3 perltoot - Tom's object-oriented tutorial for perl
7 Object-oriented programming is a big seller these days. Some managers
8 would rather have objects than sliced bread. Why is that? What's so
9 special about an object? Just what I<is> an object anyway?
11 An object is nothing but a way of tucking away complex behaviours into
12 a neat little easy-to-use bundle. (This is what professors call
13 abstraction.) Smart people who have nothing to do but sit around for
14 weeks on end figuring out really hard problems make these nifty
15 objects that even regular people can use. (This is what professors call
16 software reuse.) Users (well, programmers) can play with this little
17 bundle all they want, but they aren't to open it up and mess with the
18 insides. Just like an expensive piece of hardware, the contract says
19 that you void the warranty if you muck with the cover. So don't do that.
21 The heart of objects is the class, a protected little private namespace
22 full of data and functions. A class is a set of related routines that
23 addresses some problem area. You can think of it as a user-defined type.
24 The Perl package mechanism, also used for more traditional modules,
25 is used for class modules as well. Objects "live" in a class, meaning
26 that they belong to some package.
28 More often than not, the class provides the user with little bundles.
29 These bundles are objects. They know whose class they belong to,
30 and how to behave. Users ask the class to do something, like "give
31 me an object." Or they can ask one of these objects to do something.
32 Asking a class to do something for you is calling a I<class method>.
33 Asking an object to do something for you is calling an I<object method>.
34 Asking either a class (usually) or an object (sometimes) to give you
35 back an object is calling a I<constructor>, which is just a
38 That's all well and good, but how is an object different from any other
39 Perl data type? Just what is an object I<really>; that is, what's its
40 fundamental type? The answer to the first question is easy. An object
41 is different from any other data type in Perl in one and only one way:
42 you may dereference it using not merely string or numeric subscripts
43 as with simple arrays and hashes, but with named subroutine calls.
44 In a word, with I<methods>.
46 The answer to the second question is that it's a reference, and not just
47 any reference, mind you, but one whose referent has been I<bless>()ed
48 into a particular class (read: package). What kind of reference? Well,
49 the answer to that one is a bit less concrete. That's because in Perl
50 the designer of the class can employ any sort of reference they'd like
51 as the underlying intrinsic data type. It could be a scalar, an array,
52 or a hash reference. It could even be a code reference. But because
53 of its inherent flexibility, an object is usually a hash reference.
55 =head1 Creating a Class
57 Before you create a class, you need to decide what to name it. That's
58 because the class (package) name governs the name of the file used to
59 house it, just as with regular modules. Then, that class (package)
60 should provide one or more ways to generate objects. Finally, it should
61 provide mechanisms to allow users of its objects to indirectly manipulate
62 these objects from a distance.
64 For example, let's make a simple Person class module. It gets stored in
65 the file Person.pm. If it were called a Happy::Person class, it would
66 be stored in the file Happy/Person.pm, and its package would become
67 Happy::Person instead of just Person. (On a personal computer not
68 running Unix or Plan 9, but something like Mac OS or VMS, the directory
69 separator may be different, but the principle is the same.) Do not assume
70 any formal relationship between modules based on their directory names.
71 This is merely a grouping convenience, and has no effect on inheritance,
72 variable accessibility, or anything else.
74 For this module we aren't going to use Exporter, because we're
75 a well-behaved class module that doesn't export anything at all.
76 In order to manufacture objects, a class needs to have a I<constructor
77 method>. A constructor gives you back not just a regular data type,
78 but a brand-new object in that class. This magic is taken care of by
79 the bless() function, whose sole purpose is to enable its referent to
80 be used as an object. Remember: being an object really means nothing
81 more than that methods may now be called against it.
83 While a constructor may be named anything you'd like, most Perl
84 programmers seem to like to call theirs new(). However, new() is not
85 a reserved word, and a class is under no obligation to supply such.
86 Some programmers have also been known to use a function with
87 the same name as the class as the constructor.
89 =head2 Object Representation
91 By far the most common mechanism used in Perl to represent a Pascal
92 record, a C struct, or a C++ class is an anonymous hash. That's because a
93 hash has an arbitrary number of data fields, each conveniently accessed by
94 an arbitrary name of your own devising.
96 If you were just doing a simple
97 struct-like emulation, you would likely go about it something like this:
102 peers => [ "Norbert", "Rhys", "Phineas"],
105 If you felt like it, you could add a bit of visual distinction
106 by up-casing the hash keys:
111 PEERS => [ "Norbert", "Rhys", "Phineas"],
114 And so you could get at C<< $rec->{NAME} >> to find "Jason", or
115 C<< @{ $rec->{PEERS} } >> to get at "Norbert", "Rhys", and "Phineas".
116 (Have you ever noticed how many 23-year-old programmers seem to
117 be named "Jason" these days? :-)
119 This same model is often used for classes, although it is not considered
120 the pinnacle of programming propriety for folks from outside the
121 class to come waltzing into an object, brazenly accessing its data
122 members directly. Generally speaking, an object should be considered
123 an opaque cookie that you use I<object methods> to access. Visually,
124 methods look like you're dereffing a reference using a function name
125 instead of brackets or braces.
127 =head2 Class Interface
129 Some languages provide a formal syntactic interface to a class's methods,
130 but Perl does not. It relies on you to read the documentation of each
131 class. If you try to call an undefined method on an object, Perl won't
132 complain, but the program will trigger an exception while it's running.
133 Likewise, if you call a method expecting a prime number as its argument
134 with a non-prime one instead, you can't expect the compiler to catch this.
135 (Well, you can expect it all you like, but it's not going to happen.)
137 Let's suppose you have a well-educated user of your Person class,
138 someone who has read the docs that explain the prescribed
139 interface. Here's how they might use the Person class:
143 $him = Person->new();
146 $him->peers( "Norbert", "Rhys", "Phineas" );
148 push @All_Recs, $him; # save object in array for later
150 printf "%s is %d years old.\n", $him->name, $him->age;
151 print "His peers are: ", join(", ", $him->peers), "\n";
153 printf "Last rec's name is %s\n", $All_Recs[-1]->name;
155 As you can see, the user of the class doesn't know (or at least, has no
156 business paying attention to the fact) that the object has one particular
157 implementation or another. The interface to the class and its objects
158 is exclusively via methods, and that's all the user of the class should
161 =head2 Constructors and Instance Methods
163 Still, I<someone> has to know what's in the object. And that someone is
164 the class. It implements methods that the programmer uses to access
165 the object. Here's how to implement the Person class using the standard
166 hash-ref-as-an-object idiom. We'll make a class method called new() to
167 act as the constructor, and three object methods called name(), age(), and
168 peers() to get at per-object data hidden away in our anonymous hash.
173 ##################################################
174 ## the object constructor (simplistic version) ##
175 ##################################################
178 $self->{NAME} = undef;
179 $self->{AGE} = undef;
181 bless($self); # but see below
185 ##############################################
186 ## methods to access per-object data ##
188 ## With args, they set the value. Without ##
189 ## any, they only retrieve it/them. ##
190 ##############################################
194 if (@_) { $self->{NAME} = shift }
195 return $self->{NAME};
200 if (@_) { $self->{AGE} = shift }
206 if (@_) { @{ $self->{PEERS} } = @_ }
207 return @{ $self->{PEERS} };
210 1; # so the require or use succeeds
212 We've created three methods to access an object's data, name(), age(),
213 and peers(). These are all substantially similar. If called with an
214 argument, they set the appropriate field; otherwise they return the
215 value held by that field, meaning the value of that hash key.
217 =head2 Planning for the Future: Better Constructors
219 Even though at this point you may not even know what it means, someday
220 you're going to worry about inheritance. (You can safely ignore this
221 for now and worry about it later if you'd like.) To ensure that this
222 all works out smoothly, you must use the double-argument form of bless().
223 The second argument is the class into which the referent will be blessed.
224 By not assuming our own class as the default second argument and instead
225 using the class passed into us, we make our constructor inheritable.
227 While we're at it, let's make our constructor a bit more flexible.
228 Rather than being uniquely a class method, we'll set it up so that
229 it can be called as either a class method I<or> an object
230 method. That way you can say:
235 To do this, all we have to do is check whether what was passed in
236 was a reference or not. If so, we were invoked as an object method,
237 and we need to extract the package (class) using the ref() function.
238 If not, we just use the string passed in as the package name
239 for blessing our referent.
243 my $class = ref($proto) || $proto;
245 $self->{NAME} = undef;
246 $self->{AGE} = undef;
248 bless ($self, $class);
252 That's about all there is for constructors. These methods bring objects
253 to life, returning neat little opaque bundles to the user to be used in
254 subsequent method calls.
258 Every story has a beginning and an end. The beginning of the object's
259 story is its constructor, explicitly called when the object comes into
260 existence. But the ending of its story is the I<destructor>, a method
261 implicitly called when an object leaves this life. Any per-object
262 clean-up code is placed in the destructor, which must (in Perl) be called
265 If constructors can have arbitrary names, then why not destructors?
266 Because while a constructor is explicitly called, a destructor is not.
267 Destruction happens automatically via Perl's garbage collection (GC)
268 system, which is a quick but somewhat lazy reference-based GC system.
269 To know what to call, Perl insists that the destructor be named DESTROY.
270 Perl's notion of the right time to call a destructor is not well-defined
271 currently, which is why your destructors should not rely on when they are
274 Why is DESTROY in all caps? Perl on occasion uses purely uppercase
275 function names as a convention to indicate that the function will
276 be automatically called by Perl in some way. Others that are called
277 implicitly include BEGIN, END, AUTOLOAD, plus all methods used by
278 tied objects, described in L<perltie>.
280 In really good object-oriented programming languages, the user doesn't
281 care when the destructor is called. It just happens when it's supposed
282 to. In low-level languages without any GC at all, there's no way to
283 depend on this happening at the right time, so the programmer must
284 explicitly call the destructor to clean up memory and state, crossing
285 their fingers that it's the right time to do so. Unlike C++, an
286 object destructor is nearly never needed in Perl, and even when it is,
287 explicit invocation is uncalled for. In the case of our Person class,
288 we don't need a destructor because Perl takes care of simple matters
289 like memory deallocation.
291 The only situation where Perl's reference-based GC won't work is
292 when there's a circularity in the data structure, such as:
294 $this->{WHATEVER} = $this;
296 In that case, you must delete the self-reference manually if you expect
297 your program not to leak memory. While admittedly error-prone, this is
298 the best we can do right now. Nonetheless, rest assured that when your
299 program is finished, its objects' destructors are all duly called.
300 So you are guaranteed that an object I<eventually> gets properly
301 destroyed, except in the unique case of a program that never exits.
302 (If you're running Perl embedded in another application, this full GC
303 pass happens a bit more frequently--whenever a thread shuts down.)
305 =head2 Other Object Methods
307 The methods we've talked about so far have either been constructors or
308 else simple "data methods", interfaces to data stored in the object.
309 These are a bit like an object's data members in the C++ world, except
310 that strangers don't access them as data. Instead, they should only
311 access the object's data indirectly via its methods. This is an
312 important rule: in Perl, access to an object's data should I<only>
313 be made through methods.
315 Perl doesn't impose restrictions on who gets to use which methods.
316 The public-versus-private distinction is by convention, not syntax.
317 (Well, unless you use the Alias module described below in
318 L<Data Members as Variables>.) Occasionally you'll see method names beginning or ending
319 with an underscore or two. This marking is a convention indicating
320 that the methods are private to that class alone and sometimes to its
321 closest acquaintances, its immediate subclasses. But this distinction
322 is not enforced by Perl itself. It's up to the programmer to behave.
324 There's no reason to limit methods to those that simply access data.
325 Methods can do anything at all. The key point is that they're invoked
326 against an object or a class. Let's say we'd like object methods that
327 do more than fetch or set one particular field.
331 return sprintf "Hi, I'm %s, age %d, working with %s",
332 $self->{NAME}, $self->{AGE}, join(", ", @{$self->{PEERS}});
335 Or maybe even one like this:
339 return ++$self->{AGE};
342 Some might argue that one should go at these this way:
346 return sprintf "Hi, I'm %s, age %d, working with %s",
347 $self->name, $self->age, join(", ", $self->peers);
352 return $self->age( $self->age() + 1 );
355 But since these methods are all executing in the class itself, this
356 may not be critical. There are tradeoffs to be made. Using direct
357 hash access is faster (about an order of magnitude faster, in fact), and
358 it's more convenient when you want to interpolate in strings. But using
359 methods (the external interface) internally shields not just the users of
360 your class but even you yourself from changes in your data representation.
364 What about "class data", data items common to each object in a class?
365 What would you want that for? Well, in your Person class, you might
366 like to keep track of the total people alive. How do you implement that?
368 You I<could> make it a global variable called $Person::Census. But about
369 only reason you'd do that would be if you I<wanted> people to be able to
370 get at your class data directly. They could just say $Person::Census
371 and play around with it. Maybe this is ok in your design scheme.
372 You might even conceivably want to make it an exported variable. To be
373 exportable, a variable must be a (package) global. If this were a
374 traditional module rather than an object-oriented one, you might do that.
376 While this approach is expected in most traditional modules, it's
377 generally considered rather poor form in most object modules. In an
378 object module, you should set up a protective veil to separate interface
379 from implementation. So provide a class method to access class data
380 just as you provide object methods to access object data.
382 So, you I<could> still keep $Census as a package global and rely upon
383 others to honor the contract of the module and therefore not play around
384 with its implementation. You could even be supertricky and make $Census a
385 tied object as described in L<perltie>, thereby intercepting all accesses.
387 But more often than not, you just want to make your class data a
388 file-scoped lexical. To do so, simply put this at the top of the file:
392 Even though the scope of a my() normally expires when the block in which
393 it was declared is done (in this case the whole file being required or
394 used), Perl's deep binding of lexical variables guarantees that the
395 variable will not be deallocated, remaining accessible to functions
396 declared within that scope. This doesn't work with global variables
397 given temporary values via local(), though.
399 Irrespective of whether you leave $Census a package global or make
400 it instead a file-scoped lexical, you should make these
401 changes to your Person::new() constructor:
405 my $class = ref($proto) || $proto;
408 $self->{NAME} = undef;
409 $self->{AGE} = undef;
411 bless ($self, $class);
419 Now that we've done this, we certainly do need a destructor so that
420 when Person is destroyed, the $Census goes down. Here's how
423 sub DESTROY { --$Census }
425 Notice how there's no memory to deallocate in the destructor? That's
426 something that Perl takes care of for you all by itself.
428 Alternatively, you could use the Class::Data::Inheritable module from
432 =head2 Accessing Class Data
434 It turns out that this is not really a good way to go about handling
435 class data. A good scalable rule is that I<you must never reference class
436 data directly from an object method>. Otherwise you aren't building a
437 scalable, inheritable class. The object must be the rendezvous point
438 for all operations, especially from an object method. The globals
439 (class data) would in some sense be in the "wrong" package in your
440 derived classes. In Perl, methods execute in the context of the class
441 they were defined in, I<not> that of the object that triggered them.
442 Therefore, namespace visibility of package globals in methods is unrelated
445 Got that? Maybe not. Ok, let's say that some other class "borrowed"
446 (well, inherited) the DESTROY method as it was defined above. When those
447 objects are destroyed, the original $Census variable will be altered,
448 not the one in the new class's package namespace. Perhaps this is what
449 you want, but probably it isn't.
451 Here's how to fix this. We'll store a reference to the data in the
452 value accessed by the hash key "_CENSUS". Why the underscore? Well,
453 mostly because an initial underscore already conveys strong feelings
454 of magicalness to a C programmer. It's really just a mnemonic device
455 to remind ourselves that this field is special and not to be used as
456 a public data member in the same way that NAME, AGE, and PEERS are.
457 (Because we've been developing this code under the strict pragma, prior
458 to perl version 5.004 we'll have to quote the field name.)
462 my $class = ref($proto) || $proto;
464 $self->{NAME} = undef;
465 $self->{AGE} = undef;
468 $self->{"_CENSUS"} = \$Census;
469 bless ($self, $class);
470 ++ ${ $self->{"_CENSUS"} };
477 return ${ $self->{"_CENSUS"} };
485 -- ${ $self->{"_CENSUS"} };
488 =head2 Debugging Methods
490 It's common for a class to have a debugging mechanism. For example,
491 you might want to see when objects are created or destroyed. To do that,
492 add a debugging variable as a file-scoped lexical. For this, we'll pull
493 in the standard Carp module to emit our warnings and fatal messages.
494 That way messages will come out with the caller's filename and
495 line number instead of our own; if we wanted them to be from our own
496 perspective, we'd just use die() and warn() directly instead of croak()
497 and carp() respectively.
502 Now add a new class method to access the variable.
506 if (ref $class) { confess "Class method called as object method" }
507 unless (@_ == 1) { confess "usage: CLASSNAME->debug(level)" }
511 Now fix up DESTROY to murmur a bit as the moribund object expires:
515 if ($Debugging) { carp "Destroying $self " . $self->name }
516 -- ${ $self->{"_CENSUS"} };
519 One could conceivably make a per-object debug state. That
520 way you could call both of these:
522 Person->debug(1); # entire class
523 $him->debug(1); # just this object
525 To do so, we need our debugging method to be a "bimodal" one, one that
526 works on both classes I<and> objects. Therefore, adjust the debug()
527 and DESTROY methods as follows:
531 confess "usage: thing->debug(level)" unless @_ == 1;
534 $self->{"_DEBUG"} = $level; # just myself
536 $Debugging = $level; # whole class
542 if ($Debugging || $self->{"_DEBUG"}) {
543 carp "Destroying $self " . $self->name;
545 -- ${ $self->{"_CENSUS"} };
548 What happens if a derived class (which we'll call Employee) inherits
549 methods from this Person base class? Then C<< Employee->debug() >>, when called
550 as a class method, manipulates $Person::Debugging not $Employee::Debugging.
552 =head2 Class Destructors
554 The object destructor handles the death of each distinct object. But sometimes
555 you want a bit of cleanup when the entire class is shut down, which
556 currently only happens when the program exits. To make such a
557 I<class destructor>, create a function in that class's package named
558 END. This works just like the END function in traditional modules,
559 meaning that it gets called whenever your program exits unless it execs
560 or dies of an uncaught signal. For example,
564 print "All persons are going away now.\n";
568 When the program exits, all the class destructors (END functions) are
569 be called in the opposite order that they were loaded in (LIFO order).
571 =head2 Documenting the Interface
573 And there you have it: we've just shown you the I<implementation> of this
574 Person class. Its I<interface> would be its documentation. Usually this
575 means putting it in pod ("plain old documentation") format right there
576 in the same file. In our Person example, we would place the following
577 docs anywhere in the Person.pm file. Even though it looks mostly like
578 code, it's not. It's embedded documentation such as would be used by
579 the pod2man, pod2html, or pod2text programs. The Perl compiler ignores
580 pods entirely, just as the translators ignore code. Here's an example of
581 some pods describing the informal interface:
585 Person - class to implement people
595 $count = Person->population;
597 #######################
598 # object data methods #
599 #######################
609 $ob->peers( "Norbert", "Rhys", "Phineas" );
611 ########################
612 # other object methods #
613 ########################
615 $phrase = $ob->exclaim;
620 The Person class implements dah dee dah dee dah....
622 That's all there is to the matter of interface versus implementation.
623 A programmer who opens up the module and plays around with all the private
624 little shiny bits that were safely locked up behind the interface contract
625 has voided the warranty, and you shouldn't worry about their fate.
629 Suppose you later want to change the class to implement better names.
630 Perhaps you'd like to support both given names (called Christian names,
631 irrespective of one's religion) and family names (called surnames), plus
632 nicknames and titles. If users of your Person class have been properly
633 accessing it through its documented interface, then you can easily change
634 the underlying implementation. If they haven't, then they lose and
635 it's their fault for breaking the contract and voiding their warranty.
637 To do this, we'll make another class, this one called Fullname. What's
638 the Fullname class look like? To answer that question, you have to
639 first figure out how you want to use it. How about we use it this way:
641 $him = Person->new();
642 $him->fullname->title("St");
643 $him->fullname->christian("Thomas");
644 $him->fullname->surname("Aquinas");
645 $him->fullname->nickname("Tommy");
646 printf "His normal name is %s\n", $him->name;
647 printf "But his real name is %s\n", $him->fullname->as_string;
649 Ok. To do this, we'll change Person::new() so that it supports
650 a full name field this way:
654 my $class = ref($proto) || $proto;
656 $self->{FULLNAME} = Fullname->new();
657 $self->{AGE} = undef;
659 $self->{"_CENSUS"} = \$Census;
660 bless ($self, $class);
661 ++ ${ $self->{"_CENSUS"} };
667 return $self->{FULLNAME};
670 Then to support old code, define Person::name() this way:
674 return $self->{FULLNAME}->nickname(@_)
675 || $self->{FULLNAME}->christian(@_);
678 Here's the Fullname class. We'll use the same technique
679 of using a hash reference to hold data fields, and methods
680 by the appropriate name to access them:
687 my $class = ref($proto) || $proto;
694 bless ($self, $class);
700 if (@_) { $self->{CHRISTIAN} = shift }
701 return $self->{CHRISTIAN};
706 if (@_) { $self->{SURNAME} = shift }
707 return $self->{SURNAME};
712 if (@_) { $self->{NICK} = shift }
713 return $self->{NICK};
718 if (@_) { $self->{TITLE} = shift }
719 return $self->{TITLE};
724 my $name = join(" ", @$self{'CHRISTIAN', 'SURNAME'});
725 if ($self->{TITLE}) {
726 $name = $self->{TITLE} . " " . $name;
733 Finally, here's the test program:
738 sub END { show_census() }
741 printf "Current population: %d\n", Person->population;
748 my $him = Person->new();
750 $him->fullname->christian("Thomas");
751 $him->fullname->surname("Aquinas");
752 $him->fullname->nickname("Tommy");
753 $him->fullname->title("St");
756 printf "%s is really %s.\n", $him->name, $him->fullname->as_string;
757 printf "%s's age: %d.\n", $him->name, $him->age;
758 $him->happy_birthday;
759 printf "%s's age: %d.\n", $him->name, $him->age;
765 Object-oriented programming systems all support some notion of
766 inheritance. Inheritance means allowing one class to piggy-back on
767 top of another one so you don't have to write the same code again and
768 again. It's about software reuse, and therefore related to Laziness,
769 the principal virtue of a programmer. (The import/export mechanisms in
770 traditional modules are also a form of code reuse, but a simpler one than
771 the true inheritance that you find in object modules.)
773 Sometimes the syntax of inheritance is built into the core of the
774 language, and sometimes it's not. Perl has no special syntax for
775 specifying the class (or classes) to inherit from. Instead, it's all
776 strictly in the semantics. Each package can have a variable called @ISA,
777 which governs (method) inheritance. If you try to call a method on an
778 object or class, and that method is not found in that object's package,
779 Perl then looks to @ISA for other packages to go looking through in
780 search of the missing method.
782 Like the special per-package variables recognized by Exporter (such as
783 @EXPORT, @EXPORT_OK, @EXPORT_FAIL, %EXPORT_TAGS, and $VERSION), the @ISA
784 array I<must> be a package-scoped global and not a file-scoped lexical
785 created via my(). Most classes have just one item in their @ISA array.
786 In this case, we have what's called "single inheritance", or SI for short.
795 Not a lot to it, eh? All it's doing so far is loading in another
796 class and stating that this one will inherit methods from that
797 other class if need be. We have given it none of its own methods.
798 We rely upon an Employee to behave just like a Person.
800 Setting up an empty class like this is called the "empty subclass test";
801 that is, making a derived class that does nothing but inherit from a
802 base class. If the original base class has been designed properly,
803 then the new derived class can be used as a drop-in replacement for the
804 old one. This means you should be able to write a program like this:
807 my $empl = Employee->new();
808 $empl->name("Jason");
810 printf "%s is age %d.\n", $empl->name, $empl->age;
812 By proper design, we mean always using the two-argument form of bless(),
813 avoiding direct access of global data, and not exporting anything. If you
814 look back at the Person::new() function we defined above, we were careful
815 to do that. There's a bit of package data used in the constructor,
816 but the reference to this is stored on the object itself and all other
817 methods access package data via that reference, so we should be ok.
819 What do we mean by the Person::new() function -- isn't that actually
820 a method? Well, in principle, yes. A method is just a function that
821 expects as its first argument a class name (package) or object
822 (blessed reference). Person::new() is the function that both the
823 C<< Person->new() >> method and the C<< Employee->new() >> method end
824 up calling. Understand that while a method call looks a lot like a
825 function call, they aren't really quite the same, and if you treat them
826 as the same, you'll very soon be left with nothing but broken programs.
827 First, the actual underlying calling conventions are different: method
828 calls get an extra argument. Second, function calls don't do inheritance,
831 Method Call Resulting Function Call
832 ----------- ------------------------
833 Person->new() Person::new("Person")
834 Employee->new() Person::new("Employee")
836 So don't use function calls when you mean to call a method.
838 If an employee is just a Person, that's not all too very interesting.
839 So let's add some other methods. We'll give our employee
840 data fields to access their salary, their employee ID, and their
843 If you're getting a little tired of creating all these nearly identical
844 methods just to get at the object's data, do not despair. Later,
845 we'll describe several different convenience mechanisms for shortening
846 this up. Meanwhile, here's the straight-forward way:
850 if (@_) { $self->{SALARY} = shift }
851 return $self->{SALARY};
856 if (@_) { $self->{ID} = shift }
862 if (@_) { $self->{START_DATE} = shift }
863 return $self->{START_DATE};
866 =head2 Overridden Methods
868 What happens when both a derived class and its base class have the same
869 method defined? Well, then you get the derived class's version of that
870 method. For example, let's say that we want the peers() method called on
871 an employee to act a bit differently. Instead of just returning the list
872 of peer names, let's return slightly different strings. So doing this:
874 $empl->peers("Peter", "Paul", "Mary");
875 printf "His peers are: %s\n", join(", ", $empl->peers);
879 His peers are: PEON=PETER, PEON=PAUL, PEON=MARY
881 To do this, merely add this definition into the Employee.pm file:
885 if (@_) { @{ $self->{PEERS} } = @_ }
886 return map { "PEON=\U$_" } @{ $self->{PEERS} };
889 There, we've just demonstrated the high-falutin' concept known in certain
890 circles as I<polymorphism>. We've taken on the form and behaviour of
891 an existing object, and then we've altered it to suit our own purposes.
892 This is a form of Laziness. (Getting polymorphed is also what happens
893 when the wizard decides you'd look better as a frog.)
895 Every now and then you'll want to have a method call trigger both its
896 derived class (also known as "subclass") version as well as its base class
897 (also known as "superclass") version. In practice, constructors and
898 destructors are likely to want to do this, and it probably also makes
899 sense in the debug() method we showed previously.
901 To do this, add this to Employee.pm:
908 confess "usage: thing->debug(level)" unless @_ == 1;
911 $self->{"_DEBUG"} = $level;
913 $Debugging = $level; # whole class
915 Person::debug($self, $Debugging); # don't really do this
918 As you see, we turn around and call the Person package's debug() function.
919 But this is far too fragile for good design. What if Person doesn't
920 have a debug() function, but is inheriting I<its> debug() method
921 from elsewhere? It would have been slightly better to say
923 Person->debug($Debugging);
925 But even that's got too much hard-coded. It's somewhat better to say
927 $self->Person::debug($Debugging);
929 Which is a funny way to say to start looking for a debug() method up
930 in Person. This strategy is more often seen on overridden object methods
931 than on overridden class methods.
933 There is still something a bit off here. We've hard-coded our
934 superclass's name. This in particular is bad if you change which classes
935 you inherit from, or add others. Fortunately, the pseudoclass SUPER
936 comes to the rescue here.
938 $self->SUPER::debug($Debugging);
940 This way it starts looking in my class's @ISA. This only makes sense
941 from I<within> a method call, though. Don't try to access anything
942 in SUPER:: from anywhere else, because it doesn't exist outside
943 an overridden method call. Note that C<SUPER> refers to the superclass of
944 the current package, I<not> to the superclass of C<$self>.
946 Things are getting a bit complicated here. Have we done anything
947 we shouldn't? As before, one way to test whether we're designing
948 a decent class is via the empty subclass test. Since we already have
949 an Employee class that we're trying to check, we'd better get a new
950 empty subclass that can derive from Employee. Here's one:
956 And here's the test program:
963 my $boss = Boss->new();
965 $boss->fullname->title("Don");
966 $boss->fullname->surname("Pichon Alvarez");
967 $boss->fullname->christian("Federico Jesus");
968 $boss->fullname->nickname("Fred");
971 $boss->peers("Frank", "Felipe", "Faust");
973 printf "%s is age %d.\n", $boss->fullname->as_string, $boss->age;
974 printf "His peers are: %s\n", join(", ", $boss->peers);
976 Running it, we see that we're still ok. If you'd like to dump out your
977 object in a nice format, somewhat like the way the 'x' command works in
978 the debugger, you could use the Data::Dumper module from CPAN this way:
981 print "Here's the boss:\n";
984 Which shows us something like this:
991 SURNAME => 'Pichon Alvarez',
993 CHRISTIAN => 'Federico Jesus'
1003 Hm.... something's missing there. What about the salary, start date,
1004 and ID fields? Well, we never set them to anything, even undef, so they
1005 don't show up in the hash's keys. The Employee class has no new() method
1006 of its own, and the new() method in Person doesn't know about Employees.
1007 (Nor should it: proper OO design dictates that a subclass be allowed to
1008 know about its immediate superclass, but never vice-versa.) So let's
1009 fix up Employee::new() this way:
1013 my $class = ref($proto) || $proto;
1014 my $self = $class->SUPER::new();
1015 $self->{SALARY} = undef;
1016 $self->{ID} = undef;
1017 $self->{START_DATE} = undef;
1018 bless ($self, $class); # reconsecrate
1022 Now if you dump out an Employee or Boss object, you'll find
1023 that new fields show up there now.
1025 =head2 Multiple Inheritance
1027 Ok, at the risk of confusing beginners and annoying OO gurus, it's
1028 time to confess that Perl's object system includes that controversial
1029 notion known as multiple inheritance, or MI for short. All this means
1030 is that rather than having just one parent class who in turn might
1031 itself have a parent class, etc., that you can directly inherit from
1032 two or more parents. It's true that some uses of MI can get you into
1033 trouble, although hopefully not quite so much trouble with Perl as with
1034 dubiously-OO languages like C++.
1036 The way it works is actually pretty simple: just put more than one package
1037 name in your @ISA array. When it comes time for Perl to go finding
1038 methods for your object, it looks at each of these packages in order.
1039 Well, kinda. It's actually a fully recursive, depth-first order.
1040 Consider a bunch of @ISA arrays like this:
1042 @First::ISA = qw( Alpha );
1043 @Second::ISA = qw( Beta );
1044 @Third::ISA = qw( First Second );
1046 If you have an object of class Third:
1048 my $ob = Third->new();
1051 How do we find a spin() method (or a new() method for that matter)?
1052 Because the search is depth-first, classes will be looked up
1053 in the following order: Third, First, Alpha, Second, and Beta.
1055 In practice, few class modules have been seen that actually
1056 make use of MI. One nearly always chooses simple containership of
1057 one class within another over MI. That's why our Person
1058 object I<contained> a Fullname object. That doesn't mean
1061 However, there is one particular area where MI in Perl is rampant:
1062 borrowing another class's class methods. This is rather common,
1063 especially with some bundled "objectless" classes,
1064 like Exporter, DynaLoader, AutoLoader, and SelfLoader. These classes
1065 do not provide constructors; they exist only so you may inherit their
1066 class methods. (It's not entirely clear why inheritance was done
1067 here rather than traditional module importation.)
1069 For example, here is the POSIX module's @ISA:
1072 @ISA = qw(Exporter DynaLoader);
1074 The POSIX module isn't really an object module, but then,
1075 neither are Exporter or DynaLoader. They're just lending their
1076 classes' behaviours to POSIX.
1078 Why don't people use MI for object methods much? One reason is that
1079 it can have complicated side-effects. For one thing, your inheritance
1080 graph (no longer a tree) might converge back to the same base class.
1081 Although Perl guards against recursive inheritance, merely having parents
1082 who are related to each other via a common ancestor, incestuous though
1083 it sounds, is not forbidden. What if in our Third class shown above we
1084 wanted its new() method to also call both overridden constructors in its
1085 two parent classes? The SUPER notation would only find the first one.
1086 Also, what about if the Alpha and Beta classes both had a common ancestor,
1087 like Nought? If you kept climbing up the inheritance tree calling
1088 overridden methods, you'd end up calling Nought::new() twice,
1089 which might well be a bad idea.
1091 =head2 UNIVERSAL: The Root of All Objects
1093 Wouldn't it be convenient if all objects were rooted at some ultimate
1094 base class? That way you could give every object common methods without
1095 having to go and add it to each and every @ISA. Well, it turns out that
1096 you can. You don't see it, but Perl tacitly and irrevocably assumes
1097 that there's an extra element at the end of @ISA: the class UNIVERSAL.
1098 In version 5.003, there were no predefined methods there, but you could put
1099 whatever you felt like into it.
1101 However, as of version 5.004 (or some subversive releases, like 5.003_08),
1102 UNIVERSAL has some methods in it already. These are builtin to your Perl
1103 binary, so they don't take any extra time to load. Predefined methods
1104 include isa(), can(), and VERSION(). isa() tells you whether an object or
1105 class "is" another one without having to traverse the hierarchy yourself:
1107 $has_io = $fd->isa("IO::Handle");
1108 $itza_handle = IO::Socket->isa("IO::Handle");
1110 The can() method, called against that object or class, reports back
1111 whether its string argument is a callable method name in that class.
1112 In fact, it gives you back a function reference to that method:
1114 $his_print_method = $obj->can('as_string');
1116 Finally, the VERSION method checks whether the class (or the object's
1117 class) has a package global called $VERSION that's high enough, as in:
1119 Some_Module->VERSION(3.0);
1120 $his_vers = $ob->VERSION();
1122 However, we don't usually call VERSION ourselves. (Remember that an all
1123 uppercase function name is a Perl convention that indicates that the
1124 function will be automatically used by Perl in some way.) In this case,
1125 it happens when you say
1127 use Some_Module 3.0;
1129 If you wanted to add version checking to your Person class explained
1130 above, just add this to Person.pm:
1132 our $VERSION = '1.1';
1134 and then in Employee.pm could you can say
1138 And it would make sure that you have at least that version number or
1139 higher available. This is not the same as loading in that exact version
1140 number. No mechanism currently exists for concurrent installation of
1141 multiple versions of a module. Lamentably.
1143 =head1 Alternate Object Representations
1145 Nothing requires objects to be implemented as hash references. An object
1146 can be any sort of reference so long as its referent has been suitably
1147 blessed. That means scalar, array, and code references are also fair
1150 A scalar would work if the object has only one datum to hold. An array
1151 would work for most cases, but makes inheritance a bit dodgy because
1152 you have to invent new indices for the derived classes.
1154 =head2 Arrays as Objects
1156 If the user of your class honors the contract and sticks to the advertised
1157 interface, then you can change its underlying interface if you feel
1158 like it. Here's another implementation that conforms to the same
1159 interface specification. This time we'll use an array reference
1160 instead of a hash reference to represent the object.
1165 my($NAME, $AGE, $PEERS) = ( 0 .. 2 );
1167 ############################################
1168 ## the object constructor (array version) ##
1169 ############################################
1172 $self->[$NAME] = undef; # this is unnecessary
1173 $self->[$AGE] = undef; # as is this
1174 $self->[$PEERS] = []; # but this isn't, really
1181 if (@_) { $self->[$NAME] = shift }
1182 return $self->[$NAME];
1187 if (@_) { $self->[$AGE] = shift }
1188 return $self->[$AGE];
1193 if (@_) { @{ $self->[$PEERS] } = @_ }
1194 return @{ $self->[$PEERS] };
1197 1; # so the require or use succeeds
1199 You might guess that the array access would be a lot faster than the
1200 hash access, but they're actually comparable. The array is a I<little>
1201 bit faster, but not more than ten or fifteen percent, even when you
1202 replace the variables above like $AGE with literal numbers, like 1.
1203 A bigger difference between the two approaches can be found in memory use.
1204 A hash representation takes up more memory than an array representation
1205 because you have to allocate memory for the keys as well as for the values.
1206 However, it really isn't that bad, especially since as of version 5.004,
1207 memory is only allocated once for a given hash key, no matter how many
1208 hashes have that key. It's expected that sometime in the future, even
1209 these differences will fade into obscurity as more efficient underlying
1210 representations are devised.
1212 Still, the tiny edge in speed (and somewhat larger one in memory)
1213 is enough to make some programmers choose an array representation
1214 for simple classes. There's still a little problem with
1215 scalability, though, because later in life when you feel
1216 like creating subclasses, you'll find that hashes just work
1219 =head2 Closures as Objects
1221 Using a code reference to represent an object offers some fascinating
1222 possibilities. We can create a new anonymous function (closure) who
1223 alone in all the world can see the object's data. This is because we
1224 put the data into an anonymous hash that's lexically visible only to
1225 the closure we create, bless, and return as the object. This object's
1226 methods turn around and call the closure as a regular subroutine call,
1227 passing it the field we want to affect. (Yes,
1228 the double-function call is slow, but if you wanted fast, you wouldn't
1229 be using objects at all, eh? :-)
1231 Use would be similar to before:
1234 $him = Person->new();
1235 $him->name("Jason");
1237 $him->peers( [ "Norbert", "Rhys", "Phineas" ] );
1238 printf "%s is %d years old.\n", $him->name, $him->age;
1239 print "His peers are: ", join(", ", @{$him->peers}), "\n";
1241 but the implementation would be radically, perhaps even sublimely
1248 my $class = ref($that) || $that;
1256 if (@_) { $self->{$field} = shift }
1257 return $self->{$field};
1259 bless($closure, $class);
1263 sub name { &{ $_[0] }("NAME", @_[ 1 .. $#_ ] ) }
1264 sub age { &{ $_[0] }("AGE", @_[ 1 .. $#_ ] ) }
1265 sub peers { &{ $_[0] }("PEERS", @_[ 1 .. $#_ ] ) }
1269 Because this object is hidden behind a code reference, it's probably a bit
1270 mysterious to those whose background is more firmly rooted in standard
1271 procedural or object-based programming languages than in functional
1272 programming languages whence closures derive. The object
1273 created and returned by the new() method is itself not a data reference
1274 as we've seen before. It's an anonymous code reference that has within
1275 it access to a specific version (lexical binding and instantiation)
1276 of the object's data, which are stored in the private variable $self.
1277 Although this is the same function each time, it contains a different
1280 When a method like C<$him-E<gt>name("Jason")> is called, its implicit
1281 zeroth argument is the invoking object--just as it is with all method
1282 calls. But in this case, it's our code reference (something like a
1283 function pointer in C++, but with deep binding of lexical variables).
1284 There's not a lot to be done with a code reference beyond calling it, so
1285 that's just what we do when we say C<&{$_[0]}>. This is just a regular
1286 function call, not a method call. The initial argument is the string
1287 "NAME", and any remaining arguments are whatever had been passed to the
1290 Once we're executing inside the closure that had been created in new(),
1291 the $self hash reference suddenly becomes visible. The closure grabs
1292 its first argument ("NAME" in this case because that's what the name()
1293 method passed it), and uses that string to subscript into the private
1294 hash hidden in its unique version of $self.
1296 Nothing under the sun will allow anyone outside the executing method to
1297 be able to get at this hidden data. Well, nearly nothing. You I<could>
1298 single step through the program using the debugger and find out the
1299 pieces while you're in the method, but everyone else is out of luck.
1301 There, if that doesn't excite the Scheme folks, then I just don't know
1302 what will. Translation of this technique into C++, Java, or any other
1303 braindead-static language is left as a futile exercise for aficionados
1306 You could even add a bit of nosiness via the caller() function and
1307 make the closure refuse to operate unless called via its own package.
1308 This would no doubt satisfy certain fastidious concerns of programming
1309 police and related puritans.
1311 If you were wondering when Hubris, the third principle virtue of a
1312 programmer, would come into play, here you have it. (More seriously,
1313 Hubris is just the pride in craftsmanship that comes from having written
1314 a sound bit of well-designed code.)
1316 =head1 AUTOLOAD: Proxy Methods
1318 Autoloading is a way to intercept calls to undefined methods. An autoload
1319 routine may choose to create a new function on the fly, either loaded
1320 from disk or perhaps just eval()ed right there. This define-on-the-fly
1321 strategy is why it's called autoloading.
1323 But that's only one possible approach. Another one is to just
1324 have the autoloaded method itself directly provide the
1325 requested service. When used in this way, you may think
1326 of autoloaded methods as "proxy" methods.
1328 When Perl tries to call an undefined function in a particular package
1329 and that function is not defined, it looks for a function in
1330 that same package called AUTOLOAD. If one exists, it's called
1331 with the same arguments as the original function would have had.
1332 The fully-qualified name of the function is stored in that package's
1333 global variable $AUTOLOAD. Once called, the function can do anything
1334 it would like, including defining a new function by the right name, and
1335 then doing a really fancy kind of C<goto> right to it, erasing itself
1336 from the call stack.
1338 What does this have to do with objects? After all, we keep talking about
1339 functions, not methods. Well, since a method is just a function with
1340 an extra argument and some fancier semantics about where it's found,
1341 we can use autoloading for methods, too. Perl doesn't start looking
1342 for an AUTOLOAD method until it has exhausted the recursive hunt up
1343 through @ISA, though. Some programmers have even been known to define
1344 a UNIVERSAL::AUTOLOAD method to trap unresolved method calls to any
1347 =head2 Autoloaded Data Methods
1349 You probably began to get a little suspicious about the duplicated
1350 code way back earlier when we first showed you the Person class, and
1351 then later the Employee class. Each method used to access the
1352 hash fields looked virtually identical. This should have tickled
1353 that great programming virtue, Impatience, but for the time,
1354 we let Laziness win out, and so did nothing. Proxy methods can cure
1357 Instead of writing a new function every time we want a new data field,
1358 we'll use the autoload mechanism to generate (actually, mimic) methods on
1359 the fly. To verify that we're accessing a valid member, we will check
1360 against an C<_permitted> (pronounced "under-permitted") field, which
1361 is a reference to a file-scoped lexical (like a C file static) hash of permitted fields in this record
1362 called %fields. Why the underscore? For the same reason as the _CENSUS
1363 field we once used: as a marker that means "for internal use only".
1365 Here's what the module initialization code and class
1366 constructor will look like when taking this approach:
1370 our $AUTOLOAD; # it's a package global
1380 my $class = ref($that) || $that;
1382 _permitted => \%fields,
1385 bless $self, $class;
1389 If we wanted our record to have default values, we could fill those in
1390 where current we have C<undef> in the %fields hash.
1392 Notice how we saved a reference to our class data on the object itself?
1393 Remember that it's important to access class data through the object
1394 itself instead of having any method reference %fields directly, or else
1395 you won't have a decent inheritance.
1397 The real magic, though, is going to reside in our proxy method, which
1398 will handle all calls to undefined methods for objects of class Person
1399 (or subclasses of Person). It has to be called AUTOLOAD. Again, it's
1400 all caps because it's called for us implicitly by Perl itself, not by
1405 my $type = ref($self)
1406 or croak "$self is not an object";
1408 my $name = $AUTOLOAD;
1409 $name =~ s/.*://; # strip fully-qualified portion
1411 unless (exists $self->{_permitted}->{$name} ) {
1412 croak "Can't access `$name' field in class $type";
1416 return $self->{$name} = shift;
1418 return $self->{$name};
1422 Pretty nifty, eh? All we have to do to add new data fields
1423 is modify %fields. No new functions need be written.
1425 I could have avoided the C<_permitted> field entirely, but I
1426 wanted to demonstrate how to store a reference to class data on the
1427 object so you wouldn't have to access that class data
1428 directly from an object method.
1430 =head2 Inherited Autoloaded Data Methods
1432 But what about inheritance? Can we define our Employee
1433 class similarly? Yes, so long as we're careful enough.
1435 Here's how to be careful:
1440 our @ISA = qw(Person);
1449 my $class = ref($that) || $that;
1450 my $self = bless $that->SUPER::new(), $class;
1452 foreach $element (keys %fields) {
1453 $self->{_permitted}->{$element} = $fields{$element};
1455 @{$self}{keys %fields} = values %fields;
1459 Once we've done this, we don't even need to have an
1460 AUTOLOAD function in the Employee package, because
1461 we'll grab Person's version of that via inheritance,
1462 and it will all work out just fine.
1464 =head1 Metaclassical Tools
1466 Even though proxy methods can provide a more convenient approach to making
1467 more struct-like classes than tediously coding up data methods as
1468 functions, it still leaves a bit to be desired. For one thing, it means
1469 you have to handle bogus calls that you don't mean to trap via your proxy.
1470 It also means you have to be quite careful when dealing with inheritance,
1473 Perl programmers have responded to this by creating several different
1474 class construction classes. These metaclasses are classes
1475 that create other classes. A couple worth looking at are
1476 Class::Struct and Alias. These and other related metaclasses can be
1477 found in the modules directory on CPAN.
1479 =head2 Class::Struct
1481 One of the older ones is Class::Struct. In fact, its syntax and
1482 interface were sketched out long before perl5 even solidified into a
1483 real thing. What it does is provide you a way to "declare" a class
1484 as having objects whose fields are of a specific type. The function
1485 that does this is called, not surprisingly enough, struct(). Because
1486 structures or records are not base types in Perl, each time you want to
1487 create a class to provide a record-like data object, you yourself have
1488 to define a new() method, plus separate data-access methods for each of
1489 that record's fields. You'll quickly become bored with this process.
1490 The Class::Struct::struct() function alleviates this tedium.
1492 Here's a simple example of using it:
1494 use Class::Struct qw(struct);
1495 use Jobbie; # user-defined; see below
1500 profession => Jobbie, # calls Jobbie->new()
1506 $ob->many(0, "here");
1507 $ob->many(1, "you");
1509 print "Just set: ", $ob->many(2), "\n";
1511 $ob->profession->salary(10_000);
1513 You can declare types in the struct to be basic Perl types, or
1514 user-defined types (classes). User types will be initialized by calling
1515 that class's new() method.
1517 Here's a real-world example of using struct generation. Let's say you
1518 wanted to override Perl's idea of gethostbyname() and gethostbyaddr() so
1519 that they would return objects that acted like C structures. We don't
1520 care about high-falutin' OO gunk. All we want is for these objects to
1521 act like structs in the C sense.
1525 $h = gethostbyname("perl.com"); # object return
1526 printf "perl.com's real name is %s, address %s\n",
1527 $h->name, inet_ntoa($h->addr);
1529 Here's how to do this using the Class::Struct module.
1530 The crux is going to be this call:
1532 struct 'Net::hostent' => [ # note bracket
1540 Which creates object methods of those names and types.
1541 It even creates a new() method for us.
1543 We could also have implemented our object this way:
1545 struct 'Net::hostent' => { # note brace
1553 and then Class::Struct would have used an anonymous hash as the object
1554 type, instead of an anonymous array. The array is faster and smaller,
1555 but the hash works out better if you eventually want to do inheritance.
1556 Since for this struct-like object we aren't planning on inheritance,
1557 this time we'll opt for better speed and size over better flexibility.
1559 Here's the whole implementation:
1561 package Net::hostent;
1566 our @EXPORT = qw(gethostbyname gethostbyaddr gethost);
1567 our @EXPORT_OK = qw(
1569 $h_addrtype $h_length
1570 @h_addr_list $h_addr
1572 our %EXPORT_TAGS = ( FIELDS => [ @EXPORT_OK, @EXPORT ] );
1576 # Class::Struct forbids use of @ISA
1577 sub import { goto &Exporter::import }
1579 use Class::Struct qw(struct);
1580 struct 'Net::hostent' => [
1588 sub addr { shift->addr_list->[0] }
1592 my $hob = new(); # Class::Struct made this!
1593 $h_name = $hob->[0] = $_[0];
1594 @h_aliases = @{ $hob->[1] } = split ' ', $_[1];
1595 $h_addrtype = $hob->[2] = $_[2];
1596 $h_length = $hob->[3] = $_[3];
1598 @h_addr_list = @{ $hob->[4] } = @_[ (4 .. $#_) ];
1602 sub gethostbyname ($) { populate(CORE::gethostbyname(shift)) }
1604 sub gethostbyaddr ($;$) {
1605 my ($addr, $addrtype);
1607 require Socket unless @_;
1608 $addrtype = @_ ? shift : Socket::AF_INET();
1609 populate(CORE::gethostbyaddr($addr, $addrtype))
1613 if ($_[0] =~ /^\d+(?:\.\d+(?:\.\d+(?:\.\d+)?)?)?$/) {
1615 &gethostbyaddr(Socket::inet_aton(shift));
1623 We've snuck in quite a fair bit of other concepts besides just dynamic
1624 class creation, like overriding core functions, import/export bits,
1625 function prototyping, short-cut function call via C<&whatever>, and
1626 function replacement with C<goto &whatever>. These all mostly make
1627 sense from the perspective of a traditional module, but as you can see,
1628 we can also use them in an object module.
1630 You can look at other object-based, struct-like overrides of core
1631 functions in the 5.004 release of Perl in File::stat, Net::hostent,
1632 Net::netent, Net::protoent, Net::servent, Time::gmtime, Time::localtime,
1633 User::grent, and User::pwent. These modules have a final component
1634 that's all lowercase, by convention reserved for compiler pragmas,
1635 because they affect the compilation and change a builtin function.
1636 They also have the type names that a C programmer would most expect.
1638 =head2 Data Members as Variables
1640 If you're used to C++ objects, then you're accustomed to being able to
1641 get at an object's data members as simple variables from within a method.
1642 The Alias module provides for this, as well as a good bit more, such
1643 as the possibility of private methods that the object can call but folks
1644 outside the class cannot.
1646 Here's an example of creating a Person using the Alias module.
1647 When you update these magical instance variables, you automatically
1648 update value fields in the hash. Convenient, eh?
1652 # this is the same as before...
1655 my $class = ref($that) || $that;
1661 bless($self, $class);
1666 our ($NAME, $AGE, $PEERS);
1669 my $self = attr shift;
1670 if (@_) { $NAME = shift; }
1675 my $self = attr shift;
1676 if (@_) { $AGE = shift; }
1681 my $self = attr shift;
1682 if (@_) { @PEERS = @_; }
1687 my $self = attr shift;
1688 return sprintf "Hi, I'm %s, age %d, working with %s",
1689 $NAME, $AGE, join(", ", @PEERS);
1692 sub happy_birthday {
1693 my $self = attr shift;
1697 The need for the C<our> declaration is because what Alias does
1698 is play with package globals with the same name as the fields. To use
1699 globals while C<use strict> is in effect, you have to predeclare them.
1700 These package variables are localized to the block enclosing the attr()
1701 call just as if you'd used a local() on them. However, that means that
1702 they're still considered global variables with temporary values, just
1703 as with any other local().
1705 It would be nice to combine Alias with
1706 something like Class::Struct or Class::MethodMaker.
1710 =head2 Object Terminology
1712 In the various OO literature, it seems that a lot of different words
1713 are used to describe only a few different concepts. If you're not
1714 already an object programmer, then you don't need to worry about all
1715 these fancy words. But if you are, then you might like to know how to
1716 get at the same concepts in Perl.
1718 For example, it's common to call an object an I<instance> of a class
1719 and to call those objects' methods I<instance methods>. Data fields
1720 peculiar to each object are often called I<instance data> or I<object
1721 attributes>, and data fields common to all members of that class are
1722 I<class data>, I<class attributes>, or I<static data members>.
1724 Also, I<base class>, I<generic class>, and I<superclass> all describe
1725 the same notion, whereas I<derived class>, I<specific class>, and
1726 I<subclass> describe the other related one.
1728 C++ programmers have I<static methods> and I<virtual methods>,
1729 but Perl only has I<class methods> and I<object methods>.
1730 Actually, Perl only has methods. Whether a method gets used
1731 as a class or object method is by usage only. You could accidentally
1732 call a class method (one expecting a string argument) on an
1733 object (one expecting a reference), or vice versa.
1735 From the C++ perspective, all methods in Perl are virtual.
1736 This, by the way, is why they are never checked for function
1737 prototypes in the argument list as regular builtin and user-defined
1740 Because a class is itself something of an object, Perl's classes can be
1741 taken as describing both a "class as meta-object" (also called I<object
1742 factory>) philosophy and the "class as type definition" (I<declaring>
1743 behaviour, not I<defining> mechanism) idea. C++ supports the latter
1744 notion, but not the former.
1748 The following manpages will doubtless provide more
1749 background for this one:
1758 L<perlboot> is a kinder, gentler introduction to object-oriented
1761 L<perltooc> provides more detail on class data.
1763 Some modules which might prove interesting are Class::Accessor,
1764 Class::Class, Class::Contract, Class::Data::Inheritable,
1765 Class::MethodMaker and Tie::SecureHash
1768 =head1 AUTHOR AND COPYRIGHT
1770 Copyright (c) 1997, 1998 Tom Christiansen
1771 All rights reserved.
1773 This documentation is free; you can redistribute it and/or modify it
1774 under the same terms as Perl itself.
1776 Irrespective of its distribution, all code examples in this file
1777 are hereby placed into the public domain. You are permitted and
1778 encouraged to use this code in your own programs for fun
1779 or for profit as you see fit. A simple comment in the code giving
1780 credit would be courteous but is not required.
1784 =head2 Acknowledgments
1795 and many others for their helpful comments.