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)
11 use List::Util qw/first/;
12 use Scalar::Util qw/blessed/;
14 #======================================================================
16 #======================================================================
18 our $VERSION = '1.49_01';
19 $VERSION = eval $VERSION; # numify for warning-free dev releases
24 # special operators (-in, -between). May be extended/overridden by user.
25 # See section WHERE: BUILTIN SPECIAL OPERATORS below for implementation
26 my @BUILTIN_SPECIAL_OPS = (
27 {regex => qr/^(not )?between$/i, handler => \&_where_field_BETWEEN},
28 {regex => qr/^(not )?in$/i, handler => \&_where_field_IN},
31 #======================================================================
32 # DEBUGGING AND ERROR REPORTING
33 #======================================================================
36 return unless $_[0]->{debug}; shift; # a little faster
37 my $func = (caller(1))[3];
38 warn "[$func] ", @_, "\n";
42 my($func) = (caller(1))[3];
43 carp "[$func] Warning: ", @_;
47 my($func) = (caller(1))[3];
48 croak "[$func] Fatal: ", @_;
52 #======================================================================
54 #======================================================================
58 my $class = ref($self) || $self;
59 my %opt = (ref $_[0] eq 'HASH') ? %{$_[0]} : @_;
61 # choose our case by keeping an option around
62 delete $opt{case} if $opt{case} && $opt{case} ne 'lower';
64 # default logic for interpreting arrayrefs
65 $opt{logic} = uc $opt{logic} || 'OR';
67 # how to return bind vars
68 # LDNOTE: changed nwiger code : why this 'delete' ??
69 # $opt{bindtype} ||= delete($opt{bind_type}) || 'normal';
70 $opt{bindtype} ||= 'normal';
72 # default comparison is "=", but can be overridden
75 # try to recognize which are the 'equality' and 'unequality' ops
76 # (temporary quickfix, should go through a more seasoned API)
77 $opt{equality_op} = qr/^(\Q$opt{cmp}\E|is|(is\s+)?like)$/i;
78 $opt{inequality_op} = qr/^(!=|<>|(is\s+)?not(\s+like)?)$/i;
81 $opt{sqltrue} ||= '1=1';
82 $opt{sqlfalse} ||= '0=1';
85 $opt{special_ops} ||= [];
86 push @{$opt{special_ops}}, @BUILTIN_SPECIAL_OPS;
88 return bless \%opt, $class;
93 #======================================================================
95 #======================================================================
99 my $table = $self->_table(shift);
100 my $data = shift || return;
102 my $method = $self->_METHOD_FOR_refkind("_insert", $data);
103 my ($sql, @bind) = $self->$method($data);
104 $sql = join " ", $self->_sqlcase('insert into'), $table, $sql;
105 return wantarray ? ($sql, @bind) : $sql;
108 sub _insert_HASHREF { # explicit list of fields and then values
109 my ($self, $data) = @_;
111 my @fields = sort keys %$data;
114 { # get values (need temporary override of bindtype to avoid an error)
115 local $self->{bindtype} = 'normal';
116 ($sql, @bind) = $self->_insert_ARRAYREF([@{$data}{@fields}]);
119 # if necessary, transform values according to 'bindtype'
120 if ($self->{bindtype} eq 'columns') {
121 for my $i (0 .. $#fields) {
122 ($bind[$i]) = $self->_bindtype($fields[$i], $bind[$i]);
127 $_ = $self->_quote($_) foreach @fields;
128 $sql = "( ".join(", ", @fields).") ".$sql;
130 return ($sql, @bind);
133 sub _insert_ARRAYREF { # just generate values(?,?) part (no list of fields)
134 my ($self, $data) = @_;
136 # no names (arrayref) so can't generate bindtype
137 $self->{bindtype} ne 'columns'
138 or belch "can't do 'columns' bindtype when called with arrayref";
140 my (@values, @all_bind);
143 $self->_SWITCH_refkind($v, {
146 if ($self->{array_datatypes}) { # if array datatype are activated
150 else { # else literal SQL with bind
151 my ($sql, @bind) = @$v;
153 push @all_bind, @bind;
157 ARRAYREFREF => sub { # literal SQL with bind
158 my ($sql, @bind) = @${$v};
160 push @all_bind, @bind;
163 # THINK : anything useful to do with a HASHREF ?
165 SCALARREF => sub { # literal SQL without bind
169 SCALAR_or_UNDEF => sub {
178 my $sql = $self->_sqlcase('values')." ( ".join(", ", @values)." )";
179 return ($sql, @all_bind);
183 sub _insert_ARRAYREFREF { # literal SQL with bind
184 my ($self, $data) = @_;
189 sub _insert_SCALARREF { # literal SQL without bind
190 my ($self, $data) = @_;
197 #======================================================================
199 #======================================================================
204 my $table = $self->_table(shift);
205 my $data = shift || return;
208 # first build the 'SET' part of the sql statement
209 my (@set, @all_bind);
210 puke "Unsupported data type specified to \$sql->update"
211 unless ref $data eq 'HASH';
213 for my $k (sort keys %$data) {
216 my $label = $self->_quote($k);
218 $self->_SWITCH_refkind($v, {
220 if ($self->{array_datatypes}) { # array datatype
221 push @set, "$label = ?";
222 push @all_bind, $self->_bindtype($k, $v);
224 else { # literal SQL with bind
225 my ($sql, @bind) = @$v;
226 push @set, "$label = $sql";
227 push @all_bind, $self->_bindtype($k, @bind);
230 ARRAYREFREF => sub { # literal SQL with bind
231 my ($sql, @bind) = @${$v};
232 push @set, "$label = $sql";
233 push @all_bind, $self->_bindtype($k, @bind);
235 SCALARREF => sub { # literal SQL without bind
236 push @set, "$label = $$v";
238 SCALAR_or_UNDEF => sub {
239 push @set, "$label = ?";
240 push @all_bind, $self->_bindtype($k, $v);
246 my $sql = $self->_sqlcase('update') . " $table " . $self->_sqlcase('set ')
250 my($where_sql, @where_bind) = $self->where($where);
252 push @all_bind, @where_bind;
255 return wantarray ? ($sql, @all_bind) : $sql;
261 #======================================================================
263 #======================================================================
268 my $table = $self->_table(shift);
269 my $fields = shift || '*';
273 my($where_sql, @bind) = $self->where($where, $order);
275 my $f = (ref $fields eq 'ARRAY') ? join ', ', map { $self->_quote($_) } @$fields
277 my $sql = join(' ', $self->_sqlcase('select'), $f,
278 $self->_sqlcase('from'), $table)
281 return wantarray ? ($sql, @bind) : $sql;
284 #======================================================================
286 #======================================================================
291 my $table = $self->_table(shift);
295 my($where_sql, @bind) = $self->where($where);
296 my $sql = $self->_sqlcase('delete from') . " $table" . $where_sql;
298 return wantarray ? ($sql, @bind) : $sql;
302 #======================================================================
304 #======================================================================
308 # Finally, a separate routine just to handle WHERE clauses
310 my ($self, $where, $order) = @_;
313 my ($sql, @bind) = $self->_recurse_where($where);
314 $sql = $sql ? $self->_sqlcase(' where ') . "( $sql )" : '';
318 $sql .= $self->_order_by($order);
321 return wantarray ? ($sql, @bind) : $sql;
326 my ($self, $where, $logic) = @_;
328 # dispatch on appropriate method according to refkind of $where
329 my $method = $self->_METHOD_FOR_refkind("_where", $where);
332 my ($sql, @bind) = $self->$method($where, $logic);
334 # DBIx::Class directly calls _recurse_where in scalar context, so
335 # we must implement it, even if not in the official API
336 return wantarray ? ($sql, @bind) : $sql;
341 #======================================================================
342 # WHERE: top-level ARRAYREF
343 #======================================================================
346 sub _where_ARRAYREF {
347 my ($self, $where, $logic) = @_;
349 $logic = uc($logic || $self->{logic});
350 $logic eq 'AND' or $logic eq 'OR' or puke "unknown logic: $logic";
352 my @clauses = @$where;
354 # if the array starts with [-and|or => ...], recurse with that logic
355 my $first = $clauses[0] || '';
356 if ($first =~ /^-(and|or)/i) {
359 return $self->_where_ARRAYREF(\@clauses, $logic);
363 my (@sql_clauses, @all_bind);
365 # need to use while() so can shift() for pairs
366 while (my $el = shift @clauses) {
368 # switch according to kind of $el and get corresponding ($sql, @bind)
369 my ($sql, @bind) = $self->_SWITCH_refkind($el, {
371 # skip empty elements, otherwise get invalid trailing AND stuff
372 ARRAYREF => sub {$self->_recurse_where($el) if @$el},
374 HASHREF => sub {$self->_recurse_where($el, 'and') if %$el},
375 # LDNOTE : previous SQLA code for hashrefs was creating a dirty
376 # side-effect: the first hashref within an array would change
377 # the global logic to 'AND'. So [ {cond1, cond2}, [cond3, cond4] ]
378 # was interpreted as "(cond1 AND cond2) OR (cond3 AND cond4)",
379 # whereas it should be "(cond1 AND cond2) OR (cond3 OR cond4)".
381 SCALARREF => sub { ($$el); },
383 SCALAR => sub {# top-level arrayref with scalars, recurse in pairs
384 $self->_recurse_where({$el => shift(@clauses)})},
386 UNDEF => sub {puke "not supported : UNDEF in arrayref" },
390 push @sql_clauses, $sql;
391 push @all_bind, @bind;
395 return $self->_join_sql_clauses($logic, \@sql_clauses, \@all_bind);
400 #======================================================================
401 # WHERE: top-level HASHREF
402 #======================================================================
405 my ($self, $where) = @_;
406 my (@sql_clauses, @all_bind);
408 # LDNOTE : don't really know why we need to sort keys
409 for my $k (sort keys %$where) {
410 my $v = $where->{$k};
412 # ($k => $v) is either a special op or a regular hashpair
413 my ($sql, @bind) = ($k =~ /^-(.+)/) ? $self->_where_op_in_hash($1, $v)
415 my $method = $self->_METHOD_FOR_refkind("_where_hashpair", $v);
416 $self->$method($k, $v);
419 push @sql_clauses, $sql;
420 push @all_bind, @bind;
423 return $self->_join_sql_clauses('and', \@sql_clauses, \@all_bind);
427 sub _where_op_in_hash {
428 my ($self, $op, $v) = @_;
430 $op =~ /^(AND|OR|NEST)[_\d]*/i
431 or puke "unknown operator: -$op";
432 $op = uc($1); # uppercase, remove trailing digits
433 $self->_debug("OP(-$op) within hashref, recursing...");
435 $self->_SWITCH_refkind($v, {
438 # LDNOTE : should deprecate {-or => [...]} and {-and => [...]}
439 # because they are misleading; the only proper way would be
440 # -nest => [-or => ...], -nest => [-and ...]
441 return $self->_where_ARRAYREF($v, $op eq 'NEST' ? '' : $op);
446 belch "-or => {...} should be -nest => [...]";
447 return $self->_where_ARRAYREF([%$v], 'OR');
450 return $self->_where_HASHREF($v);
454 SCALARREF => sub { # literal SQL
456 or puke "-$op => \\\$scalar not supported, use -nest => ...";
460 ARRAYREFREF => sub { # literal SQL
462 or puke "-$op => \\[..] not supported, use -nest => ...";
466 SCALAR => sub { # permissively interpreted as SQL
468 or puke "-$op => 'scalar' not supported, use -nest => \\'scalar'";
469 belch "literal SQL should be -nest => \\'scalar' "
470 . "instead of -nest => 'scalar' ";
475 puke "-$op => undef not supported";
481 sub _where_hashpair_ARRAYREF {
482 my ($self, $k, $v) = @_;
485 my @v = @$v; # need copy because of shift below
486 $self->_debug("ARRAY($k) means distribute over elements");
488 # put apart first element if it is an operator (-and, -or)
489 my $op = $v[0] =~ /^-/ ? shift @v : undef;
490 $self->_debug("OP($op) reinjected into the distributed array") if $op;
492 my @distributed = map { {$k => $_} } @v;
493 unshift @distributed, $op if $op;
495 return $self->_recurse_where(\@distributed);
498 # LDNOTE : not sure of this one. What does "distribute over nothing" mean?
499 $self->_debug("empty ARRAY($k) means 0=1");
500 return ($self->{sqlfalse});
504 sub _where_hashpair_HASHREF {
505 my ($self, $k, $v) = @_;
507 my (@all_sql, @all_bind);
509 for my $op (sort keys %$v) {
512 # put the operator in canonical form
513 $op =~ s/^-//; # remove initial dash
514 $op =~ tr/_/ /; # underscores become spaces
515 $op =~ s/^\s+//; # no initial space
516 $op =~ s/\s+$//; # no final space
517 $op =~ s/\s+/ /; # multiple spaces become one
521 # CASE: special operators like -in or -between
522 my $special_op = first {$op =~ $_->{regex}} @{$self->{special_ops}};
524 ($sql, @bind) = $special_op->{handler}->($self, $k, $op, $val);
527 # CASE: col => {op => \@vals}
528 elsif (ref $val eq 'ARRAY') {
529 ($sql, @bind) = $self->_where_field_op_ARRAYREF($k, $op, $val);
532 # CASE: col => {op => \$scalar}
533 elsif (ref $val eq 'SCALAR') {
534 $sql = join ' ', $self->_convert($self->_quote($k)),
535 $self->_sqlcase($op),
539 # CASE: col => {op => undef} : sql "IS (NOT)? NULL"
540 elsif (! defined($val)) {
541 my $is = ($op =~ $self->{equality_op}) ? 'is' :
542 ($op =~ $self->{inequality_op}) ? 'is not' :
543 puke "unexpected operator '$op' with undef operand";
544 $sql = $self->_quote($k) . $self->_sqlcase(" $is null");
547 # CASE: col => {op => $scalar}
549 $sql = join ' ', $self->_convert($self->_quote($k)),
550 $self->_sqlcase($op),
551 $self->_convert('?');
552 @bind = $self->_bindtype($k, $val);
556 push @all_bind, @bind;
559 return $self->_join_sql_clauses('and', \@all_sql, \@all_bind);
564 sub _where_field_op_ARRAYREF {
565 my ($self, $k, $op, $vals) = @_;
568 $self->_debug("ARRAY($vals) means multiple elements: [ @$vals ]");
572 # LDNOTE : change the distribution logic when
573 # $op =~ $self->{inequality_op}, because of Morgan laws :
574 # with {field => {'!=' => [22, 33]}}, it would be ridiculous to generate
575 # WHERE field != 22 OR field != 33 : the user probably means
576 # WHERE field != 22 AND field != 33.
577 my $logic = ($op =~ $self->{inequality_op}) ? 'AND' : 'OR';
579 # distribute $op over each member of @$vals
580 return $self->_recurse_where([map { {$k => {$op, $_}} } @$vals], $logic);
584 # try to DWIM on equality operators
585 # LDNOTE : not 100% sure this is the correct thing to do ...
586 return ($self->{sqlfalse}) if $op =~ $self->{equality_op};
587 return ($self->{sqltrue}) if $op =~ $self->{inequality_op};
590 puke "operator '$op' applied on an empty array (field '$k')";
595 sub _where_hashpair_SCALARREF {
596 my ($self, $k, $v) = @_;
597 $self->_debug("SCALAR($k) means literal SQL: $$v");
598 my $sql = $self->_quote($k) . " " . $$v;
602 sub _where_hashpair_ARRAYREFREF {
603 my ($self, $k, $v) = @_;
604 $self->_debug("REF($k) means literal SQL: @${$v}");
605 my ($sql, @bind) = @${$v};
606 $sql = $self->_quote($k) . " " . $sql;
607 @bind = $self->_bindtype($k, @bind);
608 return ($sql, @bind );
611 sub _where_hashpair_SCALAR {
612 my ($self, $k, $v) = @_;
613 $self->_debug("NOREF($k) means simple key=val: $k $self->{cmp} $v");
614 my $sql = join ' ', $self->_convert($self->_quote($k)),
615 $self->_sqlcase($self->{cmp}),
616 $self->_convert('?');
617 my @bind = $self->_bindtype($k, $v);
618 return ( $sql, @bind);
622 sub _where_hashpair_UNDEF {
623 my ($self, $k, $v) = @_;
624 $self->_debug("UNDEF($k) means IS NULL");
625 my $sql = $self->_quote($k) . $self->_sqlcase(' is null');
629 #======================================================================
630 # WHERE: TOP-LEVEL OTHERS (SCALARREF, SCALAR, UNDEF)
631 #======================================================================
634 sub _where_SCALARREF {
635 my ($self, $where) = @_;
638 $self->_debug("SCALAR(*top) means literal SQL: $$where");
644 my ($self, $where) = @_;
647 $self->_debug("NOREF(*top) means literal SQL: $where");
658 #======================================================================
659 # WHERE: BUILTIN SPECIAL OPERATORS (-in, -between)
660 #======================================================================
663 sub _where_field_BETWEEN {
664 my ($self, $k, $op, $vals) = @_;
666 ref $vals eq 'ARRAY' && @$vals == 2
667 or puke "special op 'between' requires an arrayref of two values";
669 my ($label) = $self->_convert($self->_quote($k));
670 my ($placeholder) = $self->_convert('?');
671 my $and = $self->_sqlcase('and');
672 $op = $self->_sqlcase($op);
674 my $sql = "( $label $op $placeholder $and $placeholder )";
675 my @bind = $self->_bindtype($k, @$vals);
680 sub _where_field_IN {
681 my ($self, $k, $op, $vals) = @_;
683 # backwards compatibility : if scalar, force into an arrayref
684 $vals = [$vals] if defined $vals && ! ref $vals;
687 or puke "special op 'in' requires an arrayref";
689 my ($label) = $self->_convert($self->_quote($k));
690 my ($placeholder) = $self->_convert('?');
691 my $and = $self->_sqlcase('and');
692 $op = $self->_sqlcase($op);
694 if (@$vals) { # nonempty list
695 my $placeholders = join ", ", (($placeholder) x @$vals);
696 my $sql = "$label $op ( $placeholders )";
697 my @bind = $self->_bindtype($k, @$vals);
699 return ($sql, @bind);
701 else { # empty list : some databases won't understand "IN ()", so DWIM
702 my $sql = ($op =~ /\bnot\b/i) ? $self->{sqltrue} : $self->{sqlfalse};
712 #======================================================================
714 #======================================================================
717 my ($self, $arg) = @_;
719 # construct list of ordering instructions
720 my @order = $self->_SWITCH_refkind($arg, {
723 map {$self->_SWITCH_refkind($_, {
724 SCALAR => sub {$self->_quote($_)},
726 SCALARREF => sub {$$_}, # literal SQL, no quoting
727 HASHREF => sub {$self->_order_by_hash($_)}
731 SCALAR => sub {$self->_quote($arg)},
733 SCALARREF => sub {$$arg}, # literal SQL, no quoting
734 HASHREF => sub {$self->_order_by_hash($arg)},
739 my $order = join ', ', @order;
740 return $order ? $self->_sqlcase(' order by')." $order" : '';
745 my ($self, $hash) = @_;
747 # get first pair in hash
748 my ($key, $val) = each %$hash;
750 # check if one pair was found and no other pair in hash
751 $key && !(each %$hash)
752 or puke "hash passed to _order_by must have exactly one key (-desc or -asc)";
754 my ($order) = ($key =~ /^-(desc|asc)/i)
755 or puke "invalid key in _order_by hash : $key";
757 return $self->_quote($val) ." ". $self->_sqlcase($order);
762 #======================================================================
763 # DATASOURCE (FOR NOW, JUST PLAIN TABLE OR LIST OF TABLES)
764 #======================================================================
769 $self->_SWITCH_refkind($from, {
770 ARRAYREF => sub {join ', ', map { $self->_quote($_) } @$from;},
771 SCALAR => sub {$self->_quote($from)},
772 SCALARREF => sub {$$from},
773 ARRAYREFREF => sub {join ', ', @$from;},
778 #======================================================================
780 #======================================================================
786 $label or puke "can't quote an empty label";
788 # left and right quote characters
789 my ($ql, $qr, @other) = $self->_SWITCH_refkind($self->{quote_char}, {
790 SCALAR => sub {($self->{quote_char}, $self->{quote_char})},
791 ARRAYREF => sub {@{$self->{quote_char}}},
795 or puke "quote_char must be an arrayref of 2 values";
797 # no quoting if no quoting chars
798 $ql or return $label;
800 # no quoting for literal SQL
801 return $$label if ref($label) eq 'SCALAR';
803 # separate table / column (if applicable)
804 my $sep = $self->{name_sep} || '';
805 my @to_quote = $sep ? split /\Q$sep\E/, $label : ($label);
807 # do the quoting, except for "*" or for `table`.*
808 my @quoted = map { $_ eq '*' ? $_: $ql.$_.$qr} @to_quote;
810 # reassemble and return.
811 return join $sep, @quoted;
815 # Conversion, if applicable
817 my ($self, $arg) = @_;
819 # LDNOTE : modified the previous implementation below because
820 # it was not consistent : the first "return" is always an array,
821 # the second "return" is context-dependent. Anyway, _convert
822 # seems always used with just a single argument, so make it a
824 # return @_ unless $self->{convert};
825 # my $conv = $self->_sqlcase($self->{convert});
826 # my @ret = map { $conv.'('.$_.')' } @_;
827 # return wantarray ? @ret : $ret[0];
828 if ($self->{convert}) {
829 my $conv = $self->_sqlcase($self->{convert});
830 $arg = $conv.'('.$arg.')';
838 my($col, @vals) = @_;
840 #LDNOTE : changed original implementation below because it did not make
841 # sense when bindtype eq 'columns' and @vals > 1.
842 # return $self->{bindtype} eq 'columns' ? [ $col, @vals ] : @vals;
844 return $self->{bindtype} eq 'columns' ? map {[$col, $_]} @vals : @vals;
847 sub _join_sql_clauses {
848 my ($self, $logic, $clauses_aref, $bind_aref) = @_;
850 if (@$clauses_aref > 1) {
851 my $join = " " . $self->_sqlcase($logic) . " ";
852 my $sql = '( ' . join($join, @$clauses_aref) . ' )';
853 return ($sql, @$bind_aref);
855 elsif (@$clauses_aref) {
856 return ($clauses_aref->[0], @$bind_aref); # no parentheses
859 return (); # if no SQL, ignore @$bind_aref
864 # Fix SQL case, if so requested
868 # LDNOTE: if $self->{case} is true, then it contains 'lower', so we
869 # don't touch the argument ... crooked logic, but let's not change it!
870 return $self->{case} ? $_[0] : uc($_[0]);
874 #======================================================================
875 # DISPATCHING FROM REFKIND
876 #======================================================================
879 my ($self, $data) = @_;
885 # blessed objects are treated like scalars
886 $ref = (blessed $data) ? '' : ref $data;
887 $n_steps += 1 if $ref;
888 last if $ref ne 'REF';
892 my $base = $ref || (defined $data ? 'SCALAR' : 'UNDEF');
894 return $base . ('REF' x $n_steps);
900 my ($self, $data) = @_;
901 my @try = ($self->_refkind($data));
902 push @try, 'SCALAR_or_UNDEF' if $try[0] eq 'SCALAR' || $try[0] eq 'UNDEF';
903 push @try, 'FALLBACK';
907 sub _METHOD_FOR_refkind {
908 my ($self, $meth_prefix, $data) = @_;
909 my $method = first {$_} map {$self->can($meth_prefix."_".$_)}
910 $self->_try_refkind($data)
911 or puke "cannot dispatch on '$meth_prefix' for ".$self->_refkind($data);
916 sub _SWITCH_refkind {
917 my ($self, $data, $dispatch_table) = @_;
919 my $coderef = first {$_} map {$dispatch_table->{$_}}
920 $self->_try_refkind($data)
921 or puke "no dispatch entry for ".$self->_refkind($data);
928 #======================================================================
929 # VALUES, GENERATE, AUTOLOAD
930 #======================================================================
932 # LDNOTE: original code from nwiger, didn't touch code in that section
933 # I feel the AUTOLOAD stuff should not be the default, it should
934 # only be activated on explicit demand by user.
938 my $data = shift || return;
939 puke "Argument to ", __PACKAGE__, "->values must be a \\%hash"
940 unless ref $data eq 'HASH';
941 return map { $self->_bindtype($_, $data->{$_}) } sort keys %$data;
947 my(@sql, @sqlq, @sqlv);
951 if ($ref eq 'HASH') {
952 for my $k (sort keys %$_) {
955 my $label = $self->_quote($k);
957 # SQL included for values
959 my $sql = shift @bind;
960 push @sqlq, "$label = $sql";
961 push @sqlv, $self->_bindtype($k, @bind);
962 } elsif ($r eq 'SCALAR') {
963 # embedded literal SQL
964 push @sqlq, "$label = $$v";
966 push @sqlq, "$label = ?";
967 push @sqlv, $self->_bindtype($k, $v);
970 push @sql, $self->_sqlcase('set'), join ', ', @sqlq;
971 } elsif ($ref eq 'ARRAY') {
972 # unlike insert(), assume these are ONLY the column names, i.e. for SQL
977 push @sqlq, shift @val;
979 } elsif ($r eq 'SCALAR') {
980 # embedded literal SQL
987 push @sql, '(' . join(', ', @sqlq) . ')';
988 } elsif ($ref eq 'SCALAR') {
992 # strings get case twiddled
993 push @sql, $self->_sqlcase($_);
997 my $sql = join ' ', @sql;
999 # this is pretty tricky
1000 # if ask for an array, return ($stmt, @bind)
1001 # otherwise, s/?/shift @sqlv/ to put it inline
1003 return ($sql, @sqlv);
1005 1 while $sql =~ s/\?/my $d = shift(@sqlv);
1006 ref $d ? $d->[1] : $d/e;
1015 # This allows us to check for a local, then _form, attr
1017 my($name) = $AUTOLOAD =~ /.*::(.+)/;
1018 return $self->generate($name, @_);
1029 SQL::Abstract - Generate SQL from Perl data structures
1035 my $sql = SQL::Abstract->new;
1037 my($stmt, @bind) = $sql->select($table, \@fields, \%where, \@order);
1039 my($stmt, @bind) = $sql->insert($table, \%fieldvals || \@values);
1041 my($stmt, @bind) = $sql->update($table, \%fieldvals, \%where);
1043 my($stmt, @bind) = $sql->delete($table, \%where);
1045 # Then, use these in your DBI statements
1046 my $sth = $dbh->prepare($stmt);
1047 $sth->execute(@bind);
1049 # Just generate the WHERE clause
1050 my($stmt, @bind) = $sql->where(\%where, \@order);
1052 # Return values in the same order, for hashed queries
1053 # See PERFORMANCE section for more details
1054 my @bind = $sql->values(\%fieldvals);
1058 This module was inspired by the excellent L<DBIx::Abstract>.
1059 However, in using that module I found that what I really wanted
1060 to do was generate SQL, but still retain complete control over my
1061 statement handles and use the DBI interface. So, I set out to
1062 create an abstract SQL generation module.
1064 While based on the concepts used by L<DBIx::Abstract>, there are
1065 several important differences, especially when it comes to WHERE
1066 clauses. I have modified the concepts used to make the SQL easier
1067 to generate from Perl data structures and, IMO, more intuitive.
1068 The underlying idea is for this module to do what you mean, based
1069 on the data structures you provide it. The big advantage is that
1070 you don't have to modify your code every time your data changes,
1071 as this module figures it out.
1073 To begin with, an SQL INSERT is as easy as just specifying a hash
1074 of C<key=value> pairs:
1077 name => 'Jimbo Bobson',
1078 phone => '123-456-7890',
1079 address => '42 Sister Lane',
1080 city => 'St. Louis',
1081 state => 'Louisiana',
1084 The SQL can then be generated with this:
1086 my($stmt, @bind) = $sql->insert('people', \%data);
1088 Which would give you something like this:
1090 $stmt = "INSERT INTO people
1091 (address, city, name, phone, state)
1092 VALUES (?, ?, ?, ?, ?)";
1093 @bind = ('42 Sister Lane', 'St. Louis', 'Jimbo Bobson',
1094 '123-456-7890', 'Louisiana');
1096 These are then used directly in your DBI code:
1098 my $sth = $dbh->prepare($stmt);
1099 $sth->execute(@bind);
1101 =head2 Inserting and Updating Arrays
1103 If your database has array types (like for example Postgres),
1104 activate the special option C<< array_datatypes => 1 >>
1105 when creating the C<SQL::Abstract> object.
1106 Then you may use an arrayref to insert and update database array types:
1108 my $sql = SQL::Abstract->new(array_datatypes => 1);
1110 planets => [qw/Mercury Venus Earth Mars/]
1113 my($stmt, @bind) = $sql->insert('solar_system', \%data);
1117 $stmt = "INSERT INTO solar_system (planets) VALUES (?)"
1119 @bind = (['Mercury', 'Venus', 'Earth', 'Mars']);
1122 =head2 Inserting and Updating SQL
1124 In order to apply SQL functions to elements of your C<%data> you may
1125 specify a reference to an arrayref for the given hash value. For example,
1126 if you need to execute the Oracle C<to_date> function on a value, you can
1127 say something like this:
1131 date_entered => \["to_date(?,'MM/DD/YYYY')", "03/02/2003"],
1134 The first value in the array is the actual SQL. Any other values are
1135 optional and would be included in the bind values array. This gives
1138 my($stmt, @bind) = $sql->insert('people', \%data);
1140 $stmt = "INSERT INTO people (name, date_entered)
1141 VALUES (?, to_date(?,'MM/DD/YYYY'))";
1142 @bind = ('Bill', '03/02/2003');
1144 An UPDATE is just as easy, all you change is the name of the function:
1146 my($stmt, @bind) = $sql->update('people', \%data);
1148 Notice that your C<%data> isn't touched; the module will generate
1149 the appropriately quirky SQL for you automatically. Usually you'll
1150 want to specify a WHERE clause for your UPDATE, though, which is
1151 where handling C<%where> hashes comes in handy...
1153 =head2 Complex where statements
1155 This module can generate pretty complicated WHERE statements
1156 easily. For example, simple C<key=value> pairs are taken to mean
1157 equality, and if you want to see if a field is within a set
1158 of values, you can use an arrayref. Let's say we wanted to
1159 SELECT some data based on this criteria:
1162 requestor => 'inna',
1163 worker => ['nwiger', 'rcwe', 'sfz'],
1164 status => { '!=', 'completed' }
1167 my($stmt, @bind) = $sql->select('tickets', '*', \%where);
1169 The above would give you something like this:
1171 $stmt = "SELECT * FROM tickets WHERE
1172 ( requestor = ? ) AND ( status != ? )
1173 AND ( worker = ? OR worker = ? OR worker = ? )";
1174 @bind = ('inna', 'completed', 'nwiger', 'rcwe', 'sfz');
1176 Which you could then use in DBI code like so:
1178 my $sth = $dbh->prepare($stmt);
1179 $sth->execute(@bind);
1185 The functions are simple. There's one for each major SQL operation,
1186 and a constructor you use first. The arguments are specified in a
1187 similar order to each function (table, then fields, then a where
1188 clause) to try and simplify things.
1193 =head2 new(option => 'value')
1195 The C<new()> function takes a list of options and values, and returns
1196 a new B<SQL::Abstract> object which can then be used to generate SQL
1197 through the methods below. The options accepted are:
1203 If set to 'lower', then SQL will be generated in all lowercase. By
1204 default SQL is generated in "textbook" case meaning something like:
1206 SELECT a_field FROM a_table WHERE some_field LIKE '%someval%'
1208 Any setting other than 'lower' is ignored.
1212 This determines what the default comparison operator is. By default
1213 it is C<=>, meaning that a hash like this:
1215 %where = (name => 'nwiger', email => 'nate@wiger.org');
1217 Will generate SQL like this:
1219 WHERE name = 'nwiger' AND email = 'nate@wiger.org'
1221 However, you may want loose comparisons by default, so if you set
1222 C<cmp> to C<like> you would get SQL such as:
1224 WHERE name like 'nwiger' AND email like 'nate@wiger.org'
1226 You can also override the comparsion on an individual basis - see
1227 the huge section on L</"WHERE CLAUSES"> at the bottom.
1229 =item sqltrue, sqlfalse
1231 Expressions for inserting boolean values within SQL statements.
1232 By default these are C<1=1> and C<1=0>.
1236 This determines the default logical operator for multiple WHERE
1237 statements in arrays. By default it is "or", meaning that a WHERE
1241 event_date => {'>=', '2/13/99'},
1242 event_date => {'<=', '4/24/03'},
1245 Will generate SQL like this:
1247 WHERE event_date >= '2/13/99' OR event_date <= '4/24/03'
1249 This is probably not what you want given this query, though (look
1250 at the dates). To change the "OR" to an "AND", simply specify:
1252 my $sql = SQL::Abstract->new(logic => 'and');
1254 Which will change the above C<WHERE> to:
1256 WHERE event_date >= '2/13/99' AND event_date <= '4/24/03'
1258 The logic can also be changed locally by inserting
1259 an extra first element in the array :
1261 @where = (-and => event_date => {'>=', '2/13/99'},
1262 event_date => {'<=', '4/24/03'} );
1264 See the L</"WHERE CLAUSES"> section for explanations.
1268 This will automatically convert comparisons using the specified SQL
1269 function for both column and value. This is mostly used with an argument
1270 of C<upper> or C<lower>, so that the SQL will have the effect of
1271 case-insensitive "searches". For example, this:
1273 $sql = SQL::Abstract->new(convert => 'upper');
1274 %where = (keywords => 'MaKe iT CAse inSeNSItive');
1276 Will turn out the following SQL:
1278 WHERE upper(keywords) like upper('MaKe iT CAse inSeNSItive')
1280 The conversion can be C<upper()>, C<lower()>, or any other SQL function
1281 that can be applied symmetrically to fields (actually B<SQL::Abstract> does
1282 not validate this option; it will just pass through what you specify verbatim).
1286 This is a kludge because many databases suck. For example, you can't
1287 just bind values using DBI's C<execute()> for Oracle C<CLOB> or C<BLOB> fields.
1288 Instead, you have to use C<bind_param()>:
1290 $sth->bind_param(1, 'reg data');
1291 $sth->bind_param(2, $lots, {ora_type => ORA_CLOB});
1293 The problem is, B<SQL::Abstract> will normally just return a C<@bind> array,
1294 which loses track of which field each slot refers to. Fear not.
1296 If you specify C<bindtype> in new, you can determine how C<@bind> is returned.
1297 Currently, you can specify either C<normal> (default) or C<columns>. If you
1298 specify C<columns>, you will get an array that looks like this:
1300 my $sql = SQL::Abstract->new(bindtype => 'columns');
1301 my($stmt, @bind) = $sql->insert(...);
1304 [ 'column1', 'value1' ],
1305 [ 'column2', 'value2' ],
1306 [ 'column3', 'value3' ],
1309 You can then iterate through this manually, using DBI's C<bind_param()>.
1311 $sth->prepare($stmt);
1314 my($col, $data) = @$_;
1315 if ($col eq 'details' || $col eq 'comments') {
1316 $sth->bind_param($i, $data, {ora_type => ORA_CLOB});
1317 } elsif ($col eq 'image') {
1318 $sth->bind_param($i, $data, {ora_type => ORA_BLOB});
1320 $sth->bind_param($i, $data);
1324 $sth->execute; # execute without @bind now
1326 Now, why would you still use B<SQL::Abstract> if you have to do this crap?
1327 Basically, the advantage is still that you don't have to care which fields
1328 are or are not included. You could wrap that above C<for> loop in a simple
1329 sub called C<bind_fields()> or something and reuse it repeatedly. You still
1330 get a layer of abstraction over manual SQL specification.
1334 This is the character that a table or column name will be quoted
1335 with. By default this is an empty string, but you could set it to
1336 the character C<`>, to generate SQL like this:
1338 SELECT `a_field` FROM `a_table` WHERE `some_field` LIKE '%someval%'
1340 Alternatively, you can supply an array ref of two items, the first being the left
1341 hand quote character, and the second the right hand quote character. For
1342 example, you could supply C<['[',']']> for SQL Server 2000 compliant quotes
1343 that generates SQL like this:
1345 SELECT [a_field] FROM [a_table] WHERE [some_field] LIKE '%someval%'
1347 Quoting is useful if you have tables or columns names that are reserved
1348 words in your database's SQL dialect.
1352 This is the character that separates a table and column name. It is
1353 necessary to specify this when the C<quote_char> option is selected,
1354 so that tables and column names can be individually quoted like this:
1356 SELECT `table`.`one_field` FROM `table` WHERE `table`.`other_field` = 1
1358 =item array_datatypes
1360 When this option is true, arrayrefs in INSERT or UPDATE are
1361 interpreted as array datatypes and are passed directly
1363 When this option is false, arrayrefs are interpreted
1364 as literal SQL, just like refs to arrayrefs
1365 (but this behavior is for backwards compatibility; when writing
1366 new queries, use the "reference to arrayref" syntax
1372 Takes a reference to a list of "special operators"
1373 to extend the syntax understood by L<SQL::Abstract>.
1374 See section L</"SPECIAL OPERATORS"> for details.
1380 =head2 insert($table, \@values || \%fieldvals)
1382 This is the simplest function. You simply give it a table name
1383 and either an arrayref of values or hashref of field/value pairs.
1384 It returns an SQL INSERT statement and a list of bind values.
1385 See the sections on L</"Inserting and Updating Arrays"> and
1386 L</"Inserting and Updating SQL"> for information on how to insert
1387 with those data types.
1389 =head2 update($table, \%fieldvals, \%where)
1391 This takes a table, hashref of field/value pairs, and an optional
1392 hashref L<WHERE clause|/WHERE CLAUSES>. It returns an SQL UPDATE function and a list
1394 See the sections on L</"Inserting and Updating Arrays"> and
1395 L</"Inserting and Updating SQL"> for information on how to insert
1396 with those data types.
1398 =head2 select($source, $fields, $where, $order)
1400 This returns a SQL SELECT statement and associated list of bind values, as
1401 specified by the arguments :
1407 Specification of the 'FROM' part of the statement.
1408 The argument can be either a plain scalar (interpreted as a table
1409 name, will be quoted), or an arrayref (interpreted as a list
1410 of table names, joined by commas, quoted), or a scalarref
1411 (literal table name, not quoted), or a ref to an arrayref
1412 (list of literal table names, joined by commas, not quoted).
1416 Specification of the list of fields to retrieve from
1418 The argument can be either an arrayref (interpreted as a list
1419 of field names, will be joined by commas and quoted), or a
1420 plain scalar (literal SQL, not quoted).
1421 Please observe that this API is not as flexible as for
1422 the first argument C<$table>, for backwards compatibility reasons.
1426 Optional argument to specify the WHERE part of the query.
1427 The argument is most often a hashref, but can also be
1428 an arrayref or plain scalar --
1429 see section L<WHERE clause|/"WHERE CLAUSES"> for details.
1433 Optional argument to specify the ORDER BY part of the query.
1434 The argument can be a scalar, a hashref or an arrayref
1435 -- see section L<ORDER BY clause|/"ORDER BY CLAUSES">
1441 =head2 delete($table, \%where)
1443 This takes a table name and optional hashref L<WHERE clause|/WHERE CLAUSES>.
1444 It returns an SQL DELETE statement and list of bind values.
1446 =head2 where(\%where, \@order)
1448 This is used to generate just the WHERE clause. For example,
1449 if you have an arbitrary data structure and know what the
1450 rest of your SQL is going to look like, but want an easy way
1451 to produce a WHERE clause, use this. It returns an SQL WHERE
1452 clause and list of bind values.
1455 =head2 values(\%data)
1457 This just returns the values from the hash C<%data>, in the same
1458 order that would be returned from any of the other above queries.
1459 Using this allows you to markedly speed up your queries if you
1460 are affecting lots of rows. See below under the L</"PERFORMANCE"> section.
1462 =head2 generate($any, 'number', $of, \@data, $struct, \%types)
1464 Warning: This is an experimental method and subject to change.
1466 This returns arbitrarily generated SQL. It's a really basic shortcut.
1467 It will return two different things, depending on return context:
1469 my($stmt, @bind) = $sql->generate('create table', \$table, \@fields);
1470 my $stmt_and_val = $sql->generate('create table', \$table, \@fields);
1472 These would return the following:
1474 # First calling form
1475 $stmt = "CREATE TABLE test (?, ?)";
1476 @bind = (field1, field2);
1478 # Second calling form
1479 $stmt_and_val = "CREATE TABLE test (field1, field2)";
1481 Depending on what you're trying to do, it's up to you to choose the correct
1482 format. In this example, the second form is what you would want.
1486 $sql->generate('alter session', { nls_date_format => 'MM/YY' });
1490 ALTER SESSION SET nls_date_format = 'MM/YY'
1492 You get the idea. Strings get their case twiddled, but everything
1493 else remains verbatim.
1498 =head1 WHERE CLAUSES
1502 This module uses a variation on the idea from L<DBIx::Abstract>. It
1503 is B<NOT>, repeat I<not> 100% compatible. B<The main logic of this
1504 module is that things in arrays are OR'ed, and things in hashes
1507 The easiest way to explain is to show lots of examples. After
1508 each C<%where> hash shown, it is assumed you used:
1510 my($stmt, @bind) = $sql->where(\%where);
1512 However, note that the C<%where> hash can be used directly in any
1513 of the other functions as well, as described above.
1515 =head2 Key-value pairs
1517 So, let's get started. To begin, a simple hash:
1521 status => 'completed'
1524 Is converted to SQL C<key = val> statements:
1526 $stmt = "WHERE user = ? AND status = ?";
1527 @bind = ('nwiger', 'completed');
1529 One common thing I end up doing is having a list of values that
1530 a field can be in. To do this, simply specify a list inside of
1535 status => ['assigned', 'in-progress', 'pending'];
1538 This simple code will create the following:
1540 $stmt = "WHERE user = ? AND ( status = ? OR status = ? OR status = ? )";
1541 @bind = ('nwiger', 'assigned', 'in-progress', 'pending');
1543 An empty arrayref will be considered a logical false and
1546 =head2 Key-value pairs
1548 If you want to specify a different type of operator for your comparison,
1549 you can use a hashref for a given column:
1553 status => { '!=', 'completed' }
1556 Which would generate:
1558 $stmt = "WHERE user = ? AND status != ?";
1559 @bind = ('nwiger', 'completed');
1561 To test against multiple values, just enclose the values in an arrayref:
1563 status => { '!=', ['assigned', 'in-progress', 'pending'] };
1565 Which would give you:
1567 "WHERE status != ? AND status != ? AND status != ?"
1569 Notice that since the operator was recognized as being a 'negative'
1570 operator, the arrayref was interpreted with 'AND' logic (because
1571 of Morgan's laws). By contrast, the reverse
1573 status => { '=', ['assigned', 'in-progress', 'pending'] };
1577 "WHERE status = ? OR status = ? OR status = ?"
1580 The hashref can also contain multiple pairs, in which case it is expanded
1581 into an C<AND> of its elements:
1585 status => { '!=', 'completed', -not_like => 'pending%' }
1588 # Or more dynamically, like from a form
1589 $where{user} = 'nwiger';
1590 $where{status}{'!='} = 'completed';
1591 $where{status}{'-not_like'} = 'pending%';
1593 # Both generate this
1594 $stmt = "WHERE user = ? AND status != ? AND status NOT LIKE ?";
1595 @bind = ('nwiger', 'completed', 'pending%');
1598 To get an OR instead, you can combine it with the arrayref idea:
1602 priority => [ {'=', 2}, {'!=', 1} ]
1605 Which would generate:
1607 $stmt = "WHERE user = ? AND priority = ? OR priority != ?";
1608 @bind = ('nwiger', '2', '1');
1611 =head2 Logic and nesting operators
1613 In the example above,
1614 there is a subtle trap if you want to say something like
1615 this (notice the C<AND>):
1617 WHERE priority != ? AND priority != ?
1619 Because, in Perl you I<can't> do this:
1621 priority => { '!=', 2, '!=', 1 }
1623 As the second C<!=> key will obliterate the first. The solution
1624 is to use the special C<-modifier> form inside an arrayref:
1626 priority => [ -and => {'!=', 2},
1630 Normally, these would be joined by C<OR>, but the modifier tells it
1631 to use C<AND> instead. (Hint: You can use this in conjunction with the
1632 C<logic> option to C<new()> in order to change the way your queries
1633 work by default.) B<Important:> Note that the C<-modifier> goes
1634 B<INSIDE> the arrayref, as an extra first element. This will
1635 B<NOT> do what you think it might:
1637 priority => -and => [{'!=', 2}, {'!=', 1}] # WRONG!
1639 Here is a quick list of equivalencies, since there is some overlap:
1642 status => {'!=', 'completed', 'not like', 'pending%' }
1643 status => [ -and => {'!=', 'completed'}, {'not like', 'pending%'}]
1646 status => {'=', ['assigned', 'in-progress']}
1647 status => [ -or => {'=', 'assigned'}, {'=', 'in-progress'}]
1648 status => [ {'=', 'assigned'}, {'=', 'in-progress'} ]
1650 In addition to C<-and> and C<-or>, there is also a special C<-nest>
1651 operator which adds an additional set of parens, to create a subquery.
1652 For example, to get something like this:
1654 $stmt = "WHERE user = ? AND ( workhrs > ? OR geo = ? )";
1655 @bind = ('nwiger', '20', 'ASIA');
1661 -nest => [ workhrs => {'>', 20}, geo => 'ASIA' ],
1664 If you need several nested subexpressions, you can number
1665 the C<-nest> branches :
1675 =head2 Special operators : IN, BETWEEN, etc.
1677 You can also use the hashref format to compare a list of fields using the
1678 C<IN> comparison operator, by specifying the list as an arrayref:
1681 status => 'completed',
1682 reportid => { -in => [567, 2335, 2] }
1685 Which would generate:
1687 $stmt = "WHERE status = ? AND reportid IN (?,?,?)";
1688 @bind = ('completed', '567', '2335', '2');
1690 The reverse operator C<-not_in> generates SQL C<NOT IN> and is used in
1693 Another pair of operators is C<-between> and C<-not_between>,
1694 used with an arrayref of two values:
1698 completion_date => {
1699 -not_between => ['2002-10-01', '2003-02-06']
1705 WHERE user = ? AND completion_date NOT BETWEEN ( ? AND ? )
1707 These are the two builtin "special operators"; but the
1708 list can be expanded : see section L</"SPECIAL OPERATORS"> below.
1710 =head2 Nested conditions
1712 So far, we've seen how multiple conditions are joined with a top-level
1713 C<AND>. We can change this by putting the different conditions we want in
1714 hashes and then putting those hashes in an array. For example:
1719 status => { -like => ['pending%', 'dispatched'] },
1723 status => 'unassigned',
1727 This data structure would create the following:
1729 $stmt = "WHERE ( user = ? AND ( status LIKE ? OR status LIKE ? ) )
1730 OR ( user = ? AND status = ? ) )";
1731 @bind = ('nwiger', 'pending', 'dispatched', 'robot', 'unassigned');
1733 This can be combined with the C<-nest> operator to properly group
1740 ["-and", workhrs => {'>', 20}, geo => 'ASIA' ],
1741 ["-and", workhrs => {'<', 50}, geo => 'EURO' ]
1748 WHERE ( user = ? AND
1749 ( ( workhrs > ? AND geo = ? )
1750 OR ( workhrs < ? AND geo = ? ) ) )
1754 Finally, sometimes only literal SQL will do. If you want to include
1755 literal SQL verbatim, you can specify it as a scalar reference, namely:
1757 my $inn = 'is Not Null';
1759 priority => { '<', 2 },
1765 $stmt = "WHERE priority < ? AND requestor is Not Null";
1768 Note that in this example, you only get one bind parameter back, since
1769 the verbatim SQL is passed as part of the statement.
1771 Of course, just to prove a point, the above can also be accomplished
1775 priority => { '<', 2 },
1776 requestor => { '!=', undef },
1782 Conditions on boolean columns can be expressed in the
1783 same way, passing a reference to an empty string :
1786 priority => { '<', 2 },
1792 $stmt = "WHERE priority < ? AND is_ready";
1796 =head2 Literal SQL with placeholders and bind values (subqueries)
1798 If the literal SQL to be inserted has placeholders and bind values,
1799 use a reference to an arrayref (yes this is a double reference --
1800 not so common, but perfectly legal Perl). For example, to find a date
1801 in Postgres you can use something like this:
1804 date_column => \[q/= date '2008-09-30' - ?::integer/, 10/]
1809 $stmt = "WHERE ( date_column = date '2008-09-30' - ?::integer )"
1813 Literal SQL is especially useful for nesting parenthesized clauses in the
1814 main SQL query. Here is a first example :
1816 my ($sub_stmt, @sub_bind) = ("SELECT c1 FROM t1 WHERE c2 < ? AND c3 LIKE ?",
1820 bar => \["IN ($sub_stmt)" => @sub_bind],
1825 $stmt = "WHERE (foo = ? AND bar IN (SELECT c1 FROM t1
1826 WHERE c2 < ? AND c3 LIKE ?))";
1827 @bind = (1234, 100, "foo%");
1829 Other subquery operators, like for example C<"E<gt> ALL"> or C<"NOT IN">,
1830 are expressed in the same way. Of course the C<$sub_stmt> and
1831 its associated bind values can be generated through a former call
1834 my ($sub_stmt, @sub_bind)
1835 = $sql->select("t1", "c1", {c2 => {"<" => 100},
1836 c3 => {-like => "foo%"}});
1839 bar => \["> ALL ($sub_stmt)" => @sub_bind],
1842 In the examples above, the subquery was used as an operator on a column;
1843 but the same principle also applies for a clause within the main C<%where>
1844 hash, like an EXISTS subquery :
1846 my ($sub_stmt, @sub_bind)
1847 = $sql->select("t1", "*", {c1 => 1, c2 => \"> t0.c0"});
1850 -nest => \["EXISTS ($sub_stmt)" => @sub_bind],
1855 $stmt = "WHERE (foo = ? AND EXISTS (SELECT * FROM t1
1856 WHERE c1 = ? AND c2 > t0.c0))";
1860 Observe that the condition on C<c2> in the subquery refers to
1861 column C<t0.c0> of the main query : this is I<not> a bind
1862 value, so we have to express it through a scalar ref.
1863 Writing C<< c2 => {">" => "t0.c0"} >> would have generated
1864 C<< c2 > ? >> with bind value C<"t0.c0"> ... not exactly
1865 what we wanted here.
1867 Another use of the subquery technique is when some SQL clauses need
1868 parentheses, as it often occurs with some proprietary SQL extensions
1869 like for example fulltext expressions, geospatial expressions,
1870 NATIVE clauses, etc. Here is an example of a fulltext query in MySQL :
1873 -nest => \["MATCH (col1, col2) AGAINST (?)" => qw/apples/]
1876 Finally, here is an example where a subquery is used
1877 for expressing unary negation:
1879 my ($sub_stmt, @sub_bind)
1880 = $sql->where({age => [{"<" => 10}, {">" => 20}]});
1881 $sub_stmt =~ s/^ where //i; # don't want "WHERE" in the subclause
1883 lname => {like => '%son%'},
1884 -nest => \["NOT ($sub_stmt)" => @sub_bind],
1889 $stmt = "lname LIKE ? AND NOT ( age < ? OR age > ? )"
1890 @bind = ('%son%', 10, 20)
1896 These pages could go on for a while, since the nesting of the data
1897 structures this module can handle are pretty much unlimited (the
1898 module implements the C<WHERE> expansion as a recursive function
1899 internally). Your best bet is to "play around" with the module a
1900 little to see how the data structures behave, and choose the best
1901 format for your data based on that.
1903 And of course, all the values above will probably be replaced with
1904 variables gotten from forms or the command line. After all, if you
1905 knew everything ahead of time, you wouldn't have to worry about
1906 dynamically-generating SQL and could just hardwire it into your
1912 =head1 ORDER BY CLAUSES
1914 Some functions take an order by clause. This can either be a scalar (just a
1915 column name,) a hash of C<< { -desc => 'col' } >> or C<< { -asc => 'col' } >>,
1916 or an array of either of the two previous forms. Examples:
1918 Given | Will Generate
1919 ----------------------------------------------------------
1920 \'colA DESC' | ORDER BY colA DESC
1921 'colA' | ORDER BY colA
1922 [qw/colA colB/] | ORDER BY colA, colB
1923 {-asc => 'colA'} | ORDER BY colA ASC
1924 {-desc => 'colB'} | ORDER BY colB DESC
1926 {-asc => 'colA'}, | ORDER BY colA ASC, colB DESC
1929 [colA => {-asc => 'colB'}] | ORDER BY colA, colB ASC
1930 ==========================================================
1934 =head1 SPECIAL OPERATORS
1936 my $sqlmaker = SQL::Abstract->new(special_ops => [
1939 my ($self, $field, $op, $arg) = @_;
1945 A "special operator" is a SQL syntactic clause that can be
1946 applied to a field, instead of a usual binary operator.
1949 WHERE field IN (?, ?, ?)
1950 WHERE field BETWEEN ? AND ?
1951 WHERE MATCH(field) AGAINST (?, ?)
1953 Special operators IN and BETWEEN are fairly standard and therefore
1954 are builtin within C<SQL::Abstract>. For other operators,
1955 like the MATCH .. AGAINST example above which is
1956 specific to MySQL, you can write your own operator handlers :
1957 supply a C<special_ops> argument to the C<new> method.
1958 That argument takes an arrayref of operator definitions;
1959 each operator definition is a hashref with two entries
1965 the regular expression to match the operator
1969 coderef that will be called when meeting that operator
1970 in the input tree. The coderef will be called with
1971 arguments C<< ($self, $field, $op, $arg) >>, and
1972 should return a C<< ($sql, @bind) >> structure.
1976 For example, here is an implementation
1977 of the MATCH .. AGAINST syntax for MySQL
1979 my $sqlmaker = SQL::Abstract->new(special_ops => [
1981 # special op for MySql MATCH (field) AGAINST(word1, word2, ...)
1982 {regex => qr/^match$/i,
1984 my ($self, $field, $op, $arg) = @_;
1985 $arg = [$arg] if not ref $arg;
1986 my $label = $self->_quote($field);
1987 my ($placeholder) = $self->_convert('?');
1988 my $placeholders = join ", ", (($placeholder) x @$arg);
1989 my $sql = $self->_sqlcase('match') . " ($label) "
1990 . $self->_sqlcase('against') . " ($placeholders) ";
1991 my @bind = $self->_bindtype($field, @$arg);
1992 return ($sql, @bind);
2001 Thanks to some benchmarking by Mark Stosberg, it turns out that
2002 this module is many orders of magnitude faster than using C<DBIx::Abstract>.
2003 I must admit this wasn't an intentional design issue, but it's a
2004 byproduct of the fact that you get to control your C<DBI> handles
2007 To maximize performance, use a code snippet like the following:
2009 # prepare a statement handle using the first row
2010 # and then reuse it for the rest of the rows
2012 for my $href (@array_of_hashrefs) {
2013 $stmt ||= $sql->insert('table', $href);
2014 $sth ||= $dbh->prepare($stmt);
2015 $sth->execute($sql->values($href));
2018 The reason this works is because the keys in your C<$href> are sorted
2019 internally by B<SQL::Abstract>. Thus, as long as your data retains
2020 the same structure, you only have to generate the SQL the first time
2021 around. On subsequent queries, simply use the C<values> function provided
2022 by this module to return your values in the correct order.
2027 If you use my C<CGI::FormBuilder> module at all, you'll hopefully
2028 really like this part (I do, at least). Building up a complex query
2029 can be as simple as the following:
2033 use CGI::FormBuilder;
2036 my $form = CGI::FormBuilder->new(...);
2037 my $sql = SQL::Abstract->new;
2039 if ($form->submitted) {
2040 my $field = $form->field;
2041 my $id = delete $field->{id};
2042 my($stmt, @bind) = $sql->update('table', $field, {id => $id});
2045 Of course, you would still have to connect using C<DBI> to run the
2046 query, but the point is that if you make your form look like your
2047 table, the actual query script can be extremely simplistic.
2049 If you're B<REALLY> lazy (I am), check out C<HTML::QuickTable> for
2050 a fast interface to returning and formatting data. I frequently
2051 use these three modules together to write complex database query
2052 apps in under 50 lines.
2057 Version 1.50 was a major internal refactoring of C<SQL::Abstract>.
2058 Great care has been taken to preserve the I<published> behavior
2059 documented in previous versions in the 1.* family; however,
2060 some features that were previously undocumented, or behaved
2061 differently from the documentation, had to be changed in order
2062 to clarify the semantics. Hence, client code that was relying
2063 on some dark areas of C<SQL::Abstract> v1.*
2064 B<might behave differently> in v1.50.
2066 The main changes are :
2072 support for literal SQL through the C<< \ [$sql, bind] >> syntax.
2076 support for the { operator => \"..." } construct (to embed literal SQL)
2080 added -nest1, -nest2 or -nest_1, -nest_2, ...
2084 optional support for L<array datatypes|/"Inserting and Updating Arrays">
2088 defensive programming : check arguments
2092 fixed bug with global logic, which was previously implemented
2093 through global variables yielding side-effects. Prior versons would
2094 interpret C<< [ {cond1, cond2}, [cond3, cond4] ] >>
2095 as C<< "(cond1 AND cond2) OR (cond3 AND cond4)" >>.
2096 Now this is interpreted
2097 as C<< "(cond1 AND cond2) OR (cond3 OR cond4)" >>.
2101 C<-and> / C<-or> operators are no longer accepted
2102 in the middle of an arrayref : they are
2103 only admitted if in first position.
2107 changed logic for distributing an op over arrayrefs
2111 fixed semantics of _bindtype on array args
2115 dropped the C<_anoncopy> of the %where tree. No longer necessary,
2116 we just avoid shifting arrays within that tree.
2120 dropped the C<_modlogic> function
2126 =head1 ACKNOWLEDGEMENTS
2128 There are a number of individuals that have really helped out with
2129 this module. Unfortunately, most of them submitted bugs via CPAN
2130 so I have no idea who they are! But the people I do know are:
2132 Ash Berlin (order_by hash term support)
2133 Matt Trout (DBIx::Class support)
2134 Mark Stosberg (benchmarking)
2135 Chas Owens (initial "IN" operator support)
2136 Philip Collins (per-field SQL functions)
2137 Eric Kolve (hashref "AND" support)
2138 Mike Fragassi (enhancements to "BETWEEN" and "LIKE")
2139 Dan Kubb (support for "quote_char" and "name_sep")
2140 Guillermo Roditi (patch to cleanup "IN" and "BETWEEN", fix and tests for _order_by)
2141 Laurent Dami (internal refactoring, multiple -nest, extensible list of special operators, literal SQL)
2147 L<DBIx::Class>, L<DBIx::Abstract>, L<CGI::FormBuilder>, L<HTML::QuickTable>.
2151 Copyright (c) 2001-2007 Nathan Wiger <nwiger@cpan.org>. All Rights Reserved.
2153 This module is actively maintained by Matt Trout <mst@shadowcatsystems.co.uk>
2155 For support, your best bet is to try the C<DBIx::Class> users mailing list.
2156 While not an official support venue, C<DBIx::Class> makes heavy use of
2157 C<SQL::Abstract>, and as such list members there are very familiar with
2158 how to create queries.
2160 This module is free software; you may copy this under the terms of
2161 the GNU General Public License, or the Artistic License, copies of
2162 which should have accompanied your Perl kit.