6 with_assign => "+ - * / % ** << >> x .",
7 assign => "+= -= *= /= %= **= <<= >>= x= .=",
8 num_comparison => "< <= > >= == !=",
9 '3way_comparison' => "<=> cmp",
10 str_comparison => "lt le gt ge eq ne",
11 binary => '& &= | |= ^ ^=',
14 func => "atan2 cos sin exp abs log sqrt int",
15 conversion => 'bool "" 0+ qr',
18 dereferencing => '${} @{} %{} &{} *{}',
20 special => 'nomethod fallback =',
24 for $category (keys %ops) {
25 $ops_seen{$_}++ for (split /\s+/, $ops{$category});
34 *{$package . "::(("} = \&nil; # Make it findable via fetchmethod.
36 if ($_ eq 'fallback') {
37 for my $sym (*{$package . "::()"}) {
38 *$sym = \&nil; # Make it findable via fetchmethod.
42 warnings::warnif("overload arg '$_' is invalid")
46 $ {$package . "::(" . $_} = $sub;
49 #print STDERR "Setting '$ {'package'}::\cO$_' to \\&'$sub'.\n";
50 *{$package . "::(" . $_} = \&{ $sub };
56 $package = (caller())[0];
57 # *{$package . "::OVERLOAD"} = \&OVERLOAD;
59 $package->overload::OVERLOAD(@_);
63 $package = (caller())[0];
65 *{$package . "::(("} = \&nil;
67 warnings::warnif("overload arg '$_' is invalid")
69 delete $ {$package . "::"}{$_ eq 'fallback' ? '()' : "(" .$_};
75 $package = ref $package if ref $package;
76 mycan ($package, '()') || mycan ($package, '((');
81 return undef unless $globref;
82 my $sub = \&{*$globref};
84 return $sub if !ref $sub or $sub != \&nil;
85 return shift->can($ {*$globref});
88 sub OverloadedStringify {
90 $package = ref $package if ref $package;
92 ov_method mycan($package, '(""'), $package
93 or ov_method mycan($package, '(0+'), $package
94 or ov_method mycan($package, '(bool'), $package
95 or ov_method mycan($package, '(nomethod'), $package;
103 require Scalar::Util;
104 $package = Scalar::Util::blessed($package);
105 return undef if !defined $package;
107 #my $meth = $package->can('(' . shift);
108 ov_method mycan($package, '(' . shift), $package;
109 #return $meth if $meth ne \&nil;
110 #return $ {*{$meth}};
120 sub mycan { # Real can would leave stubs.
121 my ($package, $meth) = @_;
127 my $mro = mro::get_linear_isa($package);
128 foreach my $p (@$mro) {
129 my $fqmeth = $p . q{::} . $meth;
130 return \*{$fqmeth} if defined &{$fqmeth};
137 'integer' => 0x1000, # HINT_NEW_INTEGER
138 'float' => 0x2000, # HINT_NEW_FLOAT
139 'binary' => 0x4000, # HINT_NEW_BINARY
140 'q' => 0x8000, # HINT_NEW_STRING
141 'qr' => 0x10000, # HINT_NEW_RE
144 use warnings::register;
146 # Arguments: what, sub
149 warnings::warnif ("Odd number of arguments for overload::constant");
152 elsif (!exists $constants {$_ [0]}) {
153 warnings::warnif ("'$_[0]' is not an overloadable type");
155 elsif (!ref $_ [1] || "$_[1]" !~ /(^|=)CODE\(0x[0-9a-f]+\)$/) {
156 # Can't use C<ref $_[1] eq "CODE"> above as code references can be
157 # blessed, and C<ref> would return the package the ref is blessed into.
158 if (warnings::enabled) {
159 $_ [1] = "undef" unless defined $_ [1];
160 warnings::warn ("'$_[1]' is not a code reference");
165 $^H |= $constants{$_[0]};
171 sub remove_constant {
172 # Arguments: what, sub
175 $^H &= ~ $constants{$_[0]};
186 overload - Package for overloading Perl operations
199 $a = SomeThing->new( 57 );
202 if (overload::Overloaded $b) {...}
204 $strval = overload::StrVal $b;
208 This pragma allows overloading of Perl's operators for a class.
209 To overload built-in functions, see L<perlsub/Overriding Built-in Functions> instead.
215 Arguments of the C<use overload> directive are (key, value) pairs.
216 For the full set of legal keys, see L<Overloadable Operations> below.
218 Operator implementations (the values) can be subroutines,
219 references to subroutines, or anonymous subroutines
220 - in other words, anything legal inside a C<&{ ... }> call.
221 Values specified as strings are interpreted as method names.
228 '""' => sub { ...; };
230 declares that subtraction is to be implemented by method C<minus()>
231 in the class C<Number> (or one of its base classes),
232 and that the function C<Number::muas()> is to be used for the
233 assignment form of multiplication, C<*=>.
234 It also defines an anonymous subroutine to implement stringification:
235 this is called whenever an object blessed into the package C<Number>
236 is used in a string context (this subroutine might, for example,
237 return the number as a Roman numeral).
239 =head3 Calling Conventions and Magic Autogeneration
241 The following sample implementation of C<minus()> (which assumes
242 that C<Number> objects are simply blessed references to scalars)
243 illustrates the calling conventions:
247 my ($self, $other, $swap) = @_;
248 my $result = $$self - $other; # *
249 $result = -$result if $swap;
250 ref $result ? $result : bless \$result;
252 # * may recurse once - see table below
254 Three arguments are passed to all subroutines specified in the
255 C<use overload> directive (with one exception - see L</nomethod>).
256 The first of these is the operand providing the overloaded
257 operator implementation -
258 in this case, the object whose C<minus()> method is being called.
260 The second argument is the other operand, or C<undef> in the
261 case of a unary operator.
263 The third argument is set to TRUE if (and only if) the two
264 operands have been swapped. Perl may do this to ensure that the
265 first argument (C<$self>) is an object implementing the overloaded
266 operation, in line with general object calling conventions.
267 For example, if C<$x> and C<$y> are C<Number>s:
269 operation | generates a call to
270 ============|======================
271 $x - $y | minus($x, $y, '')
272 $x - 7 | minus($x, 7, '')
273 7 - $x | minus($x, 7, 1)
275 Perl may also use C<minus()> to implement other operators which
276 have not been specified in the C<use overload> directive,
277 according to the rules for L<Magic Autogeneration> described later.
278 For example, the C<use overload> above declared no subroutine
279 for any of the operators C<-->, C<neg> (the overload key for
280 unary minus), or C<-=>. Thus
282 operation | generates a call to
283 ============|======================
284 -$x | minus($x, 0, 1)
285 $x-- | minus($x, 1, undef)
286 $x -= 3 | minus($x, 3, undef)
289 where autogeneration results in the method for a standard
290 operator which does not change either of its operands, such
291 as C<->, being used to implement an operator which changes
292 the operand ("mutators": here, C<--> and C<-=>),
293 Perl passes undef as the third argument.
294 This still evaluates as FALSE, consistent with the fact that
295 the operands have not been swapped, but gives the subroutine
296 a chance to alter its behaviour in these cases.
298 In all the above examples, C<minus()> is required
299 only to return the result of the subtraction:
300 Perl takes care of the assignment to $x.
301 In fact, such methods should I<not> modify their operands,
302 even if C<undef> is passed as the third argument
303 (see L<Overloadable Operations>).
305 The same is not true of implementations of C<++> and C<-->:
306 these are expected to modify their operand.
307 An appropriate implementation of C<--> might look like
309 use overload '--' => "decr",
311 sub decr { --${$_[0]}; }
313 =head3 Mathemagic, Mutators, and Copy Constructors
315 The term 'mathemagic' describes the overloaded implementation
316 of mathematical operators.
317 Mathemagical operations raise an issue.
323 If C<$a> and C<$b> are scalars then after these statements
327 An object, however, is a reference to blessed data, so if
328 C<$a> and C<$b> are objects then the assignment C<$a = $b>
329 copies only the reference, leaving C<$a> and C<$b> referring
330 to the same object data.
331 One might therefore expect the operation C<--$a> to decrement
332 C<$b> as well as C<$a>.
333 However, this would not be consistent with how we expect the
334 mathematical operators to work.
336 Perl resolves this dilemma by transparently calling a copy
337 constructor before calling a method defined to implement
338 a mutator (C<-->, C<+=>, and so on.).
339 In the above example, when Perl reaches the decrement
340 statement, it makes a copy of the object data in C<$a> and
341 assigns to C<$a> a reference to the copied data.
342 Only then does it call C<decr()>, which alters the copied
343 data, leaving C<$b> unchanged.
344 Thus the object metaphor is preserved as far as possible,
345 while mathemagical operations still work according to the
348 Note: the preceding paragraph describes what happens when
349 Perl autogenerates the copy constructor for an object based
351 For other cases, see L<Copy Constructor>.
353 =head2 Overloadable Operations
355 The complete list of keys that can be specified in the C<use overload>
356 directive are given, separated by spaces, in the values of the
357 hash C<%overload::ops>:
359 with_assign => '+ - * / % ** << >> x .',
360 assign => '+= -= *= /= %= **= <<= >>= x= .=',
361 num_comparison => '< <= > >= == !=',
362 '3way_comparison'=> '<=> cmp',
363 str_comparison => 'lt le gt ge eq ne',
364 binary => '& &= | |= ^ ^=',
367 func => 'atan2 cos sin exp abs log sqrt int',
368 conversion => 'bool "" 0+ qr',
371 dereferencing => '${} @{} %{} &{} *{}',
373 special => 'nomethod fallback ='
375 Most of the overloadable operators map one-to-one to these keys.
376 Exceptions, including additional overloadable operations not
377 apparent from this hash, are included in the notes which follow.
379 A warning is issued if an attempt is made to register an operator not found
386 The operator C<not> is not a valid key for C<use overload>.
387 However, if the operator C<!> is overloaded then the same
388 implementation will be used for C<not>
389 (since the two operators differ only in precedence).
393 The key C<neg> is used for unary minus to disambiguate it from
398 Assuming they are to behave analogously to Perl's C<++> and C<-->,
399 overloaded implementations of these operators are required to
400 mutate their operands.
402 No distinction is made between prefix and postfix forms of the
403 increment and decrement operators: these differ only in the
404 point at which Perl calls the associated subroutine when
405 evaluating an expression.
407 =item * I<Assignments>
409 += -= *= /= %= **= <<= >>= x= .=
412 Simple assignment is not overloadable (the C<'='> key is used
413 for the L<Copy Constructor>).
414 Perl does have a way to make assignments to an object do whatever
415 you want, but this involves using tie(), not overload -
416 see L<perlfunc/tie> and the L</COOKBOOK> examples below.
418 The subroutine for the assignment variant of an operator is
419 required only to return the result of the operation.
420 It is permitted to change the value of its operand
421 (this is safe because Perl calls the copy constructor first),
422 but this is optional since Perl assigns the returned value to
423 the left-hand operand anyway.
425 An object that overloads an assignment operator does so only in
426 respect of assignments to that object.
427 In other words, Perl never calls the corresponding methods with
428 the third argument (the "swap" argument) set to TRUE.
429 For example, the operation
433 cannot lead to C<$b>'s implementation of C<*=> being called,
434 even if C<$a> is a scalar.
435 (It can, however, generate a call to C<$b>'s method for C<*>).
437 =item * I<Non-mutators with a mutator variant>
439 + - * / % ** << >> x .
442 As described L<above|"Calling Conventions and Magic Autogeneration">,
443 Perl may call methods for operators like C<+> and C<&> in the course
444 of implementing missing operations like C<++>, C<+=>, and C<&=>.
445 While these methods may detect this usage by testing the definedness
446 of the third argument, they should in all cases avoid changing their
448 This is because Perl does not call the copy constructor before
449 invoking these methods.
453 Traditionally, the Perl function C<int> rounds to 0
454 (see L<perlfunc/int>), and so for floating-point-like types one
455 should follow the same semantic.
457 =item * I<String, numeric, boolean, and regexp conversions>
461 These conversions are invoked according to context as necessary.
462 For example, the subroutine for C<'""'> (stringify) may be used
463 where the overloaded object is passed as an argument to C<print>,
464 and that for C<'bool'> where it is tested in the condition of a flow
465 control statement (like C<while>) or the ternary C<?:> operation.
467 Of course, in contexts like, for example, C<$obj + 1>, Perl will
468 invoke C<$obj>'s implementation of C<+> rather than (in this
469 example) converting C<$obj> to a number using the numify method
470 C<'0+'> (an exception to this is when no method has been provided
471 for C<'+'> and L</fallback> is set to TRUE).
473 The subroutines for C<'""'>, C<'0+'>, and C<'bool'> can return
474 any arbitrary Perl value.
475 If the corresponding operation for this value is overloaded too,
476 the operation will be called again with this value.
478 As a special case if the overload returns the object itself then it will
479 be used directly. An overloaded conversion returning the object is
480 probably a bug, because you're likely to get something that looks like
481 C<YourPackage=HASH(0x8172b34)>.
485 The subroutine for C<'qr'> is used wherever the object is
486 interpolated into or used as a regexp, including when it
487 appears on the RHS of a C<=~> or C<!~> operator.
489 C<qr> must return a compiled regexp, or a ref to a compiled regexp
490 (such as C<qr//> returns), and any further overloading on the return
491 value will be ignored.
495 If C<E<lt>E<gt>> is overloaded then the same implementation is used
496 for both the I<read-filehandle> syntax C<E<lt>$varE<gt>> and
497 I<globbing> syntax C<E<lt>${var}E<gt>>.
499 B<BUGS> Even in list context, the iterator is currently called only
500 once and with scalar context.
502 =item * I<File tests>
504 The key C<'-X'> is used to specify a subroutine to handle all the
505 filetest operators (C<-f>, C<-x>, and so on: see L<perlfunc/-X> for
507 it is not possible to overload any filetest operator individually.
508 To distinguish them, the letter following the '-' is passed as the
509 second argument (that is, in the slot that for binary operators
510 is used to pass the second operand).
512 Calling an overloaded filetest operator does not affect the stat value
513 associated with the special filehandle C<_>. It still refers to the
514 result of the last C<stat>, C<lstat> or unoverloaded filetest.
516 This overload was introduced in Perl 5.12.
520 The key C<"~~"> allows you to override the smart matching logic used by
521 the C<~~> operator and the switch construct (C<given>/C<when>). See
522 L<perlsyn/Switch Statements> and L<feature>.
524 Unusually, the overloaded implementation of the smart match operator
525 does not get full control of the smart match behaviour.
526 In particular, in the following code:
529 use overload '~~' => 'match';
531 my $obj = Foo->new();
534 the smart match does I<not> invoke the method call like this:
536 $obj->match([1,2,3],0);
538 rather, the smart match distributive rule takes precedence, so $obj is
539 smart matched against each array element in turn until a match is found,
540 so you may see between one and three of these calls instead:
546 Consult the match table in L<perlop/"Smartmatch Operator"> for
547 details of when overloading is invoked.
549 =item * I<Dereferencing>
553 If these operators are not explicitly overloaded then they
554 work in the normal way, yielding the underlying scalar,
555 array, or whatever stores the object data (or the appropriate
556 error message if the dereference operator doesn't match it).
557 Defining a catch-all C<'nomethod'> (see L<below|/nomethod>)
558 makes no difference to this as the catch-all function will
559 not be called to implement a missing dereference operator.
561 If a dereference operator is overloaded then it must return a
562 I<reference> of the appropriate type (for example, the
563 subroutine for key C<'${}'> should return a reference to a
564 scalar, not a scalar), or another object which overloads the
565 operator: that is, the subroutine only determines what is
566 dereferenced and the actual dereferencing is left to Perl.
567 As a special case, if the subroutine returns the object itself
568 then it will not be called again - avoiding infinite recursion.
574 See L<Special Keys for C<use overload>>.
578 =head2 Magic Autogeneration
580 If a method for an operation is not found then Perl tries to
581 autogenerate a substitute implementation from the operations
582 that have been defined.
584 Note: the behaviour described in this section can be disabled
585 by setting C<fallback> to FALSE (see L</fallback>).
587 In the following tables, numbers indicate priority.
588 For example, the table below states that,
589 if no implementation for C<'!'> has been defined then Perl will
590 implement it using C<'bool'> (that is, by inverting the value
591 returned by the method for C<'bool'>);
592 if boolean conversion is also unimplemented then Perl will
593 use C<'0+'> or, failing that, C<'""'>.
595 operator | can be autogenerated from
598 =========|==========================
612 Note: The iterator (C<'E<lt>E<gt>'>) and file test (C<'-X'>)
613 operators work as normal: if the operand is not a blessed glob or
614 IO reference then it is converted to a string (using the method
615 for C<'""'>, C<'0+'>, or C<'bool'>) to be interpreted as a glob
618 operator | can be autogenerated from
621 =========|==========================
625 abs | a1 a2 b1 b2 [*]
633 * one from [a1, a2] and one from [b1, b2]
635 Just as numeric comparisons can be autogenerated from the method
636 for C<< '<=>' >>, string comparisons can be autogenerated from
639 operators | can be autogenerated from
640 ====================|===========================
641 lt gt le ge eq ne | cmp
643 Similarly, autogeneration for keys C<'+='> and C<'++'> is analogous
644 to C<'-='> and C<'--'> above:
646 operator | can be autogenerated from
649 =========|==========================
653 And other assignment variations are analogous to
654 C<'+='> and C<'-='> (and similar to C<'.='> and C<'x='> above):
656 operator || *= /= %= **= <<= >>= &= ^= |=
657 -------------------||--------------------------------
658 autogenerated from || * / % ** << >> & ^ |
660 Note also that the copy constructor (key C<'='>) may be
661 autogenerated, but only for objects based on scalars.
662 See L<Copy Constructor>.
664 =head3 Minimal Set of Overloaded Operations
666 Since some operations can be automatically generated from others, there is
667 a minimal set of operations that need to be overloaded in order to have
668 the complete set of overloaded operations at one's disposal.
669 Of course, the autogenerated operations may not do exactly what the user
670 expects. The minimal set is:
675 atan2 cos sin exp log sqrt int
679 Of the conversions, only one of string, boolean or numeric is
680 needed because each can be generated from either of the other two.
682 =head2 Special Keys for C<use overload>
686 The C<'nomethod'> key is used to specify a catch-all function to
687 be called for any operator that is not individually overloaded.
688 The specified function will be passed four parameters.
689 The first three arguments coincide with those that would have been
690 passed to the corresponding method if it had been defined.
691 The fourth argument is the C<use overload> key for that missing
694 For example, if C<$a> is an object blessed into a package declaring
696 use overload 'nomethod' => 'catch_all', # ...
702 could (unless a method is specifically declared for the key
703 C<'+'>) result in a call
705 catch_all($a, 3, 1, '+')
707 See L<How Perl Chooses an Operator Implementation>.
711 The value assigned to the key C<'fallback'> tells Perl how hard
712 it should try to find an alternative way to implement a missing
717 =item * defined, but FALSE
719 use overload "fallback" => 0, # ... ;
721 This disables L<Magic Autogeneration>.
725 In the default case where no value is explicitly assigned to
726 C<fallback>, magic autogeneration is enabled.
730 The same as for C<undef>, but if a missing operator cannot be
731 autogenerated then, instead of issuing an error message, Perl
732 is allowed to revert to what it would have done for that
733 operator if there had been no C<use overload> directive.
735 Note: in most cases, particularly the L<Copy Constructor>,
736 this is unlikely to be appropriate behaviour.
740 See L<How Perl Chooses an Operator Implementation>.
742 =head3 Copy Constructor
744 As mentioned L<above|"Mathemagic, Mutators, and Copy Constructors">,
745 this operation is called when a mutator is applied to a reference
746 that shares its object with some other reference.
747 For example, if C<$b> is mathemagical, and C<'++'> is overloaded
748 with C<'incr'>, and C<'='> is overloaded with C<'clone'>, then the
752 # ... (other code which does not modify $a or $b) ...
755 would be executed in a manner equivalent to
759 $b = $b->clone(undef, "");
768 The subroutine for C<'='> does not overload the Perl assignment
769 operator: it is used only to allow mutators to work as described
770 here. (See L</Assignments> above.)
774 As for other operations, the subroutine implementing '=' is passed
775 three arguments, though the last two are always C<undef> and C<''>.
779 The copy constructor is called only before a call to a function
780 declared to implement a mutator, for example, if C<++$b;> in the
781 code above is effected via a method declared for key C<'++'>
782 (or 'nomethod', passed C<'++'> as the fourth argument) or, by
783 autogeneration, C<'+='>.
784 It is not called if the increment operation is effected by a call
785 to the method for C<'+'> since, in the equivalent code,
790 the data referred to by C<$a> is unchanged by the assignment to
791 C<$b> of a reference to new object data.
795 The copy constructor is not called if Perl determines that it is
796 unnecessary because there is no other reference to the data being
801 If C<'fallback'> is undefined or TRUE then a copy constructor
802 can be autogenerated, but only for objects based on scalars.
803 In other cases it needs to be defined explicitly.
804 Where an object's data is stored as, for example, an array of
805 scalars, the following might be appropriate:
807 use overload '=' => sub { bless [ @{$_[0]} ] }, # ...
811 If C<'fallback'> is TRUE and no copy constructor is defined then,
812 for objects not based on scalars, Perl may silently fall back on
813 simple assignment - that is, assignment of the object reference.
814 In effect, this disables the copy constructor mechanism since
815 no new copy of the object data is created.
816 This is almost certainly not what you want.
817 (It is, however, consistent: for example, Perl's fallback for the
818 C<++> operator is to increment the reference itself.)
822 =head2 How Perl Chooses an Operator Implementation
824 Which is checked first, C<nomethod> or C<fallback>?
825 If the two operands of an operator are of different types and
826 both overload the operator, which implementation is used?
827 The following are the precedence rules:
833 If the first operand has declared a subroutine to overload the
834 operator then use that implementation.
838 Otherwise, if fallback is TRUE or undefined for the
839 first operand then see if the
840 L<rules for autogeneration|"Magic Autogeneration">
841 allows another of its operators to be used instead.
845 Unless the operator is an assignment (C<+=>, C<-=>, etc.),
846 repeat step (1) in respect of the second operand.
850 Repeat Step (2) in respect of the second operand.
854 If the first operand has a "nomethod" method then use that.
858 If the second operand has a "nomethod" method then use that.
862 If C<fallback> is TRUE for both operands
863 then perform the usual operation for the operator,
864 treating the operands as numbers, strings, or booleans
865 as appropriate for the operator (see note).
869 Nothing worked - die.
873 Where there is only one operand (or only one operand with
874 overloading) the checks in respect of the other operand above are
877 There are exceptions to the above rules for dereference operations
878 (which, if Step 1 fails, always fall back to the normal, built-in
879 implementations - see Dereferencing), and for C<~~> (which has its
880 own set of rules - see C<Matching> under L</Overloadable Operations>
883 Note on Step 7: some operators have a different semantic depending
884 on the type of their operands.
885 As there is no way to instruct Perl to treat the operands as, e.g.,
886 numbers instead of strings, the result here may not be what you
888 See L<BUGS AND PITFALLS>.
890 =head2 Losing Overloading
892 The restriction for the comparison operation is that even if, for example,
893 C<cmp> should return a blessed reference, the autogenerated C<lt>
894 function will produce only a standard logical value based on the
895 numerical value of the result of C<cmp>. In particular, a working
896 numeric conversion is needed in this case (possibly expressed in terms of
899 Similarly, C<.=> and C<x=> operators lose their mathemagical properties
900 if the string conversion substitution is applied.
902 When you chop() a mathemagical object it is promoted to a string and its
903 mathemagical properties are lost. The same can happen with other
906 =head2 Inheritance and Overloading
908 Overloading respects inheritance via the @ISA hierarchy.
909 Inheritance interacts with overloading in two ways.
913 =item Method names in the C<use overload> directive
917 use overload key => value;
919 is a string, it is interpreted as a method name - which may
920 (in the usual way) be inherited from another class.
922 =item Overloading of an operation is inherited by derived classes
924 Any class derived from an overloaded class is also overloaded
925 and inherits its operator implementations.
926 If the same operator is overloaded in more than one ancestor
927 then the implementation is determined by the usual inheritance
930 For example, if C<A> inherits from C<B> and C<C> (in that order),
931 C<B> overloads C<+> with C<\&D::plus_sub>, and C<C> overloads
932 C<+> by C<"plus_meth">, then the subroutine C<D::plus_sub> will
933 be called to implement operation C<+> for an object in package C<A>.
937 Note that in Perl version prior to 5.18 inheritance of the C<fallback> key
938 was not governed by the above rules. The value of C<fallback> in the first
939 overloaded ancestor was used. This was fixed in 5.18 to follow the usual
940 rules of inheritance.
942 =head2 Run-time Overloading
944 Since all C<use> directives are executed at compile-time, the only way to
945 change overloading during run-time is to
947 eval 'use overload "+" => \&addmethod';
951 eval 'no overload "+", "--", "<="';
953 though the use of these constructs during run-time is questionable.
955 =head2 Public Functions
957 Package C<overload.pm> provides the following public functions:
961 =item overload::StrVal(arg)
963 Gives the string value of C<arg> as in the
964 absence of stringify overloading. If you
965 are using this to get the address of a reference (useful for checking if two
966 references point to the same thing) then you may be better off using
967 C<Scalar::Util::refaddr()>, which is faster.
969 =item overload::Overloaded(arg)
971 Returns true if C<arg> is subject to overloading of some operations.
973 =item overload::Method(obj,op)
975 Returns C<undef> or a reference to the method that implements C<op>.
979 =head2 Overloading Constants
981 For some applications, the Perl parser mangles constants too much.
982 It is possible to hook into this process via C<overload::constant()>
983 and C<overload::remove_constant()> functions.
985 These functions take a hash as an argument. The recognized keys of this hash
992 to overload integer constants,
996 to overload floating point constants,
1000 to overload octal and hexadecimal constants,
1004 to overload C<q>-quoted strings, constant pieces of C<qq>- and C<qx>-quoted
1005 strings and here-documents,
1009 to overload constant pieces of regular expressions.
1013 The corresponding values are references to functions which take three arguments:
1014 the first one is the I<initial> string form of the constant, the second one
1015 is how Perl interprets this constant, the third one is how the constant is used.
1016 Note that the initial string form does not
1017 contain string delimiters, and has backslashes in backslash-delimiter
1018 combinations stripped (thus the value of delimiter is not relevant for
1019 processing of this string). The return value of this function is how this
1020 constant is going to be interpreted by Perl. The third argument is undefined
1021 unless for overloaded C<q>- and C<qr>- constants, it is C<q> in single-quote
1022 context (comes from strings, regular expressions, and single-quote HERE
1023 documents), it is C<tr> for arguments of C<tr>/C<y> operators,
1024 it is C<s> for right-hand side of C<s>-operator, and it is C<qq> otherwise.
1026 Since an expression C<"ab$cd,,"> is just a shortcut for C<'ab' . $cd . ',,'>,
1027 it is expected that overloaded constant strings are equipped with reasonable
1028 overloaded catenation operator, otherwise absurd results will result.
1029 Similarly, negative numbers are considered as negations of positive constants.
1031 Note that it is probably meaningless to call the functions overload::constant()
1032 and overload::remove_constant() from anywhere but import() and unimport() methods.
1033 From these methods they may be called as
1038 die "unknown import: @_" unless @_ == 1 and $_[0] eq ':constant';
1039 overload::constant integer => sub {Math::BigInt->new(shift)};
1042 =head1 IMPLEMENTATION
1044 What follows is subject to change RSN.
1046 The table of methods for all operations is cached in magic for the
1047 symbol table hash for the package. The cache is invalidated during
1048 processing of C<use overload>, C<no overload>, new function
1049 definitions, and changes in @ISA.
1051 (Every SVish thing has a magic queue, and magic is an entry in that
1052 queue. This is how a single variable may participate in multiple
1053 forms of magic simultaneously. For instance, environment variables
1054 regularly have two forms at once: their %ENV magic and their taint
1055 magic. However, the magic which implements overloading is applied to
1056 the stashes, which are rarely used directly, thus should not slow down
1059 If a package uses overload, it carries a special flag. This flag is also
1060 set when new function are defined or @ISA is modified. There will be a
1061 slight speed penalty on the very first operation thereafter that supports
1062 overloading, while the overload tables are updated. If there is no
1063 overloading present, the flag is turned off. Thus the only speed penalty
1064 thereafter is the checking of this flag.
1066 It is expected that arguments to methods that are not explicitly supposed
1067 to be changed are constant (but this is not enforced).
1071 Please add examples to what follows!
1073 =head2 Two-face Scalars
1075 Put this in F<two_face.pm> in your Perl library directory:
1077 package two_face; # Scalars with separate string and
1079 sub new { my $p = shift; bless [@_], $p }
1080 use overload '""' => \&str, '0+' => \&num, fallback => 1;
1081 sub num {shift->[1]}
1082 sub str {shift->[0]}
1087 my $seven = two_face->new("vii", 7);
1088 printf "seven=$seven, seven=%d, eight=%d\n", $seven, $seven+1;
1089 print "seven contains 'i'\n" if $seven =~ /i/;
1091 (The second line creates a scalar which has both a string value, and a
1092 numeric value.) This prints:
1094 seven=vii, seven=7, eight=8
1097 =head2 Two-face References
1099 Suppose you want to create an object which is accessible as both an
1100 array reference and a hash reference.
1103 use overload '%{}' => \&gethash, '@{}' => sub { $ {shift()} };
1111 tie %h, ref $self, $self;
1115 sub TIEHASH { my $p = shift; bless \ shift, $p }
1118 $fields{$_} = $i++ foreach qw{zero one two three};
1120 my $self = ${shift()};
1121 my $key = $fields{shift()};
1122 defined $key or die "Out of band access";
1123 $$self->[$key] = shift;
1126 my $self = ${shift()};
1127 my $key = $fields{shift()};
1128 defined $key or die "Out of band access";
1132 Now one can access an object using both the array and hash syntax:
1134 my $bar = two_refs->new(3,4,5,6);
1136 $bar->{two} == 11 or die 'bad hash fetch';
1138 Note several important features of this example. First of all, the
1139 I<actual> type of $bar is a scalar reference, and we do not overload
1140 the scalar dereference. Thus we can get the I<actual> non-overloaded
1141 contents of $bar by just using C<$$bar> (what we do in functions which
1142 overload dereference). Similarly, the object returned by the
1143 TIEHASH() method is a scalar reference.
1145 Second, we create a new tied hash each time the hash syntax is used.
1146 This allows us not to worry about a possibility of a reference loop,
1147 which would lead to a memory leak.
1149 Both these problems can be cured. Say, if we want to overload hash
1150 dereference on a reference to an object which is I<implemented> as a
1151 hash itself, the only problem one has to circumvent is how to access
1152 this I<actual> hash (as opposed to the I<virtual> hash exhibited by the
1153 overloaded dereference operator). Here is one possible fetching routine:
1156 my ($self, $key) = (shift, shift);
1157 my $class = ref $self;
1158 bless $self, 'overload::dummy'; # Disable overloading of %{}
1159 my $out = $self->{$key};
1160 bless $self, $class; # Restore overloading
1164 To remove creation of the tied hash on each access, one may an extra
1165 level of indirection which allows a non-circular structure of references:
1168 use overload '%{}' => sub { ${shift()}->[1] },
1169 '@{}' => sub { ${shift()}->[0] };
1175 bless \ [$a, \%h], $p;
1180 tie %h, ref $self, $self;
1184 sub TIEHASH { my $p = shift; bless \ shift, $p }
1187 $fields{$_} = $i++ foreach qw{zero one two three};
1190 my $key = $fields{shift()};
1191 defined $key or die "Out of band access";
1196 my $key = $fields{shift()};
1197 defined $key or die "Out of band access";
1201 Now if $baz is overloaded like this, then C<$baz> is a reference to a
1202 reference to the intermediate array, which keeps a reference to an
1203 actual array, and the access hash. The tie()ing object for the access
1204 hash is a reference to a reference to the actual array, so
1210 There are no loops of references.
1214 Both "objects" which are blessed into the class C<two_refs1> are
1215 references to a reference to an array, thus references to a I<scalar>.
1216 Thus the accessor expression C<$$foo-E<gt>[$ind]> involves no
1217 overloaded operations.
1221 =head2 Symbolic Calculator
1223 Put this in F<symbolic.pm> in your Perl library directory:
1225 package symbolic; # Primitive symbolic calculator
1226 use overload nomethod => \&wrap;
1228 sub new { shift; bless ['n', @_] }
1230 my ($obj, $other, $inv, $meth) = @_;
1231 ($obj, $other) = ($other, $obj) if $inv;
1232 bless [$meth, $obj, $other];
1235 This module is very unusual as overloaded modules go: it does not
1236 provide any usual overloaded operators, instead it provides an
1237 implementation for L<C<nomethod>>. In this example the C<nomethod>
1238 subroutine returns an object which encapsulates operations done over
1239 the objects: C<< symbolic->new(3) >> contains C<['n', 3]>, C<< 2 +
1240 symbolic->new(3) >> contains C<['+', 2, ['n', 3]]>.
1242 Here is an example of the script which "calculates" the side of
1243 circumscribed octagon using the above package:
1246 my $iter = 1; # 2**($iter+2) = 8
1247 my $side = symbolic->new(1);
1251 $side = (sqrt(1 + $side**2) - 1)/$side;
1255 The value of $side is
1257 ['/', ['-', ['sqrt', ['+', 1, ['**', ['n', 1], 2]],
1258 undef], 1], ['n', 1]]
1260 Note that while we obtained this value using a nice little script,
1261 there is no simple way to I<use> this value. In fact this value may
1262 be inspected in debugger (see L<perldebug>), but only if
1263 C<bareStringify> B<O>ption is set, and not via C<p> command.
1265 If one attempts to print this value, then the overloaded operator
1266 C<""> will be called, which will call C<nomethod> operator. The
1267 result of this operator will be stringified again, but this result is
1268 again of type C<symbolic>, which will lead to an infinite loop.
1270 Add a pretty-printer method to the module F<symbolic.pm>:
1273 my ($meth, $a, $b) = @{+shift};
1274 $a = 'u' unless defined $a;
1275 $b = 'u' unless defined $b;
1276 $a = $a->pretty if ref $a;
1277 $b = $b->pretty if ref $b;
1281 Now one can finish the script by
1283 print "side = ", $side->pretty, "\n";
1285 The method C<pretty> is doing object-to-string conversion, so it
1286 is natural to overload the operator C<""> using this method. However,
1287 inside such a method it is not necessary to pretty-print the
1288 I<components> $a and $b of an object. In the above subroutine
1289 C<"[$meth $a $b]"> is a catenation of some strings and components $a
1290 and $b. If these components use overloading, the catenation operator
1291 will look for an overloaded operator C<.>; if not present, it will
1292 look for an overloaded operator C<"">. Thus it is enough to use
1294 use overload nomethod => \&wrap, '""' => \&str;
1296 my ($meth, $a, $b) = @{+shift};
1297 $a = 'u' unless defined $a;
1298 $b = 'u' unless defined $b;
1302 Now one can change the last line of the script to
1304 print "side = $side\n";
1308 side = [/ [- [sqrt [+ 1 [** [n 1 u] 2]] u] 1] [n 1 u]]
1310 and one can inspect the value in debugger using all the possible
1313 Something is still amiss: consider the loop variable $cnt of the
1314 script. It was a number, not an object. We cannot make this value of
1315 type C<symbolic>, since then the loop will not terminate.
1317 Indeed, to terminate the cycle, the $cnt should become false.
1318 However, the operator C<bool> for checking falsity is overloaded (this
1319 time via overloaded C<"">), and returns a long string, thus any object
1320 of type C<symbolic> is true. To overcome this, we need a way to
1321 compare an object to 0. In fact, it is easier to write a numeric
1324 Here is the text of F<symbolic.pm> with such a routine added (and
1325 slightly modified str()):
1327 package symbolic; # Primitive symbolic calculator
1329 nomethod => \&wrap, '""' => \&str, '0+' => \#
1331 sub new { shift; bless ['n', @_] }
1333 my ($obj, $other, $inv, $meth) = @_;
1334 ($obj, $other) = ($other, $obj) if $inv;
1335 bless [$meth, $obj, $other];
1338 my ($meth, $a, $b) = @{+shift};
1339 $a = 'u' unless defined $a;
1346 my %subr = ( n => sub {$_[0]},
1347 sqrt => sub {sqrt $_[0]},
1348 '-' => sub {shift() - shift()},
1349 '+' => sub {shift() + shift()},
1350 '/' => sub {shift() / shift()},
1351 '*' => sub {shift() * shift()},
1352 '**' => sub {shift() ** shift()},
1355 my ($meth, $a, $b) = @{+shift};
1356 my $subr = $subr{$meth}
1357 or die "Do not know how to ($meth) in symbolic";
1358 $a = $a->num if ref $a eq __PACKAGE__;
1359 $b = $b->num if ref $b eq __PACKAGE__;
1363 All the work of numeric conversion is done in %subr and num(). Of
1364 course, %subr is not complete, it contains only operators used in the
1365 example below. Here is the extra-credit question: why do we need an
1366 explicit recursion in num()? (Answer is at the end of this section.)
1368 Use this module like this:
1371 my $iter = symbolic->new(2); # 16-gon
1372 my $side = symbolic->new(1);
1376 $cnt = $cnt - 1; # Mutator '--' not implemented
1377 $side = (sqrt(1 + $side**2) - 1)/$side;
1379 printf "%s=%f\n", $side, $side;
1380 printf "pi=%f\n", $side*(2**($iter+2));
1382 It prints (without so many line breaks)
1384 [/ [- [sqrt [+ 1 [** [/ [- [sqrt [+ 1 [** [n 1] 2]]] 1]
1386 [/ [- [sqrt [+ 1 [** [n 1] 2]]] 1] [n 1]]]=0.198912
1389 The above module is very primitive. It does not implement
1390 mutator methods (C<++>, C<-=> and so on), does not do deep copying
1391 (not required without mutators!), and implements only those arithmetic
1392 operations which are used in the example.
1394 To implement most arithmetic operations is easy; one should just use
1395 the tables of operations, and change the code which fills %subr to
1397 my %subr = ( 'n' => sub {$_[0]} );
1398 foreach my $op (split " ", $overload::ops{with_assign}) {
1399 $subr{$op} = $subr{"$op="} = eval "sub {shift() $op shift()}";
1401 my @bins = qw(binary 3way_comparison num_comparison str_comparison);
1402 foreach my $op (split " ", "@overload::ops{ @bins }") {
1403 $subr{$op} = eval "sub {shift() $op shift()}";
1405 foreach my $op (split " ", "@overload::ops{qw(unary func)}") {
1406 print "defining '$op'\n";
1407 $subr{$op} = eval "sub {$op shift()}";
1410 Since subroutines implementing assignment operators are not required
1411 to modify their operands (see L<Overloadable Operations> above),
1412 we do not need anything special to make C<+=> and friends work,
1413 besides adding these operators to %subr and defining a copy
1414 constructor (needed since Perl has no way to know that the
1415 implementation of C<'+='> does not mutate the argument -
1416 see L<Copy Constructor>).
1418 To implement a copy constructor, add C<< '=' => \&cpy >> to C<use overload>
1419 line, and code (this code assumes that mutators change things one level
1420 deep only, so recursive copying is not needed):
1424 bless [@$self], ref $self;
1427 To make C<++> and C<--> work, we need to implement actual mutators,
1428 either directly, or in C<nomethod>. We continue to do things inside
1429 C<nomethod>, thus add
1431 if ($meth eq '++' or $meth eq '--') {
1432 @$obj = ($meth, (bless [@$obj]), 1); # Avoid circular reference
1436 after the first line of wrap(). This is not a most effective
1437 implementation, one may consider
1439 sub inc { $_[0] = bless ['++', shift, 1]; }
1443 As a final remark, note that one can fill %subr by
1445 my %subr = ( 'n' => sub {$_[0]} );
1446 foreach my $op (split " ", $overload::ops{with_assign}) {
1447 $subr{$op} = $subr{"$op="} = eval "sub {shift() $op shift()}";
1449 my @bins = qw(binary 3way_comparison num_comparison str_comparison);
1450 foreach my $op (split " ", "@overload::ops{ @bins }") {
1451 $subr{$op} = eval "sub {shift() $op shift()}";
1453 foreach my $op (split " ", "@overload::ops{qw(unary func)}") {
1454 $subr{$op} = eval "sub {$op shift()}";
1456 $subr{'++'} = $subr{'+'};
1457 $subr{'--'} = $subr{'-'};
1459 This finishes implementation of a primitive symbolic calculator in
1460 50 lines of Perl code. Since the numeric values of subexpressions
1461 are not cached, the calculator is very slow.
1463 Here is the answer for the exercise: In the case of str(), we need no
1464 explicit recursion since the overloaded C<.>-operator will fall back
1465 to an existing overloaded operator C<"">. Overloaded arithmetic
1466 operators I<do not> fall back to numeric conversion if C<fallback> is
1467 not explicitly requested. Thus without an explicit recursion num()
1468 would convert C<['+', $a, $b]> to C<$a + $b>, which would just rebuild
1469 the argument of num().
1471 If you wonder why defaults for conversion are different for str() and
1472 num(), note how easy it was to write the symbolic calculator. This
1473 simplicity is due to an appropriate choice of defaults. One extra
1474 note: due to the explicit recursion num() is more fragile than sym():
1475 we need to explicitly check for the type of $a and $b. If components
1476 $a and $b happen to be of some related type, this may lead to problems.
1478 =head2 I<Really> Symbolic Calculator
1480 One may wonder why we call the above calculator symbolic. The reason
1481 is that the actual calculation of the value of expression is postponed
1482 until the value is I<used>.
1484 To see it in action, add a method
1489 @$obj->[0,1] = ('=', shift);
1492 to the package C<symbolic>. After this change one can do
1494 my $a = symbolic->new(3);
1495 my $b = symbolic->new(4);
1496 my $c = sqrt($a**2 + $b**2);
1498 and the numeric value of $c becomes 5. However, after calling
1500 $a->STORE(12); $b->STORE(5);
1502 the numeric value of $c becomes 13. There is no doubt now that the module
1503 symbolic provides a I<symbolic> calculator indeed.
1505 To hide the rough edges under the hood, provide a tie()d interface to the
1506 package C<symbolic>. Add methods
1508 sub TIESCALAR { my $pack = shift; $pack->new(@_) }
1510 sub nop { } # Around a bug
1512 (the bug, fixed in Perl 5.14, is described in L<"BUGS">). One can use this
1515 tie $a, 'symbolic', 3;
1516 tie $b, 'symbolic', 4;
1517 $a->nop; $b->nop; # Around a bug
1519 my $c = sqrt($a**2 + $b**2);
1521 Now numeric value of $c is 5. After C<$a = 12; $b = 5> the numeric value
1522 of $c becomes 13. To insulate the user of the module add a method
1524 sub vars { my $p = shift; tie($_, $p), $_->nop foreach @_; }
1529 symbolic->vars($a, $b);
1530 my $c = sqrt($a**2 + $b**2);
1533 printf "c5 %s=%f\n", $c, $c;
1536 printf "c13 %s=%f\n", $c, $c;
1538 shows that the numeric value of $c follows changes to the values of $a
1543 Ilya Zakharevich E<lt>F<ilya@math.mps.ohio-state.edu>E<gt>.
1547 The C<overloading> pragma can be used to enable or disable overloaded
1548 operations within a lexical scope - see L<overloading>.
1552 When Perl is run with the B<-Do> switch or its equivalent, overloading
1553 induces diagnostic messages.
1555 Using the C<m> command of Perl debugger (see L<perldebug>) one can
1556 deduce which operations are overloaded (and which ancestor triggers
1557 this overloading). Say, if C<eq> is overloaded, then the method C<(eq>
1558 is shown by debugger. The method C<()> corresponds to the C<fallback>
1559 key (in fact a presence of this method shows that this package has
1560 overloading enabled, and it is what is used by the C<Overloaded>
1561 function of module C<overload>).
1563 The module might issue the following warnings:
1567 =item Odd number of arguments for overload::constant
1569 (W) The call to overload::constant contained an odd number of arguments.
1570 The arguments should come in pairs.
1572 =item '%s' is not an overloadable type
1574 (W) You tried to overload a constant type the overload package is unaware of.
1576 =item '%s' is not a code reference
1578 (W) The second (fourth, sixth, ...) argument of overload::constant needs
1579 to be a code reference. Either an anonymous subroutine, or a reference
1582 =item overload arg '%s' is invalid
1584 (W) C<use overload> was passed an argument it did not
1585 recognize. Did you mistype an operator?
1589 =head1 BUGS AND PITFALLS
1595 A pitfall when fallback is TRUE and Perl resorts to a built-in
1596 implementation of an operator is that some operators have more
1597 than one semantic, for example C<|>:
1599 use overload '0+' => sub { $_[0]->{n}; },
1601 my $x = bless { n => 4 }, "main";
1602 my $y = bless { n => 8 }, "main";
1603 print $x | $y, "\n";
1605 You might expect this to output "12".
1606 In fact, it prints "<": the ASCII result of treating "|"
1607 as a bitwise string operator - that is, the result of treating
1608 the operands as the strings "4" and "8" rather than numbers.
1609 The fact that numify (C<0+>) is implemented but stringify
1610 (C<"">) isn't makes no difference since the latter is simply
1611 autogenerated from the former.
1613 The only way to change this is to provide your own subroutine
1618 Magic autogeneration increases the potential for inadvertently
1619 creating self-referential structures.
1620 Currently Perl will not free self-referential
1621 structures until cycles are explicitly broken.
1624 use overload '+' => 'add';
1625 sub add { bless [ \$_[0], \$_[1] ] };
1627 is asking for trouble, since
1631 will effectively become
1633 $obj = add($obj, $y, undef);
1635 with the same result as
1637 $obj = [\$obj, \$foo];
1639 Even if no I<explicit> assignment-variants of operators are present in
1640 the script, they may be generated by the optimizer.
1647 my $tmp = 'obj = ' . $obj; $tmp .= "\n";
1651 The symbol table is filled with names looking like line-noise.
1655 This bug was fixed in Perl 5.18, but may still trip you up if you are using
1658 For the purpose of inheritance every overloaded package behaves as if
1659 C<fallback> is present (possibly undefined). This may create
1660 interesting effects if some package is not overloaded, but inherits
1661 from two overloaded packages.
1665 Before Perl 5.14, the relation between overloading and tie()ing was broken.
1666 Overloading was triggered or not based on the I<previous> class of the
1669 This happened because the presence of overloading was checked
1670 too early, before any tie()d access was attempted. If the
1671 class of the value FETCH()ed from the tied variable does not
1672 change, a simple workaround for code that is to run on older Perl
1673 versions is to access the value (via C<() = $foo> or some such)
1674 immediately after tie()ing, so that after this call the I<previous> class
1675 coincides with the current one.
1679 Barewords are not covered by overloaded string constants.