_char name_sep quote_names/;
693 @{ $info{sql_maker_options} }{@sql_maker_opts} =
694 delete @attrs{@sql_maker_opts} if @sql_maker_opts;
696 $info{attributes} = \%attrs if %attrs;
701 sub _default_dbi_connect_attributes () {
706 ShowErrorStatement => 1,
712 This method is deprecated in favour of setting via L</connect_info>.
716 =head2 on_disconnect_do
718 This method is deprecated in favour of setting via L</connect_info>.
722 sub _parse_connect_do {
723 my ($self, $type) = @_;
725 my $val = $self->$type;
726 return () if not defined $val;
731 push @res, [ 'do_sql', $val ];
732 } elsif (ref($val) eq 'CODE') {
734 } elsif (ref($val) eq 'ARRAY') {
735 push @res, map { [ 'do_sql', $_ ] } @$val;
737 $self->throw_exception("Invalid type for $type: ".ref($val));
745 Arguments: ($subref | $method_name), @extra_coderef_args?
747 Execute the given $subref or $method_name using the new exception-based
748 connection management.
750 The first two arguments will be the storage object that C<dbh_do> was called
751 on and a database handle to use. Any additional arguments will be passed
752 verbatim to the called subref as arguments 2 and onwards.
754 Using this (instead of $self->_dbh or $self->dbh) ensures correct
755 exception handling and reconnection (or failover in future subclasses).
757 Your subref should have no side-effects outside of the database, as
758 there is the potential for your subref to be partially double-executed
759 if the database connection was stale/dysfunctional.
763 my @stuff = $schema->storage->dbh_do(
765 my ($storage, $dbh, @cols) = @_;
766 my $cols = join(q{, }, @cols);
767 $dbh->selectrow_array("SELECT $cols FROM foo");
778 my $dbh = $self->_get_dbh;
780 return $self->$code($dbh, @_)
781 if ( $self->{_in_do_block} || $self->{transaction_depth} );
783 local $self->{_in_do_block} = 1;
785 # take a ref instead of a copy, to preserve coderef @_ aliasing semantics
789 $self->$code ($dbh, @$args);
791 $self->throw_exception($_) if $self->connected;
793 # We were not connected - reconnect and retry, but let any
794 # exception fall right through this time
795 carp "Retrying dbh_do($code) after catching disconnected exception: $_"
796 if $ENV{DBIC_STORAGE_RETRY_DEBUG};
798 $self->_populate_dbh;
799 $self->$code($self->_dbh, @$args);
804 # connects or reconnects on pid change, necessary to grab correct txn_depth
806 local $_[0]->{_in_do_block} = 1;
807 shift->next::method(@_);
812 Our C<disconnect> method also performs a rollback first if the
813 database is not in C<AutoCommit> mode.
823 push @actions, ( $self->on_disconnect_call || () );
824 push @actions, $self->_parse_connect_do ('on_disconnect_do');
826 $self->_do_connection_actions(disconnect_call_ => $_) for @actions;
828 # stops the "implicit rollback on disconnect" warning
829 $self->_exec_txn_rollback unless $self->_dbh_autocommit;
831 %{ $self->_dbh->{CachedKids} } = ();
832 $self->_dbh->disconnect;
838 =head2 with_deferred_fk_checks
842 =item Arguments: C<$coderef>
844 =item Return Value: The return value of $coderef
848 Storage specific method to run the code ref with FK checks deferred or
849 in MySQL's case disabled entirely.
853 # Storage subclasses should override this
854 sub with_deferred_fk_checks {
855 my ($self, $sub) = @_;
863 =item Arguments: none
865 =item Return Value: 1|0
869 Verifies that the current database handle is active and ready to execute
870 an SQL statement (e.g. the connection did not get stale, server is still
871 answering, etc.) This method is used internally by L</dbh>.
877 return 0 unless $self->_seems_connected;
880 local $self->_dbh->{RaiseError} = 1;
885 sub _seems_connected {
890 my $dbh = $self->_dbh
893 return $dbh->FETCH('Active');
899 my $dbh = $self->_dbh or return 0;
904 sub ensure_connected {
907 unless ($self->connected) {
908 $self->_populate_dbh;
914 Returns a C<$dbh> - a data base handle of class L<DBI>. The returned handle
915 is guaranteed to be healthy by implicitly calling L</connected>, and if
916 necessary performing a reconnection before returning. Keep in mind that this
917 is very B<expensive> on some database engines. Consider using L</dbh_do>
925 if (not $self->_dbh) {
926 $self->_populate_dbh;
928 $self->ensure_connected;
933 # this is the internal "get dbh or connect (don't check)" method
937 $self->_populate_dbh unless $self->_dbh;
943 unless ($self->_sql_maker) {
944 my $sql_maker_class = $self->sql_maker_class;
946 my %opts = %{$self->_sql_maker_opts||{}};
950 $self->sql_limit_dialect
953 my $s_class = (ref $self) || $self;
955 "Your storage class ($s_class) does not set sql_limit_dialect and you "
956 . 'have not supplied an explicit limit_dialect in your connection_info. '
957 . 'DBIC will attempt to use the GenericSubQ dialect, which works on most '
958 . 'databases but can be (and often is) painfully slow. '
959 . "Please file an RT ticket against '$s_class' ."
966 my ($quote_char, $name_sep);
968 if ($opts{quote_names}) {
969 $quote_char = (delete $opts{quote_char}) || $self->sql_quote_char || do {
970 my $s_class = (ref $self) || $self;
972 "You requested 'quote_names' but your storage class ($s_class) does "
973 . 'not explicitly define a default sql_quote_char and you have not '
974 . 'supplied a quote_char as part of your connection_info. DBIC will '
975 .q{default to the ANSI SQL standard quote '"', which works most of }
976 . "the time. Please file an RT ticket against '$s_class'."
982 $name_sep = (delete $opts{name_sep}) || $self->sql_name_sep;
985 $self->_sql_maker($sql_maker_class->new(
987 array_datatypes => 1,
988 limit_dialect => $dialect,
989 ($quote_char ? (quote_char => $quote_char) : ()),
990 name_sep => ($name_sep || '.'),
994 return $self->_sql_maker;
997 # nothing to do by default
1004 my @info = @{$self->_dbi_connect_info || []};
1005 $self->_dbh(undef); # in case ->connected failed we might get sent here
1006 $self->_dbh_details({}); # reset everything we know
1008 $self->_dbh($self->_connect(@info));
1010 $self->_conn_pid($$) if $^O ne 'MSWin32'; # on win32 these are in fact threads
1012 $self->_determine_driver;
1014 # Always set the transaction depth on connect, since
1015 # there is no transaction in progress by definition
1016 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
1018 $self->_run_connection_actions unless $self->{_in_determine_driver};
1021 sub _run_connection_actions {
1025 push @actions, ( $self->on_connect_call || () );
1026 push @actions, $self->_parse_connect_do ('on_connect_do');
1028 $self->_do_connection_actions(connect_call_ => $_) for @actions;
1033 sub set_use_dbms_capability {
1034 $_[0]->set_inherited ($_[1], $_[2]);
1037 sub get_use_dbms_capability {
1038 my ($self, $capname) = @_;
1040 my $use = $self->get_inherited ($capname);
1043 : do { $capname =~ s/^_use_/_supports_/; $self->get_dbms_capability ($capname) }
1047 sub set_dbms_capability {
1048 $_[0]->_dbh_details->{capability}{$_[1]} = $_[2];
1051 sub get_dbms_capability {
1052 my ($self, $capname) = @_;
1054 my $cap = $self->_dbh_details->{capability}{$capname};
1056 unless (defined $cap) {
1057 if (my $meth = $self->can ("_determine$capname")) {
1058 $cap = $self->$meth ? 1 : 0;
1064 $self->set_dbms_capability ($capname, $cap);
1074 unless ($info = $self->_dbh_details->{info}) {
1078 my $server_version = try { $self->_get_server_version };
1080 if (defined $server_version) {
1081 $info->{dbms_version} = $server_version;
1083 my ($numeric_version) = $server_version =~ /^([\d\.]+)/;
1084 my @verparts = split (/\./, $numeric_version);
1090 # consider only up to 3 version parts, iff not more than 3 digits
1092 while (@verparts && @use_parts < 3) {
1093 my $p = shift @verparts;
1095 push @use_parts, $p;
1097 push @use_parts, 0 while @use_parts < 3;
1099 $info->{normalized_dbms_version} = sprintf "%d.%03d%03d", @use_parts;
1103 $self->_dbh_details->{info} = $info;
1109 sub _get_server_version {
1110 shift->_dbh_get_info(18);
1114 my ($self, $info) = @_;
1116 return try { $self->_get_dbh->get_info($info) } || undef;
1119 sub _determine_driver {
1122 if ((not $self->_driver_determined) && (not $self->{_in_determine_driver})) {
1123 my $started_connected = 0;
1124 local $self->{_in_determine_driver} = 1;
1126 if (ref($self) eq __PACKAGE__) {
1128 if ($self->_dbh) { # we are connected
1129 $driver = $self->_dbh->{Driver}{Name};
1130 $started_connected = 1;
1132 # if connect_info is a CODEREF, we have no choice but to connect
1133 if (ref $self->_dbi_connect_info->[0] &&
1134 reftype $self->_dbi_connect_info->[0] eq 'CODE') {
1135 $self->_populate_dbh;
1136 $driver = $self->_dbh->{Driver}{Name};
1139 # try to use dsn to not require being connected, the driver may still
1140 # force a connection in _rebless to determine version
1141 # (dsn may not be supplied at all if all we do is make a mock-schema)
1142 my $dsn = $self->_dbi_connect_info->[0] || $ENV{DBI_DSN} || '';
1143 ($driver) = $dsn =~ /dbi:([^:]+):/i;
1144 $driver ||= $ENV{DBI_DRIVER};
1149 my $storage_class = "DBIx::Class::Storage::DBI::${driver}";
1150 if ($self->load_optional_class($storage_class)) {
1151 mro::set_mro($storage_class, 'c3');
1152 bless $self, $storage_class;
1158 $self->_driver_determined(1);
1160 Class::C3->reinitialize() if DBIx::Class::_ENV_::OLD_MRO;
1162 $self->_init; # run driver-specific initializations
1164 $self->_run_connection_actions
1165 if !$started_connected && defined $self->_dbh;
1169 sub _do_connection_actions {
1171 my $method_prefix = shift;
1174 if (not ref($call)) {
1175 my $method = $method_prefix . $call;
1177 } elsif (ref($call) eq 'CODE') {
1179 } elsif (ref($call) eq 'ARRAY') {
1180 if (ref($call->[0]) ne 'ARRAY') {
1181 $self->_do_connection_actions($method_prefix, $_) for @$call;
1183 $self->_do_connection_actions($method_prefix, @$_) for @$call;
1186 $self->throw_exception (sprintf ("Don't know how to process conection actions of type '%s'", ref($call)) );
1192 sub connect_call_do_sql {
1194 $self->_do_query(@_);
1197 sub disconnect_call_do_sql {
1199 $self->_do_query(@_);
1202 # override in db-specific backend when necessary
1203 sub connect_call_datetime_setup { 1 }
1206 my ($self, $action) = @_;
1208 if (ref $action eq 'CODE') {
1209 $action = $action->($self);
1210 $self->_do_query($_) foreach @$action;
1213 # Most debuggers expect ($sql, @bind), so we need to exclude
1214 # the attribute hash which is the second argument to $dbh->do
1215 # furthermore the bind values are usually to be presented
1216 # as named arrayref pairs, so wrap those here too
1217 my @do_args = (ref $action eq 'ARRAY') ? (@$action) : ($action);
1218 my $sql = shift @do_args;
1219 my $attrs = shift @do_args;
1220 my @bind = map { [ undef, $_ ] } @do_args;
1222 $self->_query_start($sql, \@bind);
1223 $self->_get_dbh->do($sql, $attrs, @do_args);
1224 $self->_query_end($sql, \@bind);
1231 my ($self, @info) = @_;
1233 $self->throw_exception("You failed to provide any connection info")
1236 my ($old_connect_via, $dbh);
1238 if ($INC{'Apache/DBI.pm'} && $ENV{MOD_PERL}) {
1239 $old_connect_via = $DBI::connect_via;
1240 $DBI::connect_via = 'connect';
1244 if(ref $info[0] eq 'CODE') {
1245 $dbh = $info[0]->();
1249 $dbh = DBI->connect(@info);
1256 unless ($self->unsafe) {
1258 $self->throw_exception(
1259 'Refusing clobbering of {HandleError} installed on externally supplied '
1260 ."DBI handle $dbh. Either remove the handler or use the 'unsafe' attribute."
1261 ) if $dbh->{HandleError} and ref $dbh->{HandleError} ne '__DBIC__DBH__ERROR__HANDLER__';
1263 # Default via _default_dbi_connect_attributes is 1, hence it was an explicit
1264 # request, or an external handle. Complain and set anyway
1265 unless ($dbh->{RaiseError}) {
1266 carp( ref $info[0] eq 'CODE'
1268 ? "The 'RaiseError' of the externally supplied DBI handle is set to false. "
1269 ."DBIx::Class will toggle it back to true, unless the 'unsafe' connect "
1270 .'attribute has been supplied'
1272 : 'RaiseError => 0 supplied in your connection_info, without an explicit '
1273 .'unsafe => 1. Toggling RaiseError back to true'
1276 $dbh->{RaiseError} = 1;
1279 # this odd anonymous coderef dereference is in fact really
1280 # necessary to avoid the unwanted effect described in perl5
1283 my $weak_self = $_[0];
1286 # the coderef is blessed so we can distinguish it from externally
1287 # supplied handles (which must be preserved)
1288 $_[1]->{HandleError} = bless sub {
1290 $weak_self->throw_exception("DBI Exception: $_[0]");
1293 # the handler may be invoked by something totally out of
1295 DBIx::Class::Exception->throw("DBI Exception (unhandled by DBIC, ::Schema GCed): $_[0]");
1297 }, '__DBIC__DBH__ERROR__HANDLER__';
1302 $self->throw_exception("DBI Connection failed: $_")
1305 $DBI::connect_via = $old_connect_via if $old_connect_via;
1308 $self->_dbh_autocommit($dbh->{AutoCommit});
1315 # this means we have not yet connected and do not know the AC status
1316 # (e.g. coderef $dbh), need a full-fledged connection check
1317 if (! defined $self->_dbh_autocommit) {
1318 $self->ensure_connected;
1320 # Otherwise simply connect or re-connect on pid changes
1325 $self->next::method(@_);
1328 sub _exec_txn_begin {
1331 # if the user is utilizing txn_do - good for him, otherwise we need to
1332 # ensure that the $dbh is healthy on BEGIN.
1333 # We do this via ->dbh_do instead of ->dbh, so that the ->dbh "ping"
1334 # will be replaced by a failure of begin_work itself (which will be
1335 # then retried on reconnect)
1336 if ($self->{_in_do_block}) {
1337 $self->_dbh->begin_work;
1339 $self->dbh_do(sub { $_[1]->begin_work });
1346 $self->_verify_pid if $self->_dbh;
1347 $self->throw_exception("Unable to txn_commit() on a disconnected storage")
1350 # esoteric case for folks using external $dbh handles
1351 if (! $self->transaction_depth and ! $self->_dbh->FETCH('AutoCommit') ) {
1352 carp "Storage transaction_depth 0 does not match "
1353 ."false AutoCommit of $self->{_dbh}, attempting COMMIT anyway";
1354 $self->transaction_depth(1);
1357 $self->next::method(@_);
1359 # if AutoCommit is disabled txn_depth never goes to 0
1360 # as a new txn is started immediately on commit
1361 $self->transaction_depth(1) if (
1362 !$self->transaction_depth
1364 defined $self->_dbh_autocommit
1366 ! $self->_dbh_autocommit
1370 sub _exec_txn_commit {
1371 shift->_dbh->commit;
1377 $self->_verify_pid if $self->_dbh;
1378 $self->throw_exception("Unable to txn_rollback() on a disconnected storage")
1381 # esoteric case for folks using external $dbh handles
1382 if (! $self->transaction_depth and ! $self->_dbh->FETCH('AutoCommit') ) {
1383 carp "Storage transaction_depth 0 does not match "
1384 ."false AutoCommit of $self->{_dbh}, attempting ROLLBACK anyway";
1385 $self->transaction_depth(1);
1388 $self->next::method(@_);
1390 # if AutoCommit is disabled txn_depth never goes to 0
1391 # as a new txn is started immediately on commit
1392 $self->transaction_depth(1) if (
1393 !$self->transaction_depth
1395 defined $self->_dbh_autocommit
1397 ! $self->_dbh_autocommit
1401 sub _exec_txn_rollback {
1402 shift->_dbh->rollback;
1405 # generate some identical methods
1406 for my $meth (qw/svp_begin svp_release svp_rollback/) {
1408 *{__PACKAGE__ ."::$meth"} = subname $meth => sub {
1410 $self->_verify_pid if $self->_dbh;
1411 $self->throw_exception("Unable to $meth() on a disconnected storage")
1413 $self->next::method(@_);
1417 # This used to be the top-half of _execute. It was split out to make it
1418 # easier to override in NoBindVars without duping the rest. It takes up
1419 # all of _execute's args, and emits $sql, @bind.
1420 sub _prep_for_execute {
1421 #my ($self, $op, $ident, $args) = @_;
1422 return shift->_gen_sql_bind(@_)
1426 my ($self, $op, $ident, $args) = @_;
1428 my ($sql, @bind) = $self->sql_maker->$op(
1429 blessed($ident) ? $ident->from : $ident,
1433 my (@final_bind, $colinfos);
1434 my $resolve_bindinfo = sub {
1435 $colinfos ||= $self->_resolve_column_info($ident);
1436 if (my $col = $_[1]->{dbic_colname}) {
1437 $_[1]->{sqlt_datatype} ||= $colinfos->{$col}{data_type}
1438 if $colinfos->{$col}{data_type};
1439 $_[1]->{sqlt_size} ||= $colinfos->{$col}{size}
1440 if $colinfos->{$col}{size};
1445 for my $e (@{$args->[2]{bind}||[]}, @bind) {
1446 push @final_bind, [ do {
1447 if (ref $e ne 'ARRAY') {
1450 elsif (! defined $e->[0]) {
1453 elsif (ref $e->[0] eq 'HASH') {
1455 (first { $e->[0]{$_} } qw/dbd_attrs sqlt_datatype/) ? $e->[0] : $self->$resolve_bindinfo($e->[0]),
1459 elsif (ref $e->[0] eq 'SCALAR') {
1460 ( { sqlt_datatype => ${$e->[0]} }, $e->[1] )
1463 ( $self->$resolve_bindinfo({ dbic_colname => $e->[0] }), $e->[1] )
1468 ($sql, \@final_bind);
1471 sub _format_for_trace {
1472 #my ($self, $bind) = @_;
1474 ### Turn @bind from something like this:
1475 ### ( [ "artist", 1 ], [ \%attrs, 3 ] )
1477 ### ( "'1'", "'3'" )
1480 defined( $_ && $_->[1] )
1487 my ( $self, $sql, $bind ) = @_;
1489 $self->debugobj->query_start( $sql, $self->_format_for_trace($bind) )
1494 my ( $self, $sql, $bind ) = @_;
1496 $self->debugobj->query_end( $sql, $self->_format_for_trace($bind) )
1501 sub _dbi_attrs_for_bind {
1502 my ($self, $ident, $bind) = @_;
1504 if (! defined $sba_compat) {
1505 $self->_determine_driver;
1506 $sba_compat = $self->can('source_bind_attributes') == \&source_bind_attributes
1514 my $class = ref $self;
1516 "The source_bind_attributes() override in $class relies on a deprecated codepath. "
1517 .'You are strongly advised to switch your code to override bind_attribute_by_datatype() '
1518 .'instead. This legacy compat shim will also disappear some time before DBIC 0.09'
1521 my $sba_attrs = $self->source_bind_attributes
1526 for (map { $_->[0] } @$bind) {
1528 if (exists $_->{dbd_attrs}) {
1531 elsif($_->{sqlt_datatype}) {
1532 # cache the result in the dbh_details hash, as it can not change unless
1533 # we connect to something else
1534 my $cache = $self->_dbh_details->{_datatype_map_cache} ||= {};
1535 if (not exists $cache->{$_->{sqlt_datatype}}) {
1536 $cache->{$_->{sqlt_datatype}} = $self->bind_attribute_by_data_type($_->{sqlt_datatype}) || undef;
1538 $cache->{$_->{sqlt_datatype}};
1540 elsif ($sba_attrs and $_->{dbic_colname}) {
1541 $sba_attrs->{$_->{dbic_colname}} || undef;
1544 undef; # always push something at this position
1553 my ($self, $op, $ident, @args) = @_;
1555 my ($sql, $bind) = $self->_prep_for_execute($op, $ident, \@args);
1557 shift->dbh_do( # retry over disconnects
1561 $self->_dbi_attrs_for_bind($ident, $bind)
1566 my ($self, undef, $sql, $bind, $bind_attrs) = @_;
1568 $self->_query_start( $sql, $bind );
1569 my $sth = $self->_sth($sql);
1571 for my $i (0 .. $#$bind) {
1572 if (ref $bind->[$i][1] eq 'SCALAR') { # any scalarrefs are assumed to be bind_inouts
1573 $sth->bind_param_inout(
1574 $i + 1, # bind params counts are 1-based
1576 $bind->[$i][0]{dbd_size} || $self->_max_column_bytesize($bind->[$i][0]), # size
1583 (ref $bind->[$i][1] and overload::Method($bind->[$i][1], '""'))
1592 # Can this fail without throwing an exception anyways???
1593 my $rv = $sth->execute();
1594 $self->throw_exception(
1595 $sth->errstr || $sth->err || 'Unknown error: execute() returned false, but error flags were not set...'
1598 $self->_query_end( $sql, $bind );
1600 return (wantarray ? ($rv, $sth, @$bind) : $rv);
1603 sub _prefetch_autovalues {
1604 my ($self, $source, $to_insert) = @_;
1606 my $colinfo = $source->columns_info;
1609 for my $col (keys %$colinfo) {
1611 $colinfo->{$col}{auto_nextval}
1614 ! exists $to_insert->{$col}
1616 ref $to_insert->{$col} eq 'SCALAR'
1618 (ref $to_insert->{$col} eq 'REF' and ref ${$to_insert->{$col}} eq 'ARRAY')
1621 $values{$col} = $self->_sequence_fetch(
1623 ( $colinfo->{$col}{sequence} ||=
1624 $self->_dbh_get_autoinc_seq($self->_get_dbh, $source, $col)
1634 my ($self, $source, $to_insert) = @_;
1636 my $prefetched_values = $self->_prefetch_autovalues($source, $to_insert);
1638 # fuse the values, but keep a separate list of prefetched_values so that
1639 # they can be fused once again with the final return
1640 $to_insert = { %$to_insert, %$prefetched_values };
1642 my $col_infos = $source->columns_info;
1643 my %pcols = map { $_ => 1 } $source->primary_columns;
1645 for my $col ($source->columns) {
1646 # nothing to retrieve when explicit values are supplied
1647 next if (defined $to_insert->{$col} and ! (
1648 ref $to_insert->{$col} eq 'SCALAR'
1650 (ref $to_insert->{$col} eq 'REF' and ref ${$to_insert->{$col}} eq 'ARRAY')
1653 # the 'scalar keys' is a trick to preserve the ->columns declaration order
1654 $retrieve_cols{$col} = scalar keys %retrieve_cols if (
1657 $col_infos->{$col}{retrieve_on_insert}
1661 my ($sqla_opts, @ir_container);
1662 if (%retrieve_cols and $self->_use_insert_returning) {
1663 $sqla_opts->{returning_container} = \@ir_container
1664 if $self->_use_insert_returning_bound;
1666 $sqla_opts->{returning} = [
1667 sort { $retrieve_cols{$a} <=> $retrieve_cols{$b} } keys %retrieve_cols
1671 my ($rv, $sth) = $self->_execute('insert', $source, $to_insert, $sqla_opts);
1673 my %returned_cols = %$to_insert;
1674 if (my $retlist = $sqla_opts->{returning}) { # if IR is supported - we will get everything in one set
1675 @ir_container = try {
1676 local $SIG{__WARN__} = sub {};
1677 my @r = $sth->fetchrow_array;
1680 } unless @ir_container;
1682 @returned_cols{@$retlist} = @ir_container if @ir_container;
1685 # pull in PK if needed and then everything else
1686 if (my @missing_pri = grep { $pcols{$_} } keys %retrieve_cols) {
1688 $self->throw_exception( "Missing primary key but Storage doesn't support last_insert_id" )
1689 unless $self->can('last_insert_id');
1691 my @pri_values = $self->last_insert_id($source, @missing_pri);
1693 $self->throw_exception( "Can't get last insert id" )
1694 unless (@pri_values == @missing_pri);
1696 @returned_cols{@missing_pri} = @pri_values;
1697 delete $retrieve_cols{$_} for @missing_pri;
1700 # if there is more left to pull
1701 if (%retrieve_cols) {
1702 $self->throw_exception(
1703 'Unable to retrieve additional columns without a Primary Key on ' . $source->source_name
1706 my @left_to_fetch = sort { $retrieve_cols{$a} <=> $retrieve_cols{$b} } keys %retrieve_cols;
1708 my $cur = DBIx::Class::ResultSet->new($source, {
1709 where => { map { $_ => $returned_cols{$_} } (keys %pcols) },
1710 select => \@left_to_fetch,
1713 @returned_cols{@left_to_fetch} = $cur->next;
1715 $self->throw_exception('Duplicate row returned for PK-search after fresh insert')
1716 if scalar $cur->next;
1720 return { %$prefetched_values, %returned_cols };
1724 my ($self, $source, $cols, $data) = @_;
1726 # FIXME - perhaps this is not even needed? does DBI stringify?
1728 # forcibly stringify whatever is stringifiable
1729 for my $r (0 .. $#$data) {
1730 for my $c (0 .. $#{$data->[$r]}) {
1731 $data->[$r][$c] = "$data->[$r][$c]"
1732 if ( ref $data->[$r][$c] and overload::Method($data->[$r][$c], '""') );
1736 # check the data for consistency
1737 # report a sensible error on bad data
1739 # also create a list of dynamic binds (ones that will be changing
1742 for my $col_idx (0..$#$cols) {
1744 # the first "row" is used as a point of reference
1745 my $reference_val = $data->[0][$col_idx];
1746 my $is_literal = ref $reference_val eq 'SCALAR';
1747 my $is_literal_bind = ( !$is_literal and (
1748 ref $reference_val eq 'REF'
1750 ref $$reference_val eq 'ARRAY'
1753 $dyn_bind_idx->{$col_idx} = 1
1754 if (!$is_literal and !$is_literal_bind);
1756 # use a closure for convenience (less to pass)
1757 my $bad_slice = sub {
1758 my ($msg, $slice_idx) = @_;
1759 $self->throw_exception(sprintf "%s for column '%s' in populate slice:\n%s",
1763 require Data::Dumper::Concise;
1764 local $Data::Dumper::Maxdepth = 2;
1765 Data::Dumper::Concise::Dumper ({
1766 map { $cols->[$_] =>
1767 $data->[$slice_idx][$_]
1774 for my $row_idx (1..$#$data) { # we are comparing against what we got from [0] above, hence start from 1
1775 my $val = $data->[$row_idx][$col_idx];
1778 if (ref $val ne 'SCALAR') {
1780 "Incorrect value (expecting SCALAR-ref \\'$$reference_val')",
1784 elsif ($$val ne $$reference_val) {
1786 "Inconsistent literal SQL value (expecting \\'$$reference_val')",
1791 elsif ($is_literal_bind) {
1792 if (ref $val ne 'REF' or ref $$val ne 'ARRAY') {
1794 "Incorrect value (expecting ARRAYREF-ref \\['${$reference_val}->[0]', ... ])",
1798 elsif (${$val}->[0] ne ${$reference_val}->[0]) {
1800 "Inconsistent literal SQL-bind value (expecting \\['${$reference_val}->[0]', ... ])",
1806 if (ref $val eq 'SCALAR' or (ref $val eq 'REF' and ref $$val eq 'ARRAY') ) {
1807 $bad_slice->("Literal SQL found where a plain bind value is expected", $row_idx);
1810 $bad_slice->("$val reference found where bind expected", $row_idx);
1816 # Get the sql with bind values interpolated where necessary. For dynamic
1817 # binds convert the values of the first row into a literal+bind combo, with
1818 # extra positional info in the bind attr hashref. This will allow us to match
1819 # the order properly, and is so contrived because a user-supplied literal
1820 # bind (or something else specific to a resultsource and/or storage driver)
1821 # can inject extra binds along the way, so one can't rely on "shift
1822 # positions" ordering at all. Also we can't just hand SQLA a set of some
1823 # known "values" (e.g. hashrefs that can be later matched up by address),
1824 # because we want to supply a real value on which perhaps e.g. datatype
1825 # checks will be performed
1826 my ($sql, $proto_bind) = $self->_prep_for_execute (
1829 [ { map { $cols->[$_] => $dyn_bind_idx->{$_}
1831 { dbic_colname => $cols->[$_], _bind_data_slice_idx => $_ }
1839 if (! @$proto_bind and keys %$dyn_bind_idx) {
1840 # if the bindlist is empty and we had some dynamic binds, this means the
1841 # storage ate them away (e.g. the NoBindVars component) and interpolated
1842 # them directly into the SQL. This obviosly can't be good for multi-inserts
1843 $self->throw_exception('Cannot insert_bulk without support for placeholders');
1846 # neither _execute_array, nor _execute_inserts_with_no_binds are
1847 # atomic (even if _execute _array is a single call). Thus a safety
1849 my $guard = $self->txn_scope_guard;
1851 $self->_query_start( $sql, @$proto_bind ? [[undef => '__BULK_INSERT__' ]] : () );
1852 my $sth = $self->_sth($sql);
1855 # proto bind contains the information on which pieces of $data to pull
1856 # $cols is passed in only for prettier error-reporting
1857 $self->_execute_array( $source, $sth, $proto_bind, $cols, $data );
1860 # bind_param_array doesn't work if there are no binds
1861 $self->_dbh_execute_inserts_with_no_binds( $sth, scalar @$data );
1865 $self->_query_end( $sql, @$proto_bind ? [[ undef => '__BULK_INSERT__' ]] : () );
1869 return (wantarray ? ($rv, $sth, @$proto_bind) : $rv);
1872 sub _execute_array {
1873 my ($self, $source, $sth, $proto_bind, $cols, $data, @extra) = @_;
1875 ## This must be an arrayref, else nothing works!
1876 my $tuple_status = [];
1878 my $bind_attrs = $self->_dbi_attrs_for_bind($source, $proto_bind);
1880 # Bind the values by column slices
1881 for my $i (0 .. $#$proto_bind) {
1882 my $data_slice_idx = (
1883 ref $proto_bind->[$i][0] eq 'HASH'
1885 exists $proto_bind->[$i][0]{_bind_data_slice_idx}
1886 ) ? $proto_bind->[$i][0]{_bind_data_slice_idx} : undef;
1888 $sth->bind_param_array(
1889 $i+1, # DBI bind indexes are 1-based
1890 defined $data_slice_idx
1891 # either get a "column" of dynamic values, or just repeat the same
1892 # bind over and over
1893 ? [ map { $_->[$data_slice_idx] } @$data ]
1894 : [ ($proto_bind->[$i][1]) x @$data ]
1896 defined $bind_attrs->[$i] ? $bind_attrs->[$i] : (), # some DBDs throw up when given an undef
1902 $rv = $self->_dbh_execute_array($sth, $tuple_status, @extra);
1908 # Not all DBDs are create equal. Some throw on error, some return
1909 # an undef $rv, and some set $sth->err - try whatever we can
1910 $err = ($sth->errstr || 'UNKNOWN ERROR ($sth->errstr is unset)') if (
1913 ( !defined $rv or $sth->err )
1916 # Statement must finish even if there was an exception.
1921 $err = shift unless defined $err
1926 ++$i while $i <= $#$tuple_status && !ref $tuple_status->[$i];
1928 $self->throw_exception("Unexpected populate error: $err")
1929 if ($i > $#$tuple_status);
1931 require Data::Dumper::Concise;
1932 $self->throw_exception(sprintf "execute_array() aborted with '%s' at populate slice:\n%s",
1933 ($tuple_status->[$i][1] || $err),
1934 Data::Dumper::Concise::Dumper( { map { $cols->[$_] => $data->[$i][$_] } (0 .. $#$cols) } ),
1941 sub _dbh_execute_array {
1942 #my ($self, $sth, $tuple_status, @extra) = @_;
1943 return $_[1]->execute_array({ArrayTupleStatus => $_[2]});
1946 sub _dbh_execute_inserts_with_no_binds {
1947 my ($self, $sth, $count) = @_;
1951 my $dbh = $self->_get_dbh;
1952 local $dbh->{RaiseError} = 1;
1953 local $dbh->{PrintError} = 0;
1955 $sth->execute foreach 1..$count;
1961 # Make sure statement is finished even if there was an exception.
1966 $err = shift unless defined $err;
1969 $self->throw_exception($err) if defined $err;
1975 #my ($self, $source, @args) = @_;
1976 shift->_execute('update', @_);
1981 #my ($self, $source, @args) = @_;
1982 shift->_execute('delete', @_);
1985 # We were sent here because the $rs contains a complex search
1986 # which will require a subquery to select the correct rows
1987 # (i.e. joined or limited resultsets, or non-introspectable conditions)
1989 # Generating a single PK column subquery is trivial and supported
1990 # by all RDBMS. However if we have a multicolumn PK, things get ugly.
1991 # Look at _multipk_update_delete()
1992 sub _subq_update_delete {
1994 my ($rs, $op, $values) = @_;
1996 my $rsrc = $rs->result_source;
1998 # quick check if we got a sane rs on our hands
1999 my @pcols = $rsrc->_pri_cols;
2001 my $sel = $rs->_resolved_attrs->{select};
2002 $sel = [ $sel ] unless ref $sel eq 'ARRAY';
2005 join ("\x00", map { join '.', $rs->{attrs}{alias}, $_ } sort @pcols)
2007 join ("\x00", sort @$sel )
2009 $self->throw_exception (
2010 '_subq_update_delete can not be called on resultsets selecting columns other than the primary keys'
2017 $op eq 'update' ? $values : (),
2018 { $pcols[0] => { -in => $rs->as_query } },
2023 return $self->_multipk_update_delete (@_);
2027 # ANSI SQL does not provide a reliable way to perform a multicol-PK
2028 # resultset update/delete involving subqueries. So by default resort
2029 # to simple (and inefficient) delete_all style per-row opearations,
2030 # while allowing specific storages to override this with a faster
2033 sub _multipk_update_delete {
2034 return shift->_per_row_update_delete (@_);
2037 # This is the default loop used to delete/update rows for multi PK
2038 # resultsets, and used by mysql exclusively (because it can't do anything
2041 # We do not use $row->$op style queries, because resultset update/delete
2042 # is not expected to cascade (this is what delete_all/update_all is for).
2044 # There should be no race conditions as the entire operation is rolled
2047 sub _per_row_update_delete {
2049 my ($rs, $op, $values) = @_;
2051 my $rsrc = $rs->result_source;
2052 my @pcols = $rsrc->_pri_cols;
2054 my $guard = $self->txn_scope_guard;
2056 # emulate the return value of $sth->execute for non-selects
2057 my $row_cnt = '0E0';
2059 my $subrs_cur = $rs->cursor;
2060 my @all_pk = $subrs_cur->all;
2061 for my $pks ( @all_pk) {
2064 for my $i (0.. $#pcols) {
2065 $cond->{$pcols[$i]} = $pks->[$i];
2070 $op eq 'update' ? $values : (),
2084 $self->_execute($self->_select_args(@_));
2087 sub _select_args_to_query {
2090 # my ($op, $ident, $select, $cond, $rs_attrs, $rows, $offset)
2091 # = $self->_select_args($ident, $select, $cond, $attrs);
2092 my ($op, $ident, @args) =
2093 $self->_select_args(@_);
2095 # my ($sql, $prepared_bind) = $self->_gen_sql_bind($op, $ident, [ $select, $cond, $rs_attrs, $rows, $offset ]);
2096 my ($sql, $prepared_bind) = $self->_gen_sql_bind($op, $ident, \@args);
2097 $prepared_bind ||= [];
2100 ? ($sql, $prepared_bind)
2101 : \[ "($sql)", @$prepared_bind ]
2106 my ($self, $ident, $select, $where, $attrs) = @_;
2108 my $sql_maker = $self->sql_maker;
2109 my ($alias2source, $rs_alias) = $self->_resolve_ident_sources ($ident);
2116 $rs_alias && $alias2source->{$rs_alias}
2117 ? ( _rsroot_rsrc => $alias2source->{$rs_alias} )
2122 # Sanity check the attributes (SQLMaker does it too, but
2123 # in case of a software_limit we'll never reach there)
2124 if (defined $attrs->{offset}) {
2125 $self->throw_exception('A supplied offset attribute must be a non-negative integer')
2126 if ( $attrs->{offset} =~ /\D/ or $attrs->{offset} < 0 );
2129 if (defined $attrs->{rows}) {
2130 $self->throw_exception("The rows attribute must be a positive integer if present")
2131 if ( $attrs->{rows} =~ /\D/ or $attrs->{rows} <= 0 );
2133 elsif ($attrs->{offset}) {
2134 # MySQL actually recommends this approach. I cringe.
2135 $attrs->{rows} = $sql_maker->__max_int;
2140 # see if we need to tear the prefetch apart otherwise delegate the limiting to the
2141 # storage, unless software limit was requested
2144 ( $attrs->{rows} && keys %{$attrs->{collapse}} )
2146 # grouped prefetch (to satisfy group_by == select)
2147 ( $attrs->{group_by}
2149 @{$attrs->{group_by}}
2151 $attrs->{_prefetch_selector_range}
2154 ($ident, $select, $where, $attrs)
2155 = $self->_adjust_select_args_for_complex_prefetch ($ident, $select, $where, $attrs);
2157 elsif (! $attrs->{software_limit} ) {
2159 $attrs->{rows} || (),
2160 $attrs->{offset} || (),
2164 # try to simplify the joinmap further (prune unreferenced type-single joins)
2165 $ident = $self->_prune_unused_joins ($ident, $select, $where, $attrs);
2168 # This would be the point to deflate anything found in $where
2169 # (and leave $attrs->{bind} intact). Problem is - inflators historically
2170 # expect a row object. And all we have is a resultsource (it is trivial
2171 # to extract deflator coderefs via $alias2source above).
2173 # I don't see a way forward other than changing the way deflators are
2174 # invoked, and that's just bad...
2177 return ('select', $ident, $select, $where, $attrs, @limit);
2180 # Returns a counting SELECT for a simple count
2181 # query. Abstracted so that a storage could override
2182 # this to { count => 'firstcol' } or whatever makes
2183 # sense as a performance optimization
2185 #my ($self, $source, $rs_attrs) = @_;
2186 return { count => '*' };
2189 sub source_bind_attributes {
2190 shift->throw_exception(
2191 'source_bind_attributes() was never meant to be a callable public method - '
2192 .'please contact the DBIC dev-team and describe your use case so that a reasonable '
2193 .'solution can be provided'
2194 ."\nhttp://search.cpan.org/dist/DBIx-Class/lib/DBIx/Class.pm#GETTING_HELP/SUPPORT"
2202 =item Arguments: $ident, $select, $condition, $attrs
2206 Handle a SQL select statement.
2212 my ($ident, $select, $condition, $attrs) = @_;
2213 return $self->cursor_class->new($self, \@_, $attrs);
2218 my ($rv, $sth, @bind) = $self->_select(@_);
2219 my @row = $sth->fetchrow_array;
2220 my @nextrow = $sth->fetchrow_array if @row;
2221 if(@row && @nextrow) {
2222 carp "Query returned more than one row. SQL that returns multiple rows is DEPRECATED for ->find and ->single";
2224 # Need to call finish() to work round broken DBDs
2229 =head2 sql_limit_dialect
2231 This is an accessor for the default SQL limit dialect used by a particular
2232 storage driver. Can be overridden by supplying an explicit L</limit_dialect>
2233 to L<DBIx::Class::Schema/connect>. For a list of available limit dialects
2234 see L<DBIx::Class::SQLMaker::LimitDialects>.
2240 =item Arguments: $sql
2244 Returns a L<DBI> sth (statement handle) for the supplied SQL.
2249 my ($self, $dbh, $sql) = @_;
2251 # 3 is the if_active parameter which avoids active sth re-use
2252 my $sth = $self->disable_sth_caching
2253 ? $dbh->prepare($sql)
2254 : $dbh->prepare_cached($sql, {}, 3);
2256 # XXX You would think RaiseError would make this impossible,
2257 # but apparently that's not true :(
2258 $self->throw_exception(
2261 sprintf( "\$dbh->prepare() of '%s' through %s failed *silently* without "
2262 .'an exception and/or setting $dbh->errstr',
2264 ? substr($sql, 0, 20) . '...'
2267 'DBD::' . $dbh->{Driver}{Name},
2275 carp_unique 'sth was mistakenly marked/documented as public, stop calling it (will be removed before DBIC v0.09)';
2280 my ($self, $sql) = @_;
2281 $self->dbh_do('_dbh_sth', $sql); # retry over disconnects
2284 sub _dbh_columns_info_for {
2285 my ($self, $dbh, $table) = @_;
2287 if ($dbh->can('column_info')) {
2291 my ($schema,$tab) = $table =~ /^(.+?)\.(.+)$/ ? ($1,$2) : (undef,$table);
2292 my $sth = $dbh->column_info( undef,$schema, $tab, '%' );
2294 while ( my $info = $sth->fetchrow_hashref() ){
2296 $column_info{data_type} = $info->{TYPE_NAME};
2297 $column_info{size} = $info->{COLUMN_SIZE};
2298 $column_info{is_nullable} = $info->{NULLABLE} ? 1 : 0;
2299 $column_info{default_value} = $info->{COLUMN_DEF};
2300 my $col_name = $info->{COLUMN_NAME};
2301 $col_name =~ s/^\"(.*)\"$/$1/;
2303 $result{$col_name} = \%column_info;
2308 return \%result if !$caught && scalar keys %result;
2312 my $sth = $dbh->prepare($self->sql_maker->select($table, undef, \'1 = 0'));
2314 my @columns = @{$sth->{NAME_lc}};
2315 for my $i ( 0 .. $#columns ){
2317 $column_info{data_type} = $sth->{TYPE}->[$i];
2318 $column_info{size} = $sth->{PRECISION}->[$i];
2319 $column_info{is_nullable} = $sth->{NULLABLE}->[$i] ? 1 : 0;
2321 if ($column_info{data_type} =~ m/^(.*?)\((.*?)\)$/) {
2322 $column_info{data_type} = $1;
2323 $column_info{size} = $2;
2326 $result{$columns[$i]} = \%column_info;
2330 foreach my $col (keys %result) {
2331 my $colinfo = $result{$col};
2332 my $type_num = $colinfo->{data_type};
2334 if(defined $type_num && $dbh->can('type_info')) {
2335 my $type_info = $dbh->type_info($type_num);
2336 $type_name = $type_info->{TYPE_NAME} if $type_info;
2337 $colinfo->{data_type} = $type_name if $type_name;
2344 sub columns_info_for {
2345 my ($self, $table) = @_;
2346 $self->_dbh_columns_info_for ($self->_get_dbh, $table);
2349 =head2 last_insert_id
2351 Return the row id of the last insert.
2355 sub _dbh_last_insert_id {
2356 my ($self, $dbh, $source, $col) = @_;
2358 my $id = try { $dbh->last_insert_id (undef, undef, $source->name, $col) };
2360 return $id if defined $id;
2362 my $class = ref $self;
2363 $self->throw_exception ("No storage specific _dbh_last_insert_id() method implemented in $class, and the generic DBI::last_insert_id() failed");
2366 sub last_insert_id {
2368 $self->_dbh_last_insert_id ($self->_dbh, @_);
2371 =head2 _native_data_type
2375 =item Arguments: $type_name
2379 This API is B<EXPERIMENTAL>, will almost definitely change in the future, and
2380 currently only used by L<::AutoCast|DBIx::Class::Storage::DBI::AutoCast> and
2381 L<::Sybase::ASE|DBIx::Class::Storage::DBI::Sybase::ASE>.
2383 The default implementation returns C<undef>, implement in your Storage driver if
2384 you need this functionality.
2386 Should map types from other databases to the native RDBMS type, for example
2387 C<VARCHAR2> to C<VARCHAR>.
2389 Types with modifiers should map to the underlying data type. For example,
2390 C<INTEGER AUTO_INCREMENT> should become C<INTEGER>.
2392 Composite types should map to the container type, for example
2393 C<ENUM(foo,bar,baz)> becomes C<ENUM>.
2397 sub _native_data_type {
2398 #my ($self, $data_type) = @_;
2402 # Check if placeholders are supported at all
2403 sub _determine_supports_placeholders {
2405 my $dbh = $self->_get_dbh;
2407 # some drivers provide a $dbh attribute (e.g. Sybase and $dbh->{syb_dynamic_supported})
2408 # but it is inaccurate more often than not
2410 local $dbh->{PrintError} = 0;
2411 local $dbh->{RaiseError} = 1;
2412 $dbh->do('select ?', {}, 1);
2420 # Check if placeholders bound to non-string types throw exceptions
2422 sub _determine_supports_typeless_placeholders {
2424 my $dbh = $self->_get_dbh;
2427 local $dbh->{PrintError} = 0;
2428 local $dbh->{RaiseError} = 1;
2429 # this specifically tests a bind that is NOT a string
2430 $dbh->do('select 1 where 1 = ?', {}, 1);
2440 Returns the database driver name.
2445 shift->_get_dbh->{Driver}->{Name};
2448 =head2 bind_attribute_by_data_type
2450 Given a datatype from column info, returns a database specific bind
2451 attribute for C<< $dbh->bind_param($val,$attribute) >> or nothing if we will
2452 let the database planner just handle it.
2454 Generally only needed for special case column types, like bytea in postgres.
2458 sub bind_attribute_by_data_type {
2462 =head2 is_datatype_numeric
2464 Given a datatype from column_info, returns a boolean value indicating if
2465 the current RDBMS considers it a numeric value. This controls how
2466 L<DBIx::Class::Row/set_column> decides whether to mark the column as
2467 dirty - when the datatype is deemed numeric a C<< != >> comparison will
2468 be performed instead of the usual C<eq>.
2472 sub is_datatype_numeric {
2473 #my ($self, $dt) = @_;
2475 return 0 unless $_[1];
2478 numeric | int(?:eger)? | (?:tiny|small|medium|big)int | dec(?:imal)? | real | float | double (?: \s+ precision)? | (?:big)?serial
2483 =head2 create_ddl_dir
2487 =item Arguments: $schema \@databases, $version, $directory, $preversion, \%sqlt_args
2491 Creates a SQL file based on the Schema, for each of the specified
2492 database engines in C<\@databases> in the given directory.
2493 (note: specify L<SQL::Translator> names, not L<DBI> driver names).
2495 Given a previous version number, this will also create a file containing
2496 the ALTER TABLE statements to transform the previous schema into the
2497 current one. Note that these statements may contain C<DROP TABLE> or
2498 C<DROP COLUMN> statements that can potentially destroy data.
2500 The file names are created using the C<ddl_filename> method below, please
2501 override this method in your schema if you would like a different file
2502 name format. For the ALTER file, the same format is used, replacing
2503 $version in the name with "$preversion-$version".
2505 See L<SQL::Translator/METHODS> for a list of values for C<\%sqlt_args>.
2506 The most common value for this would be C<< { add_drop_table => 1 } >>
2507 to have the SQL produced include a C<DROP TABLE> statement for each table
2508 created. For quoting purposes supply C<quote_table_names> and
2509 C<quote_field_names>.
2511 If no arguments are passed, then the following default values are assumed:
2515 =item databases - ['MySQL', 'SQLite', 'PostgreSQL']
2517 =item version - $schema->schema_version
2519 =item directory - './'
2521 =item preversion - <none>
2525 By default, C<\%sqlt_args> will have
2527 { add_drop_table => 1, ignore_constraint_names => 1, ignore_index_names => 1 }
2529 merged with the hash passed in. To disable any of those features, pass in a
2530 hashref like the following
2532 { ignore_constraint_names => 0, # ... other options }
2535 WARNING: You are strongly advised to check all SQL files created, before applying
2540 sub create_ddl_dir {
2541 my ($self, $schema, $databases, $version, $dir, $preversion, $sqltargs) = @_;
2544 carp "No directory given, using ./\n";
2549 (require File::Path and File::Path::make_path ("$dir")) # make_path does not like objects (i.e. Path::Class::Dir)
2551 $self->throw_exception(
2552 "Failed to create '$dir': " . ($! || $@ || 'error unknown')
2556 $self->throw_exception ("Directory '$dir' does not exist\n") unless(-d $dir);
2558 $databases ||= ['MySQL', 'SQLite', 'PostgreSQL'];
2559 $databases = [ $databases ] if(ref($databases) ne 'ARRAY');
2561 my $schema_version = $schema->schema_version || '1.x';
2562 $version ||= $schema_version;
2565 add_drop_table => 1,
2566 ignore_constraint_names => 1,
2567 ignore_index_names => 1,
2571 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy')) {
2572 $self->throw_exception("Can't create a ddl file without " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2575 my $sqlt = SQL::Translator->new( $sqltargs );
2577 $sqlt->parser('SQL::Translator::Parser::DBIx::Class');
2578 my $sqlt_schema = $sqlt->translate({ data => $schema })
2579 or $self->throw_exception ($sqlt->error);
2581 foreach my $db (@$databases) {
2583 $sqlt->{schema} = $sqlt_schema;
2584 $sqlt->producer($db);
2587 my $filename = $schema->ddl_filename($db, $version, $dir);
2588 if (-e $filename && ($version eq $schema_version )) {
2589 # if we are dumping the current version, overwrite the DDL
2590 carp "Overwriting existing DDL file - $filename";
2594 my $output = $sqlt->translate;
2596 carp("Failed to translate to $db, skipping. (" . $sqlt->error . ")");
2599 if(!open($file, ">$filename")) {
2600 $self->throw_exception("Can't open $filename for writing ($!)");
2603 print $file $output;
2606 next unless ($preversion);
2608 require SQL::Translator::Diff;
2610 my $prefilename = $schema->ddl_filename($db, $preversion, $dir);
2611 if(!-e $prefilename) {
2612 carp("No previous schema file found ($prefilename)");
2616 my $difffile = $schema->ddl_filename($db, $version, $dir, $preversion);
2618 carp("Overwriting existing diff file - $difffile");
2624 my $t = SQL::Translator->new($sqltargs);
2629 or $self->throw_exception ($t->error);
2631 my $out = $t->translate( $prefilename )
2632 or $self->throw_exception ($t->error);
2634 $source_schema = $t->schema;
2636 $source_schema->name( $prefilename )
2637 unless ( $source_schema->name );
2640 # The "new" style of producers have sane normalization and can support
2641 # diffing a SQL file against a DBIC->SQLT schema. Old style ones don't
2642 # And we have to diff parsed SQL against parsed SQL.
2643 my $dest_schema = $sqlt_schema;
2645 unless ( "SQL::Translator::Producer::$db"->can('preprocess_schema') ) {
2646 my $t = SQL::Translator->new($sqltargs);
2651 or $self->throw_exception ($t->error);
2653 my $out = $t->translate( $filename )
2654 or $self->throw_exception ($t->error);
2656 $dest_schema = $t->schema;
2658 $dest_schema->name( $filename )
2659 unless $dest_schema->name;
2662 my $diff = SQL::Translator::Diff::schema_diff($source_schema, $db,
2666 if(!open $file, ">$difffile") {
2667 $self->throw_exception("Can't write to $difffile ($!)");
2675 =head2 deployment_statements
2679 =item Arguments: $schema, $type, $version, $directory, $sqlt_args
2683 Returns the statements used by L</deploy> and L<DBIx::Class::Schema/deploy>.
2685 The L<SQL::Translator> (not L<DBI>) database driver name can be explicitly
2686 provided in C<$type>, otherwise the result of L</sqlt_type> is used as default.
2688 C<$directory> is used to return statements from files in a previously created
2689 L</create_ddl_dir> directory and is optional. The filenames are constructed
2690 from L<DBIx::Class::Schema/ddl_filename>, the schema name and the C<$version>.
2692 If no C<$directory> is specified then the statements are constructed on the
2693 fly using L<SQL::Translator> and C<$version> is ignored.
2695 See L<SQL::Translator/METHODS> for a list of values for C<$sqlt_args>.
2699 sub deployment_statements {
2700 my ($self, $schema, $type, $version, $dir, $sqltargs) = @_;
2701 $type ||= $self->sqlt_type;
2702 $version ||= $schema->schema_version || '1.x';
2704 my $filename = $schema->ddl_filename($type, $version, $dir);
2707 # FIXME replace this block when a proper sane sql parser is available
2709 open($file, "<$filename")
2710 or $self->throw_exception("Can't open $filename ($!)");
2713 return join('', @rows);
2716 unless (DBIx::Class::Optional::Dependencies->req_ok_for ('deploy') ) {
2717 $self->throw_exception("Can't deploy without a ddl_dir or " . DBIx::Class::Optional::Dependencies->req_missing_for ('deploy') );
2720 # sources needs to be a parser arg, but for simplicty allow at top level
2722 $sqltargs->{parser_args}{sources} = delete $sqltargs->{sources}
2723 if exists $sqltargs->{sources};
2725 my $tr = SQL::Translator->new(
2726 producer => "SQL::Translator::Producer::${type}",
2728 parser => 'SQL::Translator::Parser::DBIx::Class',
2734 @ret = $tr->translate;
2737 $ret[0] = $tr->translate;
2740 $self->throw_exception( 'Unable to produce deployment statements: ' . $tr->error)
2741 unless (@ret && defined $ret[0]);
2743 return wantarray ? @ret : $ret[0];
2746 # FIXME deploy() currently does not accurately report sql errors
2747 # Will always return true while errors are warned
2749 my ($self, $schema, $type, $sqltargs, $dir) = @_;
2753 return if($line =~ /^--/);
2754 # next if($line =~ /^DROP/m);
2755 return if($line =~ /^BEGIN TRANSACTION/m);
2756 return if($line =~ /^COMMIT/m);
2757 return if $line =~ /^\s+$/; # skip whitespace only
2758 $self->_query_start($line);
2760 # do a dbh_do cycle here, as we need some error checking in
2761 # place (even though we will ignore errors)
2762 $self->dbh_do (sub { $_[1]->do($line) });
2764 carp qq{$_ (running "${line}")};
2766 $self->_query_end($line);
2768 my @statements = $schema->deployment_statements($type, undef, $dir, { %{ $sqltargs || {} }, no_comments => 1 } );
2769 if (@statements > 1) {
2770 foreach my $statement (@statements) {
2771 $deploy->( $statement );
2774 elsif (@statements == 1) {
2775 # split on single line comments and end of statements
2776 foreach my $line ( split(/\s*--.*\n|;\n/, $statements[0])) {
2782 =head2 datetime_parser
2784 Returns the datetime parser class
2788 sub datetime_parser {
2790 return $self->{datetime_parser} ||= do {
2791 $self->build_datetime_parser(@_);
2795 =head2 datetime_parser_type
2797 Defines the datetime parser class - currently defaults to L<DateTime::Format::MySQL>
2799 =head2 build_datetime_parser
2801 See L</datetime_parser>
2805 sub build_datetime_parser {
2807 my $type = $self->datetime_parser_type(@_);
2812 =head2 is_replicating
2814 A boolean that reports if a particular L<DBIx::Class::Storage::DBI> is set to
2815 replicate from a master database. Default is undef, which is the result
2816 returned by databases that don't support replication.
2820 sub is_replicating {
2825 =head2 lag_behind_master
2827 Returns a number that represents a certain amount of lag behind a master db
2828 when a given storage is replicating. The number is database dependent, but
2829 starts at zero and increases with the amount of lag. Default in undef
2833 sub lag_behind_master {
2837 =head2 relname_to_table_alias
2841 =item Arguments: $relname, $join_count
2845 L<DBIx::Class> uses L<DBIx::Class::Relationship> names as table aliases in
2848 This hook is to allow specific L<DBIx::Class::Storage> drivers to change the
2849 way these aliases are named.
2851 The default behavior is C<< "$relname_$join_count" if $join_count > 1 >>,
2852 otherwise C<"$relname">.
2856 sub relname_to_table_alias {
2857 my ($self, $relname, $join_count) = @_;
2859 my $alias = ($join_count && $join_count > 1 ?
2860 join('_', $relname, $join_count) : $relname);
2865 # The size in bytes to use for DBI's ->bind_param_inout, this is the generic
2866 # version and it may be necessary to amend or override it for a specific storage
2867 # if such binds are necessary.
2868 sub _max_column_bytesize {
2869 my ($self, $attr) = @_;
2873 if ($attr->{sqlt_datatype}) {
2874 my $data_type = lc($attr->{sqlt_datatype});
2876 if ($attr->{sqlt_size}) {
2878 # String/sized-binary types
2879 if ($data_type =~ /^(?:
2880 l? (?:var)? char(?:acter)? (?:\s*varying)?
2882 (?:var)? binary (?:\s*varying)?
2887 $max_size = $attr->{sqlt_size};
2889 # Other charset/unicode types, assume scale of 4
2890 elsif ($data_type =~ /^(?:
2891 national \s* character (?:\s*varying)?
2900 $max_size = $attr->{sqlt_size} * 4;
2904 if (!$max_size and !$self->_is_lob_type($data_type)) {
2905 $max_size = 100 # for all other (numeric?) datatypes
2909 $max_size || $self->_dbic_connect_attributes->{LongReadLen} || $self->_get_dbh->{LongReadLen} || 8000;
2912 # Determine if a data_type is some type of BLOB
2914 my ($self, $data_type) = @_;
2915 $data_type && ($data_type =~ /lob|bfile|text|image|bytea|memo/i
2916 || $data_type =~ /^long(?:\s+(?:raw|bit\s*varying|varbit|binary
2917 |varchar|character\s*varying|nvarchar
2918 |national\s*character\s*varying))?\z/xi);
2921 sub _is_binary_lob_type {
2922 my ($self, $data_type) = @_;
2923 $data_type && ($data_type =~ /blob|bfile|image|bytea/i
2924 || $data_type =~ /^long(?:\s+(?:raw|bit\s*varying|varbit|binary))?\z/xi);
2927 sub _is_text_lob_type {
2928 my ($self, $data_type) = @_;
2929 $data_type && ($data_type =~ /^(?:clob|memo)\z/i
2930 || $data_type =~ /^long(?:\s+(?:varchar|character\s*varying|nvarchar
2931 |national\s*character\s*varying))\z/xi);
2938 =head2 DBIx::Class and AutoCommit
2940 DBIx::Class can do some wonderful magic with handling exceptions,
2941 disconnections, and transactions when you use C<< AutoCommit => 1 >>
2942 (the default) combined with L<txn_do|DBIx::Class::Storage/txn_do> for
2943 transaction support.
2945 If you set C<< AutoCommit => 0 >> in your connect info, then you are always
2946 in an assumed transaction between commits, and you're telling us you'd
2947 like to manage that manually. A lot of the magic protections offered by
2948 this module will go away. We can't protect you from exceptions due to database
2949 disconnects because we don't know anything about how to restart your
2950 transactions. You're on your own for handling all sorts of exceptional
2951 cases if you choose the C<< AutoCommit => 0 >> path, just as you would
2957 Matt S. Trout <mst@shadowcatsystems.co.uk>
2959 Andy Grundman <andy@hybridized.org>
2963 You may distribute this code under the same terms as Perl itself.