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 To define an anonymous subroutine at runtime:
22 X<subroutine, anonymous>
24 $subref = sub BLOCK; # no proto
25 $subref = sub (PROTO) BLOCK; # with proto
26 $subref = sub : ATTRS BLOCK; # with attributes
27 $subref = sub (PROTO) : ATTRS BLOCK; # with proto and attributes
29 To import subroutines:
32 use MODULE qw(NAME1 NAME2 NAME3);
35 X<subroutine, call> X<call>
37 NAME(LIST); # & is optional with parentheses.
38 NAME LIST; # Parentheses optional if predeclared/imported.
39 &NAME(LIST); # Circumvent prototypes.
40 &NAME; # Makes current @_ visible to called subroutine.
44 Like many languages, Perl provides for user-defined subroutines.
45 These may be located anywhere in the main program, loaded in from
46 other files via the C<do>, C<require>, or C<use> keywords, or
47 generated on the fly using C<eval> or anonymous subroutines.
48 You can even call a function indirectly using a variable containing
49 its name or a CODE reference.
51 The Perl model for function call and return values is simple: all
52 functions are passed as parameters one single flat list of scalars, and
53 all functions likewise return to their caller one single flat list of
54 scalars. Any arrays or hashes in these call and return lists will
55 collapse, losing their identities--but you may always use
56 pass-by-reference instead to avoid this. Both call and return lists may
57 contain as many or as few scalar elements as you'd like. (Often a
58 function without an explicit return statement is called a subroutine, but
59 there's really no difference from Perl's perspective.)
60 X<subroutine, parameter> X<parameter>
62 Any arguments passed in show up in the array C<@_>. Therefore, if
63 you called a function with two arguments, those would be stored in
64 C<$_[0]> and C<$_[1]>. The array C<@_> is a local array, but its
65 elements are aliases for the actual scalar parameters. In particular,
66 if an element C<$_[0]> is updated, the corresponding argument is
67 updated (or an error occurs if it is not updatable). If an argument
68 is an array or hash element which did not exist when the function
69 was called, that element is created only when (and if) it is modified
70 or a reference to it is taken. (Some earlier versions of Perl
71 created the element whether or not the element was assigned to.)
72 Assigning to the whole array C<@_> removes that aliasing, and does
73 not update any arguments.
74 X<subroutine, argument> X<argument> X<@_>
76 The return value of a subroutine is the value of the last expression
77 evaluated by that sub, or the empty list in the case of an empty sub.
78 More explicitly, a C<return> statement may be used to exit the
79 subroutine, optionally specifying the returned value, which will be
80 evaluated in the appropriate context (list, scalar, or void) depending
81 on the context of the subroutine call. If you specify no return value,
82 the subroutine returns an empty list in list context, the undefined
83 value in scalar context, or nothing in void context. If you return
84 one or more aggregates (arrays and hashes), these will be flattened
85 together into one large indistinguishable list.
86 X<subroutine, return value> X<return value> X<return>
88 Perl does not have named formal parameters. In practice all you
89 do is assign to a C<my()> list of these. Variables that aren't
90 declared to be private are global variables. For gory details
91 on creating private variables, see L<"Private Variables via my()">
92 and L<"Temporary Values via local()">. To create protected
93 environments for a set of functions in a separate package (and
94 probably a separate file), see L<perlmod/"Packages">.
95 X<formal parameter> X<parameter, formal>
102 $max = $foo if $max < $foo;
106 $bestday = max($mon,$tue,$wed,$thu,$fri);
110 # get a line, combining continuation lines
111 # that start with whitespace
114 $thisline = $lookahead; # global variables!
115 LINE: while (defined($lookahead = <STDIN>)) {
116 if ($lookahead =~ /^[ \t]/) {
117 $thisline .= $lookahead;
126 $lookahead = <STDIN>; # get first line
127 while (defined($line = get_line())) {
131 Assigning to a list of private variables to name your arguments:
134 my($key, $value) = @_;
135 $Foo{$key} = $value unless $Foo{$key};
138 Because the assignment copies the values, this also has the effect
139 of turning call-by-reference into call-by-value. Otherwise a
140 function is free to do in-place modifications of C<@_> and change
142 X<call-by-reference> X<call-by-value>
144 upcase_in($v1, $v2); # this changes $v1 and $v2
146 for (@_) { tr/a-z/A-Z/ }
149 You aren't allowed to modify constants in this way, of course. If an
150 argument were actually literal and you tried to change it, you'd take a
151 (presumably fatal) exception. For example, this won't work:
152 X<call-by-reference> X<call-by-value>
154 upcase_in("frederick");
156 It would be much safer if the C<upcase_in()> function
157 were written to return a copy of its parameters instead
158 of changing them in place:
160 ($v3, $v4) = upcase($v1, $v2); # this doesn't change $v1 and $v2
162 return unless defined wantarray; # void context, do nothing
164 for (@parms) { tr/a-z/A-Z/ }
165 return wantarray ? @parms : $parms[0];
168 Notice how this (unprototyped) function doesn't care whether it was
169 passed real scalars or arrays. Perl sees all arguments as one big,
170 long, flat parameter list in C<@_>. This is one area where
171 Perl's simple argument-passing style shines. The C<upcase()>
172 function would work perfectly well without changing the C<upcase()>
173 definition even if we fed it things like this:
175 @newlist = upcase(@list1, @list2);
176 @newlist = upcase( split /:/, $var );
178 Do not, however, be tempted to do this:
180 (@a, @b) = upcase(@list1, @list2);
182 Like the flattened incoming parameter list, the return list is also
183 flattened on return. So all you have managed to do here is stored
184 everything in C<@a> and made C<@b> empty. See
185 L<Pass by Reference> for alternatives.
187 A subroutine may be called using an explicit C<&> prefix. The
188 C<&> is optional in modern Perl, as are parentheses if the
189 subroutine has been predeclared. The C<&> is I<not> optional
190 when just naming the subroutine, such as when it's used as
191 an argument to defined() or undef(). Nor is it optional when you
192 want to do an indirect subroutine call with a subroutine name or
193 reference using the C<&$subref()> or C<&{$subref}()> constructs,
194 although the C<< $subref->() >> notation solves that problem.
195 See L<perlref> for more about all that.
198 Subroutines may be called recursively. If a subroutine is called
199 using the C<&> form, the argument list is optional, and if omitted,
200 no C<@_> array is set up for the subroutine: the C<@_> array at the
201 time of the call is visible to subroutine instead. This is an
202 efficiency mechanism that new users may wish to avoid.
205 &foo(1,2,3); # pass three arguments
206 foo(1,2,3); # the same
208 foo(); # pass a null list
211 &foo; # foo() get current args, like foo(@_) !!
212 foo; # like foo() IFF sub foo predeclared, else "foo"
214 Not only does the C<&> form make the argument list optional, it also
215 disables any prototype checking on arguments you do provide. This
216 is partly for historical reasons, and partly for having a convenient way
217 to cheat if you know what you're doing. See L<Prototypes> below.
220 Subroutines whose names are in all upper case are reserved to the Perl
221 core, as are modules whose names are in all lower case. A subroutine in
222 all capitals is a loosely-held convention meaning it will be called
223 indirectly by the run-time system itself, usually due to a triggered event.
224 Subroutines that do special, pre-defined things include C<AUTOLOAD>, C<CLONE>,
225 C<DESTROY> plus all functions mentioned in L<perltie> and L<PerlIO::via>.
227 The C<BEGIN>, C<CHECK>, C<INIT> and C<END> subroutines are not so much
228 subroutines as named special code blocks, of which you can have more
229 than one in a package, and which you can B<not> call explicitly. See
230 L<perlmod/"BEGIN, CHECK, INIT and END">
232 =head2 Private Variables via my()
233 X<my> X<variable, lexical> X<lexical> X<lexical variable> X<scope, lexical>
234 X<lexical scope> X<attributes, my>
238 my $foo; # declare $foo lexically local
239 my (@wid, %get); # declare list of variables local
240 my $foo = "flurp"; # declare $foo lexical, and init it
241 my @oof = @bar; # declare @oof lexical, and init it
242 my $x : Foo = $y; # similar, with an attribute applied
244 B<WARNING>: The use of attribute lists on C<my> declarations is still
245 evolving. The current semantics and interface are subject to change.
246 See L<attributes> and L<Attribute::Handlers>.
248 The C<my> operator declares the listed variables to be lexically
249 confined to the enclosing block, conditional (C<if/unless/elsif/else>),
250 loop (C<for/foreach/while/until/continue>), subroutine, C<eval>,
251 or C<do/require/use>'d file. If more than one value is listed, the
252 list must be placed in parentheses. All listed elements must be
253 legal lvalues. Only alphanumeric identifiers may be lexically
254 scoped--magical built-ins like C<$/> must currently be C<local>ized
255 with C<local> instead.
257 Unlike dynamic variables created by the C<local> operator, lexical
258 variables declared with C<my> are totally hidden from the outside
259 world, including any called subroutines. This is true if it's the
260 same subroutine called from itself or elsewhere--every call gets
264 This doesn't mean that a C<my> variable declared in a statically
265 enclosing lexical scope would be invisible. Only dynamic scopes
266 are cut off. For example, the C<bumpx()> function below has access
267 to the lexical $x variable because both the C<my> and the C<sub>
268 occurred at the same scope, presumably file scope.
273 An C<eval()>, however, can see lexical variables of the scope it is
274 being evaluated in, so long as the names aren't hidden by declarations within
275 the C<eval()> itself. See L<perlref>.
278 The parameter list to my() may be assigned to if desired, which allows you
279 to initialize your variables. (If no initializer is given for a
280 particular variable, it is created with the undefined value.) Commonly
281 this is used to name input parameters to a subroutine. Examples:
283 $arg = "fred"; # "global" variable
285 print "$arg thinks the root is $n\n";
286 fred thinks the root is 3
289 my $arg = shift; # name doesn't matter
294 The C<my> is simply a modifier on something you might assign to. So when
295 you do assign to variables in its argument list, C<my> doesn't
296 change whether those variables are viewed as a scalar or an array. So
298 my ($foo) = <STDIN>; # WRONG?
301 both supply a list context to the right-hand side, while
305 supplies a scalar context. But the following declares only one variable:
307 my $foo, $bar = 1; # WRONG
309 That has the same effect as
314 The declared variable is not introduced (is not visible) until after
315 the current statement. Thus,
319 can be used to initialize a new $x with the value of the old $x, and
322 my $x = 123 and $x == 123
324 is false unless the old $x happened to have the value C<123>.
326 Lexical scopes of control structures are not bounded precisely by the
327 braces that delimit their controlled blocks; control expressions are
328 part of that scope, too. Thus in the loop
330 while (my $line = <>) {
336 the scope of $line extends from its declaration throughout the rest of
337 the loop construct (including the C<continue> clause), but not beyond
338 it. Similarly, in the conditional
340 if ((my $answer = <STDIN>) =~ /^yes$/i) {
342 } elsif ($answer =~ /^no$/i) {
346 die "'$answer' is neither 'yes' nor 'no'";
349 the scope of $answer extends from its declaration through the rest
350 of that conditional, including any C<elsif> and C<else> clauses,
351 but not beyond it. See L<perlsyn/"Simple statements"> for information
352 on the scope of variables in statements with modifiers.
354 The C<foreach> loop defaults to scoping its index variable dynamically
355 in the manner of C<local>. However, if the index variable is
356 prefixed with the keyword C<my>, or if there is already a lexical
357 by that name in scope, then a new lexical is created instead. Thus
361 for my $i (1, 2, 3) {
365 the scope of $i extends to the end of the loop, but not beyond it,
366 rendering the value of $i inaccessible within C<some_function()>.
369 Some users may wish to encourage the use of lexically scoped variables.
370 As an aid to catching implicit uses to package variables,
371 which are always global, if you say
375 then any variable mentioned from there to the end of the enclosing
376 block must either refer to a lexical variable, be predeclared via
377 C<our> or C<use vars>, or else must be fully qualified with the package name.
378 A compilation error results otherwise. An inner block may countermand
379 this with C<no strict 'vars'>.
381 A C<my> has both a compile-time and a run-time effect. At compile
382 time, the compiler takes notice of it. The principal usefulness
383 of this is to quiet C<use strict 'vars'>, but it is also essential
384 for generation of closures as detailed in L<perlref>. Actual
385 initialization is delayed until run time, though, so it gets executed
386 at the appropriate time, such as each time through a loop, for
389 Variables declared with C<my> are not part of any package and are therefore
390 never fully qualified with the package name. In particular, you're not
391 allowed to try to make a package variable (or other global) lexical:
393 my $pack::var; # ERROR! Illegal syntax
394 my $_; # also illegal (currently)
396 In fact, a dynamic variable (also known as package or global variables)
397 are still accessible using the fully qualified C<::> notation even while a
398 lexical of the same name is also visible:
403 print "$x and $::x\n";
405 That will print out C<20> and C<10>.
407 You may declare C<my> variables at the outermost scope of a file
408 to hide any such identifiers from the world outside that file. This
409 is similar in spirit to C's static variables when they are used at
410 the file level. To do this with a subroutine requires the use of
411 a closure (an anonymous function that accesses enclosing lexicals).
412 If you want to create a private subroutine that cannot be called
413 from outside that block, it can declare a lexical variable containing
414 an anonymous sub reference:
416 my $secret_version = '1.001-beta';
417 my $secret_sub = sub { print $secret_version };
420 As long as the reference is never returned by any function within the
421 module, no outside module can see the subroutine, because its name is not in
422 any package's symbol table. Remember that it's not I<REALLY> called
423 C<$some_pack::secret_version> or anything; it's just $secret_version,
424 unqualified and unqualifiable.
426 This does not work with object methods, however; all object methods
427 have to be in the symbol table of some package to be found. See
428 L<perlref/"Function Templates"> for something of a work-around to
431 =head2 Persistent Private Variables
432 X<static> X<variable, persistent> X<variable, static> X<closure>
434 Just because a lexical variable is lexically (also called statically)
435 scoped to its enclosing block, C<eval>, or C<do> FILE, this doesn't mean that
436 within a function it works like a C static. It normally works more
437 like a C auto, but with implicit garbage collection.
439 Unlike local variables in C or C++, Perl's lexical variables don't
440 necessarily get recycled just because their scope has exited.
441 If something more permanent is still aware of the lexical, it will
442 stick around. So long as something else references a lexical, that
443 lexical won't be freed--which is as it should be. You wouldn't want
444 memory being free until you were done using it, or kept around once you
445 were done. Automatic garbage collection takes care of this for you.
447 This means that you can pass back or save away references to lexical
448 variables, whereas to return a pointer to a C auto is a grave error.
449 It also gives us a way to simulate C's function statics. Here's a
450 mechanism for giving a function private variables with both lexical
451 scoping and a static lifetime. If you do want to create something like
452 C's static variables, just enclose the whole function in an extra block,
453 and put the static variable outside the function but in the block.
458 return ++$secret_val;
461 # $secret_val now becomes unreachable by the outside
462 # world, but retains its value between calls to gimme_another
464 If this function is being sourced in from a separate file
465 via C<require> or C<use>, then this is probably just fine. If it's
466 all in the main program, you'll need to arrange for the C<my>
467 to be executed early, either by putting the whole block above
468 your main program, or more likely, placing merely a C<BEGIN>
469 code block around it to make sure it gets executed before your program
475 return ++$secret_val;
479 See L<perlmod/"BEGIN, CHECK, INIT and END"> about the
480 special triggered code blocks, C<BEGIN>, C<CHECK>, C<INIT> and C<END>.
482 If declared at the outermost scope (the file scope), then lexicals
483 work somewhat like C's file statics. They are available to all
484 functions in that same file declared below them, but are inaccessible
485 from outside that file. This strategy is sometimes used in modules
486 to create private variables that the whole module can see.
488 =head2 Temporary Values via local()
489 X<local> X<scope, dynamic> X<dynamic scope> X<variable, local>
490 X<variable, temporary>
492 B<WARNING>: In general, you should be using C<my> instead of C<local>, because
493 it's faster and safer. Exceptions to this include the global punctuation
494 variables, global filehandles and formats, and direct manipulation of the
495 Perl symbol table itself. C<local> is mostly used when the current value
496 of a variable must be visible to called subroutines.
500 # localization of values
502 local $foo; # make $foo dynamically local
503 local (@wid, %get); # make list of variables local
504 local $foo = "flurp"; # make $foo dynamic, and init it
505 local @oof = @bar; # make @oof dynamic, and init it
507 local $hash{key} = "val"; # sets a local value for this hash entry
508 local ($cond ? $v1 : $v2); # several types of lvalues support
511 # localization of symbols
513 local *FH; # localize $FH, @FH, %FH, &FH ...
514 local *merlyn = *randal; # now $merlyn is really $randal, plus
515 # @merlyn is really @randal, etc
516 local *merlyn = 'randal'; # SAME THING: promote 'randal' to *randal
517 local *merlyn = \$randal; # just alias $merlyn, not @merlyn etc
519 A C<local> modifies its listed variables to be "local" to the
520 enclosing block, C<eval>, or C<do FILE>--and to I<any subroutine
521 called from within that block>. A C<local> just gives temporary
522 values to global (meaning package) variables. It does I<not> create
523 a local variable. This is known as dynamic scoping. Lexical scoping
524 is done with C<my>, which works more like C's auto declarations.
526 Some types of lvalues can be localized as well : hash and array elements
527 and slices, conditionals (provided that their result is always
528 localizable), and symbolic references. As for simple variables, this
529 creates new, dynamically scoped values.
531 If more than one variable or expression is given to C<local>, they must be
532 placed in parentheses. This operator works
533 by saving the current values of those variables in its argument list on a
534 hidden stack and restoring them upon exiting the block, subroutine, or
535 eval. This means that called subroutines can also reference the local
536 variable, but not the global one. The argument list may be assigned to if
537 desired, which allows you to initialize your local variables. (If no
538 initializer is given for a particular variable, it is created with an
541 Because C<local> is a run-time operator, it gets executed each time
542 through a loop. Consequently, it's more efficient to localize your
543 variables outside the loop.
545 =head3 Grammatical note on local()
548 A C<local> is simply a modifier on an lvalue expression. When you assign to
549 a C<local>ized variable, the C<local> doesn't change whether its list is viewed
550 as a scalar or an array. So
552 local($foo) = <STDIN>;
553 local @FOO = <STDIN>;
555 both supply a list context to the right-hand side, while
557 local $foo = <STDIN>;
559 supplies a scalar context.
561 =head3 Localization of special variables
562 X<local, special variable>
564 If you localize a special variable, you'll be giving a new value to it,
565 but its magic won't go away. That means that all side-effects related
566 to this magic still work with the localized value.
568 This feature allows code like this to work :
570 # Read the whole contents of FILE in $slurp
571 { local $/ = undef; $slurp = <FILE>; }
573 Note, however, that this restricts localization of some values ; for
574 example, the following statement dies, as of perl 5.9.0, with an error
575 I<Modification of a read-only value attempted>, because the $1 variable is
576 magical and read-only :
580 Similarly, but in a way more difficult to spot, the following snippet will
583 sub f { local $_ = "foo"; print }
585 # now $_ is aliased to $1, thus is magic and readonly
589 See next section for an alternative to this situation.
591 B<WARNING>: Localization of tied arrays and hashes does not currently
593 This will be fixed in a future release of Perl; in the meantime, avoid
594 code that relies on any particular behaviour of localising tied arrays
595 or hashes (localising individual elements is still okay).
596 See L<perl58delta/"Localising Tied Arrays and Hashes Is Broken"> for more
600 =head3 Localization of globs
601 X<local, glob> X<glob>
607 creates a whole new symbol table entry for the glob C<name> in the
608 current package. That means that all variables in its glob slot ($name,
609 @name, %name, &name, and the C<name> filehandle) are dynamically reset.
611 This implies, among other things, that any magic eventually carried by
612 those variables is locally lost. In other words, saying C<local */>
613 will not have any effect on the internal value of the input record
616 Notably, if you want to work with a brand new value of the default scalar
617 $_, and avoid the potential problem listed above about $_ previously
618 carrying a magic value, you should use C<local *_> instead of C<local $_>.
619 As of perl 5.9.1, you can also use the lexical form of C<$_> (declaring it
620 with C<my $_>), which avoids completely this problem.
622 =head3 Localization of elements of composite types
623 X<local, composite type element> X<local, array element> X<local, hash element>
625 It's also worth taking a moment to explain what happens when you
626 C<local>ize a member of a composite type (i.e. an array or hash element).
627 In this case, the element is C<local>ized I<by name>. This means that
628 when the scope of the C<local()> ends, the saved value will be
629 restored to the hash element whose key was named in the C<local()>, or
630 the array element whose index was named in the C<local()>. If that
631 element was deleted while the C<local()> was in effect (e.g. by a
632 C<delete()> from a hash or a C<shift()> of an array), it will spring
633 back into existence, possibly extending an array and filling in the
634 skipped elements with C<undef>. For instance, if you say
636 %hash = ( 'This' => 'is', 'a' => 'test' );
640 local($hash{'a'}) = 'drill';
641 while (my $e = pop(@ary)) {
646 $hash{'only a'} = 'test';
650 print join(' ', map { "$_ $hash{$_}" } sort keys %hash),".\n";
651 print "The array has ",scalar(@ary)," elements: ",
652 join(', ', map { defined $_ ? $_ : 'undef' } @ary),"\n";
659 This is a test only a test.
660 The array has 6 elements: 0, 1, 2, undef, undef, 5
662 The behavior of local() on non-existent members of composite
663 types is subject to change in future.
665 =head2 Lvalue subroutines
666 X<lvalue> X<subroutine, lvalue>
668 B<WARNING>: Lvalue subroutines are still experimental and the
669 implementation may change in future versions of Perl.
671 It is possible to return a modifiable value from a subroutine.
672 To do this, you have to declare the subroutine to return an lvalue.
675 sub canmod : lvalue {
676 # return $val; this doesn't work, don't say "return"
683 canmod() = 5; # assigns to $val
686 The scalar/list context for the subroutine and for the right-hand
687 side of assignment is determined as if the subroutine call is replaced
688 by a scalar. For example, consider:
690 data(2,3) = get_data(3,4);
692 Both subroutines here are called in a scalar context, while in:
694 (data(2,3)) = get_data(3,4);
698 (data(2),data(3)) = get_data(3,4);
700 all the subroutines are called in a list context.
704 =item Lvalue subroutines are EXPERIMENTAL
706 They appear to be convenient, but there are several reasons to be
709 You can't use the return keyword, you must pass out the value before
710 falling out of subroutine scope. (see comment in example above). This
711 is usually not a problem, but it disallows an explicit return out of a
712 deeply nested loop, which is sometimes a nice way out.
714 They violate encapsulation. A normal mutator can check the supplied
715 argument before setting the attribute it is protecting, an lvalue
716 subroutine never gets that chance. Consider;
718 my $some_array_ref = []; # protected by mutators ??
720 sub set_arr { # normal mutator
722 die("expected array, you supplied ", ref $val)
723 unless ref $val eq 'ARRAY';
724 $some_array_ref = $val;
726 sub set_arr_lv : lvalue { # lvalue mutator
730 # set_arr_lv cannot stop this !
731 set_arr_lv() = { a => 1 };
735 =head2 Passing Symbol Table Entries (typeglobs)
738 B<WARNING>: The mechanism described in this section was originally
739 the only way to simulate pass-by-reference in older versions of
740 Perl. While it still works fine in modern versions, the new reference
741 mechanism is generally easier to work with. See below.
743 Sometimes you don't want to pass the value of an array to a subroutine
744 but rather the name of it, so that the subroutine can modify the global
745 copy of it rather than working with a local copy. In perl you can
746 refer to all objects of a particular name by prefixing the name
747 with a star: C<*foo>. This is often known as a "typeglob", because the
748 star on the front can be thought of as a wildcard match for all the
749 funny prefix characters on variables and subroutines and such.
751 When evaluated, the typeglob produces a scalar value that represents
752 all the objects of that name, including any filehandle, format, or
753 subroutine. When assigned to, it causes the name mentioned to refer to
754 whatever C<*> value was assigned to it. Example:
757 local(*someary) = @_;
758 foreach $elem (@someary) {
765 Scalars are already passed by reference, so you can modify
766 scalar arguments without using this mechanism by referring explicitly
767 to C<$_[0]> etc. You can modify all the elements of an array by passing
768 all the elements as scalars, but you have to use the C<*> mechanism (or
769 the equivalent reference mechanism) to C<push>, C<pop>, or change the size of
770 an array. It will certainly be faster to pass the typeglob (or reference).
772 Even if you don't want to modify an array, this mechanism is useful for
773 passing multiple arrays in a single LIST, because normally the LIST
774 mechanism will merge all the array values so that you can't extract out
775 the individual arrays. For more on typeglobs, see
776 L<perldata/"Typeglobs and Filehandles">.
778 =head2 When to Still Use local()
779 X<local> X<variable, local>
781 Despite the existence of C<my>, there are still three places where the
782 C<local> operator still shines. In fact, in these three places, you
783 I<must> use C<local> instead of C<my>.
789 You need to give a global variable a temporary value, especially $_.
791 The global variables, like C<@ARGV> or the punctuation variables, must be
792 C<local>ized with C<local()>. This block reads in F</etc/motd>, and splits
793 it up into chunks separated by lines of equal signs, which are placed
797 local @ARGV = ("/etc/motd");
800 @Fields = split /^\s*=+\s*$/;
803 It particular, it's important to C<local>ize $_ in any routine that assigns
804 to it. Look out for implicit assignments in C<while> conditionals.
808 You need to create a local file or directory handle or a local function.
810 A function that needs a filehandle of its own must use
811 C<local()> on a complete typeglob. This can be used to create new symbol
815 local (*READER, *WRITER); # not my!
816 pipe (READER, WRITER) or die "pipe: $!";
817 return (*READER, *WRITER);
819 ($head, $tail) = ioqueue();
821 See the Symbol module for a way to create anonymous symbol table
824 Because assignment of a reference to a typeglob creates an alias, this
825 can be used to create what is effectively a local function, or at least,
829 local *grow = \&shrink; # only until this block exists
830 grow(); # really calls shrink()
831 move(); # if move() grow()s, it shrink()s too
833 grow(); # get the real grow() again
835 See L<perlref/"Function Templates"> for more about manipulating
836 functions by name in this way.
840 You want to temporarily change just one element of an array or hash.
842 You can C<local>ize just one element of an aggregate. Usually this
846 local $SIG{INT} = 'IGNORE';
847 funct(); # uninterruptible
849 # interruptibility automatically restored here
851 But it also works on lexically declared aggregates. Prior to 5.005,
852 this operation could on occasion misbehave.
856 =head2 Pass by Reference
857 X<pass by reference> X<pass-by-reference> X<reference>
859 If you want to pass more than one array or hash into a function--or
860 return them from it--and have them maintain their integrity, then
861 you're going to have to use an explicit pass-by-reference. Before you
862 do that, you need to understand references as detailed in L<perlref>.
863 This section may not make much sense to you otherwise.
865 Here are a few simple examples. First, let's pass in several arrays
866 to a function and have it C<pop> all of then, returning a new list
867 of all their former last elements:
869 @tailings = popmany ( \@a, \@b, \@c, \@d );
874 foreach $aref ( @_ ) {
875 push @retlist, pop @$aref;
880 Here's how you might write a function that returns a
881 list of keys occurring in all the hashes passed to it:
883 @common = inter( \%foo, \%bar, \%joe );
885 my ($k, $href, %seen); # locals
887 while ( $k = each %$href ) {
891 return grep { $seen{$_} == @_ } keys %seen;
894 So far, we're using just the normal list return mechanism.
895 What happens if you want to pass or return a hash? Well,
896 if you're using only one of them, or you don't mind them
897 concatenating, then the normal calling convention is ok, although
900 Where people get into trouble is here:
902 (@a, @b) = func(@c, @d);
904 (%a, %b) = func(%c, %d);
906 That syntax simply won't work. It sets just C<@a> or C<%a> and
907 clears the C<@b> or C<%b>. Plus the function didn't get passed
908 into two separate arrays or hashes: it got one long list in C<@_>,
911 If you can arrange for everyone to deal with this through references, it's
912 cleaner code, although not so nice to look at. Here's a function that
913 takes two array references as arguments, returning the two array elements
914 in order of how many elements they have in them:
916 ($aref, $bref) = func(\@c, \@d);
917 print "@$aref has more than @$bref\n";
919 my ($cref, $dref) = @_;
920 if (@$cref > @$dref) {
921 return ($cref, $dref);
923 return ($dref, $cref);
927 It turns out that you can actually do this also:
929 (*a, *b) = func(\@c, \@d);
930 print "@a has more than @b\n";
940 Here we're using the typeglobs to do symbol table aliasing. It's
941 a tad subtle, though, and also won't work if you're using C<my>
942 variables, because only globals (even in disguise as C<local>s)
943 are in the symbol table.
945 If you're passing around filehandles, you could usually just use the bare
946 typeglob, like C<*STDOUT>, but typeglobs references work, too.
952 print $fh "her um well a hmmm\n";
955 $rec = get_rec(\*STDIN);
961 If you're planning on generating new filehandles, you could do this.
962 Notice to pass back just the bare *FH, not its reference.
967 return open (FH, $path) ? *FH : undef;
971 X<prototype> X<subroutine, prototype>
973 Perl supports a very limited kind of compile-time argument checking
974 using function prototyping. If you declare
978 then C<mypush()> takes arguments exactly like C<push()> does. The
979 function declaration must be visible at compile time. The prototype
980 affects only interpretation of new-style calls to the function,
981 where new-style is defined as not using the C<&> character. In
982 other words, if you call it like a built-in function, then it behaves
983 like a built-in function. If you call it like an old-fashioned
984 subroutine, then it behaves like an old-fashioned subroutine. It
985 naturally falls out from this rule that prototypes have no influence
986 on subroutine references like C<\&foo> or on indirect subroutine
987 calls like C<&{$subref}> or C<< $subref->() >>.
989 Method calls are not influenced by prototypes either, because the
990 function to be called is indeterminate at compile time, since
991 the exact code called depends on inheritance.
993 Because the intent of this feature is primarily to let you define
994 subroutines that work like built-in functions, here are prototypes
995 for some other functions that parse almost exactly like the
996 corresponding built-in.
998 Declared as Called as
1000 sub mylink ($$) mylink $old, $new
1001 sub myvec ($$$) myvec $var, $offset, 1
1002 sub myindex ($$;$) myindex &getstring, "substr"
1003 sub mysyswrite ($$$;$) mysyswrite $buf, 0, length($buf) - $off, $off
1004 sub myreverse (@) myreverse $a, $b, $c
1005 sub myjoin ($@) myjoin ":", $a, $b, $c
1006 sub mypop (\@) mypop @array
1007 sub mysplice (\@$$@) mysplice @array, @array, 0, @pushme
1008 sub mykeys (\%) mykeys %{$hashref}
1009 sub myopen (*;$) myopen HANDLE, $name
1010 sub mypipe (**) mypipe READHANDLE, WRITEHANDLE
1011 sub mygrep (&@) mygrep { /foo/ } $a, $b, $c
1012 sub myrand ($) myrand 42
1013 sub mytime () mytime
1015 Any backslashed prototype character represents an actual argument
1016 that absolutely must start with that character. The value passed
1017 as part of C<@_> will be a reference to the actual argument given
1018 in the subroutine call, obtained by applying C<\> to that argument.
1020 You can also backslash several argument types simultaneously by using
1021 the C<\[]> notation:
1023 sub myref (\[$@%&*])
1025 will allow calling myref() as
1033 and the first argument of myref() will be a reference to
1034 a scalar, an array, a hash, a code, or a glob.
1036 Unbackslashed prototype characters have special meanings. Any
1037 unbackslashed C<@> or C<%> eats all remaining arguments, and forces
1038 list context. An argument represented by C<$> forces scalar context. An
1039 C<&> requires an anonymous subroutine, which, if passed as the first
1040 argument, does not require the C<sub> keyword or a subsequent comma.
1042 A C<*> allows the subroutine to accept a bareword, constant, scalar expression,
1043 typeglob, or a reference to a typeglob in that slot. The value will be
1044 available to the subroutine either as a simple scalar, or (in the latter
1045 two cases) as a reference to the typeglob. If you wish to always convert
1046 such arguments to a typeglob reference, use Symbol::qualify_to_ref() as
1049 use Symbol 'qualify_to_ref';
1052 my $fh = qualify_to_ref(shift, caller);
1056 A semicolon separates mandatory arguments from optional arguments.
1057 It is redundant before C<@> or C<%>, which gobble up everything else.
1059 Note how the last three examples in the table above are treated
1060 specially by the parser. C<mygrep()> is parsed as a true list
1061 operator, C<myrand()> is parsed as a true unary operator with unary
1062 precedence the same as C<rand()>, and C<mytime()> is truly without
1063 arguments, just like C<time()>. That is, if you say
1067 you'll get C<mytime() + 2>, not C<mytime(2)>, which is how it would be parsed
1068 without a prototype.
1070 The interesting thing about C<&> is that you can generate new syntax with it,
1071 provided it's in the initial position:
1075 my($try,$catch) = @_;
1082 sub catch (&) { $_[0] }
1087 /phooey/ and print "unphooey\n";
1090 That prints C<"unphooey">. (Yes, there are still unresolved
1091 issues having to do with visibility of C<@_>. I'm ignoring that
1092 question for the moment. (But note that if we make C<@_> lexically
1093 scoped, those anonymous subroutines can act like closures... (Gee,
1094 is this sounding a little Lispish? (Never mind.))))
1096 And here's a reimplementation of the Perl C<grep> operator:
1103 push(@result, $_) if &$code;
1108 Some folks would prefer full alphanumeric prototypes. Alphanumerics have
1109 been intentionally left out of prototypes for the express purpose of
1110 someday in the future adding named, formal parameters. The current
1111 mechanism's main goal is to let module writers provide better diagnostics
1112 for module users. Larry feels the notation quite understandable to Perl
1113 programmers, and that it will not intrude greatly upon the meat of the
1114 module, nor make it harder to read. The line noise is visually
1115 encapsulated into a small pill that's easy to swallow.
1117 If you try to use an alphanumeric sequence in a prototype you will
1118 generate an optional warning - "Illegal character in prototype...".
1119 Unfortunately earlier versions of Perl allowed the prototype to be
1120 used as long as its prefix was a valid prototype. The warning may be
1121 upgraded to a fatal error in a future version of Perl once the
1122 majority of offending code is fixed.
1124 It's probably best to prototype new functions, not retrofit prototyping
1125 into older ones. That's because you must be especially careful about
1126 silent impositions of differing list versus scalar contexts. For example,
1127 if you decide that a function should take just one parameter, like this:
1131 print "you gave me $n\n";
1134 and someone has been calling it with an array or expression
1140 Then you've just supplied an automatic C<scalar> in front of their
1141 argument, which can be more than a bit surprising. The old C<@foo>
1142 which used to hold one thing doesn't get passed in. Instead,
1143 C<func()> now gets passed in a C<1>; that is, the number of elements
1144 in C<@foo>. And the C<split> gets called in scalar context so it
1145 starts scribbling on your C<@_> parameter list. Ouch!
1147 This is all very powerful, of course, and should be used only in moderation
1148 to make the world a better place.
1150 =head2 Constant Functions
1153 Functions with a prototype of C<()> are potential candidates for
1154 inlining. If the result after optimization and constant folding
1155 is either a constant or a lexically-scoped scalar which has no other
1156 references, then it will be used in place of function calls made
1157 without C<&>. Calls made using C<&> are never inlined. (See
1158 F<constant.pm> for an easy way to declare most constants.)
1160 The following functions would all be inlined:
1162 sub pi () { 3.14159 } # Not exact, but close.
1163 sub PI () { 4 * atan2 1, 1 } # As good as it gets,
1164 # and it's inlined, too!
1168 sub FLAG_FOO () { 1 << 8 }
1169 sub FLAG_BAR () { 1 << 9 }
1170 sub FLAG_MASK () { FLAG_FOO | FLAG_BAR }
1172 sub OPT_BAZ () { not (0x1B58 & FLAG_MASK) }
1174 sub N () { int(OPT_BAZ) / 3 }
1176 sub FOO_SET () { 1 if FLAG_MASK & FLAG_FOO }
1178 Be aware that these will not be inlined; as they contain inner scopes,
1179 the constant folding doesn't reduce them to a single constant:
1181 sub foo_set () { if (FLAG_MASK & FLAG_FOO) { 1 } }
1192 If you redefine a subroutine that was eligible for inlining, you'll get
1193 a mandatory warning. (You can use this warning to tell whether or not a
1194 particular subroutine is considered constant.) The warning is
1195 considered severe enough not to be optional because previously compiled
1196 invocations of the function will still be using the old value of the
1197 function. If you need to be able to redefine the subroutine, you need to
1198 ensure that it isn't inlined, either by dropping the C<()> prototype
1199 (which changes calling semantics, so beware) or by thwarting the
1200 inlining mechanism in some other way, such as
1202 sub not_inlined () {
1206 =head2 Overriding Built-in Functions
1207 X<built-in> X<override> X<CORE> X<CORE::GLOBAL>
1209 Many built-in functions may be overridden, though this should be tried
1210 only occasionally and for good reason. Typically this might be
1211 done by a package attempting to emulate missing built-in functionality
1212 on a non-Unix system.
1214 Overriding may be done only by importing the name from a module at
1215 compile time--ordinary predeclaration isn't good enough. However, the
1216 C<use subs> pragma lets you, in effect, predeclare subs
1217 via the import syntax, and these names may then override built-in ones:
1219 use subs 'chdir', 'chroot', 'chmod', 'chown';
1223 To unambiguously refer to the built-in form, precede the
1224 built-in name with the special package qualifier C<CORE::>. For example,
1225 saying C<CORE::open()> always refers to the built-in C<open()>, even
1226 if the current package has imported some other subroutine called
1227 C<&open()> from elsewhere. Even though it looks like a regular
1228 function call, it isn't: you can't take a reference to it, such as
1229 the incorrect C<\&CORE::open> might appear to produce.
1231 Library modules should not in general export built-in names like C<open>
1232 or C<chdir> as part of their default C<@EXPORT> list, because these may
1233 sneak into someone else's namespace and change the semantics unexpectedly.
1234 Instead, if the module adds that name to C<@EXPORT_OK>, then it's
1235 possible for a user to import the name explicitly, but not implicitly.
1236 That is, they could say
1240 and it would import the C<open> override. But if they said
1244 they would get the default imports without overrides.
1246 The foregoing mechanism for overriding built-in is restricted, quite
1247 deliberately, to the package that requests the import. There is a second
1248 method that is sometimes applicable when you wish to override a built-in
1249 everywhere, without regard to namespace boundaries. This is achieved by
1250 importing a sub into the special namespace C<CORE::GLOBAL::>. Here is an
1251 example that quite brazenly replaces the C<glob> operator with something
1252 that understands regular expressions.
1257 @EXPORT_OK = 'glob';
1263 my $where = ($sym =~ s/^GLOBAL_// ? 'CORE::GLOBAL' : caller(0));
1264 $pkg->export($where, $sym, @_);
1271 if (opendir D, '.') {
1272 @got = grep /$pat/, readdir D;
1279 And here's how it could be (ab)used:
1281 #use REGlob 'GLOBAL_glob'; # override glob() in ALL namespaces
1283 use REGlob 'glob'; # override glob() in Foo:: only
1284 print for <^[a-z_]+\.pm\$>; # show all pragmatic modules
1286 The initial comment shows a contrived, even dangerous example.
1287 By overriding C<glob> globally, you would be forcing the new (and
1288 subversive) behavior for the C<glob> operator for I<every> namespace,
1289 without the complete cognizance or cooperation of the modules that own
1290 those namespaces. Naturally, this should be done with extreme caution--if
1291 it must be done at all.
1293 The C<REGlob> example above does not implement all the support needed to
1294 cleanly override perl's C<glob> operator. The built-in C<glob> has
1295 different behaviors depending on whether it appears in a scalar or list
1296 context, but our C<REGlob> doesn't. Indeed, many perl built-in have such
1297 context sensitive behaviors, and these must be adequately supported by
1298 a properly written override. For a fully functional example of overriding
1299 C<glob>, study the implementation of C<File::DosGlob> in the standard
1302 When you override a built-in, your replacement should be consistent (if
1303 possible) with the built-in native syntax. You can achieve this by using
1304 a suitable prototype. To get the prototype of an overridable built-in,
1305 use the C<prototype> function with an argument of C<"CORE::builtin_name">
1306 (see L<perlfunc/prototype>).
1308 Note however that some built-ins can't have their syntax expressed by a
1309 prototype (such as C<system> or C<chomp>). If you override them you won't
1310 be able to fully mimic their original syntax.
1312 The built-ins C<do>, C<require> and C<glob> can also be overridden, but due
1313 to special magic, their original syntax is preserved, and you don't have
1314 to define a prototype for their replacements. (You can't override the
1315 C<do BLOCK> syntax, though).
1317 C<require> has special additional dark magic: if you invoke your
1318 C<require> replacement as C<require Foo::Bar>, it will actually receive
1319 the argument C<"Foo/Bar.pm"> in @_. See L<perlfunc/require>.
1321 And, as you'll have noticed from the previous example, if you override
1322 C<glob>, the C<< <*> >> glob operator is overridden as well.
1324 In a similar fashion, overriding the C<readline> function also overrides
1325 the equivalent I/O operator C<< <FILEHANDLE> >>.
1327 Finally, some built-ins (e.g. C<exists> or C<grep>) can't be overridden.
1330 X<autoloading> X<AUTOLOAD>
1332 If you call a subroutine that is undefined, you would ordinarily
1333 get an immediate, fatal error complaining that the subroutine doesn't
1334 exist. (Likewise for subroutines being used as methods, when the
1335 method doesn't exist in any base class of the class's package.)
1336 However, if an C<AUTOLOAD> subroutine is defined in the package or
1337 packages used to locate the original subroutine, then that
1338 C<AUTOLOAD> subroutine is called with the arguments that would have
1339 been passed to the original subroutine. The fully qualified name
1340 of the original subroutine magically appears in the global $AUTOLOAD
1341 variable of the same package as the C<AUTOLOAD> routine. The name
1342 is not passed as an ordinary argument because, er, well, just
1343 because, that's why. (As an exception, a method call to a nonexistent
1344 C<import> or C<unimport> method is just skipped instead.)
1346 Many C<AUTOLOAD> routines load in a definition for the requested
1347 subroutine using eval(), then execute that subroutine using a special
1348 form of goto() that erases the stack frame of the C<AUTOLOAD> routine
1349 without a trace. (See the source to the standard module documented
1350 in L<AutoLoader>, for example.) But an C<AUTOLOAD> routine can
1351 also just emulate the routine and never define it. For example,
1352 let's pretend that a function that wasn't defined should just invoke
1353 C<system> with those arguments. All you'd do is:
1356 my $program = $AUTOLOAD;
1357 $program =~ s/.*:://;
1358 system($program, @_);
1364 In fact, if you predeclare functions you want to call that way, you don't
1365 even need parentheses:
1367 use subs qw(date who ls);
1372 A more complete example of this is the standard Shell module, which
1373 can treat undefined subroutine calls as calls to external programs.
1375 Mechanisms are available to help modules writers split their modules
1376 into autoloadable files. See the standard AutoLoader module
1377 described in L<AutoLoader> and in L<AutoSplit>, the standard
1378 SelfLoader modules in L<SelfLoader>, and the document on adding C
1379 functions to Perl code in L<perlxs>.
1381 =head2 Subroutine Attributes
1382 X<attribute> X<subroutine, attribute> X<attrs>
1384 A subroutine declaration or definition may have a list of attributes
1385 associated with it. If such an attribute list is present, it is
1386 broken up at space or colon boundaries and treated as though a
1387 C<use attributes> had been seen. See L<attributes> for details
1388 about what attributes are currently supported.
1389 Unlike the limitation with the obsolescent C<use attrs>, the
1390 C<sub : ATTRLIST> syntax works to associate the attributes with
1391 a pre-declaration, and not just with a subroutine definition.
1393 The attributes must be valid as simple identifier names (without any
1394 punctuation other than the '_' character). They may have a parameter
1395 list appended, which is only checked for whether its parentheses ('(',')')
1398 Examples of valid syntax (even though the attributes are unknown):
1400 sub fnord (&\%) : switch(10,foo(7,3)) : expensive ;
1401 sub plugh () : Ugly('\(") :Bad ;
1402 sub xyzzy : _5x5 { ... }
1404 Examples of invalid syntax:
1406 sub fnord : switch(10,foo() ; # ()-string not balanced
1407 sub snoid : Ugly('(') ; # ()-string not balanced
1408 sub xyzzy : 5x5 ; # "5x5" not a valid identifier
1409 sub plugh : Y2::north ; # "Y2::north" not a simple identifier
1410 sub snurt : foo + bar ; # "+" not a colon or space
1412 The attribute list is passed as a list of constant strings to the code
1413 which associates them with the subroutine. In particular, the second example
1414 of valid syntax above currently looks like this in terms of how it's
1417 use attributes __PACKAGE__, \&plugh, q[Ugly('\(")], 'Bad';
1419 For further details on attribute lists and their manipulation,
1420 see L<attributes> and L<Attribute::Handlers>.
1424 See L<perlref/"Function Templates"> for more about references and closures.
1425 See L<perlxs> if you'd like to learn about calling C subroutines from Perl.
1426 See L<perlembed> if you'd like to learn about calling Perl subroutines from C.
1427 See L<perlmod> to learn about bundling up your functions in separate files.
1428 See L<perlmodlib> to learn what library modules come standard on your system.
1429 See L<perltoot> to learn how to make object method calls.