1 package SQL::Abstract; # see doc at end of file
5 use Module::Runtime ();
10 use Exporter 'import';
11 our @EXPORT_OK = qw(is_plain_value is_literal_value is_undef_value);
21 *SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION = $ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}
27 #======================================================================
29 #======================================================================
31 our $VERSION = '1.86';
33 # This would confuse some packagers
34 $VERSION = eval $VERSION if $VERSION =~ /_/; # numify for warning-free dev releases
38 # special operators (-in, -between). May be extended/overridden by user.
39 # See section WHERE: BUILTIN SPECIAL OPERATORS below for implementation
40 my @BUILTIN_SPECIAL_OPS = (
41 {regex => qr/^ (?: not \s )? between $/ix, handler => sub { die "NOPE" }},
42 {regex => qr/^ is (?: \s+ not )? $/ix, handler => sub { die "NOPE" }},
43 {regex => qr/^ (?: not \s )? in $/ix, handler => sub { die "NOPE" }},
44 {regex => qr/^ ident $/ix, handler => sub { die "NOPE" }},
45 {regex => qr/^ value $/ix, handler => sub { die "NOPE" }},
48 #======================================================================
49 # DEBUGGING AND ERROR REPORTING
50 #======================================================================
53 return unless $_[0]->{debug}; shift; # a little faster
54 my $func = (caller(1))[3];
55 warn "[$func] ", @_, "\n";
59 my($func) = (caller(1))[3];
60 Carp::carp "[$func] Warning: ", @_;
64 my($func) = (caller(1))[3];
65 Carp::croak "[$func] Fatal: ", @_;
68 sub is_literal_value ($) {
69 ref $_[0] eq 'SCALAR' ? [ ${$_[0]} ]
70 : ( ref $_[0] eq 'REF' and ref ${$_[0]} eq 'ARRAY' ) ? [ @${ $_[0] } ]
74 sub is_undef_value ($) {
78 and exists $_[0]->{-value}
79 and not defined $_[0]->{-value}
83 # FIXME XSify - this can be done so much more efficiently
84 sub is_plain_value ($) {
86 ! length ref $_[0] ? \($_[0])
88 ref $_[0] eq 'HASH' and keys %{$_[0]} == 1
90 exists $_[0]->{-value}
91 ) ? \($_[0]->{-value})
93 # reuse @_ for even moar speedz
94 defined ( $_[1] = Scalar::Util::blessed $_[0] )
96 # deliberately not using Devel::OverloadInfo - the checks we are
97 # intersted in are much more limited than the fullblown thing, and
98 # this is a very hot piece of code
100 # simply using ->can('(""') can leave behind stub methods that
101 # break actually using the overload later (see L<perldiag/Stub
102 # found while resolving method "%s" overloading "%s" in package
103 # "%s"> and the source of overload::mycan())
105 # either has stringification which DBI SHOULD prefer out of the box
106 grep { *{ (qq[${_}::(""]) }{CODE} } @{ $_[2] = mro::get_linear_isa( $_[1] ) }
108 # has nummification or boolification, AND fallback is *not* disabled
110 SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION
113 grep { *{"${_}::(0+"}{CODE} } @{$_[2]}
115 grep { *{"${_}::(bool"}{CODE} } @{$_[2]}
119 # no fallback specified at all
120 ! ( ($_[3]) = grep { *{"${_}::()"}{CODE} } @{$_[2]} )
122 # fallback explicitly undef
123 ! defined ${"$_[3]::()"}
136 #======================================================================
138 #======================================================================
142 bool => '_expand_bool',
143 row => '_expand_row',
145 func => '_expand_func',
146 values => '_expand_values',
147 list => '_expand_list',
150 (map +($_ => __PACKAGE__->make_binop_expander('_expand_between')),
151 qw(between not_between)),
152 (map +($_ => __PACKAGE__->make_binop_expander('_expand_in')),
154 (map +($_ => '_expand_op_andor'), ('and', 'or')),
155 (map +($_ => '_expand_op_is'), ('is', 'is_not')),
156 (map +($_ => __PACKAGE__->make_unop_expander("_expand_${_}")),
157 qw(ident value nest)),
160 (map +($_, "_render_$_"),
161 qw(op func bind ident literal row values keyword)),
164 (map +($_ => '_render_op_between'), 'between', 'not_between'),
165 (map +($_ => '_render_op_in'), 'in', 'not_in'),
166 (map +($_ => '_render_unop_postfix'),
167 'is_null', 'is_not_null', 'asc', 'desc',
169 (not => '_render_unop_paren'),
170 (map +($_ => '_render_op_andor'), qw(and or)),
171 ',' => '_render_op_multop',
174 delete => [ qw(target where returning) ],
175 update => [ qw(target set where returning) ],
176 insert => [ qw(target fields from returning) ],
177 select => [ qw(select from where order_by) ],
180 'delete.from' => '_expand_delete_clause_target',
181 'update.update' => '_expand_update_clause_target',
182 'insert.into' => '_expand_insert_clause_target',
183 'insert.values' => '_expand_insert_clause_from',
186 'delete.target' => '_render_delete_clause_target',
187 'update.target' => '_render_update_clause_target',
188 'insert.target' => '_render_insert_clause_target',
189 'insert.fields' => '_render_insert_clause_fields',
190 'insert.from' => '_render_insert_clause_from',
194 foreach my $stmt (keys %{$Defaults{clauses_of}}) {
195 $Defaults{expand}{$stmt} = '_expand_statement';
196 $Defaults{render}{$stmt} = '_render_statement';
197 foreach my $clause (@{$Defaults{clauses_of}{$stmt}}) {
198 $Defaults{expand_clause}{"${stmt}.${clause}"}
199 = "_expand_${stmt}_clause_${clause}";
205 my $class = ref($self) || $self;
206 my %opt = (ref $_[0] eq 'HASH') ? %{$_[0]} : @_;
208 # choose our case by keeping an option around
209 delete $opt{case} if $opt{case} && $opt{case} ne 'lower';
211 # default logic for interpreting arrayrefs
212 $opt{logic} = $opt{logic} ? uc $opt{logic} : 'OR';
214 # how to return bind vars
215 $opt{bindtype} ||= 'normal';
217 # default comparison is "=", but can be overridden
220 # try to recognize which are the 'equality' and 'inequality' ops
221 # (temporary quickfix (in 2007), should go through a more seasoned API)
222 $opt{equality_op} = qr/^( \Q$opt{cmp}\E | \= )$/ix;
223 $opt{inequality_op} = qr/^( != | <> )$/ix;
225 $opt{like_op} = qr/^ (is_)?r?like $/xi;
226 $opt{not_like_op} = qr/^ (is_)?not_r?like $/xi;
229 $opt{sqltrue} ||= '1=1';
230 $opt{sqlfalse} ||= '0=1';
233 $opt{special_ops} ||= [];
236 $opt{unary_ops} ||= [];
238 # rudimentary sanity-check for user supplied bits treated as functions/operators
239 # If a purported function matches this regular expression, an exception is thrown.
240 # Literal SQL is *NOT* subject to this check, only functions (and column names
241 # when quoting is not in effect)
244 # need to guard against ()'s in column names too, but this will break tons of
245 # hacks... ideas anyone?
246 $opt{injection_guard} ||= qr/
252 $opt{expand_unary} = {};
254 foreach my $name (sort keys %Defaults) {
255 $opt{$name} = { %{$Defaults{$name}}, %{$opt{$name}||{}} };
258 if ($class ne __PACKAGE__) {
260 # check for overriden methods
262 foreach my $type (qw(insert update delete)) {
263 my $method = "_${type}_returning";
264 if (__PACKAGE__->can($method) ne $class->can($method)) {
265 my $clause = "${type}.returning";
266 $opt{expand_clause}{$clause} = sub { $_[2] },
267 $opt{render_clause}{$clause}
268 = sub { [ $_[0]->$method($_[3]) ] };
271 if (__PACKAGE__->can('_table') ne $class->can('_table')) {
272 $opt{expand_clause}{'select.from'} = sub {
273 return +{ -literal => [ $_[0]->_table($_[2]) ] };
276 if (__PACKAGE__->can('_order_by') ne $class->can('_order_by')) {
277 $opt{expand_clause}{'select.order_by'} = sub { $_[2] };
278 $opt{render_clause}{'select.order_by'} = sub {
279 [ $_[0]->_order_by($_[2]) ];
282 if (__PACKAGE__->can('_select_fields') ne $class->can('_select_fields')) {
283 $opt{expand_clause}{'select.select'} = sub { $_[2] };
284 $opt{render_clause}{'select.select'} = sub {
285 my @super = $_[0]->_select_fields($_[2]);
287 ref($super[0]) eq 'HASH'
288 ? $_[0]->render_expr($super[0])
291 return $_[0]->join_query_parts(
292 ' ', { -keyword => 'select' }, $effort
296 if ($class->isa('DBIx::Class::SQLMaker')) {
297 $opt{warn_once_on_nest} = 1;
298 $opt{disable_old_special_ops} = 1;
299 $opt{render_clause}{'select.where'} = sub {
300 my ($sql, @bind) = $_[0]->where($_[2]);
301 s/\A\s+//, s/\s+\Z// for $sql;
302 return [ $sql, @bind ];
307 if ($opt{lazy_join_sql_parts}) {
308 my $mod = Module::Runtime::use_module('SQL::Abstract::Parts');
309 $opt{join_sql_parts} ||= sub { $mod->new(@_) };
312 $opt{join_sql_parts} ||= sub { join $_[0], @_[1..$#_] };
314 return bless \%opt, $class;
318 my ($self, $name, $key, $value) = @_;
319 return $self->{$name}{$key} unless @_ > 3;
320 $self->{$name}{$key} = $value;
324 sub make_unop_expander {
325 my (undef, $exp) = @_;
327 my ($self, $name, $body, $k) = @_;
328 return $self->_expand_hashpair_cmp($k, { "-${name}" => $body })
330 return $self->$exp($name, $body);
334 sub make_binop_expander {
335 my (undef, $exp) = @_;
337 my ($self, $name, $body, $k) = @_;
338 $k = shift @{$body = [ @$body ]} unless defined $k;
339 $k = ref($k) ? $k : { -ident => $k };
340 return $self->$exp($name, $body, $k);
345 my ($self, $plugin, @args) = @_;
346 unless (ref $plugin) {
347 $plugin =~ s/\A\+/${\__PACKAGE__}::Plugin::/;
348 require(join('/', split '::', $plugin).'.pm');
350 $plugin->apply_to($self, @args);
355 foreach my $type (qw(
356 expand op_expand render op_render clause_expand clause_render
358 my $name = join '_', reverse split '_', $type;
359 my $singular = "${type}er";
361 eval qq{sub ${singular} { shift->${singular}s(\@_) }; 1 }
362 or die "Method builder failed for ${singular}: $@";
363 eval qq{sub wrap_${singular} {
364 shift->wrap_${singular}s(\@_)
365 }; 1 } or die "Method builder failed for wrap_${singular}: $@";
367 eval qq{sub ${singular}s {
368 my (\$self, \@args) = \@_;
369 while (my (\$this_key, \$this_value) = splice(\@args, 0, 2)) {
370 \$self->_ext_rw('${name}', \$this_key, \$this_value);
373 }; 1 } or die "Method builder failed for ${singular}s: $@";
374 eval qq{sub wrap_${singular}s {
375 my (\$self, \@args) = \@_;
376 while (my (\$this_key, \$this_builder) = splice(\@args, 0, 2)) {
377 my \$orig = \$self->_ext_rw('${name}', \$this_key);
379 '${name}', \$this_key,
380 \$this_builder->(\$orig, '${name}', \$this_key),
384 }; 1 } or die "Method builder failed for wrap_${singular}s: $@";
385 eval qq{sub ${singular}_list { sort keys %{\$_[0]->{\$name}} }; 1; }
386 or die "Method builder failed for ${singular}_list: $@";
388 foreach my $singular (qw(unop_expander binop_expander)) {
389 eval qq{sub ${singular} { shift->${singular}s(\@_) }; 1 }
390 or die "Method builder failed for ${singular}: $@";
391 eval qq{sub ${singular}s {
392 my (\$self, \@args) = \@_;
393 while (my (\$this_key, \$this_value) = splice(\@args, 0, 2)) {
395 expand_op => \$this_key,
396 \$self->make_${singular}(\$this_value),
400 }; 1 } or die "Method builder failed for ${singular}s: $@";
404 #sub register_op { $_[0]->{is_op}{$_[1]} = 1; $_[0] }
406 sub statement_list { sort keys %{$_[0]->{clauses_of}} }
409 my ($self, $of, @clauses) = @_;
411 return @{$self->{clauses_of}{$of}||[]};
413 if (ref($clauses[0]) eq 'CODE') {
414 @clauses = $self->${\($clauses[0])}(@{$self->{clauses_of}{$of}||[]});
416 $self->{clauses_of}{$of} = \@clauses;
425 ref($self->{$_}) eq 'HASH'
434 sub sqltrue { +{ -literal => [ $_[0]->{sqltrue} ] } }
435 sub sqlfalse { +{ -literal => [ $_[0]->{sqlfalse} ] } }
437 sub _assert_pass_injection_guard {
438 if ($_[1] =~ $_[0]->{injection_guard}) {
439 my $class = ref $_[0];
440 puke "Possible SQL injection attempt '$_[1]'. If this is indeed a part of the "
441 . "desired SQL use literal SQL ( \'...' or \[ '...' ] ) or supply your own "
442 . "{injection_guard} attribute to ${class}->new()"
447 #======================================================================
449 #======================================================================
452 my ($self, $table, $data, $options) = @_;
455 if (ref($table) eq 'HASH') {
458 my %clauses = (target => $table, values => $data, %{$options||{}});
462 my @rendered = $self->render_statement({ -insert => $stmt });
463 return wantarray ? @rendered : $rendered[0];
466 sub _expand_insert_clause_target {
467 +(target => $_[0]->expand_expr($_[2], -ident));
470 sub _expand_insert_clause_fields {
472 $_[0]->expand_expr({ -list => $_[2] }, -ident)
473 ] } if ref($_[2]) eq 'ARRAY';
474 return $_[2]; # should maybe still expand somewhat?
477 sub _expand_insert_clause_from {
478 my ($self, undef, $data) = @_;
479 if (ref($data) eq 'HASH' and (keys(%$data))[0] =~ /^-/) {
480 return $self->expand_expr($data);
482 my ($f_aqt, $v_aqt) = $self->_expand_insert_values($data);
484 from => { -values => [ $v_aqt ] },
485 ($f_aqt ? (fields => $f_aqt) : ()),
489 sub _expand_insert_clause_returning {
490 +(returning => $_[0]->expand_expr({ -list => $_[2] }, -ident));
493 sub _expand_insert_values {
494 my ($self, $data) = @_;
495 if (is_literal_value($data)) {
496 (undef, $self->expand_expr($data));
498 my ($fields, $values) = (
499 ref($data) eq 'HASH' ?
500 ([ sort keys %$data ], [ @{$data}{sort keys %$data} ])
504 # no names (arrayref) means can't generate bindtype
505 !($fields) && $self->{bindtype} eq 'columns'
506 && belch "can't do 'columns' bindtype when called with arrayref";
510 ? $self->expand_expr({ -row => $fields }, -ident)
515 local our $Cur_Col_Meta = $fields->[$_];
516 $self->_expand_insert_value($values->[$_])
523 sub _render_insert_clause_fields {
524 return $_[0]->render_aqt($_[2]);
527 sub _render_insert_clause_target {
528 my ($self, undef, $from) = @_;
529 $self->join_query_parts(' ', { -keyword => 'insert into' }, $from);
532 sub _render_insert_clause_from {
533 return $_[0]->render_aqt($_[2], 1);
536 # So that subclasses can override INSERT ... RETURNING separately from
537 # UPDATE and DELETE (e.g. DBIx::Class::SQLMaker::Oracle does this)
538 sub _insert_returning { shift->_returning(@_) }
540 sub _redispatch_returning {
541 my ($self, $type, undef, $returning) = @_;
542 [ $self->${\"_${type}_returning"}({ returning => $returning }) ];
546 my ($self, $options) = @_;
548 my $f = $options->{returning};
550 my ($sql, @bind) = @{ $self->render_aqt(
551 $self->expand_expr({ -list => $f }, -ident)
553 my $rsql = $self->_sqlcase(' returning ').$sql;
554 return wantarray ? ($rsql, @bind) : $rsql;
557 sub _expand_insert_value {
560 my $k = our $Cur_Col_Meta;
562 if (ref($v) eq 'ARRAY') {
563 if ($self->{array_datatypes}) {
564 return +{ -bind => [ $k, $v ] };
566 my ($sql, @bind) = @$v;
567 $self->_assert_bindval_matches_bindtype(@bind);
568 return +{ -literal => $v };
570 if (ref($v) eq 'HASH') {
571 if (grep !/^-/, keys %$v) {
572 belch "HASH ref as bind value in insert is not supported";
573 return +{ -bind => [ $k, $v ] };
577 return +{ -bind => [ $k, undef ] };
579 return $self->expand_expr($v);
584 #======================================================================
586 #======================================================================
589 my ($self, $table, $set, $where, $options) = @_;
592 if (ref($table) eq 'HASH') {
596 @clauses{qw(target set where)} = ($table, $set, $where);
597 puke "Unsupported data type specified to \$sql->update"
598 unless ref($clauses{set}) eq 'HASH';
599 @clauses{keys %$options} = values %$options;
603 my @rendered = $self->render_statement({ -update => $stmt });
604 return wantarray ? @rendered : $rendered[0];
607 sub _render_update_clause_target {
608 my ($self, undef, $target) = @_;
609 $self->join_query_parts(' ', { -keyword => 'update' }, $target);
612 sub _update_set_values {
613 my ($self, $data) = @_;
615 return @{ $self->render_aqt(
616 $self->_expand_update_set_values(undef, $data),
620 sub _expand_update_set_values {
621 my ($self, undef, $data) = @_;
622 $self->expand_expr({ -list => [
625 $set = { -bind => $_ } unless defined $set;
626 +{ -op => [ '=', { -ident => $k }, $set ] };
632 ? ($self->{array_datatypes}
633 ? [ $k, +{ -bind => [ $k, $v ] } ]
634 : [ $k, +{ -literal => $v } ])
636 local our $Cur_Col_Meta = $k;
637 [ $k, $self->_expand_expr($v) ]
644 sub _expand_update_clause_target {
645 my ($self, undef, $target) = @_;
646 +(target => $self->expand_expr({ -list => $target }, -ident));
649 sub _expand_update_clause_set {
650 return $_[2] if ref($_[2]) eq 'HASH' and ($_[2]->{-op}||[''])->[0] eq ',';
651 +(set => $_[0]->_expand_update_set_values($_[1], $_[2]));
654 sub _expand_update_clause_where {
655 +(where => $_[0]->expand_expr($_[2]));
658 sub _expand_update_clause_returning {
659 +(returning => $_[0]->expand_expr({ -list => $_[2] }, -ident));
662 # So that subclasses can override UPDATE ... RETURNING separately from
664 sub _update_returning { shift->_returning(@_) }
668 #======================================================================
670 #======================================================================
673 my ($self, @args) = @_;
675 if (ref(my $sel = $args[0]) eq 'HASH') {
679 @clauses{qw(from select where order_by)} = @args;
681 # This oddity is to literalify since historically SQLA doesn't quote
682 # a single identifier argument, so we convert it into a literal
684 $clauses{select} = { -literal => [ $clauses{select}||'*' ] }
685 unless ref($clauses{select});
690 my @rendered = $self->render_statement({ -select => $stmt });
691 return wantarray ? @rendered : $rendered[0];
694 sub _expand_select_clause_select {
695 my ($self, undef, $select) = @_;
696 +(select => $self->expand_expr({ -list => $select }, -ident));
699 sub _expand_select_clause_from {
700 my ($self, undef, $from) = @_;
701 +(from => $self->expand_expr({ -list => $from }, -ident));
704 sub _expand_select_clause_where {
705 my ($self, undef, $where) = @_;
708 if (my $conv = $self->{convert}) {
720 ->wrap_expander(bind => $_wrap)
721 ->wrap_op_expanders(map +($_ => $_wrap), qw(ident value))
722 ->wrap_expander(func => sub {
725 my ($self, $type, $thing) = @_;
726 if (ref($thing) eq 'ARRAY' and $thing->[0] eq $conv
727 and @$thing == 2 and ref($thing->[1]) eq 'HASH'
730 or $thing->[1]{-value}
731 or $thing->[1]{-bind})
733 return { -func => $thing }; # already went through our expander
735 return $self->$orig($type, $thing);
743 return +(where => $sqla->expand_expr($where));
746 sub _expand_select_clause_order_by {
747 my ($self, undef, $order_by) = @_;
748 +(order_by => $self->_expand_order_by($order_by));
752 my ($self, $fields) = @_;
753 return $fields unless ref($fields);
754 my ($sql, @bind) = @{ $self->render_aqt(
755 $self->expand_expr({ -list => $fields }, '-ident')
757 return wantarray ? ($sql, @bind) : $sql;
760 #======================================================================
762 #======================================================================
765 my ($self, $table, $where, $options) = @_;
768 if (ref($table) eq 'HASH') {
771 my %clauses = (target => $table, where => $where, %{$options||{}});
775 my @rendered = $self->render_statement({ -delete => $stmt });
776 return wantarray ? @rendered : $rendered[0];
779 # So that subclasses can override DELETE ... RETURNING separately from
781 sub _delete_returning { shift->_returning(@_) }
783 sub _expand_delete_clause_target {
784 +(target => $_[0]->expand_expr({ -list => $_[2] }, -ident));
787 sub _expand_delete_clause_where { +(where => $_[0]->expand_expr($_[2])); }
789 sub _expand_delete_clause_returning {
790 +(returning => $_[0]->expand_expr({ -list => $_[2] }, -ident));
793 sub _render_delete_clause_target {
794 my ($self, undef, $from) = @_;
795 $self->join_query_parts(' ', { -keyword => 'delete from' }, $from);
798 #======================================================================
800 #======================================================================
804 # Finally, a separate routine just to handle WHERE clauses
806 my ($self, $where, $order) = @_;
808 local $self->{convert_where} = $self->{convert};
811 my ($sql, @bind) = defined($where)
812 ? $self->_recurse_where($where)
814 $sql = (defined $sql and length $sql) ? $self->_sqlcase(' where ') . "( $sql )" : '';
818 my ($order_sql, @order_bind) = $self->_order_by($order);
820 push @bind, @order_bind;
823 return wantarray ? ($sql, @bind) : $sql;
826 { our $Default_Scalar_To = -value }
829 my ($self, $expr, $default_scalar_to) = @_;
830 local our $Default_Scalar_To = $default_scalar_to if $default_scalar_to;
831 $self->_expand_expr($expr);
835 my ($self, $aqt, $top_level) = @_;
836 my ($k, $v, @rest) = %$aqt;
838 die "Not a node type: $k" unless $k =~ s/^-//;
839 if (my $meth = $self->{render}{$k}) {
840 local our $Render_Top_Level = $top_level;
841 return $self->$meth($k, $v);
843 die "notreached: $k";
847 my ($self, $expr, $default_scalar_to) = @_;
848 return @{ $self->render_aqt(
849 $self->expand_expr($expr, $default_scalar_to)
853 sub render_statement {
854 my ($self, $expr, $default_scalar_to) = @_;
856 $self->expand_expr($expr, $default_scalar_to), 1
860 sub _expand_statement {
861 my ($self, $type, $args) = @_;
862 my $ec = $self->{expand_clause};
865 $args->{$type} = delete $args->{_}
867 my %has_clause = map +($_ => 1), @{$self->{clauses_of}{$type}};
868 return +{ "-${type}" => +{
870 my $val = $args->{$_};
871 if (defined($val) and my $exp = $ec->{"${type}.$_"}) {
872 if ((my (@exp) = $self->$exp($_ => $val)) == 1) {
877 } elsif ($has_clause{$_}) {
878 ($_ => $self->expand_expr($val))
886 sub _render_statement {
887 my ($self, $type, $args) = @_;
889 foreach my $clause (@{$self->{clauses_of}{$type}}) {
890 next unless my $clause_expr = $args->{$clause};
892 if (my $rdr = $self->{render_clause}{"${type}.${clause}"}) {
893 $self->$rdr($clause, $clause_expr, $args);
895 my $r = $self->render_aqt($clause_expr, 1);
896 next unless defined $r->[0] and length $r->[0];
897 $self->join_query_parts(' ',
898 { -keyword => $clause },
905 my $q = $self->join_query_parts(' ', @parts);
906 return $self->join_query_parts('',
907 (our $Render_Top_Level ? $q : ('(', $q, ')'))
912 my ($self, $raw) = @_;
914 return $op if grep $_->{$op}, @{$self}{qw(expand_op render_op)};
915 s/^-(?=.)//, s/\s+/_/g for $op;
920 my ($self, $expr) = @_;
921 our $Expand_Depth ||= 0; local $Expand_Depth = $Expand_Depth + 1;
922 return undef unless defined($expr);
923 if (ref($expr) eq 'HASH') {
924 return undef unless my $kc = keys %$expr;
926 return $self->_expand_op_andor(and => $expr);
928 my ($key, $value) = %$expr;
929 if ($key =~ /^-/ and $key =~ s/ [_\s]? \d+ $//x ) {
930 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
931 . "You probably wanted ...-and => [ $key => COND1, $key => COND2 ... ]";
933 return $self->_expand_hashpair($key, $value);
935 if (ref($expr) eq 'ARRAY') {
936 return $self->_expand_op_andor(lc($self->{logic}), $expr);
938 if (my $literal = is_literal_value($expr)) {
939 return +{ -literal => $literal };
941 if (!ref($expr) or Scalar::Util::blessed($expr)) {
942 return $self->_expand_scalar($expr);
947 sub _expand_hashpair {
948 my ($self, $k, $v) = @_;
949 unless (defined($k) and length($k)) {
950 if (defined($k) and my $literal = is_literal_value($v)) {
951 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
952 return { -literal => $literal };
954 puke "Supplying an empty left hand side argument is not supported";
957 return $self->_expand_hashpair_op($k, $v);
958 } elsif ($k =~ /^[^\w]/i) {
959 my ($lhs, @rhs) = ref($v) eq 'ARRAY' ? @$v : $v;
960 return $self->_expand_op(
961 -op, [ $k, $self->expand_expr($lhs, -ident), @rhs ]
964 return $self->_expand_hashpair_ident($k, $v);
967 sub _expand_hashpair_ident {
968 my ($self, $k, $v) = @_;
970 local our $Cur_Col_Meta = $k;
972 # hash with multiple or no elements is andor
974 if (ref($v) eq 'HASH' and keys %$v != 1) {
975 return $self->_expand_op_andor(and => $v, $k);
978 # undef needs to be re-sent with cmp to achieve IS/IS NOT NULL
980 if (is_undef_value($v)) {
981 return $self->_expand_hashpair_cmp($k => undef);
984 # scalars and objects get expanded as whatever requested or values
986 if (!ref($v) or Scalar::Util::blessed($v)) {
987 return $self->_expand_hashpair_scalar($k, $v);
990 # single key hashref is a hashtriple
992 if (ref($v) eq 'HASH') {
993 return $self->_expand_hashtriple($k, %$v);
996 # arrayref needs re-engineering over the elements
998 if (ref($v) eq 'ARRAY') {
999 return $self->sqlfalse unless @$v;
1000 $self->_debug("ARRAY($k) means distribute over elements");
1002 $v->[0] =~ /^-(and|or)$/i
1003 ? (shift(@{$v = [ @$v ]}), $1)
1004 : lc($self->{logic} || 'OR')
1006 return $self->_expand_op_andor(
1011 if (my $literal = is_literal_value($v)) {
1012 unless (length $k) {
1013 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
1016 my ($sql, @bind) = @$literal;
1017 if ($self->{bindtype} eq 'columns') {
1019 $self->_assert_bindval_matches_bindtype($_);
1022 return +{ -literal => [ $self->_quote($k).' '.$sql, @bind ] };
1027 sub _expand_scalar {
1028 my ($self, $expr) = @_;
1030 return $self->_expand_expr({ (our $Default_Scalar_To) => $expr });
1033 sub _expand_hashpair_scalar {
1034 my ($self, $k, $v) = @_;
1036 return $self->_expand_hashpair_cmp(
1037 $k, $self->_expand_scalar($v),
1041 sub _expand_hashpair_op {
1042 my ($self, $k, $v) = @_;
1044 $self->_assert_pass_injection_guard($k =~ /\A-(.*)\Z/s);
1046 my $op = $self->_normalize_op($k);
1048 my $wsop = join(' ', split '_', $op);
1050 my $is_special = List::Util::first { $wsop =~ $_->{regex} }
1051 @{$self->{special_ops}};
1055 # the old special op system requires illegality for top-level use
1058 (our $Expand_Depth) == 1
1062 $self->{disable_old_special_ops}
1063 and List::Util::first { $wsop =~ $_->{regex} } @BUILTIN_SPECIAL_OPS
1067 puke "Illegal use of top-level '-$wsop'"
1071 if (my $exp = $self->{expand}{$op}||$self->{expand_op}{$op}) {
1072 return $self->$exp($op, $v);
1075 if ($self->{render}{$op}) {
1076 return { "-${op}" => $v };
1079 # Ops prefixed with -not_ get converted
1081 if (my ($rest) = $op =~/^not_(.*)$/) {
1084 $self->_expand_expr({ "-${rest}", $v })
1090 # the old unary op system means we should touch nothing and let it work
1092 my $op = join(' ', split '_', $op);
1094 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1095 return { -op => [ $op, $v ] };
1099 my $type = $is_special || $self->{render_op}{$op} ? -op : -func;
1101 if ($self->{restore_old_unop_handling}) {
1108 and (keys %$v)[0] =~ /^-/
1109 and not $self->{render_op}{$op}
1118 if ($type eq -func and ref($v) eq 'ARRAY') {
1119 return $self->_expand_expr({ -func => [ $op, @$v ] });
1122 return $self->_expand_expr({ $type => [ $op, $v ] });
1125 sub _expand_hashpair_cmp {
1126 my ($self, $k, $v) = @_;
1127 $self->_expand_hashtriple($k, $self->{cmp}, $v);
1130 sub _expand_hashtriple {
1131 my ($self, $k, $vk, $vv) = @_;
1133 my $ik = $self->_expand_expr({ -ident => $k });
1135 my $op = $self->_normalize_op($vk);
1136 $self->_assert_pass_injection_guard($op);
1138 if ($op =~ s/ _? \d+ $//x ) {
1139 return $self->_expand_expr($k, { $vk, $vv });
1141 if (my $x = $self->{expand_op}{$op}) {
1142 local our $Cur_Col_Meta = $k;
1143 return $self->$x($op, $vv, $k);
1147 my $op = join(' ', split '_', $op);
1149 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}) {
1150 return { -op => [ $op, $ik, $vv ] };
1152 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1156 { -op => [ $op, $vv ] }
1160 if (ref($vv) eq 'ARRAY') {
1162 my $logic = (defined($raw[0]) and $raw[0] =~ /^-(and|or)$/i)
1163 ? (shift(@raw), lc $1) : 'or';
1164 my @values = map +{ $vk => $_ }, @raw;
1166 $op =~ $self->{inequality_op}
1167 or $op =~ $self->{not_like_op}
1169 if (lc($logic) eq 'or' and @values > 1) {
1170 belch "A multi-element arrayref as an argument to the inequality op '${\uc(join ' ', split '_', $op)}' "
1171 . 'is technically equivalent to an always-true 1=1 (you probably wanted '
1172 . "to say ...{ \$inequality_op => [ -and => \@values ] }... instead)"
1177 # try to DWIM on equality operators
1178 return ($self->_dwim_op_to_is($op,
1179 "Supplying an empty arrayref to '%s' is deprecated",
1180 "operator '%s' applied on an empty array (field '$k')"
1181 ) ? $self->sqlfalse : $self->sqltrue);
1183 return $self->_expand_op_andor($logic => \@values, $k);
1185 if (is_undef_value($vv)) {
1186 my $is = ($self->_dwim_op_to_is($op,
1187 "Supplying an undefined argument to '%s' is deprecated",
1188 "unexpected operator '%s' with undef operand",
1189 ) ? 'is' : 'is not');
1191 return $self->_expand_hashpair($k => { $is, undef });
1193 local our $Cur_Col_Meta = $k;
1197 $self->_expand_expr($vv)
1201 sub _dwim_op_to_is {
1202 my ($self, $raw, $empty, $fail) = @_;
1204 my $op = $self->_normalize_op($raw);
1206 if ($op =~ /^not$/i) {
1209 if ($op =~ $self->{equality_op}) {
1212 if ($op =~ $self->{like_op}) {
1213 belch(sprintf $empty, uc(join ' ', split '_', $op));
1216 if ($op =~ $self->{inequality_op}) {
1219 if ($op =~ $self->{not_like_op}) {
1220 belch(sprintf $empty, uc(join ' ', split '_', $op));
1223 puke(sprintf $fail, $op);
1227 my ($self, undef, $args) = @_;
1228 my ($func, @args) = @$args;
1229 return +{ -func => [ $func, map $self->expand_expr($_), @args ] };
1233 my ($self, undef, $body) = @_;
1234 unless (defined($body) or (ref($body) and ref($body) eq 'ARRAY')) {
1235 puke "-ident requires a single plain scalar argument (a quotable identifier) or an arrayref of identifier parts";
1237 my @parts = map split(/\Q${\($self->{name_sep}||'.')}\E/, $_),
1238 ref($body) ? @$body : $body;
1239 return { -ident => $parts[-1] } if $self->{_dequalify_idents};
1240 unless ($self->{quote_char}) {
1241 $self->_assert_pass_injection_guard($_) for @parts;
1243 return +{ -ident => \@parts };
1247 +{ -bind => [ our $Cur_Col_Meta, $_[2] ] };
1251 my ($self, undef, $args) = @_;
1252 +{ -row => [ map $self->expand_expr($_), @$args ] };
1256 my ($self, undef, $args) = @_;
1257 my ($op, @opargs) = @$args;
1258 if (my $exp = $self->{expand_op}{$op}) {
1259 return $self->$exp($op, \@opargs);
1261 if (List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1262 return { -op => [ $op, @opargs ] };
1264 +{ -op => [ $op, map $self->expand_expr($_), @opargs ] };
1268 my ($self, undef, $v) = @_;
1270 return $self->_expand_expr($v);
1272 puke "-bool => undef not supported" unless defined($v);
1273 return $self->_expand_expr({ -ident => $v });
1277 my ($self, undef, $expr) = @_;
1279 ',', map $self->expand_expr($_),
1280 @{$expr->{-op}}[1..$#{$expr->{-op}}]
1281 ] } if ref($expr) eq 'HASH' and ($expr->{-op}||[''])->[0] eq ',';
1282 return +{ -op => [ ',',
1283 map $self->expand_expr($_),
1284 ref($expr) eq 'ARRAY' ? @$expr : $expr
1288 sub _expand_op_andor {
1289 my ($self, $logop, $v, $k) = @_;
1291 $v = [ map +{ $k, $_ },
1293 ? (map +{ $_ => $v->{$_} }, sort keys %$v)
1297 if (ref($v) eq 'HASH') {
1298 return undef unless keys %$v;
1301 map $self->_expand_expr({ $_ => $v->{$_} }),
1305 if (ref($v) eq 'ARRAY') {
1306 $logop eq 'and' or $logop eq 'or' or puke "unknown logic: $logop";
1309 (ref($_) eq 'ARRAY' and @$_)
1310 or (ref($_) eq 'HASH' and %$_)
1316 while (my ($el) = splice @expr, 0, 1) {
1317 puke "Supplying an empty left hand side argument is not supported in array-pairs"
1318 unless defined($el) and length($el);
1319 my $elref = ref($el);
1321 local our $Expand_Depth = 0;
1322 push(@res, grep defined, $self->_expand_expr({ $el, shift(@expr) }));
1323 } elsif ($elref eq 'ARRAY') {
1324 push(@res, grep defined, $self->_expand_expr($el)) if @$el;
1325 } elsif (my $l = is_literal_value($el)) {
1326 push @res, { -literal => $l };
1327 } elsif ($elref eq 'HASH') {
1328 local our $Expand_Depth = 0;
1329 push @res, grep defined, $self->_expand_expr($el) if %$el;
1335 # return $res[0] if @res == 1;
1336 return { -op => [ $logop, @res ] };
1342 my ($self, $op, $vv, $k) = @_;
1343 ($k, $vv) = @$vv unless defined $k;
1344 puke "$op can only take undef as argument"
1348 and exists($vv->{-value})
1349 and !defined($vv->{-value})
1351 return +{ -op => [ $op.'_null', $self->expand_expr($k, -ident) ] };
1354 sub _expand_between {
1355 my ($self, $op, $vv, $k) = @_;
1356 my @rhs = map $self->_expand_expr($_),
1357 ref($vv) eq 'ARRAY' ? @$vv : $vv;
1359 (@rhs == 1 and ref($rhs[0]) eq 'HASH' and $rhs[0]->{-literal})
1361 (@rhs == 2 and defined($rhs[0]) and defined($rhs[1]))
1363 puke "Operator '${\uc($op)}' requires either an arrayref with two defined values or expressions, or a single literal scalarref/arrayref-ref";
1367 $self->expand_expr($k),
1368 map $self->expand_expr($_, -value), @rhs
1373 my ($self, $op, $vv, $k) = @_;
1374 if (my $literal = is_literal_value($vv)) {
1375 my ($sql, @bind) = @$literal;
1376 my $opened_sql = $self->_open_outer_paren($sql);
1378 $op, $self->expand_expr($k, -ident),
1379 { -literal => [ $opened_sql, @bind ] }
1383 'SQL::Abstract before v1.75 used to generate incorrect SQL when the '
1384 . "-${\uc($op)} operator was given an undef-containing list: !!!AUDIT YOUR CODE "
1385 . 'AND DATA!!! (the upcoming Data::Query-based version of SQL::Abstract '
1386 . 'will emit the logically correct SQL instead of raising this exception)'
1388 puke("Argument passed to the '${\uc($op)}' operator can not be undefined")
1390 my @rhs = map $self->expand_expr($_, -value),
1391 map { defined($_) ? $_: puke($undef_err) }
1392 (ref($vv) eq 'ARRAY' ? @$vv : $vv);
1393 return $self->${\($op =~ /^not/ ? 'sqltrue' : 'sqlfalse')} unless @rhs;
1397 $self->expand_expr($k, -ident),
1403 my ($self, undef, $v) = @_;
1404 # DBIx::Class requires a nest warning to be emitted once but the private
1405 # method it overrode to do so no longer exists
1406 if ($self->{warn_once_on_nest}) {
1407 unless (our $Nest_Warned) {
1409 "-nest in search conditions is deprecated, you most probably wanted:\n"
1410 .q|{..., -and => [ \%cond0, \@cond1, \'cond2', \[ 'cond3', [ col => bind ] ], etc. ], ... }|
1415 return $self->_expand_expr($v);
1418 sub _expand_values {
1419 my ($self, undef, $values) = @_;
1420 return { -values => [
1423 ? $self->expand_expr($_)
1424 : +{ -row => [ map $self->expand_expr($_), @$_ ] }
1425 ), ref($values) eq 'ARRAY' ? @$values : $values
1429 sub _recurse_where {
1430 my ($self, $where) = @_;
1432 # Special case: top level simple string treated as literal
1434 my $where_exp = (ref($where)
1435 ? $self->_expand_select_clause_where(undef, $where)
1436 : { -literal => [ $where ] });
1438 # dispatch expanded expression
1440 my ($sql, @bind) = defined($where_exp) ? @{ $self->render_aqt($where_exp) || [] } : ();
1441 # DBIx::Class used to call _recurse_where in scalar context
1442 # something else might too...
1444 return ($sql, @bind);
1447 belch "Calling _recurse_where in scalar context is deprecated and will go away before 2.0";
1453 my ($self, undef, $ident) = @_;
1455 return [ $self->_quote($ident) ];
1459 my ($self, undef, $values) = @_;
1460 return $self->join_query_parts('',
1462 $self->_render_op(undef, [ ',', @$values ]),
1468 my ($self, undef, $rest) = @_;
1469 my ($func, @args) = @$rest;
1470 return $self->join_query_parts('',
1471 $self->_sqlcase($func),
1472 $self->join_query_parts('',
1474 $self->join_query_parts(', ', @args),
1481 my ($self, undef, $bind) = @_;
1482 return [ '?', $self->_bindtype(@$bind) ];
1485 sub _render_literal {
1486 my ($self, undef, $literal) = @_;
1487 $self->_assert_bindval_matches_bindtype(@{$literal}[1..$#$literal]);
1491 sub _render_keyword {
1492 my ($self, undef, $keyword) = @_;
1493 return [ $self->_sqlcase(
1494 ref($keyword) ? $$keyword : join ' ', split '_', $keyword
1499 my ($self, undef, $v) = @_;
1500 my ($op, @args) = @$v;
1501 if (my $r = $self->{render_op}{$op}) {
1502 return $self->$r($op, \@args);
1507 my $op = join(' ', split '_', $op);
1509 my $ss = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}};
1510 if ($ss and @args > 1) {
1511 puke "Special op '${op}' requires first value to be identifier"
1512 unless my ($ident) = map $_->{-ident}, grep ref($_) eq 'HASH', $args[0];
1513 my $k = join(($self->{name_sep}||'.'), @$ident);
1514 local our $Expand_Depth = 1;
1515 return [ $self->${\($ss->{handler})}($k, $op, $args[1]) ];
1517 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1518 return [ $self->${\($us->{handler})}($op, $args[0]) ];
1521 return $self->_render_unop_paren($op, \@args);
1525 return $self->_render_unop_prefix($op, \@args);
1527 return $self->_render_op_multop($op, \@args);
1533 sub _render_op_between {
1534 my ($self, $op, $args) = @_;
1535 my ($left, $low, $high) = @$args;
1538 puke "Single arg to between must be a literal"
1539 unless $low->{-literal};
1542 +($low, { -keyword => 'and' }, $high);
1545 return $self->join_query_parts(' ',
1546 '(', $left, { -keyword => $op }, @rh, ')',
1551 my ($self, $op, $args) = @_;
1552 my ($lhs, @rhs) = @$args;
1554 return $self->join_query_parts(' ',
1556 { -keyword => $op },
1557 $self->join_query_parts(' ',
1559 $self->join_query_parts(', ', @rhs),
1565 sub _render_op_andor {
1566 my ($self, $op, $args) = @_;
1567 return undef unless @$args;
1568 return $self->join_query_parts('', $args->[0]) if @$args == 1;
1569 my $inner = $self->_render_op_multop($op, $args);
1570 return undef unless defined($inner->[0]) and length($inner->[0]);
1571 return $self->join_query_parts(' ',
1576 sub _render_op_multop {
1577 my ($self, $op, $args) = @_;
1579 return undef unless @parts;
1580 return $self->render_aqt($parts[0]) if @parts == 1;
1581 my $join = ($op eq ','
1583 : { -keyword => " ${op} " }
1585 return $self->join_query_parts($join, @parts);
1588 sub _render_values {
1589 my ($self, undef, $values) = @_;
1590 my $inner = $self->join_query_parts(' ',
1591 { -keyword => 'values' },
1592 $self->join_query_parts(', ',
1593 ref($values) eq 'ARRAY' ? @$values : $values
1596 return $self->join_query_parts('',
1597 (our $Render_Top_Level ? $inner : ('(', $inner, ')'))
1601 sub join_query_parts {
1602 my ($self, $join, @parts) = @_;
1603 if (ref($join) eq 'HASH') {
1604 $join = $self->render_aqt($join)->[0];
1608 ? $self->render_aqt($_)
1609 : ((ref($_) eq 'ARRAY') ? $_ : [ $_ ])
1612 $self->{join_sql_parts}->(
1613 $join, grep defined && length, map $_->[0], @final
1615 (map @{$_}[1..$#$_], @final),
1619 sub _render_unop_paren {
1620 my ($self, $op, $v) = @_;
1621 return $self->join_query_parts('',
1622 '(', $self->_render_unop_prefix($op, $v), ')'
1626 sub _render_unop_prefix {
1627 my ($self, $op, $v) = @_;
1628 my $op_sql = $self->{restore_old_unop_handling}
1629 ? $self->_sqlcase($op)
1630 : { -keyword => $op };
1631 return $self->join_query_parts(' ',
1632 ($self->{restore_old_unop_handling}
1633 ? $self->_sqlcase($op)
1634 : { -keyword => \$op }),
1639 sub _render_unop_postfix {
1640 my ($self, $op, $v) = @_;
1641 return $self->join_query_parts(' ',
1642 $v->[0], { -keyword => $op },
1646 # Some databases (SQLite) treat col IN (1, 2) different from
1647 # col IN ( (1, 2) ). Use this to strip all outer parens while
1648 # adding them back in the corresponding method
1649 sub _open_outer_paren {
1650 my ($self, $sql) = @_;
1652 while (my ($inner) = $sql =~ /^ \s* \( (.*) \) \s* $/xs) {
1654 # there are closing parens inside, need the heavy duty machinery
1655 # to reevaluate the extraction starting from $sql (full reevaluation)
1656 if ($inner =~ /\)/) {
1657 require Text::Balanced;
1659 my (undef, $remainder) = do {
1660 # idiotic design - writes to $@ but *DOES NOT* throw exceptions
1662 Text::Balanced::extract_bracketed($sql, '()', qr/\s*/);
1665 # the entire expression needs to be a balanced bracketed thing
1666 # (after an extract no remainder sans trailing space)
1667 last if defined $remainder and $remainder =~ /\S/;
1677 #======================================================================
1679 #======================================================================
1681 sub _expand_order_by {
1682 my ($self, $arg) = @_;
1684 return unless defined($arg) and not (ref($arg) eq 'ARRAY' and !@$arg);
1686 return $self->expand_expr({ -list => $arg })
1687 if ref($arg) eq 'HASH' and ($arg->{-op}||[''])->[0] eq ',';
1689 my $expander = sub {
1690 my ($self, $dir, $expr) = @_;
1691 my @to_expand = ref($expr) eq 'ARRAY' ? @$expr : $expr;
1692 foreach my $arg (@to_expand) {
1696 and grep /^-(asc|desc)$/, keys %$arg
1698 puke "ordering direction hash passed to order by must have exactly one key (-asc or -desc)";
1702 defined($dir) ? { -op => [ $dir =~ /^-?(.*)$/ ,=> $_ ] } : $_
1704 map $self->expand_expr($_, -ident),
1705 map ref($_) eq 'ARRAY' ? @$_ : $_, @to_expand;
1706 return undef unless @exp;
1707 return undef if @exp == 1 and not defined($exp[0]);
1708 return +{ -op => [ ',', @exp ] };
1711 local @{$self->{expand}}{qw(asc desc)} = (($expander) x 2);
1713 return $self->$expander(undef, $arg);
1717 my ($self, $arg) = @_;
1719 return '' unless defined(my $expanded = $self->_expand_order_by($arg));
1721 my ($sql, @bind) = @{ $self->render_aqt($expanded) };
1723 return '' unless length($sql);
1725 my $final_sql = $self->_sqlcase(' order by ').$sql;
1727 return $final_sql unless wantarray;
1729 return ($final_sql, @bind);
1732 # _order_by no longer needs to call this so doesn't but DBIC uses it.
1734 sub _order_by_chunks {
1735 my ($self, $arg) = @_;
1737 return () unless defined(my $expanded = $self->_expand_order_by($arg));
1739 my @res = $self->_chunkify_order_by($expanded);
1740 (ref() ? $_->[0] : $_) .= '' for @res;
1744 sub _chunkify_order_by {
1745 my ($self, $expanded) = @_;
1747 return grep length, @{ $self->render_aqt($expanded) }
1748 if $expanded->{-ident} or @{$expanded->{-literal}||[]} == 1;
1751 if (ref() eq 'HASH' and $_->{-op} and $_->{-op}[0] eq ',') {
1752 my ($comma, @list) = @{$_->{-op}};
1753 return map $self->_chunkify_order_by($_), @list;
1755 return $self->render_aqt($_);
1759 #======================================================================
1760 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
1761 #======================================================================
1767 $self->expand_expr({ -list => $from }, -ident)
1772 #======================================================================
1774 #======================================================================
1776 # highly optimized, as it's called way too often
1778 # my ($self, $label) = @_;
1780 return '' unless defined $_[1];
1781 return ${$_[1]} if ref($_[1]) eq 'SCALAR';
1782 puke 'Identifier cannot be hashref' if ref($_[1]) eq 'HASH';
1784 unless ($_[0]->{quote_char}) {
1785 if (ref($_[1]) eq 'ARRAY') {
1786 return join($_[0]->{name_sep}||'.', @{$_[1]});
1788 $_[0]->_assert_pass_injection_guard($_[1]);
1793 my $qref = ref $_[0]->{quote_char};
1795 !$qref ? ($_[0]->{quote_char}, $_[0]->{quote_char})
1796 : ($qref eq 'ARRAY') ? @{$_[0]->{quote_char}}
1797 : puke "Unsupported quote_char format: $_[0]->{quote_char}";
1799 my $esc = $_[0]->{escape_char} || $r;
1801 # parts containing * are naturally unquoted
1803 $_[0]->{name_sep}||'',
1807 : do { (my $n = $_) =~ s/(\Q$esc\E|\Q$r\E)/$esc$1/g; $l . $n . $r }
1809 (ref($_[1]) eq 'ARRAY'
1813 ? split (/\Q$_[0]->{name_sep}\E/, $_[1] )
1821 # Conversion, if applicable
1823 #my ($self, $arg) = @_;
1824 if (my $conv = $_[0]->{convert_where}) {
1825 return @{ $_[0]->join_query_parts('',
1826 $_[0]->_sqlcase($conv),
1835 #my ($self, $col, @vals) = @_;
1836 # called often - tighten code
1837 return $_[0]->{bindtype} eq 'columns'
1838 ? map {[$_[1], $_]} @_[2 .. $#_]
1843 # Dies if any element of @bind is not in [colname => value] format
1844 # if bindtype is 'columns'.
1845 sub _assert_bindval_matches_bindtype {
1846 # my ($self, @bind) = @_;
1848 if ($self->{bindtype} eq 'columns') {
1850 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
1851 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
1857 # Fix SQL case, if so requested
1859 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
1860 # don't touch the argument ... crooked logic, but let's not change it!
1861 return $_[0]->{case} ? $_[1] : uc($_[1]);
1864 #======================================================================
1865 # DISPATCHING FROM REFKIND
1866 #======================================================================
1869 my ($self, $data) = @_;
1871 return 'UNDEF' unless defined $data;
1873 # blessed objects are treated like scalars
1874 my $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1876 return 'SCALAR' unless $ref;
1879 while ($ref eq 'REF') {
1881 $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1885 return ($ref||'SCALAR') . ('REF' x $n_steps);
1889 my ($self, $data) = @_;
1890 my @try = ($self->_refkind($data));
1891 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
1892 push @try, 'FALLBACK';
1896 sub _METHOD_FOR_refkind {
1897 my ($self, $meth_prefix, $data) = @_;
1900 for (@{$self->_try_refkind($data)}) {
1901 $method = $self->can($meth_prefix."_".$_)
1905 return $method || puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
1909 sub _SWITCH_refkind {
1910 my ($self, $data, $dispatch_table) = @_;
1913 for (@{$self->_try_refkind($data)}) {
1914 $coderef = $dispatch_table->{$_}
1918 puke "no dispatch entry for ".$self->_refkind($data)
1927 #======================================================================
1928 # VALUES, GENERATE, AUTOLOAD
1929 #======================================================================
1931 # LDNOTE: original code from nwiger, didn't touch code in that section
1932 # I feel the AUTOLOAD stuff should not be the default, it should
1933 # only be activated on explicit demand by user.
1937 my $data = shift || return;
1938 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
1939 unless ref $data eq 'HASH';
1942 foreach my $k (sort keys %$data) {
1943 my $v = $data->{$k};
1944 $self->_SWITCH_refkind($v, {
1946 if ($self->{array_datatypes}) { # array datatype
1947 push @all_bind, $self->_bindtype($k, $v);
1949 else { # literal SQL with bind
1950 my ($sql, @bind) = @$v;
1951 $self->_assert_bindval_matches_bindtype(@bind);
1952 push @all_bind, @bind;
1955 ARRAYREFREF => sub { # literal SQL with bind
1956 my ($sql, @bind) = @${$v};
1957 $self->_assert_bindval_matches_bindtype(@bind);
1958 push @all_bind, @bind;
1960 SCALARREF => sub { # literal SQL without bind
1962 SCALAR_or_UNDEF => sub {
1963 push @all_bind, $self->_bindtype($k, $v);
1974 my(@sql, @sqlq, @sqlv);
1978 if ($ref eq 'HASH') {
1979 for my $k (sort keys %$_) {
1982 my $label = $self->_quote($k);
1983 if ($r eq 'ARRAY') {
1984 # literal SQL with bind
1985 my ($sql, @bind) = @$v;
1986 $self->_assert_bindval_matches_bindtype(@bind);
1987 push @sqlq, "$label = $sql";
1989 } elsif ($r eq 'SCALAR') {
1990 # literal SQL without bind
1991 push @sqlq, "$label = $$v";
1993 push @sqlq, "$label = ?";
1994 push @sqlv, $self->_bindtype($k, $v);
1997 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
1998 } elsif ($ref eq 'ARRAY') {
1999 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
2002 if ($r eq 'ARRAY') { # literal SQL with bind
2003 my ($sql, @bind) = @$v;
2004 $self->_assert_bindval_matches_bindtype(@bind);
2007 } elsif ($r eq 'SCALAR') { # literal SQL without bind
2008 # embedded literal SQL
2015 push @sql, '(' . join(', ', @sqlq) . ')';
2016 } elsif ($ref eq 'SCALAR') {
2020 # strings get case twiddled
2021 push @sql, $self->_sqlcase($_);
2025 my $sql = join ' ', @sql;
2027 # this is pretty tricky
2028 # if ask for an array, return ($stmt, @bind)
2029 # otherwise, s/?/shift @sqlv/ to put it inline
2031 return ($sql, @sqlv);
2033 1 while $sql =~ s/\?/my $d = shift(@sqlv);
2034 ref $d ? $d->[1] : $d/e;
2043 # This allows us to check for a local, then _form, attr
2045 my($name) = $AUTOLOAD =~ /.*::(.+)/;
2046 puke "AUTOLOAD invoked for method name ${name} and allow_autoload option not set" unless $self->{allow_autoload};
2047 return $self->generate($name, @_);
2058 SQL::Abstract - Generate SQL from Perl data structures
2064 my $sql = SQL::Abstract->new;
2066 my($stmt, @bind) = $sql->select($source, \@fields, \%where, $order);
2068 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
2070 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
2072 my($stmt, @bind) = $sql->delete($table, \%where);
2074 # Then, use these in your DBI statements
2075 my $sth = $dbh->prepare($stmt);
2076 $sth->execute(@bind);
2078 # Just generate the WHERE clause
2079 my($stmt, @bind) = $sql->where(\%where, $order);
2081 # Return values in the same order, for hashed queries
2082 # See PERFORMANCE section for more details
2083 my @bind = $sql->values(\%fieldvals);
2087 This module was inspired by the excellent L<DBIx::Abstract>.
2088 However, in using that module I found that what I really wanted
2089 to do was generate SQL, but still retain complete control over my
2090 statement handles and use the DBI interface. So, I set out to
2091 create an abstract SQL generation module.
2093 While based on the concepts used by L<DBIx::Abstract>, there are
2094 several important differences, especially when it comes to WHERE
2095 clauses. I have modified the concepts used to make the SQL easier
2096 to generate from Perl data structures and, IMO, more intuitive.
2097 The underlying idea is for this module to do what you mean, based
2098 on the data structures you provide it. The big advantage is that
2099 you don't have to modify your code every time your data changes,
2100 as this module figures it out.
2102 To begin with, an SQL INSERT is as easy as just specifying a hash
2103 of C<key=value> pairs:
2106 name => 'Jimbo Bobson',
2107 phone => '123-456-7890',
2108 address => '42 Sister Lane',
2109 city => 'St. Louis',
2110 state => 'Louisiana',
2113 The SQL can then be generated with this:
2115 my($stmt, @bind) = $sql->insert('people', \%data);
2117 Which would give you something like this:
2119 $stmt = "INSERT INTO people
2120 (address, city, name, phone, state)
2121 VALUES (?, ?, ?, ?, ?)";
2122 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
2123 '123-456-7890', 'Louisiana');
2125 These are then used directly in your DBI code:
2127 my $sth = $dbh->prepare($stmt);
2128 $sth->execute(@bind);
2130 =head2 Inserting and Updating Arrays
2132 If your database has array types (like for example Postgres),
2133 activate the special option C<< array_datatypes => 1 >>
2134 when creating the C<SQL::Abstract> object.
2135 Then you may use an arrayref to insert and update database array types:
2137 my $sql = SQL::Abstract->new(array_datatypes => 1);
2139 planets => [qw/Mercury Venus Earth Mars/]
2142 my($stmt, @bind) = $sql->insert('solar_system', \%data);
2146 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
2148 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
2151 =head2 Inserting and Updating SQL
2153 In order to apply SQL functions to elements of your C<%data> you may
2154 specify a reference to an arrayref for the given hash value. For example,
2155 if you need to execute the Oracle C<to_date> function on a value, you can
2156 say something like this:
2160 date_entered => \[ "to_date(?,'MM/DD/YYYY')", "03/02/2003" ],
2163 The first value in the array is the actual SQL. Any other values are
2164 optional and would be included in the bind values array. This gives
2167 my($stmt, @bind) = $sql->insert('people', \%data);
2169 $stmt = "INSERT INTO people (name, date_entered)
2170 VALUES (?, to_date(?,'MM/DD/YYYY'))";
2171 @bind = ('Bill', '03/02/2003');
2173 An UPDATE is just as easy, all you change is the name of the function:
2175 my($stmt, @bind) = $sql->update('people', \%data);
2177 Notice that your C<%data> isn't touched; the module will generate
2178 the appropriately quirky SQL for you automatically. Usually you'll
2179 want to specify a WHERE clause for your UPDATE, though, which is
2180 where handling C<%where> hashes comes in handy...
2182 =head2 Complex where statements
2184 This module can generate pretty complicated WHERE statements
2185 easily. For example, simple C<key=value> pairs are taken to mean
2186 equality, and if you want to see if a field is within a set
2187 of values, you can use an arrayref. Let's say we wanted to
2188 SELECT some data based on this criteria:
2191 requestor => 'inna',
2192 worker => ['nwiger', 'rcwe', 'sfz'],
2193 status => { '!=', 'completed' }
2196 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
2198 The above would give you something like this:
2200 $stmt = "SELECT * FROM tickets WHERE
2201 ( requestor = ? ) AND ( status != ? )
2202 AND ( worker = ? OR worker = ? OR worker = ? )";
2203 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
2205 Which you could then use in DBI code like so:
2207 my $sth = $dbh->prepare($stmt);
2208 $sth->execute(@bind);
2214 The methods are simple. There's one for every major SQL operation,
2215 and a constructor you use first. The arguments are specified in a
2216 similar order for each method (table, then fields, then a where
2217 clause) to try and simplify things.
2219 =head2 new(option => 'value')
2221 The C<new()> function takes a list of options and values, and returns
2222 a new B<SQL::Abstract> object which can then be used to generate SQL
2223 through the methods below. The options accepted are:
2229 If set to 'lower', then SQL will be generated in all lowercase. By
2230 default SQL is generated in "textbook" case meaning something like:
2232 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
2234 Any setting other than 'lower' is ignored.
2238 This determines what the default comparison operator is. By default
2239 it is C<=>, meaning that a hash like this:
2241 %where = (name => 'nwiger', email => 'nate@wiger.org');
2243 Will generate SQL like this:
2245 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
2247 However, you may want loose comparisons by default, so if you set
2248 C<cmp> to C<like> you would get SQL such as:
2250 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
2252 You can also override the comparison on an individual basis - see
2253 the huge section on L</"WHERE CLAUSES"> at the bottom.
2255 =item sqltrue, sqlfalse
2257 Expressions for inserting boolean values within SQL statements.
2258 By default these are C<1=1> and C<1=0>. They are used
2259 by the special operators C<-in> and C<-not_in> for generating
2260 correct SQL even when the argument is an empty array (see below).
2264 This determines the default logical operator for multiple WHERE
2265 statements in arrays or hashes. If absent, the default logic is "or"
2266 for arrays, and "and" for hashes. This means that a WHERE
2270 event_date => {'>=', '2/13/99'},
2271 event_date => {'<=', '4/24/03'},
2274 will generate SQL like this:
2276 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
2278 This is probably not what you want given this query, though (look
2279 at the dates). To change the "OR" to an "AND", simply specify:
2281 my $sql = SQL::Abstract->new(logic => 'and');
2283 Which will change the above C<WHERE> to:
2285 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
2287 The logic can also be changed locally by inserting
2288 a modifier in front of an arrayref:
2290 @where = (-and => [event_date => {'>=', '2/13/99'},
2291 event_date => {'<=', '4/24/03'} ]);
2293 See the L</"WHERE CLAUSES"> section for explanations.
2297 This will automatically convert comparisons using the specified SQL
2298 function for both column and value. This is mostly used with an argument
2299 of C<upper> or C<lower>, so that the SQL will have the effect of
2300 case-insensitive "searches". For example, this:
2302 $sql = SQL::Abstract->new(convert => 'upper');
2303 %where = (keywords => 'MaKe iT CAse inSeNSItive');
2305 Will turn out the following SQL:
2307 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
2309 The conversion can be C<upper()>, C<lower()>, or any other SQL function
2310 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
2311 not validate this option; it will just pass through what you specify verbatim).
2315 This is a kludge because many databases suck. For example, you can't
2316 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
2317 Instead, you have to use C<bind_param()>:
2319 $sth->bind_param(1, 'reg data');
2320 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
2322 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
2323 which loses track of which field each slot refers to. Fear not.
2325 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
2326 Currently, you can specify either C<normal> (default) or C<columns>. If you
2327 specify C<columns>, you will get an array that looks like this:
2329 my $sql = SQL::Abstract->new(bindtype => 'columns');
2330 my($stmt, @bind) = $sql->insert(...);
2333 [ 'column1', 'value1' ],
2334 [ 'column2', 'value2' ],
2335 [ 'column3', 'value3' ],
2338 You can then iterate through this manually, using DBI's C<bind_param()>.
2340 $sth->prepare($stmt);
2343 my($col, $data) = @$_;
2344 if ($col eq 'details' || $col eq 'comments') {
2345 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
2346 } elsif ($col eq 'image') {
2347 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
2349 $sth->bind_param($i, $data);
2353 $sth->execute; # execute without @bind now
2355 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
2356 Basically, the advantage is still that you don't have to care which fields
2357 are or are not included. You could wrap that above C<for> loop in a simple
2358 sub called C<bind_fields()> or something and reuse it repeatedly. You still
2359 get a layer of abstraction over manual SQL specification.
2361 Note that if you set L</bindtype> to C<columns>, the C<\[ $sql, @bind ]>
2362 construct (see L</Literal SQL with placeholders and bind values (subqueries)>)
2363 will expect the bind values in this format.
2367 This is the character that a table or column name will be quoted
2368 with. By default this is an empty string, but you could set it to
2369 the character C<`>, to generate SQL like this:
2371 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
2373 Alternatively, you can supply an array ref of two items, the first being the left
2374 hand quote character, and the second the right hand quote character. For
2375 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
2376 that generates SQL like this:
2378 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
2380 Quoting is useful if you have tables or columns names that are reserved
2381 words in your database's SQL dialect.
2385 This is the character that will be used to escape L</quote_char>s appearing
2386 in an identifier before it has been quoted.
2388 The parameter default in case of a single L</quote_char> character is the quote
2391 When opening-closing-style quoting is used (L</quote_char> is an arrayref)
2392 this parameter defaults to the B<closing (right)> L</quote_char>. Occurrences
2393 of the B<opening (left)> L</quote_char> within the identifier are currently left
2394 untouched. The default for opening-closing-style quotes may change in future
2395 versions, thus you are B<strongly encouraged> to specify the escape character
2400 This is the character that separates a table and column name. It is
2401 necessary to specify this when the C<quote_char> option is selected,
2402 so that tables and column names can be individually quoted like this:
2404 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
2406 =item injection_guard
2408 A regular expression C<qr/.../> that is applied to any C<-function> and unquoted
2409 column name specified in a query structure. This is a safety mechanism to avoid
2410 injection attacks when mishandling user input e.g.:
2412 my %condition_as_column_value_pairs = get_values_from_user();
2413 $sqla->select( ... , \%condition_as_column_value_pairs );
2415 If the expression matches an exception is thrown. Note that literal SQL
2416 supplied via C<\'...'> or C<\['...']> is B<not> checked in any way.
2418 Defaults to checking for C<;> and the C<GO> keyword (TransactSQL)
2420 =item array_datatypes
2422 When this option is true, arrayrefs in INSERT or UPDATE are
2423 interpreted as array datatypes and are passed directly
2425 When this option is false, arrayrefs are interpreted
2426 as literal SQL, just like refs to arrayrefs
2427 (but this behavior is for backwards compatibility; when writing
2428 new queries, use the "reference to arrayref" syntax
2434 Takes a reference to a list of "special operators"
2435 to extend the syntax understood by L<SQL::Abstract>.
2436 See section L</"SPECIAL OPERATORS"> for details.
2440 Takes a reference to a list of "unary operators"
2441 to extend the syntax understood by L<SQL::Abstract>.
2442 See section L</"UNARY OPERATORS"> for details.
2448 =head2 insert($table, \@values || \%fieldvals, \%options)
2450 This is the simplest function. You simply give it a table name
2451 and either an arrayref of values or hashref of field/value pairs.
2452 It returns an SQL INSERT statement and a list of bind values.
2453 See the sections on L</"Inserting and Updating Arrays"> and
2454 L</"Inserting and Updating SQL"> for information on how to insert
2455 with those data types.
2457 The optional C<\%options> hash reference may contain additional
2458 options to generate the insert SQL. Currently supported options
2465 Takes either a scalar of raw SQL fields, or an array reference of
2466 field names, and adds on an SQL C<RETURNING> statement at the end.
2467 This allows you to return data generated by the insert statement
2468 (such as row IDs) without performing another C<SELECT> statement.
2469 Note, however, this is not part of the SQL standard and may not
2470 be supported by all database engines.
2474 =head2 update($table, \%fieldvals, \%where, \%options)
2476 This takes a table, hashref of field/value pairs, and an optional
2477 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
2479 See the sections on L</"Inserting and Updating Arrays"> and
2480 L</"Inserting and Updating SQL"> for information on how to insert
2481 with those data types.
2483 The optional C<\%options> hash reference may contain additional
2484 options to generate the update SQL. Currently supported options
2491 See the C<returning> option to
2492 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2496 =head2 select($source, $fields, $where, $order)
2498 This returns a SQL SELECT statement and associated list of bind values, as
2499 specified by the arguments:
2505 Specification of the 'FROM' part of the statement.
2506 The argument can be either a plain scalar (interpreted as a table
2507 name, will be quoted), or an arrayref (interpreted as a list
2508 of table names, joined by commas, quoted), or a scalarref
2509 (literal SQL, not quoted).
2513 Specification of the list of fields to retrieve from
2515 The argument can be either an arrayref (interpreted as a list
2516 of field names, will be joined by commas and quoted), or a
2517 plain scalar (literal SQL, not quoted).
2518 Please observe that this API is not as flexible as that of
2519 the first argument C<$source>, for backwards compatibility reasons.
2523 Optional argument to specify the WHERE part of the query.
2524 The argument is most often a hashref, but can also be
2525 an arrayref or plain scalar --
2526 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
2530 Optional argument to specify the ORDER BY part of the query.
2531 The argument can be a scalar, a hashref or an arrayref
2532 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
2538 =head2 delete($table, \%where, \%options)
2540 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
2541 It returns an SQL DELETE statement and list of bind values.
2543 The optional C<\%options> hash reference may contain additional
2544 options to generate the delete SQL. Currently supported options
2551 See the C<returning> option to
2552 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2556 =head2 where(\%where, $order)
2558 This is used to generate just the WHERE clause. For example,
2559 if you have an arbitrary data structure and know what the
2560 rest of your SQL is going to look like, but want an easy way
2561 to produce a WHERE clause, use this. It returns an SQL WHERE
2562 clause and list of bind values.
2565 =head2 values(\%data)
2567 This just returns the values from the hash C<%data>, in the same
2568 order that would be returned from any of the other above queries.
2569 Using this allows you to markedly speed up your queries if you
2570 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
2572 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
2574 Warning: This is an experimental method and subject to change.
2576 This returns arbitrarily generated SQL. It's a really basic shortcut.
2577 It will return two different things, depending on return context:
2579 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
2580 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
2582 These would return the following:
2584 # First calling form
2585 $stmt = "CREATE TABLE test (?, ?)";
2586 @bind = (field1, field2);
2588 # Second calling form
2589 $stmt_and_val = "CREATE TABLE test (field1, field2)";
2591 Depending on what you're trying to do, it's up to you to choose the correct
2592 format. In this example, the second form is what you would want.
2596 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
2600 ALTER SESSION SET nls_date_format = 'MM/YY'
2602 You get the idea. Strings get their case twiddled, but everything
2603 else remains verbatim.
2605 =head1 EXPORTABLE FUNCTIONS
2607 =head2 is_plain_value
2609 Determines if the supplied argument is a plain value as understood by this
2614 =item * The value is C<undef>
2616 =item * The value is a non-reference
2618 =item * The value is an object with stringification overloading
2620 =item * The value is of the form C<< { -value => $anything } >>
2624 On failure returns C<undef>, on success returns a B<scalar> reference
2625 to the original supplied argument.
2631 The stringification overloading detection is rather advanced: it takes
2632 into consideration not only the presence of a C<""> overload, but if that
2633 fails also checks for enabled
2634 L<autogenerated versions of C<"">|overload/Magic Autogeneration>, based
2635 on either C<0+> or C<bool>.
2637 Unfortunately testing in the field indicates that this
2638 detection B<< may tickle a latent bug in perl versions before 5.018 >>,
2639 but only when very large numbers of stringifying objects are involved.
2640 At the time of writing ( Sep 2014 ) there is no clear explanation of
2641 the direct cause, nor is there a manageably small test case that reliably
2642 reproduces the problem.
2644 If you encounter any of the following exceptions in B<random places within
2645 your application stack> - this module may be to blame:
2647 Operation "ne": no method found,
2648 left argument in overloaded package <something>,
2649 right argument in overloaded package <something>
2653 Stub found while resolving method "???" overloading """" in package <something>
2655 If you fall victim to the above - please attempt to reduce the problem
2656 to something that could be sent to the L<SQL::Abstract developers
2657 |DBIx::Class/GETTING HELP/SUPPORT>
2658 (either publicly or privately). As a workaround in the meantime you can
2659 set C<$ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}> to a true
2660 value, which will most likely eliminate your problem (at the expense of
2661 not being able to properly detect exotic forms of stringification).
2663 This notice and environment variable will be removed in a future version,
2664 as soon as the underlying problem is found and a reliable workaround is
2669 =head2 is_literal_value
2671 Determines if the supplied argument is a literal value as understood by this
2676 =item * C<\$sql_string>
2678 =item * C<\[ $sql_string, @bind_values ]>
2682 On failure returns C<undef>, on success returns an B<array> reference
2683 containing the unpacked version of the supplied literal SQL and bind values.
2685 =head2 is_undef_value
2687 Tests for undef, whether expanded or not.
2689 =head1 WHERE CLAUSES
2693 This module uses a variation on the idea from L<DBIx::Abstract>. It
2694 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
2695 module is that things in arrays are OR'ed, and things in hashes
2698 The easiest way to explain is to show lots of examples. After
2699 each C<%where> hash shown, it is assumed you used:
2701 my($stmt, @bind) = $sql->where(\%where);
2703 However, note that the C<%where> hash can be used directly in any
2704 of the other functions as well, as described above.
2706 =head2 Key-value pairs
2708 So, let's get started. To begin, a simple hash:
2712 status => 'completed'
2715 Is converted to SQL C<key = val> statements:
2717 $stmt = "WHERE user = ? AND status = ?";
2718 @bind = ('nwiger', 'completed');
2720 One common thing I end up doing is having a list of values that
2721 a field can be in. To do this, simply specify a list inside of
2726 status => ['assigned', 'in-progress', 'pending'];
2729 This simple code will create the following:
2731 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
2732 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
2734 A field associated to an empty arrayref will be considered a
2735 logical false and will generate 0=1.
2737 =head2 Tests for NULL values
2739 If the value part is C<undef> then this is converted to SQL <IS NULL>
2748 $stmt = "WHERE user = ? AND status IS NULL";
2751 To test if a column IS NOT NULL:
2755 status => { '!=', undef },
2758 =head2 Specific comparison operators
2760 If you want to specify a different type of operator for your comparison,
2761 you can use a hashref for a given column:
2765 status => { '!=', 'completed' }
2768 Which would generate:
2770 $stmt = "WHERE user = ? AND status != ?";
2771 @bind = ('nwiger', 'completed');
2773 To test against multiple values, just enclose the values in an arrayref:
2775 status => { '=', ['assigned', 'in-progress', 'pending'] };
2777 Which would give you:
2779 "WHERE status = ? OR status = ? OR status = ?"
2782 The hashref can also contain multiple pairs, in which case it is expanded
2783 into an C<AND> of its elements:
2787 status => { '!=', 'completed', -not_like => 'pending%' }
2790 # Or more dynamically, like from a form
2791 $where{user} = 'nwiger';
2792 $where{status}{'!='} = 'completed';
2793 $where{status}{'-not_like'} = 'pending%';
2795 # Both generate this
2796 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
2797 @bind = ('nwiger', 'completed', 'pending%');
2800 To get an OR instead, you can combine it with the arrayref idea:
2804 priority => [ { '=', 2 }, { '>', 5 } ]
2807 Which would generate:
2809 $stmt = "WHERE ( priority = ? OR priority > ? ) AND user = ?";
2810 @bind = ('2', '5', 'nwiger');
2812 If you want to include literal SQL (with or without bind values), just use a
2813 scalar reference or reference to an arrayref as the value:
2816 date_entered => { '>' => \["to_date(?, 'MM/DD/YYYY')", "11/26/2008"] },
2817 date_expires => { '<' => \"now()" }
2820 Which would generate:
2822 $stmt = "WHERE date_entered > to_date(?, 'MM/DD/YYYY') AND date_expires < now()";
2823 @bind = ('11/26/2008');
2826 =head2 Logic and nesting operators
2828 In the example above,
2829 there is a subtle trap if you want to say something like
2830 this (notice the C<AND>):
2832 WHERE priority != ? AND priority != ?
2834 Because, in Perl you I<can't> do this:
2836 priority => { '!=' => 2, '!=' => 1 }
2838 As the second C<!=> key will obliterate the first. The solution
2839 is to use the special C<-modifier> form inside an arrayref:
2841 priority => [ -and => {'!=', 2},
2845 Normally, these would be joined by C<OR>, but the modifier tells it
2846 to use C<AND> instead. (Hint: You can use this in conjunction with the
2847 C<logic> option to C<new()> in order to change the way your queries
2848 work by default.) B<Important:> Note that the C<-modifier> goes
2849 B<INSIDE> the arrayref, as an extra first element. This will
2850 B<NOT> do what you think it might:
2852 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
2854 Here is a quick list of equivalencies, since there is some overlap:
2857 status => {'!=', 'completed', 'not like', 'pending%' }
2858 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
2861 status => {'=', ['assigned', 'in-progress']}
2862 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
2863 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
2867 =head2 Special operators: IN, BETWEEN, etc.
2869 You can also use the hashref format to compare a list of fields using the
2870 C<IN> comparison operator, by specifying the list as an arrayref:
2873 status => 'completed',
2874 reportid => { -in => [567, 2335, 2] }
2877 Which would generate:
2879 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
2880 @bind = ('completed', '567', '2335', '2');
2882 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
2885 If the argument to C<-in> is an empty array, 'sqlfalse' is generated
2886 (by default: C<1=0>). Similarly, C<< -not_in => [] >> generates
2887 'sqltrue' (by default: C<1=1>).
2889 In addition to the array you can supply a chunk of literal sql or
2890 literal sql with bind:
2893 customer => { -in => \[
2894 'SELECT cust_id FROM cust WHERE balance > ?',
2897 status => { -in => \'SELECT status_codes FROM states' },
2903 customer IN ( SELECT cust_id FROM cust WHERE balance > ? )
2904 AND status IN ( SELECT status_codes FROM states )
2908 Finally, if the argument to C<-in> is not a reference, it will be
2909 treated as a single-element array.
2911 Another pair of operators is C<-between> and C<-not_between>,
2912 used with an arrayref of two values:
2916 completion_date => {
2917 -not_between => ['2002-10-01', '2003-02-06']
2923 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
2925 Just like with C<-in> all plausible combinations of literal SQL
2929 start0 => { -between => [ 1, 2 ] },
2930 start1 => { -between => \["? AND ?", 1, 2] },
2931 start2 => { -between => \"lower(x) AND upper(y)" },
2932 start3 => { -between => [
2934 \["upper(?)", 'stuff' ],
2941 ( start0 BETWEEN ? AND ? )
2942 AND ( start1 BETWEEN ? AND ? )
2943 AND ( start2 BETWEEN lower(x) AND upper(y) )
2944 AND ( start3 BETWEEN lower(x) AND upper(?) )
2946 @bind = (1, 2, 1, 2, 'stuff');
2949 These are the two builtin "special operators"; but the
2950 list can be expanded: see section L</"SPECIAL OPERATORS"> below.
2952 =head2 Unary operators: bool
2954 If you wish to test against boolean columns or functions within your
2955 database you can use the C<-bool> and C<-not_bool> operators. For
2956 example to test the column C<is_user> being true and the column
2957 C<is_enabled> being false you would use:-
2961 -not_bool => 'is_enabled',
2966 WHERE is_user AND NOT is_enabled
2968 If a more complex combination is required, testing more conditions,
2969 then you should use the and/or operators:-
2974 -not_bool => { two=> { -rlike => 'bar' } },
2975 -not_bool => { three => [ { '=', 2 }, { '>', 5 } ] },
2986 (NOT ( three = ? OR three > ? ))
2989 =head2 Nested conditions, -and/-or prefixes
2991 So far, we've seen how multiple conditions are joined with a top-level
2992 C<AND>. We can change this by putting the different conditions we want in
2993 hashes and then putting those hashes in an array. For example:
2998 status => { -like => ['pending%', 'dispatched'] },
3002 status => 'unassigned',
3006 This data structure would create the following:
3008 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
3009 OR ( user = ? AND status = ? ) )";
3010 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
3013 Clauses in hashrefs or arrayrefs can be prefixed with an C<-and> or C<-or>
3014 to change the logic inside:
3020 -and => [ workhrs => {'>', 20}, geo => 'ASIA' ],
3021 -or => { workhrs => {'<', 50}, geo => 'EURO' },
3028 $stmt = "WHERE ( user = ?
3029 AND ( ( workhrs > ? AND geo = ? )
3030 OR ( workhrs < ? OR geo = ? ) ) )";
3031 @bind = ('nwiger', '20', 'ASIA', '50', 'EURO');
3033 =head3 Algebraic inconsistency, for historical reasons
3035 C<Important note>: when connecting several conditions, the C<-and->|C<-or>
3036 operator goes C<outside> of the nested structure; whereas when connecting
3037 several constraints on one column, the C<-and> operator goes
3038 C<inside> the arrayref. Here is an example combining both features:
3041 -and => [a => 1, b => 2],
3042 -or => [c => 3, d => 4],
3043 e => [-and => {-like => 'foo%'}, {-like => '%bar'} ]
3048 WHERE ( ( ( a = ? AND b = ? )
3049 OR ( c = ? OR d = ? )
3050 OR ( e LIKE ? AND e LIKE ? ) ) )
3052 This difference in syntax is unfortunate but must be preserved for
3053 historical reasons. So be careful: the two examples below would
3054 seem algebraically equivalent, but they are not
3057 { -like => 'foo%' },
3058 { -like => '%bar' },
3060 # yields: WHERE ( ( col LIKE ? AND col LIKE ? ) )
3063 { col => { -like => 'foo%' } },
3064 { col => { -like => '%bar' } },
3066 # yields: WHERE ( ( col LIKE ? OR col LIKE ? ) )
3069 =head2 Literal SQL and value type operators
3071 The basic premise of SQL::Abstract is that in WHERE specifications the "left
3072 side" is a column name and the "right side" is a value (normally rendered as
3073 a placeholder). This holds true for both hashrefs and arrayref pairs as you
3074 see in the L</WHERE CLAUSES> examples above. Sometimes it is necessary to
3075 alter this behavior. There are several ways of doing so.
3079 This is a virtual operator that signals the string to its right side is an
3080 identifier (a column name) and not a value. For example to compare two
3081 columns you would write:
3084 priority => { '<', 2 },
3085 requestor => { -ident => 'submitter' },
3090 $stmt = "WHERE priority < ? AND requestor = submitter";
3093 If you are maintaining legacy code you may see a different construct as
3094 described in L</Deprecated usage of Literal SQL>, please use C<-ident> in new
3099 This is a virtual operator that signals that the construct to its right side
3100 is a value to be passed to DBI. This is for example necessary when you want
3101 to write a where clause against an array (for RDBMS that support such
3102 datatypes). For example:
3105 array => { -value => [1, 2, 3] }
3110 $stmt = 'WHERE array = ?';
3111 @bind = ([1, 2, 3]);
3113 Note that if you were to simply say:
3119 the result would probably not be what you wanted:
3121 $stmt = 'WHERE array = ? OR array = ? OR array = ?';
3126 Finally, sometimes only literal SQL will do. To include a random snippet
3127 of SQL verbatim, you specify it as a scalar reference. Consider this only
3128 as a last resort. Usually there is a better way. For example:
3131 priority => { '<', 2 },
3132 requestor => { -in => \'(SELECT name FROM hitmen)' },
3137 $stmt = "WHERE priority < ? AND requestor IN (SELECT name FROM hitmen)"
3140 Note that in this example, you only get one bind parameter back, since
3141 the verbatim SQL is passed as part of the statement.
3145 Never use untrusted input as a literal SQL argument - this is a massive
3146 security risk (there is no way to check literal snippets for SQL
3147 injections and other nastyness). If you need to deal with untrusted input
3148 use literal SQL with placeholders as described next.
3150 =head3 Literal SQL with placeholders and bind values (subqueries)
3152 If the literal SQL to be inserted has placeholders and bind values,
3153 use a reference to an arrayref (yes this is a double reference --
3154 not so common, but perfectly legal Perl). For example, to find a date
3155 in Postgres you can use something like this:
3158 date_column => \[ "= date '2008-09-30' - ?::integer", 10 ]
3163 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
3166 Note that you must pass the bind values in the same format as they are returned
3167 by L<where|/where(\%where, $order)>. This means that if you set L</bindtype>
3168 to C<columns>, you must provide the bind values in the
3169 C<< [ column_meta => value ] >> format, where C<column_meta> is an opaque
3170 scalar value; most commonly the column name, but you can use any scalar value
3171 (including references and blessed references), L<SQL::Abstract> will simply
3172 pass it through intact. So if C<bindtype> is set to C<columns> the above
3173 example will look like:
3176 date_column => \[ "= date '2008-09-30' - ?::integer", [ {} => 10 ] ]
3179 Literal SQL is especially useful for nesting parenthesized clauses in the
3180 main SQL query. Here is a first example:
3182 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
3186 bar => \["IN ($sub_stmt)" => @sub_bind],
3191 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
3192 WHERE c2 < ? AND c3 LIKE ?))";
3193 @bind = (1234, 100, "foo%");
3195 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
3196 are expressed in the same way. Of course the C<$sub_stmt> and
3197 its associated bind values can be generated through a former call
3200 my ($sub_stmt, @sub_bind)
3201 = $sql->select("t1", "c1", {c2 => {"<" => 100},
3202 c3 => {-like => "foo%"}});
3205 bar => \["> ALL ($sub_stmt)" => @sub_bind],
3208 In the examples above, the subquery was used as an operator on a column;
3209 but the same principle also applies for a clause within the main C<%where>
3210 hash, like an EXISTS subquery:
3212 my ($sub_stmt, @sub_bind)
3213 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
3214 my %where = ( -and => [
3216 \["EXISTS ($sub_stmt)" => @sub_bind],
3221 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
3222 WHERE c1 = ? AND c2 > t0.c0))";
3226 Observe that the condition on C<c2> in the subquery refers to
3227 column C<t0.c0> of the main query: this is I<not> a bind
3228 value, so we have to express it through a scalar ref.
3229 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
3230 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
3231 what we wanted here.
3233 Finally, here is an example where a subquery is used
3234 for expressing unary negation:
3236 my ($sub_stmt, @sub_bind)
3237 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
3238 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
3240 lname => {like => '%son%'},
3241 \["NOT ($sub_stmt)" => @sub_bind],
3246 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
3247 @bind = ('%son%', 10, 20)
3249 =head3 Deprecated usage of Literal SQL
3251 Below are some examples of archaic use of literal SQL. It is shown only as
3252 reference for those who deal with legacy code. Each example has a much
3253 better, cleaner and safer alternative that users should opt for in new code.
3259 my %where = ( requestor => \'IS NOT NULL' )
3261 $stmt = "WHERE requestor IS NOT NULL"
3263 This used to be the way of generating NULL comparisons, before the handling
3264 of C<undef> got formalized. For new code please use the superior syntax as
3265 described in L</Tests for NULL values>.
3269 my %where = ( requestor => \'= submitter' )
3271 $stmt = "WHERE requestor = submitter"
3273 This used to be the only way to compare columns. Use the superior L</-ident>
3274 method for all new code. For example an identifier declared in such a way
3275 will be properly quoted if L</quote_char> is properly set, while the legacy
3276 form will remain as supplied.
3280 my %where = ( is_ready => \"", completed => { '>', '2012-12-21' } )
3282 $stmt = "WHERE completed > ? AND is_ready"
3283 @bind = ('2012-12-21')
3285 Using an empty string literal used to be the only way to express a boolean.
3286 For all new code please use the much more readable
3287 L<-bool|/Unary operators: bool> operator.
3293 These pages could go on for a while, since the nesting of the data
3294 structures this module can handle are pretty much unlimited (the
3295 module implements the C<WHERE> expansion as a recursive function
3296 internally). Your best bet is to "play around" with the module a
3297 little to see how the data structures behave, and choose the best
3298 format for your data based on that.
3300 And of course, all the values above will probably be replaced with
3301 variables gotten from forms or the command line. After all, if you
3302 knew everything ahead of time, you wouldn't have to worry about
3303 dynamically-generating SQL and could just hardwire it into your
3306 =head1 ORDER BY CLAUSES
3308 Some functions take an order by clause. This can either be a scalar (just a
3309 column name), a hashref of C<< { -desc => 'col' } >> or C<< { -asc => 'col' }
3310 >>, a scalarref, an arrayref-ref, or an arrayref of any of the previous
3313 Given | Will Generate
3314 ---------------------------------------------------------------
3316 'colA' | ORDER BY colA
3318 [qw/colA colB/] | ORDER BY colA, colB
3320 {-asc => 'colA'} | ORDER BY colA ASC
3322 {-desc => 'colB'} | ORDER BY colB DESC
3324 ['colA', {-asc => 'colB'}] | ORDER BY colA, colB ASC
3326 { -asc => [qw/colA colB/] } | ORDER BY colA ASC, colB ASC
3328 \'colA DESC' | ORDER BY colA DESC
3330 \[ 'FUNC(colA, ?)', $x ] | ORDER BY FUNC(colA, ?)
3331 | /* ...with $x bound to ? */
3334 { -asc => 'colA' }, | colA ASC,
3335 { -desc => [qw/colB/] }, | colB DESC,
3336 { -asc => [qw/colC colD/] },| colC ASC, colD ASC,
3337 \'colE DESC', | colE DESC,
3338 \[ 'FUNC(colF, ?)', $x ], | FUNC(colF, ?)
3339 ] | /* ...with $x bound to ? */
3340 ===============================================================
3344 =head1 OLD EXTENSION SYSTEM
3346 =head2 SPECIAL OPERATORS
3348 my $sqlmaker = SQL::Abstract->new(special_ops => [
3352 my ($self, $field, $op, $arg) = @_;
3358 handler => 'method_name',
3362 A "special operator" is a SQL syntactic clause that can be
3363 applied to a field, instead of a usual binary operator.
3366 WHERE field IN (?, ?, ?)
3367 WHERE field BETWEEN ? AND ?
3368 WHERE MATCH(field) AGAINST (?, ?)
3370 Special operators IN and BETWEEN are fairly standard and therefore
3371 are builtin within C<SQL::Abstract> (as the overridable methods
3372 C<_where_field_IN> and C<_where_field_BETWEEN>). For other operators,
3373 like the MATCH .. AGAINST example above which is specific to MySQL,
3374 you can write your own operator handlers - supply a C<special_ops>
3375 argument to the C<new> method. That argument takes an arrayref of
3376 operator definitions; each operator definition is a hashref with two
3383 the regular expression to match the operator
3387 Either a coderef or a plain scalar method name. In both cases
3388 the expected return is C<< ($sql, @bind) >>.
3390 When supplied with a method name, it is simply called on the
3391 L<SQL::Abstract> object as:
3393 $self->$method_name($field, $op, $arg)
3397 $field is the LHS of the operator
3398 $op is the part that matched the handler regex
3401 When supplied with a coderef, it is called as:
3403 $coderef->($self, $field, $op, $arg)
3408 For example, here is an implementation
3409 of the MATCH .. AGAINST syntax for MySQL
3411 my $sqlmaker = SQL::Abstract->new(special_ops => [
3413 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
3414 {regex => qr/^match$/i,
3416 my ($self, $field, $op, $arg) = @_;
3417 $arg = [$arg] if not ref $arg;
3418 my $label = $self->_quote($field);
3419 my ($placeholder) = $self->_convert('?');
3420 my $placeholders = join ", ", (($placeholder) x @$arg);
3421 my $sql = $self->_sqlcase('match') . " ($label) "
3422 . $self->_sqlcase('against') . " ($placeholders) ";
3423 my @bind = $self->_bindtype($field, @$arg);
3424 return ($sql, @bind);
3431 =head2 UNARY OPERATORS
3433 my $sqlmaker = SQL::Abstract->new(unary_ops => [
3437 my ($self, $op, $arg) = @_;
3443 handler => 'method_name',
3447 A "unary operator" is a SQL syntactic clause that can be
3448 applied to a field - the operator goes before the field
3450 You can write your own operator handlers - supply a C<unary_ops>
3451 argument to the C<new> method. That argument takes an arrayref of
3452 operator definitions; each operator definition is a hashref with two
3459 the regular expression to match the operator
3463 Either a coderef or a plain scalar method name. In both cases
3464 the expected return is C<< $sql >>.
3466 When supplied with a method name, it is simply called on the
3467 L<SQL::Abstract> object as:
3469 $self->$method_name($op, $arg)
3473 $op is the part that matched the handler regex
3474 $arg is the RHS or argument of the operator
3476 When supplied with a coderef, it is called as:
3478 $coderef->($self, $op, $arg)
3483 =head1 NEW METHODS (EXPERIMENTAL)
3485 See L<SQL::Abstract::Reference> for the C<expr> versus C<aqt> concept and
3486 an explanation of what the below extensions are extending.
3490 $sqla->plugin('+Foo');
3492 Enables plugin SQL::Abstract::Plugin::Foo.
3496 my ($sql, @bind) = $sqla->render_expr($expr);
3498 =head2 render_statement
3500 Use this if you may be rendering a top level statement so e.g. a SELECT
3501 query doesn't get wrapped in parens
3503 my ($sql, @bind) = $sqla->render_statement($expr);
3507 Expression expansion with optional default for scalars.
3509 my $aqt = $self->expand_expr($expr);
3510 my $aqt = $self->expand_expr($expr, -ident);
3514 Top level means avoid parens on statement AQT.
3516 my $res = $self->render_aqt($aqt, $top_level);
3517 my ($sql, @bind) = @$res;
3519 =head2 join_query_parts
3521 Similar to join() but will render hashrefs as nodes for both join and parts,
3522 and treats arrayref as a nested C<[ $join, @parts ]> structure.
3524 my $part = $self->join_query_parts($join, @parts);
3526 =head1 NEW EXTENSION SYSTEM
3530 my $sqla2 = $sqla->clone;
3532 Performs a semi-shallow copy such that extension methods won't leak state
3533 but excessive depth is avoided.
3543 =head2 clause_expander
3545 =head2 clause_expanders
3547 $sqla->expander('name' => sub { ... });
3548 $sqla->expanders('name1' => sub { ... }, 'name2' => sub { ... });
3550 =head2 expander_list
3552 =head2 op_expander_list
3554 =head2 clause_expander_list
3556 my @names = $sqla->expander_list;
3558 =head2 wrap_expander
3560 =head2 wrap_expanders
3562 =head2 wrap_op_expander
3564 =head2 wrap_op_expanders
3566 =head2 wrap_clause_expander
3568 =head2 wrap_clause_expanders
3570 $sqla->wrap_expander('name' => sub { my ($orig) = @_; sub { ... } });
3571 $sqla->wrap_expanders(
3572 'name1' => sub { my ($orig1) = @_; sub { ... } },
3573 'name2' => sub { my ($orig2) = @_; sub { ... } },
3584 =head2 clause_renderer
3586 =head2 clause_renderers
3588 $sqla->renderer('name' => sub { ... });
3589 $sqla->renderers('name1' => sub { ... }, 'name2' => sub { ... });
3591 =head2 renderer_list
3593 =head2 op_renderer_list
3595 =head2 clause_renderer_list
3597 my @names = $sqla->renderer_list;
3599 =head2 wrap_renderer
3601 =head2 wrap_renderers
3603 =head2 wrap_op_renderer
3605 =head2 wrap_op_renderers
3607 =head2 wrap_clause_renderer
3609 =head2 wrap_clause_renderers
3611 $sqla->wrap_renderer('name' => sub { my ($orig) = @_; sub { ... } });
3612 $sqla->wrap_renderers(
3613 'name1' => sub { my ($orig1) = @_; sub { ... } },
3614 'name2' => sub { my ($orig2) = @_; sub { ... } },
3619 my @clauses = $sqla->clauses_of('select');
3620 $sqla->clauses_of(select => \@new_clauses);
3621 $sqla->clauses_of(select => sub {
3622 my (undef, @old_clauses) = @_;
3624 return @new_clauses;
3627 =head2 statement_list
3629 my @list = $sqla->statement_list;
3631 =head2 make_unop_expander
3633 my $exp = $sqla->make_unop_expander(sub { ... });
3635 If the op is found as a binop, assumes it wants a default comparison, so
3636 the inner expander sub can reliably operate as
3638 sub { my ($self, $name, $body) = @_; ... }
3640 =head2 make_binop_expander
3642 my $exp = $sqla->make_binop_expander(sub { ... });
3644 If the op is found as a unop, assumes the value will be an arrayref with the
3645 LHS as the first entry, and converts that to an ident node if it's a simple
3646 scalar. So the inner expander sub looks like
3649 my ($self, $name, $body, $k) = @_;
3650 { -blah => [ map $self->expand_expr($_), $k, $body ] }
3653 =head2 unop_expander
3655 =head2 unop_expanders
3657 =head2 binop_expander
3659 =head2 binop_expanders
3661 The above methods operate exactly like the op_ versions but wrap the coderef
3662 using the appropriate make_ method first.
3666 Thanks to some benchmarking by Mark Stosberg, it turns out that
3667 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
3668 I must admit this wasn't an intentional design issue, but it's a
3669 byproduct of the fact that you get to control your C<DBI> handles
3672 To maximize performance, use a code snippet like the following:
3674 # prepare a statement handle using the first row
3675 # and then reuse it for the rest of the rows
3677 for my $href (@array_of_hashrefs) {
3678 $stmt ||= $sql->insert('table', $href);
3679 $sth ||= $dbh->prepare($stmt);
3680 $sth->execute($sql->values($href));
3683 The reason this works is because the keys in your C<$href> are sorted
3684 internally by B<SQL::Abstract>. Thus, as long as your data retains
3685 the same structure, you only have to generate the SQL the first time
3686 around. On subsequent queries, simply use the C<values> function provided
3687 by this module to return your values in the correct order.
3689 However this depends on the values having the same type - if, for
3690 example, the values of a where clause may either have values
3691 (resulting in sql of the form C<column = ?> with a single bind
3692 value), or alternatively the values might be C<undef> (resulting in
3693 sql of the form C<column IS NULL> with no bind value) then the
3694 caching technique suggested will not work.
3698 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
3699 really like this part (I do, at least). Building up a complex query
3700 can be as simple as the following:
3707 use CGI::FormBuilder;
3710 my $form = CGI::FormBuilder->new(...);
3711 my $sql = SQL::Abstract->new;
3713 if ($form->submitted) {
3714 my $field = $form->field;
3715 my $id = delete $field->{id};
3716 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
3719 Of course, you would still have to connect using C<DBI> to run the
3720 query, but the point is that if you make your form look like your
3721 table, the actual query script can be extremely simplistic.
3723 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
3724 a fast interface to returning and formatting data. I frequently
3725 use these three modules together to write complex database query
3726 apps in under 50 lines.
3728 =head1 HOW TO CONTRIBUTE
3730 Contributions are always welcome, in all usable forms (we especially
3731 welcome documentation improvements). The delivery methods include git-
3732 or unified-diff formatted patches, GitHub pull requests, or plain bug
3733 reports either via RT or the Mailing list. Contributors are generally
3734 granted full access to the official repository after their first several
3735 patches pass successful review.
3737 This project is maintained in a git repository. The code and related tools are
3738 accessible at the following locations:
3742 =item * Official repo: L<git://git.shadowcat.co.uk/dbsrgits/SQL-Abstract.git>
3744 =item * Official gitweb: L<http://git.shadowcat.co.uk/gitweb/gitweb.cgi?p=dbsrgits/SQL-Abstract.git>
3746 =item * GitHub mirror: L<https://github.com/dbsrgits/sql-abstract>
3748 =item * Authorized committers: L<ssh://dbsrgits@git.shadowcat.co.uk/SQL-Abstract.git>
3754 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
3755 Great care has been taken to preserve the I<published> behavior
3756 documented in previous versions in the 1.* family; however,
3757 some features that were previously undocumented, or behaved
3758 differently from the documentation, had to be changed in order
3759 to clarify the semantics. Hence, client code that was relying
3760 on some dark areas of C<SQL::Abstract> v1.*
3761 B<might behave differently> in v1.50.
3763 The main changes are:
3769 support for literal SQL through the C<< \ [ $sql, @bind ] >> syntax.
3773 support for the { operator => \"..." } construct (to embed literal SQL)
3777 support for the { operator => \["...", @bind] } construct (to embed literal SQL with bind values)
3781 optional support for L<array datatypes|/"Inserting and Updating Arrays">
3785 defensive programming: check arguments
3789 fixed bug with global logic, which was previously implemented
3790 through global variables yielding side-effects. Prior versions would
3791 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
3792 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
3793 Now this is interpreted
3794 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
3799 fixed semantics of _bindtype on array args
3803 dropped the C<_anoncopy> of the %where tree. No longer necessary,
3804 we just avoid shifting arrays within that tree.
3808 dropped the C<_modlogic> function
3812 =head1 ACKNOWLEDGEMENTS
3814 There are a number of individuals that have really helped out with
3815 this module. Unfortunately, most of them submitted bugs via CPAN
3816 so I have no idea who they are! But the people I do know are:
3818 Ash Berlin (order_by hash term support)
3819 Matt Trout (DBIx::Class support)
3820 Mark Stosberg (benchmarking)
3821 Chas Owens (initial "IN" operator support)
3822 Philip Collins (per-field SQL functions)
3823 Eric Kolve (hashref "AND" support)
3824 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
3825 Dan Kubb (support for "quote_char" and "name_sep")
3826 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
3827 Laurent Dami (internal refactoring, extensible list of special operators, literal SQL)
3828 Norbert Buchmuller (support for literal SQL in hashpair, misc. fixes & tests)
3829 Peter Rabbitson (rewrite of SQLA::Test, misc. fixes & tests)
3830 Oliver Charles (support for "RETURNING" after "INSERT")
3836 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
3840 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
3842 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
3844 For support, your best bet is to try the C<DBIx::Class> users mailing list.
3845 While not an official support venue, C<DBIx::Class> makes heavy use of
3846 C<SQL::Abstract>, and as such list members there are very familiar with
3847 how to create queries.
3851 This module is free software; you may copy this under the same
3852 terms as perl itself (either the GNU General Public License or
3853 the Artistic License)