3 perlxs - XS language reference manual
9 XS is a language used to create an extension interface
10 between Perl and some C library which one wishes to use with
11 Perl. The XS interface is combined with the library to
12 create a new library which can be linked to Perl. An B<XSUB>
13 is a function in the XS language and is the core component
14 of the Perl application interface.
16 The XS compiler is called B<xsubpp>. This compiler will embed
17 the constructs necessary to let an XSUB, which is really a C
18 function in disguise, manipulate Perl values and creates the
19 glue necessary to let Perl access the XSUB. The compiler
20 uses B<typemaps> to determine how to map C function parameters
21 and variables to Perl values. The default typemap handles
22 many common C types. A supplement typemap must be created
23 to handle special structures and types for the library being
26 See L<perlxstut> for a tutorial on the whole extension creation process.
30 Many of the examples which follow will concentrate on creating an interface
31 between Perl and the ONC+ RPC bind library functions. The rpcb_gettime()
32 function is used to demonstrate many features of the XS language. This
33 function has two parameters; the first is an input parameter and the second
34 is an output parameter. The function also returns a status value.
36 bool_t rpcb_gettime(const char *host, time_t *timep);
38 From C this function will be called with the following
44 status = rpcb_gettime( "localhost", &timep );
46 If an XSUB is created to offer a direct translation between this function
47 and Perl, then this XSUB will be used from Perl with the following code.
48 The $status and $timep variables will contain the output of the function.
51 $status = rpcb_gettime( "localhost", $timep );
53 The following XS file shows an XS subroutine, or XSUB, which
54 demonstrates one possible interface to the rpcb_gettime()
55 function. This XSUB represents a direct translation between
56 C and Perl and so preserves the interface even from Perl.
57 This XSUB will be invoked from Perl with the usage shown
58 above. Note that the first three #include statements, for
59 C<EXTERN.h>, C<perl.h>, and C<XSUB.h>, will always be present at the
60 beginning of an XS file. This approach and others will be
61 expanded later in this document.
68 MODULE = RPC PACKAGE = RPC
71 rpcb_gettime(host,timep)
77 Any extension to Perl, including those containing XSUBs,
78 should have a Perl module to serve as the bootstrap which
79 pulls the extension into Perl. This module will export the
80 extension's functions and variables to the Perl program and
81 will cause the extension's XSUBs to be linked into Perl.
82 The following module will be used for most of the examples
83 in this document and should be used from Perl with the C<use>
84 command as shown earlier. Perl modules are explained in
85 more detail later in this document.
91 @ISA = qw(Exporter DynaLoader);
92 @EXPORT = qw( rpcb_gettime );
97 Throughout this document a variety of interfaces to the rpcb_gettime()
98 XSUB will be explored. The XSUBs will take their parameters in different
99 orders or will take different numbers of parameters. In each case the
100 XSUB is an abstraction between Perl and the real C rpcb_gettime()
101 function, and the XSUB must always ensure that the real rpcb_gettime()
102 function is called with the correct parameters. This abstraction will
103 allow the programmer to create a more Perl-like interface to the C
106 =head2 The Anatomy of an XSUB
108 The following XSUB allows a Perl program to access a C library function
109 called sin(). The XSUB will imitate the C function which takes a single
110 argument and returns a single value.
116 When using C pointers the indirection operator C<*> should be considered
117 part of the type and the address operator C<&> should be considered part of
118 the variable, as is demonstrated in the rpcb_gettime() function above. See
119 the section on typemaps for more about handling qualifiers and unary
120 operators in C types.
122 The function name and the return type must be placed on
131 The function body may be indented or left-adjusted. The following example
132 shows a function with its body left-adjusted. Most examples in this
133 document will indent the body.
141 =head2 The Argument Stack
143 The argument stack is used to store the values which are
144 sent as parameters to the XSUB and to store the XSUB's
145 return value. In reality all Perl functions keep their
146 values on this stack at the same time, each limited to its
147 own range of positions on the stack. In this document the
148 first position on that stack which belongs to the active
149 function will be referred to as position 0 for that function.
151 XSUBs refer to their stack arguments with the macro B<ST(x)>, where I<x>
152 refers to a position in this XSUB's part of the stack. Position 0 for that
153 function would be known to the XSUB as ST(0). The XSUB's incoming
154 parameters and outgoing return values always begin at ST(0). For many
155 simple cases the B<xsubpp> compiler will generate the code necessary to
156 handle the argument stack by embedding code fragments found in the
157 typemaps. In more complex cases the programmer must supply the code.
159 =head2 The RETVAL Variable
161 The RETVAL variable is a magic variable which always matches
162 the return type of the C library function. The B<xsubpp> compiler will
163 supply this variable in each XSUB and by default will use it to hold the
164 return value of the C library function being called. In simple cases the
165 value of RETVAL will be placed in ST(0) of the argument stack where it can
166 be received by Perl as the return value of the XSUB.
168 If the XSUB has a return type of C<void> then the compiler will
169 not supply a RETVAL variable for that function. When using
170 the PPCODE: directive the RETVAL variable is not needed, unless used
173 If PPCODE: directive is not used, C<void> return value should be used
174 only for subroutines which do not return a value, I<even if> CODE:
175 directive is used which sets ST(0) explicitly.
177 Older versions of this document recommended to use C<void> return
178 value in such cases. It was discovered that this could lead to
179 segfaults in cases when XSUB was I<truely> C<void>. This practice is
180 now deprecated, and may be not supported at some future version. Use
181 the return value C<SV *> in such cases. (Currently C<xsubpp> contains
182 some heuristic code which tries to disambiguate between "truely-void"
183 and "old-practice-declared-as-void" functions. Hence your code is at
184 mercy of this heuristics unless you use C<SV *> as return value.)
186 =head2 The MODULE Keyword
188 The MODULE keyword is used to start the XS code and to
189 specify the package of the functions which are being
190 defined. All text preceding the first MODULE keyword is
191 considered C code and is passed through to the output
192 untouched. Every XS module will have a bootstrap function
193 which is used to hook the XSUBs into Perl. The package name
194 of this bootstrap function will match the value of the last
195 MODULE statement in the XS source files. The value of
196 MODULE should always remain constant within the same XS
197 file, though this is not required.
199 The following example will start the XS code and will place
200 all functions in a package named RPC.
204 =head2 The PACKAGE Keyword
206 When functions within an XS source file must be separated into packages
207 the PACKAGE keyword should be used. This keyword is used with the MODULE
208 keyword and must follow immediately after it when used.
210 MODULE = RPC PACKAGE = RPC
212 [ XS code in package RPC ]
214 MODULE = RPC PACKAGE = RPCB
216 [ XS code in package RPCB ]
218 MODULE = RPC PACKAGE = RPC
220 [ XS code in package RPC ]
222 Although this keyword is optional and in some cases provides redundant
223 information it should always be used. This keyword will ensure that the
224 XSUBs appear in the desired package.
226 =head2 The PREFIX Keyword
228 The PREFIX keyword designates prefixes which should be
229 removed from the Perl function names. If the C function is
230 C<rpcb_gettime()> and the PREFIX value is C<rpcb_> then Perl will
231 see this function as C<gettime()>.
233 This keyword should follow the PACKAGE keyword when used.
234 If PACKAGE is not used then PREFIX should follow the MODULE
237 MODULE = RPC PREFIX = rpc_
239 MODULE = RPC PACKAGE = RPCB PREFIX = rpcb_
241 =head2 The OUTPUT: Keyword
243 The OUTPUT: keyword indicates that certain function parameters should be
244 updated (new values made visible to Perl) when the XSUB terminates or that
245 certain values should be returned to the calling Perl function. For
246 simple functions, such as the sin() function above, the RETVAL variable is
247 automatically designated as an output value. In more complex functions
248 the B<xsubpp> compiler will need help to determine which variables are output
251 This keyword will normally be used to complement the CODE: keyword.
252 The RETVAL variable is not recognized as an output variable when the
253 CODE: keyword is present. The OUTPUT: keyword is used in this
254 situation to tell the compiler that RETVAL really is an output
257 The OUTPUT: keyword can also be used to indicate that function parameters
258 are output variables. This may be necessary when a parameter has been
259 modified within the function and the programmer would like the update to
263 rpcb_gettime(host,timep)
269 The OUTPUT: keyword will also allow an output parameter to
270 be mapped to a matching piece of code rather than to a
274 rpcb_gettime(host,timep)
278 timep sv_setnv(ST(1), (double)timep);
280 B<xsubpp> emits an automatic C<SvSETMAGIC()> for all parameters in the
281 OUTPUT section of the XSUB, except RETVAL. This is the usually desired
282 behavior, as it takes care of properly invoking 'set' magic on output
283 parameters (needed for hash or array element parameters that must be
284 created if they didn't exist). If for some reason, this behavior is
285 not desired, the OUTPUT section may contain a C<SETMAGIC: DISABLE> line
286 to disable it for the remainder of the parameters in the OUTPUT section.
287 Likewise, C<SETMAGIC: ENABLE> can be used to reenable it for the
288 remainder of the OUTPUT section. See L<perlguts> for more details
291 =head2 The CODE: Keyword
293 This keyword is used in more complicated XSUBs which require
294 special handling for the C function. The RETVAL variable is
295 available but will not be returned unless it is specified
296 under the OUTPUT: keyword.
298 The following XSUB is for a C function which requires special handling of
299 its parameters. The Perl usage is given first.
301 $status = rpcb_gettime( "localhost", $timep );
306 rpcb_gettime(host,timep)
310 RETVAL = rpcb_gettime( host, &timep );
315 =head2 The INIT: Keyword
317 The INIT: keyword allows initialization to be inserted into the XSUB before
318 the compiler generates the call to the C function. Unlike the CODE: keyword
319 above, this keyword does not affect the way the compiler handles RETVAL.
322 rpcb_gettime(host,timep)
326 printf("# Host is %s\n", host );
330 =head2 The NO_INIT Keyword
332 The NO_INIT keyword is used to indicate that a function
333 parameter is being used only as an output value. The B<xsubpp>
334 compiler will normally generate code to read the values of
335 all function parameters from the argument stack and assign
336 them to C variables upon entry to the function. NO_INIT
337 will tell the compiler that some parameters will be used for
338 output rather than for input and that they will be handled
339 before the function terminates.
341 The following example shows a variation of the rpcb_gettime() function.
342 This function uses the timep variable only as an output variable and does
343 not care about its initial contents.
346 rpcb_gettime(host,timep)
348 time_t &timep = NO_INIT
352 =head2 Initializing Function Parameters
354 Function parameters are normally initialized with their
355 values from the argument stack. The typemaps contain the
356 code segments which are used to transfer the Perl values to
357 the C parameters. The programmer, however, is allowed to
358 override the typemaps and supply alternate initialization
361 The following code demonstrates how to supply initialization code for
362 function parameters. The initialization code is eval'd by the compiler
363 before it is added to the output so anything which should be interpreted
364 literally, such as double quotes, must be protected with backslashes.
367 rpcb_gettime(host,timep)
368 char *host = (char *)SvPV(ST(0),na);
373 This should not be used to supply default values for parameters. One
374 would normally use this when a function parameter must be processed by
375 another library function before it can be used. Default parameters are
376 covered in the next section.
378 =head2 Default Parameter Values
380 Default values can be specified for function parameters by
381 placing an assignment statement in the parameter list. The
382 default value may be a number or a string. Defaults should
383 always be used on the right-most parameters only.
385 To allow the XSUB for rpcb_gettime() to have a default host
386 value the parameters to the XSUB could be rearranged. The
387 XSUB will then call the real rpcb_gettime() function with
388 the parameters in the correct order. Perl will call this
389 XSUB with either of the following statements.
391 $status = rpcb_gettime( $timep, $host );
393 $status = rpcb_gettime( $timep );
395 The XSUB will look like the code which follows. A CODE:
396 block is used to call the real rpcb_gettime() function with
397 the parameters in the correct order for that function.
400 rpcb_gettime(timep,host="localhost")
402 time_t timep = NO_INIT
404 RETVAL = rpcb_gettime( host, &timep );
409 =head2 The PREINIT: Keyword
411 The PREINIT: keyword allows extra variables to be declared before the
412 typemaps are expanded. If a variable is declared in a CODE: block then that
413 variable will follow any typemap code. This may result in a C syntax
414 error. To force the variable to be declared before the typemap code, place
415 it into a PREINIT: block. The PREINIT: keyword may be used one or more
416 times within an XSUB.
418 The following examples are equivalent, but if the code is using complex
419 typemaps then the first example is safer.
423 time_t timep = NO_INIT
425 char *host = "localhost";
427 RETVAL = rpcb_gettime( host, &timep );
432 A correct, but error-prone example.
436 time_t timep = NO_INIT
438 char *host = "localhost";
439 RETVAL = rpcb_gettime( host, &timep );
444 =head2 The SCOPE: Keyword
446 The SCOPE: keyword allows scoping to be enabled for a particular XSUB. If
447 enabled, the XSUB will invoke ENTER and LEAVE automatically.
449 To support potentially complex type mappings, if a typemap entry used
450 by this XSUB contains a comment like C</*scope*/> then scoping will
451 automatically be enabled for that XSUB.
461 =head2 The INPUT: Keyword
463 The XSUB's parameters are usually evaluated immediately after entering the
464 XSUB. The INPUT: keyword can be used to force those parameters to be
465 evaluated a little later. The INPUT: keyword can be used multiple times
466 within an XSUB and can be used to list one or more input variables. This
467 keyword is used with the PREINIT: keyword.
469 The following example shows how the input parameter C<timep> can be
470 evaluated late, after a PREINIT.
473 rpcb_gettime(host,timep)
480 RETVAL = rpcb_gettime( host, &tt );
486 The next example shows each input parameter evaluated late.
489 rpcb_gettime(host,timep)
500 RETVAL = rpcb_gettime( h, &tt );
506 =head2 Variable-length Parameter Lists
508 XSUBs can have variable-length parameter lists by specifying an ellipsis
509 C<(...)> in the parameter list. This use of the ellipsis is similar to that
510 found in ANSI C. The programmer is able to determine the number of
511 arguments passed to the XSUB by examining the C<items> variable which the
512 B<xsubpp> compiler supplies for all XSUBs. By using this mechanism one can
513 create an XSUB which accepts a list of parameters of unknown length.
515 The I<host> parameter for the rpcb_gettime() XSUB can be
516 optional so the ellipsis can be used to indicate that the
517 XSUB will take a variable number of parameters. Perl should
518 be able to call this XSUB with either of the following statements.
520 $status = rpcb_gettime( $timep, $host );
522 $status = rpcb_gettime( $timep );
524 The XS code, with ellipsis, follows.
527 rpcb_gettime(timep, ...)
528 time_t timep = NO_INIT
530 char *host = "localhost";
533 host = (char *)SvPV(ST(1), na);
534 RETVAL = rpcb_gettime( host, &timep );
539 =head2 The PPCODE: Keyword
541 The PPCODE: keyword is an alternate form of the CODE: keyword and is used
542 to tell the B<xsubpp> compiler that the programmer is supplying the code to
543 control the argument stack for the XSUBs return values. Occasionally one
544 will want an XSUB to return a list of values rather than a single value.
545 In these cases one must use PPCODE: and then explicitly push the list of
546 values on the stack. The PPCODE: and CODE: keywords are not used
547 together within the same XSUB.
549 The following XSUB will call the C rpcb_gettime() function
550 and will return its two output values, timep and status, to
551 Perl as a single list.
560 status = rpcb_gettime( host, &timep );
562 PUSHs(sv_2mortal(newSViv(status)));
563 PUSHs(sv_2mortal(newSViv(timep)));
565 Notice that the programmer must supply the C code necessary
566 to have the real rpcb_gettime() function called and to have
567 the return values properly placed on the argument stack.
569 The C<void> return type for this function tells the B<xsubpp> compiler that
570 the RETVAL variable is not needed or used and that it should not be created.
571 In most scenarios the void return type should be used with the PPCODE:
574 The EXTEND() macro is used to make room on the argument
575 stack for 2 return values. The PPCODE: directive causes the
576 B<xsubpp> compiler to create a stack pointer available as C<SP>, and it
577 is this pointer which is being used in the EXTEND() macro.
578 The values are then pushed onto the stack with the PUSHs()
581 Now the rpcb_gettime() function can be used from Perl with
582 the following statement.
584 ($status, $timep) = rpcb_gettime("localhost");
586 When handling output parameters with a PPCODE section, be sure to handle
587 'set' magic properly. See L<perlguts> for details about 'set' magic.
589 =head2 Returning Undef And Empty Lists
591 Occasionally the programmer will want to return simply
592 C<undef> or an empty list if a function fails rather than a
593 separate status value. The rpcb_gettime() function offers
594 just this situation. If the function succeeds we would like
595 to have it return the time and if it fails we would like to
596 have undef returned. In the following Perl code the value
597 of $timep will either be undef or it will be a valid time.
599 $timep = rpcb_gettime( "localhost" );
601 The following XSUB uses the C<SV *> return type as a mneumonic only,
602 and uses a CODE: block to indicate to the compiler
603 that the programmer has supplied all the necessary code. The
604 sv_newmortal() call will initialize the return value to undef, making that
605 the default return value.
614 ST(0) = sv_newmortal();
615 if( rpcb_gettime( host, &timep ) )
616 sv_setnv( ST(0), (double)timep);
618 The next example demonstrates how one would place an explicit undef in the
619 return value, should the need arise.
628 ST(0) = sv_newmortal();
629 if( rpcb_gettime( host, &timep ) ){
630 sv_setnv( ST(0), (double)timep);
636 To return an empty list one must use a PPCODE: block and
637 then not push return values on the stack.
645 if( rpcb_gettime( host, &timep ) )
646 PUSHs(sv_2mortal(newSViv(timep)));
648 /* Nothing pushed on stack, so an empty */
649 /* list is implicitly returned. */
652 Some people may be inclined to include an explicit C<return> in the above
653 XSUB, rather than letting control fall through to the end. In those
654 situations C<XSRETURN_EMPTY> should be used, instead. This will ensure that
655 the XSUB stack is properly adjusted. Consult L<perlguts/"API LISTING"> for
656 other C<XSRETURN> macros.
658 =head2 The REQUIRE: Keyword
660 The REQUIRE: keyword is used to indicate the minimum version of the
661 B<xsubpp> compiler needed to compile the XS module. An XS module which
662 contains the following statement will compile with only B<xsubpp> version
667 =head2 The CLEANUP: Keyword
669 This keyword can be used when an XSUB requires special cleanup procedures
670 before it terminates. When the CLEANUP: keyword is used it must follow
671 any CODE:, PPCODE:, or OUTPUT: blocks which are present in the XSUB. The
672 code specified for the cleanup block will be added as the last statements
675 =head2 The BOOT: Keyword
677 The BOOT: keyword is used to add code to the extension's bootstrap
678 function. The bootstrap function is generated by the B<xsubpp> compiler and
679 normally holds the statements necessary to register any XSUBs with Perl.
680 With the BOOT: keyword the programmer can tell the compiler to add extra
681 statements to the bootstrap function.
683 This keyword may be used any time after the first MODULE keyword and should
684 appear on a line by itself. The first blank line after the keyword will
685 terminate the code block.
688 # The following message will be printed when the
689 # bootstrap function executes.
690 printf("Hello from the bootstrap!\n");
692 =head2 The VERSIONCHECK: Keyword
694 The VERSIONCHECK: keyword corresponds to B<xsubpp>'s C<-versioncheck> and
695 C<-noversioncheck> options. This keyword overrides the command line
696 options. Version checking is enabled by default. When version checking is
697 enabled the XS module will attempt to verify that its version matches the
698 version of the PM module.
700 To enable version checking:
704 To disable version checking:
706 VERSIONCHECK: DISABLE
708 =head2 The PROTOTYPES: Keyword
710 The PROTOTYPES: keyword corresponds to B<xsubpp>'s C<-prototypes> and
711 C<-noprototypes> options. This keyword overrides the command line options.
712 Prototypes are enabled by default. When prototypes are enabled XSUBs will
713 be given Perl prototypes. This keyword may be used multiple times in an XS
714 module to enable and disable prototypes for different parts of the module.
716 To enable prototypes:
720 To disable prototypes:
724 =head2 The PROTOTYPE: Keyword
726 This keyword is similar to the PROTOTYPES: keyword above but can be used to
727 force B<xsubpp> to use a specific prototype for the XSUB. This keyword
728 overrides all other prototype options and keywords but affects only the
729 current XSUB. Consult L<perlsub/Prototypes> for information about Perl
733 rpcb_gettime(timep, ...)
734 time_t timep = NO_INIT
737 char *host = "localhost";
740 host = (char *)SvPV(ST(1), na);
741 RETVAL = rpcb_gettime( host, &timep );
746 =head2 The ALIAS: Keyword
748 The ALIAS: keyword allows an XSUB to have two more unique Perl names
749 and to know which of those names was used when it was invoked. The Perl
750 names may be fully-qualified with package names. Each alias is given an
751 index. The compiler will setup a variable called C<ix> which contain the
752 index of the alias which was used. When the XSUB is called with its
753 declared name C<ix> will be 0.
755 The following example will create aliases C<FOO::gettime()> and
756 C<BAR::getit()> for this function.
759 rpcb_gettime(host,timep)
766 printf("# ix = %d\n", ix );
770 =head2 The INCLUDE: Keyword
772 This keyword can be used to pull other files into the XS module. The other
773 files may have XS code. INCLUDE: can also be used to run a command to
774 generate the XS code to be pulled into the module.
776 The file F<Rpcb1.xsh> contains our C<rpcb_gettime()> function:
779 rpcb_gettime(host,timep)
785 The XS module can use INCLUDE: to pull that file into it.
789 If the parameters to the INCLUDE: keyword are followed by a pipe (C<|>) then
790 the compiler will interpret the parameters as a command.
792 INCLUDE: cat Rpcb1.xsh |
794 =head2 The CASE: Keyword
796 The CASE: keyword allows an XSUB to have multiple distinct parts with each
797 part acting as a virtual XSUB. CASE: is greedy and if it is used then all
798 other XS keywords must be contained within a CASE:. This means nothing may
799 precede the first CASE: in the XSUB and anything following the last CASE: is
800 included in that case.
802 A CASE: might switch via a parameter of the XSUB, via the C<ix> ALIAS:
803 variable (see L<"The ALIAS: Keyword">), or maybe via the C<items> variable
804 (see L<"Variable-length Parameter Lists">). The last CASE: becomes the
805 B<default> case if it is not associated with a conditional. The following
806 example shows CASE switched via C<ix> with a function C<rpcb_gettime()>
807 having an alias C<x_gettime()>. When the function is called as
808 C<rpcb_gettime()> its parameters are the usual C<(char *host, time_t *timep)>,
809 but when the function is called as C<x_gettime()> its parameters are
810 reversed, C<(time_t *timep, char *host)>.
818 # 'a' is timep, 'b' is host
822 RETVAL = rpcb_gettime( b, &a );
827 # 'a' is host, 'b' is timep
834 That function can be called with either of the following statements. Note
835 the different argument lists.
837 $status = rpcb_gettime( $host, $timep );
839 $status = x_gettime( $timep, $host );
841 =head2 The & Unary Operator
843 The & unary operator is used to tell the compiler that it should dereference
844 the object when it calls the C function. This is used when a CODE: block is
845 not used and the object is a not a pointer type (the object is an C<int> or
846 C<long> but not a C<int*> or C<long*>).
848 The following XSUB will generate incorrect C code. The xsubpp compiler will
849 turn this into code which calls C<rpcb_gettime()> with parameters C<(char
850 *host, time_t timep)>, but the real C<rpcb_gettime()> wants the C<timep>
851 parameter to be of type C<time_t*> rather than C<time_t>.
854 rpcb_gettime(host,timep)
860 That problem is corrected by using the C<&> operator. The xsubpp compiler
861 will now turn this into code which calls C<rpcb_gettime()> correctly with
862 parameters C<(char *host, time_t *timep)>. It does this by carrying the
863 C<&> through, so the function call looks like C<rpcb_gettime(host, &timep)>.
866 rpcb_gettime(host,timep)
872 =head2 Inserting Comments and C Preprocessor Directives
874 C preprocessor directives are allowed within BOOT:, PREINIT: INIT:,
875 CODE:, PPCODE:, and CLEANUP: blocks, as well as outside the functions.
876 Comments are allowed anywhere after the MODULE keyword. The compiler
877 will pass the preprocessor directives through untouched and will remove
880 Comments can be added to XSUBs by placing a C<#> as the first
881 non-whitespace of a line. Care should be taken to avoid making the
882 comment look like a C preprocessor directive, lest it be interpreted as
883 such. The simplest way to prevent this is to put whitespace in front of
886 If you use preprocessor directives to choose one of two
887 versions of a function, use
890 #else /* ... version2 */
900 because otherwise xsubpp will believe that you made a duplicate
901 definition of the function. Also, put a blank line before the
902 #else/#endif so it will not be seen as part of the function body.
904 =head2 Using XS With C++
906 If a function is defined as a C++ method then it will assume
907 its first argument is an object pointer. The object pointer
908 will be stored in a variable called THIS. The object should
909 have been created by C++ with the new() function and should
910 be blessed by Perl with the sv_setref_pv() macro. The
911 blessing of the object by Perl can be handled by a typemap. An example
912 typemap is shown at the end of this section.
914 If the method is defined as static it will call the C++
915 function using the class::method() syntax. If the method is not static
916 the function will be called using the THIS-E<gt>method() syntax.
918 The next examples will use the following C++ class.
925 void set_blue( int );
931 The XSUBs for the blue() and set_blue() methods are defined with the class
932 name but the parameter for the object (THIS, or "self") is implicit and is
939 color::set_blue( val )
942 Both functions will expect an object as the first parameter. The xsubpp
943 compiler will call that object C<THIS> and will use it to call the specified
944 method. So in the C++ code the blue() and set_blue() methods will be called
945 in the following manner.
947 RETVAL = THIS->blue();
949 THIS->set_blue( val );
951 If the function's name is B<DESTROY> then the C++ C<delete> function will be
952 called and C<THIS> will be given as its parameter.
957 The C++ code will call C<delete>.
961 If the function's name is B<new> then the C++ C<new> function will be called
962 to create a dynamic C++ object. The XSUB will expect the class name, which
963 will be kept in a variable called C<CLASS>, to be given as the first
969 The C++ code will call C<new>.
971 RETVAL = new color();
973 The following is an example of a typemap that could be used for this C++
980 # The Perl object is blessed into 'CLASS', which should be a
981 # char* having the name of the package for the blessing.
983 sv_setref_pv( $arg, CLASS, (void*)$var );
987 if( sv_isobject($arg) && (SvTYPE(SvRV($arg)) == SVt_PVMG) )
988 $var = ($type)SvIV((SV*)SvRV( $arg ));
990 warn( \"${Package}::$func_name() -- $var is not a blessed SV reference\" );
994 =head2 Interface Strategy
996 When designing an interface between Perl and a C library a straight
997 translation from C to XS is often sufficient. The interface will often be
998 very C-like and occasionally nonintuitive, especially when the C function
999 modifies one of its parameters. In cases where the programmer wishes to
1000 create a more Perl-like interface the following strategy may help to
1001 identify the more critical parts of the interface.
1003 Identify the C functions which modify their parameters. The XSUBs for
1004 these functions may be able to return lists to Perl, or may be
1005 candidates to return undef or an empty list in case of failure.
1007 Identify which values are used by only the C and XSUB functions
1008 themselves. If Perl does not need to access the contents of the value
1009 then it may not be necessary to provide a translation for that value
1012 Identify the pointers in the C function parameter lists and return
1013 values. Some pointers can be handled in XS with the & unary operator on
1014 the variable name while others will require the use of the * operator on
1015 the type name. In general it is easier to work with the & operator.
1017 Identify the structures used by the C functions. In many
1018 cases it may be helpful to use the T_PTROBJ typemap for
1019 these structures so they can be manipulated by Perl as
1022 =head2 Perl Objects And C Structures
1024 When dealing with C structures one should select either
1025 B<T_PTROBJ> or B<T_PTRREF> for the XS type. Both types are
1026 designed to handle pointers to complex objects. The
1027 T_PTRREF type will allow the Perl object to be unblessed
1028 while the T_PTROBJ type requires that the object be blessed.
1029 By using T_PTROBJ one can achieve a form of type-checking
1030 because the XSUB will attempt to verify that the Perl object
1031 is of the expected type.
1033 The following XS code shows the getnetconfigent() function which is used
1034 with ONC+ TIRPC. The getnetconfigent() function will return a pointer to a
1035 C structure and has the C prototype shown below. The example will
1036 demonstrate how the C pointer will become a Perl reference. Perl will
1037 consider this reference to be a pointer to a blessed object and will
1038 attempt to call a destructor for the object. A destructor will be
1039 provided in the XS source to free the memory used by getnetconfigent().
1040 Destructors in XS can be created by specifying an XSUB function whose name
1041 ends with the word B<DESTROY>. XS destructors can be used to free memory
1042 which may have been malloc'd by another XSUB.
1044 struct netconfig *getnetconfigent(const char *netid);
1046 A C<typedef> will be created for C<struct netconfig>. The Perl
1047 object will be blessed in a class matching the name of the C
1048 type, with the tag C<Ptr> appended, and the name should not
1049 have embedded spaces if it will be a Perl package name. The
1050 destructor will be placed in a class corresponding to the
1051 class of the object and the PREFIX keyword will be used to
1052 trim the name to the word DESTROY as Perl will expect.
1054 typedef struct netconfig Netconfig;
1056 MODULE = RPC PACKAGE = RPC
1059 getnetconfigent(netid)
1062 MODULE = RPC PACKAGE = NetconfigPtr PREFIX = rpcb_
1065 rpcb_DESTROY(netconf)
1068 printf("Now in NetconfigPtr::DESTROY\n");
1071 This example requires the following typemap entry. Consult the typemap
1072 section for more information about adding new typemaps for an extension.
1075 Netconfig * T_PTROBJ
1077 This example will be used with the following Perl statements.
1080 $netconf = getnetconfigent("udp");
1082 When Perl destroys the object referenced by $netconf it will send the
1083 object to the supplied XSUB DESTROY function. Perl cannot determine, and
1084 does not care, that this object is a C struct and not a Perl object. In
1085 this sense, there is no difference between the object created by the
1086 getnetconfigent() XSUB and an object created by a normal Perl subroutine.
1090 The typemap is a collection of code fragments which are used by the B<xsubpp>
1091 compiler to map C function parameters and values to Perl values. The
1092 typemap file may consist of three sections labeled C<TYPEMAP>, C<INPUT>, and
1093 C<OUTPUT>. The INPUT section tells the compiler how to translate Perl values
1094 into variables of certain C types. The OUTPUT section tells the compiler
1095 how to translate the values from certain C types into values Perl can
1096 understand. The TYPEMAP section tells the compiler which of the INPUT and
1097 OUTPUT code fragments should be used to map a given C type to a Perl value.
1098 Each of the sections of the typemap must be preceded by one of the TYPEMAP,
1099 INPUT, or OUTPUT keywords.
1101 The default typemap in the C<ext> directory of the Perl source contains many
1102 useful types which can be used by Perl extensions. Some extensions define
1103 additional typemaps which they keep in their own directory. These
1104 additional typemaps may reference INPUT and OUTPUT maps in the main
1105 typemap. The B<xsubpp> compiler will allow the extension's own typemap to
1106 override any mappings which are in the default typemap.
1108 Most extensions which require a custom typemap will need only the TYPEMAP
1109 section of the typemap file. The custom typemap used in the
1110 getnetconfigent() example shown earlier demonstrates what may be the typical
1111 use of extension typemaps. That typemap is used to equate a C structure
1112 with the T_PTROBJ typemap. The typemap used by getnetconfigent() is shown
1113 here. Note that the C type is separated from the XS type with a tab and
1114 that the C unary operator C<*> is considered to be a part of the C type name.
1117 Netconfig *<tab>T_PTROBJ
1119 Here's a more complicated example: suppose that you wanted C<struct
1120 netconfig> to be blessed into the class C<Net::Config>. One way to do
1121 this is to use underscores (_) to separate package names, as follows:
1123 typedef struct netconfig * Net_Config;
1125 And then provide a typemap entry C<T_PTROBJ_SPECIAL> that maps underscores to
1126 double-colons (::), and declare C<Net_Config> to be of that type:
1130 Net_Config T_PTROBJ_SPECIAL
1134 if (sv_derived_from($arg, \"${(my $ntt=$ntype)=~s/_/::/g;\$ntt}\")) {
1135 IV tmp = SvIV((SV*)SvRV($arg));
1139 croak(\"$var is not of type ${(my $ntt=$ntype)=~s/_/::/g;\$ntt}\")
1143 sv_setref_pv($arg, \"${(my $ntt=$ntype)=~s/_/::/g;\$ntt}\",
1146 The INPUT and OUTPUT sections substitute underscores for double-colons
1147 on the fly, giving the desired effect. This example demonstrates some
1148 of the power and versatility of the typemap facility.
1152 File C<RPC.xs>: Interface to some ONC+ RPC bind library functions.
1158 #include <rpc/rpc.h>
1160 typedef struct netconfig Netconfig;
1162 MODULE = RPC PACKAGE = RPC
1165 rpcb_gettime(host="localhost")
1170 ST(0) = sv_newmortal();
1171 if( rpcb_gettime( host, &timep ) )
1172 sv_setnv( ST(0), (double)timep );
1175 getnetconfigent(netid="udp")
1178 MODULE = RPC PACKAGE = NetconfigPtr PREFIX = rpcb_
1181 rpcb_DESTROY(netconf)
1184 printf("NetconfigPtr::DESTROY\n");
1187 File C<typemap>: Custom typemap for RPC.xs.
1190 Netconfig * T_PTROBJ
1192 File C<RPC.pm>: Perl module for the RPC extension.
1198 @ISA = qw(Exporter DynaLoader);
1199 @EXPORT = qw(rpcb_gettime getnetconfigent);
1204 File C<rpctest.pl>: Perl test program for the RPC extension.
1208 $netconf = getnetconfigent();
1209 $a = rpcb_gettime();
1210 print "time = $a\n";
1211 print "netconf = $netconf\n";
1213 $netconf = getnetconfigent("tcp");
1214 $a = rpcb_gettime("poplar");
1215 print "time = $a\n";
1216 print "netconf = $netconf\n";
1221 This document covers features supported by C<xsubpp> 1.935.
1225 Dean Roehrich <F<roehrich@cray.com>>