1 package SQL::Abstract; # see doc at end of file
10 our @EXPORT_OK = qw(is_plain_value is_literal_value is_undef_value);
20 *SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION = $ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}
26 #======================================================================
28 #======================================================================
30 our $VERSION = '2.000001';
32 # This would confuse some packagers
33 $VERSION = eval $VERSION if $VERSION =~ /_/; # numify for warning-free dev releases
37 # special operators (-in, -between). May be extended/overridden by user.
38 # See section WHERE: BUILTIN SPECIAL OPERATORS below for implementation
39 my @BUILTIN_SPECIAL_OPS = (
40 {regex => qr/^ (?: not \s )? between $/ix, handler => sub { die "NOPE" }},
41 {regex => qr/^ is (?: \s+ not )? $/ix, handler => sub { die "NOPE" }},
42 {regex => qr/^ (?: not \s )? in $/ix, handler => sub { die "NOPE" }},
43 {regex => qr/^ ident $/ix, handler => sub { die "NOPE" }},
44 {regex => qr/^ value $/ix, handler => sub { die "NOPE" }},
47 #======================================================================
48 # DEBUGGING AND ERROR REPORTING
49 #======================================================================
52 return unless $_[0]->{debug}; shift; # a little faster
53 my $func = (caller(1))[3];
54 warn "[$func] ", @_, "\n";
58 my($func) = (caller(1))[3];
59 Carp::carp "[$func] Warning: ", @_;
63 my($func) = (caller(1))[3];
64 Carp::croak "[$func] Fatal: ", @_;
67 sub is_literal_value ($) {
68 ref $_[0] eq 'SCALAR' ? [ ${$_[0]} ]
69 : ( ref $_[0] eq 'REF' and ref ${$_[0]} eq 'ARRAY' ) ? [ @${ $_[0] } ]
73 sub is_undef_value ($) {
77 and exists $_[0]->{-value}
78 and not defined $_[0]->{-value}
82 # FIXME XSify - this can be done so much more efficiently
83 sub is_plain_value ($) {
85 ! length ref $_[0] ? \($_[0])
87 ref $_[0] eq 'HASH' and keys %{$_[0]} == 1
89 exists $_[0]->{-value}
90 ) ? \($_[0]->{-value})
92 # reuse @_ for even moar speedz
93 defined ( $_[1] = Scalar::Util::blessed $_[0] )
95 # deliberately not using Devel::OverloadInfo - the checks we are
96 # intersted in are much more limited than the fullblown thing, and
97 # this is a very hot piece of code
99 # simply using ->can('(""') can leave behind stub methods that
100 # break actually using the overload later (see L<perldiag/Stub
101 # found while resolving method "%s" overloading "%s" in package
102 # "%s"> and the source of overload::mycan())
104 # either has stringification which DBI SHOULD prefer out of the box
105 grep { *{ (qq[${_}::(""]) }{CODE} } @{ $_[2] = mro::get_linear_isa( $_[1] ) }
107 # has nummification or boolification, AND fallback is *not* disabled
109 SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION
112 grep { *{"${_}::(0+"}{CODE} } @{$_[2]}
114 grep { *{"${_}::(bool"}{CODE} } @{$_[2]}
118 # no fallback specified at all
119 ! ( ($_[3]) = grep { *{"${_}::()"}{CODE} } @{$_[2]} )
121 # fallback explicitly undef
122 ! defined ${"$_[3]::()"}
135 #======================================================================
137 #======================================================================
141 bool => '_expand_bool',
142 row => '_expand_row',
144 func => '_expand_func',
145 values => '_expand_values',
146 list => '_expand_list',
149 (map +($_ => __PACKAGE__->make_binop_expander('_expand_between')),
150 qw(between not_between)),
151 (map +($_ => __PACKAGE__->make_binop_expander('_expand_in')),
153 (map +($_ => '_expand_op_andor'), ('and', 'or')),
154 (map +($_ => '_expand_op_is'), ('is', 'is_not')),
155 (map +($_ => __PACKAGE__->make_unop_expander("_expand_${_}")),
156 qw(ident value nest)),
157 bind => __PACKAGE__->make_unop_expander(sub { +{ -bind => $_[2] } }),
160 (map +($_, "_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 foreach my $type (qw(in between)) {
297 my $meth = "_where_field_".uc($type);
298 if (__PACKAGE__->can($meth) ne $class->can($meth)) {
300 my ($self, $op, $v, $k) = @_;
301 $op = join ' ', split '_', $op;
302 return +{ -literal => [
303 $self->$meth($k, $op, $v)
306 $opt{expand_op}{$_} = $exp for $type, "not_${type}";
309 if ($class->isa('DBIx::Class::SQLMaker')) {
310 $opt{warn_once_on_nest} = 1;
311 $opt{disable_old_special_ops} = 1;
312 $opt{render_clause}{'select.where'} = sub {
313 my ($sql, @bind) = $_[0]->where($_[2]);
314 s/\A\s+//, s/\s+\Z// for $sql;
315 return [ $sql, @bind ];
317 $opt{expand_op}{ident} = $class->make_unop_expander(sub {
318 my ($self, undef, $body) = @_;
319 $body = $body->from if Scalar::Util::blessed($body);
320 $self->_expand_ident(ident => $body);
323 if ($class->isa('SQL::Abstract::More')) {
324 my $orig = $opt{expand_op}{or};
325 $opt{expand_op}{or} = sub {
326 my ($self, $logop, $v, $k) = @_;
327 if ($k and ref($v) eq 'ARRAY') {
328 my ($type, $val) = @$v;
331 ref($type) eq 'HASH' and ref($val) eq 'HASH'
332 and keys %$type == 1 and keys %$val == 1
333 and (keys %$type)[0] eq (keys %$val)[0]
336 ($type) = values %$type;
337 ($val) = values %$val;
339 if ($self->is_bind_value_with_type(my $v = [ $type, $val ])) {
340 return $self->expand_expr(
341 { $k, map +($op ? { $op => $_ } : $_), { -bind => $v } }
345 return $self->$orig($logop, $v, $k);
347 $opt{render}{bind} = sub {
348 return [ '?', map +(ref($_->[0]) ? $_ : $_->[1]), $_[2] ]
353 if ($opt{lazy_join_sql_parts}) {
354 require SQL::Abstract::Parts;
355 $opt{join_sql_parts} ||= sub { SQL::Abstract::Parts->new(@_) };
358 $opt{join_sql_parts} ||= sub { join $_[0], @_[1..$#_] };
360 return bless \%opt, $class;
364 my ($self, $name, $key, $value) = @_;
365 return $self->{$name}{$key} unless @_ > 3;
366 $self->{$name}{$key} = $value;
370 sub make_unop_expander {
371 my (undef, $exp) = @_;
373 my ($self, $name, $body, $k) = @_;
374 return $self->_expand_hashpair_cmp($k, { "-${name}" => $body })
376 return $self->$exp($name, $body);
380 sub make_binop_expander {
381 my (undef, $exp) = @_;
383 my ($self, $name, $body, $k) = @_;
384 $k = shift @{$body = [ @$body ]} unless defined $k;
385 $k = ref($k) ? $k : { -ident => $k };
386 return $self->$exp($name, $body, $k);
391 my ($self, $plugin, @args) = @_;
392 unless (ref $plugin) {
393 $plugin =~ s/\A\+/${\__PACKAGE__}::Plugin::/;
394 require(join('/', split '::', $plugin).'.pm');
396 $plugin->apply_to($self, @args);
401 foreach my $type (qw(
402 expand op_expand render op_render clause_expand clause_render
404 my $name = join '_', reverse split '_', $type;
405 my $singular = "${type}er";
407 eval qq{sub ${singular} {
409 return \$self->_ext_rw('${name}', \@_) if \@_ == 1;
410 return \$self->${singular}s(\@_)
411 }; 1 } or die "Method builder failed for ${singular}: $@";
412 eval qq{sub wrap_${singular} {
413 shift->wrap_${singular}s(\@_)
414 }; 1 } or die "Method builder failed for wrap_${singular}: $@";
416 eval qq{sub ${singular}s {
417 my (\$self, \@args) = \@_;
418 while (my (\$this_key, \$this_value) = splice(\@args, 0, 2)) {
419 \$self->_ext_rw('${name}', \$this_key, \$this_value);
422 }; 1 } or die "Method builder failed for ${singular}s: $@";
423 eval qq{sub wrap_${singular}s {
424 my (\$self, \@args) = \@_;
425 while (my (\$this_key, \$this_builder) = splice(\@args, 0, 2)) {
426 my \$orig = \$self->_ext_rw('${name}', \$this_key);
428 '${name}', \$this_key,
429 \$this_builder->(\$orig, '${name}', \$this_key),
433 }; 1 } or die "Method builder failed for wrap_${singular}s: $@";
434 eval qq{sub ${singular}_list { sort keys %{\$_[0]->{\$name}} }; 1; }
435 or die "Method builder failed for ${singular}_list: $@";
437 foreach my $singular (qw(unop_expander binop_expander)) {
438 eval qq{sub ${singular} { shift->${singular}s(\@_) }; 1 }
439 or die "Method builder failed for ${singular}: $@";
440 eval qq{sub ${singular}s {
441 my (\$self, \@args) = \@_;
442 while (my (\$this_key, \$this_value) = splice(\@args, 0, 2)) {
444 expand_op => \$this_key,
445 \$self->make_${singular}(\$this_value),
449 }; 1 } or die "Method builder failed for ${singular}s: $@";
453 #sub register_op { $_[0]->{is_op}{$_[1]} = 1; $_[0] }
455 sub statement_list { sort keys %{$_[0]->{clauses_of}} }
458 my ($self, $of, @clauses) = @_;
460 return @{$self->{clauses_of}{$of}||[]};
462 if (ref($clauses[0]) eq 'CODE') {
463 @clauses = $self->${\($clauses[0])}(@{$self->{clauses_of}{$of}||[]});
465 $self->{clauses_of}{$of} = \@clauses;
474 ref($self->{$_}) eq 'HASH'
483 sub sqltrue { +{ -literal => [ $_[0]->{sqltrue} ] } }
484 sub sqlfalse { +{ -literal => [ $_[0]->{sqlfalse} ] } }
486 sub _assert_pass_injection_guard {
487 if ($_[1] =~ $_[0]->{injection_guard}) {
488 my $class = ref $_[0];
489 puke "Possible SQL injection attempt '$_[1]'. If this is indeed a part of the "
490 . "desired SQL use literal SQL ( \'...' or \[ '...' ] ) or supply your own "
491 . "{injection_guard} attribute to ${class}->new()"
496 #======================================================================
498 #======================================================================
501 my ($self, $table, $data, $options) = @_;
504 if (ref($table) eq 'HASH') {
507 my %clauses = (target => $table, values => $data, %{$options||{}});
511 my @rendered = $self->render_statement({ -insert => $stmt });
512 return wantarray ? @rendered : $rendered[0];
515 sub _expand_insert_clause_target {
516 +(target => $_[0]->expand_expr($_[2], -ident));
519 sub _expand_insert_clause_fields {
521 $_[0]->expand_expr({ -list => $_[2] }, -ident)
522 ] } if ref($_[2]) eq 'ARRAY';
523 return $_[2]; # should maybe still expand somewhat?
526 sub _expand_insert_clause_from {
527 my ($self, undef, $data) = @_;
528 if (ref($data) eq 'HASH' and (keys(%$data))[0] =~ /^-/) {
529 return $self->expand_expr($data);
531 my ($f_aqt, $v_aqt) = $self->_expand_insert_values($data);
533 from => { -values => [ $v_aqt ] },
534 ($f_aqt ? (fields => $f_aqt) : ()),
538 sub _expand_insert_clause_returning {
539 +(returning => $_[0]->expand_expr({ -list => $_[2] }, -ident));
542 sub _expand_insert_values {
543 my ($self, $data) = @_;
544 if (is_literal_value($data)) {
545 (undef, $self->expand_expr($data));
547 my ($fields, $values) = (
548 ref($data) eq 'HASH' ?
549 ([ sort keys %$data ], [ @{$data}{sort keys %$data} ])
553 # no names (arrayref) means can't generate bindtype
554 !($fields) && $self->{bindtype} eq 'columns'
555 && belch "can't do 'columns' bindtype when called with arrayref";
559 ? $self->expand_expr({ -row => $fields }, -ident)
564 local our $Cur_Col_Meta = $fields->[$_];
565 $self->_expand_insert_value($values->[$_])
572 sub _render_insert_clause_fields {
573 return $_[0]->render_aqt($_[2]);
576 sub _render_insert_clause_target {
577 my ($self, undef, $from) = @_;
578 $self->join_query_parts(' ', { -keyword => 'insert into' }, $from);
581 sub _render_insert_clause_from {
582 return $_[0]->render_aqt($_[2], 1);
585 # So that subclasses can override INSERT ... RETURNING separately from
586 # UPDATE and DELETE (e.g. DBIx::Class::SQLMaker::Oracle does this)
587 sub _insert_returning { shift->_returning(@_) }
589 sub _redispatch_returning {
590 my ($self, $type, undef, $returning) = @_;
591 [ $self->${\"_${type}_returning"}({ returning => $returning }) ];
595 my ($self, $options) = @_;
597 my $f = $options->{returning};
599 my ($sql, @bind) = @{ $self->render_aqt(
600 $self->expand_expr({ -list => $f }, -ident)
602 my $rsql = $self->_sqlcase(' returning ').$sql;
603 return wantarray ? ($rsql, @bind) : $rsql;
606 sub _expand_insert_value {
609 my $k = our $Cur_Col_Meta;
611 if (ref($v) eq 'ARRAY') {
612 if ($self->{array_datatypes}) {
613 return +{ -bind => [ $k, $v ] };
615 my ($sql, @bind) = @$v;
616 $self->_assert_bindval_matches_bindtype(@bind);
617 return +{ -literal => $v };
619 if (ref($v) eq 'HASH') {
620 if (grep !/^-/, keys %$v) {
621 belch "HASH ref as bind value in insert is not supported";
622 return +{ -bind => [ $k, $v ] };
626 return +{ -bind => [ $k, undef ] };
628 return $self->expand_expr($v);
633 #======================================================================
635 #======================================================================
638 my ($self, $table, $set, $where, $options) = @_;
641 if (ref($table) eq 'HASH') {
645 @clauses{qw(target set where)} = ($table, $set, $where);
646 puke "Unsupported data type specified to \$sql->update"
647 unless ref($clauses{set}) eq 'HASH';
648 @clauses{keys %$options} = values %$options;
652 my @rendered = $self->render_statement({ -update => $stmt });
653 return wantarray ? @rendered : $rendered[0];
656 sub _render_update_clause_target {
657 my ($self, undef, $target) = @_;
658 $self->join_query_parts(' ', { -keyword => 'update' }, $target);
661 sub _update_set_values {
662 my ($self, $data) = @_;
664 return @{ $self->render_aqt(
665 $self->_expand_update_set_values(undef, $data),
669 sub _expand_update_set_values {
670 my ($self, undef, $data) = @_;
671 $self->expand_expr({ -list => [
674 $set = { -bind => $_ } unless defined $set;
675 +{ -op => [ '=', { -ident => $k }, $set ] };
681 ? ($self->{array_datatypes}
682 ? [ $k, +{ -bind => [ $k, $v ] } ]
683 : [ $k, +{ -literal => $v } ])
685 local our $Cur_Col_Meta = $k;
686 [ $k, $self->_expand_expr($v) ]
693 sub _expand_update_clause_target {
694 my ($self, undef, $target) = @_;
695 +(target => $self->expand_expr({ -list => $target }, -ident));
698 sub _expand_update_clause_set {
699 return $_[2] if ref($_[2]) eq 'HASH' and ($_[2]->{-op}||[''])->[0] eq ',';
700 +(set => $_[0]->_expand_update_set_values($_[1], $_[2]));
703 sub _expand_update_clause_where {
704 +(where => $_[0]->expand_expr($_[2]));
707 sub _expand_update_clause_returning {
708 +(returning => $_[0]->expand_expr({ -list => $_[2] }, -ident));
711 # So that subclasses can override UPDATE ... RETURNING separately from
713 sub _update_returning { shift->_returning(@_) }
717 #======================================================================
719 #======================================================================
722 my ($self, @args) = @_;
724 if (ref(my $sel = $args[0]) eq 'HASH') {
728 @clauses{qw(from select where order_by)} = @args;
730 # This oddity is to literalify since historically SQLA doesn't quote
731 # a single identifier argument, so we convert it into a literal
733 $clauses{select} = { -literal => [ $clauses{select}||'*' ] }
734 unless ref($clauses{select});
739 my @rendered = $self->render_statement({ -select => $stmt });
740 return wantarray ? @rendered : $rendered[0];
743 sub _expand_select_clause_select {
744 my ($self, undef, $select) = @_;
745 +(select => $self->expand_expr({ -list => $select }, -ident));
748 sub _expand_select_clause_from {
749 my ($self, undef, $from) = @_;
750 +(from => $self->expand_expr({ -list => $from }, -ident));
753 sub _expand_select_clause_where {
754 my ($self, undef, $where) = @_;
757 if (my $conv = $self->{convert}) {
769 ->wrap_expander(bind => $_wrap)
770 ->wrap_op_expanders(map +($_ => $_wrap), qw(ident value))
771 ->wrap_expander(func => sub {
774 my ($self, $type, $thing) = @_;
775 if (ref($thing) eq 'ARRAY' and $thing->[0] eq $conv
776 and @$thing == 2 and ref($thing->[1]) eq 'HASH'
779 or $thing->[1]{-value}
780 or $thing->[1]{-bind})
782 return { -func => $thing }; # already went through our expander
784 return $self->$orig($type, $thing);
792 return +(where => $sqla->expand_expr($where));
795 sub _expand_select_clause_order_by {
796 my ($self, undef, $order_by) = @_;
797 +(order_by => $self->_expand_order_by($order_by));
801 my ($self, $fields) = @_;
802 return $fields unless ref($fields);
803 my ($sql, @bind) = @{ $self->render_aqt(
804 $self->expand_expr({ -list => $fields }, '-ident')
806 return wantarray ? ($sql, @bind) : $sql;
809 #======================================================================
811 #======================================================================
814 my ($self, $table, $where, $options) = @_;
817 if (ref($table) eq 'HASH') {
820 my %clauses = (target => $table, where => $where, %{$options||{}});
824 my @rendered = $self->render_statement({ -delete => $stmt });
825 return wantarray ? @rendered : $rendered[0];
828 # So that subclasses can override DELETE ... RETURNING separately from
830 sub _delete_returning { shift->_returning(@_) }
832 sub _expand_delete_clause_target {
833 +(target => $_[0]->expand_expr({ -list => $_[2] }, -ident));
836 sub _expand_delete_clause_where { +(where => $_[0]->expand_expr($_[2])); }
838 sub _expand_delete_clause_returning {
839 +(returning => $_[0]->expand_expr({ -list => $_[2] }, -ident));
842 sub _render_delete_clause_target {
843 my ($self, undef, $from) = @_;
844 $self->join_query_parts(' ', { -keyword => 'delete from' }, $from);
847 #======================================================================
849 #======================================================================
853 # Finally, a separate routine just to handle WHERE clauses
855 my ($self, $where, $order) = @_;
857 local $self->{convert_where} = $self->{convert};
860 my ($sql, @bind) = defined($where)
861 ? $self->_recurse_where($where)
863 $sql = (defined $sql and length $sql) ? $self->_sqlcase(' where ') . "( $sql )" : '';
867 my ($order_sql, @order_bind) = $self->_order_by($order);
869 push @bind, @order_bind;
872 return wantarray ? ($sql, @bind) : $sql;
875 { our $Default_Scalar_To = -value }
878 my ($self, $expr, $default_scalar_to) = @_;
879 local our $Default_Scalar_To = $default_scalar_to if $default_scalar_to;
880 $self->_expand_expr($expr);
884 my ($self, $aqt, $top_level) = @_;
885 my ($k, $v, @rest) = %$aqt;
887 die "Not a node type: $k" unless $k =~ s/^-//;
888 if (my $meth = $self->{render}{$k}) {
889 local our $Render_Top_Level = $top_level;
890 return $self->$meth($k, $v)||[];
892 die "notreached: $k";
896 my ($self, $expr, $default_scalar_to) = @_;
897 return @{ $self->render_aqt(
898 $self->expand_expr($expr, $default_scalar_to)
902 sub render_statement {
903 my ($self, $expr, $default_scalar_to) = @_;
905 $self->expand_expr($expr, $default_scalar_to), 1
909 sub _expand_statement {
910 my ($self, $type, $args) = @_;
911 my $ec = $self->{expand_clause};
914 $args->{$type} = delete $args->{_}
916 my %has_clause = map +($_ => 1), @{$self->{clauses_of}{$type}};
917 return +{ "-${type}" => +{
919 my $val = $args->{$_};
920 if (defined($val) and my $exp = $ec->{"${type}.$_"}) {
921 if ((my (@exp) = $self->$exp($_ => $val)) == 1) {
926 } elsif ($has_clause{$_}) {
927 ($_ => $self->expand_expr($val))
935 sub _render_statement {
936 my ($self, $type, $args) = @_;
938 foreach my $clause (@{$self->{clauses_of}{$type}}) {
939 next unless my $clause_expr = $args->{$clause};
941 if (my $rdr = $self->{render_clause}{"${type}.${clause}"}) {
942 $self->$rdr($clause, $clause_expr, $args);
944 my $r = $self->render_aqt($clause_expr, 1);
945 next unless defined $r->[0] and length $r->[0];
946 $self->join_query_parts(' ',
947 { -keyword => $clause },
954 my $q = $self->join_query_parts(' ', @parts);
955 return $self->join_query_parts('',
956 (our $Render_Top_Level ? $q : ('(', $q, ')'))
961 my ($self, $raw) = @_;
963 return $op if grep $_->{$op}, @{$self}{qw(expand_op render_op)};
964 s/^-(?=.)//, s/\s+/_/g for $op;
969 my ($self, $expr) = @_;
970 our $Expand_Depth ||= 0; local $Expand_Depth = $Expand_Depth + 1;
971 return undef unless defined($expr);
972 if (ref($expr) eq 'HASH') {
973 return undef unless my $kc = keys %$expr;
975 return $self->_expand_logop(and => $expr);
977 my ($key, $value) = %$expr;
978 if ($key =~ /^-/ and $key =~ s/ [_\s]? \d+ $//x ) {
979 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
980 . "You probably wanted ...-and => [ $key => COND1, $key => COND2 ... ]";
982 return $self->_expand_hashpair($key, $value);
984 if (ref($expr) eq 'ARRAY') {
985 return $self->_expand_logop(lc($self->{logic}), $expr);
987 if (my $literal = is_literal_value($expr)) {
988 return +{ -literal => $literal };
990 if (!ref($expr) or Scalar::Util::blessed($expr)) {
991 return $self->_expand_scalar($expr);
996 sub _expand_hashpair {
997 my ($self, $k, $v) = @_;
998 unless (defined($k) and length($k)) {
999 if (defined($k) and my $literal = is_literal_value($v)) {
1000 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
1001 return { -literal => $literal };
1003 puke "Supplying an empty left hand side argument is not supported";
1006 return $self->_expand_hashpair_op($k, $v);
1007 } elsif ($k =~ /^\W+$/) {
1008 my ($lhs, @rhs) = ref($v) eq 'ARRAY' ? @$v : $v;
1009 return $self->_expand_op(
1010 -op, [ $k, $self->expand_expr($lhs, -ident), @rhs ]
1013 return $self->_expand_hashpair_ident($k, $v);
1016 sub _expand_hashpair_ident {
1017 my ($self, $k, $v) = @_;
1019 local our $Cur_Col_Meta = $k;
1021 # hash with multiple or no elements is andor
1023 if (ref($v) eq 'HASH' and keys %$v != 1) {
1024 return $self->_expand_logop(and => $v, $k);
1027 # undef needs to be re-sent with cmp to achieve IS/IS NOT NULL
1029 if (is_undef_value($v)) {
1030 return $self->_expand_hashpair_cmp($k => undef);
1033 # scalars and objects get expanded as whatever requested or values
1035 if (!ref($v) or Scalar::Util::blessed($v)) {
1036 return $self->_expand_hashpair_scalar($k, $v);
1039 # single key hashref is a hashtriple
1041 if (ref($v) eq 'HASH') {
1042 return $self->_expand_hashtriple($k, %$v);
1045 # arrayref needs re-engineering over the elements
1047 if (ref($v) eq 'ARRAY') {
1048 return $self->sqlfalse unless @$v;
1049 $self->_debug("ARRAY($k) means distribute over elements");
1051 ($v->[0]||'') =~ /^-(and|or)$/i
1052 ? (shift(@{$v = [ @$v ]}), $1)
1053 : lc($self->{logic} || 'OR')
1055 return $self->_expand_logop(
1060 if (my $literal = is_literal_value($v)) {
1061 unless (length $k) {
1062 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
1065 my ($sql, @bind) = @$literal;
1066 if ($self->{bindtype} eq 'columns') {
1068 $self->_assert_bindval_matches_bindtype($_);
1071 return +{ -literal => [ $self->_quote($k).' '.$sql, @bind ] };
1076 sub _expand_scalar {
1077 my ($self, $expr) = @_;
1079 return $self->_expand_expr({ (our $Default_Scalar_To) => $expr });
1082 sub _expand_hashpair_scalar {
1083 my ($self, $k, $v) = @_;
1085 return $self->_expand_hashpair_cmp(
1086 $k, $self->_expand_scalar($v),
1090 sub _expand_hashpair_op {
1091 my ($self, $k, $v) = @_;
1093 $self->_assert_pass_injection_guard($k =~ /\A-(.*)\Z/s);
1095 my $op = $self->_normalize_op($k);
1097 my $wsop = join(' ', split '_', $op);
1099 my $is_special = List::Util::first { $wsop =~ $_->{regex} }
1100 @{$self->{special_ops}};
1104 # the old special op system requires illegality for top-level use
1107 (our $Expand_Depth) == 1
1111 $self->{disable_old_special_ops}
1112 and List::Util::first { $wsop =~ $_->{regex} } @BUILTIN_SPECIAL_OPS
1116 puke "Illegal use of top-level '-$wsop'"
1120 if (my $exp = $self->{expand}{$op}||$self->{expand_op}{$op}) {
1121 return $self->$exp($op, $v);
1124 if ($self->{render}{$op}) {
1125 return { "-${op}" => $v };
1128 # Ops prefixed with -not_ get converted
1130 if (my ($rest) = $op =~/^not_(.*)$/) {
1133 $self->_expand_expr({ "-${rest}", $v })
1139 # the old unary op system means we should touch nothing and let it work
1141 my $op = join(' ', split '_', $op);
1143 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1144 return { -op => [ $op, $v ] };
1148 my $type = $is_special || $self->{render_op}{$op} ? -op : -func;
1150 if ($self->{restore_old_unop_handling}) {
1157 and (keys %$v)[0] =~ /^-/
1158 and not $self->{render_op}{$op}
1167 if ($type eq -func and ref($v) eq 'ARRAY') {
1168 return $self->_expand_expr({ -func => [ $op, @$v ] });
1171 return $self->_expand_expr({ $type => [ $op, $v ] });
1174 sub _expand_hashpair_cmp {
1175 my ($self, $k, $v) = @_;
1176 $self->_expand_hashtriple($k, $self->{cmp}, $v);
1179 sub _expand_hashtriple {
1180 my ($self, $k, $vk, $vv) = @_;
1182 my $ik = $self->_expand_expr({ -ident => $k });
1184 my $op = $self->_normalize_op($vk);
1185 $self->_assert_pass_injection_guard($op);
1187 if ($op =~ s/ _? \d+ $//x ) {
1188 return $self->_expand_expr($k, { $vk, $vv });
1190 if (my $x = $self->{expand_op}{$op}) {
1191 local our $Cur_Col_Meta = $k;
1192 return $self->$x($op, $vv, $k);
1196 my $op = join(' ', split '_', $op);
1198 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}) {
1199 return { -op => [ $op, $ik, $vv ] };
1201 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1205 { -op => [ $op, $vv ] }
1209 if (ref($vv) eq 'ARRAY') {
1211 my $logic = (defined($raw[0]) and $raw[0] =~ /^-(and|or)$/i)
1212 ? (shift(@raw), lc $1) : 'or';
1213 my @values = map +{ $vk => $_ }, @raw;
1215 $op =~ $self->{inequality_op}
1216 or $op =~ $self->{not_like_op}
1218 if (lc($logic) eq 'or' and @values > 1) {
1219 belch "A multi-element arrayref as an argument to the inequality op '${\uc(join ' ', split '_', $op)}' "
1220 . 'is technically equivalent to an always-true 1=1 (you probably wanted '
1221 . "to say ...{ \$inequality_op => [ -and => \@values ] }... instead)"
1226 # try to DWIM on equality operators
1227 return ($self->_dwim_op_to_is($op,
1228 "Supplying an empty arrayref to '%s' is deprecated",
1229 "operator '%s' applied on an empty array (field '$k')"
1230 ) ? $self->sqlfalse : $self->sqltrue);
1232 return $self->_expand_logop($logic => \@values, $k);
1234 if (is_undef_value($vv)) {
1235 my $is = ($self->_dwim_op_to_is($op,
1236 "Supplying an undefined argument to '%s' is deprecated",
1237 "unexpected operator '%s' with undef operand",
1238 ) ? 'is' : 'is not');
1240 return $self->_expand_hashpair($k => { $is, undef });
1242 local our $Cur_Col_Meta = $k;
1246 $self->_expand_expr($vv)
1250 sub _dwim_op_to_is {
1251 my ($self, $raw, $empty, $fail) = @_;
1253 my $op = $self->_normalize_op($raw);
1255 if ($op =~ /^not$/i) {
1258 if ($op =~ $self->{equality_op}) {
1261 if ($op =~ $self->{like_op}) {
1262 belch(sprintf $empty, uc(join ' ', split '_', $op));
1265 if ($op =~ $self->{inequality_op}) {
1268 if ($op =~ $self->{not_like_op}) {
1269 belch(sprintf $empty, uc(join ' ', split '_', $op));
1272 puke(sprintf $fail, $op);
1276 my ($self, undef, $args) = @_;
1277 my ($func, @args) = @$args;
1278 return +{ -func => [ $func, map $self->expand_expr($_), @args ] };
1282 my ($self, undef, $body) = @_;
1283 unless (defined($body) or (ref($body) and ref($body) eq 'ARRAY')) {
1284 puke "-ident requires a single plain scalar argument (a quotable identifier) or an arrayref of identifier parts";
1286 my ($sep) = map +(defined() ? $_ : '.') , $self->{name_sep};
1287 my @parts = map +($sep
1288 ? map split(/\Q${sep}\E/, $_), @$_
1290 ), ref($body) ? $body : [ $body ];
1291 return { -ident => $parts[-1] } if $self->{_dequalify_idents};
1292 unless ($self->{quote_char}) {
1293 $self->_assert_pass_injection_guard($_) for @parts;
1295 return +{ -ident => \@parts };
1299 +{ -bind => [ our $Cur_Col_Meta, $_[2] ] };
1303 my ($self, undef, $args) = @_;
1304 +{ -row => [ map $self->expand_expr($_), @$args ] };
1308 my ($self, undef, $args) = @_;
1309 my ($op, @opargs) = @$args;
1310 if (my $exp = $self->{expand_op}{$op}) {
1311 return $self->$exp($op, \@opargs);
1313 if (List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1314 return { -op => [ $op, @opargs ] };
1316 +{ -op => [ $op, map $self->expand_expr($_), @opargs ] };
1320 my ($self, undef, $v) = @_;
1322 return $self->_expand_expr($v);
1324 puke "-bool => undef not supported" unless defined($v);
1325 return $self->_expand_expr({ -ident => $v });
1329 my ($self, undef, $expr) = @_;
1331 ',', map $self->expand_expr($_),
1332 @{$expr->{-op}}[1..$#{$expr->{-op}}]
1333 ] } if ref($expr) eq 'HASH' and ($expr->{-op}||[''])->[0] eq ',';
1334 return +{ -op => [ ',',
1335 map $self->expand_expr($_),
1336 ref($expr) eq 'ARRAY' ? @$expr : $expr
1341 my ($self, $logop, $v, $k) = @_;
1342 $self->${\$self->{expand_op}{$logop}}($logop, $v, $k);
1345 sub _expand_op_andor {
1346 my ($self, $logop, $v, $k) = @_;
1348 $v = [ map +{ $k, $_ },
1350 ? (map +{ $_ => $v->{$_} }, sort keys %$v)
1354 if (ref($v) eq 'HASH') {
1355 return undef unless keys %$v;
1358 map $self->_expand_expr({ $_ => $v->{$_} }),
1362 if (ref($v) eq 'ARRAY') {
1363 $logop eq 'and' or $logop eq 'or' or puke "unknown logic: $logop";
1366 (ref($_) eq 'ARRAY' and @$_)
1367 or (ref($_) eq 'HASH' and %$_)
1373 while (my ($el) = splice @expr, 0, 1) {
1374 puke "Supplying an empty left hand side argument is not supported in array-pairs"
1375 unless defined($el) and length($el);
1376 my $elref = ref($el);
1378 local our $Expand_Depth = 0;
1379 push(@res, grep defined, $self->_expand_expr({ $el, shift(@expr) }));
1380 } elsif ($elref eq 'ARRAY') {
1381 push(@res, grep defined, $self->_expand_expr($el)) if @$el;
1382 } elsif (my $l = is_literal_value($el)) {
1383 push @res, { -literal => $l };
1384 } elsif ($elref eq 'HASH') {
1385 local our $Expand_Depth = 0;
1386 push @res, grep defined, $self->_expand_expr($el) if %$el;
1392 # return $res[0] if @res == 1;
1393 return { -op => [ $logop, @res ] };
1399 my ($self, $op, $vv, $k) = @_;
1400 ($k, $vv) = @$vv unless defined $k;
1401 puke "$op can only take undef as argument"
1405 and exists($vv->{-value})
1406 and !defined($vv->{-value})
1408 return +{ -op => [ $op.'_null', $self->expand_expr($k, -ident) ] };
1411 sub _expand_between {
1412 my ($self, $op, $vv, $k) = @_;
1413 my @rhs = map $self->_expand_expr($_),
1414 ref($vv) eq 'ARRAY' ? @$vv : $vv;
1416 (@rhs == 1 and ref($rhs[0]) eq 'HASH' and $rhs[0]->{-literal})
1418 (@rhs == 2 and defined($rhs[0]) and defined($rhs[1]))
1420 puke "Operator '${\uc($op)}' requires either an arrayref with two defined values or expressions, or a single literal scalarref/arrayref-ref";
1424 $self->expand_expr($k),
1425 map $self->expand_expr($_, -value), @rhs
1430 my ($self, $op, $vv, $k) = @_;
1431 if (my $literal = is_literal_value($vv)) {
1432 my ($sql, @bind) = @$literal;
1433 my $opened_sql = $self->_open_outer_paren($sql);
1435 $op, $self->expand_expr($k, -ident),
1436 { -literal => [ $opened_sql, @bind ] }
1440 'SQL::Abstract before v1.75 used to generate incorrect SQL when the '
1441 . "-${\uc($op)} operator was given an undef-containing list: !!!AUDIT YOUR CODE "
1442 . 'AND DATA!!! (the upcoming Data::Query-based version of SQL::Abstract '
1443 . 'will emit the logically correct SQL instead of raising this exception)'
1445 puke("Argument passed to the '${\uc($op)}' operator can not be undefined")
1447 my @rhs = map $self->expand_expr($_, -value),
1448 map { defined($_) ? $_: puke($undef_err) }
1449 (ref($vv) eq 'ARRAY' ? @$vv : $vv);
1450 return $self->${\($op =~ /^not/ ? 'sqltrue' : 'sqlfalse')} unless @rhs;
1454 $self->expand_expr($k, -ident),
1460 my ($self, undef, $v) = @_;
1461 # DBIx::Class requires a nest warning to be emitted once but the private
1462 # method it overrode to do so no longer exists
1463 if ($self->{warn_once_on_nest}) {
1464 unless (our $Nest_Warned) {
1466 "-nest in search conditions is deprecated, you most probably wanted:\n"
1467 .q|{..., -and => [ \%cond0, \@cond1, \'cond2', \[ 'cond3', [ col => bind ] ], etc. ], ... }|
1472 return $self->_expand_expr($v);
1475 sub _expand_values {
1476 my ($self, undef, $values) = @_;
1477 return { -values => [
1480 ? $self->expand_expr($_)
1481 : +{ -row => [ map $self->expand_expr($_), @$_ ] }
1482 ), ref($values) eq 'ARRAY' ? @$values : $values
1486 sub _recurse_where {
1487 my ($self, $where) = @_;
1489 # Special case: top level simple string treated as literal
1491 my $where_exp = (ref($where)
1492 ? $self->_expand_select_clause_where(undef, $where)
1493 : { -literal => [ $where ] });
1495 # dispatch expanded expression
1497 my ($sql, @bind) = defined($where_exp) ? @{ $self->render_aqt($where_exp) || [] } : ();
1498 # DBIx::Class used to call _recurse_where in scalar context
1499 # something else might too...
1501 return ($sql, @bind);
1504 belch "Calling _recurse_where in scalar context is deprecated and will go away before 2.0";
1510 my ($self, undef, $ident) = @_;
1512 return [ $self->_quote($ident) ];
1516 my ($self, undef, $values) = @_;
1517 return $self->join_query_parts('',
1519 $self->_render_op(undef, [ ',', @$values ]),
1525 my ($self, undef, $rest) = @_;
1526 my ($func, @args) = @$rest;
1527 return $self->join_query_parts('',
1528 $self->_sqlcase($func),
1529 $self->join_query_parts('',
1531 $self->join_query_parts(', ', @args),
1538 my ($self, undef, $bind) = @_;
1539 return [ '?', $self->_bindtype(@$bind) ];
1542 sub _render_literal {
1543 my ($self, undef, $literal) = @_;
1544 $self->_assert_bindval_matches_bindtype(@{$literal}[1..$#$literal]);
1548 sub _render_keyword {
1549 my ($self, undef, $keyword) = @_;
1550 return [ $self->_sqlcase(
1551 ref($keyword) ? $$keyword : join ' ', split '_', $keyword
1556 my ($self, undef, $v) = @_;
1557 my ($op, @args) = @$v;
1558 if (my $r = $self->{render_op}{$op}) {
1559 return $self->$r($op, \@args);
1564 my $op = join(' ', split '_', $op);
1566 my $ss = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}};
1567 if ($ss and @args > 1) {
1568 puke "Special op '${op}' requires first value to be identifier"
1569 unless my ($ident) = map $_->{-ident}, grep ref($_) eq 'HASH', $args[0];
1570 my $k = join(($self->{name_sep}||'.'), @$ident);
1571 local our $Expand_Depth = 1;
1572 return [ $self->${\($ss->{handler})}($k, $op, $args[1]) ];
1574 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1575 return [ $self->${\($us->{handler})}($op, $args[0]) ];
1578 return $self->_render_unop_paren($op, \@args);
1582 return $self->_render_unop_prefix($op, \@args);
1584 return $self->_render_op_multop($op, \@args);
1590 sub _render_op_between {
1591 my ($self, $op, $args) = @_;
1592 my ($left, $low, $high) = @$args;
1595 puke "Single arg to between must be a literal"
1596 unless $low->{-literal};
1599 +($low, { -keyword => 'and' }, $high);
1602 return $self->join_query_parts(' ',
1603 '(', $left, { -keyword => $op }, @rh, ')',
1608 my ($self, $op, $args) = @_;
1609 my ($lhs, @rhs) = @$args;
1611 return $self->join_query_parts(' ',
1613 { -keyword => $op },
1614 $self->join_query_parts(' ',
1616 $self->join_query_parts(', ', @rhs),
1622 sub _render_op_andor {
1623 my ($self, $op, $args) = @_;
1624 return undef unless @$args;
1625 return $self->join_query_parts('', $args->[0]) if @$args == 1;
1626 my $inner = $self->_render_op_multop($op, $args);
1627 return undef unless defined($inner->[0]) and length($inner->[0]);
1628 return $self->join_query_parts(' ',
1633 sub _render_op_multop {
1634 my ($self, $op, $args) = @_;
1636 return undef unless @parts;
1637 return $self->render_aqt($parts[0]) if @parts == 1;
1638 my $join = ($op eq ','
1640 : { -keyword => " ${op} " }
1642 return $self->join_query_parts($join, @parts);
1645 sub _render_values {
1646 my ($self, undef, $values) = @_;
1647 my $inner = $self->join_query_parts(' ',
1648 { -keyword => 'values' },
1649 $self->join_query_parts(', ',
1650 ref($values) eq 'ARRAY' ? @$values : $values
1653 return $self->join_query_parts('',
1654 (our $Render_Top_Level ? $inner : ('(', $inner, ')'))
1658 sub join_query_parts {
1659 my ($self, $join, @parts) = @_;
1660 if (ref($join) eq 'HASH') {
1661 $join = $self->render_aqt($join)->[0];
1665 ? $self->render_aqt($_)
1666 : ((ref($_) eq 'ARRAY') ? $_ : [ $_ ])
1669 $self->{join_sql_parts}->(
1670 $join, grep defined && length, map $_->[0], @final
1672 (map @{$_}[1..$#$_], @final),
1676 sub _render_unop_paren {
1677 my ($self, $op, $v) = @_;
1678 return $self->join_query_parts('',
1679 '(', $self->_render_unop_prefix($op, $v), ')'
1683 sub _render_unop_prefix {
1684 my ($self, $op, $v) = @_;
1685 my $op_sql = $self->{restore_old_unop_handling}
1686 ? $self->_sqlcase($op)
1687 : { -keyword => $op };
1688 return $self->join_query_parts(' ',
1689 ($self->{restore_old_unop_handling}
1690 ? $self->_sqlcase($op)
1691 : { -keyword => \$op }),
1696 sub _render_unop_postfix {
1697 my ($self, $op, $v) = @_;
1698 return $self->join_query_parts(' ',
1699 $v->[0], { -keyword => $op },
1703 # Some databases (SQLite) treat col IN (1, 2) different from
1704 # col IN ( (1, 2) ). Use this to strip all outer parens while
1705 # adding them back in the corresponding method
1706 sub _open_outer_paren {
1707 my ($self, $sql) = @_;
1709 while (my ($inner) = $sql =~ /^ \s* \( (.*) \) \s* $/xs) {
1711 # there are closing parens inside, need the heavy duty machinery
1712 # to reevaluate the extraction starting from $sql (full reevaluation)
1713 if ($inner =~ /\)/) {
1714 require Text::Balanced;
1716 my (undef, $remainder) = do {
1717 # idiotic design - writes to $@ but *DOES NOT* throw exceptions
1719 Text::Balanced::extract_bracketed($sql, '()', qr/\s*/);
1722 # the entire expression needs to be a balanced bracketed thing
1723 # (after an extract no remainder sans trailing space)
1724 last if defined $remainder and $remainder =~ /\S/;
1733 sub _where_field_IN {
1734 my ($self, $k, $op, $vals) = @_;
1735 @{$self->_render_op_in(
1738 $self->expand_expr($k, -ident),
1739 map $self->expand_expr($_, -value),
1740 ref($vals) eq 'ARRAY' ? @$vals : $vals
1745 sub _where_field_BETWEEN {
1746 my ($self, $k, $op, $vals) = @_;
1747 @{$self->_render_op_between(
1749 [ $self->expand_expr($k, -ident), ref($vals) eq 'ARRAY' ? @$vals : $vals ]
1753 #======================================================================
1755 #======================================================================
1757 sub _expand_order_by {
1758 my ($self, $arg) = @_;
1760 return unless defined($arg) and not (ref($arg) eq 'ARRAY' and !@$arg);
1762 return $self->expand_expr({ -list => $arg })
1763 if ref($arg) eq 'HASH' and ($arg->{-op}||[''])->[0] eq ',';
1765 my $expander = sub {
1766 my ($self, $dir, $expr) = @_;
1767 my @to_expand = ref($expr) eq 'ARRAY' ? @$expr : $expr;
1768 foreach my $arg (@to_expand) {
1772 and grep /^-(asc|desc)$/, keys %$arg
1774 puke "ordering direction hash passed to order by must have exactly one key (-asc or -desc)";
1778 defined($dir) ? { -op => [ $dir =~ /^-?(.*)$/ ,=> $_ ] } : $_
1780 map $self->expand_expr($_, -ident),
1781 map ref($_) eq 'ARRAY' ? @$_ : $_, @to_expand;
1782 return undef unless @exp;
1783 return undef if @exp == 1 and not defined($exp[0]);
1784 return +{ -op => [ ',', @exp ] };
1787 local @{$self->{expand}}{qw(asc desc)} = (($expander) x 2);
1789 return $self->$expander(undef, $arg);
1793 my ($self, $arg) = @_;
1795 return '' unless defined(my $expanded = $self->_expand_order_by($arg));
1797 my ($sql, @bind) = @{ $self->render_aqt($expanded) };
1799 return '' unless length($sql);
1801 my $final_sql = $self->_sqlcase(' order by ').$sql;
1803 return $final_sql unless wantarray;
1805 return ($final_sql, @bind);
1808 # _order_by no longer needs to call this so doesn't but DBIC uses it.
1810 sub _order_by_chunks {
1811 my ($self, $arg) = @_;
1813 return () unless defined(my $expanded = $self->_expand_order_by($arg));
1815 my @res = $self->_chunkify_order_by($expanded);
1816 (ref() ? $_->[0] : $_) .= '' for @res;
1820 sub _chunkify_order_by {
1821 my ($self, $expanded) = @_;
1823 return grep length, @{ $self->render_aqt($expanded) }
1824 if $expanded->{-ident} or @{$expanded->{-literal}||[]} == 1;
1827 if (ref() eq 'HASH' and $_->{-op} and $_->{-op}[0] eq ',') {
1828 my ($comma, @list) = @{$_->{-op}};
1829 return map $self->_chunkify_order_by($_), @list;
1831 return $self->render_aqt($_);
1835 #======================================================================
1836 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
1837 #======================================================================
1843 $self->expand_expr({ -list => $from }, -ident)
1848 #======================================================================
1850 #======================================================================
1852 # highly optimized, as it's called way too often
1854 # my ($self, $label) = @_;
1856 return '' unless defined $_[1];
1857 return ${$_[1]} if ref($_[1]) eq 'SCALAR';
1858 puke 'Identifier cannot be hashref' if ref($_[1]) eq 'HASH';
1860 unless ($_[0]->{quote_char}) {
1861 if (ref($_[1]) eq 'ARRAY') {
1862 return join($_[0]->{name_sep}||'.', @{$_[1]});
1864 $_[0]->_assert_pass_injection_guard($_[1]);
1869 my $qref = ref $_[0]->{quote_char};
1871 !$qref ? ($_[0]->{quote_char}, $_[0]->{quote_char})
1872 : ($qref eq 'ARRAY') ? @{$_[0]->{quote_char}}
1873 : puke "Unsupported quote_char format: $_[0]->{quote_char}";
1875 my $esc = $_[0]->{escape_char} || $r;
1877 # parts containing * are naturally unquoted
1879 $_[0]->{name_sep}||'',
1883 : do { (my $n = $_) =~ s/(\Q$esc\E|\Q$r\E)/$esc$1/g; $l . $n . $r }
1885 (ref($_[1]) eq 'ARRAY'
1889 ? split (/\Q$_[0]->{name_sep}\E/, $_[1] )
1897 # Conversion, if applicable
1899 #my ($self, $arg) = @_;
1900 if (my $conv = $_[0]->{convert_where}) {
1901 return @{ $_[0]->join_query_parts('',
1902 $_[0]->_sqlcase($conv),
1911 #my ($self, $col, @vals) = @_;
1912 # called often - tighten code
1913 return $_[0]->{bindtype} eq 'columns'
1914 ? map {[$_[1], $_]} @_[2 .. $#_]
1919 # Dies if any element of @bind is not in [colname => value] format
1920 # if bindtype is 'columns'.
1921 sub _assert_bindval_matches_bindtype {
1922 # my ($self, @bind) = @_;
1924 if ($self->{bindtype} eq 'columns') {
1926 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
1927 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
1933 # Fix SQL case, if so requested
1935 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
1936 # don't touch the argument ... crooked logic, but let's not change it!
1937 return $_[0]->{case} ? $_[1] : uc($_[1]);
1940 #======================================================================
1941 # DISPATCHING FROM REFKIND
1942 #======================================================================
1945 my ($self, $data) = @_;
1947 return 'UNDEF' unless defined $data;
1949 # blessed objects are treated like scalars
1950 my $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1952 return 'SCALAR' unless $ref;
1955 while ($ref eq 'REF') {
1957 $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1961 return ($ref||'SCALAR') . ('REF' x $n_steps);
1965 my ($self, $data) = @_;
1966 my @try = ($self->_refkind($data));
1967 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
1968 push @try, 'FALLBACK';
1972 sub _METHOD_FOR_refkind {
1973 my ($self, $meth_prefix, $data) = @_;
1976 for (@{$self->_try_refkind($data)}) {
1977 $method = $self->can($meth_prefix."_".$_)
1981 return $method || puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
1985 sub _SWITCH_refkind {
1986 my ($self, $data, $dispatch_table) = @_;
1989 for (@{$self->_try_refkind($data)}) {
1990 $coderef = $dispatch_table->{$_}
1994 puke "no dispatch entry for ".$self->_refkind($data)
2003 #======================================================================
2004 # VALUES, GENERATE, AUTOLOAD
2005 #======================================================================
2007 # LDNOTE: original code from nwiger, didn't touch code in that section
2008 # I feel the AUTOLOAD stuff should not be the default, it should
2009 # only be activated on explicit demand by user.
2013 my $data = shift || return;
2014 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
2015 unless ref $data eq 'HASH';
2018 foreach my $k (sort keys %$data) {
2019 my $v = $data->{$k};
2020 $self->_SWITCH_refkind($v, {
2022 if ($self->{array_datatypes}) { # array datatype
2023 push @all_bind, $self->_bindtype($k, $v);
2025 else { # literal SQL with bind
2026 my ($sql, @bind) = @$v;
2027 $self->_assert_bindval_matches_bindtype(@bind);
2028 push @all_bind, @bind;
2031 ARRAYREFREF => sub { # literal SQL with bind
2032 my ($sql, @bind) = @${$v};
2033 $self->_assert_bindval_matches_bindtype(@bind);
2034 push @all_bind, @bind;
2036 SCALARREF => sub { # literal SQL without bind
2038 SCALAR_or_UNDEF => sub {
2039 push @all_bind, $self->_bindtype($k, $v);
2050 my(@sql, @sqlq, @sqlv);
2054 if ($ref eq 'HASH') {
2055 for my $k (sort keys %$_) {
2058 my $label = $self->_quote($k);
2059 if ($r eq 'ARRAY') {
2060 # literal SQL with bind
2061 my ($sql, @bind) = @$v;
2062 $self->_assert_bindval_matches_bindtype(@bind);
2063 push @sqlq, "$label = $sql";
2065 } elsif ($r eq 'SCALAR') {
2066 # literal SQL without bind
2067 push @sqlq, "$label = $$v";
2069 push @sqlq, "$label = ?";
2070 push @sqlv, $self->_bindtype($k, $v);
2073 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
2074 } elsif ($ref eq 'ARRAY') {
2075 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
2078 if ($r eq 'ARRAY') { # literal SQL with bind
2079 my ($sql, @bind) = @$v;
2080 $self->_assert_bindval_matches_bindtype(@bind);
2083 } elsif ($r eq 'SCALAR') { # literal SQL without bind
2084 # embedded literal SQL
2091 push @sql, '(' . join(', ', @sqlq) . ')';
2092 } elsif ($ref eq 'SCALAR') {
2096 # strings get case twiddled
2097 push @sql, $self->_sqlcase($_);
2101 my $sql = join ' ', @sql;
2103 # this is pretty tricky
2104 # if ask for an array, return ($stmt, @bind)
2105 # otherwise, s/?/shift @sqlv/ to put it inline
2107 return ($sql, @sqlv);
2109 1 while $sql =~ s/\?/my $d = shift(@sqlv);
2110 ref $d ? $d->[1] : $d/e;
2119 # This allows us to check for a local, then _form, attr
2121 my($name) = $AUTOLOAD =~ /.*::(.+)/;
2122 puke "AUTOLOAD invoked for method name ${name} and allow_autoload option not set" unless $self->{allow_autoload};
2123 return $self->generate($name, @_);
2134 SQL::Abstract - Generate SQL from Perl data structures
2140 my $sql = SQL::Abstract->new;
2142 my($stmt, @bind) = $sql->select($source, \@fields, \%where, $order);
2144 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
2146 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
2148 my($stmt, @bind) = $sql->delete($table, \%where);
2150 # Then, use these in your DBI statements
2151 my $sth = $dbh->prepare($stmt);
2152 $sth->execute(@bind);
2154 # Just generate the WHERE clause
2155 my($stmt, @bind) = $sql->where(\%where, $order);
2157 # Return values in the same order, for hashed queries
2158 # See PERFORMANCE section for more details
2159 my @bind = $sql->values(\%fieldvals);
2163 This module was inspired by the excellent L<DBIx::Abstract>.
2164 However, in using that module I found that what I really wanted
2165 to do was generate SQL, but still retain complete control over my
2166 statement handles and use the DBI interface. So, I set out to
2167 create an abstract SQL generation module.
2169 While based on the concepts used by L<DBIx::Abstract>, there are
2170 several important differences, especially when it comes to WHERE
2171 clauses. I have modified the concepts used to make the SQL easier
2172 to generate from Perl data structures and, IMO, more intuitive.
2173 The underlying idea is for this module to do what you mean, based
2174 on the data structures you provide it. The big advantage is that
2175 you don't have to modify your code every time your data changes,
2176 as this module figures it out.
2178 To begin with, an SQL INSERT is as easy as just specifying a hash
2179 of C<key=value> pairs:
2182 name => 'Jimbo Bobson',
2183 phone => '123-456-7890',
2184 address => '42 Sister Lane',
2185 city => 'St. Louis',
2186 state => 'Louisiana',
2189 The SQL can then be generated with this:
2191 my($stmt, @bind) = $sql->insert('people', \%data);
2193 Which would give you something like this:
2195 $stmt = "INSERT INTO people
2196 (address, city, name, phone, state)
2197 VALUES (?, ?, ?, ?, ?)";
2198 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
2199 '123-456-7890', 'Louisiana');
2201 These are then used directly in your DBI code:
2203 my $sth = $dbh->prepare($stmt);
2204 $sth->execute(@bind);
2206 =head2 Inserting and Updating Arrays
2208 If your database has array types (like for example Postgres),
2209 activate the special option C<< array_datatypes => 1 >>
2210 when creating the C<SQL::Abstract> object.
2211 Then you may use an arrayref to insert and update database array types:
2213 my $sql = SQL::Abstract->new(array_datatypes => 1);
2215 planets => [qw/Mercury Venus Earth Mars/]
2218 my($stmt, @bind) = $sql->insert('solar_system', \%data);
2222 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
2224 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
2227 =head2 Inserting and Updating SQL
2229 In order to apply SQL functions to elements of your C<%data> you may
2230 specify a reference to an arrayref for the given hash value. For example,
2231 if you need to execute the Oracle C<to_date> function on a value, you can
2232 say something like this:
2236 date_entered => \[ "to_date(?,'MM/DD/YYYY')", "03/02/2003" ],
2239 The first value in the array is the actual SQL. Any other values are
2240 optional and would be included in the bind values array. This gives
2243 my($stmt, @bind) = $sql->insert('people', \%data);
2245 $stmt = "INSERT INTO people (name, date_entered)
2246 VALUES (?, to_date(?,'MM/DD/YYYY'))";
2247 @bind = ('Bill', '03/02/2003');
2249 An UPDATE is just as easy, all you change is the name of the function:
2251 my($stmt, @bind) = $sql->update('people', \%data);
2253 Notice that your C<%data> isn't touched; the module will generate
2254 the appropriately quirky SQL for you automatically. Usually you'll
2255 want to specify a WHERE clause for your UPDATE, though, which is
2256 where handling C<%where> hashes comes in handy...
2258 =head2 Complex where statements
2260 This module can generate pretty complicated WHERE statements
2261 easily. For example, simple C<key=value> pairs are taken to mean
2262 equality, and if you want to see if a field is within a set
2263 of values, you can use an arrayref. Let's say we wanted to
2264 SELECT some data based on this criteria:
2267 requestor => 'inna',
2268 worker => ['nwiger', 'rcwe', 'sfz'],
2269 status => { '!=', 'completed' }
2272 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
2274 The above would give you something like this:
2276 $stmt = "SELECT * FROM tickets WHERE
2277 ( requestor = ? ) AND ( status != ? )
2278 AND ( worker = ? OR worker = ? OR worker = ? )";
2279 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
2281 Which you could then use in DBI code like so:
2283 my $sth = $dbh->prepare($stmt);
2284 $sth->execute(@bind);
2290 The methods are simple. There's one for every major SQL operation,
2291 and a constructor you use first. The arguments are specified in a
2292 similar order for each method (table, then fields, then a where
2293 clause) to try and simplify things.
2295 =head2 new(option => 'value')
2297 The C<new()> function takes a list of options and values, and returns
2298 a new B<SQL::Abstract> object which can then be used to generate SQL
2299 through the methods below. The options accepted are:
2305 If set to 'lower', then SQL will be generated in all lowercase. By
2306 default SQL is generated in "textbook" case meaning something like:
2308 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
2310 Any setting other than 'lower' is ignored.
2314 This determines what the default comparison operator is. By default
2315 it is C<=>, meaning that a hash like this:
2317 %where = (name => 'nwiger', email => 'nate@wiger.org');
2319 Will generate SQL like this:
2321 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
2323 However, you may want loose comparisons by default, so if you set
2324 C<cmp> to C<like> you would get SQL such as:
2326 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
2328 You can also override the comparison on an individual basis - see
2329 the huge section on L</"WHERE CLAUSES"> at the bottom.
2331 =item sqltrue, sqlfalse
2333 Expressions for inserting boolean values within SQL statements.
2334 By default these are C<1=1> and C<1=0>. They are used
2335 by the special operators C<-in> and C<-not_in> for generating
2336 correct SQL even when the argument is an empty array (see below).
2340 This determines the default logical operator for multiple WHERE
2341 statements in arrays or hashes. If absent, the default logic is "or"
2342 for arrays, and "and" for hashes. This means that a WHERE
2346 event_date => {'>=', '2/13/99'},
2347 event_date => {'<=', '4/24/03'},
2350 will generate SQL like this:
2352 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
2354 This is probably not what you want given this query, though (look
2355 at the dates). To change the "OR" to an "AND", simply specify:
2357 my $sql = SQL::Abstract->new(logic => 'and');
2359 Which will change the above C<WHERE> to:
2361 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
2363 The logic can also be changed locally by inserting
2364 a modifier in front of an arrayref:
2366 @where = (-and => [event_date => {'>=', '2/13/99'},
2367 event_date => {'<=', '4/24/03'} ]);
2369 See the L</"WHERE CLAUSES"> section for explanations.
2373 This will automatically convert comparisons using the specified SQL
2374 function for both column and value. This is mostly used with an argument
2375 of C<upper> or C<lower>, so that the SQL will have the effect of
2376 case-insensitive "searches". For example, this:
2378 $sql = SQL::Abstract->new(convert => 'upper');
2379 %where = (keywords => 'MaKe iT CAse inSeNSItive');
2381 Will turn out the following SQL:
2383 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
2385 The conversion can be C<upper()>, C<lower()>, or any other SQL function
2386 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
2387 not validate this option; it will just pass through what you specify verbatim).
2391 This is a kludge because many databases suck. For example, you can't
2392 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
2393 Instead, you have to use C<bind_param()>:
2395 $sth->bind_param(1, 'reg data');
2396 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
2398 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
2399 which loses track of which field each slot refers to. Fear not.
2401 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
2402 Currently, you can specify either C<normal> (default) or C<columns>. If you
2403 specify C<columns>, you will get an array that looks like this:
2405 my $sql = SQL::Abstract->new(bindtype => 'columns');
2406 my($stmt, @bind) = $sql->insert(...);
2409 [ 'column1', 'value1' ],
2410 [ 'column2', 'value2' ],
2411 [ 'column3', 'value3' ],
2414 You can then iterate through this manually, using DBI's C<bind_param()>.
2416 $sth->prepare($stmt);
2419 my($col, $data) = @$_;
2420 if ($col eq 'details' || $col eq 'comments') {
2421 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
2422 } elsif ($col eq 'image') {
2423 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
2425 $sth->bind_param($i, $data);
2429 $sth->execute; # execute without @bind now
2431 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
2432 Basically, the advantage is still that you don't have to care which fields
2433 are or are not included. You could wrap that above C<for> loop in a simple
2434 sub called C<bind_fields()> or something and reuse it repeatedly. You still
2435 get a layer of abstraction over manual SQL specification.
2437 Note that if you set L</bindtype> to C<columns>, the C<\[ $sql, @bind ]>
2438 construct (see L</Literal SQL with placeholders and bind values (subqueries)>)
2439 will expect the bind values in this format.
2443 This is the character that a table or column name will be quoted
2444 with. By default this is an empty string, but you could set it to
2445 the character C<`>, to generate SQL like this:
2447 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
2449 Alternatively, you can supply an array ref of two items, the first being the left
2450 hand quote character, and the second the right hand quote character. For
2451 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
2452 that generates SQL like this:
2454 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
2456 Quoting is useful if you have tables or columns names that are reserved
2457 words in your database's SQL dialect.
2461 This is the character that will be used to escape L</quote_char>s appearing
2462 in an identifier before it has been quoted.
2464 The parameter default in case of a single L</quote_char> character is the quote
2467 When opening-closing-style quoting is used (L</quote_char> is an arrayref)
2468 this parameter defaults to the B<closing (right)> L</quote_char>. Occurrences
2469 of the B<opening (left)> L</quote_char> within the identifier are currently left
2470 untouched. The default for opening-closing-style quotes may change in future
2471 versions, thus you are B<strongly encouraged> to specify the escape character
2476 This is the character that separates a table and column name. It is
2477 necessary to specify this when the C<quote_char> option is selected,
2478 so that tables and column names can be individually quoted like this:
2480 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
2482 =item injection_guard
2484 A regular expression C<qr/.../> that is applied to any C<-function> and unquoted
2485 column name specified in a query structure. This is a safety mechanism to avoid
2486 injection attacks when mishandling user input e.g.:
2488 my %condition_as_column_value_pairs = get_values_from_user();
2489 $sqla->select( ... , \%condition_as_column_value_pairs );
2491 If the expression matches an exception is thrown. Note that literal SQL
2492 supplied via C<\'...'> or C<\['...']> is B<not> checked in any way.
2494 Defaults to checking for C<;> and the C<GO> keyword (TransactSQL)
2496 =item array_datatypes
2498 When this option is true, arrayrefs in INSERT or UPDATE are
2499 interpreted as array datatypes and are passed directly
2501 When this option is false, arrayrefs are interpreted
2502 as literal SQL, just like refs to arrayrefs
2503 (but this behavior is for backwards compatibility; when writing
2504 new queries, use the "reference to arrayref" syntax
2510 Takes a reference to a list of "special operators"
2511 to extend the syntax understood by L<SQL::Abstract>.
2512 See section L</"SPECIAL OPERATORS"> for details.
2516 Takes a reference to a list of "unary operators"
2517 to extend the syntax understood by L<SQL::Abstract>.
2518 See section L</"UNARY OPERATORS"> for details.
2524 =head2 insert($table, \@values || \%fieldvals, \%options)
2526 This is the simplest function. You simply give it a table name
2527 and either an arrayref of values or hashref of field/value pairs.
2528 It returns an SQL INSERT statement and a list of bind values.
2529 See the sections on L</"Inserting and Updating Arrays"> and
2530 L</"Inserting and Updating SQL"> for information on how to insert
2531 with those data types.
2533 The optional C<\%options> hash reference may contain additional
2534 options to generate the insert SQL. Currently supported options
2541 Takes either a scalar of raw SQL fields, or an array reference of
2542 field names, and adds on an SQL C<RETURNING> statement at the end.
2543 This allows you to return data generated by the insert statement
2544 (such as row IDs) without performing another C<SELECT> statement.
2545 Note, however, this is not part of the SQL standard and may not
2546 be supported by all database engines.
2550 =head2 update($table, \%fieldvals, \%where, \%options)
2552 This takes a table, hashref of field/value pairs, and an optional
2553 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
2555 See the sections on L</"Inserting and Updating Arrays"> and
2556 L</"Inserting and Updating SQL"> for information on how to insert
2557 with those data types.
2559 The optional C<\%options> hash reference may contain additional
2560 options to generate the update SQL. Currently supported options
2567 See the C<returning> option to
2568 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2572 =head2 select($source, $fields, $where, $order)
2574 This returns a SQL SELECT statement and associated list of bind values, as
2575 specified by the arguments:
2581 Specification of the 'FROM' part of the statement.
2582 The argument can be either a plain scalar (interpreted as a table
2583 name, will be quoted), or an arrayref (interpreted as a list
2584 of table names, joined by commas, quoted), or a scalarref
2585 (literal SQL, not quoted).
2589 Specification of the list of fields to retrieve from
2591 The argument can be either an arrayref (interpreted as a list
2592 of field names, will be joined by commas and quoted), or a
2593 plain scalar (literal SQL, not quoted).
2594 Please observe that this API is not as flexible as that of
2595 the first argument C<$source>, for backwards compatibility reasons.
2599 Optional argument to specify the WHERE part of the query.
2600 The argument is most often a hashref, but can also be
2601 an arrayref or plain scalar --
2602 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
2606 Optional argument to specify the ORDER BY part of the query.
2607 The argument can be a scalar, a hashref or an arrayref
2608 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
2614 =head2 delete($table, \%where, \%options)
2616 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
2617 It returns an SQL DELETE statement and list of bind values.
2619 The optional C<\%options> hash reference may contain additional
2620 options to generate the delete SQL. Currently supported options
2627 See the C<returning> option to
2628 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2632 =head2 where(\%where, $order)
2634 This is used to generate just the WHERE clause. For example,
2635 if you have an arbitrary data structure and know what the
2636 rest of your SQL is going to look like, but want an easy way
2637 to produce a WHERE clause, use this. It returns an SQL WHERE
2638 clause and list of bind values.
2641 =head2 values(\%data)
2643 This just returns the values from the hash C<%data>, in the same
2644 order that would be returned from any of the other above queries.
2645 Using this allows you to markedly speed up your queries if you
2646 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
2648 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
2650 Warning: This is an experimental method and subject to change.
2652 This returns arbitrarily generated SQL. It's a really basic shortcut.
2653 It will return two different things, depending on return context:
2655 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
2656 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
2658 These would return the following:
2660 # First calling form
2661 $stmt = "CREATE TABLE test (?, ?)";
2662 @bind = (field1, field2);
2664 # Second calling form
2665 $stmt_and_val = "CREATE TABLE test (field1, field2)";
2667 Depending on what you're trying to do, it's up to you to choose the correct
2668 format. In this example, the second form is what you would want.
2672 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
2676 ALTER SESSION SET nls_date_format = 'MM/YY'
2678 You get the idea. Strings get their case twiddled, but everything
2679 else remains verbatim.
2681 =head1 EXPORTABLE FUNCTIONS
2683 =head2 is_plain_value
2685 Determines if the supplied argument is a plain value as understood by this
2690 =item * The value is C<undef>
2692 =item * The value is a non-reference
2694 =item * The value is an object with stringification overloading
2696 =item * The value is of the form C<< { -value => $anything } >>
2700 On failure returns C<undef>, on success returns a B<scalar> reference
2701 to the original supplied argument.
2707 The stringification overloading detection is rather advanced: it takes
2708 into consideration not only the presence of a C<""> overload, but if that
2709 fails also checks for enabled
2710 L<autogenerated versions of C<"">|overload/Magic Autogeneration>, based
2711 on either C<0+> or C<bool>.
2713 Unfortunately testing in the field indicates that this
2714 detection B<< may tickle a latent bug in perl versions before 5.018 >>,
2715 but only when very large numbers of stringifying objects are involved.
2716 At the time of writing ( Sep 2014 ) there is no clear explanation of
2717 the direct cause, nor is there a manageably small test case that reliably
2718 reproduces the problem.
2720 If you encounter any of the following exceptions in B<random places within
2721 your application stack> - this module may be to blame:
2723 Operation "ne": no method found,
2724 left argument in overloaded package <something>,
2725 right argument in overloaded package <something>
2729 Stub found while resolving method "???" overloading """" in package <something>
2731 If you fall victim to the above - please attempt to reduce the problem
2732 to something that could be sent to the L<SQL::Abstract developers
2733 |DBIx::Class/GETTING HELP/SUPPORT>
2734 (either publicly or privately). As a workaround in the meantime you can
2735 set C<$ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}> to a true
2736 value, which will most likely eliminate your problem (at the expense of
2737 not being able to properly detect exotic forms of stringification).
2739 This notice and environment variable will be removed in a future version,
2740 as soon as the underlying problem is found and a reliable workaround is
2745 =head2 is_literal_value
2747 Determines if the supplied argument is a literal value as understood by this
2752 =item * C<\$sql_string>
2754 =item * C<\[ $sql_string, @bind_values ]>
2758 On failure returns C<undef>, on success returns an B<array> reference
2759 containing the unpacked version of the supplied literal SQL and bind values.
2761 =head2 is_undef_value
2763 Tests for undef, whether expanded or not.
2765 =head1 WHERE CLAUSES
2769 This module uses a variation on the idea from L<DBIx::Abstract>. It
2770 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
2771 module is that things in arrays are OR'ed, and things in hashes
2774 The easiest way to explain is to show lots of examples. After
2775 each C<%where> hash shown, it is assumed you used:
2777 my($stmt, @bind) = $sql->where(\%where);
2779 However, note that the C<%where> hash can be used directly in any
2780 of the other functions as well, as described above.
2782 =head2 Key-value pairs
2784 So, let's get started. To begin, a simple hash:
2788 status => 'completed'
2791 Is converted to SQL C<key = val> statements:
2793 $stmt = "WHERE user = ? AND status = ?";
2794 @bind = ('nwiger', 'completed');
2796 One common thing I end up doing is having a list of values that
2797 a field can be in. To do this, simply specify a list inside of
2802 status => ['assigned', 'in-progress', 'pending'];
2805 This simple code will create the following:
2807 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
2808 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
2810 A field associated to an empty arrayref will be considered a
2811 logical false and will generate 0=1.
2813 =head2 Tests for NULL values
2815 If the value part is C<undef> then this is converted to SQL <IS NULL>
2824 $stmt = "WHERE user = ? AND status IS NULL";
2827 To test if a column IS NOT NULL:
2831 status => { '!=', undef },
2834 =head2 Specific comparison operators
2836 If you want to specify a different type of operator for your comparison,
2837 you can use a hashref for a given column:
2841 status => { '!=', 'completed' }
2844 Which would generate:
2846 $stmt = "WHERE user = ? AND status != ?";
2847 @bind = ('nwiger', 'completed');
2849 To test against multiple values, just enclose the values in an arrayref:
2851 status => { '=', ['assigned', 'in-progress', 'pending'] };
2853 Which would give you:
2855 "WHERE status = ? OR status = ? OR status = ?"
2858 The hashref can also contain multiple pairs, in which case it is expanded
2859 into an C<AND> of its elements:
2863 status => { '!=', 'completed', -not_like => 'pending%' }
2866 # Or more dynamically, like from a form
2867 $where{user} = 'nwiger';
2868 $where{status}{'!='} = 'completed';
2869 $where{status}{'-not_like'} = 'pending%';
2871 # Both generate this
2872 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
2873 @bind = ('nwiger', 'completed', 'pending%');
2876 To get an OR instead, you can combine it with the arrayref idea:
2880 priority => [ { '=', 2 }, { '>', 5 } ]
2883 Which would generate:
2885 $stmt = "WHERE ( priority = ? OR priority > ? ) AND user = ?";
2886 @bind = ('2', '5', 'nwiger');
2888 If you want to include literal SQL (with or without bind values), just use a
2889 scalar reference or reference to an arrayref as the value:
2892 date_entered => { '>' => \["to_date(?, 'MM/DD/YYYY')", "11/26/2008"] },
2893 date_expires => { '<' => \"now()" }
2896 Which would generate:
2898 $stmt = "WHERE date_entered > to_date(?, 'MM/DD/YYYY') AND date_expires < now()";
2899 @bind = ('11/26/2008');
2902 =head2 Logic and nesting operators
2904 In the example above,
2905 there is a subtle trap if you want to say something like
2906 this (notice the C<AND>):
2908 WHERE priority != ? AND priority != ?
2910 Because, in Perl you I<can't> do this:
2912 priority => { '!=' => 2, '!=' => 1 }
2914 As the second C<!=> key will obliterate the first. The solution
2915 is to use the special C<-modifier> form inside an arrayref:
2917 priority => [ -and => {'!=', 2},
2921 Normally, these would be joined by C<OR>, but the modifier tells it
2922 to use C<AND> instead. (Hint: You can use this in conjunction with the
2923 C<logic> option to C<new()> in order to change the way your queries
2924 work by default.) B<Important:> Note that the C<-modifier> goes
2925 B<INSIDE> the arrayref, as an extra first element. This will
2926 B<NOT> do what you think it might:
2928 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
2930 Here is a quick list of equivalencies, since there is some overlap:
2933 status => {'!=', 'completed', 'not like', 'pending%' }
2934 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
2937 status => {'=', ['assigned', 'in-progress']}
2938 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
2939 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
2943 =head2 Special operators: IN, BETWEEN, etc.
2945 You can also use the hashref format to compare a list of fields using the
2946 C<IN> comparison operator, by specifying the list as an arrayref:
2949 status => 'completed',
2950 reportid => { -in => [567, 2335, 2] }
2953 Which would generate:
2955 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
2956 @bind = ('completed', '567', '2335', '2');
2958 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
2961 If the argument to C<-in> is an empty array, 'sqlfalse' is generated
2962 (by default: C<1=0>). Similarly, C<< -not_in => [] >> generates
2963 'sqltrue' (by default: C<1=1>).
2965 In addition to the array you can supply a chunk of literal sql or
2966 literal sql with bind:
2969 customer => { -in => \[
2970 'SELECT cust_id FROM cust WHERE balance > ?',
2973 status => { -in => \'SELECT status_codes FROM states' },
2979 customer IN ( SELECT cust_id FROM cust WHERE balance > ? )
2980 AND status IN ( SELECT status_codes FROM states )
2984 Finally, if the argument to C<-in> is not a reference, it will be
2985 treated as a single-element array.
2987 Another pair of operators is C<-between> and C<-not_between>,
2988 used with an arrayref of two values:
2992 completion_date => {
2993 -not_between => ['2002-10-01', '2003-02-06']
2999 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
3001 Just like with C<-in> all plausible combinations of literal SQL
3005 start0 => { -between => [ 1, 2 ] },
3006 start1 => { -between => \["? AND ?", 1, 2] },
3007 start2 => { -between => \"lower(x) AND upper(y)" },
3008 start3 => { -between => [
3010 \["upper(?)", 'stuff' ],
3017 ( start0 BETWEEN ? AND ? )
3018 AND ( start1 BETWEEN ? AND ? )
3019 AND ( start2 BETWEEN lower(x) AND upper(y) )
3020 AND ( start3 BETWEEN lower(x) AND upper(?) )
3022 @bind = (1, 2, 1, 2, 'stuff');
3025 These are the two builtin "special operators"; but the
3026 list can be expanded: see section L</"SPECIAL OPERATORS"> below.
3028 =head2 Unary operators: bool
3030 If you wish to test against boolean columns or functions within your
3031 database you can use the C<-bool> and C<-not_bool> operators. For
3032 example to test the column C<is_user> being true and the column
3033 C<is_enabled> being false you would use:-
3037 -not_bool => 'is_enabled',
3042 WHERE is_user AND NOT is_enabled
3044 If a more complex combination is required, testing more conditions,
3045 then you should use the and/or operators:-
3050 -not_bool => { two=> { -rlike => 'bar' } },
3051 -not_bool => { three => [ { '=', 2 }, { '>', 5 } ] },
3062 (NOT ( three = ? OR three > ? ))
3065 =head2 Nested conditions, -and/-or prefixes
3067 So far, we've seen how multiple conditions are joined with a top-level
3068 C<AND>. We can change this by putting the different conditions we want in
3069 hashes and then putting those hashes in an array. For example:
3074 status => { -like => ['pending%', 'dispatched'] },
3078 status => 'unassigned',
3082 This data structure would create the following:
3084 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
3085 OR ( user = ? AND status = ? ) )";
3086 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
3089 Clauses in hashrefs or arrayrefs can be prefixed with an C<-and> or C<-or>
3090 to change the logic inside:
3096 -and => [ workhrs => {'>', 20}, geo => 'ASIA' ],
3097 -or => { workhrs => {'<', 50}, geo => 'EURO' },
3104 $stmt = "WHERE ( user = ?
3105 AND ( ( workhrs > ? AND geo = ? )
3106 OR ( workhrs < ? OR geo = ? ) ) )";
3107 @bind = ('nwiger', '20', 'ASIA', '50', 'EURO');
3109 =head3 Algebraic inconsistency, for historical reasons
3111 C<Important note>: when connecting several conditions, the C<-and->|C<-or>
3112 operator goes C<outside> of the nested structure; whereas when connecting
3113 several constraints on one column, the C<-and> operator goes
3114 C<inside> the arrayref. Here is an example combining both features:
3117 -and => [a => 1, b => 2],
3118 -or => [c => 3, d => 4],
3119 e => [-and => {-like => 'foo%'}, {-like => '%bar'} ]
3124 WHERE ( ( ( a = ? AND b = ? )
3125 OR ( c = ? OR d = ? )
3126 OR ( e LIKE ? AND e LIKE ? ) ) )
3128 This difference in syntax is unfortunate but must be preserved for
3129 historical reasons. So be careful: the two examples below would
3130 seem algebraically equivalent, but they are not
3133 { -like => 'foo%' },
3134 { -like => '%bar' },
3136 # yields: WHERE ( ( col LIKE ? AND col LIKE ? ) )
3139 { col => { -like => 'foo%' } },
3140 { col => { -like => '%bar' } },
3142 # yields: WHERE ( ( col LIKE ? OR col LIKE ? ) )
3145 =head2 Literal SQL and value type operators
3147 The basic premise of SQL::Abstract is that in WHERE specifications the "left
3148 side" is a column name and the "right side" is a value (normally rendered as
3149 a placeholder). This holds true for both hashrefs and arrayref pairs as you
3150 see in the L</WHERE CLAUSES> examples above. Sometimes it is necessary to
3151 alter this behavior. There are several ways of doing so.
3155 This is a virtual operator that signals the string to its right side is an
3156 identifier (a column name) and not a value. For example to compare two
3157 columns you would write:
3160 priority => { '<', 2 },
3161 requestor => { -ident => 'submitter' },
3166 $stmt = "WHERE priority < ? AND requestor = submitter";
3169 If you are maintaining legacy code you may see a different construct as
3170 described in L</Deprecated usage of Literal SQL>, please use C<-ident> in new
3175 This is a virtual operator that signals that the construct to its right side
3176 is a value to be passed to DBI. This is for example necessary when you want
3177 to write a where clause against an array (for RDBMS that support such
3178 datatypes). For example:
3181 array => { -value => [1, 2, 3] }
3186 $stmt = 'WHERE array = ?';
3187 @bind = ([1, 2, 3]);
3189 Note that if you were to simply say:
3195 the result would probably not be what you wanted:
3197 $stmt = 'WHERE array = ? OR array = ? OR array = ?';
3202 Finally, sometimes only literal SQL will do. To include a random snippet
3203 of SQL verbatim, you specify it as a scalar reference. Consider this only
3204 as a last resort. Usually there is a better way. For example:
3207 priority => { '<', 2 },
3208 requestor => { -in => \'(SELECT name FROM hitmen)' },
3213 $stmt = "WHERE priority < ? AND requestor IN (SELECT name FROM hitmen)"
3216 Note that in this example, you only get one bind parameter back, since
3217 the verbatim SQL is passed as part of the statement.
3221 Never use untrusted input as a literal SQL argument - this is a massive
3222 security risk (there is no way to check literal snippets for SQL
3223 injections and other nastyness). If you need to deal with untrusted input
3224 use literal SQL with placeholders as described next.
3226 =head3 Literal SQL with placeholders and bind values (subqueries)
3228 If the literal SQL to be inserted has placeholders and bind values,
3229 use a reference to an arrayref (yes this is a double reference --
3230 not so common, but perfectly legal Perl). For example, to find a date
3231 in Postgres you can use something like this:
3234 date_column => \[ "= date '2008-09-30' - ?::integer", 10 ]
3239 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
3242 Note that you must pass the bind values in the same format as they are returned
3243 by L<where|/where(\%where, $order)>. This means that if you set L</bindtype>
3244 to C<columns>, you must provide the bind values in the
3245 C<< [ column_meta => value ] >> format, where C<column_meta> is an opaque
3246 scalar value; most commonly the column name, but you can use any scalar value
3247 (including references and blessed references), L<SQL::Abstract> will simply
3248 pass it through intact. So if C<bindtype> is set to C<columns> the above
3249 example will look like:
3252 date_column => \[ "= date '2008-09-30' - ?::integer", [ {} => 10 ] ]
3255 Literal SQL is especially useful for nesting parenthesized clauses in the
3256 main SQL query. Here is a first example:
3258 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
3262 bar => \["IN ($sub_stmt)" => @sub_bind],
3267 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
3268 WHERE c2 < ? AND c3 LIKE ?))";
3269 @bind = (1234, 100, "foo%");
3271 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
3272 are expressed in the same way. Of course the C<$sub_stmt> and
3273 its associated bind values can be generated through a former call
3276 my ($sub_stmt, @sub_bind)
3277 = $sql->select("t1", "c1", {c2 => {"<" => 100},
3278 c3 => {-like => "foo%"}});
3281 bar => \["> ALL ($sub_stmt)" => @sub_bind],
3284 In the examples above, the subquery was used as an operator on a column;
3285 but the same principle also applies for a clause within the main C<%where>
3286 hash, like an EXISTS subquery:
3288 my ($sub_stmt, @sub_bind)
3289 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
3290 my %where = ( -and => [
3292 \["EXISTS ($sub_stmt)" => @sub_bind],
3297 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
3298 WHERE c1 = ? AND c2 > t0.c0))";
3302 Observe that the condition on C<c2> in the subquery refers to
3303 column C<t0.c0> of the main query: this is I<not> a bind
3304 value, so we have to express it through a scalar ref.
3305 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
3306 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
3307 what we wanted here.
3309 Finally, here is an example where a subquery is used
3310 for expressing unary negation:
3312 my ($sub_stmt, @sub_bind)
3313 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
3314 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
3316 lname => {like => '%son%'},
3317 \["NOT ($sub_stmt)" => @sub_bind],
3322 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
3323 @bind = ('%son%', 10, 20)
3325 =head3 Deprecated usage of Literal SQL
3327 Below are some examples of archaic use of literal SQL. It is shown only as
3328 reference for those who deal with legacy code. Each example has a much
3329 better, cleaner and safer alternative that users should opt for in new code.
3335 my %where = ( requestor => \'IS NOT NULL' )
3337 $stmt = "WHERE requestor IS NOT NULL"
3339 This used to be the way of generating NULL comparisons, before the handling
3340 of C<undef> got formalized. For new code please use the superior syntax as
3341 described in L</Tests for NULL values>.
3345 my %where = ( requestor => \'= submitter' )
3347 $stmt = "WHERE requestor = submitter"
3349 This used to be the only way to compare columns. Use the superior L</-ident>
3350 method for all new code. For example an identifier declared in such a way
3351 will be properly quoted if L</quote_char> is properly set, while the legacy
3352 form will remain as supplied.
3356 my %where = ( is_ready => \"", completed => { '>', '2012-12-21' } )
3358 $stmt = "WHERE completed > ? AND is_ready"
3359 @bind = ('2012-12-21')
3361 Using an empty string literal used to be the only way to express a boolean.
3362 For all new code please use the much more readable
3363 L<-bool|/Unary operators: bool> operator.
3369 These pages could go on for a while, since the nesting of the data
3370 structures this module can handle are pretty much unlimited (the
3371 module implements the C<WHERE> expansion as a recursive function
3372 internally). Your best bet is to "play around" with the module a
3373 little to see how the data structures behave, and choose the best
3374 format for your data based on that.
3376 And of course, all the values above will probably be replaced with
3377 variables gotten from forms or the command line. After all, if you
3378 knew everything ahead of time, you wouldn't have to worry about
3379 dynamically-generating SQL and could just hardwire it into your
3382 =head1 ORDER BY CLAUSES
3384 Some functions take an order by clause. This can either be a scalar (just a
3385 column name), a hashref of C<< { -desc => 'col' } >> or C<< { -asc => 'col' }
3386 >>, a scalarref, an arrayref-ref, or an arrayref of any of the previous
3389 Given | Will Generate
3390 ---------------------------------------------------------------
3392 'colA' | ORDER BY colA
3394 [qw/colA colB/] | ORDER BY colA, colB
3396 {-asc => 'colA'} | ORDER BY colA ASC
3398 {-desc => 'colB'} | ORDER BY colB DESC
3400 ['colA', {-asc => 'colB'}] | ORDER BY colA, colB ASC
3402 { -asc => [qw/colA colB/] } | ORDER BY colA ASC, colB ASC
3404 \'colA DESC' | ORDER BY colA DESC
3406 \[ 'FUNC(colA, ?)', $x ] | ORDER BY FUNC(colA, ?)
3407 | /* ...with $x bound to ? */
3410 { -asc => 'colA' }, | colA ASC,
3411 { -desc => [qw/colB/] }, | colB DESC,
3412 { -asc => [qw/colC colD/] },| colC ASC, colD ASC,
3413 \'colE DESC', | colE DESC,
3414 \[ 'FUNC(colF, ?)', $x ], | FUNC(colF, ?)
3415 ] | /* ...with $x bound to ? */
3416 ===============================================================
3420 =head1 OLD EXTENSION SYSTEM
3422 =head2 SPECIAL OPERATORS
3424 my $sqlmaker = SQL::Abstract->new(special_ops => [
3428 my ($self, $field, $op, $arg) = @_;
3434 handler => 'method_name',
3438 A "special operator" is a SQL syntactic clause that can be
3439 applied to a field, instead of a usual binary operator.
3442 WHERE field IN (?, ?, ?)
3443 WHERE field BETWEEN ? AND ?
3444 WHERE MATCH(field) AGAINST (?, ?)
3446 Special operators IN and BETWEEN are fairly standard and therefore
3447 are builtin within C<SQL::Abstract> (as the overridable methods
3448 C<_where_field_IN> and C<_where_field_BETWEEN>). For other operators,
3449 like the MATCH .. AGAINST example above which is specific to MySQL,
3450 you can write your own operator handlers - supply a C<special_ops>
3451 argument to the C<new> method. That argument takes an arrayref of
3452 operator definitions; each operator definition is a hashref with two
3459 the regular expression to match the operator
3463 Either a coderef or a plain scalar method name. In both cases
3464 the expected return is C<< ($sql, @bind) >>.
3466 When supplied with a method name, it is simply called on the
3467 L<SQL::Abstract> object as:
3469 $self->$method_name($field, $op, $arg)
3473 $field is the LHS of the operator
3474 $op is the part that matched the handler regex
3477 When supplied with a coderef, it is called as:
3479 $coderef->($self, $field, $op, $arg)
3484 For example, here is an implementation
3485 of the MATCH .. AGAINST syntax for MySQL
3487 my $sqlmaker = SQL::Abstract->new(special_ops => [
3489 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
3490 {regex => qr/^match$/i,
3492 my ($self, $field, $op, $arg) = @_;
3493 $arg = [$arg] if not ref $arg;
3494 my $label = $self->_quote($field);
3495 my ($placeholder) = $self->_convert('?');
3496 my $placeholders = join ", ", (($placeholder) x @$arg);
3497 my $sql = $self->_sqlcase('match') . " ($label) "
3498 . $self->_sqlcase('against') . " ($placeholders) ";
3499 my @bind = $self->_bindtype($field, @$arg);
3500 return ($sql, @bind);
3507 =head2 UNARY OPERATORS
3509 my $sqlmaker = SQL::Abstract->new(unary_ops => [
3513 my ($self, $op, $arg) = @_;
3519 handler => 'method_name',
3523 A "unary operator" is a SQL syntactic clause that can be
3524 applied to a field - the operator goes before the field
3526 You can write your own operator handlers - supply a C<unary_ops>
3527 argument to the C<new> method. That argument takes an arrayref of
3528 operator definitions; each operator definition is a hashref with two
3535 the regular expression to match the operator
3539 Either a coderef or a plain scalar method name. In both cases
3540 the expected return is C<< $sql >>.
3542 When supplied with a method name, it is simply called on the
3543 L<SQL::Abstract> object as:
3545 $self->$method_name($op, $arg)
3549 $op is the part that matched the handler regex
3550 $arg is the RHS or argument of the operator
3552 When supplied with a coderef, it is called as:
3554 $coderef->($self, $op, $arg)
3559 =head1 NEW METHODS (EXPERIMENTAL)
3561 See L<SQL::Abstract::Reference> for the C<expr> versus C<aqt> concept and
3562 an explanation of what the below extensions are extending.
3566 $sqla->plugin('+Foo');
3568 Enables plugin SQL::Abstract::Plugin::Foo.
3572 my ($sql, @bind) = $sqla->render_expr($expr);
3574 =head2 render_statement
3576 Use this if you may be rendering a top level statement so e.g. a SELECT
3577 query doesn't get wrapped in parens
3579 my ($sql, @bind) = $sqla->render_statement($expr);
3583 Expression expansion with optional default for scalars.
3585 my $aqt = $self->expand_expr($expr);
3586 my $aqt = $self->expand_expr($expr, -ident);
3590 Top level means avoid parens on statement AQT.
3592 my $res = $self->render_aqt($aqt, $top_level);
3593 my ($sql, @bind) = @$res;
3595 =head2 join_query_parts
3597 Similar to join() but will render hashrefs as nodes for both join and parts,
3598 and treats arrayref as a nested C<[ $join, @parts ]> structure.
3600 my $part = $self->join_query_parts($join, @parts);
3602 =head1 NEW EXTENSION SYSTEM
3606 my $sqla2 = $sqla->clone;
3608 Performs a semi-shallow copy such that extension methods won't leak state
3609 but excessive depth is avoided.
3619 =head2 clause_expander
3621 =head2 clause_expanders
3623 $sqla->expander('name' => sub { ... });
3624 $sqla->expanders('name1' => sub { ... }, 'name2' => sub { ... });
3626 =head2 expander_list
3628 =head2 op_expander_list
3630 =head2 clause_expander_list
3632 my @names = $sqla->expander_list;
3634 =head2 wrap_expander
3636 =head2 wrap_expanders
3638 =head2 wrap_op_expander
3640 =head2 wrap_op_expanders
3642 =head2 wrap_clause_expander
3644 =head2 wrap_clause_expanders
3646 $sqla->wrap_expander('name' => sub { my ($orig) = @_; sub { ... } });
3647 $sqla->wrap_expanders(
3648 'name1' => sub { my ($orig1) = @_; sub { ... } },
3649 'name2' => sub { my ($orig2) = @_; sub { ... } },
3660 =head2 clause_renderer
3662 =head2 clause_renderers
3664 $sqla->renderer('name' => sub { ... });
3665 $sqla->renderers('name1' => sub { ... }, 'name2' => sub { ... });
3667 =head2 renderer_list
3669 =head2 op_renderer_list
3671 =head2 clause_renderer_list
3673 my @names = $sqla->renderer_list;
3675 =head2 wrap_renderer
3677 =head2 wrap_renderers
3679 =head2 wrap_op_renderer
3681 =head2 wrap_op_renderers
3683 =head2 wrap_clause_renderer
3685 =head2 wrap_clause_renderers
3687 $sqla->wrap_renderer('name' => sub { my ($orig) = @_; sub { ... } });
3688 $sqla->wrap_renderers(
3689 'name1' => sub { my ($orig1) = @_; sub { ... } },
3690 'name2' => sub { my ($orig2) = @_; sub { ... } },
3695 my @clauses = $sqla->clauses_of('select');
3696 $sqla->clauses_of(select => \@new_clauses);
3697 $sqla->clauses_of(select => sub {
3698 my (undef, @old_clauses) = @_;
3700 return @new_clauses;
3703 =head2 statement_list
3705 my @list = $sqla->statement_list;
3707 =head2 make_unop_expander
3709 my $exp = $sqla->make_unop_expander(sub { ... });
3711 If the op is found as a binop, assumes it wants a default comparison, so
3712 the inner expander sub can reliably operate as
3714 sub { my ($self, $name, $body) = @_; ... }
3716 =head2 make_binop_expander
3718 my $exp = $sqla->make_binop_expander(sub { ... });
3720 If the op is found as a unop, assumes the value will be an arrayref with the
3721 LHS as the first entry, and converts that to an ident node if it's a simple
3722 scalar. So the inner expander sub looks like
3725 my ($self, $name, $body, $k) = @_;
3726 { -blah => [ map $self->expand_expr($_), $k, $body ] }
3729 =head2 unop_expander
3731 =head2 unop_expanders
3733 =head2 binop_expander
3735 =head2 binop_expanders
3737 The above methods operate exactly like the op_ versions but wrap the coderef
3738 using the appropriate make_ method first.
3742 Thanks to some benchmarking by Mark Stosberg, it turns out that
3743 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
3744 I must admit this wasn't an intentional design issue, but it's a
3745 byproduct of the fact that you get to control your C<DBI> handles
3748 To maximize performance, use a code snippet like the following:
3750 # prepare a statement handle using the first row
3751 # and then reuse it for the rest of the rows
3753 for my $href (@array_of_hashrefs) {
3754 $stmt ||= $sql->insert('table', $href);
3755 $sth ||= $dbh->prepare($stmt);
3756 $sth->execute($sql->values($href));
3759 The reason this works is because the keys in your C<$href> are sorted
3760 internally by B<SQL::Abstract>. Thus, as long as your data retains
3761 the same structure, you only have to generate the SQL the first time
3762 around. On subsequent queries, simply use the C<values> function provided
3763 by this module to return your values in the correct order.
3765 However this depends on the values having the same type - if, for
3766 example, the values of a where clause may either have values
3767 (resulting in sql of the form C<column = ?> with a single bind
3768 value), or alternatively the values might be C<undef> (resulting in
3769 sql of the form C<column IS NULL> with no bind value) then the
3770 caching technique suggested will not work.
3774 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
3775 really like this part (I do, at least). Building up a complex query
3776 can be as simple as the following:
3783 use CGI::FormBuilder;
3786 my $form = CGI::FormBuilder->new(...);
3787 my $sql = SQL::Abstract->new;
3789 if ($form->submitted) {
3790 my $field = $form->field;
3791 my $id = delete $field->{id};
3792 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
3795 Of course, you would still have to connect using C<DBI> to run the
3796 query, but the point is that if you make your form look like your
3797 table, the actual query script can be extremely simplistic.
3799 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
3800 a fast interface to returning and formatting data. I frequently
3801 use these three modules together to write complex database query
3802 apps in under 50 lines.
3804 =head1 HOW TO CONTRIBUTE
3806 Contributions are always welcome, in all usable forms (we especially
3807 welcome documentation improvements). The delivery methods include git-
3808 or unified-diff formatted patches, GitHub pull requests, or plain bug
3809 reports either via RT or the Mailing list. Contributors are generally
3810 granted full access to the official repository after their first several
3811 patches pass successful review.
3813 This project is maintained in a git repository. The code and related tools are
3814 accessible at the following locations:
3818 =item * Official repo: L<git://git.shadowcat.co.uk/dbsrgits/SQL-Abstract.git>
3820 =item * Official gitweb: L<http://git.shadowcat.co.uk/gitweb/gitweb.cgi?p=dbsrgits/SQL-Abstract.git>
3822 =item * GitHub mirror: L<https://github.com/dbsrgits/sql-abstract>
3824 =item * Authorized committers: L<ssh://dbsrgits@git.shadowcat.co.uk/SQL-Abstract.git>
3830 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
3831 Great care has been taken to preserve the I<published> behavior
3832 documented in previous versions in the 1.* family; however,
3833 some features that were previously undocumented, or behaved
3834 differently from the documentation, had to be changed in order
3835 to clarify the semantics. Hence, client code that was relying
3836 on some dark areas of C<SQL::Abstract> v1.*
3837 B<might behave differently> in v1.50.
3839 The main changes are:
3845 support for literal SQL through the C<< \ [ $sql, @bind ] >> syntax.
3849 support for the { operator => \"..." } construct (to embed literal SQL)
3853 support for the { operator => \["...", @bind] } construct (to embed literal SQL with bind values)
3857 optional support for L<array datatypes|/"Inserting and Updating Arrays">
3861 defensive programming: check arguments
3865 fixed bug with global logic, which was previously implemented
3866 through global variables yielding side-effects. Prior versions would
3867 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
3868 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
3869 Now this is interpreted
3870 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
3875 fixed semantics of _bindtype on array args
3879 dropped the C<_anoncopy> of the %where tree. No longer necessary,
3880 we just avoid shifting arrays within that tree.
3884 dropped the C<_modlogic> function
3888 =head1 ACKNOWLEDGEMENTS
3890 There are a number of individuals that have really helped out with
3891 this module. Unfortunately, most of them submitted bugs via CPAN
3892 so I have no idea who they are! But the people I do know are:
3894 Ash Berlin (order_by hash term support)
3895 Matt Trout (DBIx::Class support)
3896 Mark Stosberg (benchmarking)
3897 Chas Owens (initial "IN" operator support)
3898 Philip Collins (per-field SQL functions)
3899 Eric Kolve (hashref "AND" support)
3900 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
3901 Dan Kubb (support for "quote_char" and "name_sep")
3902 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
3903 Laurent Dami (internal refactoring, extensible list of special operators, literal SQL)
3904 Norbert Buchmuller (support for literal SQL in hashpair, misc. fixes & tests)
3905 Peter Rabbitson (rewrite of SQLA::Test, misc. fixes & tests)
3906 Oliver Charles (support for "RETURNING" after "INSERT")
3912 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
3916 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
3918 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
3920 For support, your best bet is to try the C<DBIx::Class> users mailing list.
3921 While not an official support venue, C<DBIx::Class> makes heavy use of
3922 C<SQL::Abstract>, and as such list members there are very familiar with
3923 how to create queries.
3927 This module is free software; you may copy this under the same
3928 terms as perl itself (either the GNU General Public License or
3929 the Artistic License)