1 package DBIx::Class::Storage::DBI;
2 # -*- mode: cperl; cperl-indent-level: 2 -*-
4 use base 'DBIx::Class::Storage';
8 use Carp::Clan qw/^DBIx::Class/;
10 use SQL::Abstract::Limit;
11 use DBIx::Class::Storage::DBI::Cursor;
12 use DBIx::Class::Storage::Statistics;
13 use Scalar::Util qw/blessed weaken/;
15 __PACKAGE__->mk_group_accessors('simple' =>
16 qw/_connect_info _dbi_connect_info _dbh _sql_maker _sql_maker_opts
17 _conn_pid _conn_tid transaction_depth _dbh_autocommit savepoints/
20 # the values for these accessors are picked out (and deleted) from
21 # the attribute hashref passed to connect_info
22 my @storage_options = qw/
23 on_connect_do on_disconnect_do disable_sth_caching unsafe auto_savepoint
25 __PACKAGE__->mk_group_accessors('simple' => @storage_options);
28 # default cursor class, overridable in connect_info attributes
29 __PACKAGE__->cursor_class('DBIx::Class::Storage::DBI::Cursor');
31 __PACKAGE__->mk_group_accessors('inherited' => qw/sql_maker_class/);
32 __PACKAGE__->sql_maker_class('DBIC::SQL::Abstract');
36 package # Hide from PAUSE
37 DBIC::SQL::Abstract; # Would merge upstream, but nate doesn't reply :(
39 use base qw/SQL::Abstract::Limit/;
41 # This prevents the caching of $dbh in S::A::L, I believe
43 my $self = shift->SUPER::new(@_);
45 # If limit_dialect is a ref (like a $dbh), go ahead and replace
46 # it with what it resolves to:
47 $self->{limit_dialect} = $self->_find_syntax($self->{limit_dialect})
48 if ref $self->{limit_dialect};
54 my ($self, $sql, $order, $rows, $offset ) = @_;
57 my $last = $rows + $offset;
58 my ( $order_by ) = $self->_order_by( $order );
63 SELECT Q1.*, ROW_NUMBER() OVER( ) AS ROW_NUM FROM (
68 WHERE ROW_NUM BETWEEN $offset AND $last
74 # While we're at it, this should make LIMIT queries more efficient,
75 # without digging into things too deeply
76 use Scalar::Util 'blessed';
78 my ($self, $syntax) = @_;
79 my $dbhname = blessed($syntax) ? $syntax->{Driver}{Name} : $syntax;
80 if(ref($self) && $dbhname && $dbhname eq 'DB2') {
81 return 'RowNumberOver';
84 $self->{_cached_syntax} ||= $self->SUPER::_find_syntax($syntax);
88 my ($self, $table, $fields, $where, $order, @rest) = @_;
89 $table = $self->_quote($table) unless ref($table);
90 local $self->{rownum_hack_count} = 1
91 if (defined $rest[0] && $self->{limit_dialect} eq 'RowNum');
92 @rest = (-1) unless defined $rest[0];
93 die "LIMIT 0 Does Not Compute" if $rest[0] == 0;
94 # and anyway, SQL::Abstract::Limit will cause a barf if we don't first
95 local $self->{having_bind} = [];
96 my ($sql, @ret) = $self->SUPER::select(
97 $table, $self->_recurse_fields($fields), $where, $order, @rest
102 $self->{for} eq 'update' ? ' FOR UPDATE' :
103 $self->{for} eq 'shared' ? ' FOR SHARE' :
108 return wantarray ? ($sql, @ret, @{$self->{having_bind}}) : $sql;
114 $table = $self->_quote($table) unless ref($table);
115 $self->SUPER::insert($table, @_);
121 $table = $self->_quote($table) unless ref($table);
122 $self->SUPER::update($table, @_);
128 $table = $self->_quote($table) unless ref($table);
129 $self->SUPER::delete($table, @_);
135 return $_[1].$self->_order_by($_[2]);
137 return $self->SUPER::_emulate_limit(@_);
141 sub _recurse_fields {
142 my ($self, $fields, $params) = @_;
143 my $ref = ref $fields;
144 return $self->_quote($fields) unless $ref;
145 return $$fields if $ref eq 'SCALAR';
147 if ($ref eq 'ARRAY') {
148 return join(', ', map {
149 $self->_recurse_fields($_)
150 .(exists $self->{rownum_hack_count} && !($params && $params->{no_rownum_hack})
151 ? ' AS col'.$self->{rownum_hack_count}++
154 } elsif ($ref eq 'HASH') {
155 foreach my $func (keys %$fields) {
156 return $self->_sqlcase($func)
157 .'( '.$self->_recurse_fields($fields->{$func}).' )';
166 if (ref $_[0] eq 'HASH') {
167 if (defined $_[0]->{group_by}) {
168 $ret = $self->_sqlcase(' group by ')
169 .$self->_recurse_fields($_[0]->{group_by}, { no_rownum_hack => 1 });
171 if (defined $_[0]->{having}) {
173 ($frag, @extra) = $self->_recurse_where($_[0]->{having});
174 push(@{$self->{having_bind}}, @extra);
175 $ret .= $self->_sqlcase(' having ').$frag;
177 if (defined $_[0]->{order_by}) {
178 $ret .= $self->_order_by($_[0]->{order_by});
180 } elsif (ref $_[0] eq 'SCALAR') {
181 $ret = $self->_sqlcase(' order by ').${ $_[0] };
182 } elsif (ref $_[0] eq 'ARRAY' && @{$_[0]}) {
183 my @order = @{+shift};
184 $ret = $self->_sqlcase(' order by ')
186 my $r = $self->_order_by($_, @_);
187 $r =~ s/^ ?ORDER BY //i;
191 $ret = $self->SUPER::_order_by(@_);
196 sub _order_directions {
197 my ($self, $order) = @_;
198 $order = $order->{order_by} if ref $order eq 'HASH';
199 return $self->SUPER::_order_directions($order);
203 my ($self, $from) = @_;
204 if (ref $from eq 'ARRAY') {
205 return $self->_recurse_from(@$from);
206 } elsif (ref $from eq 'HASH') {
207 return $self->_make_as($from);
209 return $from; # would love to quote here but _table ends up getting called
210 # twice during an ->select without a limit clause due to
211 # the way S::A::Limit->select works. should maybe consider
212 # bypassing this and doing S::A::select($self, ...) in
213 # our select method above. meantime, quoting shims have
214 # been added to select/insert/update/delete here
219 my ($self, $from, @join) = @_;
221 push(@sqlf, $self->_make_as($from));
222 foreach my $j (@join) {
225 # check whether a join type exists
226 my $join_clause = '';
227 my $to_jt = ref($to) eq 'ARRAY' ? $to->[0] : $to;
228 if (ref($to_jt) eq 'HASH' and exists($to_jt->{-join_type})) {
229 $join_clause = ' '.uc($to_jt->{-join_type}).' JOIN ';
231 $join_clause = ' JOIN ';
233 push(@sqlf, $join_clause);
235 if (ref $to eq 'ARRAY') {
236 push(@sqlf, '(', $self->_recurse_from(@$to), ')');
238 push(@sqlf, $self->_make_as($to));
240 push(@sqlf, ' ON ', $self->_join_condition($on));
242 return join('', @sqlf);
246 my ($self, $from) = @_;
247 return join(' ', map { (ref $_ eq 'SCALAR' ? $$_ : $self->_quote($_)) }
248 reverse each %{$self->_skip_options($from)});
252 my ($self, $hash) = @_;
254 $clean_hash->{$_} = $hash->{$_}
255 for grep {!/^-/} keys %$hash;
259 sub _join_condition {
260 my ($self, $cond) = @_;
261 if (ref $cond eq 'HASH') {
266 # XXX no throw_exception() in this package and croak() fails with strange results
267 Carp::croak(ref($v) . qq{ reference arguments are not supported in JOINS - try using \"..." instead'})
268 if ref($v) ne 'SCALAR';
272 my $x = '= '.$self->_quote($v); $j{$_} = \$x;
275 return scalar($self->_recurse_where(\%j));
276 } elsif (ref $cond eq 'ARRAY') {
277 return join(' OR ', map { $self->_join_condition($_) } @$cond);
279 die "Can't handle this yet!";
284 my ($self, $label) = @_;
285 return '' unless defined $label;
286 return "*" if $label eq '*';
287 return $label unless $self->{quote_char};
288 if(ref $self->{quote_char} eq "ARRAY"){
289 return $self->{quote_char}->[0] . $label . $self->{quote_char}->[1]
290 if !defined $self->{name_sep};
291 my $sep = $self->{name_sep};
292 return join($self->{name_sep},
293 map { $self->{quote_char}->[0] . $_ . $self->{quote_char}->[1] }
294 split(/\Q$sep\E/,$label));
296 return $self->SUPER::_quote($label);
301 $self->{limit_dialect} = shift if @_;
302 return $self->{limit_dialect};
307 $self->{quote_char} = shift if @_;
308 return $self->{quote_char};
313 $self->{name_sep} = shift if @_;
314 return $self->{name_sep};
317 } # End of BEGIN block
321 DBIx::Class::Storage::DBI - DBI storage handler
325 my $schema = MySchema->connect('dbi:SQLite:my.db');
327 $schema->storage->debug(1);
328 $schema->dbh_do("DROP TABLE authors");
330 $schema->resultset('Book')->search({
331 written_on => $schema->storage->datetime_parser(DateTime->now)
336 This class represents the connection to an RDBMS via L<DBI>. See
337 L<DBIx::Class::Storage> for general information. This pod only
338 documents DBI-specific methods and behaviors.
345 my $new = shift->next::method(@_);
347 $new->transaction_depth(0);
348 $new->_sql_maker_opts({});
349 $new->{savepoints} = [];
350 $new->{_in_dbh_do} = 0;
351 $new->{_dbh_gen} = 0;
358 This method is normally called by L<DBIx::Class::Schema/connection>, which
359 encapsulates its argument list in an arrayref before passing them here.
361 The argument list may contain:
367 The same 4-element argument set one would normally pass to
368 L<DBI/connect>, optionally followed by L<extra attributes|/DBIx::Class
369 specific connection attributes> recognized by DBIx::Class:
371 $connect_info_args = [ $dsn, $user, $password, \%dbi_attributes?, \%extra_attributes? ];
375 A single code reference which returns a connected L<DBI database
376 handle|DBI/connect> optionally followed by L<extra
377 attributes|/DBIx::Class specific connection attributes> recognized by
380 $connect_info_args = [ sub { DBI->connect (...) }, \%extra_attributes? ];
384 A single hashref with all the attributes and the dsn/user/password
387 $connect_info_args = [{
395 This is particularly useful for L<Catalyst> based applications, allowing the
396 following config (in L<Config::General> style):
401 dsn dbi:mysql:database=test
410 Please note that the L<DBI> docs recommend that you always explicitly
411 set C<AutoCommit> to either I<0> or I<1>. L<DBIx::Class> further
412 recommends that it be set to I<1>, and that you perform transactions
413 via our L</txn_do> method. L<DBIx::Class> will set it to I<1> if you
414 do not do explicitly set it to zero. This is the default for most
415 DBDs. See L</DBIx::Class and AutoCommit> for details.
417 =head3 DBIx::Class specific connection attributes
419 In addition to the standard L<DBI|DBI/ATTRIBUTES_COMMON_TO_ALL_HANDLES>
420 L<connection|DBI/Database_Handle_Attributes> attributes, DBIx::Class recognizes
421 the following connection options. These options can be mixed in with your other
422 L<DBI> connection attributes, or placed in a seperate hashref
423 (C<\%extra_attributes>) as shown above.
425 Every time C<connect_info> is invoked, any previous settings for
426 these options will be cleared before setting the new ones, regardless of
427 whether any options are specified in the new C<connect_info>.
434 Specifies things to do immediately after connecting or re-connecting to
435 the database. Its value may contain:
439 =item an array reference
441 This contains SQL statements to execute in order. Each element contains
442 a string or a code reference that returns a string.
444 =item a code reference
446 This contains some code to execute. Unlike code references within an
447 array reference, its return value is ignored.
451 =item on_disconnect_do
453 Takes arguments in the same form as L</on_connect_do> and executes them
454 immediately before disconnecting from the database.
456 Note, this only runs if you explicitly call L</disconnect> on the
459 =item disable_sth_caching
461 If set to a true value, this option will disable the caching of
462 statement handles via L<DBI/prepare_cached>.
466 Sets the limit dialect. This is useful for JDBC-bridge among others
467 where the remote SQL-dialect cannot be determined by the name of the
468 driver alone. See also L<SQL::Abstract::Limit>.
472 Specifies what characters to use to quote table and column names. If
473 you use this you will want to specify L</name_sep> as well.
475 C<quote_char> expects either a single character, in which case is it
476 is placed on either side of the table/column name, or an arrayref of length
477 2 in which case the table/column name is placed between the elements.
479 For example under MySQL you should use C<< quote_char => '`' >>, and for
480 SQL Server you should use C<< quote_char => [qw/[ ]/] >>.
484 This only needs to be used in conjunction with L<quote_char>, and is used to
485 specify the charecter that seperates elements (schemas, tables, columns) from
486 each other. In most cases this is simply a C<.>.
488 The consequences of not supplying this value is that L<SQL::Abstract>
489 will assume DBIx::Class' uses of aliases to be complete column
490 names. The output will look like I<"me.name"> when it should actually
495 This Storage driver normally installs its own C<HandleError>, sets
496 C<RaiseError> and C<ShowErrorStatement> on, and sets C<PrintError> off on
497 all database handles, including those supplied by a coderef. It does this
498 so that it can have consistent and useful error behavior.
500 If you set this option to a true value, Storage will not do its usual
501 modifications to the database handle's attributes, and instead relies on
502 the settings in your connect_info DBI options (or the values you set in
503 your connection coderef, in the case that you are connecting via coderef).
505 Note that your custom settings can cause Storage to malfunction,
506 especially if you set a C<HandleError> handler that suppresses exceptions
507 and/or disable C<RaiseError>.
511 If this option is true, L<DBIx::Class> will use savepoints when nesting
512 transactions, making it possible to recover from failure in the inner
513 transaction without having to abort all outer transactions.
517 Use this argument to supply a cursor class other than the default
518 L<DBIx::Class::Storage::DBI::Cursor>.
522 Some real-life examples of arguments to L</connect_info> and
523 L<DBIx::Class::Schema/connect>
525 # Simple SQLite connection
526 ->connect_info([ 'dbi:SQLite:./foo.db' ]);
529 ->connect_info([ sub { DBI->connect(...) } ]);
531 # A bit more complicated
538 { quote_char => q{"}, name_sep => q{.} },
542 # Equivalent to the previous example
548 { AutoCommit => 1, quote_char => q{"}, name_sep => q{.} },
552 # Same, but with hashref as argument
553 # See parse_connect_info for explanation
556 dsn => 'dbi:Pg:dbname=foo',
558 password => 'my_pg_password',
565 # Subref + DBIx::Class-specific connection options
568 sub { DBI->connect(...) },
572 on_connect_do => ['SET search_path TO myschema,otherschema,public'],
573 disable_sth_caching => 1,
583 my ($self, $info_arg) = @_;
585 return $self->_connect_info if !$info_arg;
587 my @args = @$info_arg; # take a shallow copy for further mutilation
588 $self->_connect_info([@args]); # copy for _connect_info
591 # combine/pre-parse arguments depending on invocation style
594 if (ref $args[0] eq 'CODE') { # coderef with optional \%extra_attributes
595 %attrs = %{ $args[1] || {} };
598 elsif (ref $args[0] eq 'HASH') { # single hashref (i.e. Catalyst config)
599 %attrs = %{$args[0]};
601 for (qw/password user dsn/) {
602 unshift @args, delete $attrs{$_};
605 else { # otherwise assume dsn/user/password + \%attrs + \%extra_attrs
607 % { $args[3] || {} },
608 % { $args[4] || {} },
610 @args = @args[0,1,2];
613 # Kill sql_maker/_sql_maker_opts, so we get a fresh one with only
614 # the new set of options
615 $self->_sql_maker(undef);
616 $self->_sql_maker_opts({});
619 for my $storage_opt (@storage_options, 'cursor_class') { # @storage_options is declared at the top of the module
620 if(my $value = delete $attrs{$storage_opt}) {
621 $self->$storage_opt($value);
624 for my $sql_maker_opt (qw/limit_dialect quote_char name_sep/) {
625 if(my $opt_val = delete $attrs{$sql_maker_opt}) {
626 $self->_sql_maker_opts->{$sql_maker_opt} = $opt_val;
631 %attrs = () if (ref $args[0] eq 'CODE'); # _connect() never looks past $args[0] in this case
633 $self->_dbi_connect_info([@args, keys %attrs ? \%attrs : ()]);
634 $self->_connect_info;
639 This method is deprecated in favour of setting via L</connect_info>.
644 Arguments: ($subref | $method_name), @extra_coderef_args?
646 Execute the given $subref or $method_name using the new exception-based
647 connection management.
649 The first two arguments will be the storage object that C<dbh_do> was called
650 on and a database handle to use. Any additional arguments will be passed
651 verbatim to the called subref as arguments 2 and onwards.
653 Using this (instead of $self->_dbh or $self->dbh) ensures correct
654 exception handling and reconnection (or failover in future subclasses).
656 Your subref should have no side-effects outside of the database, as
657 there is the potential for your subref to be partially double-executed
658 if the database connection was stale/dysfunctional.
662 my @stuff = $schema->storage->dbh_do(
664 my ($storage, $dbh, @cols) = @_;
665 my $cols = join(q{, }, @cols);
666 $dbh->selectrow_array("SELECT $cols FROM foo");
677 my $dbh = $self->_dbh;
679 return $self->$code($dbh, @_) if $self->{_in_dbh_do}
680 || $self->{transaction_depth};
682 local $self->{_in_dbh_do} = 1;
685 my $want_array = wantarray;
688 $self->_verify_pid if $dbh;
690 $self->_populate_dbh;
695 @result = $self->$code($dbh, @_);
697 elsif(defined $want_array) {
698 $result[0] = $self->$code($dbh, @_);
701 $self->$code($dbh, @_);
706 if(!$exception) { return $want_array ? @result : $result[0] }
708 $self->throw_exception($exception) if $self->connected;
710 # We were not connected - reconnect and retry, but let any
711 # exception fall right through this time
712 $self->_populate_dbh;
713 $self->$code($self->_dbh, @_);
716 # This is basically a blend of dbh_do above and DBIx::Class::Storage::txn_do.
717 # It also informs dbh_do to bypass itself while under the direction of txn_do,
718 # via $self->{_in_dbh_do} (this saves some redundant eval and errorcheck, etc)
723 ref $coderef eq 'CODE' or $self->throw_exception
724 ('$coderef must be a CODE reference');
726 return $coderef->(@_) if $self->{transaction_depth} && ! $self->auto_savepoint;
728 local $self->{_in_dbh_do} = 1;
731 my $want_array = wantarray;
736 $self->_verify_pid if $self->_dbh;
737 $self->_populate_dbh if !$self->_dbh;
741 @result = $coderef->(@_);
743 elsif(defined $want_array) {
744 $result[0] = $coderef->(@_);
753 if(!$exception) { return $want_array ? @result : $result[0] }
755 if($tried++ > 0 || $self->connected) {
756 eval { $self->txn_rollback };
757 my $rollback_exception = $@;
758 if($rollback_exception) {
759 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
760 $self->throw_exception($exception) # propagate nested rollback
761 if $rollback_exception =~ /$exception_class/;
763 $self->throw_exception(
764 "Transaction aborted: ${exception}. "
765 . "Rollback failed: ${rollback_exception}"
768 $self->throw_exception($exception)
771 # We were not connected, and was first try - reconnect and retry
773 $self->_populate_dbh;
779 Our C<disconnect> method also performs a rollback first if the
780 database is not in C<AutoCommit> mode.
787 if( $self->connected ) {
788 my $connection_do = $self->on_disconnect_do;
789 $self->_do_connection_actions($connection_do) if ref($connection_do);
791 $self->_dbh->rollback unless $self->_dbh_autocommit;
792 $self->_dbh->disconnect;
798 =head2 with_deferred_fk_checks
802 =item Arguments: C<$coderef>
804 =item Return Value: The return value of $coderef
808 Storage specific method to run the code ref with FK checks deferred or
809 in MySQL's case disabled entirely.
813 # Storage subclasses should override this
814 sub with_deferred_fk_checks {
815 my ($self, $sub) = @_;
823 if(my $dbh = $self->_dbh) {
824 if(defined $self->_conn_tid && $self->_conn_tid != threads->tid) {
831 return 0 if !$self->_dbh;
833 return ($dbh->FETCH('Active') && $dbh->ping);
839 # handle pid changes correctly
840 # NOTE: assumes $self->_dbh is a valid $dbh
844 return if defined $self->_conn_pid && $self->_conn_pid == $$;
846 $self->_dbh->{InactiveDestroy} = 1;
853 sub ensure_connected {
856 unless ($self->connected) {
857 $self->_populate_dbh;
863 Returns the dbh - a data base handle of class L<DBI>.
870 $self->ensure_connected;
874 sub _sql_maker_args {
877 return ( bindtype=>'columns', limit_dialect => $self->dbh, %{$self->_sql_maker_opts} );
882 unless ($self->_sql_maker) {
883 my $sql_maker_class = $self->sql_maker_class;
884 $self->_sql_maker($sql_maker_class->new( $self->_sql_maker_args ));
886 return $self->_sql_maker;
893 my @info = @{$self->_dbi_connect_info || []};
894 $self->_dbh($self->_connect(@info));
896 # Always set the transaction depth on connect, since
897 # there is no transaction in progress by definition
898 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
900 if(ref $self eq 'DBIx::Class::Storage::DBI') {
901 my $driver = $self->_dbh->{Driver}->{Name};
902 if ($self->load_optional_class("DBIx::Class::Storage::DBI::${driver}")) {
903 bless $self, "DBIx::Class::Storage::DBI::${driver}";
908 my $connection_do = $self->on_connect_do;
909 $self->_do_connection_actions($connection_do) if ref($connection_do);
911 $self->_conn_pid($$);
912 $self->_conn_tid(threads->tid) if $INC{'threads.pm'};
915 sub _do_connection_actions {
917 my $connection_do = shift;
919 if (ref $connection_do eq 'ARRAY') {
920 $self->_do_query($_) foreach @$connection_do;
922 elsif (ref $connection_do eq 'CODE') {
930 my ($self, $action) = @_;
932 if (ref $action eq 'CODE') {
933 $action = $action->($self);
934 $self->_do_query($_) foreach @$action;
937 my @to_run = (ref $action eq 'ARRAY') ? (@$action) : ($action);
938 $self->_query_start(@to_run);
939 $self->_dbh->do(@to_run);
940 $self->_query_end(@to_run);
947 my ($self, @info) = @_;
949 $self->throw_exception("You failed to provide any connection info")
952 my ($old_connect_via, $dbh);
954 if ($INC{'Apache/DBI.pm'} && $ENV{MOD_PERL}) {
955 $old_connect_via = $DBI::connect_via;
956 $DBI::connect_via = 'connect';
960 if(ref $info[0] eq 'CODE') {
964 $dbh = DBI->connect(@info);
967 if($dbh && !$self->unsafe) {
968 my $weak_self = $self;
970 $dbh->{HandleError} = sub {
972 $weak_self->throw_exception("DBI Exception: $_[0]");
975 croak ("DBI Exception: $_[0]");
978 $dbh->{ShowErrorStatement} = 1;
979 $dbh->{RaiseError} = 1;
980 $dbh->{PrintError} = 0;
984 $DBI::connect_via = $old_connect_via if $old_connect_via;
986 $self->throw_exception("DBI Connection failed: " . ($@||$DBI::errstr))
989 $self->_dbh_autocommit($dbh->{AutoCommit});
995 my ($self, $name) = @_;
997 $name = $self->_svp_generate_name
998 unless defined $name;
1000 $self->throw_exception ("You can't use savepoints outside a transaction")
1001 if $self->{transaction_depth} == 0;
1003 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1004 unless $self->can('_svp_begin');
1006 push @{ $self->{savepoints} }, $name;
1008 $self->debugobj->svp_begin($name) if $self->debug;
1010 return $self->_svp_begin($name);
1014 my ($self, $name) = @_;
1016 $self->throw_exception ("You can't use savepoints outside a transaction")
1017 if $self->{transaction_depth} == 0;
1019 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1020 unless $self->can('_svp_release');
1022 if (defined $name) {
1023 $self->throw_exception ("Savepoint '$name' does not exist")
1024 unless grep { $_ eq $name } @{ $self->{savepoints} };
1026 # Dig through the stack until we find the one we are releasing. This keeps
1027 # the stack up to date.
1030 do { $svp = pop @{ $self->{savepoints} } } while $svp ne $name;
1032 $name = pop @{ $self->{savepoints} };
1035 $self->debugobj->svp_release($name) if $self->debug;
1037 return $self->_svp_release($name);
1041 my ($self, $name) = @_;
1043 $self->throw_exception ("You can't use savepoints outside a transaction")
1044 if $self->{transaction_depth} == 0;
1046 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1047 unless $self->can('_svp_rollback');
1049 if (defined $name) {
1050 # If they passed us a name, verify that it exists in the stack
1051 unless(grep({ $_ eq $name } @{ $self->{savepoints} })) {
1052 $self->throw_exception("Savepoint '$name' does not exist!");
1055 # Dig through the stack until we find the one we are releasing. This keeps
1056 # the stack up to date.
1057 while(my $s = pop(@{ $self->{savepoints} })) {
1058 last if($s eq $name);
1060 # Add the savepoint back to the stack, as a rollback doesn't remove the
1061 # named savepoint, only everything after it.
1062 push(@{ $self->{savepoints} }, $name);
1064 # We'll assume they want to rollback to the last savepoint
1065 $name = $self->{savepoints}->[-1];
1068 $self->debugobj->svp_rollback($name) if $self->debug;
1070 return $self->_svp_rollback($name);
1073 sub _svp_generate_name {
1076 return 'savepoint_'.scalar(@{ $self->{'savepoints'} });
1081 $self->ensure_connected();
1082 if($self->{transaction_depth} == 0) {
1083 $self->debugobj->txn_begin()
1085 # this isn't ->_dbh-> because
1086 # we should reconnect on begin_work
1087 # for AutoCommit users
1088 $self->dbh->begin_work;
1089 } elsif ($self->auto_savepoint) {
1092 $self->{transaction_depth}++;
1097 if ($self->{transaction_depth} == 1) {
1098 my $dbh = $self->_dbh;
1099 $self->debugobj->txn_commit()
1102 $self->{transaction_depth} = 0
1103 if $self->_dbh_autocommit;
1105 elsif($self->{transaction_depth} > 1) {
1106 $self->{transaction_depth}--;
1108 if $self->auto_savepoint;
1114 my $dbh = $self->_dbh;
1116 if ($self->{transaction_depth} == 1) {
1117 $self->debugobj->txn_rollback()
1119 $self->{transaction_depth} = 0
1120 if $self->_dbh_autocommit;
1123 elsif($self->{transaction_depth} > 1) {
1124 $self->{transaction_depth}--;
1125 if ($self->auto_savepoint) {
1126 $self->svp_rollback;
1131 die DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION->new;
1136 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
1137 $error =~ /$exception_class/ and $self->throw_exception($error);
1138 # ensure that a failed rollback resets the transaction depth
1139 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
1140 $self->throw_exception($error);
1144 # This used to be the top-half of _execute. It was split out to make it
1145 # easier to override in NoBindVars without duping the rest. It takes up
1146 # all of _execute's args, and emits $sql, @bind.
1147 sub _prep_for_execute {
1148 my ($self, $op, $extra_bind, $ident, $args) = @_;
1150 my ($sql, @bind) = $self->sql_maker->$op($ident, @$args);
1152 map { ref $_ eq 'ARRAY' ? $_ : [ '!!dummy', $_ ] } @$extra_bind)
1155 return ($sql, \@bind);
1158 sub _fix_bind_params {
1159 my ($self, @bind) = @_;
1161 ### Turn @bind from something like this:
1162 ### ( [ "artist", 1 ], [ "cdid", 1, 3 ] )
1164 ### ( "'1'", "'1'", "'3'" )
1167 if ( defined( $_ && $_->[1] ) ) {
1168 map { qq{'$_'}; } @{$_}[ 1 .. $#$_ ];
1175 my ( $self, $sql, @bind ) = @_;
1177 if ( $self->debug ) {
1178 @bind = $self->_fix_bind_params(@bind);
1180 $self->debugobj->query_start( $sql, @bind );
1185 my ( $self, $sql, @bind ) = @_;
1187 if ( $self->debug ) {
1188 @bind = $self->_fix_bind_params(@bind);
1189 $self->debugobj->query_end( $sql, @bind );
1194 my ($self, $dbh, $op, $extra_bind, $ident, $bind_attributes, @args) = @_;
1196 if( blessed($ident) && $ident->isa("DBIx::Class::ResultSource") ) {
1197 $ident = $ident->from();
1200 my ($sql, $bind) = $self->_prep_for_execute($op, $extra_bind, $ident, \@args);
1202 $self->_query_start( $sql, @$bind );
1204 my $sth = $self->sth($sql,$op);
1206 my $placeholder_index = 1;
1208 foreach my $bound (@$bind) {
1209 my $attributes = {};
1210 my($column_name, @data) = @$bound;
1212 if ($bind_attributes) {
1213 $attributes = $bind_attributes->{$column_name}
1214 if defined $bind_attributes->{$column_name};
1217 foreach my $data (@data) {
1218 $data = ref $data ? ''.$data : $data; # stringify args
1220 $sth->bind_param($placeholder_index, $data, $attributes);
1221 $placeholder_index++;
1225 # Can this fail without throwing an exception anyways???
1226 my $rv = $sth->execute();
1227 $self->throw_exception($sth->errstr) if !$rv;
1229 $self->_query_end( $sql, @$bind );
1231 return (wantarray ? ($rv, $sth, @$bind) : $rv);
1236 $self->dbh_do('_dbh_execute', @_)
1240 my ($self, $source, $to_insert) = @_;
1242 my $ident = $source->from;
1243 my $bind_attributes = $self->source_bind_attributes($source);
1245 $self->ensure_connected;
1246 foreach my $col ( $source->columns ) {
1247 if ( !defined $to_insert->{$col} ) {
1248 my $col_info = $source->column_info($col);
1250 if ( $col_info->{auto_nextval} ) {
1251 $to_insert->{$col} = $self->_sequence_fetch( 'nextval', $col_info->{sequence} || $self->_dbh_get_autoinc_seq($self->dbh, $source) );
1256 $self->_execute('insert' => [], $source, $bind_attributes, $to_insert);
1261 ## Still not quite perfect, and EXPERIMENTAL
1262 ## Currently it is assumed that all values passed will be "normal", i.e. not
1263 ## scalar refs, or at least, all the same type as the first set, the statement is
1264 ## only prepped once.
1266 my ($self, $source, $cols, $data) = @_;
1268 my $table = $source->from;
1269 @colvalues{@$cols} = (0..$#$cols);
1270 my ($sql, @bind) = $self->sql_maker->insert($table, \%colvalues);
1272 $self->_query_start( $sql, @bind );
1273 my $sth = $self->sth($sql);
1275 # @bind = map { ref $_ ? ''.$_ : $_ } @bind; # stringify args
1277 ## This must be an arrayref, else nothing works!
1279 my $tuple_status = [];
1282 ##print STDERR Dumper( $data, $sql, [@bind] );
1286 ## Get the bind_attributes, if any exist
1287 my $bind_attributes = $self->source_bind_attributes($source);
1289 ## Bind the values and execute
1290 my $placeholder_index = 1;
1292 foreach my $bound (@bind) {
1294 my $attributes = {};
1295 my ($column_name, $data_index) = @$bound;
1297 if( $bind_attributes ) {
1298 $attributes = $bind_attributes->{$column_name}
1299 if defined $bind_attributes->{$column_name};
1302 my @data = map { $_->[$data_index] } @$data;
1304 $sth->bind_param_array( $placeholder_index, [@data], $attributes );
1305 $placeholder_index++;
1307 my $rv = $sth->execute_array({ArrayTupleStatus => $tuple_status});
1308 $self->throw_exception($sth->errstr) if !$rv;
1310 $self->_query_end( $sql, @bind );
1311 return (wantarray ? ($rv, $sth, @bind) : $rv);
1315 my $self = shift @_;
1316 my $source = shift @_;
1317 my $bind_attributes = $self->source_bind_attributes($source);
1319 return $self->_execute('update' => [], $source, $bind_attributes, @_);
1324 my $self = shift @_;
1325 my $source = shift @_;
1327 my $bind_attrs = {}; ## If ever it's needed...
1329 return $self->_execute('delete' => [], $source, $bind_attrs, @_);
1333 my ($self, $ident, $select, $condition, $attrs) = @_;
1334 my $order = $attrs->{order_by};
1336 if (ref $condition eq 'SCALAR') {
1337 my $unwrap = ${$condition};
1338 if ($unwrap =~ s/ORDER BY (.*)$//i) {
1340 $condition = \$unwrap;
1344 my $for = delete $attrs->{for};
1345 my $sql_maker = $self->sql_maker;
1346 local $sql_maker->{for} = $for;
1348 if (exists $attrs->{group_by} || $attrs->{having}) {
1350 group_by => $attrs->{group_by},
1351 having => $attrs->{having},
1352 ($order ? (order_by => $order) : ())
1355 my $bind_attrs = {}; ## Future support
1356 my @args = ('select', $attrs->{bind}, $ident, $bind_attrs, $select, $condition, $order);
1357 if ($attrs->{software_limit} ||
1358 $self->sql_maker->_default_limit_syntax eq "GenericSubQ") {
1359 $attrs->{software_limit} = 1;
1361 $self->throw_exception("rows attribute must be positive if present")
1362 if (defined($attrs->{rows}) && !($attrs->{rows} > 0));
1364 # MySQL actually recommends this approach. I cringe.
1365 $attrs->{rows} = 2**48 if not defined $attrs->{rows} and defined $attrs->{offset};
1366 push @args, $attrs->{rows}, $attrs->{offset};
1369 return $self->_execute(@args);
1372 sub source_bind_attributes {
1373 my ($self, $source) = @_;
1375 my $bind_attributes;
1376 foreach my $column ($source->columns) {
1378 my $data_type = $source->column_info($column)->{data_type} || '';
1379 $bind_attributes->{$column} = $self->bind_attribute_by_data_type($data_type)
1383 return $bind_attributes;
1390 =item Arguments: $ident, $select, $condition, $attrs
1394 Handle a SQL select statement.
1400 my ($ident, $select, $condition, $attrs) = @_;
1401 return $self->cursor_class->new($self, \@_, $attrs);
1406 my ($rv, $sth, @bind) = $self->_select(@_);
1407 my @row = $sth->fetchrow_array;
1408 if(@row && $sth->fetchrow_array) {
1409 carp "Query returned more than one row. SQL that returns multiple rows is DEPRECATED for ->find and ->single";
1411 # Need to call finish() to work round broken DBDs
1420 =item Arguments: $sql
1424 Returns a L<DBI> sth (statement handle) for the supplied SQL.
1429 my ($self, $dbh, $sql) = @_;
1431 # 3 is the if_active parameter which avoids active sth re-use
1432 my $sth = $self->disable_sth_caching
1433 ? $dbh->prepare($sql)
1434 : $dbh->prepare_cached($sql, {}, 3);
1436 # XXX You would think RaiseError would make this impossible,
1437 # but apparently that's not true :(
1438 $self->throw_exception($dbh->errstr) if !$sth;
1444 my ($self, $sql) = @_;
1445 $self->dbh_do('_dbh_sth', $sql);
1448 sub _dbh_columns_info_for {
1449 my ($self, $dbh, $table) = @_;
1451 if ($dbh->can('column_info')) {
1454 my ($schema,$tab) = $table =~ /^(.+?)\.(.+)$/ ? ($1,$2) : (undef,$table);
1455 my $sth = $dbh->column_info( undef,$schema, $tab, '%' );
1457 while ( my $info = $sth->fetchrow_hashref() ){
1459 $column_info{data_type} = $info->{TYPE_NAME};
1460 $column_info{size} = $info->{COLUMN_SIZE};
1461 $column_info{is_nullable} = $info->{NULLABLE} ? 1 : 0;
1462 $column_info{default_value} = $info->{COLUMN_DEF};
1463 my $col_name = $info->{COLUMN_NAME};
1464 $col_name =~ s/^\"(.*)\"$/$1/;
1466 $result{$col_name} = \%column_info;
1469 return \%result if !$@ && scalar keys %result;
1473 my $sth = $dbh->prepare($self->sql_maker->select($table, undef, \'1 = 0'));
1475 my @columns = @{$sth->{NAME_lc}};
1476 for my $i ( 0 .. $#columns ){
1478 $column_info{data_type} = $sth->{TYPE}->[$i];
1479 $column_info{size} = $sth->{PRECISION}->[$i];
1480 $column_info{is_nullable} = $sth->{NULLABLE}->[$i] ? 1 : 0;
1482 if ($column_info{data_type} =~ m/^(.*?)\((.*?)\)$/) {
1483 $column_info{data_type} = $1;
1484 $column_info{size} = $2;
1487 $result{$columns[$i]} = \%column_info;
1491 foreach my $col (keys %result) {
1492 my $colinfo = $result{$col};
1493 my $type_num = $colinfo->{data_type};
1495 if(defined $type_num && $dbh->can('type_info')) {
1496 my $type_info = $dbh->type_info($type_num);
1497 $type_name = $type_info->{TYPE_NAME} if $type_info;
1498 $colinfo->{data_type} = $type_name if $type_name;
1505 sub columns_info_for {
1506 my ($self, $table) = @_;
1507 $self->dbh_do('_dbh_columns_info_for', $table);
1510 =head2 last_insert_id
1512 Return the row id of the last insert.
1516 sub _dbh_last_insert_id {
1517 my ($self, $dbh, $source, $col) = @_;
1518 # XXX This is a SQLite-ism as a default... is there a DBI-generic way?
1519 $dbh->func('last_insert_rowid');
1522 sub last_insert_id {
1524 $self->dbh_do('_dbh_last_insert_id', @_);
1529 Returns the database driver name.
1533 sub sqlt_type { shift->dbh->{Driver}->{Name} }
1535 =head2 bind_attribute_by_data_type
1537 Given a datatype from column info, returns a database specific bind
1538 attribute for $dbh->bind_param($val,$attribute) or nothing if we will
1539 let the database planner just handle it.
1541 Generally only needed for special case column types, like bytea in postgres.
1545 sub bind_attribute_by_data_type {
1549 =head2 create_ddl_dir
1553 =item Arguments: $schema \@databases, $version, $directory, $preversion, \%sqlt_args
1557 Creates a SQL file based on the Schema, for each of the specified
1558 database types, in the given directory.
1560 By default, C<\%sqlt_args> will have
1562 { add_drop_table => 1, ignore_constraint_names => 1, ignore_index_names => 1 }
1564 merged with the hash passed in. To disable any of those features, pass in a
1565 hashref like the following
1567 { ignore_constraint_names => 0, # ... other options }
1571 sub create_ddl_dir {
1572 my ($self, $schema, $databases, $version, $dir, $preversion, $sqltargs) = @_;
1574 if(!$dir || !-d $dir) {
1575 warn "No directory given, using ./\n";
1578 $databases ||= ['MySQL', 'SQLite', 'PostgreSQL'];
1579 $databases = [ $databases ] if(ref($databases) ne 'ARRAY');
1580 $version ||= $schema->VERSION || '1.x';
1582 add_drop_table => 1,
1583 ignore_constraint_names => 1,
1584 ignore_index_names => 1,
1588 $self->throw_exception(q{Can't create a ddl file without SQL::Translator 0.09: '}
1589 . $self->_check_sqlt_message . q{'})
1590 if !$self->_check_sqlt_version;
1592 my $sqlt = SQL::Translator->new( $sqltargs );
1594 $sqlt->parser('SQL::Translator::Parser::DBIx::Class');
1595 my $sqlt_schema = $sqlt->translate({ data => $schema }) or die $sqlt->error;
1597 foreach my $db (@$databases) {
1599 $sqlt = $self->configure_sqlt($sqlt, $db);
1600 $sqlt->{schema} = $sqlt_schema;
1601 $sqlt->producer($db);
1604 my $filename = $schema->ddl_filename($db, $version, $dir);
1605 if (-e $filename && (!$version || ($version == $schema->schema_version()))) {
1606 # if we are dumping the current version, overwrite the DDL
1607 warn "Overwriting existing DDL file - $filename";
1611 my $output = $sqlt->translate;
1613 warn("Failed to translate to $db, skipping. (" . $sqlt->error . ")");
1616 if(!open($file, ">$filename")) {
1617 $self->throw_exception("Can't open $filename for writing ($!)");
1620 print $file $output;
1623 next unless ($preversion);
1625 require SQL::Translator::Diff;
1627 my $prefilename = $schema->ddl_filename($db, $preversion, $dir);
1628 if(!-e $prefilename) {
1629 warn("No previous schema file found ($prefilename)");
1633 my $difffile = $schema->ddl_filename($db, $version, $dir, $preversion);
1635 warn("Overwriting existing diff file - $difffile");
1641 my $t = SQL::Translator->new($sqltargs);
1644 $t->parser( $db ) or die $t->error;
1645 $t = $self->configure_sqlt($t, $db);
1646 my $out = $t->translate( $prefilename ) or die $t->error;
1647 $source_schema = $t->schema;
1648 unless ( $source_schema->name ) {
1649 $source_schema->name( $prefilename );
1653 # The "new" style of producers have sane normalization and can support
1654 # diffing a SQL file against a DBIC->SQLT schema. Old style ones don't
1655 # And we have to diff parsed SQL against parsed SQL.
1656 my $dest_schema = $sqlt_schema;
1658 unless ( "SQL::Translator::Producer::$db"->can('preprocess_schema') ) {
1659 my $t = SQL::Translator->new($sqltargs);
1662 $t->parser( $db ) or die $t->error;
1663 $t = $self->configure_sqlt($t, $db);
1664 my $out = $t->translate( $filename ) or die $t->error;
1665 $dest_schema = $t->schema;
1666 $dest_schema->name( $filename )
1667 unless $dest_schema->name;
1670 my $diff = SQL::Translator::Diff::schema_diff($source_schema, $db,
1674 if(!open $file, ">$difffile") {
1675 $self->throw_exception("Can't write to $difffile ($!)");
1683 sub configure_sqlt() {
1686 my $db = shift || $self->sqlt_type;
1687 if ($db eq 'PostgreSQL') {
1688 $tr->quote_table_names(0);
1689 $tr->quote_field_names(0);
1694 =head2 deployment_statements
1698 =item Arguments: $schema, $type, $version, $directory, $sqlt_args
1702 Returns the statements used by L</deploy> and L<DBIx::Class::Schema/deploy>.
1703 The database driver name is given by C<$type>, though the value from
1704 L</sqlt_type> is used if it is not specified.
1706 C<$directory> is used to return statements from files in a previously created
1707 L</create_ddl_dir> directory and is optional. The filenames are constructed
1708 from L<DBIx::Class::Schema/ddl_filename>, the schema name and the C<$version>.
1710 If no C<$directory> is specified then the statements are constructed on the
1711 fly using L<SQL::Translator> and C<$version> is ignored.
1713 See L<SQL::Translator/METHODS> for a list of values for C<$sqlt_args>.
1717 sub deployment_statements {
1718 my ($self, $schema, $type, $version, $dir, $sqltargs) = @_;
1719 # Need to be connected to get the correct sqlt_type
1720 $self->ensure_connected() unless $type;
1721 $type ||= $self->sqlt_type;
1722 $version ||= $schema->VERSION || '1.x';
1724 my $filename = $schema->ddl_filename($type, $dir, $version);
1728 open($file, "<$filename")
1729 or $self->throw_exception("Can't open $filename ($!)");
1732 return join('', @rows);
1735 $self->throw_exception(q{Can't deploy without SQL::Translator 0.09: '}
1736 . $self->_check_sqlt_message . q{'})
1737 if !$self->_check_sqlt_version;
1739 require SQL::Translator::Parser::DBIx::Class;
1740 eval qq{use SQL::Translator::Producer::${type}};
1741 $self->throw_exception($@) if $@;
1743 # sources needs to be a parser arg, but for simplicty allow at top level
1745 $sqltargs->{parser_args}{sources} = delete $sqltargs->{sources}
1746 if exists $sqltargs->{sources};
1748 my $tr = SQL::Translator->new(%$sqltargs);
1749 SQL::Translator::Parser::DBIx::Class::parse( $tr, $schema );
1750 return "SQL::Translator::Producer::${type}"->can('produce')->($tr);
1754 my ($self, $schema, $type, $sqltargs, $dir) = @_;
1755 foreach my $statement ( $self->deployment_statements($schema, $type, undef, $dir, { no_comments => 1, %{ $sqltargs || {} } } ) ) {
1756 foreach my $line ( split(";\n", $statement)) {
1757 next if($line =~ /^--/);
1759 # next if($line =~ /^DROP/m);
1760 next if($line =~ /^BEGIN TRANSACTION/m);
1761 next if($line =~ /^COMMIT/m);
1762 next if $line =~ /^\s+$/; # skip whitespace only
1763 $self->_query_start($line);
1765 $self->dbh->do($line); # shouldn't be using ->dbh ?
1768 warn qq{$@ (running "${line}")};
1770 $self->_query_end($line);
1775 =head2 datetime_parser
1777 Returns the datetime parser class
1781 sub datetime_parser {
1783 return $self->{datetime_parser} ||= do {
1784 $self->ensure_connected;
1785 $self->build_datetime_parser(@_);
1789 =head2 datetime_parser_type
1791 Defines (returns) the datetime parser class - currently hardwired to
1792 L<DateTime::Format::MySQL>
1796 sub datetime_parser_type { "DateTime::Format::MySQL"; }
1798 =head2 build_datetime_parser
1800 See L</datetime_parser>
1804 sub build_datetime_parser {
1806 my $type = $self->datetime_parser_type(@_);
1808 $self->throw_exception("Couldn't load ${type}: $@") if $@;
1813 my $_check_sqlt_version; # private
1814 my $_check_sqlt_message; # private
1815 sub _check_sqlt_version {
1816 return $_check_sqlt_version if defined $_check_sqlt_version;
1817 eval 'use SQL::Translator "0.09"';
1818 $_check_sqlt_message = $@ || '';
1819 $_check_sqlt_version = !$@;
1822 sub _check_sqlt_message {
1823 _check_sqlt_version if !defined $_check_sqlt_message;
1824 $_check_sqlt_message;
1828 =head2 is_replicating
1830 A boolean that reports if a particular L<DBIx::Class::Storage::DBI> is set to
1831 replicate from a master database. Default is undef, which is the result
1832 returned by databases that don't support replication.
1836 sub is_replicating {
1841 =head2 lag_behind_master
1843 Returns a number that represents a certain amount of lag behind a master db
1844 when a given storage is replicating. The number is database dependent, but
1845 starts at zero and increases with the amount of lag. Default in undef
1849 sub lag_behind_master {
1855 return if !$self->_dbh;
1864 =head2 DBIx::Class and AutoCommit
1866 DBIx::Class can do some wonderful magic with handling exceptions,
1867 disconnections, and transactions when you use C<< AutoCommit => 1 >>
1868 combined with C<txn_do> for transaction support.
1870 If you set C<< AutoCommit => 0 >> in your connect info, then you are always
1871 in an assumed transaction between commits, and you're telling us you'd
1872 like to manage that manually. A lot of the magic protections offered by
1873 this module will go away. We can't protect you from exceptions due to database
1874 disconnects because we don't know anything about how to restart your
1875 transactions. You're on your own for handling all sorts of exceptional
1876 cases if you choose the C<< AutoCommit => 0 >> path, just as you would
1882 The module defines a set of methods within the DBIC::SQL::Abstract
1883 namespace. These build on L<SQL::Abstract::Limit> to provide the
1884 SQL query functions.
1886 The following methods are extended:-
1900 See L</connect_info> for details.
1901 For setting, this method is deprecated in favor of L</connect_info>.
1905 See L</connect_info> for details.
1906 For setting, this method is deprecated in favor of L</connect_info>.
1910 See L</connect_info> for details.
1911 For setting, this method is deprecated in favor of L</connect_info>.
1917 Matt S. Trout <mst@shadowcatsystems.co.uk>
1919 Andy Grundman <andy@hybridized.org>
1923 You may distribute this code under the same terms as Perl itself.