1 package SQL::Abstract; # see doc at end of file
3 # LDNOTE : this code is heavy refactoring from original SQLA.
4 # Several design decisions will need discussion during
5 # the test / diffusion / acceptance phase; those are marked with flag
6 # 'LDNOTE' (note by laurent.dami AT free.fr)
14 #======================================================================
16 #======================================================================
18 our $VERSION = '1.62';
20 # This would confuse some packagers
21 #$VERSION = eval $VERSION; # numify for warning-free dev releases
25 # special operators (-in, -between). May be extended/overridden by user.
26 # See section WHERE: BUILTIN SPECIAL OPERATORS below for implementation
27 my @BUILTIN_SPECIAL_OPS = (
28 {regex => qr/^(not )?between$/i, handler => '_where_field_BETWEEN'},
29 {regex => qr/^(not )?in$/i, handler => '_where_field_IN'},
32 # unaryish operators - key maps to handler
33 my @BUILTIN_UNARY_OPS = (
34 # the digits are backcompat stuff
35 { regex => qr/^and (?: \s? \d+ )? $/xi, handler => '_where_op_ANDOR' },
36 { regex => qr/^or (?: \s? \d+ )? $/xi, handler => '_where_op_ANDOR' },
37 { regex => qr/^nest (?: \s? \d+ )? $/xi, handler => '_where_op_NEST' },
38 { regex => qr/^ (?: not \s )? bool $/xi, handler => '_where_op_BOOL' },
41 #======================================================================
42 # DEBUGGING AND ERROR REPORTING
43 #======================================================================
46 return unless $_[0]->{debug}; shift; # a little faster
47 my $func = (caller(1))[3];
48 warn "[$func] ", @_, "\n";
52 my($func) = (caller(1))[3];
53 carp "[$func] Warning: ", @_;
57 my($func) = (caller(1))[3];
58 croak "[$func] Fatal: ", @_;
62 #======================================================================
64 #======================================================================
68 my $class = ref($self) || $self;
69 my %opt = (ref $_[0] eq 'HASH') ? %{$_[0]} : @_;
71 # choose our case by keeping an option around
72 delete $opt{case} if $opt{case} && $opt{case} ne 'lower';
74 # default logic for interpreting arrayrefs
75 $opt{logic} = $opt{logic} ? uc $opt{logic} : 'OR';
77 # how to return bind vars
78 # LDNOTE: changed nwiger code : why this 'delete' ??
79 # $opt{bindtype} ||= delete($opt{bind_type}) || 'normal';
80 $opt{bindtype} ||= 'normal';
82 # default comparison is "=", but can be overridden
85 # generic SQL comparison operators
86 my $anchored_cmp_ops = join ('|', map { '^' . $_ . '$' } (
87 '(?:is \s+)? (?:not \s+)? i? like',
89 (map { quotemeta($_) } (qw/ < > != <> = <= >= /) ),
91 $opt{cmp_ops} = qr/$anchored_cmp_ops/ix;
93 # try to recognize which are the 'equality' and 'unequality' ops
94 # (temporary quickfix, should go through a more seasoned API)
95 $opt{equality_op} = qr/^(\Q$opt{cmp}\E|is|(is\s+)?like)$/i;
96 $opt{inequality_op} = qr/^(!=|<>|(is\s+)?not(\s+like)?)$/i;
99 $opt{sqltrue} ||= '1=1';
100 $opt{sqlfalse} ||= '0=1';
103 $opt{special_ops} ||= [];
104 push @{$opt{special_ops}}, @BUILTIN_SPECIAL_OPS;
107 $opt{unary_ops} ||= [];
108 push @{$opt{unary_ops}}, @BUILTIN_UNARY_OPS;
110 return bless \%opt, $class;
115 #======================================================================
117 #======================================================================
121 my $table = $self->_table(shift);
122 my $data = shift || return;
125 my $method = $self->_METHOD_FOR_refkind("_insert", $data);
126 my ($sql, @bind) = $self->$method($data);
127 $sql = join " ", $self->_sqlcase('insert into'), $table, $sql;
129 if (my $ret = $options->{returning}) {
130 $sql .= $self->_insert_returning ($ret);
133 return wantarray ? ($sql, @bind) : $sql;
136 sub _insert_returning {
137 my ($self, $fields) = @_;
139 my $f = $self->_SWITCH_refkind($fields, {
140 ARRAYREF => sub {join ', ', map { $self->_quote($_) } @$fields;},
141 SCALAR => sub {$self->_quote($fields)},
142 SCALARREF => sub {$$fields},
144 return join (' ', $self->_sqlcase(' returning'), $f);
147 sub _insert_HASHREF { # explicit list of fields and then values
148 my ($self, $data) = @_;
150 my @fields = sort keys %$data;
152 my ($sql, @bind) = $self->_insert_values($data);
155 $_ = $self->_quote($_) foreach @fields;
156 $sql = "( ".join(", ", @fields).") ".$sql;
158 return ($sql, @bind);
161 sub _insert_ARRAYREF { # just generate values(?,?) part (no list of fields)
162 my ($self, $data) = @_;
164 # no names (arrayref) so can't generate bindtype
165 $self->{bindtype} ne 'columns'
166 or belch "can't do 'columns' bindtype when called with arrayref";
168 # fold the list of values into a hash of column name - value pairs
169 # (where the column names are artificially generated, and their
170 # lexicographical ordering keep the ordering of the original list)
171 my $i = "a"; # incremented values will be in lexicographical order
172 my $data_in_hash = { map { ($i++ => $_) } @$data };
174 return $self->_insert_values($data_in_hash);
177 sub _insert_ARRAYREFREF { # literal SQL with bind
178 my ($self, $data) = @_;
180 my ($sql, @bind) = @${$data};
181 $self->_assert_bindval_matches_bindtype(@bind);
183 return ($sql, @bind);
187 sub _insert_SCALARREF { # literal SQL without bind
188 my ($self, $data) = @_;
194 my ($self, $data) = @_;
196 my (@values, @all_bind);
197 foreach my $column (sort keys %$data) {
198 my $v = $data->{$column};
200 $self->_SWITCH_refkind($v, {
203 if ($self->{array_datatypes}) { # if array datatype are activated
205 push @all_bind, $self->_bindtype($column, $v);
207 else { # else literal SQL with bind
208 my ($sql, @bind) = @$v;
209 $self->_assert_bindval_matches_bindtype(@bind);
211 push @all_bind, @bind;
215 ARRAYREFREF => sub { # literal SQL with bind
216 my ($sql, @bind) = @${$v};
217 $self->_assert_bindval_matches_bindtype(@bind);
219 push @all_bind, @bind;
222 # THINK : anything useful to do with a HASHREF ?
223 HASHREF => sub { # (nothing, but old SQLA passed it through)
224 #TODO in SQLA >= 2.0 it will die instead
225 belch "HASH ref as bind value in insert is not supported";
227 push @all_bind, $self->_bindtype($column, $v);
230 SCALARREF => sub { # literal SQL without bind
234 SCALAR_or_UNDEF => sub {
236 push @all_bind, $self->_bindtype($column, $v);
243 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
244 return ($sql, @all_bind);
249 #======================================================================
251 #======================================================================
256 my $table = $self->_table(shift);
257 my $data = shift || return;
260 # first build the 'SET' part of the sql statement
261 my (@set, @all_bind);
262 puke "Unsupported data type specified to \$sql->update"
263 unless ref $data eq 'HASH';
265 for my $k (sort keys %$data) {
268 my $label = $self->_quote($k);
270 $self->_SWITCH_refkind($v, {
272 if ($self->{array_datatypes}) { # array datatype
273 push @set, "$label = ?";
274 push @all_bind, $self->_bindtype($k, $v);
276 else { # literal SQL with bind
277 my ($sql, @bind) = @$v;
278 $self->_assert_bindval_matches_bindtype(@bind);
279 push @set, "$label = $sql";
280 push @all_bind, @bind;
283 ARRAYREFREF => sub { # literal SQL with bind
284 my ($sql, @bind) = @${$v};
285 $self->_assert_bindval_matches_bindtype(@bind);
286 push @set, "$label = $sql";
287 push @all_bind, @bind;
289 SCALARREF => sub { # literal SQL without bind
290 push @set, "$label = $$v";
292 SCALAR_or_UNDEF => sub {
293 push @set, "$label = ?";
294 push @all_bind, $self->_bindtype($k, $v);
300 my $sql = $self->_sqlcase('update') . " $table " . $self->_sqlcase('set ')
304 my($where_sql, @where_bind) = $self->where($where);
306 push @all_bind, @where_bind;
309 return wantarray ? ($sql, @all_bind) : $sql;
315 #======================================================================
317 #======================================================================
322 my $table = $self->_table(shift);
323 my $fields = shift || '*';
327 my($where_sql, @bind) = $self->where($where, $order);
329 my $f = (ref $fields eq 'ARRAY') ? join ', ', map { $self->_quote($_) } @$fields
331 my $sql = join(' ', $self->_sqlcase('select'), $f,
332 $self->_sqlcase('from'), $table)
335 return wantarray ? ($sql, @bind) : $sql;
338 #======================================================================
340 #======================================================================
345 my $table = $self->_table(shift);
349 my($where_sql, @bind) = $self->where($where);
350 my $sql = $self->_sqlcase('delete from') . " $table" . $where_sql;
352 return wantarray ? ($sql, @bind) : $sql;
356 #======================================================================
358 #======================================================================
362 # Finally, a separate routine just to handle WHERE clauses
364 my ($self, $where, $order) = @_;
367 my ($sql, @bind) = $self->_recurse_where($where);
368 $sql = $sql ? $self->_sqlcase(' where ') . "( $sql )" : '';
372 $sql .= $self->_order_by($order);
375 return wantarray ? ($sql, @bind) : $sql;
380 my ($self, $where, $logic) = @_;
382 # dispatch on appropriate method according to refkind of $where
383 my $method = $self->_METHOD_FOR_refkind("_where", $where);
386 my ($sql, @bind) = $self->$method($where, $logic);
388 # DBIx::Class directly calls _recurse_where in scalar context, so
389 # we must implement it, even if not in the official API
390 return wantarray ? ($sql, @bind) : $sql;
395 #======================================================================
396 # WHERE: top-level ARRAYREF
397 #======================================================================
400 sub _where_ARRAYREF {
401 my ($self, $where, $logic) = @_;
403 $logic = uc($logic || $self->{logic});
404 $logic eq 'AND' or $logic eq 'OR' or puke "unknown logic: $logic";
406 my @clauses = @$where;
408 my (@sql_clauses, @all_bind);
409 # need to use while() so can shift() for pairs
410 while (my $el = shift @clauses) {
412 # switch according to kind of $el and get corresponding ($sql, @bind)
413 my ($sql, @bind) = $self->_SWITCH_refkind($el, {
415 # skip empty elements, otherwise get invalid trailing AND stuff
416 ARRAYREF => sub {$self->_recurse_where($el) if @$el},
418 ARRAYREFREF => sub { @{${$el}} if @{${$el}}},
420 HASHREF => sub {$self->_recurse_where($el, 'and') if %$el},
421 # LDNOTE : previous SQLA code for hashrefs was creating a dirty
422 # side-effect: the first hashref within an array would change
423 # the global logic to 'AND'. So [ {cond1, cond2}, [cond3, cond4] ]
424 # was interpreted as "(cond1 AND cond2) OR (cond3 AND cond4)",
425 # whereas it should be "(cond1 AND cond2) OR (cond3 OR cond4)".
427 SCALARREF => sub { ($$el); },
429 SCALAR => sub {# top-level arrayref with scalars, recurse in pairs
430 $self->_recurse_where({$el => shift(@clauses)})},
432 UNDEF => sub {puke "not supported : UNDEF in arrayref" },
436 push @sql_clauses, $sql;
437 push @all_bind, @bind;
441 return $self->_join_sql_clauses($logic, \@sql_clauses, \@all_bind);
444 #======================================================================
445 # WHERE: top-level ARRAYREFREF
446 #======================================================================
448 sub _where_ARRAYREFREF {
449 my ($self, $where) = @_;
450 my ($sql, @bind) = @{${$where}};
452 return ($sql, @bind);
455 #======================================================================
456 # WHERE: top-level HASHREF
457 #======================================================================
460 my ($self, $where) = @_;
461 my (@sql_clauses, @all_bind);
463 for my $k (sort keys %$where) {
464 my $v = $where->{$k};
466 # ($k => $v) is either a special unary op or a regular hashpair
467 my ($sql, @bind) = do {
469 # put the operator in canonical form
471 $op =~ s/^-//; # remove initial dash
472 $op =~ s/[_\t ]+/ /g; # underscores and whitespace become single spaces
473 $op =~ s/^\s+|\s+$//g;# remove leading/trailing space
475 $self->_debug("Unary OP(-$op) within hashref, recursing...");
477 my $op_entry = List::Util::first {$op =~ $_->{regex}} @{$self->{unary_ops}};
478 if (my $handler = $op_entry->{handler}) {
479 if (not ref $handler) {
480 if ($op =~ s/\s?\d+$//) {
481 belch 'Use of [and|or|nest]_N modifiers is deprecated and will be removed in SQLA v2.0. '
482 . "You probably wanted ...-and => [ -$op => COND1, -$op => COND2 ... ]";
484 $self->$handler ($op, $v);
486 elsif (ref $handler eq 'CODE') {
487 $handler->($self, $op, $v);
490 puke "Illegal handler for operator $k - expecting a method name or a coderef";
494 $self->debug("Generic unary OP: $k - recursing as function");
495 $self->_where_func_generic ($op, $v);
499 my $method = $self->_METHOD_FOR_refkind("_where_hashpair", $v);
500 $self->$method($k, $v);
504 push @sql_clauses, $sql;
505 push @all_bind, @bind;
508 return $self->_join_sql_clauses('and', \@sql_clauses, \@all_bind);
511 sub _where_func_generic {
512 my ($self, $op, $rhs) = @_;
514 my ($sql, @bind) = $self->_SWITCH_refkind ($rhs, {
516 puke "Illegal use of top-level '$op'"
517 unless $self->{_nested_func_lhs};
520 $self->_convert('?'),
521 $self->_bindtype($self->{_nested_func_lhs}, $rhs)
525 $self->_recurse_where ($rhs)
529 $sql = sprintf ('%s%s',
530 $self->_sqlcase($op),
531 ($op =~ $self->{cmp_ops}) ? " $sql" : "( $sql )",
534 return ($sql, @bind);
537 sub _where_op_ANDOR {
538 my ($self, $op, $v) = @_;
540 $self->_SWITCH_refkind($v, {
542 return $self->_where_ARRAYREF($v, $op);
546 return ( $op =~ /^or/i )
547 ? $self->_where_ARRAYREF( [ map { $_ => $v->{$_} } ( sort keys %$v ) ], $op )
548 : $self->_where_HASHREF($v);
552 puke "-$op => \\\$scalar not supported, use -nest => ...";
556 puke "-$op => \\[..] not supported, use -nest => ...";
559 SCALAR => sub { # permissively interpreted as SQL
560 puke "-$op => 'scalar' not supported, use -nest => \\'scalar'";
564 puke "-$op => undef not supported";
570 my ($self, $op, $v) = @_;
572 $self->_SWITCH_refkind($v, {
574 SCALAR => sub { # permissively interpreted as SQL
575 belch "literal SQL should be -nest => \\'scalar' "
576 . "instead of -nest => 'scalar' ";
581 puke "-$op => undef not supported";
585 $self->_recurse_where ($v);
593 my ($self, $op, $v) = @_;
595 my ( $prefix, $suffix ) = ( $op =~ /\bnot\b/i )
599 my ($sql, @bind) = do {
600 $self->_SWITCH_refkind($v, {
601 SCALAR => sub { # interpreted as SQL column
602 $self->_convert($self->_quote($v));
606 puke "-$op => undef not supported";
610 $self->_recurse_where ($v);
616 join ('', $prefix, $sql, $suffix),
622 sub _where_hashpair_ARRAYREF {
623 my ($self, $k, $v) = @_;
626 my @v = @$v; # need copy because of shift below
627 $self->_debug("ARRAY($k) means distribute over elements");
629 # put apart first element if it is an operator (-and, -or)
631 (defined $v[0] && $v[0] =~ /^ - (?: AND|OR ) $/ix)
635 my @distributed = map { {$k => $_} } @v;
638 $self->_debug("OP($op) reinjected into the distributed array");
639 unshift @distributed, $op;
642 my $logic = $op ? substr($op, 1) : '';
644 return $self->_recurse_where(\@distributed, $logic);
647 # LDNOTE : not sure of this one. What does "distribute over nothing" mean?
648 $self->_debug("empty ARRAY($k) means 0=1");
649 return ($self->{sqlfalse});
653 sub _where_hashpair_HASHREF {
654 my ($self, $k, $v, $logic) = @_;
657 local $self->{_nested_func_lhs} = $self->{_nested_func_lhs};
659 my ($all_sql, @all_bind);
661 for my $orig_op (sort keys %$v) {
662 my $val = $v->{$orig_op};
664 # put the operator in canonical form
666 $op =~ s/^-//; # remove initial dash
667 $op =~ s/[_\t ]+/ /g; # underscores and whitespace become single spaces
668 $op =~ s/^\s+|\s+$//g;# remove leading/trailing space
672 # CASE: col-value logic modifiers
673 if ( $orig_op =~ /^ \- (and|or) $/xi ) {
674 ($sql, @bind) = $self->_where_hashpair_HASHREF($k, $val, $1);
676 # CASE: special operators like -in or -between
677 elsif ( my $special_op = List::Util::first {$op =~ $_->{regex}} @{$self->{special_ops}} ) {
678 my $handler = $special_op->{handler};
680 puke "No handler supplied for special operator $orig_op";
682 elsif (not ref $handler) {
683 ($sql, @bind) = $self->$handler ($k, $op, $val);
685 elsif (ref $handler eq 'CODE') {
686 ($sql, @bind) = $handler->($self, $k, $op, $val);
689 puke "Illegal handler for special operator $orig_op - expecting a method name or a coderef";
693 $self->_SWITCH_refkind($val, {
695 ARRAYREF => sub { # CASE: col => {op => \@vals}
696 ($sql, @bind) = $self->_where_field_op_ARRAYREF($k, $op, $val);
699 ARRAYREFREF => sub { # CASE: col => {op => \[$sql, @bind]} (literal SQL with bind)
700 my ($sub_sql, @sub_bind) = @$$val;
701 $self->_assert_bindval_matches_bindtype(@sub_bind);
702 $sql = join ' ', $self->_convert($self->_quote($k)),
703 $self->_sqlcase($op),
708 UNDEF => sub { # CASE: col => {op => undef} : sql "IS (NOT)? NULL"
709 my $is = ($op =~ $self->{equality_op}) ? 'is' :
710 ($op =~ $self->{inequality_op}) ? 'is not' :
711 puke "unexpected operator '$orig_op' with undef operand";
712 $sql = $self->_quote($k) . $self->_sqlcase(" $is null");
715 FALLBACK => sub { # CASE: col => {op/func => $stuff}
717 # if we are starting to nest and the first func is not a cmp op
720 unless ($self->{_nested_func_lhs}) {
721 $self->{_nested_func_lhs} = $k;
722 $prefix = $self->{cmp} unless $op =~ $self->{cmp_ops};
725 ($sql, @bind) = $self->_where_func_generic ($op, $val);
726 $sql = join ' ', $self->_convert($self->_quote($k)), $prefix||(), $sql;
731 ($all_sql) = (defined $all_sql and $all_sql) ? $self->_join_sql_clauses($logic, [$all_sql, $sql], []) : $sql;
732 push @all_bind, @bind;
734 return ($all_sql, @all_bind);
739 sub _where_field_op_ARRAYREF {
740 my ($self, $k, $op, $vals) = @_;
742 my @vals = @$vals; #always work on a copy
745 $self->_debug(sprintf '%s means multiple elements: [ %s ]',
747 join (', ', map { defined $_ ? "'$_'" : 'NULL' } @vals ),
750 # see if the first element is an -and/-or op
752 if (defined $vals[0] && $vals[0] =~ /^ - ( AND|OR ) $/ix) {
757 # distribute $op over each remaining member of @vals, append logic if exists
758 return $self->_recurse_where([map { {$k => {$op, $_}} } @vals], $logic);
760 # LDNOTE : had planned to change the distribution logic when
761 # $op =~ $self->{inequality_op}, because of Morgan laws :
762 # with {field => {'!=' => [22, 33]}}, it would be ridiculous to generate
763 # WHERE field != 22 OR field != 33 : the user probably means
764 # WHERE field != 22 AND field != 33.
765 # To do this, replace the above to roughly :
766 # my $logic = ($op =~ $self->{inequality_op}) ? 'AND' : 'OR';
767 # return $self->_recurse_where([map { {$k => {$op, $_}} } @vals], $logic);
771 # try to DWIM on equality operators
772 # LDNOTE : not 100% sure this is the correct thing to do ...
773 return ($self->{sqlfalse}) if $op =~ $self->{equality_op};
774 return ($self->{sqltrue}) if $op =~ $self->{inequality_op};
777 puke "operator '$op' applied on an empty array (field '$k')";
782 sub _where_hashpair_SCALARREF {
783 my ($self, $k, $v) = @_;
784 $self->_debug("SCALAR($k) means literal SQL: $$v");
785 my $sql = $self->_quote($k) . " " . $$v;
789 # literal SQL with bind
790 sub _where_hashpair_ARRAYREFREF {
791 my ($self, $k, $v) = @_;
792 $self->_debug("REF($k) means literal SQL: @${$v}");
793 my ($sql, @bind) = @${$v};
794 $self->_assert_bindval_matches_bindtype(@bind);
795 $sql = $self->_quote($k) . " " . $sql;
796 return ($sql, @bind );
799 # literal SQL without bind
800 sub _where_hashpair_SCALAR {
801 my ($self, $k, $v) = @_;
802 $self->_debug("NOREF($k) means simple key=val: $k $self->{cmp} $v");
803 my $sql = join ' ', $self->_convert($self->_quote($k)),
804 $self->_sqlcase($self->{cmp}),
805 $self->_convert('?');
806 my @bind = $self->_bindtype($k, $v);
807 return ( $sql, @bind);
811 sub _where_hashpair_UNDEF {
812 my ($self, $k, $v) = @_;
813 $self->_debug("UNDEF($k) means IS NULL");
814 my $sql = $self->_quote($k) . $self->_sqlcase(' is null');
818 #======================================================================
819 # WHERE: TOP-LEVEL OTHERS (SCALARREF, SCALAR, UNDEF)
820 #======================================================================
823 sub _where_SCALARREF {
824 my ($self, $where) = @_;
827 $self->_debug("SCALAR(*top) means literal SQL: $$where");
833 my ($self, $where) = @_;
836 $self->_debug("NOREF(*top) means literal SQL: $where");
847 #======================================================================
848 # WHERE: BUILTIN SPECIAL OPERATORS (-in, -between)
849 #======================================================================
852 sub _where_field_BETWEEN {
853 my ($self, $k, $op, $vals) = @_;
855 my ($label, $and, $placeholder);
856 $label = $self->_convert($self->_quote($k));
857 $and = ' ' . $self->_sqlcase('and') . ' ';
858 $placeholder = $self->_convert('?');
859 $op = $self->_sqlcase($op);
861 my ($clause, @bind) = $self->_SWITCH_refkind($vals, {
869 puke "special op 'between' accepts an arrayref with exactly two values"
872 my (@all_sql, @all_bind);
873 foreach my $val (@$vals) {
874 my ($sql, @bind) = $self->_SWITCH_refkind($val, {
876 return ($placeholder, ($val));
879 return ($self->_convert($$val), ());
882 my ($sql, @bind) = @$$val;
883 return ($self->_convert($sql), @bind);
887 push @all_bind, @bind;
891 (join $and, @all_sql),
892 $self->_bindtype($k, @all_bind),
896 puke "special op 'between' accepts an arrayref with two values, or a single literal scalarref/arrayref-ref";
900 my $sql = "( $label $op $clause )";
905 sub _where_field_IN {
906 my ($self, $k, $op, $vals) = @_;
908 # backwards compatibility : if scalar, force into an arrayref
909 $vals = [$vals] if defined $vals && ! ref $vals;
911 my ($label) = $self->_convert($self->_quote($k));
912 my ($placeholder) = $self->_convert('?');
913 $op = $self->_sqlcase($op);
915 my ($sql, @bind) = $self->_SWITCH_refkind($vals, {
916 ARRAYREF => sub { # list of choices
917 if (@$vals) { # nonempty list
918 my $placeholders = join ", ", (($placeholder) x @$vals);
919 my $sql = "$label $op ( $placeholders )";
920 my @bind = $self->_bindtype($k, @$vals);
922 return ($sql, @bind);
924 else { # empty list : some databases won't understand "IN ()", so DWIM
925 my $sql = ($op =~ /\bnot\b/i) ? $self->{sqltrue} : $self->{sqlfalse};
930 SCALARREF => sub { # literal SQL
931 my $sql = $self->_open_outer_paren ($$vals);
932 return ("$label $op ( $sql )");
934 ARRAYREFREF => sub { # literal SQL with bind
935 my ($sql, @bind) = @$$vals;
936 $self->_assert_bindval_matches_bindtype(@bind);
937 $sql = $self->_open_outer_paren ($sql);
938 return ("$label $op ( $sql )", @bind);
942 puke "special op 'in' requires an arrayref (or scalarref/arrayref-ref)";
946 return ($sql, @bind);
949 # Some databases (SQLite) treat col IN (1, 2) different from
950 # col IN ( (1, 2) ). Use this to strip all outer parens while
951 # adding them back in the corresponding method
952 sub _open_outer_paren {
953 my ($self, $sql) = @_;
954 $sql = $1 while $sql =~ /^ \s* \( (.*) \) \s* $/xs;
959 #======================================================================
961 #======================================================================
964 my ($self, $arg) = @_;
967 for my $c ($self->_order_by_chunks ($arg) ) {
968 $self->_SWITCH_refkind ($c, {
969 SCALAR => sub { push @sql, $c },
970 ARRAYREF => sub { push @sql, shift @$c; push @bind, @$c },
976 $self->_sqlcase(' order by'),
982 return wantarray ? ($sql, @bind) : $sql;
985 sub _order_by_chunks {
986 my ($self, $arg) = @_;
988 return $self->_SWITCH_refkind($arg, {
991 map { $self->_order_by_chunks ($_ ) } @$arg;
994 ARRAYREFREF => sub { [ @$$arg ] },
996 SCALAR => sub {$self->_quote($arg)},
998 UNDEF => sub {return () },
1000 SCALARREF => sub {$$arg}, # literal SQL, no quoting
1003 # get first pair in hash
1004 my ($key, $val) = each %$arg;
1006 return () unless $key;
1008 if ( (keys %$arg) > 1 or not $key =~ /^-(desc|asc)/i ) {
1009 puke "hash passed to _order_by must have exactly one key (-desc or -asc)";
1015 for my $c ($self->_order_by_chunks ($val)) {
1018 $self->_SWITCH_refkind ($c, {
1023 ($sql, @bind) = @$c;
1027 $sql = $sql . ' ' . $self->_sqlcase($direction);
1029 push @ret, [ $sql, @bind];
1038 #======================================================================
1039 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
1040 #======================================================================
1045 $self->_SWITCH_refkind($from, {
1046 ARRAYREF => sub {join ', ', map { $self->_quote($_) } @$from;},
1047 SCALAR => sub {$self->_quote($from)},
1048 SCALARREF => sub {$$from},
1049 ARRAYREFREF => sub {join ', ', @$from;},
1054 #======================================================================
1056 #======================================================================
1062 $label or puke "can't quote an empty label";
1064 # left and right quote characters
1065 my ($ql, $qr, @other) = $self->_SWITCH_refkind($self->{quote_char}, {
1066 SCALAR => sub {($self->{quote_char}, $self->{quote_char})},
1067 ARRAYREF => sub {@{$self->{quote_char}}},
1071 or puke "quote_char must be an arrayref of 2 values";
1073 # no quoting if no quoting chars
1074 $ql or return $label;
1076 # no quoting for literal SQL
1077 return $$label if ref($label) eq 'SCALAR';
1079 # separate table / column (if applicable)
1080 my $sep = $self->{name_sep} || '';
1081 my @to_quote = $sep ? split /\Q$sep\E/, $label : ($label);
1083 # do the quoting, except for "*" or for `table`.*
1084 my @quoted = map { $_ eq '*' ? $_: $ql.$_.$qr} @to_quote;
1086 # reassemble and return.
1087 return join $sep, @quoted;
1091 # Conversion, if applicable
1093 my ($self, $arg) = @_;
1095 # LDNOTE : modified the previous implementation below because
1096 # it was not consistent : the first "return" is always an array,
1097 # the second "return" is context-dependent. Anyway, _convert
1098 # seems always used with just a single argument, so make it a
1100 # return @_ unless $self->{convert};
1101 # my $conv = $self->_sqlcase($self->{convert});
1102 # my @ret = map { $conv.'('.$_.')' } @_;
1103 # return wantarray ? @ret : $ret[0];
1104 if ($self->{convert}) {
1105 my $conv = $self->_sqlcase($self->{convert});
1106 $arg = $conv.'('.$arg.')';
1114 my($col, @vals) = @_;
1116 #LDNOTE : changed original implementation below because it did not make
1117 # sense when bindtype eq 'columns' and @vals > 1.
1118 # return $self->{bindtype} eq 'columns' ? [ $col, @vals ] : @vals;
1120 return $self->{bindtype} eq 'columns' ? map {[$col, $_]} @vals : @vals;
1123 # Dies if any element of @bind is not in [colname => value] format
1124 # if bindtype is 'columns'.
1125 sub _assert_bindval_matches_bindtype {
1126 my ($self, @bind) = @_;
1128 if ($self->{bindtype} eq 'columns') {
1129 foreach my $val (@bind) {
1130 if (!defined $val || ref($val) ne 'ARRAY' || @$val != 2) {
1131 die "bindtype 'columns' selected, you need to pass: [column_name => bind_value]"
1137 sub _join_sql_clauses {
1138 my ($self, $logic, $clauses_aref, $bind_aref) = @_;
1140 if (@$clauses_aref > 1) {
1141 my $join = " " . $self->_sqlcase($logic) . " ";
1142 my $sql = '( ' . join($join, @$clauses_aref) . ' )';
1143 return ($sql, @$bind_aref);
1145 elsif (@$clauses_aref) {
1146 return ($clauses_aref->[0], @$bind_aref); # no parentheses
1149 return (); # if no SQL, ignore @$bind_aref
1154 # Fix SQL case, if so requested
1158 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
1159 # don't touch the argument ... crooked logic, but let's not change it!
1160 return $self->{case} ? $_[0] : uc($_[0]);
1164 #======================================================================
1165 # DISPATCHING FROM REFKIND
1166 #======================================================================
1169 my ($self, $data) = @_;
1175 # blessed objects are treated like scalars
1176 $ref = (Scalar::Util::blessed $data) ? '' : ref $data;
1177 $n_steps += 1 if $ref;
1178 last if $ref ne 'REF';
1182 my $base = $ref || (defined $data ? 'SCALAR' : 'UNDEF');
1184 return $base . ('REF' x $n_steps);
1190 my ($self, $data) = @_;
1191 my @try = ($self->_refkind($data));
1192 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
1193 push @try, 'FALLBACK';
1197 sub _METHOD_FOR_refkind {
1198 my ($self, $meth_prefix, $data) = @_;
1201 for ($self->_try_refkind($data)) {
1202 $method = $self->can($meth_prefix."_".$_)
1206 return $method || puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
1210 sub _SWITCH_refkind {
1211 my ($self, $data, $dispatch_table) = @_;
1214 for ($self->_try_refkind($data)) {
1215 $coderef = $dispatch_table->{$_}
1219 puke "no dispatch entry for ".$self->_refkind($data)
1228 #======================================================================
1229 # VALUES, GENERATE, AUTOLOAD
1230 #======================================================================
1232 # LDNOTE: original code from nwiger, didn't touch code in that section
1233 # I feel the AUTOLOAD stuff should not be the default, it should
1234 # only be activated on explicit demand by user.
1238 my $data = shift || return;
1239 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
1240 unless ref $data eq 'HASH';
1243 foreach my $k ( sort keys %$data ) {
1244 my $v = $data->{$k};
1245 $self->_SWITCH_refkind($v, {
1247 if ($self->{array_datatypes}) { # array datatype
1248 push @all_bind, $self->_bindtype($k, $v);
1250 else { # literal SQL with bind
1251 my ($sql, @bind) = @$v;
1252 $self->_assert_bindval_matches_bindtype(@bind);
1253 push @all_bind, @bind;
1256 ARRAYREFREF => sub { # literal SQL with bind
1257 my ($sql, @bind) = @${$v};
1258 $self->_assert_bindval_matches_bindtype(@bind);
1259 push @all_bind, @bind;
1261 SCALARREF => sub { # literal SQL without bind
1263 SCALAR_or_UNDEF => sub {
1264 push @all_bind, $self->_bindtype($k, $v);
1275 my(@sql, @sqlq, @sqlv);
1279 if ($ref eq 'HASH') {
1280 for my $k (sort keys %$_) {
1283 my $label = $self->_quote($k);
1284 if ($r eq 'ARRAY') {
1285 # literal SQL with bind
1286 my ($sql, @bind) = @$v;
1287 $self->_assert_bindval_matches_bindtype(@bind);
1288 push @sqlq, "$label = $sql";
1290 } elsif ($r eq 'SCALAR') {
1291 # literal SQL without bind
1292 push @sqlq, "$label = $$v";
1294 push @sqlq, "$label = ?";
1295 push @sqlv, $self->_bindtype($k, $v);
1298 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
1299 } elsif ($ref eq 'ARRAY') {
1300 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
1303 if ($r eq 'ARRAY') { # literal SQL with bind
1304 my ($sql, @bind) = @$v;
1305 $self->_assert_bindval_matches_bindtype(@bind);
1308 } elsif ($r eq 'SCALAR') { # literal SQL without bind
1309 # embedded literal SQL
1316 push @sql, '(' . join(', ', @sqlq) . ')';
1317 } elsif ($ref eq 'SCALAR') {
1321 # strings get case twiddled
1322 push @sql, $self->_sqlcase($_);
1326 my $sql = join ' ', @sql;
1328 # this is pretty tricky
1329 # if ask for an array, return ($stmt, @bind)
1330 # otherwise, s/?/shift @sqlv/ to put it inline
1332 return ($sql, @sqlv);
1334 1 while $sql =~ s/\?/my $d = shift(@sqlv);
1335 ref $d ? $d->[1] : $d/e;
1344 # This allows us to check for a local, then _form, attr
1346 my($name) = $AUTOLOAD =~ /.*::(.+)/;
1347 return $self->generate($name, @_);
1358 SQL::Abstract - Generate SQL from Perl data structures
1364 my $sql = SQL::Abstract->new;
1366 my($stmt, @bind) = $sql->select($table, \@fields, \%where, \@order);
1368 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
1370 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
1372 my($stmt, @bind) = $sql->delete($table, \%where);
1374 # Then, use these in your DBI statements
1375 my $sth = $dbh->prepare($stmt);
1376 $sth->execute(@bind);
1378 # Just generate the WHERE clause
1379 my($stmt, @bind) = $sql->where(\%where, \@order);
1381 # Return values in the same order, for hashed queries
1382 # See PERFORMANCE section for more details
1383 my @bind = $sql->values(\%fieldvals);
1387 This module was inspired by the excellent L<DBIx::Abstract>.
1388 However, in using that module I found that what I really wanted
1389 to do was generate SQL, but still retain complete control over my
1390 statement handles and use the DBI interface. So, I set out to
1391 create an abstract SQL generation module.
1393 While based on the concepts used by L<DBIx::Abstract>, there are
1394 several important differences, especially when it comes to WHERE
1395 clauses. I have modified the concepts used to make the SQL easier
1396 to generate from Perl data structures and, IMO, more intuitive.
1397 The underlying idea is for this module to do what you mean, based
1398 on the data structures you provide it. The big advantage is that
1399 you don't have to modify your code every time your data changes,
1400 as this module figures it out.
1402 To begin with, an SQL INSERT is as easy as just specifying a hash
1403 of C<key=value> pairs:
1406 name => 'Jimbo Bobson',
1407 phone => '123-456-7890',
1408 address => '42 Sister Lane',
1409 city => 'St. Louis',
1410 state => 'Louisiana',
1413 The SQL can then be generated with this:
1415 my($stmt, @bind) = $sql->insert('people', \%data);
1417 Which would give you something like this:
1419 $stmt = "INSERT INTO people
1420 (address, city, name, phone, state)
1421 VALUES (?, ?, ?, ?, ?)";
1422 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
1423 '123-456-7890', 'Louisiana');
1425 These are then used directly in your DBI code:
1427 my $sth = $dbh->prepare($stmt);
1428 $sth->execute(@bind);
1430 =head2 Inserting and Updating Arrays
1432 If your database has array types (like for example Postgres),
1433 activate the special option C<< array_datatypes => 1 >>
1434 when creating the C<SQL::Abstract> object.
1435 Then you may use an arrayref to insert and update database array types:
1437 my $sql = SQL::Abstract->new(array_datatypes => 1);
1439 planets => [qw/Mercury Venus Earth Mars/]
1442 my($stmt, @bind) = $sql->insert('solar_system', \%data);
1446 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
1448 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
1451 =head2 Inserting and Updating SQL
1453 In order to apply SQL functions to elements of your C<%data> you may
1454 specify a reference to an arrayref for the given hash value. For example,
1455 if you need to execute the Oracle C<to_date> function on a value, you can
1456 say something like this:
1460 date_entered => \["to_date(?,'MM/DD/YYYY')", "03/02/2003"],
1463 The first value in the array is the actual SQL. Any other values are
1464 optional and would be included in the bind values array. This gives
1467 my($stmt, @bind) = $sql->insert('people', \%data);
1469 $stmt = "INSERT INTO people (name, date_entered)
1470 VALUES (?, to_date(?,'MM/DD/YYYY'))";
1471 @bind = ('Bill', '03/02/2003');
1473 An UPDATE is just as easy, all you change is the name of the function:
1475 my($stmt, @bind) = $sql->update('people', \%data);
1477 Notice that your C<%data> isn't touched; the module will generate
1478 the appropriately quirky SQL for you automatically. Usually you'll
1479 want to specify a WHERE clause for your UPDATE, though, which is
1480 where handling C<%where> hashes comes in handy...
1482 =head2 Complex where statements
1484 This module can generate pretty complicated WHERE statements
1485 easily. For example, simple C<key=value> pairs are taken to mean
1486 equality, and if you want to see if a field is within a set
1487 of values, you can use an arrayref. Let's say we wanted to
1488 SELECT some data based on this criteria:
1491 requestor => 'inna',
1492 worker => ['nwiger', 'rcwe', 'sfz'],
1493 status => { '!=', 'completed' }
1496 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
1498 The above would give you something like this:
1500 $stmt = "SELECT * FROM tickets WHERE
1501 ( requestor = ? ) AND ( status != ? )
1502 AND ( worker = ? OR worker = ? OR worker = ? )";
1503 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
1505 Which you could then use in DBI code like so:
1507 my $sth = $dbh->prepare($stmt);
1508 $sth->execute(@bind);
1514 The functions are simple. There's one for each major SQL operation,
1515 and a constructor you use first. The arguments are specified in a
1516 similar order to each function (table, then fields, then a where
1517 clause) to try and simplify things.
1522 =head2 new(option => 'value')
1524 The C<new()> function takes a list of options and values, and returns
1525 a new B<SQL::Abstract> object which can then be used to generate SQL
1526 through the methods below. The options accepted are:
1532 If set to 'lower', then SQL will be generated in all lowercase. By
1533 default SQL is generated in "textbook" case meaning something like:
1535 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
1537 Any setting other than 'lower' is ignored.
1541 This determines what the default comparison operator is. By default
1542 it is C<=>, meaning that a hash like this:
1544 %where = (name => 'nwiger', email => 'nate@wiger.org');
1546 Will generate SQL like this:
1548 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
1550 However, you may want loose comparisons by default, so if you set
1551 C<cmp> to C<like> you would get SQL such as:
1553 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
1555 You can also override the comparsion on an individual basis - see
1556 the huge section on L</"WHERE CLAUSES"> at the bottom.
1558 =item sqltrue, sqlfalse
1560 Expressions for inserting boolean values within SQL statements.
1561 By default these are C<1=1> and C<1=0>. They are used
1562 by the special operators C<-in> and C<-not_in> for generating
1563 correct SQL even when the argument is an empty array (see below).
1567 This determines the default logical operator for multiple WHERE
1568 statements in arrays or hashes. If absent, the default logic is "or"
1569 for arrays, and "and" for hashes. This means that a WHERE
1573 event_date => {'>=', '2/13/99'},
1574 event_date => {'<=', '4/24/03'},
1577 will generate SQL like this:
1579 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
1581 This is probably not what you want given this query, though (look
1582 at the dates). To change the "OR" to an "AND", simply specify:
1584 my $sql = SQL::Abstract->new(logic => 'and');
1586 Which will change the above C<WHERE> to:
1588 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
1590 The logic can also be changed locally by inserting
1591 a modifier in front of an arrayref :
1593 @where = (-and => [event_date => {'>=', '2/13/99'},
1594 event_date => {'<=', '4/24/03'} ]);
1596 See the L</"WHERE CLAUSES"> section for explanations.
1600 This will automatically convert comparisons using the specified SQL
1601 function for both column and value. This is mostly used with an argument
1602 of C<upper> or C<lower>, so that the SQL will have the effect of
1603 case-insensitive "searches". For example, this:
1605 $sql = SQL::Abstract->new(convert => 'upper');
1606 %where = (keywords => 'MaKe iT CAse inSeNSItive');
1608 Will turn out the following SQL:
1610 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
1612 The conversion can be C<upper()>, C<lower()>, or any other SQL function
1613 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
1614 not validate this option; it will just pass through what you specify verbatim).
1618 This is a kludge because many databases suck. For example, you can't
1619 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
1620 Instead, you have to use C<bind_param()>:
1622 $sth->bind_param(1, 'reg data');
1623 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
1625 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
1626 which loses track of which field each slot refers to. Fear not.
1628 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
1629 Currently, you can specify either C<normal> (default) or C<columns>. If you
1630 specify C<columns>, you will get an array that looks like this:
1632 my $sql = SQL::Abstract->new(bindtype => 'columns');
1633 my($stmt, @bind) = $sql->insert(...);
1636 [ 'column1', 'value1' ],
1637 [ 'column2', 'value2' ],
1638 [ 'column3', 'value3' ],
1641 You can then iterate through this manually, using DBI's C<bind_param()>.
1643 $sth->prepare($stmt);
1646 my($col, $data) = @$_;
1647 if ($col eq 'details' || $col eq 'comments') {
1648 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
1649 } elsif ($col eq 'image') {
1650 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
1652 $sth->bind_param($i, $data);
1656 $sth->execute; # execute without @bind now
1658 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
1659 Basically, the advantage is still that you don't have to care which fields
1660 are or are not included. You could wrap that above C<for> loop in a simple
1661 sub called C<bind_fields()> or something and reuse it repeatedly. You still
1662 get a layer of abstraction over manual SQL specification.
1664 Note that if you set L</bindtype> to C<columns>, the C<\[$sql, @bind]>
1665 construct (see L</Literal SQL with placeholders and bind values (subqueries)>)
1666 will expect the bind values in this format.
1670 This is the character that a table or column name will be quoted
1671 with. By default this is an empty string, but you could set it to
1672 the character C<`>, to generate SQL like this:
1674 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
1676 Alternatively, you can supply an array ref of two items, the first being the left
1677 hand quote character, and the second the right hand quote character. For
1678 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
1679 that generates SQL like this:
1681 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
1683 Quoting is useful if you have tables or columns names that are reserved
1684 words in your database's SQL dialect.
1688 This is the character that separates a table and column name. It is
1689 necessary to specify this when the C<quote_char> option is selected,
1690 so that tables and column names can be individually quoted like this:
1692 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
1694 =item array_datatypes
1696 When this option is true, arrayrefs in INSERT or UPDATE are
1697 interpreted as array datatypes and are passed directly
1699 When this option is false, arrayrefs are interpreted
1700 as literal SQL, just like refs to arrayrefs
1701 (but this behavior is for backwards compatibility; when writing
1702 new queries, use the "reference to arrayref" syntax
1708 Takes a reference to a list of "special operators"
1709 to extend the syntax understood by L<SQL::Abstract>.
1710 See section L</"SPECIAL OPERATORS"> for details.
1714 Takes a reference to a list of "unary operators"
1715 to extend the syntax understood by L<SQL::Abstract>.
1716 See section L</"UNARY OPERATORS"> for details.
1722 =head2 insert($table, \@values || \%fieldvals, \%options)
1724 This is the simplest function. You simply give it a table name
1725 and either an arrayref of values or hashref of field/value pairs.
1726 It returns an SQL INSERT statement and a list of bind values.
1727 See the sections on L</"Inserting and Updating Arrays"> and
1728 L</"Inserting and Updating SQL"> for information on how to insert
1729 with those data types.
1731 The optional C<\%options> hash reference may contain additional
1732 options to generate the insert SQL. Currently supported options
1739 Takes either a scalar of raw SQL fields, or an array reference of
1740 field names, and adds on an SQL C<RETURNING> statement at the end.
1741 This allows you to return data generated by the insert statement
1742 (such as row IDs) without performing another C<SELECT> statement.
1743 Note, however, this is not part of the SQL standard and may not
1744 be supported by all database engines.
1748 =head2 update($table, \%fieldvals, \%where)
1750 This takes a table, hashref of field/value pairs, and an optional
1751 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
1753 See the sections on L</"Inserting and Updating Arrays"> and
1754 L</"Inserting and Updating SQL"> for information on how to insert
1755 with those data types.
1757 =head2 select($source, $fields, $where, $order)
1759 This returns a SQL SELECT statement and associated list of bind values, as
1760 specified by the arguments :
1766 Specification of the 'FROM' part of the statement.
1767 The argument can be either a plain scalar (interpreted as a table
1768 name, will be quoted), or an arrayref (interpreted as a list
1769 of table names, joined by commas, quoted), or a scalarref
1770 (literal table name, not quoted), or a ref to an arrayref
1771 (list of literal table names, joined by commas, not quoted).
1775 Specification of the list of fields to retrieve from
1777 The argument can be either an arrayref (interpreted as a list
1778 of field names, will be joined by commas and quoted), or a
1779 plain scalar (literal SQL, not quoted).
1780 Please observe that this API is not as flexible as for
1781 the first argument C<$table>, for backwards compatibility reasons.
1785 Optional argument to specify the WHERE part of the query.
1786 The argument is most often a hashref, but can also be
1787 an arrayref or plain scalar --
1788 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
1792 Optional argument to specify the ORDER BY part of the query.
1793 The argument can be a scalar, a hashref or an arrayref
1794 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
1800 =head2 delete($table, \%where)
1802 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
1803 It returns an SQL DELETE statement and list of bind values.
1805 =head2 where(\%where, \@order)
1807 This is used to generate just the WHERE clause. For example,
1808 if you have an arbitrary data structure and know what the
1809 rest of your SQL is going to look like, but want an easy way
1810 to produce a WHERE clause, use this. It returns an SQL WHERE
1811 clause and list of bind values.
1814 =head2 values(\%data)
1816 This just returns the values from the hash C<%data>, in the same
1817 order that would be returned from any of the other above queries.
1818 Using this allows you to markedly speed up your queries if you
1819 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
1821 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
1823 Warning: This is an experimental method and subject to change.
1825 This returns arbitrarily generated SQL. It's a really basic shortcut.
1826 It will return two different things, depending on return context:
1828 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
1829 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
1831 These would return the following:
1833 # First calling form
1834 $stmt = "CREATE TABLE test (?, ?)";
1835 @bind = (field1, field2);
1837 # Second calling form
1838 $stmt_and_val = "CREATE TABLE test (field1, field2)";
1840 Depending on what you're trying to do, it's up to you to choose the correct
1841 format. In this example, the second form is what you would want.
1845 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
1849 ALTER SESSION SET nls_date_format = 'MM/YY'
1851 You get the idea. Strings get their case twiddled, but everything
1852 else remains verbatim.
1857 =head1 WHERE CLAUSES
1861 This module uses a variation on the idea from L<DBIx::Abstract>. It
1862 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
1863 module is that things in arrays are OR'ed, and things in hashes
1866 The easiest way to explain is to show lots of examples. After
1867 each C<%where> hash shown, it is assumed you used:
1869 my($stmt, @bind) = $sql->where(\%where);
1871 However, note that the C<%where> hash can be used directly in any
1872 of the other functions as well, as described above.
1874 =head2 Key-value pairs
1876 So, let's get started. To begin, a simple hash:
1880 status => 'completed'
1883 Is converted to SQL C<key = val> statements:
1885 $stmt = "WHERE user = ? AND status = ?";
1886 @bind = ('nwiger', 'completed');
1888 One common thing I end up doing is having a list of values that
1889 a field can be in. To do this, simply specify a list inside of
1894 status => ['assigned', 'in-progress', 'pending'];
1897 This simple code will create the following:
1899 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
1900 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
1902 A field associated to an empty arrayref will be considered a
1903 logical false and will generate 0=1.
1905 =head2 Specific comparison operators
1907 If you want to specify a different type of operator for your comparison,
1908 you can use a hashref for a given column:
1912 status => { '!=', 'completed' }
1915 Which would generate:
1917 $stmt = "WHERE user = ? AND status != ?";
1918 @bind = ('nwiger', 'completed');
1920 To test against multiple values, just enclose the values in an arrayref:
1922 status => { '=', ['assigned', 'in-progress', 'pending'] };
1924 Which would give you:
1926 "WHERE status = ? OR status = ? OR status = ?"
1929 The hashref can also contain multiple pairs, in which case it is expanded
1930 into an C<AND> of its elements:
1934 status => { '!=', 'completed', -not_like => 'pending%' }
1937 # Or more dynamically, like from a form
1938 $where{user} = 'nwiger';
1939 $where{status}{'!='} = 'completed';
1940 $where{status}{'-not_like'} = 'pending%';
1942 # Both generate this
1943 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
1944 @bind = ('nwiger', 'completed', 'pending%');
1947 To get an OR instead, you can combine it with the arrayref idea:
1951 priority => [ {'=', 2}, {'!=', 1} ]
1954 Which would generate:
1956 $stmt = "WHERE user = ? AND priority = ? OR priority != ?";
1957 @bind = ('nwiger', '2', '1');
1959 If you want to include literal SQL (with or without bind values), just use a
1960 scalar reference or array reference as the value:
1963 date_entered => { '>' => \["to_date(?, 'MM/DD/YYYY')", "11/26/2008"] },
1964 date_expires => { '<' => \"now()" }
1967 Which would generate:
1969 $stmt = "WHERE date_entered > "to_date(?, 'MM/DD/YYYY') AND date_expires < now()";
1970 @bind = ('11/26/2008');
1973 =head2 Logic and nesting operators
1975 In the example above,
1976 there is a subtle trap if you want to say something like
1977 this (notice the C<AND>):
1979 WHERE priority != ? AND priority != ?
1981 Because, in Perl you I<can't> do this:
1983 priority => { '!=', 2, '!=', 1 }
1985 As the second C<!=> key will obliterate the first. The solution
1986 is to use the special C<-modifier> form inside an arrayref:
1988 priority => [ -and => {'!=', 2},
1992 Normally, these would be joined by C<OR>, but the modifier tells it
1993 to use C<AND> instead. (Hint: You can use this in conjunction with the
1994 C<logic> option to C<new()> in order to change the way your queries
1995 work by default.) B<Important:> Note that the C<-modifier> goes
1996 B<INSIDE> the arrayref, as an extra first element. This will
1997 B<NOT> do what you think it might:
1999 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
2001 Here is a quick list of equivalencies, since there is some overlap:
2004 status => {'!=', 'completed', 'not like', 'pending%' }
2005 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
2008 status => {'=', ['assigned', 'in-progress']}
2009 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
2010 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
2014 =head2 Special operators : IN, BETWEEN, etc.
2016 You can also use the hashref format to compare a list of fields using the
2017 C<IN> comparison operator, by specifying the list as an arrayref:
2020 status => 'completed',
2021 reportid => { -in => [567, 2335, 2] }
2024 Which would generate:
2026 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
2027 @bind = ('completed', '567', '2335', '2');
2029 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
2032 If the argument to C<-in> is an empty array, 'sqlfalse' is generated
2033 (by default : C<1=0>). Similarly, C<< -not_in => [] >> generates
2034 'sqltrue' (by default : C<1=1>).
2036 In addition to the array you can supply a chunk of literal sql or
2037 literal sql with bind:
2040 customer => { -in => \[
2041 'SELECT cust_id FROM cust WHERE balance > ?',
2044 status => { -in => \'SELECT status_codes FROM states' },
2050 customer IN ( SELECT cust_id FROM cust WHERE balance > ? )
2051 AND status IN ( SELECT status_codes FROM states )
2057 Another pair of operators is C<-between> and C<-not_between>,
2058 used with an arrayref of two values:
2062 completion_date => {
2063 -not_between => ['2002-10-01', '2003-02-06']
2069 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
2071 Just like with C<-in> all plausible combinations of literal SQL
2075 start0 => { -between => [ 1, 2 ] },
2076 start1 => { -between => \["? AND ?", 1, 2] },
2077 start2 => { -between => \"lower(x) AND upper(y)" },
2078 start3 => { -between => [
2080 \["upper(?)", 'stuff' ],
2087 ( start0 BETWEEN ? AND ? )
2088 AND ( start1 BETWEEN ? AND ? )
2089 AND ( start2 BETWEEN lower(x) AND upper(y) )
2090 AND ( start3 BETWEEN lower(x) AND upper(?) )
2092 @bind = (1, 2, 1, 2, 'stuff');
2095 These are the two builtin "special operators"; but the
2096 list can be expanded : see section L</"SPECIAL OPERATORS"> below.
2098 =head2 Unary operators: bool
2100 If you wish to test against boolean columns or functions within your
2101 database you can use the C<-bool> and C<-not_bool> operators. For
2102 example to test the column C<is_user> being true and the column
2103 <is_enabled> being false you would use:-
2107 -not_bool => 'is_enabled',
2112 WHERE is_user AND NOT is_enabled
2114 If a more complex combination is required, testing more conditions,
2115 then you should use the and/or operators:-
2122 -not_bool => 'four',
2128 WHERE one AND two AND three AND NOT four
2131 =head2 Nested conditions, -and/-or prefixes
2133 So far, we've seen how multiple conditions are joined with a top-level
2134 C<AND>. We can change this by putting the different conditions we want in
2135 hashes and then putting those hashes in an array. For example:
2140 status => { -like => ['pending%', 'dispatched'] },
2144 status => 'unassigned',
2148 This data structure would create the following:
2150 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
2151 OR ( user = ? AND status = ? ) )";
2152 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
2155 There is also a special C<-nest>
2156 operator which adds an additional set of parens, to create a subquery.
2157 For example, to get something like this:
2159 $stmt = "WHERE user = ? AND ( workhrs > ? OR geo = ? )";
2160 @bind = ('nwiger', '20', 'ASIA');
2166 -nest => [ workhrs => {'>', 20}, geo => 'ASIA' ],
2170 Finally, clauses in hashrefs or arrayrefs can be
2171 prefixed with an C<-and> or C<-or> to change the logic
2178 -and => [workhrs => {'>', 20}, geo => 'ASIA' ],
2179 -and => [workhrs => {'<', 50}, geo => 'EURO' ]
2186 WHERE ( user = ? AND
2187 ( ( workhrs > ? AND geo = ? )
2188 OR ( workhrs < ? AND geo = ? ) ) )
2191 =head2 Algebraic inconsistency, for historical reasons
2193 C<Important note>: when connecting several conditions, the C<-and->|C<-or>
2194 operator goes C<outside> of the nested structure; whereas when connecting
2195 several constraints on one column, the C<-and> operator goes
2196 C<inside> the arrayref. Here is an example combining both features :
2199 -and => [a => 1, b => 2],
2200 -or => [c => 3, d => 4],
2201 e => [-and => {-like => 'foo%'}, {-like => '%bar'} ]
2206 WHERE ( ( ( a = ? AND b = ? )
2207 OR ( c = ? OR d = ? )
2208 OR ( e LIKE ? AND e LIKE ? ) ) )
2210 This difference in syntax is unfortunate but must be preserved for
2211 historical reasons. So be careful : the two examples below would
2212 seem algebraically equivalent, but they are not
2214 {col => [-and => {-like => 'foo%'}, {-like => '%bar'}]}
2215 # yields : WHERE ( ( col LIKE ? AND col LIKE ? ) )
2217 [-and => {col => {-like => 'foo%'}, {col => {-like => '%bar'}}]]
2218 # yields : WHERE ( ( col LIKE ? OR col LIKE ? ) )
2223 Finally, sometimes only literal SQL will do. If you want to include
2224 literal SQL verbatim, you can specify it as a scalar reference, namely:
2226 my $inn = 'is Not Null';
2228 priority => { '<', 2 },
2234 $stmt = "WHERE priority < ? AND requestor is Not Null";
2237 Note that in this example, you only get one bind parameter back, since
2238 the verbatim SQL is passed as part of the statement.
2240 Of course, just to prove a point, the above can also be accomplished
2244 priority => { '<', 2 },
2245 requestor => { '!=', undef },
2251 Conditions on boolean columns can be expressed in the same way, passing
2252 a reference to an empty string, however using liternal SQL in this way
2253 is deprecated - the preferred method is to use the boolean operators -
2254 see L</"Unary operators: bool"> :
2257 priority => { '<', 2 },
2263 $stmt = "WHERE priority < ? AND is_ready";
2267 =head2 Literal SQL with placeholders and bind values (subqueries)
2269 If the literal SQL to be inserted has placeholders and bind values,
2270 use a reference to an arrayref (yes this is a double reference --
2271 not so common, but perfectly legal Perl). For example, to find a date
2272 in Postgres you can use something like this:
2275 date_column => \[q/= date '2008-09-30' - ?::integer/, 10/]
2280 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
2283 Note that you must pass the bind values in the same format as they are returned
2284 by L</where>. That means that if you set L</bindtype> to C<columns>, you must
2285 provide the bind values in the C<< [ column_meta => value ] >> format, where
2286 C<column_meta> is an opaque scalar value; most commonly the column name, but
2287 you can use any scalar value (including references and blessed references),
2288 L<SQL::Abstract> will simply pass it through intact. So if C<bindtype> is set
2289 to C<columns> the above example will look like:
2292 date_column => \[q/= date '2008-09-30' - ?::integer/, [ dummy => 10 ]/]
2295 Literal SQL is especially useful for nesting parenthesized clauses in the
2296 main SQL query. Here is a first example :
2298 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
2302 bar => \["IN ($sub_stmt)" => @sub_bind],
2307 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
2308 WHERE c2 < ? AND c3 LIKE ?))";
2309 @bind = (1234, 100, "foo%");
2311 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
2312 are expressed in the same way. Of course the C<$sub_stmt> and
2313 its associated bind values can be generated through a former call
2316 my ($sub_stmt, @sub_bind)
2317 = $sql->select("t1", "c1", {c2 => {"<" => 100},
2318 c3 => {-like => "foo%"}});
2321 bar => \["> ALL ($sub_stmt)" => @sub_bind],
2324 In the examples above, the subquery was used as an operator on a column;
2325 but the same principle also applies for a clause within the main C<%where>
2326 hash, like an EXISTS subquery :
2328 my ($sub_stmt, @sub_bind)
2329 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
2332 -nest => \["EXISTS ($sub_stmt)" => @sub_bind],
2337 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
2338 WHERE c1 = ? AND c2 > t0.c0))";
2342 Observe that the condition on C<c2> in the subquery refers to
2343 column C<t0.c0> of the main query : this is I<not> a bind
2344 value, so we have to express it through a scalar ref.
2345 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
2346 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
2347 what we wanted here.
2349 Another use of the subquery technique is when some SQL clauses need
2350 parentheses, as it often occurs with some proprietary SQL extensions
2351 like for example fulltext expressions, geospatial expressions,
2352 NATIVE clauses, etc. Here is an example of a fulltext query in MySQL :
2355 -nest => \["MATCH (col1, col2) AGAINST (?)" => qw/apples/]
2358 Finally, here is an example where a subquery is used
2359 for expressing unary negation:
2361 my ($sub_stmt, @sub_bind)
2362 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
2363 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
2365 lname => {like => '%son%'},
2366 -nest => \["NOT ($sub_stmt)" => @sub_bind],
2371 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
2372 @bind = ('%son%', 10, 20)
2378 These pages could go on for a while, since the nesting of the data
2379 structures this module can handle are pretty much unlimited (the
2380 module implements the C<WHERE> expansion as a recursive function
2381 internally). Your best bet is to "play around" with the module a
2382 little to see how the data structures behave, and choose the best
2383 format for your data based on that.
2385 And of course, all the values above will probably be replaced with
2386 variables gotten from forms or the command line. After all, if you
2387 knew everything ahead of time, you wouldn't have to worry about
2388 dynamically-generating SQL and could just hardwire it into your
2394 =head1 ORDER BY CLAUSES
2396 Some functions take an order by clause. This can either be a scalar (just a
2397 column name,) a hash of C<< { -desc => 'col' } >> or C<< { -asc => 'col' } >>,
2398 or an array of either of the two previous forms. Examples:
2400 Given | Will Generate
2401 ----------------------------------------------------------
2403 \'colA DESC' | ORDER BY colA DESC
2405 'colA' | ORDER BY colA
2407 [qw/colA colB/] | ORDER BY colA, colB
2409 {-asc => 'colA'} | ORDER BY colA ASC
2411 {-desc => 'colB'} | ORDER BY colB DESC
2413 ['colA', {-asc => 'colB'}] | ORDER BY colA, colB ASC
2415 { -asc => [qw/colA colB] } | ORDER BY colA ASC, colB ASC
2418 { -asc => 'colA' }, | ORDER BY colA ASC, colB DESC,
2419 { -desc => [qw/colB/], | colC ASC, colD ASC
2420 { -asc => [qw/colC colD/],|
2422 ===========================================================
2426 =head1 SPECIAL OPERATORS
2428 my $sqlmaker = SQL::Abstract->new(special_ops => [
2432 my ($self, $field, $op, $arg) = @_;
2438 handler => 'method_name',
2442 A "special operator" is a SQL syntactic clause that can be
2443 applied to a field, instead of a usual binary operator.
2446 WHERE field IN (?, ?, ?)
2447 WHERE field BETWEEN ? AND ?
2448 WHERE MATCH(field) AGAINST (?, ?)
2450 Special operators IN and BETWEEN are fairly standard and therefore
2451 are builtin within C<SQL::Abstract> (as the overridable methods
2452 C<_where_field_IN> and C<_where_field_BETWEEN>). For other operators,
2453 like the MATCH .. AGAINST example above which is specific to MySQL,
2454 you can write your own operator handlers - supply a C<special_ops>
2455 argument to the C<new> method. That argument takes an arrayref of
2456 operator definitions; each operator definition is a hashref with two
2463 the regular expression to match the operator
2467 Either a coderef or a plain scalar method name. In both cases
2468 the expected return is C<< ($sql, @bind) >>.
2470 When supplied with a method name, it is simply called on the
2471 L<SQL::Abstract/> object as:
2473 $self->$method_name ($field, $op, $arg)
2477 $op is the part that matched the handler regex
2478 $field is the LHS of the operator
2481 When supplied with a coderef, it is called as:
2483 $coderef->($self, $field, $op, $arg)
2488 For example, here is an implementation
2489 of the MATCH .. AGAINST syntax for MySQL
2491 my $sqlmaker = SQL::Abstract->new(special_ops => [
2493 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
2494 {regex => qr/^match$/i,
2496 my ($self, $field, $op, $arg) = @_;
2497 $arg = [$arg] if not ref $arg;
2498 my $label = $self->_quote($field);
2499 my ($placeholder) = $self->_convert('?');
2500 my $placeholders = join ", ", (($placeholder) x @$arg);
2501 my $sql = $self->_sqlcase('match') . " ($label) "
2502 . $self->_sqlcase('against') . " ($placeholders) ";
2503 my @bind = $self->_bindtype($field, @$arg);
2504 return ($sql, @bind);
2511 =head1 UNARY OPERATORS
2513 my $sqlmaker = SQL::Abstract->new(unary_ops => [
2517 my ($self, $op, $arg) = @_;
2523 handler => 'method_name',
2527 A "unary operator" is a SQL syntactic clause that can be
2528 applied to a field - the operator goes before the field
2530 You can write your own operator handlers - supply a C<unary_ops>
2531 argument to the C<new> method. That argument takes an arrayref of
2532 operator definitions; each operator definition is a hashref with two
2539 the regular expression to match the operator
2543 Either a coderef or a plain scalar method name. In both cases
2544 the expected return is C<< $sql >>.
2546 When supplied with a method name, it is simply called on the
2547 L<SQL::Abstract/> object as:
2549 $self->$method_name ($op, $arg)
2553 $op is the part that matched the handler regex
2554 $arg is the RHS or argument of the operator
2556 When supplied with a coderef, it is called as:
2558 $coderef->($self, $op, $arg)
2566 Thanks to some benchmarking by Mark Stosberg, it turns out that
2567 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
2568 I must admit this wasn't an intentional design issue, but it's a
2569 byproduct of the fact that you get to control your C<DBI> handles
2572 To maximize performance, use a code snippet like the following:
2574 # prepare a statement handle using the first row
2575 # and then reuse it for the rest of the rows
2577 for my $href (@array_of_hashrefs) {
2578 $stmt ||= $sql->insert('table', $href);
2579 $sth ||= $dbh->prepare($stmt);
2580 $sth->execute($sql->values($href));
2583 The reason this works is because the keys in your C<$href> are sorted
2584 internally by B<SQL::Abstract>. Thus, as long as your data retains
2585 the same structure, you only have to generate the SQL the first time
2586 around. On subsequent queries, simply use the C<values> function provided
2587 by this module to return your values in the correct order.
2592 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
2593 really like this part (I do, at least). Building up a complex query
2594 can be as simple as the following:
2598 use CGI::FormBuilder;
2601 my $form = CGI::FormBuilder->new(...);
2602 my $sql = SQL::Abstract->new;
2604 if ($form->submitted) {
2605 my $field = $form->field;
2606 my $id = delete $field->{id};
2607 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
2610 Of course, you would still have to connect using C<DBI> to run the
2611 query, but the point is that if you make your form look like your
2612 table, the actual query script can be extremely simplistic.
2614 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
2615 a fast interface to returning and formatting data. I frequently
2616 use these three modules together to write complex database query
2617 apps in under 50 lines.
2622 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
2623 Great care has been taken to preserve the I<published> behavior
2624 documented in previous versions in the 1.* family; however,
2625 some features that were previously undocumented, or behaved
2626 differently from the documentation, had to be changed in order
2627 to clarify the semantics. Hence, client code that was relying
2628 on some dark areas of C<SQL::Abstract> v1.*
2629 B<might behave differently> in v1.50.
2631 The main changes are :
2637 support for literal SQL through the C<< \ [$sql, bind] >> syntax.
2641 support for the { operator => \"..." } construct (to embed literal SQL)
2645 support for the { operator => \["...", @bind] } construct (to embed literal SQL with bind values)
2649 optional support for L<array datatypes|/"Inserting and Updating Arrays">
2653 defensive programming : check arguments
2657 fixed bug with global logic, which was previously implemented
2658 through global variables yielding side-effects. Prior versions would
2659 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
2660 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
2661 Now this is interpreted
2662 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
2667 fixed semantics of _bindtype on array args
2671 dropped the C<_anoncopy> of the %where tree. No longer necessary,
2672 we just avoid shifting arrays within that tree.
2676 dropped the C<_modlogic> function
2682 =head1 ACKNOWLEDGEMENTS
2684 There are a number of individuals that have really helped out with
2685 this module. Unfortunately, most of them submitted bugs via CPAN
2686 so I have no idea who they are! But the people I do know are:
2688 Ash Berlin (order_by hash term support)
2689 Matt Trout (DBIx::Class support)
2690 Mark Stosberg (benchmarking)
2691 Chas Owens (initial "IN" operator support)
2692 Philip Collins (per-field SQL functions)
2693 Eric Kolve (hashref "AND" support)
2694 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
2695 Dan Kubb (support for "quote_char" and "name_sep")
2696 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
2697 Laurent Dami (internal refactoring, multiple -nest, extensible list of special operators, literal SQL)
2698 Norbert Buchmuller (support for literal SQL in hashpair, misc. fixes & tests)
2699 Peter Rabbitson (rewrite of SQLA::Test, misc. fixes & tests)
2700 Oliver Charles (support for "RETURNING" after "INSERT")
2706 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
2710 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
2712 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
2714 For support, your best bet is to try the C<DBIx::Class> users mailing list.
2715 While not an official support venue, C<DBIx::Class> makes heavy use of
2716 C<SQL::Abstract>, and as such list members there are very familiar with
2717 how to create queries.
2721 This module is free software; you may copy this under the same
2722 terms as perl itself (either the GNU General Public License or
2723 the Artistic License)