Commit | Line | Data |
a0d0e21e |
1 | =head1 NAME |
2 | |
f102b883 |
3 | perlmod - Perl modules (packages and symbol tables) |
a0d0e21e |
4 | |
5 | =head1 DESCRIPTION |
6 | |
7 | =head2 Packages |
8 | |
748a9306 |
9 | Perl provides a mechanism for alternative namespaces to protect packages |
5a964f20 |
10 | from stomping on each other's variables. In fact, there's really no such |
11 | thing as a global variable in Perl (although some identifiers default |
12 | to the main package instead of the current one). The package statement |
13 | declares the compilation unit as |
f102b883 |
14 | being in the given namespace. The scope of the package declaration |
15 | is from the declaration itself through the end of the enclosing block, |
16 | C<eval>, C<sub>, or end of file, whichever comes first (the same scope |
17 | as the my() and local() operators). All further unqualified dynamic |
5a964f20 |
18 | identifiers will be in this namespace. A package statement only affects |
19 | dynamic variables--including those you've used local() on--but |
f102b883 |
20 | I<not> lexical variables created with my(). Typically it would be |
21 | the first declaration in a file to be included by the C<require> or |
22 | C<use> operator. You can switch into a package in more than one place; |
5a964f20 |
23 | it merely influences which symbol table is used by the compiler for the |
f102b883 |
24 | rest of that block. You can refer to variables and filehandles in other |
25 | packages by prefixing the identifier with the package name and a double |
26 | colon: C<$Package::Variable>. If the package name is null, the C<main> |
27 | package is assumed. That is, C<$::sail> is equivalent to C<$main::sail>. |
a0d0e21e |
28 | |
d3ebb66b |
29 | The old package delimiter was a single quote, but double colon is now the |
30 | preferred delimiter, in part because it's more readable to humans, and |
31 | in part because it's more readable to B<emacs> macros. It also makes C++ |
32 | programmers feel like they know what's going on--as opposed to using the |
33 | single quote as separator, which was there to make Ada programmers feel |
34 | like they knew what's going on. Because the old-fashioned syntax is still |
35 | supported for backwards compatibility, if you try to use a string like |
36 | C<"This is $owner's house">, you'll be accessing C<$owner::s>; that is, |
37 | the $s variable in package C<owner>, which is probably not what you meant. |
38 | Use braces to disambiguate, as in C<"This is ${owner}'s house">. |
a0d0e21e |
39 | |
40 | Packages may be nested inside other packages: C<$OUTER::INNER::var>. This |
41 | implies nothing about the order of name lookups, however. All symbols |
42 | are either local to the current package, or must be fully qualified |
43 | from the outer package name down. For instance, there is nowhere |
44 | within package C<OUTER> that C<$INNER::var> refers to C<$OUTER::INNER::var>. |
45 | It would treat package C<INNER> as a totally separate global package. |
46 | |
47 | Only identifiers starting with letters (or underscore) are stored in a |
cb1a09d0 |
48 | package's symbol table. All other symbols are kept in package C<main>, |
5a964f20 |
49 | including all of the punctuation variables like $_. In addition, when |
50 | unqualified, the identifiers STDIN, STDOUT, STDERR, ARGV, ARGVOUT, ENV, |
51 | INC, and SIG are forced to be in package C<main>, even when used for other |
52 | purposes than their builtin one. Note also that, if you have a package |
53 | called C<m>, C<s>, or C<y>, then you can't use the qualified form of an |
54 | identifier because it will be interpreted instead as a pattern match, |
55 | a substitution, or a transliteration. |
a0d0e21e |
56 | |
57 | (Variables beginning with underscore used to be forced into package |
58 | main, but we decided it was more useful for package writers to be able |
cb1a09d0 |
59 | to use leading underscore to indicate private variables and method names. |
60 | $_ is still global though.) |
a0d0e21e |
61 | |
62 | Eval()ed strings are compiled in the package in which the eval() was |
63 | compiled. (Assignments to C<$SIG{}>, however, assume the signal |
748a9306 |
64 | handler specified is in the C<main> package. Qualify the signal handler |
a0d0e21e |
65 | name if you wish to have a signal handler in a package.) For an |
66 | example, examine F<perldb.pl> in the Perl library. It initially switches |
67 | to the C<DB> package so that the debugger doesn't interfere with variables |
68 | in the script you are trying to debug. At various points, however, it |
69 | temporarily switches back to the C<main> package to evaluate various |
70 | expressions in the context of the C<main> package (or wherever you came |
71 | from). See L<perldebug>. |
72 | |
f102b883 |
73 | The special symbol C<__PACKAGE__> contains the current package, but cannot |
74 | (easily) be used to construct variables. |
75 | |
5f05dabc |
76 | See L<perlsub> for other scoping issues related to my() and local(), |
f102b883 |
77 | and L<perlref> regarding closures. |
cb1a09d0 |
78 | |
a0d0e21e |
79 | =head2 Symbol Tables |
80 | |
aa689395 |
81 | The symbol table for a package happens to be stored in the hash of that |
82 | name with two colons appended. The main symbol table's name is thus |
83 | C<%main::>, or C<%::> for short. Likewise symbol table for the nested |
84 | package mentioned earlier is named C<%OUTER::INNER::>. |
85 | |
86 | The value in each entry of the hash is what you are referring to when you |
87 | use the C<*name> typeglob notation. In fact, the following have the same |
88 | effect, though the first is more efficient because it does the symbol |
89 | table lookups at compile time: |
a0d0e21e |
90 | |
f102b883 |
91 | local *main::foo = *main::bar; |
92 | local $main::{foo} = $main::{bar}; |
a0d0e21e |
93 | |
94 | You can use this to print out all the variables in a package, for |
5a964f20 |
95 | instance. The standard F<dumpvar.pl> library and the CPAN module |
96 | Devel::Symdump make use of this. |
a0d0e21e |
97 | |
cb1a09d0 |
98 | Assignment to a typeglob performs an aliasing operation, i.e., |
a0d0e21e |
99 | |
100 | *dick = *richard; |
101 | |
5a964f20 |
102 | causes variables, subroutines, formats, and file and directory handles |
103 | accessible via the identifier C<richard> also to be accessible via the |
104 | identifier C<dick>. If you want to alias only a particular variable or |
105 | subroutine, you can assign a reference instead: |
a0d0e21e |
106 | |
107 | *dick = \$richard; |
108 | |
5a964f20 |
109 | Which makes $richard and $dick the same variable, but leaves |
a0d0e21e |
110 | @richard and @dick as separate arrays. Tricky, eh? |
111 | |
cb1a09d0 |
112 | This mechanism may be used to pass and return cheap references |
113 | into or from subroutines if you won't want to copy the whole |
5a964f20 |
114 | thing. It only works when assigning to dynamic variables, not |
115 | lexicals. |
cb1a09d0 |
116 | |
5a964f20 |
117 | %some_hash = (); # can't be my() |
cb1a09d0 |
118 | *some_hash = fn( \%another_hash ); |
119 | sub fn { |
120 | local *hashsym = shift; |
121 | # now use %hashsym normally, and you |
122 | # will affect the caller's %another_hash |
123 | my %nhash = (); # do what you want |
5f05dabc |
124 | return \%nhash; |
cb1a09d0 |
125 | } |
126 | |
5f05dabc |
127 | On return, the reference will overwrite the hash slot in the |
cb1a09d0 |
128 | symbol table specified by the *some_hash typeglob. This |
c36e9b62 |
129 | is a somewhat tricky way of passing around references cheaply |
cb1a09d0 |
130 | when you won't want to have to remember to dereference variables |
131 | explicitly. |
132 | |
133 | Another use of symbol tables is for making "constant" scalars. |
134 | |
135 | *PI = \3.14159265358979; |
136 | |
137 | Now you cannot alter $PI, which is probably a good thing all in all. |
5a964f20 |
138 | This isn't the same as a constant subroutine, which is subject to |
139 | optimization at compile-time. This isn't. A constant subroutine is one |
140 | prototyped to take no arguments and to return a constant expression. |
141 | See L<perlsub> for details on these. The C<use constant> pragma is a |
142 | convenient shorthand for these. |
cb1a09d0 |
143 | |
55497cff |
144 | You can say C<*foo{PACKAGE}> and C<*foo{NAME}> to find out what name and |
145 | package the *foo symbol table entry comes from. This may be useful |
5a964f20 |
146 | in a subroutine that gets passed typeglobs as arguments: |
55497cff |
147 | |
148 | sub identify_typeglob { |
149 | my $glob = shift; |
150 | print 'You gave me ', *{$glob}{PACKAGE}, '::', *{$glob}{NAME}, "\n"; |
151 | } |
152 | identify_typeglob *foo; |
153 | identify_typeglob *bar::baz; |
154 | |
155 | This prints |
156 | |
157 | You gave me main::foo |
158 | You gave me bar::baz |
159 | |
160 | The *foo{THING} notation can also be used to obtain references to the |
161 | individual elements of *foo, see L<perlref>. |
162 | |
a0d0e21e |
163 | =head2 Package Constructors and Destructors |
164 | |
165 | There are two special subroutine definitions that function as package |
166 | constructors and destructors. These are the C<BEGIN> and C<END> |
167 | routines. The C<sub> is optional for these routines. |
168 | |
f102b883 |
169 | A C<BEGIN> subroutine is executed as soon as possible, that is, the moment |
170 | it is completely defined, even before the rest of the containing file |
171 | is parsed. You may have multiple C<BEGIN> blocks within a file--they |
172 | will execute in order of definition. Because a C<BEGIN> block executes |
173 | immediately, it can pull in definitions of subroutines and such from other |
174 | files in time to be visible to the rest of the file. Once a C<BEGIN> |
175 | has run, it is immediately undefined and any code it used is returned to |
176 | Perl's memory pool. This means you can't ever explicitly call a C<BEGIN>. |
a0d0e21e |
177 | |
5a964f20 |
178 | An C<END> subroutine is executed as late as possible, that is, when |
179 | the interpreter is being exited, even if it is exiting as a result of |
180 | a die() function. (But not if it's polymorphing into another program |
181 | via C<exec>, or being blown out of the water by a signal--you have to |
182 | trap that yourself (if you can).) You may have multiple C<END> blocks |
183 | within a file--they will execute in reverse order of definition; that is: |
184 | last in, first out (LIFO). |
a0d0e21e |
185 | |
5a964f20 |
186 | Inside an C<END> subroutine, C<$?> contains the value that the script is |
c36e9b62 |
187 | going to pass to C<exit()>. You can modify C<$?> to change the exit |
f102b883 |
188 | value of the script. Beware of changing C<$?> by accident (e.g. by |
c36e9b62 |
189 | running something via C<system>). |
190 | |
5a964f20 |
191 | Note that when you use the B<-n> and B<-p> switches to Perl, C<BEGIN> and |
192 | C<END> work just as they do in B<awk>, as a degenerate case. As currently |
193 | implemented (and subject to change, since its inconvenient at best), |
194 | both C<BEGIN> I<and> C<END> blocks are run when you use the B<-c> switch |
195 | for a compile-only syntax check, although your main code is not. |
a0d0e21e |
196 | |
197 | =head2 Perl Classes |
198 | |
4633a7c4 |
199 | There is no special class syntax in Perl, but a package may function |
5a964f20 |
200 | as a class if it provides subroutines to act as methods. Such a |
201 | package may also derive some of its methods from another class (package) |
202 | by listing the other package name in its global @ISA array (which |
203 | must be a package global, not a lexical). |
4633a7c4 |
204 | |
f102b883 |
205 | For more on this, see L<perltoot> and L<perlobj>. |
a0d0e21e |
206 | |
207 | =head2 Perl Modules |
208 | |
c07a80fd |
209 | A module is just a package that is defined in a library file of |
a0d0e21e |
210 | the same name, and is designed to be reusable. It may do this by |
211 | providing a mechanism for exporting some of its symbols into the symbol |
212 | table of any package using it. Or it may function as a class |
213 | definition and make its semantics available implicitly through method |
214 | calls on the class and its objects, without explicit exportation of any |
215 | symbols. Or it can do a little of both. |
216 | |
9607fc9c |
217 | For example, to start a normal module called Some::Module, create |
218 | a file called Some/Module.pm and start with this template: |
219 | |
220 | package Some::Module; # assumes Some/Module.pm |
221 | |
222 | use strict; |
223 | |
224 | BEGIN { |
225 | use Exporter (); |
226 | use vars qw($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS); |
227 | |
228 | # set the version for version checking |
229 | $VERSION = 1.00; |
230 | # if using RCS/CVS, this may be preferred |
231 | $VERSION = do { my @r = (q$Revision: 2.21 $ =~ /\d+/g); sprintf "%d."."%02d" x $#r, @r }; # must be all one line, for MakeMaker |
232 | |
233 | @ISA = qw(Exporter); |
234 | @EXPORT = qw(&func1 &func2 &func4); |
235 | %EXPORT_TAGS = ( ); # eg: TAG => [ qw!name1 name2! ], |
236 | |
237 | # your exported package globals go here, |
238 | # as well as any optionally exported functions |
239 | @EXPORT_OK = qw($Var1 %Hashit &func3); |
240 | } |
241 | use vars @EXPORT_OK; |
242 | |
243 | # non-exported package globals go here |
244 | use vars qw(@more $stuff); |
245 | |
246 | # initalize package globals, first exported ones |
247 | $Var1 = ''; |
248 | %Hashit = (); |
249 | |
250 | # then the others (which are still accessible as $Some::Module::stuff) |
251 | $stuff = ''; |
252 | @more = (); |
253 | |
254 | # all file-scoped lexicals must be created before |
255 | # the functions below that use them. |
256 | |
257 | # file-private lexicals go here |
258 | my $priv_var = ''; |
259 | my %secret_hash = (); |
260 | |
261 | # here's a file-private function as a closure, |
262 | # callable as &$priv_func; it cannot be prototyped. |
263 | my $priv_func = sub { |
264 | # stuff goes here. |
265 | }; |
266 | |
267 | # make all your functions, whether exported or not; |
268 | # remember to put something interesting in the {} stubs |
269 | sub func1 {} # no prototype |
270 | sub func2() {} # proto'd void |
271 | sub func3($$) {} # proto'd to 2 scalars |
272 | |
273 | # this one isn't exported, but could be called! |
274 | sub func4(\%) {} # proto'd to 1 hash ref |
275 | |
276 | END { } # module clean-up code here (global destructor) |
4633a7c4 |
277 | |
278 | Then go on to declare and use your variables in functions |
279 | without any qualifications. |
f102b883 |
280 | See L<Exporter> and the L<perlmodlib> for details on |
4633a7c4 |
281 | mechanics and style issues in module creation. |
282 | |
283 | Perl modules are included into your program by saying |
a0d0e21e |
284 | |
285 | use Module; |
286 | |
287 | or |
288 | |
289 | use Module LIST; |
290 | |
291 | This is exactly equivalent to |
292 | |
5a964f20 |
293 | BEGIN { require Module; import Module; } |
a0d0e21e |
294 | |
295 | or |
296 | |
5a964f20 |
297 | BEGIN { require Module; import Module LIST; } |
a0d0e21e |
298 | |
cb1a09d0 |
299 | As a special case |
300 | |
301 | use Module (); |
302 | |
303 | is exactly equivalent to |
304 | |
5a964f20 |
305 | BEGIN { require Module; } |
cb1a09d0 |
306 | |
a0d0e21e |
307 | All Perl module files have the extension F<.pm>. C<use> assumes this so |
308 | that you don't have to spell out "F<Module.pm>" in quotes. This also |
309 | helps to differentiate new modules from old F<.pl> and F<.ph> files. |
310 | Module names are also capitalized unless they're functioning as pragmas, |
311 | "Pragmas" are in effect compiler directives, and are sometimes called |
312 | "pragmatic modules" (or even "pragmata" if you're a classicist). |
313 | |
5a964f20 |
314 | The two statements: |
315 | |
316 | require SomeModule; |
317 | require "SomeModule.pm"; |
318 | |
319 | differ from each other in two ways. In the first case, any double |
320 | colons in the module name, such as C<Some::Module>, are translated |
321 | into your system's directory separator, usually "/". The second |
322 | case does not, and would have to be specified literally. The other difference |
323 | is that seeing the first C<require> clues in the compiler that uses of |
324 | indirect object notation involving "SomeModule", as in C<$ob = purge SomeModule>, |
325 | are method calls, not function calls. (Yes, this really can make a difference.) |
326 | |
a0d0e21e |
327 | Because the C<use> statement implies a C<BEGIN> block, the importation |
328 | of semantics happens at the moment the C<use> statement is compiled, |
329 | before the rest of the file is compiled. This is how it is able |
330 | to function as a pragma mechanism, and also how modules are able to |
331 | declare subroutines that are then visible as list operators for |
332 | the rest of the current file. This will not work if you use C<require> |
cb1a09d0 |
333 | instead of C<use>. With require you can get into this problem: |
a0d0e21e |
334 | |
335 | require Cwd; # make Cwd:: accessible |
54310121 |
336 | $here = Cwd::getcwd(); |
a0d0e21e |
337 | |
5f05dabc |
338 | use Cwd; # import names from Cwd:: |
a0d0e21e |
339 | $here = getcwd(); |
340 | |
341 | require Cwd; # make Cwd:: accessible |
342 | $here = getcwd(); # oops! no main::getcwd() |
343 | |
5a964f20 |
344 | In general, C<use Module ()> is recommended over C<require Module>, |
345 | because it determines module availability at compile time, not in the |
346 | middle of your program's execution. An exception would be if two modules |
347 | each tried to C<use> each other, and each also called a function from |
348 | that other module. In that case, it's easy to use C<require>s instead. |
cb1a09d0 |
349 | |
a0d0e21e |
350 | Perl packages may be nested inside other package names, so we can have |
351 | package names containing C<::>. But if we used that package name |
352 | directly as a filename it would makes for unwieldy or impossible |
353 | filenames on some systems. Therefore, if a module's name is, say, |
354 | C<Text::Soundex>, then its definition is actually found in the library |
355 | file F<Text/Soundex.pm>. |
356 | |
357 | Perl modules always have a F<.pm> file, but there may also be dynamically |
358 | linked executables or autoloaded subroutine definitions associated with |
359 | the module. If so, these will be entirely transparent to the user of |
360 | the module. It is the responsibility of the F<.pm> file to load (or |
361 | arrange to autoload) any additional functionality. The POSIX module |
362 | happens to do both dynamic loading and autoloading, but the user can |
5f05dabc |
363 | say just C<use POSIX> to get it all. |
a0d0e21e |
364 | |
f102b883 |
365 | For more information on writing extension modules, see L<perlxstut> |
a0d0e21e |
366 | and L<perlguts>. |
367 | |
f102b883 |
368 | =head1 SEE ALSO |
cb1a09d0 |
369 | |
f102b883 |
370 | See L<perlmodlib> for general style issues related to building Perl |
371 | modules and classes as well as descriptions of the standard library and |
372 | CPAN, L<Exporter> for how Perl's standard import/export mechanism works, |
373 | L<perltoot> for an in-depth tutorial on creating classes, L<perlobj> |
374 | for a hard-core reference document on objects, and L<perlsub> for an |
375 | explanation of functions and scoping. |