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/
187 $opt{node_types} = +{
188 map +("-$_" => '_render_'.$_),
189 qw(op func bind ident literal list)
192 $opt{expand_unary} = {};
194 return bless \%opt, $class;
197 sub sqltrue { +{ -literal => [ $_[0]->{sqltrue} ] } }
198 sub sqlfalse { +{ -literal => [ $_[0]->{sqlfalse} ] } }
200 sub _assert_pass_injection_guard {
201 if ($_[1] =~ $_[0]->{injection_guard}) {
202 my $class = ref $_[0];
203 puke "Possible SQL injection attempt '$_[1]'. If this is indeed a part of the "
204 . "desired SQL use literal SQL ( \'...' or \[ '...' ] ) or supply your own "
205 . "{injection_guard} attribute to ${class}->new()"
210 #======================================================================
212 #======================================================================
216 my $table = $self->_table(shift);
217 my $data = shift || return;
220 my $method = $self->_METHOD_FOR_refkind("_insert", $data);
221 my ($sql, @bind) = $self->$method($data);
222 $sql = join " ", $self->_sqlcase('insert into'), $table, $sql;
224 if ($options->{returning}) {
225 my ($s, @b) = $self->_insert_returning($options);
230 return wantarray ? ($sql, @bind) : $sql;
233 # So that subclasses can override INSERT ... RETURNING separately from
234 # UPDATE and DELETE (e.g. DBIx::Class::SQLMaker::Oracle does this)
235 sub _insert_returning { shift->_returning(@_) }
238 my ($self, $options) = @_;
240 my $f = $options->{returning};
242 my ($sql, @bind) = $self->render_aqt(
243 $self->_expand_maybe_list_expr($f, undef, -ident)
246 ? $self->_sqlcase(' returning ') . $sql
247 : ($self->_sqlcase(' returning ').$sql, @bind);
250 sub _insert_HASHREF { # explicit list of fields and then values
251 my ($self, $data) = @_;
253 my @fields = sort keys %$data;
255 my ($sql, @bind) = $self->_insert_values($data);
258 $_ = $self->_quote($_) foreach @fields;
259 $sql = "( ".join(", ", @fields).") ".$sql;
261 return ($sql, @bind);
264 sub _insert_ARRAYREF { # just generate values(?,?) part (no list of fields)
265 my ($self, $data) = @_;
267 # no names (arrayref) so can't generate bindtype
268 $self->{bindtype} ne 'columns'
269 or belch "can't do 'columns' bindtype when called with arrayref";
271 my (@values, @all_bind);
272 foreach my $value (@$data) {
273 my ($values, @bind) = $self->_insert_value(undef, $value);
274 push @values, $values;
275 push @all_bind, @bind;
277 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
278 return ($sql, @all_bind);
281 sub _insert_ARRAYREFREF { # literal SQL with bind
282 my ($self, $data) = @_;
284 my ($sql, @bind) = @${$data};
285 $self->_assert_bindval_matches_bindtype(@bind);
287 return ($sql, @bind);
291 sub _insert_SCALARREF { # literal SQL without bind
292 my ($self, $data) = @_;
298 my ($self, $data) = @_;
300 my (@values, @all_bind);
301 foreach my $column (sort keys %$data) {
302 my ($values, @bind) = $self->_insert_value($column, $data->{$column});
303 push @values, $values;
304 push @all_bind, @bind;
306 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
307 return ($sql, @all_bind);
311 my ($self, $column, $v) = @_;
313 return $self->render_aqt(
314 $self->_expand_insert_value($column, $v)
318 sub _expand_insert_value {
319 my ($self, $column, $v) = @_;
321 if (ref($v) eq 'ARRAY') {
322 if ($self->{array_datatypes}) {
323 return +{ -bind => [ $column, $v ] };
325 my ($sql, @bind) = @$v;
326 $self->_assert_bindval_matches_bindtype(@bind);
327 return +{ -literal => $v };
329 if (ref($v) eq 'HASH') {
330 if (grep !/^-/, keys %$v) {
331 belch "HASH ref as bind value in insert is not supported";
332 return +{ -bind => [ $column, $v ] };
336 return +{ -bind => [ $column, undef ] };
338 local our $Cur_Col_Meta = $column;
339 return $self->expand_expr($v);
344 #======================================================================
346 #======================================================================
351 my $table = $self->_table(shift);
352 my $data = shift || return;
356 # first build the 'SET' part of the sql statement
357 puke "Unsupported data type specified to \$sql->update"
358 unless ref $data eq 'HASH';
360 my ($sql, @all_bind) = $self->_update_set_values($data);
361 $sql = $self->_sqlcase('update ') . $table . $self->_sqlcase(' set ')
365 my($where_sql, @where_bind) = $self->where($where);
367 push @all_bind, @where_bind;
370 if ($options->{returning}) {
371 my ($returning_sql, @returning_bind) = $self->_update_returning($options);
372 $sql .= $returning_sql;
373 push @all_bind, @returning_bind;
376 return wantarray ? ($sql, @all_bind) : $sql;
379 sub _update_set_values {
380 my ($self, $data) = @_;
382 return $self->render_aqt(
383 $self->_expand_update_set_values($data),
387 sub _expand_update_set_values {
388 my ($self, $data) = @_;
389 $self->_expand_maybe_list_expr( [
392 $set = { -bind => $_ } unless defined $set;
393 +{ -op => [ '=', { -ident => $k }, $set ] };
399 ? ($self->{array_datatypes}
400 ? [ $k, +{ -bind => [ $k, $v ] } ]
401 : [ $k, +{ -literal => $v } ])
403 local our $Cur_Col_Meta = $k;
404 [ $k, $self->_expand_expr($v) ]
411 # So that subclasses can override UPDATE ... RETURNING separately from
413 sub _update_returning { shift->_returning(@_) }
417 #======================================================================
419 #======================================================================
424 my $table = $self->_table(shift);
425 my $fields = shift || '*';
429 my ($fields_sql, @bind) = $self->_select_fields($fields);
431 my ($where_sql, @where_bind) = $self->where($where, $order);
432 push @bind, @where_bind;
434 my $sql = join(' ', $self->_sqlcase('select'), $fields_sql,
435 $self->_sqlcase('from'), $table)
438 return wantarray ? ($sql, @bind) : $sql;
442 my ($self, $fields) = @_;
443 return $fields unless ref($fields);
444 return $self->render_aqt(
445 $self->_expand_maybe_list_expr($fields, undef, '-ident')
449 #======================================================================
451 #======================================================================
456 my $table = $self->_table(shift);
460 my($where_sql, @bind) = $self->where($where);
461 my $sql = $self->_sqlcase('delete from ') . $table . $where_sql;
463 if ($options->{returning}) {
464 my ($returning_sql, @returning_bind) = $self->_delete_returning($options);
465 $sql .= $returning_sql;
466 push @bind, @returning_bind;
469 return wantarray ? ($sql, @bind) : $sql;
472 # So that subclasses can override DELETE ... RETURNING separately from
474 sub _delete_returning { shift->_returning(@_) }
478 #======================================================================
480 #======================================================================
484 # Finally, a separate routine just to handle WHERE clauses
486 my ($self, $where, $order) = @_;
488 local $self->{convert_where} = $self->{convert};
491 my ($sql, @bind) = defined($where)
492 ? $self->_recurse_where($where)
494 $sql = (defined $sql and length $sql) ? $self->_sqlcase(' where ') . "( $sql )" : '';
498 my ($order_sql, @order_bind) = $self->_order_by($order);
500 push @bind, @order_bind;
503 return wantarray ? ($sql, @bind) : $sql;
507 my ($self, $expr, $default) = @_;
508 $self->_expand_expr($expr, undef, $default);
512 my ($self, $aqt) = @_;
513 my ($k, $v, @rest) = %$aqt;
515 if (my $meth = $self->{node_types}{$k}) {
516 return $self->$meth($v);
518 die "notreached: $k";
522 my ($self, $expr, $logic, $default_scalar_to) = @_;
523 local our $Default_Scalar_To = $default_scalar_to if $default_scalar_to;
524 our $Expand_Depth ||= 0; local $Expand_Depth = $Expand_Depth + 1;
525 return undef unless defined($expr);
526 if (ref($expr) eq 'HASH') {
527 if (keys %$expr > 1) {
531 map $self->_expand_expr_hashpair($_ => $expr->{$_}, $logic),
535 return undef unless keys %$expr;
536 return $self->_expand_expr_hashpair(%$expr, $logic);
538 if (ref($expr) eq 'ARRAY') {
539 my $logic = lc($logic || $self->{logic});
540 $logic eq 'and' or $logic eq 'or' or puke "unknown logic: $logic";
543 (ref($_) eq 'ARRAY' and @$_)
544 or (ref($_) eq 'HASH' and %$_)
550 while (my ($el) = splice @expr, 0, 1) {
551 puke "Supplying an empty left hand side argument is not supported in array-pairs"
552 unless defined($el) and length($el);
553 my $elref = ref($el);
555 local $Expand_Depth = 0;
556 push(@res, grep defined, $self->_expand_expr({ $el, shift(@expr) }));
557 } elsif ($elref eq 'ARRAY') {
558 push(@res, grep defined, $self->_expand_expr($el)) if @$el;
559 } elsif (my $l = is_literal_value($el)) {
560 push @res, { -literal => $l };
561 } elsif ($elref eq 'HASH') {
562 local $Expand_Depth = 0;
563 push @res, grep defined, $self->_expand_expr($el) if %$el;
568 return { -op => [ $logic, @res ] };
570 if (my $literal = is_literal_value($expr)) {
571 return +{ -literal => $literal };
573 if (!ref($expr) or Scalar::Util::blessed($expr)) {
574 if (my $d = $Default_Scalar_To) {
575 return $self->_expand_expr({ $d => $expr });
577 if (my $m = our $Cur_Col_Meta) {
578 return +{ -bind => [ $m, $expr ] };
580 return +{ -bind => [ undef, $expr ] };
585 sub _expand_expr_hashpair {
586 my ($self, $k, $v, $logic) = @_;
587 unless (defined($k) and length($k)) {
588 if (defined($k) and my $literal = is_literal_value($v)) {
589 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
590 return { -literal => $literal };
592 puke "Supplying an empty left hand side argument is not supported";
595 $self->_assert_pass_injection_guard($k =~ /^-(.*)$/s);
596 if ($k =~ s/ [_\s]? \d+ $//x ) {
597 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
598 . "You probably wanted ...-and => [ $k => COND1, $k => COND2 ... ]";
601 # DBIx::Class requires a nest warning to be emitted once but the private
602 # method it overrode to do so no longer exists
603 if ($self->{is_dbic_sqlmaker}) {
604 unless (our $Nest_Warned) {
606 "-nest in search conditions is deprecated, you most probably wanted:\n"
607 .q|{..., -and => [ \%cond0, \@cond1, \'cond2', \[ 'cond3', [ col => bind ] ], etc. ], ... }|
612 return $self->_expand_expr($v);
616 return $self->_expand_expr($v);
618 puke "-bool => undef not supported" unless defined($v);
619 return { -ident => $v };
622 return { -op => [ 'not', $self->_expand_expr($v) ] };
624 if (my ($rest) = $k =~/^-not[_ ](.*)$/) {
627 $self->_expand_expr_hashpair("-${rest}", $v, $logic)
630 if (my ($logic) = $k =~ /^-(and|or)$/i) {
631 if (ref($v) eq 'HASH') {
632 return $self->_expand_expr($v, $logic);
634 if (ref($v) eq 'ARRAY') {
635 return $self->_expand_expr($v, $logic);
640 $op =~ s/^-// if length($op) > 1;
642 # top level special ops are illegal in general
643 # note that, arguably, if it makes no sense at top level, it also
644 # makes no sense on the other side of an = sign or similar but DBIC
645 # gets disappointingly upset if I disallow it
647 (our $Expand_Depth) == 1
648 and List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}
650 puke "Illegal use of top-level '-$op'"
652 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
653 return { -op => [ $op, $v ] };
656 if ($k eq '-value') {
657 return +{ -bind => [ our $Cur_Col_Meta, $v ] };
659 if (my $custom = $self->{expand_unary}{$k}) {
660 return $self->$custom($v);
662 if ($self->{node_types}{$k}) {
668 and (keys %$v)[0] =~ /^-/
670 my ($func) = $k =~ /^-(.*)$/;
671 if (List::Util::first { $func =~ $_->{regex} } @{$self->{special_ops}}) {
672 return +{ -op => [ $func, $self->_expand_expr($v) ] };
674 return +{ -func => [ $func, $self->_expand_expr($v) ] };
676 if (!ref($v) or is_literal_value($v)) {
677 return +{ -op => [ $k =~ /^-(.*)$/, $self->_expand_expr($v) ] };
684 and exists $v->{-value}
685 and not defined $v->{-value}
688 return $self->_expand_expr_hashpair($k => { $self->{cmp} => undef });
690 if (!ref($v) or Scalar::Util::blessed($v)) {
691 my $d = our $Default_Scalar_To;
696 ($d ? { $d => $v } : { -bind => [ $k, $v ] })
700 if (ref($v) eq 'HASH') {
704 map $self->_expand_expr_hashpair($k => { $_ => $v->{$_} }),
708 return undef unless keys %$v;
712 $self->_assert_pass_injection_guard($vk);
713 if ($vk =~ s/ [_\s]? \d+ $//x ) {
714 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
715 . "You probably wanted ...-and => [ -$vk => COND1, -$vk => COND2 ... ]";
717 if ($vk =~ /^(?:not[ _])?between$/) {
718 local our $Cur_Col_Meta = $k;
719 my @rhs = map $self->_expand_expr($_),
720 ref($vv) eq 'ARRAY' ? @$vv : $vv;
722 (@rhs == 1 and ref($rhs[0]) eq 'HASH' and $rhs[0]->{-literal})
724 (@rhs == 2 and defined($rhs[0]) and defined($rhs[1]))
726 puke "Operator '${\uc($vk)}' requires either an arrayref with two defined values or expressions, or a single literal scalarref/arrayref-ref";
729 join(' ', split '_', $vk),
734 if ($vk =~ /^(?:not[ _])?in$/) {
735 if (my $literal = is_literal_value($vv)) {
736 my ($sql, @bind) = @$literal;
737 my $opened_sql = $self->_open_outer_paren($sql);
739 $vk, { -ident => $k },
740 [ { -literal => [ $opened_sql, @bind ] } ]
744 'SQL::Abstract before v1.75 used to generate incorrect SQL when the '
745 . "-${\uc($vk)} operator was given an undef-containing list: !!!AUDIT YOUR CODE "
746 . 'AND DATA!!! (the upcoming Data::Query-based version of SQL::Abstract '
747 . 'will emit the logically correct SQL instead of raising this exception)'
749 puke("Argument passed to the '${\uc($vk)}' operator can not be undefined")
751 my @rhs = map $self->_expand_expr($_),
752 map { ref($_) ? $_ : { -bind => [ $k, $_ ] } }
753 map { defined($_) ? $_: puke($undef_err) }
754 (ref($vv) eq 'ARRAY' ? @$vv : $vv);
755 return $self->${\($vk =~ /^not/ ? 'sqltrue' : 'sqlfalse')} unless @rhs;
758 join(' ', split '_', $vk),
763 if ($vk eq 'ident') {
764 if (! defined $vv or (ref($vv) and ref($vv) eq 'ARRAY')) {
765 puke "-$vk requires a single plain scalar argument (a quotable identifier) or an arrayref of identifier parts";
773 if ($vk eq 'value') {
774 return $self->_expand_expr_hashpair($k, undef) unless defined($vv);
778 { -bind => [ $k, $vv ] }
781 if ($vk =~ /^is(?:[ _]not)?$/) {
782 puke "$vk can only take undef as argument"
786 and exists($vv->{-value})
787 and !defined($vv->{-value})
790 return +{ -op => [ $vk.' null', { -ident => $k } ] };
792 if ($vk =~ /^(and|or)$/) {
793 if (ref($vv) eq 'HASH') {
796 map $self->_expand_expr_hashpair($k, { $_ => $vv->{$_} }),
801 if (my $us = List::Util::first { $vk =~ $_->{regex} } @{$self->{special_ops}}) {
802 return { -op => [ $vk, { -ident => $k }, $vv ] };
804 if (my $us = List::Util::first { $vk =~ $_->{regex} } @{$self->{unary_ops}}) {
808 { -op => [ $vk, $vv ] }
811 if (ref($vv) eq 'ARRAY') {
812 my ($logic, @values) = (
813 (defined($vv->[0]) and $vv->[0] =~ /^-(and|or)$/i)
818 $vk =~ $self->{inequality_op}
819 or join(' ', split '_', $vk) =~ $self->{not_like_op}
821 if (lc($logic) eq '-or' and @values > 1) {
822 my $op = uc join ' ', split '_', $vk;
823 belch "A multi-element arrayref as an argument to the inequality op '$op' "
824 . 'is technically equivalent to an always-true 1=1 (you probably wanted '
825 . "to say ...{ \$inequality_op => [ -and => \@values ] }... instead)"
830 # try to DWIM on equality operators
831 my $op = join ' ', split '_', $vk;
833 $op =~ $self->{equality_op} ? $self->sqlfalse
834 : $op =~ $self->{like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->sqlfalse
835 : $op =~ $self->{inequality_op} ? $self->sqltrue
836 : $op =~ $self->{not_like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->sqltrue
837 : puke "operator '$op' applied on an empty array (field '$k')";
841 map $self->_expand_expr_hashpair($k => { $vk => $_ }),
849 and exists $vv->{-value}
850 and not defined $vv->{-value}
853 my $op = join ' ', split '_', $vk;
855 $op =~ /^not$/i ? 'is not' # legacy
856 : $op =~ $self->{equality_op} ? 'is'
857 : $op =~ $self->{like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is'
858 : $op =~ $self->{inequality_op} ? 'is not'
859 : $op =~ $self->{not_like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is not'
860 : puke "unexpected operator '$op' with undef operand";
861 return +{ -op => [ $is.' null', { -ident => $k } ] };
863 local our $Cur_Col_Meta = $k;
867 $self->_expand_expr($vv)
870 if (ref($v) eq 'ARRAY') {
871 return $self->sqlfalse unless @$v;
872 $self->_debug("ARRAY($k) means distribute over elements");
874 $v->[0] =~ /^-((?:and|or))$/i
875 ? ($v = [ @{$v}[1..$#$v] ], $1)
876 : ($self->{logic} || 'or')
880 map $self->_expand_expr({ $k => $_ }, $this_logic), @$v
883 if (my $literal = is_literal_value($v)) {
885 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
888 my ($sql, @bind) = @$literal;
889 if ($self->{bindtype} eq 'columns') {
891 $self->_assert_bindval_matches_bindtype($_);
894 return +{ -literal => [ $self->_quote($k).' '.$sql, @bind ] };
900 my ($self, $where, $logic) = @_;
902 # Special case: top level simple string treated as literal
904 my $where_exp = (ref($where)
905 ? $self->expand_expr($where, $logic)
906 : { -literal => [ $where ] });
908 # dispatch expanded expression
910 my ($sql, @bind) = defined($where_exp) ? $self->render_aqt($where_exp) : (undef);
911 # DBIx::Class used to call _recurse_where in scalar context
912 # something else might too...
914 return ($sql, @bind);
917 belch "Calling _recurse_where in scalar context is deprecated and will go away before 2.0";
923 my ($self, $ident) = @_;
925 return $self->_convert($self->_quote($ident));
928 my %unop_postfix = map +($_ => 1),
929 'is null', 'is not null',
937 my ($self, $args) = @_;
938 my ($left, $low, $high) = @$args;
939 my ($rhsql, @rhbind) = do {
941 puke "Single arg to between must be a literal"
942 unless $low->{-literal};
945 my ($l, $h) = map [ $self->render_aqt($_) ], $low, $high;
946 (join(' ', $l->[0], $self->_sqlcase('and'), $h->[0]),
947 @{$l}[1..$#$l], @{$h}[1..$#$h])
950 my ($lhsql, @lhbind) = $self->render_aqt($left);
952 join(' ', '(', $lhsql, $self->_sqlcase($op), $rhsql, ')'),
956 }), 'between', 'not between'),
960 my ($self, $args) = @_;
961 my ($lhs, $rhs) = @$args;
964 my ($sql, @bind) = $self->render_aqt($_);
965 push @in_bind, @bind;
968 my ($lhsql, @lbind) = $self->render_aqt($lhs);
970 $lhsql.' '.$self->_sqlcase($op).' ( '
981 my ($op, @args) = @$v;
982 $op =~ s/^-// if length($op) > 1;
984 if (my $h = $special{$op}) {
985 return $self->$h(\@args);
987 my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}};
988 if ($us and @args > 1) {
989 puke "Special op '${op}' requires first value to be identifier"
990 unless my ($k) = map $_->{-ident}, grep ref($_) eq 'HASH', $args[0];
991 local our $Expand_Depth = 1;
992 return $self->${\($us->{handler})}($k, $op, $args[1]);
994 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
995 return $self->${\($us->{handler})}($op, $args[0]);
997 my $final_op = $op =~ /^(?:is|not)_/ ? join(' ', split '_', $op) : $op;
998 if (@args == 1 and $op !~ /^(and|or)$/) {
999 my ($expr_sql, @bind) = $self->render_aqt($args[0]);
1000 my $op_sql = $self->_sqlcase($final_op);
1002 $unop_postfix{lc($final_op)}
1003 ? "${expr_sql} ${op_sql}"
1004 : "${op_sql} ${expr_sql}"
1006 return (($op eq 'not' || $us ? '('.$final_sql.')' : $final_sql), @bind);
1008 my @parts = grep length($_->[0]), map [ $self->render_aqt($_) ], @args;
1009 return '' unless @parts;
1010 my $is_andor = !!($op =~ /^(and|or)$/);
1011 return @{$parts[0]} if $is_andor and @parts == 1;
1012 my ($final_sql) = map +($is_andor ? "( ${_} )" : $_), join(
1013 ' '.$self->_sqlcase($final_op).' ',
1018 map @{$_}[1..$#$_], @parts
1025 my ($self, $list) = @_;
1026 my @parts = grep length($_->[0]), map [ $self->render_aqt($_) ], @$list;
1027 return join(', ', map $_->[0], @parts), map @{$_}[1..$#$_], @parts;
1031 my ($self, $rest) = @_;
1032 my ($func, @args) = @$rest;
1036 push @arg_sql, shift @x;
1038 } map [ $self->render_aqt($_) ], @args;
1039 return ($self->_sqlcase($func).'('.join(', ', @arg_sql).')', @bind);
1043 my ($self, $bind) = @_;
1044 return ($self->_convert('?'), $self->_bindtype(@$bind));
1047 sub _render_literal {
1048 my ($self, $literal) = @_;
1049 $self->_assert_bindval_matches_bindtype(@{$literal}[1..$#$literal]);
1053 # Some databases (SQLite) treat col IN (1, 2) different from
1054 # col IN ( (1, 2) ). Use this to strip all outer parens while
1055 # adding them back in the corresponding method
1056 sub _open_outer_paren {
1057 my ($self, $sql) = @_;
1059 while (my ($inner) = $sql =~ /^ \s* \( (.*) \) \s* $/xs) {
1061 # there are closing parens inside, need the heavy duty machinery
1062 # to reevaluate the extraction starting from $sql (full reevaluation)
1063 if ($inner =~ /\)/) {
1064 require Text::Balanced;
1066 my (undef, $remainder) = do {
1067 # idiotic design - writes to $@ but *DOES NOT* throw exceptions
1069 Text::Balanced::extract_bracketed($sql, '()', qr/\s*/);
1072 # the entire expression needs to be a balanced bracketed thing
1073 # (after an extract no remainder sans trailing space)
1074 last if defined $remainder and $remainder =~ /\S/;
1084 #======================================================================
1086 #======================================================================
1088 sub _expand_order_by {
1089 my ($self, $arg) = @_;
1091 return unless defined($arg) and not (ref($arg) eq 'ARRAY' and !@$arg);
1093 my $expander = sub {
1094 my ($self, $dir, $expr) = @_;
1095 my @to_expand = ref($expr) eq 'ARRAY' ? @$expr : $expr;
1096 foreach my $arg (@to_expand) {
1100 and grep /^-(asc|desc)$/, keys %$arg
1102 puke "ordering direction hash passed to order by must have exactly one key (-asc or -desc)";
1105 my @exp = map +(defined($dir) ? { -op => [ $dir => $_ ] } : $_),
1106 map $self->expand_expr($_, -ident),
1107 map ref($_) eq 'ARRAY' ? @$_ : $_, @to_expand;
1108 return (@exp > 1 ? { -list => \@exp } : $exp[0]);
1111 local @{$self->{expand_unary}}{qw(-asc -desc)} = (
1112 sub { shift->$expander(asc => @_) },
1113 sub { shift->$expander(desc => @_) },
1116 return $self->$expander(undef, $arg);
1120 my ($self, $arg) = @_;
1122 return '' unless defined(my $expanded = $self->_expand_order_by($arg));
1124 my ($sql, @bind) = $self->render_aqt($expanded);
1126 return '' unless length($sql);
1128 my $final_sql = $self->_sqlcase(' order by ').$sql;
1130 return wantarray ? ($final_sql, @bind) : $final_sql;
1133 # _order_by no longer needs to call this so doesn't but DBIC uses it.
1135 sub _order_by_chunks {
1136 my ($self, $arg) = @_;
1138 return () unless defined(my $expanded = $self->_expand_order_by($arg));
1140 return $self->_chunkify_order_by($expanded);
1143 sub _chunkify_order_by {
1144 my ($self, $expanded) = @_;
1146 return grep length, $self->render_aqt($expanded)
1147 if $expanded->{-ident} or @{$expanded->{-literal}||[]} == 1;
1150 if (ref() eq 'HASH' and my $l = $_->{-list}) {
1151 return map $self->_chunkify_order_by($_), @$l;
1153 return [ $self->render_aqt($_) ];
1157 #======================================================================
1158 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
1159 #======================================================================
1165 $self->_expand_maybe_list_expr($from, undef, -ident)
1170 #======================================================================
1172 #======================================================================
1174 sub _expand_maybe_list_expr {
1175 my ($self, $expr, $logic, $default) = @_;
1177 if (ref($expr) eq 'ARRAY') {
1179 map $self->_expand_expr($_, $logic, $default), @$expr
1186 return $self->_expand_expr($e, $logic, $default);
1189 # highly optimized, as it's called way too often
1191 # my ($self, $label) = @_;
1193 return '' unless defined $_[1];
1194 return ${$_[1]} if ref($_[1]) eq 'SCALAR';
1195 puke 'Identifier cannot be hashref' if ref($_[1]) eq 'HASH';
1197 unless ($_[0]->{quote_char}) {
1198 if (ref($_[1]) eq 'ARRAY') {
1199 return join($_[0]->{name_sep}||'.', @{$_[1]});
1201 $_[0]->_assert_pass_injection_guard($_[1]);
1206 my $qref = ref $_[0]->{quote_char};
1208 !$qref ? ($_[0]->{quote_char}, $_[0]->{quote_char})
1209 : ($qref eq 'ARRAY') ? @{$_[0]->{quote_char}}
1210 : puke "Unsupported quote_char format: $_[0]->{quote_char}";
1212 my $esc = $_[0]->{escape_char} || $r;
1214 # parts containing * are naturally unquoted
1216 $_[0]->{name_sep}||'',
1220 : do { (my $n = $_) =~ s/(\Q$esc\E|\Q$r\E)/$esc$1/g; $l . $n . $r }
1222 (ref($_[1]) eq 'ARRAY'
1226 ? split (/\Q$_[0]->{name_sep}\E/, $_[1] )
1234 # Conversion, if applicable
1236 #my ($self, $arg) = @_;
1237 if ($_[0]->{convert_where}) {
1238 return $_[0]->_sqlcase($_[0]->{convert_where}) .'(' . $_[1] . ')';
1245 #my ($self, $col, @vals) = @_;
1246 # called often - tighten code
1247 return $_[0]->{bindtype} eq 'columns'
1248 ? map {[$_[1], $_]} @_[2 .. $#_]
1253 # Dies if any element of @bind is not in [colname => value] format
1254 # if bindtype is 'columns'.
1255 sub _assert_bindval_matches_bindtype {
1256 # my ($self, @bind) = @_;
1258 if ($self->{bindtype} eq 'columns') {
1260 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
1261 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
1267 sub _join_sql_clauses {
1268 my ($self, $logic, $clauses_aref, $bind_aref) = @_;
1270 if (@$clauses_aref > 1) {
1271 my $join = " " . $self->_sqlcase($logic) . " ";
1272 my $sql = '( ' . join($join, @$clauses_aref) . ' )';
1273 return ($sql, @$bind_aref);
1275 elsif (@$clauses_aref) {
1276 return ($clauses_aref->[0], @$bind_aref); # no parentheses
1279 return (); # if no SQL, ignore @$bind_aref
1284 # Fix SQL case, if so requested
1286 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
1287 # don't touch the argument ... crooked logic, but let's not change it!
1288 return $_[0]->{case} ? $_[1] : uc($_[1]);
1292 #======================================================================
1293 # DISPATCHING FROM REFKIND
1294 #======================================================================
1297 my ($self, $data) = @_;
1299 return 'UNDEF' unless defined $data;
1301 # blessed objects are treated like scalars
1302 my $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1304 return 'SCALAR' unless $ref;
1307 while ($ref eq 'REF') {
1309 $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1313 return ($ref||'SCALAR') . ('REF' x $n_steps);
1317 my ($self, $data) = @_;
1318 my @try = ($self->_refkind($data));
1319 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
1320 push @try, 'FALLBACK';
1324 sub _METHOD_FOR_refkind {
1325 my ($self, $meth_prefix, $data) = @_;
1328 for (@{$self->_try_refkind($data)}) {
1329 $method = $self->can($meth_prefix."_".$_)
1333 return $method || puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
1337 sub _SWITCH_refkind {
1338 my ($self, $data, $dispatch_table) = @_;
1341 for (@{$self->_try_refkind($data)}) {
1342 $coderef = $dispatch_table->{$_}
1346 puke "no dispatch entry for ".$self->_refkind($data)
1355 #======================================================================
1356 # VALUES, GENERATE, AUTOLOAD
1357 #======================================================================
1359 # LDNOTE: original code from nwiger, didn't touch code in that section
1360 # I feel the AUTOLOAD stuff should not be the default, it should
1361 # only be activated on explicit demand by user.
1365 my $data = shift || return;
1366 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
1367 unless ref $data eq 'HASH';
1370 foreach my $k (sort keys %$data) {
1371 my $v = $data->{$k};
1372 $self->_SWITCH_refkind($v, {
1374 if ($self->{array_datatypes}) { # array datatype
1375 push @all_bind, $self->_bindtype($k, $v);
1377 else { # literal SQL with bind
1378 my ($sql, @bind) = @$v;
1379 $self->_assert_bindval_matches_bindtype(@bind);
1380 push @all_bind, @bind;
1383 ARRAYREFREF => sub { # literal SQL with bind
1384 my ($sql, @bind) = @${$v};
1385 $self->_assert_bindval_matches_bindtype(@bind);
1386 push @all_bind, @bind;
1388 SCALARREF => sub { # literal SQL without bind
1390 SCALAR_or_UNDEF => sub {
1391 push @all_bind, $self->_bindtype($k, $v);
1402 my(@sql, @sqlq, @sqlv);
1406 if ($ref eq 'HASH') {
1407 for my $k (sort keys %$_) {
1410 my $label = $self->_quote($k);
1411 if ($r eq 'ARRAY') {
1412 # literal SQL with bind
1413 my ($sql, @bind) = @$v;
1414 $self->_assert_bindval_matches_bindtype(@bind);
1415 push @sqlq, "$label = $sql";
1417 } elsif ($r eq 'SCALAR') {
1418 # literal SQL without bind
1419 push @sqlq, "$label = $$v";
1421 push @sqlq, "$label = ?";
1422 push @sqlv, $self->_bindtype($k, $v);
1425 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
1426 } elsif ($ref eq 'ARRAY') {
1427 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
1430 if ($r eq 'ARRAY') { # literal SQL with bind
1431 my ($sql, @bind) = @$v;
1432 $self->_assert_bindval_matches_bindtype(@bind);
1435 } elsif ($r eq 'SCALAR') { # literal SQL without bind
1436 # embedded literal SQL
1443 push @sql, '(' . join(', ', @sqlq) . ')';
1444 } elsif ($ref eq 'SCALAR') {
1448 # strings get case twiddled
1449 push @sql, $self->_sqlcase($_);
1453 my $sql = join ' ', @sql;
1455 # this is pretty tricky
1456 # if ask for an array, return ($stmt, @bind)
1457 # otherwise, s/?/shift @sqlv/ to put it inline
1459 return ($sql, @sqlv);
1461 1 while $sql =~ s/\?/my $d = shift(@sqlv);
1462 ref $d ? $d->[1] : $d/e;
1471 # This allows us to check for a local, then _form, attr
1473 my($name) = $AUTOLOAD =~ /.*::(.+)/;
1474 return $self->generate($name, @_);
1485 SQL::Abstract - Generate SQL from Perl data structures
1491 my $sql = SQL::Abstract->new;
1493 my($stmt, @bind) = $sql->select($source, \@fields, \%where, $order);
1495 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
1497 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
1499 my($stmt, @bind) = $sql->delete($table, \%where);
1501 # Then, use these in your DBI statements
1502 my $sth = $dbh->prepare($stmt);
1503 $sth->execute(@bind);
1505 # Just generate the WHERE clause
1506 my($stmt, @bind) = $sql->where(\%where, $order);
1508 # Return values in the same order, for hashed queries
1509 # See PERFORMANCE section for more details
1510 my @bind = $sql->values(\%fieldvals);
1514 This module was inspired by the excellent L<DBIx::Abstract>.
1515 However, in using that module I found that what I really wanted
1516 to do was generate SQL, but still retain complete control over my
1517 statement handles and use the DBI interface. So, I set out to
1518 create an abstract SQL generation module.
1520 While based on the concepts used by L<DBIx::Abstract>, there are
1521 several important differences, especially when it comes to WHERE
1522 clauses. I have modified the concepts used to make the SQL easier
1523 to generate from Perl data structures and, IMO, more intuitive.
1524 The underlying idea is for this module to do what you mean, based
1525 on the data structures you provide it. The big advantage is that
1526 you don't have to modify your code every time your data changes,
1527 as this module figures it out.
1529 To begin with, an SQL INSERT is as easy as just specifying a hash
1530 of C<key=value> pairs:
1533 name => 'Jimbo Bobson',
1534 phone => '123-456-7890',
1535 address => '42 Sister Lane',
1536 city => 'St. Louis',
1537 state => 'Louisiana',
1540 The SQL can then be generated with this:
1542 my($stmt, @bind) = $sql->insert('people', \%data);
1544 Which would give you something like this:
1546 $stmt = "INSERT INTO people
1547 (address, city, name, phone, state)
1548 VALUES (?, ?, ?, ?, ?)";
1549 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
1550 '123-456-7890', 'Louisiana');
1552 These are then used directly in your DBI code:
1554 my $sth = $dbh->prepare($stmt);
1555 $sth->execute(@bind);
1557 =head2 Inserting and Updating Arrays
1559 If your database has array types (like for example Postgres),
1560 activate the special option C<< array_datatypes => 1 >>
1561 when creating the C<SQL::Abstract> object.
1562 Then you may use an arrayref to insert and update database array types:
1564 my $sql = SQL::Abstract->new(array_datatypes => 1);
1566 planets => [qw/Mercury Venus Earth Mars/]
1569 my($stmt, @bind) = $sql->insert('solar_system', \%data);
1573 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
1575 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
1578 =head2 Inserting and Updating SQL
1580 In order to apply SQL functions to elements of your C<%data> you may
1581 specify a reference to an arrayref for the given hash value. For example,
1582 if you need to execute the Oracle C<to_date> function on a value, you can
1583 say something like this:
1587 date_entered => \[ "to_date(?,'MM/DD/YYYY')", "03/02/2003" ],
1590 The first value in the array is the actual SQL. Any other values are
1591 optional and would be included in the bind values array. This gives
1594 my($stmt, @bind) = $sql->insert('people', \%data);
1596 $stmt = "INSERT INTO people (name, date_entered)
1597 VALUES (?, to_date(?,'MM/DD/YYYY'))";
1598 @bind = ('Bill', '03/02/2003');
1600 An UPDATE is just as easy, all you change is the name of the function:
1602 my($stmt, @bind) = $sql->update('people', \%data);
1604 Notice that your C<%data> isn't touched; the module will generate
1605 the appropriately quirky SQL for you automatically. Usually you'll
1606 want to specify a WHERE clause for your UPDATE, though, which is
1607 where handling C<%where> hashes comes in handy...
1609 =head2 Complex where statements
1611 This module can generate pretty complicated WHERE statements
1612 easily. For example, simple C<key=value> pairs are taken to mean
1613 equality, and if you want to see if a field is within a set
1614 of values, you can use an arrayref. Let's say we wanted to
1615 SELECT some data based on this criteria:
1618 requestor => 'inna',
1619 worker => ['nwiger', 'rcwe', 'sfz'],
1620 status => { '!=', 'completed' }
1623 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
1625 The above would give you something like this:
1627 $stmt = "SELECT * FROM tickets WHERE
1628 ( requestor = ? ) AND ( status != ? )
1629 AND ( worker = ? OR worker = ? OR worker = ? )";
1630 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
1632 Which you could then use in DBI code like so:
1634 my $sth = $dbh->prepare($stmt);
1635 $sth->execute(@bind);
1641 The methods are simple. There's one for every major SQL operation,
1642 and a constructor you use first. The arguments are specified in a
1643 similar order for each method (table, then fields, then a where
1644 clause) to try and simplify things.
1646 =head2 new(option => 'value')
1648 The C<new()> function takes a list of options and values, and returns
1649 a new B<SQL::Abstract> object which can then be used to generate SQL
1650 through the methods below. The options accepted are:
1656 If set to 'lower', then SQL will be generated in all lowercase. By
1657 default SQL is generated in "textbook" case meaning something like:
1659 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
1661 Any setting other than 'lower' is ignored.
1665 This determines what the default comparison operator is. By default
1666 it is C<=>, meaning that a hash like this:
1668 %where = (name => 'nwiger', email => 'nate@wiger.org');
1670 Will generate SQL like this:
1672 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
1674 However, you may want loose comparisons by default, so if you set
1675 C<cmp> to C<like> you would get SQL such as:
1677 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
1679 You can also override the comparison on an individual basis - see
1680 the huge section on L</"WHERE CLAUSES"> at the bottom.
1682 =item sqltrue, sqlfalse
1684 Expressions for inserting boolean values within SQL statements.
1685 By default these are C<1=1> and C<1=0>. They are used
1686 by the special operators C<-in> and C<-not_in> for generating
1687 correct SQL even when the argument is an empty array (see below).
1691 This determines the default logical operator for multiple WHERE
1692 statements in arrays or hashes. If absent, the default logic is "or"
1693 for arrays, and "and" for hashes. This means that a WHERE
1697 event_date => {'>=', '2/13/99'},
1698 event_date => {'<=', '4/24/03'},
1701 will generate SQL like this:
1703 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
1705 This is probably not what you want given this query, though (look
1706 at the dates). To change the "OR" to an "AND", simply specify:
1708 my $sql = SQL::Abstract->new(logic => 'and');
1710 Which will change the above C<WHERE> to:
1712 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
1714 The logic can also be changed locally by inserting
1715 a modifier in front of an arrayref:
1717 @where = (-and => [event_date => {'>=', '2/13/99'},
1718 event_date => {'<=', '4/24/03'} ]);
1720 See the L</"WHERE CLAUSES"> section for explanations.
1724 This will automatically convert comparisons using the specified SQL
1725 function for both column and value. This is mostly used with an argument
1726 of C<upper> or C<lower>, so that the SQL will have the effect of
1727 case-insensitive "searches". For example, this:
1729 $sql = SQL::Abstract->new(convert => 'upper');
1730 %where = (keywords => 'MaKe iT CAse inSeNSItive');
1732 Will turn out the following SQL:
1734 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
1736 The conversion can be C<upper()>, C<lower()>, or any other SQL function
1737 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
1738 not validate this option; it will just pass through what you specify verbatim).
1742 This is a kludge because many databases suck. For example, you can't
1743 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
1744 Instead, you have to use C<bind_param()>:
1746 $sth->bind_param(1, 'reg data');
1747 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
1749 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
1750 which loses track of which field each slot refers to. Fear not.
1752 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
1753 Currently, you can specify either C<normal> (default) or C<columns>. If you
1754 specify C<columns>, you will get an array that looks like this:
1756 my $sql = SQL::Abstract->new(bindtype => 'columns');
1757 my($stmt, @bind) = $sql->insert(...);
1760 [ 'column1', 'value1' ],
1761 [ 'column2', 'value2' ],
1762 [ 'column3', 'value3' ],
1765 You can then iterate through this manually, using DBI's C<bind_param()>.
1767 $sth->prepare($stmt);
1770 my($col, $data) = @$_;
1771 if ($col eq 'details' || $col eq 'comments') {
1772 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
1773 } elsif ($col eq 'image') {
1774 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
1776 $sth->bind_param($i, $data);
1780 $sth->execute; # execute without @bind now
1782 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
1783 Basically, the advantage is still that you don't have to care which fields
1784 are or are not included. You could wrap that above C<for> loop in a simple
1785 sub called C<bind_fields()> or something and reuse it repeatedly. You still
1786 get a layer of abstraction over manual SQL specification.
1788 Note that if you set L</bindtype> to C<columns>, the C<\[ $sql, @bind ]>
1789 construct (see L</Literal SQL with placeholders and bind values (subqueries)>)
1790 will expect the bind values in this format.
1794 This is the character that a table or column name will be quoted
1795 with. By default this is an empty string, but you could set it to
1796 the character C<`>, to generate SQL like this:
1798 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
1800 Alternatively, you can supply an array ref of two items, the first being the left
1801 hand quote character, and the second the right hand quote character. For
1802 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
1803 that generates SQL like this:
1805 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
1807 Quoting is useful if you have tables or columns names that are reserved
1808 words in your database's SQL dialect.
1812 This is the character that will be used to escape L</quote_char>s appearing
1813 in an identifier before it has been quoted.
1815 The parameter default in case of a single L</quote_char> character is the quote
1818 When opening-closing-style quoting is used (L</quote_char> is an arrayref)
1819 this parameter defaults to the B<closing (right)> L</quote_char>. Occurrences
1820 of the B<opening (left)> L</quote_char> within the identifier are currently left
1821 untouched. The default for opening-closing-style quotes may change in future
1822 versions, thus you are B<strongly encouraged> to specify the escape character
1827 This is the character that separates a table and column name. It is
1828 necessary to specify this when the C<quote_char> option is selected,
1829 so that tables and column names can be individually quoted like this:
1831 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
1833 =item injection_guard
1835 A regular expression C<qr/.../> that is applied to any C<-function> and unquoted
1836 column name specified in a query structure. This is a safety mechanism to avoid
1837 injection attacks when mishandling user input e.g.:
1839 my %condition_as_column_value_pairs = get_values_from_user();
1840 $sqla->select( ... , \%condition_as_column_value_pairs );
1842 If the expression matches an exception is thrown. Note that literal SQL
1843 supplied via C<\'...'> or C<\['...']> is B<not> checked in any way.
1845 Defaults to checking for C<;> and the C<GO> keyword (TransactSQL)
1847 =item array_datatypes
1849 When this option is true, arrayrefs in INSERT or UPDATE are
1850 interpreted as array datatypes and are passed directly
1852 When this option is false, arrayrefs are interpreted
1853 as literal SQL, just like refs to arrayrefs
1854 (but this behavior is for backwards compatibility; when writing
1855 new queries, use the "reference to arrayref" syntax
1861 Takes a reference to a list of "special operators"
1862 to extend the syntax understood by L<SQL::Abstract>.
1863 See section L</"SPECIAL OPERATORS"> for details.
1867 Takes a reference to a list of "unary operators"
1868 to extend the syntax understood by L<SQL::Abstract>.
1869 See section L</"UNARY OPERATORS"> for details.
1875 =head2 insert($table, \@values || \%fieldvals, \%options)
1877 This is the simplest function. You simply give it a table name
1878 and either an arrayref of values or hashref of field/value pairs.
1879 It returns an SQL INSERT statement and a list of bind values.
1880 See the sections on L</"Inserting and Updating Arrays"> and
1881 L</"Inserting and Updating SQL"> for information on how to insert
1882 with those data types.
1884 The optional C<\%options> hash reference may contain additional
1885 options to generate the insert SQL. Currently supported options
1892 Takes either a scalar of raw SQL fields, or an array reference of
1893 field names, and adds on an SQL C<RETURNING> statement at the end.
1894 This allows you to return data generated by the insert statement
1895 (such as row IDs) without performing another C<SELECT> statement.
1896 Note, however, this is not part of the SQL standard and may not
1897 be supported by all database engines.
1901 =head2 update($table, \%fieldvals, \%where, \%options)
1903 This takes a table, hashref of field/value pairs, and an optional
1904 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
1906 See the sections on L</"Inserting and Updating Arrays"> and
1907 L</"Inserting and Updating SQL"> for information on how to insert
1908 with those data types.
1910 The optional C<\%options> hash reference may contain additional
1911 options to generate the update SQL. Currently supported options
1918 See the C<returning> option to
1919 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
1923 =head2 select($source, $fields, $where, $order)
1925 This returns a SQL SELECT statement and associated list of bind values, as
1926 specified by the arguments:
1932 Specification of the 'FROM' part of the statement.
1933 The argument can be either a plain scalar (interpreted as a table
1934 name, will be quoted), or an arrayref (interpreted as a list
1935 of table names, joined by commas, quoted), or a scalarref
1936 (literal SQL, not quoted).
1940 Specification of the list of fields to retrieve from
1942 The argument can be either an arrayref (interpreted as a list
1943 of field names, will be joined by commas and quoted), or a
1944 plain scalar (literal SQL, not quoted).
1945 Please observe that this API is not as flexible as that of
1946 the first argument C<$source>, for backwards compatibility reasons.
1950 Optional argument to specify the WHERE part of the query.
1951 The argument is most often a hashref, but can also be
1952 an arrayref or plain scalar --
1953 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
1957 Optional argument to specify the ORDER BY part of the query.
1958 The argument can be a scalar, a hashref or an arrayref
1959 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
1965 =head2 delete($table, \%where, \%options)
1967 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
1968 It returns an SQL DELETE statement and list of bind values.
1970 The optional C<\%options> hash reference may contain additional
1971 options to generate the delete SQL. Currently supported options
1978 See the C<returning> option to
1979 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
1983 =head2 where(\%where, $order)
1985 This is used to generate just the WHERE clause. For example,
1986 if you have an arbitrary data structure and know what the
1987 rest of your SQL is going to look like, but want an easy way
1988 to produce a WHERE clause, use this. It returns an SQL WHERE
1989 clause and list of bind values.
1992 =head2 values(\%data)
1994 This just returns the values from the hash C<%data>, in the same
1995 order that would be returned from any of the other above queries.
1996 Using this allows you to markedly speed up your queries if you
1997 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
1999 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
2001 Warning: This is an experimental method and subject to change.
2003 This returns arbitrarily generated SQL. It's a really basic shortcut.
2004 It will return two different things, depending on return context:
2006 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
2007 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
2009 These would return the following:
2011 # First calling form
2012 $stmt = "CREATE TABLE test (?, ?)";
2013 @bind = (field1, field2);
2015 # Second calling form
2016 $stmt_and_val = "CREATE TABLE test (field1, field2)";
2018 Depending on what you're trying to do, it's up to you to choose the correct
2019 format. In this example, the second form is what you would want.
2023 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
2027 ALTER SESSION SET nls_date_format = 'MM/YY'
2029 You get the idea. Strings get their case twiddled, but everything
2030 else remains verbatim.
2032 =head1 EXPORTABLE FUNCTIONS
2034 =head2 is_plain_value
2036 Determines if the supplied argument is a plain value as understood by this
2041 =item * The value is C<undef>
2043 =item * The value is a non-reference
2045 =item * The value is an object with stringification overloading
2047 =item * The value is of the form C<< { -value => $anything } >>
2051 On failure returns C<undef>, on success returns a B<scalar> reference
2052 to the original supplied argument.
2058 The stringification overloading detection is rather advanced: it takes
2059 into consideration not only the presence of a C<""> overload, but if that
2060 fails also checks for enabled
2061 L<autogenerated versions of C<"">|overload/Magic Autogeneration>, based
2062 on either C<0+> or C<bool>.
2064 Unfortunately testing in the field indicates that this
2065 detection B<< may tickle a latent bug in perl versions before 5.018 >>,
2066 but only when very large numbers of stringifying objects are involved.
2067 At the time of writing ( Sep 2014 ) there is no clear explanation of
2068 the direct cause, nor is there a manageably small test case that reliably
2069 reproduces the problem.
2071 If you encounter any of the following exceptions in B<random places within
2072 your application stack> - this module may be to blame:
2074 Operation "ne": no method found,
2075 left argument in overloaded package <something>,
2076 right argument in overloaded package <something>
2080 Stub found while resolving method "???" overloading """" in package <something>
2082 If you fall victim to the above - please attempt to reduce the problem
2083 to something that could be sent to the L<SQL::Abstract developers
2084 |DBIx::Class/GETTING HELP/SUPPORT>
2085 (either publicly or privately). As a workaround in the meantime you can
2086 set C<$ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}> to a true
2087 value, which will most likely eliminate your problem (at the expense of
2088 not being able to properly detect exotic forms of stringification).
2090 This notice and environment variable will be removed in a future version,
2091 as soon as the underlying problem is found and a reliable workaround is
2096 =head2 is_literal_value
2098 Determines if the supplied argument is a literal value as understood by this
2103 =item * C<\$sql_string>
2105 =item * C<\[ $sql_string, @bind_values ]>
2109 On failure returns C<undef>, on success returns an B<array> reference
2110 containing the unpacked version of the supplied literal SQL and bind values.
2112 =head1 WHERE CLAUSES
2116 This module uses a variation on the idea from L<DBIx::Abstract>. It
2117 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
2118 module is that things in arrays are OR'ed, and things in hashes
2121 The easiest way to explain is to show lots of examples. After
2122 each C<%where> hash shown, it is assumed you used:
2124 my($stmt, @bind) = $sql->where(\%where);
2126 However, note that the C<%where> hash can be used directly in any
2127 of the other functions as well, as described above.
2129 =head2 Key-value pairs
2131 So, let's get started. To begin, a simple hash:
2135 status => 'completed'
2138 Is converted to SQL C<key = val> statements:
2140 $stmt = "WHERE user = ? AND status = ?";
2141 @bind = ('nwiger', 'completed');
2143 One common thing I end up doing is having a list of values that
2144 a field can be in. To do this, simply specify a list inside of
2149 status => ['assigned', 'in-progress', 'pending'];
2152 This simple code will create the following:
2154 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
2155 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
2157 A field associated to an empty arrayref will be considered a
2158 logical false and will generate 0=1.
2160 =head2 Tests for NULL values
2162 If the value part is C<undef> then this is converted to SQL <IS NULL>
2171 $stmt = "WHERE user = ? AND status IS NULL";
2174 To test if a column IS NOT NULL:
2178 status => { '!=', undef },
2181 =head2 Specific comparison operators
2183 If you want to specify a different type of operator for your comparison,
2184 you can use a hashref for a given column:
2188 status => { '!=', 'completed' }
2191 Which would generate:
2193 $stmt = "WHERE user = ? AND status != ?";
2194 @bind = ('nwiger', 'completed');
2196 To test against multiple values, just enclose the values in an arrayref:
2198 status => { '=', ['assigned', 'in-progress', 'pending'] };
2200 Which would give you:
2202 "WHERE status = ? OR status = ? OR status = ?"
2205 The hashref can also contain multiple pairs, in which case it is expanded
2206 into an C<AND> of its elements:
2210 status => { '!=', 'completed', -not_like => 'pending%' }
2213 # Or more dynamically, like from a form
2214 $where{user} = 'nwiger';
2215 $where{status}{'!='} = 'completed';
2216 $where{status}{'-not_like'} = 'pending%';
2218 # Both generate this
2219 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
2220 @bind = ('nwiger', 'completed', 'pending%');
2223 To get an OR instead, you can combine it with the arrayref idea:
2227 priority => [ { '=', 2 }, { '>', 5 } ]
2230 Which would generate:
2232 $stmt = "WHERE ( priority = ? OR priority > ? ) AND user = ?";
2233 @bind = ('2', '5', 'nwiger');
2235 If you want to include literal SQL (with or without bind values), just use a
2236 scalar reference or reference to an arrayref as the value:
2239 date_entered => { '>' => \["to_date(?, 'MM/DD/YYYY')", "11/26/2008"] },
2240 date_expires => { '<' => \"now()" }
2243 Which would generate:
2245 $stmt = "WHERE date_entered > to_date(?, 'MM/DD/YYYY') AND date_expires < now()";
2246 @bind = ('11/26/2008');
2249 =head2 Logic and nesting operators
2251 In the example above,
2252 there is a subtle trap if you want to say something like
2253 this (notice the C<AND>):
2255 WHERE priority != ? AND priority != ?
2257 Because, in Perl you I<can't> do this:
2259 priority => { '!=' => 2, '!=' => 1 }
2261 As the second C<!=> key will obliterate the first. The solution
2262 is to use the special C<-modifier> form inside an arrayref:
2264 priority => [ -and => {'!=', 2},
2268 Normally, these would be joined by C<OR>, but the modifier tells it
2269 to use C<AND> instead. (Hint: You can use this in conjunction with the
2270 C<logic> option to C<new()> in order to change the way your queries
2271 work by default.) B<Important:> Note that the C<-modifier> goes
2272 B<INSIDE> the arrayref, as an extra first element. This will
2273 B<NOT> do what you think it might:
2275 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
2277 Here is a quick list of equivalencies, since there is some overlap:
2280 status => {'!=', 'completed', 'not like', 'pending%' }
2281 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
2284 status => {'=', ['assigned', 'in-progress']}
2285 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
2286 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
2290 =head2 Special operators: IN, BETWEEN, etc.
2292 You can also use the hashref format to compare a list of fields using the
2293 C<IN> comparison operator, by specifying the list as an arrayref:
2296 status => 'completed',
2297 reportid => { -in => [567, 2335, 2] }
2300 Which would generate:
2302 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
2303 @bind = ('completed', '567', '2335', '2');
2305 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
2308 If the argument to C<-in> is an empty array, 'sqlfalse' is generated
2309 (by default: C<1=0>). Similarly, C<< -not_in => [] >> generates
2310 'sqltrue' (by default: C<1=1>).
2312 In addition to the array you can supply a chunk of literal sql or
2313 literal sql with bind:
2316 customer => { -in => \[
2317 'SELECT cust_id FROM cust WHERE balance > ?',
2320 status => { -in => \'SELECT status_codes FROM states' },
2326 customer IN ( SELECT cust_id FROM cust WHERE balance > ? )
2327 AND status IN ( SELECT status_codes FROM states )
2331 Finally, if the argument to C<-in> is not a reference, it will be
2332 treated as a single-element array.
2334 Another pair of operators is C<-between> and C<-not_between>,
2335 used with an arrayref of two values:
2339 completion_date => {
2340 -not_between => ['2002-10-01', '2003-02-06']
2346 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
2348 Just like with C<-in> all plausible combinations of literal SQL
2352 start0 => { -between => [ 1, 2 ] },
2353 start1 => { -between => \["? AND ?", 1, 2] },
2354 start2 => { -between => \"lower(x) AND upper(y)" },
2355 start3 => { -between => [
2357 \["upper(?)", 'stuff' ],
2364 ( start0 BETWEEN ? AND ? )
2365 AND ( start1 BETWEEN ? AND ? )
2366 AND ( start2 BETWEEN lower(x) AND upper(y) )
2367 AND ( start3 BETWEEN lower(x) AND upper(?) )
2369 @bind = (1, 2, 1, 2, 'stuff');
2372 These are the two builtin "special operators"; but the
2373 list can be expanded: see section L</"SPECIAL OPERATORS"> below.
2375 =head2 Unary operators: bool
2377 If you wish to test against boolean columns or functions within your
2378 database you can use the C<-bool> and C<-not_bool> operators. For
2379 example to test the column C<is_user> being true and the column
2380 C<is_enabled> being false you would use:-
2384 -not_bool => 'is_enabled',
2389 WHERE is_user AND NOT is_enabled
2391 If a more complex combination is required, testing more conditions,
2392 then you should use the and/or operators:-
2397 -not_bool => { two=> { -rlike => 'bar' } },
2398 -not_bool => { three => [ { '=', 2 }, { '>', 5 } ] },
2409 (NOT ( three = ? OR three > ? ))
2412 =head2 Nested conditions, -and/-or prefixes
2414 So far, we've seen how multiple conditions are joined with a top-level
2415 C<AND>. We can change this by putting the different conditions we want in
2416 hashes and then putting those hashes in an array. For example:
2421 status => { -like => ['pending%', 'dispatched'] },
2425 status => 'unassigned',
2429 This data structure would create the following:
2431 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
2432 OR ( user = ? AND status = ? ) )";
2433 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
2436 Clauses in hashrefs or arrayrefs can be prefixed with an C<-and> or C<-or>
2437 to change the logic inside:
2443 -and => [ workhrs => {'>', 20}, geo => 'ASIA' ],
2444 -or => { workhrs => {'<', 50}, geo => 'EURO' },
2451 $stmt = "WHERE ( user = ?
2452 AND ( ( workhrs > ? AND geo = ? )
2453 OR ( workhrs < ? OR geo = ? ) ) )";
2454 @bind = ('nwiger', '20', 'ASIA', '50', 'EURO');
2456 =head3 Algebraic inconsistency, for historical reasons
2458 C<Important note>: when connecting several conditions, the C<-and->|C<-or>
2459 operator goes C<outside> of the nested structure; whereas when connecting
2460 several constraints on one column, the C<-and> operator goes
2461 C<inside> the arrayref. Here is an example combining both features:
2464 -and => [a => 1, b => 2],
2465 -or => [c => 3, d => 4],
2466 e => [-and => {-like => 'foo%'}, {-like => '%bar'} ]
2471 WHERE ( ( ( a = ? AND b = ? )
2472 OR ( c = ? OR d = ? )
2473 OR ( e LIKE ? AND e LIKE ? ) ) )
2475 This difference in syntax is unfortunate but must be preserved for
2476 historical reasons. So be careful: the two examples below would
2477 seem algebraically equivalent, but they are not
2480 { -like => 'foo%' },
2481 { -like => '%bar' },
2483 # yields: WHERE ( ( col LIKE ? AND col LIKE ? ) )
2486 { col => { -like => 'foo%' } },
2487 { col => { -like => '%bar' } },
2489 # yields: WHERE ( ( col LIKE ? OR col LIKE ? ) )
2492 =head2 Literal SQL and value type operators
2494 The basic premise of SQL::Abstract is that in WHERE specifications the "left
2495 side" is a column name and the "right side" is a value (normally rendered as
2496 a placeholder). This holds true for both hashrefs and arrayref pairs as you
2497 see in the L</WHERE CLAUSES> examples above. Sometimes it is necessary to
2498 alter this behavior. There are several ways of doing so.
2502 This is a virtual operator that signals the string to its right side is an
2503 identifier (a column name) and not a value. For example to compare two
2504 columns you would write:
2507 priority => { '<', 2 },
2508 requestor => { -ident => 'submitter' },
2513 $stmt = "WHERE priority < ? AND requestor = submitter";
2516 If you are maintaining legacy code you may see a different construct as
2517 described in L</Deprecated usage of Literal SQL>, please use C<-ident> in new
2522 This is a virtual operator that signals that the construct to its right side
2523 is a value to be passed to DBI. This is for example necessary when you want
2524 to write a where clause against an array (for RDBMS that support such
2525 datatypes). For example:
2528 array => { -value => [1, 2, 3] }
2533 $stmt = 'WHERE array = ?';
2534 @bind = ([1, 2, 3]);
2536 Note that if you were to simply say:
2542 the result would probably not be what you wanted:
2544 $stmt = 'WHERE array = ? OR array = ? OR array = ?';
2549 Finally, sometimes only literal SQL will do. To include a random snippet
2550 of SQL verbatim, you specify it as a scalar reference. Consider this only
2551 as a last resort. Usually there is a better way. For example:
2554 priority => { '<', 2 },
2555 requestor => { -in => \'(SELECT name FROM hitmen)' },
2560 $stmt = "WHERE priority < ? AND requestor IN (SELECT name FROM hitmen)"
2563 Note that in this example, you only get one bind parameter back, since
2564 the verbatim SQL is passed as part of the statement.
2568 Never use untrusted input as a literal SQL argument - this is a massive
2569 security risk (there is no way to check literal snippets for SQL
2570 injections and other nastyness). If you need to deal with untrusted input
2571 use literal SQL with placeholders as described next.
2573 =head3 Literal SQL with placeholders and bind values (subqueries)
2575 If the literal SQL to be inserted has placeholders and bind values,
2576 use a reference to an arrayref (yes this is a double reference --
2577 not so common, but perfectly legal Perl). For example, to find a date
2578 in Postgres you can use something like this:
2581 date_column => \[ "= date '2008-09-30' - ?::integer", 10 ]
2586 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
2589 Note that you must pass the bind values in the same format as they are returned
2590 by L<where|/where(\%where, $order)>. This means that if you set L</bindtype>
2591 to C<columns>, you must provide the bind values in the
2592 C<< [ column_meta => value ] >> format, where C<column_meta> is an opaque
2593 scalar value; most commonly the column name, but you can use any scalar value
2594 (including references and blessed references), L<SQL::Abstract> will simply
2595 pass it through intact. So if C<bindtype> is set to C<columns> the above
2596 example will look like:
2599 date_column => \[ "= date '2008-09-30' - ?::integer", [ {} => 10 ] ]
2602 Literal SQL is especially useful for nesting parenthesized clauses in the
2603 main SQL query. Here is a first example:
2605 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
2609 bar => \["IN ($sub_stmt)" => @sub_bind],
2614 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
2615 WHERE c2 < ? AND c3 LIKE ?))";
2616 @bind = (1234, 100, "foo%");
2618 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
2619 are expressed in the same way. Of course the C<$sub_stmt> and
2620 its associated bind values can be generated through a former call
2623 my ($sub_stmt, @sub_bind)
2624 = $sql->select("t1", "c1", {c2 => {"<" => 100},
2625 c3 => {-like => "foo%"}});
2628 bar => \["> ALL ($sub_stmt)" => @sub_bind],
2631 In the examples above, the subquery was used as an operator on a column;
2632 but the same principle also applies for a clause within the main C<%where>
2633 hash, like an EXISTS subquery:
2635 my ($sub_stmt, @sub_bind)
2636 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
2637 my %where = ( -and => [
2639 \["EXISTS ($sub_stmt)" => @sub_bind],
2644 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
2645 WHERE c1 = ? AND c2 > t0.c0))";
2649 Observe that the condition on C<c2> in the subquery refers to
2650 column C<t0.c0> of the main query: this is I<not> a bind
2651 value, so we have to express it through a scalar ref.
2652 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
2653 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
2654 what we wanted here.
2656 Finally, here is an example where a subquery is used
2657 for expressing unary negation:
2659 my ($sub_stmt, @sub_bind)
2660 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
2661 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
2663 lname => {like => '%son%'},
2664 \["NOT ($sub_stmt)" => @sub_bind],
2669 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
2670 @bind = ('%son%', 10, 20)
2672 =head3 Deprecated usage of Literal SQL
2674 Below are some examples of archaic use of literal SQL. It is shown only as
2675 reference for those who deal with legacy code. Each example has a much
2676 better, cleaner and safer alternative that users should opt for in new code.
2682 my %where = ( requestor => \'IS NOT NULL' )
2684 $stmt = "WHERE requestor IS NOT NULL"
2686 This used to be the way of generating NULL comparisons, before the handling
2687 of C<undef> got formalized. For new code please use the superior syntax as
2688 described in L</Tests for NULL values>.
2692 my %where = ( requestor => \'= submitter' )
2694 $stmt = "WHERE requestor = submitter"
2696 This used to be the only way to compare columns. Use the superior L</-ident>
2697 method for all new code. For example an identifier declared in such a way
2698 will be properly quoted if L</quote_char> is properly set, while the legacy
2699 form will remain as supplied.
2703 my %where = ( is_ready => \"", completed => { '>', '2012-12-21' } )
2705 $stmt = "WHERE completed > ? AND is_ready"
2706 @bind = ('2012-12-21')
2708 Using an empty string literal used to be the only way to express a boolean.
2709 For all new code please use the much more readable
2710 L<-bool|/Unary operators: bool> operator.
2716 These pages could go on for a while, since the nesting of the data
2717 structures this module can handle are pretty much unlimited (the
2718 module implements the C<WHERE> expansion as a recursive function
2719 internally). Your best bet is to "play around" with the module a
2720 little to see how the data structures behave, and choose the best
2721 format for your data based on that.
2723 And of course, all the values above will probably be replaced with
2724 variables gotten from forms or the command line. After all, if you
2725 knew everything ahead of time, you wouldn't have to worry about
2726 dynamically-generating SQL and could just hardwire it into your
2729 =head1 ORDER BY CLAUSES
2731 Some functions take an order by clause. This can either be a scalar (just a
2732 column name), a hashref of C<< { -desc => 'col' } >> or C<< { -asc => 'col' }
2733 >>, a scalarref, an arrayref-ref, or an arrayref of any of the previous
2736 Given | Will Generate
2737 ---------------------------------------------------------------
2739 'colA' | ORDER BY colA
2741 [qw/colA colB/] | ORDER BY colA, colB
2743 {-asc => 'colA'} | ORDER BY colA ASC
2745 {-desc => 'colB'} | ORDER BY colB DESC
2747 ['colA', {-asc => 'colB'}] | ORDER BY colA, colB ASC
2749 { -asc => [qw/colA colB/] } | ORDER BY colA ASC, colB ASC
2751 \'colA DESC' | ORDER BY colA DESC
2753 \[ 'FUNC(colA, ?)', $x ] | ORDER BY FUNC(colA, ?)
2754 | /* ...with $x bound to ? */
2757 { -asc => 'colA' }, | colA ASC,
2758 { -desc => [qw/colB/] }, | colB DESC,
2759 { -asc => [qw/colC colD/] },| colC ASC, colD ASC,
2760 \'colE DESC', | colE DESC,
2761 \[ 'FUNC(colF, ?)', $x ], | FUNC(colF, ?)
2762 ] | /* ...with $x bound to ? */
2763 ===============================================================
2767 =head1 SPECIAL OPERATORS
2769 my $sqlmaker = SQL::Abstract->new(special_ops => [
2773 my ($self, $field, $op, $arg) = @_;
2779 handler => 'method_name',
2783 A "special operator" is a SQL syntactic clause that can be
2784 applied to a field, instead of a usual binary operator.
2787 WHERE field IN (?, ?, ?)
2788 WHERE field BETWEEN ? AND ?
2789 WHERE MATCH(field) AGAINST (?, ?)
2791 Special operators IN and BETWEEN are fairly standard and therefore
2792 are builtin within C<SQL::Abstract> (as the overridable methods
2793 C<_where_field_IN> and C<_where_field_BETWEEN>). For other operators,
2794 like the MATCH .. AGAINST example above which is specific to MySQL,
2795 you can write your own operator handlers - supply a C<special_ops>
2796 argument to the C<new> method. That argument takes an arrayref of
2797 operator definitions; each operator definition is a hashref with two
2804 the regular expression to match the operator
2808 Either a coderef or a plain scalar method name. In both cases
2809 the expected return is C<< ($sql, @bind) >>.
2811 When supplied with a method name, it is simply called on the
2812 L<SQL::Abstract> object as:
2814 $self->$method_name($field, $op, $arg)
2818 $field is the LHS of the operator
2819 $op is the part that matched the handler regex
2822 When supplied with a coderef, it is called as:
2824 $coderef->($self, $field, $op, $arg)
2829 For example, here is an implementation
2830 of the MATCH .. AGAINST syntax for MySQL
2832 my $sqlmaker = SQL::Abstract->new(special_ops => [
2834 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
2835 {regex => qr/^match$/i,
2837 my ($self, $field, $op, $arg) = @_;
2838 $arg = [$arg] if not ref $arg;
2839 my $label = $self->_quote($field);
2840 my ($placeholder) = $self->_convert('?');
2841 my $placeholders = join ", ", (($placeholder) x @$arg);
2842 my $sql = $self->_sqlcase('match') . " ($label) "
2843 . $self->_sqlcase('against') . " ($placeholders) ";
2844 my @bind = $self->_bindtype($field, @$arg);
2845 return ($sql, @bind);
2852 =head1 UNARY OPERATORS
2854 my $sqlmaker = SQL::Abstract->new(unary_ops => [
2858 my ($self, $op, $arg) = @_;
2864 handler => 'method_name',
2868 A "unary operator" is a SQL syntactic clause that can be
2869 applied to a field - the operator goes before the field
2871 You can write your own operator handlers - supply a C<unary_ops>
2872 argument to the C<new> method. That argument takes an arrayref of
2873 operator definitions; each operator definition is a hashref with two
2880 the regular expression to match the operator
2884 Either a coderef or a plain scalar method name. In both cases
2885 the expected return is C<< $sql >>.
2887 When supplied with a method name, it is simply called on the
2888 L<SQL::Abstract> object as:
2890 $self->$method_name($op, $arg)
2894 $op is the part that matched the handler regex
2895 $arg is the RHS or argument of the operator
2897 When supplied with a coderef, it is called as:
2899 $coderef->($self, $op, $arg)
2907 Thanks to some benchmarking by Mark Stosberg, it turns out that
2908 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
2909 I must admit this wasn't an intentional design issue, but it's a
2910 byproduct of the fact that you get to control your C<DBI> handles
2913 To maximize performance, use a code snippet like the following:
2915 # prepare a statement handle using the first row
2916 # and then reuse it for the rest of the rows
2918 for my $href (@array_of_hashrefs) {
2919 $stmt ||= $sql->insert('table', $href);
2920 $sth ||= $dbh->prepare($stmt);
2921 $sth->execute($sql->values($href));
2924 The reason this works is because the keys in your C<$href> are sorted
2925 internally by B<SQL::Abstract>. Thus, as long as your data retains
2926 the same structure, you only have to generate the SQL the first time
2927 around. On subsequent queries, simply use the C<values> function provided
2928 by this module to return your values in the correct order.
2930 However this depends on the values having the same type - if, for
2931 example, the values of a where clause may either have values
2932 (resulting in sql of the form C<column = ?> with a single bind
2933 value), or alternatively the values might be C<undef> (resulting in
2934 sql of the form C<column IS NULL> with no bind value) then the
2935 caching technique suggested will not work.
2939 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
2940 really like this part (I do, at least). Building up a complex query
2941 can be as simple as the following:
2948 use CGI::FormBuilder;
2951 my $form = CGI::FormBuilder->new(...);
2952 my $sql = SQL::Abstract->new;
2954 if ($form->submitted) {
2955 my $field = $form->field;
2956 my $id = delete $field->{id};
2957 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
2960 Of course, you would still have to connect using C<DBI> to run the
2961 query, but the point is that if you make your form look like your
2962 table, the actual query script can be extremely simplistic.
2964 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
2965 a fast interface to returning and formatting data. I frequently
2966 use these three modules together to write complex database query
2967 apps in under 50 lines.
2969 =head1 HOW TO CONTRIBUTE
2971 Contributions are always welcome, in all usable forms (we especially
2972 welcome documentation improvements). The delivery methods include git-
2973 or unified-diff formatted patches, GitHub pull requests, or plain bug
2974 reports either via RT or the Mailing list. Contributors are generally
2975 granted full access to the official repository after their first several
2976 patches pass successful review.
2978 This project is maintained in a git repository. The code and related tools are
2979 accessible at the following locations:
2983 =item * Official repo: L<git://git.shadowcat.co.uk/dbsrgits/SQL-Abstract.git>
2985 =item * Official gitweb: L<http://git.shadowcat.co.uk/gitweb/gitweb.cgi?p=dbsrgits/SQL-Abstract.git>
2987 =item * GitHub mirror: L<https://github.com/dbsrgits/sql-abstract>
2989 =item * Authorized committers: L<ssh://dbsrgits@git.shadowcat.co.uk/SQL-Abstract.git>
2995 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
2996 Great care has been taken to preserve the I<published> behavior
2997 documented in previous versions in the 1.* family; however,
2998 some features that were previously undocumented, or behaved
2999 differently from the documentation, had to be changed in order
3000 to clarify the semantics. Hence, client code that was relying
3001 on some dark areas of C<SQL::Abstract> v1.*
3002 B<might behave differently> in v1.50.
3004 The main changes are:
3010 support for literal SQL through the C<< \ [ $sql, @bind ] >> syntax.
3014 support for the { operator => \"..." } construct (to embed literal SQL)
3018 support for the { operator => \["...", @bind] } construct (to embed literal SQL with bind values)
3022 optional support for L<array datatypes|/"Inserting and Updating Arrays">
3026 defensive programming: check arguments
3030 fixed bug with global logic, which was previously implemented
3031 through global variables yielding side-effects. Prior versions would
3032 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
3033 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
3034 Now this is interpreted
3035 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
3040 fixed semantics of _bindtype on array args
3044 dropped the C<_anoncopy> of the %where tree. No longer necessary,
3045 we just avoid shifting arrays within that tree.
3049 dropped the C<_modlogic> function
3053 =head1 ACKNOWLEDGEMENTS
3055 There are a number of individuals that have really helped out with
3056 this module. Unfortunately, most of them submitted bugs via CPAN
3057 so I have no idea who they are! But the people I do know are:
3059 Ash Berlin (order_by hash term support)
3060 Matt Trout (DBIx::Class support)
3061 Mark Stosberg (benchmarking)
3062 Chas Owens (initial "IN" operator support)
3063 Philip Collins (per-field SQL functions)
3064 Eric Kolve (hashref "AND" support)
3065 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
3066 Dan Kubb (support for "quote_char" and "name_sep")
3067 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
3068 Laurent Dami (internal refactoring, extensible list of special operators, literal SQL)
3069 Norbert Buchmuller (support for literal SQL in hashpair, misc. fixes & tests)
3070 Peter Rabbitson (rewrite of SQLA::Test, misc. fixes & tests)
3071 Oliver Charles (support for "RETURNING" after "INSERT")
3077 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
3081 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
3083 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
3085 For support, your best bet is to try the C<DBIx::Class> users mailing list.
3086 While not an official support venue, C<DBIx::Class> makes heavy use of
3087 C<SQL::Abstract>, and as such list members there are very familiar with
3088 how to create queries.
3092 This module is free software; you may copy this under the same
3093 terms as perl itself (either the GNU General Public License or
3094 the Artistic License)