3 perlsub - Perl subroutines
7 To declare subroutines:
9 sub NAME; # A "forward" declaration.
10 sub NAME(PROTO); # ditto, but with prototypes
11 sub NAME : ATTRS; # with attributes
12 sub NAME(PROTO) : ATTRS; # with attributes and prototypes
14 sub NAME BLOCK # A declaration and a definition.
15 sub NAME(PROTO) BLOCK # ditto, but with prototypes
16 sub NAME : ATTRS BLOCK # with attributes
17 sub NAME(PROTO) : ATTRS BLOCK # with prototypes and attributes
19 To define an anonymous subroutine at runtime:
21 $subref = sub BLOCK; # no proto
22 $subref = sub (PROTO) BLOCK; # with proto
23 $subref = sub : ATTRS BLOCK; # with attributes
24 $subref = sub (PROTO) : ATTRS BLOCK; # with proto and attributes
26 To import subroutines:
28 use MODULE qw(NAME1 NAME2 NAME3);
32 NAME(LIST); # & is optional with parentheses.
33 NAME LIST; # Parentheses optional if predeclared/imported.
34 &NAME(LIST); # Circumvent prototypes.
35 &NAME; # Makes current @_ visible to called subroutine.
39 Like many languages, Perl provides for user-defined subroutines.
40 These may be located anywhere in the main program, loaded in from
41 other files via the C<do>, C<require>, or C<use> keywords, or
42 generated on the fly using C<eval> or anonymous subroutines.
43 You can even call a function indirectly using a variable containing
44 its name or a CODE reference.
46 The Perl model for function call and return values is simple: all
47 functions are passed as parameters one single flat list of scalars, and
48 all functions likewise return to their caller one single flat list of
49 scalars. Any arrays or hashes in these call and return lists will
50 collapse, losing their identities--but you may always use
51 pass-by-reference instead to avoid this. Both call and return lists may
52 contain as many or as few scalar elements as you'd like. (Often a
53 function without an explicit return statement is called a subroutine, but
54 there's really no difference from Perl's perspective.)
56 Any arguments passed in show up in the array C<@_>. Therefore, if
57 you called a function with two arguments, those would be stored in
58 C<$_[0]> and C<$_[1]>. The array C<@_> is a local array, but its
59 elements are aliases for the actual scalar parameters. In particular,
60 if an element C<$_[0]> is updated, the corresponding argument is
61 updated (or an error occurs if it is not updatable). If an argument
62 is an array or hash element which did not exist when the function
63 was called, that element is created only when (and if) it is modified
64 or a reference to it is taken. (Some earlier versions of Perl
65 created the element whether or not the element was assigned to.)
66 Assigning to the whole array C<@_> removes that aliasing, and does
67 not update any arguments.
69 The return value of a subroutine is the value of the last expression
70 evaluated by that sub, or the empty list in the case of an empty sub.
71 More explicitly, a C<return> statement may be used to exit the
72 subroutine, optionally specifying the returned value, which will be
73 evaluated in the appropriate context (list, scalar, or void) depending
74 on the context of the subroutine call. If you specify no return value,
75 the subroutine returns an empty list in list context, the undefined
76 value in scalar context, or nothing in void context. If you return
77 one or more aggregates (arrays and hashes), these will be flattened
78 together into one large indistinguishable list.
80 Perl does not have named formal parameters. In practice all you
81 do is assign to a C<my()> list of these. Variables that aren't
82 declared to be private are global variables. For gory details
83 on creating private variables, see L<"Private Variables via my()">
84 and L<"Temporary Values via local()">. To create protected
85 environments for a set of functions in a separate package (and
86 probably a separate file), see L<perlmod/"Packages">.
93 $max = $foo if $max < $foo;
97 $bestday = max($mon,$tue,$wed,$thu,$fri);
101 # get a line, combining continuation lines
102 # that start with whitespace
105 $thisline = $lookahead; # global variables!
106 LINE: while (defined($lookahead = <STDIN>)) {
107 if ($lookahead =~ /^[ \t]/) {
108 $thisline .= $lookahead;
117 $lookahead = <STDIN>; # get first line
118 while (defined($line = get_line())) {
122 Assigning to a list of private variables to name your arguments:
125 my($key, $value) = @_;
126 $Foo{$key} = $value unless $Foo{$key};
129 Because the assignment copies the values, this also has the effect
130 of turning call-by-reference into call-by-value. Otherwise a
131 function is free to do in-place modifications of C<@_> and change
134 upcase_in($v1, $v2); # this changes $v1 and $v2
136 for (@_) { tr/a-z/A-Z/ }
139 You aren't allowed to modify constants in this way, of course. If an
140 argument were actually literal and you tried to change it, you'd take a
141 (presumably fatal) exception. For example, this won't work:
143 upcase_in("frederick");
145 It would be much safer if the C<upcase_in()> function
146 were written to return a copy of its parameters instead
147 of changing them in place:
149 ($v3, $v4) = upcase($v1, $v2); # this doesn't change $v1 and $v2
151 return unless defined wantarray; # void context, do nothing
153 for (@parms) { tr/a-z/A-Z/ }
154 return wantarray ? @parms : $parms[0];
157 Notice how this (unprototyped) function doesn't care whether it was
158 passed real scalars or arrays. Perl sees all arguments as one big,
159 long, flat parameter list in C<@_>. This is one area where
160 Perl's simple argument-passing style shines. The C<upcase()>
161 function would work perfectly well without changing the C<upcase()>
162 definition even if we fed it things like this:
164 @newlist = upcase(@list1, @list2);
165 @newlist = upcase( split /:/, $var );
167 Do not, however, be tempted to do this:
169 (@a, @b) = upcase(@list1, @list2);
171 Like the flattened incoming parameter list, the return list is also
172 flattened on return. So all you have managed to do here is stored
173 everything in C<@a> and made C<@b> empty. See
174 L<Pass by Reference> for alternatives.
176 A subroutine may be called using an explicit C<&> prefix. The
177 C<&> is optional in modern Perl, as are parentheses if the
178 subroutine has been predeclared. The C<&> is I<not> optional
179 when just naming the subroutine, such as when it's used as
180 an argument to defined() or undef(). Nor is it optional when you
181 want to do an indirect subroutine call with a subroutine name or
182 reference using the C<&$subref()> or C<&{$subref}()> constructs,
183 although the C<< $subref->() >> notation solves that problem.
184 See L<perlref> for more about all that.
186 Subroutines may be called recursively. If a subroutine is called
187 using the C<&> form, the argument list is optional, and if omitted,
188 no C<@_> array is set up for the subroutine: the C<@_> array at the
189 time of the call is visible to subroutine instead. This is an
190 efficiency mechanism that new users may wish to avoid.
192 &foo(1,2,3); # pass three arguments
193 foo(1,2,3); # the same
195 foo(); # pass a null list
198 &foo; # foo() get current args, like foo(@_) !!
199 foo; # like foo() IFF sub foo predeclared, else "foo"
201 Not only does the C<&> form make the argument list optional, it also
202 disables any prototype checking on arguments you do provide. This
203 is partly for historical reasons, and partly for having a convenient way
204 to cheat if you know what you're doing. See L<Prototypes> below.
206 Subroutines whose names are in all upper case are reserved to the Perl
207 core, as are modules whose names are in all lower case. A subroutine in
208 all capitals is a loosely-held convention meaning it will be called
209 indirectly by the run-time system itself, usually due to a triggered event.
210 Subroutines that do special, pre-defined things include C<AUTOLOAD>, C<CLONE>,
211 C<DESTROY> plus all functions mentioned in L<perltie> and L<PerlIO::via>.
213 The C<BEGIN>, C<CHECK>, C<INIT> and C<END> subroutines are not so much
214 subroutines as named special code blocks, of which you can have more
215 than one in a package, and which you can B<not> call explicitly. See
216 L<perlmod/"BEGIN, CHECK, INIT and END">
218 =head2 Private Variables via my()
222 my $foo; # declare $foo lexically local
223 my (@wid, %get); # declare list of variables local
224 my $foo = "flurp"; # declare $foo lexical, and init it
225 my @oof = @bar; # declare @oof lexical, and init it
226 my $x : Foo = $y; # similar, with an attribute applied
228 B<WARNING>: The use of attribute lists on C<my> declarations is still
229 evolving. The current semantics and interface are subject to change.
230 See L<attributes> and L<Attribute::Handlers>.
232 The C<my> operator declares the listed variables to be lexically
233 confined to the enclosing block, conditional (C<if/unless/elsif/else>),
234 loop (C<for/foreach/while/until/continue>), subroutine, C<eval>,
235 or C<do/require/use>'d file. If more than one value is listed, the
236 list must be placed in parentheses. All listed elements must be
237 legal lvalues. Only alphanumeric identifiers may be lexically
238 scoped--magical built-ins like C<$/> must currently be C<local>ized
239 with C<local> instead.
241 Unlike dynamic variables created by the C<local> operator, lexical
242 variables declared with C<my> are totally hidden from the outside
243 world, including any called subroutines. This is true if it's the
244 same subroutine called from itself or elsewhere--every call gets
247 This doesn't mean that a C<my> variable declared in a statically
248 enclosing lexical scope would be invisible. Only dynamic scopes
249 are cut off. For example, the C<bumpx()> function below has access
250 to the lexical $x variable because both the C<my> and the C<sub>
251 occurred at the same scope, presumably file scope.
256 An C<eval()>, however, can see lexical variables of the scope it is
257 being evaluated in, so long as the names aren't hidden by declarations within
258 the C<eval()> itself. See L<perlref>.
260 The parameter list to my() may be assigned to if desired, which allows you
261 to initialize your variables. (If no initializer is given for a
262 particular variable, it is created with the undefined value.) Commonly
263 this is used to name input parameters to a subroutine. Examples:
265 $arg = "fred"; # "global" variable
267 print "$arg thinks the root is $n\n";
268 fred thinks the root is 3
271 my $arg = shift; # name doesn't matter
276 The C<my> is simply a modifier on something you might assign to. So when
277 you do assign to variables in its argument list, C<my> doesn't
278 change whether those variables are viewed as a scalar or an array. So
280 my ($foo) = <STDIN>; # WRONG?
283 both supply a list context to the right-hand side, while
287 supplies a scalar context. But the following declares only one variable:
289 my $foo, $bar = 1; # WRONG
291 That has the same effect as
296 The declared variable is not introduced (is not visible) until after
297 the current statement. Thus,
301 can be used to initialize a new $x with the value of the old $x, and
304 my $x = 123 and $x == 123
306 is false unless the old $x happened to have the value C<123>.
308 Lexical scopes of control structures are not bounded precisely by the
309 braces that delimit their controlled blocks; control expressions are
310 part of that scope, too. Thus in the loop
312 while (my $line = <>) {
318 the scope of $line extends from its declaration throughout the rest of
319 the loop construct (including the C<continue> clause), but not beyond
320 it. Similarly, in the conditional
322 if ((my $answer = <STDIN>) =~ /^yes$/i) {
324 } elsif ($answer =~ /^no$/i) {
328 die "'$answer' is neither 'yes' nor 'no'";
331 the scope of $answer extends from its declaration through the rest
332 of that conditional, including any C<elsif> and C<else> clauses,
333 but not beyond it. See L<perlsyn/"Simple statements"> for information
334 on the scope of variables in statements with modifiers.
336 The C<foreach> loop defaults to scoping its index variable dynamically
337 in the manner of C<local>. However, if the index variable is
338 prefixed with the keyword C<my>, or if there is already a lexical
339 by that name in scope, then a new lexical is created instead. Thus
342 for my $i (1, 2, 3) {
346 the scope of $i extends to the end of the loop, but not beyond it,
347 rendering the value of $i inaccessible within C<some_function()>.
349 Some users may wish to encourage the use of lexically scoped variables.
350 As an aid to catching implicit uses to package variables,
351 which are always global, if you say
355 then any variable mentioned from there to the end of the enclosing
356 block must either refer to a lexical variable, be predeclared via
357 C<our> or C<use vars>, or else must be fully qualified with the package name.
358 A compilation error results otherwise. An inner block may countermand
359 this with C<no strict 'vars'>.
361 A C<my> has both a compile-time and a run-time effect. At compile
362 time, the compiler takes notice of it. The principal usefulness
363 of this is to quiet C<use strict 'vars'>, but it is also essential
364 for generation of closures as detailed in L<perlref>. Actual
365 initialization is delayed until run time, though, so it gets executed
366 at the appropriate time, such as each time through a loop, for
369 Variables declared with C<my> are not part of any package and are therefore
370 never fully qualified with the package name. In particular, you're not
371 allowed to try to make a package variable (or other global) lexical:
373 my $pack::var; # ERROR! Illegal syntax
374 my $_; # also illegal (currently)
376 In fact, a dynamic variable (also known as package or global variables)
377 are still accessible using the fully qualified C<::> notation even while a
378 lexical of the same name is also visible:
383 print "$x and $::x\n";
385 That will print out C<20> and C<10>.
387 You may declare C<my> variables at the outermost scope of a file
388 to hide any such identifiers from the world outside that file. This
389 is similar in spirit to C's static variables when they are used at
390 the file level. To do this with a subroutine requires the use of
391 a closure (an anonymous function that accesses enclosing lexicals).
392 If you want to create a private subroutine that cannot be called
393 from outside that block, it can declare a lexical variable containing
394 an anonymous sub reference:
396 my $secret_version = '1.001-beta';
397 my $secret_sub = sub { print $secret_version };
400 As long as the reference is never returned by any function within the
401 module, no outside module can see the subroutine, because its name is not in
402 any package's symbol table. Remember that it's not I<REALLY> called
403 C<$some_pack::secret_version> or anything; it's just $secret_version,
404 unqualified and unqualifiable.
406 This does not work with object methods, however; all object methods
407 have to be in the symbol table of some package to be found. See
408 L<perlref/"Function Templates"> for something of a work-around to
411 =head2 Persistent Private Variables
413 Just because a lexical variable is lexically (also called statically)
414 scoped to its enclosing block, C<eval>, or C<do> FILE, this doesn't mean that
415 within a function it works like a C static. It normally works more
416 like a C auto, but with implicit garbage collection.
418 Unlike local variables in C or C++, Perl's lexical variables don't
419 necessarily get recycled just because their scope has exited.
420 If something more permanent is still aware of the lexical, it will
421 stick around. So long as something else references a lexical, that
422 lexical won't be freed--which is as it should be. You wouldn't want
423 memory being free until you were done using it, or kept around once you
424 were done. Automatic garbage collection takes care of this for you.
426 This means that you can pass back or save away references to lexical
427 variables, whereas to return a pointer to a C auto is a grave error.
428 It also gives us a way to simulate C's function statics. Here's a
429 mechanism for giving a function private variables with both lexical
430 scoping and a static lifetime. If you do want to create something like
431 C's static variables, just enclose the whole function in an extra block,
432 and put the static variable outside the function but in the block.
437 return ++$secret_val;
440 # $secret_val now becomes unreachable by the outside
441 # world, but retains its value between calls to gimme_another
443 If this function is being sourced in from a separate file
444 via C<require> or C<use>, then this is probably just fine. If it's
445 all in the main program, you'll need to arrange for the C<my>
446 to be executed early, either by putting the whole block above
447 your main program, or more likely, placing merely a C<BEGIN>
448 code block around it to make sure it gets executed before your program
454 return ++$secret_val;
458 See L<perlmod/"BEGIN, CHECK, INIT and END"> about the
459 special triggered code blocks, C<BEGIN>, C<CHECK>, C<INIT> and C<END>.
461 If declared at the outermost scope (the file scope), then lexicals
462 work somewhat like C's file statics. They are available to all
463 functions in that same file declared below them, but are inaccessible
464 from outside that file. This strategy is sometimes used in modules
465 to create private variables that the whole module can see.
467 =head2 Temporary Values via local()
469 B<WARNING>: In general, you should be using C<my> instead of C<local>, because
470 it's faster and safer. Exceptions to this include the global punctuation
471 variables, global filehandles and formats, and direct manipulation of the
472 Perl symbol table itself. C<local> is mostly used when the current value
473 of a variable must be visible to called subroutines.
477 # localization of values
479 local $foo; # make $foo dynamically local
480 local (@wid, %get); # make list of variables local
481 local $foo = "flurp"; # make $foo dynamic, and init it
482 local @oof = @bar; # make @oof dynamic, and init it
484 local $hash{key} = "val"; # sets a local value for this hash entry
485 local ($cond ? $v1 : $v2); # several types of lvalues support
488 # localization of symbols
490 local *FH; # localize $FH, @FH, %FH, &FH ...
491 local *merlyn = *randal; # now $merlyn is really $randal, plus
492 # @merlyn is really @randal, etc
493 local *merlyn = 'randal'; # SAME THING: promote 'randal' to *randal
494 local *merlyn = \$randal; # just alias $merlyn, not @merlyn etc
496 A C<local> modifies its listed variables to be "local" to the
497 enclosing block, C<eval>, or C<do FILE>--and to I<any subroutine
498 called from within that block>. A C<local> just gives temporary
499 values to global (meaning package) variables. It does I<not> create
500 a local variable. This is known as dynamic scoping. Lexical scoping
501 is done with C<my>, which works more like C's auto declarations.
503 Some types of lvalues can be localized as well : hash and array elements
504 and slices, conditionals (provided that their result is always
505 localizable), and symbolic references. As for simple variables, this
506 creates new, dynamically scoped values.
508 If more than one variable or expression is given to C<local>, they must be
509 placed in parentheses. This operator works
510 by saving the current values of those variables in its argument list on a
511 hidden stack and restoring them upon exiting the block, subroutine, or
512 eval. This means that called subroutines can also reference the local
513 variable, but not the global one. The argument list may be assigned to if
514 desired, which allows you to initialize your local variables. (If no
515 initializer is given for a particular variable, it is created with an
518 Because C<local> is a run-time operator, it gets executed each time
519 through a loop. Consequently, it's more efficient to localize your
520 variables outside the loop.
522 =head3 Grammatical note on local()
524 A C<local> is simply a modifier on an lvalue expression. When you assign to
525 a C<local>ized variable, the C<local> doesn't change whether its list is viewed
526 as a scalar or an array. So
528 local($foo) = <STDIN>;
529 local @FOO = <STDIN>;
531 both supply a list context to the right-hand side, while
533 local $foo = <STDIN>;
535 supplies a scalar context.
537 =head3 Localization of special variables
539 If you localize a special variable, you'll be giving a new value to it,
540 but its magic won't go away. That means that all side-effects related
541 to this magic still work with the localized value.
543 This feature allows code like this to work :
545 # Read the whole contents of FILE in $slurp
546 { local $/ = undef; $slurp = <FILE>; }
548 Note, however, that this restricts localization of some values ; for
549 example, the following statement dies, as of perl 5.9.0, with an error
550 I<Modification of a read-only value attempted>, because the $1 variable is
551 magical and read-only :
555 Similarly, but in a way more difficult to spot, the following snippet will
558 sub f { local $_ = "foo"; print }
560 # now $_ is aliased to $1, thus is magic and readonly
564 See next section for an alternative to this situation.
566 B<WARNING>: Localization of tied arrays and hashes does not currently
568 This will be fixed in a future release of Perl; in the meantime, avoid
569 code that relies on any particular behaviour of localising tied arrays
570 or hashes (localising individual elements is still okay).
571 See L<perl58delta/"Localising Tied Arrays and Hashes Is Broken"> for more
574 =head3 Localization of globs
580 creates a whole new symbol table entry for the glob C<name> in the
581 current package. That means that all variables in its glob slot ($name,
582 @name, %name, &name, and the C<name> filehandle) are dynamically reset.
584 This implies, among other things, that any magic eventually carried by
585 those variables is locally lost. In other words, saying C<local */>
586 will not have any effect on the internal value of the input record
589 Notably, if you want to work with a brand new value of the default scalar
590 $_, and avoid the potential problem listed above about $_ previously
591 carrying a magic value, you should use C<local *_> instead of C<local $_>.
592 As of perl 5.9.1, you can also use the lexical form of C<$_> (declaring it
593 with C<my $_>), which avoids completely this problem.
595 =head3 Localization of elements of composite types
597 It's also worth taking a moment to explain what happens when you
598 C<local>ize a member of a composite type (i.e. an array or hash element).
599 In this case, the element is C<local>ized I<by name>. This means that
600 when the scope of the C<local()> ends, the saved value will be
601 restored to the hash element whose key was named in the C<local()>, or
602 the array element whose index was named in the C<local()>. If that
603 element was deleted while the C<local()> was in effect (e.g. by a
604 C<delete()> from a hash or a C<shift()> of an array), it will spring
605 back into existence, possibly extending an array and filling in the
606 skipped elements with C<undef>. For instance, if you say
608 %hash = ( 'This' => 'is', 'a' => 'test' );
612 local($hash{'a'}) = 'drill';
613 while (my $e = pop(@ary)) {
618 $hash{'only a'} = 'test';
622 print join(' ', map { "$_ $hash{$_}" } sort keys %hash),".\n";
623 print "The array has ",scalar(@ary)," elements: ",
624 join(', ', map { defined $_ ? $_ : 'undef' } @ary),"\n";
631 This is a test only a test.
632 The array has 6 elements: 0, 1, 2, undef, undef, 5
634 The behavior of local() on non-existent members of composite
635 types is subject to change in future.
637 =head2 Lvalue subroutines
639 B<WARNING>: Lvalue subroutines are still experimental and the
640 implementation may change in future versions of Perl.
642 It is possible to return a modifiable value from a subroutine.
643 To do this, you have to declare the subroutine to return an lvalue.
646 sub canmod : lvalue {
647 # return $val; this doesn't work, don't say "return"
654 canmod() = 5; # assigns to $val
657 The scalar/list context for the subroutine and for the right-hand
658 side of assignment is determined as if the subroutine call is replaced
659 by a scalar. For example, consider:
661 data(2,3) = get_data(3,4);
663 Both subroutines here are called in a scalar context, while in:
665 (data(2,3)) = get_data(3,4);
669 (data(2),data(3)) = get_data(3,4);
671 all the subroutines are called in a list context.
675 =item Lvalue subroutines are EXPERIMENTAL
677 They appear to be convenient, but there are several reasons to be
680 You can't use the return keyword, you must pass out the value before
681 falling out of subroutine scope. (see comment in example above). This
682 is usually not a problem, but it disallows an explicit return out of a
683 deeply nested loop, which is sometimes a nice way out.
685 They violate encapsulation. A normal mutator can check the supplied
686 argument before setting the attribute it is protecting, an lvalue
687 subroutine never gets that chance. Consider;
689 my $some_array_ref = []; # protected by mutators ??
691 sub set_arr { # normal mutator
693 die("expected array, you supplied ", ref $val)
694 unless ref $val eq 'ARRAY';
695 $some_array_ref = $val;
697 sub set_arr_lv : lvalue { # lvalue mutator
701 # set_arr_lv cannot stop this !
702 set_arr_lv() = { a => 1 };
706 =head2 Passing Symbol Table Entries (typeglobs)
708 B<WARNING>: The mechanism described in this section was originally
709 the only way to simulate pass-by-reference in older versions of
710 Perl. While it still works fine in modern versions, the new reference
711 mechanism is generally easier to work with. See below.
713 Sometimes you don't want to pass the value of an array to a subroutine
714 but rather the name of it, so that the subroutine can modify the global
715 copy of it rather than working with a local copy. In perl you can
716 refer to all objects of a particular name by prefixing the name
717 with a star: C<*foo>. This is often known as a "typeglob", because the
718 star on the front can be thought of as a wildcard match for all the
719 funny prefix characters on variables and subroutines and such.
721 When evaluated, the typeglob produces a scalar value that represents
722 all the objects of that name, including any filehandle, format, or
723 subroutine. When assigned to, it causes the name mentioned to refer to
724 whatever C<*> value was assigned to it. Example:
727 local(*someary) = @_;
728 foreach $elem (@someary) {
735 Scalars are already passed by reference, so you can modify
736 scalar arguments without using this mechanism by referring explicitly
737 to C<$_[0]> etc. You can modify all the elements of an array by passing
738 all the elements as scalars, but you have to use the C<*> mechanism (or
739 the equivalent reference mechanism) to C<push>, C<pop>, or change the size of
740 an array. It will certainly be faster to pass the typeglob (or reference).
742 Even if you don't want to modify an array, this mechanism is useful for
743 passing multiple arrays in a single LIST, because normally the LIST
744 mechanism will merge all the array values so that you can't extract out
745 the individual arrays. For more on typeglobs, see
746 L<perldata/"Typeglobs and Filehandles">.
748 =head2 When to Still Use local()
750 Despite the existence of C<my>, there are still three places where the
751 C<local> operator still shines. In fact, in these three places, you
752 I<must> use C<local> instead of C<my>.
758 You need to give a global variable a temporary value, especially $_.
760 The global variables, like C<@ARGV> or the punctuation variables, must be
761 C<local>ized with C<local()>. This block reads in F</etc/motd>, and splits
762 it up into chunks separated by lines of equal signs, which are placed
766 local @ARGV = ("/etc/motd");
769 @Fields = split /^\s*=+\s*$/;
772 It particular, it's important to C<local>ize $_ in any routine that assigns
773 to it. Look out for implicit assignments in C<while> conditionals.
777 You need to create a local file or directory handle or a local function.
779 A function that needs a filehandle of its own must use
780 C<local()> on a complete typeglob. This can be used to create new symbol
784 local (*READER, *WRITER); # not my!
785 pipe (READER, WRITER) or die "pipe: $!";
786 return (*READER, *WRITER);
788 ($head, $tail) = ioqueue();
790 See the Symbol module for a way to create anonymous symbol table
793 Because assignment of a reference to a typeglob creates an alias, this
794 can be used to create what is effectively a local function, or at least,
798 local *grow = \&shrink; # only until this block exists
799 grow(); # really calls shrink()
800 move(); # if move() grow()s, it shrink()s too
802 grow(); # get the real grow() again
804 See L<perlref/"Function Templates"> for more about manipulating
805 functions by name in this way.
809 You want to temporarily change just one element of an array or hash.
811 You can C<local>ize just one element of an aggregate. Usually this
815 local $SIG{INT} = 'IGNORE';
816 funct(); # uninterruptible
818 # interruptibility automatically restored here
820 But it also works on lexically declared aggregates. Prior to 5.005,
821 this operation could on occasion misbehave.
825 =head2 Pass by Reference
827 If you want to pass more than one array or hash into a function--or
828 return them from it--and have them maintain their integrity, then
829 you're going to have to use an explicit pass-by-reference. Before you
830 do that, you need to understand references as detailed in L<perlref>.
831 This section may not make much sense to you otherwise.
833 Here are a few simple examples. First, let's pass in several arrays
834 to a function and have it C<pop> all of then, returning a new list
835 of all their former last elements:
837 @tailings = popmany ( \@a, \@b, \@c, \@d );
842 foreach $aref ( @_ ) {
843 push @retlist, pop @$aref;
848 Here's how you might write a function that returns a
849 list of keys occurring in all the hashes passed to it:
851 @common = inter( \%foo, \%bar, \%joe );
853 my ($k, $href, %seen); # locals
855 while ( $k = each %$href ) {
859 return grep { $seen{$_} == @_ } keys %seen;
862 So far, we're using just the normal list return mechanism.
863 What happens if you want to pass or return a hash? Well,
864 if you're using only one of them, or you don't mind them
865 concatenating, then the normal calling convention is ok, although
868 Where people get into trouble is here:
870 (@a, @b) = func(@c, @d);
872 (%a, %b) = func(%c, %d);
874 That syntax simply won't work. It sets just C<@a> or C<%a> and
875 clears the C<@b> or C<%b>. Plus the function didn't get passed
876 into two separate arrays or hashes: it got one long list in C<@_>,
879 If you can arrange for everyone to deal with this through references, it's
880 cleaner code, although not so nice to look at. Here's a function that
881 takes two array references as arguments, returning the two array elements
882 in order of how many elements they have in them:
884 ($aref, $bref) = func(\@c, \@d);
885 print "@$aref has more than @$bref\n";
887 my ($cref, $dref) = @_;
888 if (@$cref > @$dref) {
889 return ($cref, $dref);
891 return ($dref, $cref);
895 It turns out that you can actually do this also:
897 (*a, *b) = func(\@c, \@d);
898 print "@a has more than @b\n";
908 Here we're using the typeglobs to do symbol table aliasing. It's
909 a tad subtle, though, and also won't work if you're using C<my>
910 variables, because only globals (even in disguise as C<local>s)
911 are in the symbol table.
913 If you're passing around filehandles, you could usually just use the bare
914 typeglob, like C<*STDOUT>, but typeglobs references work, too.
920 print $fh "her um well a hmmm\n";
923 $rec = get_rec(\*STDIN);
929 If you're planning on generating new filehandles, you could do this.
930 Notice to pass back just the bare *FH, not its reference.
935 return open (FH, $path) ? *FH : undef;
940 Perl supports a very limited kind of compile-time argument checking
941 using function prototyping. If you declare
945 then C<mypush()> takes arguments exactly like C<push()> does. The
946 function declaration must be visible at compile time. The prototype
947 affects only interpretation of new-style calls to the function,
948 where new-style is defined as not using the C<&> character. In
949 other words, if you call it like a built-in function, then it behaves
950 like a built-in function. If you call it like an old-fashioned
951 subroutine, then it behaves like an old-fashioned subroutine. It
952 naturally falls out from this rule that prototypes have no influence
953 on subroutine references like C<\&foo> or on indirect subroutine
954 calls like C<&{$subref}> or C<< $subref->() >>.
956 Method calls are not influenced by prototypes either, because the
957 function to be called is indeterminate at compile time, since
958 the exact code called depends on inheritance.
960 Because the intent of this feature is primarily to let you define
961 subroutines that work like built-in functions, here are prototypes
962 for some other functions that parse almost exactly like the
963 corresponding built-in.
965 Declared as Called as
967 sub mylink ($$) mylink $old, $new
968 sub myvec ($$$) myvec $var, $offset, 1
969 sub myindex ($$;$) myindex &getstring, "substr"
970 sub mysyswrite ($$$;$) mysyswrite $buf, 0, length($buf) - $off, $off
971 sub myreverse (@) myreverse $a, $b, $c
972 sub myjoin ($@) myjoin ":", $a, $b, $c
973 sub mypop (\@) mypop @array
974 sub mysplice (\@$$@) mysplice @array, @array, 0, @pushme
975 sub mykeys (\%) mykeys %{$hashref}
976 sub myopen (*;$) myopen HANDLE, $name
977 sub mypipe (**) mypipe READHANDLE, WRITEHANDLE
978 sub mygrep (&@) mygrep { /foo/ } $a, $b, $c
979 sub myrand ($) myrand 42
982 Any backslashed prototype character represents an actual argument
983 that absolutely must start with that character. The value passed
984 as part of C<@_> will be a reference to the actual argument given
985 in the subroutine call, obtained by applying C<\> to that argument.
987 You can also backslash several argument types simultaneously by using
992 will allow calling myref() as
1000 and the first argument of myref() will be a reference to
1001 a scalar, an array, a hash, a code, or a glob.
1003 Unbackslashed prototype characters have special meanings. Any
1004 unbackslashed C<@> or C<%> eats all remaining arguments, and forces
1005 list context. An argument represented by C<$> forces scalar context. An
1006 C<&> requires an anonymous subroutine, which, if passed as the first
1007 argument, does not require the C<sub> keyword or a subsequent comma.
1009 A C<*> allows the subroutine to accept a bareword, constant, scalar expression,
1010 typeglob, or a reference to a typeglob in that slot. The value will be
1011 available to the subroutine either as a simple scalar, or (in the latter
1012 two cases) as a reference to the typeglob. If you wish to always convert
1013 such arguments to a typeglob reference, use Symbol::qualify_to_ref() as
1016 use Symbol 'qualify_to_ref';
1019 my $fh = qualify_to_ref(shift, caller);
1023 A semicolon separates mandatory arguments from optional arguments.
1024 It is redundant before C<@> or C<%>, which gobble up everything else.
1026 Note how the last three examples in the table above are treated
1027 specially by the parser. C<mygrep()> is parsed as a true list
1028 operator, C<myrand()> is parsed as a true unary operator with unary
1029 precedence the same as C<rand()>, and C<mytime()> is truly without
1030 arguments, just like C<time()>. That is, if you say
1034 you'll get C<mytime() + 2>, not C<mytime(2)>, which is how it would be parsed
1035 without a prototype.
1037 The interesting thing about C<&> is that you can generate new syntax with it,
1038 provided it's in the initial position:
1041 my($try,$catch) = @_;
1048 sub catch (&) { $_[0] }
1053 /phooey/ and print "unphooey\n";
1056 That prints C<"unphooey">. (Yes, there are still unresolved
1057 issues having to do with visibility of C<@_>. I'm ignoring that
1058 question for the moment. (But note that if we make C<@_> lexically
1059 scoped, those anonymous subroutines can act like closures... (Gee,
1060 is this sounding a little Lispish? (Never mind.))))
1062 And here's a reimplementation of the Perl C<grep> operator:
1068 push(@result, $_) if &$code;
1073 Some folks would prefer full alphanumeric prototypes. Alphanumerics have
1074 been intentionally left out of prototypes for the express purpose of
1075 someday in the future adding named, formal parameters. The current
1076 mechanism's main goal is to let module writers provide better diagnostics
1077 for module users. Larry feels the notation quite understandable to Perl
1078 programmers, and that it will not intrude greatly upon the meat of the
1079 module, nor make it harder to read. The line noise is visually
1080 encapsulated into a small pill that's easy to swallow.
1082 If you try to use an alphanumeric sequence in a prototype you will
1083 generate an optional warning - "Illegal character in prototype...".
1084 Unfortunately earlier versions of Perl allowed the prototype to be
1085 used as long as its prefix was a valid prototype. The warning may be
1086 upgraded to a fatal error in a future version of Perl once the
1087 majority of offending code is fixed.
1089 It's probably best to prototype new functions, not retrofit prototyping
1090 into older ones. That's because you must be especially careful about
1091 silent impositions of differing list versus scalar contexts. For example,
1092 if you decide that a function should take just one parameter, like this:
1096 print "you gave me $n\n";
1099 and someone has been calling it with an array or expression
1105 Then you've just supplied an automatic C<scalar> in front of their
1106 argument, which can be more than a bit surprising. The old C<@foo>
1107 which used to hold one thing doesn't get passed in. Instead,
1108 C<func()> now gets passed in a C<1>; that is, the number of elements
1109 in C<@foo>. And the C<split> gets called in scalar context so it
1110 starts scribbling on your C<@_> parameter list. Ouch!
1112 This is all very powerful, of course, and should be used only in moderation
1113 to make the world a better place.
1115 =head2 Constant Functions
1117 Functions with a prototype of C<()> are potential candidates for
1118 inlining. If the result after optimization and constant folding
1119 is either a constant or a lexically-scoped scalar which has no other
1120 references, then it will be used in place of function calls made
1121 without C<&>. Calls made using C<&> are never inlined. (See
1122 F<constant.pm> for an easy way to declare most constants.)
1124 The following functions would all be inlined:
1126 sub pi () { 3.14159 } # Not exact, but close.
1127 sub PI () { 4 * atan2 1, 1 } # As good as it gets,
1128 # and it's inlined, too!
1132 sub FLAG_FOO () { 1 << 8 }
1133 sub FLAG_BAR () { 1 << 9 }
1134 sub FLAG_MASK () { FLAG_FOO | FLAG_BAR }
1136 sub OPT_BAZ () { not (0x1B58 & FLAG_MASK) }
1138 sub N () { int(OPT_BAZ) / 3 }
1140 sub FOO_SET () { 1 if FLAG_MASK & FLAG_FOO }
1142 Be aware that these will not be inlined; as they contain inner scopes,
1143 the constant folding doesn't reduce them to a single constant:
1145 sub foo_set () { if (FLAG_MASK & FLAG_FOO) { 1 } }
1156 If you redefine a subroutine that was eligible for inlining, you'll get
1157 a mandatory warning. (You can use this warning to tell whether or not a
1158 particular subroutine is considered constant.) The warning is
1159 considered severe enough not to be optional because previously compiled
1160 invocations of the function will still be using the old value of the
1161 function. If you need to be able to redefine the subroutine, you need to
1162 ensure that it isn't inlined, either by dropping the C<()> prototype
1163 (which changes calling semantics, so beware) or by thwarting the
1164 inlining mechanism in some other way, such as
1166 sub not_inlined () {
1170 =head2 Overriding Built-in Functions
1172 Many built-in functions may be overridden, though this should be tried
1173 only occasionally and for good reason. Typically this might be
1174 done by a package attempting to emulate missing built-in functionality
1175 on a non-Unix system.
1177 Overriding may be done only by importing the name from a module at
1178 compile time--ordinary predeclaration isn't good enough. However, the
1179 C<use subs> pragma lets you, in effect, predeclare subs
1180 via the import syntax, and these names may then override built-in ones:
1182 use subs 'chdir', 'chroot', 'chmod', 'chown';
1186 To unambiguously refer to the built-in form, precede the
1187 built-in name with the special package qualifier C<CORE::>. For example,
1188 saying C<CORE::open()> always refers to the built-in C<open()>, even
1189 if the current package has imported some other subroutine called
1190 C<&open()> from elsewhere. Even though it looks like a regular
1191 function call, it isn't: you can't take a reference to it, such as
1192 the incorrect C<\&CORE::open> might appear to produce.
1194 Library modules should not in general export built-in names like C<open>
1195 or C<chdir> as part of their default C<@EXPORT> list, because these may
1196 sneak into someone else's namespace and change the semantics unexpectedly.
1197 Instead, if the module adds that name to C<@EXPORT_OK>, then it's
1198 possible for a user to import the name explicitly, but not implicitly.
1199 That is, they could say
1203 and it would import the C<open> override. But if they said
1207 they would get the default imports without overrides.
1209 The foregoing mechanism for overriding built-in is restricted, quite
1210 deliberately, to the package that requests the import. There is a second
1211 method that is sometimes applicable when you wish to override a built-in
1212 everywhere, without regard to namespace boundaries. This is achieved by
1213 importing a sub into the special namespace C<CORE::GLOBAL::>. Here is an
1214 example that quite brazenly replaces the C<glob> operator with something
1215 that understands regular expressions.
1220 @EXPORT_OK = 'glob';
1226 my $where = ($sym =~ s/^GLOBAL_// ? 'CORE::GLOBAL' : caller(0));
1227 $pkg->export($where, $sym, @_);
1234 if (opendir D, '.') {
1235 @got = grep /$pat/, readdir D;
1242 And here's how it could be (ab)used:
1244 #use REGlob 'GLOBAL_glob'; # override glob() in ALL namespaces
1246 use REGlob 'glob'; # override glob() in Foo:: only
1247 print for <^[a-z_]+\.pm\$>; # show all pragmatic modules
1249 The initial comment shows a contrived, even dangerous example.
1250 By overriding C<glob> globally, you would be forcing the new (and
1251 subversive) behavior for the C<glob> operator for I<every> namespace,
1252 without the complete cognizance or cooperation of the modules that own
1253 those namespaces. Naturally, this should be done with extreme caution--if
1254 it must be done at all.
1256 The C<REGlob> example above does not implement all the support needed to
1257 cleanly override perl's C<glob> operator. The built-in C<glob> has
1258 different behaviors depending on whether it appears in a scalar or list
1259 context, but our C<REGlob> doesn't. Indeed, many perl built-in have such
1260 context sensitive behaviors, and these must be adequately supported by
1261 a properly written override. For a fully functional example of overriding
1262 C<glob>, study the implementation of C<File::DosGlob> in the standard
1265 When you override a built-in, your replacement should be consistent (if
1266 possible) with the built-in native syntax. You can achieve this by using
1267 a suitable prototype. To get the prototype of an overridable built-in,
1268 use the C<prototype> function with an argument of C<"CORE::builtin_name">
1269 (see L<perlfunc/prototype>).
1271 Note however that some built-ins can't have their syntax expressed by a
1272 prototype (such as C<system> or C<chomp>). If you override them you won't
1273 be able to fully mimic their original syntax.
1275 The built-ins C<do>, C<require> and C<glob> can also be overridden, but due
1276 to special magic, their original syntax is preserved, and you don't have
1277 to define a prototype for their replacements. (You can't override the
1278 C<do BLOCK> syntax, though).
1280 C<require> has special additional dark magic: if you invoke your
1281 C<require> replacement as C<require Foo::Bar>, it will actually receive
1282 the argument C<"Foo/Bar.pm"> in @_. See L<perlfunc/require>.
1284 And, as you'll have noticed from the previous example, if you override
1285 C<glob>, the C<< <*> >> glob operator is overridden as well.
1287 In a similar fashion, overriding the C<readline> function also overrides
1288 the equivalent I/O operator C<< <FILEHANDLE> >>.
1290 Finally, some built-ins (e.g. C<exists> or C<grep>) can't be overridden.
1294 If you call a subroutine that is undefined, you would ordinarily
1295 get an immediate, fatal error complaining that the subroutine doesn't
1296 exist. (Likewise for subroutines being used as methods, when the
1297 method doesn't exist in any base class of the class's package.)
1298 However, if an C<AUTOLOAD> subroutine is defined in the package or
1299 packages used to locate the original subroutine, then that
1300 C<AUTOLOAD> subroutine is called with the arguments that would have
1301 been passed to the original subroutine. The fully qualified name
1302 of the original subroutine magically appears in the global $AUTOLOAD
1303 variable of the same package as the C<AUTOLOAD> routine. The name
1304 is not passed as an ordinary argument because, er, well, just
1305 because, that's why. (As an exception, a method call to a nonexistent
1306 C<import> or C<unimport> method is just skipped instead.)
1308 Many C<AUTOLOAD> routines load in a definition for the requested
1309 subroutine using eval(), then execute that subroutine using a special
1310 form of goto() that erases the stack frame of the C<AUTOLOAD> routine
1311 without a trace. (See the source to the standard module documented
1312 in L<AutoLoader>, for example.) But an C<AUTOLOAD> routine can
1313 also just emulate the routine and never define it. For example,
1314 let's pretend that a function that wasn't defined should just invoke
1315 C<system> with those arguments. All you'd do is:
1318 my $program = $AUTOLOAD;
1319 $program =~ s/.*:://;
1320 system($program, @_);
1326 In fact, if you predeclare functions you want to call that way, you don't
1327 even need parentheses:
1329 use subs qw(date who ls);
1334 A more complete example of this is the standard Shell module, which
1335 can treat undefined subroutine calls as calls to external programs.
1337 Mechanisms are available to help modules writers split their modules
1338 into autoloadable files. See the standard AutoLoader module
1339 described in L<AutoLoader> and in L<AutoSplit>, the standard
1340 SelfLoader modules in L<SelfLoader>, and the document on adding C
1341 functions to Perl code in L<perlxs>.
1343 =head2 Subroutine Attributes
1345 A subroutine declaration or definition may have a list of attributes
1346 associated with it. If such an attribute list is present, it is
1347 broken up at space or colon boundaries and treated as though a
1348 C<use attributes> had been seen. See L<attributes> for details
1349 about what attributes are currently supported.
1350 Unlike the limitation with the obsolescent C<use attrs>, the
1351 C<sub : ATTRLIST> syntax works to associate the attributes with
1352 a pre-declaration, and not just with a subroutine definition.
1354 The attributes must be valid as simple identifier names (without any
1355 punctuation other than the '_' character). They may have a parameter
1356 list appended, which is only checked for whether its parentheses ('(',')')
1359 Examples of valid syntax (even though the attributes are unknown):
1361 sub fnord (&\%) : switch(10,foo(7,3)) : expensive ;
1362 sub plugh () : Ugly('\(") :Bad ;
1363 sub xyzzy : _5x5 { ... }
1365 Examples of invalid syntax:
1367 sub fnord : switch(10,foo() ; # ()-string not balanced
1368 sub snoid : Ugly('(') ; # ()-string not balanced
1369 sub xyzzy : 5x5 ; # "5x5" not a valid identifier
1370 sub plugh : Y2::north ; # "Y2::north" not a simple identifier
1371 sub snurt : foo + bar ; # "+" not a colon or space
1373 The attribute list is passed as a list of constant strings to the code
1374 which associates them with the subroutine. In particular, the second example
1375 of valid syntax above currently looks like this in terms of how it's
1378 use attributes __PACKAGE__, \&plugh, q[Ugly('\(")], 'Bad';
1380 For further details on attribute lists and their manipulation,
1381 see L<attributes> and L<Attribute::Handlers>.
1385 See L<perlref/"Function Templates"> for more about references and closures.
1386 See L<perlxs> if you'd like to learn about calling C subroutines from Perl.
1387 See L<perlembed> if you'd like to learn about calling Perl subroutines from C.
1388 See L<perlmod> to learn about bundling up your functions in separate files.
1389 See L<perlmodlib> to learn what library modules come standard on your system.
1390 See L<perltoot> to learn how to make object method calls.