3 perlguts - Introduction to the Perl API
7 This document attempts to describe how to use the Perl API, as well as
8 to provide some info on the basic workings of the Perl core. It is far
9 from complete and probably contains many errors. Please refer any
10 questions or comments to the author below.
16 Perl has three typedefs that handle Perl's three main data types:
22 Each typedef has specific routines that manipulate the various data types.
24 =head2 What is an "IV"?
26 Perl uses a special typedef IV which is a simple signed integer type that is
27 guaranteed to be large enough to hold a pointer (as well as an integer).
28 Additionally, there is the UV, which is simply an unsigned IV.
30 Perl also uses two special typedefs, I32 and I16, which will always be at
31 least 32-bits and 16-bits long, respectively. (Again, there are U32 and U16,
32 as well.) They will usually be exactly 32 and 16 bits long, but on Crays
33 they will both be 64 bits.
35 =head2 Working with SVs
37 An SV can be created and loaded with one command. There are five types of
38 values that can be loaded: an integer value (IV), an unsigned integer
39 value (UV), a double (NV), a string (PV), and another scalar (SV).
41 The seven routines are:
46 SV* newSVpv(const char*, STRLEN);
47 SV* newSVpvn(const char*, STRLEN);
48 SV* newSVpvf(const char*, ...);
51 C<STRLEN> is an integer type (Size_t, usually defined as size_t in
52 F<config.h>) guaranteed to be large enough to represent the size of
53 any string that perl can handle.
55 In the unlikely case of a SV requiring more complex initialisation, you
56 can create an empty SV with newSV(len). If C<len> is 0 an empty SV of
57 type NULL is returned, else an SV of type PV is returned with len + 1 (for
58 the NUL) bytes of storage allocated, accessible via SvPVX. In both cases
59 the SV has value undef.
61 SV *sv = newSV(0); /* no storage allocated */
62 SV *sv = newSV(10); /* 10 (+1) bytes of uninitialised storage allocated */
64 To change the value of an I<already-existing> SV, there are eight routines:
66 void sv_setiv(SV*, IV);
67 void sv_setuv(SV*, UV);
68 void sv_setnv(SV*, double);
69 void sv_setpv(SV*, const char*);
70 void sv_setpvn(SV*, const char*, STRLEN)
71 void sv_setpvf(SV*, const char*, ...);
72 void sv_vsetpvfn(SV*, const char*, STRLEN, va_list *, SV **, I32, bool *);
73 void sv_setsv(SV*, SV*);
75 Notice that you can choose to specify the length of the string to be
76 assigned by using C<sv_setpvn>, C<newSVpvn>, or C<newSVpv>, or you may
77 allow Perl to calculate the length by using C<sv_setpv> or by specifying
78 0 as the second argument to C<newSVpv>. Be warned, though, that Perl will
79 determine the string's length by using C<strlen>, which depends on the
80 string terminating with a NUL character.
82 The arguments of C<sv_setpvf> are processed like C<sprintf>, and the
83 formatted output becomes the value.
85 C<sv_vsetpvfn> is an analogue of C<vsprintf>, but it allows you to specify
86 either a pointer to a variable argument list or the address and length of
87 an array of SVs. The last argument points to a boolean; on return, if that
88 boolean is true, then locale-specific information has been used to format
89 the string, and the string's contents are therefore untrustworthy (see
90 L<perlsec>). This pointer may be NULL if that information is not
91 important. Note that this function requires you to specify the length of
94 The C<sv_set*()> functions are not generic enough to operate on values
95 that have "magic". See L<Magic Virtual Tables> later in this document.
97 All SVs that contain strings should be terminated with a NUL character.
98 If it is not NUL-terminated there is a risk of
99 core dumps and corruptions from code which passes the string to C
100 functions or system calls which expect a NUL-terminated string.
101 Perl's own functions typically add a trailing NUL for this reason.
102 Nevertheless, you should be very careful when you pass a string stored
103 in an SV to a C function or system call.
105 To access the actual value that an SV points to, you can use the macros:
110 SvPV(SV*, STRLEN len)
113 which will automatically coerce the actual scalar type into an IV, UV, double,
116 In the C<SvPV> macro, the length of the string returned is placed into the
117 variable C<len> (this is a macro, so you do I<not> use C<&len>). If you do
118 not care what the length of the data is, use the C<SvPV_nolen> macro.
119 Historically the C<SvPV> macro with the global variable C<PL_na> has been
120 used in this case. But that can be quite inefficient because C<PL_na> must
121 be accessed in thread-local storage in threaded Perl. In any case, remember
122 that Perl allows arbitrary strings of data that may both contain NULs and
123 might not be terminated by a NUL.
125 Also remember that C doesn't allow you to safely say C<foo(SvPV(s, len),
126 len);>. It might work with your compiler, but it won't work for everyone.
127 Break this sort of statement up into separate assignments:
135 If you want to know if the scalar value is TRUE, you can use:
139 Although Perl will automatically grow strings for you, if you need to force
140 Perl to allocate more memory for your SV, you can use the macro
142 SvGROW(SV*, STRLEN newlen)
144 which will determine if more memory needs to be allocated. If so, it will
145 call the function C<sv_grow>. Note that C<SvGROW> can only increase, not
146 decrease, the allocated memory of an SV and that it does not automatically
147 add a byte for the a trailing NUL (perl's own string functions typically do
148 C<SvGROW(sv, len + 1)>).
150 If you have an SV and want to know what kind of data Perl thinks is stored
151 in it, you can use the following macros to check the type of SV you have.
157 You can get and set the current length of the string stored in an SV with
158 the following macros:
161 SvCUR_set(SV*, I32 val)
163 You can also get a pointer to the end of the string stored in the SV
168 But note that these last three macros are valid only if C<SvPOK()> is true.
170 If you want to append something to the end of string stored in an C<SV*>,
171 you can use the following functions:
173 void sv_catpv(SV*, const char*);
174 void sv_catpvn(SV*, const char*, STRLEN);
175 void sv_catpvf(SV*, const char*, ...);
176 void sv_vcatpvfn(SV*, const char*, STRLEN, va_list *, SV **, I32, bool);
177 void sv_catsv(SV*, SV*);
179 The first function calculates the length of the string to be appended by
180 using C<strlen>. In the second, you specify the length of the string
181 yourself. The third function processes its arguments like C<sprintf> and
182 appends the formatted output. The fourth function works like C<vsprintf>.
183 You can specify the address and length of an array of SVs instead of the
184 va_list argument. The fifth function extends the string stored in the first
185 SV with the string stored in the second SV. It also forces the second SV
186 to be interpreted as a string.
188 The C<sv_cat*()> functions are not generic enough to operate on values that
189 have "magic". See L<Magic Virtual Tables> later in this document.
191 If you know the name of a scalar variable, you can get a pointer to its SV
192 by using the following:
194 SV* get_sv("package::varname", FALSE);
196 This returns NULL if the variable does not exist.
198 If you want to know if this variable (or any other SV) is actually C<defined>,
203 The scalar C<undef> value is stored in an SV instance called C<PL_sv_undef>.
205 Its address can be used whenever an C<SV*> is needed. Make sure that
206 you don't try to compare a random sv with C<&PL_sv_undef>. For example
207 when interfacing Perl code, it'll work correctly for:
211 But won't work when called as:
216 So to repeat always use SvOK() to check whether an sv is defined.
218 Also you have to be careful when using C<&PL_sv_undef> as a value in
219 AVs or HVs (see L<AVs, HVs and undefined values>).
221 There are also the two values C<PL_sv_yes> and C<PL_sv_no>, which contain
222 boolean TRUE and FALSE values, respectively. Like C<PL_sv_undef>, their
223 addresses can be used whenever an C<SV*> is needed.
225 Do not be fooled into thinking that C<(SV *) 0> is the same as C<&PL_sv_undef>.
229 if (I-am-to-return-a-real-value) {
230 sv = sv_2mortal(newSViv(42));
234 This code tries to return a new SV (which contains the value 42) if it should
235 return a real value, or undef otherwise. Instead it has returned a NULL
236 pointer which, somewhere down the line, will cause a segmentation violation,
237 bus error, or just weird results. Change the zero to C<&PL_sv_undef> in the
238 first line and all will be well.
240 To free an SV that you've created, call C<SvREFCNT_dec(SV*)>. Normally this
241 call is not necessary (see L<Reference Counts and Mortality>).
245 Perl provides the function C<sv_chop> to efficiently remove characters
246 from the beginning of a string; you give it an SV and a pointer to
247 somewhere inside the PV, and it discards everything before the
248 pointer. The efficiency comes by means of a little hack: instead of
249 actually removing the characters, C<sv_chop> sets the flag C<OOK>
250 (offset OK) to signal to other functions that the offset hack is in
251 effect, and it puts the number of bytes chopped off into the IV field
252 of the SV. It then moves the PV pointer (called C<SvPVX>) forward that
253 many bytes, and adjusts C<SvCUR> and C<SvLEN>.
255 Hence, at this point, the start of the buffer that we allocated lives
256 at C<SvPVX(sv) - SvIV(sv)> in memory and the PV pointer is pointing
257 into the middle of this allocated storage.
259 This is best demonstrated by example:
261 % ./perl -Ilib -MDevel::Peek -le '$a="12345"; $a=~s/.//; Dump($a)'
262 SV = PVIV(0x8128450) at 0x81340f0
264 FLAGS = (POK,OOK,pPOK)
266 PV = 0x8135781 ( "1" . ) "2345"\0
270 Here the number of bytes chopped off (1) is put into IV, and
271 C<Devel::Peek::Dump> helpfully reminds us that this is an offset. The
272 portion of the string between the "real" and the "fake" beginnings is
273 shown in parentheses, and the values of C<SvCUR> and C<SvLEN> reflect
274 the fake beginning, not the real one.
276 Something similar to the offset hack is performed on AVs to enable
277 efficient shifting and splicing off the beginning of the array; while
278 C<AvARRAY> points to the first element in the array that is visible from
279 Perl, C<AvALLOC> points to the real start of the C array. These are
280 usually the same, but a C<shift> operation can be carried out by
281 increasing C<AvARRAY> by one and decreasing C<AvFILL> and C<AvLEN>.
282 Again, the location of the real start of the C array only comes into
283 play when freeing the array. See C<av_shift> in F<av.c>.
285 =head2 What's Really Stored in an SV?
287 Recall that the usual method of determining the type of scalar you have is
288 to use C<Sv*OK> macros. Because a scalar can be both a number and a string,
289 usually these macros will always return TRUE and calling the C<Sv*V>
290 macros will do the appropriate conversion of string to integer/double or
291 integer/double to string.
293 If you I<really> need to know if you have an integer, double, or string
294 pointer in an SV, you can use the following three macros instead:
300 These will tell you if you truly have an integer, double, or string pointer
301 stored in your SV. The "p" stands for private.
303 The are various ways in which the private and public flags may differ.
304 For example, a tied SV may have a valid underlying value in the IV slot
305 (so SvIOKp is true), but the data should be accessed via the FETCH
306 routine rather than directly, so SvIOK is false. Another is when
307 numeric conversion has occurred and precision has been lost: only the
308 private flag is set on 'lossy' values. So when an NV is converted to an
309 IV with loss, SvIOKp, SvNOKp and SvNOK will be set, while SvIOK wont be.
311 In general, though, it's best to use the C<Sv*V> macros.
313 =head2 Working with AVs
315 There are two ways to create and load an AV. The first method creates an
320 The second method both creates the AV and initially populates it with SVs:
322 AV* av_make(I32 num, SV **ptr);
324 The second argument points to an array containing C<num> C<SV*>'s. Once the
325 AV has been created, the SVs can be destroyed, if so desired.
327 Once the AV has been created, the following operations are possible on AVs:
329 void av_push(AV*, SV*);
332 void av_unshift(AV*, I32 num);
334 These should be familiar operations, with the exception of C<av_unshift>.
335 This routine adds C<num> elements at the front of the array with the C<undef>
336 value. You must then use C<av_store> (described below) to assign values
337 to these new elements.
339 Here are some other functions:
342 SV** av_fetch(AV*, I32 key, I32 lval);
343 SV** av_store(AV*, I32 key, SV* val);
345 The C<av_len> function returns the highest index value in array (just
346 like $#array in Perl). If the array is empty, -1 is returned. The
347 C<av_fetch> function returns the value at index C<key>, but if C<lval>
348 is non-zero, then C<av_fetch> will store an undef value at that index.
349 The C<av_store> function stores the value C<val> at index C<key>, and does
350 not increment the reference count of C<val>. Thus the caller is responsible
351 for taking care of that, and if C<av_store> returns NULL, the caller will
352 have to decrement the reference count to avoid a memory leak. Note that
353 C<av_fetch> and C<av_store> both return C<SV**>'s, not C<SV*>'s as their
358 void av_extend(AV*, I32 key);
360 The C<av_clear> function deletes all the elements in the AV* array, but
361 does not actually delete the array itself. The C<av_undef> function will
362 delete all the elements in the array plus the array itself. The
363 C<av_extend> function extends the array so that it contains at least C<key+1>
364 elements. If C<key+1> is less than the currently allocated length of the array,
365 then nothing is done.
367 If you know the name of an array variable, you can get a pointer to its AV
368 by using the following:
370 AV* get_av("package::varname", FALSE);
372 This returns NULL if the variable does not exist.
374 See L<Understanding the Magic of Tied Hashes and Arrays> for more
375 information on how to use the array access functions on tied arrays.
377 =head2 Working with HVs
379 To create an HV, you use the following routine:
383 Once the HV has been created, the following operations are possible on HVs:
385 SV** hv_store(HV*, const char* key, U32 klen, SV* val, U32 hash);
386 SV** hv_fetch(HV*, const char* key, U32 klen, I32 lval);
388 The C<klen> parameter is the length of the key being passed in (Note that
389 you cannot pass 0 in as a value of C<klen> to tell Perl to measure the
390 length of the key). The C<val> argument contains the SV pointer to the
391 scalar being stored, and C<hash> is the precomputed hash value (zero if
392 you want C<hv_store> to calculate it for you). The C<lval> parameter
393 indicates whether this fetch is actually a part of a store operation, in
394 which case a new undefined value will be added to the HV with the supplied
395 key and C<hv_fetch> will return as if the value had already existed.
397 Remember that C<hv_store> and C<hv_fetch> return C<SV**>'s and not just
398 C<SV*>. To access the scalar value, you must first dereference the return
399 value. However, you should check to make sure that the return value is
400 not NULL before dereferencing it.
402 These two functions check if a hash table entry exists, and deletes it.
404 bool hv_exists(HV*, const char* key, U32 klen);
405 SV* hv_delete(HV*, const char* key, U32 klen, I32 flags);
407 If C<flags> does not include the C<G_DISCARD> flag then C<hv_delete> will
408 create and return a mortal copy of the deleted value.
410 And more miscellaneous functions:
415 Like their AV counterparts, C<hv_clear> deletes all the entries in the hash
416 table but does not actually delete the hash table. The C<hv_undef> deletes
417 both the entries and the hash table itself.
419 Perl keeps the actual data in linked list of structures with a typedef of HE.
420 These contain the actual key and value pointers (plus extra administrative
421 overhead). The key is a string pointer; the value is an C<SV*>. However,
422 once you have an C<HE*>, to get the actual key and value, use the routines
425 I32 hv_iterinit(HV*);
426 /* Prepares starting point to traverse hash table */
427 HE* hv_iternext(HV*);
428 /* Get the next entry, and return a pointer to a
429 structure that has both the key and value */
430 char* hv_iterkey(HE* entry, I32* retlen);
431 /* Get the key from an HE structure and also return
432 the length of the key string */
433 SV* hv_iterval(HV*, HE* entry);
434 /* Return an SV pointer to the value of the HE
436 SV* hv_iternextsv(HV*, char** key, I32* retlen);
437 /* This convenience routine combines hv_iternext,
438 hv_iterkey, and hv_iterval. The key and retlen
439 arguments are return values for the key and its
440 length. The value is returned in the SV* argument */
442 If you know the name of a hash variable, you can get a pointer to its HV
443 by using the following:
445 HV* get_hv("package::varname", FALSE);
447 This returns NULL if the variable does not exist.
449 The hash algorithm is defined in the C<PERL_HASH(hash, key, klen)> macro:
453 hash = (hash * 33) + *key++;
454 hash = hash + (hash >> 5); /* after 5.6 */
456 The last step was added in version 5.6 to improve distribution of
457 lower bits in the resulting hash value.
459 See L<Understanding the Magic of Tied Hashes and Arrays> for more
460 information on how to use the hash access functions on tied hashes.
462 =head2 Hash API Extensions
464 Beginning with version 5.004, the following functions are also supported:
466 HE* hv_fetch_ent (HV* tb, SV* key, I32 lval, U32 hash);
467 HE* hv_store_ent (HV* tb, SV* key, SV* val, U32 hash);
469 bool hv_exists_ent (HV* tb, SV* key, U32 hash);
470 SV* hv_delete_ent (HV* tb, SV* key, I32 flags, U32 hash);
472 SV* hv_iterkeysv (HE* entry);
474 Note that these functions take C<SV*> keys, which simplifies writing
475 of extension code that deals with hash structures. These functions
476 also allow passing of C<SV*> keys to C<tie> functions without forcing
477 you to stringify the keys (unlike the previous set of functions).
479 They also return and accept whole hash entries (C<HE*>), making their
480 use more efficient (since the hash number for a particular string
481 doesn't have to be recomputed every time). See L<perlapi> for detailed
484 The following macros must always be used to access the contents of hash
485 entries. Note that the arguments to these macros must be simple
486 variables, since they may get evaluated more than once. See
487 L<perlapi> for detailed descriptions of these macros.
489 HePV(HE* he, STRLEN len)
493 HeSVKEY_force(HE* he)
494 HeSVKEY_set(HE* he, SV* sv)
496 These two lower level macros are defined, but must only be used when
497 dealing with keys that are not C<SV*>s:
502 Note that both C<hv_store> and C<hv_store_ent> do not increment the
503 reference count of the stored C<val>, which is the caller's responsibility.
504 If these functions return a NULL value, the caller will usually have to
505 decrement the reference count of C<val> to avoid a memory leak.
507 =head2 AVs, HVs and undefined values
509 Sometimes you have to store undefined values in AVs or HVs. Although
510 this may be a rare case, it can be tricky. That's because you're
511 used to using C<&PL_sv_undef> if you need an undefined SV.
513 For example, intuition tells you that this XS code:
516 av_store( av, 0, &PL_sv_undef );
518 is equivalent to this Perl code:
523 Unfortunately, this isn't true. AVs use C<&PL_sv_undef> as a marker
524 for indicating that an array element has not yet been initialized.
525 Thus, C<exists $av[0]> would be true for the above Perl code, but
526 false for the array generated by the XS code.
528 Other problems can occur when storing C<&PL_sv_undef> in HVs:
530 hv_store( hv, "key", 3, &PL_sv_undef, 0 );
532 This will indeed make the value C<undef>, but if you try to modify
533 the value of C<key>, you'll get the following error:
535 Modification of non-creatable hash value attempted
537 In perl 5.8.0, C<&PL_sv_undef> was also used to mark placeholders
538 in restricted hashes. This caused such hash entries not to appear
539 when iterating over the hash or when checking for the keys
540 with the C<hv_exists> function.
542 You can run into similar problems when you store C<&PL_sv_true> or
543 C<&PL_sv_false> into AVs or HVs. Trying to modify such elements
544 will give you the following error:
546 Modification of a read-only value attempted
548 To make a long story short, you can use the special variables
549 C<&PL_sv_undef>, C<&PL_sv_true> and C<&PL_sv_false> with AVs and
550 HVs, but you have to make sure you know what you're doing.
552 Generally, if you want to store an undefined value in an AV
553 or HV, you should not use C<&PL_sv_undef>, but rather create a
554 new undefined value using the C<newSV> function, for example:
556 av_store( av, 42, newSV(0) );
557 hv_store( hv, "foo", 3, newSV(0), 0 );
561 References are a special type of scalar that point to other data types
562 (including references).
564 To create a reference, use either of the following functions:
566 SV* newRV_inc((SV*) thing);
567 SV* newRV_noinc((SV*) thing);
569 The C<thing> argument can be any of an C<SV*>, C<AV*>, or C<HV*>. The
570 functions are identical except that C<newRV_inc> increments the reference
571 count of the C<thing>, while C<newRV_noinc> does not. For historical
572 reasons, C<newRV> is a synonym for C<newRV_inc>.
574 Once you have a reference, you can use the following macro to dereference
579 then call the appropriate routines, casting the returned C<SV*> to either an
580 C<AV*> or C<HV*>, if required.
582 To determine if an SV is a reference, you can use the following macro:
586 To discover what type of value the reference refers to, use the following
587 macro and then check the return value.
591 The most useful types that will be returned are:
600 SVt_PVGV Glob (possible a file handle)
601 SVt_PVMG Blessed or Magical Scalar
603 See the sv.h header file for more details.
605 =head2 Blessed References and Class Objects
607 References are also used to support object-oriented programming. In perl's
608 OO lexicon, an object is simply a reference that has been blessed into a
609 package (or class). Once blessed, the programmer may now use the reference
610 to access the various methods in the class.
612 A reference can be blessed into a package with the following function:
614 SV* sv_bless(SV* sv, HV* stash);
616 The C<sv> argument must be a reference value. The C<stash> argument
617 specifies which class the reference will belong to. See
618 L<Stashes and Globs> for information on converting class names into stashes.
620 /* Still under construction */
622 Upgrades rv to reference if not already one. Creates new SV for rv to
623 point to. If C<classname> is non-null, the SV is blessed into the specified
624 class. SV is returned.
626 SV* newSVrv(SV* rv, const char* classname);
628 Copies integer, unsigned integer or double into an SV whose reference is C<rv>. SV is blessed
629 if C<classname> is non-null.
631 SV* sv_setref_iv(SV* rv, const char* classname, IV iv);
632 SV* sv_setref_uv(SV* rv, const char* classname, UV uv);
633 SV* sv_setref_nv(SV* rv, const char* classname, NV iv);
635 Copies the pointer value (I<the address, not the string!>) into an SV whose
636 reference is rv. SV is blessed if C<classname> is non-null.
638 SV* sv_setref_pv(SV* rv, const char* classname, PV iv);
640 Copies string into an SV whose reference is C<rv>. Set length to 0 to let
641 Perl calculate the string length. SV is blessed if C<classname> is non-null.
643 SV* sv_setref_pvn(SV* rv, const char* classname, PV iv, STRLEN length);
645 Tests whether the SV is blessed into the specified class. It does not
646 check inheritance relationships.
648 int sv_isa(SV* sv, const char* name);
650 Tests whether the SV is a reference to a blessed object.
652 int sv_isobject(SV* sv);
654 Tests whether the SV is derived from the specified class. SV can be either
655 a reference to a blessed object or a string containing a class name. This
656 is the function implementing the C<UNIVERSAL::isa> functionality.
658 bool sv_derived_from(SV* sv, const char* name);
660 To check if you've got an object derived from a specific class you have
663 if (sv_isobject(sv) && sv_derived_from(sv, class)) { ... }
665 =head2 Creating New Variables
667 To create a new Perl variable with an undef value which can be accessed from
668 your Perl script, use the following routines, depending on the variable type.
670 SV* get_sv("package::varname", TRUE);
671 AV* get_av("package::varname", TRUE);
672 HV* get_hv("package::varname", TRUE);
674 Notice the use of TRUE as the second parameter. The new variable can now
675 be set, using the routines appropriate to the data type.
677 There are additional macros whose values may be bitwise OR'ed with the
678 C<TRUE> argument to enable certain extra features. Those bits are:
684 Marks the variable as multiply defined, thus preventing the:
686 Name <varname> used only once: possible typo
694 Had to create <varname> unexpectedly
696 if the variable did not exist before the function was called.
700 If you do not specify a package name, the variable is created in the current
703 =head2 Reference Counts and Mortality
705 Perl uses a reference count-driven garbage collection mechanism. SVs,
706 AVs, or HVs (xV for short in the following) start their life with a
707 reference count of 1. If the reference count of an xV ever drops to 0,
708 then it will be destroyed and its memory made available for reuse.
710 This normally doesn't happen at the Perl level unless a variable is
711 undef'ed or the last variable holding a reference to it is changed or
712 overwritten. At the internal level, however, reference counts can be
713 manipulated with the following macros:
715 int SvREFCNT(SV* sv);
716 SV* SvREFCNT_inc(SV* sv);
717 void SvREFCNT_dec(SV* sv);
719 However, there is one other function which manipulates the reference
720 count of its argument. The C<newRV_inc> function, you will recall,
721 creates a reference to the specified argument. As a side effect,
722 it increments the argument's reference count. If this is not what
723 you want, use C<newRV_noinc> instead.
725 For example, imagine you want to return a reference from an XSUB function.
726 Inside the XSUB routine, you create an SV which initially has a reference
727 count of one. Then you call C<newRV_inc>, passing it the just-created SV.
728 This returns the reference as a new SV, but the reference count of the
729 SV you passed to C<newRV_inc> has been incremented to two. Now you
730 return the reference from the XSUB routine and forget about the SV.
731 But Perl hasn't! Whenever the returned reference is destroyed, the
732 reference count of the original SV is decreased to one and nothing happens.
733 The SV will hang around without any way to access it until Perl itself
734 terminates. This is a memory leak.
736 The correct procedure, then, is to use C<newRV_noinc> instead of
737 C<newRV_inc>. Then, if and when the last reference is destroyed,
738 the reference count of the SV will go to zero and it will be destroyed,
739 stopping any memory leak.
741 There are some convenience functions available that can help with the
742 destruction of xVs. These functions introduce the concept of "mortality".
743 An xV that is mortal has had its reference count marked to be decremented,
744 but not actually decremented, until "a short time later". Generally the
745 term "short time later" means a single Perl statement, such as a call to
746 an XSUB function. The actual determinant for when mortal xVs have their
747 reference count decremented depends on two macros, SAVETMPS and FREETMPS.
748 See L<perlcall> and L<perlxs> for more details on these macros.
750 "Mortalization" then is at its simplest a deferred C<SvREFCNT_dec>.
751 However, if you mortalize a variable twice, the reference count will
752 later be decremented twice.
754 "Mortal" SVs are mainly used for SVs that are placed on perl's stack.
755 For example an SV which is created just to pass a number to a called sub
756 is made mortal to have it cleaned up automatically when it's popped off
757 the stack. Similarly, results returned by XSUBs (which are pushed on the
758 stack) are often made mortal.
760 To create a mortal variable, use the functions:
764 SV* sv_mortalcopy(SV*)
766 The first call creates a mortal SV (with no value), the second converts an existing
767 SV to a mortal SV (and thus defers a call to C<SvREFCNT_dec>), and the
768 third creates a mortal copy of an existing SV.
769 Because C<sv_newmortal> gives the new SV no value,it must normally be given one
770 via C<sv_setpv>, C<sv_setiv>, etc. :
772 SV *tmp = sv_newmortal();
773 sv_setiv(tmp, an_integer);
775 As that is multiple C statements it is quite common so see this idiom instead:
777 SV *tmp = sv_2mortal(newSViv(an_integer));
780 You should be careful about creating mortal variables. Strange things
781 can happen if you make the same value mortal within multiple contexts,
782 or if you make a variable mortal multiple times. Thinking of "Mortalization"
783 as deferred C<SvREFCNT_dec> should help to minimize such problems.
784 For example if you are passing an SV which you I<know> has high enough REFCNT
785 to survive its use on the stack you need not do any mortalization.
786 If you are not sure then doing an C<SvREFCNT_inc> and C<sv_2mortal>, or
787 making a C<sv_mortalcopy> is safer.
789 The mortal routines are not just for SVs -- AVs and HVs can be
790 made mortal by passing their address (type-casted to C<SV*>) to the
791 C<sv_2mortal> or C<sv_mortalcopy> routines.
793 =head2 Stashes and Globs
795 A B<stash> is a hash that contains all variables that are defined
796 within a package. Each key of the stash is a symbol
797 name (shared by all the different types of objects that have the same
798 name), and each value in the hash table is a GV (Glob Value). This GV
799 in turn contains references to the various objects of that name,
800 including (but not limited to) the following:
809 There is a single stash called C<PL_defstash> that holds the items that exist
810 in the C<main> package. To get at the items in other packages, append the
811 string "::" to the package name. The items in the C<Foo> package are in
812 the stash C<Foo::> in PL_defstash. The items in the C<Bar::Baz> package are
813 in the stash C<Baz::> in C<Bar::>'s stash.
815 To get the stash pointer for a particular package, use the function:
817 HV* gv_stashpv(const char* name, I32 create)
818 HV* gv_stashsv(SV*, I32 create)
820 The first function takes a literal string, the second uses the string stored
821 in the SV. Remember that a stash is just a hash table, so you get back an
822 C<HV*>. The C<create> flag will create a new package if it is set.
824 The name that C<gv_stash*v> wants is the name of the package whose symbol table
825 you want. The default package is called C<main>. If you have multiply nested
826 packages, pass their names to C<gv_stash*v>, separated by C<::> as in the Perl
829 Alternately, if you have an SV that is a blessed reference, you can find
830 out the stash pointer by using:
832 HV* SvSTASH(SvRV(SV*));
834 then use the following to get the package name itself:
836 char* HvNAME(HV* stash);
838 If you need to bless or re-bless an object you can use the following
841 SV* sv_bless(SV*, HV* stash)
843 where the first argument, an C<SV*>, must be a reference, and the second
844 argument is a stash. The returned C<SV*> can now be used in the same way
847 For more information on references and blessings, consult L<perlref>.
849 =head2 Double-Typed SVs
851 Scalar variables normally contain only one type of value, an integer,
852 double, pointer, or reference. Perl will automatically convert the
853 actual scalar data from the stored type into the requested type.
855 Some scalar variables contain more than one type of scalar data. For
856 example, the variable C<$!> contains either the numeric value of C<errno>
857 or its string equivalent from either C<strerror> or C<sys_errlist[]>.
859 To force multiple data values into an SV, you must do two things: use the
860 C<sv_set*v> routines to add the additional scalar type, then set a flag
861 so that Perl will believe it contains more than one type of data. The
862 four macros to set the flags are:
869 The particular macro you must use depends on which C<sv_set*v> routine
870 you called first. This is because every C<sv_set*v> routine turns on
871 only the bit for the particular type of data being set, and turns off
874 For example, to create a new Perl variable called "dberror" that contains
875 both the numeric and descriptive string error values, you could use the
879 extern char *dberror_list;
881 SV* sv = get_sv("dberror", TRUE);
882 sv_setiv(sv, (IV) dberror);
883 sv_setpv(sv, dberror_list[dberror]);
886 If the order of C<sv_setiv> and C<sv_setpv> had been reversed, then the
887 macro C<SvPOK_on> would need to be called instead of C<SvIOK_on>.
889 =head2 Magic Variables
891 [This section still under construction. Ignore everything here. Post no
892 bills. Everything not permitted is forbidden.]
894 Any SV may be magical, that is, it has special features that a normal
895 SV does not have. These features are stored in the SV structure in a
896 linked list of C<struct magic>'s, typedef'ed to C<MAGIC>.
909 Note this is current as of patchlevel 0, and could change at any time.
911 =head2 Assigning Magic
913 Perl adds magic to an SV using the sv_magic function:
915 void sv_magic(SV* sv, SV* obj, int how, const char* name, I32 namlen);
917 The C<sv> argument is a pointer to the SV that is to acquire a new magical
920 If C<sv> is not already magical, Perl uses the C<SvUPGRADE> macro to
921 convert C<sv> to type C<SVt_PVMG>. Perl then continues by adding new magic
922 to the beginning of the linked list of magical features. Any prior entry
923 of the same type of magic is deleted. Note that this can be overridden,
924 and multiple instances of the same type of magic can be associated with an
927 The C<name> and C<namlen> arguments are used to associate a string with
928 the magic, typically the name of a variable. C<namlen> is stored in the
929 C<mg_len> field and if C<name> is non-null then either a C<savepvn> copy of
930 C<name> or C<name> itself is stored in the C<mg_ptr> field, depending on
931 whether C<namlen> is greater than zero or equal to zero respectively. As a
932 special case, if C<(name && namlen == HEf_SVKEY)> then C<name> is assumed
933 to contain an C<SV*> and is stored as-is with its REFCNT incremented.
935 The sv_magic function uses C<how> to determine which, if any, predefined
936 "Magic Virtual Table" should be assigned to the C<mg_virtual> field.
937 See the L<Magic Virtual Tables> section below. The C<how> argument is also
938 stored in the C<mg_type> field. The value of C<how> should be chosen
939 from the set of macros C<PERL_MAGIC_foo> found in F<perl.h>. Note that before
940 these macros were added, Perl internals used to directly use character
941 literals, so you may occasionally come across old code or documentation
942 referring to 'U' magic rather than C<PERL_MAGIC_uvar> for example.
944 The C<obj> argument is stored in the C<mg_obj> field of the C<MAGIC>
945 structure. If it is not the same as the C<sv> argument, the reference
946 count of the C<obj> object is incremented. If it is the same, or if
947 the C<how> argument is C<PERL_MAGIC_arylen>, or if it is a NULL pointer,
948 then C<obj> is merely stored, without the reference count being incremented.
950 See also C<sv_magicext> in L<perlapi> for a more flexible way to add magic
953 There is also a function to add magic to an C<HV>:
955 void hv_magic(HV *hv, GV *gv, int how);
957 This simply calls C<sv_magic> and coerces the C<gv> argument into an C<SV>.
959 To remove the magic from an SV, call the function sv_unmagic:
961 void sv_unmagic(SV *sv, int type);
963 The C<type> argument should be equal to the C<how> value when the C<SV>
964 was initially made magical.
966 =head2 Magic Virtual Tables
968 The C<mg_virtual> field in the C<MAGIC> structure is a pointer to an
969 C<MGVTBL>, which is a structure of function pointers and stands for
970 "Magic Virtual Table" to handle the various operations that might be
971 applied to that variable.
973 The C<MGVTBL> has five pointers to the following routine types:
975 int (*svt_get)(SV* sv, MAGIC* mg);
976 int (*svt_set)(SV* sv, MAGIC* mg);
977 U32 (*svt_len)(SV* sv, MAGIC* mg);
978 int (*svt_clear)(SV* sv, MAGIC* mg);
979 int (*svt_free)(SV* sv, MAGIC* mg);
981 This MGVTBL structure is set at compile-time in F<perl.h> and there are
982 currently 19 types (or 21 with overloading turned on). These different
983 structures contain pointers to various routines that perform additional
984 actions depending on which function is being called.
986 Function pointer Action taken
987 ---------------- ------------
988 svt_get Do something before the value of the SV is retrieved.
989 svt_set Do something after the SV is assigned a value.
990 svt_len Report on the SV's length.
991 svt_clear Clear something the SV represents.
992 svt_free Free any extra storage associated with the SV.
994 For instance, the MGVTBL structure called C<vtbl_sv> (which corresponds
995 to an C<mg_type> of C<PERL_MAGIC_sv>) contains:
997 { magic_get, magic_set, magic_len, 0, 0 }
999 Thus, when an SV is determined to be magical and of type C<PERL_MAGIC_sv>,
1000 if a get operation is being performed, the routine C<magic_get> is
1001 called. All the various routines for the various magical types begin
1002 with C<magic_>. NOTE: the magic routines are not considered part of
1003 the Perl API, and may not be exported by the Perl library.
1005 The current kinds of Magic Virtual Tables are:
1008 (old-style char and macro) MGVTBL Type of magic
1009 -------------------------- ------ ----------------------------
1010 \0 PERL_MAGIC_sv vtbl_sv Special scalar variable
1011 A PERL_MAGIC_overload vtbl_amagic %OVERLOAD hash
1012 a PERL_MAGIC_overload_elem vtbl_amagicelem %OVERLOAD hash element
1013 c PERL_MAGIC_overload_table (none) Holds overload table (AMT)
1015 B PERL_MAGIC_bm vtbl_bm Boyer-Moore (fast string search)
1016 D PERL_MAGIC_regdata vtbl_regdata Regex match position data
1018 d PERL_MAGIC_regdatum vtbl_regdatum Regex match position data
1020 E PERL_MAGIC_env vtbl_env %ENV hash
1021 e PERL_MAGIC_envelem vtbl_envelem %ENV hash element
1022 f PERL_MAGIC_fm vtbl_fm Formline ('compiled' format)
1023 g PERL_MAGIC_regex_global vtbl_mglob m//g target / study()ed string
1024 I PERL_MAGIC_isa vtbl_isa @ISA array
1025 i PERL_MAGIC_isaelem vtbl_isaelem @ISA array element
1026 k PERL_MAGIC_nkeys vtbl_nkeys scalar(keys()) lvalue
1027 L PERL_MAGIC_dbfile (none) Debugger %_<filename
1028 l PERL_MAGIC_dbline vtbl_dbline Debugger %_<filename element
1029 m PERL_MAGIC_mutex vtbl_mutex ???
1030 o PERL_MAGIC_collxfrm vtbl_collxfrm Locale collate transformation
1031 P PERL_MAGIC_tied vtbl_pack Tied array or hash
1032 p PERL_MAGIC_tiedelem vtbl_packelem Tied array or hash element
1033 q PERL_MAGIC_tiedscalar vtbl_packelem Tied scalar or handle
1034 r PERL_MAGIC_qr vtbl_qr precompiled qr// regex
1035 S PERL_MAGIC_sig vtbl_sig %SIG hash
1036 s PERL_MAGIC_sigelem vtbl_sigelem %SIG hash element
1037 t PERL_MAGIC_taint vtbl_taint Taintedness
1038 U PERL_MAGIC_uvar vtbl_uvar Available for use by extensions
1039 v PERL_MAGIC_vec vtbl_vec vec() lvalue
1040 V PERL_MAGIC_vstring (none) v-string scalars
1041 w PERL_MAGIC_utf8 vtbl_utf8 UTF-8 length+offset cache
1042 x PERL_MAGIC_substr vtbl_substr substr() lvalue
1043 y PERL_MAGIC_defelem vtbl_defelem Shadow "foreach" iterator
1044 variable / smart parameter
1046 * PERL_MAGIC_glob vtbl_glob GV (typeglob)
1047 # PERL_MAGIC_arylen vtbl_arylen Array length ($#ary)
1048 . PERL_MAGIC_pos vtbl_pos pos() lvalue
1049 < PERL_MAGIC_backref vtbl_backref back pointer to a weak ref
1050 ~ PERL_MAGIC_ext (none) Available for use by extensions
1051 : PERL_MAGIC_symtab (none) hash used as symbol table
1052 % PERL_MAGIC_rhash (none) hash used as restricted hash
1053 @ PERL_MAGIC_arylen_p vtbl_arylen_p pointer to $#a from @a
1056 When an uppercase and lowercase letter both exist in the table, then the
1057 uppercase letter is typically used to represent some kind of composite type
1058 (a list or a hash), and the lowercase letter is used to represent an element
1059 of that composite type. Some internals code makes use of this case
1060 relationship. However, 'v' and 'V' (vec and v-string) are in no way related.
1062 The C<PERL_MAGIC_ext> and C<PERL_MAGIC_uvar> magic types are defined
1063 specifically for use by extensions and will not be used by perl itself.
1064 Extensions can use C<PERL_MAGIC_ext> magic to 'attach' private information
1065 to variables (typically objects). This is especially useful because
1066 there is no way for normal perl code to corrupt this private information
1067 (unlike using extra elements of a hash object).
1069 Similarly, C<PERL_MAGIC_uvar> magic can be used much like tie() to call a
1070 C function any time a scalar's value is used or changed. The C<MAGIC>'s
1071 C<mg_ptr> field points to a C<ufuncs> structure:
1074 I32 (*uf_val)(pTHX_ IV, SV*);
1075 I32 (*uf_set)(pTHX_ IV, SV*);
1079 When the SV is read from or written to, the C<uf_val> or C<uf_set>
1080 function will be called with C<uf_index> as the first arg and a pointer to
1081 the SV as the second. A simple example of how to add C<PERL_MAGIC_uvar>
1082 magic is shown below. Note that the ufuncs structure is copied by
1083 sv_magic, so you can safely allocate it on the stack.
1091 uf.uf_val = &my_get_fn;
1092 uf.uf_set = &my_set_fn;
1094 sv_magic(sv, 0, PERL_MAGIC_uvar, (char*)&uf, sizeof(uf));
1096 Note that because multiple extensions may be using C<PERL_MAGIC_ext>
1097 or C<PERL_MAGIC_uvar> magic, it is important for extensions to take
1098 extra care to avoid conflict. Typically only using the magic on
1099 objects blessed into the same class as the extension is sufficient.
1100 For C<PERL_MAGIC_ext> magic, it may also be appropriate to add an I32
1101 'signature' at the top of the private data area and check that.
1103 Also note that the C<sv_set*()> and C<sv_cat*()> functions described
1104 earlier do B<not> invoke 'set' magic on their targets. This must
1105 be done by the user either by calling the C<SvSETMAGIC()> macro after
1106 calling these functions, or by using one of the C<sv_set*_mg()> or
1107 C<sv_cat*_mg()> functions. Similarly, generic C code must call the
1108 C<SvGETMAGIC()> macro to invoke any 'get' magic if they use an SV
1109 obtained from external sources in functions that don't handle magic.
1110 See L<perlapi> for a description of these functions.
1111 For example, calls to the C<sv_cat*()> functions typically need to be
1112 followed by C<SvSETMAGIC()>, but they don't need a prior C<SvGETMAGIC()>
1113 since their implementation handles 'get' magic.
1115 =head2 Finding Magic
1117 MAGIC* mg_find(SV*, int type); /* Finds the magic pointer of that type */
1119 This routine returns a pointer to the C<MAGIC> structure stored in the SV.
1120 If the SV does not have that magical feature, C<NULL> is returned. Also,
1121 if the SV is not of type SVt_PVMG, Perl may core dump.
1123 int mg_copy(SV* sv, SV* nsv, const char* key, STRLEN klen);
1125 This routine checks to see what types of magic C<sv> has. If the mg_type
1126 field is an uppercase letter, then the mg_obj is copied to C<nsv>, but
1127 the mg_type field is changed to be the lowercase letter.
1129 =head2 Understanding the Magic of Tied Hashes and Arrays
1131 Tied hashes and arrays are magical beasts of the C<PERL_MAGIC_tied>
1134 WARNING: As of the 5.004 release, proper usage of the array and hash
1135 access functions requires understanding a few caveats. Some
1136 of these caveats are actually considered bugs in the API, to be fixed
1137 in later releases, and are bracketed with [MAYCHANGE] below. If
1138 you find yourself actually applying such information in this section, be
1139 aware that the behavior may change in the future, umm, without warning.
1141 The perl tie function associates a variable with an object that implements
1142 the various GET, SET, etc methods. To perform the equivalent of the perl
1143 tie function from an XSUB, you must mimic this behaviour. The code below
1144 carries out the necessary steps - firstly it creates a new hash, and then
1145 creates a second hash which it blesses into the class which will implement
1146 the tie methods. Lastly it ties the two hashes together, and returns a
1147 reference to the new tied hash. Note that the code below does NOT call the
1148 TIEHASH method in the MyTie class -
1149 see L<Calling Perl Routines from within C Programs> for details on how
1160 tie = newRV_noinc((SV*)newHV());
1161 stash = gv_stashpv("MyTie", TRUE);
1162 sv_bless(tie, stash);
1163 hv_magic(hash, (GV*)tie, PERL_MAGIC_tied);
1164 RETVAL = newRV_noinc(hash);
1168 The C<av_store> function, when given a tied array argument, merely
1169 copies the magic of the array onto the value to be "stored", using
1170 C<mg_copy>. It may also return NULL, indicating that the value did not
1171 actually need to be stored in the array. [MAYCHANGE] After a call to
1172 C<av_store> on a tied array, the caller will usually need to call
1173 C<mg_set(val)> to actually invoke the perl level "STORE" method on the
1174 TIEARRAY object. If C<av_store> did return NULL, a call to
1175 C<SvREFCNT_dec(val)> will also be usually necessary to avoid a memory
1178 The previous paragraph is applicable verbatim to tied hash access using the
1179 C<hv_store> and C<hv_store_ent> functions as well.
1181 C<av_fetch> and the corresponding hash functions C<hv_fetch> and
1182 C<hv_fetch_ent> actually return an undefined mortal value whose magic
1183 has been initialized using C<mg_copy>. Note the value so returned does not
1184 need to be deallocated, as it is already mortal. [MAYCHANGE] But you will
1185 need to call C<mg_get()> on the returned value in order to actually invoke
1186 the perl level "FETCH" method on the underlying TIE object. Similarly,
1187 you may also call C<mg_set()> on the return value after possibly assigning
1188 a suitable value to it using C<sv_setsv>, which will invoke the "STORE"
1189 method on the TIE object. [/MAYCHANGE]
1192 In other words, the array or hash fetch/store functions don't really
1193 fetch and store actual values in the case of tied arrays and hashes. They
1194 merely call C<mg_copy> to attach magic to the values that were meant to be
1195 "stored" or "fetched". Later calls to C<mg_get> and C<mg_set> actually
1196 do the job of invoking the TIE methods on the underlying objects. Thus
1197 the magic mechanism currently implements a kind of lazy access to arrays
1200 Currently (as of perl version 5.004), use of the hash and array access
1201 functions requires the user to be aware of whether they are operating on
1202 "normal" hashes and arrays, or on their tied variants. The API may be
1203 changed to provide more transparent access to both tied and normal data
1204 types in future versions.
1207 You would do well to understand that the TIEARRAY and TIEHASH interfaces
1208 are mere sugar to invoke some perl method calls while using the uniform hash
1209 and array syntax. The use of this sugar imposes some overhead (typically
1210 about two to four extra opcodes per FETCH/STORE operation, in addition to
1211 the creation of all the mortal variables required to invoke the methods).
1212 This overhead will be comparatively small if the TIE methods are themselves
1213 substantial, but if they are only a few statements long, the overhead
1214 will not be insignificant.
1216 =head2 Localizing changes
1218 Perl has a very handy construction
1225 This construction is I<approximately> equivalent to
1234 The biggest difference is that the first construction would
1235 reinstate the initial value of $var, irrespective of how control exits
1236 the block: C<goto>, C<return>, C<die>/C<eval>, etc. It is a little bit
1237 more efficient as well.
1239 There is a way to achieve a similar task from C via Perl API: create a
1240 I<pseudo-block>, and arrange for some changes to be automatically
1241 undone at the end of it, either explicit, or via a non-local exit (via
1242 die()). A I<block>-like construct is created by a pair of
1243 C<ENTER>/C<LEAVE> macros (see L<perlcall/"Returning a Scalar">).
1244 Such a construct may be created specially for some important localized
1245 task, or an existing one (like boundaries of enclosing Perl
1246 subroutine/block, or an existing pair for freeing TMPs) may be
1247 used. (In the second case the overhead of additional localization must
1248 be almost negligible.) Note that any XSUB is automatically enclosed in
1249 an C<ENTER>/C<LEAVE> pair.
1251 Inside such a I<pseudo-block> the following service is available:
1255 =item C<SAVEINT(int i)>
1257 =item C<SAVEIV(IV i)>
1259 =item C<SAVEI32(I32 i)>
1261 =item C<SAVELONG(long i)>
1263 These macros arrange things to restore the value of integer variable
1264 C<i> at the end of enclosing I<pseudo-block>.
1266 =item C<SAVESPTR(s)>
1268 =item C<SAVEPPTR(p)>
1270 These macros arrange things to restore the value of pointers C<s> and
1271 C<p>. C<s> must be a pointer of a type which survives conversion to
1272 C<SV*> and back, C<p> should be able to survive conversion to C<char*>
1275 =item C<SAVEFREESV(SV *sv)>
1277 The refcount of C<sv> would be decremented at the end of
1278 I<pseudo-block>. This is similar to C<sv_2mortal> in that it is also a
1279 mechanism for doing a delayed C<SvREFCNT_dec>. However, while C<sv_2mortal>
1280 extends the lifetime of C<sv> until the beginning of the next statement,
1281 C<SAVEFREESV> extends it until the end of the enclosing scope. These
1282 lifetimes can be wildly different.
1284 Also compare C<SAVEMORTALIZESV>.
1286 =item C<SAVEMORTALIZESV(SV *sv)>
1288 Just like C<SAVEFREESV>, but mortalizes C<sv> at the end of the current
1289 scope instead of decrementing its reference count. This usually has the
1290 effect of keeping C<sv> alive until the statement that called the currently
1291 live scope has finished executing.
1293 =item C<SAVEFREEOP(OP *op)>
1295 The C<OP *> is op_free()ed at the end of I<pseudo-block>.
1297 =item C<SAVEFREEPV(p)>
1299 The chunk of memory which is pointed to by C<p> is Safefree()ed at the
1300 end of I<pseudo-block>.
1302 =item C<SAVECLEARSV(SV *sv)>
1304 Clears a slot in the current scratchpad which corresponds to C<sv> at
1305 the end of I<pseudo-block>.
1307 =item C<SAVEDELETE(HV *hv, char *key, I32 length)>
1309 The key C<key> of C<hv> is deleted at the end of I<pseudo-block>. The
1310 string pointed to by C<key> is Safefree()ed. If one has a I<key> in
1311 short-lived storage, the corresponding string may be reallocated like
1314 SAVEDELETE(PL_defstash, savepv(tmpbuf), strlen(tmpbuf));
1316 =item C<SAVEDESTRUCTOR(DESTRUCTORFUNC_NOCONTEXT_t f, void *p)>
1318 At the end of I<pseudo-block> the function C<f> is called with the
1321 =item C<SAVEDESTRUCTOR_X(DESTRUCTORFUNC_t f, void *p)>
1323 At the end of I<pseudo-block> the function C<f> is called with the
1324 implicit context argument (if any), and C<p>.
1326 =item C<SAVESTACK_POS()>
1328 The current offset on the Perl internal stack (cf. C<SP>) is restored
1329 at the end of I<pseudo-block>.
1333 The following API list contains functions, thus one needs to
1334 provide pointers to the modifiable data explicitly (either C pointers,
1335 or Perlish C<GV *>s). Where the above macros take C<int>, a similar
1336 function takes C<int *>.
1340 =item C<SV* save_scalar(GV *gv)>
1342 Equivalent to Perl code C<local $gv>.
1344 =item C<AV* save_ary(GV *gv)>
1346 =item C<HV* save_hash(GV *gv)>
1348 Similar to C<save_scalar>, but localize C<@gv> and C<%gv>.
1350 =item C<void save_item(SV *item)>
1352 Duplicates the current value of C<SV>, on the exit from the current
1353 C<ENTER>/C<LEAVE> I<pseudo-block> will restore the value of C<SV>
1354 using the stored value. It doesn't handle magic. Use C<save_scalar> if
1357 =item C<void save_list(SV **sarg, I32 maxsarg)>
1359 A variant of C<save_item> which takes multiple arguments via an array
1360 C<sarg> of C<SV*> of length C<maxsarg>.
1362 =item C<SV* save_svref(SV **sptr)>
1364 Similar to C<save_scalar>, but will reinstate an C<SV *>.
1366 =item C<void save_aptr(AV **aptr)>
1368 =item C<void save_hptr(HV **hptr)>
1370 Similar to C<save_svref>, but localize C<AV *> and C<HV *>.
1374 The C<Alias> module implements localization of the basic types within the
1375 I<caller's scope>. People who are interested in how to localize things in
1376 the containing scope should take a look there too.
1380 =head2 XSUBs and the Argument Stack
1382 The XSUB mechanism is a simple way for Perl programs to access C subroutines.
1383 An XSUB routine will have a stack that contains the arguments from the Perl
1384 program, and a way to map from the Perl data structures to a C equivalent.
1386 The stack arguments are accessible through the C<ST(n)> macro, which returns
1387 the C<n>'th stack argument. Argument 0 is the first argument passed in the
1388 Perl subroutine call. These arguments are C<SV*>, and can be used anywhere
1391 Most of the time, output from the C routine can be handled through use of
1392 the RETVAL and OUTPUT directives. However, there are some cases where the
1393 argument stack is not already long enough to handle all the return values.
1394 An example is the POSIX tzname() call, which takes no arguments, but returns
1395 two, the local time zone's standard and summer time abbreviations.
1397 To handle this situation, the PPCODE directive is used and the stack is
1398 extended using the macro:
1402 where C<SP> is the macro that represents the local copy of the stack pointer,
1403 and C<num> is the number of elements the stack should be extended by.
1405 Now that there is room on the stack, values can be pushed on it using C<PUSHs>
1406 macro. The pushed values will often need to be "mortal" (See
1407 L</Reference Counts and Mortality>):
1409 PUSHs(sv_2mortal(newSViv(an_integer)))
1410 PUSHs(sv_2mortal(newSVuv(an_unsigned_integer)))
1411 PUSHs(sv_2mortal(newSVnv(a_double)))
1412 PUSHs(sv_2mortal(newSVpv("Some String",0)))
1414 And now the Perl program calling C<tzname>, the two values will be assigned
1417 ($standard_abbrev, $summer_abbrev) = POSIX::tzname;
1419 An alternate (and possibly simpler) method to pushing values on the stack is
1424 This macro automatically adjust the stack for you, if needed. Thus, you
1425 do not need to call C<EXTEND> to extend the stack.
1427 Despite their suggestions in earlier versions of this document the macros
1428 C<(X)PUSH[iunp]> are I<not> suited to XSUBs which return multiple results.
1429 For that, either stick to the C<(X)PUSHs> macros shown above, or use the new
1430 C<m(X)PUSH[iunp]> macros instead; see L</Putting a C value on Perl stack>.
1432 For more information, consult L<perlxs> and L<perlxstut>.
1434 =head2 Calling Perl Routines from within C Programs
1436 There are four routines that can be used to call a Perl subroutine from
1437 within a C program. These four are:
1439 I32 call_sv(SV*, I32);
1440 I32 call_pv(const char*, I32);
1441 I32 call_method(const char*, I32);
1442 I32 call_argv(const char*, I32, register char**);
1444 The routine most often used is C<call_sv>. The C<SV*> argument
1445 contains either the name of the Perl subroutine to be called, or a
1446 reference to the subroutine. The second argument consists of flags
1447 that control the context in which the subroutine is called, whether
1448 or not the subroutine is being passed arguments, how errors should be
1449 trapped, and how to treat return values.
1451 All four routines return the number of arguments that the subroutine returned
1454 These routines used to be called C<perl_call_sv>, etc., before Perl v5.6.0,
1455 but those names are now deprecated; macros of the same name are provided for
1458 When using any of these routines (except C<call_argv>), the programmer
1459 must manipulate the Perl stack. These include the following macros and
1474 For a detailed description of calling conventions from C to Perl,
1475 consult L<perlcall>.
1477 =head2 Memory Allocation
1481 All memory meant to be used with the Perl API functions should be manipulated
1482 using the macros described in this section. The macros provide the necessary
1483 transparency between differences in the actual malloc implementation that is
1486 It is suggested that you enable the version of malloc that is distributed
1487 with Perl. It keeps pools of various sizes of unallocated memory in
1488 order to satisfy allocation requests more quickly. However, on some
1489 platforms, it may cause spurious malloc or free errors.
1491 The following three macros are used to initially allocate memory :
1493 Newx(pointer, number, type);
1494 Newxc(pointer, number, type, cast);
1495 Newxz(pointer, number, type);
1497 The first argument C<pointer> should be the name of a variable that will
1498 point to the newly allocated memory.
1500 The second and third arguments C<number> and C<type> specify how many of
1501 the specified type of data structure should be allocated. The argument
1502 C<type> is passed to C<sizeof>. The final argument to C<Newxc>, C<cast>,
1503 should be used if the C<pointer> argument is different from the C<type>
1506 Unlike the C<Newx> and C<Newxc> macros, the C<Newxz> macro calls C<memzero>
1507 to zero out all the newly allocated memory.
1511 Renew(pointer, number, type);
1512 Renewc(pointer, number, type, cast);
1515 These three macros are used to change a memory buffer size or to free a
1516 piece of memory no longer needed. The arguments to C<Renew> and C<Renewc>
1517 match those of C<New> and C<Newc> with the exception of not needing the
1518 "magic cookie" argument.
1522 Move(source, dest, number, type);
1523 Copy(source, dest, number, type);
1524 Zero(dest, number, type);
1526 These three macros are used to move, copy, or zero out previously allocated
1527 memory. The C<source> and C<dest> arguments point to the source and
1528 destination starting points. Perl will move, copy, or zero out C<number>
1529 instances of the size of the C<type> data structure (using the C<sizeof>
1534 The most recent development releases of Perl has been experimenting with
1535 removing Perl's dependency on the "normal" standard I/O suite and allowing
1536 other stdio implementations to be used. This involves creating a new
1537 abstraction layer that then calls whichever implementation of stdio Perl
1538 was compiled with. All XSUBs should now use the functions in the PerlIO
1539 abstraction layer and not make any assumptions about what kind of stdio
1542 For a complete description of the PerlIO abstraction, consult L<perlapio>.
1544 =head2 Putting a C value on Perl stack
1546 A lot of opcodes (this is an elementary operation in the internal perl
1547 stack machine) put an SV* on the stack. However, as an optimization
1548 the corresponding SV is (usually) not recreated each time. The opcodes
1549 reuse specially assigned SVs (I<target>s) which are (as a corollary)
1550 not constantly freed/created.
1552 Each of the targets is created only once (but see
1553 L<Scratchpads and recursion> below), and when an opcode needs to put
1554 an integer, a double, or a string on stack, it just sets the
1555 corresponding parts of its I<target> and puts the I<target> on stack.
1557 The macro to put this target on stack is C<PUSHTARG>, and it is
1558 directly used in some opcodes, as well as indirectly in zillions of
1559 others, which use it via C<(X)PUSH[iunp]>.
1561 Because the target is reused, you must be careful when pushing multiple
1562 values on the stack. The following code will not do what you think:
1567 This translates as "set C<TARG> to 10, push a pointer to C<TARG> onto
1568 the stack; set C<TARG> to 20, push a pointer to C<TARG> onto the stack".
1569 At the end of the operation, the stack does not contain the values 10
1570 and 20, but actually contains two pointers to C<TARG>, which we have set
1573 If you need to push multiple different values then you should either use
1574 the C<(X)PUSHs> macros, or else use the new C<m(X)PUSH[iunp]> macros,
1575 none of which make use of C<TARG>. The C<(X)PUSHs> macros simply push an
1576 SV* on the stack, which, as noted under L</XSUBs and the Argument Stack>,
1577 will often need to be "mortal". The new C<m(X)PUSH[iunp]> macros make
1578 this a little easier to achieve by creating a new mortal for you (via
1579 C<(X)PUSHmortal>), pushing that onto the stack (extending it if necessary
1580 in the case of the C<mXPUSH[iunp]> macros), and then setting its value.
1581 Thus, instead of writing this to "fix" the example above:
1583 XPUSHs(sv_2mortal(newSViv(10)))
1584 XPUSHs(sv_2mortal(newSViv(20)))
1586 you can simply write:
1591 On a related note, if you do use C<(X)PUSH[iunp]>, then you're going to
1592 need a C<dTARG> in your variable declarations so that the C<*PUSH*>
1593 macros can make use of the local variable C<TARG>. See also C<dTARGET>
1598 The question remains on when the SVs which are I<target>s for opcodes
1599 are created. The answer is that they are created when the current unit --
1600 a subroutine or a file (for opcodes for statements outside of
1601 subroutines) -- is compiled. During this time a special anonymous Perl
1602 array is created, which is called a scratchpad for the current
1605 A scratchpad keeps SVs which are lexicals for the current unit and are
1606 targets for opcodes. One can deduce that an SV lives on a scratchpad
1607 by looking on its flags: lexicals have C<SVs_PADMY> set, and
1608 I<target>s have C<SVs_PADTMP> set.
1610 The correspondence between OPs and I<target>s is not 1-to-1. Different
1611 OPs in the compile tree of the unit can use the same target, if this
1612 would not conflict with the expected life of the temporary.
1614 =head2 Scratchpads and recursion
1616 In fact it is not 100% true that a compiled unit contains a pointer to
1617 the scratchpad AV. In fact it contains a pointer to an AV of
1618 (initially) one element, and this element is the scratchpad AV. Why do
1619 we need an extra level of indirection?
1621 The answer is B<recursion>, and maybe B<threads>. Both
1622 these can create several execution pointers going into the same
1623 subroutine. For the subroutine-child not write over the temporaries
1624 for the subroutine-parent (lifespan of which covers the call to the
1625 child), the parent and the child should have different
1626 scratchpads. (I<And> the lexicals should be separate anyway!)
1628 So each subroutine is born with an array of scratchpads (of length 1).
1629 On each entry to the subroutine it is checked that the current
1630 depth of the recursion is not more than the length of this array, and
1631 if it is, new scratchpad is created and pushed into the array.
1633 The I<target>s on this scratchpad are C<undef>s, but they are already
1634 marked with correct flags.
1636 =head1 Compiled code
1640 Here we describe the internal form your code is converted to by
1641 Perl. Start with a simple example:
1645 This is converted to a tree similar to this one:
1653 (but slightly more complicated). This tree reflects the way Perl
1654 parsed your code, but has nothing to do with the execution order.
1655 There is an additional "thread" going through the nodes of the tree
1656 which shows the order of execution of the nodes. In our simplified
1657 example above it looks like:
1659 $b ---> $c ---> + ---> $a ---> assign-to
1661 But with the actual compile tree for C<$a = $b + $c> it is different:
1662 some nodes I<optimized away>. As a corollary, though the actual tree
1663 contains more nodes than our simplified example, the execution order
1664 is the same as in our example.
1666 =head2 Examining the tree
1668 If you have your perl compiled for debugging (usually done with
1669 C<-DDEBUGGING> on the C<Configure> command line), you may examine the
1670 compiled tree by specifying C<-Dx> on the Perl command line. The
1671 output takes several lines per node, and for C<$b+$c> it looks like
1676 FLAGS = (SCALAR,KIDS)
1678 TYPE = null ===> (4)
1680 FLAGS = (SCALAR,KIDS)
1682 3 TYPE = gvsv ===> 4
1688 TYPE = null ===> (5)
1690 FLAGS = (SCALAR,KIDS)
1692 4 TYPE = gvsv ===> 5
1698 This tree has 5 nodes (one per C<TYPE> specifier), only 3 of them are
1699 not optimized away (one per number in the left column). The immediate
1700 children of the given node correspond to C<{}> pairs on the same level
1701 of indentation, thus this listing corresponds to the tree:
1709 The execution order is indicated by C<===E<gt>> marks, thus it is C<3
1710 4 5 6> (node C<6> is not included into above listing), i.e.,
1711 C<gvsv gvsv add whatever>.
1713 Each of these nodes represents an op, a fundamental operation inside the
1714 Perl core. The code which implements each operation can be found in the
1715 F<pp*.c> files; the function which implements the op with type C<gvsv>
1716 is C<pp_gvsv>, and so on. As the tree above shows, different ops have
1717 different numbers of children: C<add> is a binary operator, as one would
1718 expect, and so has two children. To accommodate the various different
1719 numbers of children, there are various types of op data structure, and
1720 they link together in different ways.
1722 The simplest type of op structure is C<OP>: this has no children. Unary
1723 operators, C<UNOP>s, have one child, and this is pointed to by the
1724 C<op_first> field. Binary operators (C<BINOP>s) have not only an
1725 C<op_first> field but also an C<op_last> field. The most complex type of
1726 op is a C<LISTOP>, which has any number of children. In this case, the
1727 first child is pointed to by C<op_first> and the last child by
1728 C<op_last>. The children in between can be found by iteratively
1729 following the C<op_sibling> pointer from the first child to the last.
1731 There are also two other op types: a C<PMOP> holds a regular expression,
1732 and has no children, and a C<LOOP> may or may not have children. If the
1733 C<op_children> field is non-zero, it behaves like a C<LISTOP>. To
1734 complicate matters, if a C<UNOP> is actually a C<null> op after
1735 optimization (see L</Compile pass 2: context propagation>) it will still
1736 have children in accordance with its former type.
1738 Another way to examine the tree is to use a compiler back-end module, such
1741 =head2 Compile pass 1: check routines
1743 The tree is created by the compiler while I<yacc> code feeds it
1744 the constructions it recognizes. Since I<yacc> works bottom-up, so does
1745 the first pass of perl compilation.
1747 What makes this pass interesting for perl developers is that some
1748 optimization may be performed on this pass. This is optimization by
1749 so-called "check routines". The correspondence between node names
1750 and corresponding check routines is described in F<opcode.pl> (do not
1751 forget to run C<make regen_headers> if you modify this file).
1753 A check routine is called when the node is fully constructed except
1754 for the execution-order thread. Since at this time there are no
1755 back-links to the currently constructed node, one can do most any
1756 operation to the top-level node, including freeing it and/or creating
1757 new nodes above/below it.
1759 The check routine returns the node which should be inserted into the
1760 tree (if the top-level node was not modified, check routine returns
1763 By convention, check routines have names C<ck_*>. They are usually
1764 called from C<new*OP> subroutines (or C<convert>) (which in turn are
1765 called from F<perly.y>).
1767 =head2 Compile pass 1a: constant folding
1769 Immediately after the check routine is called the returned node is
1770 checked for being compile-time executable. If it is (the value is
1771 judged to be constant) it is immediately executed, and a I<constant>
1772 node with the "return value" of the corresponding subtree is
1773 substituted instead. The subtree is deleted.
1775 If constant folding was not performed, the execution-order thread is
1778 =head2 Compile pass 2: context propagation
1780 When a context for a part of compile tree is known, it is propagated
1781 down through the tree. At this time the context can have 5 values
1782 (instead of 2 for runtime context): void, boolean, scalar, list, and
1783 lvalue. In contrast with the pass 1 this pass is processed from top
1784 to bottom: a node's context determines the context for its children.
1786 Additional context-dependent optimizations are performed at this time.
1787 Since at this moment the compile tree contains back-references (via
1788 "thread" pointers), nodes cannot be free()d now. To allow
1789 optimized-away nodes at this stage, such nodes are null()ified instead
1790 of free()ing (i.e. their type is changed to OP_NULL).
1792 =head2 Compile pass 3: peephole optimization
1794 After the compile tree for a subroutine (or for an C<eval> or a file)
1795 is created, an additional pass over the code is performed. This pass
1796 is neither top-down or bottom-up, but in the execution order (with
1797 additional complications for conditionals). These optimizations are
1798 done in the subroutine peep(). Optimizations performed at this stage
1799 are subject to the same restrictions as in the pass 2.
1801 =head2 Pluggable runops
1803 The compile tree is executed in a runops function. There are two runops
1804 functions, in F<run.c> and in F<dump.c>. C<Perl_runops_debug> is used
1805 with DEBUGGING and C<Perl_runops_standard> is used otherwise. For fine
1806 control over the execution of the compile tree it is possible to provide
1807 your own runops function.
1809 It's probably best to copy one of the existing runops functions and
1810 change it to suit your needs. Then, in the BOOT section of your XS
1813 PL_runops = my_runops;
1815 This function should be as efficient as possible to keep your programs
1816 running as fast as possible.
1818 =head1 Examining internal data structures with the C<dump> functions
1820 To aid debugging, the source file F<dump.c> contains a number of
1821 functions which produce formatted output of internal data structures.
1823 The most commonly used of these functions is C<Perl_sv_dump>; it's used
1824 for dumping SVs, AVs, HVs, and CVs. The C<Devel::Peek> module calls
1825 C<sv_dump> to produce debugging output from Perl-space, so users of that
1826 module should already be familiar with its format.
1828 C<Perl_op_dump> can be used to dump an C<OP> structure or any of its
1829 derivatives, and produces output similar to C<perl -Dx>; in fact,
1830 C<Perl_dump_eval> will dump the main root of the code being evaluated,
1831 exactly like C<-Dx>.
1833 Other useful functions are C<Perl_dump_sub>, which turns a C<GV> into an
1834 op tree, C<Perl_dump_packsubs> which calls C<Perl_dump_sub> on all the
1835 subroutines in a package like so: (Thankfully, these are all xsubs, so
1836 there is no op tree)
1838 (gdb) print Perl_dump_packsubs(PL_defstash)
1840 SUB attributes::bootstrap = (xsub 0x811fedc 0)
1842 SUB UNIVERSAL::can = (xsub 0x811f50c 0)
1844 SUB UNIVERSAL::isa = (xsub 0x811f304 0)
1846 SUB UNIVERSAL::VERSION = (xsub 0x811f7ac 0)
1848 SUB DynaLoader::boot_DynaLoader = (xsub 0x805b188 0)
1850 and C<Perl_dump_all>, which dumps all the subroutines in the stash and
1851 the op tree of the main root.
1853 =head1 How multiple interpreters and concurrency are supported
1855 =head2 Background and PERL_IMPLICIT_CONTEXT
1857 The Perl interpreter can be regarded as a closed box: it has an API
1858 for feeding it code or otherwise making it do things, but it also has
1859 functions for its own use. This smells a lot like an object, and
1860 there are ways for you to build Perl so that you can have multiple
1861 interpreters, with one interpreter represented either as a C structure,
1862 or inside a thread-specific structure. These structures contain all
1863 the context, the state of that interpreter.
1865 Two macros control the major Perl build flavors: MULTIPLICITY and
1866 USE_5005THREADS. The MULTIPLICITY build has a C structure
1867 that packages all the interpreter state, and there is a similar thread-specific
1868 data structure under USE_5005THREADS. In both cases,
1869 PERL_IMPLICIT_CONTEXT is also normally defined, and enables the
1870 support for passing in a "hidden" first argument that represents all three
1873 Two other "encapsulation" macros are the PERL_GLOBAL_STRUCT and
1874 PERL_GLOBAL_STRUCT_PRIVATE (the latter turns on the former, and the
1875 former turns on MULTIPLICITY.) The PERL_GLOBAL_STRUCT causes all the
1876 internal variables of Perl to be wrapped inside a single global struct,
1877 struct perl_vars, accessible as (globals) &PL_Vars or PL_VarsPtr or
1878 the function Perl_GetVars(). The PERL_GLOBAL_STRUCT_PRIVATE goes
1879 one step further, there is still a single struct (allocated in main()
1880 either from heap or from stack) but there are no global data symbols
1881 pointing to it. In either case the global struct should be initialised
1882 as the very first thing in main() using Perl_init_global_struct() and
1883 correspondingly tear it down after perl_free() using Perl_free_global_struct(),
1884 please see F<miniperlmain.c> for usage details. You may also need
1885 to use C<dVAR> in your coding to "declare the global variables"
1886 when you are using them. dTHX does this for you automatically.
1888 For backward compatibility reasons defining just PERL_GLOBAL_STRUCT
1889 doesn't actually hide all symbols inside a big global struct: some
1890 PerlIO_xxx vtables are left visible. The PERL_GLOBAL_STRUCT_PRIVATE
1891 then hides everything (see how the PERLIO_FUNCS_DECL is used).
1893 All this obviously requires a way for the Perl internal functions to be
1894 either subroutines taking some kind of structure as the first
1895 argument, or subroutines taking nothing as the first argument. To
1896 enable these two very different ways of building the interpreter,
1897 the Perl source (as it does in so many other situations) makes heavy
1898 use of macros and subroutine naming conventions.
1900 First problem: deciding which functions will be public API functions and
1901 which will be private. All functions whose names begin C<S_> are private
1902 (think "S" for "secret" or "static"). All other functions begin with
1903 "Perl_", but just because a function begins with "Perl_" does not mean it is
1904 part of the API. (See L</Internal Functions>.) The easiest way to be B<sure> a
1905 function is part of the API is to find its entry in L<perlapi>.
1906 If it exists in L<perlapi>, it's part of the API. If it doesn't, and you
1907 think it should be (i.e., you need it for your extension), send mail via
1908 L<perlbug> explaining why you think it should be.
1910 Second problem: there must be a syntax so that the same subroutine
1911 declarations and calls can pass a structure as their first argument,
1912 or pass nothing. To solve this, the subroutines are named and
1913 declared in a particular way. Here's a typical start of a static
1914 function used within the Perl guts:
1917 S_incline(pTHX_ char *s)
1919 STATIC becomes "static" in C, and may be #define'd to nothing in some
1920 configurations in future.
1922 A public function (i.e. part of the internal API, but not necessarily
1923 sanctioned for use in extensions) begins like this:
1926 Perl_sv_setiv(pTHX_ SV* dsv, IV num)
1928 C<pTHX_> is one of a number of macros (in perl.h) that hide the
1929 details of the interpreter's context. THX stands for "thread", "this",
1930 or "thingy", as the case may be. (And no, George Lucas is not involved. :-)
1931 The first character could be 'p' for a B<p>rototype, 'a' for B<a>rgument,
1932 or 'd' for B<d>eclaration, so we have C<pTHX>, C<aTHX> and C<dTHX>, and
1935 When Perl is built without options that set PERL_IMPLICIT_CONTEXT, there is no
1936 first argument containing the interpreter's context. The trailing underscore
1937 in the pTHX_ macro indicates that the macro expansion needs a comma
1938 after the context argument because other arguments follow it. If
1939 PERL_IMPLICIT_CONTEXT is not defined, pTHX_ will be ignored, and the
1940 subroutine is not prototyped to take the extra argument. The form of the
1941 macro without the trailing underscore is used when there are no additional
1944 When a core function calls another, it must pass the context. This
1945 is normally hidden via macros. Consider C<sv_setiv>. It expands into
1946 something like this:
1948 #ifdef PERL_IMPLICIT_CONTEXT
1949 #define sv_setiv(a,b) Perl_sv_setiv(aTHX_ a, b)
1950 /* can't do this for vararg functions, see below */
1952 #define sv_setiv Perl_sv_setiv
1955 This works well, and means that XS authors can gleefully write:
1959 and still have it work under all the modes Perl could have been
1962 This doesn't work so cleanly for varargs functions, though, as macros
1963 imply that the number of arguments is known in advance. Instead we
1964 either need to spell them out fully, passing C<aTHX_> as the first
1965 argument (the Perl core tends to do this with functions like
1966 Perl_warner), or use a context-free version.
1968 The context-free version of Perl_warner is called
1969 Perl_warner_nocontext, and does not take the extra argument. Instead
1970 it does dTHX; to get the context from thread-local storage. We
1971 C<#define warner Perl_warner_nocontext> so that extensions get source
1972 compatibility at the expense of performance. (Passing an arg is
1973 cheaper than grabbing it from thread-local storage.)
1975 You can ignore [pad]THXx when browsing the Perl headers/sources.
1976 Those are strictly for use within the core. Extensions and embedders
1977 need only be aware of [pad]THX.
1979 =head2 So what happened to dTHR?
1981 C<dTHR> was introduced in perl 5.005 to support the older thread model.
1982 The older thread model now uses the C<THX> mechanism to pass context
1983 pointers around, so C<dTHR> is not useful any more. Perl 5.6.0 and
1984 later still have it for backward source compatibility, but it is defined
1987 =head2 How do I use all this in extensions?
1989 When Perl is built with PERL_IMPLICIT_CONTEXT, extensions that call
1990 any functions in the Perl API will need to pass the initial context
1991 argument somehow. The kicker is that you will need to write it in
1992 such a way that the extension still compiles when Perl hasn't been
1993 built with PERL_IMPLICIT_CONTEXT enabled.
1995 There are three ways to do this. First, the easy but inefficient way,
1996 which is also the default, in order to maintain source compatibility
1997 with extensions: whenever XSUB.h is #included, it redefines the aTHX
1998 and aTHX_ macros to call a function that will return the context.
1999 Thus, something like:
2003 in your extension will translate to this when PERL_IMPLICIT_CONTEXT is
2006 Perl_sv_setiv(Perl_get_context(), sv, num);
2008 or to this otherwise:
2010 Perl_sv_setiv(sv, num);
2012 You have to do nothing new in your extension to get this; since
2013 the Perl library provides Perl_get_context(), it will all just
2016 The second, more efficient way is to use the following template for
2019 #define PERL_NO_GET_CONTEXT /* we want efficiency */
2024 static my_private_function(int arg1, int arg2);
2027 my_private_function(int arg1, int arg2)
2029 dTHX; /* fetch context */
2030 ... call many Perl API functions ...
2035 MODULE = Foo PACKAGE = Foo
2043 my_private_function(arg, 10);
2045 Note that the only two changes from the normal way of writing an
2046 extension is the addition of a C<#define PERL_NO_GET_CONTEXT> before
2047 including the Perl headers, followed by a C<dTHX;> declaration at
2048 the start of every function that will call the Perl API. (You'll
2049 know which functions need this, because the C compiler will complain
2050 that there's an undeclared identifier in those functions.) No changes
2051 are needed for the XSUBs themselves, because the XS() macro is
2052 correctly defined to pass in the implicit context if needed.
2054 The third, even more efficient way is to ape how it is done within
2058 #define PERL_NO_GET_CONTEXT /* we want efficiency */
2063 /* pTHX_ only needed for functions that call Perl API */
2064 static my_private_function(pTHX_ int arg1, int arg2);
2067 my_private_function(pTHX_ int arg1, int arg2)
2069 /* dTHX; not needed here, because THX is an argument */
2070 ... call Perl API functions ...
2075 MODULE = Foo PACKAGE = Foo
2083 my_private_function(aTHX_ arg, 10);
2085 This implementation never has to fetch the context using a function
2086 call, since it is always passed as an extra argument. Depending on
2087 your needs for simplicity or efficiency, you may mix the previous
2088 two approaches freely.
2090 Never add a comma after C<pTHX> yourself--always use the form of the
2091 macro with the underscore for functions that take explicit arguments,
2092 or the form without the argument for functions with no explicit arguments.
2094 If one is compiling Perl with the C<-DPERL_GLOBAL_STRUCT> the C<dVAR>
2095 definition is needed if the Perl global variables (see F<perlvars.h>
2096 or F<globvar.sym>) are accessed in the function and C<dTHX> is not
2097 used (the C<dTHX> includes the C<dVAR> if necessary). One notices
2098 the need for C<dVAR> only with the said compile-time define, because
2099 otherwise the Perl global variables are visible as-is.
2101 =head2 Should I do anything special if I call perl from multiple threads?
2103 If you create interpreters in one thread and then proceed to call them in
2104 another, you need to make sure perl's own Thread Local Storage (TLS) slot is
2105 initialized correctly in each of those threads.
2107 The C<perl_alloc> and C<perl_clone> API functions will automatically set
2108 the TLS slot to the interpreter they created, so that there is no need to do
2109 anything special if the interpreter is always accessed in the same thread that
2110 created it, and that thread did not create or call any other interpreters
2111 afterwards. If that is not the case, you have to set the TLS slot of the
2112 thread before calling any functions in the Perl API on that particular
2113 interpreter. This is done by calling the C<PERL_SET_CONTEXT> macro in that
2114 thread as the first thing you do:
2116 /* do this before doing anything else with some_perl */
2117 PERL_SET_CONTEXT(some_perl);
2119 ... other Perl API calls on some_perl go here ...
2121 =head2 Future Plans and PERL_IMPLICIT_SYS
2123 Just as PERL_IMPLICIT_CONTEXT provides a way to bundle up everything
2124 that the interpreter knows about itself and pass it around, so too are
2125 there plans to allow the interpreter to bundle up everything it knows
2126 about the environment it's running on. This is enabled with the
2127 PERL_IMPLICIT_SYS macro. Currently it only works with USE_ITHREADS
2128 and USE_5005THREADS on Windows (see inside iperlsys.h).
2130 This allows the ability to provide an extra pointer (called the "host"
2131 environment) for all the system calls. This makes it possible for
2132 all the system stuff to maintain their own state, broken down into
2133 seven C structures. These are thin wrappers around the usual system
2134 calls (see win32/perllib.c) for the default perl executable, but for a
2135 more ambitious host (like the one that would do fork() emulation) all
2136 the extra work needed to pretend that different interpreters are
2137 actually different "processes", would be done here.
2139 The Perl engine/interpreter and the host are orthogonal entities.
2140 There could be one or more interpreters in a process, and one or
2141 more "hosts", with free association between them.
2143 =head1 Internal Functions
2145 All of Perl's internal functions which will be exposed to the outside
2146 world are prefixed by C<Perl_> so that they will not conflict with XS
2147 functions or functions used in a program in which Perl is embedded.
2148 Similarly, all global variables begin with C<PL_>. (By convention,
2149 static functions start with C<S_>.)
2151 Inside the Perl core, you can get at the functions either with or
2152 without the C<Perl_> prefix, thanks to a bunch of defines that live in
2153 F<embed.h>. This header file is generated automatically from
2154 F<embed.pl> and F<embed.fnc>. F<embed.pl> also creates the prototyping
2155 header files for the internal functions, generates the documentation
2156 and a lot of other bits and pieces. It's important that when you add
2157 a new function to the core or change an existing one, you change the
2158 data in the table in F<embed.fnc> as well. Here's a sample entry from
2161 Apd |SV** |av_fetch |AV* ar|I32 key|I32 lval
2163 The second column is the return type, the third column the name. Columns
2164 after that are the arguments. The first column is a set of flags:
2170 This function is a part of the public API.
2174 This function has a C<Perl_> prefix; i.e. it is defined as C<Perl_av_fetch>
2178 This function has documentation using the C<apidoc> feature which we'll
2179 look at in a second.
2183 Other available flags are:
2189 This is a static function and is defined as C<S_whatever>, and usually
2190 called within the sources as C<whatever(...)>.
2194 This does not use C<aTHX_> and C<pTHX> to pass interpreter context. (See
2195 L<perlguts/Background and PERL_IMPLICIT_CONTEXT>.)
2199 This function never returns; C<croak>, C<exit> and friends.
2203 This function takes a variable number of arguments, C<printf> style.
2204 The argument list should end with C<...>, like this:
2206 Afprd |void |croak |const char* pat|...
2210 This function is part of the experimental development API, and may change
2211 or disappear without notice.
2215 This function should not have a compatibility macro to define, say,
2216 C<Perl_parse> to C<parse>. It must be called as C<Perl_parse>.
2220 This function isn't exported out of the Perl core.
2224 This is implemented as a macro.
2228 This function is explicitly exported.
2232 This function is visible to extensions included in the Perl core.
2236 Binary backward compatibility; this function is a macro but also has
2237 a C<Perl_> implementation (which is exported).
2241 If you edit F<embed.pl> or F<embed.fnc>, you will need to run
2242 C<make regen_headers> to force a rebuild of F<embed.h> and other
2243 auto-generated files.
2245 =head2 Formatted Printing of IVs, UVs, and NVs
2247 If you are printing IVs, UVs, or NVS instead of the stdio(3) style
2248 formatting codes like C<%d>, C<%ld>, C<%f>, you should use the
2249 following macros for portability
2254 UVxf UV in hexadecimal
2259 These will take care of 64-bit integers and long doubles.
2262 printf("IV is %"IVdf"\n", iv);
2264 The IVdf will expand to whatever is the correct format for the IVs.
2266 If you are printing addresses of pointers, use UVxf combined
2267 with PTR2UV(), do not use %lx or %p.
2269 =head2 Pointer-To-Integer and Integer-To-Pointer
2271 Because pointer size does not necessarily equal integer size,
2272 use the follow macros to do it right.
2277 INT2PTR(pointertotype, integer)
2282 SV *sv = INT2PTR(SV*, iv);
2289 =head2 Exception Handling
2291 There are a couple of macros to do very basic exception handling in XS
2292 modules. You have to define C<NO_XSLOCKS> before including F<XSUB.h> to
2293 be able to use these macros:
2298 You can use these macros if you call code that may croak, but you need
2299 to do some cleanup before giving control back to Perl. For example:
2301 dXCPT; /* set up necessary variables */
2304 code_that_may_croak();
2309 /* do cleanup here */
2313 Note that you always have to rethrow an exception that has been
2314 caught. Using these macros, it is not possible to just catch the
2315 exception and ignore it. If you have to ignore the exception, you
2316 have to use the C<call_*> function.
2318 The advantage of using the above macros is that you don't have
2319 to setup an extra function for C<call_*>, and that using these
2320 macros is faster than using C<call_*>.
2322 =head2 Source Documentation
2324 There's an effort going on to document the internal functions and
2325 automatically produce reference manuals from them - L<perlapi> is one
2326 such manual which details all the functions which are available to XS
2327 writers. L<perlintern> is the autogenerated manual for the functions
2328 which are not part of the API and are supposedly for internal use only.
2330 Source documentation is created by putting POD comments into the C
2334 =for apidoc sv_setiv
2336 Copies an integer into the given SV. Does not handle 'set' magic. See
2342 Please try and supply some documentation if you add functions to the
2345 =head2 Backwards compatibility
2347 The Perl API changes over time. New functions are added or the interfaces
2348 of existing functions are changed. The C<Devel::PPPort> module tries to
2349 provide compatibility code for some of these changes, so XS writers don't
2350 have to code it themselves when supporting multiple versions of Perl.
2352 C<Devel::PPPort> generates a C header file F<ppport.h> that can also
2353 be run as a Perl script. To generate F<ppport.h>, run:
2355 perl -MDevel::PPPort -eDevel::PPPort::WriteFile
2357 Besides checking existing XS code, the script can also be used to retrieve
2358 compatibility information for various API calls using the C<--api-info>
2359 command line switch. For example:
2361 % perl ppport.h --api-info=sv_magicext
2363 For details, see C<perldoc ppport.h>.
2365 =head1 Unicode Support
2367 Perl 5.6.0 introduced Unicode support. It's important for porters and XS
2368 writers to understand this support and make sure that the code they
2369 write does not corrupt Unicode data.
2371 =head2 What B<is> Unicode, anyway?
2373 In the olden, less enlightened times, we all used to use ASCII. Most of
2374 us did, anyway. The big problem with ASCII is that it's American. Well,
2375 no, that's not actually the problem; the problem is that it's not
2376 particularly useful for people who don't use the Roman alphabet. What
2377 used to happen was that particular languages would stick their own
2378 alphabet in the upper range of the sequence, between 128 and 255. Of
2379 course, we then ended up with plenty of variants that weren't quite
2380 ASCII, and the whole point of it being a standard was lost.
2382 Worse still, if you've got a language like Chinese or
2383 Japanese that has hundreds or thousands of characters, then you really
2384 can't fit them into a mere 256, so they had to forget about ASCII
2385 altogether, and build their own systems using pairs of numbers to refer
2388 To fix this, some people formed Unicode, Inc. and
2389 produced a new character set containing all the characters you can
2390 possibly think of and more. There are several ways of representing these
2391 characters, and the one Perl uses is called UTF-8. UTF-8 uses
2392 a variable number of bytes to represent a character, instead of just
2393 one. You can learn more about Unicode at http://www.unicode.org/
2395 =head2 How can I recognise a UTF-8 string?
2397 You can't. This is because UTF-8 data is stored in bytes just like
2398 non-UTF-8 data. The Unicode character 200, (C<0xC8> for you hex types)
2399 capital E with a grave accent, is represented by the two bytes
2400 C<v196.172>. Unfortunately, the non-Unicode string C<chr(196).chr(172)>
2401 has that byte sequence as well. So you can't tell just by looking - this
2402 is what makes Unicode input an interesting problem.
2404 The API function C<is_utf8_string> can help; it'll tell you if a string
2405 contains only valid UTF-8 characters. However, it can't do the work for
2406 you. On a character-by-character basis, C<is_utf8_char> will tell you
2407 whether the current character in a string is valid UTF-8.
2409 =head2 How does UTF-8 represent Unicode characters?
2411 As mentioned above, UTF-8 uses a variable number of bytes to store a
2412 character. Characters with values 1...128 are stored in one byte, just
2413 like good ol' ASCII. Character 129 is stored as C<v194.129>; this
2414 continues up to character 191, which is C<v194.191>. Now we've run out of
2415 bits (191 is binary C<10111111>) so we move on; 192 is C<v195.128>. And
2416 so it goes on, moving to three bytes at character 2048.
2418 Assuming you know you're dealing with a UTF-8 string, you can find out
2419 how long the first character in it is with the C<UTF8SKIP> macro:
2421 char *utf = "\305\233\340\240\201";
2424 len = UTF8SKIP(utf); /* len is 2 here */
2426 len = UTF8SKIP(utf); /* len is 3 here */
2428 Another way to skip over characters in a UTF-8 string is to use
2429 C<utf8_hop>, which takes a string and a number of characters to skip
2430 over. You're on your own about bounds checking, though, so don't use it
2433 All bytes in a multi-byte UTF-8 character will have the high bit set,
2434 so you can test if you need to do something special with this
2435 character like this (the UTF8_IS_INVARIANT() is a macro that tests
2436 whether the byte can be encoded as a single byte even in UTF-8):
2439 UV uv; /* Note: a UV, not a U8, not a char */
2441 if (!UTF8_IS_INVARIANT(*utf))
2442 /* Must treat this as UTF-8 */
2443 uv = utf8_to_uv(utf);
2445 /* OK to treat this character as a byte */
2448 You can also see in that example that we use C<utf8_to_uv> to get the
2449 value of the character; the inverse function C<uv_to_utf8> is available
2450 for putting a UV into UTF-8:
2452 if (!UTF8_IS_INVARIANT(uv))
2453 /* Must treat this as UTF8 */
2454 utf8 = uv_to_utf8(utf8, uv);
2456 /* OK to treat this character as a byte */
2459 You B<must> convert characters to UVs using the above functions if
2460 you're ever in a situation where you have to match UTF-8 and non-UTF-8
2461 characters. You may not skip over UTF-8 characters in this case. If you
2462 do this, you'll lose the ability to match hi-bit non-UTF-8 characters;
2463 for instance, if your UTF-8 string contains C<v196.172>, and you skip
2464 that character, you can never match a C<chr(200)> in a non-UTF-8 string.
2467 =head2 How does Perl store UTF-8 strings?
2469 Currently, Perl deals with Unicode strings and non-Unicode strings
2470 slightly differently. If a string has been identified as being UTF-8
2471 encoded, Perl will set a flag in the SV, C<SVf_UTF8>. You can check and
2472 manipulate this flag with the following macros:
2478 This flag has an important effect on Perl's treatment of the string: if
2479 Unicode data is not properly distinguished, regular expressions,
2480 C<length>, C<substr> and other string handling operations will have
2481 undesirable results.
2483 The problem comes when you have, for instance, a string that isn't
2484 flagged is UTF-8, and contains a byte sequence that could be UTF-8 -
2485 especially when combining non-UTF-8 and UTF-8 strings.
2487 Never forget that the C<SVf_UTF8> flag is separate to the PV value; you
2488 need be sure you don't accidentally knock it off while you're
2489 manipulating SVs. More specifically, you cannot expect to do this:
2498 nsv = newSVpvn(p, len);
2500 The C<char*> string does not tell you the whole story, and you can't
2501 copy or reconstruct an SV just by copying the string value. Check if the
2502 old SV has the UTF-8 flag set, and act accordingly:
2506 nsv = newSVpvn(p, len);
2510 In fact, your C<frobnicate> function should be made aware of whether or
2511 not it's dealing with UTF-8 data, so that it can handle the string
2514 Since just passing an SV to an XS function and copying the data of
2515 the SV is not enough to copy the UTF-8 flags, even less right is just
2516 passing a C<char *> to an XS function.
2518 =head2 How do I convert a string to UTF-8?
2520 If you're mixing UTF-8 and non-UTF-8 strings, you might find it necessary
2521 to upgrade one of the strings to UTF-8. If you've got an SV, the easiest
2524 sv_utf8_upgrade(sv);
2526 However, you must not do this, for example:
2529 sv_utf8_upgrade(left);
2531 If you do this in a binary operator, you will actually change one of the
2532 strings that came into the operator, and, while it shouldn't be noticeable
2533 by the end user, it can cause problems.
2535 Instead, C<bytes_to_utf8> will give you a UTF-8-encoded B<copy> of its
2536 string argument. This is useful for having the data available for
2537 comparisons and so on, without harming the original SV. There's also
2538 C<utf8_to_bytes> to go the other way, but naturally, this will fail if
2539 the string contains any characters above 255 that can't be represented
2542 =head2 Is there anything else I need to know?
2544 Not really. Just remember these things:
2550 There's no way to tell if a string is UTF-8 or not. You can tell if an SV
2551 is UTF-8 by looking at is C<SvUTF8> flag. Don't forget to set the flag if
2552 something should be UTF-8. Treat the flag as part of the PV, even though
2553 it's not - if you pass on the PV to somewhere, pass on the flag too.
2557 If a string is UTF-8, B<always> use C<utf8_to_uv> to get at the value,
2558 unless C<UTF8_IS_INVARIANT(*s)> in which case you can use C<*s>.
2562 When writing a character C<uv> to a UTF-8 string, B<always> use
2563 C<uv_to_utf8>, unless C<UTF8_IS_INVARIANT(uv))> in which case
2564 you can use C<*s = uv>.
2568 Mixing UTF-8 and non-UTF-8 strings is tricky. Use C<bytes_to_utf8> to get
2569 a new string which is UTF-8 encoded. There are tricks you can use to
2570 delay deciding whether you need to use a UTF-8 string until you get to a
2571 high character - C<HALF_UPGRADE> is one of those.
2575 =head1 Custom Operators
2577 Custom operator support is a new experimental feature that allows you to
2578 define your own ops. This is primarily to allow the building of
2579 interpreters for other languages in the Perl core, but it also allows
2580 optimizations through the creation of "macro-ops" (ops which perform the
2581 functions of multiple ops which are usually executed together, such as
2582 C<gvsv, gvsv, add>.)
2584 This feature is implemented as a new op type, C<OP_CUSTOM>. The Perl
2585 core does not "know" anything special about this op type, and so it will
2586 not be involved in any optimizations. This also means that you can
2587 define your custom ops to be any op structure - unary, binary, list and
2590 It's important to know what custom operators won't do for you. They
2591 won't let you add new syntax to Perl, directly. They won't even let you
2592 add new keywords, directly. In fact, they won't change the way Perl
2593 compiles a program at all. You have to do those changes yourself, after
2594 Perl has compiled the program. You do this either by manipulating the op
2595 tree using a C<CHECK> block and the C<B::Generate> module, or by adding
2596 a custom peephole optimizer with the C<optimize> module.
2598 When you do this, you replace ordinary Perl ops with custom ops by
2599 creating ops with the type C<OP_CUSTOM> and the C<pp_addr> of your own
2600 PP function. This should be defined in XS code, and should look like
2601 the PP ops in C<pp_*.c>. You are responsible for ensuring that your op
2602 takes the appropriate number of values from the stack, and you are
2603 responsible for adding stack marks if necessary.
2605 You should also "register" your op with the Perl interpreter so that it
2606 can produce sensible error and warning messages. Since it is possible to
2607 have multiple custom ops within the one "logical" op type C<OP_CUSTOM>,
2608 Perl uses the value of C<< o->op_ppaddr >> as a key into the
2609 C<PL_custom_op_descs> and C<PL_custom_op_names> hashes. This means you
2610 need to enter a name and description for your op at the appropriate
2611 place in the C<PL_custom_op_names> and C<PL_custom_op_descs> hashes.
2613 Forthcoming versions of C<B::Generate> (version 1.0 and above) should
2614 directly support the creation of custom ops by name.
2618 Until May 1997, this document was maintained by Jeff Okamoto
2619 E<lt>okamoto@corp.hp.comE<gt>. It is now maintained as part of Perl
2620 itself by the Perl 5 Porters E<lt>perl5-porters@perl.orgE<gt>.
2622 With lots of help and suggestions from Dean Roehrich, Malcolm Beattie,
2623 Andreas Koenig, Paul Hudson, Ilya Zakharevich, Paul Marquess, Neil
2624 Bowers, Matthew Green, Tim Bunce, Spider Boardman, Ulrich Pfeifer,
2625 Stephen McCamant, and Gurusamy Sarathy.
2629 perlapi(1), perlintern(1), perlxs(1), perlembed(1)