1 package SQL::Abstract; # see doc at end of file
10 our @EXPORT_OK = qw(is_plain_value is_literal_value);
20 *SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION = $ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}
26 #======================================================================
28 #======================================================================
30 our $VERSION = '1.87';
32 # This would confuse some packagers
33 $VERSION = eval $VERSION if $VERSION =~ /_/; # numify for warning-free dev releases
37 # special operators (-in, -between). May be extended/overridden by user.
38 # See section WHERE: BUILTIN SPECIAL OPERATORS below for implementation
39 my @BUILTIN_SPECIAL_OPS = (
40 {regex => qr/^ (?: not \s )? between $/ix, handler => sub { die "NOPE" }},
41 {regex => qr/^ (?: not \s )? in $/ix, handler => sub { die "NOPE" }},
42 {regex => qr/^ ident $/ix, handler => sub { die "NOPE" }},
43 {regex => qr/^ value $/ix, handler => sub { die "NOPE" }},
44 {regex => qr/^ is (?: \s+ not )? $/ix, handler => sub { die "NOPE" }},
47 # unaryish operators - key maps to handler
48 my @BUILTIN_UNARY_OPS = (
49 # the digits are backcompat stuff
50 { regex => qr/^ and (?: [_\s]? \d+ )? $/xi, handler => '_where_op_ANDOR' },
51 { regex => qr/^ or (?: [_\s]? \d+ )? $/xi, handler => '_where_op_ANDOR' },
52 { regex => qr/^ nest (?: [_\s]? \d+ )? $/xi, handler => '_where_op_NEST' },
53 { regex => qr/^ (?: not \s )? bool $/xi, handler => '_where_op_BOOL' },
54 { regex => qr/^ ident $/xi, handler => '_where_op_IDENT' },
55 { regex => qr/^ value $/xi, handler => '_where_op_VALUE' },
56 { regex => qr/^ op $/xi, handler => '_where_op_OP' },
57 { regex => qr/^ bind $/xi, handler => '_where_op_BIND' },
58 { regex => qr/^ literal $/xi, handler => '_where_op_LITERAL' },
59 { regex => qr/^ func $/xi, handler => '_where_op_FUNC' },
62 #======================================================================
63 # DEBUGGING AND ERROR REPORTING
64 #======================================================================
67 return unless $_[0]->{debug}; shift; # a little faster
68 my $func = (caller(1))[3];
69 warn "[$func] ", @_, "\n";
73 my($func) = (caller(1))[3];
74 Carp::carp "[$func] Warning: ", @_;
78 my($func) = (caller(1))[3];
79 Carp::croak "[$func] Fatal: ", @_;
82 sub is_literal_value ($) {
83 ref $_[0] eq 'SCALAR' ? [ ${$_[0]} ]
84 : ( ref $_[0] eq 'REF' and ref ${$_[0]} eq 'ARRAY' ) ? [ @${ $_[0] } ]
88 # FIXME XSify - this can be done so much more efficiently
89 sub is_plain_value ($) {
91 ! length ref $_[0] ? \($_[0])
93 ref $_[0] eq 'HASH' and keys %{$_[0]} == 1
95 exists $_[0]->{-value}
96 ) ? \($_[0]->{-value})
98 # reuse @_ for even moar speedz
99 defined ( $_[1] = Scalar::Util::blessed $_[0] )
101 # deliberately not using Devel::OverloadInfo - the checks we are
102 # intersted in are much more limited than the fullblown thing, and
103 # this is a very hot piece of code
105 # simply using ->can('(""') can leave behind stub methods that
106 # break actually using the overload later (see L<perldiag/Stub
107 # found while resolving method "%s" overloading "%s" in package
108 # "%s"> and the source of overload::mycan())
110 # either has stringification which DBI SHOULD prefer out of the box
111 grep { *{ (qq[${_}::(""]) }{CODE} } @{ $_[2] = mro::get_linear_isa( $_[1] ) }
113 # has nummification or boolification, AND fallback is *not* disabled
115 SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION
118 grep { *{"${_}::(0+"}{CODE} } @{$_[2]}
120 grep { *{"${_}::(bool"}{CODE} } @{$_[2]}
124 # no fallback specified at all
125 ! ( ($_[3]) = grep { *{"${_}::()"}{CODE} } @{$_[2]} )
127 # fallback explicitly undef
128 ! defined ${"$_[3]::()"}
141 #======================================================================
143 #======================================================================
147 my $class = ref($self) || $self;
148 my %opt = (ref $_[0] eq 'HASH') ? %{$_[0]} : @_;
150 # choose our case by keeping an option around
151 delete $opt{case} if $opt{case} && $opt{case} ne 'lower';
153 # default logic for interpreting arrayrefs
154 $opt{logic} = $opt{logic} ? uc $opt{logic} : 'OR';
156 # how to return bind vars
157 $opt{bindtype} ||= 'normal';
159 # default comparison is "=", but can be overridden
162 # try to recognize which are the 'equality' and 'inequality' ops
163 # (temporary quickfix (in 2007), should go through a more seasoned API)
164 $opt{equality_op} = qr/^( \Q$opt{cmp}\E | \= )$/ix;
165 $opt{inequality_op} = qr/^( != | <> )$/ix;
167 $opt{like_op} = qr/^ (is\s+)? r?like $/xi;
168 $opt{not_like_op} = qr/^ (is\s+)? not \s+ r?like $/xi;
171 $opt{sqltrue} ||= '1=1';
172 $opt{sqlfalse} ||= '0=1';
175 $opt{user_special_ops} = [ @{$opt{special_ops} ||= []} ];
176 # regexes are applied in order, thus push after user-defines
177 push @{$opt{special_ops}}, @BUILTIN_SPECIAL_OPS;
180 $opt{unary_ops} ||= [];
181 push @{$opt{unary_ops}}, @BUILTIN_UNARY_OPS;
183 # rudimentary sanity-check for user supplied bits treated as functions/operators
184 # If a purported function matches this regular expression, an exception is thrown.
185 # Literal SQL is *NOT* subject to this check, only functions (and column names
186 # when quoting is not in effect)
189 # need to guard against ()'s in column names too, but this will break tons of
190 # hacks... ideas anyone?
191 $opt{injection_guard} ||= qr/
197 return bless \%opt, $class;
201 sub _assert_pass_injection_guard {
202 if ($_[1] =~ $_[0]->{injection_guard}) {
203 my $class = ref $_[0];
204 puke "Possible SQL injection attempt '$_[1]'. If this is indeed a part of the "
205 . "desired SQL use literal SQL ( \'...' or \[ '...' ] ) or supply your own "
206 . "{injection_guard} attribute to ${class}->new()"
211 #======================================================================
213 #======================================================================
217 my $table = $self->_table(shift);
218 my $data = shift || return;
221 my $method = $self->_METHOD_FOR_refkind("_insert", $data);
222 my ($sql, @bind) = $self->$method($data);
223 $sql = join " ", $self->_sqlcase('insert into'), $table, $sql;
225 if ($options->{returning}) {
226 my ($s, @b) = $self->_insert_returning($options);
231 return wantarray ? ($sql, @bind) : $sql;
234 # So that subclasses can override INSERT ... RETURNING separately from
235 # UPDATE and DELETE (e.g. DBIx::Class::SQLMaker::Oracle does this)
236 sub _insert_returning { shift->_returning(@_) }
239 my ($self, $options) = @_;
241 my $f = $options->{returning};
243 my $fieldlist = $self->_SWITCH_refkind($f, {
244 ARRAYREF => sub {join ', ', map { $self->_quote($_) } @$f;},
245 SCALAR => sub {$self->_quote($f)},
246 SCALARREF => sub {$$f},
248 return $self->_sqlcase(' returning ') . $fieldlist;
251 sub _insert_HASHREF { # explicit list of fields and then values
252 my ($self, $data) = @_;
254 my @fields = sort keys %$data;
256 my ($sql, @bind) = $self->_insert_values($data);
259 $_ = $self->_quote($_) foreach @fields;
260 $sql = "( ".join(", ", @fields).") ".$sql;
262 return ($sql, @bind);
265 sub _insert_ARRAYREF { # just generate values(?,?) part (no list of fields)
266 my ($self, $data) = @_;
268 # no names (arrayref) so can't generate bindtype
269 $self->{bindtype} ne 'columns'
270 or belch "can't do 'columns' bindtype when called with arrayref";
272 my (@values, @all_bind);
273 foreach my $value (@$data) {
274 my ($values, @bind) = $self->_insert_value(undef, $value);
275 push @values, $values;
276 push @all_bind, @bind;
278 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
279 return ($sql, @all_bind);
282 sub _insert_ARRAYREFREF { # literal SQL with bind
283 my ($self, $data) = @_;
285 my ($sql, @bind) = @${$data};
286 $self->_assert_bindval_matches_bindtype(@bind);
288 return ($sql, @bind);
292 sub _insert_SCALARREF { # literal SQL without bind
293 my ($self, $data) = @_;
299 my ($self, $data) = @_;
301 my (@values, @all_bind);
302 foreach my $column (sort keys %$data) {
303 my ($values, @bind) = $self->_insert_value($column, $data->{$column});
304 push @values, $values;
305 push @all_bind, @bind;
307 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
308 return ($sql, @all_bind);
312 my ($self, $column, $v) = @_;
314 my (@values, @all_bind);
315 $self->_SWITCH_refkind($v, {
318 if ($self->{array_datatypes}) { # if array datatype are activated
320 push @all_bind, $self->_bindtype($column, $v);
322 else { # else literal SQL with bind
323 my ($sql, @bind) = @$v;
324 $self->_assert_bindval_matches_bindtype(@bind);
326 push @all_bind, @bind;
330 ARRAYREFREF => sub { # literal SQL with bind
331 my ($sql, @bind) = @${$v};
332 $self->_assert_bindval_matches_bindtype(@bind);
334 push @all_bind, @bind;
337 # THINK: anything useful to do with a HASHREF ?
338 HASHREF => sub { # (nothing, but old SQLA passed it through)
339 #TODO in SQLA >= 2.0 it will die instead
340 belch "HASH ref as bind value in insert is not supported";
342 push @all_bind, $self->_bindtype($column, $v);
345 SCALARREF => sub { # literal SQL without bind
349 SCALAR_or_UNDEF => sub {
351 push @all_bind, $self->_bindtype($column, $v);
356 my $sql = join(", ", @values);
357 return ($sql, @all_bind);
362 #======================================================================
364 #======================================================================
369 my $table = $self->_table(shift);
370 my $data = shift || return;
374 # first build the 'SET' part of the sql statement
375 puke "Unsupported data type specified to \$sql->update"
376 unless ref $data eq 'HASH';
378 my ($sql, @all_bind) = $self->_update_set_values($data);
379 $sql = $self->_sqlcase('update ') . $table . $self->_sqlcase(' set ')
383 my($where_sql, @where_bind) = $self->where($where);
385 push @all_bind, @where_bind;
388 if ($options->{returning}) {
389 my ($returning_sql, @returning_bind) = $self->_update_returning($options);
390 $sql .= $returning_sql;
391 push @all_bind, @returning_bind;
394 return wantarray ? ($sql, @all_bind) : $sql;
397 sub _update_set_values {
398 my ($self, $data) = @_;
400 my (@set, @all_bind);
401 for my $k (sort keys %$data) {
404 my $label = $self->_quote($k);
406 $self->_SWITCH_refkind($v, {
408 if ($self->{array_datatypes}) { # array datatype
409 push @set, "$label = ?";
410 push @all_bind, $self->_bindtype($k, $v);
412 else { # literal SQL with bind
413 my ($sql, @bind) = @$v;
414 $self->_assert_bindval_matches_bindtype(@bind);
415 push @set, "$label = $sql";
416 push @all_bind, @bind;
419 ARRAYREFREF => sub { # literal SQL with bind
420 my ($sql, @bind) = @${$v};
421 $self->_assert_bindval_matches_bindtype(@bind);
422 push @set, "$label = $sql";
423 push @all_bind, @bind;
425 SCALARREF => sub { # literal SQL without bind
426 push @set, "$label = $$v";
429 my ($op, $arg, @rest) = %$v;
431 puke 'Operator calls in update must be in the form { -op => $arg }'
432 if (@rest or not $op =~ /^\-(.+)/);
434 local $self->{_nested_func_lhs} = $k;
435 my ($sql, @bind) = $self->_where_unary_op($1, $arg);
437 push @set, "$label = $sql";
438 push @all_bind, @bind;
440 SCALAR_or_UNDEF => sub {
441 push @set, "$label = ?";
442 push @all_bind, $self->_bindtype($k, $v);
448 my $sql = join ', ', @set;
450 return ($sql, @all_bind);
453 # So that subclasses can override UPDATE ... RETURNING separately from
455 sub _update_returning { shift->_returning(@_) }
459 #======================================================================
461 #======================================================================
466 my $table = $self->_table(shift);
467 my $fields = shift || '*';
471 my ($fields_sql, @bind) = $self->_select_fields($fields);
473 my ($where_sql, @where_bind) = $self->where($where, $order);
474 push @bind, @where_bind;
476 my $sql = join(' ', $self->_sqlcase('select'), $fields_sql,
477 $self->_sqlcase('from'), $table)
480 return wantarray ? ($sql, @bind) : $sql;
484 my ($self, $fields) = @_;
485 return ref $fields eq 'ARRAY' ? join ', ', map { $self->_quote($_) } @$fields
489 #======================================================================
491 #======================================================================
496 my $table = $self->_table(shift);
500 my($where_sql, @bind) = $self->where($where);
501 my $sql = $self->_sqlcase('delete from ') . $table . $where_sql;
503 if ($options->{returning}) {
504 my ($returning_sql, @returning_bind) = $self->_delete_returning($options);
505 $sql .= $returning_sql;
506 push @bind, @returning_bind;
509 return wantarray ? ($sql, @bind) : $sql;
512 # So that subclasses can override DELETE ... RETURNING separately from
514 sub _delete_returning { shift->_returning(@_) }
518 #======================================================================
520 #======================================================================
524 # Finally, a separate routine just to handle WHERE clauses
526 my ($self, $where, $order) = @_;
529 my ($sql, @bind) = $self->_recurse_where($where);
530 $sql = (defined $sql and length $sql) ? $self->_sqlcase(' where ') . "( $sql )" : '';
534 my ($order_sql, @order_bind) = $self->_order_by($order);
536 push @bind, @order_bind;
539 return wantarray ? ($sql, @bind) : $sql;
543 my ($self, $expr, $logic) = @_;
544 return undef unless defined($expr);
545 if (ref($expr) eq 'HASH') {
546 if (keys %$expr > 1) {
548 return +{ "-${logic}" => [
549 map $self->_expand_expr_hashpair($_ => $expr->{$_}, $logic),
553 return unless %$expr;
554 return $self->_expand_expr_hashpair(%$expr, $logic);
556 if (ref($expr) eq 'ARRAY') {
557 my $logic = lc($logic || $self->{logic});
558 $logic eq 'and' or $logic eq 'or' or puke "unknown logic: $logic";
564 while (my ($el) = splice @expr, 0, 1) {
565 puke "Supplying an empty left hand side argument is not supported in array-pairs"
566 unless defined($el) and length($el);
567 my $elref = ref($el);
569 push(@res, $self->_expand_expr({ $el, shift(@expr) }));
570 } elsif ($elref eq 'ARRAY') {
571 push(@res, $self->_expand_expr($el)) if @$el;
572 } elsif (is_literal_value($el)) {
574 } elsif ($elref eq 'HASH') {
575 push @res, $self->_expand_expr($el);
580 return { '-'.$logic => \@res };
582 if (my $literal = is_literal_value($expr)) {
583 return +{ -literal => $literal };
585 if (!ref($expr) or Scalar::Util::blessed($expr)) {
586 if (my $m = our $Cur_Col_Meta) {
587 return +{ -bind => [ $m, $expr ] };
589 return +{ -value => $expr };
594 sub _expand_expr_hashpair {
595 my ($self, $k, $v, $logic) = @_;
596 unless (defined($k) and length($k)) {
597 if (defined($k) and my $literal = is_literal_value($v)) {
598 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
599 return { -literal => $literal };
601 puke "Supplying an empty left hand side argument is not supported";
604 $self->_assert_pass_injection_guard($k =~ /^-(.*)$/s);
605 if ($k =~ s/ [_\s]? \d+ $//x ) {
606 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
607 . "You probably wanted ...-and => [ $k => COND1, $k => COND2 ... ]";
610 return $self->_expand_expr($v);
614 return $self->_expand_expr($v);
616 puke "-bool => undef not supported" unless defined($v);
617 return { -ident => $v };
620 return { -not => $self->_expand_expr($v) };
622 if (my ($rest) = $k =~/^-not[_ ](.*)$/) {
624 $self->_expand_expr_hashpair("-${rest}", $v, $logic)
627 if (my ($logic) = $k =~ /^-(and|or)$/i) {
628 if (ref($v) eq 'HASH') {
629 return $self->_expand_expr($v, $logic);
631 if (ref($v) eq 'ARRAY') {
632 return $self->_expand_expr($v, $logic);
637 $op =~ s/^-// if length($op) > 1;
639 # top level special ops are illegal in general
640 puke "Illegal use of top-level '-$op'"
641 if !(defined $self->{_nested_func_lhs})
642 and List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}
643 and not List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}};
645 if ($k eq '-value' and my $m = our $Cur_Col_Meta) {
646 return +{ -bind => [ $m, $v ] };
648 if ($k eq '-op' or $k eq '-ident' or $k eq '-value' or $k eq '-bind' or $k eq '-literal' or $k eq '-func') {
654 and (keys %$v)[0] =~ /^-/
656 my ($func) = $k =~ /^-(.*)$/;
657 return +{ -func => [ $func, $self->_expand_expr($v) ] };
659 if (!ref($v) or is_literal_value($v)) {
660 return +{ -op => [ $k =~ /^-(.*)$/, $self->_expand_expr($v) ] };
667 and exists $v->{-value}
668 and not defined $v->{-value}
671 return $self->_expand_expr_hashpair($k => { $self->{cmp} => undef });
673 if (!ref($v) or Scalar::Util::blessed($v)) {
678 { -bind => [ $k, $v ] }
682 if (ref($v) eq 'HASH') {
685 map $self->_expand_expr_hashpair($k => { $_ => $v->{$_} }),
692 $self->_assert_pass_injection_guard($vk);
693 if ($vk =~ s/ [_\s]? \d+ $//x ) {
694 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
695 . "You probably wanted ...-and => [ -$vk => COND1, -$vk => COND2 ... ]";
697 if ($vk =~ /^(?:not[ _])?between$/) {
698 local our $Cur_Col_Meta = $k;
699 my @rhs = map $self->_expand_expr($_),
700 ref($vv) eq 'ARRAY' ? @$vv : $vv;
702 (@rhs == 1 and ref($rhs[0]) eq 'HASH' and $rhs[0]->{-literal})
704 (@rhs == 2 and defined($rhs[0]) and defined($rhs[1]))
706 puke "Operator '${\uc($vk)}' requires either an arrayref with two defined values or expressions, or a single literal scalarref/arrayref-ref";
709 join(' ', split '_', $vk),
714 if ($vk =~ /^(?:not[ _])?in$/) {
715 if (my $literal = is_literal_value($vv)) {
716 my ($sql, @bind) = @$literal;
717 my $opened_sql = $self->_open_outer_paren($sql);
719 $vk, { -ident => $k },
720 [ { -literal => [ $opened_sql, @bind ] } ]
724 'SQL::Abstract before v1.75 used to generate incorrect SQL when the '
725 . "-${\uc($vk)} operator was given an undef-containing list: !!!AUDIT YOUR CODE "
726 . 'AND DATA!!! (the upcoming Data::Query-based version of SQL::Abstract '
727 . 'will emit the logically correct SQL instead of raising this exception)'
729 puke("Argument passed to the '${\uc($vk)}' operator can not be undefined")
731 my @rhs = map $self->_expand_expr($_),
732 map { ref($_) ? $_ : { -bind => [ $k, $_ ] } }
733 map { defined($_) ? $_: puke($undef_err) }
734 (ref($vv) eq 'ARRAY' ? @$vv : $vv);
736 -literal => [ $self->{$vk =~ /^not/ ? 'sqltrue' : 'sqlfalse'} ]
740 join(' ', split '_', $vk),
745 if ($vk eq 'ident') {
746 if (! defined $vv or ref $vv) {
747 puke "-$vk requires a single plain scalar argument (a quotable identifier)";
755 if ($vk eq 'value') {
756 return $self->_expand_expr_hashpair($k, undef) unless defined($vv);
760 { -bind => [ $k, $vv ] }
763 if ($vk =~ /^is(?:[ _]not)?$/) {
764 puke "$vk can only take undef as argument"
768 and exists($vv->{-value})
769 and !defined($vv->{-value})
772 return +{ -op => [ $vk.' null', { -ident => $k } ] };
774 if ($vk =~ /^(and|or)$/) {
775 if (ref($vv) eq 'HASH') {
776 return +{ "-${vk}" => [
777 map $self->_expand_expr_hashpair($k, { $_ => $vv->{$_} }),
782 if (my $us = List::Util::first { $vk =~ $_->{regex} } @{$self->{user_special_ops}}) {
783 return { -op => [ $vk, { -ident => $k }, $vv ] };
785 if (ref($vv) eq 'ARRAY') {
786 my ($logic, @values) = (
787 (defined($vv->[0]) and $vv->[0] =~ /^-(and|or)$/i)
792 $vk =~ $self->{inequality_op}
793 or join(' ', split '_', $vk) =~ $self->{not_like_op}
795 if (lc($logic) eq '-or' and @values > 1) {
796 my $op = uc join ' ', split '_', $vk;
797 belch "A multi-element arrayref as an argument to the inequality op '$op' "
798 . 'is technically equivalent to an always-true 1=1 (you probably wanted '
799 . "to say ...{ \$inequality_op => [ -and => \@values ] }... instead)"
804 # try to DWIM on equality operators
805 my $op = join ' ', split '_', $vk;
807 $op =~ $self->{equality_op} ? $self->{sqlfalse}
808 : $op =~ $self->{like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->{sqlfalse}
809 : $op =~ $self->{inequality_op} ? $self->{sqltrue}
810 : $op =~ $self->{not_like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->{sqltrue}
811 : puke "operator '$op' applied on an empty array (field '$k')";
813 return +{ $logic => [
814 map $self->_expand_expr_hashpair($k => { $vk => $_ }),
822 and exists $vv->{-value}
823 and not defined $vv->{-value}
826 my $op = join ' ', split '_', $vk;
828 $op =~ /^not$/i ? 'is not' # legacy
829 : $op =~ $self->{equality_op} ? 'is'
830 : $op =~ $self->{like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is'
831 : $op =~ $self->{inequality_op} ? 'is not'
832 : $op =~ $self->{not_like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is not'
833 : puke "unexpected operator '$op' with undef operand";
834 return +{ -op => [ $is.' null', { -ident => $k } ] };
836 local our $Cur_Col_Meta = $k;
840 $self->_expand_expr($vv)
843 if (ref($v) eq 'ARRAY') {
844 return $self->{sqlfalse} unless @$v;
845 $self->_debug("ARRAY($k) means distribute over elements");
847 $v->[0] =~ /^-((?:and|or))$/i
848 ? ($v = [ @{$v}[1..$#$v] ], $1)
849 : ($self->{logic} || 'or')
851 return +{ "-${this_logic}" => [ map $self->_expand_expr({ $k => $_ }, $this_logic), @$v ] };
853 if (my $literal = is_literal_value($v)) {
855 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
858 my ($sql, @bind) = @$literal;
859 if ($self->{bindtype} eq 'columns') {
861 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
862 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
866 return +{ -literal => [ $self->_quote($k).' '.$sql, @bind ] };
872 my ($self, $where, $logic) = @_;
874 #print STDERR Data::Dumper::Concise::Dumper([ $where, $logic ]);
876 my $where_exp = $self->_expand_expr($where, $logic);
878 #print STDERR Data::Dumper::Concise::Dumper([ EXP => $where_exp ]);
880 # dispatch on appropriate method according to refkind of $where
881 my $method = $self->_METHOD_FOR_refkind("_where", $where_exp);
883 my ($sql, @bind) = $self->$method($where_exp, $logic);
885 # DBIx::Class used to call _recurse_where in scalar context
886 # something else might too...
888 return ($sql, @bind);
891 belch "Calling _recurse_where in scalar context is deprecated and will go away before 2.0";
898 #======================================================================
899 # WHERE: top-level ARRAYREF
900 #======================================================================
903 sub _where_ARRAYREF {
904 my ($self, $where, $logic) = @_;
906 $logic = uc($logic || $self->{logic});
907 $logic eq 'AND' or $logic eq 'OR' or puke "unknown logic: $logic";
909 my @clauses = @$where;
911 my (@sql_clauses, @all_bind);
912 # need to use while() so can shift() for pairs
914 my $el = shift @clauses;
916 $el = undef if (defined $el and ! length $el);
918 # switch according to kind of $el and get corresponding ($sql, @bind)
919 my ($sql, @bind) = $self->_SWITCH_refkind($el, {
921 # skip empty elements, otherwise get invalid trailing AND stuff
922 ARRAYREF => sub {$self->_recurse_where($el) if @$el},
926 $self->_assert_bindval_matches_bindtype(@b);
930 HASHREF => sub {$self->_recurse_where($el, 'and') if %$el},
932 SCALARREF => sub { ($$el); },
935 # top-level arrayref with scalars, recurse in pairs
936 $self->_recurse_where({$el => shift(@clauses)})
939 UNDEF => sub {puke "Supplying an empty left hand side argument is not supported in array-pairs" },
943 push @sql_clauses, $sql;
944 push @all_bind, @bind;
948 return $self->_join_sql_clauses($logic, \@sql_clauses, \@all_bind);
951 #======================================================================
952 # WHERE: top-level ARRAYREFREF
953 #======================================================================
955 sub _where_ARRAYREFREF {
956 my ($self, $where) = @_;
957 my ($sql, @bind) = @$$where;
958 $self->_assert_bindval_matches_bindtype(@bind);
959 return ($sql, @bind);
962 #======================================================================
963 # WHERE: top-level HASHREF
964 #======================================================================
967 my ($self, $where) = @_;
968 my (@sql_clauses, @all_bind);
970 for my $k (sort keys %$where) {
971 my $v = $where->{$k};
973 # ($k => $v) is either a special unary op or a regular hashpair
974 my ($sql, @bind) = do {
976 # put the operator in canonical form
978 $op = substr $op, 1; # remove initial dash
979 $op =~ s/^\s+|\s+$//g;# remove leading/trailing space
980 $op =~ s/\s+/ /g; # compress whitespace
982 # so that -not_foo works correctly
983 $op =~ s/^not_/NOT /i;
985 $self->_debug("Unary OP(-$op) within hashref, recursing...");
986 my ($s, @b) = $self->_where_unary_op($op, $v);
988 # top level vs nested
989 # we assume that handled unary ops will take care of their ()s
991 List::Util::first {$op =~ $_->{regex}} @{$self->{unary_ops}}
993 ( defined $self->{_nested_func_lhs} and $self->{_nested_func_lhs} eq $k )
999 if (is_literal_value ($v) ) {
1000 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
1003 puke "Supplying an empty left hand side argument is not supported in hash-pairs";
1007 my $method = $self->_METHOD_FOR_refkind("_where_hashpair", $v);
1008 $self->$method($k, $v);
1012 push @sql_clauses, $sql;
1013 push @all_bind, @bind;
1016 return $self->_join_sql_clauses('and', \@sql_clauses, \@all_bind);
1019 sub _where_unary_op {
1020 my ($self, $op, $rhs) = @_;
1022 $op =~ s/^-// if length($op) > 1;
1024 # top level special ops are illegal in general
1025 puke "Illegal use of top-level '-$op'"
1026 if !(defined $self->{_nested_func_lhs})
1027 and List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}
1028 and not List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}};
1030 if (my $op_entry = List::Util::first { $op =~ $_->{regex} } @{$self->{unary_ops}}) {
1031 my $handler = $op_entry->{handler};
1033 if (not ref $handler) {
1034 if ($op =~ s/ [_\s]? \d+ $//x ) {
1035 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
1036 . "You probably wanted ...-and => [ -$op => COND1, -$op => COND2 ... ]";
1038 return $self->$handler($op, $rhs);
1040 elsif (ref $handler eq 'CODE') {
1041 return $handler->($self, $op, $rhs);
1044 puke "Illegal handler for operator $op - expecting a method name or a coderef";
1048 $self->_debug("Generic unary OP: $op - recursing as function");
1050 $self->_assert_pass_injection_guard($op);
1052 my ($sql, @bind) = $self->_SWITCH_refkind($rhs, {
1054 puke "Illegal use of top-level '-$op'"
1055 unless defined $self->{_nested_func_lhs};
1058 $self->_convert('?'),
1059 $self->_bindtype($self->{_nested_func_lhs}, $rhs)
1063 $self->_recurse_where($rhs)
1067 $sql = sprintf('%s %s',
1068 $self->_sqlcase($op),
1072 return ($sql, @bind);
1075 sub _where_op_ANDOR {
1076 my ($self, $op, $v) = @_;
1078 $self->_SWITCH_refkind($v, {
1080 return $self->_where_ARRAYREF($v, $op);
1084 return ($op =~ /^or/i)
1085 ? $self->_where_ARRAYREF([ map { $_ => $v->{$_} } (sort keys %$v) ], $op)
1086 : $self->_where_HASHREF($v);
1090 puke "-$op => \\\$scalar makes little sense, use " .
1092 ? '[ \$scalar, \%rest_of_conditions ] instead'
1093 : '-and => [ \$scalar, \%rest_of_conditions ] instead'
1097 ARRAYREFREF => sub {
1098 puke "-$op => \\[...] makes little sense, use " .
1100 ? '[ \[...], \%rest_of_conditions ] instead'
1101 : '-and => [ \[...], \%rest_of_conditions ] instead'
1105 SCALAR => sub { # permissively interpreted as SQL
1106 puke "-$op => \$value makes little sense, use -bool => \$value instead";
1110 puke "-$op => undef not supported";
1115 sub _where_op_NEST {
1116 my ($self, $op, $v) = @_;
1118 $self->_SWITCH_refkind($v, {
1120 SCALAR => sub { # permissively interpreted as SQL
1121 belch "literal SQL should be -nest => \\'scalar' "
1122 . "instead of -nest => 'scalar' ";
1127 puke "-$op => undef not supported";
1131 $self->_recurse_where($v);
1138 sub _where_op_BOOL {
1139 my ($self, $op, $v) = @_;
1141 my ($s, @b) = $self->_SWITCH_refkind($v, {
1142 SCALAR => sub { # interpreted as SQL column
1143 $self->_convert($self->_quote($v));
1147 puke "-$op => undef not supported";
1151 $self->_recurse_where($v);
1155 $s = "(NOT $s)" if $op =~ /^not/i;
1160 sub _where_op_IDENT {
1162 my ($op, $rhs) = splice @_, -2;
1163 if (! defined $rhs or length ref $rhs) {
1164 puke "-$op requires a single plain scalar argument (a quotable identifier)";
1167 # in case we are called as a top level special op (no '=')
1168 my $has_lhs = my $lhs = shift;
1170 $_ = $self->_convert($self->_quote($_)) for ($lhs, $rhs);
1178 sub _where_op_VALUE {
1180 my ($op, $rhs) = splice @_, -2;
1182 # in case we are called as a top level special op (no '=')
1186 if (! defined $rhs) {
1188 ? $self->_where_hashpair_HASHREF($lhs, { -is => undef })
1195 (defined $lhs ? $lhs : $self->{_nested_func_lhs}),
1202 $self->_convert($self->_quote($lhs)) . ' = ' . $self->_convert('?'),
1206 $self->_convert('?'),
1213 my %unop_postfix = map +($_ => 1), 'is null', 'is not null';
1219 my ($self, $args) = @_;
1220 my ($left, $low, $high) = @$args;
1221 my ($rhsql, @rhbind) = do {
1223 puke "Single arg to between must be a literal"
1224 unless $low->{-literal};
1227 local $self->{_nested_func_lhs} = $left->{-ident}
1228 if ref($left) eq 'HASH' and $left->{-ident};
1229 my ($l, $h) = map [ $self->_where_unary_op(%$_) ], $low, $high;
1230 (join(' ', $l->[0], $self->_sqlcase('and'), $h->[0]),
1231 @{$l}[1..$#$l], @{$h}[1..$#$h])
1234 my ($lhsql, @lhbind) = $self->_recurse_where($left);
1236 join(' ', '(', $lhsql, $self->_sqlcase($op), $rhsql, ')'),
1240 }), 'between', 'not between'),
1244 my ($self, $args) = @_;
1245 my ($lhs, $rhs) = @$args;
1248 local $self->{_nested_func_lhs} = $lhs->{-ident}
1249 if ref($lhs) eq 'HASH' and $lhs->{-ident};
1250 my ($sql, @bind) = $self->_where_unary_op(%$_);
1251 push @in_bind, @bind;
1254 my ($lhsql, @lbind) = $self->_recurse_where($lhs);
1256 $lhsql.' '.$self->_sqlcase($op).' ( '
1257 .join(', ', @in_sql)
1262 }), 'in', 'not in'),
1266 my ($self, undef, $v) = @_;
1267 my ($op, @args) = @$v;
1268 $op =~ s/^-// if length($op) > 1;
1269 local $self->{_nested_func_lhs};
1270 if (my $h = $special{$op}) {
1271 return $self->$h(\@args);
1273 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{user_special_ops}}) {
1274 puke "Special op '${op}' requires first value to be identifier"
1275 unless my ($k) = map $_->{-ident}, grep ref($_) eq 'HASH', $args[0];
1276 return $self->${\($us->{handler})}($k, $op, $args[1]);
1278 my $final_op = $op =~ /^(?:is|not)_/ ? join(' ', split '_', $op) : $op;
1280 my ($expr_sql, @bind) = $self->_recurse_where($args[0]);
1281 my $op_sql = $self->_sqlcase($final_op);
1283 $unop_postfix{lc($final_op)}
1284 ? "${expr_sql} ${op_sql}"
1285 : "${op_sql} ${expr_sql}"
1287 return ($final_sql, @bind);
1288 } elsif (@args == 2) {
1289 my ($l, $r) = map [ $self->_recurse_where($_) ], @args;
1291 $l->[0].' '.$self->_sqlcase($final_op).' '.$r->[0],
1292 @{$l}[1..$#$l], @{$r}[1..$#$r]
1298 sub _where_op_FUNC {
1299 my ($self, undef, $rest) = @_;
1300 my ($func, @args) = @$rest;
1304 push @arg_sql, shift @x;
1306 } map [ $self->_recurse_where($_) ], @args;
1307 return ($self->_sqlcase($func).'('.join(', ', @arg_sql).')', @bind);
1310 sub _where_op_BIND {
1311 my ($self, undef, $bind) = @_;
1312 return ($self->_convert('?'), $self->_bindtype(@$bind));
1315 sub _where_op_LITERAL {
1316 my ($self, undef, $literal) = @_;
1317 $self->_assert_bindval_matches_bindtype(@{$literal}[1..$#$literal]);
1321 sub _where_hashpair_ARRAYREF {
1322 my ($self, $k, $v) = @_;
1325 my @v = @$v; # need copy because of shift below
1326 $self->_debug("ARRAY($k) means distribute over elements");
1328 # put apart first element if it is an operator (-and, -or)
1330 (defined $v[0] && $v[0] =~ /^ - (?: AND|OR ) $/ix)
1334 my @distributed = map { {$k => $_} } @v;
1337 $self->_debug("OP($op) reinjected into the distributed array");
1338 unshift @distributed, $op;
1341 my $logic = $op ? substr($op, 1) : '';
1343 return $self->_recurse_where(\@distributed, $logic);
1346 $self->_debug("empty ARRAY($k) means 0=1");
1347 return ($self->{sqlfalse});
1351 sub _where_hashpair_HASHREF {
1352 my ($self, $k, $v, $logic) = @_;
1355 local $self->{_nested_func_lhs} = defined $self->{_nested_func_lhs}
1356 ? $self->{_nested_func_lhs}
1360 my ($all_sql, @all_bind);
1362 for my $orig_op (sort keys %$v) {
1363 my $val = $v->{$orig_op};
1365 # put the operator in canonical form
1368 # FIXME - we need to phase out dash-less ops
1369 $op =~ s/^-//; # remove possible initial dash
1370 $op =~ s/^\s+|\s+$//g;# remove leading/trailing space
1371 $op =~ s/\s+/ /g; # compress whitespace
1373 $self->_assert_pass_injection_guard($op);
1376 $op =~ s/^is_not/IS NOT/i;
1378 # so that -not_foo works correctly
1379 $op =~ s/^not_/NOT /i;
1381 # another retarded special case: foo => { $op => { -value => undef } }
1382 if (ref $val eq 'HASH' and keys %$val == 1 and exists $val->{-value} and ! defined $val->{-value} ) {
1388 # CASE: col-value logic modifiers
1389 if ($orig_op =~ /^ \- (and|or) $/xi) {
1390 ($sql, @bind) = $self->_where_hashpair_HASHREF($k, $val, $1);
1392 # CASE: special operators like -in or -between
1393 elsif (my $special_op = List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}}) {
1394 my $handler = $special_op->{handler};
1396 puke "No handler supplied for special operator $orig_op";
1398 elsif (not ref $handler) {
1399 ($sql, @bind) = $self->$handler($k, $op, $val);
1401 elsif (ref $handler eq 'CODE') {
1402 ($sql, @bind) = $handler->($self, $k, $op, $val);
1405 puke "Illegal handler for special operator $orig_op - expecting a method name or a coderef";
1409 $self->_SWITCH_refkind($val, {
1411 ARRAYREF => sub { # CASE: col => {op => \@vals}
1412 ($sql, @bind) = $self->_where_field_op_ARRAYREF($k, $op, $val);
1415 ARRAYREFREF => sub { # CASE: col => {op => \[$sql, @bind]} (literal SQL with bind)
1416 my ($sub_sql, @sub_bind) = @$$val;
1417 $self->_assert_bindval_matches_bindtype(@sub_bind);
1418 $sql = join ' ', $self->_convert($self->_quote($k)),
1419 $self->_sqlcase($op),
1424 UNDEF => sub { # CASE: col => {op => undef} : sql "IS (NOT)? NULL"
1426 $op =~ /^not$/i ? 'is not' # legacy
1427 : $op =~ $self->{equality_op} ? 'is'
1428 : $op =~ $self->{like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is'
1429 : $op =~ $self->{inequality_op} ? 'is not'
1430 : $op =~ $self->{not_like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is not'
1431 : puke "unexpected operator '$orig_op' with undef operand";
1433 $sql = $self->_quote($k) . $self->_sqlcase(" $is null");
1436 FALLBACK => sub { # CASE: col => {op/func => $stuff}
1437 ($sql, @bind) = $self->_where_unary_op($op, $val);
1440 $self->_convert($self->_quote($k)),
1441 $self->{_nested_func_lhs} eq $k ? $sql : "($sql)", # top level vs nested
1447 ($all_sql) = (defined $all_sql and $all_sql) ? $self->_join_sql_clauses($logic, [$all_sql, $sql], []) : $sql;
1448 push @all_bind, @bind;
1450 return ($all_sql, @all_bind);
1453 sub _where_field_IS {
1454 my ($self, $k, $op, $v) = @_;
1456 my ($s) = $self->_SWITCH_refkind($v, {
1459 $self->_convert($self->_quote($k)),
1460 map { $self->_sqlcase($_)} ($op, 'null')
1463 puke "$op can only take undef as argument";
1470 sub _where_field_op_ARRAYREF {
1471 my ($self, $k, $op, $vals) = @_;
1473 my @vals = @$vals; #always work on a copy
1476 $self->_debug(sprintf '%s means multiple elements: [ %s ]',
1478 join(', ', map { defined $_ ? "'$_'" : 'NULL' } @vals ),
1481 # see if the first element is an -and/-or op
1483 if (defined $vals[0] && $vals[0] =~ /^ - (AND|OR) $/ix) {
1488 # a long standing API wart - an attempt to change this behavior during
1489 # the 1.50 series failed *spectacularly*. Warn instead and leave the
1494 (!$logic or $logic eq 'OR')
1496 ($op =~ $self->{inequality_op} or $op =~ $self->{not_like_op})
1499 belch "A multi-element arrayref as an argument to the inequality op '$o' "
1500 . 'is technically equivalent to an always-true 1=1 (you probably wanted '
1501 . "to say ...{ \$inequality_op => [ -and => \@values ] }... instead)"
1505 # distribute $op over each remaining member of @vals, append logic if exists
1506 return $self->_recurse_where([map { {$k => {$op, $_}} } @vals], $logic);
1510 # try to DWIM on equality operators
1512 $op =~ $self->{equality_op} ? $self->{sqlfalse}
1513 : $op =~ $self->{like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->{sqlfalse}
1514 : $op =~ $self->{inequality_op} ? $self->{sqltrue}
1515 : $op =~ $self->{not_like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->{sqltrue}
1516 : puke "operator '$op' applied on an empty array (field '$k')";
1521 sub _where_hashpair_SCALARREF {
1522 my ($self, $k, $v) = @_;
1523 $self->_debug("SCALAR($k) means literal SQL: $$v");
1524 my $sql = $self->_quote($k) . " " . $$v;
1528 # literal SQL with bind
1529 sub _where_hashpair_ARRAYREFREF {
1530 my ($self, $k, $v) = @_;
1531 $self->_debug("REF($k) means literal SQL: @${$v}");
1532 my ($sql, @bind) = @$$v;
1533 $self->_assert_bindval_matches_bindtype(@bind);
1534 $sql = $self->_quote($k) . " " . $sql;
1535 return ($sql, @bind );
1538 # literal SQL without bind
1539 sub _where_hashpair_SCALAR {
1540 my ($self, $k, $v) = @_;
1541 $self->_debug("NOREF($k) means simple key=val: $k $self->{cmp} $v");
1542 return ($self->_where_hashpair_HASHREF($k, { $self->{cmp} => $v }));
1546 sub _where_hashpair_UNDEF {
1547 my ($self, $k, $v) = @_;
1548 $self->_debug("UNDEF($k) means IS NULL");
1549 return $self->_where_hashpair_HASHREF($k, { -is => undef });
1552 #======================================================================
1553 # WHERE: TOP-LEVEL OTHERS (SCALARREF, SCALAR, UNDEF)
1554 #======================================================================
1557 sub _where_SCALARREF {
1558 my ($self, $where) = @_;
1561 $self->_debug("SCALAR(*top) means literal SQL: $$where");
1567 my ($self, $where) = @_;
1570 $self->_debug("NOREF(*top) means literal SQL: $where");
1581 #======================================================================
1582 # WHERE: BUILTIN SPECIAL OPERATORS (-in, -between)
1583 #======================================================================
1586 sub _where_field_BETWEEN {
1587 my ($self, $k, $op, $vals) = @_;
1589 my ($label, $and, $placeholder);
1590 $label = $self->_convert($self->_quote($k));
1591 $and = ' ' . $self->_sqlcase('and') . ' ';
1592 $placeholder = $self->_convert('?');
1593 $op = $self->_sqlcase($op);
1595 my $invalid_args = "Operator '$op' requires either an arrayref with two defined values or expressions, or a single literal scalarref/arrayref-ref";
1597 my ($clause, @bind) = $self->_SWITCH_refkind($vals, {
1598 ARRAYREFREF => sub {
1599 my ($s, @b) = @$$vals;
1600 $self->_assert_bindval_matches_bindtype(@b);
1607 puke $invalid_args if @$vals != 2;
1609 my (@all_sql, @all_bind);
1610 foreach my $val (@$vals) {
1611 my ($sql, @bind) = $self->_SWITCH_refkind($val, {
1613 return ($placeholder, $self->_bindtype($k, $val) );
1618 ARRAYREFREF => sub {
1619 my ($sql, @bind) = @$$val;
1620 $self->_assert_bindval_matches_bindtype(@bind);
1621 return ($sql, @bind);
1624 my ($func, $arg, @rest) = %$val;
1625 puke "Only simple { -func => arg } functions accepted as sub-arguments to BETWEEN"
1626 if (@rest or $func !~ /^ \- (.+)/x);
1627 $self->_where_unary_op($1 => $arg);
1633 push @all_sql, $sql;
1634 push @all_bind, @bind;
1638 (join $and, @all_sql),
1647 my $sql = "( $label $op $clause )";
1648 return ($sql, @bind)
1652 sub _where_field_IN {
1653 my ($self, $k, $op, $vals) = @_;
1655 # backwards compatibility: if scalar, force into an arrayref
1656 $vals = [$vals] if defined $vals && ! ref $vals;
1658 my ($label) = $self->_convert($self->_quote($k));
1659 my ($placeholder) = $self->_convert('?');
1660 $op = $self->_sqlcase($op);
1662 my ($sql, @bind) = $self->_SWITCH_refkind($vals, {
1663 ARRAYREF => sub { # list of choices
1664 if (@$vals) { # nonempty list
1665 my (@all_sql, @all_bind);
1667 for my $val (@$vals) {
1668 my ($sql, @bind) = $self->_SWITCH_refkind($val, {
1670 return ($placeholder, $val);
1675 ARRAYREFREF => sub {
1676 my ($sql, @bind) = @$$val;
1677 $self->_assert_bindval_matches_bindtype(@bind);
1678 return ($sql, @bind);
1681 my ($func, $arg, @rest) = %$val;
1682 puke "Only simple { -func => arg } functions accepted as sub-arguments to IN"
1683 if (@rest or $func !~ /^ \- (.+)/x);
1684 $self->_where_unary_op($1 => $arg);
1688 'SQL::Abstract before v1.75 used to generate incorrect SQL when the '
1689 . "-$op operator was given an undef-containing list: !!!AUDIT YOUR CODE "
1690 . 'AND DATA!!! (the upcoming Data::Query-based version of SQL::Abstract '
1691 . 'will emit the logically correct SQL instead of raising this exception)'
1695 push @all_sql, $sql;
1696 push @all_bind, @bind;
1700 sprintf('%s %s ( %s )',
1703 join(', ', @all_sql)
1705 $self->_bindtype($k, @all_bind),
1708 else { # empty list: some databases won't understand "IN ()", so DWIM
1709 my $sql = ($op =~ /\bnot\b/i) ? $self->{sqltrue} : $self->{sqlfalse};
1714 SCALARREF => sub { # literal SQL
1715 my $sql = $self->_open_outer_paren($$vals);
1716 return ("$label $op ( $sql )");
1718 ARRAYREFREF => sub { # literal SQL with bind
1719 my ($sql, @bind) = @$$vals;
1720 $self->_assert_bindval_matches_bindtype(@bind);
1721 $sql = $self->_open_outer_paren($sql);
1722 return ("$label $op ( $sql )", @bind);
1726 puke "Argument passed to the '$op' operator can not be undefined";
1730 puke "special op $op requires an arrayref (or scalarref/arrayref-ref)";
1734 return ($sql, @bind);
1737 # Some databases (SQLite) treat col IN (1, 2) different from
1738 # col IN ( (1, 2) ). Use this to strip all outer parens while
1739 # adding them back in the corresponding method
1740 sub _open_outer_paren {
1741 my ($self, $sql) = @_;
1743 while (my ($inner) = $sql =~ /^ \s* \( (.*) \) \s* $/xs) {
1745 # there are closing parens inside, need the heavy duty machinery
1746 # to reevaluate the extraction starting from $sql (full reevaluation)
1747 if ($inner =~ /\)/) {
1748 require Text::Balanced;
1750 my (undef, $remainder) = do {
1751 # idiotic design - writes to $@ but *DOES NOT* throw exceptions
1753 Text::Balanced::extract_bracketed($sql, '()', qr/\s*/);
1756 # the entire expression needs to be a balanced bracketed thing
1757 # (after an extract no remainder sans trailing space)
1758 last if defined $remainder and $remainder =~ /\S/;
1768 #======================================================================
1770 #======================================================================
1773 my ($self, $arg) = @_;
1776 for my $c ($self->_order_by_chunks($arg) ) {
1777 $self->_SWITCH_refkind($c, {
1778 SCALAR => sub { push @sql, $c },
1779 ARRAYREF => sub { push @sql, shift @$c; push @bind, @$c },
1785 $self->_sqlcase(' order by'),
1791 return wantarray ? ($sql, @bind) : $sql;
1794 sub _order_by_chunks {
1795 my ($self, $arg) = @_;
1797 return $self->_SWITCH_refkind($arg, {
1800 map { $self->_order_by_chunks($_ ) } @$arg;
1803 ARRAYREFREF => sub {
1804 my ($s, @b) = @$$arg;
1805 $self->_assert_bindval_matches_bindtype(@b);
1809 SCALAR => sub {$self->_quote($arg)},
1811 UNDEF => sub {return () },
1813 SCALARREF => sub {$$arg}, # literal SQL, no quoting
1816 # get first pair in hash
1817 my ($key, $val, @rest) = %$arg;
1819 return () unless $key;
1821 if (@rest or not $key =~ /^-(desc|asc)/i) {
1822 puke "hash passed to _order_by must have exactly one key (-desc or -asc)";
1828 for my $c ($self->_order_by_chunks($val)) {
1831 $self->_SWITCH_refkind($c, {
1836 ($sql, @bind) = @$c;
1840 $sql = $sql . ' ' . $self->_sqlcase($direction);
1842 push @ret, [ $sql, @bind];
1851 #======================================================================
1852 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
1853 #======================================================================
1858 $self->_SWITCH_refkind($from, {
1859 ARRAYREF => sub {join ', ', map { $self->_quote($_) } @$from;},
1860 SCALAR => sub {$self->_quote($from)},
1861 SCALARREF => sub {$$from},
1866 #======================================================================
1868 #======================================================================
1870 # highly optimized, as it's called way too often
1872 # my ($self, $label) = @_;
1874 return '' unless defined $_[1];
1875 return ${$_[1]} if ref($_[1]) eq 'SCALAR';
1877 $_[0]->{quote_char} or
1878 ($_[0]->_assert_pass_injection_guard($_[1]), return $_[1]);
1880 my $qref = ref $_[0]->{quote_char};
1882 !$qref ? ($_[0]->{quote_char}, $_[0]->{quote_char})
1883 : ($qref eq 'ARRAY') ? @{$_[0]->{quote_char}}
1884 : puke "Unsupported quote_char format: $_[0]->{quote_char}";
1886 my $esc = $_[0]->{escape_char} || $r;
1888 # parts containing * are naturally unquoted
1889 return join($_[0]->{name_sep}||'', map
1890 +( $_ eq '*' ? $_ : do { (my $n = $_) =~ s/(\Q$esc\E|\Q$r\E)/$esc$1/g; $l . $n . $r } ),
1891 ( $_[0]->{name_sep} ? split (/\Q$_[0]->{name_sep}\E/, $_[1] ) : $_[1] )
1896 # Conversion, if applicable
1898 #my ($self, $arg) = @_;
1899 if ($_[0]->{convert}) {
1900 return $_[0]->_sqlcase($_[0]->{convert}) .'(' . $_[1] . ')';
1907 #my ($self, $col, @vals) = @_;
1908 # called often - tighten code
1909 return $_[0]->{bindtype} eq 'columns'
1910 ? map {[$_[1], $_]} @_[2 .. $#_]
1915 # Dies if any element of @bind is not in [colname => value] format
1916 # if bindtype is 'columns'.
1917 sub _assert_bindval_matches_bindtype {
1918 # my ($self, @bind) = @_;
1920 if ($self->{bindtype} eq 'columns') {
1922 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
1923 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
1929 sub _join_sql_clauses {
1930 my ($self, $logic, $clauses_aref, $bind_aref) = @_;
1932 if (@$clauses_aref > 1) {
1933 my $join = " " . $self->_sqlcase($logic) . " ";
1934 my $sql = '( ' . join($join, @$clauses_aref) . ' )';
1935 return ($sql, @$bind_aref);
1937 elsif (@$clauses_aref) {
1938 return ($clauses_aref->[0], @$bind_aref); # no parentheses
1941 return (); # if no SQL, ignore @$bind_aref
1946 # Fix SQL case, if so requested
1948 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
1949 # don't touch the argument ... crooked logic, but let's not change it!
1950 return $_[0]->{case} ? $_[1] : uc($_[1]);
1954 #======================================================================
1955 # DISPATCHING FROM REFKIND
1956 #======================================================================
1959 my ($self, $data) = @_;
1961 return 'UNDEF' unless defined $data;
1963 # blessed objects are treated like scalars
1964 my $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1966 return 'SCALAR' unless $ref;
1969 while ($ref eq 'REF') {
1971 $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1975 return ($ref||'SCALAR') . ('REF' x $n_steps);
1979 my ($self, $data) = @_;
1980 my @try = ($self->_refkind($data));
1981 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
1982 push @try, 'FALLBACK';
1986 sub _METHOD_FOR_refkind {
1987 my ($self, $meth_prefix, $data) = @_;
1990 for (@{$self->_try_refkind($data)}) {
1991 $method = $self->can($meth_prefix."_".$_)
1995 return $method || puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
1999 sub _SWITCH_refkind {
2000 my ($self, $data, $dispatch_table) = @_;
2003 for (@{$self->_try_refkind($data)}) {
2004 $coderef = $dispatch_table->{$_}
2008 puke "no dispatch entry for ".$self->_refkind($data)
2017 #======================================================================
2018 # VALUES, GENERATE, AUTOLOAD
2019 #======================================================================
2021 # LDNOTE: original code from nwiger, didn't touch code in that section
2022 # I feel the AUTOLOAD stuff should not be the default, it should
2023 # only be activated on explicit demand by user.
2027 my $data = shift || return;
2028 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
2029 unless ref $data eq 'HASH';
2032 foreach my $k (sort keys %$data) {
2033 my $v = $data->{$k};
2034 $self->_SWITCH_refkind($v, {
2036 if ($self->{array_datatypes}) { # array datatype
2037 push @all_bind, $self->_bindtype($k, $v);
2039 else { # literal SQL with bind
2040 my ($sql, @bind) = @$v;
2041 $self->_assert_bindval_matches_bindtype(@bind);
2042 push @all_bind, @bind;
2045 ARRAYREFREF => sub { # literal SQL with bind
2046 my ($sql, @bind) = @${$v};
2047 $self->_assert_bindval_matches_bindtype(@bind);
2048 push @all_bind, @bind;
2050 SCALARREF => sub { # literal SQL without bind
2052 SCALAR_or_UNDEF => sub {
2053 push @all_bind, $self->_bindtype($k, $v);
2064 my(@sql, @sqlq, @sqlv);
2068 if ($ref eq 'HASH') {
2069 for my $k (sort keys %$_) {
2072 my $label = $self->_quote($k);
2073 if ($r eq 'ARRAY') {
2074 # literal SQL with bind
2075 my ($sql, @bind) = @$v;
2076 $self->_assert_bindval_matches_bindtype(@bind);
2077 push @sqlq, "$label = $sql";
2079 } elsif ($r eq 'SCALAR') {
2080 # literal SQL without bind
2081 push @sqlq, "$label = $$v";
2083 push @sqlq, "$label = ?";
2084 push @sqlv, $self->_bindtype($k, $v);
2087 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
2088 } elsif ($ref eq 'ARRAY') {
2089 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
2092 if ($r eq 'ARRAY') { # literal SQL with bind
2093 my ($sql, @bind) = @$v;
2094 $self->_assert_bindval_matches_bindtype(@bind);
2097 } elsif ($r eq 'SCALAR') { # literal SQL without bind
2098 # embedded literal SQL
2105 push @sql, '(' . join(', ', @sqlq) . ')';
2106 } elsif ($ref eq 'SCALAR') {
2110 # strings get case twiddled
2111 push @sql, $self->_sqlcase($_);
2115 my $sql = join ' ', @sql;
2117 # this is pretty tricky
2118 # if ask for an array, return ($stmt, @bind)
2119 # otherwise, s/?/shift @sqlv/ to put it inline
2121 return ($sql, @sqlv);
2123 1 while $sql =~ s/\?/my $d = shift(@sqlv);
2124 ref $d ? $d->[1] : $d/e;
2133 # This allows us to check for a local, then _form, attr
2135 my($name) = $AUTOLOAD =~ /.*::(.+)/;
2136 return $self->generate($name, @_);
2147 SQL::Abstract - Generate SQL from Perl data structures
2153 my $sql = SQL::Abstract->new;
2155 my($stmt, @bind) = $sql->select($source, \@fields, \%where, $order);
2157 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
2159 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
2161 my($stmt, @bind) = $sql->delete($table, \%where);
2163 # Then, use these in your DBI statements
2164 my $sth = $dbh->prepare($stmt);
2165 $sth->execute(@bind);
2167 # Just generate the WHERE clause
2168 my($stmt, @bind) = $sql->where(\%where, $order);
2170 # Return values in the same order, for hashed queries
2171 # See PERFORMANCE section for more details
2172 my @bind = $sql->values(\%fieldvals);
2176 This module was inspired by the excellent L<DBIx::Abstract>.
2177 However, in using that module I found that what I really wanted
2178 to do was generate SQL, but still retain complete control over my
2179 statement handles and use the DBI interface. So, I set out to
2180 create an abstract SQL generation module.
2182 While based on the concepts used by L<DBIx::Abstract>, there are
2183 several important differences, especially when it comes to WHERE
2184 clauses. I have modified the concepts used to make the SQL easier
2185 to generate from Perl data structures and, IMO, more intuitive.
2186 The underlying idea is for this module to do what you mean, based
2187 on the data structures you provide it. The big advantage is that
2188 you don't have to modify your code every time your data changes,
2189 as this module figures it out.
2191 To begin with, an SQL INSERT is as easy as just specifying a hash
2192 of C<key=value> pairs:
2195 name => 'Jimbo Bobson',
2196 phone => '123-456-7890',
2197 address => '42 Sister Lane',
2198 city => 'St. Louis',
2199 state => 'Louisiana',
2202 The SQL can then be generated with this:
2204 my($stmt, @bind) = $sql->insert('people', \%data);
2206 Which would give you something like this:
2208 $stmt = "INSERT INTO people
2209 (address, city, name, phone, state)
2210 VALUES (?, ?, ?, ?, ?)";
2211 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
2212 '123-456-7890', 'Louisiana');
2214 These are then used directly in your DBI code:
2216 my $sth = $dbh->prepare($stmt);
2217 $sth->execute(@bind);
2219 =head2 Inserting and Updating Arrays
2221 If your database has array types (like for example Postgres),
2222 activate the special option C<< array_datatypes => 1 >>
2223 when creating the C<SQL::Abstract> object.
2224 Then you may use an arrayref to insert and update database array types:
2226 my $sql = SQL::Abstract->new(array_datatypes => 1);
2228 planets => [qw/Mercury Venus Earth Mars/]
2231 my($stmt, @bind) = $sql->insert('solar_system', \%data);
2235 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
2237 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
2240 =head2 Inserting and Updating SQL
2242 In order to apply SQL functions to elements of your C<%data> you may
2243 specify a reference to an arrayref for the given hash value. For example,
2244 if you need to execute the Oracle C<to_date> function on a value, you can
2245 say something like this:
2249 date_entered => \[ "to_date(?,'MM/DD/YYYY')", "03/02/2003" ],
2252 The first value in the array is the actual SQL. Any other values are
2253 optional and would be included in the bind values array. This gives
2256 my($stmt, @bind) = $sql->insert('people', \%data);
2258 $stmt = "INSERT INTO people (name, date_entered)
2259 VALUES (?, to_date(?,'MM/DD/YYYY'))";
2260 @bind = ('Bill', '03/02/2003');
2262 An UPDATE is just as easy, all you change is the name of the function:
2264 my($stmt, @bind) = $sql->update('people', \%data);
2266 Notice that your C<%data> isn't touched; the module will generate
2267 the appropriately quirky SQL for you automatically. Usually you'll
2268 want to specify a WHERE clause for your UPDATE, though, which is
2269 where handling C<%where> hashes comes in handy...
2271 =head2 Complex where statements
2273 This module can generate pretty complicated WHERE statements
2274 easily. For example, simple C<key=value> pairs are taken to mean
2275 equality, and if you want to see if a field is within a set
2276 of values, you can use an arrayref. Let's say we wanted to
2277 SELECT some data based on this criteria:
2280 requestor => 'inna',
2281 worker => ['nwiger', 'rcwe', 'sfz'],
2282 status => { '!=', 'completed' }
2285 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
2287 The above would give you something like this:
2289 $stmt = "SELECT * FROM tickets WHERE
2290 ( requestor = ? ) AND ( status != ? )
2291 AND ( worker = ? OR worker = ? OR worker = ? )";
2292 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
2294 Which you could then use in DBI code like so:
2296 my $sth = $dbh->prepare($stmt);
2297 $sth->execute(@bind);
2303 The methods are simple. There's one for every major SQL operation,
2304 and a constructor you use first. The arguments are specified in a
2305 similar order for each method (table, then fields, then a where
2306 clause) to try and simplify things.
2308 =head2 new(option => 'value')
2310 The C<new()> function takes a list of options and values, and returns
2311 a new B<SQL::Abstract> object which can then be used to generate SQL
2312 through the methods below. The options accepted are:
2318 If set to 'lower', then SQL will be generated in all lowercase. By
2319 default SQL is generated in "textbook" case meaning something like:
2321 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
2323 Any setting other than 'lower' is ignored.
2327 This determines what the default comparison operator is. By default
2328 it is C<=>, meaning that a hash like this:
2330 %where = (name => 'nwiger', email => 'nate@wiger.org');
2332 Will generate SQL like this:
2334 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
2336 However, you may want loose comparisons by default, so if you set
2337 C<cmp> to C<like> you would get SQL such as:
2339 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
2341 You can also override the comparison on an individual basis - see
2342 the huge section on L</"WHERE CLAUSES"> at the bottom.
2344 =item sqltrue, sqlfalse
2346 Expressions for inserting boolean values within SQL statements.
2347 By default these are C<1=1> and C<1=0>. They are used
2348 by the special operators C<-in> and C<-not_in> for generating
2349 correct SQL even when the argument is an empty array (see below).
2353 This determines the default logical operator for multiple WHERE
2354 statements in arrays or hashes. If absent, the default logic is "or"
2355 for arrays, and "and" for hashes. This means that a WHERE
2359 event_date => {'>=', '2/13/99'},
2360 event_date => {'<=', '4/24/03'},
2363 will generate SQL like this:
2365 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
2367 This is probably not what you want given this query, though (look
2368 at the dates). To change the "OR" to an "AND", simply specify:
2370 my $sql = SQL::Abstract->new(logic => 'and');
2372 Which will change the above C<WHERE> to:
2374 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
2376 The logic can also be changed locally by inserting
2377 a modifier in front of an arrayref:
2379 @where = (-and => [event_date => {'>=', '2/13/99'},
2380 event_date => {'<=', '4/24/03'} ]);
2382 See the L</"WHERE CLAUSES"> section for explanations.
2386 This will automatically convert comparisons using the specified SQL
2387 function for both column and value. This is mostly used with an argument
2388 of C<upper> or C<lower>, so that the SQL will have the effect of
2389 case-insensitive "searches". For example, this:
2391 $sql = SQL::Abstract->new(convert => 'upper');
2392 %where = (keywords => 'MaKe iT CAse inSeNSItive');
2394 Will turn out the following SQL:
2396 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
2398 The conversion can be C<upper()>, C<lower()>, or any other SQL function
2399 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
2400 not validate this option; it will just pass through what you specify verbatim).
2404 This is a kludge because many databases suck. For example, you can't
2405 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
2406 Instead, you have to use C<bind_param()>:
2408 $sth->bind_param(1, 'reg data');
2409 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
2411 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
2412 which loses track of which field each slot refers to. Fear not.
2414 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
2415 Currently, you can specify either C<normal> (default) or C<columns>. If you
2416 specify C<columns>, you will get an array that looks like this:
2418 my $sql = SQL::Abstract->new(bindtype => 'columns');
2419 my($stmt, @bind) = $sql->insert(...);
2422 [ 'column1', 'value1' ],
2423 [ 'column2', 'value2' ],
2424 [ 'column3', 'value3' ],
2427 You can then iterate through this manually, using DBI's C<bind_param()>.
2429 $sth->prepare($stmt);
2432 my($col, $data) = @$_;
2433 if ($col eq 'details' || $col eq 'comments') {
2434 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
2435 } elsif ($col eq 'image') {
2436 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
2438 $sth->bind_param($i, $data);
2442 $sth->execute; # execute without @bind now
2444 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
2445 Basically, the advantage is still that you don't have to care which fields
2446 are or are not included. You could wrap that above C<for> loop in a simple
2447 sub called C<bind_fields()> or something and reuse it repeatedly. You still
2448 get a layer of abstraction over manual SQL specification.
2450 Note that if you set L</bindtype> to C<columns>, the C<\[ $sql, @bind ]>
2451 construct (see L</Literal SQL with placeholders and bind values (subqueries)>)
2452 will expect the bind values in this format.
2456 This is the character that a table or column name will be quoted
2457 with. By default this is an empty string, but you could set it to
2458 the character C<`>, to generate SQL like this:
2460 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
2462 Alternatively, you can supply an array ref of two items, the first being the left
2463 hand quote character, and the second the right hand quote character. For
2464 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
2465 that generates SQL like this:
2467 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
2469 Quoting is useful if you have tables or columns names that are reserved
2470 words in your database's SQL dialect.
2474 This is the character that will be used to escape L</quote_char>s appearing
2475 in an identifier before it has been quoted.
2477 The parameter default in case of a single L</quote_char> character is the quote
2480 When opening-closing-style quoting is used (L</quote_char> is an arrayref)
2481 this parameter defaults to the B<closing (right)> L</quote_char>. Occurrences
2482 of the B<opening (left)> L</quote_char> within the identifier are currently left
2483 untouched. The default for opening-closing-style quotes may change in future
2484 versions, thus you are B<strongly encouraged> to specify the escape character
2489 This is the character that separates a table and column name. It is
2490 necessary to specify this when the C<quote_char> option is selected,
2491 so that tables and column names can be individually quoted like this:
2493 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
2495 =item injection_guard
2497 A regular expression C<qr/.../> that is applied to any C<-function> and unquoted
2498 column name specified in a query structure. This is a safety mechanism to avoid
2499 injection attacks when mishandling user input e.g.:
2501 my %condition_as_column_value_pairs = get_values_from_user();
2502 $sqla->select( ... , \%condition_as_column_value_pairs );
2504 If the expression matches an exception is thrown. Note that literal SQL
2505 supplied via C<\'...'> or C<\['...']> is B<not> checked in any way.
2507 Defaults to checking for C<;> and the C<GO> keyword (TransactSQL)
2509 =item array_datatypes
2511 When this option is true, arrayrefs in INSERT or UPDATE are
2512 interpreted as array datatypes and are passed directly
2514 When this option is false, arrayrefs are interpreted
2515 as literal SQL, just like refs to arrayrefs
2516 (but this behavior is for backwards compatibility; when writing
2517 new queries, use the "reference to arrayref" syntax
2523 Takes a reference to a list of "special operators"
2524 to extend the syntax understood by L<SQL::Abstract>.
2525 See section L</"SPECIAL OPERATORS"> for details.
2529 Takes a reference to a list of "unary operators"
2530 to extend the syntax understood by L<SQL::Abstract>.
2531 See section L</"UNARY OPERATORS"> for details.
2537 =head2 insert($table, \@values || \%fieldvals, \%options)
2539 This is the simplest function. You simply give it a table name
2540 and either an arrayref of values or hashref of field/value pairs.
2541 It returns an SQL INSERT statement and a list of bind values.
2542 See the sections on L</"Inserting and Updating Arrays"> and
2543 L</"Inserting and Updating SQL"> for information on how to insert
2544 with those data types.
2546 The optional C<\%options> hash reference may contain additional
2547 options to generate the insert SQL. Currently supported options
2554 Takes either a scalar of raw SQL fields, or an array reference of
2555 field names, and adds on an SQL C<RETURNING> statement at the end.
2556 This allows you to return data generated by the insert statement
2557 (such as row IDs) without performing another C<SELECT> statement.
2558 Note, however, this is not part of the SQL standard and may not
2559 be supported by all database engines.
2563 =head2 update($table, \%fieldvals, \%where, \%options)
2565 This takes a table, hashref of field/value pairs, and an optional
2566 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
2568 See the sections on L</"Inserting and Updating Arrays"> and
2569 L</"Inserting and Updating SQL"> for information on how to insert
2570 with those data types.
2572 The optional C<\%options> hash reference may contain additional
2573 options to generate the update SQL. Currently supported options
2580 See the C<returning> option to
2581 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2585 =head2 select($source, $fields, $where, $order)
2587 This returns a SQL SELECT statement and associated list of bind values, as
2588 specified by the arguments:
2594 Specification of the 'FROM' part of the statement.
2595 The argument can be either a plain scalar (interpreted as a table
2596 name, will be quoted), or an arrayref (interpreted as a list
2597 of table names, joined by commas, quoted), or a scalarref
2598 (literal SQL, not quoted).
2602 Specification of the list of fields to retrieve from
2604 The argument can be either an arrayref (interpreted as a list
2605 of field names, will be joined by commas and quoted), or a
2606 plain scalar (literal SQL, not quoted).
2607 Please observe that this API is not as flexible as that of
2608 the first argument C<$source>, for backwards compatibility reasons.
2612 Optional argument to specify the WHERE part of the query.
2613 The argument is most often a hashref, but can also be
2614 an arrayref or plain scalar --
2615 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
2619 Optional argument to specify the ORDER BY part of the query.
2620 The argument can be a scalar, a hashref or an arrayref
2621 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
2627 =head2 delete($table, \%where, \%options)
2629 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
2630 It returns an SQL DELETE statement and list of bind values.
2632 The optional C<\%options> hash reference may contain additional
2633 options to generate the delete SQL. Currently supported options
2640 See the C<returning> option to
2641 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
2645 =head2 where(\%where, $order)
2647 This is used to generate just the WHERE clause. For example,
2648 if you have an arbitrary data structure and know what the
2649 rest of your SQL is going to look like, but want an easy way
2650 to produce a WHERE clause, use this. It returns an SQL WHERE
2651 clause and list of bind values.
2654 =head2 values(\%data)
2656 This just returns the values from the hash C<%data>, in the same
2657 order that would be returned from any of the other above queries.
2658 Using this allows you to markedly speed up your queries if you
2659 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
2661 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
2663 Warning: This is an experimental method and subject to change.
2665 This returns arbitrarily generated SQL. It's a really basic shortcut.
2666 It will return two different things, depending on return context:
2668 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
2669 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
2671 These would return the following:
2673 # First calling form
2674 $stmt = "CREATE TABLE test (?, ?)";
2675 @bind = (field1, field2);
2677 # Second calling form
2678 $stmt_and_val = "CREATE TABLE test (field1, field2)";
2680 Depending on what you're trying to do, it's up to you to choose the correct
2681 format. In this example, the second form is what you would want.
2685 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
2689 ALTER SESSION SET nls_date_format = 'MM/YY'
2691 You get the idea. Strings get their case twiddled, but everything
2692 else remains verbatim.
2694 =head1 EXPORTABLE FUNCTIONS
2696 =head2 is_plain_value
2698 Determines if the supplied argument is a plain value as understood by this
2703 =item * The value is C<undef>
2705 =item * The value is a non-reference
2707 =item * The value is an object with stringification overloading
2709 =item * The value is of the form C<< { -value => $anything } >>
2713 On failure returns C<undef>, on success returns a B<scalar> reference
2714 to the original supplied argument.
2720 The stringification overloading detection is rather advanced: it takes
2721 into consideration not only the presence of a C<""> overload, but if that
2722 fails also checks for enabled
2723 L<autogenerated versions of C<"">|overload/Magic Autogeneration>, based
2724 on either C<0+> or C<bool>.
2726 Unfortunately testing in the field indicates that this
2727 detection B<< may tickle a latent bug in perl versions before 5.018 >>,
2728 but only when very large numbers of stringifying objects are involved.
2729 At the time of writing ( Sep 2014 ) there is no clear explanation of
2730 the direct cause, nor is there a manageably small test case that reliably
2731 reproduces the problem.
2733 If you encounter any of the following exceptions in B<random places within
2734 your application stack> - this module may be to blame:
2736 Operation "ne": no method found,
2737 left argument in overloaded package <something>,
2738 right argument in overloaded package <something>
2742 Stub found while resolving method "???" overloading """" in package <something>
2744 If you fall victim to the above - please attempt to reduce the problem
2745 to something that could be sent to the L<SQL::Abstract developers
2746 |DBIx::Class/GETTING HELP/SUPPORT>
2747 (either publicly or privately). As a workaround in the meantime you can
2748 set C<$ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}> to a true
2749 value, which will most likely eliminate your problem (at the expense of
2750 not being able to properly detect exotic forms of stringification).
2752 This notice and environment variable will be removed in a future version,
2753 as soon as the underlying problem is found and a reliable workaround is
2758 =head2 is_literal_value
2760 Determines if the supplied argument is a literal value as understood by this
2765 =item * C<\$sql_string>
2767 =item * C<\[ $sql_string, @bind_values ]>
2771 On failure returns C<undef>, on success returns an B<array> reference
2772 containing the unpacked version of the supplied literal SQL and bind values.
2774 =head1 WHERE CLAUSES
2778 This module uses a variation on the idea from L<DBIx::Abstract>. It
2779 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
2780 module is that things in arrays are OR'ed, and things in hashes
2783 The easiest way to explain is to show lots of examples. After
2784 each C<%where> hash shown, it is assumed you used:
2786 my($stmt, @bind) = $sql->where(\%where);
2788 However, note that the C<%where> hash can be used directly in any
2789 of the other functions as well, as described above.
2791 =head2 Key-value pairs
2793 So, let's get started. To begin, a simple hash:
2797 status => 'completed'
2800 Is converted to SQL C<key = val> statements:
2802 $stmt = "WHERE user = ? AND status = ?";
2803 @bind = ('nwiger', 'completed');
2805 One common thing I end up doing is having a list of values that
2806 a field can be in. To do this, simply specify a list inside of
2811 status => ['assigned', 'in-progress', 'pending'];
2814 This simple code will create the following:
2816 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
2817 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
2819 A field associated to an empty arrayref will be considered a
2820 logical false and will generate 0=1.
2822 =head2 Tests for NULL values
2824 If the value part is C<undef> then this is converted to SQL <IS NULL>
2833 $stmt = "WHERE user = ? AND status IS NULL";
2836 To test if a column IS NOT NULL:
2840 status => { '!=', undef },
2843 =head2 Specific comparison operators
2845 If you want to specify a different type of operator for your comparison,
2846 you can use a hashref for a given column:
2850 status => { '!=', 'completed' }
2853 Which would generate:
2855 $stmt = "WHERE user = ? AND status != ?";
2856 @bind = ('nwiger', 'completed');
2858 To test against multiple values, just enclose the values in an arrayref:
2860 status => { '=', ['assigned', 'in-progress', 'pending'] };
2862 Which would give you:
2864 "WHERE status = ? OR status = ? OR status = ?"
2867 The hashref can also contain multiple pairs, in which case it is expanded
2868 into an C<AND> of its elements:
2872 status => { '!=', 'completed', -not_like => 'pending%' }
2875 # Or more dynamically, like from a form
2876 $where{user} = 'nwiger';
2877 $where{status}{'!='} = 'completed';
2878 $where{status}{'-not_like'} = 'pending%';
2880 # Both generate this
2881 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
2882 @bind = ('nwiger', 'completed', 'pending%');
2885 To get an OR instead, you can combine it with the arrayref idea:
2889 priority => [ { '=', 2 }, { '>', 5 } ]
2892 Which would generate:
2894 $stmt = "WHERE ( priority = ? OR priority > ? ) AND user = ?";
2895 @bind = ('2', '5', 'nwiger');
2897 If you want to include literal SQL (with or without bind values), just use a
2898 scalar reference or reference to an arrayref as the value:
2901 date_entered => { '>' => \["to_date(?, 'MM/DD/YYYY')", "11/26/2008"] },
2902 date_expires => { '<' => \"now()" }
2905 Which would generate:
2907 $stmt = "WHERE date_entered > to_date(?, 'MM/DD/YYYY') AND date_expires < now()";
2908 @bind = ('11/26/2008');
2911 =head2 Logic and nesting operators
2913 In the example above,
2914 there is a subtle trap if you want to say something like
2915 this (notice the C<AND>):
2917 WHERE priority != ? AND priority != ?
2919 Because, in Perl you I<can't> do this:
2921 priority => { '!=' => 2, '!=' => 1 }
2923 As the second C<!=> key will obliterate the first. The solution
2924 is to use the special C<-modifier> form inside an arrayref:
2926 priority => [ -and => {'!=', 2},
2930 Normally, these would be joined by C<OR>, but the modifier tells it
2931 to use C<AND> instead. (Hint: You can use this in conjunction with the
2932 C<logic> option to C<new()> in order to change the way your queries
2933 work by default.) B<Important:> Note that the C<-modifier> goes
2934 B<INSIDE> the arrayref, as an extra first element. This will
2935 B<NOT> do what you think it might:
2937 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
2939 Here is a quick list of equivalencies, since there is some overlap:
2942 status => {'!=', 'completed', 'not like', 'pending%' }
2943 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
2946 status => {'=', ['assigned', 'in-progress']}
2947 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
2948 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
2952 =head2 Special operators: IN, BETWEEN, etc.
2954 You can also use the hashref format to compare a list of fields using the
2955 C<IN> comparison operator, by specifying the list as an arrayref:
2958 status => 'completed',
2959 reportid => { -in => [567, 2335, 2] }
2962 Which would generate:
2964 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
2965 @bind = ('completed', '567', '2335', '2');
2967 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
2970 If the argument to C<-in> is an empty array, 'sqlfalse' is generated
2971 (by default: C<1=0>). Similarly, C<< -not_in => [] >> generates
2972 'sqltrue' (by default: C<1=1>).
2974 In addition to the array you can supply a chunk of literal sql or
2975 literal sql with bind:
2978 customer => { -in => \[
2979 'SELECT cust_id FROM cust WHERE balance > ?',
2982 status => { -in => \'SELECT status_codes FROM states' },
2988 customer IN ( SELECT cust_id FROM cust WHERE balance > ? )
2989 AND status IN ( SELECT status_codes FROM states )
2993 Finally, if the argument to C<-in> is not a reference, it will be
2994 treated as a single-element array.
2996 Another pair of operators is C<-between> and C<-not_between>,
2997 used with an arrayref of two values:
3001 completion_date => {
3002 -not_between => ['2002-10-01', '2003-02-06']
3008 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
3010 Just like with C<-in> all plausible combinations of literal SQL
3014 start0 => { -between => [ 1, 2 ] },
3015 start1 => { -between => \["? AND ?", 1, 2] },
3016 start2 => { -between => \"lower(x) AND upper(y)" },
3017 start3 => { -between => [
3019 \["upper(?)", 'stuff' ],
3026 ( start0 BETWEEN ? AND ? )
3027 AND ( start1 BETWEEN ? AND ? )
3028 AND ( start2 BETWEEN lower(x) AND upper(y) )
3029 AND ( start3 BETWEEN lower(x) AND upper(?) )
3031 @bind = (1, 2, 1, 2, 'stuff');
3034 These are the two builtin "special operators"; but the
3035 list can be expanded: see section L</"SPECIAL OPERATORS"> below.
3037 =head2 Unary operators: bool
3039 If you wish to test against boolean columns or functions within your
3040 database you can use the C<-bool> and C<-not_bool> operators. For
3041 example to test the column C<is_user> being true and the column
3042 C<is_enabled> being false you would use:-
3046 -not_bool => 'is_enabled',
3051 WHERE is_user AND NOT is_enabled
3053 If a more complex combination is required, testing more conditions,
3054 then you should use the and/or operators:-
3059 -not_bool => { two=> { -rlike => 'bar' } },
3060 -not_bool => { three => [ { '=', 2 }, { '>', 5 } ] },
3071 (NOT ( three = ? OR three > ? ))
3074 =head2 Nested conditions, -and/-or prefixes
3076 So far, we've seen how multiple conditions are joined with a top-level
3077 C<AND>. We can change this by putting the different conditions we want in
3078 hashes and then putting those hashes in an array. For example:
3083 status => { -like => ['pending%', 'dispatched'] },
3087 status => 'unassigned',
3091 This data structure would create the following:
3093 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
3094 OR ( user = ? AND status = ? ) )";
3095 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
3098 Clauses in hashrefs or arrayrefs can be prefixed with an C<-and> or C<-or>
3099 to change the logic inside:
3105 -and => [ workhrs => {'>', 20}, geo => 'ASIA' ],
3106 -or => { workhrs => {'<', 50}, geo => 'EURO' },
3113 $stmt = "WHERE ( user = ?
3114 AND ( ( workhrs > ? AND geo = ? )
3115 OR ( workhrs < ? OR geo = ? ) ) )";
3116 @bind = ('nwiger', '20', 'ASIA', '50', 'EURO');
3118 =head3 Algebraic inconsistency, for historical reasons
3120 C<Important note>: when connecting several conditions, the C<-and->|C<-or>
3121 operator goes C<outside> of the nested structure; whereas when connecting
3122 several constraints on one column, the C<-and> operator goes
3123 C<inside> the arrayref. Here is an example combining both features:
3126 -and => [a => 1, b => 2],
3127 -or => [c => 3, d => 4],
3128 e => [-and => {-like => 'foo%'}, {-like => '%bar'} ]
3133 WHERE ( ( ( a = ? AND b = ? )
3134 OR ( c = ? OR d = ? )
3135 OR ( e LIKE ? AND e LIKE ? ) ) )
3137 This difference in syntax is unfortunate but must be preserved for
3138 historical reasons. So be careful: the two examples below would
3139 seem algebraically equivalent, but they are not
3142 { -like => 'foo%' },
3143 { -like => '%bar' },
3145 # yields: WHERE ( ( col LIKE ? AND col LIKE ? ) )
3148 { col => { -like => 'foo%' } },
3149 { col => { -like => '%bar' } },
3151 # yields: WHERE ( ( col LIKE ? OR col LIKE ? ) )
3154 =head2 Literal SQL and value type operators
3156 The basic premise of SQL::Abstract is that in WHERE specifications the "left
3157 side" is a column name and the "right side" is a value (normally rendered as
3158 a placeholder). This holds true for both hashrefs and arrayref pairs as you
3159 see in the L</WHERE CLAUSES> examples above. Sometimes it is necessary to
3160 alter this behavior. There are several ways of doing so.
3164 This is a virtual operator that signals the string to its right side is an
3165 identifier (a column name) and not a value. For example to compare two
3166 columns you would write:
3169 priority => { '<', 2 },
3170 requestor => { -ident => 'submitter' },
3175 $stmt = "WHERE priority < ? AND requestor = submitter";
3178 If you are maintaining legacy code you may see a different construct as
3179 described in L</Deprecated usage of Literal SQL>, please use C<-ident> in new
3184 This is a virtual operator that signals that the construct to its right side
3185 is a value to be passed to DBI. This is for example necessary when you want
3186 to write a where clause against an array (for RDBMS that support such
3187 datatypes). For example:
3190 array => { -value => [1, 2, 3] }
3195 $stmt = 'WHERE array = ?';
3196 @bind = ([1, 2, 3]);
3198 Note that if you were to simply say:
3204 the result would probably not be what you wanted:
3206 $stmt = 'WHERE array = ? OR array = ? OR array = ?';
3211 Finally, sometimes only literal SQL will do. To include a random snippet
3212 of SQL verbatim, you specify it as a scalar reference. Consider this only
3213 as a last resort. Usually there is a better way. For example:
3216 priority => { '<', 2 },
3217 requestor => { -in => \'(SELECT name FROM hitmen)' },
3222 $stmt = "WHERE priority < ? AND requestor IN (SELECT name FROM hitmen)"
3225 Note that in this example, you only get one bind parameter back, since
3226 the verbatim SQL is passed as part of the statement.
3230 Never use untrusted input as a literal SQL argument - this is a massive
3231 security risk (there is no way to check literal snippets for SQL
3232 injections and other nastyness). If you need to deal with untrusted input
3233 use literal SQL with placeholders as described next.
3235 =head3 Literal SQL with placeholders and bind values (subqueries)
3237 If the literal SQL to be inserted has placeholders and bind values,
3238 use a reference to an arrayref (yes this is a double reference --
3239 not so common, but perfectly legal Perl). For example, to find a date
3240 in Postgres you can use something like this:
3243 date_column => \[ "= date '2008-09-30' - ?::integer", 10 ]
3248 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
3251 Note that you must pass the bind values in the same format as they are returned
3252 by L<where|/where(\%where, $order)>. This means that if you set L</bindtype>
3253 to C<columns>, you must provide the bind values in the
3254 C<< [ column_meta => value ] >> format, where C<column_meta> is an opaque
3255 scalar value; most commonly the column name, but you can use any scalar value
3256 (including references and blessed references), L<SQL::Abstract> will simply
3257 pass it through intact. So if C<bindtype> is set to C<columns> the above
3258 example will look like:
3261 date_column => \[ "= date '2008-09-30' - ?::integer", [ {} => 10 ] ]
3264 Literal SQL is especially useful for nesting parenthesized clauses in the
3265 main SQL query. Here is a first example:
3267 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
3271 bar => \["IN ($sub_stmt)" => @sub_bind],
3276 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
3277 WHERE c2 < ? AND c3 LIKE ?))";
3278 @bind = (1234, 100, "foo%");
3280 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
3281 are expressed in the same way. Of course the C<$sub_stmt> and
3282 its associated bind values can be generated through a former call
3285 my ($sub_stmt, @sub_bind)
3286 = $sql->select("t1", "c1", {c2 => {"<" => 100},
3287 c3 => {-like => "foo%"}});
3290 bar => \["> ALL ($sub_stmt)" => @sub_bind],
3293 In the examples above, the subquery was used as an operator on a column;
3294 but the same principle also applies for a clause within the main C<%where>
3295 hash, like an EXISTS subquery:
3297 my ($sub_stmt, @sub_bind)
3298 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
3299 my %where = ( -and => [
3301 \["EXISTS ($sub_stmt)" => @sub_bind],
3306 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
3307 WHERE c1 = ? AND c2 > t0.c0))";
3311 Observe that the condition on C<c2> in the subquery refers to
3312 column C<t0.c0> of the main query: this is I<not> a bind
3313 value, so we have to express it through a scalar ref.
3314 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
3315 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
3316 what we wanted here.
3318 Finally, here is an example where a subquery is used
3319 for expressing unary negation:
3321 my ($sub_stmt, @sub_bind)
3322 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
3323 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
3325 lname => {like => '%son%'},
3326 \["NOT ($sub_stmt)" => @sub_bind],
3331 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
3332 @bind = ('%son%', 10, 20)
3334 =head3 Deprecated usage of Literal SQL
3336 Below are some examples of archaic use of literal SQL. It is shown only as
3337 reference for those who deal with legacy code. Each example has a much
3338 better, cleaner and safer alternative that users should opt for in new code.
3344 my %where = ( requestor => \'IS NOT NULL' )
3346 $stmt = "WHERE requestor IS NOT NULL"
3348 This used to be the way of generating NULL comparisons, before the handling
3349 of C<undef> got formalized. For new code please use the superior syntax as
3350 described in L</Tests for NULL values>.
3354 my %where = ( requestor => \'= submitter' )
3356 $stmt = "WHERE requestor = submitter"
3358 This used to be the only way to compare columns. Use the superior L</-ident>
3359 method for all new code. For example an identifier declared in such a way
3360 will be properly quoted if L</quote_char> is properly set, while the legacy
3361 form will remain as supplied.
3365 my %where = ( is_ready => \"", completed => { '>', '2012-12-21' } )
3367 $stmt = "WHERE completed > ? AND is_ready"
3368 @bind = ('2012-12-21')
3370 Using an empty string literal used to be the only way to express a boolean.
3371 For all new code please use the much more readable
3372 L<-bool|/Unary operators: bool> operator.
3378 These pages could go on for a while, since the nesting of the data
3379 structures this module can handle are pretty much unlimited (the
3380 module implements the C<WHERE> expansion as a recursive function
3381 internally). Your best bet is to "play around" with the module a
3382 little to see how the data structures behave, and choose the best
3383 format for your data based on that.
3385 And of course, all the values above will probably be replaced with
3386 variables gotten from forms or the command line. After all, if you
3387 knew everything ahead of time, you wouldn't have to worry about
3388 dynamically-generating SQL and could just hardwire it into your
3391 =head1 ORDER BY CLAUSES
3393 Some functions take an order by clause. This can either be a scalar (just a
3394 column name), a hashref of C<< { -desc => 'col' } >> or C<< { -asc => 'col' }
3395 >>, a scalarref, an arrayref-ref, or an arrayref of any of the previous
3398 Given | Will Generate
3399 ---------------------------------------------------------------
3401 'colA' | ORDER BY colA
3403 [qw/colA colB/] | ORDER BY colA, colB
3405 {-asc => 'colA'} | ORDER BY colA ASC
3407 {-desc => 'colB'} | ORDER BY colB DESC
3409 ['colA', {-asc => 'colB'}] | ORDER BY colA, colB ASC
3411 { -asc => [qw/colA colB/] } | ORDER BY colA ASC, colB ASC
3413 \'colA DESC' | ORDER BY colA DESC
3415 \[ 'FUNC(colA, ?)', $x ] | ORDER BY FUNC(colA, ?)
3416 | /* ...with $x bound to ? */
3419 { -asc => 'colA' }, | colA ASC,
3420 { -desc => [qw/colB/] }, | colB DESC,
3421 { -asc => [qw/colC colD/] },| colC ASC, colD ASC,
3422 \'colE DESC', | colE DESC,
3423 \[ 'FUNC(colF, ?)', $x ], | FUNC(colF, ?)
3424 ] | /* ...with $x bound to ? */
3425 ===============================================================
3429 =head1 SPECIAL OPERATORS
3431 my $sqlmaker = SQL::Abstract->new(special_ops => [
3435 my ($self, $field, $op, $arg) = @_;
3441 handler => 'method_name',
3445 A "special operator" is a SQL syntactic clause that can be
3446 applied to a field, instead of a usual binary operator.
3449 WHERE field IN (?, ?, ?)
3450 WHERE field BETWEEN ? AND ?
3451 WHERE MATCH(field) AGAINST (?, ?)
3453 Special operators IN and BETWEEN are fairly standard and therefore
3454 are builtin within C<SQL::Abstract> (as the overridable methods
3455 C<_where_field_IN> and C<_where_field_BETWEEN>). For other operators,
3456 like the MATCH .. AGAINST example above which is specific to MySQL,
3457 you can write your own operator handlers - supply a C<special_ops>
3458 argument to the C<new> method. That argument takes an arrayref of
3459 operator definitions; each operator definition is a hashref with two
3466 the regular expression to match the operator
3470 Either a coderef or a plain scalar method name. In both cases
3471 the expected return is C<< ($sql, @bind) >>.
3473 When supplied with a method name, it is simply called on the
3474 L<SQL::Abstract> object as:
3476 $self->$method_name($field, $op, $arg)
3480 $field is the LHS of the operator
3481 $op is the part that matched the handler regex
3484 When supplied with a coderef, it is called as:
3486 $coderef->($self, $field, $op, $arg)
3491 For example, here is an implementation
3492 of the MATCH .. AGAINST syntax for MySQL
3494 my $sqlmaker = SQL::Abstract->new(special_ops => [
3496 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
3497 {regex => qr/^match$/i,
3499 my ($self, $field, $op, $arg) = @_;
3500 $arg = [$arg] if not ref $arg;
3501 my $label = $self->_quote($field);
3502 my ($placeholder) = $self->_convert('?');
3503 my $placeholders = join ", ", (($placeholder) x @$arg);
3504 my $sql = $self->_sqlcase('match') . " ($label) "
3505 . $self->_sqlcase('against') . " ($placeholders) ";
3506 my @bind = $self->_bindtype($field, @$arg);
3507 return ($sql, @bind);
3514 =head1 UNARY OPERATORS
3516 my $sqlmaker = SQL::Abstract->new(unary_ops => [
3520 my ($self, $op, $arg) = @_;
3526 handler => 'method_name',
3530 A "unary operator" is a SQL syntactic clause that can be
3531 applied to a field - the operator goes before the field
3533 You can write your own operator handlers - supply a C<unary_ops>
3534 argument to the C<new> method. That argument takes an arrayref of
3535 operator definitions; each operator definition is a hashref with two
3542 the regular expression to match the operator
3546 Either a coderef or a plain scalar method name. In both cases
3547 the expected return is C<< $sql >>.
3549 When supplied with a method name, it is simply called on the
3550 L<SQL::Abstract> object as:
3552 $self->$method_name($op, $arg)
3556 $op is the part that matched the handler regex
3557 $arg is the RHS or argument of the operator
3559 When supplied with a coderef, it is called as:
3561 $coderef->($self, $op, $arg)
3569 Thanks to some benchmarking by Mark Stosberg, it turns out that
3570 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
3571 I must admit this wasn't an intentional design issue, but it's a
3572 byproduct of the fact that you get to control your C<DBI> handles
3575 To maximize performance, use a code snippet like the following:
3577 # prepare a statement handle using the first row
3578 # and then reuse it for the rest of the rows
3580 for my $href (@array_of_hashrefs) {
3581 $stmt ||= $sql->insert('table', $href);
3582 $sth ||= $dbh->prepare($stmt);
3583 $sth->execute($sql->values($href));
3586 The reason this works is because the keys in your C<$href> are sorted
3587 internally by B<SQL::Abstract>. Thus, as long as your data retains
3588 the same structure, you only have to generate the SQL the first time
3589 around. On subsequent queries, simply use the C<values> function provided
3590 by this module to return your values in the correct order.
3592 However this depends on the values having the same type - if, for
3593 example, the values of a where clause may either have values
3594 (resulting in sql of the form C<column = ?> with a single bind
3595 value), or alternatively the values might be C<undef> (resulting in
3596 sql of the form C<column IS NULL> with no bind value) then the
3597 caching technique suggested will not work.
3601 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
3602 really like this part (I do, at least). Building up a complex query
3603 can be as simple as the following:
3610 use CGI::FormBuilder;
3613 my $form = CGI::FormBuilder->new(...);
3614 my $sql = SQL::Abstract->new;
3616 if ($form->submitted) {
3617 my $field = $form->field;
3618 my $id = delete $field->{id};
3619 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
3622 Of course, you would still have to connect using C<DBI> to run the
3623 query, but the point is that if you make your form look like your
3624 table, the actual query script can be extremely simplistic.
3626 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
3627 a fast interface to returning and formatting data. I frequently
3628 use these three modules together to write complex database query
3629 apps in under 50 lines.
3631 =head1 HOW TO CONTRIBUTE
3633 Contributions are always welcome, in all usable forms (we especially
3634 welcome documentation improvements). The delivery methods include git-
3635 or unified-diff formatted patches, GitHub pull requests, or plain bug
3636 reports either via RT or the Mailing list. Contributors are generally
3637 granted full access to the official repository after their first several
3638 patches pass successful review.
3640 This project is maintained in a git repository. The code and related tools are
3641 accessible at the following locations:
3645 =item * Official repo: L<git://git.shadowcat.co.uk/dbsrgits/SQL-Abstract.git>
3647 =item * Official gitweb: L<http://git.shadowcat.co.uk/gitweb/gitweb.cgi?p=dbsrgits/SQL-Abstract.git>
3649 =item * GitHub mirror: L<https://github.com/dbsrgits/sql-abstract>
3651 =item * Authorized committers: L<ssh://dbsrgits@git.shadowcat.co.uk/SQL-Abstract.git>
3657 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
3658 Great care has been taken to preserve the I<published> behavior
3659 documented in previous versions in the 1.* family; however,
3660 some features that were previously undocumented, or behaved
3661 differently from the documentation, had to be changed in order
3662 to clarify the semantics. Hence, client code that was relying
3663 on some dark areas of C<SQL::Abstract> v1.*
3664 B<might behave differently> in v1.50.
3666 The main changes are:
3672 support for literal SQL through the C<< \ [ $sql, @bind ] >> syntax.
3676 support for the { operator => \"..." } construct (to embed literal SQL)
3680 support for the { operator => \["...", @bind] } construct (to embed literal SQL with bind values)
3684 optional support for L<array datatypes|/"Inserting and Updating Arrays">
3688 defensive programming: check arguments
3692 fixed bug with global logic, which was previously implemented
3693 through global variables yielding side-effects. Prior versions would
3694 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
3695 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
3696 Now this is interpreted
3697 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
3702 fixed semantics of _bindtype on array args
3706 dropped the C<_anoncopy> of the %where tree. No longer necessary,
3707 we just avoid shifting arrays within that tree.
3711 dropped the C<_modlogic> function
3715 =head1 ACKNOWLEDGEMENTS
3717 There are a number of individuals that have really helped out with
3718 this module. Unfortunately, most of them submitted bugs via CPAN
3719 so I have no idea who they are! But the people I do know are:
3721 Ash Berlin (order_by hash term support)
3722 Matt Trout (DBIx::Class support)
3723 Mark Stosberg (benchmarking)
3724 Chas Owens (initial "IN" operator support)
3725 Philip Collins (per-field SQL functions)
3726 Eric Kolve (hashref "AND" support)
3727 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
3728 Dan Kubb (support for "quote_char" and "name_sep")
3729 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
3730 Laurent Dami (internal refactoring, extensible list of special operators, literal SQL)
3731 Norbert Buchmuller (support for literal SQL in hashpair, misc. fixes & tests)
3732 Peter Rabbitson (rewrite of SQLA::Test, misc. fixes & tests)
3733 Oliver Charles (support for "RETURNING" after "INSERT")
3739 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
3743 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
3745 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
3747 For support, your best bet is to try the C<DBIx::Class> users mailing list.
3748 While not an official support venue, C<DBIx::Class> makes heavy use of
3749 C<SQL::Abstract>, and as such list members there are very familiar with
3750 how to create queries.
3754 This module is free software; you may copy this under the same
3755 terms as perl itself (either the GNU General Public License or
3756 the Artistic License)