7 First you need to understand what references are in Perl.
8 See L<perlref> for that. Second, if you still find the following
9 reference work too complicated, a tutorial on object-oriented programming
10 in Perl can be found in L<perltoot> and L<perltooc>.
12 If you're still with us, then
13 here are three very simple definitions that you should find reassuring.
19 An object is simply a reference that happens to know which class it
24 A class is simply a package that happens to provide methods to deal
25 with object references.
29 A method is simply a subroutine that expects an object reference (or
30 a package name, for class methods) as the first argument.
34 We'll cover these points now in more depth.
36 =head2 An Object is Simply a Reference
38 Unlike say C++, Perl doesn't provide any special syntax for
39 constructors. A constructor is merely a subroutine that returns a
40 reference to something "blessed" into a class, generally the
41 class that the subroutine is defined in. Here is a typical
47 That word C<new> isn't special. You could have written
48 a construct this way, too:
51 sub spawn { bless {} }
53 This might even be preferable, because the C++ programmers won't
54 be tricked into thinking that C<new> works in Perl as it does in C++.
55 It doesn't. We recommend that you name your constructors whatever
56 makes sense in the context of the problem you're solving. For example,
57 constructors in the Tk extension to Perl are named after the widgets
60 One thing that's different about Perl constructors compared with those in
61 C++ is that in Perl, they have to allocate their own memory. (The other
62 things is that they don't automatically call overridden base-class
63 constructors.) The C<{}> allocates an anonymous hash containing no
64 key/value pairs, and returns it The bless() takes that reference and
65 tells the object it references that it's now a Critter, and returns
66 the reference. This is for convenience, because the referenced object
67 itself knows that it has been blessed, and the reference to it could
68 have been returned directly, like this:
76 You often see such a thing in more complicated constructors
77 that wish to call methods in the class as part of the construction:
86 If you care about inheritance (and you should; see
87 L<perlmodlib/"Modules: Creation, Use, and Abuse">),
88 then you want to use the two-arg form of bless
89 so that your constructors may be inherited:
99 Or if you expect people to call not just C<< CLASS->new() >> but also
100 C<< $obj->new() >>, then use something like this. The initialize()
101 method used will be of whatever $class we blessed the
106 my $class = ref($this) || $this;
113 Within the class package, the methods will typically deal with the
114 reference as an ordinary reference. Outside the class package,
115 the reference is generally treated as an opaque value that may
116 be accessed only through the class's methods.
118 Although a constructor can in theory re-bless a referenced object
119 currently belonging to another class, this is almost certainly going
120 to get you into trouble. The new class is responsible for all
121 cleanup later. The previous blessing is forgotten, as an object
122 may belong to only one class at a time. (Although of course it's
123 free to inherit methods from many classes.) If you find yourself
124 having to do this, the parent class is probably misbehaving, though.
126 A clarification: Perl objects are blessed. References are not. Objects
127 know which package they belong to. References do not. The bless()
128 function uses the reference to find the object. Consider
129 the following example:
134 print "\$b is a ", ref($b), "\n";
136 This reports $b as being a BLAH, so obviously bless()
137 operated on the object and not on the reference.
139 =head2 A Class is Simply a Package
141 Unlike say C++, Perl doesn't provide any special syntax for class
142 definitions. You use a package as a class by putting method
143 definitions into the class.
145 There is a special array within each package called @ISA, which says
146 where else to look for a method if you can't find it in the current
147 package. This is how Perl implements inheritance. Each element of the
148 @ISA array is just the name of another package that happens to be a
149 class package. The classes are searched (depth first) for missing
150 methods in the order that they occur in @ISA. The classes accessible
151 through @ISA are known as base classes of the current class.
153 All classes implicitly inherit from class C<UNIVERSAL> as their
154 last base class. Several commonly used methods are automatically
155 supplied in the UNIVERSAL class; see L<"Default UNIVERSAL methods"> for
158 If a missing method is found in a base class, it is cached
159 in the current class for efficiency. Changing @ISA or defining new
160 subroutines invalidates the cache and causes Perl to do the lookup again.
162 If neither the current class, its named base classes, nor the UNIVERSAL
163 class contains the requested method, these three places are searched
164 all over again, this time looking for a method named AUTOLOAD(). If an
165 AUTOLOAD is found, this method is called on behalf of the missing method,
166 setting the package global $AUTOLOAD to be the fully qualified name of
167 the method that was intended to be called.
169 If none of that works, Perl finally gives up and complains.
171 If you want to stop the AUTOLOAD inheritance say simply
175 and the call will die using the name of the sub being called.
177 Perl classes do method inheritance only. Data inheritance is left up
178 to the class itself. By and large, this is not a problem in Perl,
179 because most classes model the attributes of their object using an
180 anonymous hash, which serves as its own little namespace to be carved up
181 by the various classes that might want to do something with the object.
182 The only problem with this is that you can't sure that you aren't using
183 a piece of the hash that isn't already used. A reasonable workaround
184 is to prepend your fieldname in the hash with the package name.
188 $self->{ __PACKAGE__ . ".count"}++;
191 =head2 A Method is Simply a Subroutine
193 Unlike say C++, Perl doesn't provide any special syntax for method
194 definition. (It does provide a little syntax for method invocation
195 though. More on that later.) A method expects its first argument
196 to be the object (reference) or package (string) it is being invoked
197 on. There are two ways of calling methods, which we'll call class
198 methods and instance methods.
200 A class method expects a class name as the first argument. It
201 provides functionality for the class as a whole, not for any
202 individual object belonging to the class. Constructors are often
203 class methods, but see L<perltoot> and L<perltooc> for alternatives.
204 Many class methods simply ignore their first argument, because they
205 already know what package they're in and don't care what package
206 they were invoked via. (These aren't necessarily the same, because
207 class methods follow the inheritance tree just like ordinary instance
208 methods.) Another typical use for class methods is to look up an
212 my ($class, $name) = @_;
216 An instance method expects an object reference as its first argument.
217 Typically it shifts the first argument into a "self" or "this" variable,
218 and then uses that as an ordinary reference.
222 my @keys = @_ ? @_ : sort keys %$self;
223 foreach $key (@keys) {
224 print "\t$key => $self->{$key}\n";
228 =head2 Method Invocation
230 For various historical and other reasons, Perl offers two equivalent
231 ways to write a method call. The simpler and more common way is to use
234 my $fred = Critter->find("Fred");
235 $fred->display("Height", "Weight");
237 You should already be familiar with the use of the C<< -> >> operator with
238 references. In fact, since C<$fred> above is a reference to an object,
239 you could think of the method call as just another form of
242 Whatever is on the left side of the arrow, whether a reference or a
243 class name, is passed to the method subroutine as its first argument.
244 So the above code is mostly equivalent to:
246 my $fred = Critter::find("Critter", "Fred");
247 Critter::display($fred, "Height", "Weight");
249 How does Perl know which package the subroutine is in? By looking at
250 the left side of the arrow, which must be either a package name or a
251 reference to an object, i.e. something that has been blessed to a
252 package. Either way, that's the package where Perl starts looking. If
253 that package has no subroutine with that name, Perl starts looking for
254 it in any base classes of that package, and so on.
256 If you need to, you I<can> force Perl to start looking in some other package:
258 my $barney = MyCritter->Critter::find("Barney");
259 $barney->Critter::display("Height", "Weight");
261 Here C<MyCritter> is presumably a subclass of C<Critter> that defines
262 its own versions of find() and display(). We haven't specified what
263 those methods do, but that doesn't matter above since we've forced Perl
264 to start looking for the subroutines in C<Critter>.
266 As a special case of the above, you may use the C<SUPER> pseudo-class to
267 tell Perl to start looking for the method in the packages named in the
268 current class's C<@ISA> list.
271 use base 'Critter'; # sets @MyCritter::ISA = ('Critter');
274 my ($self, @args) = @_;
275 $self->SUPER::display("Name", @args);
278 It is important to note that C<SUPER> refers to the superclass(es) of the
279 I<current package> and not to the superclass(es) of the object. Also, the
280 C<SUPER> pseudo-class can only currently be used as a modifier to a method
281 name, but not in any of the other ways that class names are normally used,
284 something->SUPER::method(...); # OK
285 SUPER::method(...); # WRONG
286 SUPER->method(...); # WRONG
288 Instead of a class name or an object reference, you can also use any
289 expression that returns either of those on the left side of the arrow.
290 So the following statement is valid:
292 Critter->find("Fred")->display("Height", "Weight");
294 and so is the following:
296 my $fred = (reverse "rettirC")->find(reverse "derF");
298 =head2 Indirect Object Syntax
300 The other way to invoke a method is by using the so-called "indirect
301 object" notation. This syntax was available in Perl 4 long before
302 objects were introduced, and is still used with filehandles like this:
304 print STDERR "help!!!\n";
306 The same syntax can be used to call either object or class methods.
308 my $fred = find Critter "Fred";
309 display $fred "Height", "Weight";
311 Notice that there is no comma between the object or class name and the
312 parameters. This is how Perl can tell you want an indirect method call
313 instead of an ordinary subroutine call.
315 But what if there are no arguments? In that case, Perl must guess what
316 you want. Even worse, it must make that guess I<at compile time>.
317 Usually Perl gets it right, but when it doesn't you get a function
318 call compiled as a method, or vice versa. This can introduce subtle bugs
319 that are hard to detect.
321 For example, a call to a method C<new> in indirect notation -- as C++
322 programmers are wont to make -- can be miscompiled into a subroutine
323 call if there's already a C<new> function in scope. You'd end up
324 calling the current package's C<new> as a subroutine, rather than the
325 desired class's method. The compiler tries to cheat by remembering
326 bareword C<require>s, but the grief when it messes up just isn't worth the
327 years of debugging it will take you to track down such subtle bugs.
329 There is another problem with this syntax: the indirect object is
330 limited to a name, a scalar variable, or a block, because it would have
331 to do too much lookahead otherwise, just like any other postfix
332 dereference in the language. (These are the same quirky rules as are
333 used for the filehandle slot in functions like C<print> and C<printf>.)
334 This can lead to horribly confusing precedence problems, as in these
337 move $obj->{FIELD}; # probably wrong!
338 move $ary[$i]; # probably wrong!
340 Those actually parse as the very surprising:
342 $obj->move->{FIELD}; # Well, lookee here
343 $ary->move([$i]); # Didn't expect this one, eh?
345 Rather than what you might have expected:
347 $obj->{FIELD}->move(); # You should be so lucky.
348 $ary[$i]->move; # Yeah, sure.
350 To get the correct behavior with indirect object syntax, you would have
351 to use a block around the indirect object:
353 move {$obj->{FIELD}};
356 Even then, you still have the same potential problem if there happens to
357 be a function named C<move> in the current package. B<The C<< -> >>
358 notation suffers from neither of these disturbing ambiguities, so we
359 recommend you use it exclusively.> However, you may still end up having
360 to read code using the indirect object notation, so it's important to be
363 =head2 Default UNIVERSAL methods
365 The C<UNIVERSAL> package automatically contains the following methods that
366 are inherited by all other classes:
372 C<isa> returns I<true> if its object is blessed into a subclass of C<CLASS>
374 You can also call C<UNIVERSAL::isa> as a subroutine with two arguments.
375 The first does not need to be an object or even a reference. This
376 allows you to check what a reference points to, or whether
377 something is a reference of a given type. Example
379 if(UNIVERSAL::isa($ref, 'ARRAY')) {
383 To determine if a reference is a blessed object, you can write
385 print "It's an object\n" if UNIVERSAL::isa($val, 'UNIVERSAL');
389 C<can> checks to see if its object has a method called C<METHOD>,
390 if it does then a reference to the sub is returned, if it does not then
391 I<undef> is returned.
393 C<UNIVERSAL::can> can also be called as a subroutine with two arguments.
394 It'll always return I<undef> if its first argument isn't an object or a
395 class name. So here's another way to check if a reference is a
398 print "It's still an object\n" if UNIVERSAL::can($val, 'can');
400 You can also use the C<blessed> function of Scalar::Util:
402 use Scalar::Util 'blessed';
404 my $blessing = blessed $suspected_object;
406 C<blessed> returns the name of the package the argument has been
407 blessed into, or C<undef>.
409 =item VERSION( [NEED] )
411 C<VERSION> returns the version number of the class (package). If the
412 NEED argument is given then it will check that the current version (as
413 defined by the $VERSION variable in the given package) not less than
414 NEED; it will die if this is not the case. This method is normally
415 called as a class method. This method is called automatically by the
416 C<VERSION> form of C<use>.
418 use A 1.2 qw(some imported subs);
424 B<NOTE:> C<can> directly uses Perl's internal code for method lookup, and
425 C<isa> uses a very similar method and cache-ing strategy. This may cause
426 strange effects if the Perl code dynamically changes @ISA in any package.
428 You may add other methods to the UNIVERSAL class via Perl or XS code.
429 You do not need to C<use UNIVERSAL> to make these methods
430 available to your program (and you should not do so).
434 When the last reference to an object goes away, the object is
435 automatically destroyed. (This may even be after you exit, if you've
436 stored references in global variables.) If you want to capture control
437 just before the object is freed, you may define a DESTROY method in
438 your class. It will automatically be called at the appropriate moment,
439 and you can do any extra cleanup you need to do. Perl passes a reference
440 to the object under destruction as the first (and only) argument. Beware
441 that the reference is a read-only value, and cannot be modified by
442 manipulating C<$_[0]> within the destructor. The object itself (i.e.
443 the thingy the reference points to, namely C<${$_[0]}>, C<@{$_[0]}>,
444 C<%{$_[0]}> etc.) is not similarly constrained.
446 If you arrange to re-bless the reference before the destructor returns,
447 perl will again call the DESTROY method for the re-blessed object after
448 the current one returns. This can be used for clean delegation of
449 object destruction, or for ensuring that destructors in the base classes
450 of your choosing get called. Explicitly calling DESTROY is also possible,
451 but is usually never needed.
453 Do not confuse the previous discussion with how objects I<CONTAINED> in the current
454 one are destroyed. Such objects will be freed and destroyed automatically
455 when the current object is freed, provided no other references to them exist
460 That's about all there is to it. Now you need just to go off and buy a
461 book about object-oriented design methodology, and bang your forehead
462 with it for the next six months or so.
464 =head2 Two-Phased Garbage Collection
466 For most purposes, Perl uses a fast and simple, reference-based
467 garbage collection system. That means there's an extra
468 dereference going on at some level, so if you haven't built
469 your Perl executable using your C compiler's C<-O> flag, performance
470 will suffer. If you I<have> built Perl with C<cc -O>, then this
471 probably won't matter.
473 A more serious concern is that unreachable memory with a non-zero
474 reference count will not normally get freed. Therefore, this is a bad
482 Even thought $a I<should> go away, it can't. When building recursive data
483 structures, you'll have to break the self-reference yourself explicitly
484 if you don't care to leak. For example, here's a self-referential
485 node such as one might use in a sophisticated tree structure:
489 my $class = ref($self) || $self;
491 $node->{LEFT} = $node->{RIGHT} = $node;
492 $node->{DATA} = [ @_ ];
493 return bless $node => $class;
496 If you create nodes like that, they (currently) won't go away unless you
497 break their self reference yourself. (In other words, this is not to be
498 construed as a feature, and you shouldn't depend on it.)
502 When an interpreter thread finally shuts down (usually when your program
503 exits), then a rather costly but complete mark-and-sweep style of garbage
504 collection is performed, and everything allocated by that thread gets
505 destroyed. This is essential to support Perl as an embedded or a
506 multithreadable language. For example, this program demonstrates Perl's
507 two-phased garbage collection:
515 warn "CREATING " . \$test;
521 warn "DESTROYING $self";
526 warn "starting program";
530 $$a = 0; # break selfref
531 warn "leaving block";
534 warn "just exited block";
535 warn "time to die...";
538 When run as F</foo/test>, the following output is produced:
540 starting program at /foo/test line 18.
541 CREATING SCALAR(0x8e5b8) at /foo/test line 7.
542 CREATING SCALAR(0x8e57c) at /foo/test line 7.
543 leaving block at /foo/test line 23.
544 DESTROYING Subtle=SCALAR(0x8e5b8) at /foo/test line 13.
545 just exited block at /foo/test line 26.
546 time to die... at /foo/test line 27.
547 DESTROYING Subtle=SCALAR(0x8e57c) during global destruction.
549 Notice that "global destruction" bit there? That's the thread
550 garbage collector reaching the unreachable.
552 Objects are always destructed, even when regular refs aren't. Objects
553 are destructed in a separate pass before ordinary refs just to
554 prevent object destructors from using refs that have been themselves
555 destructed. Plain refs are only garbage-collected if the destruct level
556 is greater than 0. You can test the higher levels of global destruction
557 by setting the PERL_DESTRUCT_LEVEL environment variable, presuming
558 C<-DDEBUGGING> was enabled during perl build time.
559 See L<perlhack/PERL_DESTRUCT_LEVEL> for more information.
561 A more complete garbage collection strategy will be implemented
564 In the meantime, the best solution is to create a non-recursive container
565 class that holds a pointer to the self-referential data structure.
566 Define a DESTROY method for the containing object's class that manually
567 breaks the circularities in the self-referential structure.
571 A kinder, gentler tutorial on object-oriented programming in Perl can
572 be found in L<perltoot>, L<perlboot> and L<perltooc>. You should
573 also check out L<perlbot> for other object tricks, traps, and tips, as
574 well as L<perlmodlib> for some style guides on constructing both