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 $subref = sub (PROTO) : ATTRS BLOCK; # with proto and attributes
26 To import subroutines:
28 use MODULE qw(NAME1 NAME2 NAME3);
32 NAME(LIST); # & is optional with parentheses.
33 NAME LIST; # Parentheses optional if predeclared/imported.
34 &NAME(LIST); # Circumvent prototypes.
35 &NAME; # Makes current @_ visible to called subroutine.
39 Like many languages, Perl provides for user-defined subroutines.
40 These may be located anywhere in the main program, loaded in from
41 other files via the C<do>, C<require>, or C<use> keywords, or
42 generated on the fly using C<eval> or anonymous subroutines (closures).
43 You can even call a function indirectly using a variable containing
44 its name or a CODE reference.
46 The Perl model for function call and return values is simple: all
47 functions are passed as parameters one single flat list of scalars, and
48 all functions likewise return to their caller one single flat list of
49 scalars. Any arrays or hashes in these call and return lists will
50 collapse, losing their identities--but you may always use
51 pass-by-reference instead to avoid this. Both call and return lists may
52 contain as many or as few scalar elements as you'd like. (Often a
53 function without an explicit return statement is called a subroutine, but
54 there's really no difference from Perl's perspective.)
56 Any arguments passed in show up in the array C<@_>. Therefore, if
57 you called a function with two arguments, those would be stored in
58 C<$_[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. More explicitly, a C<return> statement may be used to exit the
71 subroutine, optionally specifying the returned value, which will be
72 evaluated in the appropriate context (list, scalar, or void) depending
73 on the context of the subroutine call. If you specify no return value,
74 the subroutine returns an empty list in list context, the undefined
75 value in scalar context, or nothing in void context. If you return
76 one or more aggregates (arrays and hashes), these will be flattened
77 together into one large indistinguishable list.
79 Perl does not have named formal parameters. In practice all you
80 do is assign to a C<my()> list of these. Variables that aren't
81 declared to be private are global variables. For gory details
82 on creating private variables, see L<"Private Variables via my()">
83 and L<"Temporary Values via local()">. To create protected
84 environments for a set of functions in a separate package (and
85 probably a separate file), see L<perlmod/"Packages">.
92 $max = $foo if $max < $foo;
96 $bestday = max($mon,$tue,$wed,$thu,$fri);
100 # get a line, combining continuation lines
101 # that start with whitespace
104 $thisline = $lookahead; # global variables!
105 LINE: while (defined($lookahead = <STDIN>)) {
106 if ($lookahead =~ /^[ \t]/) {
107 $thisline .= $lookahead;
116 $lookahead = <STDIN>; # get first line
117 while (defined($line = get_line())) {
121 Assigning to a list of private variables to name your arguments:
124 my($key, $value) = @_;
125 $Foo{$key} = $value unless $Foo{$key};
128 Because the assignment copies the values, this also has the effect
129 of turning call-by-reference into call-by-value. Otherwise a
130 function is free to do in-place modifications of C<@_> and change
133 upcase_in($v1, $v2); # this changes $v1 and $v2
135 for (@_) { tr/a-z/A-Z/ }
138 You aren't allowed to modify constants in this way, of course. If an
139 argument were actually literal and you tried to change it, you'd take a
140 (presumably fatal) exception. For example, this won't work:
142 upcase_in("frederick");
144 It would be much safer if the C<upcase_in()> function
145 were written to return a copy of its parameters instead
146 of changing them in place:
148 ($v3, $v4) = upcase($v1, $v2); # this doesn't change $v1 and $v2
150 return unless defined wantarray; # void context, do nothing
152 for (@parms) { tr/a-z/A-Z/ }
153 return wantarray ? @parms : $parms[0];
156 Notice how this (unprototyped) function doesn't care whether it was
157 passed real scalars or arrays. Perl sees all arugments as one big,
158 long, flat parameter list in C<@_>. This is one area where
159 Perl's simple argument-passing style shines. The C<upcase()>
160 function would work perfectly well without changing the C<upcase()>
161 definition even if we fed it things like this:
163 @newlist = upcase(@list1, @list2);
164 @newlist = upcase( split /:/, $var );
166 Do not, however, be tempted to do this:
168 (@a, @b) = upcase(@list1, @list2);
170 Like the flattened incoming parameter list, the return list is also
171 flattened on return. So all you have managed to do here is stored
172 everything in C<@a> and made C<@b> an empty list. See L<Pass by
173 Reference> for alternatives.
175 A subroutine may be called using an explicit C<&> prefix. The
176 C<&> is optional in modern Perl, as are parentheses if the
177 subroutine has been predeclared. The C<&> is I<not> optional
178 when just naming the subroutine, such as when it's used as
179 an argument to defined() or undef(). Nor is it optional when you
180 want to do an indirect subroutine call with a subroutine name or
181 reference using the C<&$subref()> or C<&{$subref}()> constructs,
182 although the C<$subref-E<gt>()> notation solves that problem.
183 See L<perlref> for more about all that.
185 Subroutines may be called recursively. If a subroutine is called
186 using the C<&> form, the argument list is optional, and if omitted,
187 no C<@_> array is set up for the subroutine: the C<@_> array at the
188 time of the call is visible to subroutine instead. This is an
189 efficiency mechanism that new users may wish to avoid.
191 &foo(1,2,3); # pass three arguments
192 foo(1,2,3); # the same
194 foo(); # pass a null list
197 &foo; # foo() get current args, like foo(@_) !!
198 foo; # like foo() IFF sub foo predeclared, else "foo"
200 Not only does the C<&> form make the argument list optional, it also
201 disables any prototype checking on arguments you do provide. This
202 is partly for historical reasons, and partly for having a convenient way
203 to cheat if you know what you're doing. See L<Prototypes> below.
205 Functions whose names are in all upper case are reserved to the Perl
206 core, as are modules whose names are in all lower case. A
207 function in all capitals is a loosely-held convention meaning it
208 will be called indirectly by the run-time system itself, usually
209 due to a triggered event. Functions that do special, pre-defined
210 things include C<BEGIN>, C<END>, C<AUTOLOAD>, and C<DESTROY>--plus
211 all functions mentioned in L<perltie>. The 5.005 release adds
212 C<INIT> to this list.
214 =head2 Private Variables via my()
218 my $foo; # declare $foo lexically local
219 my (@wid, %get); # declare list of variables local
220 my $foo = "flurp"; # declare $foo lexical, and init it
221 my @oof = @bar; # declare @oof lexical, and init it
222 my $x : Foo = $y; # similar, with an attribute applied
224 B<WARNING>: The use of attribute lists on C<my> declarations is
225 experimental. This feature should not be relied upon. It may
226 change or disappear in future releases of Perl. See L<attributes>.
228 The C<my> operator declares the listed variables to be lexically
229 confined to the enclosing block, conditional (C<if/unless/elsif/else>),
230 loop (C<for/foreach/while/until/continue>), subroutine, C<eval>,
231 or C<do/require/use>'d file. If more than one value is listed, the
232 list must be placed in parentheses. All listed elements must be
233 legal lvalues. Only alphanumeric identifiers may be lexically
234 scoped--magical built-ins like C<$/> must currently be C<local>ize
235 with C<local> instead.
237 Unlike dynamic variables created by the C<local> operator, lexical
238 variables declared with C<my> are totally hidden from the outside
239 world, including any called subroutines. This is true if it's the
240 same subroutine called from itself or elsewhere--every call gets
243 This doesn't mean that a C<my> variable declared in a statically
244 enclosing lexical scope would be invisible. Only dynamic scopes
245 are cut off. For example, the C<bumpx()> function below has access
246 to the lexical $x variable because both the C<my> and the C<sub>
247 occurred at the same scope, presumably file scope.
252 An C<eval()>, however, can see lexical variables of the scope it is
253 being evaluated in, so long as the names aren't hidden by declarations within
254 the C<eval()> itself. See L<perlref>.
256 The parameter list to my() may be assigned to if desired, which allows you
257 to initialize your variables. (If no initializer is given for a
258 particular variable, it is created with the undefined value.) Commonly
259 this is used to name input parameters to a subroutine. Examples:
261 $arg = "fred"; # "global" variable
263 print "$arg thinks the root is $n\n";
264 fred thinks the root is 3
267 my $arg = shift; # name doesn't matter
272 The C<my> is simply a modifier on something you might assign to. So when
273 you do assign to variables in its argument list, C<my> doesn't
274 change whether those variables are viewed as a scalar or an array. So
276 my ($foo) = <STDIN>; # WRONG?
279 both supply a list context to the right-hand side, while
283 supplies a scalar context. But the following declares only one variable:
285 my $foo, $bar = 1; # WRONG
287 That has the same effect as
292 The declared variable is not introduced (is not visible) until after
293 the current statement. Thus,
297 can be used to initialize a new $x with the value of the old $x, and
300 my $x = 123 and $x == 123
302 is false unless the old $x happened to have the value C<123>.
304 Lexical scopes of control structures are not bounded precisely by the
305 braces that delimit their controlled blocks; control expressions are
306 part of that scope, too. Thus in the loop
308 while (my $line = <>) {
314 the scope of $line extends from its declaration throughout the rest of
315 the loop construct (including the C<continue> clause), but not beyond
316 it. Similarly, in the conditional
318 if ((my $answer = <STDIN>) =~ /^yes$/i) {
320 } elsif ($answer =~ /^no$/i) {
324 die "'$answer' is neither 'yes' nor 'no'";
327 the scope of $answer extends from its declaration through the rest
328 of that conditional, including any C<elsif> and C<else> clauses,
331 None of the foregoing text applies to C<if/unless> or C<while/until>
332 modifiers appended to simple statements. Such modifiers are not
333 control structures and have no effect on scoping.
335 The C<foreach> loop defaults to scoping its index variable dynamically
336 in the manner of C<local>. However, if the index variable is
337 prefixed with the keyword C<my>, or if there is already a lexical
338 by that name in scope, then a new lexical is created instead. Thus
341 for my $i (1, 2, 3) {
345 the scope of $i extends to the end of the loop, but not beyond it,
346 rendering the value of $i inaccessible within C<some_function()>.
348 Some users may wish to encourage the use of lexically scoped variables.
349 As an aid to catching implicit uses to package variables,
350 which are always global, if you say
354 then any variable mentioned from there to the end of the enclosing
355 block must either refer to a lexical variable, be predeclared via
356 C<use vars>, or else must be fully qualified with the package name.
357 A compilation error results otherwise. An inner block may countermand
358 this with C<no strict 'vars'>.
360 A C<my> has both a compile-time and a run-time effect. At compile
361 time, the compiler takes notice of it. The principle usefulness
362 of this is to quiet C<use strict 'vars'>, but it is also essential
363 for generation of closures as detailed in L<perlref>. Actual
364 initialization is delayed until run time, though, so it gets executed
365 at the appropriate time, such as each time through a loop, for
368 Variables declared with C<my> are not part of any package and are therefore
369 never fully qualified with the package name. In particular, you're not
370 allowed to try to make a package variable (or other global) lexical:
372 my $pack::var; # ERROR! Illegal syntax
373 my $_; # also illegal (currently)
375 In fact, a dynamic variable (also known as package or global variables)
376 are still accessible using the fully qualified C<::> notation even while a
377 lexical of the same name is also visible:
382 print "$x and $::x\n";
384 That will print out C<20> and C<10>.
386 You may declare C<my> variables at the outermost scope of a file
387 to hide any such identifiers from the world outside that file. This
388 is similar in spirit to C's static variables when they are used at
389 the file level. To do this with a subroutine requires the use of
390 a closure (an anonymous function that accesses enclosing lexicals).
391 If you want to create a private subroutine that cannot be called
392 from outside that block, it can declare a lexical variable containing
393 an anonymous sub reference:
395 my $secret_version = '1.001-beta';
396 my $secret_sub = sub { print $secret_version };
399 As long as the reference is never returned by any function within the
400 module, no outside module can see the subroutine, because its name is not in
401 any package's symbol table. Remember that it's not I<REALLY> called
402 C<$some_pack::secret_version> or anything; it's just $secret_version,
403 unqualified and unqualifiable.
405 This does not work with object methods, however; all object methods
406 have to be in the symbol table of some package to be found. See
407 L<perlref/"Function Templates"> for something of a work-around to
410 =head2 Persistent Private Variables
412 Just because a lexical variable is lexically (also called statically)
413 scoped to its enclosing block, C<eval>, or C<do> FILE, this doesn't mean that
414 within a function it works like a C static. It normally works more
415 like a C auto, but with implicit garbage collection.
417 Unlike local variables in C or C++, Perl's lexical variables don't
418 necessarily get recycled just because their scope has exited.
419 If something more permanent is still aware of the lexical, it will
420 stick around. So long as something else references a lexical, that
421 lexical won't be freed--which is as it should be. You wouldn't want
422 memory being free until you were done using it, or kept around once you
423 were done. Automatic garbage collection takes care of this for you.
425 This means that you can pass back or save away references to lexical
426 variables, whereas to return a pointer to a C auto is a grave error.
427 It also gives us a way to simulate C's function statics. Here's a
428 mechanism for giving a function private variables with both lexical
429 scoping and a static lifetime. If you do want to create something like
430 C's static variables, just enclose the whole function in an extra block,
431 and put the static variable outside the function but in the block.
436 return ++$secret_val;
439 # $secret_val now becomes unreachable by the outside
440 # world, but retains its value between calls to gimme_another
442 If this function is being sourced in from a separate file
443 via C<require> or C<use>, then this is probably just fine. If it's
444 all in the main program, you'll need to arrange for the C<my>
445 to be executed early, either by putting the whole block above
446 your main program, or more likely, placing merely a C<BEGIN>
447 sub around it to make sure it gets executed before your program
453 return ++$secret_val;
457 See L<perlmod/"Package Constructors and Destructors"> about the
458 special triggered functions, C<BEGIN> and C<INIT>.
460 If declared at the outermost scope (the file scope), then lexicals
461 work somewhat like C's file statics. They are available to all
462 functions in that same file declared below them, but are inaccessible
463 from outside that file. This strategy is sometimes used in modules
464 to create private variables that the whole module can see.
466 =head2 Temporary Values via local()
468 B<WARNING>: In general, you should be using C<my> instead of C<local>, because
469 it's faster and safer. Exceptions to this include the global punctuation
470 variables, filehandles and formats, and direct manipulation of the Perl
471 symbol table itself. Format variables often use C<local> though, as do
472 other variables whose current value must be visible to called
477 local $foo; # declare $foo dynamically local
478 local (@wid, %get); # declare list of variables local
479 local $foo = "flurp"; # declare $foo dynamic, and init it
480 local @oof = @bar; # declare @oof dynamic, and init it
482 local *FH; # localize $FH, @FH, %FH, &FH ...
483 local *merlyn = *randal; # now $merlyn is really $randal, plus
484 # @merlyn is really @randal, etc
485 local *merlyn = 'randal'; # SAME THING: promote 'randal' to *randal
486 local *merlyn = \$randal; # just alias $merlyn, not @merlyn etc
488 A C<local> modifies its listed variables to be "local" to the
489 enclosing block, C<eval>, or C<do FILE>--and to I<any subroutine
490 called from within that block>. A C<local> just gives temporary
491 values to global (meaning package) variables. It does I<not> create
492 a local variable. This is known as dynamic scoping. Lexical scoping
493 is done with C<my>, which works more like C's auto declarations.
495 If more than one variable is given to C<local>, they must be placed in
496 parentheses. All listed elements must be legal lvalues. This operator works
497 by saving the current values of those variables in its argument list on a
498 hidden stack and restoring them upon exiting the block, subroutine, or
499 eval. This means that called subroutines can also reference the local
500 variable, but not the global one. The argument list may be assigned to if
501 desired, which allows you to initialize your local variables. (If no
502 initializer is given for a particular variable, it is created with an
503 undefined value.) Commonly this is used to name the parameters to a
504 subroutine. Examples:
509 # assume this function uses global %digits hash
512 # now temporarily add to %digits hash
514 # (NOTE: not claiming this is efficient!)
515 local %digits = (%digits, 't' => 10, 'e' => 11);
516 parse_num(); # parse_num gets this new %digits!
518 # old %digits restored here
520 Because C<local> is a run-time operator, it gets executed each time
521 through a loop. In releases of Perl previous to 5.0, this used more stack
522 storage each time until the loop was exited. Perl now reclaims the space
523 each time through, but it's still more efficient to declare your variables
526 A C<local> is simply a modifier on an lvalue expression. When you assign to
527 a C<local>ized variable, the C<local> doesn't change whether its list is viewed
528 as a scalar or an array. So
530 local($foo) = <STDIN>;
531 local @FOO = <STDIN>;
533 both supply a list context to the right-hand side, while
535 local $foo = <STDIN>;
537 supplies a scalar context.
539 A note about C<local()> and composite types is in order. Something
540 like C<local(%foo)> works by temporarily placing a brand new hash in
541 the symbol table. The old hash is left alone, but is hidden "behind"
544 This means the old variable is completely invisible via the symbol
545 table (i.e. the hash entry in the C<*foo> typeglob) for the duration
546 of the dynamic scope within which the C<local()> was seen. This
547 has the effect of allowing one to temporarily occlude any magic on
548 composite types. For instance, this will briefly alter a tied
549 hash to some other implementation:
551 tie %ahash, 'APackage';
555 tie %ahash, 'BPackage';
556 [..called code will see %ahash tied to 'BPackage'..]
559 [..%ahash is a normal (untied) hash here..]
562 [..%ahash back to its initial tied self again..]
564 As another example, a custom implementation of C<%ENV> might look
569 tie %ENV, 'MyOwnEnv';
570 [..do your own fancy %ENV manipulation here..]
572 [..normal %ENV behavior here..]
574 It's also worth taking a moment to explain what happens when you
575 C<local>ize a member of a composite type (i.e. an array or hash element).
576 In this case, the element is C<local>ized I<by name>. This means that
577 when the scope of the C<local()> ends, the saved value will be
578 restored to the hash element whose key was named in the C<local()>, or
579 the array element whose index was named in the C<local()>. If that
580 element was deleted while the C<local()> was in effect (e.g. by a
581 C<delete()> from a hash or a C<shift()> of an array), it will spring
582 back into existence, possibly extending an array and filling in the
583 skipped elements with C<undef>. For instance, if you say
585 %hash = ( 'This' => 'is', 'a' => 'test' );
589 local($hash{'a'}) = 'drill';
590 while (my $e = pop(@ary)) {
595 $hash{'only a'} = 'test';
599 print join(' ', map { "$_ $hash{$_}" } sort keys %hash),".\n";
600 print "The array has ",scalar(@ary)," elements: ",
601 join(', ', map { defined $_ ? $_ : 'undef' } @ary),"\n";
608 This is a test only a test.
609 The array has 6 elements: 0, 1, 2, undef, undef, 5
611 The behavior of local() on non-existent members of composite
612 types is subject to change in future.
614 =head2 Lvalue subroutines
616 B<WARNING>: Lvalue subroutines are still experimental and the implementation
617 may change in future versions of Perl.
619 It is possible to return a modifiable value from a subroutine.
620 To do this, you have to declare the subroutine to return an lvalue.
623 sub canmod : lvalue {
630 canmod() = 5; # assigns to $val
633 The scalar/list context for the subroutine and for the right-hand
634 side of assignment is determined as if the subroutine call is replaced
635 by a scalar. For example, consider:
637 data(2,3) = get_data(3,4);
639 Both subroutines here are called in a scalar context, while in:
641 (data(2,3)) = get_data(3,4);
645 (data(2),data(3)) = get_data(3,4);
647 all the subroutines are called in a list context.
649 The current implementation does not allow arrays and hashes to be
650 returned from lvalue subroutines directly. You may return a
651 reference instead. This restriction may be lifted in future.
653 =head2 Passing Symbol Table Entries (typeglobs)
655 B<WARNING>: The mechanism described in this section was originally
656 the only way to simulate pass-by-reference in older versions of
657 Perl. While it still works fine in modern versions, the new reference
658 mechanism is generally easier to work with. See below.
660 Sometimes you don't want to pass the value of an array to a subroutine
661 but rather the name of it, so that the subroutine can modify the global
662 copy of it rather than working with a local copy. In perl you can
663 refer to all objects of a particular name by prefixing the name
664 with a star: C<*foo>. This is often known as a "typeglob", because the
665 star on the front can be thought of as a wildcard match for all the
666 funny prefix characters on variables and subroutines and such.
668 When evaluated, the typeglob produces a scalar value that represents
669 all the objects of that name, including any filehandle, format, or
670 subroutine. When assigned to, it causes the name mentioned to refer to
671 whatever C<*> value was assigned to it. Example:
674 local(*someary) = @_;
675 foreach $elem (@someary) {
682 Scalars are already passed by reference, so you can modify
683 scalar arguments without using this mechanism by referring explicitly
684 to C<$_[0]> etc. You can modify all the elements of an array by passing
685 all the elements as scalars, but you have to use the C<*> mechanism (or
686 the equivalent reference mechanism) to C<push>, C<pop>, or change the size of
687 an array. It will certainly be faster to pass the typeglob (or reference).
689 Even if you don't want to modify an array, this mechanism is useful for
690 passing multiple arrays in a single LIST, because normally the LIST
691 mechanism will merge all the array values so that you can't extract out
692 the individual arrays. For more on typeglobs, see
693 L<perldata/"Typeglobs and Filehandles">.
695 =head2 When to Still Use local()
697 Despite the existence of C<my>, there are still three places where the
698 C<local> operator still shines. In fact, in these three places, you
699 I<must> use C<local> instead of C<my>.
703 =item 1. You need to give a global variable a temporary value, especially $_.
705 The global variables, like C<@ARGV> or the punctuation variables, must be
706 C<local>ized with C<local()>. This block reads in F</etc/motd>, and splits
707 it up into chunks separated by lines of equal signs, which are placed
711 local @ARGV = ("/etc/motd");
714 @Fields = split /^\s*=+\s*$/;
717 It particular, it's important to C<local>ize $_ in any routine that assigns
718 to it. Look out for implicit assignments in C<while> conditionals.
720 =item 2. You need to create a local file or directory handle or a local function.
722 A function that needs a filehandle of its own must use
723 C<local()> on a complete typeglob. This can be used to create new symbol
727 local (*READER, *WRITER); # not my!
728 pipe (READER, WRITER); or die "pipe: $!";
729 return (*READER, *WRITER);
731 ($head, $tail) = ioqueue();
733 See the Symbol module for a way to create anonymous symbol table
736 Because assignment of a reference to a typeglob creates an alias, this
737 can be used to create what is effectively a local function, or at least,
741 local *grow = \&shrink; # only until this block exists
742 grow(); # really calls shrink()
743 move(); # if move() grow()s, it shrink()s too
745 grow(); # get the real grow() again
747 See L<perlref/"Function Templates"> for more about manipulating
748 functions by name in this way.
750 =item 3. You want to temporarily change just one element of an array or hash.
752 You can C<local>ize just one element of an aggregate. Usually this
756 local $SIG{INT} = 'IGNORE';
757 funct(); # uninterruptible
759 # interruptibility automatically restored here
761 But it also works on lexically declared aggregates. Prior to 5.005,
762 this operation could on occasion misbehave.
766 =head2 Pass by Reference
768 If you want to pass more than one array or hash into a function--or
769 return them from it--and have them maintain their integrity, then
770 you're going to have to use an explicit pass-by-reference. Before you
771 do that, you need to understand references as detailed in L<perlref>.
772 This section may not make much sense to you otherwise.
774 Here are a few simple examples. First, let's pass in several arrays
775 to a function and have it C<pop> all of then, returning a new list
776 of all their former last elements:
778 @tailings = popmany ( \@a, \@b, \@c, \@d );
783 foreach $aref ( @_ ) {
784 push @retlist, pop @$aref;
789 Here's how you might write a function that returns a
790 list of keys occurring in all the hashes passed to it:
792 @common = inter( \%foo, \%bar, \%joe );
794 my ($k, $href, %seen); # locals
796 while ( $k = each %$href ) {
800 return grep { $seen{$_} == @_ } keys %seen;
803 So far, we're using just the normal list return mechanism.
804 What happens if you want to pass or return a hash? Well,
805 if you're using only one of them, or you don't mind them
806 concatenating, then the normal calling convention is ok, although
809 Where people get into trouble is here:
811 (@a, @b) = func(@c, @d);
813 (%a, %b) = func(%c, %d);
815 That syntax simply won't work. It sets just C<@a> or C<%a> and
816 clears the C<@b> or C<%b>. Plus the function didn't get passed
817 into two separate arrays or hashes: it got one long list in C<@_>,
820 If you can arrange for everyone to deal with this through references, it's
821 cleaner code, although not so nice to look at. Here's a function that
822 takes two array references as arguments, returning the two array elements
823 in order of how many elements they have in them:
825 ($aref, $bref) = func(\@c, \@d);
826 print "@$aref has more than @$bref\n";
828 my ($cref, $dref) = @_;
829 if (@$cref > @$dref) {
830 return ($cref, $dref);
832 return ($dref, $cref);
836 It turns out that you can actually do this also:
838 (*a, *b) = func(\@c, \@d);
839 print "@a has more than @b\n";
849 Here we're using the typeglobs to do symbol table aliasing. It's
850 a tad subtle, though, and also won't work if you're using C<my>
851 variables, because only globals (even in disguise as C<local>s)
852 are in the symbol table.
854 If you're passing around filehandles, you could usually just use the bare
855 typeglob, like C<*STDOUT>, but typeglobs references work, too.
861 print $fh "her um well a hmmm\n";
864 $rec = get_rec(\*STDIN);
870 If you're planning on generating new filehandles, you could do this.
871 Notice to pass back just the bare *FH, not its reference.
876 return open (FH, $path) ? *FH : undef;
881 Perl supports a very limited kind of compile-time argument checking
882 using function prototyping. If you declare
886 then C<mypush()> takes arguments exactly like C<push()> does. The
887 function declaration must be visible at compile time. The prototype
888 affects only interpretation of new-style calls to the function,
889 where new-style is defined as not using the C<&> character. In
890 other words, if you call it like a built-in function, then it behaves
891 like a built-in function. If you call it like an old-fashioned
892 subroutine, then it behaves like an old-fashioned subroutine. It
893 naturally falls out from this rule that prototypes have no influence
894 on subroutine references like C<\&foo> or on indirect subroutine
895 calls like C<&{$subref}> or C<$subref-E<gt>()>.
897 Method calls are not influenced by prototypes either, because the
898 function to be called is indeterminate at compile time, since
899 the exact code called depends on inheritance.
901 Because the intent of this feature is primarily to let you define
902 subroutines that work like built-in functions, here are prototypes
903 for some other functions that parse almost exactly like the
904 corresponding built-in.
906 Declared as Called as
908 sub mylink ($$) mylink $old, $new
909 sub myvec ($$$) myvec $var, $offset, 1
910 sub myindex ($$;$) myindex &getstring, "substr"
911 sub mysyswrite ($$$;$) mysyswrite $buf, 0, length($buf) - $off, $off
912 sub myreverse (@) myreverse $a, $b, $c
913 sub myjoin ($@) myjoin ":", $a, $b, $c
914 sub mypop (\@) mypop @array
915 sub mysplice (\@$$@) mysplice @array, @array, 0, @pushme
916 sub mykeys (\%) mykeys %{$hashref}
917 sub myopen (*;$) myopen HANDLE, $name
918 sub mypipe (**) mypipe READHANDLE, WRITEHANDLE
919 sub mygrep (&@) mygrep { /foo/ } $a, $b, $c
920 sub myrand ($) myrand 42
923 Any backslashed prototype character represents an actual argument
924 that absolutely must start with that character. The value passed
925 as part of C<@_> will be a reference to the actual argument given
926 in the subroutine call, obtained by applying C<\> to that argument.
928 Unbackslashed prototype characters have special meanings. Any
929 unbackslashed C<@> or C<%> eats all remaining arguments, and forces
930 list context. An argument represented by C<$> forces scalar context. An
931 C<&> requires an anonymous subroutine, which, if passed as the first
932 argument, does not require the C<sub> keyword or a subsequent comma. A
933 C<*> allows the subroutine to accept a bareword, constant, scalar expression,
934 typeglob, or a reference to a typeglob in that slot. The value will be
935 available to the subroutine either as a simple scalar, or (in the latter
936 two cases) as a reference to the typeglob.
938 A semicolon separates mandatory arguments from optional arguments.
939 It is redundant before C<@> or C<%>, which gobble up everything else.
941 Note how the last three examples in the table above are treated
942 specially by the parser. C<mygrep()> is parsed as a true list
943 operator, C<myrand()> is parsed as a true unary operator with unary
944 precedence the same as C<rand()>, and C<mytime()> is truly without
945 arguments, just like C<time()>. That is, if you say
949 you'll get C<mytime() + 2>, not C<mytime(2)>, which is how it would be parsed
952 The interesting thing about C<&> is that you can generate new syntax with it,
953 provided it's in the initial position:
956 my($try,$catch) = @_;
963 sub catch (&) { $_[0] }
968 /phooey/ and print "unphooey\n";
971 That prints C<"unphooey">. (Yes, there are still unresolved
972 issues having to do with visibility of C<@_>. I'm ignoring that
973 question for the moment. (But note that if we make C<@_> lexically
974 scoped, those anonymous subroutines can act like closures... (Gee,
975 is this sounding a little Lispish? (Never mind.))))
977 And here's a reimplementation of the Perl C<grep> operator:
983 push(@result, $_) if &$code;
988 Some folks would prefer full alphanumeric prototypes. Alphanumerics have
989 been intentionally left out of prototypes for the express purpose of
990 someday in the future adding named, formal parameters. The current
991 mechanism's main goal is to let module writers provide better diagnostics
992 for module users. Larry feels the notation quite understandable to Perl
993 programmers, and that it will not intrude greatly upon the meat of the
994 module, nor make it harder to read. The line noise is visually
995 encapsulated into a small pill that's easy to swallow.
997 It's probably best to prototype new functions, not retrofit prototyping
998 into older ones. That's because you must be especially careful about
999 silent impositions of differing list versus scalar contexts. For example,
1000 if you decide that a function should take just one parameter, like this:
1004 print "you gave me $n\n";
1007 and someone has been calling it with an array or expression
1013 Then you've just supplied an automatic C<scalar> in front of their
1014 argument, which can be more than a bit surprising. The old C<@foo>
1015 which used to hold one thing doesn't get passed in. Instead,
1016 C<func()> now gets passed in a C<1>; that is, the number of elements
1017 in C<@foo>. And the C<split> gets called in scalar context so it
1018 starts scribbling on your C<@_> parameter list. Ouch!
1020 This is all very powerful, of course, and should be used only in moderation
1021 to make the world a better place.
1023 =head2 Constant Functions
1025 Functions with a prototype of C<()> are potential candidates for
1026 inlining. If the result after optimization and constant folding
1027 is either a constant or a lexically-scoped scalar which has no other
1028 references, then it will be used in place of function calls made
1029 without C<&>. Calls made using C<&> are never inlined. (See
1030 F<constant.pm> for an easy way to declare most constants.)
1032 The following functions would all be inlined:
1034 sub pi () { 3.14159 } # Not exact, but close.
1035 sub PI () { 4 * atan2 1, 1 } # As good as it gets,
1036 # and it's inlined, too!
1040 sub FLAG_FOO () { 1 << 8 }
1041 sub FLAG_BAR () { 1 << 9 }
1042 sub FLAG_MASK () { FLAG_FOO | FLAG_BAR }
1044 sub OPT_BAZ () { not (0x1B58 & FLAG_MASK) }
1054 sub N () { int(BAZ_VAL) / 3 }
1057 for (1..N) { $prod *= $_ }
1058 sub N_FACTORIAL () { $prod }
1061 If you redefine a subroutine that was eligible for inlining, you'll get
1062 a mandatory warning. (You can use this warning to tell whether or not a
1063 particular subroutine is considered constant.) The warning is
1064 considered severe enough not to be optional because previously compiled
1065 invocations of the function will still be using the old value of the
1066 function. If you need to be able to redefine the subroutine, you need to
1067 ensure that it isn't inlined, either by dropping the C<()> prototype
1068 (which changes calling semantics, so beware) or by thwarting the
1069 inlining mechanism in some other way, such as
1071 sub not_inlined () {
1075 =head2 Overriding Built-in Functions
1077 Many built-in functions may be overridden, though this should be tried
1078 only occasionally and for good reason. Typically this might be
1079 done by a package attempting to emulate missing built-in functionality
1080 on a non-Unix system.
1082 Overriding may be done only by importing the name from a
1083 module--ordinary predeclaration isn't good enough. However, the
1084 C<use subs> pragma lets you, in effect, predeclare subs
1085 via the import syntax, and these names may then override built-in ones:
1087 use subs 'chdir', 'chroot', 'chmod', 'chown';
1091 To unambiguously refer to the built-in form, precede the
1092 built-in name with the special package qualifier C<CORE::>. For example,
1093 saying C<CORE::open()> always refers to the built-in C<open()>, even
1094 if the current package has imported some other subroutine called
1095 C<&open()> from elsewhere. Even though it looks like a regular
1096 function call, it isn't: you can't take a reference to it, such as
1097 the incorrect C<\&CORE::open> might appear to produce.
1099 Library modules should not in general export built-in names like C<open>
1100 or C<chdir> as part of their default C<@EXPORT> list, because these may
1101 sneak into someone else's namespace and change the semantics unexpectedly.
1102 Instead, if the module adds that name to C<@EXPORT_OK>, then it's
1103 possible for a user to import the name explicitly, but not implicitly.
1104 That is, they could say
1108 and it would import the C<open> override. But if they said
1112 they would get the default imports without overrides.
1114 The foregoing mechanism for overriding built-in is restricted, quite
1115 deliberately, to the package that requests the import. There is a second
1116 method that is sometimes applicable when you wish to override a built-in
1117 everywhere, without regard to namespace boundaries. This is achieved by
1118 importing a sub into the special namespace C<CORE::GLOBAL::>. Here is an
1119 example that quite brazenly replaces the C<glob> operator with something
1120 that understands regular expressions.
1125 @EXPORT_OK = 'glob';
1131 my $where = ($sym =~ s/^GLOBAL_// ? 'CORE::GLOBAL' : caller(0));
1132 $pkg->export($where, $sym, @_);
1139 if (opendir D, '.') {
1140 @got = grep /$pat/, readdir D;
1147 And here's how it could be (ab)used:
1149 #use REGlob 'GLOBAL_glob'; # override glob() in ALL namespaces
1151 use REGlob 'glob'; # override glob() in Foo:: only
1152 print for <^[a-z_]+\.pm\$>; # show all pragmatic modules
1154 The initial comment shows a contrived, even dangerous example.
1155 By overriding C<glob> globally, you would be forcing the new (and
1156 subversive) behavior for the C<glob> operator for I<every> namespace,
1157 without the complete cognizance or cooperation of the modules that own
1158 those namespaces. Naturally, this should be done with extreme caution--if
1159 it must be done at all.
1161 The C<REGlob> example above does not implement all the support needed to
1162 cleanly override perl's C<glob> operator. The built-in C<glob> has
1163 different behaviors depending on whether it appears in a scalar or list
1164 context, but our C<REGlob> doesn't. Indeed, many perl built-in have such
1165 context sensitive behaviors, and these must be adequately supported by
1166 a properly written override. For a fully functional example of overriding
1167 C<glob>, study the implementation of C<File::DosGlob> in the standard
1172 If you call a subroutine that is undefined, you would ordinarily
1173 get an immediate, fatal error complaining that the subroutine doesn't
1174 exist. (Likewise for subroutines being used as methods, when the
1175 method doesn't exist in any base class of the class's package.)
1176 However, if an C<AUTOLOAD> subroutine is defined in the package or
1177 packages used to locate the original subroutine, then that
1178 C<AUTOLOAD> subroutine is called with the arguments that would have
1179 been passed to the original subroutine. The fully qualified name
1180 of the original subroutine magically appears in the global $AUTOLOAD
1181 variable of the same package as the C<AUTOLOAD> routine. The name
1182 is not passed as an ordinary argument because, er, well, just
1183 because, that's why...
1185 Many C<AUTOLOAD> routines load in a definition for the requested
1186 subroutine using eval(), then execute that subroutine using a special
1187 form of goto() that erases the stack frame of the C<AUTOLOAD> routine
1188 without a trace. (See the source to the standard module documented
1189 in L<AutoLoader>, for example.) But an C<AUTOLOAD> routine can
1190 also just emulate the routine and never define it. For example,
1191 let's pretend that a function that wasn't defined should just invoke
1192 C<system> with those arguments. All you'd do is:
1195 my $program = $AUTOLOAD;
1196 $program =~ s/.*:://;
1197 system($program, @_);
1203 In fact, if you predeclare functions you want to call that way, you don't
1204 even need parentheses:
1206 use subs qw(date who ls);
1211 A more complete example of this is the standard Shell module, which
1212 can treat undefined subroutine calls as calls to external programs.
1214 Mechanisms are available to help modules writers split their modules
1215 into autoloadable files. See the standard AutoLoader module
1216 described in L<AutoLoader> and in L<AutoSplit>, the standard
1217 SelfLoader modules in L<SelfLoader>, and the document on adding C
1218 functions to Perl code in L<perlxs>.
1220 =head2 Subroutine Attributes
1222 A subroutine declaration or definition may have a list of attributes
1223 associated with it. If such an attribute list is present, it is
1224 broken up at space or comma boundaries and treated as though a
1225 C<use attributes> had been seen. See L<attributes> for details
1226 about what attributes are currently supported.
1227 Unlike the limitation with the obsolescent C<use attrs>, the
1228 C<sub : ATTRLIST> syntax works to associate the attributes with
1229 a pre-declaration, and not just with a subroutine definition.
1231 The attributes must be valid as simple identifier names (without any
1232 punctuation other than the '_' character). They may have a parameter
1233 list appended, which is only checked for whether its parentheses ('(',')')
1236 Examples of valid syntax (even though the attributes are unknown):
1238 sub fnord (&\%) : switch(10,foo(7,3)) , , expensive ;
1239 sub plugh () : Ugly('\(") , Bad ;
1240 sub xyzzy : _5x5 { ... }
1242 Examples of invalid syntax:
1244 sub fnord : switch(10,foo() ; # ()-string not balanced
1245 sub snoid : Ugly('(') ; # ()-string not balanced
1246 sub xyzzy : 5x5 ; # "5x5" not a valid identifier
1247 sub plugh : Y2::north ; # "Y2::north" not a simple identifier
1248 sub snurt : foo + bar ; # "+" not a comma or space
1250 The attribute list is passed as a list of constant strings to the code
1251 which associates them with the subroutine. In particular, the second example
1252 of valid syntax above currently looks like this in terms of how it's
1255 use attributes __PACKAGE__, \&plugh, q[Ugly('\(")], 'Bad';
1257 For further details on attribute lists and their manipulation,
1262 See L<perlref/"Function Templates"> for more about references and closures.
1263 See L<perlxs> if you'd like to learn about calling C subroutines from Perl.
1264 See L<perlembed> if you'd like to learn about calling PErl subroutines from C.
1265 See L<perlmod> to learn about bundling up your functions in separate files.
1266 See L<perlmodlib> to learn what library modules come standard on your system.
1267 See L<perltoot> to learn how to make object method calls.