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).
40 ("PV" stands for "Pointer Value". You might think that it is misnamed
41 because it is described as pointing only to strings. However, it is
42 possible to have it point to other things. For example, it could point
43 to an array of UVs. But,
44 using it for non-strings requires care, as the underlying assumption of
45 much of the internals is that PVs are just for strings. Often, for
46 example, a trailing C<NUL> is tacked on automatically. The non-string use
47 is documented only in this paragraph.)
49 The seven routines are:
54 SV* newSVpv(const char*, STRLEN);
55 SV* newSVpvn(const char*, STRLEN);
56 SV* newSVpvf(const char*, ...);
59 C<STRLEN> is an integer type (Size_t, usually defined as size_t in
60 F<config.h>) guaranteed to be large enough to represent the size of
61 any string that perl can handle.
63 In the unlikely case of a SV requiring more complex initialization, you
64 can create an empty SV with newSV(len). If C<len> is 0 an empty SV of
65 type NULL is returned, else an SV of type PV is returned with len + 1 (for
66 the C<NUL>) bytes of storage allocated, accessible via SvPVX. In both cases
67 the SV has the undef value.
69 SV *sv = newSV(0); /* no storage allocated */
70 SV *sv = newSV(10); /* 10 (+1) bytes of uninitialised storage
73 To change the value of an I<already-existing> SV, there are eight routines:
75 void sv_setiv(SV*, IV);
76 void sv_setuv(SV*, UV);
77 void sv_setnv(SV*, double);
78 void sv_setpv(SV*, const char*);
79 void sv_setpvn(SV*, const char*, STRLEN)
80 void sv_setpvf(SV*, const char*, ...);
81 void sv_vsetpvfn(SV*, const char*, STRLEN, va_list *,
83 void sv_setsv(SV*, SV*);
85 Notice that you can choose to specify the length of the string to be
86 assigned by using C<sv_setpvn>, C<newSVpvn>, or C<newSVpv>, or you may
87 allow Perl to calculate the length by using C<sv_setpv> or by specifying
88 0 as the second argument to C<newSVpv>. Be warned, though, that Perl will
89 determine the string's length by using C<strlen>, which depends on the
90 string terminating with a C<NUL> character, and not otherwise containing
93 The arguments of C<sv_setpvf> are processed like C<sprintf>, and the
94 formatted output becomes the value.
96 C<sv_vsetpvfn> is an analogue of C<vsprintf>, but it allows you to specify
97 either a pointer to a variable argument list or the address and length of
98 an array of SVs. The last argument points to a boolean; on return, if that
99 boolean is true, then locale-specific information has been used to format
100 the string, and the string's contents are therefore untrustworthy (see
101 L<perlsec>). This pointer may be NULL if that information is not
102 important. Note that this function requires you to specify the length of
105 The C<sv_set*()> functions are not generic enough to operate on values
106 that have "magic". See L</Magic Virtual Tables> later in this document.
108 All SVs that contain strings should be terminated with a C<NUL> character.
109 If it is not C<NUL>-terminated there is a risk of
110 core dumps and corruptions from code which passes the string to C
111 functions or system calls which expect a C<NUL>-terminated string.
112 Perl's own functions typically add a trailing C<NUL> for this reason.
113 Nevertheless, you should be very careful when you pass a string stored
114 in an SV to a C function or system call.
116 To access the actual value that an SV points to, you can use the macros:
121 SvPV(SV*, STRLEN len)
124 which will automatically coerce the actual scalar type into an IV, UV, double,
127 In the C<SvPV> macro, the length of the string returned is placed into the
128 variable C<len> (this is a macro, so you do I<not> use C<&len>). If you do
129 not care what the length of the data is, use the C<SvPV_nolen> macro.
130 Historically the C<SvPV> macro with the global variable C<PL_na> has been
131 used in this case. But that can be quite inefficient because C<PL_na> must
132 be accessed in thread-local storage in threaded Perl. In any case, remember
133 that Perl allows arbitrary strings of data that may both contain NULs and
134 might not be terminated by a C<NUL>.
136 Also remember that C doesn't allow you to safely say C<foo(SvPV(s, len),
137 len);>. It might work with your
138 compiler, but it won't work for everyone.
139 Break this sort of statement up into separate assignments:
147 If you want to know if the scalar value is TRUE, you can use:
151 Although Perl will automatically grow strings for you, if you need to force
152 Perl to allocate more memory for your SV, you can use the macro
154 SvGROW(SV*, STRLEN newlen)
156 which will determine if more memory needs to be allocated. If so, it will
157 call the function C<sv_grow>. Note that C<SvGROW> can only increase, not
158 decrease, the allocated memory of an SV and that it does not automatically
159 add space for the trailing C<NUL> byte (perl's own string functions typically do
160 C<SvGROW(sv, len + 1)>).
162 If you want to write to an existing SV's buffer and set its value to a
163 string, use SvPV_force() or one of its variants to force the SV to be
164 a PV. This will remove any of various types of non-stringness from
165 the SV while preserving the content of the SV in the PV. This can be
166 used, for example, to append data from an API function to a buffer
167 without extra copying:
169 (void)SvPVbyte_force(sv, len);
170 s = SvGROW(sv, len + needlen + 1);
171 /* something that modifies up to needlen bytes at s+len, but
172 modifies newlen bytes
173 eg. newlen = read(fd, s + len, needlen);
174 ignoring errors for these examples
176 s[len + newlen] = '\0';
177 SvCUR_set(sv, len + newlen);
181 If you already have the data in memory or if you want to keep your
182 code simple, you can use one of the sv_cat*() variants, such as
183 sv_catpvn(). If you want to insert anywhere in the string you can use
184 sv_insert() or sv_insert_flags().
186 If you don't need the existing content of the SV, you can avoid some
189 sv_setpvn(sv, "", 0);
190 s = SvGROW(sv, needlen + 1);
191 /* something that modifies up to needlen bytes at s, but modifies
193 eg. newlen = read(fd, s. needlen);
196 SvCUR_set(sv, newlen);
197 SvPOK_only(sv); /* also clears SVf_UTF8 */
200 Again, if you already have the data in memory or want to avoid the
201 complexity of the above, you can use sv_setpvn().
203 If you have a buffer allocated with Newx() and want to set that as the
204 SV's value, you can use sv_usepvn_flags(). That has some requirements
205 if you want to avoid perl re-allocating the buffer to fit the trailing
208 Newx(buf, somesize+1, char);
209 /* ... fill in buf ... */
210 buf[somesize] = '\0';
211 sv_usepvn_flags(sv, buf, somesize, SV_SMAGIC | SV_HAS_TRAILING_NUL);
212 /* buf now belongs to perl, don't release it */
214 If you have an SV and want to know what kind of data Perl thinks is stored
215 in it, you can use the following macros to check the type of SV you have.
221 You can get and set the current length of the string stored in an SV with
222 the following macros:
225 SvCUR_set(SV*, I32 val)
227 You can also get a pointer to the end of the string stored in the SV
232 But note that these last three macros are valid only if C<SvPOK()> is true.
234 If you want to append something to the end of string stored in an C<SV*>,
235 you can use the following functions:
237 void sv_catpv(SV*, const char*);
238 void sv_catpvn(SV*, const char*, STRLEN);
239 void sv_catpvf(SV*, const char*, ...);
240 void sv_vcatpvfn(SV*, const char*, STRLEN, va_list *, SV **,
242 void sv_catsv(SV*, SV*);
244 The first function calculates the length of the string to be appended by
245 using C<strlen>. In the second, you specify the length of the string
246 yourself. The third function processes its arguments like C<sprintf> and
247 appends the formatted output. The fourth function works like C<vsprintf>.
248 You can specify the address and length of an array of SVs instead of the
249 va_list argument. The fifth function
250 extends the string stored in the first
251 SV with the string stored in the second SV. It also forces the second SV
252 to be interpreted as a string.
254 The C<sv_cat*()> functions are not generic enough to operate on values that
255 have "magic". See L</Magic Virtual Tables> later in this document.
257 If you know the name of a scalar variable, you can get a pointer to its SV
258 by using the following:
260 SV* get_sv("package::varname", 0);
262 This returns NULL if the variable does not exist.
264 If you want to know if this variable (or any other SV) is actually C<defined>,
269 The scalar C<undef> value is stored in an SV instance called C<PL_sv_undef>.
271 Its address can be used whenever an C<SV*> is needed. Make sure that
272 you don't try to compare a random sv with C<&PL_sv_undef>. For example
273 when interfacing Perl code, it'll work correctly for:
277 But won't work when called as:
282 So to repeat always use SvOK() to check whether an sv is defined.
284 Also you have to be careful when using C<&PL_sv_undef> as a value in
285 AVs or HVs (see L</AVs, HVs and undefined values>).
287 There are also the two values C<PL_sv_yes> and C<PL_sv_no>, which contain
288 boolean TRUE and FALSE values, respectively. Like C<PL_sv_undef>, their
289 addresses can be used whenever an C<SV*> is needed.
291 Do not be fooled into thinking that C<(SV *) 0> is the same as C<&PL_sv_undef>.
295 if (I-am-to-return-a-real-value) {
296 sv = sv_2mortal(newSViv(42));
300 This code tries to return a new SV (which contains the value 42) if it should
301 return a real value, or undef otherwise. Instead it has returned a NULL
302 pointer which, somewhere down the line, will cause a segmentation violation,
303 bus error, or just weird results. Change the zero to C<&PL_sv_undef> in the
304 first line and all will be well.
306 To free an SV that you've created, call C<SvREFCNT_dec(SV*)>. Normally this
307 call is not necessary (see L</Reference Counts and Mortality>).
311 Perl provides the function C<sv_chop> to efficiently remove characters
312 from the beginning of a string; you give it an SV and a pointer to
313 somewhere inside the PV, and it discards everything before the
314 pointer. The efficiency comes by means of a little hack: instead of
315 actually removing the characters, C<sv_chop> sets the flag C<OOK>
316 (offset OK) to signal to other functions that the offset hack is in
317 effect, and it moves the PV pointer (called C<SvPVX>) forward
318 by the number of bytes chopped off, and adjusts C<SvCUR> and C<SvLEN>
319 accordingly. (A portion of the space between the old and new PV
320 pointers is used to store the count of chopped bytes.)
322 Hence, at this point, the start of the buffer that we allocated lives
323 at C<SvPVX(sv) - SvIV(sv)> in memory and the PV pointer is pointing
324 into the middle of this allocated storage.
326 This is best demonstrated by example. Normally copy-on-write will prevent
327 the substitution from operator from using this hack, but if you can craft a
328 string for which copy-on-write is not possible, you can see it in play. In
329 the current implementation, the final byte of a string buffer is used as a
330 copy-on-write reference count. If the buffer is not big enough, then
331 copy-on-write is skipped. First have a look at an empty string:
333 % ./perl -Ilib -MDevel::Peek -le '$a=""; $a .= ""; Dump $a'
334 SV = PV(0x7ffb7c008a70) at 0x7ffb7c030390
337 PV = 0x7ffb7bc05b50 ""\0
341 Notice here the LEN is 10. (It may differ on your platform.) Extend the
342 length of the string to one less than 10, and do a substitution:
344 % ./perl -Ilib -MDevel::Peek -le '$a=""; $a.="123456789"; $a=~s/.//; \
346 SV = PV(0x7ffa04008a70) at 0x7ffa04030390
348 FLAGS = (POK,OOK,pPOK)
350 PV = 0x7ffa03c05b61 ( "\1" . ) "23456789"\0
354 Here the number of bytes chopped off (1) is shown next as the OFFSET. The
355 portion of the string between the "real" and the "fake" beginnings is
356 shown in parentheses, and the values of C<SvCUR> and C<SvLEN> reflect
357 the fake beginning, not the real one. (The first character of the string
358 buffer happens to have changed to "\1" here, not "1", because the current
359 implementation stores the offset count in the string buffer. This is
362 Something similar to the offset hack is performed on AVs to enable
363 efficient shifting and splicing off the beginning of the array; while
364 C<AvARRAY> points to the first element in the array that is visible from
365 Perl, C<AvALLOC> points to the real start of the C array. These are
366 usually the same, but a C<shift> operation can be carried out by
367 increasing C<AvARRAY> by one and decreasing C<AvFILL> and C<AvMAX>.
368 Again, the location of the real start of the C array only comes into
369 play when freeing the array. See C<av_shift> in F<av.c>.
371 =head2 What's Really Stored in an SV?
373 Recall that the usual method of determining the type of scalar you have is
374 to use C<Sv*OK> macros. Because a scalar can be both a number and a string,
375 usually these macros will always return TRUE and calling the C<Sv*V>
376 macros will do the appropriate conversion of string to integer/double or
377 integer/double to string.
379 If you I<really> need to know if you have an integer, double, or string
380 pointer in an SV, you can use the following three macros instead:
386 These will tell you if you truly have an integer, double, or string pointer
387 stored in your SV. The "p" stands for private.
389 There are various ways in which the private and public flags may differ.
390 For example, in perl 5.16 and earlier a tied SV may have a valid
391 underlying value in the IV slot (so SvIOKp is true), but the data
392 should be accessed via the FETCH routine rather than directly,
393 so SvIOK is false. (In perl 5.18 onwards, tied scalars use
394 the flags the same way as untied scalars.) Another is when
395 numeric conversion has occurred and precision has been lost: only the
396 private flag is set on 'lossy' values. So when an NV is converted to an
397 IV with loss, SvIOKp, SvNOKp and SvNOK will be set, while SvIOK wont be.
399 In general, though, it's best to use the C<Sv*V> macros.
401 =head2 Working with AVs
403 There are two ways to create and load an AV. The first method creates an
408 The second method both creates the AV and initially populates it with SVs:
410 AV* av_make(SSize_t num, SV **ptr);
412 The second argument points to an array containing C<num> C<SV*>'s. Once the
413 AV has been created, the SVs can be destroyed, if so desired.
415 Once the AV has been created, the following operations are possible on it:
417 void av_push(AV*, SV*);
420 void av_unshift(AV*, SSize_t num);
422 These should be familiar operations, with the exception of C<av_unshift>.
423 This routine adds C<num> elements at the front of the array with the C<undef>
424 value. You must then use C<av_store> (described below) to assign values
425 to these new elements.
427 Here are some other functions:
429 SSize_t av_top_index(AV*);
430 SV** av_fetch(AV*, SSize_t key, I32 lval);
431 SV** av_store(AV*, SSize_t key, SV* val);
433 The C<av_top_index> function returns the highest index value in an array (just
434 like $#array in Perl). If the array is empty, -1 is returned. The
435 C<av_fetch> function returns the value at index C<key>, but if C<lval>
436 is non-zero, then C<av_fetch> will store an undef value at that index.
437 The C<av_store> function stores the value C<val> at index C<key>, and does
438 not increment the reference count of C<val>. Thus the caller is responsible
439 for taking care of that, and if C<av_store> returns NULL, the caller will
440 have to decrement the reference count to avoid a memory leak. Note that
441 C<av_fetch> and C<av_store> both return C<SV**>'s, not C<SV*>'s as their
448 void av_extend(AV*, SSize_t key);
450 The C<av_clear> function deletes all the elements in the AV* array, but
451 does not actually delete the array itself. The C<av_undef> function will
452 delete all the elements in the array plus the array itself. The
453 C<av_extend> function extends the array so that it contains at least C<key+1>
454 elements. If C<key+1> is less than the currently allocated length of the array,
455 then nothing is done.
457 If you know the name of an array variable, you can get a pointer to its AV
458 by using the following:
460 AV* get_av("package::varname", 0);
462 This returns NULL if the variable does not exist.
464 See L</Understanding the Magic of Tied Hashes and Arrays> for more
465 information on how to use the array access functions on tied arrays.
467 =head2 Working with HVs
469 To create an HV, you use the following routine:
473 Once the HV has been created, the following operations are possible on it:
475 SV** hv_store(HV*, const char* key, U32 klen, SV* val, U32 hash);
476 SV** hv_fetch(HV*, const char* key, U32 klen, I32 lval);
478 The C<klen> parameter is the length of the key being passed in (Note that
479 you cannot pass 0 in as a value of C<klen> to tell Perl to measure the
480 length of the key). The C<val> argument contains the SV pointer to the
481 scalar being stored, and C<hash> is the precomputed hash value (zero if
482 you want C<hv_store> to calculate it for you). The C<lval> parameter
483 indicates whether this fetch is actually a part of a store operation, in
484 which case a new undefined value will be added to the HV with the supplied
485 key and C<hv_fetch> will return as if the value had already existed.
487 Remember that C<hv_store> and C<hv_fetch> return C<SV**>'s and not just
488 C<SV*>. To access the scalar value, you must first dereference the return
489 value. However, you should check to make sure that the return value is
490 not NULL before dereferencing it.
492 The first of these two functions checks if a hash table entry exists, and the
495 bool hv_exists(HV*, const char* key, U32 klen);
496 SV* hv_delete(HV*, const char* key, U32 klen, I32 flags);
498 If C<flags> does not include the C<G_DISCARD> flag then C<hv_delete> will
499 create and return a mortal copy of the deleted value.
501 And more miscellaneous functions:
506 Like their AV counterparts, C<hv_clear> deletes all the entries in the hash
507 table but does not actually delete the hash table. The C<hv_undef> deletes
508 both the entries and the hash table itself.
510 Perl keeps the actual data in a linked list of structures with a typedef of HE.
511 These contain the actual key and value pointers (plus extra administrative
512 overhead). The key is a string pointer; the value is an C<SV*>. However,
513 once you have an C<HE*>, to get the actual key and value, use the routines
516 I32 hv_iterinit(HV*);
517 /* Prepares starting point to traverse hash table */
518 HE* hv_iternext(HV*);
519 /* Get the next entry, and return a pointer to a
520 structure that has both the key and value */
521 char* hv_iterkey(HE* entry, I32* retlen);
522 /* Get the key from an HE structure and also return
523 the length of the key string */
524 SV* hv_iterval(HV*, HE* entry);
525 /* Return an SV pointer to the value of the HE
527 SV* hv_iternextsv(HV*, char** key, I32* retlen);
528 /* This convenience routine combines hv_iternext,
529 hv_iterkey, and hv_iterval. The key and retlen
530 arguments are return values for the key and its
531 length. The value is returned in the SV* argument */
533 If you know the name of a hash variable, you can get a pointer to its HV
534 by using the following:
536 HV* get_hv("package::varname", 0);
538 This returns NULL if the variable does not exist.
540 The hash algorithm is defined in the C<PERL_HASH> macro:
542 PERL_HASH(hash, key, klen)
544 The exact implementation of this macro varies by architecture and version
545 of perl, and the return value may change per invocation, so the value
546 is only valid for the duration of a single perl process.
548 See L</Understanding the Magic of Tied Hashes and Arrays> for more
549 information on how to use the hash access functions on tied hashes.
551 =head2 Hash API Extensions
553 Beginning with version 5.004, the following functions are also supported:
555 HE* hv_fetch_ent (HV* tb, SV* key, I32 lval, U32 hash);
556 HE* hv_store_ent (HV* tb, SV* key, SV* val, U32 hash);
558 bool hv_exists_ent (HV* tb, SV* key, U32 hash);
559 SV* hv_delete_ent (HV* tb, SV* key, I32 flags, U32 hash);
561 SV* hv_iterkeysv (HE* entry);
563 Note that these functions take C<SV*> keys, which simplifies writing
564 of extension code that deals with hash structures. These functions
565 also allow passing of C<SV*> keys to C<tie> functions without forcing
566 you to stringify the keys (unlike the previous set of functions).
568 They also return and accept whole hash entries (C<HE*>), making their
569 use more efficient (since the hash number for a particular string
570 doesn't have to be recomputed every time). See L<perlapi> for detailed
573 The following macros must always be used to access the contents of hash
574 entries. Note that the arguments to these macros must be simple
575 variables, since they may get evaluated more than once. See
576 L<perlapi> for detailed descriptions of these macros.
578 HePV(HE* he, STRLEN len)
582 HeSVKEY_force(HE* he)
583 HeSVKEY_set(HE* he, SV* sv)
585 These two lower level macros are defined, but must only be used when
586 dealing with keys that are not C<SV*>s:
591 Note that both C<hv_store> and C<hv_store_ent> do not increment the
592 reference count of the stored C<val>, which is the caller's responsibility.
593 If these functions return a NULL value, the caller will usually have to
594 decrement the reference count of C<val> to avoid a memory leak.
596 =head2 AVs, HVs and undefined values
598 Sometimes you have to store undefined values in AVs or HVs. Although
599 this may be a rare case, it can be tricky. That's because you're
600 used to using C<&PL_sv_undef> if you need an undefined SV.
602 For example, intuition tells you that this XS code:
605 av_store( av, 0, &PL_sv_undef );
607 is equivalent to this Perl code:
612 Unfortunately, this isn't true. In perl 5.18 and earlier, AVs use C<&PL_sv_undef> as a marker
613 for indicating that an array element has not yet been initialized.
614 Thus, C<exists $av[0]> would be true for the above Perl code, but
615 false for the array generated by the XS code. In perl 5.20, storing
616 &PL_sv_undef will create a read-only element, because the scalar
617 &PL_sv_undef itself is stored, not a copy.
619 Similar problems can occur when storing C<&PL_sv_undef> in HVs:
621 hv_store( hv, "key", 3, &PL_sv_undef, 0 );
623 This will indeed make the value C<undef>, but if you try to modify
624 the value of C<key>, you'll get the following error:
626 Modification of non-creatable hash value attempted
628 In perl 5.8.0, C<&PL_sv_undef> was also used to mark placeholders
629 in restricted hashes. This caused such hash entries not to appear
630 when iterating over the hash or when checking for the keys
631 with the C<hv_exists> function.
633 You can run into similar problems when you store C<&PL_sv_yes> or
634 C<&PL_sv_no> into AVs or HVs. Trying to modify such elements
635 will give you the following error:
637 Modification of a read-only value attempted
639 To make a long story short, you can use the special variables
640 C<&PL_sv_undef>, C<&PL_sv_yes> and C<&PL_sv_no> with AVs and
641 HVs, but you have to make sure you know what you're doing.
643 Generally, if you want to store an undefined value in an AV
644 or HV, you should not use C<&PL_sv_undef>, but rather create a
645 new undefined value using the C<newSV> function, for example:
647 av_store( av, 42, newSV(0) );
648 hv_store( hv, "foo", 3, newSV(0), 0 );
652 References are a special type of scalar that point to other data types
653 (including other references).
655 To create a reference, use either of the following functions:
657 SV* newRV_inc((SV*) thing);
658 SV* newRV_noinc((SV*) thing);
660 The C<thing> argument can be any of an C<SV*>, C<AV*>, or C<HV*>. The
661 functions are identical except that C<newRV_inc> increments the reference
662 count of the C<thing>, while C<newRV_noinc> does not. For historical
663 reasons, C<newRV> is a synonym for C<newRV_inc>.
665 Once you have a reference, you can use the following macro to dereference
670 then call the appropriate routines, casting the returned C<SV*> to either an
671 C<AV*> or C<HV*>, if required.
673 To determine if an SV is a reference, you can use the following macro:
677 To discover what type of value the reference refers to, use the following
678 macro and then check the return value.
682 The most useful types that will be returned are:
688 SVt_PVGV Glob (possibly a file handle)
690 See L<perlapi/svtype> for more details.
692 =head2 Blessed References and Class Objects
694 References are also used to support object-oriented programming. In perl's
695 OO lexicon, an object is simply a reference that has been blessed into a
696 package (or class). Once blessed, the programmer may now use the reference
697 to access the various methods in the class.
699 A reference can be blessed into a package with the following function:
701 SV* sv_bless(SV* sv, HV* stash);
703 The C<sv> argument must be a reference value. The C<stash> argument
704 specifies which class the reference will belong to. See
705 L</Stashes and Globs> for information on converting class names into stashes.
707 /* Still under construction */
709 The following function upgrades rv to reference if not already one.
710 Creates a new SV for rv to point to. If C<classname> is non-null, the SV
711 is blessed into the specified class. SV is returned.
713 SV* newSVrv(SV* rv, const char* classname);
715 The following three functions copy integer, unsigned integer or double
716 into an SV whose reference is C<rv>. SV is blessed if C<classname> is
719 SV* sv_setref_iv(SV* rv, const char* classname, IV iv);
720 SV* sv_setref_uv(SV* rv, const char* classname, UV uv);
721 SV* sv_setref_nv(SV* rv, const char* classname, NV iv);
723 The following function copies the pointer value (I<the address, not the
724 string!>) into an SV whose reference is rv. SV is blessed if C<classname>
727 SV* sv_setref_pv(SV* rv, const char* classname, void* pv);
729 The following function copies a string into an SV whose reference is C<rv>.
730 Set length to 0 to let Perl calculate the string length. SV is blessed if
731 C<classname> is non-null.
733 SV* sv_setref_pvn(SV* rv, const char* classname, char* pv,
736 The following function tests whether the SV is blessed into the specified
737 class. It does not check inheritance relationships.
739 int sv_isa(SV* sv, const char* name);
741 The following function tests whether the SV is a reference to a blessed object.
743 int sv_isobject(SV* sv);
745 The following function tests whether the SV is derived from the specified
746 class. SV can be either a reference to a blessed object or a string
747 containing a class name. This is the function implementing the
748 C<UNIVERSAL::isa> functionality.
750 bool sv_derived_from(SV* sv, const char* name);
752 To check if you've got an object derived from a specific class you have
755 if (sv_isobject(sv) && sv_derived_from(sv, class)) { ... }
757 =head2 Creating New Variables
759 To create a new Perl variable with an undef value which can be accessed from
760 your Perl script, use the following routines, depending on the variable type.
762 SV* get_sv("package::varname", GV_ADD);
763 AV* get_av("package::varname", GV_ADD);
764 HV* get_hv("package::varname", GV_ADD);
766 Notice the use of GV_ADD as the second parameter. The new variable can now
767 be set, using the routines appropriate to the data type.
769 There are additional macros whose values may be bitwise OR'ed with the
770 C<GV_ADD> argument to enable certain extra features. Those bits are:
776 Marks the variable as multiply defined, thus preventing the:
778 Name <varname> used only once: possible typo
786 Had to create <varname> unexpectedly
788 if the variable did not exist before the function was called.
792 If you do not specify a package name, the variable is created in the current
795 =head2 Reference Counts and Mortality
797 Perl uses a reference count-driven garbage collection mechanism. SVs,
798 AVs, or HVs (xV for short in the following) start their life with a
799 reference count of 1. If the reference count of an xV ever drops to 0,
800 then it will be destroyed and its memory made available for reuse.
802 This normally doesn't happen at the Perl level unless a variable is
803 undef'ed or the last variable holding a reference to it is changed or
804 overwritten. At the internal level, however, reference counts can be
805 manipulated with the following macros:
807 int SvREFCNT(SV* sv);
808 SV* SvREFCNT_inc(SV* sv);
809 void SvREFCNT_dec(SV* sv);
811 However, there is one other function which manipulates the reference
812 count of its argument. The C<newRV_inc> function, you will recall,
813 creates a reference to the specified argument. As a side effect,
814 it increments the argument's reference count. If this is not what
815 you want, use C<newRV_noinc> instead.
817 For example, imagine you want to return a reference from an XSUB function.
818 Inside the XSUB routine, you create an SV which initially has a reference
819 count of one. Then you call C<newRV_inc>, passing it the just-created SV.
820 This returns the reference as a new SV, but the reference count of the
821 SV you passed to C<newRV_inc> has been incremented to two. Now you
822 return the reference from the XSUB routine and forget about the SV.
823 But Perl hasn't! Whenever the returned reference is destroyed, the
824 reference count of the original SV is decreased to one and nothing happens.
825 The SV will hang around without any way to access it until Perl itself
826 terminates. This is a memory leak.
828 The correct procedure, then, is to use C<newRV_noinc> instead of
829 C<newRV_inc>. Then, if and when the last reference is destroyed,
830 the reference count of the SV will go to zero and it will be destroyed,
831 stopping any memory leak.
833 There are some convenience functions available that can help with the
834 destruction of xVs. These functions introduce the concept of "mortality".
835 An xV that is mortal has had its reference count marked to be decremented,
836 but not actually decremented, until "a short time later". Generally the
837 term "short time later" means a single Perl statement, such as a call to
838 an XSUB function. The actual determinant for when mortal xVs have their
839 reference count decremented depends on two macros, SAVETMPS and FREETMPS.
840 See L<perlcall> and L<perlxs> for more details on these macros.
842 "Mortalization" then is at its simplest a deferred C<SvREFCNT_dec>.
843 However, if you mortalize a variable twice, the reference count will
844 later be decremented twice.
846 "Mortal" SVs are mainly used for SVs that are placed on perl's stack.
847 For example an SV which is created just to pass a number to a called sub
848 is made mortal to have it cleaned up automatically when it's popped off
849 the stack. Similarly, results returned by XSUBs (which are pushed on the
850 stack) are often made mortal.
852 To create a mortal variable, use the functions:
856 SV* sv_mortalcopy(SV*)
858 The first call creates a mortal SV (with no value), the second converts an existing
859 SV to a mortal SV (and thus defers a call to C<SvREFCNT_dec>), and the
860 third creates a mortal copy of an existing SV.
861 Because C<sv_newmortal> gives the new SV no value, it must normally be given one
862 via C<sv_setpv>, C<sv_setiv>, etc. :
864 SV *tmp = sv_newmortal();
865 sv_setiv(tmp, an_integer);
867 As that is multiple C statements it is quite common so see this idiom instead:
869 SV *tmp = sv_2mortal(newSViv(an_integer));
872 You should be careful about creating mortal variables. Strange things
873 can happen if you make the same value mortal within multiple contexts,
874 or if you make a variable mortal multiple
875 times. Thinking of "Mortalization"
876 as deferred C<SvREFCNT_dec> should help to minimize such problems.
877 For example if you are passing an SV which you I<know> has a high enough REFCNT
878 to survive its use on the stack you need not do any mortalization.
879 If you are not sure then doing an C<SvREFCNT_inc> and C<sv_2mortal>, or
880 making a C<sv_mortalcopy> is safer.
882 The mortal routines are not just for SVs; AVs and HVs can be
883 made mortal by passing their address (type-casted to C<SV*>) to the
884 C<sv_2mortal> or C<sv_mortalcopy> routines.
886 =head2 Stashes and Globs
888 A B<stash> is a hash that contains all variables that are defined
889 within a package. Each key of the stash is a symbol
890 name (shared by all the different types of objects that have the same
891 name), and each value in the hash table is a GV (Glob Value). This GV
892 in turn contains references to the various objects of that name,
893 including (but not limited to) the following:
902 There is a single stash called C<PL_defstash> that holds the items that exist
903 in the C<main> package. To get at the items in other packages, append the
904 string "::" to the package name. The items in the C<Foo> package are in
905 the stash C<Foo::> in PL_defstash. The items in the C<Bar::Baz> package are
906 in the stash C<Baz::> in C<Bar::>'s stash.
908 To get the stash pointer for a particular package, use the function:
910 HV* gv_stashpv(const char* name, I32 flags)
911 HV* gv_stashsv(SV*, I32 flags)
913 The first function takes a literal string, the second uses the string stored
914 in the SV. Remember that a stash is just a hash table, so you get back an
915 C<HV*>. The C<flags> flag will create a new package if it is set to GV_ADD.
917 The name that C<gv_stash*v> wants is the name of the package whose symbol table
918 you want. The default package is called C<main>. If you have multiply nested
919 packages, pass their names to C<gv_stash*v>, separated by C<::> as in the Perl
922 Alternately, if you have an SV that is a blessed reference, you can find
923 out the stash pointer by using:
925 HV* SvSTASH(SvRV(SV*));
927 then use the following to get the package name itself:
929 char* HvNAME(HV* stash);
931 If you need to bless or re-bless an object you can use the following
934 SV* sv_bless(SV*, HV* stash)
936 where the first argument, an C<SV*>, must be a reference, and the second
937 argument is a stash. The returned C<SV*> can now be used in the same way
940 For more information on references and blessings, consult L<perlref>.
942 =head2 Double-Typed SVs
944 Scalar variables normally contain only one type of value, an integer,
945 double, pointer, or reference. Perl will automatically convert the
946 actual scalar data from the stored type into the requested type.
948 Some scalar variables contain more than one type of scalar data. For
949 example, the variable C<$!> contains either the numeric value of C<errno>
950 or its string equivalent from either C<strerror> or C<sys_errlist[]>.
952 To force multiple data values into an SV, you must do two things: use the
953 C<sv_set*v> routines to add the additional scalar type, then set a flag
954 so that Perl will believe it contains more than one type of data. The
955 four macros to set the flags are:
962 The particular macro you must use depends on which C<sv_set*v> routine
963 you called first. This is because every C<sv_set*v> routine turns on
964 only the bit for the particular type of data being set, and turns off
967 For example, to create a new Perl variable called "dberror" that contains
968 both the numeric and descriptive string error values, you could use the
972 extern char *dberror_list;
974 SV* sv = get_sv("dberror", GV_ADD);
975 sv_setiv(sv, (IV) dberror);
976 sv_setpv(sv, dberror_list[dberror]);
979 If the order of C<sv_setiv> and C<sv_setpv> had been reversed, then the
980 macro C<SvPOK_on> would need to be called instead of C<SvIOK_on>.
982 =head2 Read-Only Values
984 In Perl 5.16 and earlier, copy-on-write (see the next section) shared a
985 flag bit with read-only scalars. So the only way to test whether
986 C<sv_setsv>, etc., will raise a "Modification of a read-only value" error
987 in those versions is:
989 SvREADONLY(sv) && !SvIsCOW(sv)
991 Under Perl 5.18 and later, SvREADONLY only applies to read-only variables,
992 and, under 5.20, copy-on-write scalars can also be read-only, so the above
993 check is incorrect. You just want:
997 If you need to do this check often, define your own macro like this:
999 #if PERL_VERSION >= 18
1000 # define SvTRULYREADONLY(sv) SvREADONLY(sv)
1002 # define SvTRULYREADONLY(sv) (SvREADONLY(sv) && !SvIsCOW(sv))
1005 =head2 Copy on Write
1007 Perl implements a copy-on-write (COW) mechanism for scalars, in which
1008 string copies are not immediately made when requested, but are deferred
1009 until made necessary by one or the other scalar changing. This is mostly
1010 transparent, but one must take care not to modify string buffers that are
1011 shared by multiple SVs.
1013 You can test whether an SV is using copy-on-write with C<SvIsCOW(sv)>.
1015 You can force an SV to make its own copy of its string buffer by calling C<sv_force_normal(sv)> or SvPV_force_nolen(sv).
1017 If you want to make the SV drop its string buffer, use
1018 C<sv_force_normal_flags(sv, SV_COW_DROP_PV)> or simply
1019 C<sv_setsv(sv, NULL)>.
1021 All of these functions will croak on read-only scalars (see the previous
1022 section for more on those).
1024 To test that your code is behaving correctly and not modifying COW buffers,
1025 on systems that support L<mmap(2)> (i.e., Unix) you can configure perl with
1026 C<-Accflags=-DPERL_DEBUG_READONLY_COW> and it will turn buffer violations
1027 into crashes. You will find it to be marvellously slow, so you may want to
1028 skip perl's own tests.
1030 =head2 Magic Variables
1032 [This section still under construction. Ignore everything here. Post no
1033 bills. Everything not permitted is forbidden.]
1035 Any SV may be magical, that is, it has special features that a normal
1036 SV does not have. These features are stored in the SV structure in a
1037 linked list of C<struct magic>'s, typedef'ed to C<MAGIC>.
1040 MAGIC* mg_moremagic;
1050 Note this is current as of patchlevel 0, and could change at any time.
1052 =head2 Assigning Magic
1054 Perl adds magic to an SV using the sv_magic function:
1056 void sv_magic(SV* sv, SV* obj, int how, const char* name, I32 namlen);
1058 The C<sv> argument is a pointer to the SV that is to acquire a new magical
1061 If C<sv> is not already magical, Perl uses the C<SvUPGRADE> macro to
1062 convert C<sv> to type C<SVt_PVMG>.
1063 Perl then continues by adding new magic
1064 to the beginning of the linked list of magical features. Any prior entry
1065 of the same type of magic is deleted. Note that this can be overridden,
1066 and multiple instances of the same type of magic can be associated with an
1069 The C<name> and C<namlen> arguments are used to associate a string with
1070 the magic, typically the name of a variable. C<namlen> is stored in the
1071 C<mg_len> field and if C<name> is non-null then either a C<savepvn> copy of
1072 C<name> or C<name> itself is stored in the C<mg_ptr> field, depending on
1073 whether C<namlen> is greater than zero or equal to zero respectively. As a
1074 special case, if C<(name && namlen == HEf_SVKEY)> then C<name> is assumed
1075 to contain an C<SV*> and is stored as-is with its REFCNT incremented.
1077 The sv_magic function uses C<how> to determine which, if any, predefined
1078 "Magic Virtual Table" should be assigned to the C<mg_virtual> field.
1079 See the L</Magic Virtual Tables> section below. The C<how> argument is also
1080 stored in the C<mg_type> field. The value of
1081 C<how> should be chosen from the set of macros
1082 C<PERL_MAGIC_foo> found in F<perl.h>. Note that before
1083 these macros were added, Perl internals used to directly use character
1084 literals, so you may occasionally come across old code or documentation
1085 referring to 'U' magic rather than C<PERL_MAGIC_uvar> for example.
1087 The C<obj> argument is stored in the C<mg_obj> field of the C<MAGIC>
1088 structure. If it is not the same as the C<sv> argument, the reference
1089 count of the C<obj> object is incremented. If it is the same, or if
1090 the C<how> argument is C<PERL_MAGIC_arylen>, or if it is a NULL pointer,
1091 then C<obj> is merely stored, without the reference count being incremented.
1093 See also C<sv_magicext> in L<perlapi> for a more flexible way to add magic
1096 There is also a function to add magic to an C<HV>:
1098 void hv_magic(HV *hv, GV *gv, int how);
1100 This simply calls C<sv_magic> and coerces the C<gv> argument into an C<SV>.
1102 To remove the magic from an SV, call the function sv_unmagic:
1104 int sv_unmagic(SV *sv, int type);
1106 The C<type> argument should be equal to the C<how> value when the C<SV>
1107 was initially made magical.
1109 However, note that C<sv_unmagic> removes all magic of a certain C<type> from the
1110 C<SV>. If you want to remove only certain
1111 magic of a C<type> based on the magic
1112 virtual table, use C<sv_unmagicext> instead:
1114 int sv_unmagicext(SV *sv, int type, MGVTBL *vtbl);
1116 =head2 Magic Virtual Tables
1118 The C<mg_virtual> field in the C<MAGIC> structure is a pointer to an
1119 C<MGVTBL>, which is a structure of function pointers and stands for
1120 "Magic Virtual Table" to handle the various operations that might be
1121 applied to that variable.
1123 The C<MGVTBL> has five (or sometimes eight) pointers to the following
1126 int (*svt_get)(SV* sv, MAGIC* mg);
1127 int (*svt_set)(SV* sv, MAGIC* mg);
1128 U32 (*svt_len)(SV* sv, MAGIC* mg);
1129 int (*svt_clear)(SV* sv, MAGIC* mg);
1130 int (*svt_free)(SV* sv, MAGIC* mg);
1132 int (*svt_copy)(SV *sv, MAGIC* mg, SV *nsv,
1133 const char *name, I32 namlen);
1134 int (*svt_dup)(MAGIC *mg, CLONE_PARAMS *param);
1135 int (*svt_local)(SV *nsv, MAGIC *mg);
1138 This MGVTBL structure is set at compile-time in F<perl.h> and there are
1139 currently 32 types. These different structures contain pointers to various
1140 routines that perform additional actions depending on which function is
1143 Function pointer Action taken
1144 ---------------- ------------
1145 svt_get Do something before the value of the SV is
1147 svt_set Do something after the SV is assigned a value.
1148 svt_len Report on the SV's length.
1149 svt_clear Clear something the SV represents.
1150 svt_free Free any extra storage associated with the SV.
1152 svt_copy copy tied variable magic to a tied element
1153 svt_dup duplicate a magic structure during thread cloning
1154 svt_local copy magic to local value during 'local'
1156 For instance, the MGVTBL structure called C<vtbl_sv> (which corresponds
1157 to an C<mg_type> of C<PERL_MAGIC_sv>) contains:
1159 { magic_get, magic_set, magic_len, 0, 0 }
1161 Thus, when an SV is determined to be magical and of type C<PERL_MAGIC_sv>,
1162 if a get operation is being performed, the routine C<magic_get> is
1163 called. All the various routines for the various magical types begin
1164 with C<magic_>. NOTE: the magic routines are not considered part of
1165 the Perl API, and may not be exported by the Perl library.
1167 The last three slots are a recent addition, and for source code
1168 compatibility they are only checked for if one of the three flags
1169 MGf_COPY, MGf_DUP or MGf_LOCAL is set in mg_flags.
1170 This means that most code can continue declaring
1171 a vtable as a 5-element value. These three are
1172 currently used exclusively by the threading code, and are highly subject
1175 The current kinds of Magic Virtual Tables are:
1178 This table is generated by regen/mg_vtable.pl. Any changes made here
1181 =for mg_vtable.pl begin
1184 (old-style char and macro) MGVTBL Type of magic
1185 -------------------------- ------ -------------
1186 \0 PERL_MAGIC_sv vtbl_sv Special scalar variable
1187 # PERL_MAGIC_arylen vtbl_arylen Array length ($#ary)
1188 % PERL_MAGIC_rhash (none) Extra data for restricted
1190 * PERL_MAGIC_debugvar vtbl_debugvar $DB::single, signal, trace
1192 . PERL_MAGIC_pos vtbl_pos pos() lvalue
1193 : PERL_MAGIC_symtab (none) Extra data for symbol
1195 < PERL_MAGIC_backref vtbl_backref For weak ref data
1196 @ PERL_MAGIC_arylen_p (none) To move arylen out of XPVAV
1197 B PERL_MAGIC_bm vtbl_regexp Boyer-Moore
1198 (fast string search)
1199 c PERL_MAGIC_overload_table vtbl_ovrld Holds overload table
1201 D PERL_MAGIC_regdata vtbl_regdata Regex match position data
1203 d PERL_MAGIC_regdatum vtbl_regdatum Regex match position data
1205 E PERL_MAGIC_env vtbl_env %ENV hash
1206 e PERL_MAGIC_envelem vtbl_envelem %ENV hash element
1207 f PERL_MAGIC_fm vtbl_regexp Formline
1209 g PERL_MAGIC_regex_global vtbl_mglob m//g target
1210 H PERL_MAGIC_hints vtbl_hints %^H hash
1211 h PERL_MAGIC_hintselem vtbl_hintselem %^H hash element
1212 I PERL_MAGIC_isa vtbl_isa @ISA array
1213 i PERL_MAGIC_isaelem vtbl_isaelem @ISA array element
1214 k PERL_MAGIC_nkeys vtbl_nkeys scalar(keys()) lvalue
1215 L PERL_MAGIC_dbfile (none) Debugger %_<filename
1216 l PERL_MAGIC_dbline vtbl_dbline Debugger %_<filename
1218 N PERL_MAGIC_shared (none) Shared between threads
1219 n PERL_MAGIC_shared_scalar (none) Shared between threads
1220 o PERL_MAGIC_collxfrm vtbl_collxfrm Locale transformation
1221 P PERL_MAGIC_tied vtbl_pack Tied array or hash
1222 p PERL_MAGIC_tiedelem vtbl_packelem Tied array or hash element
1223 q PERL_MAGIC_tiedscalar vtbl_packelem Tied scalar or handle
1224 r PERL_MAGIC_qr vtbl_regexp Precompiled qr// regex
1225 S PERL_MAGIC_sig (none) %SIG hash
1226 s PERL_MAGIC_sigelem vtbl_sigelem %SIG hash element
1227 t PERL_MAGIC_taint vtbl_taint Taintedness
1228 U PERL_MAGIC_uvar vtbl_uvar Available for use by
1230 u PERL_MAGIC_uvar_elem (none) Reserved for use by
1232 V PERL_MAGIC_vstring (none) SV was vstring literal
1233 v PERL_MAGIC_vec vtbl_vec vec() lvalue
1234 w PERL_MAGIC_utf8 vtbl_utf8 Cached UTF-8 information
1235 x PERL_MAGIC_substr vtbl_substr substr() lvalue
1236 y PERL_MAGIC_defelem vtbl_defelem Shadow "foreach" iterator
1237 variable / smart parameter
1239 \ PERL_MAGIC_lvref vtbl_lvref Lvalue reference
1241 ] PERL_MAGIC_checkcall vtbl_checkcall Inlining/mutation of call
1243 ~ PERL_MAGIC_ext (none) Available for use by
1246 =for mg_vtable.pl end
1248 When an uppercase and lowercase letter both exist in the table, then the
1249 uppercase letter is typically used to represent some kind of composite type
1250 (a list or a hash), and the lowercase letter is used to represent an element
1251 of that composite type. Some internals code makes use of this case
1252 relationship. However, 'v' and 'V' (vec and v-string) are in no way related.
1254 The C<PERL_MAGIC_ext> and C<PERL_MAGIC_uvar> magic types are defined
1255 specifically for use by extensions and will not be used by perl itself.
1256 Extensions can use C<PERL_MAGIC_ext> magic to 'attach' private information
1257 to variables (typically objects). This is especially useful because
1258 there is no way for normal perl code to corrupt this private information
1259 (unlike using extra elements of a hash object).
1261 Similarly, C<PERL_MAGIC_uvar> magic can be used much like tie() to call a
1262 C function any time a scalar's value is used or changed. The C<MAGIC>'s
1263 C<mg_ptr> field points to a C<ufuncs> structure:
1266 I32 (*uf_val)(pTHX_ IV, SV*);
1267 I32 (*uf_set)(pTHX_ IV, SV*);
1271 When the SV is read from or written to, the C<uf_val> or C<uf_set>
1272 function will be called with C<uf_index> as the first arg and a pointer to
1273 the SV as the second. A simple example of how to add C<PERL_MAGIC_uvar>
1274 magic is shown below. Note that the ufuncs structure is copied by
1275 sv_magic, so you can safely allocate it on the stack.
1283 uf.uf_val = &my_get_fn;
1284 uf.uf_set = &my_set_fn;
1286 sv_magic(sv, 0, PERL_MAGIC_uvar, (char*)&uf, sizeof(uf));
1288 Attaching C<PERL_MAGIC_uvar> to arrays is permissible but has no effect.
1290 For hashes there is a specialized hook that gives control over hash
1291 keys (but not values). This hook calls C<PERL_MAGIC_uvar> 'get' magic
1292 if the "set" function in the C<ufuncs> structure is NULL. The hook
1293 is activated whenever the hash is accessed with a key specified as
1294 an C<SV> through the functions C<hv_store_ent>, C<hv_fetch_ent>,
1295 C<hv_delete_ent>, and C<hv_exists_ent>. Accessing the key as a string
1296 through the functions without the C<..._ent> suffix circumvents the
1297 hook. See L<Hash::Util::FieldHash/GUTS> for a detailed description.
1299 Note that because multiple extensions may be using C<PERL_MAGIC_ext>
1300 or C<PERL_MAGIC_uvar> magic, it is important for extensions to take
1301 extra care to avoid conflict. Typically only using the magic on
1302 objects blessed into the same class as the extension is sufficient.
1303 For C<PERL_MAGIC_ext> magic, it is usually a good idea to define an
1304 C<MGVTBL>, even if all its fields will be C<0>, so that individual
1305 C<MAGIC> pointers can be identified as a particular kind of magic
1306 using their magic virtual table. C<mg_findext> provides an easy way
1309 STATIC MGVTBL my_vtbl = { 0, 0, 0, 0, 0, 0, 0, 0 };
1312 if ((mg = mg_findext(sv, PERL_MAGIC_ext, &my_vtbl))) {
1313 /* this is really ours, not another module's PERL_MAGIC_ext */
1314 my_priv_data_t *priv = (my_priv_data_t *)mg->mg_ptr;
1318 Also note that the C<sv_set*()> and C<sv_cat*()> functions described
1319 earlier do B<not> invoke 'set' magic on their targets. This must
1320 be done by the user either by calling the C<SvSETMAGIC()> macro after
1321 calling these functions, or by using one of the C<sv_set*_mg()> or
1322 C<sv_cat*_mg()> functions. Similarly, generic C code must call the
1323 C<SvGETMAGIC()> macro to invoke any 'get' magic if they use an SV
1324 obtained from external sources in functions that don't handle magic.
1325 See L<perlapi> for a description of these functions.
1326 For example, calls to the C<sv_cat*()> functions typically need to be
1327 followed by C<SvSETMAGIC()>, but they don't need a prior C<SvGETMAGIC()>
1328 since their implementation handles 'get' magic.
1330 =head2 Finding Magic
1332 MAGIC *mg_find(SV *sv, int type); /* Finds the magic pointer of that
1335 This routine returns a pointer to a C<MAGIC> structure stored in the SV.
1336 If the SV does not have that magical
1337 feature, C<NULL> is returned. If the
1338 SV has multiple instances of that magical feature, the first one will be
1339 returned. C<mg_findext> can be used
1340 to find a C<MAGIC> structure of an SV
1341 based on both its magic type and its magic virtual table:
1343 MAGIC *mg_findext(SV *sv, int type, MGVTBL *vtbl);
1345 Also, if the SV passed to C<mg_find> or C<mg_findext> is not of type
1346 SVt_PVMG, Perl may core dump.
1348 int mg_copy(SV* sv, SV* nsv, const char* key, STRLEN klen);
1350 This routine checks to see what types of magic C<sv> has. If the mg_type
1351 field is an uppercase letter, then the mg_obj is copied to C<nsv>, but
1352 the mg_type field is changed to be the lowercase letter.
1354 =head2 Understanding the Magic of Tied Hashes and Arrays
1356 Tied hashes and arrays are magical beasts of the C<PERL_MAGIC_tied>
1359 WARNING: As of the 5.004 release, proper usage of the array and hash
1360 access functions requires understanding a few caveats. Some
1361 of these caveats are actually considered bugs in the API, to be fixed
1362 in later releases, and are bracketed with [MAYCHANGE] below. If
1363 you find yourself actually applying such information in this section, be
1364 aware that the behavior may change in the future, umm, without warning.
1366 The perl tie function associates a variable with an object that implements
1367 the various GET, SET, etc methods. To perform the equivalent of the perl
1368 tie function from an XSUB, you must mimic this behaviour. The code below
1369 carries out the necessary steps -- firstly it creates a new hash, and then
1370 creates a second hash which it blesses into the class which will implement
1371 the tie methods. Lastly it ties the two hashes together, and returns a
1372 reference to the new tied hash. Note that the code below does NOT call the
1373 TIEHASH method in the MyTie class -
1374 see L</Calling Perl Routines from within C Programs> for details on how
1385 tie = newRV_noinc((SV*)newHV());
1386 stash = gv_stashpv("MyTie", GV_ADD);
1387 sv_bless(tie, stash);
1388 hv_magic(hash, (GV*)tie, PERL_MAGIC_tied);
1389 RETVAL = newRV_noinc(hash);
1393 The C<av_store> function, when given a tied array argument, merely
1394 copies the magic of the array onto the value to be "stored", using
1395 C<mg_copy>. It may also return NULL, indicating that the value did not
1396 actually need to be stored in the array. [MAYCHANGE] After a call to
1397 C<av_store> on a tied array, the caller will usually need to call
1398 C<mg_set(val)> to actually invoke the perl level "STORE" method on the
1399 TIEARRAY object. If C<av_store> did return NULL, a call to
1400 C<SvREFCNT_dec(val)> will also be usually necessary to avoid a memory
1403 The previous paragraph is applicable verbatim to tied hash access using the
1404 C<hv_store> and C<hv_store_ent> functions as well.
1406 C<av_fetch> and the corresponding hash functions C<hv_fetch> and
1407 C<hv_fetch_ent> actually return an undefined mortal value whose magic
1408 has been initialized using C<mg_copy>. Note the value so returned does not
1409 need to be deallocated, as it is already mortal. [MAYCHANGE] But you will
1410 need to call C<mg_get()> on the returned value in order to actually invoke
1411 the perl level "FETCH" method on the underlying TIE object. Similarly,
1412 you may also call C<mg_set()> on the return value after possibly assigning
1413 a suitable value to it using C<sv_setsv>, which will invoke the "STORE"
1414 method on the TIE object. [/MAYCHANGE]
1417 In other words, the array or hash fetch/store functions don't really
1418 fetch and store actual values in the case of tied arrays and hashes. They
1419 merely call C<mg_copy> to attach magic to the values that were meant to be
1420 "stored" or "fetched". Later calls to C<mg_get> and C<mg_set> actually
1421 do the job of invoking the TIE methods on the underlying objects. Thus
1422 the magic mechanism currently implements a kind of lazy access to arrays
1425 Currently (as of perl version 5.004), use of the hash and array access
1426 functions requires the user to be aware of whether they are operating on
1427 "normal" hashes and arrays, or on their tied variants. The API may be
1428 changed to provide more transparent access to both tied and normal data
1429 types in future versions.
1432 You would do well to understand that the TIEARRAY and TIEHASH interfaces
1433 are mere sugar to invoke some perl method calls while using the uniform hash
1434 and array syntax. The use of this sugar imposes some overhead (typically
1435 about two to four extra opcodes per FETCH/STORE operation, in addition to
1436 the creation of all the mortal variables required to invoke the methods).
1437 This overhead will be comparatively small if the TIE methods are themselves
1438 substantial, but if they are only a few statements long, the overhead
1439 will not be insignificant.
1441 =head2 Localizing changes
1443 Perl has a very handy construction
1450 This construction is I<approximately> equivalent to
1459 The biggest difference is that the first construction would
1460 reinstate the initial value of $var, irrespective of how control exits
1461 the block: C<goto>, C<return>, C<die>/C<eval>, etc. It is a little bit
1462 more efficient as well.
1464 There is a way to achieve a similar task from C via Perl API: create a
1465 I<pseudo-block>, and arrange for some changes to be automatically
1466 undone at the end of it, either explicit, or via a non-local exit (via
1467 die()). A I<block>-like construct is created by a pair of
1468 C<ENTER>/C<LEAVE> macros (see L<perlcall/"Returning a Scalar">).
1469 Such a construct may be created specially for some important localized
1470 task, or an existing one (like boundaries of enclosing Perl
1471 subroutine/block, or an existing pair for freeing TMPs) may be
1472 used. (In the second case the overhead of additional localization must
1473 be almost negligible.) Note that any XSUB is automatically enclosed in
1474 an C<ENTER>/C<LEAVE> pair.
1476 Inside such a I<pseudo-block> the following service is available:
1480 =item C<SAVEINT(int i)>
1482 =item C<SAVEIV(IV i)>
1484 =item C<SAVEI32(I32 i)>
1486 =item C<SAVELONG(long i)>
1488 These macros arrange things to restore the value of integer variable
1489 C<i> at the end of enclosing I<pseudo-block>.
1491 =item C<SAVESPTR(s)>
1493 =item C<SAVEPPTR(p)>
1495 These macros arrange things to restore the value of pointers C<s> and
1496 C<p>. C<s> must be a pointer of a type which survives conversion to
1497 C<SV*> and back, C<p> should be able to survive conversion to C<char*>
1500 =item C<SAVEFREESV(SV *sv)>
1502 The refcount of C<sv> will be decremented at the end of
1503 I<pseudo-block>. This is similar to C<sv_2mortal> in that it is also a
1504 mechanism for doing a delayed C<SvREFCNT_dec>. However, while C<sv_2mortal>
1505 extends the lifetime of C<sv> until the beginning of the next statement,
1506 C<SAVEFREESV> extends it until the end of the enclosing scope. These
1507 lifetimes can be wildly different.
1509 Also compare C<SAVEMORTALIZESV>.
1511 =item C<SAVEMORTALIZESV(SV *sv)>
1513 Just like C<SAVEFREESV>, but mortalizes C<sv> at the end of the current
1514 scope instead of decrementing its reference count. This usually has the
1515 effect of keeping C<sv> alive until the statement that called the currently
1516 live scope has finished executing.
1518 =item C<SAVEFREEOP(OP *op)>
1520 The C<OP *> is op_free()ed at the end of I<pseudo-block>.
1522 =item C<SAVEFREEPV(p)>
1524 The chunk of memory which is pointed to by C<p> is Safefree()ed at the
1525 end of I<pseudo-block>.
1527 =item C<SAVECLEARSV(SV *sv)>
1529 Clears a slot in the current scratchpad which corresponds to C<sv> at
1530 the end of I<pseudo-block>.
1532 =item C<SAVEDELETE(HV *hv, char *key, I32 length)>
1534 The key C<key> of C<hv> is deleted at the end of I<pseudo-block>. The
1535 string pointed to by C<key> is Safefree()ed. If one has a I<key> in
1536 short-lived storage, the corresponding string may be reallocated like
1539 SAVEDELETE(PL_defstash, savepv(tmpbuf), strlen(tmpbuf));
1541 =item C<SAVEDESTRUCTOR(DESTRUCTORFUNC_NOCONTEXT_t f, void *p)>
1543 At the end of I<pseudo-block> the function C<f> is called with the
1546 =item C<SAVEDESTRUCTOR_X(DESTRUCTORFUNC_t f, void *p)>
1548 At the end of I<pseudo-block> the function C<f> is called with the
1549 implicit context argument (if any), and C<p>.
1551 =item C<SAVESTACK_POS()>
1553 The current offset on the Perl internal stack (cf. C<SP>) is restored
1554 at the end of I<pseudo-block>.
1558 The following API list contains functions, thus one needs to
1559 provide pointers to the modifiable data explicitly (either C pointers,
1560 or Perlish C<GV *>s). Where the above macros take C<int>, a similar
1561 function takes C<int *>.
1565 =item C<SV* save_scalar(GV *gv)>
1567 Equivalent to Perl code C<local $gv>.
1569 =item C<AV* save_ary(GV *gv)>
1571 =item C<HV* save_hash(GV *gv)>
1573 Similar to C<save_scalar>, but localize C<@gv> and C<%gv>.
1575 =item C<void save_item(SV *item)>
1577 Duplicates the current value of C<SV>, on the exit from the current
1578 C<ENTER>/C<LEAVE> I<pseudo-block> will restore the value of C<SV>
1579 using the stored value. It doesn't handle magic. Use C<save_scalar> if
1582 =item C<void save_list(SV **sarg, I32 maxsarg)>
1584 A variant of C<save_item> which takes multiple arguments via an array
1585 C<sarg> of C<SV*> of length C<maxsarg>.
1587 =item C<SV* save_svref(SV **sptr)>
1589 Similar to C<save_scalar>, but will reinstate an C<SV *>.
1591 =item C<void save_aptr(AV **aptr)>
1593 =item C<void save_hptr(HV **hptr)>
1595 Similar to C<save_svref>, but localize C<AV *> and C<HV *>.
1599 The C<Alias> module implements localization of the basic types within the
1600 I<caller's scope>. People who are interested in how to localize things in
1601 the containing scope should take a look there too.
1605 =head2 XSUBs and the Argument Stack
1607 The XSUB mechanism is a simple way for Perl programs to access C subroutines.
1608 An XSUB routine will have a stack that contains the arguments from the Perl
1609 program, and a way to map from the Perl data structures to a C equivalent.
1611 The stack arguments are accessible through the C<ST(n)> macro, which returns
1612 the C<n>'th stack argument. Argument 0 is the first argument passed in the
1613 Perl subroutine call. These arguments are C<SV*>, and can be used anywhere
1616 Most of the time, output from the C routine can be handled through use of
1617 the RETVAL and OUTPUT directives. However, there are some cases where the
1618 argument stack is not already long enough to handle all the return values.
1619 An example is the POSIX tzname() call, which takes no arguments, but returns
1620 two, the local time zone's standard and summer time abbreviations.
1622 To handle this situation, the PPCODE directive is used and the stack is
1623 extended using the macro:
1627 where C<SP> is the macro that represents the local copy of the stack pointer,
1628 and C<num> is the number of elements the stack should be extended by.
1630 Now that there is room on the stack, values can be pushed on it using C<PUSHs>
1631 macro. The pushed values will often need to be "mortal" (See
1632 L</Reference Counts and Mortality>):
1634 PUSHs(sv_2mortal(newSViv(an_integer)))
1635 PUSHs(sv_2mortal(newSVuv(an_unsigned_integer)))
1636 PUSHs(sv_2mortal(newSVnv(a_double)))
1637 PUSHs(sv_2mortal(newSVpv("Some String",0)))
1638 /* Although the last example is better written as the more
1640 PUSHs(newSVpvs_flags("Some String", SVs_TEMP))
1642 And now the Perl program calling C<tzname>, the two values will be assigned
1645 ($standard_abbrev, $summer_abbrev) = POSIX::tzname;
1647 An alternate (and possibly simpler) method to pushing values on the stack is
1652 This macro automatically adjusts the stack for you, if needed. Thus, you
1653 do not need to call C<EXTEND> to extend the stack.
1655 Despite their suggestions in earlier versions of this document the macros
1656 C<(X)PUSH[iunp]> are I<not> suited to XSUBs which return multiple results.
1657 For that, either stick to the C<(X)PUSHs> macros shown above, or use the new
1658 C<m(X)PUSH[iunp]> macros instead; see L</Putting a C value on Perl stack>.
1660 For more information, consult L<perlxs> and L<perlxstut>.
1662 =head2 Autoloading with XSUBs
1664 If an AUTOLOAD routine is an XSUB, as with Perl subroutines, Perl puts the
1665 fully-qualified name of the autoloaded subroutine in the $AUTOLOAD variable
1666 of the XSUB's package.
1668 But it also puts the same information in certain fields of the XSUB itself:
1670 HV *stash = CvSTASH(cv);
1671 const char *subname = SvPVX(cv);
1672 STRLEN name_length = SvCUR(cv); /* in bytes */
1673 U32 is_utf8 = SvUTF8(cv);
1675 C<SvPVX(cv)> contains just the sub name itself, not including the package.
1676 For an AUTOLOAD routine in UNIVERSAL or one of its superclasses,
1677 C<CvSTASH(cv)> returns NULL during a method call on a nonexistent package.
1679 B<Note>: Setting $AUTOLOAD stopped working in 5.6.1, which did not support
1680 XS AUTOLOAD subs at all. Perl 5.8.0 introduced the use of fields in the
1681 XSUB itself. Perl 5.16.0 restored the setting of $AUTOLOAD. If you need
1682 to support 5.8-5.14, use the XSUB's fields.
1684 =head2 Calling Perl Routines from within C Programs
1686 There are four routines that can be used to call a Perl subroutine from
1687 within a C program. These four are:
1689 I32 call_sv(SV*, I32);
1690 I32 call_pv(const char*, I32);
1691 I32 call_method(const char*, I32);
1692 I32 call_argv(const char*, I32, char**);
1694 The routine most often used is C<call_sv>. The C<SV*> argument
1695 contains either the name of the Perl subroutine to be called, or a
1696 reference to the subroutine. The second argument consists of flags
1697 that control the context in which the subroutine is called, whether
1698 or not the subroutine is being passed arguments, how errors should be
1699 trapped, and how to treat return values.
1701 All four routines return the number of arguments that the subroutine returned
1704 These routines used to be called C<perl_call_sv>, etc., before Perl v5.6.0,
1705 but those names are now deprecated; macros of the same name are provided for
1708 When using any of these routines (except C<call_argv>), the programmer
1709 must manipulate the Perl stack. These include the following macros and
1724 For a detailed description of calling conventions from C to Perl,
1725 consult L<perlcall>.
1727 =head2 Putting a C value on Perl stack
1729 A lot of opcodes (this is an elementary operation in the internal perl
1730 stack machine) put an SV* on the stack. However, as an optimization
1731 the corresponding SV is (usually) not recreated each time. The opcodes
1732 reuse specially assigned SVs (I<target>s) which are (as a corollary)
1733 not constantly freed/created.
1735 Each of the targets is created only once (but see
1736 L</Scratchpads and recursion> below), and when an opcode needs to put
1737 an integer, a double, or a string on stack, it just sets the
1738 corresponding parts of its I<target> and puts the I<target> on stack.
1740 The macro to put this target on stack is C<PUSHTARG>, and it is
1741 directly used in some opcodes, as well as indirectly in zillions of
1742 others, which use it via C<(X)PUSH[iunp]>.
1744 Because the target is reused, you must be careful when pushing multiple
1745 values on the stack. The following code will not do what you think:
1750 This translates as "set C<TARG> to 10, push a pointer to C<TARG> onto
1751 the stack; set C<TARG> to 20, push a pointer to C<TARG> onto the stack".
1752 At the end of the operation, the stack does not contain the values 10
1753 and 20, but actually contains two pointers to C<TARG>, which we have set
1756 If you need to push multiple different values then you should either use
1757 the C<(X)PUSHs> macros, or else use the new C<m(X)PUSH[iunp]> macros,
1758 none of which make use of C<TARG>. The C<(X)PUSHs> macros simply push an
1759 SV* on the stack, which, as noted under L</XSUBs and the Argument Stack>,
1760 will often need to be "mortal". The new C<m(X)PUSH[iunp]> macros make
1761 this a little easier to achieve by creating a new mortal for you (via
1762 C<(X)PUSHmortal>), pushing that onto the stack (extending it if necessary
1763 in the case of the C<mXPUSH[iunp]> macros), and then setting its value.
1764 Thus, instead of writing this to "fix" the example above:
1766 XPUSHs(sv_2mortal(newSViv(10)))
1767 XPUSHs(sv_2mortal(newSViv(20)))
1769 you can simply write:
1774 On a related note, if you do use C<(X)PUSH[iunp]>, then you're going to
1775 need a C<dTARG> in your variable declarations so that the C<*PUSH*>
1776 macros can make use of the local variable C<TARG>. See also C<dTARGET>
1781 The question remains on when the SVs which are I<target>s for opcodes
1782 are created. The answer is that they are created when the current
1783 unit--a subroutine or a file (for opcodes for statements outside of
1784 subroutines)--is compiled. During this time a special anonymous Perl
1785 array is created, which is called a scratchpad for the current unit.
1787 A scratchpad keeps SVs which are lexicals for the current unit and are
1788 targets for opcodes. A previous version of this document
1789 stated that one can deduce that an SV lives on a scratchpad
1790 by looking on its flags: lexicals have C<SVs_PADMY> set, and
1791 I<target>s have C<SVs_PADTMP> set. But this has never been fully true.
1792 C<SVs_PADMY> could be set on a variable that no longer resides in any pad.
1793 While I<target>s do have C<SVs_PADTMP> set, it can also be set on variables
1794 that have never resided in a pad, but nonetheless act like I<target>s. As
1795 of perl 5.21.5, the C<SVs_PADMY> flag is no longer used and is defined as
1796 0. C<SvPADMY()> now returns true for anything without C<SVs_PADTMP>.
1798 The correspondence between OPs and I<target>s is not 1-to-1. Different
1799 OPs in the compile tree of the unit can use the same target, if this
1800 would not conflict with the expected life of the temporary.
1802 =head2 Scratchpads and recursion
1804 In fact it is not 100% true that a compiled unit contains a pointer to
1805 the scratchpad AV. In fact it contains a pointer to an AV of
1806 (initially) one element, and this element is the scratchpad AV. Why do
1807 we need an extra level of indirection?
1809 The answer is B<recursion>, and maybe B<threads>. Both
1810 these can create several execution pointers going into the same
1811 subroutine. For the subroutine-child not write over the temporaries
1812 for the subroutine-parent (lifespan of which covers the call to the
1813 child), the parent and the child should have different
1814 scratchpads. (I<And> the lexicals should be separate anyway!)
1816 So each subroutine is born with an array of scratchpads (of length 1).
1817 On each entry to the subroutine it is checked that the current
1818 depth of the recursion is not more than the length of this array, and
1819 if it is, new scratchpad is created and pushed into the array.
1821 The I<target>s on this scratchpad are C<undef>s, but they are already
1822 marked with correct flags.
1824 =head1 Memory Allocation
1828 All memory meant to be used with the Perl API functions should be manipulated
1829 using the macros described in this section. The macros provide the necessary
1830 transparency between differences in the actual malloc implementation that is
1833 It is suggested that you enable the version of malloc that is distributed
1834 with Perl. It keeps pools of various sizes of unallocated memory in
1835 order to satisfy allocation requests more quickly. However, on some
1836 platforms, it may cause spurious malloc or free errors.
1838 The following three macros are used to initially allocate memory :
1840 Newx(pointer, number, type);
1841 Newxc(pointer, number, type, cast);
1842 Newxz(pointer, number, type);
1844 The first argument C<pointer> should be the name of a variable that will
1845 point to the newly allocated memory.
1847 The second and third arguments C<number> and C<type> specify how many of
1848 the specified type of data structure should be allocated. The argument
1849 C<type> is passed to C<sizeof>. The final argument to C<Newxc>, C<cast>,
1850 should be used if the C<pointer> argument is different from the C<type>
1853 Unlike the C<Newx> and C<Newxc> macros, the C<Newxz> macro calls C<memzero>
1854 to zero out all the newly allocated memory.
1858 Renew(pointer, number, type);
1859 Renewc(pointer, number, type, cast);
1862 These three macros are used to change a memory buffer size or to free a
1863 piece of memory no longer needed. The arguments to C<Renew> and C<Renewc>
1864 match those of C<New> and C<Newc> with the exception of not needing the
1865 "magic cookie" argument.
1869 Move(source, dest, number, type);
1870 Copy(source, dest, number, type);
1871 Zero(dest, number, type);
1873 These three macros are used to move, copy, or zero out previously allocated
1874 memory. The C<source> and C<dest> arguments point to the source and
1875 destination starting points. Perl will move, copy, or zero out C<number>
1876 instances of the size of the C<type> data structure (using the C<sizeof>
1881 The most recent development releases of Perl have been experimenting with
1882 removing Perl's dependency on the "normal" standard I/O suite and allowing
1883 other stdio implementations to be used. This involves creating a new
1884 abstraction layer that then calls whichever implementation of stdio Perl
1885 was compiled with. All XSUBs should now use the functions in the PerlIO
1886 abstraction layer and not make any assumptions about what kind of stdio
1889 For a complete description of the PerlIO abstraction, consult L<perlapio>.
1891 =head1 Compiled code
1895 Here we describe the internal form your code is converted to by
1896 Perl. Start with a simple example:
1900 This is converted to a tree similar to this one:
1908 (but slightly more complicated). This tree reflects the way Perl
1909 parsed your code, but has nothing to do with the execution order.
1910 There is an additional "thread" going through the nodes of the tree
1911 which shows the order of execution of the nodes. In our simplified
1912 example above it looks like:
1914 $b ---> $c ---> + ---> $a ---> assign-to
1916 But with the actual compile tree for C<$a = $b + $c> it is different:
1917 some nodes I<optimized away>. As a corollary, though the actual tree
1918 contains more nodes than our simplified example, the execution order
1919 is the same as in our example.
1921 =head2 Examining the tree
1923 If you have your perl compiled for debugging (usually done with
1924 C<-DDEBUGGING> on the C<Configure> command line), you may examine the
1925 compiled tree by specifying C<-Dx> on the Perl command line. The
1926 output takes several lines per node, and for C<$b+$c> it looks like
1931 FLAGS = (SCALAR,KIDS)
1933 TYPE = null ===> (4)
1935 FLAGS = (SCALAR,KIDS)
1937 3 TYPE = gvsv ===> 4
1943 TYPE = null ===> (5)
1945 FLAGS = (SCALAR,KIDS)
1947 4 TYPE = gvsv ===> 5
1953 This tree has 5 nodes (one per C<TYPE> specifier), only 3 of them are
1954 not optimized away (one per number in the left column). The immediate
1955 children of the given node correspond to C<{}> pairs on the same level
1956 of indentation, thus this listing corresponds to the tree:
1964 The execution order is indicated by C<===E<gt>> marks, thus it is C<3
1965 4 5 6> (node C<6> is not included into above listing), i.e.,
1966 C<gvsv gvsv add whatever>.
1968 Each of these nodes represents an op, a fundamental operation inside the
1969 Perl core. The code which implements each operation can be found in the
1970 F<pp*.c> files; the function which implements the op with type C<gvsv>
1971 is C<pp_gvsv>, and so on. As the tree above shows, different ops have
1972 different numbers of children: C<add> is a binary operator, as one would
1973 expect, and so has two children. To accommodate the various different
1974 numbers of children, there are various types of op data structure, and
1975 they link together in different ways.
1977 The simplest type of op structure is C<OP>: this has no children. Unary
1978 operators, C<UNOP>s, have one child, and this is pointed to by the
1979 C<op_first> field. Binary operators (C<BINOP>s) have not only an
1980 C<op_first> field but also an C<op_last> field. The most complex type of
1981 op is a C<LISTOP>, which has any number of children. In this case, the
1982 first child is pointed to by C<op_first> and the last child by
1983 C<op_last>. The children in between can be found by iteratively
1984 following the C<OpSIBLING> pointer from the first child to the last (but
1987 There are also some other op types: a C<PMOP> holds a regular expression,
1988 and has no children, and a C<LOOP> may or may not have children. If the
1989 C<op_children> field is non-zero, it behaves like a C<LISTOP>. To
1990 complicate matters, if a C<UNOP> is actually a C<null> op after
1991 optimization (see L</Compile pass 2: context propagation>) it will still
1992 have children in accordance with its former type.
1994 Finally, there is a C<LOGOP>, or logic op. Like a C<LISTOP>, this has one
1995 or more children, but it doesn't have an C<op_last> field: so you have to
1996 follow C<op_first> and then the C<OpSIBLING> chain itself to find the
1997 last child. Instead it has an C<op_other> field, which is comparable to
1998 the C<op_next> field described below, and represents an alternate
1999 execution path. Operators like C<and>, C<or> and C<?> are C<LOGOP>s. Note
2000 that in general, C<op_other> may not point to any of the direct children
2003 Starting in version 5.21.2, perls built with the experimental
2004 define C<-DPERL_OP_PARENT> add an extra boolean flag for each op,
2005 C<op_moresib>. When not set, this indicates that this is the last op in an
2006 C<OpSIBLING> chain. This frees up the C<op_sibling> field on the last
2007 sibling to point back to the parent op. Under this build, that field is
2008 also renamed C<op_sibparent> to reflect its joint role. The macro
2009 C<OpSIBLING(o)> wraps this special behaviour, and always returns NULL on
2010 the last sibling. With this build the C<op_parent(o)> function can be
2011 used to find the parent of any op. Thus for forward compatibility, you
2012 should always use the C<OpSIBLING(o)> macro rather than accessing
2013 C<op_sibling> directly.
2015 Another way to examine the tree is to use a compiler back-end module, such
2018 =head2 Compile pass 1: check routines
2020 The tree is created by the compiler while I<yacc> code feeds it
2021 the constructions it recognizes. Since I<yacc> works bottom-up, so does
2022 the first pass of perl compilation.
2024 What makes this pass interesting for perl developers is that some
2025 optimization may be performed on this pass. This is optimization by
2026 so-called "check routines". The correspondence between node names
2027 and corresponding check routines is described in F<opcode.pl> (do not
2028 forget to run C<make regen_headers> if you modify this file).
2030 A check routine is called when the node is fully constructed except
2031 for the execution-order thread. Since at this time there are no
2032 back-links to the currently constructed node, one can do most any
2033 operation to the top-level node, including freeing it and/or creating
2034 new nodes above/below it.
2036 The check routine returns the node which should be inserted into the
2037 tree (if the top-level node was not modified, check routine returns
2040 By convention, check routines have names C<ck_*>. They are usually
2041 called from C<new*OP> subroutines (or C<convert>) (which in turn are
2042 called from F<perly.y>).
2044 =head2 Compile pass 1a: constant folding
2046 Immediately after the check routine is called the returned node is
2047 checked for being compile-time executable. If it is (the value is
2048 judged to be constant) it is immediately executed, and a I<constant>
2049 node with the "return value" of the corresponding subtree is
2050 substituted instead. The subtree is deleted.
2052 If constant folding was not performed, the execution-order thread is
2055 =head2 Compile pass 2: context propagation
2057 When a context for a part of compile tree is known, it is propagated
2058 down through the tree. At this time the context can have 5 values
2059 (instead of 2 for runtime context): void, boolean, scalar, list, and
2060 lvalue. In contrast with the pass 1 this pass is processed from top
2061 to bottom: a node's context determines the context for its children.
2063 Additional context-dependent optimizations are performed at this time.
2064 Since at this moment the compile tree contains back-references (via
2065 "thread" pointers), nodes cannot be free()d now. To allow
2066 optimized-away nodes at this stage, such nodes are null()ified instead
2067 of free()ing (i.e. their type is changed to OP_NULL).
2069 =head2 Compile pass 3: peephole optimization
2071 After the compile tree for a subroutine (or for an C<eval> or a file)
2072 is created, an additional pass over the code is performed. This pass
2073 is neither top-down or bottom-up, but in the execution order (with
2074 additional complications for conditionals). Optimizations performed
2075 at this stage are subject to the same restrictions as in the pass 2.
2077 Peephole optimizations are done by calling the function pointed to
2078 by the global variable C<PL_peepp>. By default, C<PL_peepp> just
2079 calls the function pointed to by the global variable C<PL_rpeepp>.
2080 By default, that performs some basic op fixups and optimisations along
2081 the execution-order op chain, and recursively calls C<PL_rpeepp> for
2082 each side chain of ops (resulting from conditionals). Extensions may
2083 provide additional optimisations or fixups, hooking into either the
2084 per-subroutine or recursive stage, like this:
2086 static peep_t prev_peepp;
2087 static void my_peep(pTHX_ OP *o)
2089 /* custom per-subroutine optimisation goes here */
2090 prev_peepp(aTHX_ o);
2091 /* custom per-subroutine optimisation may also go here */
2094 prev_peepp = PL_peepp;
2097 static peep_t prev_rpeepp;
2098 static void my_rpeep(pTHX_ OP *o)
2101 for(; o; o = o->op_next) {
2102 /* custom per-op optimisation goes here */
2104 prev_rpeepp(aTHX_ orig_o);
2107 prev_rpeepp = PL_rpeepp;
2108 PL_rpeepp = my_rpeep;
2110 =head2 Pluggable runops
2112 The compile tree is executed in a runops function. There are two runops
2113 functions, in F<run.c> and in F<dump.c>. C<Perl_runops_debug> is used
2114 with DEBUGGING and C<Perl_runops_standard> is used otherwise. For fine
2115 control over the execution of the compile tree it is possible to provide
2116 your own runops function.
2118 It's probably best to copy one of the existing runops functions and
2119 change it to suit your needs. Then, in the BOOT section of your XS
2122 PL_runops = my_runops;
2124 This function should be as efficient as possible to keep your programs
2125 running as fast as possible.
2127 =head2 Compile-time scope hooks
2129 As of perl 5.14 it is possible to hook into the compile-time lexical
2130 scope mechanism using C<Perl_blockhook_register>. This is used like
2133 STATIC void my_start_hook(pTHX_ int full);
2134 STATIC BHK my_hooks;
2137 BhkENTRY_set(&my_hooks, bhk_start, my_start_hook);
2138 Perl_blockhook_register(aTHX_ &my_hooks);
2140 This will arrange to have C<my_start_hook> called at the start of
2141 compiling every lexical scope. The available hooks are:
2145 =item C<void bhk_start(pTHX_ int full)>
2147 This is called just after starting a new lexical scope. Note that Perl
2152 creates two scopes: the first starts at the C<(> and has C<full == 1>,
2153 the second starts at the C<{> and has C<full == 0>. Both end at the
2154 C<}>, so calls to C<start> and C<pre>/C<post_end> will match. Anything
2155 pushed onto the save stack by this hook will be popped just before the
2156 scope ends (between the C<pre_> and C<post_end> hooks, in fact).
2158 =item C<void bhk_pre_end(pTHX_ OP **o)>
2160 This is called at the end of a lexical scope, just before unwinding the
2161 stack. I<o> is the root of the optree representing the scope; it is a
2162 double pointer so you can replace the OP if you need to.
2164 =item C<void bhk_post_end(pTHX_ OP **o)>
2166 This is called at the end of a lexical scope, just after unwinding the
2167 stack. I<o> is as above. Note that it is possible for calls to C<pre_>
2168 and C<post_end> to nest, if there is something on the save stack that
2171 =item C<void bhk_eval(pTHX_ OP *const o)>
2173 This is called just before starting to compile an C<eval STRING>, C<do
2174 FILE>, C<require> or C<use>, after the eval has been set up. I<o> is the
2175 OP that requested the eval, and will normally be an C<OP_ENTEREVAL>,
2176 C<OP_DOFILE> or C<OP_REQUIRE>.
2180 Once you have your hook functions, you need a C<BHK> structure to put
2181 them in. It's best to allocate it statically, since there is no way to
2182 free it once it's registered. The function pointers should be inserted
2183 into this structure using the C<BhkENTRY_set> macro, which will also set
2184 flags indicating which entries are valid. If you do need to allocate
2185 your C<BHK> dynamically for some reason, be sure to zero it before you
2188 Once registered, there is no mechanism to switch these hooks off, so if
2189 that is necessary you will need to do this yourself. An entry in C<%^H>
2190 is probably the best way, so the effect is lexically scoped; however it
2191 is also possible to use the C<BhkDISABLE> and C<BhkENABLE> macros to
2192 temporarily switch entries on and off. You should also be aware that
2193 generally speaking at least one scope will have opened before your
2194 extension is loaded, so you will see some C<pre>/C<post_end> pairs that
2195 didn't have a matching C<start>.
2197 =head1 Examining internal data structures with the C<dump> functions
2199 To aid debugging, the source file F<dump.c> contains a number of
2200 functions which produce formatted output of internal data structures.
2202 The most commonly used of these functions is C<Perl_sv_dump>; it's used
2203 for dumping SVs, AVs, HVs, and CVs. The C<Devel::Peek> module calls
2204 C<sv_dump> to produce debugging output from Perl-space, so users of that
2205 module should already be familiar with its format.
2207 C<Perl_op_dump> can be used to dump an C<OP> structure or any of its
2208 derivatives, and produces output similar to C<perl -Dx>; in fact,
2209 C<Perl_dump_eval> will dump the main root of the code being evaluated,
2210 exactly like C<-Dx>.
2212 Other useful functions are C<Perl_dump_sub>, which turns a C<GV> into an
2213 op tree, C<Perl_dump_packsubs> which calls C<Perl_dump_sub> on all the
2214 subroutines in a package like so: (Thankfully, these are all xsubs, so
2215 there is no op tree)
2217 (gdb) print Perl_dump_packsubs(PL_defstash)
2219 SUB attributes::bootstrap = (xsub 0x811fedc 0)
2221 SUB UNIVERSAL::can = (xsub 0x811f50c 0)
2223 SUB UNIVERSAL::isa = (xsub 0x811f304 0)
2225 SUB UNIVERSAL::VERSION = (xsub 0x811f7ac 0)
2227 SUB DynaLoader::boot_DynaLoader = (xsub 0x805b188 0)
2229 and C<Perl_dump_all>, which dumps all the subroutines in the stash and
2230 the op tree of the main root.
2232 =head1 How multiple interpreters and concurrency are supported
2234 =head2 Background and PERL_IMPLICIT_CONTEXT
2236 The Perl interpreter can be regarded as a closed box: it has an API
2237 for feeding it code or otherwise making it do things, but it also has
2238 functions for its own use. This smells a lot like an object, and
2239 there are ways for you to build Perl so that you can have multiple
2240 interpreters, with one interpreter represented either as a C structure,
2241 or inside a thread-specific structure. These structures contain all
2242 the context, the state of that interpreter.
2244 One macro controls the major Perl build flavor: MULTIPLICITY. The
2245 MULTIPLICITY build has a C structure that packages all the interpreter
2246 state. With multiplicity-enabled perls, PERL_IMPLICIT_CONTEXT is also
2247 normally defined, and enables the support for passing in a "hidden" first
2248 argument that represents all three data structures. MULTIPLICITY makes
2249 multi-threaded perls possible (with the ithreads threading model, related
2250 to the macro USE_ITHREADS.)
2252 Two other "encapsulation" macros are the PERL_GLOBAL_STRUCT and
2253 PERL_GLOBAL_STRUCT_PRIVATE (the latter turns on the former, and the
2254 former turns on MULTIPLICITY.) The PERL_GLOBAL_STRUCT causes all the
2255 internal variables of Perl to be wrapped inside a single global struct,
2256 struct perl_vars, accessible as (globals) &PL_Vars or PL_VarsPtr or
2257 the function Perl_GetVars(). The PERL_GLOBAL_STRUCT_PRIVATE goes
2258 one step further, there is still a single struct (allocated in main()
2259 either from heap or from stack) but there are no global data symbols
2260 pointing to it. In either case the global struct should be initialized
2261 as the very first thing in main() using Perl_init_global_struct() and
2262 correspondingly tear it down after perl_free() using Perl_free_global_struct(),
2263 please see F<miniperlmain.c> for usage details. You may also need
2264 to use C<dVAR> in your coding to "declare the global variables"
2265 when you are using them. dTHX does this for you automatically.
2267 To see whether you have non-const data you can use a BSD (or GNU)
2270 nm libperl.a | grep -v ' [TURtr] '
2272 If this displays any C<D> or C<d> symbols (or possibly C<C> or C<c>),
2273 you have non-const data. The symbols the C<grep> removed are as follows:
2274 C<Tt> are I<text>, or code, the C<Rr> are I<read-only> (const) data,
2275 and the C<U> is <undefined>, external symbols referred to.
2277 The test F<t/porting/libperl.t> does this kind of symbol sanity
2278 checking on C<libperl.a>.
2280 For backward compatibility reasons defining just PERL_GLOBAL_STRUCT
2281 doesn't actually hide all symbols inside a big global struct: some
2282 PerlIO_xxx vtables are left visible. The PERL_GLOBAL_STRUCT_PRIVATE
2283 then hides everything (see how the PERLIO_FUNCS_DECL is used).
2285 All this obviously requires a way for the Perl internal functions to be
2286 either subroutines taking some kind of structure as the first
2287 argument, or subroutines taking nothing as the first argument. To
2288 enable these two very different ways of building the interpreter,
2289 the Perl source (as it does in so many other situations) makes heavy
2290 use of macros and subroutine naming conventions.
2292 First problem: deciding which functions will be public API functions and
2293 which will be private. All functions whose names begin C<S_> are private
2294 (think "S" for "secret" or "static"). All other functions begin with
2295 "Perl_", but just because a function begins with "Perl_" does not mean it is
2296 part of the API. (See L</Internal
2297 Functions>.) The easiest way to be B<sure> a
2298 function is part of the API is to find its entry in L<perlapi>.
2299 If it exists in L<perlapi>, it's part of the API. If it doesn't, and you
2300 think it should be (i.e., you need it for your extension), send mail via
2301 L<perlbug> explaining why you think it should be.
2303 Second problem: there must be a syntax so that the same subroutine
2304 declarations and calls can pass a structure as their first argument,
2305 or pass nothing. To solve this, the subroutines are named and
2306 declared in a particular way. Here's a typical start of a static
2307 function used within the Perl guts:
2310 S_incline(pTHX_ char *s)
2312 STATIC becomes "static" in C, and may be #define'd to nothing in some
2313 configurations in the future.
2315 A public function (i.e. part of the internal API, but not necessarily
2316 sanctioned for use in extensions) begins like this:
2319 Perl_sv_setiv(pTHX_ SV* dsv, IV num)
2321 C<pTHX_> is one of a number of macros (in F<perl.h>) that hide the
2322 details of the interpreter's context. THX stands for "thread", "this",
2323 or "thingy", as the case may be. (And no, George Lucas is not involved. :-)
2324 The first character could be 'p' for a B<p>rototype, 'a' for B<a>rgument,
2325 or 'd' for B<d>eclaration, so we have C<pTHX>, C<aTHX> and C<dTHX>, and
2328 When Perl is built without options that set PERL_IMPLICIT_CONTEXT, there is no
2329 first argument containing the interpreter's context. The trailing underscore
2330 in the pTHX_ macro indicates that the macro expansion needs a comma
2331 after the context argument because other arguments follow it. If
2332 PERL_IMPLICIT_CONTEXT is not defined, pTHX_ will be ignored, and the
2333 subroutine is not prototyped to take the extra argument. The form of the
2334 macro without the trailing underscore is used when there are no additional
2337 When a core function calls another, it must pass the context. This
2338 is normally hidden via macros. Consider C<sv_setiv>. It expands into
2339 something like this:
2341 #ifdef PERL_IMPLICIT_CONTEXT
2342 #define sv_setiv(a,b) Perl_sv_setiv(aTHX_ a, b)
2343 /* can't do this for vararg functions, see below */
2345 #define sv_setiv Perl_sv_setiv
2348 This works well, and means that XS authors can gleefully write:
2352 and still have it work under all the modes Perl could have been
2355 This doesn't work so cleanly for varargs functions, though, as macros
2356 imply that the number of arguments is known in advance. Instead we
2357 either need to spell them out fully, passing C<aTHX_> as the first
2358 argument (the Perl core tends to do this with functions like
2359 Perl_warner), or use a context-free version.
2361 The context-free version of Perl_warner is called
2362 Perl_warner_nocontext, and does not take the extra argument. Instead
2363 it does dTHX; to get the context from thread-local storage. We
2364 C<#define warner Perl_warner_nocontext> so that extensions get source
2365 compatibility at the expense of performance. (Passing an arg is
2366 cheaper than grabbing it from thread-local storage.)
2368 You can ignore [pad]THXx when browsing the Perl headers/sources.
2369 Those are strictly for use within the core. Extensions and embedders
2370 need only be aware of [pad]THX.
2372 =head2 So what happened to dTHR?
2374 C<dTHR> was introduced in perl 5.005 to support the older thread model.
2375 The older thread model now uses the C<THX> mechanism to pass context
2376 pointers around, so C<dTHR> is not useful any more. Perl 5.6.0 and
2377 later still have it for backward source compatibility, but it is defined
2380 =head2 How do I use all this in extensions?
2382 When Perl is built with PERL_IMPLICIT_CONTEXT, extensions that call
2383 any functions in the Perl API will need to pass the initial context
2384 argument somehow. The kicker is that you will need to write it in
2385 such a way that the extension still compiles when Perl hasn't been
2386 built with PERL_IMPLICIT_CONTEXT enabled.
2388 There are three ways to do this. First, the easy but inefficient way,
2389 which is also the default, in order to maintain source compatibility
2390 with extensions: whenever F<XSUB.h> is #included, it redefines the aTHX
2391 and aTHX_ macros to call a function that will return the context.
2392 Thus, something like:
2396 in your extension will translate to this when PERL_IMPLICIT_CONTEXT is
2399 Perl_sv_setiv(Perl_get_context(), sv, num);
2401 or to this otherwise:
2403 Perl_sv_setiv(sv, num);
2405 You don't have to do anything new in your extension to get this; since
2406 the Perl library provides Perl_get_context(), it will all just
2409 The second, more efficient way is to use the following template for
2412 #define PERL_NO_GET_CONTEXT /* we want efficiency */
2417 STATIC void my_private_function(int arg1, int arg2);
2420 my_private_function(int arg1, int arg2)
2422 dTHX; /* fetch context */
2423 ... call many Perl API functions ...
2428 MODULE = Foo PACKAGE = Foo
2436 my_private_function(arg, 10);
2438 Note that the only two changes from the normal way of writing an
2439 extension is the addition of a C<#define PERL_NO_GET_CONTEXT> before
2440 including the Perl headers, followed by a C<dTHX;> declaration at
2441 the start of every function that will call the Perl API. (You'll
2442 know which functions need this, because the C compiler will complain
2443 that there's an undeclared identifier in those functions.) No changes
2444 are needed for the XSUBs themselves, because the XS() macro is
2445 correctly defined to pass in the implicit context if needed.
2447 The third, even more efficient way is to ape how it is done within
2451 #define PERL_NO_GET_CONTEXT /* we want efficiency */
2456 /* pTHX_ only needed for functions that call Perl API */
2457 STATIC void my_private_function(pTHX_ int arg1, int arg2);
2460 my_private_function(pTHX_ int arg1, int arg2)
2462 /* dTHX; not needed here, because THX is an argument */
2463 ... call Perl API functions ...
2468 MODULE = Foo PACKAGE = Foo
2476 my_private_function(aTHX_ arg, 10);
2478 This implementation never has to fetch the context using a function
2479 call, since it is always passed as an extra argument. Depending on
2480 your needs for simplicity or efficiency, you may mix the previous
2481 two approaches freely.
2483 Never add a comma after C<pTHX> yourself--always use the form of the
2484 macro with the underscore for functions that take explicit arguments,
2485 or the form without the argument for functions with no explicit arguments.
2487 If one is compiling Perl with the C<-DPERL_GLOBAL_STRUCT> the C<dVAR>
2488 definition is needed if the Perl global variables (see F<perlvars.h>
2489 or F<globvar.sym>) are accessed in the function and C<dTHX> is not
2490 used (the C<dTHX> includes the C<dVAR> if necessary). One notices
2491 the need for C<dVAR> only with the said compile-time define, because
2492 otherwise the Perl global variables are visible as-is.
2494 =head2 Should I do anything special if I call perl from multiple threads?
2496 If you create interpreters in one thread and then proceed to call them in
2497 another, you need to make sure perl's own Thread Local Storage (TLS) slot is
2498 initialized correctly in each of those threads.
2500 The C<perl_alloc> and C<perl_clone> API functions will automatically set
2501 the TLS slot to the interpreter they created, so that there is no need to do
2502 anything special if the interpreter is always accessed in the same thread that
2503 created it, and that thread did not create or call any other interpreters
2504 afterwards. If that is not the case, you have to set the TLS slot of the
2505 thread before calling any functions in the Perl API on that particular
2506 interpreter. This is done by calling the C<PERL_SET_CONTEXT> macro in that
2507 thread as the first thing you do:
2509 /* do this before doing anything else with some_perl */
2510 PERL_SET_CONTEXT(some_perl);
2512 ... other Perl API calls on some_perl go here ...
2514 =head2 Future Plans and PERL_IMPLICIT_SYS
2516 Just as PERL_IMPLICIT_CONTEXT provides a way to bundle up everything
2517 that the interpreter knows about itself and pass it around, so too are
2518 there plans to allow the interpreter to bundle up everything it knows
2519 about the environment it's running on. This is enabled with the
2520 PERL_IMPLICIT_SYS macro. Currently it only works with USE_ITHREADS on
2523 This allows the ability to provide an extra pointer (called the "host"
2524 environment) for all the system calls. This makes it possible for
2525 all the system stuff to maintain their own state, broken down into
2526 seven C structures. These are thin wrappers around the usual system
2527 calls (see F<win32/perllib.c>) for the default perl executable, but for a
2528 more ambitious host (like the one that would do fork() emulation) all
2529 the extra work needed to pretend that different interpreters are
2530 actually different "processes", would be done here.
2532 The Perl engine/interpreter and the host are orthogonal entities.
2533 There could be one or more interpreters in a process, and one or
2534 more "hosts", with free association between them.
2536 =head1 Internal Functions
2538 All of Perl's internal functions which will be exposed to the outside
2539 world are prefixed by C<Perl_> so that they will not conflict with XS
2540 functions or functions used in a program in which Perl is embedded.
2541 Similarly, all global variables begin with C<PL_>. (By convention,
2542 static functions start with C<S_>.)
2544 Inside the Perl core (C<PERL_CORE> defined), you can get at the functions
2545 either with or without the C<Perl_> prefix, thanks to a bunch of defines
2546 that live in F<embed.h>. Note that extension code should I<not> set
2547 C<PERL_CORE>; this exposes the full perl internals, and is likely to cause
2548 breakage of the XS in each new perl release.
2550 The file F<embed.h> is generated automatically from
2551 F<embed.pl> and F<embed.fnc>. F<embed.pl> also creates the prototyping
2552 header files for the internal functions, generates the documentation
2553 and a lot of other bits and pieces. It's important that when you add
2554 a new function to the core or change an existing one, you change the
2555 data in the table in F<embed.fnc> as well. Here's a sample entry from
2558 Apd |SV** |av_fetch |AV* ar|I32 key|I32 lval
2560 The second column is the return type, the third column the name. Columns
2561 after that are the arguments. The first column is a set of flags:
2567 This function is a part of the public
2568 API. All such functions should also
2569 have 'd', very few do not.
2573 This function has a C<Perl_> prefix; i.e. it is defined as
2578 This function has documentation using the C<apidoc> feature which we'll
2579 look at in a second. Some functions have 'd' but not 'A'; docs are good.
2583 Other available flags are:
2589 This is a static function and is defined as C<STATIC S_whatever>, and
2590 usually called within the sources as C<whatever(...)>.
2594 This does not need an interpreter context, so the definition has no
2595 C<pTHX>, and it follows that callers don't use C<aTHX>. (See
2596 L</Background and PERL_IMPLICIT_CONTEXT>.)
2600 This function never returns; C<croak>, C<exit> and friends.
2604 This function takes a variable number of arguments, C<printf> style.
2605 The argument list should end with C<...>, like this:
2607 Afprd |void |croak |const char* pat|...
2611 This function is part of the experimental development API, and may change
2612 or disappear without notice.
2616 This function should not have a compatibility macro to define, say,
2617 C<Perl_parse> to C<parse>. It must be called as C<Perl_parse>.
2621 This function isn't exported out of the Perl core.
2625 This is implemented as a macro.
2629 This function is explicitly exported.
2633 This function is visible to extensions included in the Perl core.
2637 Binary backward compatibility; this function is a macro but also has
2638 a C<Perl_> implementation (which is exported).
2642 See the comments at the top of C<embed.fnc> for others.
2646 If you edit F<embed.pl> or F<embed.fnc>, you will need to run
2647 C<make regen_headers> to force a rebuild of F<embed.h> and other
2648 auto-generated files.
2650 =head2 Formatted Printing of IVs, UVs, and NVs
2652 If you are printing IVs, UVs, or NVS instead of the stdio(3) style
2653 formatting codes like C<%d>, C<%ld>, C<%f>, you should use the
2654 following macros for portability
2659 UVxf UV in hexadecimal
2664 These will take care of 64-bit integers and long doubles.
2667 printf("IV is %"IVdf"\n", iv);
2669 The IVdf will expand to whatever is the correct format for the IVs.
2671 Note that there are different "long doubles": Perl will use
2672 whatever the compiler has.
2674 If you are printing addresses of pointers, use UVxf combined
2675 with PTR2UV(), do not use %lx or %p.
2677 =head2 Pointer-To-Integer and Integer-To-Pointer
2679 Because pointer size does not necessarily equal integer size,
2680 use the follow macros to do it right.
2685 INT2PTR(pointertotype, integer)
2690 SV *sv = INT2PTR(SV*, iv);
2697 =head2 Exception Handling
2699 There are a couple of macros to do very basic exception handling in XS
2700 modules. You have to define C<NO_XSLOCKS> before including F<XSUB.h> to
2701 be able to use these macros:
2706 You can use these macros if you call code that may croak, but you need
2707 to do some cleanup before giving control back to Perl. For example:
2709 dXCPT; /* set up necessary variables */
2712 code_that_may_croak();
2717 /* do cleanup here */
2721 Note that you always have to rethrow an exception that has been
2722 caught. Using these macros, it is not possible to just catch the
2723 exception and ignore it. If you have to ignore the exception, you
2724 have to use the C<call_*> function.
2726 The advantage of using the above macros is that you don't have
2727 to setup an extra function for C<call_*>, and that using these
2728 macros is faster than using C<call_*>.
2730 =head2 Source Documentation
2732 There's an effort going on to document the internal functions and
2733 automatically produce reference manuals from them -- L<perlapi> is one
2734 such manual which details all the functions which are available to XS
2735 writers. L<perlintern> is the autogenerated manual for the functions
2736 which are not part of the API and are supposedly for internal use only.
2738 Source documentation is created by putting POD comments into the C
2742 =for apidoc sv_setiv
2744 Copies an integer into the given SV. Does not handle 'set' magic. See
2745 L<perlapi/sv_setiv_mg>.
2750 Please try and supply some documentation if you add functions to the
2753 =head2 Backwards compatibility
2755 The Perl API changes over time. New functions are
2756 added or the interfaces of existing functions are
2757 changed. The C<Devel::PPPort> module tries to
2758 provide compatibility code for some of these changes, so XS writers don't
2759 have to code it themselves when supporting multiple versions of Perl.
2761 C<Devel::PPPort> generates a C header file F<ppport.h> that can also
2762 be run as a Perl script. To generate F<ppport.h>, run:
2764 perl -MDevel::PPPort -eDevel::PPPort::WriteFile
2766 Besides checking existing XS code, the script can also be used to retrieve
2767 compatibility information for various API calls using the C<--api-info>
2768 command line switch. For example:
2770 % perl ppport.h --api-info=sv_magicext
2772 For details, see C<perldoc ppport.h>.
2774 =head1 Unicode Support
2776 Perl 5.6.0 introduced Unicode support. It's important for porters and XS
2777 writers to understand this support and make sure that the code they
2778 write does not corrupt Unicode data.
2780 =head2 What B<is> Unicode, anyway?
2782 In the olden, less enlightened times, we all used to use ASCII. Most of
2783 us did, anyway. The big problem with ASCII is that it's American. Well,
2784 no, that's not actually the problem; the problem is that it's not
2785 particularly useful for people who don't use the Roman alphabet. What
2786 used to happen was that particular languages would stick their own
2787 alphabet in the upper range of the sequence, between 128 and 255. Of
2788 course, we then ended up with plenty of variants that weren't quite
2789 ASCII, and the whole point of it being a standard was lost.
2791 Worse still, if you've got a language like Chinese or
2792 Japanese that has hundreds or thousands of characters, then you really
2793 can't fit them into a mere 256, so they had to forget about ASCII
2794 altogether, and build their own systems using pairs of numbers to refer
2797 To fix this, some people formed Unicode, Inc. and
2798 produced a new character set containing all the characters you can
2799 possibly think of and more. There are several ways of representing these
2800 characters, and the one Perl uses is called UTF-8. UTF-8 uses
2801 a variable number of bytes to represent a character. You can learn more
2802 about Unicode and Perl's Unicode model in L<perlunicode>.
2804 (On EBCDIC platforms, Perl uses instead UTF-EBCDIC, which is a form of
2805 UTF-8 adapted for EBCDIC platforms. Below, we just talk about UTF-8.
2806 UTF-EBCDIC is like UTF-8, but the details are different. The macros
2807 hide the differences from you, just remember that the particular numbers
2808 and bit patterns presented below will differ in UTF-EBCDIC.)
2810 =head2 How can I recognise a UTF-8 string?
2812 You can't. This is because UTF-8 data is stored in bytes just like
2813 non-UTF-8 data. The Unicode character 200, (C<0xC8> for you hex types)
2814 capital E with a grave accent, is represented by the two bytes
2815 C<v196.172>. Unfortunately, the non-Unicode string C<chr(196).chr(172)>
2816 has that byte sequence as well. So you can't tell just by looking -- this
2817 is what makes Unicode input an interesting problem.
2819 In general, you either have to know what you're dealing with, or you
2820 have to guess. The API function C<is_utf8_string> can help; it'll tell
2821 you if a string contains only valid UTF-8 characters, and the chances
2822 of a non-UTF-8 string looking like valid UTF-8 become very small very
2823 quickly with increasing string length. On a character-by-character
2824 basis, C<isUTF8_CHAR>
2825 will tell you whether the current character in a string is valid UTF-8.
2827 =head2 How does UTF-8 represent Unicode characters?
2829 As mentioned above, UTF-8 uses a variable number of bytes to store a
2830 character. Characters with values 0...127 are stored in one
2831 byte, just like good ol' ASCII. Character 128 is stored as
2832 C<v194.128>; this continues up to character 191, which is
2833 C<v194.191>. Now we've run out of bits (191 is binary
2834 C<10111111>) so we move on; character 192 is C<v195.128>. And
2835 so it goes on, moving to three bytes at character 2048.
2836 L<perlunicode/Unicode Encodings> has pictures of how this works.
2838 Assuming you know you're dealing with a UTF-8 string, you can find out
2839 how long the first character in it is with the C<UTF8SKIP> macro:
2841 char *utf = "\305\233\340\240\201";
2844 len = UTF8SKIP(utf); /* len is 2 here */
2846 len = UTF8SKIP(utf); /* len is 3 here */
2848 Another way to skip over characters in a UTF-8 string is to use
2849 C<utf8_hop>, which takes a string and a number of characters to skip
2850 over. You're on your own about bounds checking, though, so don't use it
2853 All bytes in a multi-byte UTF-8 character will have the high bit set,
2854 so you can test if you need to do something special with this
2855 character like this (the C<UTF8_IS_INVARIANT()> is a macro that tests
2856 whether the byte is encoded as a single byte even in UTF-8):
2859 U8 *utf_end; /* 1 beyond buffer pointed to by utf */
2860 UV uv; /* Note: a UV, not a U8, not a char */
2861 STRLEN len; /* length of character in bytes */
2863 if (!UTF8_IS_INVARIANT(*utf))
2864 /* Must treat this as UTF-8 */
2865 uv = utf8_to_uvchr_buf(utf, utf_end, &len);
2867 /* OK to treat this character as a byte */
2870 You can also see in that example that we use C<utf8_to_uvchr_buf> to get the
2871 value of the character; the inverse function C<uvchr_to_utf8> is available
2872 for putting a UV into UTF-8:
2874 if (!UVCHR_IS_INVARIANT(uv))
2875 /* Must treat this as UTF8 */
2876 utf8 = uvchr_to_utf8(utf8, uv);
2878 /* OK to treat this character as a byte */
2881 You B<must> convert characters to UVs using the above functions if
2882 you're ever in a situation where you have to match UTF-8 and non-UTF-8
2883 characters. You may not skip over UTF-8 characters in this case. If you
2884 do this, you'll lose the ability to match hi-bit non-UTF-8 characters;
2885 for instance, if your UTF-8 string contains C<v196.172>, and you skip
2886 that character, you can never match a C<chr(200)> in a non-UTF-8 string.
2889 (Note that we don't have to test for invariant characters in the
2890 examples above. The functions work on any well-formed UTF-8 input.
2891 It's just that its faster to avoid the function overhead when it's not
2894 =head2 How does Perl store UTF-8 strings?
2896 Currently, Perl deals with UTF-8 strings and non-UTF-8 strings
2897 slightly differently. A flag in the SV, C<SVf_UTF8>, indicates that the
2898 string is internally encoded as UTF-8. Without it, the byte value is the
2899 codepoint number and vice versa. This flag is only meaningful if the SV
2900 is C<SvPOK> or immediately after stringification via C<SvPV> or a
2901 similar macro. You can check and manipulate this flag with the
2908 This flag has an important effect on Perl's treatment of the string: if
2909 UTF-8 data is not properly distinguished, regular expressions,
2910 C<length>, C<substr> and other string handling operations will have
2911 undesirable (wrong) results.
2913 The problem comes when you have, for instance, a string that isn't
2914 flagged as UTF-8, and contains a byte sequence that could be UTF-8 --
2915 especially when combining non-UTF-8 and UTF-8 strings.
2917 Never forget that the C<SVf_UTF8> flag is separate from the PV value; you
2918 need to be sure you don't accidentally knock it off while you're
2919 manipulating SVs. More specifically, you cannot expect to do this:
2928 nsv = newSVpvn(p, len);
2930 The C<char*> string does not tell you the whole story, and you can't
2931 copy or reconstruct an SV just by copying the string value. Check if the
2932 old SV has the UTF8 flag set (I<after> the C<SvPV> call), and act
2936 is_utf8 = SvUTF8(sv);
2937 frobnicate(p, is_utf8);
2938 nsv = newSVpvn(p, len);
2942 In the above, your C<frobnicate> function has been changed to be made
2943 aware of whether or not it's dealing with UTF-8 data, so that it can
2944 handle the string appropriately.
2946 Since just passing an SV to an XS function and copying the data of
2947 the SV is not enough to copy the UTF8 flags, even less right is just
2948 passing a S<C<char *>> to an XS function.
2950 For full generality, use the L<C<DO_UTF8>|perlapi/DO_UTF8> macro to see if the
2951 string in an SV is to be I<treated> as UTF-8. This takes into account
2952 if the call to the XS function is being made from within the scope of
2953 L<S<C<use bytes>>|bytes>. If so, the underlying bytes that comprise the
2954 UTF-8 string are to be exposed, rather than the character they
2955 represent. But this pragma should only really be used for debugging and
2956 perhaps low-level testing at the byte level. Hence most XS code need
2957 not concern itself with this, but various areas of the perl core do need
2960 And this isn't the whole story. Starting in Perl v5.12, strings that
2961 aren't encoded in UTF-8 may also be treated as Unicode under various
2962 conditions (see L<perlunicode/ASCII Rules versus Unicode Rules>).
2963 This is only really a problem for characters whose ordinals are between
2964 128 and 255, and their behavior varies under ASCII versus Unicode rules
2965 in ways that your code cares about (see L<perlunicode/The "Unicode Bug">).
2966 There is no published API for dealing with this, as it is subject to
2967 change, but you can look at the code for C<pp_lc> in F<pp.c> for an
2968 example as to how it's currently done.
2970 =head2 How do I convert a string to UTF-8?
2972 If you're mixing UTF-8 and non-UTF-8 strings, it is necessary to upgrade
2973 the non-UTF-8 strings to UTF-8. If you've got an SV, the easiest way to do
2976 sv_utf8_upgrade(sv);
2978 However, you must not do this, for example:
2981 sv_utf8_upgrade(left);
2983 If you do this in a binary operator, you will actually change one of the
2984 strings that came into the operator, and, while it shouldn't be noticeable
2985 by the end user, it can cause problems in deficient code.
2987 Instead, C<bytes_to_utf8> will give you a UTF-8-encoded B<copy> of its
2988 string argument. This is useful for having the data available for
2989 comparisons and so on, without harming the original SV. There's also
2990 C<utf8_to_bytes> to go the other way, but naturally, this will fail if
2991 the string contains any characters above 255 that can't be represented
2994 =head2 How do I compare strings?
2996 L<perlapi/sv_cmp> and L<perlapi/sv_cmp_flags> do a lexigraphic
2997 comparison of two SV's, and handle UTF-8ness properly. Note, however,
2998 that Unicode specifies a much fancier mechanism for collation, available
2999 via the L<Unicode::Collate> module.
3001 To just compare two strings for equality/non-equality, you can just use
3002 L<C<memEQ()>|perlapi/memEQ> and L<C<memNE()>|perlapi/memEQ> as usual,
3003 except the strings must be both UTF-8 or not UTF-8 encoded.
3005 To compare two strings case-insensitively, use
3006 L<C<foldEQ_utf8()>|perlapi/foldEQ_utf8> (the strings don't have to have
3007 the same UTF-8ness).
3009 =head2 Is there anything else I need to know?
3011 Not really. Just remember these things:
3017 There's no way to tell if a S<C<char *>> or S<C<U8 *>> string is UTF-8
3018 or not. But you can tell if an SV is to be treated as UTF-8 by calling
3019 C<DO_UTF8> on it, after stringifying it with C<SvPV> or a similar
3020 macro. And, you can tell if SV is actually UTF-8 (even if it is not to
3021 be treated as such) by looking at its C<SvUTF8> flag (again after
3022 stringifying it). Don't forget to set the flag if something should be
3024 Treat the flag as part of the PV, even though it's not -- if you pass on
3025 the PV to somewhere, pass on the flag too.
3029 If a string is UTF-8, B<always> use C<utf8_to_uvchr_buf> to get at the value,
3030 unless C<UTF8_IS_INVARIANT(*s)> in which case you can use C<*s>.
3034 When writing a character UV to a UTF-8 string, B<always> use
3035 C<uvchr_to_utf8>, unless C<UVCHR_IS_INVARIANT(uv))> in which case
3036 you can use C<*s = uv>.
3040 Mixing UTF-8 and non-UTF-8 strings is
3041 tricky. Use C<bytes_to_utf8> to get
3042 a new string which is UTF-8 encoded, and then combine them.
3046 =head1 Custom Operators
3048 Custom operator support is an experimental feature that allows you to
3049 define your own ops. This is primarily to allow the building of
3050 interpreters for other languages in the Perl core, but it also allows
3051 optimizations through the creation of "macro-ops" (ops which perform the
3052 functions of multiple ops which are usually executed together, such as
3053 C<gvsv, gvsv, add>.)
3055 This feature is implemented as a new op type, C<OP_CUSTOM>. The Perl
3056 core does not "know" anything special about this op type, and so it will
3057 not be involved in any optimizations. This also means that you can
3058 define your custom ops to be any op structure -- unary, binary, list and
3061 It's important to know what custom operators won't do for you. They
3062 won't let you add new syntax to Perl, directly. They won't even let you
3063 add new keywords, directly. In fact, they won't change the way Perl
3064 compiles a program at all. You have to do those changes yourself, after
3065 Perl has compiled the program. You do this either by manipulating the op
3066 tree using a C<CHECK> block and the C<B::Generate> module, or by adding
3067 a custom peephole optimizer with the C<optimize> module.
3069 When you do this, you replace ordinary Perl ops with custom ops by
3070 creating ops with the type C<OP_CUSTOM> and the C<op_ppaddr> of your own
3071 PP function. This should be defined in XS code, and should look like
3072 the PP ops in C<pp_*.c>. You are responsible for ensuring that your op
3073 takes the appropriate number of values from the stack, and you are
3074 responsible for adding stack marks if necessary.
3076 You should also "register" your op with the Perl interpreter so that it
3077 can produce sensible error and warning messages. Since it is possible to
3078 have multiple custom ops within the one "logical" op type C<OP_CUSTOM>,
3079 Perl uses the value of C<< o->op_ppaddr >> to determine which custom op
3080 it is dealing with. You should create an C<XOP> structure for each
3081 ppaddr you use, set the properties of the custom op with
3082 C<XopENTRY_set>, and register the structure against the ppaddr using
3083 C<Perl_custom_op_register>. A trivial example might look like:
3086 static OP *my_pp(pTHX);
3089 XopENTRY_set(&my_xop, xop_name, "myxop");
3090 XopENTRY_set(&my_xop, xop_desc, "Useless custom op");
3091 Perl_custom_op_register(aTHX_ my_pp, &my_xop);
3093 The available fields in the structure are:
3099 A short name for your op. This will be included in some error messages,
3100 and will also be returned as C<< $op->name >> by the L<B|B> module, so
3101 it will appear in the output of module like L<B::Concise|B::Concise>.
3105 A short description of the function of the op.
3109 Which of the various C<*OP> structures this op uses. This should be one of
3110 the C<OA_*> constants from F<op.h>, namely
3130 =item OA_PVOP_OR_SVOP
3132 This should be interpreted as 'C<PVOP>' only. The C<_OR_SVOP> is because
3133 the only core C<PVOP>, C<OP_TRANS>, can sometimes be a C<SVOP> instead.
3141 The other C<OA_*> constants should not be used.
3145 This member is of type C<Perl_cpeep_t>, which expands to C<void
3146 (*Perl_cpeep_t)(aTHX_ OP *o, OP *oldop)>. If it is set, this function
3147 will be called from C<Perl_rpeep> when ops of this type are encountered
3148 by the peephole optimizer. I<o> is the OP that needs optimizing;
3149 I<oldop> is the previous OP optimized, whose C<op_next> points to I<o>.
3153 C<B::Generate> directly supports the creation of custom ops by name.
3156 =head1 Dynamic Scope and the Context Stack
3158 B<Note:> this section describes a non-public internal API that is subject
3159 to change without notice.
3161 =head2 Introduction to the context stack
3163 In Perl, dynamic scoping refers to the runtime nesting of things like
3164 subroutine calls, evals etc, as well as the entering and exiting of block
3165 scopes. For example, the restoring of a C<local>ised variable is
3166 determined by the dynamic scope.
3168 Perl tracks the dynamic scope by a data structure called the context
3169 stack, which is an array of C<PERL_CONTEXT> structures, and which is
3170 itself a big union for all the types of context. Whenever a new scope is
3171 entered (such as a block, a C<for> loop, or a subroutine call), a new
3172 context entry is pushed onto the stack. Similarly when leaving a block or
3173 returning from a subroutine call etc. a context is popped. Since the
3174 context stack represents the current dynamic scope, it can be searched.
3175 For example, C<next LABEL> searches back through the stack looking for a
3176 loop context that matches the label; C<return> pops contexts until it
3177 finds a sub or eval context or similar; C<caller> examines sub contexts on
3180 Each context entry is labelled with a context type, C<cx_type>. Typical
3181 context types are C<CXt_SUB>, C<CXt_EVAL> etc., as well as C<CXt_BLOCK>
3182 and C<CXt_NULL> which represent a basic scope (as pushed by C<pp_enter>)
3183 and a sort block. The type determines which part of the context union are
3186 The main division in the context struct is between a substitution scope
3187 (C<CXt_SUBST>) and block scopes, which are everything else. The former is
3188 just used while executing C<s///e>, and won't be discussed further
3191 All the block scope types share a common base, which corresponds to
3192 C<CXt_BLOCK>. This stores the old values of various scope-related
3193 variables like C<PL_curpm>, as well as information about the current
3194 scope, such as C<gimme>. On scope exit, the old variables are restored.
3196 Particular block scope types store extra per-type information. For
3197 example, C<CXt_SUB> stores the currently executing CV, while the various
3198 for loop types might hold the original loop variable SV. On scope exit,
3199 the per-type data is processed; for example the CV has its reference count
3200 decremented, and the original loop variable is restored.
3202 The macro C<cxstack> returns the base of the current context stack, while
3203 C<cxstack_ix> is the index of the current frame within that stack.
3205 In fact, the context stack is actually part of a stack-of-stacks system;
3206 whenever something unusual is done such as calling a C<DESTROY> or tie
3207 handler, a new stack is pushed, then popped at the end.
3209 Note that the API described here changed considerably in perl 5.24; prior
3210 to that, big macros like C<PUSHBLOCK> and C<POPSUB> were used; in 5.24
3211 they were replaced by the inline static functions described below. In
3212 addition, the ordering and detail of how these macros/function work
3213 changed in many ways, often subtly. In particular they didn't handle
3214 saving the savestack and temps stack positions, and required additional
3215 C<ENTER>, C<SAVETMPS> and C<LEAVE> compared to the new functions. The
3216 old-style macros will not be described further.
3219 =head2 Pushing contexts
3221 For pushing a new context, the two basic functions are
3222 C<cx = cx_pushblock()>, which pushes a new basic context block and returns
3223 its address, and a family of similar functions with names like
3224 C<cx_pushsub(cx)> which populate the additional type-dependent fields in
3225 the C<cx> struct. Note that C<CXt_NULL> and C<CXt_BLOCK> don't have their
3226 own push functions, as they don't store any data beyond that pushed by
3229 The fields of the context struct and the arguments to the C<cx_*>
3230 functions are subject to change between perl releases, representing
3231 whatever is convenient or efficient for that release.
3233 A typical context stack pushing can be found in C<pp_entersub>; the
3234 following shows a simplified and stripped-down example of a non-XS call,
3235 along with comments showing roughly what each function does.
3239 bool hasargs = cBOOL(PL_op->op_flags & OPf_STACKED);
3240 OP *retop = PL_op->op_next;
3241 I32 old_ss_ix = PL_savestack_ix;
3244 /* ... make mortal copies of stack args which are PADTMPs here ... */
3246 /* ... do any additional savestack pushes here ... */
3248 /* Now push a new context entry of type 'CXt_SUB'; initially just
3249 * doing the actions common to all block types: */
3251 cx = cx_pushblock(CXt_SUB, gimme, MARK, old_ss_ix);
3253 /* this does (approximately):
3254 CXINC; /* cxstack_ix++ (grow if necessary) */
3255 cx = CX_CUR(); /* and get the address of new frame */
3256 cx->cx_type = CXt_SUB;
3257 cx->blk_gimme = gimme;
3258 cx->blk_oldsp = MARK - PL_stack_base;
3259 cx->blk_oldsaveix = old_ss_ix;
3260 cx->blk_oldcop = PL_curcop;
3261 cx->blk_oldmarksp = PL_markstack_ptr - PL_markstack;
3262 cx->blk_oldscopesp = PL_scopestack_ix;
3263 cx->blk_oldpm = PL_curpm;
3264 cx->blk_old_tmpsfloor = PL_tmps_floor;
3266 PL_tmps_floor = PL_tmps_ix;
3270 /* then update the new context frame with subroutine-specific info,
3271 * such as the CV about to be executed: */
3273 cx_pushsub(cx, cv, retop, hasargs);
3275 /* this does (approximately):
3276 cx->blk_sub.cv = cv;
3277 cx->blk_sub.olddepth = CvDEPTH(cv);
3278 cx->blk_sub.prevcomppad = PL_comppad;
3279 cx->cx_type |= (hasargs) ? CXp_HASARGS : 0;
3280 cx->blk_sub.retop = retop;
3281 SvREFCNT_inc_simple_void_NN(cv);
3284 Note that C<cx_pushblock()> sets two new floors: for the args stack (to
3285 C<MARK>) and the temps stack (to C<PL_tmps_ix>). While executing at this
3286 scope level, every C<nextstate> (amongst others) will reset the args and
3287 tmps stack levels to these floors. Note that since C<cx_pushblock> uses
3288 the current value of C<PL_tmps_ix> rather than it being passed as an arg,
3289 this dictates at what point C<cx_pushblock> should be called. In
3290 particular, any new mortals which should be freed only on scope exit
3291 (rather than at the next C<nextstate>) should be created first.
3293 Most callers of C<cx_pushblock> simply set the new args stack floor to the
3294 top of the previous stack frame, but for C<CXt_LOOP_LIST> it stores the
3295 items being iterated over on the stack, and so sets C<blk_oldsp> to the
3296 top of these items instead. Note that, contrary to its name, C<blk_oldsp>
3297 doesn't always represent the value to restore C<PL_stack_sp> to on scope
3300 Note the early capture of C<PL_savestack_ix> to C<old_ss_ix>, which is
3301 later passed as an arg to C<cx_pushblock>. In the case of C<pp_entersub>,
3302 this is because, although most values needing saving are stored in fields
3303 of the context struct, an extra value needs saving only when the debugger
3304 is running, and it doesn't make sense to bloat the struct for this rare
3305 case. So instead it is saved on the savestack. Since this value gets
3306 calculated and saved before the context is pushed, it is necessary to pass
3307 the old value of C<PL_savestack_ix> to C<cx_pushblock>, to ensure that the
3308 saved value gets freed during scope exit. For most users of
3309 C<cx_pushblock>, where nothing needs pushing on the save stack,
3310 C<PL_savestack_ix> is just passed directly as an arg to C<cx_pushblock>.
3312 Note that where possible, values should be saved in the context struct
3313 rather than on the save stack; it's much faster that way.
3315 Normally C<cx_pushblock> should be immediately followed by the appropriate
3316 C<cx_pushfoo>, with nothing between them; this is because if code
3317 in-between could die (e.g. a warning upgraded to fatal), then the context
3318 stack unwinding code in C<dounwind> would see (in the example above) a
3319 C<CXt_SUB> context frame, but without all the subroutine-specific fields
3320 set, and crashes would soon ensue.
3322 Where the two must be separate, initially set the type to C<CXt_NULL> or
3323 C<CXt_BLOCK>, and later change it to C<CXt_foo> when doing the
3324 C<cx_pushfoo>. This is exactly what C<pp_enteriter> does, once it's
3325 determined which type of loop it's pushing.
3327 =head2 Popping contexts
3329 Contexts are popped using C<cx_popsub()> etc. and C<cx_popblock()>. Note
3330 however, that unlike C<cx_pushblock>, neither of these functions actually
3331 decrement the current context stack index; this is done separately using
3334 There are two main ways that contexts are popped. During normal execution
3335 as scopes are exited, functions like C<pp_leave>, C<pp_leaveloop> and
3336 C<pp_leavesub> process and pop just one context using C<cx_popfoo> and
3337 C<cx_popblock>. On the other hand, things like C<pp_return> and C<next>
3338 may have to pop back several scopes until a sub or loop context is found,
3339 and exceptions (such as C<die>) need to pop back contexts until an eval
3340 context is found. Both of these are accomplished by C<dounwind()>, which
3341 is capable of processing and popping all contexts above the target one.
3343 Here is a typical example of context popping, as found in C<pp_leavesub>
3344 (simplified slightly):
3353 gimme = cx->blk_gimme;
3354 oldsp = PL_stack_base + cx->blk_oldsp; /* last arg of previous frame */
3356 if (gimme == G_VOID)
3357 PL_stack_sp = oldsp;
3359 leave_adjust_stacks(oldsp, oldsp, gimme, 0);
3364 retop = cx->blk_sub.retop;
3369 The steps above are in a very specific order, designed to be the reverse
3370 order of when the context was pushed. The first thing to do is to copy
3371 and/or protect any any return arguments and free any temps in the current
3372 scope. Scope exits like an rvalue sub normally return a mortal copy of
3373 their return args (as opposed to lvalue subs). It is important to make
3374 this copy before the save stack is popped or variables are restored, or
3375 bad things like the following can happen:
3377 sub f { my $x =...; $x } # $x freed before we get to copy it
3378 sub f { /(...)/; $1 } # PL_curpm restored before $1 copied
3380 Although we wish to free any temps at the same time, we have to be careful
3381 not to free any temps which are keeping return args alive; nor to free the
3382 temps we have just created while mortal copying return args. Fortunately,
3383 C<leave_adjust_stacks()> is capable of making mortal copies of return args,
3384 shifting args down the stack, and only processing those entries on the
3385 temps stack that are safe to do so.
3387 In void context no args are returned, so it's more efficient to skip
3388 calling C<leave_adjust_stacks()>. Also in void context, a C<nextstate> op
3389 is likely to be imminently called which will do a C<FREETMPS>, so there's
3390 no need to do that either.
3392 The next step is to pop savestack entries: C<CX_LEAVE_SCOPE(cx)> is just
3393 defined as C<<LEAVE_SCOPE(cx->blk_oldsaveix)>>. Note that during the
3394 popping, it's possible for perl to call destructors, call C<STORE> to undo
3395 localisations of tied vars, and so on. Any of these can die or call
3396 C<exit()>. In this case, C<dounwind()> will be called, and the current
3397 context stack frame will be re-processed. Thus it is vital that all steps
3398 in popping a context are done in such a way to support reentrancy. The
3399 other alternative, of decrementing C<cxstack_ix> I<before> processing the
3400 frame, would lead to leaks and the like if something died halfway through,
3401 or overwriting of the current frame.
3403 C<CX_LEAVE_SCOPE> itself is safely re-entrant: if only half the savestack
3404 items have been popped before dying and getting trapped by eval, then the
3405 C<CX_LEAVE_SCOPE>s in C<dounwind> or C<pp_leaveeval> will continue where
3406 the first one left off.
3408 The next step is the type-specific context processing; in this case
3409 C<cx_popsub>. In part, this looks like:
3411 cv = cx->blk_sub.cv;
3412 CvDEPTH(cv) = cx->blk_sub.olddepth;
3413 cx->blk_sub.cv = NULL;
3416 where its processing the just-executed CV. Note that before it decrements
3417 the CV's reference count, it nulls the C<blk_sub.cv>. This means that if
3418 it re-enters, the CV won't be freed twice. It also means that you can't
3419 rely on such type-specific fields having useful values after the return
3422 Next, C<cx_popblock> restores all the various interpreter vars to their
3423 previous values or previous high water marks; it expands to:
3425 PL_markstack_ptr = PL_markstack + cx->blk_oldmarksp;
3426 PL_scopestack_ix = cx->blk_oldscopesp;
3427 PL_curpm = cx->blk_oldpm;
3428 PL_curcop = cx->blk_oldcop;
3429 PL_tmps_floor = cx->blk_old_tmpsfloor;
3431 Note that it I<doesn't> restore C<PL_stack_sp>; as mentioned earlier,
3432 which value to restore it to depends on the context type (specifically
3433 C<for (list) {}>), and what args (if any) it returns; and that will
3434 already have been sorted out earlier by C<leave_adjust_stacks()>.
3436 Finally, the context stack pointer is actually decremented by C<CX_POP(cx)>.
3437 After this point, it's possible that that the current context frame could
3438 be overwritten by other contexts being pushed. Although things like ties
3439 and C<DESTROY> are supposed to work within a new context stack, it's best
3440 not to assume this. Indeed on debugging builds, C<CX_POP(cx)> deliberately
3441 sets C<cx> to null to detect code that is still relying on the field
3442 values in that context frame. Note in the C<pp_leavesub()> example above,
3443 we grab C<blk_sub.retop> I<before> calling C<CX_POP>.
3445 =head2 Redoing contexts
3447 Finally, there is C<cx_topblock(cx)>, which acts like a super-C<nextstate>
3448 as regards to resetting various vars to their base values. It is used in
3449 places like C<pp_next>, C<pp_redo> and C<pp_goto> where rather than
3450 exiting a scope, we want to re-initialise the scope. As well as resetting
3451 C<PL_stack_sp> like C<nextstate>, it also resets C<PL_markstack_ptr>,
3452 C<PL_scopestack_ix> and C<PL_curpm>. Note that it doesn't do a
3458 Until May 1997, this document was maintained by Jeff Okamoto
3459 E<lt>okamoto@corp.hp.comE<gt>. It is now maintained as part of Perl
3460 itself by the Perl 5 Porters E<lt>perl5-porters@perl.orgE<gt>.
3462 With lots of help and suggestions from Dean Roehrich, Malcolm Beattie,
3463 Andreas Koenig, Paul Hudson, Ilya Zakharevich, Paul Marquess, Neil
3464 Bowers, Matthew Green, Tim Bunce, Spider Boardman, Ulrich Pfeifer,
3465 Stephen McCamant, and Gurusamy Sarathy.
3469 L<perlapi>, L<perlintern>, L<perlxs>, L<perlembed>
https://perl5.git.perl.org / perl5.git / blob