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/^ is (?: \s+ not )? $/ix, handler => sub { die "NOPE" }},
42 {regex => qr/^ (?: not \s )? in $/ix, handler => sub { die "NOPE" }},
43 {regex => qr/^ ident $/ix, handler => sub { die "NOPE" }},
44 {regex => qr/^ value $/ix, handler => sub { die "NOPE" }},
47 #======================================================================
48 # DEBUGGING AND ERROR REPORTING
49 #======================================================================
52 return unless $_[0]->{debug}; shift; # a little faster
53 my $func = (caller(1))[3];
54 warn "[$func] ", @_, "\n";
58 my($func) = (caller(1))[3];
59 Carp::carp "[$func] Warning: ", @_;
63 my($func) = (caller(1))[3];
64 Carp::croak "[$func] Fatal: ", @_;
67 sub is_literal_value ($) {
68 ref $_[0] eq 'SCALAR' ? [ ${$_[0]} ]
69 : ( ref $_[0] eq 'REF' and ref ${$_[0]} eq 'ARRAY' ) ? [ @${ $_[0] } ]
73 sub is_undef_value ($) {
77 and exists $_[0]->{-value}
78 and not defined $_[0]->{-value}
82 # FIXME XSify - this can be done so much more efficiently
83 sub is_plain_value ($) {
85 ! length ref $_[0] ? \($_[0])
87 ref $_[0] eq 'HASH' and keys %{$_[0]} == 1
89 exists $_[0]->{-value}
90 ) ? \($_[0]->{-value})
92 # reuse @_ for even moar speedz
93 defined ( $_[1] = Scalar::Util::blessed $_[0] )
95 # deliberately not using Devel::OverloadInfo - the checks we are
96 # intersted in are much more limited than the fullblown thing, and
97 # this is a very hot piece of code
99 # simply using ->can('(""') can leave behind stub methods that
100 # break actually using the overload later (see L<perldiag/Stub
101 # found while resolving method "%s" overloading "%s" in package
102 # "%s"> and the source of overload::mycan())
104 # either has stringification which DBI SHOULD prefer out of the box
105 grep { *{ (qq[${_}::(""]) }{CODE} } @{ $_[2] = mro::get_linear_isa( $_[1] ) }
107 # has nummification or boolification, AND fallback is *not* disabled
109 SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION
112 grep { *{"${_}::(0+"}{CODE} } @{$_[2]}
114 grep { *{"${_}::(bool"}{CODE} } @{$_[2]}
118 # no fallback specified at all
119 ! ( ($_[3]) = grep { *{"${_}::()"}{CODE} } @{$_[2]} )
121 # fallback explicitly undef
122 ! defined ${"$_[3]::()"}
135 #======================================================================
137 #======================================================================
141 my $class = ref($self) || $self;
142 my %opt = (ref $_[0] eq 'HASH') ? %{$_[0]} : @_;
144 # choose our case by keeping an option around
145 delete $opt{case} if $opt{case} && $opt{case} ne 'lower';
147 # default logic for interpreting arrayrefs
148 $opt{logic} = $opt{logic} ? uc $opt{logic} : 'OR';
150 # how to return bind vars
151 $opt{bindtype} ||= 'normal';
153 # default comparison is "=", but can be overridden
156 # try to recognize which are the 'equality' and 'inequality' ops
157 # (temporary quickfix (in 2007), should go through a more seasoned API)
158 $opt{equality_op} = qr/^( \Q$opt{cmp}\E | \= )$/ix;
159 $opt{inequality_op} = qr/^( != | <> )$/ix;
161 $opt{like_op} = qr/^ (is_)?r?like $/xi;
162 $opt{not_like_op} = qr/^ (is_)?not_r?like $/xi;
165 $opt{sqltrue} ||= '1=1';
166 $opt{sqlfalse} ||= '0=1';
169 $opt{special_ops} ||= [];
171 if ($class->isa('DBIx::Class::SQLMaker')) {
172 $opt{warn_once_on_nest} = 1;
173 $opt{disable_old_special_ops} = 1;
177 $opt{unary_ops} ||= [];
179 # rudimentary sanity-check for user supplied bits treated as functions/operators
180 # If a purported function matches this regular expression, an exception is thrown.
181 # Literal SQL is *NOT* subject to this check, only functions (and column names
182 # when quoting is not in effect)
185 # need to guard against ()'s in column names too, but this will break tons of
186 # hacks... ideas anyone?
187 $opt{injection_guard} ||= qr/
193 $opt{expand_unary} = {};
196 not => '_expand_not',
197 bool => '_expand_bool',
198 and => '_expand_op_andor',
199 or => '_expand_op_andor',
200 nest => '_expand_nest',
201 bind => '_expand_bind',
203 not_in => '_expand_in',
204 row => '_expand_row',
205 between => '_expand_between',
206 not_between => '_expand_between',
208 (map +($_ => '_expand_op_is'), ('is', 'is_not')),
209 ident => '_expand_ident',
210 value => '_expand_value',
214 'between' => '_expand_between',
215 'not_between' => '_expand_between',
216 'in' => '_expand_in',
217 'not_in' => '_expand_in',
218 'nest' => '_expand_nest',
219 (map +($_ => '_expand_op_andor'), ('and', 'or')),
220 (map +($_ => '_expand_op_is'), ('is', 'is_not')),
221 'ident' => '_expand_ident',
222 'value' => '_expand_value',
226 (map +($_, "_render_$_"), qw(op func bind ident literal row)),
231 (map +($_ => '_render_op_between'), 'between', 'not_between'),
232 (map +($_ => '_render_op_in'), 'in', 'not_in'),
233 (map +($_ => '_render_unop_postfix'),
234 'is_null', 'is_not_null', 'asc', 'desc',
236 (not => '_render_unop_paren'),
237 (map +($_ => '_render_op_andor'), qw(and or)),
238 ',' => '_render_op_multop',
241 return bless \%opt, $class;
244 sub sqltrue { +{ -literal => [ $_[0]->{sqltrue} ] } }
245 sub sqlfalse { +{ -literal => [ $_[0]->{sqlfalse} ] } }
247 sub _assert_pass_injection_guard {
248 if ($_[1] =~ $_[0]->{injection_guard}) {
249 my $class = ref $_[0];
250 puke "Possible SQL injection attempt '$_[1]'. If this is indeed a part of the "
251 . "desired SQL use literal SQL ( \'...' or \[ '...' ] ) or supply your own "
252 . "{injection_guard} attribute to ${class}->new()"
257 #======================================================================
259 #======================================================================
263 my $table = $self->_table(shift);
264 my $data = shift || return;
270 if (is_literal_value($data)) {
271 $self->expand_expr($data);
273 ($fields, my $values) = (
274 ref($data) eq 'HASH' ?
275 ([ sort keys %$data ], [ @{$data}{sort keys %$data} ])
279 # no names (arrayref) means can't generate bindtype
280 !($fields) && $self->{bindtype} eq 'columns'
281 && belch "can't do 'columns' bindtype when called with arrayref";
285 local our $Cur_Col_Meta = $fields->[$_];
286 $self->_expand_insert_value($values->[$_])
292 my $f_aqt = (@$fields
293 ? $self->expand_expr({ -row => $fields }, -ident)
297 my @parts = ([ $self->_sqlcase('insert into').' '.$table ]);
298 push @parts, [ $self->render_aqt($f_aqt) ] if $f_aqt;
299 push @parts, [ $self->render_aqt($v_aqt) ];
300 $parts[-1][0] =~ s/^/VALUES /;
302 if ($options->{returning}) {
303 push @parts, [ $self->_insert_returning($options) ];
306 my ($sql, @bind) = $self->_join_parts(' ', @parts);
308 return wantarray ? ($sql, @bind) : $sql;
311 # So that subclasses can override INSERT ... RETURNING separately from
312 # UPDATE and DELETE (e.g. DBIx::Class::SQLMaker::Oracle does this)
313 sub _insert_returning { shift->_returning(@_) }
316 my ($self, $options) = @_;
318 my $f = $options->{returning};
320 my ($sql, @bind) = $self->render_aqt(
321 $self->_expand_maybe_list_expr($f, -ident)
324 ? $self->_sqlcase(' returning ') . $sql
325 : ($self->_sqlcase(' returning ').$sql, @bind);
328 sub _expand_insert_value {
331 my $k = our $Cur_Col_Meta;
333 if (ref($v) eq 'ARRAY') {
334 if ($self->{array_datatypes}) {
335 return +{ -bind => [ $k, $v ] };
337 my ($sql, @bind) = @$v;
338 $self->_assert_bindval_matches_bindtype(@bind);
339 return +{ -literal => $v };
341 if (ref($v) eq 'HASH') {
342 if (grep !/^-/, keys %$v) {
343 belch "HASH ref as bind value in insert is not supported";
344 return +{ -bind => [ $k, $v ] };
348 return +{ -bind => [ $k, undef ] };
350 return $self->expand_expr($v);
355 #======================================================================
357 #======================================================================
362 my $table = $self->_table(shift);
363 my $data = shift || return;
367 # first build the 'SET' part of the sql statement
368 puke "Unsupported data type specified to \$sql->update"
369 unless ref $data eq 'HASH';
371 my ($sql, @all_bind) = $self->_update_set_values($data);
372 $sql = $self->_sqlcase('update ') . $table . $self->_sqlcase(' set ')
376 my($where_sql, @where_bind) = $self->where($where);
378 push @all_bind, @where_bind;
381 if ($options->{returning}) {
382 my ($returning_sql, @returning_bind) = $self->_update_returning($options);
383 $sql .= $returning_sql;
384 push @all_bind, @returning_bind;
387 return wantarray ? ($sql, @all_bind) : $sql;
390 sub _update_set_values {
391 my ($self, $data) = @_;
393 return $self->render_aqt(
394 $self->_expand_update_set_values($data),
398 sub _expand_update_set_values {
399 my ($self, $data) = @_;
400 $self->_expand_maybe_list_expr( [
403 $set = { -bind => $_ } unless defined $set;
404 +{ -op => [ '=', $self->_expand_ident(-ident => $k), $set ] };
410 ? ($self->{array_datatypes}
411 ? [ $k, +{ -bind => [ $k, $v ] } ]
412 : [ $k, +{ -literal => $v } ])
414 local our $Cur_Col_Meta = $k;
415 [ $k, $self->_expand_expr($v) ]
422 # So that subclasses can override UPDATE ... RETURNING separately from
424 sub _update_returning { shift->_returning(@_) }
428 #======================================================================
430 #======================================================================
435 my $table = $self->_table(shift);
436 my $fields = shift || '*';
440 my ($fields_sql, @bind) = $self->_select_fields($fields);
442 my ($where_sql, @where_bind) = $self->where($where, $order);
443 push @bind, @where_bind;
445 my $sql = join(' ', $self->_sqlcase('select'), $fields_sql,
446 $self->_sqlcase('from'), $table)
449 return wantarray ? ($sql, @bind) : $sql;
453 my ($self, $fields) = @_;
454 return $fields unless ref($fields);
455 return $self->render_aqt(
456 $self->_expand_maybe_list_expr($fields, '-ident')
460 #======================================================================
462 #======================================================================
467 my $table = $self->_table(shift);
471 my($where_sql, @bind) = $self->where($where);
472 my $sql = $self->_sqlcase('delete from ') . $table . $where_sql;
474 if ($options->{returning}) {
475 my ($returning_sql, @returning_bind) = $self->_delete_returning($options);
476 $sql .= $returning_sql;
477 push @bind, @returning_bind;
480 return wantarray ? ($sql, @bind) : $sql;
483 # So that subclasses can override DELETE ... RETURNING separately from
485 sub _delete_returning { shift->_returning(@_) }
489 #======================================================================
491 #======================================================================
495 # Finally, a separate routine just to handle WHERE clauses
497 my ($self, $where, $order) = @_;
499 local $self->{convert_where} = $self->{convert};
502 my ($sql, @bind) = defined($where)
503 ? $self->_recurse_where($where)
505 $sql = (defined $sql and length $sql) ? $self->_sqlcase(' where ') . "( $sql )" : '';
509 my ($order_sql, @order_bind) = $self->_order_by($order);
511 push @bind, @order_bind;
514 return wantarray ? ($sql, @bind) : $sql;
517 { our $Default_Scalar_To = -value }
520 my ($self, $expr, $default_scalar_to) = @_;
521 local our $Default_Scalar_To = $default_scalar_to if $default_scalar_to;
522 $self->_expand_expr($expr);
526 my ($self, $aqt) = @_;
527 my ($k, $v, @rest) = %$aqt;
529 die "Not a node type: $k" unless $k =~ s/^-//;
530 if (my $meth = $self->{render}{$k}) {
531 return $self->$meth($v);
533 die "notreached: $k";
537 my ($self, $expr, $default_scalar_to) = @_;
538 $self->render_aqt($self->expand_expr($expr, $default_scalar_to));
542 my ($self, $raw) = @_;
543 s/^-(?=.)//, s/\s+/_/g for my $op = lc $raw;
548 my ($self, $expr) = @_;
549 our $Expand_Depth ||= 0; local $Expand_Depth = $Expand_Depth + 1;
550 return undef unless defined($expr);
551 if (ref($expr) eq 'HASH') {
552 return undef unless my $kc = keys %$expr;
554 return $self->_expand_op_andor(and => $expr);
556 my ($key, $value) = %$expr;
557 if ($key =~ /^-/ and $key =~ s/ [_\s]? \d+ $//x ) {
558 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
559 . "You probably wanted ...-and => [ $key => COND1, $key => COND2 ... ]";
561 return $self->_expand_hashpair($key, $value);
563 if (ref($expr) eq 'ARRAY') {
564 return $self->_expand_op_andor(lc($self->{logic}), $expr);
566 if (my $literal = is_literal_value($expr)) {
567 return +{ -literal => $literal };
569 if (!ref($expr) or Scalar::Util::blessed($expr)) {
570 return $self->_expand_scalar($expr);
575 sub _expand_hashpair {
576 my ($self, $k, $v) = @_;
577 unless (defined($k) and length($k)) {
578 if (defined($k) and my $literal = is_literal_value($v)) {
579 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
580 return { -literal => $literal };
582 puke "Supplying an empty left hand side argument is not supported";
585 return $self->_expand_hashpair_op($k, $v);
587 return $self->_expand_hashpair_ident($k, $v);
590 sub _expand_hashpair_ident {
591 my ($self, $k, $v) = @_;
593 local our $Cur_Col_Meta = $k;
595 # hash with multiple or no elements is andor
597 if (ref($v) eq 'HASH' and keys %$v != 1) {
598 return $self->_expand_op_andor(and => $v, $k);
601 # undef needs to be re-sent with cmp to achieve IS/IS NOT NULL
603 if (is_undef_value($v)) {
604 return $self->_expand_hashpair_cmp($k => undef);
607 # scalars and objects get expanded as whatever requested or values
609 if (!ref($v) or Scalar::Util::blessed($v)) {
610 return $self->_expand_hashpair_scalar($k, $v);
613 # single key hashref is a hashtriple
615 if (ref($v) eq 'HASH') {
616 return $self->_expand_hashtriple($k, %$v);
619 # arrayref needs re-engineering over the elements
621 if (ref($v) eq 'ARRAY') {
622 return $self->sqlfalse unless @$v;
623 $self->_debug("ARRAY($k) means distribute over elements");
625 $v->[0] =~ /^-(and|or)$/i
626 ? (shift(@{$v = [ @$v ]}), $1)
627 : lc($self->{logic} || 'OR')
629 return $self->_expand_op_andor(
634 if (my $literal = is_literal_value($v)) {
636 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
639 my ($sql, @bind) = @$literal;
640 if ($self->{bindtype} eq 'columns') {
642 $self->_assert_bindval_matches_bindtype($_);
645 return +{ -literal => [ $self->_quote($k).' '.$sql, @bind ] };
651 my ($self, $expr) = @_;
653 return $self->_expand_expr({ (our $Default_Scalar_To) => $expr });
656 sub _expand_hashpair_scalar {
657 my ($self, $k, $v) = @_;
659 return $self->_expand_hashpair_cmp(
660 $k, $self->_expand_scalar($v),
664 sub _expand_hashpair_op {
665 my ($self, $k, $v) = @_;
667 $self->_assert_pass_injection_guard($k =~ /\A-(.*)\Z/s);
669 my $op = $self->_normalize_op($k);
671 if (my $exp = $self->{expand}{$op}) {
672 return $self->$exp($op, $v);
675 # Ops prefixed with -not_ get converted
677 if (my ($rest) = $op =~/^not_(.*)$/) {
680 $self->_expand_expr({ "-${rest}", $v })
686 my $op = join(' ', split '_', $op);
688 # the old special op system requires illegality for top-level use
691 (our $Expand_Depth) == 1
693 List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}
695 $self->{disable_old_special_ops}
696 and List::Util::first { $op =~ $_->{regex} } @BUILTIN_SPECIAL_OPS
700 puke "Illegal use of top-level '-$op'"
703 # the old unary op system means we should touch nothing and let it work
705 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
706 return { -op => [ $op, $v ] };
710 # an explicit node type is currently assumed to be expanded (this is almost
711 # certainly wrong and there should be expansion anyway)
713 if ($self->{render}{$op}) {
717 # hashref RHS values get expanded and used as op/func args
722 and (keys %$v)[0] =~ /^-/
724 my ($func) = $k =~ /^-(.*)$/;
726 if (List::Util::first { $func =~ $_->{regex} } @{$self->{special_ops}}) {
727 return +{ -op => [ $func, $self->_expand_expr($v) ] };
730 return +{ -func => [ $func, $self->_expand_expr($v) ] };
733 # scalars and literals get simply expanded
735 if (!ref($v) or is_literal_value($v)) {
736 return +{ -op => [ $op, $self->_expand_expr($v) ] };
742 sub _expand_hashpair_cmp {
743 my ($self, $k, $v) = @_;
744 $self->_expand_hashtriple($k, $self->{cmp}, $v);
747 sub _expand_hashtriple {
748 my ($self, $k, $vk, $vv) = @_;
750 my $ik = $self->_expand_ident(-ident => $k);
752 my $op = $self->_normalize_op($vk);
753 $self->_assert_pass_injection_guard($op);
755 if ($op =~ s/ _? \d+ $//x ) {
756 return $self->_expand_expr($k, { $vk, $vv });
758 if (my $x = $self->{expand_op}{$op}) {
759 local our $Cur_Col_Meta = $k;
760 return $self->$x($op, $vv, $k);
764 my $op = join(' ', split '_', $op);
766 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}) {
767 return { -op => [ $op, $ik, $vv ] };
769 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
773 { -op => [ $op, $vv ] }
777 if (ref($vv) eq 'ARRAY') {
779 my $logic = (defined($raw[0]) and $raw[0] =~ /^-(and|or)$/i)
780 ? (shift(@raw), $1) : 'or';
781 my @values = map +{ $vk => $_ }, @raw;
783 $op =~ $self->{inequality_op}
784 or $op =~ $self->{not_like_op}
786 if (lc($logic) eq 'or' and @values > 1) {
787 belch "A multi-element arrayref as an argument to the inequality op '${\uc(join ' ', split '_', $op)}' "
788 . 'is technically equivalent to an always-true 1=1 (you probably wanted '
789 . "to say ...{ \$inequality_op => [ -and => \@values ] }... instead)"
794 # try to DWIM on equality operators
795 return ($self->_dwim_op_to_is($op,
796 "Supplying an empty arrayref to '%s' is deprecated",
797 "operator '%s' applied on an empty array (field '$k')"
798 ) ? $self->sqlfalse : $self->sqltrue);
800 return $self->_expand_op_andor($logic => \@values, $k);
802 if (is_undef_value($vv)) {
803 my $is = ($self->_dwim_op_to_is($op,
804 "Supplying an undefined argument to '%s' is deprecated",
805 "unexpected operator '%s' with undef operand",
806 ) ? 'is' : 'is not');
808 return $self->_expand_hashpair($k => { $is, undef });
810 local our $Cur_Col_Meta = $k;
814 $self->_expand_expr($vv)
819 my ($self, $raw, $empty, $fail) = @_;
821 my $op = $self->_normalize_op($raw);
823 if ($op =~ /^not$/i) {
826 if ($op =~ $self->{equality_op}) {
829 if ($op =~ $self->{like_op}) {
830 belch(sprintf $empty, uc(join ' ', split '_', $op));
833 if ($op =~ $self->{inequality_op}) {
836 if ($op =~ $self->{not_like_op}) {
837 belch(sprintf $empty, uc(join ' ', split '_', $op));
840 puke(sprintf $fail, $op);
844 my ($self, undef, $body, $k) = @_;
845 return $self->_expand_hashpair_cmp(
846 $k, { -ident => $body }
848 unless (defined($body) or (ref($body) and ref($body) eq 'ARRAY')) {
849 puke "-ident requires a single plain scalar argument (a quotable identifier) or an arrayref of identifier parts";
851 my @parts = map split(/\Q${\($self->{name_sep}||'.')}\E/, $_),
852 ref($body) ? @$body : $body;
853 return { -ident => $parts[-1] } if $self->{_dequalify_idents};
854 unless ($self->{quote_char}) {
855 $self->_assert_pass_injection_guard($_) for @parts;
857 return +{ -ident => \@parts };
861 return $_[0]->_expand_hashpair_cmp(
862 $_[3], { -value => $_[2] },
864 +{ -bind => [ our $Cur_Col_Meta, $_[2] ] };
868 +{ -op => [ 'not', $_[0]->_expand_expr($_[2]) ] };
872 my ($self, undef, $args) = @_;
873 +{ -row => [ map $self->expand_expr($_), @$args ] };
877 my ($self, undef, $args) = @_;
878 my ($op, @opargs) = @$args;
879 if (my $exp = $self->{expand_op}{$op}) {
880 return $self->$exp($op, \@opargs);
882 +{ -op => [ $op, map $self->expand_expr($_), @opargs ] };
886 my ($self, undef, $v) = @_;
888 return $self->_expand_expr($v);
890 puke "-bool => undef not supported" unless defined($v);
891 return $self->_expand_ident(-ident => $v);
894 sub _expand_op_andor {
895 my ($self, $logop, $v, $k) = @_;
897 $v = [ map +{ $k, $_ },
899 ? (map +{ $_ => $v->{$_} }, sort keys %$v)
903 if (ref($v) eq 'HASH') {
904 return undef unless keys %$v;
907 map $self->_expand_expr({ $_ => $v->{$_} }),
911 if (ref($v) eq 'ARRAY') {
912 $logop eq 'and' or $logop eq 'or' or puke "unknown logic: $logop";
915 (ref($_) eq 'ARRAY' and @$_)
916 or (ref($_) eq 'HASH' and %$_)
922 while (my ($el) = splice @expr, 0, 1) {
923 puke "Supplying an empty left hand side argument is not supported in array-pairs"
924 unless defined($el) and length($el);
925 my $elref = ref($el);
927 local our $Expand_Depth = 0;
928 push(@res, grep defined, $self->_expand_expr({ $el, shift(@expr) }));
929 } elsif ($elref eq 'ARRAY') {
930 push(@res, grep defined, $self->_expand_expr($el)) if @$el;
931 } elsif (my $l = is_literal_value($el)) {
932 push @res, { -literal => $l };
933 } elsif ($elref eq 'HASH') {
934 local our $Expand_Depth = 0;
935 push @res, grep defined, $self->_expand_expr($el) if %$el;
941 # return $res[0] if @res == 1;
942 return { -op => [ $logop, @res ] };
948 my ($self, $op, $vv, $k) = @_;
949 ($k, $vv) = @$vv unless defined $k;
950 puke "$op can only take undef as argument"
954 and exists($vv->{-value})
955 and !defined($vv->{-value})
957 return +{ -op => [ $op.'_null', $self->expand_expr($k, -ident) ] };
960 sub _expand_between {
961 my ($self, $op, $vv, $k) = @_;
962 $k = shift @{$vv = [ @$vv ]} unless defined $k;
963 my @rhs = map $self->_expand_expr($_),
964 ref($vv) eq 'ARRAY' ? @$vv : $vv;
966 (@rhs == 1 and ref($rhs[0]) eq 'HASH' and $rhs[0]->{-literal})
968 (@rhs == 2 and defined($rhs[0]) and defined($rhs[1]))
970 puke "Operator '${\uc($op)}' requires either an arrayref with two defined values or expressions, or a single literal scalarref/arrayref-ref";
974 $self->expand_expr(ref($k) ? $k : { -ident => $k }),
980 my ($self, $op, $vv, $k) = @_;
981 $k = shift @{$vv = [ @$vv ]} unless defined $k;
982 if (my $literal = is_literal_value($vv)) {
983 my ($sql, @bind) = @$literal;
984 my $opened_sql = $self->_open_outer_paren($sql);
986 $op, $self->expand_expr($k, -ident),
987 { -literal => [ $opened_sql, @bind ] }
991 'SQL::Abstract before v1.75 used to generate incorrect SQL when the '
992 . "-${\uc($op)} operator was given an undef-containing list: !!!AUDIT YOUR CODE "
993 . 'AND DATA!!! (the upcoming Data::Query-based version of SQL::Abstract '
994 . 'will emit the logically correct SQL instead of raising this exception)'
996 puke("Argument passed to the '${\uc($op)}' operator can not be undefined")
998 my @rhs = map $self->expand_expr($_, -value),
999 map { defined($_) ? $_: puke($undef_err) }
1000 (ref($vv) eq 'ARRAY' ? @$vv : $vv);
1001 return $self->${\($op =~ /^not/ ? 'sqltrue' : 'sqlfalse')} unless @rhs;
1005 $self->expand_expr($k, -ident),
1011 my ($self, undef, $v) = @_;
1012 # DBIx::Class requires a nest warning to be emitted once but the private
1013 # method it overrode to do so no longer exists
1014 if ($self->{warn_once_on_nest}) {
1015 unless (our $Nest_Warned) {
1017 "-nest in search conditions is deprecated, you most probably wanted:\n"
1018 .q|{..., -and => [ \%cond0, \@cond1, \'cond2', \[ 'cond3', [ col => bind ] ], etc. ], ... }|
1023 return $self->_expand_expr($v);
1027 my ($self, undef, $bind) = @_;
1028 return { -bind => $bind };
1031 sub _recurse_where {
1032 my ($self, $where, $logic) = @_;
1034 # Special case: top level simple string treated as literal
1036 my $where_exp = (ref($where)
1037 ? $self->_expand_expr($where, $logic)
1038 : { -literal => [ $where ] });
1040 # dispatch expanded expression
1042 my ($sql, @bind) = defined($where_exp) ? $self->render_aqt($where_exp) : (undef);
1043 # DBIx::Class used to call _recurse_where in scalar context
1044 # something else might too...
1046 return ($sql, @bind);
1049 belch "Calling _recurse_where in scalar context is deprecated and will go away before 2.0";
1055 my ($self, $ident) = @_;
1057 return $self->_convert($self->_quote($ident));
1061 my ($self, $values) = @_;
1062 my ($sql, @bind) = $self->_render_op([ ',', @$values ]);
1063 return "($sql)", @bind;
1067 my ($self, $rest) = @_;
1068 my ($func, @args) = @$rest;
1072 push @arg_sql, shift @x;
1074 } map [ $self->render_aqt($_) ], @args;
1075 return ($self->_sqlcase($func).'('.join(', ', @arg_sql).')', @bind);
1079 my ($self, $bind) = @_;
1080 return ($self->_convert('?'), $self->_bindtype(@$bind));
1083 sub _render_literal {
1084 my ($self, $literal) = @_;
1085 $self->_assert_bindval_matches_bindtype(@{$literal}[1..$#$literal]);
1090 my ($self, $v) = @_;
1091 my ($op, @args) = @$v;
1092 if (my $r = $self->{render_op}{$op}) {
1093 return $self->$r($op, \@args);
1098 my $op = join(' ', split '_', $op);
1100 my $ss = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}};
1101 if ($ss and @args > 1) {
1102 puke "Special op '${op}' requires first value to be identifier"
1103 unless my ($ident) = map $_->{-ident}, grep ref($_) eq 'HASH', $args[0];
1104 my $k = join(($self->{name_sep}||'.'), @$ident);
1105 local our $Expand_Depth = 1;
1106 return $self->${\($ss->{handler})}($k, $op, $args[1]);
1108 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1109 return $self->${\($us->{handler})}($op, $args[0]);
1112 return $self->_render_unop_paren($op, \@args);
1116 return $self->_render_unop_prefix($op, \@args);
1118 return $self->_render_op_multop($op, \@args);
1124 sub _render_op_between {
1125 my ($self, $op, $args) = @_;
1126 my ($left, $low, $high) = @$args;
1127 my ($rhsql, @rhbind) = do {
1129 puke "Single arg to between must be a literal"
1130 unless $low->{-literal};
1133 my ($l, $h) = map [ $self->render_aqt($_) ], $low, $high;
1134 (join(' ', $l->[0], $self->_sqlcase('and'), $h->[0]),
1135 @{$l}[1..$#$l], @{$h}[1..$#$h])
1138 my ($lhsql, @lhbind) = $self->render_aqt($left);
1142 $self->_sqlcase(join ' ', split '_', $op),
1150 my ($self, $op, $args) = @_;
1151 my ($lhs, @rhs) = @$args;
1154 my ($sql, @bind) = $self->render_aqt($_);
1155 push @in_bind, @bind;
1158 my ($lhsql, @lbind) = $self->render_aqt($lhs);
1160 $lhsql.' '.$self->_sqlcase(join ' ', split '_', $op).' ( '
1161 .join(', ', @in_sql)
1167 sub _render_op_andor {
1168 my ($self, $op, $args) = @_;
1169 my @parts = grep length($_->[0]), map [ $self->render_aqt($_) ], @$args;
1170 return '' unless @parts;
1171 return @{$parts[0]} if @parts == 1;
1172 my ($sql, @bind) = $self->_join_parts(' '.$self->_sqlcase($op).' ', @parts);
1173 return '( '.$sql.' )', @bind;
1176 sub _render_op_multop {
1177 my ($self, $op, $args) = @_;
1178 my @parts = grep length($_->[0]), map [ $self->render_aqt($_) ], @$args;
1179 return '' unless @parts;
1180 return @{$parts[0]} if @parts == 1;
1181 my $join = ($op eq ','
1183 : ' '.$self->_sqlcase(join ' ', split '_', $op).' '
1185 return $self->_join_parts($join, @parts);
1189 my ($self, $join, @parts) = @_;
1191 join($join, map $_->[0], @parts),
1192 map @{$_}[1..$#$_], @parts
1196 sub _render_unop_paren {
1197 my ($self, $op, $v) = @_;
1198 my ($sql, @bind) = $self->_render_unop_prefix($op, $v);
1199 return "(${sql})", @bind;
1202 sub _render_unop_prefix {
1203 my ($self, $op, $v) = @_;
1204 my ($expr_sql, @bind) = $self->render_aqt($v->[0]);
1206 my $op_sql = $self->_sqlcase($op); # join ' ', split '_', $op);
1207 return ("${op_sql} ${expr_sql}", @bind);
1210 sub _render_unop_postfix {
1211 my ($self, $op, $v) = @_;
1212 my ($expr_sql, @bind) = $self->render_aqt($v->[0]);
1213 my $op_sql = $self->_sqlcase(join ' ', split '_', $op);
1214 return ($expr_sql.' '.$op_sql, @bind);
1217 # Some databases (SQLite) treat col IN (1, 2) different from
1218 # col IN ( (1, 2) ). Use this to strip all outer parens while
1219 # adding them back in the corresponding method
1220 sub _open_outer_paren {
1221 my ($self, $sql) = @_;
1223 while (my ($inner) = $sql =~ /^ \s* \( (.*) \) \s* $/xs) {
1225 # there are closing parens inside, need the heavy duty machinery
1226 # to reevaluate the extraction starting from $sql (full reevaluation)
1227 if ($inner =~ /\)/) {
1228 require Text::Balanced;
1230 my (undef, $remainder) = do {
1231 # idiotic design - writes to $@ but *DOES NOT* throw exceptions
1233 Text::Balanced::extract_bracketed($sql, '()', qr/\s*/);
1236 # the entire expression needs to be a balanced bracketed thing
1237 # (after an extract no remainder sans trailing space)
1238 last if defined $remainder and $remainder =~ /\S/;
1248 #======================================================================
1250 #======================================================================
1252 sub _expand_order_by {
1253 my ($self, $arg) = @_;
1255 return unless defined($arg) and not (ref($arg) eq 'ARRAY' and !@$arg);
1257 my $expander = sub {
1258 my ($self, $dir, $expr) = @_;
1259 my @to_expand = ref($expr) eq 'ARRAY' ? @$expr : $expr;
1260 foreach my $arg (@to_expand) {
1264 and grep /^-(asc|desc)$/, keys %$arg
1266 puke "ordering direction hash passed to order by must have exactly one key (-asc or -desc)";
1270 defined($dir) ? { -op => [ $dir =~ /^-?(.*)$/ ,=> $_ ] } : $_
1272 map $self->expand_expr($_, -ident),
1273 map ref($_) eq 'ARRAY' ? @$_ : $_, @to_expand;
1274 return undef unless @exp;
1275 return undef if @exp == 1 and not defined($exp[0]);
1276 return +{ -op => [ ',', @exp ] };
1279 local @{$self->{expand}}{qw(asc desc)} = (($expander) x 2);
1281 return $self->$expander(undef, $arg);
1285 my ($self, $arg) = @_;
1287 return '' unless defined(my $expanded = $self->_expand_order_by($arg));
1289 my ($sql, @bind) = $self->render_aqt($expanded);
1291 return '' unless length($sql);
1293 my $final_sql = $self->_sqlcase(' order by ').$sql;
1295 return wantarray ? ($final_sql, @bind) : $final_sql;
1298 # _order_by no longer needs to call this so doesn't but DBIC uses it.
1300 sub _order_by_chunks {
1301 my ($self, $arg) = @_;
1303 return () unless defined(my $expanded = $self->_expand_order_by($arg));
1305 return $self->_chunkify_order_by($expanded);
1308 sub _chunkify_order_by {
1309 my ($self, $expanded) = @_;
1311 return grep length, $self->render_aqt($expanded)
1312 if $expanded->{-ident} or @{$expanded->{-literal}||[]} == 1;
1315 if (ref() eq 'HASH' and $_->{-op} and $_->{-op}[0] eq ',') {
1316 my ($comma, @list) = @{$_->{-op}};
1317 return map $self->_chunkify_order_by($_), @list;
1319 return [ $self->render_aqt($_) ];
1323 #======================================================================
1324 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
1325 #======================================================================
1331 $self->_expand_maybe_list_expr($from, -ident)
1336 #======================================================================
1338 #======================================================================
1340 sub _expand_maybe_list_expr {
1341 my ($self, $expr, $default) = @_;
1342 return +{ -op => [ ',',
1343 map $self->expand_expr($_, $default),
1344 ref($expr) eq 'ARRAY' ? @$expr : $expr
1348 # highly optimized, as it's called way too often
1350 # my ($self, $label) = @_;
1352 return '' unless defined $_[1];
1353 return ${$_[1]} if ref($_[1]) eq 'SCALAR';
1354 puke 'Identifier cannot be hashref' if ref($_[1]) eq 'HASH';
1356 unless ($_[0]->{quote_char}) {
1357 if (ref($_[1]) eq 'ARRAY') {
1358 return join($_[0]->{name_sep}||'.', @{$_[1]});
1360 $_[0]->_assert_pass_injection_guard($_[1]);
1365 my $qref = ref $_[0]->{quote_char};
1367 !$qref ? ($_[0]->{quote_char}, $_[0]->{quote_char})
1368 : ($qref eq 'ARRAY') ? @{$_[0]->{quote_char}}
1369 : puke "Unsupported quote_char format: $_[0]->{quote_char}";
1371 my $esc = $_[0]->{escape_char} || $r;
1373 # parts containing * are naturally unquoted
1375 $_[0]->{name_sep}||'',
1379 : do { (my $n = $_) =~ s/(\Q$esc\E|\Q$r\E)/$esc$1/g; $l . $n . $r }
1381 (ref($_[1]) eq 'ARRAY'
1385 ? split (/\Q$_[0]->{name_sep}\E/, $_[1] )
1393 # Conversion, if applicable
1395 #my ($self, $arg) = @_;
1396 if ($_[0]->{convert_where}) {
1397 return $_[0]->_sqlcase($_[0]->{convert_where}) .'(' . $_[1] . ')';
1404 #my ($self, $col, @vals) = @_;
1405 # called often - tighten code
1406 return $_[0]->{bindtype} eq 'columns'
1407 ? map {[$_[1], $_]} @_[2 .. $#_]
1412 # Dies if any element of @bind is not in [colname => value] format
1413 # if bindtype is 'columns'.
1414 sub _assert_bindval_matches_bindtype {
1415 # my ($self, @bind) = @_;
1417 if ($self->{bindtype} eq 'columns') {
1419 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
1420 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
1426 sub _join_sql_clauses {
1427 my ($self, $logic, $clauses_aref, $bind_aref) = @_;
1429 if (@$clauses_aref > 1) {
1430 my $join = " " . $self->_sqlcase($logic) . " ";
1431 my $sql = '( ' . join($join, @$clauses_aref) . ' )';
1432 return ($sql, @$bind_aref);
1434 elsif (@$clauses_aref) {
1435 return ($clauses_aref->[0], @$bind_aref); # no parentheses
1438 return (); # if no SQL, ignore @$bind_aref
1443 # Fix SQL case, if so requested
1445 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
1446 # don't touch the argument ... crooked logic, but let's not change it!
1447 return $_[0]->{case} ? $_[1] : uc($_[1]);
1451 #======================================================================
1452 # DISPATCHING FROM REFKIND
1453 #======================================================================
1456 my ($self, $data) = @_;
1458 return 'UNDEF' unless defined $data;
1460 # blessed objects are treated like scalars
1461 my $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1463 return 'SCALAR' unless $ref;
1466 while ($ref eq 'REF') {
1468 $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1472 return ($ref||'SCALAR') . ('REF' x $n_steps);
1476 my ($self, $data) = @_;
1477 my @try = ($self->_refkind($data));
1478 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
1479 push @try, 'FALLBACK';
1483 sub _METHOD_FOR_refkind {
1484 my ($self, $meth_prefix, $data) = @_;
1487 for (@{$self->_try_refkind($data)}) {
1488 $method = $self->can($meth_prefix."_".$_)
1492 return $method || puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
1496 sub _SWITCH_refkind {
1497 my ($self, $data, $dispatch_table) = @_;
1500 for (@{$self->_try_refkind($data)}) {
1501 $coderef = $dispatch_table->{$_}
1505 puke "no dispatch entry for ".$self->_refkind($data)
1514 #======================================================================
1515 # VALUES, GENERATE, AUTOLOAD
1516 #======================================================================
1518 # LDNOTE: original code from nwiger, didn't touch code in that section
1519 # I feel the AUTOLOAD stuff should not be the default, it should
1520 # only be activated on explicit demand by user.
1524 my $data = shift || return;
1525 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
1526 unless ref $data eq 'HASH';
1529 foreach my $k (sort keys %$data) {
1530 my $v = $data->{$k};
1531 $self->_SWITCH_refkind($v, {
1533 if ($self->{array_datatypes}) { # array datatype
1534 push @all_bind, $self->_bindtype($k, $v);
1536 else { # literal SQL with bind
1537 my ($sql, @bind) = @$v;
1538 $self->_assert_bindval_matches_bindtype(@bind);
1539 push @all_bind, @bind;
1542 ARRAYREFREF => sub { # literal SQL with bind
1543 my ($sql, @bind) = @${$v};
1544 $self->_assert_bindval_matches_bindtype(@bind);
1545 push @all_bind, @bind;
1547 SCALARREF => sub { # literal SQL without bind
1549 SCALAR_or_UNDEF => sub {
1550 push @all_bind, $self->_bindtype($k, $v);
1561 my(@sql, @sqlq, @sqlv);
1565 if ($ref eq 'HASH') {
1566 for my $k (sort keys %$_) {
1569 my $label = $self->_quote($k);
1570 if ($r eq 'ARRAY') {
1571 # literal SQL with bind
1572 my ($sql, @bind) = @$v;
1573 $self->_assert_bindval_matches_bindtype(@bind);
1574 push @sqlq, "$label = $sql";
1576 } elsif ($r eq 'SCALAR') {
1577 # literal SQL without bind
1578 push @sqlq, "$label = $$v";
1580 push @sqlq, "$label = ?";
1581 push @sqlv, $self->_bindtype($k, $v);
1584 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
1585 } elsif ($ref eq 'ARRAY') {
1586 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
1589 if ($r eq 'ARRAY') { # literal SQL with bind
1590 my ($sql, @bind) = @$v;
1591 $self->_assert_bindval_matches_bindtype(@bind);
1594 } elsif ($r eq 'SCALAR') { # literal SQL without bind
1595 # embedded literal SQL
1602 push @sql, '(' . join(', ', @sqlq) . ')';
1603 } elsif ($ref eq 'SCALAR') {
1607 # strings get case twiddled
1608 push @sql, $self->_sqlcase($_);
1612 my $sql = join ' ', @sql;
1614 # this is pretty tricky
1615 # if ask for an array, return ($stmt, @bind)
1616 # otherwise, s/?/shift @sqlv/ to put it inline
1618 return ($sql, @sqlv);
1620 1 while $sql =~ s/\?/my $d = shift(@sqlv);
1621 ref $d ? $d->[1] : $d/e;
1630 # This allows us to check for a local, then _form, attr
1632 my($name) = $AUTOLOAD =~ /.*::(.+)/;
1633 return $self->generate($name, @_);
1644 SQL::Abstract - Generate SQL from Perl data structures
1650 my $sql = SQL::Abstract->new;
1652 my($stmt, @bind) = $sql->select($source, \@fields, \%where, $order);
1654 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
1656 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
1658 my($stmt, @bind) = $sql->delete($table, \%where);
1660 # Then, use these in your DBI statements
1661 my $sth = $dbh->prepare($stmt);
1662 $sth->execute(@bind);
1664 # Just generate the WHERE clause
1665 my($stmt, @bind) = $sql->where(\%where, $order);
1667 # Return values in the same order, for hashed queries
1668 # See PERFORMANCE section for more details
1669 my @bind = $sql->values(\%fieldvals);
1673 This module was inspired by the excellent L<DBIx::Abstract>.
1674 However, in using that module I found that what I really wanted
1675 to do was generate SQL, but still retain complete control over my
1676 statement handles and use the DBI interface. So, I set out to
1677 create an abstract SQL generation module.
1679 While based on the concepts used by L<DBIx::Abstract>, there are
1680 several important differences, especially when it comes to WHERE
1681 clauses. I have modified the concepts used to make the SQL easier
1682 to generate from Perl data structures and, IMO, more intuitive.
1683 The underlying idea is for this module to do what you mean, based
1684 on the data structures you provide it. The big advantage is that
1685 you don't have to modify your code every time your data changes,
1686 as this module figures it out.
1688 To begin with, an SQL INSERT is as easy as just specifying a hash
1689 of C<key=value> pairs:
1692 name => 'Jimbo Bobson',
1693 phone => '123-456-7890',
1694 address => '42 Sister Lane',
1695 city => 'St. Louis',
1696 state => 'Louisiana',
1699 The SQL can then be generated with this:
1701 my($stmt, @bind) = $sql->insert('people', \%data);
1703 Which would give you something like this:
1705 $stmt = "INSERT INTO people
1706 (address, city, name, phone, state)
1707 VALUES (?, ?, ?, ?, ?)";
1708 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
1709 '123-456-7890', 'Louisiana');
1711 These are then used directly in your DBI code:
1713 my $sth = $dbh->prepare($stmt);
1714 $sth->execute(@bind);
1716 =head2 Inserting and Updating Arrays
1718 If your database has array types (like for example Postgres),
1719 activate the special option C<< array_datatypes => 1 >>
1720 when creating the C<SQL::Abstract> object.
1721 Then you may use an arrayref to insert and update database array types:
1723 my $sql = SQL::Abstract->new(array_datatypes => 1);
1725 planets => [qw/Mercury Venus Earth Mars/]
1728 my($stmt, @bind) = $sql->insert('solar_system', \%data);
1732 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
1734 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
1737 =head2 Inserting and Updating SQL
1739 In order to apply SQL functions to elements of your C<%data> you may
1740 specify a reference to an arrayref for the given hash value. For example,
1741 if you need to execute the Oracle C<to_date> function on a value, you can
1742 say something like this:
1746 date_entered => \[ "to_date(?,'MM/DD/YYYY')", "03/02/2003" ],
1749 The first value in the array is the actual SQL. Any other values are
1750 optional and would be included in the bind values array. This gives
1753 my($stmt, @bind) = $sql->insert('people', \%data);
1755 $stmt = "INSERT INTO people (name, date_entered)
1756 VALUES (?, to_date(?,'MM/DD/YYYY'))";
1757 @bind = ('Bill', '03/02/2003');
1759 An UPDATE is just as easy, all you change is the name of the function:
1761 my($stmt, @bind) = $sql->update('people', \%data);
1763 Notice that your C<%data> isn't touched; the module will generate
1764 the appropriately quirky SQL for you automatically. Usually you'll
1765 want to specify a WHERE clause for your UPDATE, though, which is
1766 where handling C<%where> hashes comes in handy...
1768 =head2 Complex where statements
1770 This module can generate pretty complicated WHERE statements
1771 easily. For example, simple C<key=value> pairs are taken to mean
1772 equality, and if you want to see if a field is within a set
1773 of values, you can use an arrayref. Let's say we wanted to
1774 SELECT some data based on this criteria:
1777 requestor => 'inna',
1778 worker => ['nwiger', 'rcwe', 'sfz'],
1779 status => { '!=', 'completed' }
1782 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
1784 The above would give you something like this:
1786 $stmt = "SELECT * FROM tickets WHERE
1787 ( requestor = ? ) AND ( status != ? )
1788 AND ( worker = ? OR worker = ? OR worker = ? )";
1789 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
1791 Which you could then use in DBI code like so:
1793 my $sth = $dbh->prepare($stmt);
1794 $sth->execute(@bind);
1800 The methods are simple. There's one for every major SQL operation,
1801 and a constructor you use first. The arguments are specified in a
1802 similar order for each method (table, then fields, then a where
1803 clause) to try and simplify things.
1805 =head2 new(option => 'value')
1807 The C<new()> function takes a list of options and values, and returns
1808 a new B<SQL::Abstract> object which can then be used to generate SQL
1809 through the methods below. The options accepted are:
1815 If set to 'lower', then SQL will be generated in all lowercase. By
1816 default SQL is generated in "textbook" case meaning something like:
1818 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
1820 Any setting other than 'lower' is ignored.
1824 This determines what the default comparison operator is. By default
1825 it is C<=>, meaning that a hash like this:
1827 %where = (name => 'nwiger', email => 'nate@wiger.org');
1829 Will generate SQL like this:
1831 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
1833 However, you may want loose comparisons by default, so if you set
1834 C<cmp> to C<like> you would get SQL such as:
1836 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
1838 You can also override the comparison on an individual basis - see
1839 the huge section on L</"WHERE CLAUSES"> at the bottom.
1841 =item sqltrue, sqlfalse
1843 Expressions for inserting boolean values within SQL statements.
1844 By default these are C<1=1> and C<1=0>. They are used
1845 by the special operators C<-in> and C<-not_in> for generating
1846 correct SQL even when the argument is an empty array (see below).
1850 This determines the default logical operator for multiple WHERE
1851 statements in arrays or hashes. If absent, the default logic is "or"
1852 for arrays, and "and" for hashes. This means that a WHERE
1856 event_date => {'>=', '2/13/99'},
1857 event_date => {'<=', '4/24/03'},
1860 will generate SQL like this:
1862 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
1864 This is probably not what you want given this query, though (look
1865 at the dates). To change the "OR" to an "AND", simply specify:
1867 my $sql = SQL::Abstract->new(logic => 'and');
1869 Which will change the above C<WHERE> to:
1871 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
1873 The logic can also be changed locally by inserting
1874 a modifier in front of an arrayref:
1876 @where = (-and => [event_date => {'>=', '2/13/99'},
1877 event_date => {'<=', '4/24/03'} ]);
1879 See the L</"WHERE CLAUSES"> section for explanations.
1883 This will automatically convert comparisons using the specified SQL
1884 function for both column and value. This is mostly used with an argument
1885 of C<upper> or C<lower>, so that the SQL will have the effect of
1886 case-insensitive "searches". For example, this:
1888 $sql = SQL::Abstract->new(convert => 'upper');
1889 %where = (keywords => 'MaKe iT CAse inSeNSItive');
1891 Will turn out the following SQL:
1893 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
1895 The conversion can be C<upper()>, C<lower()>, or any other SQL function
1896 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
1897 not validate this option; it will just pass through what you specify verbatim).
1901 This is a kludge because many databases suck. For example, you can't
1902 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
1903 Instead, you have to use C<bind_param()>:
1905 $sth->bind_param(1, 'reg data');
1906 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
1908 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
1909 which loses track of which field each slot refers to. Fear not.
1911 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
1912 Currently, you can specify either C<normal> (default) or C<columns>. If you
1913 specify C<columns>, you will get an array that looks like this:
1915 my $sql = SQL::Abstract->new(bindtype => 'columns');
1916 my($stmt, @bind) = $sql->insert(...);
1919 [ 'column1', 'value1' ],
1920 [ 'column2', 'value2' ],
1921 [ 'column3', 'value3' ],
1924 You can then iterate through this manually, using DBI's C<bind_param()>.
1926 $sth->prepare($stmt);
1929 my($col, $data) = @$_;
1930 if ($col eq 'details' || $col eq 'comments') {
1931 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
1932 } elsif ($col eq 'image') {
1933 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
1935 $sth->bind_param($i, $data);
1939 $sth->execute; # execute without @bind now
1941 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
1942 Basically, the advantage is still that you don't have to care which fields
1943 are or are not included. You could wrap that above C<for> loop in a simple
1944 sub called C<bind_fields()> or something and reuse it repeatedly. You still
1945 get a layer of abstraction over manual SQL specification.
1947 Note that if you set L</bindtype> to C<columns>, the C<\[ $sql, @bind ]>
1948 construct (see L</Literal SQL with placeholders and bind values (subqueries)>)
1949 will expect the bind values in this format.
1953 This is the character that a table or column name will be quoted
1954 with. By default this is an empty string, but you could set it to
1955 the character C<`>, to generate SQL like this:
1957 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
1959 Alternatively, you can supply an array ref of two items, the first being the left
1960 hand quote character, and the second the right hand quote character. For
1961 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
1962 that generates SQL like this:
1964 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
1966 Quoting is useful if you have tables or columns names that are reserved
1967 words in your database's SQL dialect.
1971 This is the character that will be used to escape L</quote_char>s appearing
1972 in an identifier before it has been quoted.
1974 The parameter default in case of a single L</quote_char> character is the quote
1977 When opening-closing-style quoting is used (L</quote_char> is an arrayref)
1978 this parameter defaults to the B<closing (right)> L</quote_char>. Occurrences
1979 of the B<opening (left)> L</quote_char> within the identifier are currently left
1980 untouched. The default for opening-closing-style quotes may change in future
1981 versions, thus you are B<strongly encouraged> to specify the escape character
1986 This is the character that separates a table and column name. It is
1987 necessary to specify this when the C<quote_char> option is selected,
1988 so that tables and column names can be individually quoted like this:
1990 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
1992 =item injection_guard
1994 A regular expression C<qr/.../> that is applied to any C<-function> and unquoted
1995 column name specified in a query structure. This is a safety mechanism to avoid
1996 injection attacks when mishandling user input e.g.:
1998 my %condition_as_column_value_pairs = get_values_from_user();
1999 $sqla->select( ... , \%condition_as_column_value_pairs );
2001 If the expression matches an exception is thrown. Note that literal SQL
2002 supplied via C<\'...'> or C<\['...']> is B<not> checked in any way.
2004 Defaults to checking for C<;> and the C<GO> keyword (TransactSQL)
2006 =item array_datatypes
2008 When this option is true, arrayrefs in INSERT or UPDATE are
2009 interpreted as array datatypes and are passed directly
2011 When this option is false, arrayrefs are interpreted
2012 as literal SQL, just like refs to arrayrefs
2013 (but this behavior is for backwards compatibility; when writing
2014 new queries, use the "reference to arrayref" syntax
2020 Takes a reference to a list of "special operators"
2021 to extend the syntax understood by L<SQL::Abstract>.
2022 See section L</"SPECIAL OPERATORS"> for details.
2026 Takes a reference to a list of "unary operators"
2027 to extend the syntax understood by L<SQL::Abstract>.
2028 See section L</"UNARY OPERATORS"> for details.
2034 =head2 insert($table, \@values || \%fieldvals, \%options)
2036 This is the simplest function. You simply give it a table name
2037 and either an arrayref of values or hashref of field/value pairs.
2038 It returns an SQL INSERT statement and a list of bind values.
2039 See the sections on L</"Inserting and Updating Arrays"> and
2040 L</"Inserting and Updating SQL"> for information on how to insert
2041 with those data types.
2043 The optional C<\%options> hash reference may contain additional
2044 options to generate the insert SQL. Currently supported options
2051 Takes either a scalar of raw SQL fields, or an array reference of
2052 field names, and adds on an SQL C<RETURNING> statement at the end.
2053 This allows you to return data generated by the insert statement
2054 (such as row IDs) without performing another C<SELECT> statement.
2055 Note, however, this is not part of the SQL standard and may not
2056 be supported by all database engines.
2060 =head2 update($table, \%fieldvals, \%where, \%options)
2062 This takes a table, hashref of field/value pairs, and an optional
2063 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
2065 See the sections on L</"Inserting and Updating Arrays"> and
2066 L</"Inserting and Updating SQL"> for information on how to insert
2067 with those data types.
2069 The optional C<\%options> hash reference may contain additional
2070 options to generate the update SQL. Currently supported options
2077 See the C<returning> option to
2078 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2082 =head2 select($source, $fields, $where, $order)
2084 This returns a SQL SELECT statement and associated list of bind values, as
2085 specified by the arguments:
2091 Specification of the 'FROM' part of the statement.
2092 The argument can be either a plain scalar (interpreted as a table
2093 name, will be quoted), or an arrayref (interpreted as a list
2094 of table names, joined by commas, quoted), or a scalarref
2095 (literal SQL, not quoted).
2099 Specification of the list of fields to retrieve from
2101 The argument can be either an arrayref (interpreted as a list
2102 of field names, will be joined by commas and quoted), or a
2103 plain scalar (literal SQL, not quoted).
2104 Please observe that this API is not as flexible as that of
2105 the first argument C<$source>, for backwards compatibility reasons.
2109 Optional argument to specify the WHERE part of the query.
2110 The argument is most often a hashref, but can also be
2111 an arrayref or plain scalar --
2112 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
2116 Optional argument to specify the ORDER BY part of the query.
2117 The argument can be a scalar, a hashref or an arrayref
2118 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
2124 =head2 delete($table, \%where, \%options)
2126 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
2127 It returns an SQL DELETE statement and list of bind values.
2129 The optional C<\%options> hash reference may contain additional
2130 options to generate the delete SQL. Currently supported options
2137 See the C<returning> option to
2138 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2142 =head2 where(\%where, $order)
2144 This is used to generate just the WHERE clause. For example,
2145 if you have an arbitrary data structure and know what the
2146 rest of your SQL is going to look like, but want an easy way
2147 to produce a WHERE clause, use this. It returns an SQL WHERE
2148 clause and list of bind values.
2151 =head2 values(\%data)
2153 This just returns the values from the hash C<%data>, in the same
2154 order that would be returned from any of the other above queries.
2155 Using this allows you to markedly speed up your queries if you
2156 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
2158 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
2160 Warning: This is an experimental method and subject to change.
2162 This returns arbitrarily generated SQL. It's a really basic shortcut.
2163 It will return two different things, depending on return context:
2165 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
2166 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
2168 These would return the following:
2170 # First calling form
2171 $stmt = "CREATE TABLE test (?, ?)";
2172 @bind = (field1, field2);
2174 # Second calling form
2175 $stmt_and_val = "CREATE TABLE test (field1, field2)";
2177 Depending on what you're trying to do, it's up to you to choose the correct
2178 format. In this example, the second form is what you would want.
2182 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
2186 ALTER SESSION SET nls_date_format = 'MM/YY'
2188 You get the idea. Strings get their case twiddled, but everything
2189 else remains verbatim.
2191 =head1 EXPORTABLE FUNCTIONS
2193 =head2 is_plain_value
2195 Determines if the supplied argument is a plain value as understood by this
2200 =item * The value is C<undef>
2202 =item * The value is a non-reference
2204 =item * The value is an object with stringification overloading
2206 =item * The value is of the form C<< { -value => $anything } >>
2210 On failure returns C<undef>, on success returns a B<scalar> reference
2211 to the original supplied argument.
2217 The stringification overloading detection is rather advanced: it takes
2218 into consideration not only the presence of a C<""> overload, but if that
2219 fails also checks for enabled
2220 L<autogenerated versions of C<"">|overload/Magic Autogeneration>, based
2221 on either C<0+> or C<bool>.
2223 Unfortunately testing in the field indicates that this
2224 detection B<< may tickle a latent bug in perl versions before 5.018 >>,
2225 but only when very large numbers of stringifying objects are involved.
2226 At the time of writing ( Sep 2014 ) there is no clear explanation of
2227 the direct cause, nor is there a manageably small test case that reliably
2228 reproduces the problem.
2230 If you encounter any of the following exceptions in B<random places within
2231 your application stack> - this module may be to blame:
2233 Operation "ne": no method found,
2234 left argument in overloaded package <something>,
2235 right argument in overloaded package <something>
2239 Stub found while resolving method "???" overloading """" in package <something>
2241 If you fall victim to the above - please attempt to reduce the problem
2242 to something that could be sent to the L<SQL::Abstract developers
2243 |DBIx::Class/GETTING HELP/SUPPORT>
2244 (either publicly or privately). As a workaround in the meantime you can
2245 set C<$ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}> to a true
2246 value, which will most likely eliminate your problem (at the expense of
2247 not being able to properly detect exotic forms of stringification).
2249 This notice and environment variable will be removed in a future version,
2250 as soon as the underlying problem is found and a reliable workaround is
2255 =head2 is_literal_value
2257 Determines if the supplied argument is a literal value as understood by this
2262 =item * C<\$sql_string>
2264 =item * C<\[ $sql_string, @bind_values ]>
2268 On failure returns C<undef>, on success returns an B<array> reference
2269 containing the unpacked version of the supplied literal SQL and bind values.
2271 =head1 WHERE CLAUSES
2275 This module uses a variation on the idea from L<DBIx::Abstract>. It
2276 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
2277 module is that things in arrays are OR'ed, and things in hashes
2280 The easiest way to explain is to show lots of examples. After
2281 each C<%where> hash shown, it is assumed you used:
2283 my($stmt, @bind) = $sql->where(\%where);
2285 However, note that the C<%where> hash can be used directly in any
2286 of the other functions as well, as described above.
2288 =head2 Key-value pairs
2290 So, let's get started. To begin, a simple hash:
2294 status => 'completed'
2297 Is converted to SQL C<key = val> statements:
2299 $stmt = "WHERE user = ? AND status = ?";
2300 @bind = ('nwiger', 'completed');
2302 One common thing I end up doing is having a list of values that
2303 a field can be in. To do this, simply specify a list inside of
2308 status => ['assigned', 'in-progress', 'pending'];
2311 This simple code will create the following:
2313 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
2314 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
2316 A field associated to an empty arrayref will be considered a
2317 logical false and will generate 0=1.
2319 =head2 Tests for NULL values
2321 If the value part is C<undef> then this is converted to SQL <IS NULL>
2330 $stmt = "WHERE user = ? AND status IS NULL";
2333 To test if a column IS NOT NULL:
2337 status => { '!=', undef },
2340 =head2 Specific comparison operators
2342 If you want to specify a different type of operator for your comparison,
2343 you can use a hashref for a given column:
2347 status => { '!=', 'completed' }
2350 Which would generate:
2352 $stmt = "WHERE user = ? AND status != ?";
2353 @bind = ('nwiger', 'completed');
2355 To test against multiple values, just enclose the values in an arrayref:
2357 status => { '=', ['assigned', 'in-progress', 'pending'] };
2359 Which would give you:
2361 "WHERE status = ? OR status = ? OR status = ?"
2364 The hashref can also contain multiple pairs, in which case it is expanded
2365 into an C<AND> of its elements:
2369 status => { '!=', 'completed', -not_like => 'pending%' }
2372 # Or more dynamically, like from a form
2373 $where{user} = 'nwiger';
2374 $where{status}{'!='} = 'completed';
2375 $where{status}{'-not_like'} = 'pending%';
2377 # Both generate this
2378 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
2379 @bind = ('nwiger', 'completed', 'pending%');
2382 To get an OR instead, you can combine it with the arrayref idea:
2386 priority => [ { '=', 2 }, { '>', 5 } ]
2389 Which would generate:
2391 $stmt = "WHERE ( priority = ? OR priority > ? ) AND user = ?";
2392 @bind = ('2', '5', 'nwiger');
2394 If you want to include literal SQL (with or without bind values), just use a
2395 scalar reference or reference to an arrayref as the value:
2398 date_entered => { '>' => \["to_date(?, 'MM/DD/YYYY')", "11/26/2008"] },
2399 date_expires => { '<' => \"now()" }
2402 Which would generate:
2404 $stmt = "WHERE date_entered > to_date(?, 'MM/DD/YYYY') AND date_expires < now()";
2405 @bind = ('11/26/2008');
2408 =head2 Logic and nesting operators
2410 In the example above,
2411 there is a subtle trap if you want to say something like
2412 this (notice the C<AND>):
2414 WHERE priority != ? AND priority != ?
2416 Because, in Perl you I<can't> do this:
2418 priority => { '!=' => 2, '!=' => 1 }
2420 As the second C<!=> key will obliterate the first. The solution
2421 is to use the special C<-modifier> form inside an arrayref:
2423 priority => [ -and => {'!=', 2},
2427 Normally, these would be joined by C<OR>, but the modifier tells it
2428 to use C<AND> instead. (Hint: You can use this in conjunction with the
2429 C<logic> option to C<new()> in order to change the way your queries
2430 work by default.) B<Important:> Note that the C<-modifier> goes
2431 B<INSIDE> the arrayref, as an extra first element. This will
2432 B<NOT> do what you think it might:
2434 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
2436 Here is a quick list of equivalencies, since there is some overlap:
2439 status => {'!=', 'completed', 'not like', 'pending%' }
2440 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
2443 status => {'=', ['assigned', 'in-progress']}
2444 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
2445 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
2449 =head2 Special operators: IN, BETWEEN, etc.
2451 You can also use the hashref format to compare a list of fields using the
2452 C<IN> comparison operator, by specifying the list as an arrayref:
2455 status => 'completed',
2456 reportid => { -in => [567, 2335, 2] }
2459 Which would generate:
2461 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
2462 @bind = ('completed', '567', '2335', '2');
2464 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
2467 If the argument to C<-in> is an empty array, 'sqlfalse' is generated
2468 (by default: C<1=0>). Similarly, C<< -not_in => [] >> generates
2469 'sqltrue' (by default: C<1=1>).
2471 In addition to the array you can supply a chunk of literal sql or
2472 literal sql with bind:
2475 customer => { -in => \[
2476 'SELECT cust_id FROM cust WHERE balance > ?',
2479 status => { -in => \'SELECT status_codes FROM states' },
2485 customer IN ( SELECT cust_id FROM cust WHERE balance > ? )
2486 AND status IN ( SELECT status_codes FROM states )
2490 Finally, if the argument to C<-in> is not a reference, it will be
2491 treated as a single-element array.
2493 Another pair of operators is C<-between> and C<-not_between>,
2494 used with an arrayref of two values:
2498 completion_date => {
2499 -not_between => ['2002-10-01', '2003-02-06']
2505 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
2507 Just like with C<-in> all plausible combinations of literal SQL
2511 start0 => { -between => [ 1, 2 ] },
2512 start1 => { -between => \["? AND ?", 1, 2] },
2513 start2 => { -between => \"lower(x) AND upper(y)" },
2514 start3 => { -between => [
2516 \["upper(?)", 'stuff' ],
2523 ( start0 BETWEEN ? AND ? )
2524 AND ( start1 BETWEEN ? AND ? )
2525 AND ( start2 BETWEEN lower(x) AND upper(y) )
2526 AND ( start3 BETWEEN lower(x) AND upper(?) )
2528 @bind = (1, 2, 1, 2, 'stuff');
2531 These are the two builtin "special operators"; but the
2532 list can be expanded: see section L</"SPECIAL OPERATORS"> below.
2534 =head2 Unary operators: bool
2536 If you wish to test against boolean columns or functions within your
2537 database you can use the C<-bool> and C<-not_bool> operators. For
2538 example to test the column C<is_user> being true and the column
2539 C<is_enabled> being false you would use:-
2543 -not_bool => 'is_enabled',
2548 WHERE is_user AND NOT is_enabled
2550 If a more complex combination is required, testing more conditions,
2551 then you should use the and/or operators:-
2556 -not_bool => { two=> { -rlike => 'bar' } },
2557 -not_bool => { three => [ { '=', 2 }, { '>', 5 } ] },
2568 (NOT ( three = ? OR three > ? ))
2571 =head2 Nested conditions, -and/-or prefixes
2573 So far, we've seen how multiple conditions are joined with a top-level
2574 C<AND>. We can change this by putting the different conditions we want in
2575 hashes and then putting those hashes in an array. For example:
2580 status => { -like => ['pending%', 'dispatched'] },
2584 status => 'unassigned',
2588 This data structure would create the following:
2590 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
2591 OR ( user = ? AND status = ? ) )";
2592 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
2595 Clauses in hashrefs or arrayrefs can be prefixed with an C<-and> or C<-or>
2596 to change the logic inside:
2602 -and => [ workhrs => {'>', 20}, geo => 'ASIA' ],
2603 -or => { workhrs => {'<', 50}, geo => 'EURO' },
2610 $stmt = "WHERE ( user = ?
2611 AND ( ( workhrs > ? AND geo = ? )
2612 OR ( workhrs < ? OR geo = ? ) ) )";
2613 @bind = ('nwiger', '20', 'ASIA', '50', 'EURO');
2615 =head3 Algebraic inconsistency, for historical reasons
2617 C<Important note>: when connecting several conditions, the C<-and->|C<-or>
2618 operator goes C<outside> of the nested structure; whereas when connecting
2619 several constraints on one column, the C<-and> operator goes
2620 C<inside> the arrayref. Here is an example combining both features:
2623 -and => [a => 1, b => 2],
2624 -or => [c => 3, d => 4],
2625 e => [-and => {-like => 'foo%'}, {-like => '%bar'} ]
2630 WHERE ( ( ( a = ? AND b = ? )
2631 OR ( c = ? OR d = ? )
2632 OR ( e LIKE ? AND e LIKE ? ) ) )
2634 This difference in syntax is unfortunate but must be preserved for
2635 historical reasons. So be careful: the two examples below would
2636 seem algebraically equivalent, but they are not
2639 { -like => 'foo%' },
2640 { -like => '%bar' },
2642 # yields: WHERE ( ( col LIKE ? AND col LIKE ? ) )
2645 { col => { -like => 'foo%' } },
2646 { col => { -like => '%bar' } },
2648 # yields: WHERE ( ( col LIKE ? OR col LIKE ? ) )
2651 =head2 Literal SQL and value type operators
2653 The basic premise of SQL::Abstract is that in WHERE specifications the "left
2654 side" is a column name and the "right side" is a value (normally rendered as
2655 a placeholder). This holds true for both hashrefs and arrayref pairs as you
2656 see in the L</WHERE CLAUSES> examples above. Sometimes it is necessary to
2657 alter this behavior. There are several ways of doing so.
2661 This is a virtual operator that signals the string to its right side is an
2662 identifier (a column name) and not a value. For example to compare two
2663 columns you would write:
2666 priority => { '<', 2 },
2667 requestor => { -ident => 'submitter' },
2672 $stmt = "WHERE priority < ? AND requestor = submitter";
2675 If you are maintaining legacy code you may see a different construct as
2676 described in L</Deprecated usage of Literal SQL>, please use C<-ident> in new
2681 This is a virtual operator that signals that the construct to its right side
2682 is a value to be passed to DBI. This is for example necessary when you want
2683 to write a where clause against an array (for RDBMS that support such
2684 datatypes). For example:
2687 array => { -value => [1, 2, 3] }
2692 $stmt = 'WHERE array = ?';
2693 @bind = ([1, 2, 3]);
2695 Note that if you were to simply say:
2701 the result would probably not be what you wanted:
2703 $stmt = 'WHERE array = ? OR array = ? OR array = ?';
2708 Finally, sometimes only literal SQL will do. To include a random snippet
2709 of SQL verbatim, you specify it as a scalar reference. Consider this only
2710 as a last resort. Usually there is a better way. For example:
2713 priority => { '<', 2 },
2714 requestor => { -in => \'(SELECT name FROM hitmen)' },
2719 $stmt = "WHERE priority < ? AND requestor IN (SELECT name FROM hitmen)"
2722 Note that in this example, you only get one bind parameter back, since
2723 the verbatim SQL is passed as part of the statement.
2727 Never use untrusted input as a literal SQL argument - this is a massive
2728 security risk (there is no way to check literal snippets for SQL
2729 injections and other nastyness). If you need to deal with untrusted input
2730 use literal SQL with placeholders as described next.
2732 =head3 Literal SQL with placeholders and bind values (subqueries)
2734 If the literal SQL to be inserted has placeholders and bind values,
2735 use a reference to an arrayref (yes this is a double reference --
2736 not so common, but perfectly legal Perl). For example, to find a date
2737 in Postgres you can use something like this:
2740 date_column => \[ "= date '2008-09-30' - ?::integer", 10 ]
2745 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
2748 Note that you must pass the bind values in the same format as they are returned
2749 by L<where|/where(\%where, $order)>. This means that if you set L</bindtype>
2750 to C<columns>, you must provide the bind values in the
2751 C<< [ column_meta => value ] >> format, where C<column_meta> is an opaque
2752 scalar value; most commonly the column name, but you can use any scalar value
2753 (including references and blessed references), L<SQL::Abstract> will simply
2754 pass it through intact. So if C<bindtype> is set to C<columns> the above
2755 example will look like:
2758 date_column => \[ "= date '2008-09-30' - ?::integer", [ {} => 10 ] ]
2761 Literal SQL is especially useful for nesting parenthesized clauses in the
2762 main SQL query. Here is a first example:
2764 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
2768 bar => \["IN ($sub_stmt)" => @sub_bind],
2773 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
2774 WHERE c2 < ? AND c3 LIKE ?))";
2775 @bind = (1234, 100, "foo%");
2777 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
2778 are expressed in the same way. Of course the C<$sub_stmt> and
2779 its associated bind values can be generated through a former call
2782 my ($sub_stmt, @sub_bind)
2783 = $sql->select("t1", "c1", {c2 => {"<" => 100},
2784 c3 => {-like => "foo%"}});
2787 bar => \["> ALL ($sub_stmt)" => @sub_bind],
2790 In the examples above, the subquery was used as an operator on a column;
2791 but the same principle also applies for a clause within the main C<%where>
2792 hash, like an EXISTS subquery:
2794 my ($sub_stmt, @sub_bind)
2795 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
2796 my %where = ( -and => [
2798 \["EXISTS ($sub_stmt)" => @sub_bind],
2803 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
2804 WHERE c1 = ? AND c2 > t0.c0))";
2808 Observe that the condition on C<c2> in the subquery refers to
2809 column C<t0.c0> of the main query: this is I<not> a bind
2810 value, so we have to express it through a scalar ref.
2811 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
2812 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
2813 what we wanted here.
2815 Finally, here is an example where a subquery is used
2816 for expressing unary negation:
2818 my ($sub_stmt, @sub_bind)
2819 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
2820 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
2822 lname => {like => '%son%'},
2823 \["NOT ($sub_stmt)" => @sub_bind],
2828 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
2829 @bind = ('%son%', 10, 20)
2831 =head3 Deprecated usage of Literal SQL
2833 Below are some examples of archaic use of literal SQL. It is shown only as
2834 reference for those who deal with legacy code. Each example has a much
2835 better, cleaner and safer alternative that users should opt for in new code.
2841 my %where = ( requestor => \'IS NOT NULL' )
2843 $stmt = "WHERE requestor IS NOT NULL"
2845 This used to be the way of generating NULL comparisons, before the handling
2846 of C<undef> got formalized. For new code please use the superior syntax as
2847 described in L</Tests for NULL values>.
2851 my %where = ( requestor => \'= submitter' )
2853 $stmt = "WHERE requestor = submitter"
2855 This used to be the only way to compare columns. Use the superior L</-ident>
2856 method for all new code. For example an identifier declared in such a way
2857 will be properly quoted if L</quote_char> is properly set, while the legacy
2858 form will remain as supplied.
2862 my %where = ( is_ready => \"", completed => { '>', '2012-12-21' } )
2864 $stmt = "WHERE completed > ? AND is_ready"
2865 @bind = ('2012-12-21')
2867 Using an empty string literal used to be the only way to express a boolean.
2868 For all new code please use the much more readable
2869 L<-bool|/Unary operators: bool> operator.
2875 These pages could go on for a while, since the nesting of the data
2876 structures this module can handle are pretty much unlimited (the
2877 module implements the C<WHERE> expansion as a recursive function
2878 internally). Your best bet is to "play around" with the module a
2879 little to see how the data structures behave, and choose the best
2880 format for your data based on that.
2882 And of course, all the values above will probably be replaced with
2883 variables gotten from forms or the command line. After all, if you
2884 knew everything ahead of time, you wouldn't have to worry about
2885 dynamically-generating SQL and could just hardwire it into your
2888 =head1 ORDER BY CLAUSES
2890 Some functions take an order by clause. This can either be a scalar (just a
2891 column name), a hashref of C<< { -desc => 'col' } >> or C<< { -asc => 'col' }
2892 >>, a scalarref, an arrayref-ref, or an arrayref of any of the previous
2895 Given | Will Generate
2896 ---------------------------------------------------------------
2898 'colA' | ORDER BY colA
2900 [qw/colA colB/] | ORDER BY colA, colB
2902 {-asc => 'colA'} | ORDER BY colA ASC
2904 {-desc => 'colB'} | ORDER BY colB DESC
2906 ['colA', {-asc => 'colB'}] | ORDER BY colA, colB ASC
2908 { -asc => [qw/colA colB/] } | ORDER BY colA ASC, colB ASC
2910 \'colA DESC' | ORDER BY colA DESC
2912 \[ 'FUNC(colA, ?)', $x ] | ORDER BY FUNC(colA, ?)
2913 | /* ...with $x bound to ? */
2916 { -asc => 'colA' }, | colA ASC,
2917 { -desc => [qw/colB/] }, | colB DESC,
2918 { -asc => [qw/colC colD/] },| colC ASC, colD ASC,
2919 \'colE DESC', | colE DESC,
2920 \[ 'FUNC(colF, ?)', $x ], | FUNC(colF, ?)
2921 ] | /* ...with $x bound to ? */
2922 ===============================================================
2926 =head1 SPECIAL OPERATORS
2928 my $sqlmaker = SQL::Abstract->new(special_ops => [
2932 my ($self, $field, $op, $arg) = @_;
2938 handler => 'method_name',
2942 A "special operator" is a SQL syntactic clause that can be
2943 applied to a field, instead of a usual binary operator.
2946 WHERE field IN (?, ?, ?)
2947 WHERE field BETWEEN ? AND ?
2948 WHERE MATCH(field) AGAINST (?, ?)
2950 Special operators IN and BETWEEN are fairly standard and therefore
2951 are builtin within C<SQL::Abstract> (as the overridable methods
2952 C<_where_field_IN> and C<_where_field_BETWEEN>). For other operators,
2953 like the MATCH .. AGAINST example above which is specific to MySQL,
2954 you can write your own operator handlers - supply a C<special_ops>
2955 argument to the C<new> method. That argument takes an arrayref of
2956 operator definitions; each operator definition is a hashref with two
2963 the regular expression to match the operator
2967 Either a coderef or a plain scalar method name. In both cases
2968 the expected return is C<< ($sql, @bind) >>.
2970 When supplied with a method name, it is simply called on the
2971 L<SQL::Abstract> object as:
2973 $self->$method_name($field, $op, $arg)
2977 $field is the LHS of the operator
2978 $op is the part that matched the handler regex
2981 When supplied with a coderef, it is called as:
2983 $coderef->($self, $field, $op, $arg)
2988 For example, here is an implementation
2989 of the MATCH .. AGAINST syntax for MySQL
2991 my $sqlmaker = SQL::Abstract->new(special_ops => [
2993 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
2994 {regex => qr/^match$/i,
2996 my ($self, $field, $op, $arg) = @_;
2997 $arg = [$arg] if not ref $arg;
2998 my $label = $self->_quote($field);
2999 my ($placeholder) = $self->_convert('?');
3000 my $placeholders = join ", ", (($placeholder) x @$arg);
3001 my $sql = $self->_sqlcase('match') . " ($label) "
3002 . $self->_sqlcase('against') . " ($placeholders) ";
3003 my @bind = $self->_bindtype($field, @$arg);
3004 return ($sql, @bind);
3011 =head1 UNARY OPERATORS
3013 my $sqlmaker = SQL::Abstract->new(unary_ops => [
3017 my ($self, $op, $arg) = @_;
3023 handler => 'method_name',
3027 A "unary operator" is a SQL syntactic clause that can be
3028 applied to a field - the operator goes before the field
3030 You can write your own operator handlers - supply a C<unary_ops>
3031 argument to the C<new> method. That argument takes an arrayref of
3032 operator definitions; each operator definition is a hashref with two
3039 the regular expression to match the operator
3043 Either a coderef or a plain scalar method name. In both cases
3044 the expected return is C<< $sql >>.
3046 When supplied with a method name, it is simply called on the
3047 L<SQL::Abstract> object as:
3049 $self->$method_name($op, $arg)
3053 $op is the part that matched the handler regex
3054 $arg is the RHS or argument of the operator
3056 When supplied with a coderef, it is called as:
3058 $coderef->($self, $op, $arg)
3066 Thanks to some benchmarking by Mark Stosberg, it turns out that
3067 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
3068 I must admit this wasn't an intentional design issue, but it's a
3069 byproduct of the fact that you get to control your C<DBI> handles
3072 To maximize performance, use a code snippet like the following:
3074 # prepare a statement handle using the first row
3075 # and then reuse it for the rest of the rows
3077 for my $href (@array_of_hashrefs) {
3078 $stmt ||= $sql->insert('table', $href);
3079 $sth ||= $dbh->prepare($stmt);
3080 $sth->execute($sql->values($href));
3083 The reason this works is because the keys in your C<$href> are sorted
3084 internally by B<SQL::Abstract>. Thus, as long as your data retains
3085 the same structure, you only have to generate the SQL the first time
3086 around. On subsequent queries, simply use the C<values> function provided
3087 by this module to return your values in the correct order.
3089 However this depends on the values having the same type - if, for
3090 example, the values of a where clause may either have values
3091 (resulting in sql of the form C<column = ?> with a single bind
3092 value), or alternatively the values might be C<undef> (resulting in
3093 sql of the form C<column IS NULL> with no bind value) then the
3094 caching technique suggested will not work.
3098 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
3099 really like this part (I do, at least). Building up a complex query
3100 can be as simple as the following:
3107 use CGI::FormBuilder;
3110 my $form = CGI::FormBuilder->new(...);
3111 my $sql = SQL::Abstract->new;
3113 if ($form->submitted) {
3114 my $field = $form->field;
3115 my $id = delete $field->{id};
3116 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
3119 Of course, you would still have to connect using C<DBI> to run the
3120 query, but the point is that if you make your form look like your
3121 table, the actual query script can be extremely simplistic.
3123 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
3124 a fast interface to returning and formatting data. I frequently
3125 use these three modules together to write complex database query
3126 apps in under 50 lines.
3128 =head1 HOW TO CONTRIBUTE
3130 Contributions are always welcome, in all usable forms (we especially
3131 welcome documentation improvements). The delivery methods include git-
3132 or unified-diff formatted patches, GitHub pull requests, or plain bug
3133 reports either via RT or the Mailing list. Contributors are generally
3134 granted full access to the official repository after their first several
3135 patches pass successful review.
3137 This project is maintained in a git repository. The code and related tools are
3138 accessible at the following locations:
3142 =item * Official repo: L<git://git.shadowcat.co.uk/dbsrgits/SQL-Abstract.git>
3144 =item * Official gitweb: L<http://git.shadowcat.co.uk/gitweb/gitweb.cgi?p=dbsrgits/SQL-Abstract.git>
3146 =item * GitHub mirror: L<https://github.com/dbsrgits/sql-abstract>
3148 =item * Authorized committers: L<ssh://dbsrgits@git.shadowcat.co.uk/SQL-Abstract.git>
3154 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
3155 Great care has been taken to preserve the I<published> behavior
3156 documented in previous versions in the 1.* family; however,
3157 some features that were previously undocumented, or behaved
3158 differently from the documentation, had to be changed in order
3159 to clarify the semantics. Hence, client code that was relying
3160 on some dark areas of C<SQL::Abstract> v1.*
3161 B<might behave differently> in v1.50.
3163 The main changes are:
3169 support for literal SQL through the C<< \ [ $sql, @bind ] >> syntax.
3173 support for the { operator => \"..." } construct (to embed literal SQL)
3177 support for the { operator => \["...", @bind] } construct (to embed literal SQL with bind values)
3181 optional support for L<array datatypes|/"Inserting and Updating Arrays">
3185 defensive programming: check arguments
3189 fixed bug with global logic, which was previously implemented
3190 through global variables yielding side-effects. Prior versions would
3191 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
3192 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
3193 Now this is interpreted
3194 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
3199 fixed semantics of _bindtype on array args
3203 dropped the C<_anoncopy> of the %where tree. No longer necessary,
3204 we just avoid shifting arrays within that tree.
3208 dropped the C<_modlogic> function
3212 =head1 ACKNOWLEDGEMENTS
3214 There are a number of individuals that have really helped out with
3215 this module. Unfortunately, most of them submitted bugs via CPAN
3216 so I have no idea who they are! But the people I do know are:
3218 Ash Berlin (order_by hash term support)
3219 Matt Trout (DBIx::Class support)
3220 Mark Stosberg (benchmarking)
3221 Chas Owens (initial "IN" operator support)
3222 Philip Collins (per-field SQL functions)
3223 Eric Kolve (hashref "AND" support)
3224 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
3225 Dan Kubb (support for "quote_char" and "name_sep")
3226 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
3227 Laurent Dami (internal refactoring, extensible list of special operators, literal SQL)
3228 Norbert Buchmuller (support for literal SQL in hashpair, misc. fixes & tests)
3229 Peter Rabbitson (rewrite of SQLA::Test, misc. fixes & tests)
3230 Oliver Charles (support for "RETURNING" after "INSERT")
3236 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
3240 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
3242 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
3244 For support, your best bet is to try the C<DBIx::Class> users mailing list.
3245 While not an official support venue, C<DBIx::Class> makes heavy use of
3246 C<SQL::Abstract>, and as such list members there are very familiar with
3247 how to create queries.
3251 This module is free software; you may copy this under the same
3252 terms as perl itself (either the GNU General Public License or
3253 the Artistic License)