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 = '1.86';
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 => '_expand_value',
197 -not => '_expand_not',
198 -bool => '_expand_bool',
201 return bless \%opt, $class;
204 sub sqltrue { +{ -literal => [ $_[0]->{sqltrue} ] } }
205 sub sqlfalse { +{ -literal => [ $_[0]->{sqlfalse} ] } }
207 sub _assert_pass_injection_guard {
208 if ($_[1] =~ $_[0]->{injection_guard}) {
209 my $class = ref $_[0];
210 puke "Possible SQL injection attempt '$_[1]'. If this is indeed a part of the "
211 . "desired SQL use literal SQL ( \'...' or \[ '...' ] ) or supply your own "
212 . "{injection_guard} attribute to ${class}->new()"
217 #======================================================================
219 #======================================================================
223 my $table = $self->_table(shift);
224 my $data = shift || return;
227 my $method = $self->_METHOD_FOR_refkind("_insert", $data);
228 my ($sql, @bind) = $self->$method($data);
229 $sql = join " ", $self->_sqlcase('insert into'), $table, $sql;
231 if ($options->{returning}) {
232 my ($s, @b) = $self->_insert_returning($options);
237 return wantarray ? ($sql, @bind) : $sql;
240 # So that subclasses can override INSERT ... RETURNING separately from
241 # UPDATE and DELETE (e.g. DBIx::Class::SQLMaker::Oracle does this)
242 sub _insert_returning { shift->_returning(@_) }
245 my ($self, $options) = @_;
247 my $f = $options->{returning};
249 my ($sql, @bind) = $self->render_aqt(
250 $self->_expand_maybe_list_expr($f, undef, -ident)
253 ? $self->_sqlcase(' returning ') . $sql
254 : ($self->_sqlcase(' returning ').$sql, @bind);
257 sub _insert_HASHREF { # explicit list of fields and then values
258 my ($self, $data) = @_;
260 my @fields = sort keys %$data;
262 my ($sql, @bind) = $self->_insert_values($data);
265 $_ = $self->_quote($_) foreach @fields;
266 $sql = "( ".join(", ", @fields).") ".$sql;
268 return ($sql, @bind);
271 sub _insert_ARRAYREF { # just generate values(?,?) part (no list of fields)
272 my ($self, $data) = @_;
274 # no names (arrayref) so can't generate bindtype
275 $self->{bindtype} ne 'columns'
276 or belch "can't do 'columns' bindtype when called with arrayref";
278 my (@values, @all_bind);
279 foreach my $value (@$data) {
280 my ($values, @bind) = $self->_insert_value(undef, $value);
281 push @values, $values;
282 push @all_bind, @bind;
284 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
285 return ($sql, @all_bind);
288 sub _insert_ARRAYREFREF { # literal SQL with bind
289 my ($self, $data) = @_;
291 my ($sql, @bind) = @${$data};
292 $self->_assert_bindval_matches_bindtype(@bind);
294 return ($sql, @bind);
298 sub _insert_SCALARREF { # literal SQL without bind
299 my ($self, $data) = @_;
305 my ($self, $data) = @_;
307 my (@values, @all_bind);
308 foreach my $column (sort keys %$data) {
309 my ($values, @bind) = $self->_insert_value($column, $data->{$column});
310 push @values, $values;
311 push @all_bind, @bind;
313 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
314 return ($sql, @all_bind);
318 my ($self, $column, $v) = @_;
320 return $self->render_aqt(
321 $self->_expand_insert_value($column, $v)
325 sub _expand_insert_value {
326 my ($self, $column, $v) = @_;
328 if (ref($v) eq 'ARRAY') {
329 if ($self->{array_datatypes}) {
330 return +{ -bind => [ $column, $v ] };
332 my ($sql, @bind) = @$v;
333 $self->_assert_bindval_matches_bindtype(@bind);
334 return +{ -literal => $v };
336 if (ref($v) eq 'HASH') {
337 if (grep !/^-/, keys %$v) {
338 belch "HASH ref as bind value in insert is not supported";
339 return +{ -bind => [ $column, $v ] };
343 return +{ -bind => [ $column, undef ] };
345 local our $Cur_Col_Meta = $column;
346 return $self->expand_expr($v);
351 #======================================================================
353 #======================================================================
358 my $table = $self->_table(shift);
359 my $data = shift || return;
363 # first build the 'SET' part of the sql statement
364 puke "Unsupported data type specified to \$sql->update"
365 unless ref $data eq 'HASH';
367 my ($sql, @all_bind) = $self->_update_set_values($data);
368 $sql = $self->_sqlcase('update ') . $table . $self->_sqlcase(' set ')
372 my($where_sql, @where_bind) = $self->where($where);
374 push @all_bind, @where_bind;
377 if ($options->{returning}) {
378 my ($returning_sql, @returning_bind) = $self->_update_returning($options);
379 $sql .= $returning_sql;
380 push @all_bind, @returning_bind;
383 return wantarray ? ($sql, @all_bind) : $sql;
386 sub _update_set_values {
387 my ($self, $data) = @_;
389 return $self->render_aqt(
390 $self->_expand_update_set_values($data),
394 sub _expand_update_set_values {
395 my ($self, $data) = @_;
396 $self->_expand_maybe_list_expr( [
399 $set = { -bind => $_ } unless defined $set;
400 +{ -op => [ '=', $self->_expand_ident(-ident => $k), $set ] };
406 ? ($self->{array_datatypes}
407 ? [ $k, +{ -bind => [ $k, $v ] } ]
408 : [ $k, +{ -literal => $v } ])
410 local our $Cur_Col_Meta = $k;
411 [ $k, $self->_expand_expr($v) ]
418 # So that subclasses can override UPDATE ... RETURNING separately from
420 sub _update_returning { shift->_returning(@_) }
424 #======================================================================
426 #======================================================================
431 my $table = $self->_table(shift);
432 my $fields = shift || '*';
436 my ($fields_sql, @bind) = $self->_select_fields($fields);
438 my ($where_sql, @where_bind) = $self->where($where, $order);
439 push @bind, @where_bind;
441 my $sql = join(' ', $self->_sqlcase('select'), $fields_sql,
442 $self->_sqlcase('from'), $table)
445 return wantarray ? ($sql, @bind) : $sql;
449 my ($self, $fields) = @_;
450 return $fields unless ref($fields);
451 return $self->render_aqt(
452 $self->_expand_maybe_list_expr($fields, undef, '-ident')
456 #======================================================================
458 #======================================================================
463 my $table = $self->_table(shift);
467 my($where_sql, @bind) = $self->where($where);
468 my $sql = $self->_sqlcase('delete from ') . $table . $where_sql;
470 if ($options->{returning}) {
471 my ($returning_sql, @returning_bind) = $self->_delete_returning($options);
472 $sql .= $returning_sql;
473 push @bind, @returning_bind;
476 return wantarray ? ($sql, @bind) : $sql;
479 # So that subclasses can override DELETE ... RETURNING separately from
481 sub _delete_returning { shift->_returning(@_) }
485 #======================================================================
487 #======================================================================
491 # Finally, a separate routine just to handle WHERE clauses
493 my ($self, $where, $order) = @_;
495 local $self->{convert_where} = $self->{convert};
498 my ($sql, @bind) = defined($where)
499 ? $self->_recurse_where($where)
501 $sql = (defined $sql and length $sql) ? $self->_sqlcase(' where ') . "( $sql )" : '';
505 my ($order_sql, @order_bind) = $self->_order_by($order);
507 push @bind, @order_bind;
510 return wantarray ? ($sql, @bind) : $sql;
514 my ($self, $expr, $default_scalar_to) = @_;
515 local our $Default_Scalar_To = $default_scalar_to if $default_scalar_to;
516 $self->_expand_expr($expr);
520 my ($self, $aqt) = @_;
521 my ($k, $v, @rest) = %$aqt;
523 if (my $meth = $self->{render}{$k}) {
524 return $self->$meth($v);
526 die "notreached: $k";
530 my ($self, $expr) = @_;
531 $self->render_aqt($self->expand_expr($expr));
535 my ($self, $expr, $logic) = @_;
536 our $Expand_Depth ||= 0; local $Expand_Depth = $Expand_Depth + 1;
537 return undef unless defined($expr);
538 if (ref($expr) eq 'HASH') {
539 return undef unless my $kc = keys %$expr;
544 map $self->_expand_expr({ $_ => $expr->{$_} }, $logic),
548 my ($key, $value) = %$expr;
549 if (my $exp = $self->{expand}{$key}) {
550 return $self->$exp($key, $value);
552 return $self->_expand_expr_hashpair($key, $value, $logic);
554 if (ref($expr) eq 'ARRAY') {
555 my $logic = lc($logic || $self->{logic});
556 $logic eq 'and' or $logic eq 'or' or puke "unknown logic: $logic";
559 (ref($_) eq 'ARRAY' and @$_)
560 or (ref($_) eq 'HASH' and %$_)
566 while (my ($el) = splice @expr, 0, 1) {
567 puke "Supplying an empty left hand side argument is not supported in array-pairs"
568 unless defined($el) and length($el);
569 my $elref = ref($el);
571 local $Expand_Depth = 0;
572 push(@res, grep defined, $self->_expand_expr({ $el, shift(@expr) }));
573 } elsif ($elref eq 'ARRAY') {
574 push(@res, grep defined, $self->_expand_expr($el)) if @$el;
575 } elsif (my $l = is_literal_value($el)) {
576 push @res, { -literal => $l };
577 } elsif ($elref eq 'HASH') {
578 local $Expand_Depth = 0;
579 push @res, grep defined, $self->_expand_expr($el) if %$el;
585 # return $res[0] if @res == 1;
586 return { -op => [ $logic, @res ] };
588 if (my $literal = is_literal_value($expr)) {
589 return +{ -literal => $literal };
591 if (!ref($expr) or Scalar::Util::blessed($expr)) {
592 if (my $d = our $Default_Scalar_To) {
593 return $self->_expand_expr({ $d => $expr });
595 if (my $m = our $Cur_Col_Meta) {
596 return +{ -bind => [ $m, $expr ] };
598 return +{ -bind => [ undef, $expr ] };
603 sub _expand_expr_hashpair {
604 my ($self, $k, $v, $logic) = @_;
605 unless (defined($k) and length($k)) {
606 if (defined($k) and my $literal = is_literal_value($v)) {
607 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
608 return { -literal => $literal };
610 puke "Supplying an empty left hand side argument is not supported";
613 $self->_assert_pass_injection_guard($k =~ /^-(.*)$/s);
614 if ($k =~ s/ [_\s]? \d+ $//x ) {
615 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
616 . "You probably wanted ...-and => [ $k => COND1, $k => COND2 ... ]";
619 # DBIx::Class requires a nest warning to be emitted once but the private
620 # method it overrode to do so no longer exists
621 if ($self->{is_dbic_sqlmaker}) {
622 unless (our $Nest_Warned) {
624 "-nest in search conditions is deprecated, you most probably wanted:\n"
625 .q|{..., -and => [ \%cond0, \@cond1, \'cond2', \[ 'cond3', [ col => bind ] ], etc. ], ... }|
630 return $self->_expand_expr($v);
632 if (my ($rest) = $k =~/^-not[_ ](.*)$/) {
635 $self->_expand_expr({ "-${rest}", $v }, $logic)
638 if (my ($logic) = $k =~ /^-(and|or)$/i) {
639 if (ref($v) eq 'HASH') {
640 return $self->_expand_expr($v, $logic);
642 if (ref($v) eq 'ARRAY') {
643 return $self->_expand_expr($v, $logic);
648 $op =~ s/^-// if length($op) > 1;
650 # top level special ops are illegal in general
651 # note that, arguably, if it makes no sense at top level, it also
652 # makes no sense on the other side of an = sign or similar but DBIC
653 # gets disappointingly upset if I disallow it
655 (our $Expand_Depth) == 1
656 and List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}
658 puke "Illegal use of top-level '-$op'"
660 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
661 return { -op => [ $op, $v ] };
664 if (my $custom = $self->{expand_unary}{$k}) {
665 return $self->$custom($v);
667 if ($self->{render}{$k}) {
673 and (keys %$v)[0] =~ /^-/
675 my ($func) = $k =~ /^-(.*)$/;
676 if (List::Util::first { $func =~ $_->{regex} } @{$self->{special_ops}}) {
677 return +{ -op => [ $func, $self->_expand_expr($v) ] };
679 return +{ -func => [ $func, $self->_expand_expr($v) ] };
681 if (!ref($v) or is_literal_value($v)) {
682 return +{ -op => [ $k =~ /^-(.*)$/, $self->_expand_expr($v) ] };
689 and exists $v->{-value}
690 and not defined $v->{-value}
693 return $self->_expand_expr({ $k => { $self->{cmp} => undef } });
695 if (!ref($v) or Scalar::Util::blessed($v)) {
696 my $d = our $Default_Scalar_To;
700 $self->_expand_ident(-ident => $k),
702 ? $self->_expand_expr($d => $v)
703 : { -bind => [ $k, $v ] }
708 if (ref($v) eq 'HASH') {
712 map $self->_expand_expr({ $k => { $_ => $v->{$_} } }),
716 return undef unless keys %$v;
718 my $op = join ' ', split '_', (map lc, $vk =~ /^-?(.*)$/)[0];
719 $self->_assert_pass_injection_guard($op);
720 if ($op =~ s/ [_\s]? \d+ $//x ) {
721 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
722 . "You probably wanted ...-and => [ -$op => COND1, -$op => COND2 ... ]";
724 if ($op =~ /^(?:not )?between$/) {
725 local our $Cur_Col_Meta = $k;
726 my @rhs = map $self->_expand_expr($_),
727 ref($vv) eq 'ARRAY' ? @$vv : $vv;
729 (@rhs == 1 and ref($rhs[0]) eq 'HASH' and $rhs[0]->{-literal})
731 (@rhs == 2 and defined($rhs[0]) and defined($rhs[1]))
733 puke "Operator '${\uc($op)}' requires either an arrayref with two defined values or expressions, or a single literal scalarref/arrayref-ref";
737 $self->_expand_ident(-ident => $k),
741 if ($op =~ /^(?:not )?in$/) {
742 if (my $literal = is_literal_value($vv)) {
743 my ($sql, @bind) = @$literal;
744 my $opened_sql = $self->_open_outer_paren($sql);
746 $op, $self->_expand_ident(-ident => $k),
747 [ { -literal => [ $opened_sql, @bind ] } ]
751 'SQL::Abstract before v1.75 used to generate incorrect SQL when the '
752 . "-${\uc($op)} operator was given an undef-containing list: !!!AUDIT YOUR CODE "
753 . 'AND DATA!!! (the upcoming Data::Query-based version of SQL::Abstract '
754 . 'will emit the logically correct SQL instead of raising this exception)'
756 puke("Argument passed to the '${\uc($op)}' operator can not be undefined")
758 my @rhs = map $self->_expand_expr($_),
759 map { ref($_) ? $_ : { -bind => [ $k, $_ ] } }
760 map { defined($_) ? $_: puke($undef_err) }
761 (ref($vv) eq 'ARRAY' ? @$vv : $vv);
762 return $self->${\($op =~ /^not/ ? 'sqltrue' : 'sqlfalse')} unless @rhs;
766 $self->_expand_ident(-ident => $k),
770 if ($op eq 'ident') {
771 if (! defined $vv or (ref($vv) and ref($vv) eq 'ARRAY')) {
772 puke "-$op requires a single plain scalar argument (a quotable identifier) or an arrayref of identifier parts";
776 $self->_expand_ident(-ident => $k),
777 $self->_expand_ident(-ident => $vv),
780 if ($op eq 'value') {
781 return $self->_expand_expr({ $k, undef }) unless defined($vv);
784 $self->_expand_ident(-ident => $k),
785 { -bind => [ $k, $vv ] }
788 if ($op =~ /^is(?: not)?$/) {
789 puke "$op can only take undef as argument"
793 and exists($vv->{-value})
794 and !defined($vv->{-value})
796 return +{ -op => [ $op.' null', $self->_expand_ident(-ident => $k) ] };
798 if ($op =~ /^(and|or)$/) {
799 if (ref($vv) eq 'HASH') {
802 map $self->_expand_expr({ $k, { $_ => $vv->{$_} } }),
807 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}) {
808 return { -op => [ $op, $self->_expand_ident(-ident => $k), $vv ] };
810 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
813 $self->_expand_ident(-ident => $k),
814 { -op => [ $op, $vv ] }
817 if (ref($vv) eq 'ARRAY') {
818 my ($logic, @values) = (
819 (defined($vv->[0]) and $vv->[0] =~ /^-(and|or)$/i)
824 $op =~ $self->{inequality_op}
825 or $op =~ $self->{not_like_op}
827 if (lc($logic) eq '-or' and @values > 1) {
828 belch "A multi-element arrayref as an argument to the inequality op '${\uc($op)}' "
829 . 'is technically equivalent to an always-true 1=1 (you probably wanted '
830 . "to say ...{ \$inequality_op => [ -and => \@values ] }... instead)"
835 # try to DWIM on equality operators
837 $op =~ $self->{equality_op} ? $self->sqlfalse
838 : $op =~ $self->{like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->sqlfalse
839 : $op =~ $self->{inequality_op} ? $self->sqltrue
840 : $op =~ $self->{not_like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->sqltrue
841 : puke "operator '$op' applied on an empty array (field '$k')";
845 map $self->_expand_expr({ $k => { $vk => $_ } }),
853 and exists $vv->{-value}
854 and not defined $vv->{-value}
858 $op =~ /^not$/i ? 'is not' # legacy
859 : $op =~ $self->{equality_op} ? 'is'
860 : $op =~ $self->{like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is'
861 : $op =~ $self->{inequality_op} ? 'is not'
862 : $op =~ $self->{not_like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is not'
863 : puke "unexpected operator '$op' with undef operand";
864 return +{ -op => [ $is.' null', $self->_expand_ident(-ident => $k) ] };
866 local our $Cur_Col_Meta = $k;
869 $self->_expand_ident(-ident => $k),
870 $self->_expand_expr($vv)
873 if (ref($v) eq 'ARRAY') {
874 return $self->sqlfalse unless @$v;
875 $self->_debug("ARRAY($k) means distribute over elements");
877 $v->[0] =~ /^-((?:and|or))$/i
878 ? ($v = [ @{$v}[1..$#$v] ], $1)
879 : ($self->{logic} || 'or')
883 map $self->_expand_expr({ $k => $_ }, $this_logic), @$v
886 if (my $literal = is_literal_value($v)) {
888 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
891 my ($sql, @bind) = @$literal;
892 if ($self->{bindtype} eq 'columns') {
894 $self->_assert_bindval_matches_bindtype($_);
897 return +{ -literal => [ $self->_quote($k).' '.$sql, @bind ] };
903 my ($self, undef, $body) = @_;
904 my @parts = map split(/\Q${\($self->{name_sep}||'.')}\E/, $_),
905 ref($body) ? @$body : $body;
906 return { -ident => $parts[-1] } if $self->{_dequalify_idents};
907 unless ($self->{quote_char}) {
908 $self->_assert_pass_injection_guard($_) for @parts;
910 return +{ -ident => \@parts };
914 +{ -bind => [ our $Cur_Col_Meta, $_[2] ] };
918 +{ -op => [ 'not', $_[0]->_expand_expr($_[2]) ] };
922 my ($self, undef, $v) = @_;
924 return $self->_expand_expr($v);
926 puke "-bool => undef not supported" unless defined($v);
927 return $self->_expand_ident(-ident => $v);
931 my ($self, $where, $logic) = @_;
933 # Special case: top level simple string treated as literal
935 my $where_exp = (ref($where)
936 ? $self->_expand_expr($where, $logic)
937 : { -literal => [ $where ] });
939 # dispatch expanded expression
941 my ($sql, @bind) = defined($where_exp) ? $self->render_aqt($where_exp) : (undef);
942 # DBIx::Class used to call _recurse_where in scalar context
943 # something else might too...
945 return ($sql, @bind);
948 belch "Calling _recurse_where in scalar context is deprecated and will go away before 2.0";
954 my ($self, $ident) = @_;
956 return $self->_convert($self->_quote($ident));
959 my %unop_postfix = map +($_ => 1),
960 'is null', 'is not null',
968 my ($self, $args) = @_;
969 my ($left, $low, $high) = @$args;
970 my ($rhsql, @rhbind) = do {
972 puke "Single arg to between must be a literal"
973 unless $low->{-literal};
976 my ($l, $h) = map [ $self->render_aqt($_) ], $low, $high;
977 (join(' ', $l->[0], $self->_sqlcase('and'), $h->[0]),
978 @{$l}[1..$#$l], @{$h}[1..$#$h])
981 my ($lhsql, @lhbind) = $self->render_aqt($left);
983 join(' ', '(', $lhsql, $self->_sqlcase($op), $rhsql, ')'),
987 }), 'between', 'not between'),
991 my ($self, $args) = @_;
992 my ($lhs, $rhs) = @$args;
995 my ($sql, @bind) = $self->render_aqt($_);
996 push @in_bind, @bind;
999 my ($lhsql, @lbind) = $self->render_aqt($lhs);
1001 $lhsql.' '.$self->_sqlcase($op).' ( '
1002 .join(', ', @in_sql)
1007 }), 'in', 'not in'),
1011 my ($self, $v) = @_;
1012 my ($op, @args) = @$v;
1013 $op =~ s/^-// if length($op) > 1;
1015 if (my $h = $special{$op}) {
1016 return $self->$h(\@args);
1018 my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}};
1019 if ($us and @args > 1) {
1020 puke "Special op '${op}' requires first value to be identifier"
1021 unless my ($ident) = map $_->{-ident}, grep ref($_) eq 'HASH', $args[0];
1022 my $k = join(($self->{name_sep}||'.'), @$ident);
1023 local our $Expand_Depth = 1;
1024 return $self->${\($us->{handler})}($k, $op, $args[1]);
1026 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1027 return $self->${\($us->{handler})}($op, $args[0]);
1029 my $final_op = $op =~ /^(?:is|not)_/ ? join(' ', split '_', $op) : $op;
1030 if (@args == 1 and $op !~ /^(and|or)$/) {
1031 my ($expr_sql, @bind) = $self->render_aqt($args[0]);
1032 my $op_sql = $self->_sqlcase($final_op);
1034 $unop_postfix{lc($final_op)}
1035 ? "${expr_sql} ${op_sql}"
1036 : "${op_sql} ${expr_sql}"
1038 return (($op eq 'not' || $us ? '('.$final_sql.')' : $final_sql), @bind);
1040 my @parts = grep length($_->[0]), map [ $self->render_aqt($_) ], @args;
1041 return '' unless @parts;
1042 my $is_andor = !!($op =~ /^(and|or)$/);
1043 return @{$parts[0]} if $is_andor and @parts == 1;
1044 my ($final_sql) = map +($is_andor ? "( ${_} )" : $_), join(
1045 ' '.$self->_sqlcase($final_op).' ',
1050 map @{$_}[1..$#$_], @parts
1057 my ($self, $list) = @_;
1058 my @parts = grep length($_->[0]), map [ $self->render_aqt($_) ], @$list;
1059 return join(', ', map $_->[0], @parts), map @{$_}[1..$#$_], @parts;
1063 my ($self, $rest) = @_;
1064 my ($func, @args) = @$rest;
1068 push @arg_sql, shift @x;
1070 } map [ $self->render_aqt($_) ], @args;
1071 return ($self->_sqlcase($func).'('.join(', ', @arg_sql).')', @bind);
1075 my ($self, $bind) = @_;
1076 return ($self->_convert('?'), $self->_bindtype(@$bind));
1079 sub _render_literal {
1080 my ($self, $literal) = @_;
1081 $self->_assert_bindval_matches_bindtype(@{$literal}[1..$#$literal]);
1085 # Some databases (SQLite) treat col IN (1, 2) different from
1086 # col IN ( (1, 2) ). Use this to strip all outer parens while
1087 # adding them back in the corresponding method
1088 sub _open_outer_paren {
1089 my ($self, $sql) = @_;
1091 while (my ($inner) = $sql =~ /^ \s* \( (.*) \) \s* $/xs) {
1093 # there are closing parens inside, need the heavy duty machinery
1094 # to reevaluate the extraction starting from $sql (full reevaluation)
1095 if ($inner =~ /\)/) {
1096 require Text::Balanced;
1098 my (undef, $remainder) = do {
1099 # idiotic design - writes to $@ but *DOES NOT* throw exceptions
1101 Text::Balanced::extract_bracketed($sql, '()', qr/\s*/);
1104 # the entire expression needs to be a balanced bracketed thing
1105 # (after an extract no remainder sans trailing space)
1106 last if defined $remainder and $remainder =~ /\S/;
1116 #======================================================================
1118 #======================================================================
1120 sub _expand_order_by {
1121 my ($self, $arg) = @_;
1123 return unless defined($arg) and not (ref($arg) eq 'ARRAY' and !@$arg);
1125 my $expander = sub {
1126 my ($self, $dir, $expr) = @_;
1127 my @to_expand = ref($expr) eq 'ARRAY' ? @$expr : $expr;
1128 foreach my $arg (@to_expand) {
1132 and grep /^-(asc|desc)$/, keys %$arg
1134 puke "ordering direction hash passed to order by must have exactly one key (-asc or -desc)";
1137 my @exp = map +(defined($dir) ? { -op => [ $dir => $_ ] } : $_),
1138 map $self->expand_expr($_, -ident),
1139 map ref($_) eq 'ARRAY' ? @$_ : $_, @to_expand;
1140 return (@exp > 1 ? { -list => \@exp } : $exp[0]);
1143 local @{$self->{expand_unary}}{qw(-asc -desc)} = (
1144 sub { shift->$expander(asc => @_) },
1145 sub { shift->$expander(desc => @_) },
1148 return $self->$expander(undef, $arg);
1152 my ($self, $arg) = @_;
1154 return '' unless defined(my $expanded = $self->_expand_order_by($arg));
1156 my ($sql, @bind) = $self->render_aqt($expanded);
1158 return '' unless length($sql);
1160 my $final_sql = $self->_sqlcase(' order by ').$sql;
1162 return wantarray ? ($final_sql, @bind) : $final_sql;
1165 # _order_by no longer needs to call this so doesn't but DBIC uses it.
1167 sub _order_by_chunks {
1168 my ($self, $arg) = @_;
1170 return () unless defined(my $expanded = $self->_expand_order_by($arg));
1172 return $self->_chunkify_order_by($expanded);
1175 sub _chunkify_order_by {
1176 my ($self, $expanded) = @_;
1178 return grep length, $self->render_aqt($expanded)
1179 if $expanded->{-ident} or @{$expanded->{-literal}||[]} == 1;
1182 if (ref() eq 'HASH' and my $l = $_->{-list}) {
1183 return map $self->_chunkify_order_by($_), @$l;
1185 return [ $self->render_aqt($_) ];
1189 #======================================================================
1190 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
1191 #======================================================================
1197 $self->_expand_maybe_list_expr($from, undef, -ident)
1202 #======================================================================
1204 #======================================================================
1206 sub _expand_maybe_list_expr {
1207 my ($self, $expr, $logic, $default) = @_;
1209 if (ref($expr) eq 'ARRAY') {
1211 map $self->expand_expr($_, $default), @$expr
1218 return $self->expand_expr($e, $default);
1221 # highly optimized, as it's called way too often
1223 # my ($self, $label) = @_;
1225 return '' unless defined $_[1];
1226 return ${$_[1]} if ref($_[1]) eq 'SCALAR';
1227 puke 'Identifier cannot be hashref' if ref($_[1]) eq 'HASH';
1229 unless ($_[0]->{quote_char}) {
1230 if (ref($_[1]) eq 'ARRAY') {
1231 return join($_[0]->{name_sep}||'.', @{$_[1]});
1233 $_[0]->_assert_pass_injection_guard($_[1]);
1238 my $qref = ref $_[0]->{quote_char};
1240 !$qref ? ($_[0]->{quote_char}, $_[0]->{quote_char})
1241 : ($qref eq 'ARRAY') ? @{$_[0]->{quote_char}}
1242 : puke "Unsupported quote_char format: $_[0]->{quote_char}";
1244 my $esc = $_[0]->{escape_char} || $r;
1246 # parts containing * are naturally unquoted
1248 $_[0]->{name_sep}||'',
1252 : do { (my $n = $_) =~ s/(\Q$esc\E|\Q$r\E)/$esc$1/g; $l . $n . $r }
1254 (ref($_[1]) eq 'ARRAY'
1258 ? split (/\Q$_[0]->{name_sep}\E/, $_[1] )
1266 # Conversion, if applicable
1268 #my ($self, $arg) = @_;
1269 if ($_[0]->{convert_where}) {
1270 return $_[0]->_sqlcase($_[0]->{convert_where}) .'(' . $_[1] . ')';
1277 #my ($self, $col, @vals) = @_;
1278 # called often - tighten code
1279 return $_[0]->{bindtype} eq 'columns'
1280 ? map {[$_[1], $_]} @_[2 .. $#_]
1285 # Dies if any element of @bind is not in [colname => value] format
1286 # if bindtype is 'columns'.
1287 sub _assert_bindval_matches_bindtype {
1288 # my ($self, @bind) = @_;
1290 if ($self->{bindtype} eq 'columns') {
1292 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
1293 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
1299 sub _join_sql_clauses {
1300 my ($self, $logic, $clauses_aref, $bind_aref) = @_;
1302 if (@$clauses_aref > 1) {
1303 my $join = " " . $self->_sqlcase($logic) . " ";
1304 my $sql = '( ' . join($join, @$clauses_aref) . ' )';
1305 return ($sql, @$bind_aref);
1307 elsif (@$clauses_aref) {
1308 return ($clauses_aref->[0], @$bind_aref); # no parentheses
1311 return (); # if no SQL, ignore @$bind_aref
1316 # Fix SQL case, if so requested
1318 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
1319 # don't touch the argument ... crooked logic, but let's not change it!
1320 return $_[0]->{case} ? $_[1] : uc($_[1]);
1324 #======================================================================
1325 # DISPATCHING FROM REFKIND
1326 #======================================================================
1329 my ($self, $data) = @_;
1331 return 'UNDEF' unless defined $data;
1333 # blessed objects are treated like scalars
1334 my $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1336 return 'SCALAR' unless $ref;
1339 while ($ref eq 'REF') {
1341 $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1345 return ($ref||'SCALAR') . ('REF' x $n_steps);
1349 my ($self, $data) = @_;
1350 my @try = ($self->_refkind($data));
1351 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
1352 push @try, 'FALLBACK';
1356 sub _METHOD_FOR_refkind {
1357 my ($self, $meth_prefix, $data) = @_;
1360 for (@{$self->_try_refkind($data)}) {
1361 $method = $self->can($meth_prefix."_".$_)
1365 return $method || puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
1369 sub _SWITCH_refkind {
1370 my ($self, $data, $dispatch_table) = @_;
1373 for (@{$self->_try_refkind($data)}) {
1374 $coderef = $dispatch_table->{$_}
1378 puke "no dispatch entry for ".$self->_refkind($data)
1387 #======================================================================
1388 # VALUES, GENERATE, AUTOLOAD
1389 #======================================================================
1391 # LDNOTE: original code from nwiger, didn't touch code in that section
1392 # I feel the AUTOLOAD stuff should not be the default, it should
1393 # only be activated on explicit demand by user.
1397 my $data = shift || return;
1398 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
1399 unless ref $data eq 'HASH';
1402 foreach my $k (sort keys %$data) {
1403 my $v = $data->{$k};
1404 $self->_SWITCH_refkind($v, {
1406 if ($self->{array_datatypes}) { # array datatype
1407 push @all_bind, $self->_bindtype($k, $v);
1409 else { # literal SQL with bind
1410 my ($sql, @bind) = @$v;
1411 $self->_assert_bindval_matches_bindtype(@bind);
1412 push @all_bind, @bind;
1415 ARRAYREFREF => sub { # literal SQL with bind
1416 my ($sql, @bind) = @${$v};
1417 $self->_assert_bindval_matches_bindtype(@bind);
1418 push @all_bind, @bind;
1420 SCALARREF => sub { # literal SQL without bind
1422 SCALAR_or_UNDEF => sub {
1423 push @all_bind, $self->_bindtype($k, $v);
1434 my(@sql, @sqlq, @sqlv);
1438 if ($ref eq 'HASH') {
1439 for my $k (sort keys %$_) {
1442 my $label = $self->_quote($k);
1443 if ($r eq 'ARRAY') {
1444 # literal SQL with bind
1445 my ($sql, @bind) = @$v;
1446 $self->_assert_bindval_matches_bindtype(@bind);
1447 push @sqlq, "$label = $sql";
1449 } elsif ($r eq 'SCALAR') {
1450 # literal SQL without bind
1451 push @sqlq, "$label = $$v";
1453 push @sqlq, "$label = ?";
1454 push @sqlv, $self->_bindtype($k, $v);
1457 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
1458 } elsif ($ref eq 'ARRAY') {
1459 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
1462 if ($r eq 'ARRAY') { # literal SQL with bind
1463 my ($sql, @bind) = @$v;
1464 $self->_assert_bindval_matches_bindtype(@bind);
1467 } elsif ($r eq 'SCALAR') { # literal SQL without bind
1468 # embedded literal SQL
1475 push @sql, '(' . join(', ', @sqlq) . ')';
1476 } elsif ($ref eq 'SCALAR') {
1480 # strings get case twiddled
1481 push @sql, $self->_sqlcase($_);
1485 my $sql = join ' ', @sql;
1487 # this is pretty tricky
1488 # if ask for an array, return ($stmt, @bind)
1489 # otherwise, s/?/shift @sqlv/ to put it inline
1491 return ($sql, @sqlv);
1493 1 while $sql =~ s/\?/my $d = shift(@sqlv);
1494 ref $d ? $d->[1] : $d/e;
1503 # This allows us to check for a local, then _form, attr
1505 my($name) = $AUTOLOAD =~ /.*::(.+)/;
1506 return $self->generate($name, @_);
1517 SQL::Abstract - Generate SQL from Perl data structures
1523 my $sql = SQL::Abstract->new;
1525 my($stmt, @bind) = $sql->select($source, \@fields, \%where, $order);
1527 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
1529 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
1531 my($stmt, @bind) = $sql->delete($table, \%where);
1533 # Then, use these in your DBI statements
1534 my $sth = $dbh->prepare($stmt);
1535 $sth->execute(@bind);
1537 # Just generate the WHERE clause
1538 my($stmt, @bind) = $sql->where(\%where, $order);
1540 # Return values in the same order, for hashed queries
1541 # See PERFORMANCE section for more details
1542 my @bind = $sql->values(\%fieldvals);
1546 This module was inspired by the excellent L<DBIx::Abstract>.
1547 However, in using that module I found that what I really wanted
1548 to do was generate SQL, but still retain complete control over my
1549 statement handles and use the DBI interface. So, I set out to
1550 create an abstract SQL generation module.
1552 While based on the concepts used by L<DBIx::Abstract>, there are
1553 several important differences, especially when it comes to WHERE
1554 clauses. I have modified the concepts used to make the SQL easier
1555 to generate from Perl data structures and, IMO, more intuitive.
1556 The underlying idea is for this module to do what you mean, based
1557 on the data structures you provide it. The big advantage is that
1558 you don't have to modify your code every time your data changes,
1559 as this module figures it out.
1561 To begin with, an SQL INSERT is as easy as just specifying a hash
1562 of C<key=value> pairs:
1565 name => 'Jimbo Bobson',
1566 phone => '123-456-7890',
1567 address => '42 Sister Lane',
1568 city => 'St. Louis',
1569 state => 'Louisiana',
1572 The SQL can then be generated with this:
1574 my($stmt, @bind) = $sql->insert('people', \%data);
1576 Which would give you something like this:
1578 $stmt = "INSERT INTO people
1579 (address, city, name, phone, state)
1580 VALUES (?, ?, ?, ?, ?)";
1581 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
1582 '123-456-7890', 'Louisiana');
1584 These are then used directly in your DBI code:
1586 my $sth = $dbh->prepare($stmt);
1587 $sth->execute(@bind);
1589 =head2 Inserting and Updating Arrays
1591 If your database has array types (like for example Postgres),
1592 activate the special option C<< array_datatypes => 1 >>
1593 when creating the C<SQL::Abstract> object.
1594 Then you may use an arrayref to insert and update database array types:
1596 my $sql = SQL::Abstract->new(array_datatypes => 1);
1598 planets => [qw/Mercury Venus Earth Mars/]
1601 my($stmt, @bind) = $sql->insert('solar_system', \%data);
1605 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
1607 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
1610 =head2 Inserting and Updating SQL
1612 In order to apply SQL functions to elements of your C<%data> you may
1613 specify a reference to an arrayref for the given hash value. For example,
1614 if you need to execute the Oracle C<to_date> function on a value, you can
1615 say something like this:
1619 date_entered => \[ "to_date(?,'MM/DD/YYYY')", "03/02/2003" ],
1622 The first value in the array is the actual SQL. Any other values are
1623 optional and would be included in the bind values array. This gives
1626 my($stmt, @bind) = $sql->insert('people', \%data);
1628 $stmt = "INSERT INTO people (name, date_entered)
1629 VALUES (?, to_date(?,'MM/DD/YYYY'))";
1630 @bind = ('Bill', '03/02/2003');
1632 An UPDATE is just as easy, all you change is the name of the function:
1634 my($stmt, @bind) = $sql->update('people', \%data);
1636 Notice that your C<%data> isn't touched; the module will generate
1637 the appropriately quirky SQL for you automatically. Usually you'll
1638 want to specify a WHERE clause for your UPDATE, though, which is
1639 where handling C<%where> hashes comes in handy...
1641 =head2 Complex where statements
1643 This module can generate pretty complicated WHERE statements
1644 easily. For example, simple C<key=value> pairs are taken to mean
1645 equality, and if you want to see if a field is within a set
1646 of values, you can use an arrayref. Let's say we wanted to
1647 SELECT some data based on this criteria:
1650 requestor => 'inna',
1651 worker => ['nwiger', 'rcwe', 'sfz'],
1652 status => { '!=', 'completed' }
1655 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
1657 The above would give you something like this:
1659 $stmt = "SELECT * FROM tickets WHERE
1660 ( requestor = ? ) AND ( status != ? )
1661 AND ( worker = ? OR worker = ? OR worker = ? )";
1662 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
1664 Which you could then use in DBI code like so:
1666 my $sth = $dbh->prepare($stmt);
1667 $sth->execute(@bind);
1673 The methods are simple. There's one for every major SQL operation,
1674 and a constructor you use first. The arguments are specified in a
1675 similar order for each method (table, then fields, then a where
1676 clause) to try and simplify things.
1678 =head2 new(option => 'value')
1680 The C<new()> function takes a list of options and values, and returns
1681 a new B<SQL::Abstract> object which can then be used to generate SQL
1682 through the methods below. The options accepted are:
1688 If set to 'lower', then SQL will be generated in all lowercase. By
1689 default SQL is generated in "textbook" case meaning something like:
1691 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
1693 Any setting other than 'lower' is ignored.
1697 This determines what the default comparison operator is. By default
1698 it is C<=>, meaning that a hash like this:
1700 %where = (name => 'nwiger', email => 'nate@wiger.org');
1702 Will generate SQL like this:
1704 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
1706 However, you may want loose comparisons by default, so if you set
1707 C<cmp> to C<like> you would get SQL such as:
1709 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
1711 You can also override the comparison on an individual basis - see
1712 the huge section on L</"WHERE CLAUSES"> at the bottom.
1714 =item sqltrue, sqlfalse
1716 Expressions for inserting boolean values within SQL statements.
1717 By default these are C<1=1> and C<1=0>. They are used
1718 by the special operators C<-in> and C<-not_in> for generating
1719 correct SQL even when the argument is an empty array (see below).
1723 This determines the default logical operator for multiple WHERE
1724 statements in arrays or hashes. If absent, the default logic is "or"
1725 for arrays, and "and" for hashes. This means that a WHERE
1729 event_date => {'>=', '2/13/99'},
1730 event_date => {'<=', '4/24/03'},
1733 will generate SQL like this:
1735 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
1737 This is probably not what you want given this query, though (look
1738 at the dates). To change the "OR" to an "AND", simply specify:
1740 my $sql = SQL::Abstract->new(logic => 'and');
1742 Which will change the above C<WHERE> to:
1744 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
1746 The logic can also be changed locally by inserting
1747 a modifier in front of an arrayref:
1749 @where = (-and => [event_date => {'>=', '2/13/99'},
1750 event_date => {'<=', '4/24/03'} ]);
1752 See the L</"WHERE CLAUSES"> section for explanations.
1756 This will automatically convert comparisons using the specified SQL
1757 function for both column and value. This is mostly used with an argument
1758 of C<upper> or C<lower>, so that the SQL will have the effect of
1759 case-insensitive "searches". For example, this:
1761 $sql = SQL::Abstract->new(convert => 'upper');
1762 %where = (keywords => 'MaKe iT CAse inSeNSItive');
1764 Will turn out the following SQL:
1766 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
1768 The conversion can be C<upper()>, C<lower()>, or any other SQL function
1769 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
1770 not validate this option; it will just pass through what you specify verbatim).
1774 This is a kludge because many databases suck. For example, you can't
1775 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
1776 Instead, you have to use C<bind_param()>:
1778 $sth->bind_param(1, 'reg data');
1779 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
1781 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
1782 which loses track of which field each slot refers to. Fear not.
1784 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
1785 Currently, you can specify either C<normal> (default) or C<columns>. If you
1786 specify C<columns>, you will get an array that looks like this:
1788 my $sql = SQL::Abstract->new(bindtype => 'columns');
1789 my($stmt, @bind) = $sql->insert(...);
1792 [ 'column1', 'value1' ],
1793 [ 'column2', 'value2' ],
1794 [ 'column3', 'value3' ],
1797 You can then iterate through this manually, using DBI's C<bind_param()>.
1799 $sth->prepare($stmt);
1802 my($col, $data) = @$_;
1803 if ($col eq 'details' || $col eq 'comments') {
1804 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
1805 } elsif ($col eq 'image') {
1806 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
1808 $sth->bind_param($i, $data);
1812 $sth->execute; # execute without @bind now
1814 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
1815 Basically, the advantage is still that you don't have to care which fields
1816 are or are not included. You could wrap that above C<for> loop in a simple
1817 sub called C<bind_fields()> or something and reuse it repeatedly. You still
1818 get a layer of abstraction over manual SQL specification.
1820 Note that if you set L</bindtype> to C<columns>, the C<\[ $sql, @bind ]>
1821 construct (see L</Literal SQL with placeholders and bind values (subqueries)>)
1822 will expect the bind values in this format.
1826 This is the character that a table or column name will be quoted
1827 with. By default this is an empty string, but you could set it to
1828 the character C<`>, to generate SQL like this:
1830 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
1832 Alternatively, you can supply an array ref of two items, the first being the left
1833 hand quote character, and the second the right hand quote character. For
1834 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
1835 that generates SQL like this:
1837 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
1839 Quoting is useful if you have tables or columns names that are reserved
1840 words in your database's SQL dialect.
1844 This is the character that will be used to escape L</quote_char>s appearing
1845 in an identifier before it has been quoted.
1847 The parameter default in case of a single L</quote_char> character is the quote
1850 When opening-closing-style quoting is used (L</quote_char> is an arrayref)
1851 this parameter defaults to the B<closing (right)> L</quote_char>. Occurrences
1852 of the B<opening (left)> L</quote_char> within the identifier are currently left
1853 untouched. The default for opening-closing-style quotes may change in future
1854 versions, thus you are B<strongly encouraged> to specify the escape character
1859 This is the character that separates a table and column name. It is
1860 necessary to specify this when the C<quote_char> option is selected,
1861 so that tables and column names can be individually quoted like this:
1863 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
1865 =item injection_guard
1867 A regular expression C<qr/.../> that is applied to any C<-function> and unquoted
1868 column name specified in a query structure. This is a safety mechanism to avoid
1869 injection attacks when mishandling user input e.g.:
1871 my %condition_as_column_value_pairs = get_values_from_user();
1872 $sqla->select( ... , \%condition_as_column_value_pairs );
1874 If the expression matches an exception is thrown. Note that literal SQL
1875 supplied via C<\'...'> or C<\['...']> is B<not> checked in any way.
1877 Defaults to checking for C<;> and the C<GO> keyword (TransactSQL)
1879 =item array_datatypes
1881 When this option is true, arrayrefs in INSERT or UPDATE are
1882 interpreted as array datatypes and are passed directly
1884 When this option is false, arrayrefs are interpreted
1885 as literal SQL, just like refs to arrayrefs
1886 (but this behavior is for backwards compatibility; when writing
1887 new queries, use the "reference to arrayref" syntax
1893 Takes a reference to a list of "special operators"
1894 to extend the syntax understood by L<SQL::Abstract>.
1895 See section L</"SPECIAL OPERATORS"> for details.
1899 Takes a reference to a list of "unary operators"
1900 to extend the syntax understood by L<SQL::Abstract>.
1901 See section L</"UNARY OPERATORS"> for details.
1907 =head2 insert($table, \@values || \%fieldvals, \%options)
1909 This is the simplest function. You simply give it a table name
1910 and either an arrayref of values or hashref of field/value pairs.
1911 It returns an SQL INSERT statement and a list of bind values.
1912 See the sections on L</"Inserting and Updating Arrays"> and
1913 L</"Inserting and Updating SQL"> for information on how to insert
1914 with those data types.
1916 The optional C<\%options> hash reference may contain additional
1917 options to generate the insert SQL. Currently supported options
1924 Takes either a scalar of raw SQL fields, or an array reference of
1925 field names, and adds on an SQL C<RETURNING> statement at the end.
1926 This allows you to return data generated by the insert statement
1927 (such as row IDs) without performing another C<SELECT> statement.
1928 Note, however, this is not part of the SQL standard and may not
1929 be supported by all database engines.
1933 =head2 update($table, \%fieldvals, \%where, \%options)
1935 This takes a table, hashref of field/value pairs, and an optional
1936 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
1938 See the sections on L</"Inserting and Updating Arrays"> and
1939 L</"Inserting and Updating SQL"> for information on how to insert
1940 with those data types.
1942 The optional C<\%options> hash reference may contain additional
1943 options to generate the update SQL. Currently supported options
1950 See the C<returning> option to
1951 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
1955 =head2 select($source, $fields, $where, $order)
1957 This returns a SQL SELECT statement and associated list of bind values, as
1958 specified by the arguments:
1964 Specification of the 'FROM' part of the statement.
1965 The argument can be either a plain scalar (interpreted as a table
1966 name, will be quoted), or an arrayref (interpreted as a list
1967 of table names, joined by commas, quoted), or a scalarref
1968 (literal SQL, not quoted).
1972 Specification of the list of fields to retrieve from
1974 The argument can be either an arrayref (interpreted as a list
1975 of field names, will be joined by commas and quoted), or a
1976 plain scalar (literal SQL, not quoted).
1977 Please observe that this API is not as flexible as that of
1978 the first argument C<$source>, for backwards compatibility reasons.
1982 Optional argument to specify the WHERE part of the query.
1983 The argument is most often a hashref, but can also be
1984 an arrayref or plain scalar --
1985 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
1989 Optional argument to specify the ORDER BY part of the query.
1990 The argument can be a scalar, a hashref or an arrayref
1991 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
1997 =head2 delete($table, \%where, \%options)
1999 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
2000 It returns an SQL DELETE statement and list of bind values.
2002 The optional C<\%options> hash reference may contain additional
2003 options to generate the delete SQL. Currently supported options
2010 See the C<returning> option to
2011 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2015 =head2 where(\%where, $order)
2017 This is used to generate just the WHERE clause. For example,
2018 if you have an arbitrary data structure and know what the
2019 rest of your SQL is going to look like, but want an easy way
2020 to produce a WHERE clause, use this. It returns an SQL WHERE
2021 clause and list of bind values.
2024 =head2 values(\%data)
2026 This just returns the values from the hash C<%data>, in the same
2027 order that would be returned from any of the other above queries.
2028 Using this allows you to markedly speed up your queries if you
2029 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
2031 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
2033 Warning: This is an experimental method and subject to change.
2035 This returns arbitrarily generated SQL. It's a really basic shortcut.
2036 It will return two different things, depending on return context:
2038 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
2039 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
2041 These would return the following:
2043 # First calling form
2044 $stmt = "CREATE TABLE test (?, ?)";
2045 @bind = (field1, field2);
2047 # Second calling form
2048 $stmt_and_val = "CREATE TABLE test (field1, field2)";
2050 Depending on what you're trying to do, it's up to you to choose the correct
2051 format. In this example, the second form is what you would want.
2055 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
2059 ALTER SESSION SET nls_date_format = 'MM/YY'
2061 You get the idea. Strings get their case twiddled, but everything
2062 else remains verbatim.
2064 =head1 EXPORTABLE FUNCTIONS
2066 =head2 is_plain_value
2068 Determines if the supplied argument is a plain value as understood by this
2073 =item * The value is C<undef>
2075 =item * The value is a non-reference
2077 =item * The value is an object with stringification overloading
2079 =item * The value is of the form C<< { -value => $anything } >>
2083 On failure returns C<undef>, on success returns a B<scalar> reference
2084 to the original supplied argument.
2090 The stringification overloading detection is rather advanced: it takes
2091 into consideration not only the presence of a C<""> overload, but if that
2092 fails also checks for enabled
2093 L<autogenerated versions of C<"">|overload/Magic Autogeneration>, based
2094 on either C<0+> or C<bool>.
2096 Unfortunately testing in the field indicates that this
2097 detection B<< may tickle a latent bug in perl versions before 5.018 >>,
2098 but only when very large numbers of stringifying objects are involved.
2099 At the time of writing ( Sep 2014 ) there is no clear explanation of
2100 the direct cause, nor is there a manageably small test case that reliably
2101 reproduces the problem.
2103 If you encounter any of the following exceptions in B<random places within
2104 your application stack> - this module may be to blame:
2106 Operation "ne": no method found,
2107 left argument in overloaded package <something>,
2108 right argument in overloaded package <something>
2112 Stub found while resolving method "???" overloading """" in package <something>
2114 If you fall victim to the above - please attempt to reduce the problem
2115 to something that could be sent to the L<SQL::Abstract developers
2116 |DBIx::Class/GETTING HELP/SUPPORT>
2117 (either publicly or privately). As a workaround in the meantime you can
2118 set C<$ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}> to a true
2119 value, which will most likely eliminate your problem (at the expense of
2120 not being able to properly detect exotic forms of stringification).
2122 This notice and environment variable will be removed in a future version,
2123 as soon as the underlying problem is found and a reliable workaround is
2128 =head2 is_literal_value
2130 Determines if the supplied argument is a literal value as understood by this
2135 =item * C<\$sql_string>
2137 =item * C<\[ $sql_string, @bind_values ]>
2141 On failure returns C<undef>, on success returns an B<array> reference
2142 containing the unpacked version of the supplied literal SQL and bind values.
2144 =head1 WHERE CLAUSES
2148 This module uses a variation on the idea from L<DBIx::Abstract>. It
2149 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
2150 module is that things in arrays are OR'ed, and things in hashes
2153 The easiest way to explain is to show lots of examples. After
2154 each C<%where> hash shown, it is assumed you used:
2156 my($stmt, @bind) = $sql->where(\%where);
2158 However, note that the C<%where> hash can be used directly in any
2159 of the other functions as well, as described above.
2161 =head2 Key-value pairs
2163 So, let's get started. To begin, a simple hash:
2167 status => 'completed'
2170 Is converted to SQL C<key = val> statements:
2172 $stmt = "WHERE user = ? AND status = ?";
2173 @bind = ('nwiger', 'completed');
2175 One common thing I end up doing is having a list of values that
2176 a field can be in. To do this, simply specify a list inside of
2181 status => ['assigned', 'in-progress', 'pending'];
2184 This simple code will create the following:
2186 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
2187 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
2189 A field associated to an empty arrayref will be considered a
2190 logical false and will generate 0=1.
2192 =head2 Tests for NULL values
2194 If the value part is C<undef> then this is converted to SQL <IS NULL>
2203 $stmt = "WHERE user = ? AND status IS NULL";
2206 To test if a column IS NOT NULL:
2210 status => { '!=', undef },
2213 =head2 Specific comparison operators
2215 If you want to specify a different type of operator for your comparison,
2216 you can use a hashref for a given column:
2220 status => { '!=', 'completed' }
2223 Which would generate:
2225 $stmt = "WHERE user = ? AND status != ?";
2226 @bind = ('nwiger', 'completed');
2228 To test against multiple values, just enclose the values in an arrayref:
2230 status => { '=', ['assigned', 'in-progress', 'pending'] };
2232 Which would give you:
2234 "WHERE status = ? OR status = ? OR status = ?"
2237 The hashref can also contain multiple pairs, in which case it is expanded
2238 into an C<AND> of its elements:
2242 status => { '!=', 'completed', -not_like => 'pending%' }
2245 # Or more dynamically, like from a form
2246 $where{user} = 'nwiger';
2247 $where{status}{'!='} = 'completed';
2248 $where{status}{'-not_like'} = 'pending%';
2250 # Both generate this
2251 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
2252 @bind = ('nwiger', 'completed', 'pending%');
2255 To get an OR instead, you can combine it with the arrayref idea:
2259 priority => [ { '=', 2 }, { '>', 5 } ]
2262 Which would generate:
2264 $stmt = "WHERE ( priority = ? OR priority > ? ) AND user = ?";
2265 @bind = ('2', '5', 'nwiger');
2267 If you want to include literal SQL (with or without bind values), just use a
2268 scalar reference or reference to an arrayref as the value:
2271 date_entered => { '>' => \["to_date(?, 'MM/DD/YYYY')", "11/26/2008"] },
2272 date_expires => { '<' => \"now()" }
2275 Which would generate:
2277 $stmt = "WHERE date_entered > to_date(?, 'MM/DD/YYYY') AND date_expires < now()";
2278 @bind = ('11/26/2008');
2281 =head2 Logic and nesting operators
2283 In the example above,
2284 there is a subtle trap if you want to say something like
2285 this (notice the C<AND>):
2287 WHERE priority != ? AND priority != ?
2289 Because, in Perl you I<can't> do this:
2291 priority => { '!=' => 2, '!=' => 1 }
2293 As the second C<!=> key will obliterate the first. The solution
2294 is to use the special C<-modifier> form inside an arrayref:
2296 priority => [ -and => {'!=', 2},
2300 Normally, these would be joined by C<OR>, but the modifier tells it
2301 to use C<AND> instead. (Hint: You can use this in conjunction with the
2302 C<logic> option to C<new()> in order to change the way your queries
2303 work by default.) B<Important:> Note that the C<-modifier> goes
2304 B<INSIDE> the arrayref, as an extra first element. This will
2305 B<NOT> do what you think it might:
2307 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
2309 Here is a quick list of equivalencies, since there is some overlap:
2312 status => {'!=', 'completed', 'not like', 'pending%' }
2313 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
2316 status => {'=', ['assigned', 'in-progress']}
2317 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
2318 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
2322 =head2 Special operators: IN, BETWEEN, etc.
2324 You can also use the hashref format to compare a list of fields using the
2325 C<IN> comparison operator, by specifying the list as an arrayref:
2328 status => 'completed',
2329 reportid => { -in => [567, 2335, 2] }
2332 Which would generate:
2334 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
2335 @bind = ('completed', '567', '2335', '2');
2337 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
2340 If the argument to C<-in> is an empty array, 'sqlfalse' is generated
2341 (by default: C<1=0>). Similarly, C<< -not_in => [] >> generates
2342 'sqltrue' (by default: C<1=1>).
2344 In addition to the array you can supply a chunk of literal sql or
2345 literal sql with bind:
2348 customer => { -in => \[
2349 'SELECT cust_id FROM cust WHERE balance > ?',
2352 status => { -in => \'SELECT status_codes FROM states' },
2358 customer IN ( SELECT cust_id FROM cust WHERE balance > ? )
2359 AND status IN ( SELECT status_codes FROM states )
2363 Finally, if the argument to C<-in> is not a reference, it will be
2364 treated as a single-element array.
2366 Another pair of operators is C<-between> and C<-not_between>,
2367 used with an arrayref of two values:
2371 completion_date => {
2372 -not_between => ['2002-10-01', '2003-02-06']
2378 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
2380 Just like with C<-in> all plausible combinations of literal SQL
2384 start0 => { -between => [ 1, 2 ] },
2385 start1 => { -between => \["? AND ?", 1, 2] },
2386 start2 => { -between => \"lower(x) AND upper(y)" },
2387 start3 => { -between => [
2389 \["upper(?)", 'stuff' ],
2396 ( start0 BETWEEN ? AND ? )
2397 AND ( start1 BETWEEN ? AND ? )
2398 AND ( start2 BETWEEN lower(x) AND upper(y) )
2399 AND ( start3 BETWEEN lower(x) AND upper(?) )
2401 @bind = (1, 2, 1, 2, 'stuff');
2404 These are the two builtin "special operators"; but the
2405 list can be expanded: see section L</"SPECIAL OPERATORS"> below.
2407 =head2 Unary operators: bool
2409 If you wish to test against boolean columns or functions within your
2410 database you can use the C<-bool> and C<-not_bool> operators. For
2411 example to test the column C<is_user> being true and the column
2412 C<is_enabled> being false you would use:-
2416 -not_bool => 'is_enabled',
2421 WHERE is_user AND NOT is_enabled
2423 If a more complex combination is required, testing more conditions,
2424 then you should use the and/or operators:-
2429 -not_bool => { two=> { -rlike => 'bar' } },
2430 -not_bool => { three => [ { '=', 2 }, { '>', 5 } ] },
2441 (NOT ( three = ? OR three > ? ))
2444 =head2 Nested conditions, -and/-or prefixes
2446 So far, we've seen how multiple conditions are joined with a top-level
2447 C<AND>. We can change this by putting the different conditions we want in
2448 hashes and then putting those hashes in an array. For example:
2453 status => { -like => ['pending%', 'dispatched'] },
2457 status => 'unassigned',
2461 This data structure would create the following:
2463 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
2464 OR ( user = ? AND status = ? ) )";
2465 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
2468 Clauses in hashrefs or arrayrefs can be prefixed with an C<-and> or C<-or>
2469 to change the logic inside:
2475 -and => [ workhrs => {'>', 20}, geo => 'ASIA' ],
2476 -or => { workhrs => {'<', 50}, geo => 'EURO' },
2483 $stmt = "WHERE ( user = ?
2484 AND ( ( workhrs > ? AND geo = ? )
2485 OR ( workhrs < ? OR geo = ? ) ) )";
2486 @bind = ('nwiger', '20', 'ASIA', '50', 'EURO');
2488 =head3 Algebraic inconsistency, for historical reasons
2490 C<Important note>: when connecting several conditions, the C<-and->|C<-or>
2491 operator goes C<outside> of the nested structure; whereas when connecting
2492 several constraints on one column, the C<-and> operator goes
2493 C<inside> the arrayref. Here is an example combining both features:
2496 -and => [a => 1, b => 2],
2497 -or => [c => 3, d => 4],
2498 e => [-and => {-like => 'foo%'}, {-like => '%bar'} ]
2503 WHERE ( ( ( a = ? AND b = ? )
2504 OR ( c = ? OR d = ? )
2505 OR ( e LIKE ? AND e LIKE ? ) ) )
2507 This difference in syntax is unfortunate but must be preserved for
2508 historical reasons. So be careful: the two examples below would
2509 seem algebraically equivalent, but they are not
2512 { -like => 'foo%' },
2513 { -like => '%bar' },
2515 # yields: WHERE ( ( col LIKE ? AND col LIKE ? ) )
2518 { col => { -like => 'foo%' } },
2519 { col => { -like => '%bar' } },
2521 # yields: WHERE ( ( col LIKE ? OR col LIKE ? ) )
2524 =head2 Literal SQL and value type operators
2526 The basic premise of SQL::Abstract is that in WHERE specifications the "left
2527 side" is a column name and the "right side" is a value (normally rendered as
2528 a placeholder). This holds true for both hashrefs and arrayref pairs as you
2529 see in the L</WHERE CLAUSES> examples above. Sometimes it is necessary to
2530 alter this behavior. There are several ways of doing so.
2534 This is a virtual operator that signals the string to its right side is an
2535 identifier (a column name) and not a value. For example to compare two
2536 columns you would write:
2539 priority => { '<', 2 },
2540 requestor => { -ident => 'submitter' },
2545 $stmt = "WHERE priority < ? AND requestor = submitter";
2548 If you are maintaining legacy code you may see a different construct as
2549 described in L</Deprecated usage of Literal SQL>, please use C<-ident> in new
2554 This is a virtual operator that signals that the construct to its right side
2555 is a value to be passed to DBI. This is for example necessary when you want
2556 to write a where clause against an array (for RDBMS that support such
2557 datatypes). For example:
2560 array => { -value => [1, 2, 3] }
2565 $stmt = 'WHERE array = ?';
2566 @bind = ([1, 2, 3]);
2568 Note that if you were to simply say:
2574 the result would probably not be what you wanted:
2576 $stmt = 'WHERE array = ? OR array = ? OR array = ?';
2581 Finally, sometimes only literal SQL will do. To include a random snippet
2582 of SQL verbatim, you specify it as a scalar reference. Consider this only
2583 as a last resort. Usually there is a better way. For example:
2586 priority => { '<', 2 },
2587 requestor => { -in => \'(SELECT name FROM hitmen)' },
2592 $stmt = "WHERE priority < ? AND requestor IN (SELECT name FROM hitmen)"
2595 Note that in this example, you only get one bind parameter back, since
2596 the verbatim SQL is passed as part of the statement.
2600 Never use untrusted input as a literal SQL argument - this is a massive
2601 security risk (there is no way to check literal snippets for SQL
2602 injections and other nastyness). If you need to deal with untrusted input
2603 use literal SQL with placeholders as described next.
2605 =head3 Literal SQL with placeholders and bind values (subqueries)
2607 If the literal SQL to be inserted has placeholders and bind values,
2608 use a reference to an arrayref (yes this is a double reference --
2609 not so common, but perfectly legal Perl). For example, to find a date
2610 in Postgres you can use something like this:
2613 date_column => \[ "= date '2008-09-30' - ?::integer", 10 ]
2618 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
2621 Note that you must pass the bind values in the same format as they are returned
2622 by L<where|/where(\%where, $order)>. This means that if you set L</bindtype>
2623 to C<columns>, you must provide the bind values in the
2624 C<< [ column_meta => value ] >> format, where C<column_meta> is an opaque
2625 scalar value; most commonly the column name, but you can use any scalar value
2626 (including references and blessed references), L<SQL::Abstract> will simply
2627 pass it through intact. So if C<bindtype> is set to C<columns> the above
2628 example will look like:
2631 date_column => \[ "= date '2008-09-30' - ?::integer", [ {} => 10 ] ]
2634 Literal SQL is especially useful for nesting parenthesized clauses in the
2635 main SQL query. Here is a first example:
2637 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
2641 bar => \["IN ($sub_stmt)" => @sub_bind],
2646 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
2647 WHERE c2 < ? AND c3 LIKE ?))";
2648 @bind = (1234, 100, "foo%");
2650 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
2651 are expressed in the same way. Of course the C<$sub_stmt> and
2652 its associated bind values can be generated through a former call
2655 my ($sub_stmt, @sub_bind)
2656 = $sql->select("t1", "c1", {c2 => {"<" => 100},
2657 c3 => {-like => "foo%"}});
2660 bar => \["> ALL ($sub_stmt)" => @sub_bind],
2663 In the examples above, the subquery was used as an operator on a column;
2664 but the same principle also applies for a clause within the main C<%where>
2665 hash, like an EXISTS subquery:
2667 my ($sub_stmt, @sub_bind)
2668 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
2669 my %where = ( -and => [
2671 \["EXISTS ($sub_stmt)" => @sub_bind],
2676 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
2677 WHERE c1 = ? AND c2 > t0.c0))";
2681 Observe that the condition on C<c2> in the subquery refers to
2682 column C<t0.c0> of the main query: this is I<not> a bind
2683 value, so we have to express it through a scalar ref.
2684 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
2685 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
2686 what we wanted here.
2688 Finally, here is an example where a subquery is used
2689 for expressing unary negation:
2691 my ($sub_stmt, @sub_bind)
2692 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
2693 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
2695 lname => {like => '%son%'},
2696 \["NOT ($sub_stmt)" => @sub_bind],
2701 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
2702 @bind = ('%son%', 10, 20)
2704 =head3 Deprecated usage of Literal SQL
2706 Below are some examples of archaic use of literal SQL. It is shown only as
2707 reference for those who deal with legacy code. Each example has a much
2708 better, cleaner and safer alternative that users should opt for in new code.
2714 my %where = ( requestor => \'IS NOT NULL' )
2716 $stmt = "WHERE requestor IS NOT NULL"
2718 This used to be the way of generating NULL comparisons, before the handling
2719 of C<undef> got formalized. For new code please use the superior syntax as
2720 described in L</Tests for NULL values>.
2724 my %where = ( requestor => \'= submitter' )
2726 $stmt = "WHERE requestor = submitter"
2728 This used to be the only way to compare columns. Use the superior L</-ident>
2729 method for all new code. For example an identifier declared in such a way
2730 will be properly quoted if L</quote_char> is properly set, while the legacy
2731 form will remain as supplied.
2735 my %where = ( is_ready => \"", completed => { '>', '2012-12-21' } )
2737 $stmt = "WHERE completed > ? AND is_ready"
2738 @bind = ('2012-12-21')
2740 Using an empty string literal used to be the only way to express a boolean.
2741 For all new code please use the much more readable
2742 L<-bool|/Unary operators: bool> operator.
2748 These pages could go on for a while, since the nesting of the data
2749 structures this module can handle are pretty much unlimited (the
2750 module implements the C<WHERE> expansion as a recursive function
2751 internally). Your best bet is to "play around" with the module a
2752 little to see how the data structures behave, and choose the best
2753 format for your data based on that.
2755 And of course, all the values above will probably be replaced with
2756 variables gotten from forms or the command line. After all, if you
2757 knew everything ahead of time, you wouldn't have to worry about
2758 dynamically-generating SQL and could just hardwire it into your
2761 =head1 ORDER BY CLAUSES
2763 Some functions take an order by clause. This can either be a scalar (just a
2764 column name), a hashref of C<< { -desc => 'col' } >> or C<< { -asc => 'col' }
2765 >>, a scalarref, an arrayref-ref, or an arrayref of any of the previous
2768 Given | Will Generate
2769 ---------------------------------------------------------------
2771 'colA' | ORDER BY colA
2773 [qw/colA colB/] | ORDER BY colA, colB
2775 {-asc => 'colA'} | ORDER BY colA ASC
2777 {-desc => 'colB'} | ORDER BY colB DESC
2779 ['colA', {-asc => 'colB'}] | ORDER BY colA, colB ASC
2781 { -asc => [qw/colA colB/] } | ORDER BY colA ASC, colB ASC
2783 \'colA DESC' | ORDER BY colA DESC
2785 \[ 'FUNC(colA, ?)', $x ] | ORDER BY FUNC(colA, ?)
2786 | /* ...with $x bound to ? */
2789 { -asc => 'colA' }, | colA ASC,
2790 { -desc => [qw/colB/] }, | colB DESC,
2791 { -asc => [qw/colC colD/] },| colC ASC, colD ASC,
2792 \'colE DESC', | colE DESC,
2793 \[ 'FUNC(colF, ?)', $x ], | FUNC(colF, ?)
2794 ] | /* ...with $x bound to ? */
2795 ===============================================================
2799 =head1 SPECIAL OPERATORS
2801 my $sqlmaker = SQL::Abstract->new(special_ops => [
2805 my ($self, $field, $op, $arg) = @_;
2811 handler => 'method_name',
2815 A "special operator" is a SQL syntactic clause that can be
2816 applied to a field, instead of a usual binary operator.
2819 WHERE field IN (?, ?, ?)
2820 WHERE field BETWEEN ? AND ?
2821 WHERE MATCH(field) AGAINST (?, ?)
2823 Special operators IN and BETWEEN are fairly standard and therefore
2824 are builtin within C<SQL::Abstract> (as the overridable methods
2825 C<_where_field_IN> and C<_where_field_BETWEEN>). For other operators,
2826 like the MATCH .. AGAINST example above which is specific to MySQL,
2827 you can write your own operator handlers - supply a C<special_ops>
2828 argument to the C<new> method. That argument takes an arrayref of
2829 operator definitions; each operator definition is a hashref with two
2836 the regular expression to match the operator
2840 Either a coderef or a plain scalar method name. In both cases
2841 the expected return is C<< ($sql, @bind) >>.
2843 When supplied with a method name, it is simply called on the
2844 L<SQL::Abstract> object as:
2846 $self->$method_name($field, $op, $arg)
2850 $field is the LHS of the operator
2851 $op is the part that matched the handler regex
2854 When supplied with a coderef, it is called as:
2856 $coderef->($self, $field, $op, $arg)
2861 For example, here is an implementation
2862 of the MATCH .. AGAINST syntax for MySQL
2864 my $sqlmaker = SQL::Abstract->new(special_ops => [
2866 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
2867 {regex => qr/^match$/i,
2869 my ($self, $field, $op, $arg) = @_;
2870 $arg = [$arg] if not ref $arg;
2871 my $label = $self->_quote($field);
2872 my ($placeholder) = $self->_convert('?');
2873 my $placeholders = join ", ", (($placeholder) x @$arg);
2874 my $sql = $self->_sqlcase('match') . " ($label) "
2875 . $self->_sqlcase('against') . " ($placeholders) ";
2876 my @bind = $self->_bindtype($field, @$arg);
2877 return ($sql, @bind);
2884 =head1 UNARY OPERATORS
2886 my $sqlmaker = SQL::Abstract->new(unary_ops => [
2890 my ($self, $op, $arg) = @_;
2896 handler => 'method_name',
2900 A "unary operator" is a SQL syntactic clause that can be
2901 applied to a field - the operator goes before the field
2903 You can write your own operator handlers - supply a C<unary_ops>
2904 argument to the C<new> method. That argument takes an arrayref of
2905 operator definitions; each operator definition is a hashref with two
2912 the regular expression to match the operator
2916 Either a coderef or a plain scalar method name. In both cases
2917 the expected return is C<< $sql >>.
2919 When supplied with a method name, it is simply called on the
2920 L<SQL::Abstract> object as:
2922 $self->$method_name($op, $arg)
2926 $op is the part that matched the handler regex
2927 $arg is the RHS or argument of the operator
2929 When supplied with a coderef, it is called as:
2931 $coderef->($self, $op, $arg)
2939 Thanks to some benchmarking by Mark Stosberg, it turns out that
2940 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
2941 I must admit this wasn't an intentional design issue, but it's a
2942 byproduct of the fact that you get to control your C<DBI> handles
2945 To maximize performance, use a code snippet like the following:
2947 # prepare a statement handle using the first row
2948 # and then reuse it for the rest of the rows
2950 for my $href (@array_of_hashrefs) {
2951 $stmt ||= $sql->insert('table', $href);
2952 $sth ||= $dbh->prepare($stmt);
2953 $sth->execute($sql->values($href));
2956 The reason this works is because the keys in your C<$href> are sorted
2957 internally by B<SQL::Abstract>. Thus, as long as your data retains
2958 the same structure, you only have to generate the SQL the first time
2959 around. On subsequent queries, simply use the C<values> function provided
2960 by this module to return your values in the correct order.
2962 However this depends on the values having the same type - if, for
2963 example, the values of a where clause may either have values
2964 (resulting in sql of the form C<column = ?> with a single bind
2965 value), or alternatively the values might be C<undef> (resulting in
2966 sql of the form C<column IS NULL> with no bind value) then the
2967 caching technique suggested will not work.
2971 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
2972 really like this part (I do, at least). Building up a complex query
2973 can be as simple as the following:
2980 use CGI::FormBuilder;
2983 my $form = CGI::FormBuilder->new(...);
2984 my $sql = SQL::Abstract->new;
2986 if ($form->submitted) {
2987 my $field = $form->field;
2988 my $id = delete $field->{id};
2989 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
2992 Of course, you would still have to connect using C<DBI> to run the
2993 query, but the point is that if you make your form look like your
2994 table, the actual query script can be extremely simplistic.
2996 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
2997 a fast interface to returning and formatting data. I frequently
2998 use these three modules together to write complex database query
2999 apps in under 50 lines.
3001 =head1 HOW TO CONTRIBUTE
3003 Contributions are always welcome, in all usable forms (we especially
3004 welcome documentation improvements). The delivery methods include git-
3005 or unified-diff formatted patches, GitHub pull requests, or plain bug
3006 reports either via RT or the Mailing list. Contributors are generally
3007 granted full access to the official repository after their first several
3008 patches pass successful review.
3010 This project is maintained in a git repository. The code and related tools are
3011 accessible at the following locations:
3015 =item * Official repo: L<git://git.shadowcat.co.uk/dbsrgits/SQL-Abstract.git>
3017 =item * Official gitweb: L<http://git.shadowcat.co.uk/gitweb/gitweb.cgi?p=dbsrgits/SQL-Abstract.git>
3019 =item * GitHub mirror: L<https://github.com/dbsrgits/sql-abstract>
3021 =item * Authorized committers: L<ssh://dbsrgits@git.shadowcat.co.uk/SQL-Abstract.git>
3027 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
3028 Great care has been taken to preserve the I<published> behavior
3029 documented in previous versions in the 1.* family; however,
3030 some features that were previously undocumented, or behaved
3031 differently from the documentation, had to be changed in order
3032 to clarify the semantics. Hence, client code that was relying
3033 on some dark areas of C<SQL::Abstract> v1.*
3034 B<might behave differently> in v1.50.
3036 The main changes are:
3042 support for literal SQL through the C<< \ [ $sql, @bind ] >> syntax.
3046 support for the { operator => \"..." } construct (to embed literal SQL)
3050 support for the { operator => \["...", @bind] } construct (to embed literal SQL with bind values)
3054 optional support for L<array datatypes|/"Inserting and Updating Arrays">
3058 defensive programming: check arguments
3062 fixed bug with global logic, which was previously implemented
3063 through global variables yielding side-effects. Prior versions would
3064 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
3065 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
3066 Now this is interpreted
3067 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
3072 fixed semantics of _bindtype on array args
3076 dropped the C<_anoncopy> of the %where tree. No longer necessary,
3077 we just avoid shifting arrays within that tree.
3081 dropped the C<_modlogic> function
3085 =head1 ACKNOWLEDGEMENTS
3087 There are a number of individuals that have really helped out with
3088 this module. Unfortunately, most of them submitted bugs via CPAN
3089 so I have no idea who they are! But the people I do know are:
3091 Ash Berlin (order_by hash term support)
3092 Matt Trout (DBIx::Class support)
3093 Mark Stosberg (benchmarking)
3094 Chas Owens (initial "IN" operator support)
3095 Philip Collins (per-field SQL functions)
3096 Eric Kolve (hashref "AND" support)
3097 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
3098 Dan Kubb (support for "quote_char" and "name_sep")
3099 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
3100 Laurent Dami (internal refactoring, extensible list of special operators, literal SQL)
3101 Norbert Buchmuller (support for literal SQL in hashpair, misc. fixes & tests)
3102 Peter Rabbitson (rewrite of SQLA::Test, misc. fixes & tests)
3103 Oliver Charles (support for "RETURNING" after "INSERT")
3109 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
3113 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
3115 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
3117 For support, your best bet is to try the C<DBIx::Class> users mailing list.
3118 While not an official support venue, C<DBIx::Class> makes heavy use of
3119 C<SQL::Abstract>, and as such list members there are very familiar with
3120 how to create queries.
3124 This module is free software; you may copy this under the same
3125 terms as perl itself (either the GNU General Public License or
3126 the Artistic License)