1 package SQL::Abstract; # see doc at end of file
10 our @EXPORT_OK = qw(is_plain_value is_literal_value);
20 *SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION = $ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}
26 #======================================================================
28 #======================================================================
30 our $VERSION = '1.87';
32 # This would confuse some packagers
33 $VERSION = eval $VERSION if $VERSION =~ /_/; # numify for warning-free dev releases
37 # special operators (-in, -between). May be extended/overridden by user.
38 # See section WHERE: BUILTIN SPECIAL OPERATORS below for implementation
39 my @BUILTIN_SPECIAL_OPS = (
40 {regex => qr/^ (?: not \s )? between $/ix, handler => sub { die "NOPE" }},
41 {regex => qr/^ (?: not \s )? in $/ix, handler => sub { die "NOPE" }},
42 {regex => qr/^ is (?: \s+ not )? $/ix, handler => sub { die "NOPE" }},
45 #======================================================================
46 # DEBUGGING AND ERROR REPORTING
47 #======================================================================
50 return unless $_[0]->{debug}; shift; # a little faster
51 my $func = (caller(1))[3];
52 warn "[$func] ", @_, "\n";
56 my($func) = (caller(1))[3];
57 Carp::carp "[$func] Warning: ", @_;
61 my($func) = (caller(1))[3];
62 Carp::croak "[$func] Fatal: ", @_;
65 sub is_literal_value ($) {
66 ref $_[0] eq 'SCALAR' ? [ ${$_[0]} ]
67 : ( ref $_[0] eq 'REF' and ref ${$_[0]} eq 'ARRAY' ) ? [ @${ $_[0] } ]
71 # FIXME XSify - this can be done so much more efficiently
72 sub is_plain_value ($) {
74 ! length ref $_[0] ? \($_[0])
76 ref $_[0] eq 'HASH' and keys %{$_[0]} == 1
78 exists $_[0]->{-value}
79 ) ? \($_[0]->{-value})
81 # reuse @_ for even moar speedz
82 defined ( $_[1] = Scalar::Util::blessed $_[0] )
84 # deliberately not using Devel::OverloadInfo - the checks we are
85 # intersted in are much more limited than the fullblown thing, and
86 # this is a very hot piece of code
88 # simply using ->can('(""') can leave behind stub methods that
89 # break actually using the overload later (see L<perldiag/Stub
90 # found while resolving method "%s" overloading "%s" in package
91 # "%s"> and the source of overload::mycan())
93 # either has stringification which DBI SHOULD prefer out of the box
94 grep { *{ (qq[${_}::(""]) }{CODE} } @{ $_[2] = mro::get_linear_isa( $_[1] ) }
96 # has nummification or boolification, AND fallback is *not* disabled
98 SQL::Abstract::_ENV_::DETECT_AUTOGENERATED_STRINGIFICATION
101 grep { *{"${_}::(0+"}{CODE} } @{$_[2]}
103 grep { *{"${_}::(bool"}{CODE} } @{$_[2]}
107 # no fallback specified at all
108 ! ( ($_[3]) = grep { *{"${_}::()"}{CODE} } @{$_[2]} )
110 # fallback explicitly undef
111 ! defined ${"$_[3]::()"}
124 #======================================================================
126 #======================================================================
130 my $class = ref($self) || $self;
131 my %opt = (ref $_[0] eq 'HASH') ? %{$_[0]} : @_;
133 # choose our case by keeping an option around
134 delete $opt{case} if $opt{case} && $opt{case} ne 'lower';
136 # default logic for interpreting arrayrefs
137 $opt{logic} = $opt{logic} ? uc $opt{logic} : 'OR';
139 # how to return bind vars
140 $opt{bindtype} ||= 'normal';
142 # default comparison is "=", but can be overridden
145 # try to recognize which are the 'equality' and 'inequality' ops
146 # (temporary quickfix (in 2007), should go through a more seasoned API)
147 $opt{equality_op} = qr/^( \Q$opt{cmp}\E | \= )$/ix;
148 $opt{inequality_op} = qr/^( != | <> )$/ix;
150 $opt{like_op} = qr/^ (is\s+)? r?like $/xi;
151 $opt{not_like_op} = qr/^ (is\s+)? not \s+ r?like $/xi;
154 $opt{sqltrue} ||= '1=1';
155 $opt{sqlfalse} ||= '0=1';
158 $opt{user_special_ops} = [ @{$opt{special_ops} ||= []} ];
159 # regexes are applied in order, thus push after user-defines
160 push @{$opt{special_ops}}, @BUILTIN_SPECIAL_OPS;
163 $opt{unary_ops} ||= [];
165 # rudimentary sanity-check for user supplied bits treated as functions/operators
166 # If a purported function matches this regular expression, an exception is thrown.
167 # Literal SQL is *NOT* subject to this check, only functions (and column names
168 # when quoting is not in effect)
171 # need to guard against ()'s in column names too, but this will break tons of
172 # hacks... ideas anyone?
173 $opt{injection_guard} ||= qr/
179 return bless \%opt, $class;
182 sub sqltrue { +{ -literal => [ $_[0]->{sqltrue} ] } }
183 sub sqlfalse { +{ -literal => [ $_[0]->{sqlfalse} ] } }
185 sub _assert_pass_injection_guard {
186 if ($_[1] =~ $_[0]->{injection_guard}) {
187 my $class = ref $_[0];
188 puke "Possible SQL injection attempt '$_[1]'. If this is indeed a part of the "
189 . "desired SQL use literal SQL ( \'...' or \[ '...' ] ) or supply your own "
190 . "{injection_guard} attribute to ${class}->new()"
195 #======================================================================
197 #======================================================================
201 my $table = $self->_table(shift);
202 my $data = shift || return;
205 my $method = $self->_METHOD_FOR_refkind("_insert", $data);
206 my ($sql, @bind) = $self->$method($data);
207 $sql = join " ", $self->_sqlcase('insert into'), $table, $sql;
209 if ($options->{returning}) {
210 my ($s, @b) = $self->_insert_returning($options);
215 return wantarray ? ($sql, @bind) : $sql;
218 # So that subclasses can override INSERT ... RETURNING separately from
219 # UPDATE and DELETE (e.g. DBIx::Class::SQLMaker::Oracle does this)
220 sub _insert_returning { shift->_returning(@_) }
223 my ($self, $options) = @_;
225 my $f = $options->{returning};
227 my ($sql, @bind) = $self->_render_expr(
228 $self->_expand_maybe_list_expr($f, undef, -ident)
231 ? $self->_sqlcase(' returning ') . $sql
232 : ($self->_sqlcase(' returning ').$sql, @bind);
235 sub _insert_HASHREF { # explicit list of fields and then values
236 my ($self, $data) = @_;
238 my @fields = sort keys %$data;
240 my ($sql, @bind) = $self->_insert_values($data);
243 $_ = $self->_quote($_) foreach @fields;
244 $sql = "( ".join(", ", @fields).") ".$sql;
246 return ($sql, @bind);
249 sub _insert_ARRAYREF { # just generate values(?,?) part (no list of fields)
250 my ($self, $data) = @_;
252 # no names (arrayref) so can't generate bindtype
253 $self->{bindtype} ne 'columns'
254 or belch "can't do 'columns' bindtype when called with arrayref";
256 my (@values, @all_bind);
257 foreach my $value (@$data) {
258 my ($values, @bind) = $self->_insert_value(undef, $value);
259 push @values, $values;
260 push @all_bind, @bind;
262 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
263 return ($sql, @all_bind);
266 sub _insert_ARRAYREFREF { # literal SQL with bind
267 my ($self, $data) = @_;
269 my ($sql, @bind) = @${$data};
270 $self->_assert_bindval_matches_bindtype(@bind);
272 return ($sql, @bind);
276 sub _insert_SCALARREF { # literal SQL without bind
277 my ($self, $data) = @_;
283 my ($self, $data) = @_;
285 my (@values, @all_bind);
286 foreach my $column (sort keys %$data) {
287 my ($values, @bind) = $self->_insert_value($column, $data->{$column});
288 push @values, $values;
289 push @all_bind, @bind;
291 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
292 return ($sql, @all_bind);
296 my ($self, $column, $v) = @_;
298 my (@values, @all_bind);
299 $self->_SWITCH_refkind($v, {
302 if ($self->{array_datatypes}) { # if array datatype are activated
304 push @all_bind, $self->_bindtype($column, $v);
306 else { # else literal SQL with bind
307 my ($sql, @bind) = @$v;
308 $self->_assert_bindval_matches_bindtype(@bind);
310 push @all_bind, @bind;
314 ARRAYREFREF => sub { # literal SQL with bind
315 my ($sql, @bind) = @${$v};
316 $self->_assert_bindval_matches_bindtype(@bind);
318 push @all_bind, @bind;
321 # THINK: anything useful to do with a HASHREF ?
322 HASHREF => sub { # (nothing, but old SQLA passed it through)
323 #TODO in SQLA >= 2.0 it will die instead
324 belch "HASH ref as bind value in insert is not supported";
326 push @all_bind, $self->_bindtype($column, $v);
329 SCALARREF => sub { # literal SQL without bind
333 SCALAR_or_UNDEF => sub {
335 push @all_bind, $self->_bindtype($column, $v);
340 my $sql = join(", ", @values);
341 return ($sql, @all_bind);
346 #======================================================================
348 #======================================================================
353 my $table = $self->_table(shift);
354 my $data = shift || return;
358 # first build the 'SET' part of the sql statement
359 puke "Unsupported data type specified to \$sql->update"
360 unless ref $data eq 'HASH';
362 my ($sql, @all_bind) = $self->_update_set_values($data);
363 $sql = $self->_sqlcase('update ') . $table . $self->_sqlcase(' set ')
367 my($where_sql, @where_bind) = $self->where($where);
369 push @all_bind, @where_bind;
372 if ($options->{returning}) {
373 my ($returning_sql, @returning_bind) = $self->_update_returning($options);
374 $sql .= $returning_sql;
375 push @all_bind, @returning_bind;
378 return wantarray ? ($sql, @all_bind) : $sql;
381 sub _update_set_values {
382 my ($self, $data) = @_;
384 my (@set, @all_bind);
385 for my $k (sort keys %$data) {
388 my $label = $self->_quote($k);
390 $self->_SWITCH_refkind($v, {
392 if ($self->{array_datatypes}) { # array datatype
393 push @set, "$label = ?";
394 push @all_bind, $self->_bindtype($k, $v);
396 else { # literal SQL with bind
397 my ($sql, @bind) = @$v;
398 $self->_assert_bindval_matches_bindtype(@bind);
399 push @set, "$label = $sql";
400 push @all_bind, @bind;
403 ARRAYREFREF => sub { # literal SQL with bind
404 my ($sql, @bind) = @${$v};
405 $self->_assert_bindval_matches_bindtype(@bind);
406 push @set, "$label = $sql";
407 push @all_bind, @bind;
409 SCALARREF => sub { # literal SQL without bind
410 push @set, "$label = $$v";
413 my ($op, $arg, @rest) = %$v;
415 puke 'Operator calls in update must be in the form { -op => $arg }'
416 if (@rest or not $op =~ /^\-(.+)/);
418 local our $Cur_Col_Meta = $k;
419 my ($sql, @bind) = $self->_render_expr(
420 $self->_expand_expr_hashpair($op, $arg)
423 push @set, "$label = $sql";
424 push @all_bind, @bind;
426 SCALAR_or_UNDEF => sub {
427 push @set, "$label = ?";
428 push @all_bind, $self->_bindtype($k, $v);
434 my $sql = join ', ', @set;
436 return ($sql, @all_bind);
439 # So that subclasses can override UPDATE ... RETURNING separately from
441 sub _update_returning { shift->_returning(@_) }
445 #======================================================================
447 #======================================================================
452 my $table = $self->_table(shift);
453 my $fields = shift || '*';
457 my ($fields_sql, @bind) = $self->_select_fields($fields);
459 my ($where_sql, @where_bind) = $self->where($where, $order);
460 push @bind, @where_bind;
462 my $sql = join(' ', $self->_sqlcase('select'), $fields_sql,
463 $self->_sqlcase('from'), $table)
466 return wantarray ? ($sql, @bind) : $sql;
470 my ($self, $fields) = @_;
471 return $self->_render_expr(
472 $self->_expand_maybe_list_expr($fields, undef, '-ident')
476 #======================================================================
478 #======================================================================
483 my $table = $self->_table(shift);
487 my($where_sql, @bind) = $self->where($where);
488 my $sql = $self->_sqlcase('delete from ') . $table . $where_sql;
490 if ($options->{returning}) {
491 my ($returning_sql, @returning_bind) = $self->_delete_returning($options);
492 $sql .= $returning_sql;
493 push @bind, @returning_bind;
496 return wantarray ? ($sql, @bind) : $sql;
499 # So that subclasses can override DELETE ... RETURNING separately from
501 sub _delete_returning { shift->_returning(@_) }
505 #======================================================================
507 #======================================================================
511 # Finally, a separate routine just to handle WHERE clauses
513 my ($self, $where, $order) = @_;
515 local $self->{convert_where} = $self->{convert};
518 my ($sql, @bind) = defined($where)
519 ? $self->_recurse_where($where)
521 $sql = (defined $sql and length $sql) ? $self->_sqlcase(' where ') . "( $sql )" : '';
525 my ($order_sql, @order_bind) = $self->_order_by($order);
527 push @bind, @order_bind;
530 return wantarray ? ($sql, @bind) : $sql;
534 my ($self, $expr, $logic, $default_scalar_to) = @_;
535 local our $Default_Scalar_To = $default_scalar_to if $default_scalar_to;
536 return undef unless defined($expr);
537 if (ref($expr) eq 'HASH') {
538 if (keys %$expr > 1) {
542 map $self->_expand_expr_hashpair($_ => $expr->{$_}, $logic),
546 return unless %$expr;
547 return $self->_expand_expr_hashpair(%$expr, $logic);
549 if (ref($expr) eq 'ARRAY') {
550 my $logic = lc($logic || $self->{logic});
551 $logic eq 'and' or $logic eq 'or' or puke "unknown logic: $logic";
557 while (my ($el) = splice @expr, 0, 1) {
558 puke "Supplying an empty left hand side argument is not supported in array-pairs"
559 unless defined($el) and length($el);
560 my $elref = ref($el);
562 push(@res, $self->_expand_expr({ $el, shift(@expr) }));
563 } elsif ($elref eq 'ARRAY') {
564 push(@res, $self->_expand_expr($el)) if @$el;
565 } elsif (my $l = is_literal_value($el)) {
566 push @res, { -literal => $l };
567 } elsif ($elref eq 'HASH') {
568 push @res, $self->_expand_expr($el);
573 return { -op => [ $logic, @res ] };
575 if (my $literal = is_literal_value($expr)) {
576 return +{ -literal => $literal };
578 if (!ref($expr) or Scalar::Util::blessed($expr)) {
579 if (my $d = $Default_Scalar_To) {
580 return +{ $d => $expr };
582 if (my $m = our $Cur_Col_Meta) {
583 return +{ -bind => [ $m, $expr ] };
585 return +{ -value => $expr };
590 sub _expand_expr_hashpair {
591 my ($self, $k, $v, $logic) = @_;
592 unless (defined($k) and length($k)) {
593 if (defined($k) and my $literal = is_literal_value($v)) {
594 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
595 return { -literal => $literal };
597 puke "Supplying an empty left hand side argument is not supported";
600 $self->_assert_pass_injection_guard($k =~ /^-(.*)$/s);
601 if ($k =~ s/ [_\s]? \d+ $//x ) {
602 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
603 . "You probably wanted ...-and => [ $k => COND1, $k => COND2 ... ]";
606 return $self->_expand_expr($v);
610 return $self->_expand_expr($v);
612 puke "-bool => undef not supported" unless defined($v);
613 return { -ident => $v };
616 return { -op => [ 'not', $self->_expand_expr($v) ] };
618 if (my ($rest) = $k =~/^-not[_ ](.*)$/) {
621 $self->_expand_expr_hashpair("-${rest}", $v, $logic)
624 if (my ($logic) = $k =~ /^-(and|or)$/i) {
625 if (ref($v) eq 'HASH') {
626 return $self->_expand_expr($v, $logic);
628 if (ref($v) eq 'ARRAY') {
629 return $self->_expand_expr($v, $logic);
634 $op =~ s/^-// if length($op) > 1;
636 # top level special ops are illegal in general
637 puke "Illegal use of top-level '-$op'"
638 if List::Util::first { $op =~ $_->{regex} } @{$self->{special_ops}};
640 if ($k eq '-value' and my $m = our $Cur_Col_Meta) {
641 return +{ -bind => [ $m, $v ] };
643 if ($k eq '-op' or $k eq '-ident' or $k eq '-value' or $k eq '-bind' or $k eq '-literal' or $k eq '-func') {
646 if (my $custom = $self->{custom_expansions}{($k =~ /^-(.*)$/)[0]}) {
647 return $self->$custom($v);
652 and (keys %$v)[0] =~ /^-/
654 my ($func) = $k =~ /^-(.*)$/;
655 return +{ -func => [ $func, $self->_expand_expr($v) ] };
657 if (!ref($v) or is_literal_value($v)) {
658 return +{ -op => [ $k =~ /^-(.*)$/, $self->_expand_expr($v) ] };
665 and exists $v->{-value}
666 and not defined $v->{-value}
669 return $self->_expand_expr_hashpair($k => { $self->{cmp} => undef });
671 if (!ref($v) or Scalar::Util::blessed($v)) {
676 { -bind => [ $k, $v ] }
680 if (ref($v) eq 'HASH') {
684 map $self->_expand_expr_hashpair($k => { $_ => $v->{$_} }),
691 $self->_assert_pass_injection_guard($vk);
692 if ($vk =~ s/ [_\s]? \d+ $//x ) {
693 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
694 . "You probably wanted ...-and => [ -$vk => COND1, -$vk => COND2 ... ]";
696 if ($vk =~ /^(?:not[ _])?between$/) {
697 local our $Cur_Col_Meta = $k;
698 my @rhs = map $self->_expand_expr($_),
699 ref($vv) eq 'ARRAY' ? @$vv : $vv;
701 (@rhs == 1 and ref($rhs[0]) eq 'HASH' and $rhs[0]->{-literal})
703 (@rhs == 2 and defined($rhs[0]) and defined($rhs[1]))
705 puke "Operator '${\uc($vk)}' requires either an arrayref with two defined values or expressions, or a single literal scalarref/arrayref-ref";
708 join(' ', split '_', $vk),
713 if ($vk =~ /^(?:not[ _])?in$/) {
714 if (my $literal = is_literal_value($vv)) {
715 my ($sql, @bind) = @$literal;
716 my $opened_sql = $self->_open_outer_paren($sql);
718 $vk, { -ident => $k },
719 [ { -literal => [ $opened_sql, @bind ] } ]
723 'SQL::Abstract before v1.75 used to generate incorrect SQL when the '
724 . "-${\uc($vk)} operator was given an undef-containing list: !!!AUDIT YOUR CODE "
725 . 'AND DATA!!! (the upcoming Data::Query-based version of SQL::Abstract '
726 . 'will emit the logically correct SQL instead of raising this exception)'
728 puke("Argument passed to the '${\uc($vk)}' operator can not be undefined")
730 my @rhs = map $self->_expand_expr($_),
731 map { ref($_) ? $_ : { -bind => [ $k, $_ ] } }
732 map { defined($_) ? $_: puke($undef_err) }
733 (ref($vv) eq 'ARRAY' ? @$vv : $vv);
734 return $self->${\($vk =~ /^not/ ? 'sqltrue' : 'sqlfalse')} unless @rhs;
737 join(' ', split '_', $vk),
742 if ($vk eq 'ident') {
743 if (! defined $vv or ref $vv) {
744 puke "-$vk requires a single plain scalar argument (a quotable identifier)";
752 if ($vk eq 'value') {
753 return $self->_expand_expr_hashpair($k, undef) unless defined($vv);
757 { -bind => [ $k, $vv ] }
760 if ($vk =~ /^is(?:[ _]not)?$/) {
761 puke "$vk can only take undef as argument"
765 and exists($vv->{-value})
766 and !defined($vv->{-value})
769 return +{ -op => [ $vk.' null', { -ident => $k } ] };
771 if ($vk =~ /^(and|or)$/) {
772 if (ref($vv) eq 'HASH') {
775 map $self->_expand_expr_hashpair($k, { $_ => $vv->{$_} }),
780 if (my $us = List::Util::first { $vk =~ $_->{regex} } @{$self->{user_special_ops}}) {
781 return { -op => [ $vk, { -ident => $k }, $vv ] };
783 if (ref($vv) eq 'ARRAY') {
784 my ($logic, @values) = (
785 (defined($vv->[0]) and $vv->[0] =~ /^-(and|or)$/i)
790 $vk =~ $self->{inequality_op}
791 or join(' ', split '_', $vk) =~ $self->{not_like_op}
793 if (lc($logic) eq '-or' and @values > 1) {
794 my $op = uc join ' ', split '_', $vk;
795 belch "A multi-element arrayref as an argument to the inequality op '$op' "
796 . 'is technically equivalent to an always-true 1=1 (you probably wanted '
797 . "to say ...{ \$inequality_op => [ -and => \@values ] }... instead)"
802 # try to DWIM on equality operators
803 my $op = join ' ', split '_', $vk;
805 $op =~ $self->{equality_op} ? $self->sqlfalse
806 : $op =~ $self->{like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->sqlfalse
807 : $op =~ $self->{inequality_op} ? $self->sqltrue
808 : $op =~ $self->{not_like_op} ? belch("Supplying an empty arrayref to '@{[ uc $op]}' is deprecated") && $self->sqltrue
809 : puke "operator '$op' applied on an empty array (field '$k')";
813 map $self->_expand_expr_hashpair($k => { $vk => $_ }),
821 and exists $vv->{-value}
822 and not defined $vv->{-value}
825 my $op = join ' ', split '_', $vk;
827 $op =~ /^not$/i ? 'is not' # legacy
828 : $op =~ $self->{equality_op} ? 'is'
829 : $op =~ $self->{like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is'
830 : $op =~ $self->{inequality_op} ? 'is not'
831 : $op =~ $self->{not_like_op} ? belch("Supplying an undefined argument to '@{[ uc $op]}' is deprecated") && 'is not'
832 : puke "unexpected operator '$op' with undef operand";
833 return +{ -op => [ $is.' null', { -ident => $k } ] };
835 local our $Cur_Col_Meta = $k;
839 $self->_expand_expr($vv)
842 if (ref($v) eq 'ARRAY') {
843 return $self->sqlfalse unless @$v;
844 $self->_debug("ARRAY($k) means distribute over elements");
846 $v->[0] =~ /^-((?:and|or))$/i
847 ? ($v = [ @{$v}[1..$#$v] ], $1)
848 : ($self->{logic} || 'or')
852 map $self->_expand_expr({ $k => $_ }, $this_logic), @$v
855 if (my $literal = is_literal_value($v)) {
857 belch 'Hash-pairs consisting of an empty string with a literal are deprecated, and will be removed in 2.0: use -and => [ $literal ] instead';
860 my ($sql, @bind) = @$literal;
861 if ($self->{bindtype} eq 'columns') {
863 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
864 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
868 return +{ -literal => [ $self->_quote($k).' '.$sql, @bind ] };
874 my ($self, $expr) = @_;
875 my ($k, $v, @rest) = %$expr;
877 my %op = map +("-$_" => '_render_'.$_),
878 qw(op func value bind ident literal);
879 if (my $meth = $op{$k}) {
880 return $self->$meth($v);
882 die "notreached: $k";
886 my ($self, $where, $logic) = @_;
888 #print STDERR Data::Dumper::Concise::Dumper([ $where, $logic ]);
890 my $where_exp = $self->_expand_expr($where, $logic);
892 #print STDERR Data::Dumper::Concise::Dumper([ EXP => $where_exp ]);
894 # dispatch on appropriate method according to refkind of $where
895 # my $method = $self->_METHOD_FOR_refkind("_where", $where_exp);
897 # my ($sql, @bind) = $self->$method($where_exp, $logic);
899 my ($sql, @bind) = defined($where_exp) ? $self->_render_expr($where_exp) : (undef);
901 # DBIx::Class used to call _recurse_where in scalar context
902 # something else might too...
904 return ($sql, @bind);
907 belch "Calling _recurse_where in scalar context is deprecated and will go away before 2.0";
913 my ($self, $ident) = @_;
915 return $self->_convert($self->_quote($ident));
919 my ($self, $value) = @_;
921 return ($self->_convert('?'), $self->_bindtype(undef, $value));
924 my %unop_postfix = map +($_ => 1),
925 'is null', 'is not null',
933 my ($self, $args) = @_;
934 my ($left, $low, $high) = @$args;
935 my ($rhsql, @rhbind) = do {
937 puke "Single arg to between must be a literal"
938 unless $low->{-literal};
941 my ($l, $h) = map [ $self->_render_expr($_) ], $low, $high;
942 (join(' ', $l->[0], $self->_sqlcase('and'), $h->[0]),
943 @{$l}[1..$#$l], @{$h}[1..$#$h])
946 my ($lhsql, @lhbind) = $self->_render_expr($left);
948 join(' ', '(', $lhsql, $self->_sqlcase($op), $rhsql, ')'),
952 }), 'between', 'not between'),
956 my ($self, $args) = @_;
957 my ($lhs, $rhs) = @$args;
960 my ($sql, @bind) = $self->_render_expr($_);
961 push @in_bind, @bind;
964 my ($lhsql, @lbind) = $self->_render_expr($lhs);
966 $lhsql.' '.$self->_sqlcase($op).' ( '
977 my ($op, @args) = @$v;
978 $op =~ s/^-// if length($op) > 1;
980 if (my $h = $special{$op}) {
981 return $self->$h(\@args);
983 if (my $us = List::Util::first { $op =~ $_->{regex} } @{$self->{user_special_ops}}) {
984 puke "Special op '${op}' requires first value to be identifier"
985 unless my ($k) = map $_->{-ident}, grep ref($_) eq 'HASH', $args[0];
986 return $self->${\($us->{handler})}($k, $op, $args[1]);
988 my $final_op = $op =~ /^(?:is|not)_/ ? join(' ', split '_', $op) : $op;
989 if (@args == 1 and $op !~ /^(and|or)$/) {
990 my ($expr_sql, @bind) = $self->_render_expr($args[0]);
991 my $op_sql = $self->_sqlcase($final_op);
993 $unop_postfix{lc($final_op)}
994 ? "${expr_sql} ${op_sql}"
995 : "${op_sql} ${expr_sql}"
997 return (($op eq 'not' ? '('.$final_sql.')' : $final_sql), @bind);
999 my @parts = map [ $self->_render_expr($_) ], @args;
1000 my ($final_sql) = map +($op =~ /^(and|or)$/ ? "(${_})" : $_), join(
1001 ($final_op eq ',' ? '' : ' ').$self->_sqlcase($final_op).' ',
1006 map @{$_}[1..$#$_], @parts
1013 my ($self, $rest) = @_;
1014 my ($func, @args) = @$rest;
1018 push @arg_sql, shift @x;
1020 } map [ $self->_render_expr($_) ], @args;
1021 return ($self->_sqlcase($func).'('.join(', ', @arg_sql).')', @bind);
1025 my ($self, $bind) = @_;
1026 return ($self->_convert('?'), $self->_bindtype(@$bind));
1029 sub _render_literal {
1030 my ($self, $literal) = @_;
1031 $self->_assert_bindval_matches_bindtype(@{$literal}[1..$#$literal]);
1035 # Some databases (SQLite) treat col IN (1, 2) different from
1036 # col IN ( (1, 2) ). Use this to strip all outer parens while
1037 # adding them back in the corresponding method
1038 sub _open_outer_paren {
1039 my ($self, $sql) = @_;
1041 while (my ($inner) = $sql =~ /^ \s* \( (.*) \) \s* $/xs) {
1043 # there are closing parens inside, need the heavy duty machinery
1044 # to reevaluate the extraction starting from $sql (full reevaluation)
1045 if ($inner =~ /\)/) {
1046 require Text::Balanced;
1048 my (undef, $remainder) = do {
1049 # idiotic design - writes to $@ but *DOES NOT* throw exceptions
1051 Text::Balanced::extract_bracketed($sql, '()', qr/\s*/);
1054 # the entire expression needs to be a balanced bracketed thing
1055 # (after an extract no remainder sans trailing space)
1056 last if defined $remainder and $remainder =~ /\S/;
1066 #======================================================================
1068 #======================================================================
1071 my ($self, $arg) = @_;
1073 return '' unless defined($arg) and not (ref($arg) eq 'ARRAY' and !@$arg);
1075 my $expander = sub {
1076 my ($self, $dir, $expr) = @_;
1077 my @exp = map +(defined($dir) ? { -op => [ $dir => $_ ] } : $_),
1078 map $self->_expand_expr($_, undef, -ident),
1079 ref($expr) eq 'ARRAY' ? @$expr : $expr;
1080 return (@exp > 1 ? { -op => [ ',', @exp ] } : $exp[0]);
1083 local $self->{custom_expansions} = {
1084 asc => sub { shift->$expander(asc => @_) },
1085 desc => sub { shift->$expander(desc => @_) },
1088 my $expanded = $self->$expander(undef, $arg);
1090 my ($sql, @bind) = $self->_render_expr($expanded);
1092 my $final_sql = $self->_sqlcase(' order by ').$sql;
1094 return wantarray ? ($final_sql, @bind) : $final_sql;
1097 #======================================================================
1098 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
1099 #======================================================================
1104 ($self->_render_expr(
1105 $self->_expand_maybe_list_expr($from, undef, -ident)
1110 #======================================================================
1112 #======================================================================
1114 sub _expand_maybe_list_expr {
1115 my ($self, $expr, $logic, $default) = @_;
1117 if (ref($expr) eq 'ARRAY') {
1119 ',', map $self->_expand_expr($_, $logic, $default), @$expr
1126 return $self->_expand_expr($e, $logic, $default);
1129 # highly optimized, as it's called way too often
1131 # my ($self, $label) = @_;
1133 return '' unless defined $_[1];
1134 return ${$_[1]} if ref($_[1]) eq 'SCALAR';
1136 $_[0]->{quote_char} or
1137 ($_[0]->_assert_pass_injection_guard($_[1]), return $_[1]);
1139 my $qref = ref $_[0]->{quote_char};
1141 !$qref ? ($_[0]->{quote_char}, $_[0]->{quote_char})
1142 : ($qref eq 'ARRAY') ? @{$_[0]->{quote_char}}
1143 : puke "Unsupported quote_char format: $_[0]->{quote_char}";
1145 my $esc = $_[0]->{escape_char} || $r;
1147 # parts containing * are naturally unquoted
1148 return join($_[0]->{name_sep}||'', map
1149 +( $_ eq '*' ? $_ : do { (my $n = $_) =~ s/(\Q$esc\E|\Q$r\E)/$esc$1/g; $l . $n . $r } ),
1150 ( $_[0]->{name_sep} ? split (/\Q$_[0]->{name_sep}\E/, $_[1] ) : $_[1] )
1155 # Conversion, if applicable
1157 #my ($self, $arg) = @_;
1158 if ($_[0]->{convert_where}) {
1159 return $_[0]->_sqlcase($_[0]->{convert_where}) .'(' . $_[1] . ')';
1166 #my ($self, $col, @vals) = @_;
1167 # called often - tighten code
1168 return $_[0]->{bindtype} eq 'columns'
1169 ? map {[$_[1], $_]} @_[2 .. $#_]
1174 # Dies if any element of @bind is not in [colname => value] format
1175 # if bindtype is 'columns'.
1176 sub _assert_bindval_matches_bindtype {
1177 # my ($self, @bind) = @_;
1179 if ($self->{bindtype} eq 'columns') {
1181 if (!defined $_ || ref($_) ne 'ARRAY' || @$_ != 2) {
1182 puke "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
1188 sub _join_sql_clauses {
1189 my ($self, $logic, $clauses_aref, $bind_aref) = @_;
1191 if (@$clauses_aref > 1) {
1192 my $join = " " . $self->_sqlcase($logic) . " ";
1193 my $sql = '( ' . join($join, @$clauses_aref) . ' )';
1194 return ($sql, @$bind_aref);
1196 elsif (@$clauses_aref) {
1197 return ($clauses_aref->[0], @$bind_aref); # no parentheses
1200 return (); # if no SQL, ignore @$bind_aref
1205 # Fix SQL case, if so requested
1207 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
1208 # don't touch the argument ... crooked logic, but let's not change it!
1209 return $_[0]->{case} ? $_[1] : uc($_[1]);
1213 #======================================================================
1214 # DISPATCHING FROM REFKIND
1215 #======================================================================
1218 my ($self, $data) = @_;
1220 return 'UNDEF' unless defined $data;
1222 # blessed objects are treated like scalars
1223 my $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1225 return 'SCALAR' unless $ref;
1228 while ($ref eq 'REF') {
1230 $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1234 return ($ref||'SCALAR') . ('REF' x $n_steps);
1238 my ($self, $data) = @_;
1239 my @try = ($self->_refkind($data));
1240 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
1241 push @try, 'FALLBACK';
1245 sub _METHOD_FOR_refkind {
1246 my ($self, $meth_prefix, $data) = @_;
1249 for (@{$self->_try_refkind($data)}) {
1250 $method = $self->can($meth_prefix."_".$_)
1254 return $method || puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
1258 sub _SWITCH_refkind {
1259 my ($self, $data, $dispatch_table) = @_;
1262 for (@{$self->_try_refkind($data)}) {
1263 $coderef = $dispatch_table->{$_}
1267 puke "no dispatch entry for ".$self->_refkind($data)
1276 #======================================================================
1277 # VALUES, GENERATE, AUTOLOAD
1278 #======================================================================
1280 # LDNOTE: original code from nwiger, didn't touch code in that section
1281 # I feel the AUTOLOAD stuff should not be the default, it should
1282 # only be activated on explicit demand by user.
1286 my $data = shift || return;
1287 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
1288 unless ref $data eq 'HASH';
1291 foreach my $k (sort keys %$data) {
1292 my $v = $data->{$k};
1293 $self->_SWITCH_refkind($v, {
1295 if ($self->{array_datatypes}) { # array datatype
1296 push @all_bind, $self->_bindtype($k, $v);
1298 else { # literal SQL with bind
1299 my ($sql, @bind) = @$v;
1300 $self->_assert_bindval_matches_bindtype(@bind);
1301 push @all_bind, @bind;
1304 ARRAYREFREF => sub { # literal SQL with bind
1305 my ($sql, @bind) = @${$v};
1306 $self->_assert_bindval_matches_bindtype(@bind);
1307 push @all_bind, @bind;
1309 SCALARREF => sub { # literal SQL without bind
1311 SCALAR_or_UNDEF => sub {
1312 push @all_bind, $self->_bindtype($k, $v);
1323 my(@sql, @sqlq, @sqlv);
1327 if ($ref eq 'HASH') {
1328 for my $k (sort keys %$_) {
1331 my $label = $self->_quote($k);
1332 if ($r eq 'ARRAY') {
1333 # literal SQL with bind
1334 my ($sql, @bind) = @$v;
1335 $self->_assert_bindval_matches_bindtype(@bind);
1336 push @sqlq, "$label = $sql";
1338 } elsif ($r eq 'SCALAR') {
1339 # literal SQL without bind
1340 push @sqlq, "$label = $$v";
1342 push @sqlq, "$label = ?";
1343 push @sqlv, $self->_bindtype($k, $v);
1346 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
1347 } elsif ($ref eq 'ARRAY') {
1348 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
1351 if ($r eq 'ARRAY') { # literal SQL with bind
1352 my ($sql, @bind) = @$v;
1353 $self->_assert_bindval_matches_bindtype(@bind);
1356 } elsif ($r eq 'SCALAR') { # literal SQL without bind
1357 # embedded literal SQL
1364 push @sql, '(' . join(', ', @sqlq) . ')';
1365 } elsif ($ref eq 'SCALAR') {
1369 # strings get case twiddled
1370 push @sql, $self->_sqlcase($_);
1374 my $sql = join ' ', @sql;
1376 # this is pretty tricky
1377 # if ask for an array, return ($stmt, @bind)
1378 # otherwise, s/?/shift @sqlv/ to put it inline
1380 return ($sql, @sqlv);
1382 1 while $sql =~ s/\?/my $d = shift(@sqlv);
1383 ref $d ? $d->[1] : $d/e;
1392 # This allows us to check for a local, then _form, attr
1394 my($name) = $AUTOLOAD =~ /.*::(.+)/;
1395 return $self->generate($name, @_);
1406 SQL::Abstract - Generate SQL from Perl data structures
1412 my $sql = SQL::Abstract->new;
1414 my($stmt, @bind) = $sql->select($source, \@fields, \%where, $order);
1416 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
1418 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
1420 my($stmt, @bind) = $sql->delete($table, \%where);
1422 # Then, use these in your DBI statements
1423 my $sth = $dbh->prepare($stmt);
1424 $sth->execute(@bind);
1426 # Just generate the WHERE clause
1427 my($stmt, @bind) = $sql->where(\%where, $order);
1429 # Return values in the same order, for hashed queries
1430 # See PERFORMANCE section for more details
1431 my @bind = $sql->values(\%fieldvals);
1435 This module was inspired by the excellent L<DBIx::Abstract>.
1436 However, in using that module I found that what I really wanted
1437 to do was generate SQL, but still retain complete control over my
1438 statement handles and use the DBI interface. So, I set out to
1439 create an abstract SQL generation module.
1441 While based on the concepts used by L<DBIx::Abstract>, there are
1442 several important differences, especially when it comes to WHERE
1443 clauses. I have modified the concepts used to make the SQL easier
1444 to generate from Perl data structures and, IMO, more intuitive.
1445 The underlying idea is for this module to do what you mean, based
1446 on the data structures you provide it. The big advantage is that
1447 you don't have to modify your code every time your data changes,
1448 as this module figures it out.
1450 To begin with, an SQL INSERT is as easy as just specifying a hash
1451 of C<key=value> pairs:
1454 name => 'Jimbo Bobson',
1455 phone => '123-456-7890',
1456 address => '42 Sister Lane',
1457 city => 'St. Louis',
1458 state => 'Louisiana',
1461 The SQL can then be generated with this:
1463 my($stmt, @bind) = $sql->insert('people', \%data);
1465 Which would give you something like this:
1467 $stmt = "INSERT INTO people
1468 (address, city, name, phone, state)
1469 VALUES (?, ?, ?, ?, ?)";
1470 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
1471 '123-456-7890', 'Louisiana');
1473 These are then used directly in your DBI code:
1475 my $sth = $dbh->prepare($stmt);
1476 $sth->execute(@bind);
1478 =head2 Inserting and Updating Arrays
1480 If your database has array types (like for example Postgres),
1481 activate the special option C<< array_datatypes => 1 >>
1482 when creating the C<SQL::Abstract> object.
1483 Then you may use an arrayref to insert and update database array types:
1485 my $sql = SQL::Abstract->new(array_datatypes => 1);
1487 planets => [qw/Mercury Venus Earth Mars/]
1490 my($stmt, @bind) = $sql->insert('solar_system', \%data);
1494 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
1496 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
1499 =head2 Inserting and Updating SQL
1501 In order to apply SQL functions to elements of your C<%data> you may
1502 specify a reference to an arrayref for the given hash value. For example,
1503 if you need to execute the Oracle C<to_date> function on a value, you can
1504 say something like this:
1508 date_entered => \[ "to_date(?,'MM/DD/YYYY')", "03/02/2003" ],
1511 The first value in the array is the actual SQL. Any other values are
1512 optional and would be included in the bind values array. This gives
1515 my($stmt, @bind) = $sql->insert('people', \%data);
1517 $stmt = "INSERT INTO people (name, date_entered)
1518 VALUES (?, to_date(?,'MM/DD/YYYY'))";
1519 @bind = ('Bill', '03/02/2003');
1521 An UPDATE is just as easy, all you change is the name of the function:
1523 my($stmt, @bind) = $sql->update('people', \%data);
1525 Notice that your C<%data> isn't touched; the module will generate
1526 the appropriately quirky SQL for you automatically. Usually you'll
1527 want to specify a WHERE clause for your UPDATE, though, which is
1528 where handling C<%where> hashes comes in handy...
1530 =head2 Complex where statements
1532 This module can generate pretty complicated WHERE statements
1533 easily. For example, simple C<key=value> pairs are taken to mean
1534 equality, and if you want to see if a field is within a set
1535 of values, you can use an arrayref. Let's say we wanted to
1536 SELECT some data based on this criteria:
1539 requestor => 'inna',
1540 worker => ['nwiger', 'rcwe', 'sfz'],
1541 status => { '!=', 'completed' }
1544 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
1546 The above would give you something like this:
1548 $stmt = "SELECT * FROM tickets WHERE
1549 ( requestor = ? ) AND ( status != ? )
1550 AND ( worker = ? OR worker = ? OR worker = ? )";
1551 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
1553 Which you could then use in DBI code like so:
1555 my $sth = $dbh->prepare($stmt);
1556 $sth->execute(@bind);
1562 The methods are simple. There's one for every major SQL operation,
1563 and a constructor you use first. The arguments are specified in a
1564 similar order for each method (table, then fields, then a where
1565 clause) to try and simplify things.
1567 =head2 new(option => 'value')
1569 The C<new()> function takes a list of options and values, and returns
1570 a new B<SQL::Abstract> object which can then be used to generate SQL
1571 through the methods below. The options accepted are:
1577 If set to 'lower', then SQL will be generated in all lowercase. By
1578 default SQL is generated in "textbook" case meaning something like:
1580 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
1582 Any setting other than 'lower' is ignored.
1586 This determines what the default comparison operator is. By default
1587 it is C<=>, meaning that a hash like this:
1589 %where = (name => 'nwiger', email => 'nate@wiger.org');
1591 Will generate SQL like this:
1593 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
1595 However, you may want loose comparisons by default, so if you set
1596 C<cmp> to C<like> you would get SQL such as:
1598 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
1600 You can also override the comparison on an individual basis - see
1601 the huge section on L</"WHERE CLAUSES"> at the bottom.
1603 =item sqltrue, sqlfalse
1605 Expressions for inserting boolean values within SQL statements.
1606 By default these are C<1=1> and C<1=0>. They are used
1607 by the special operators C<-in> and C<-not_in> for generating
1608 correct SQL even when the argument is an empty array (see below).
1612 This determines the default logical operator for multiple WHERE
1613 statements in arrays or hashes. If absent, the default logic is "or"
1614 for arrays, and "and" for hashes. This means that a WHERE
1618 event_date => {'>=', '2/13/99'},
1619 event_date => {'<=', '4/24/03'},
1622 will generate SQL like this:
1624 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
1626 This is probably not what you want given this query, though (look
1627 at the dates). To change the "OR" to an "AND", simply specify:
1629 my $sql = SQL::Abstract->new(logic => 'and');
1631 Which will change the above C<WHERE> to:
1633 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
1635 The logic can also be changed locally by inserting
1636 a modifier in front of an arrayref:
1638 @where = (-and => [event_date => {'>=', '2/13/99'},
1639 event_date => {'<=', '4/24/03'} ]);
1641 See the L</"WHERE CLAUSES"> section for explanations.
1645 This will automatically convert comparisons using the specified SQL
1646 function for both column and value. This is mostly used with an argument
1647 of C<upper> or C<lower>, so that the SQL will have the effect of
1648 case-insensitive "searches". For example, this:
1650 $sql = SQL::Abstract->new(convert => 'upper');
1651 %where = (keywords => 'MaKe iT CAse inSeNSItive');
1653 Will turn out the following SQL:
1655 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
1657 The conversion can be C<upper()>, C<lower()>, or any other SQL function
1658 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
1659 not validate this option; it will just pass through what you specify verbatim).
1663 This is a kludge because many databases suck. For example, you can't
1664 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
1665 Instead, you have to use C<bind_param()>:
1667 $sth->bind_param(1, 'reg data');
1668 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
1670 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
1671 which loses track of which field each slot refers to. Fear not.
1673 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
1674 Currently, you can specify either C<normal> (default) or C<columns>. If you
1675 specify C<columns>, you will get an array that looks like this:
1677 my $sql = SQL::Abstract->new(bindtype => 'columns');
1678 my($stmt, @bind) = $sql->insert(...);
1681 [ 'column1', 'value1' ],
1682 [ 'column2', 'value2' ],
1683 [ 'column3', 'value3' ],
1686 You can then iterate through this manually, using DBI's C<bind_param()>.
1688 $sth->prepare($stmt);
1691 my($col, $data) = @$_;
1692 if ($col eq 'details' || $col eq 'comments') {
1693 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
1694 } elsif ($col eq 'image') {
1695 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
1697 $sth->bind_param($i, $data);
1701 $sth->execute; # execute without @bind now
1703 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
1704 Basically, the advantage is still that you don't have to care which fields
1705 are or are not included. You could wrap that above C<for> loop in a simple
1706 sub called C<bind_fields()> or something and reuse it repeatedly. You still
1707 get a layer of abstraction over manual SQL specification.
1709 Note that if you set L</bindtype> to C<columns>, the C<\[ $sql, @bind ]>
1710 construct (see L</Literal SQL with placeholders and bind values (subqueries)>)
1711 will expect the bind values in this format.
1715 This is the character that a table or column name will be quoted
1716 with. By default this is an empty string, but you could set it to
1717 the character C<`>, to generate SQL like this:
1719 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
1721 Alternatively, you can supply an array ref of two items, the first being the left
1722 hand quote character, and the second the right hand quote character. For
1723 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
1724 that generates SQL like this:
1726 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
1728 Quoting is useful if you have tables or columns names that are reserved
1729 words in your database's SQL dialect.
1733 This is the character that will be used to escape L</quote_char>s appearing
1734 in an identifier before it has been quoted.
1736 The parameter default in case of a single L</quote_char> character is the quote
1739 When opening-closing-style quoting is used (L</quote_char> is an arrayref)
1740 this parameter defaults to the B<closing (right)> L</quote_char>. Occurrences
1741 of the B<opening (left)> L</quote_char> within the identifier are currently left
1742 untouched. The default for opening-closing-style quotes may change in future
1743 versions, thus you are B<strongly encouraged> to specify the escape character
1748 This is the character that separates a table and column name. It is
1749 necessary to specify this when the C<quote_char> option is selected,
1750 so that tables and column names can be individually quoted like this:
1752 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
1754 =item injection_guard
1756 A regular expression C<qr/.../> that is applied to any C<-function> and unquoted
1757 column name specified in a query structure. This is a safety mechanism to avoid
1758 injection attacks when mishandling user input e.g.:
1760 my %condition_as_column_value_pairs = get_values_from_user();
1761 $sqla->select( ... , \%condition_as_column_value_pairs );
1763 If the expression matches an exception is thrown. Note that literal SQL
1764 supplied via C<\'...'> or C<\['...']> is B<not> checked in any way.
1766 Defaults to checking for C<;> and the C<GO> keyword (TransactSQL)
1768 =item array_datatypes
1770 When this option is true, arrayrefs in INSERT or UPDATE are
1771 interpreted as array datatypes and are passed directly
1773 When this option is false, arrayrefs are interpreted
1774 as literal SQL, just like refs to arrayrefs
1775 (but this behavior is for backwards compatibility; when writing
1776 new queries, use the "reference to arrayref" syntax
1782 Takes a reference to a list of "special operators"
1783 to extend the syntax understood by L<SQL::Abstract>.
1784 See section L</"SPECIAL OPERATORS"> for details.
1788 Takes a reference to a list of "unary operators"
1789 to extend the syntax understood by L<SQL::Abstract>.
1790 See section L</"UNARY OPERATORS"> for details.
1796 =head2 insert($table, \@values || \%fieldvals, \%options)
1798 This is the simplest function. You simply give it a table name
1799 and either an arrayref of values or hashref of field/value pairs.
1800 It returns an SQL INSERT statement and a list of bind values.
1801 See the sections on L</"Inserting and Updating Arrays"> and
1802 L</"Inserting and Updating SQL"> for information on how to insert
1803 with those data types.
1805 The optional C<\%options> hash reference may contain additional
1806 options to generate the insert SQL. Currently supported options
1813 Takes either a scalar of raw SQL fields, or an array reference of
1814 field names, and adds on an SQL C<RETURNING> statement at the end.
1815 This allows you to return data generated by the insert statement
1816 (such as row IDs) without performing another C<SELECT> statement.
1817 Note, however, this is not part of the SQL standard and may not
1818 be supported by all database engines.
1822 =head2 update($table, \%fieldvals, \%where, \%options)
1824 This takes a table, hashref of field/value pairs, and an optional
1825 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
1827 See the sections on L</"Inserting and Updating Arrays"> and
1828 L</"Inserting and Updating SQL"> for information on how to insert
1829 with those data types.
1831 The optional C<\%options> hash reference may contain additional
1832 options to generate the update SQL. Currently supported options
1839 See the C<returning> option to
1840 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
1844 =head2 select($source, $fields, $where, $order)
1846 This returns a SQL SELECT statement and associated list of bind values, as
1847 specified by the arguments:
1853 Specification of the 'FROM' part of the statement.
1854 The argument can be either a plain scalar (interpreted as a table
1855 name, will be quoted), or an arrayref (interpreted as a list
1856 of table names, joined by commas, quoted), or a scalarref
1857 (literal SQL, not quoted).
1861 Specification of the list of fields to retrieve from
1863 The argument can be either an arrayref (interpreted as a list
1864 of field names, will be joined by commas and quoted), or a
1865 plain scalar (literal SQL, not quoted).
1866 Please observe that this API is not as flexible as that of
1867 the first argument C<$source>, for backwards compatibility reasons.
1871 Optional argument to specify the WHERE part of the query.
1872 The argument is most often a hashref, but can also be
1873 an arrayref or plain scalar --
1874 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
1878 Optional argument to specify the ORDER BY part of the query.
1879 The argument can be a scalar, a hashref or an arrayref
1880 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
1886 =head2 delete($table, \%where, \%options)
1888 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
1889 It returns an SQL DELETE statement and list of bind values.
1891 The optional C<\%options> hash reference may contain additional
1892 options to generate the delete SQL. Currently supported options
1899 See the C<returning> option to
1900 L<insert|/insert($table, \@values || \%fieldvals, \%options)>.
1904 =head2 where(\%where, $order)
1906 This is used to generate just the WHERE clause. For example,
1907 if you have an arbitrary data structure and know what the
1908 rest of your SQL is going to look like, but want an easy way
1909 to produce a WHERE clause, use this. It returns an SQL WHERE
1910 clause and list of bind values.
1913 =head2 values(\%data)
1915 This just returns the values from the hash C<%data>, in the same
1916 order that would be returned from any of the other above queries.
1917 Using this allows you to markedly speed up your queries if you
1918 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
1920 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
1922 Warning: This is an experimental method and subject to change.
1924 This returns arbitrarily generated SQL. It's a really basic shortcut.
1925 It will return two different things, depending on return context:
1927 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
1928 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
1930 These would return the following:
1932 # First calling form
1933 $stmt = "CREATE TABLE test (?, ?)";
1934 @bind = (field1, field2);
1936 # Second calling form
1937 $stmt_and_val = "CREATE TABLE test (field1, field2)";
1939 Depending on what you're trying to do, it's up to you to choose the correct
1940 format. In this example, the second form is what you would want.
1944 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
1948 ALTER SESSION SET nls_date_format = 'MM/YY'
1950 You get the idea. Strings get their case twiddled, but everything
1951 else remains verbatim.
1953 =head1 EXPORTABLE FUNCTIONS
1955 =head2 is_plain_value
1957 Determines if the supplied argument is a plain value as understood by this
1962 =item * The value is C<undef>
1964 =item * The value is a non-reference
1966 =item * The value is an object with stringification overloading
1968 =item * The value is of the form C<< { -value => $anything } >>
1972 On failure returns C<undef>, on success returns a B<scalar> reference
1973 to the original supplied argument.
1979 The stringification overloading detection is rather advanced: it takes
1980 into consideration not only the presence of a C<""> overload, but if that
1981 fails also checks for enabled
1982 L<autogenerated versions of C<"">|overload/Magic Autogeneration>, based
1983 on either C<0+> or C<bool>.
1985 Unfortunately testing in the field indicates that this
1986 detection B<< may tickle a latent bug in perl versions before 5.018 >>,
1987 but only when very large numbers of stringifying objects are involved.
1988 At the time of writing ( Sep 2014 ) there is no clear explanation of
1989 the direct cause, nor is there a manageably small test case that reliably
1990 reproduces the problem.
1992 If you encounter any of the following exceptions in B<random places within
1993 your application stack> - this module may be to blame:
1995 Operation "ne": no method found,
1996 left argument in overloaded package <something>,
1997 right argument in overloaded package <something>
2001 Stub found while resolving method "???" overloading """" in package <something>
2003 If you fall victim to the above - please attempt to reduce the problem
2004 to something that could be sent to the L<SQL::Abstract developers
2005 |DBIx::Class/GETTING HELP/SUPPORT>
2006 (either publicly or privately). As a workaround in the meantime you can
2007 set C<$ENV{SQLA_ISVALUE_IGNORE_AUTOGENERATED_STRINGIFICATION}> to a true
2008 value, which will most likely eliminate your problem (at the expense of
2009 not being able to properly detect exotic forms of stringification).
2011 This notice and environment variable will be removed in a future version,
2012 as soon as the underlying problem is found and a reliable workaround is
2017 =head2 is_literal_value
2019 Determines if the supplied argument is a literal value as understood by this
2024 =item * C<\$sql_string>
2026 =item * C<\[ $sql_string, @bind_values ]>
2030 On failure returns C<undef>, on success returns an B<array> reference
2031 containing the unpacked version of the supplied literal SQL and bind values.
2033 =head1 WHERE CLAUSES
2037 This module uses a variation on the idea from L<DBIx::Abstract>. It
2038 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
2039 module is that things in arrays are OR'ed, and things in hashes
2042 The easiest way to explain is to show lots of examples. After
2043 each C<%where> hash shown, it is assumed you used:
2045 my($stmt, @bind) = $sql->where(\%where);
2047 However, note that the C<%where> hash can be used directly in any
2048 of the other functions as well, as described above.
2050 =head2 Key-value pairs
2052 So, let's get started. To begin, a simple hash:
2056 status => 'completed'
2059 Is converted to SQL C<key = val> statements:
2061 $stmt = "WHERE user = ? AND status = ?";
2062 @bind = ('nwiger', 'completed');
2064 One common thing I end up doing is having a list of values that
2065 a field can be in. To do this, simply specify a list inside of
2070 status => ['assigned', 'in-progress', 'pending'];
2073 This simple code will create the following:
2075 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
2076 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
2078 A field associated to an empty arrayref will be considered a
2079 logical false and will generate 0=1.
2081 =head2 Tests for NULL values
2083 If the value part is C<undef> then this is converted to SQL <IS NULL>
2092 $stmt = "WHERE user = ? AND status IS NULL";
2095 To test if a column IS NOT NULL:
2099 status => { '!=', undef },
2102 =head2 Specific comparison operators
2104 If you want to specify a different type of operator for your comparison,
2105 you can use a hashref for a given column:
2109 status => { '!=', 'completed' }
2112 Which would generate:
2114 $stmt = "WHERE user = ? AND status != ?";
2115 @bind = ('nwiger', 'completed');
2117 To test against multiple values, just enclose the values in an arrayref:
2119 status => { '=', ['assigned', 'in-progress', 'pending'] };
2121 Which would give you:
2123 "WHERE status = ? OR status = ? OR status = ?"
2126 The hashref can also contain multiple pairs, in which case it is expanded
2127 into an C<AND> of its elements:
2131 status => { '!=', 'completed', -not_like => 'pending%' }
2134 # Or more dynamically, like from a form
2135 $where{user} = 'nwiger';
2136 $where{status}{'!='} = 'completed';
2137 $where{status}{'-not_like'} = 'pending%';
2139 # Both generate this
2140 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
2141 @bind = ('nwiger', 'completed', 'pending%');
2144 To get an OR instead, you can combine it with the arrayref idea:
2148 priority => [ { '=', 2 }, { '>', 5 } ]
2151 Which would generate:
2153 $stmt = "WHERE ( priority = ? OR priority > ? ) AND user = ?";
2154 @bind = ('2', '5', 'nwiger');
2156 If you want to include literal SQL (with or without bind values), just use a
2157 scalar reference or reference to an arrayref as the value:
2160 date_entered => { '>' => \["to_date(?, 'MM/DD/YYYY')", "11/26/2008"] },
2161 date_expires => { '<' => \"now()" }
2164 Which would generate:
2166 $stmt = "WHERE date_entered > to_date(?, 'MM/DD/YYYY') AND date_expires < now()";
2167 @bind = ('11/26/2008');
2170 =head2 Logic and nesting operators
2172 In the example above,
2173 there is a subtle trap if you want to say something like
2174 this (notice the C<AND>):
2176 WHERE priority != ? AND priority != ?
2178 Because, in Perl you I<can't> do this:
2180 priority => { '!=' => 2, '!=' => 1 }
2182 As the second C<!=> key will obliterate the first. The solution
2183 is to use the special C<-modifier> form inside an arrayref:
2185 priority => [ -and => {'!=', 2},
2189 Normally, these would be joined by C<OR>, but the modifier tells it
2190 to use C<AND> instead. (Hint: You can use this in conjunction with the
2191 C<logic> option to C<new()> in order to change the way your queries
2192 work by default.) B<Important:> Note that the C<-modifier> goes
2193 B<INSIDE> the arrayref, as an extra first element. This will
2194 B<NOT> do what you think it might:
2196 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
2198 Here is a quick list of equivalencies, since there is some overlap:
2201 status => {'!=', 'completed', 'not like', 'pending%' }
2202 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
2205 status => {'=', ['assigned', 'in-progress']}
2206 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
2207 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
2211 =head2 Special operators: IN, BETWEEN, etc.
2213 You can also use the hashref format to compare a list of fields using the
2214 C<IN> comparison operator, by specifying the list as an arrayref:
2217 status => 'completed',
2218 reportid => { -in => [567, 2335, 2] }
2221 Which would generate:
2223 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
2224 @bind = ('completed', '567', '2335', '2');
2226 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
2229 If the argument to C<-in> is an empty array, 'sqlfalse' is generated
2230 (by default: C<1=0>). Similarly, C<< -not_in => [] >> generates
2231 'sqltrue' (by default: C<1=1>).
2233 In addition to the array you can supply a chunk of literal sql or
2234 literal sql with bind:
2237 customer => { -in => \[
2238 'SELECT cust_id FROM cust WHERE balance > ?',
2241 status => { -in => \'SELECT status_codes FROM states' },
2247 customer IN ( SELECT cust_id FROM cust WHERE balance > ? )
2248 AND status IN ( SELECT status_codes FROM states )
2252 Finally, if the argument to C<-in> is not a reference, it will be
2253 treated as a single-element array.
2255 Another pair of operators is C<-between> and C<-not_between>,
2256 used with an arrayref of two values:
2260 completion_date => {
2261 -not_between => ['2002-10-01', '2003-02-06']
2267 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
2269 Just like with C<-in> all plausible combinations of literal SQL
2273 start0 => { -between => [ 1, 2 ] },
2274 start1 => { -between => \["? AND ?", 1, 2] },
2275 start2 => { -between => \"lower(x) AND upper(y)" },
2276 start3 => { -between => [
2278 \["upper(?)", 'stuff' ],
2285 ( start0 BETWEEN ? AND ? )
2286 AND ( start1 BETWEEN ? AND ? )
2287 AND ( start2 BETWEEN lower(x) AND upper(y) )
2288 AND ( start3 BETWEEN lower(x) AND upper(?) )
2290 @bind = (1, 2, 1, 2, 'stuff');
2293 These are the two builtin "special operators"; but the
2294 list can be expanded: see section L</"SPECIAL OPERATORS"> below.
2296 =head2 Unary operators: bool
2298 If you wish to test against boolean columns or functions within your
2299 database you can use the C<-bool> and C<-not_bool> operators. For
2300 example to test the column C<is_user> being true and the column
2301 C<is_enabled> being false you would use:-
2305 -not_bool => 'is_enabled',
2310 WHERE is_user AND NOT is_enabled
2312 If a more complex combination is required, testing more conditions,
2313 then you should use the and/or operators:-
2318 -not_bool => { two=> { -rlike => 'bar' } },
2319 -not_bool => { three => [ { '=', 2 }, { '>', 5 } ] },
2330 (NOT ( three = ? OR three > ? ))
2333 =head2 Nested conditions, -and/-or prefixes
2335 So far, we've seen how multiple conditions are joined with a top-level
2336 C<AND>. We can change this by putting the different conditions we want in
2337 hashes and then putting those hashes in an array. For example:
2342 status => { -like => ['pending%', 'dispatched'] },
2346 status => 'unassigned',
2350 This data structure would create the following:
2352 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
2353 OR ( user = ? AND status = ? ) )";
2354 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
2357 Clauses in hashrefs or arrayrefs can be prefixed with an C<-and> or C<-or>
2358 to change the logic inside:
2364 -and => [ workhrs => {'>', 20}, geo => 'ASIA' ],
2365 -or => { workhrs => {'<', 50}, geo => 'EURO' },
2372 $stmt = "WHERE ( user = ?
2373 AND ( ( workhrs > ? AND geo = ? )
2374 OR ( workhrs < ? OR geo = ? ) ) )";
2375 @bind = ('nwiger', '20', 'ASIA', '50', 'EURO');
2377 =head3 Algebraic inconsistency, for historical reasons
2379 C<Important note>: when connecting several conditions, the C<-and->|C<-or>
2380 operator goes C<outside> of the nested structure; whereas when connecting
2381 several constraints on one column, the C<-and> operator goes
2382 C<inside> the arrayref. Here is an example combining both features:
2385 -and => [a => 1, b => 2],
2386 -or => [c => 3, d => 4],
2387 e => [-and => {-like => 'foo%'}, {-like => '%bar'} ]
2392 WHERE ( ( ( a = ? AND b = ? )
2393 OR ( c = ? OR d = ? )
2394 OR ( e LIKE ? AND e LIKE ? ) ) )
2396 This difference in syntax is unfortunate but must be preserved for
2397 historical reasons. So be careful: the two examples below would
2398 seem algebraically equivalent, but they are not
2401 { -like => 'foo%' },
2402 { -like => '%bar' },
2404 # yields: WHERE ( ( col LIKE ? AND col LIKE ? ) )
2407 { col => { -like => 'foo%' } },
2408 { col => { -like => '%bar' } },
2410 # yields: WHERE ( ( col LIKE ? OR col LIKE ? ) )
2413 =head2 Literal SQL and value type operators
2415 The basic premise of SQL::Abstract is that in WHERE specifications the "left
2416 side" is a column name and the "right side" is a value (normally rendered as
2417 a placeholder). This holds true for both hashrefs and arrayref pairs as you
2418 see in the L</WHERE CLAUSES> examples above. Sometimes it is necessary to
2419 alter this behavior. There are several ways of doing so.
2423 This is a virtual operator that signals the string to its right side is an
2424 identifier (a column name) and not a value. For example to compare two
2425 columns you would write:
2428 priority => { '<', 2 },
2429 requestor => { -ident => 'submitter' },
2434 $stmt = "WHERE priority < ? AND requestor = submitter";
2437 If you are maintaining legacy code you may see a different construct as
2438 described in L</Deprecated usage of Literal SQL>, please use C<-ident> in new
2443 This is a virtual operator that signals that the construct to its right side
2444 is a value to be passed to DBI. This is for example necessary when you want
2445 to write a where clause against an array (for RDBMS that support such
2446 datatypes). For example:
2449 array => { -value => [1, 2, 3] }
2454 $stmt = 'WHERE array = ?';
2455 @bind = ([1, 2, 3]);
2457 Note that if you were to simply say:
2463 the result would probably not be what you wanted:
2465 $stmt = 'WHERE array = ? OR array = ? OR array = ?';
2470 Finally, sometimes only literal SQL will do. To include a random snippet
2471 of SQL verbatim, you specify it as a scalar reference. Consider this only
2472 as a last resort. Usually there is a better way. For example:
2475 priority => { '<', 2 },
2476 requestor => { -in => \'(SELECT name FROM hitmen)' },
2481 $stmt = "WHERE priority < ? AND requestor IN (SELECT name FROM hitmen)"
2484 Note that in this example, you only get one bind parameter back, since
2485 the verbatim SQL is passed as part of the statement.
2489 Never use untrusted input as a literal SQL argument - this is a massive
2490 security risk (there is no way to check literal snippets for SQL
2491 injections and other nastyness). If you need to deal with untrusted input
2492 use literal SQL with placeholders as described next.
2494 =head3 Literal SQL with placeholders and bind values (subqueries)
2496 If the literal SQL to be inserted has placeholders and bind values,
2497 use a reference to an arrayref (yes this is a double reference --
2498 not so common, but perfectly legal Perl). For example, to find a date
2499 in Postgres you can use something like this:
2502 date_column => \[ "= date '2008-09-30' - ?::integer", 10 ]
2507 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
2510 Note that you must pass the bind values in the same format as they are returned
2511 by L<where|/where(\%where, $order)>. This means that if you set L</bindtype>
2512 to C<columns>, you must provide the bind values in the
2513 C<< [ column_meta => value ] >> format, where C<column_meta> is an opaque
2514 scalar value; most commonly the column name, but you can use any scalar value
2515 (including references and blessed references), L<SQL::Abstract> will simply
2516 pass it through intact. So if C<bindtype> is set to C<columns> the above
2517 example will look like:
2520 date_column => \[ "= date '2008-09-30' - ?::integer", [ {} => 10 ] ]
2523 Literal SQL is especially useful for nesting parenthesized clauses in the
2524 main SQL query. Here is a first example:
2526 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
2530 bar => \["IN ($sub_stmt)" => @sub_bind],
2535 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
2536 WHERE c2 < ? AND c3 LIKE ?))";
2537 @bind = (1234, 100, "foo%");
2539 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
2540 are expressed in the same way. Of course the C<$sub_stmt> and
2541 its associated bind values can be generated through a former call
2544 my ($sub_stmt, @sub_bind)
2545 = $sql->select("t1", "c1", {c2 => {"<" => 100},
2546 c3 => {-like => "foo%"}});
2549 bar => \["> ALL ($sub_stmt)" => @sub_bind],
2552 In the examples above, the subquery was used as an operator on a column;
2553 but the same principle also applies for a clause within the main C<%where>
2554 hash, like an EXISTS subquery:
2556 my ($sub_stmt, @sub_bind)
2557 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
2558 my %where = ( -and => [
2560 \["EXISTS ($sub_stmt)" => @sub_bind],
2565 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
2566 WHERE c1 = ? AND c2 > t0.c0))";
2570 Observe that the condition on C<c2> in the subquery refers to
2571 column C<t0.c0> of the main query: this is I<not> a bind
2572 value, so we have to express it through a scalar ref.
2573 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
2574 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
2575 what we wanted here.
2577 Finally, here is an example where a subquery is used
2578 for expressing unary negation:
2580 my ($sub_stmt, @sub_bind)
2581 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
2582 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
2584 lname => {like => '%son%'},
2585 \["NOT ($sub_stmt)" => @sub_bind],
2590 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
2591 @bind = ('%son%', 10, 20)
2593 =head3 Deprecated usage of Literal SQL
2595 Below are some examples of archaic use of literal SQL. It is shown only as
2596 reference for those who deal with legacy code. Each example has a much
2597 better, cleaner and safer alternative that users should opt for in new code.
2603 my %where = ( requestor => \'IS NOT NULL' )
2605 $stmt = "WHERE requestor IS NOT NULL"
2607 This used to be the way of generating NULL comparisons, before the handling
2608 of C<undef> got formalized. For new code please use the superior syntax as
2609 described in L</Tests for NULL values>.
2613 my %where = ( requestor => \'= submitter' )
2615 $stmt = "WHERE requestor = submitter"
2617 This used to be the only way to compare columns. Use the superior L</-ident>
2618 method for all new code. For example an identifier declared in such a way
2619 will be properly quoted if L</quote_char> is properly set, while the legacy
2620 form will remain as supplied.
2624 my %where = ( is_ready => \"", completed => { '>', '2012-12-21' } )
2626 $stmt = "WHERE completed > ? AND is_ready"
2627 @bind = ('2012-12-21')
2629 Using an empty string literal used to be the only way to express a boolean.
2630 For all new code please use the much more readable
2631 L<-bool|/Unary operators: bool> operator.
2637 These pages could go on for a while, since the nesting of the data
2638 structures this module can handle are pretty much unlimited (the
2639 module implements the C<WHERE> expansion as a recursive function
2640 internally). Your best bet is to "play around" with the module a
2641 little to see how the data structures behave, and choose the best
2642 format for your data based on that.
2644 And of course, all the values above will probably be replaced with
2645 variables gotten from forms or the command line. After all, if you
2646 knew everything ahead of time, you wouldn't have to worry about
2647 dynamically-generating SQL and could just hardwire it into your
2650 =head1 ORDER BY CLAUSES
2652 Some functions take an order by clause. This can either be a scalar (just a
2653 column name), a hashref of C<< { -desc => 'col' } >> or C<< { -asc => 'col' }
2654 >>, a scalarref, an arrayref-ref, or an arrayref of any of the previous
2657 Given | Will Generate
2658 ---------------------------------------------------------------
2660 'colA' | ORDER BY colA
2662 [qw/colA colB/] | ORDER BY colA, colB
2664 {-asc => 'colA'} | ORDER BY colA ASC
2666 {-desc => 'colB'} | ORDER BY colB DESC
2668 ['colA', {-asc => 'colB'}] | ORDER BY colA, colB ASC
2670 { -asc => [qw/colA colB/] } | ORDER BY colA ASC, colB ASC
2672 \'colA DESC' | ORDER BY colA DESC
2674 \[ 'FUNC(colA, ?)', $x ] | ORDER BY FUNC(colA, ?)
2675 | /* ...with $x bound to ? */
2678 { -asc => 'colA' }, | colA ASC,
2679 { -desc => [qw/colB/] }, | colB DESC,
2680 { -asc => [qw/colC colD/] },| colC ASC, colD ASC,
2681 \'colE DESC', | colE DESC,
2682 \[ 'FUNC(colF, ?)', $x ], | FUNC(colF, ?)
2683 ] | /* ...with $x bound to ? */
2684 ===============================================================
2688 =head1 SPECIAL OPERATORS
2690 my $sqlmaker = SQL::Abstract->new(special_ops => [
2694 my ($self, $field, $op, $arg) = @_;
2700 handler => 'method_name',
2704 A "special operator" is a SQL syntactic clause that can be
2705 applied to a field, instead of a usual binary operator.
2708 WHERE field IN (?, ?, ?)
2709 WHERE field BETWEEN ? AND ?
2710 WHERE MATCH(field) AGAINST (?, ?)
2712 Special operators IN and BETWEEN are fairly standard and therefore
2713 are builtin within C<SQL::Abstract> (as the overridable methods
2714 C<_where_field_IN> and C<_where_field_BETWEEN>). For other operators,
2715 like the MATCH .. AGAINST example above which is specific to MySQL,
2716 you can write your own operator handlers - supply a C<special_ops>
2717 argument to the C<new> method. That argument takes an arrayref of
2718 operator definitions; each operator definition is a hashref with two
2725 the regular expression to match the operator
2729 Either a coderef or a plain scalar method name. In both cases
2730 the expected return is C<< ($sql, @bind) >>.
2732 When supplied with a method name, it is simply called on the
2733 L<SQL::Abstract> object as:
2735 $self->$method_name($field, $op, $arg)
2739 $field is the LHS of the operator
2740 $op is the part that matched the handler regex
2743 When supplied with a coderef, it is called as:
2745 $coderef->($self, $field, $op, $arg)
2750 For example, here is an implementation
2751 of the MATCH .. AGAINST syntax for MySQL
2753 my $sqlmaker = SQL::Abstract->new(special_ops => [
2755 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
2756 {regex => qr/^match$/i,
2758 my ($self, $field, $op, $arg) = @_;
2759 $arg = [$arg] if not ref $arg;
2760 my $label = $self->_quote($field);
2761 my ($placeholder) = $self->_convert('?');
2762 my $placeholders = join ", ", (($placeholder) x @$arg);
2763 my $sql = $self->_sqlcase('match') . " ($label) "
2764 . $self->_sqlcase('against') . " ($placeholders) ";
2765 my @bind = $self->_bindtype($field, @$arg);
2766 return ($sql, @bind);
2773 =head1 UNARY OPERATORS
2775 my $sqlmaker = SQL::Abstract->new(unary_ops => [
2779 my ($self, $op, $arg) = @_;
2785 handler => 'method_name',
2789 A "unary operator" is a SQL syntactic clause that can be
2790 applied to a field - the operator goes before the field
2792 You can write your own operator handlers - supply a C<unary_ops>
2793 argument to the C<new> method. That argument takes an arrayref of
2794 operator definitions; each operator definition is a hashref with two
2801 the regular expression to match the operator
2805 Either a coderef or a plain scalar method name. In both cases
2806 the expected return is C<< $sql >>.
2808 When supplied with a method name, it is simply called on the
2809 L<SQL::Abstract> object as:
2811 $self->$method_name($op, $arg)
2815 $op is the part that matched the handler regex
2816 $arg is the RHS or argument of the operator
2818 When supplied with a coderef, it is called as:
2820 $coderef->($self, $op, $arg)
2828 Thanks to some benchmarking by Mark Stosberg, it turns out that
2829 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
2830 I must admit this wasn't an intentional design issue, but it's a
2831 byproduct of the fact that you get to control your C<DBI> handles
2834 To maximize performance, use a code snippet like the following:
2836 # prepare a statement handle using the first row
2837 # and then reuse it for the rest of the rows
2839 for my $href (@array_of_hashrefs) {
2840 $stmt ||= $sql->insert('table', $href);
2841 $sth ||= $dbh->prepare($stmt);
2842 $sth->execute($sql->values($href));
2845 The reason this works is because the keys in your C<$href> are sorted
2846 internally by B<SQL::Abstract>. Thus, as long as your data retains
2847 the same structure, you only have to generate the SQL the first time
2848 around. On subsequent queries, simply use the C<values> function provided
2849 by this module to return your values in the correct order.
2851 However this depends on the values having the same type - if, for
2852 example, the values of a where clause may either have values
2853 (resulting in sql of the form C<column = ?> with a single bind
2854 value), or alternatively the values might be C<undef> (resulting in
2855 sql of the form C<column IS NULL> with no bind value) then the
2856 caching technique suggested will not work.
2860 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
2861 really like this part (I do, at least). Building up a complex query
2862 can be as simple as the following:
2869 use CGI::FormBuilder;
2872 my $form = CGI::FormBuilder->new(...);
2873 my $sql = SQL::Abstract->new;
2875 if ($form->submitted) {
2876 my $field = $form->field;
2877 my $id = delete $field->{id};
2878 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
2881 Of course, you would still have to connect using C<DBI> to run the
2882 query, but the point is that if you make your form look like your
2883 table, the actual query script can be extremely simplistic.
2885 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
2886 a fast interface to returning and formatting data. I frequently
2887 use these three modules together to write complex database query
2888 apps in under 50 lines.
2890 =head1 HOW TO CONTRIBUTE
2892 Contributions are always welcome, in all usable forms (we especially
2893 welcome documentation improvements). The delivery methods include git-
2894 or unified-diff formatted patches, GitHub pull requests, or plain bug
2895 reports either via RT or the Mailing list. Contributors are generally
2896 granted full access to the official repository after their first several
2897 patches pass successful review.
2899 This project is maintained in a git repository. The code and related tools are
2900 accessible at the following locations:
2904 =item * Official repo: L<git://git.shadowcat.co.uk/dbsrgits/SQL-Abstract.git>
2906 =item * Official gitweb: L<http://git.shadowcat.co.uk/gitweb/gitweb.cgi?p=dbsrgits/SQL-Abstract.git>
2908 =item * GitHub mirror: L<https://github.com/dbsrgits/sql-abstract>
2910 =item * Authorized committers: L<ssh://dbsrgits@git.shadowcat.co.uk/SQL-Abstract.git>
2916 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
2917 Great care has been taken to preserve the I<published> behavior
2918 documented in previous versions in the 1.* family; however,
2919 some features that were previously undocumented, or behaved
2920 differently from the documentation, had to be changed in order
2921 to clarify the semantics. Hence, client code that was relying
2922 on some dark areas of C<SQL::Abstract> v1.*
2923 B<might behave differently> in v1.50.
2925 The main changes are:
2931 support for literal SQL through the C<< \ [ $sql, @bind ] >> syntax.
2935 support for the { operator => \"..." } construct (to embed literal SQL)
2939 support for the { operator => \["...", @bind] } construct (to embed literal SQL with bind values)
2943 optional support for L<array datatypes|/"Inserting and Updating Arrays">
2947 defensive programming: check arguments
2951 fixed bug with global logic, which was previously implemented
2952 through global variables yielding side-effects. Prior versions would
2953 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
2954 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
2955 Now this is interpreted
2956 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
2961 fixed semantics of _bindtype on array args
2965 dropped the C<_anoncopy> of the %where tree. No longer necessary,
2966 we just avoid shifting arrays within that tree.
2970 dropped the C<_modlogic> function
2974 =head1 ACKNOWLEDGEMENTS
2976 There are a number of individuals that have really helped out with
2977 this module. Unfortunately, most of them submitted bugs via CPAN
2978 so I have no idea who they are! But the people I do know are:
2980 Ash Berlin (order_by hash term support)
2981 Matt Trout (DBIx::Class support)
2982 Mark Stosberg (benchmarking)
2983 Chas Owens (initial "IN" operator support)
2984 Philip Collins (per-field SQL functions)
2985 Eric Kolve (hashref "AND" support)
2986 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
2987 Dan Kubb (support for "quote_char" and "name_sep")
2988 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
2989 Laurent Dami (internal refactoring, extensible list of special operators, literal SQL)
2990 Norbert Buchmuller (support for literal SQL in hashpair, misc. fixes & tests)
2991 Peter Rabbitson (rewrite of SQLA::Test, misc. fixes & tests)
2992 Oliver Charles (support for "RETURNING" after "INSERT")
2998 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
3002 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
3004 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
3006 For support, your best bet is to try the C<DBIx::Class> users mailing list.
3007 While not an official support venue, C<DBIx::Class> makes heavy use of
3008 C<SQL::Abstract>, and as such list members there are very familiar with
3009 how to create queries.
3013 This module is free software; you may copy this under the same
3014 terms as perl itself (either the GNU General Public License or
3015 the Artistic License)