2 X<subroutine> X<function>
4 perlsub - Perl subroutines
8 To declare subroutines:
9 X<subroutine, declaration> X<sub>
11 sub NAME; # A "forward" declaration.
12 sub NAME(PROTO); # ditto, but with prototypes
13 sub NAME : ATTRS; # with attributes
14 sub NAME(PROTO) : ATTRS; # with attributes and prototypes
16 sub NAME BLOCK # A declaration and a definition.
17 sub NAME(PROTO) BLOCK # ditto, but with prototypes
18 sub NAME : ATTRS BLOCK # with attributes
19 sub NAME(PROTO) : ATTRS BLOCK # with prototypes and attributes
21 use feature 'signatures';
22 sub NAME(SIG) BLOCK # with signature
23 sub NAME :ATTRS (SIG) BLOCK # with signature, attributes
24 sub NAME :prototype(PROTO) (SIG) BLOCK # with signature, prototype
26 To define an anonymous subroutine at runtime:
27 X<subroutine, anonymous>
29 $subref = sub BLOCK; # no proto
30 $subref = sub (PROTO) BLOCK; # with proto
31 $subref = sub : ATTRS BLOCK; # with attributes
32 $subref = sub (PROTO) : ATTRS BLOCK; # with proto and attributes
34 use feature 'signatures';
35 $subref = sub (SIG) BLOCK; # with signature
36 $subref = sub : ATTRS(SIG) BLOCK; # with signature, attributes
38 To import subroutines:
41 use MODULE qw(NAME1 NAME2 NAME3);
44 X<subroutine, call> X<call>
46 NAME(LIST); # & is optional with parentheses.
47 NAME LIST; # Parentheses optional if predeclared/imported.
48 &NAME(LIST); # Circumvent prototypes.
49 &NAME; # Makes current @_ visible to called subroutine.
53 Like many languages, Perl provides for user-defined subroutines.
54 These may be located anywhere in the main program, loaded in from
55 other files via the C<do>, C<require>, or C<use> keywords, or
56 generated on the fly using C<eval> or anonymous subroutines.
57 You can even call a function indirectly using a variable containing
58 its name or a CODE reference.
60 The Perl model for function call and return values is simple: all
61 functions are passed as parameters one single flat list of scalars, and
62 all functions likewise return to their caller one single flat list of
63 scalars. Any arrays or hashes in these call and return lists will
64 collapse, losing their identities--but you may always use
65 pass-by-reference instead to avoid this. Both call and return lists may
66 contain as many or as few scalar elements as you'd like. (Often a
67 function without an explicit return statement is called a subroutine, but
68 there's really no difference from Perl's perspective.)
69 X<subroutine, parameter> X<parameter>
71 Any arguments passed in show up in the array C<@_>.
72 (They may also show up in lexical variables introduced by a signature;
73 see L</Signatures> below.) Therefore, if
74 you called a function with two arguments, those would be stored in
75 C<$_[0]> and C<$_[1]>. The array C<@_> is a local array, but its
76 elements are aliases for the actual scalar parameters. In particular,
77 if an element C<$_[0]> is updated, the corresponding argument is
78 updated (or an error occurs if it is not updatable). If an argument
79 is an array or hash element which did not exist when the function
80 was called, that element is created only when (and if) it is modified
81 or a reference to it is taken. (Some earlier versions of Perl
82 created the element whether or not the element was assigned to.)
83 Assigning to the whole array C<@_> removes that aliasing, and does
84 not update any arguments.
85 X<subroutine, argument> X<argument> X<@_>
87 A C<return> statement may be used to exit a subroutine, optionally
88 specifying the returned value, which will be evaluated in the
89 appropriate context (list, scalar, or void) depending on the context of
90 the subroutine call. If you specify no return value, the subroutine
91 returns an empty list in list context, the undefined value in scalar
92 context, or nothing in void context. If you return one or more
93 aggregates (arrays and hashes), these will be flattened together into
94 one large indistinguishable list.
96 If no C<return> is found and if the last statement is an expression, its
97 value is returned. If the last statement is a loop control structure
98 like a C<foreach> or a C<while>, the returned value is unspecified. The
99 empty sub returns the empty list.
100 X<subroutine, return value> X<return value> X<return>
102 Aside from an experimental facility (see L</Signatures> below),
103 Perl does not have named formal parameters. In practice all you
104 do is assign to a C<my()> list of these. Variables that aren't
105 declared to be private are global variables. For gory details
106 on creating private variables, see L</"Private Variables via my()">
107 and L</"Temporary Values via local()">. To create protected
108 environments for a set of functions in a separate package (and
109 probably a separate file), see L<perlmod/"Packages">.
110 X<formal parameter> X<parameter, formal>
117 $max = $foo if $max < $foo;
121 $bestday = max($mon,$tue,$wed,$thu,$fri);
125 # get a line, combining continuation lines
126 # that start with whitespace
129 $thisline = $lookahead; # global variables!
130 LINE: while (defined($lookahead = <STDIN>)) {
131 if ($lookahead =~ /^[ \t]/) {
132 $thisline .= $lookahead;
141 $lookahead = <STDIN>; # get first line
142 while (defined($line = get_line())) {
146 Assigning to a list of private variables to name your arguments:
149 my($key, $value) = @_;
150 $Foo{$key} = $value unless $Foo{$key};
153 Because the assignment copies the values, this also has the effect
154 of turning call-by-reference into call-by-value. Otherwise a
155 function is free to do in-place modifications of C<@_> and change
157 X<call-by-reference> X<call-by-value>
159 upcase_in($v1, $v2); # this changes $v1 and $v2
161 for (@_) { tr/a-z/A-Z/ }
164 You aren't allowed to modify constants in this way, of course. If an
165 argument were actually literal and you tried to change it, you'd take a
166 (presumably fatal) exception. For example, this won't work:
167 X<call-by-reference> X<call-by-value>
169 upcase_in("frederick");
171 It would be much safer if the C<upcase_in()> function
172 were written to return a copy of its parameters instead
173 of changing them in place:
175 ($v3, $v4) = upcase($v1, $v2); # this doesn't change $v1 and $v2
177 return unless defined wantarray; # void context, do nothing
179 for (@parms) { tr/a-z/A-Z/ }
180 return wantarray ? @parms : $parms[0];
183 Notice how this (unprototyped) function doesn't care whether it was
184 passed real scalars or arrays. Perl sees all arguments as one big,
185 long, flat parameter list in C<@_>. This is one area where
186 Perl's simple argument-passing style shines. The C<upcase()>
187 function would work perfectly well without changing the C<upcase()>
188 definition even if we fed it things like this:
190 @newlist = upcase(@list1, @list2);
191 @newlist = upcase( split /:/, $var );
193 Do not, however, be tempted to do this:
195 (@a, @b) = upcase(@list1, @list2);
197 Like the flattened incoming parameter list, the return list is also
198 flattened on return. So all you have managed to do here is stored
199 everything in C<@a> and made C<@b> empty. See
200 L</Pass by Reference> for alternatives.
202 A subroutine may be called using an explicit C<&> prefix. The
203 C<&> is optional in modern Perl, as are parentheses if the
204 subroutine has been predeclared. The C<&> is I<not> optional
205 when just naming the subroutine, such as when it's used as
206 an argument to defined() or undef(). Nor is it optional when you
207 want to do an indirect subroutine call with a subroutine name or
208 reference using the C<&$subref()> or C<&{$subref}()> constructs,
209 although the C<< $subref->() >> notation solves that problem.
210 See L<perlref> for more about all that.
213 Subroutines may be called recursively. If a subroutine is called
214 using the C<&> form, the argument list is optional, and if omitted,
215 no C<@_> array is set up for the subroutine: the C<@_> array at the
216 time of the call is visible to subroutine instead. This is an
217 efficiency mechanism that new users may wish to avoid.
220 &foo(1,2,3); # pass three arguments
221 foo(1,2,3); # the same
223 foo(); # pass a null list
226 &foo; # foo() get current args, like foo(@_) !!
228 foo; # like foo() iff sub foo predeclared, else
229 # a compile-time error
231 foo; # like foo() iff sub foo predeclared, else
232 # a literal string "foo"
234 Not only does the C<&> form make the argument list optional, it also
235 disables any prototype checking on arguments you do provide. This
236 is partly for historical reasons, and partly for having a convenient way
237 to cheat if you know what you're doing. See L</Prototypes> below.
240 Since Perl 5.16.0, the C<__SUB__> token is available under C<use feature
241 'current_sub'> and C<use 5.16.0>. It will evaluate to a reference to the
242 currently-running sub, which allows for recursive calls without knowing
243 your subroutine's name.
246 my $factorial = sub {
249 return($x * __SUB__->( $x - 1 ) );
252 The behavior of C<__SUB__> within a regex code block (such as C</(?{...})/>)
253 is subject to change.
255 Subroutines whose names are in all upper case are reserved to the Perl
256 core, as are modules whose names are in all lower case. A subroutine in
257 all capitals is a loosely-held convention meaning it will be called
258 indirectly by the run-time system itself, usually due to a triggered event.
259 Subroutines whose name start with a left parenthesis are also reserved the
260 same way. The following is a list of some subroutines that currently do
261 special, pre-defined things.
265 =item documented later in this document
269 =item documented in L<perlmod>
271 C<CLONE>, C<CLONE_SKIP>
273 =item documented in L<perlobj>
277 =item documented in L<perltie>
279 C<BINMODE>, C<CLEAR>, C<CLOSE>, C<DELETE>, C<DESTROY>, C<EOF>, C<EXISTS>,
280 C<EXTEND>, C<FETCH>, C<FETCHSIZE>, C<FILENO>, C<FIRSTKEY>, C<GETC>,
281 C<NEXTKEY>, C<OPEN>, C<POP>, C<PRINT>, C<PRINTF>, C<PUSH>, C<READ>,
282 C<READLINE>, C<SCALAR>, C<SEEK>, C<SHIFT>, C<SPLICE>, C<STORE>,
283 C<STORESIZE>, C<TELL>, C<TIEARRAY>, C<TIEHANDLE>, C<TIEHASH>,
284 C<TIESCALAR>, C<UNSHIFT>, C<UNTIE>, C<WRITE>
286 =item documented in L<PerlIO::via>
288 C<BINMODE>, C<CLEARERR>, C<CLOSE>, C<EOF>, C<ERROR>, C<FDOPEN>, C<FILENO>,
289 C<FILL>, C<FLUSH>, C<OPEN>, C<POPPED>, C<PUSHED>, C<READ>, C<SEEK>,
290 C<SETLINEBUF>, C<SYSOPEN>, C<TELL>, C<UNREAD>, C<UTF8>, C<WRITE>
292 =item documented in L<perlfunc>
294 L<< C<import> | perlfunc/use >>, L<< C<unimport> | perlfunc/use >>,
295 L<< C<INC> | perlfunc/require >>
297 =item documented in L<UNIVERSAL>
301 =item documented in L<perldebguts>
303 C<DB::DB>, C<DB::sub>, C<DB::lsub>, C<DB::goto>, C<DB::postponed>
305 =item undocumented, used internally by the L<overload> feature
307 any starting with C<(>
311 The C<BEGIN>, C<UNITCHECK>, C<CHECK>, C<INIT> and C<END> subroutines
312 are not so much subroutines as named special code blocks, of which you
313 can have more than one in a package, and which you can B<not> call
314 explicitly. See L<perlmod/"BEGIN, UNITCHECK, CHECK, INIT and END">
318 B<WARNING>: Subroutine signatures are experimental. The feature may be
319 modified or removed in future versions of Perl.
321 Perl has an experimental facility to allow a subroutine's formal
322 parameters to be introduced by special syntax, separate from the
323 procedural code of the subroutine body. The formal parameter list
324 is known as a I<signature>. The facility must be enabled first by a
325 pragmatic declaration, C<use feature 'signatures'>, and it will produce
326 a warning unless the "experimental::signatures" warnings category is
329 The signature is part of a subroutine's body. Normally the body of a
330 subroutine is simply a braced block of code, but when using a signature,
331 the signature is a parenthesised list that goes immediately before the
332 block, after any name or attributes.
336 sub foo :lvalue ($a, $b = 1, @c) { .... }
338 The signature declares lexical variables that are
339 in scope for the block. When the subroutine is called, the signature
340 takes control first. It populates the signature variables from the
341 list of arguments that were passed. If the argument list doesn't meet
342 the requirements of the signature, then it will throw an exception.
343 When the signature processing is complete, control passes to the block.
345 Positional parameters are handled by simply naming scalar variables in
346 the signature. For example,
348 sub foo ($left, $right) {
349 return $left + $right;
352 takes two positional parameters, which must be filled at runtime by
353 two arguments. By default the parameters are mandatory, and it is
354 not permitted to pass more arguments than expected. So the above is
358 die "Too many arguments for subroutine" unless @_ <= 2;
359 die "Too few arguments for subroutine" unless @_ >= 2;
362 return $left + $right;
365 An argument can be ignored by omitting the main part of the name from
366 a parameter declaration, leaving just a bare C<$> sigil. For example,
368 sub foo ($first, $, $third) {
369 return "first=$first, third=$third";
372 Although the ignored argument doesn't go into a variable, it is still
373 mandatory for the caller to pass it.
375 A positional parameter is made optional by giving a default value,
376 separated from the parameter name by C<=>:
378 sub foo ($left, $right = 0) {
379 return $left + $right;
382 The above subroutine may be called with either one or two arguments.
383 The default value expression is evaluated when the subroutine is called,
384 so it may provide different default values for different calls. It is
385 only evaluated if the argument was actually omitted from the call.
389 sub foo ($thing, $id = $auto_id++) {
390 print "$thing has ID $id";
393 automatically assigns distinct sequential IDs to things for which no
394 ID was supplied by the caller. A default value expression may also
395 refer to parameters earlier in the signature, making the default for
396 one parameter vary according to the earlier parameters. For example,
398 sub foo ($first_name, $surname, $nickname = $first_name) {
399 print "$first_name $surname is known as \"$nickname\"";
402 An optional parameter can be nameless just like a mandatory parameter.
405 sub foo ($thing, $ = 1) {
409 The parameter's default value will still be evaluated if the corresponding
410 argument isn't supplied, even though the value won't be stored anywhere.
411 This is in case evaluating it has important side effects. However, it
412 will be evaluated in void context, so if it doesn't have side effects
413 and is not trivial it will generate a warning if the "void" warning
414 category is enabled. If a nameless optional parameter's default value
415 is not important, it may be omitted just as the parameter's name was:
417 sub foo ($thing, $=) {
421 Optional positional parameters must come after all mandatory positional
422 parameters. (If there are no mandatory positional parameters then an
423 optional positional parameters can be the first thing in the signature.)
424 If there are multiple optional positional parameters and not enough
425 arguments are supplied to fill them all, they will be filled from left
428 After positional parameters, additional arguments may be captured in a
429 slurpy parameter. The simplest form of this is just an array variable:
431 sub foo ($filter, @inputs) {
432 print $filter->($_) foreach @inputs;
435 With a slurpy parameter in the signature, there is no upper limit on how
436 many arguments may be passed. A slurpy array parameter may be nameless
437 just like a positional parameter, in which case its only effect is to
438 turn off the argument limit that would otherwise apply:
440 sub foo ($thing, @) {
444 A slurpy parameter may instead be a hash, in which case the arguments
445 available to it are interpreted as alternating keys and values.
446 There must be as many keys as values: if there is an odd argument then
447 an exception will be thrown. Keys will be stringified, and if there are
448 duplicates then the later instance takes precedence over the earlier,
449 as with standard hash construction.
451 sub foo ($filter, %inputs) {
452 print $filter->($_, $inputs{$_}) foreach sort keys %inputs;
455 A slurpy hash parameter may be nameless just like other kinds of
456 parameter. It still insists that the number of arguments available to
457 it be even, even though they're not being put into a variable.
459 sub foo ($thing, %) {
463 A slurpy parameter, either array or hash, must be the last thing in the
464 signature. It may follow mandatory and optional positional parameters;
465 it may also be the only thing in the signature. Slurpy parameters cannot
466 have default values: if no arguments are supplied for them then you get
467 an empty array or empty hash.
469 A signature may be entirely empty, in which case all it does is check
470 that the caller passed no arguments:
476 When using a signature, the arguments are still available in the special
477 array variable C<@_>, in addition to the lexical variables of the
478 signature. There is a difference between the two ways of accessing the
479 arguments: C<@_> I<aliases> the arguments, but the signature variables
480 get I<copies> of the arguments. So writing to a signature variable
481 only changes that variable, and has no effect on the caller's variables,
482 but writing to an element of C<@_> modifies whatever the caller used to
483 supply that argument.
485 There is a potential syntactic ambiguity between signatures and prototypes
486 (see L</Prototypes>), because both start with an opening parenthesis and
487 both can appear in some of the same places, such as just after the name
488 in a subroutine declaration. For historical reasons, when signatures
489 are not enabled, any opening parenthesis in such a context will trigger
490 very forgiving prototype parsing. Most signatures will be interpreted
491 as prototypes in those circumstances, but won't be valid prototypes.
492 (A valid prototype cannot contain any alphabetic character.) This will
493 lead to somewhat confusing error messages.
495 To avoid ambiguity, when signatures are enabled the special syntax
496 for prototypes is disabled. There is no attempt to guess whether a
497 parenthesised group was intended to be a prototype or a signature.
498 To give a subroutine a prototype under these circumstances, use a
499 L<prototype attribute|attributes/Built-in Attributes>. For example,
501 sub foo :prototype($) { $_[0] }
503 It is entirely possible for a subroutine to have both a prototype and
504 a signature. They do different jobs: the prototype affects compilation
505 of calls to the subroutine, and the signature puts argument values into
506 lexical variables at runtime. You can therefore write
508 sub foo :prototype($$) ($left, $right) {
509 return $left + $right;
512 The prototype attribute, and any other attributes, must come before
513 the signature. The signature always immediately precedes the block of
514 the subroutine's body.
516 =head2 Private Variables via my()
517 X<my> X<variable, lexical> X<lexical> X<lexical variable> X<scope, lexical>
518 X<lexical scope> X<attributes, my>
522 my $foo; # declare $foo lexically local
523 my (@wid, %get); # declare list of variables local
524 my $foo = "flurp"; # declare $foo lexical, and init it
525 my @oof = @bar; # declare @oof lexical, and init it
526 my $x : Foo = $y; # similar, with an attribute applied
528 B<WARNING>: The use of attribute lists on C<my> declarations is still
529 evolving. The current semantics and interface are subject to change.
530 See L<attributes> and L<Attribute::Handlers>.
532 The C<my> operator declares the listed variables to be lexically
533 confined to the enclosing block, conditional
534 (C<if>/C<unless>/C<elsif>/C<else>), loop
535 (C<for>/C<foreach>/C<while>/C<until>/C<continue>), subroutine, C<eval>,
536 or C<do>/C<require>/C<use>'d file. If more than one value is listed, the
537 list must be placed in parentheses. All listed elements must be
538 legal lvalues. Only alphanumeric identifiers may be lexically
539 scoped--magical built-ins like C<$/> must currently be C<local>ized
540 with C<local> instead.
542 Unlike dynamic variables created by the C<local> operator, lexical
543 variables declared with C<my> are totally hidden from the outside
544 world, including any called subroutines. This is true if it's the
545 same subroutine called from itself or elsewhere--every call gets
549 This doesn't mean that a C<my> variable declared in a statically
550 enclosing lexical scope would be invisible. Only dynamic scopes
551 are cut off. For example, the C<bumpx()> function below has access
552 to the lexical $x variable because both the C<my> and the C<sub>
553 occurred at the same scope, presumably file scope.
558 An C<eval()>, however, can see lexical variables of the scope it is
559 being evaluated in, so long as the names aren't hidden by declarations within
560 the C<eval()> itself. See L<perlref>.
563 The parameter list to my() may be assigned to if desired, which allows you
564 to initialize your variables. (If no initializer is given for a
565 particular variable, it is created with the undefined value.) Commonly
566 this is used to name input parameters to a subroutine. Examples:
568 $arg = "fred"; # "global" variable
570 print "$arg thinks the root is $n\n";
571 fred thinks the root is 3
574 my $arg = shift; # name doesn't matter
579 The C<my> is simply a modifier on something you might assign to. So when
580 you do assign to variables in its argument list, C<my> doesn't
581 change whether those variables are viewed as a scalar or an array. So
583 my ($foo) = <STDIN>; # WRONG?
586 both supply a list context to the right-hand side, while
590 supplies a scalar context. But the following declares only one variable:
592 my $foo, $bar = 1; # WRONG
594 That has the same effect as
599 The declared variable is not introduced (is not visible) until after
600 the current statement. Thus,
604 can be used to initialize a new $x with the value of the old $x, and
607 my $x = 123 and $x == 123
609 is false unless the old $x happened to have the value C<123>.
611 Lexical scopes of control structures are not bounded precisely by the
612 braces that delimit their controlled blocks; control expressions are
613 part of that scope, too. Thus in the loop
615 while (my $line = <>) {
621 the scope of $line extends from its declaration throughout the rest of
622 the loop construct (including the C<continue> clause), but not beyond
623 it. Similarly, in the conditional
625 if ((my $answer = <STDIN>) =~ /^yes$/i) {
627 } elsif ($answer =~ /^no$/i) {
631 die "'$answer' is neither 'yes' nor 'no'";
634 the scope of $answer extends from its declaration through the rest
635 of that conditional, including any C<elsif> and C<else> clauses,
636 but not beyond it. See L<perlsyn/"Simple Statements"> for information
637 on the scope of variables in statements with modifiers.
639 The C<foreach> loop defaults to scoping its index variable dynamically
640 in the manner of C<local>. However, if the index variable is
641 prefixed with the keyword C<my>, or if there is already a lexical
642 by that name in scope, then a new lexical is created instead. Thus
646 for my $i (1, 2, 3) {
650 the scope of $i extends to the end of the loop, but not beyond it,
651 rendering the value of $i inaccessible within C<some_function()>.
654 Some users may wish to encourage the use of lexically scoped variables.
655 As an aid to catching implicit uses to package variables,
656 which are always global, if you say
660 then any variable mentioned from there to the end of the enclosing
661 block must either refer to a lexical variable, be predeclared via
662 C<our> or C<use vars>, or else must be fully qualified with the package name.
663 A compilation error results otherwise. An inner block may countermand
664 this with C<no strict 'vars'>.
666 A C<my> has both a compile-time and a run-time effect. At compile
667 time, the compiler takes notice of it. The principal usefulness
668 of this is to quiet C<use strict 'vars'>, but it is also essential
669 for generation of closures as detailed in L<perlref>. Actual
670 initialization is delayed until run time, though, so it gets executed
671 at the appropriate time, such as each time through a loop, for
674 Variables declared with C<my> are not part of any package and are therefore
675 never fully qualified with the package name. In particular, you're not
676 allowed to try to make a package variable (or other global) lexical:
678 my $pack::var; # ERROR! Illegal syntax
680 In fact, a dynamic variable (also known as package or global variables)
681 are still accessible using the fully qualified C<::> notation even while a
682 lexical of the same name is also visible:
687 print "$x and $::x\n";
689 That will print out C<20> and C<10>.
691 You may declare C<my> variables at the outermost scope of a file
692 to hide any such identifiers from the world outside that file. This
693 is similar in spirit to C's static variables when they are used at
694 the file level. To do this with a subroutine requires the use of
695 a closure (an anonymous function that accesses enclosing lexicals).
696 If you want to create a private subroutine that cannot be called
697 from outside that block, it can declare a lexical variable containing
698 an anonymous sub reference:
700 my $secret_version = '1.001-beta';
701 my $secret_sub = sub { print $secret_version };
704 As long as the reference is never returned by any function within the
705 module, no outside module can see the subroutine, because its name is not in
706 any package's symbol table. Remember that it's not I<REALLY> called
707 C<$some_pack::secret_version> or anything; it's just $secret_version,
708 unqualified and unqualifiable.
710 This does not work with object methods, however; all object methods
711 have to be in the symbol table of some package to be found. See
712 L<perlref/"Function Templates"> for something of a work-around to
715 =head2 Persistent Private Variables
716 X<state> X<state variable> X<static> X<variable, persistent> X<variable, static> X<closure>
718 There are two ways to build persistent private variables in Perl 5.10.
719 First, you can simply use the C<state> feature. Or, you can use closures,
720 if you want to stay compatible with releases older than 5.10.
722 =head3 Persistent variables via state()
724 Beginning with Perl 5.10.0, you can declare variables with the C<state>
725 keyword in place of C<my>. For that to work, though, you must have
726 enabled that feature beforehand, either by using the C<feature> pragma, or
727 by using C<-E> on one-liners (see L<feature>). Beginning with Perl 5.16,
728 the C<CORE::state> form does not require the
731 The C<state> keyword creates a lexical variable (following the same scoping
732 rules as C<my>) that persists from one subroutine call to the next. If a
733 state variable resides inside an anonymous subroutine, then each copy of
734 the subroutine has its own copy of the state variable. However, the value
735 of the state variable will still persist between calls to the same copy of
736 the anonymous subroutine. (Don't forget that C<sub { ... }> creates a new
737 subroutine each time it is executed.)
739 For example, the following code maintains a private counter, incremented
740 each time the gimme_another() function is called:
743 sub gimme_another { state $x; return ++$x }
745 And this example uses anonymous subroutines to create separate counters:
749 return sub { state $x; return ++$x }
752 Also, since C<$x> is lexical, it can't be reached or modified by any Perl
755 When combined with variable declaration, simple assignment to C<state>
756 variables (as in C<state $x = 42>) is executed only the first time. When such
757 statements are evaluated subsequent times, the assignment is ignored. The
758 behavior of assignment to C<state> declarations where the left hand side
759 of the assignment involves any parentheses is currently undefined.
761 =head3 Persistent variables with closures
763 Just because a lexical variable is lexically (also called statically)
764 scoped to its enclosing block, C<eval>, or C<do> FILE, this doesn't mean that
765 within a function it works like a C static. It normally works more
766 like a C auto, but with implicit garbage collection.
768 Unlike local variables in C or C++, Perl's lexical variables don't
769 necessarily get recycled just because their scope has exited.
770 If something more permanent is still aware of the lexical, it will
771 stick around. So long as something else references a lexical, that
772 lexical won't be freed--which is as it should be. You wouldn't want
773 memory being free until you were done using it, or kept around once you
774 were done. Automatic garbage collection takes care of this for you.
776 This means that you can pass back or save away references to lexical
777 variables, whereas to return a pointer to a C auto is a grave error.
778 It also gives us a way to simulate C's function statics. Here's a
779 mechanism for giving a function private variables with both lexical
780 scoping and a static lifetime. If you do want to create something like
781 C's static variables, just enclose the whole function in an extra block,
782 and put the static variable outside the function but in the block.
787 return ++$secret_val;
790 # $secret_val now becomes unreachable by the outside
791 # world, but retains its value between calls to gimme_another
793 If this function is being sourced in from a separate file
794 via C<require> or C<use>, then this is probably just fine. If it's
795 all in the main program, you'll need to arrange for the C<my>
796 to be executed early, either by putting the whole block above
797 your main program, or more likely, placing merely a C<BEGIN>
798 code block around it to make sure it gets executed before your program
804 return ++$secret_val;
808 See L<perlmod/"BEGIN, UNITCHECK, CHECK, INIT and END"> about the
809 special triggered code blocks, C<BEGIN>, C<UNITCHECK>, C<CHECK>,
812 If declared at the outermost scope (the file scope), then lexicals
813 work somewhat like C's file statics. They are available to all
814 functions in that same file declared below them, but are inaccessible
815 from outside that file. This strategy is sometimes used in modules
816 to create private variables that the whole module can see.
818 =head2 Temporary Values via local()
819 X<local> X<scope, dynamic> X<dynamic scope> X<variable, local>
820 X<variable, temporary>
822 B<WARNING>: In general, you should be using C<my> instead of C<local>, because
823 it's faster and safer. Exceptions to this include the global punctuation
824 variables, global filehandles and formats, and direct manipulation of the
825 Perl symbol table itself. C<local> is mostly used when the current value
826 of a variable must be visible to called subroutines.
830 # localization of values
832 local $foo; # make $foo dynamically local
833 local (@wid, %get); # make list of variables local
834 local $foo = "flurp"; # make $foo dynamic, and init it
835 local @oof = @bar; # make @oof dynamic, and init it
837 local $hash{key} = "val"; # sets a local value for this hash entry
838 delete local $hash{key}; # delete this entry for the current block
839 local ($cond ? $v1 : $v2); # several types of lvalues support
842 # localization of symbols
844 local *FH; # localize $FH, @FH, %FH, &FH ...
845 local *merlyn = *randal; # now $merlyn is really $randal, plus
846 # @merlyn is really @randal, etc
847 local *merlyn = 'randal'; # SAME THING: promote 'randal' to *randal
848 local *merlyn = \$randal; # just alias $merlyn, not @merlyn etc
850 A C<local> modifies its listed variables to be "local" to the
851 enclosing block, C<eval>, or C<do FILE>--and to I<any subroutine
852 called from within that block>. A C<local> just gives temporary
853 values to global (meaning package) variables. It does I<not> create
854 a local variable. This is known as dynamic scoping. Lexical scoping
855 is done with C<my>, which works more like C's auto declarations.
857 Some types of lvalues can be localized as well: hash and array elements
858 and slices, conditionals (provided that their result is always
859 localizable), and symbolic references. As for simple variables, this
860 creates new, dynamically scoped values.
862 If more than one variable or expression is given to C<local>, they must be
863 placed in parentheses. This operator works
864 by saving the current values of those variables in its argument list on a
865 hidden stack and restoring them upon exiting the block, subroutine, or
866 eval. This means that called subroutines can also reference the local
867 variable, but not the global one. The argument list may be assigned to if
868 desired, which allows you to initialize your local variables. (If no
869 initializer is given for a particular variable, it is created with an
872 Because C<local> is a run-time operator, it gets executed each time
873 through a loop. Consequently, it's more efficient to localize your
874 variables outside the loop.
876 =head3 Grammatical note on local()
879 A C<local> is simply a modifier on an lvalue expression. When you assign to
880 a C<local>ized variable, the C<local> doesn't change whether its list is viewed
881 as a scalar or an array. So
883 local($foo) = <STDIN>;
884 local @FOO = <STDIN>;
886 both supply a list context to the right-hand side, while
888 local $foo = <STDIN>;
890 supplies a scalar context.
892 =head3 Localization of special variables
893 X<local, special variable>
895 If you localize a special variable, you'll be giving a new value to it,
896 but its magic won't go away. That means that all side-effects related
897 to this magic still work with the localized value.
899 This feature allows code like this to work :
901 # Read the whole contents of FILE in $slurp
902 { local $/ = undef; $slurp = <FILE>; }
904 Note, however, that this restricts localization of some values ; for
905 example, the following statement dies, as of perl 5.10.0, with an error
906 I<Modification of a read-only value attempted>, because the $1 variable is
907 magical and read-only :
911 One exception is the default scalar variable: starting with perl 5.14
912 C<local($_)> will always strip all magic from $_, to make it possible
913 to safely reuse $_ in a subroutine.
915 B<WARNING>: Localization of tied arrays and hashes does not currently
917 This will be fixed in a future release of Perl; in the meantime, avoid
918 code that relies on any particular behavior of localising tied arrays
919 or hashes (localising individual elements is still okay).
920 See L<perl58delta/"Localising Tied Arrays and Hashes Is Broken"> for more
924 =head3 Localization of globs
925 X<local, glob> X<glob>
931 creates a whole new symbol table entry for the glob C<name> in the
932 current package. That means that all variables in its glob slot ($name,
933 @name, %name, &name, and the C<name> filehandle) are dynamically reset.
935 This implies, among other things, that any magic eventually carried by
936 those variables is locally lost. In other words, saying C<local */>
937 will not have any effect on the internal value of the input record
940 =head3 Localization of elements of composite types
941 X<local, composite type element> X<local, array element> X<local, hash element>
943 It's also worth taking a moment to explain what happens when you
944 C<local>ize a member of a composite type (i.e. an array or hash element).
945 In this case, the element is C<local>ized I<by name>. This means that
946 when the scope of the C<local()> ends, the saved value will be
947 restored to the hash element whose key was named in the C<local()>, or
948 the array element whose index was named in the C<local()>. If that
949 element was deleted while the C<local()> was in effect (e.g. by a
950 C<delete()> from a hash or a C<shift()> of an array), it will spring
951 back into existence, possibly extending an array and filling in the
952 skipped elements with C<undef>. For instance, if you say
954 %hash = ( 'This' => 'is', 'a' => 'test' );
958 local($hash{'a'}) = 'drill';
959 while (my $e = pop(@ary)) {
964 $hash{'only a'} = 'test';
968 print join(' ', map { "$_ $hash{$_}" } sort keys %hash),".\n";
969 print "The array has ",scalar(@ary)," elements: ",
970 join(', ', map { defined $_ ? $_ : 'undef' } @ary),"\n";
977 This is a test only a test.
978 The array has 6 elements: 0, 1, 2, undef, undef, 5
980 The behavior of local() on non-existent members of composite
981 types is subject to change in future. The behavior of local()
982 on array elements specified using negative indexes is particularly
983 surprising, and is very likely to change.
985 =head3 Localized deletion of elements of composite types
986 X<delete> X<local, composite type element> X<local, array element> X<local, hash element>
988 You can use the C<delete local $array[$idx]> and C<delete local $hash{key}>
989 constructs to delete a composite type entry for the current block and restore
990 it when it ends. They return the array/hash value before the localization,
991 which means that they are respectively equivalent to
994 my $val = $array[$idx];
1003 my $val = $hash{key};
1009 except that for those the C<local> is
1010 scoped to the C<do> block. Slices are
1019 my $a = delete local $hash{a};
1024 my @nums = delete local @$a[0, 2]
1026 # $a is [ undef, 8 ]
1028 $a[0] = 999; # will be erased when the scope ends
1030 # $a is back to [ 7, 8, 9 ]
1033 # %hash is back to its original state
1035 =head2 Lvalue subroutines
1036 X<lvalue> X<subroutine, lvalue>
1038 It is possible to return a modifiable value from a subroutine.
1039 To do this, you have to declare the subroutine to return an lvalue.
1042 sub canmod : lvalue {
1043 $val; # or: return $val;
1049 canmod() = 5; # assigns to $val
1050 nomod() = 5; # ERROR
1052 The scalar/list context for the subroutine and for the right-hand
1053 side of assignment is determined as if the subroutine call is replaced
1054 by a scalar. For example, consider:
1056 data(2,3) = get_data(3,4);
1058 Both subroutines here are called in a scalar context, while in:
1060 (data(2,3)) = get_data(3,4);
1064 (data(2),data(3)) = get_data(3,4);
1066 all the subroutines are called in a list context.
1068 Lvalue subroutines are convenient, but you have to keep in mind that,
1069 when used with objects, they may violate encapsulation. A normal
1070 mutator can check the supplied argument before setting the attribute
1071 it is protecting, an lvalue subroutine cannot. If you require any
1072 special processing when storing and retrieving the values, consider
1073 using the CPAN module Sentinel or something similar.
1075 =head2 Lexical Subroutines
1076 X<my sub> X<state sub> X<our sub> X<subroutine, lexical>
1078 Beginning with Perl 5.18, you can declare a private subroutine with C<my>
1079 or C<state>. As with state variables, the C<state> keyword is only
1080 available under C<use feature 'state'> or C<use 5.010> or higher.
1082 Prior to Perl 5.26, lexical subroutines were deemed experimental and were
1083 available only under the C<use feature 'lexical_subs'> pragma. They also
1084 produced a warning unless the "experimental::lexical_subs" warnings
1085 category was disabled.
1087 These subroutines are only visible within the block in which they are
1088 declared, and only after that declaration:
1090 # Include these two lines if your code is intended to run under Perl
1091 # versions earlier than 5.26.
1092 no warnings "experimental::lexical_subs";
1093 use feature 'lexical_subs';
1095 foo(); # calls the package/global subroutine
1097 foo(); # also calls the package subroutine
1099 foo(); # calls "state" sub
1100 my $ref = \&foo; # take a reference to "state" sub
1103 bar(); # calls "my" sub
1105 You can't (directly) write a recursive lexical subroutine:
1112 This example fails because C<baz()> refers to the package/global subroutine
1113 C<baz>, not the lexical subroutine currently being defined.
1115 The solution is to use L<C<__SUB__>|perlfunc/__SUB__>:
1118 __SUB__->(); # calls itself
1121 It is possible to predeclare a lexical subroutine. The C<sub foo {...}>
1122 subroutine definition syntax respects any previous C<my sub;> or C<state sub;>
1123 declaration. Using this to define recursive subroutines is a bad idea,
1126 my sub baz; # predeclaration
1127 sub baz { # define the "my" sub
1128 baz(); # WRONG: calls itself, but leaks memory
1131 Just like C<< my $f; $f = sub { $f->() } >>, this example leaks memory. The
1132 name C<baz> is a reference to the subroutine, and the subroutine uses the name
1133 C<baz>; they keep each other alive (see L<perlref/Circular References>).
1135 =head3 C<state sub> vs C<my sub>
1137 What is the difference between "state" subs and "my" subs? Each time that
1138 execution enters a block when "my" subs are declared, a new copy of each
1139 sub is created. "State" subroutines persist from one execution of the
1140 containing block to the next.
1142 So, in general, "state" subroutines are faster. But "my" subs are
1143 necessary if you want to create closures:
1148 ... do something with $x ...
1153 In this example, a new C<$x> is created when C<whatever> is called, and
1154 also a new C<inner>, which can see the new C<$x>. A "state" sub will only
1155 see the C<$x> from the first call to C<whatever>.
1157 =head3 C<our> subroutines
1159 Like C<our $variable>, C<our sub> creates a lexical alias to the package
1160 subroutine of the same name.
1162 The two main uses for this are to switch back to using the package sub
1163 inside an inner scope:
1170 # need to use the outer foo here
1176 and to make a subroutine visible to other packages in the same scope:
1178 package MySneakyModule;
1180 our sub do_something { ... }
1182 sub do_something_with_caller {
1184 () = caller 1; # sets @DB::args
1185 do_something(@args); # uses MySneakyModule::do_something
1188 =head2 Passing Symbol Table Entries (typeglobs)
1191 B<WARNING>: The mechanism described in this section was originally
1192 the only way to simulate pass-by-reference in older versions of
1193 Perl. While it still works fine in modern versions, the new reference
1194 mechanism is generally easier to work with. See below.
1196 Sometimes you don't want to pass the value of an array to a subroutine
1197 but rather the name of it, so that the subroutine can modify the global
1198 copy of it rather than working with a local copy. In perl you can
1199 refer to all objects of a particular name by prefixing the name
1200 with a star: C<*foo>. This is often known as a "typeglob", because the
1201 star on the front can be thought of as a wildcard match for all the
1202 funny prefix characters on variables and subroutines and such.
1204 When evaluated, the typeglob produces a scalar value that represents
1205 all the objects of that name, including any filehandle, format, or
1206 subroutine. When assigned to, it causes the name mentioned to refer to
1207 whatever C<*> value was assigned to it. Example:
1210 local(*someary) = @_;
1211 foreach $elem (@someary) {
1218 Scalars are already passed by reference, so you can modify
1219 scalar arguments without using this mechanism by referring explicitly
1220 to C<$_[0]> etc. You can modify all the elements of an array by passing
1221 all the elements as scalars, but you have to use the C<*> mechanism (or
1222 the equivalent reference mechanism) to C<push>, C<pop>, or change the size of
1223 an array. It will certainly be faster to pass the typeglob (or reference).
1225 Even if you don't want to modify an array, this mechanism is useful for
1226 passing multiple arrays in a single LIST, because normally the LIST
1227 mechanism will merge all the array values so that you can't extract out
1228 the individual arrays. For more on typeglobs, see
1229 L<perldata/"Typeglobs and Filehandles">.
1231 =head2 When to Still Use local()
1232 X<local> X<variable, local>
1234 Despite the existence of C<my>, there are still three places where the
1235 C<local> operator still shines. In fact, in these three places, you
1236 I<must> use C<local> instead of C<my>.
1242 You need to give a global variable a temporary value, especially $_.
1244 The global variables, like C<@ARGV> or the punctuation variables, must be
1245 C<local>ized with C<local()>. This block reads in F</etc/motd>, and splits
1246 it up into chunks separated by lines of equal signs, which are placed
1250 local @ARGV = ("/etc/motd");
1253 @Fields = split /^\s*=+\s*$/;
1256 It particular, it's important to C<local>ize $_ in any routine that assigns
1257 to it. Look out for implicit assignments in C<while> conditionals.
1261 You need to create a local file or directory handle or a local function.
1263 A function that needs a filehandle of its own must use
1264 C<local()> on a complete typeglob. This can be used to create new symbol
1268 local (*READER, *WRITER); # not my!
1269 pipe (READER, WRITER) or die "pipe: $!";
1270 return (*READER, *WRITER);
1272 ($head, $tail) = ioqueue();
1274 See the Symbol module for a way to create anonymous symbol table
1277 Because assignment of a reference to a typeglob creates an alias, this
1278 can be used to create what is effectively a local function, or at least,
1282 local *grow = \&shrink; # only until this block exits
1283 grow(); # really calls shrink()
1284 move(); # if move() grow()s, it shrink()s too
1286 grow(); # get the real grow() again
1288 See L<perlref/"Function Templates"> for more about manipulating
1289 functions by name in this way.
1293 You want to temporarily change just one element of an array or hash.
1295 You can C<local>ize just one element of an aggregate. Usually this
1296 is done on dynamics:
1299 local $SIG{INT} = 'IGNORE';
1300 funct(); # uninterruptible
1302 # interruptibility automatically restored here
1304 But it also works on lexically declared aggregates.
1308 =head2 Pass by Reference
1309 X<pass by reference> X<pass-by-reference> X<reference>
1311 If you want to pass more than one array or hash into a function--or
1312 return them from it--and have them maintain their integrity, then
1313 you're going to have to use an explicit pass-by-reference. Before you
1314 do that, you need to understand references as detailed in L<perlref>.
1315 This section may not make much sense to you otherwise.
1317 Here are a few simple examples. First, let's pass in several arrays
1318 to a function and have it C<pop> all of then, returning a new list
1319 of all their former last elements:
1321 @tailings = popmany ( \@a, \@b, \@c, \@d );
1326 foreach $aref ( @_ ) {
1327 push @retlist, pop @$aref;
1332 Here's how you might write a function that returns a
1333 list of keys occurring in all the hashes passed to it:
1335 @common = inter( \%foo, \%bar, \%joe );
1337 my ($k, $href, %seen); # locals
1338 foreach $href (@_) {
1339 while ( $k = each %$href ) {
1343 return grep { $seen{$_} == @_ } keys %seen;
1346 So far, we're using just the normal list return mechanism.
1347 What happens if you want to pass or return a hash? Well,
1348 if you're using only one of them, or you don't mind them
1349 concatenating, then the normal calling convention is ok, although
1352 Where people get into trouble is here:
1354 (@a, @b) = func(@c, @d);
1356 (%a, %b) = func(%c, %d);
1358 That syntax simply won't work. It sets just C<@a> or C<%a> and
1359 clears the C<@b> or C<%b>. Plus the function didn't get passed
1360 into two separate arrays or hashes: it got one long list in C<@_>,
1363 If you can arrange for everyone to deal with this through references, it's
1364 cleaner code, although not so nice to look at. Here's a function that
1365 takes two array references as arguments, returning the two array elements
1366 in order of how many elements they have in them:
1368 ($aref, $bref) = func(\@c, \@d);
1369 print "@$aref has more than @$bref\n";
1371 my ($cref, $dref) = @_;
1372 if (@$cref > @$dref) {
1373 return ($cref, $dref);
1375 return ($dref, $cref);
1379 It turns out that you can actually do this also:
1381 (*a, *b) = func(\@c, \@d);
1382 print "@a has more than @b\n";
1384 local (*c, *d) = @_;
1392 Here we're using the typeglobs to do symbol table aliasing. It's
1393 a tad subtle, though, and also won't work if you're using C<my>
1394 variables, because only globals (even in disguise as C<local>s)
1395 are in the symbol table.
1397 If you're passing around filehandles, you could usually just use the bare
1398 typeglob, like C<*STDOUT>, but typeglobs references work, too.
1404 print $fh "her um well a hmmm\n";
1407 $rec = get_rec(\*STDIN);
1410 return scalar <$fh>;
1413 If you're planning on generating new filehandles, you could do this.
1414 Notice to pass back just the bare *FH, not its reference.
1419 return open (FH, $path) ? *FH : undef;
1423 X<prototype> X<subroutine, prototype>
1425 Perl supports a very limited kind of compile-time argument checking
1426 using function prototyping. This can be declared in either the PROTO
1427 section or with a L<prototype attribute|attributes/Built-in Attributes>.
1428 If you declare either of
1431 sub mypush :prototype(\@@)
1433 then C<mypush()> takes arguments exactly like C<push()> does.
1435 If subroutine signatures are enabled (see L</Signatures>), then
1436 the shorter PROTO syntax is unavailable, because it would clash with
1437 signatures. In that case, a prototype can only be declared in the form
1441 function declaration must be visible at compile time. The prototype
1442 affects only interpretation of new-style calls to the function,
1443 where new-style is defined as not using the C<&> character. In
1444 other words, if you call it like a built-in function, then it behaves
1445 like a built-in function. If you call it like an old-fashioned
1446 subroutine, then it behaves like an old-fashioned subroutine. It
1447 naturally falls out from this rule that prototypes have no influence
1448 on subroutine references like C<\&foo> or on indirect subroutine
1449 calls like C<&{$subref}> or C<< $subref->() >>.
1451 Method calls are not influenced by prototypes either, because the
1452 function to be called is indeterminate at compile time, since
1453 the exact code called depends on inheritance.
1455 Because the intent of this feature is primarily to let you define
1456 subroutines that work like built-in functions, here are prototypes
1457 for some other functions that parse almost exactly like the
1458 corresponding built-in.
1460 Declared as Called as
1462 sub mylink ($$) mylink $old, $new
1463 sub myvec ($$$) myvec $var, $offset, 1
1464 sub myindex ($$;$) myindex &getstring, "substr"
1465 sub mysyswrite ($$$;$) mysyswrite $buf, 0, length($buf) - $off, $off
1466 sub myreverse (@) myreverse $a, $b, $c
1467 sub myjoin ($@) myjoin ":", $a, $b, $c
1468 sub mypop (\@) mypop @array
1469 sub mysplice (\@$$@) mysplice @array, 0, 2, @pushme
1470 sub mykeys (\[%@]) mykeys %{$hashref}
1471 sub myopen (*;$) myopen HANDLE, $name
1472 sub mypipe (**) mypipe READHANDLE, WRITEHANDLE
1473 sub mygrep (&@) mygrep { /foo/ } $a, $b, $c
1474 sub myrand (;$) myrand 42
1475 sub mytime () mytime
1477 Any backslashed prototype character represents an actual argument
1478 that must start with that character (optionally preceded by C<my>,
1479 C<our> or C<local>), with the exception of C<$>, which will
1480 accept any scalar lvalue expression, such as C<$foo = 7> or
1481 C<< my_function()->[0] >>. The value passed as part of C<@_> will be a
1482 reference to the actual argument given in the subroutine call,
1483 obtained by applying C<\> to that argument.
1485 You can use the C<\[]> backslash group notation to specify more than one
1486 allowed argument type. For example:
1488 sub myref (\[$@%&*])
1490 will allow calling myref() as
1498 and the first argument of myref() will be a reference to
1499 a scalar, an array, a hash, a code, or a glob.
1501 Unbackslashed prototype characters have special meanings. Any
1502 unbackslashed C<@> or C<%> eats all remaining arguments, and forces
1503 list context. An argument represented by C<$> forces scalar context. An
1504 C<&> requires an anonymous subroutine, which, if passed as the first
1505 argument, does not require the C<sub> keyword or a subsequent comma.
1507 A C<*> allows the subroutine to accept a bareword, constant, scalar expression,
1508 typeglob, or a reference to a typeglob in that slot. The value will be
1509 available to the subroutine either as a simple scalar, or (in the latter
1510 two cases) as a reference to the typeglob. If you wish to always convert
1511 such arguments to a typeglob reference, use Symbol::qualify_to_ref() as
1514 use Symbol 'qualify_to_ref';
1517 my $fh = qualify_to_ref(shift, caller);
1521 The C<+> prototype is a special alternative to C<$> that will act like
1522 C<\[@%]> when given a literal array or hash variable, but will otherwise
1523 force scalar context on the argument. This is useful for functions which
1524 should accept either a literal array or an array reference as the argument:
1528 die "Not an array or arrayref" unless ref $aref eq 'ARRAY';
1532 When using the C<+> prototype, your function must check that the argument
1533 is of an acceptable type.
1535 A semicolon (C<;>) separates mandatory arguments from optional arguments.
1536 It is redundant before C<@> or C<%>, which gobble up everything else.
1538 As the last character of a prototype, or just before a semicolon, a C<@>
1539 or a C<%>, you can use C<_> in place of C<$>: if this argument is not
1540 provided, C<$_> will be used instead.
1542 Note how the last three examples in the table above are treated
1543 specially by the parser. C<mygrep()> is parsed as a true list
1544 operator, C<myrand()> is parsed as a true unary operator with unary
1545 precedence the same as C<rand()>, and C<mytime()> is truly without
1546 arguments, just like C<time()>. That is, if you say
1550 you'll get C<mytime() + 2>, not C<mytime(2)>, which is how it would be parsed
1551 without a prototype. If you want to force a unary function to have the
1552 same precedence as a list operator, add C<;> to the end of the prototype:
1554 sub mygetprotobynumber($;);
1555 mygetprotobynumber $a > $b; # parsed as mygetprotobynumber($a > $b)
1557 The interesting thing about C<&> is that you can generate new syntax with it,
1558 provided it's in the initial position:
1562 my($try,$catch) = @_;
1569 sub catch (&) { $_[0] }
1574 /phooey/ and print "unphooey\n";
1577 That prints C<"unphooey">. (Yes, there are still unresolved
1578 issues having to do with visibility of C<@_>. I'm ignoring that
1579 question for the moment. (But note that if we make C<@_> lexically
1580 scoped, those anonymous subroutines can act like closures... (Gee,
1581 is this sounding a little Lispish? (Never mind.))))
1583 And here's a reimplementation of the Perl C<grep> operator:
1590 push(@result, $_) if &$code;
1595 Some folks would prefer full alphanumeric prototypes. Alphanumerics have
1596 been intentionally left out of prototypes for the express purpose of
1597 someday in the future adding named, formal parameters. The current
1598 mechanism's main goal is to let module writers provide better diagnostics
1599 for module users. Larry feels the notation quite understandable to Perl
1600 programmers, and that it will not intrude greatly upon the meat of the
1601 module, nor make it harder to read. The line noise is visually
1602 encapsulated into a small pill that's easy to swallow.
1604 If you try to use an alphanumeric sequence in a prototype you will
1605 generate an optional warning - "Illegal character in prototype...".
1606 Unfortunately earlier versions of Perl allowed the prototype to be
1607 used as long as its prefix was a valid prototype. The warning may be
1608 upgraded to a fatal error in a future version of Perl once the
1609 majority of offending code is fixed.
1611 It's probably best to prototype new functions, not retrofit prototyping
1612 into older ones. That's because you must be especially careful about
1613 silent impositions of differing list versus scalar contexts. For example,
1614 if you decide that a function should take just one parameter, like this:
1618 print "you gave me $n\n";
1621 and someone has been calling it with an array or expression
1625 func( $text =~ /\w+/g );
1627 Then you've just supplied an automatic C<scalar> in front of their
1628 argument, which can be more than a bit surprising. The old C<@foo>
1629 which used to hold one thing doesn't get passed in. Instead,
1630 C<func()> now gets passed in a C<1>; that is, the number of elements
1631 in C<@foo>. And the C<m//g> gets called in scalar context so instead of a
1632 list of words it returns a boolean result and advances C<pos($text)>. Ouch!
1634 If a sub has both a PROTO and a BLOCK, the prototype is not applied
1635 until after the BLOCK is completely defined. This means that a recursive
1636 function with a prototype has to be predeclared for the prototype to take
1644 This is all very powerful, of course, and should be used only in moderation
1645 to make the world a better place.
1647 =head2 Constant Functions
1650 Functions with a prototype of C<()> are potential candidates for
1651 inlining. If the result after optimization and constant folding
1652 is either a constant or a lexically-scoped scalar which has no other
1653 references, then it will be used in place of function calls made
1654 without C<&>. Calls made using C<&> are never inlined. (See
1655 F<constant.pm> for an easy way to declare most constants.)
1657 The following functions would all be inlined:
1659 sub pi () { 3.14159 } # Not exact, but close.
1660 sub PI () { 4 * atan2 1, 1 } # As good as it gets,
1661 # and it's inlined, too!
1665 sub FLAG_FOO () { 1 << 8 }
1666 sub FLAG_BAR () { 1 << 9 }
1667 sub FLAG_MASK () { FLAG_FOO | FLAG_BAR }
1669 sub OPT_BAZ () { not (0x1B58 & FLAG_MASK) }
1671 sub N () { int(OPT_BAZ) / 3 }
1673 sub FOO_SET () { 1 if FLAG_MASK & FLAG_FOO }
1674 sub FOO_SET2 () { if (FLAG_MASK & FLAG_FOO) { 1 } }
1676 (Be aware that the last example was not always inlined in Perl 5.20 and
1677 earlier, which did not behave consistently with subroutines containing
1678 inner scopes.) You can countermand inlining by using an explicit
1689 sub bonk_val () { return 12345 }
1691 As alluded to earlier you can also declare inlined subs dynamically at
1692 BEGIN time if their body consists of a lexically-scoped scalar which
1693 has no other references. Only the first example here will be inlined:
1698 *INLINED = sub () { $var };
1705 *NOT_INLINED = sub () { $var };
1708 A not so obvious caveat with this (see [RT #79908]) is that the
1709 variable will be immediately inlined, and will stop behaving like a
1710 normal lexical variable, e.g. this will print C<79907>, not C<79908>:
1714 *RT_79908 = sub () { $x };
1717 print RT_79908(); # prints 79907
1719 As of Perl 5.22, this buggy behavior, while preserved for backward
1720 compatibility, is detected and emits a deprecation warning. If you want
1721 the subroutine to be inlined (with no warning), make sure the variable is
1722 not used in a context where it could be modified aside from where it is
1728 *INLINED = sub () { $x };
1730 # Warns. Future Perl versions will stop inlining it.
1734 *ALSO_INLINED = sub () { $x };
1737 Perl 5.22 also introduces the experimental "const" attribute as an
1738 alternative. (Disable the "experimental::const_attr" warnings if you want
1739 to use it.) When applied to an anonymous subroutine, it forces the sub to
1740 be called when the C<sub> expression is evaluated. The return value is
1741 captured and turned into a constant subroutine:
1744 *INLINED = sub : const { $x };
1747 The return value of C<INLINED> in this example will always be 54321,
1748 regardless of later modifications to $x. You can also put any arbitrary
1749 code inside the sub, at it will be executed immediately and its return
1750 value captured the same way.
1752 If you really want a subroutine with a C<()> prototype that returns a
1753 lexical variable you can easily force it to not be inlined by adding
1754 an explicit C<return>:
1758 *RT_79908 = sub () { return $x };
1761 print RT_79908(); # prints 79908
1763 The easiest way to tell if a subroutine was inlined is by using
1764 L<B::Deparse>. Consider this example of two subroutines returning
1765 C<1>, one with a C<()> prototype causing it to be inlined, and one
1766 without (with deparse output truncated for clarity):
1768 $ perl -MO=Deparse -le 'sub ONE { 1 } if (ONE) { print ONE if ONE }'
1773 print ONE() if ONE ;
1775 $ perl -MO=Deparse -le 'sub ONE () { 1 } if (ONE) { print ONE if ONE }'
1781 If you redefine a subroutine that was eligible for inlining, you'll
1782 get a warning by default. You can use this warning to tell whether or
1783 not a particular subroutine is considered inlinable, since it's
1784 different than the warning for overriding non-inlined subroutines:
1786 $ perl -e 'sub one () {1} sub one () {2}'
1787 Constant subroutine one redefined at -e line 1.
1788 $ perl -we 'sub one {1} sub one {2}'
1789 Subroutine one redefined at -e line 1.
1791 The warning is considered severe enough not to be affected by the
1792 B<-w> switch (or its absence) because previously compiled invocations
1793 of the function will still be using the old value of the function. If
1794 you need to be able to redefine the subroutine, you need to ensure
1795 that it isn't inlined, either by dropping the C<()> prototype (which
1796 changes calling semantics, so beware) or by thwarting the inlining
1797 mechanism in some other way, e.g. by adding an explicit C<return>, as
1800 sub not_inlined () { return 23 }
1802 =head2 Overriding Built-in Functions
1803 X<built-in> X<override> X<CORE> X<CORE::GLOBAL>
1805 Many built-in functions may be overridden, though this should be tried
1806 only occasionally and for good reason. Typically this might be
1807 done by a package attempting to emulate missing built-in functionality
1808 on a non-Unix system.
1810 Overriding may be done only by importing the name from a module at
1811 compile time--ordinary predeclaration isn't good enough. However, the
1812 C<use subs> pragma lets you, in effect, predeclare subs
1813 via the import syntax, and these names may then override built-in ones:
1815 use subs 'chdir', 'chroot', 'chmod', 'chown';
1819 To unambiguously refer to the built-in form, precede the
1820 built-in name with the special package qualifier C<CORE::>. For example,
1821 saying C<CORE::open()> always refers to the built-in C<open()>, even
1822 if the current package has imported some other subroutine called
1823 C<&open()> from elsewhere. Even though it looks like a regular
1824 function call, it isn't: the CORE:: prefix in that case is part of Perl's
1825 syntax, and works for any keyword, regardless of what is in the CORE
1826 package. Taking a reference to it, that is, C<\&CORE::open>, only works
1827 for some keywords. See L<CORE>.
1829 Library modules should not in general export built-in names like C<open>
1830 or C<chdir> as part of their default C<@EXPORT> list, because these may
1831 sneak into someone else's namespace and change the semantics unexpectedly.
1832 Instead, if the module adds that name to C<@EXPORT_OK>, then it's
1833 possible for a user to import the name explicitly, but not implicitly.
1834 That is, they could say
1838 and it would import the C<open> override. But if they said
1842 they would get the default imports without overrides.
1844 The foregoing mechanism for overriding built-in is restricted, quite
1845 deliberately, to the package that requests the import. There is a second
1846 method that is sometimes applicable when you wish to override a built-in
1847 everywhere, without regard to namespace boundaries. This is achieved by
1848 importing a sub into the special namespace C<CORE::GLOBAL::>. Here is an
1849 example that quite brazenly replaces the C<glob> operator with something
1850 that understands regular expressions.
1855 @EXPORT_OK = 'glob';
1861 my $where = ($sym =~ s/^GLOBAL_// ? 'CORE::GLOBAL' : caller(0));
1862 $pkg->export($where, $sym, @_);
1868 if (opendir my $d, '.') {
1869 @got = grep /$pat/, readdir $d;
1876 And here's how it could be (ab)used:
1878 #use REGlob 'GLOBAL_glob'; # override glob() in ALL namespaces
1880 use REGlob 'glob'; # override glob() in Foo:: only
1881 print for <^[a-z_]+\.pm\$>; # show all pragmatic modules
1883 The initial comment shows a contrived, even dangerous example.
1884 By overriding C<glob> globally, you would be forcing the new (and
1885 subversive) behavior for the C<glob> operator for I<every> namespace,
1886 without the complete cognizance or cooperation of the modules that own
1887 those namespaces. Naturally, this should be done with extreme caution--if
1888 it must be done at all.
1890 The C<REGlob> example above does not implement all the support needed to
1891 cleanly override perl's C<glob> operator. The built-in C<glob> has
1892 different behaviors depending on whether it appears in a scalar or list
1893 context, but our C<REGlob> doesn't. Indeed, many perl built-in have such
1894 context sensitive behaviors, and these must be adequately supported by
1895 a properly written override. For a fully functional example of overriding
1896 C<glob>, study the implementation of C<File::DosGlob> in the standard
1899 When you override a built-in, your replacement should be consistent (if
1900 possible) with the built-in native syntax. You can achieve this by using
1901 a suitable prototype. To get the prototype of an overridable built-in,
1902 use the C<prototype> function with an argument of C<"CORE::builtin_name">
1903 (see L<perlfunc/prototype>).
1905 Note however that some built-ins can't have their syntax expressed by a
1906 prototype (such as C<system> or C<chomp>). If you override them you won't
1907 be able to fully mimic their original syntax.
1909 The built-ins C<do>, C<require> and C<glob> can also be overridden, but due
1910 to special magic, their original syntax is preserved, and you don't have
1911 to define a prototype for their replacements. (You can't override the
1912 C<do BLOCK> syntax, though).
1914 C<require> has special additional dark magic: if you invoke your
1915 C<require> replacement as C<require Foo::Bar>, it will actually receive
1916 the argument C<"Foo/Bar.pm"> in @_. See L<perlfunc/require>.
1918 And, as you'll have noticed from the previous example, if you override
1919 C<glob>, the C<< <*> >> glob operator is overridden as well.
1921 In a similar fashion, overriding the C<readline> function also overrides
1922 the equivalent I/O operator C<< <FILEHANDLE> >>. Also, overriding
1923 C<readpipe> also overrides the operators C<``> and C<qx//>.
1925 Finally, some built-ins (e.g. C<exists> or C<grep>) can't be overridden.
1928 X<autoloading> X<AUTOLOAD>
1930 If you call a subroutine that is undefined, you would ordinarily
1931 get an immediate, fatal error complaining that the subroutine doesn't
1932 exist. (Likewise for subroutines being used as methods, when the
1933 method doesn't exist in any base class of the class's package.)
1934 However, if an C<AUTOLOAD> subroutine is defined in the package or
1935 packages used to locate the original subroutine, then that
1936 C<AUTOLOAD> subroutine is called with the arguments that would have
1937 been passed to the original subroutine. The fully qualified name
1938 of the original subroutine magically appears in the global $AUTOLOAD
1939 variable of the same package as the C<AUTOLOAD> routine. The name
1940 is not passed as an ordinary argument because, er, well, just
1941 because, that's why. (As an exception, a method call to a nonexistent
1942 C<import> or C<unimport> method is just skipped instead. Also, if
1943 the AUTOLOAD subroutine is an XSUB, there are other ways to retrieve the
1944 subroutine name. See L<perlguts/Autoloading with XSUBs> for details.)
1947 Many C<AUTOLOAD> routines load in a definition for the requested
1948 subroutine using eval(), then execute that subroutine using a special
1949 form of goto() that erases the stack frame of the C<AUTOLOAD> routine
1950 without a trace. (See the source to the standard module documented
1951 in L<AutoLoader>, for example.) But an C<AUTOLOAD> routine can
1952 also just emulate the routine and never define it. For example,
1953 let's pretend that a function that wasn't defined should just invoke
1954 C<system> with those arguments. All you'd do is:
1957 our $AUTOLOAD; # keep 'use strict' happy
1958 my $program = $AUTOLOAD;
1959 $program =~ s/.*:://;
1960 system($program, @_);
1966 In fact, if you predeclare functions you want to call that way, you don't
1967 even need parentheses:
1969 use subs qw(date who ls);
1974 A more complete example of this is the Shell module on CPAN, which
1975 can treat undefined subroutine calls as calls to external programs.
1977 Mechanisms are available to help modules writers split their modules
1978 into autoloadable files. See the standard AutoLoader module
1979 described in L<AutoLoader> and in L<AutoSplit>, the standard
1980 SelfLoader modules in L<SelfLoader>, and the document on adding C
1981 functions to Perl code in L<perlxs>.
1983 =head2 Subroutine Attributes
1984 X<attribute> X<subroutine, attribute> X<attrs>
1986 A subroutine declaration or definition may have a list of attributes
1987 associated with it. If such an attribute list is present, it is
1988 broken up at space or colon boundaries and treated as though a
1989 C<use attributes> had been seen. See L<attributes> for details
1990 about what attributes are currently supported.
1991 Unlike the limitation with the obsolescent C<use attrs>, the
1992 C<sub : ATTRLIST> syntax works to associate the attributes with
1993 a pre-declaration, and not just with a subroutine definition.
1995 The attributes must be valid as simple identifier names (without any
1996 punctuation other than the '_' character). They may have a parameter
1997 list appended, which is only checked for whether its parentheses ('(',')')
2000 Examples of valid syntax (even though the attributes are unknown):
2002 sub fnord (&\%) : switch(10,foo(7,3)) : expensive;
2003 sub plugh () : Ugly('\(") :Bad;
2004 sub xyzzy : _5x5 { ... }
2006 Examples of invalid syntax:
2008 sub fnord : switch(10,foo(); # ()-string not balanced
2009 sub snoid : Ugly('('); # ()-string not balanced
2010 sub xyzzy : 5x5; # "5x5" not a valid identifier
2011 sub plugh : Y2::north; # "Y2::north" not a simple identifier
2012 sub snurt : foo + bar; # "+" not a colon or space
2014 The attribute list is passed as a list of constant strings to the code
2015 which associates them with the subroutine. In particular, the second example
2016 of valid syntax above currently looks like this in terms of how it's
2019 use attributes __PACKAGE__, \&plugh, q[Ugly('\(")], 'Bad';
2021 For further details on attribute lists and their manipulation,
2022 see L<attributes> and L<Attribute::Handlers>.
2026 See L<perlref/"Function Templates"> for more about references and closures.
2027 See L<perlxs> if you'd like to learn about calling C subroutines from Perl.
2028 See L<perlembed> if you'd like to learn about calling Perl subroutines from C.
2029 See L<perlmod> to learn about bundling up your functions in separate files.
2030 See L<perlmodlib> to learn what library modules come standard on your system.
2031 See L<perlootut> to learn how to make object method calls.