1 package SQL::Abstract; # see doc at end of file
10 our @EXPORT_OK = qw(is_plain_value is_literal_value);
20 *SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION = $ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}
26 #======================================================================
28 #======================================================================
30 our $VERSION = '1.87';
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/^ (?: not \s )? in $/ix, handler => sub { die "NOPE" }},
42 {regex => qr/^ is (?: \s+ not )? $/ix, handler => sub { die "NOPE" }},
45 #======================================================================
46 # DEBUGGING AND ERROR REPORTING
47 #======================================================================
50 return unless $_[0]->{debug}; shift; # a little faster
51 my $func = (caller(1))[3];
52 warn "[$func] ", @_, "\n";
56 my($func) = (caller(1))[3];
57 Carp::carp "[$func] Warning: ", @_;
61 my($func) = (caller(1))[3];
62 Carp::croak "[$func] Fatal: ", @_;
65 sub is_literal_value ($) {
66 ref $_[0] eq 'SCALAR' ? [ ${$_[0]} ]
67 : ( ref $_[0] eq 'REF' and ref ${$_[0]} eq 'ARRAY' ) ? [ @${ $_[0] } ]
71 # FIXME XSify - this can be done so much more efficiently
72 sub is_plain_value ($) {
74 ! length ref $_[0] ? \($_[0])
76 ref $_[0] eq 'HASH' and keys %{$_[0]} == 1
78 exists $_[0]->{-value}
79 ) ? \($_[0]->{-value})
81 # reuse @_ for even moar speedz
82 defined ( $_[1] = Scalar::Util::blessed $_[0] )
84 # deliberately not using Devel::OverloadInfo - the checks we are
85 # intersted in are much more limited than the fullblown thing, and
86 # this is a very hot piece of code
88 # simply using ->can('(""') can leave behind stub methods that
89 # break actually using the overload later (see L<perldiag/Stub
90 # found while resolving method "%s" overloading "%s" in package
91 # "%s"> and the source of overload::mycan())
93 # either has stringification which DBI SHOULD prefer out of the box
94 grep { *{ (qq[${_}::(""]) }{CODE} } @{ $_[2] = mro::get_linear_isa( $_[1] ) }
96 # has nummification or boolification, AND fallback is *not* disabled
98 SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION
101 grep { *{"${_}::(0+"}{CODE} } @{$_[2]}
103 grep { *{"${_}::(bool"}{CODE} } @{$_[2]}
107 # no fallback specified at all
108 ! ( ($_[3]) = grep { *{"${_}::()"}{CODE} } @{$_[2]} )
110 # fallback explicitly undef
111 ! defined ${"$_[3]::()"}
124 #======================================================================
126 #======================================================================
130 my $class = ref($self) || $self;
131 my %opt = (ref $_[0] eq 'HASH') ? %{$_[0]} : @_;
133 # choose our case by keeping an option around
134 delete $opt{case} if $opt{case} && $opt{case} ne 'lower';
136 # default logic for interpreting arrayrefs
137 $opt{logic} = $opt{logic} ? uc $opt{logic} : 'OR';
139 # how to return bind vars
140 $opt{bindtype} ||= 'normal';
142 # default comparison is "=", but can be overridden
145 # try to recognize which are the 'equality' and 'inequality' ops
146 # (temporary quickfix (in 2007), should go through a more seasoned API)
147 $opt{equality_op} = qr/^( \Q$opt{cmp}\E | \= )$/ix;
148 $opt{inequality_op} = qr/^( != | <> )$/ix;
150 $opt{like_op} = qr/^ (is\s+)? r?like $/xi;
151 $opt{not_like_op} = qr/^ (is\s+)? not \s+ r?like $/xi;
154 $opt{sqltrue} ||= '1=1';
155 $opt{sqlfalse} ||= '0=1';
158 $opt{special_ops} ||= [];
160 # regexes are applied in order, thus push after user-defines
161 push @{$opt{special_ops}}, @BUILTIN_SPECIAL_OPS;
163 if ($class->isa('DBIx::Class::SQLMaker')) {
164 push @{$opt{special_ops}}, our $DBIC_Compat_Op ||= {
165 regex => qr/^(?:ident|value)$/i, handler => sub { die "NOPE" }
167 $opt{is_dbic_sqlmaker} = 1;
171 $opt{unary_ops} ||= [];
173 # rudimentary sanity-check for user supplied bits treated as functions/operators
174 # If a purported function matches this regular expression, an exception is thrown.
175 # Literal SQL is *NOT* subject to this check, only functions (and column names
176 # when quoting is not in effect)
179 # need to guard against ()'s in column names too, but this will break tons of
180 # hacks... ideas anyone?
181 $opt{injection_guard} ||= qr/
188 (map +("-$_", "_render_$_"), qw(op func bind ident literal list)),
192 $opt{expand_unary} = {};
195 -ident => '_expand_ident',
196 -value => sub { +{ -bind => [ our $Cur_Col_Meta, $_[2] ] } },
197 -not => sub { +{ -op => [ 'not', $_[0]->_expand_expr($_[2]) ] } },
199 my ($self, undef, $v) = @_;
201 return $self->_expand_expr($v);
203 puke "-bool => undef not supported" unless defined($v);
204 return $self->_expand_ident(-ident => $v);
208 return bless \%opt, $class;
211 sub sqltrue { +{ -literal => [ $_[0]->{sqltrue} ] } }
212 sub sqlfalse { +{ -literal => [ $_[0]->{sqlfalse} ] } }
214 sub _assert_pass_injection_guard {
215 if ($_[1] =~ $_[0]->{injection_guard}) {
216 my $class = ref $_[0];
217 puke "Possible SQL injection attempt '$_[1]'. If this is indeed a part of the "
218 . "desired SQL use literal SQL ( \'...' or \[ '...' ] ) or supply your own "
219 . "{injection_guard} attribute to ${class}->new()"
224 #======================================================================
226 #======================================================================
230 my $table = $self->_table(shift);
231 my $data = shift || return;
234 my $method = $self->_METHOD_FOR_refkind("_insert", $data);
235 my ($sql, @bind) = $self->$method($data);
236 $sql = join " ", $self->_sqlcase('insert into'), $table, $sql;
238 if ($options->{returning}) {
239 my ($s, @b) = $self->_insert_returning($options);
244 return wantarray ? ($sql, @bind) : $sql;
247 # So that subclasses can override INSERT ... RETURNING separately from
248 # UPDATE and DELETE (e.g. DBIx::Class::SQLMaker::Oracle does this)
249 sub _insert_returning { shift->_returning(@_) }
252 my ($self, $options) = @_;
254 my $f = $options->{returning};
256 my ($sql, @bind) = $self->render_aqt(
257 $self->_expand_maybe_list_expr($f, undef, -ident)
260 ? $self->_sqlcase(' returning ') . $sql
261 : ($self->_sqlcase(' returning ').$sql, @bind);
264 sub _insert_HASHREF { # explicit list of fields and then values
265 my ($self, $data) = @_;
267 my @fields = sort keys %$data;
269 my ($sql, @bind) = $self->_insert_values($data);
272 $_ = $self->_quote($_) foreach @fields;
273 $sql = "( ".join(", ", @fields).") ".$sql;
275 return ($sql, @bind);
278 sub _insert_ARRAYREF { # just generate values(?,?) part (no list of fields)
279 my ($self, $data) = @_;
281 # no names (arrayref) so can't generate bindtype
282 $self->{bindtype} ne 'columns'
283 or belch "can't do 'columns' bindtype when called with arrayref";
285 my (@values, @all_bind);
286 foreach my $value (@$data) {
287 my ($values, @bind) = $self->_insert_value(undef, $value);
288 push @values, $values;
289 push @all_bind, @bind;
291 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
292 return ($sql, @all_bind);
295 sub _insert_ARRAYREFREF { # literal SQL with bind
296 my ($self, $data) = @_;
298 my ($sql, @bind) = @${$data};
299 $self->_assert_bindval_matches_bindtype(@bind);
301 return ($sql, @bind);
305 sub _insert_SCALARREF { # literal SQL without bind
306 my ($self, $data) = @_;
312 my ($self, $data) = @_;
314 my (@values, @all_bind);
315 foreach my $column (sort keys %$data) {
316 my ($values, @bind) = $self->_insert_value($column, $data->{$column});
317 push @values, $values;
318 push @all_bind, @bind;
320 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
321 return ($sql, @all_bind);
325 my ($self, $column, $v) = @_;
327 return $self->render_aqt(
328 $self->_expand_insert_value($column, $v)
332 sub _expand_insert_value {
333 my ($self, $column, $v) = @_;
335 if (ref($v) eq 'ARRAY') {
336 if ($self->{array_datatypes}) {
337 return +{ -bind => [ $column, $v ] };
339 my ($sql, @bind) = @$v;
340 $self->_assert_bindval_matches_bindtype(@bind);
341 return +{ -literal => $v };
343 if (ref($v) eq 'HASH') {
344 if (grep !/^-/, keys %$v) {
345 belch "HASH ref as bind value in insert is not supported";
346 return +{ -bind => [ $column, $v ] };
350 return +{ -bind => [ $column, undef ] };
352 local our $Cur_Col_Meta = $column;
353 return $self->expand_expr($v);
358 #======================================================================
360 #======================================================================
365 my $table = $self->_table(shift);
366 my $data = shift || return;
370 # first build the 'SET' part of the sql statement
371 puke "Unsupported data type specified to \$sql->update"
372 unless ref $data eq 'HASH';
374 my ($sql, @all_bind) = $self->_update_set_values($data);
375 $sql = $self->_sqlcase('update ') . $table . $self->_sqlcase(' set ')
379 my($where_sql, @where_bind) = $self->where($where);
381 push @all_bind, @where_bind;
384 if ($options->{returning}) {
385 my ($returning_sql, @returning_bind) = $self->_update_returning($options);
386 $sql .= $returning_sql;
387 push @all_bind, @returning_bind;
390 return wantarray ? ($sql, @all_bind) : $sql;
393 sub _update_set_values {
394 my ($self, $data) = @_;
396 return $self->render_aqt(
397 $self->_expand_update_set_values($data),
401 sub _expand_update_set_values {
402 my ($self, $data) = @_;
403 $self->_expand_maybe_list_expr( [
406 $set = { -bind => $_ } unless defined $set;
407 +{ -op => [ '=', $self->_expand_ident(-ident => $k), $set ] };
413 ? ($self->{array_datatypes}
414 ? [ $k, +{ -bind => [ $k, $v ] } ]
415 : [ $k, +{ -literal => $v } ])
417 local our $Cur_Col_Meta = $k;
418 [ $k, $self->_expand_expr($v) ]
425 # So that subclasses can override UPDATE ... RETURNING separately from
427 sub _update_returning { shift->_returning(@_) }
431 #======================================================================
433 #======================================================================
438 my $table = $self->_table(shift);
439 my $fields = shift || '*';
443 my ($fields_sql, @bind) = $self->_select_fields($fields);
445 my ($where_sql, @where_bind) = $self->where($where, $order);
446 push @bind, @where_bind;
448 my $sql = join(' ', $self->_sqlcase('select'), $fields_sql,
449 $self->_sqlcase('from'), $table)
452 return wantarray ? ($sql, @bind) : $sql;
456 my ($self, $fields) = @_;
457 return $fields unless ref($fields);
458 return $self->render_aqt(
459 $self->_expand_maybe_list_expr($fields, undef, '-ident')
463 #======================================================================
465 #======================================================================
470 my $table = $self->_table(shift);
474 my($where_sql, @bind) = $self->where($where);
475 my $sql = $self->_sqlcase('delete from ') . $table . $where_sql;
477 if ($options->{returning}) {
478 my ($returning_sql, @returning_bind) = $self->_delete_returning($options);
479 $sql .= $returning_sql;
480 push @bind, @returning_bind;
483 return wantarray ? ($sql, @bind) : $sql;
486 # So that subclasses can override DELETE ... RETURNING separately from
488 sub _delete_returning { shift->_returning(@_) }
492 #======================================================================
494 #======================================================================
498 # Finally, a separate routine just to handle WHERE clauses
500 my ($self, $where, $order) = @_;
502 local $self->{convert_where} = $self->{convert};
505 my ($sql, @bind) = defined($where)
506 ? $self->_recurse_where($where)
508 $sql = (defined $sql and length $sql) ? $self->_sqlcase(' where ') . "( $sql )" : '';
512 my ($order_sql, @order_bind) = $self->_order_by($order);
514 push @bind, @order_bind;
517 return wantarray ? ($sql, @bind) : $sql;
521 my ($self, $expr, $default_scalar_to) = @_;
522 local our $Default_Scalar_To = $default_scalar_to if $default_scalar_to;
523 $self->_expand_expr($expr);
527 my ($self, $aqt) = @_;
528 my ($k, $v, @rest) = %$aqt;
530 if (my $meth = $self->{render}{$k}) {
531 return $self->$meth($v);
533 die "notreached: $k";
537 my ($self, $expr) = @_;
538 $self->render_aqt($self->expand_expr($expr));
542 my ($self, $expr, $logic) = @_;
543 our $Expand_Depth ||= 0; local $Expand_Depth = $Expand_Depth + 1;
544 return undef unless defined($expr);
545 if (ref($expr) eq 'HASH') {
546 return undef unless my $kc = keys %$expr;
551 map $self->_expand_expr({ $_ => $expr->{$_} }, $logic),
555 my ($key, $value) = %$expr;
556 if (my $exp = $self->{expand}{$key}) {
557 return $self->$exp($key, $value);
559 return $self->_expand_expr_hashpair($key, $value, $logic);
561 if (ref($expr) eq 'ARRAY') {
562 my $logic = lc($logic || $self->{logic});
563 $logic eq 'and' or $logic eq 'or' or puke "unknown logic: $logic";
566 (ref($_) eq 'ARRAY' and @$_)
567 or (ref($_) eq 'HASH' and %$_)
573 while (my ($el) = splice @expr, 0, 1) {
574 puke "Supplying an empty left hand side argument is not supported in array-pairs"
575 unless defined($el) and length($el);
576 my $elref = ref($el);
578 local $Expand_Depth = 0;
579 push(@res, grep defined, $self->_expand_expr({ $el, shift(@expr) }));
580 } elsif ($elref eq 'ARRAY') {
581 push(@res, grep defined, $self->_expand_expr($el)) if @$el;
582 } elsif (my $l = is_literal_value($el)) {
583 push @res, { -literal => $l };
584 } elsif ($elref eq 'HASH') {
585 local $Expand_Depth = 0;
586 push @res, grep defined, $self->_expand_expr($el) if %$el;
592 # return $res[0] if @res == 1;
593 return { -op => [ $logic, @res ] };
595 if (my $literal = is_literal_value($expr)) {
596 return +{ -literal => $literal };
598 if (!ref($expr) or Scalar::Util::blessed($expr)) {
599 if (my $d = our $Default_Scalar_To) {
600 return $self->_expand_expr({ $d => $expr });
602 if (my $m = our $Cur_Col_Meta) {
603 return +{ -bind => [ $m, $expr ] };
605 return +{ -bind => [ undef, $expr ] };
610 sub _expand_expr_hashpair {
611 my ($self, $k, $v, $logic) = @_;
612 unless (defined($k) and length($k)) {
613 if (defined($k) and my $literal = is_literal_value($v)) {
614 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
615 return { -literal => $literal };
617 puke "Supplying an empty left hand side argument is not supported";
620 $self->_assert_pass_injection_guard($k =~ /^-(.*)$/s);
621 if ($k =~ s/ [_\s]? \d+ $//x ) {
622 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
623 . "You probably wanted ...-and => [ $k => COND1, $k => COND2 ... ]";
626 # DBIx::Class requires a nest warning to be emitted once but the private
627 # method it overrode to do so no longer exists
628 if ($self->{is_dbic_sqlmaker}) {
629 unless (our $Nest_Warned) {
631 "-nest in search conditions is deprecated, you most probably wanted:\n"
632 .q|{..., -and => [ \%cond0, \@cond1, \'cond2', \[ 'cond3', [ col => bind ] ], etc. ], ... }|
637 return $self->_expand_expr($v);
639 if (my ($rest) = $k =~/^-not[_ ](.*)$/) {
642 $self->_expand_expr({ "-${rest}", $v }, $logic)
645 if (my ($logic) = $k =~ /^-(and|or)$/i) {
646 if (ref($v) eq 'HASH') {
647 return $self->_expand_expr($v, $logic);
649 if (ref($v) eq 'ARRAY') {
650 return $self->_expand_expr($v, $logic);
655 $op =~ s/^-// if length($op) > 1;
657 # top level special ops are illegal in general
658 # note that, arguably, if it makes no sense at top level, it also
659 # makes no sense on the other side of an = sign or similar but DBIC
660 # gets disappointingly upset if I disallow it
662 (our $Expand_Depth) == 1
663 and List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}
665 puke "Illegal use of top-level '-$op'"
667 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
668 return { -op => [ $op, $v ] };
671 if (my $custom = $self->{expand_unary}{$k}) {
672 return $self->$custom($v);
674 if ($self->{render}{$k}) {
680 and (keys %$v)[0] =~ /^-/
682 my ($func) = $k =~ /^-(.*)$/;
683 if (List::Util::first { $func =~ $_->{regex} } @{$self->{special_ops}}) {
684 return +{ -op => [ $func, $self->_expand_expr($v) ] };
686 return +{ -func => [ $func, $self->_expand_expr($v) ] };
688 if (!ref($v) or is_literal_value($v)) {
689 return +{ -op => [ $k =~ /^-(.*)$/, $self->_expand_expr($v) ] };
696 and exists $v->{-value}
697 and not defined $v->{-value}
700 return $self->_expand_expr({ $k => { $self->{cmp} => undef } });
702 if (!ref($v) or Scalar::Util::blessed($v)) {
703 my $d = our $Default_Scalar_To;
707 $self->_expand_ident(-ident => $k),
709 ? $self->_expand_expr($d => $v)
710 : { -bind => [ $k, $v ] }
715 if (ref($v) eq 'HASH') {
719 map $self->_expand_expr({ $k => { $_ => $v->{$_} } }),
723 return undef unless keys %$v;
725 my $op = join ' ', split '_', (map lc, $vk =~ /^-?(.*)$/)[0];
726 $self->_assert_pass_injection_guard($op);
727 if ($op =~ s/ [_\s]? \d+ $//x ) {
728 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
729 . "You probably wanted ...-and => [ -$op => COND1, -$op => COND2 ... ]";
731 if ($op =~ /^(?:not )?between$/) {
732 local our $Cur_Col_Meta = $k;
733 my @rhs = map $self->_expand_expr($_),
734 ref($vv) eq 'ARRAY' ? @$vv : $vv;
736 (@rhs == 1 and ref($rhs[0]) eq 'HASH' and $rhs[0]->{-literal})
738 (@rhs == 2 and defined($rhs[0]) and defined($rhs[1]))
740 puke "Operator '${\uc($op)}' requires either an arrayref with two defined values or expressions, or a single literal scalarref/arrayref-ref";
744 $self->_expand_ident(-ident => $k),
748 if ($op =~ /^(?:not )?in$/) {
749 if (my $literal = is_literal_value($vv)) {
750 my ($sql, @bind) = @$literal;
751 my $opened_sql = $self->_open_outer_paren($sql);
753 $op, $self->_expand_ident(-ident => $k),
754 [ { -literal => [ $opened_sql, @bind ] } ]
758 'SQL::Abstract before v1.75 used to generate incorrect SQL when the '
759 . "-${\uc($op)} operator was given an undef-containing list: !!!AUDIT YOUR CODE "
760 . 'AND DATA!!! (the upcoming Data::Query-based version of SQL::Abstract '
761 . 'will emit the logically correct SQL instead of raising this exception)'
763 puke("Argument passed to the '${\uc($op)}' operator can not be undefined")
765 my @rhs = map $self->_expand_expr($_),
766 map { ref($_) ? $_ : { -bind => [ $k, $_ ] } }
767 map { defined($_) ? $_: puke($undef_err) }
768 (ref($vv) eq 'ARRAY' ? @$vv : $vv);
769 return $self->${\($op =~ /^not/ ? 'sqltrue' : 'sqlfalse')} unless @rhs;
773 $self->_expand_ident(-ident => $k),
777 if ($op eq 'ident') {
778 if (! defined $vv or (ref($vv) and ref($vv) eq 'ARRAY')) {
779 puke "-$op requires a single plain scalar argument (a quotable identifier) or an arrayref of identifier parts";
783 $self->_expand_ident(-ident => $k),
784 $self->_expand_ident(-ident => $vv),
787 if ($op eq 'value') {
788 return $self->_expand_expr({ $k, undef }) unless defined($vv);
791 $self->_expand_ident(-ident => $k),
792 { -bind => [ $k, $vv ] }
795 if ($op =~ /^is(?: not)?$/) {
796 puke "$op can only take undef as argument"
800 and exists($vv->{-value})
801 and !defined($vv->{-value})
803 return +{ -op => [ $op.' null', $self->_expand_ident(-ident => $k) ] };
805 if ($op =~ /^(and|or)$/) {
806 if (ref($vv) eq 'HASH') {
809 map $self->_expand_expr({ $k, { $_ => $vv->{$_} } }),
814 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}) {
815 return { -op => [ $op, $self->_expand_ident(-ident => $k), $vv ] };
817 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
820 $self->_expand_ident(-ident => $k),
821 { -op => [ $op, $vv ] }
824 if (ref($vv) eq 'ARRAY') {
825 my ($logic, @values) = (
826 (defined($vv->[0]) and $vv->[0] =~ /^-(and|or)$/i)
831 $op =~ $self->{inequality_op}
832 or $op =~ $self->{not_like_op}
834 if (lc($logic) eq '-or' and @values > 1) {
835 belch "A multi-element arrayref as an argument to the inequality op '${\uc($op)}' "
836 . 'is technically equivalent to an always-true 1=1 (you probably wanted '
837 . "to say ...{ \$inequality_op => [ -and => \@values ] }... instead)"
842 # try to DWIM on equality operators
844 $op =~ $self->{equality_op} ? $self->sqlfalse
845 : $op =~ $self->{like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->sqlfalse
846 : $op =~ $self->{inequality_op} ? $self->sqltrue
847 : $op =~ $self->{not_like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->sqltrue
848 : puke "operator '$op' applied on an empty array (field '$k')";
852 map $self->_expand_expr({ $k => { $vk => $_ } }),
860 and exists $vv->{-value}
861 and not defined $vv->{-value}
865 $op =~ /^not$/i ? 'is not' # legacy
866 : $op =~ $self->{equality_op} ? 'is'
867 : $op =~ $self->{like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is'
868 : $op =~ $self->{inequality_op} ? 'is not'
869 : $op =~ $self->{not_like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is not'
870 : puke "unexpected operator '$op' with undef operand";
871 return +{ -op => [ $is.' null', $self->_expand_ident(-ident => $k) ] };
873 local our $Cur_Col_Meta = $k;
876 $self->_expand_ident(-ident => $k),
877 $self->_expand_expr($vv)
880 if (ref($v) eq 'ARRAY') {
881 return $self->sqlfalse unless @$v;
882 $self->_debug("ARRAY($k) means distribute over elements");
884 $v->[0] =~ /^-((?:and|or))$/i
885 ? ($v = [ @{$v}[1..$#$v] ], $1)
886 : ($self->{logic} || 'or')
890 map $self->_expand_expr({ $k => $_ }, $this_logic), @$v
893 if (my $literal = is_literal_value($v)) {
895 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
898 my ($sql, @bind) = @$literal;
899 if ($self->{bindtype} eq 'columns') {
901 $self->_assert_bindval_matches_bindtype($_);
904 return +{ -literal => [ $self->_quote($k).' '.$sql, @bind ] };
910 my ($self, undef, $body) = @_;
911 my @parts = map split(/\Q${\($self->{name_sep}||'.')}\E/, $_),
912 ref($body) ? @$body : $body;
913 return { -ident => $parts[-1] } if $self->{_dequalify_idents};
914 unless ($self->{quote_char}) {
915 $self->_assert_pass_injection_guard($_) for @parts;
917 return +{ -ident => \@parts };
921 my ($self, $where, $logic) = @_;
923 # Special case: top level simple string treated as literal
925 my $where_exp = (ref($where)
926 ? $self->_expand_expr($where, $logic)
927 : { -literal => [ $where ] });
929 # dispatch expanded expression
931 my ($sql, @bind) = defined($where_exp) ? $self->render_aqt($where_exp) : (undef);
932 # DBIx::Class used to call _recurse_where in scalar context
933 # something else might too...
935 return ($sql, @bind);
938 belch "Calling _recurse_where in scalar context is deprecated and will go away before 2.0";
944 my ($self, $ident) = @_;
946 return $self->_convert($self->_quote($ident));
949 my %unop_postfix = map +($_ => 1),
950 'is null', 'is not null',
958 my ($self, $args) = @_;
959 my ($left, $low, $high) = @$args;
960 my ($rhsql, @rhbind) = do {
962 puke "Single arg to between must be a literal"
963 unless $low->{-literal};
966 my ($l, $h) = map [ $self->render_aqt($_) ], $low, $high;
967 (join(' ', $l->[0], $self->_sqlcase('and'), $h->[0]),
968 @{$l}[1..$#$l], @{$h}[1..$#$h])
971 my ($lhsql, @lhbind) = $self->render_aqt($left);
973 join(' ', '(', $lhsql, $self->_sqlcase($op), $rhsql, ')'),
977 }), 'between', 'not between'),
981 my ($self, $args) = @_;
982 my ($lhs, $rhs) = @$args;
985 my ($sql, @bind) = $self->render_aqt($_);
986 push @in_bind, @bind;
989 my ($lhsql, @lbind) = $self->render_aqt($lhs);
991 $lhsql.' '.$self->_sqlcase($op).' ( '
1001 my ($self, $v) = @_;
1002 my ($op, @args) = @$v;
1003 $op =~ s/^-// if length($op) > 1;
1005 if (my $h = $special{$op}) {
1006 return $self->$h(\@args);
1008 my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}};
1009 if ($us and @args > 1) {
1010 puke "Special op '${op}' requires first value to be identifier"
1011 unless my ($ident) = map $_->{-ident}, grep ref($_) eq 'HASH', $args[0];
1012 my $k = join(($self->{name_sep}||'.'), @$ident);
1013 local our $Expand_Depth = 1;
1014 return $self->${\($us->{handler})}($k, $op, $args[1]);
1016 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1017 return $self->${\($us->{handler})}($op, $args[0]);
1019 my $final_op = $op =~ /^(?:is|not)_/ ? join(' ', split '_', $op) : $op;
1020 if (@args == 1 and $op !~ /^(and|or)$/) {
1021 my ($expr_sql, @bind) = $self->render_aqt($args[0]);
1022 my $op_sql = $self->_sqlcase($final_op);
1024 $unop_postfix{lc($final_op)}
1025 ? "${expr_sql} ${op_sql}"
1026 : "${op_sql} ${expr_sql}"
1028 return (($op eq 'not' || $us ? '('.$final_sql.')' : $final_sql), @bind);
1030 my @parts = grep length($_->[0]), map [ $self->render_aqt($_) ], @args;
1031 return '' unless @parts;
1032 my $is_andor = !!($op =~ /^(and|or)$/);
1033 return @{$parts[0]} if $is_andor and @parts == 1;
1034 my ($final_sql) = map +($is_andor ? "( ${_} )" : $_), join(
1035 ' '.$self->_sqlcase($final_op).' ',
1040 map @{$_}[1..$#$_], @parts
1047 my ($self, $list) = @_;
1048 my @parts = grep length($_->[0]), map [ $self->render_aqt($_) ], @$list;
1049 return join(', ', map $_->[0], @parts), map @{$_}[1..$#$_], @parts;
1053 my ($self, $rest) = @_;
1054 my ($func, @args) = @$rest;
1058 push @arg_sql, shift @x;
1060 } map [ $self->render_aqt($_) ], @args;
1061 return ($self->_sqlcase($func).'('.join(', ', @arg_sql).')', @bind);
1065 my ($self, $bind) = @_;
1066 return ($self->_convert('?'), $self->_bindtype(@$bind));
1069 sub _render_literal {
1070 my ($self, $literal) = @_;
1071 $self->_assert_bindval_matches_bindtype(@{$literal}[1..$#$literal]);
1075 # Some databases (SQLite) treat col IN (1, 2) different from
1076 # col IN ( (1, 2) ). Use this to strip all outer parens while
1077 # adding them back in the corresponding method
1078 sub _open_outer_paren {
1079 my ($self, $sql) = @_;
1081 while (my ($inner) = $sql =~ /^ \s* \( (.*) \) \s* $/xs) {
1083 # there are closing parens inside, need the heavy duty machinery
1084 # to reevaluate the extraction starting from $sql (full reevaluation)
1085 if ($inner =~ /\)/) {
1086 require Text::Balanced;
1088 my (undef, $remainder) = do {
1089 # idiotic design - writes to $@ but *DOES NOT* throw exceptions
1091 Text::Balanced::extract_bracketed($sql, '()', qr/\s*/);
1094 # the entire expression needs to be a balanced bracketed thing
1095 # (after an extract no remainder sans trailing space)
1096 last if defined $remainder and $remainder =~ /\S/;
1106 #======================================================================
1108 #======================================================================
1110 sub _expand_order_by {
1111 my ($self, $arg) = @_;
1113 return unless defined($arg) and not (ref($arg) eq 'ARRAY' and !@$arg);
1115 my $expander = sub {
1116 my ($self, $dir, $expr) = @_;
1117 my @to_expand = ref($expr) eq 'ARRAY' ? @$expr : $expr;
1118 foreach my $arg (@to_expand) {
1122 and grep /^-(asc|desc)$/, keys %$arg
1124 puke "ordering direction hash passed to order by must have exactly one key (-asc or -desc)";
1127 my @exp = map +(defined($dir) ? { -op => [ $dir => $_ ] } : $_),
1128 map $self->expand_expr($_, -ident),
1129 map ref($_) eq 'ARRAY' ? @$_ : $_, @to_expand;
1130 return (@exp > 1 ? { -list => \@exp } : $exp[0]);
1133 local @{$self->{expand_unary}}{qw(-asc -desc)} = (
1134 sub { shift->$expander(asc => @_) },
1135 sub { shift->$expander(desc => @_) },
1138 return $self->$expander(undef, $arg);
1142 my ($self, $arg) = @_;
1144 return '' unless defined(my $expanded = $self->_expand_order_by($arg));
1146 my ($sql, @bind) = $self->render_aqt($expanded);
1148 return '' unless length($sql);
1150 my $final_sql = $self->_sqlcase(' order by ').$sql;
1152 return wantarray ? ($final_sql, @bind) : $final_sql;
1155 # _order_by no longer needs to call this so doesn't but DBIC uses it.
1157 sub _order_by_chunks {
1158 my ($self, $arg) = @_;
1160 return () unless defined(my $expanded = $self->_expand_order_by($arg));
1162 return $self->_chunkify_order_by($expanded);
1165 sub _chunkify_order_by {
1166 my ($self, $expanded) = @_;
1168 return grep length, $self->render_aqt($expanded)
1169 if $expanded->{-ident} or @{$expanded->{-literal}||[]} == 1;
1172 if (ref() eq 'HASH' and my $l = $_->{-list}) {
1173 return map $self->_chunkify_order_by($_), @$l;
1175 return [ $self->render_aqt($_) ];
1179 #======================================================================
1180 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
1181 #======================================================================
1187 $self->_expand_maybe_list_expr($from, undef, -ident)
1192 #======================================================================
1194 #======================================================================
1196 sub _expand_maybe_list_expr {
1197 my ($self, $expr, $logic, $default) = @_;
1199 if (ref($expr) eq 'ARRAY') {
1201 map $self->expand_expr($_, $default), @$expr
1208 return $self->expand_expr($e, $default);
1211 # highly optimized, as it's called way too often
1213 # my ($self, $label) = @_;
1215 return '' unless defined $_[1];
1216 return ${$_[1]} if ref($_[1]) eq 'SCALAR';
1217 puke 'Identifier cannot be hashref' if ref($_[1]) eq 'HASH';
1219 unless ($_[0]->{quote_char}) {
1220 if (ref($_[1]) eq 'ARRAY') {
1221 return join($_[0]->{name_sep}||'.', @{$_[1]});
1223 $_[0]->_assert_pass_injection_guard($_[1]);
1228 my $qref = ref $_[0]->{quote_char};
1230 !$qref ? ($_[0]->{quote_char}, $_[0]->{quote_char})
1231 : ($qref eq 'ARRAY') ? @{$_[0]->{quote_char}}
1232 : puke "Unsupported quote_char format: $_[0]->{quote_char}";
1234 my $esc = $_[0]->{escape_char} || $r;
1236 # parts containing * are naturally unquoted
1238 $_[0]->{name_sep}||'',
1242 : do { (my $n = $_) =~ s/(\Q$esc\E|\Q$r\E)/$esc$1/g; $l . $n . $r }
1244 (ref($_[1]) eq 'ARRAY'
1248 ? split (/\Q$_[0]->{name_sep}\E/, $_[1] )
1256 # Conversion, if applicable
1258 #my ($self, $arg) = @_;
1259 if ($_[0]->{convert_where}) {
1260 return $_[0]->_sqlcase($_[0]->{convert_where}) .'(' . $_[1] . ')';
1267 #my ($self, $col, @vals) = @_;
1268 # called often - tighten code
1269 return $_[0]->{bindtype} eq 'columns'
1270 ? map {[$_[1], $_]} @_[2 .. $#_]
1275 # Dies if any element of @bind is not in [colname => value] format
1276 # if bindtype is 'columns'.
1277 sub _assert_bindval_matches_bindtype {
1278 # my ($self, @bind) = @_;
1280 if ($self->{bindtype} eq 'columns') {
1282 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
1283 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
1289 sub _join_sql_clauses {
1290 my ($self, $logic, $clauses_aref, $bind_aref) = @_;
1292 if (@$clauses_aref > 1) {
1293 my $join = " " . $self->_sqlcase($logic) . " ";
1294 my $sql = '( ' . join($join, @$clauses_aref) . ' )';
1295 return ($sql, @$bind_aref);
1297 elsif (@$clauses_aref) {
1298 return ($clauses_aref->[0], @$bind_aref); # no parentheses
1301 return (); # if no SQL, ignore @$bind_aref
1306 # Fix SQL case, if so requested
1308 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
1309 # don't touch the argument ... crooked logic, but let's not change it!
1310 return $_[0]->{case} ? $_[1] : uc($_[1]);
1314 #======================================================================
1315 # DISPATCHING FROM REFKIND
1316 #======================================================================
1319 my ($self, $data) = @_;
1321 return 'UNDEF' unless defined $data;
1323 # blessed objects are treated like scalars
1324 my $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1326 return 'SCALAR' unless $ref;
1329 while ($ref eq 'REF') {
1331 $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1335 return ($ref||'SCALAR') . ('REF' x $n_steps);
1339 my ($self, $data) = @_;
1340 my @try = ($self->_refkind($data));
1341 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
1342 push @try, 'FALLBACK';
1346 sub _METHOD_FOR_refkind {
1347 my ($self, $meth_prefix, $data) = @_;
1350 for (@{$self->_try_refkind($data)}) {
1351 $method = $self->can($meth_prefix."_".$_)
1355 return $method || puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
1359 sub _SWITCH_refkind {
1360 my ($self, $data, $dispatch_table) = @_;
1363 for (@{$self->_try_refkind($data)}) {
1364 $coderef = $dispatch_table->{$_}
1368 puke "no dispatch entry for ".$self->_refkind($data)
1377 #======================================================================
1378 # VALUES, GENERATE, AUTOLOAD
1379 #======================================================================
1381 # LDNOTE: original code from nwiger, didn't touch code in that section
1382 # I feel the AUTOLOAD stuff should not be the default, it should
1383 # only be activated on explicit demand by user.
1387 my $data = shift || return;
1388 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
1389 unless ref $data eq 'HASH';
1392 foreach my $k (sort keys %$data) {
1393 my $v = $data->{$k};
1394 $self->_SWITCH_refkind($v, {
1396 if ($self->{array_datatypes}) { # array datatype
1397 push @all_bind, $self->_bindtype($k, $v);
1399 else { # literal SQL with bind
1400 my ($sql, @bind) = @$v;
1401 $self->_assert_bindval_matches_bindtype(@bind);
1402 push @all_bind, @bind;
1405 ARRAYREFREF => sub { # literal SQL with bind
1406 my ($sql, @bind) = @${$v};
1407 $self->_assert_bindval_matches_bindtype(@bind);
1408 push @all_bind, @bind;
1410 SCALARREF => sub { # literal SQL without bind
1412 SCALAR_or_UNDEF => sub {
1413 push @all_bind, $self->_bindtype($k, $v);
1424 my(@sql, @sqlq, @sqlv);
1428 if ($ref eq 'HASH') {
1429 for my $k (sort keys %$_) {
1432 my $label = $self->_quote($k);
1433 if ($r eq 'ARRAY') {
1434 # literal SQL with bind
1435 my ($sql, @bind) = @$v;
1436 $self->_assert_bindval_matches_bindtype(@bind);
1437 push @sqlq, "$label = $sql";
1439 } elsif ($r eq 'SCALAR') {
1440 # literal SQL without bind
1441 push @sqlq, "$label = $$v";
1443 push @sqlq, "$label = ?";
1444 push @sqlv, $self->_bindtype($k, $v);
1447 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
1448 } elsif ($ref eq 'ARRAY') {
1449 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
1452 if ($r eq 'ARRAY') { # literal SQL with bind
1453 my ($sql, @bind) = @$v;
1454 $self->_assert_bindval_matches_bindtype(@bind);
1457 } elsif ($r eq 'SCALAR') { # literal SQL without bind
1458 # embedded literal SQL
1465 push @sql, '(' . join(', ', @sqlq) . ')';
1466 } elsif ($ref eq 'SCALAR') {
1470 # strings get case twiddled
1471 push @sql, $self->_sqlcase($_);
1475 my $sql = join ' ', @sql;
1477 # this is pretty tricky
1478 # if ask for an array, return ($stmt, @bind)
1479 # otherwise, s/?/shift @sqlv/ to put it inline
1481 return ($sql, @sqlv);
1483 1 while $sql =~ s/\?/my $d = shift(@sqlv);
1484 ref $d ? $d->[1] : $d/e;
1493 # This allows us to check for a local, then _form, attr
1495 my($name) = $AUTOLOAD =~ /.*::(.+)/;
1496 return $self->generate($name, @_);
1507 SQL::Abstract - Generate SQL from Perl data structures
1513 my $sql = SQL::Abstract->new;
1515 my($stmt, @bind) = $sql->select($source, \@fields, \%where, $order);
1517 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
1519 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
1521 my($stmt, @bind) = $sql->delete($table, \%where);
1523 # Then, use these in your DBI statements
1524 my $sth = $dbh->prepare($stmt);
1525 $sth->execute(@bind);
1527 # Just generate the WHERE clause
1528 my($stmt, @bind) = $sql->where(\%where, $order);
1530 # Return values in the same order, for hashed queries
1531 # See PERFORMANCE section for more details
1532 my @bind = $sql->values(\%fieldvals);
1536 This module was inspired by the excellent L<DBIx::Abstract>.
1537 However, in using that module I found that what I really wanted
1538 to do was generate SQL, but still retain complete control over my
1539 statement handles and use the DBI interface. So, I set out to
1540 create an abstract SQL generation module.
1542 While based on the concepts used by L<DBIx::Abstract>, there are
1543 several important differences, especially when it comes to WHERE
1544 clauses. I have modified the concepts used to make the SQL easier
1545 to generate from Perl data structures and, IMO, more intuitive.
1546 The underlying idea is for this module to do what you mean, based
1547 on the data structures you provide it. The big advantage is that
1548 you don't have to modify your code every time your data changes,
1549 as this module figures it out.
1551 To begin with, an SQL INSERT is as easy as just specifying a hash
1552 of C<key=value> pairs:
1555 name => 'Jimbo Bobson',
1556 phone => '123-456-7890',
1557 address => '42 Sister Lane',
1558 city => 'St. Louis',
1559 state => 'Louisiana',
1562 The SQL can then be generated with this:
1564 my($stmt, @bind) = $sql->insert('people', \%data);
1566 Which would give you something like this:
1568 $stmt = "INSERT INTO people
1569 (address, city, name, phone, state)
1570 VALUES (?, ?, ?, ?, ?)";
1571 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
1572 '123-456-7890', 'Louisiana');
1574 These are then used directly in your DBI code:
1576 my $sth = $dbh->prepare($stmt);
1577 $sth->execute(@bind);
1579 =head2 Inserting and Updating Arrays
1581 If your database has array types (like for example Postgres),
1582 activate the special option C<< array_datatypes => 1 >>
1583 when creating the C<SQL::Abstract> object.
1584 Then you may use an arrayref to insert and update database array types:
1586 my $sql = SQL::Abstract->new(array_datatypes => 1);
1588 planets => [qw/Mercury Venus Earth Mars/]
1591 my($stmt, @bind) = $sql->insert('solar_system', \%data);
1595 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
1597 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
1600 =head2 Inserting and Updating SQL
1602 In order to apply SQL functions to elements of your C<%data> you may
1603 specify a reference to an arrayref for the given hash value. For example,
1604 if you need to execute the Oracle C<to_date> function on a value, you can
1605 say something like this:
1609 date_entered => \[ "to_date(?,'MM/DD/YYYY')", "03/02/2003" ],
1612 The first value in the array is the actual SQL. Any other values are
1613 optional and would be included in the bind values array. This gives
1616 my($stmt, @bind) = $sql->insert('people', \%data);
1618 $stmt = "INSERT INTO people (name, date_entered)
1619 VALUES (?, to_date(?,'MM/DD/YYYY'))";
1620 @bind = ('Bill', '03/02/2003');
1622 An UPDATE is just as easy, all you change is the name of the function:
1624 my($stmt, @bind) = $sql->update('people', \%data);
1626 Notice that your C<%data> isn't touched; the module will generate
1627 the appropriately quirky SQL for you automatically. Usually you'll
1628 want to specify a WHERE clause for your UPDATE, though, which is
1629 where handling C<%where> hashes comes in handy...
1631 =head2 Complex where statements
1633 This module can generate pretty complicated WHERE statements
1634 easily. For example, simple C<key=value> pairs are taken to mean
1635 equality, and if you want to see if a field is within a set
1636 of values, you can use an arrayref. Let's say we wanted to
1637 SELECT some data based on this criteria:
1640 requestor => 'inna',
1641 worker => ['nwiger', 'rcwe', 'sfz'],
1642 status => { '!=', 'completed' }
1645 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
1647 The above would give you something like this:
1649 $stmt = "SELECT * FROM tickets WHERE
1650 ( requestor = ? ) AND ( status != ? )
1651 AND ( worker = ? OR worker = ? OR worker = ? )";
1652 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
1654 Which you could then use in DBI code like so:
1656 my $sth = $dbh->prepare($stmt);
1657 $sth->execute(@bind);
1663 The methods are simple. There's one for every major SQL operation,
1664 and a constructor you use first. The arguments are specified in a
1665 similar order for each method (table, then fields, then a where
1666 clause) to try and simplify things.
1668 =head2 new(option => 'value')
1670 The C<new()> function takes a list of options and values, and returns
1671 a new B<SQL::Abstract> object which can then be used to generate SQL
1672 through the methods below. The options accepted are:
1678 If set to 'lower', then SQL will be generated in all lowercase. By
1679 default SQL is generated in "textbook" case meaning something like:
1681 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
1683 Any setting other than 'lower' is ignored.
1687 This determines what the default comparison operator is. By default
1688 it is C<=>, meaning that a hash like this:
1690 %where = (name => 'nwiger', email => 'nate@wiger.org');
1692 Will generate SQL like this:
1694 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
1696 However, you may want loose comparisons by default, so if you set
1697 C<cmp> to C<like> you would get SQL such as:
1699 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
1701 You can also override the comparison on an individual basis - see
1702 the huge section on L</"WHERE CLAUSES"> at the bottom.
1704 =item sqltrue, sqlfalse
1706 Expressions for inserting boolean values within SQL statements.
1707 By default these are C<1=1> and C<1=0>. They are used
1708 by the special operators C<-in> and C<-not_in> for generating
1709 correct SQL even when the argument is an empty array (see below).
1713 This determines the default logical operator for multiple WHERE
1714 statements in arrays or hashes. If absent, the default logic is "or"
1715 for arrays, and "and" for hashes. This means that a WHERE
1719 event_date => {'>=', '2/13/99'},
1720 event_date => {'<=', '4/24/03'},
1723 will generate SQL like this:
1725 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
1727 This is probably not what you want given this query, though (look
1728 at the dates). To change the "OR" to an "AND", simply specify:
1730 my $sql = SQL::Abstract->new(logic => 'and');
1732 Which will change the above C<WHERE> to:
1734 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
1736 The logic can also be changed locally by inserting
1737 a modifier in front of an arrayref:
1739 @where = (-and => [event_date => {'>=', '2/13/99'},
1740 event_date => {'<=', '4/24/03'} ]);
1742 See the L</"WHERE CLAUSES"> section for explanations.
1746 This will automatically convert comparisons using the specified SQL
1747 function for both column and value. This is mostly used with an argument
1748 of C<upper> or C<lower>, so that the SQL will have the effect of
1749 case-insensitive "searches". For example, this:
1751 $sql = SQL::Abstract->new(convert => 'upper');
1752 %where = (keywords => 'MaKe iT CAse inSeNSItive');
1754 Will turn out the following SQL:
1756 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
1758 The conversion can be C<upper()>, C<lower()>, or any other SQL function
1759 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
1760 not validate this option; it will just pass through what you specify verbatim).
1764 This is a kludge because many databases suck. For example, you can't
1765 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
1766 Instead, you have to use C<bind_param()>:
1768 $sth->bind_param(1, 'reg data');
1769 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
1771 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
1772 which loses track of which field each slot refers to. Fear not.
1774 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
1775 Currently, you can specify either C<normal> (default) or C<columns>. If you
1776 specify C<columns>, you will get an array that looks like this:
1778 my $sql = SQL::Abstract->new(bindtype => 'columns');
1779 my($stmt, @bind) = $sql->insert(...);
1782 [ 'column1', 'value1' ],
1783 [ 'column2', 'value2' ],
1784 [ 'column3', 'value3' ],
1787 You can then iterate through this manually, using DBI's C<bind_param()>.
1789 $sth->prepare($stmt);
1792 my($col, $data) = @$_;
1793 if ($col eq 'details' || $col eq 'comments') {
1794 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
1795 } elsif ($col eq 'image') {
1796 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
1798 $sth->bind_param($i, $data);
1802 $sth->execute; # execute without @bind now
1804 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
1805 Basically, the advantage is still that you don't have to care which fields
1806 are or are not included. You could wrap that above C<for> loop in a simple
1807 sub called C<bind_fields()> or something and reuse it repeatedly. You still
1808 get a layer of abstraction over manual SQL specification.
1810 Note that if you set L</bindtype> to C<columns>, the C<\[ $sql, @bind ]>
1811 construct (see L</Literal SQL with placeholders and bind values (subqueries)>)
1812 will expect the bind values in this format.
1816 This is the character that a table or column name will be quoted
1817 with. By default this is an empty string, but you could set it to
1818 the character C<`>, to generate SQL like this:
1820 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
1822 Alternatively, you can supply an array ref of two items, the first being the left
1823 hand quote character, and the second the right hand quote character. For
1824 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
1825 that generates SQL like this:
1827 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
1829 Quoting is useful if you have tables or columns names that are reserved
1830 words in your database's SQL dialect.
1834 This is the character that will be used to escape L</quote_char>s appearing
1835 in an identifier before it has been quoted.
1837 The parameter default in case of a single L</quote_char> character is the quote
1840 When opening-closing-style quoting is used (L</quote_char> is an arrayref)
1841 this parameter defaults to the B<closing (right)> L</quote_char>. Occurrences
1842 of the B<opening (left)> L</quote_char> within the identifier are currently left
1843 untouched. The default for opening-closing-style quotes may change in future
1844 versions, thus you are B<strongly encouraged> to specify the escape character
1849 This is the character that separates a table and column name. It is
1850 necessary to specify this when the C<quote_char> option is selected,
1851 so that tables and column names can be individually quoted like this:
1853 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
1855 =item injection_guard
1857 A regular expression C<qr/.../> that is applied to any C<-function> and unquoted
1858 column name specified in a query structure. This is a safety mechanism to avoid
1859 injection attacks when mishandling user input e.g.:
1861 my %condition_as_column_value_pairs = get_values_from_user();
1862 $sqla->select( ... , \%condition_as_column_value_pairs );
1864 If the expression matches an exception is thrown. Note that literal SQL
1865 supplied via C<\'...'> or C<\['...']> is B<not> checked in any way.
1867 Defaults to checking for C<;> and the C<GO> keyword (TransactSQL)
1869 =item array_datatypes
1871 When this option is true, arrayrefs in INSERT or UPDATE are
1872 interpreted as array datatypes and are passed directly
1874 When this option is false, arrayrefs are interpreted
1875 as literal SQL, just like refs to arrayrefs
1876 (but this behavior is for backwards compatibility; when writing
1877 new queries, use the "reference to arrayref" syntax
1883 Takes a reference to a list of "special operators"
1884 to extend the syntax understood by L<SQL::Abstract>.
1885 See section L</"SPECIAL OPERATORS"> for details.
1889 Takes a reference to a list of "unary operators"
1890 to extend the syntax understood by L<SQL::Abstract>.
1891 See section L</"UNARY OPERATORS"> for details.
1897 =head2 insert($table, \@values || \%fieldvals, \%options)
1899 This is the simplest function. You simply give it a table name
1900 and either an arrayref of values or hashref of field/value pairs.
1901 It returns an SQL INSERT statement and a list of bind values.
1902 See the sections on L</"Inserting and Updating Arrays"> and
1903 L</"Inserting and Updating SQL"> for information on how to insert
1904 with those data types.
1906 The optional C<\%options> hash reference may contain additional
1907 options to generate the insert SQL. Currently supported options
1914 Takes either a scalar of raw SQL fields, or an array reference of
1915 field names, and adds on an SQL C<RETURNING> statement at the end.
1916 This allows you to return data generated by the insert statement
1917 (such as row IDs) without performing another C<SELECT> statement.
1918 Note, however, this is not part of the SQL standard and may not
1919 be supported by all database engines.
1923 =head2 update($table, \%fieldvals, \%where, \%options)
1925 This takes a table, hashref of field/value pairs, and an optional
1926 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
1928 See the sections on L</"Inserting and Updating Arrays"> and
1929 L</"Inserting and Updating SQL"> for information on how to insert
1930 with those data types.
1932 The optional C<\%options> hash reference may contain additional
1933 options to generate the update SQL. Currently supported options
1940 See the C<returning> option to
1941 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
1945 =head2 select($source, $fields, $where, $order)
1947 This returns a SQL SELECT statement and associated list of bind values, as
1948 specified by the arguments:
1954 Specification of the 'FROM' part of the statement.
1955 The argument can be either a plain scalar (interpreted as a table
1956 name, will be quoted), or an arrayref (interpreted as a list
1957 of table names, joined by commas, quoted), or a scalarref
1958 (literal SQL, not quoted).
1962 Specification of the list of fields to retrieve from
1964 The argument can be either an arrayref (interpreted as a list
1965 of field names, will be joined by commas and quoted), or a
1966 plain scalar (literal SQL, not quoted).
1967 Please observe that this API is not as flexible as that of
1968 the first argument C<$source>, for backwards compatibility reasons.
1972 Optional argument to specify the WHERE part of the query.
1973 The argument is most often a hashref, but can also be
1974 an arrayref or plain scalar --
1975 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
1979 Optional argument to specify the ORDER BY part of the query.
1980 The argument can be a scalar, a hashref or an arrayref
1981 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
1987 =head2 delete($table, \%where, \%options)
1989 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
1990 It returns an SQL DELETE statement and list of bind values.
1992 The optional C<\%options> hash reference may contain additional
1993 options to generate the delete SQL. Currently supported options
2000 See the C<returning> option to
2001 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2005 =head2 where(\%where, $order)
2007 This is used to generate just the WHERE clause. For example,
2008 if you have an arbitrary data structure and know what the
2009 rest of your SQL is going to look like, but want an easy way
2010 to produce a WHERE clause, use this. It returns an SQL WHERE
2011 clause and list of bind values.
2014 =head2 values(\%data)
2016 This just returns the values from the hash C<%data>, in the same
2017 order that would be returned from any of the other above queries.
2018 Using this allows you to markedly speed up your queries if you
2019 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
2021 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
2023 Warning: This is an experimental method and subject to change.
2025 This returns arbitrarily generated SQL. It's a really basic shortcut.
2026 It will return two different things, depending on return context:
2028 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
2029 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
2031 These would return the following:
2033 # First calling form
2034 $stmt = "CREATE TABLE test (?, ?)";
2035 @bind = (field1, field2);
2037 # Second calling form
2038 $stmt_and_val = "CREATE TABLE test (field1, field2)";
2040 Depending on what you're trying to do, it's up to you to choose the correct
2041 format. In this example, the second form is what you would want.
2045 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
2049 ALTER SESSION SET nls_date_format = 'MM/YY'
2051 You get the idea. Strings get their case twiddled, but everything
2052 else remains verbatim.
2054 =head1 EXPORTABLE FUNCTIONS
2056 =head2 is_plain_value
2058 Determines if the supplied argument is a plain value as understood by this
2063 =item * The value is C<undef>
2065 =item * The value is a non-reference
2067 =item * The value is an object with stringification overloading
2069 =item * The value is of the form C<< { -value => $anything } >>
2073 On failure returns C<undef>, on success returns a B<scalar> reference
2074 to the original supplied argument.
2080 The stringification overloading detection is rather advanced: it takes
2081 into consideration not only the presence of a C<""> overload, but if that
2082 fails also checks for enabled
2083 L<autogenerated versions of C<"">|overload/Magic Autogeneration>, based
2084 on either C<0+> or C<bool>.
2086 Unfortunately testing in the field indicates that this
2087 detection B<< may tickle a latent bug in perl versions before 5.018 >>,
2088 but only when very large numbers of stringifying objects are involved.
2089 At the time of writing ( Sep 2014 ) there is no clear explanation of
2090 the direct cause, nor is there a manageably small test case that reliably
2091 reproduces the problem.
2093 If you encounter any of the following exceptions in B<random places within
2094 your application stack> - this module may be to blame:
2096 Operation "ne": no method found,
2097 left argument in overloaded package <something>,
2098 right argument in overloaded package <something>
2102 Stub found while resolving method "???" overloading """" in package <something>
2104 If you fall victim to the above - please attempt to reduce the problem
2105 to something that could be sent to the L<SQL::Abstract developers
2106 |DBIx::Class/GETTING HELP/SUPPORT>
2107 (either publicly or privately). As a workaround in the meantime you can
2108 set C<$ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}> to a true
2109 value, which will most likely eliminate your problem (at the expense of
2110 not being able to properly detect exotic forms of stringification).
2112 This notice and environment variable will be removed in a future version,
2113 as soon as the underlying problem is found and a reliable workaround is
2118 =head2 is_literal_value
2120 Determines if the supplied argument is a literal value as understood by this
2125 =item * C<\$sql_string>
2127 =item * C<\[ $sql_string, @bind_values ]>
2131 On failure returns C<undef>, on success returns an B<array> reference
2132 containing the unpacked version of the supplied literal SQL and bind values.
2134 =head1 WHERE CLAUSES
2138 This module uses a variation on the idea from L<DBIx::Abstract>. It
2139 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
2140 module is that things in arrays are OR'ed, and things in hashes
2143 The easiest way to explain is to show lots of examples. After
2144 each C<%where> hash shown, it is assumed you used:
2146 my($stmt, @bind) = $sql->where(\%where);
2148 However, note that the C<%where> hash can be used directly in any
2149 of the other functions as well, as described above.
2151 =head2 Key-value pairs
2153 So, let's get started. To begin, a simple hash:
2157 status => 'completed'
2160 Is converted to SQL C<key = val> statements:
2162 $stmt = "WHERE user = ? AND status = ?";
2163 @bind = ('nwiger', 'completed');
2165 One common thing I end up doing is having a list of values that
2166 a field can be in. To do this, simply specify a list inside of
2171 status => ['assigned', 'in-progress', 'pending'];
2174 This simple code will create the following:
2176 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
2177 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
2179 A field associated to an empty arrayref will be considered a
2180 logical false and will generate 0=1.
2182 =head2 Tests for NULL values
2184 If the value part is C<undef> then this is converted to SQL <IS NULL>
2193 $stmt = "WHERE user = ? AND status IS NULL";
2196 To test if a column IS NOT NULL:
2200 status => { '!=', undef },
2203 =head2 Specific comparison operators
2205 If you want to specify a different type of operator for your comparison,
2206 you can use a hashref for a given column:
2210 status => { '!=', 'completed' }
2213 Which would generate:
2215 $stmt = "WHERE user = ? AND status != ?";
2216 @bind = ('nwiger', 'completed');
2218 To test against multiple values, just enclose the values in an arrayref:
2220 status => { '=', ['assigned', 'in-progress', 'pending'] };
2222 Which would give you:
2224 "WHERE status = ? OR status = ? OR status = ?"
2227 The hashref can also contain multiple pairs, in which case it is expanded
2228 into an C<AND> of its elements:
2232 status => { '!=', 'completed', -not_like => 'pending%' }
2235 # Or more dynamically, like from a form
2236 $where{user} = 'nwiger';
2237 $where{status}{'!='} = 'completed';
2238 $where{status}{'-not_like'} = 'pending%';
2240 # Both generate this
2241 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
2242 @bind = ('nwiger', 'completed', 'pending%');
2245 To get an OR instead, you can combine it with the arrayref idea:
2249 priority => [ { '=', 2 }, { '>', 5 } ]
2252 Which would generate:
2254 $stmt = "WHERE ( priority = ? OR priority > ? ) AND user = ?";
2255 @bind = ('2', '5', 'nwiger');
2257 If you want to include literal SQL (with or without bind values), just use a
2258 scalar reference or reference to an arrayref as the value:
2261 date_entered => { '>' => \["to_date(?, 'MM/DD/YYYY')", "11/26/2008"] },
2262 date_expires => { '<' => \"now()" }
2265 Which would generate:
2267 $stmt = "WHERE date_entered > to_date(?, 'MM/DD/YYYY') AND date_expires < now()";
2268 @bind = ('11/26/2008');
2271 =head2 Logic and nesting operators
2273 In the example above,
2274 there is a subtle trap if you want to say something like
2275 this (notice the C<AND>):
2277 WHERE priority != ? AND priority != ?
2279 Because, in Perl you I<can't> do this:
2281 priority => { '!=' => 2, '!=' => 1 }
2283 As the second C<!=> key will obliterate the first. The solution
2284 is to use the special C<-modifier> form inside an arrayref:
2286 priority => [ -and => {'!=', 2},
2290 Normally, these would be joined by C<OR>, but the modifier tells it
2291 to use C<AND> instead. (Hint: You can use this in conjunction with the
2292 C<logic> option to C<new()> in order to change the way your queries
2293 work by default.) B<Important:> Note that the C<-modifier> goes
2294 B<INSIDE> the arrayref, as an extra first element. This will
2295 B<NOT> do what you think it might:
2297 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
2299 Here is a quick list of equivalencies, since there is some overlap:
2302 status => {'!=', 'completed', 'not like', 'pending%' }
2303 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
2306 status => {'=', ['assigned', 'in-progress']}
2307 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
2308 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
2312 =head2 Special operators: IN, BETWEEN, etc.
2314 You can also use the hashref format to compare a list of fields using the
2315 C<IN> comparison operator, by specifying the list as an arrayref:
2318 status => 'completed',
2319 reportid => { -in => [567, 2335, 2] }
2322 Which would generate:
2324 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
2325 @bind = ('completed', '567', '2335', '2');
2327 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
2330 If the argument to C<-in> is an empty array, 'sqlfalse' is generated
2331 (by default: C<1=0>). Similarly, C<< -not_in => [] >> generates
2332 'sqltrue' (by default: C<1=1>).
2334 In addition to the array you can supply a chunk of literal sql or
2335 literal sql with bind:
2338 customer => { -in => \[
2339 'SELECT cust_id FROM cust WHERE balance > ?',
2342 status => { -in => \'SELECT status_codes FROM states' },
2348 customer IN ( SELECT cust_id FROM cust WHERE balance > ? )
2349 AND status IN ( SELECT status_codes FROM states )
2353 Finally, if the argument to C<-in> is not a reference, it will be
2354 treated as a single-element array.
2356 Another pair of operators is C<-between> and C<-not_between>,
2357 used with an arrayref of two values:
2361 completion_date => {
2362 -not_between => ['2002-10-01', '2003-02-06']
2368 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
2370 Just like with C<-in> all plausible combinations of literal SQL
2374 start0 => { -between => [ 1, 2 ] },
2375 start1 => { -between => \["? AND ?", 1, 2] },
2376 start2 => { -between => \"lower(x) AND upper(y)" },
2377 start3 => { -between => [
2379 \["upper(?)", 'stuff' ],
2386 ( start0 BETWEEN ? AND ? )
2387 AND ( start1 BETWEEN ? AND ? )
2388 AND ( start2 BETWEEN lower(x) AND upper(y) )
2389 AND ( start3 BETWEEN lower(x) AND upper(?) )
2391 @bind = (1, 2, 1, 2, 'stuff');
2394 These are the two builtin "special operators"; but the
2395 list can be expanded: see section L</"SPECIAL OPERATORS"> below.
2397 =head2 Unary operators: bool
2399 If you wish to test against boolean columns or functions within your
2400 database you can use the C<-bool> and C<-not_bool> operators. For
2401 example to test the column C<is_user> being true and the column
2402 C<is_enabled> being false you would use:-
2406 -not_bool => 'is_enabled',
2411 WHERE is_user AND NOT is_enabled
2413 If a more complex combination is required, testing more conditions,
2414 then you should use the and/or operators:-
2419 -not_bool => { two=> { -rlike => 'bar' } },
2420 -not_bool => { three => [ { '=', 2 }, { '>', 5 } ] },
2431 (NOT ( three = ? OR three > ? ))
2434 =head2 Nested conditions, -and/-or prefixes
2436 So far, we've seen how multiple conditions are joined with a top-level
2437 C<AND>. We can change this by putting the different conditions we want in
2438 hashes and then putting those hashes in an array. For example:
2443 status => { -like => ['pending%', 'dispatched'] },
2447 status => 'unassigned',
2451 This data structure would create the following:
2453 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
2454 OR ( user = ? AND status = ? ) )";
2455 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
2458 Clauses in hashrefs or arrayrefs can be prefixed with an C<-and> or C<-or>
2459 to change the logic inside:
2465 -and => [ workhrs => {'>', 20}, geo => 'ASIA' ],
2466 -or => { workhrs => {'<', 50}, geo => 'EURO' },
2473 $stmt = "WHERE ( user = ?
2474 AND ( ( workhrs > ? AND geo = ? )
2475 OR ( workhrs < ? OR geo = ? ) ) )";
2476 @bind = ('nwiger', '20', 'ASIA', '50', 'EURO');
2478 =head3 Algebraic inconsistency, for historical reasons
2480 C<Important note>: when connecting several conditions, the C<-and->|C<-or>
2481 operator goes C<outside> of the nested structure; whereas when connecting
2482 several constraints on one column, the C<-and> operator goes
2483 C<inside> the arrayref. Here is an example combining both features:
2486 -and => [a => 1, b => 2],
2487 -or => [c => 3, d => 4],
2488 e => [-and => {-like => 'foo%'}, {-like => '%bar'} ]
2493 WHERE ( ( ( a = ? AND b = ? )
2494 OR ( c = ? OR d = ? )
2495 OR ( e LIKE ? AND e LIKE ? ) ) )
2497 This difference in syntax is unfortunate but must be preserved for
2498 historical reasons. So be careful: the two examples below would
2499 seem algebraically equivalent, but they are not
2502 { -like => 'foo%' },
2503 { -like => '%bar' },
2505 # yields: WHERE ( ( col LIKE ? AND col LIKE ? ) )
2508 { col => { -like => 'foo%' } },
2509 { col => { -like => '%bar' } },
2511 # yields: WHERE ( ( col LIKE ? OR col LIKE ? ) )
2514 =head2 Literal SQL and value type operators
2516 The basic premise of SQL::Abstract is that in WHERE specifications the "left
2517 side" is a column name and the "right side" is a value (normally rendered as
2518 a placeholder). This holds true for both hashrefs and arrayref pairs as you
2519 see in the L</WHERE CLAUSES> examples above. Sometimes it is necessary to
2520 alter this behavior. There are several ways of doing so.
2524 This is a virtual operator that signals the string to its right side is an
2525 identifier (a column name) and not a value. For example to compare two
2526 columns you would write:
2529 priority => { '<', 2 },
2530 requestor => { -ident => 'submitter' },
2535 $stmt = "WHERE priority < ? AND requestor = submitter";
2538 If you are maintaining legacy code you may see a different construct as
2539 described in L</Deprecated usage of Literal SQL>, please use C<-ident> in new
2544 This is a virtual operator that signals that the construct to its right side
2545 is a value to be passed to DBI. This is for example necessary when you want
2546 to write a where clause against an array (for RDBMS that support such
2547 datatypes). For example:
2550 array => { -value => [1, 2, 3] }
2555 $stmt = 'WHERE array = ?';
2556 @bind = ([1, 2, 3]);
2558 Note that if you were to simply say:
2564 the result would probably not be what you wanted:
2566 $stmt = 'WHERE array = ? OR array = ? OR array = ?';
2571 Finally, sometimes only literal SQL will do. To include a random snippet
2572 of SQL verbatim, you specify it as a scalar reference. Consider this only
2573 as a last resort. Usually there is a better way. For example:
2576 priority => { '<', 2 },
2577 requestor => { -in => \'(SELECT name FROM hitmen)' },
2582 $stmt = "WHERE priority < ? AND requestor IN (SELECT name FROM hitmen)"
2585 Note that in this example, you only get one bind parameter back, since
2586 the verbatim SQL is passed as part of the statement.
2590 Never use untrusted input as a literal SQL argument - this is a massive
2591 security risk (there is no way to check literal snippets for SQL
2592 injections and other nastyness). If you need to deal with untrusted input
2593 use literal SQL with placeholders as described next.
2595 =head3 Literal SQL with placeholders and bind values (subqueries)
2597 If the literal SQL to be inserted has placeholders and bind values,
2598 use a reference to an arrayref (yes this is a double reference --
2599 not so common, but perfectly legal Perl). For example, to find a date
2600 in Postgres you can use something like this:
2603 date_column => \[ "= date '2008-09-30' - ?::integer", 10 ]
2608 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
2611 Note that you must pass the bind values in the same format as they are returned
2612 by L<where|/where(\%where, $order)>. This means that if you set L</bindtype>
2613 to C<columns>, you must provide the bind values in the
2614 C<< [ column_meta => value ] >> format, where C<column_meta> is an opaque
2615 scalar value; most commonly the column name, but you can use any scalar value
2616 (including references and blessed references), L<SQL::Abstract> will simply
2617 pass it through intact. So if C<bindtype> is set to C<columns> the above
2618 example will look like:
2621 date_column => \[ "= date '2008-09-30' - ?::integer", [ {} => 10 ] ]
2624 Literal SQL is especially useful for nesting parenthesized clauses in the
2625 main SQL query. Here is a first example:
2627 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
2631 bar => \["IN ($sub_stmt)" => @sub_bind],
2636 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
2637 WHERE c2 < ? AND c3 LIKE ?))";
2638 @bind = (1234, 100, "foo%");
2640 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
2641 are expressed in the same way. Of course the C<$sub_stmt> and
2642 its associated bind values can be generated through a former call
2645 my ($sub_stmt, @sub_bind)
2646 = $sql->select("t1", "c1", {c2 => {"<" => 100},
2647 c3 => {-like => "foo%"}});
2650 bar => \["> ALL ($sub_stmt)" => @sub_bind],
2653 In the examples above, the subquery was used as an operator on a column;
2654 but the same principle also applies for a clause within the main C<%where>
2655 hash, like an EXISTS subquery:
2657 my ($sub_stmt, @sub_bind)
2658 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
2659 my %where = ( -and => [
2661 \["EXISTS ($sub_stmt)" => @sub_bind],
2666 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
2667 WHERE c1 = ? AND c2 > t0.c0))";
2671 Observe that the condition on C<c2> in the subquery refers to
2672 column C<t0.c0> of the main query: this is I<not> a bind
2673 value, so we have to express it through a scalar ref.
2674 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
2675 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
2676 what we wanted here.
2678 Finally, here is an example where a subquery is used
2679 for expressing unary negation:
2681 my ($sub_stmt, @sub_bind)
2682 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
2683 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
2685 lname => {like => '%son%'},
2686 \["NOT ($sub_stmt)" => @sub_bind],
2691 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
2692 @bind = ('%son%', 10, 20)
2694 =head3 Deprecated usage of Literal SQL
2696 Below are some examples of archaic use of literal SQL. It is shown only as
2697 reference for those who deal with legacy code. Each example has a much
2698 better, cleaner and safer alternative that users should opt for in new code.
2704 my %where = ( requestor => \'IS NOT NULL' )
2706 $stmt = "WHERE requestor IS NOT NULL"
2708 This used to be the way of generating NULL comparisons, before the handling
2709 of C<undef> got formalized. For new code please use the superior syntax as
2710 described in L</Tests for NULL values>.
2714 my %where = ( requestor => \'= submitter' )
2716 $stmt = "WHERE requestor = submitter"
2718 This used to be the only way to compare columns. Use the superior L</-ident>
2719 method for all new code. For example an identifier declared in such a way
2720 will be properly quoted if L</quote_char> is properly set, while the legacy
2721 form will remain as supplied.
2725 my %where = ( is_ready => \"", completed => { '>', '2012-12-21' } )
2727 $stmt = "WHERE completed > ? AND is_ready"
2728 @bind = ('2012-12-21')
2730 Using an empty string literal used to be the only way to express a boolean.
2731 For all new code please use the much more readable
2732 L<-bool|/Unary operators: bool> operator.
2738 These pages could go on for a while, since the nesting of the data
2739 structures this module can handle are pretty much unlimited (the
2740 module implements the C<WHERE> expansion as a recursive function
2741 internally). Your best bet is to "play around" with the module a
2742 little to see how the data structures behave, and choose the best
2743 format for your data based on that.
2745 And of course, all the values above will probably be replaced with
2746 variables gotten from forms or the command line. After all, if you
2747 knew everything ahead of time, you wouldn't have to worry about
2748 dynamically-generating SQL and could just hardwire it into your
2751 =head1 ORDER BY CLAUSES
2753 Some functions take an order by clause. This can either be a scalar (just a
2754 column name), a hashref of C<< { -desc => 'col' } >> or C<< { -asc => 'col' }
2755 >>, a scalarref, an arrayref-ref, or an arrayref of any of the previous
2758 Given | Will Generate
2759 ---------------------------------------------------------------
2761 'colA' | ORDER BY colA
2763 [qw/colA colB/] | ORDER BY colA, colB
2765 {-asc => 'colA'} | ORDER BY colA ASC
2767 {-desc => 'colB'} | ORDER BY colB DESC
2769 ['colA', {-asc => 'colB'}] | ORDER BY colA, colB ASC
2771 { -asc => [qw/colA colB/] } | ORDER BY colA ASC, colB ASC
2773 \'colA DESC' | ORDER BY colA DESC
2775 \[ 'FUNC(colA, ?)', $x ] | ORDER BY FUNC(colA, ?)
2776 | /* ...with $x bound to ? */
2779 { -asc => 'colA' }, | colA ASC,
2780 { -desc => [qw/colB/] }, | colB DESC,
2781 { -asc => [qw/colC colD/] },| colC ASC, colD ASC,
2782 \'colE DESC', | colE DESC,
2783 \[ 'FUNC(colF, ?)', $x ], | FUNC(colF, ?)
2784 ] | /* ...with $x bound to ? */
2785 ===============================================================
2789 =head1 SPECIAL OPERATORS
2791 my $sqlmaker = SQL::Abstract->new(special_ops => [
2795 my ($self, $field, $op, $arg) = @_;
2801 handler => 'method_name',
2805 A "special operator" is a SQL syntactic clause that can be
2806 applied to a field, instead of a usual binary operator.
2809 WHERE field IN (?, ?, ?)
2810 WHERE field BETWEEN ? AND ?
2811 WHERE MATCH(field) AGAINST (?, ?)
2813 Special operators IN and BETWEEN are fairly standard and therefore
2814 are builtin within C<SQL::Abstract> (as the overridable methods
2815 C<_where_field_IN> and C<_where_field_BETWEEN>). For other operators,
2816 like the MATCH .. AGAINST example above which is specific to MySQL,
2817 you can write your own operator handlers - supply a C<special_ops>
2818 argument to the C<new> method. That argument takes an arrayref of
2819 operator definitions; each operator definition is a hashref with two
2826 the regular expression to match the operator
2830 Either a coderef or a plain scalar method name. In both cases
2831 the expected return is C<< ($sql, @bind) >>.
2833 When supplied with a method name, it is simply called on the
2834 L<SQL::Abstract> object as:
2836 $self->$method_name($field, $op, $arg)
2840 $field is the LHS of the operator
2841 $op is the part that matched the handler regex
2844 When supplied with a coderef, it is called as:
2846 $coderef->($self, $field, $op, $arg)
2851 For example, here is an implementation
2852 of the MATCH .. AGAINST syntax for MySQL
2854 my $sqlmaker = SQL::Abstract->new(special_ops => [
2856 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
2857 {regex => qr/^match$/i,
2859 my ($self, $field, $op, $arg) = @_;
2860 $arg = [$arg] if not ref $arg;
2861 my $label = $self->_quote($field);
2862 my ($placeholder) = $self->_convert('?');
2863 my $placeholders = join ", ", (($placeholder) x @$arg);
2864 my $sql = $self->_sqlcase('match') . " ($label) "
2865 . $self->_sqlcase('against') . " ($placeholders) ";
2866 my @bind = $self->_bindtype($field, @$arg);
2867 return ($sql, @bind);
2874 =head1 UNARY OPERATORS
2876 my $sqlmaker = SQL::Abstract->new(unary_ops => [
2880 my ($self, $op, $arg) = @_;
2886 handler => 'method_name',
2890 A "unary operator" is a SQL syntactic clause that can be
2891 applied to a field - the operator goes before the field
2893 You can write your own operator handlers - supply a C<unary_ops>
2894 argument to the C<new> method. That argument takes an arrayref of
2895 operator definitions; each operator definition is a hashref with two
2902 the regular expression to match the operator
2906 Either a coderef or a plain scalar method name. In both cases
2907 the expected return is C<< $sql >>.
2909 When supplied with a method name, it is simply called on the
2910 L<SQL::Abstract> object as:
2912 $self->$method_name($op, $arg)
2916 $op is the part that matched the handler regex
2917 $arg is the RHS or argument of the operator
2919 When supplied with a coderef, it is called as:
2921 $coderef->($self, $op, $arg)
2929 Thanks to some benchmarking by Mark Stosberg, it turns out that
2930 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
2931 I must admit this wasn't an intentional design issue, but it's a
2932 byproduct of the fact that you get to control your C<DBI> handles
2935 To maximize performance, use a code snippet like the following:
2937 # prepare a statement handle using the first row
2938 # and then reuse it for the rest of the rows
2940 for my $href (@array_of_hashrefs) {
2941 $stmt ||= $sql->insert('table', $href);
2942 $sth ||= $dbh->prepare($stmt);
2943 $sth->execute($sql->values($href));
2946 The reason this works is because the keys in your C<$href> are sorted
2947 internally by B<SQL::Abstract>. Thus, as long as your data retains
2948 the same structure, you only have to generate the SQL the first time
2949 around. On subsequent queries, simply use the C<values> function provided
2950 by this module to return your values in the correct order.
2952 However this depends on the values having the same type - if, for
2953 example, the values of a where clause may either have values
2954 (resulting in sql of the form C<column = ?> with a single bind
2955 value), or alternatively the values might be C<undef> (resulting in
2956 sql of the form C<column IS NULL> with no bind value) then the
2957 caching technique suggested will not work.
2961 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
2962 really like this part (I do, at least). Building up a complex query
2963 can be as simple as the following:
2970 use CGI::FormBuilder;
2973 my $form = CGI::FormBuilder->new(...);
2974 my $sql = SQL::Abstract->new;
2976 if ($form->submitted) {
2977 my $field = $form->field;
2978 my $id = delete $field->{id};
2979 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
2982 Of course, you would still have to connect using C<DBI> to run the
2983 query, but the point is that if you make your form look like your
2984 table, the actual query script can be extremely simplistic.
2986 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
2987 a fast interface to returning and formatting data. I frequently
2988 use these three modules together to write complex database query
2989 apps in under 50 lines.
2991 =head1 HOW TO CONTRIBUTE
2993 Contributions are always welcome, in all usable forms (we especially
2994 welcome documentation improvements). The delivery methods include git-
2995 or unified-diff formatted patches, GitHub pull requests, or plain bug
2996 reports either via RT or the Mailing list. Contributors are generally
2997 granted full access to the official repository after their first several
2998 patches pass successful review.
3000 This project is maintained in a git repository. The code and related tools are
3001 accessible at the following locations:
3005 =item * Official repo: L<git://git.shadowcat.co.uk/dbsrgits/SQL-Abstract.git>
3007 =item * Official gitweb: L<http://git.shadowcat.co.uk/gitweb/gitweb.cgi?p=dbsrgits/SQL-Abstract.git>
3009 =item * GitHub mirror: L<https://github.com/dbsrgits/sql-abstract>
3011 =item * Authorized committers: L<ssh://dbsrgits@git.shadowcat.co.uk/SQL-Abstract.git>
3017 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
3018 Great care has been taken to preserve the I<published> behavior
3019 documented in previous versions in the 1.* family; however,
3020 some features that were previously undocumented, or behaved
3021 differently from the documentation, had to be changed in order
3022 to clarify the semantics. Hence, client code that was relying
3023 on some dark areas of C<SQL::Abstract> v1.*
3024 B<might behave differently> in v1.50.
3026 The main changes are:
3032 support for literal SQL through the C<< \ [ $sql, @bind ] >> syntax.
3036 support for the { operator => \"..." } construct (to embed literal SQL)
3040 support for the { operator => \["...", @bind] } construct (to embed literal SQL with bind values)
3044 optional support for L<array datatypes|/"Inserting and Updating Arrays">
3048 defensive programming: check arguments
3052 fixed bug with global logic, which was previously implemented
3053 through global variables yielding side-effects. Prior versions would
3054 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
3055 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
3056 Now this is interpreted
3057 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
3062 fixed semantics of _bindtype on array args
3066 dropped the C<_anoncopy> of the %where tree. No longer necessary,
3067 we just avoid shifting arrays within that tree.
3071 dropped the C<_modlogic> function
3075 =head1 ACKNOWLEDGEMENTS
3077 There are a number of individuals that have really helped out with
3078 this module. Unfortunately, most of them submitted bugs via CPAN
3079 so I have no idea who they are! But the people I do know are:
3081 Ash Berlin (order_by hash term support)
3082 Matt Trout (DBIx::Class support)
3083 Mark Stosberg (benchmarking)
3084 Chas Owens (initial "IN" operator support)
3085 Philip Collins (per-field SQL functions)
3086 Eric Kolve (hashref "AND" support)
3087 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
3088 Dan Kubb (support for "quote_char" and "name_sep")
3089 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
3090 Laurent Dami (internal refactoring, extensible list of special operators, literal SQL)
3091 Norbert Buchmuller (support for literal SQL in hashpair, misc. fixes & tests)
3092 Peter Rabbitson (rewrite of SQLA::Test, misc. fixes & tests)
3093 Oliver Charles (support for "RETURNING" after "INSERT")
3099 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
3103 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
3105 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
3107 For support, your best bet is to try the C<DBIx::Class> users mailing list.
3108 While not an official support venue, C<DBIx::Class> makes heavy use of
3109 C<SQL::Abstract>, and as such list members there are very familiar with
3110 how to create queries.
3114 This module is free software; you may copy this under the same
3115 terms as perl itself (either the GNU General Public License or
3116 the Artistic License)