This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Fix test for perl version number in Makefile.PL
[perl5.git] / sv.c
CommitLineData
a0d0e21e 1/* sv.c
79072805 2 *
4bb101f2 3 * Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
241d1a3b 4 * 2000, 2001, 2002, 2003, 2004, 2005, by Larry Wall and others
79072805
LW
5 *
6 * You may distribute under the terms of either the GNU General Public
7 * License or the Artistic License, as specified in the README file.
8 *
a0d0e21e 9 * "I wonder what the Entish is for 'yes' and 'no'," he thought.
645c22ef
DM
10 *
11 *
5e045b90
AMS
12 * This file contains the code that creates, manipulates and destroys
13 * scalar values (SVs). The other types (AV, HV, GV, etc.) reuse the
14 * structure of an SV, so their creation and destruction is handled
15 * here; higher-level functions are in av.c, hv.c, and so on. Opcode
16 * level functions (eg. substr, split, join) for each of the types are
17 * in the pp*.c files.
79072805
LW
18 */
19
20#include "EXTERN.h"
864dbfa3 21#define PERL_IN_SV_C
79072805 22#include "perl.h"
d2f185dc 23#include "regcomp.h"
79072805 24
51371543 25#define FCALL *f
2c5424a7 26
2f8ed50e
OS
27#ifdef __Lynx__
28/* Missing proto on LynxOS */
29 char *gconvert(double, int, int, char *);
30#endif
31
e23c8137
JH
32#ifdef PERL_UTF8_CACHE_ASSERT
33/* The cache element 0 is the Unicode offset;
34 * the cache element 1 is the byte offset of the element 0;
35 * the cache element 2 is the Unicode length of the substring;
36 * the cache element 3 is the byte length of the substring;
37 * The checking of the substring side would be good
38 * but substr() has enough code paths to make my head spin;
39 * if adding more checks watch out for the following tests:
40 * t/op/index.t t/op/length.t t/op/pat.t t/op/substr.t
41 * lib/utf8.t lib/Unicode/Collate/t/index.t
42 * --jhi
43 */
44#define ASSERT_UTF8_CACHE(cache) \
45 STMT_START { if (cache) { assert((cache)[0] <= (cache)[1]); } } STMT_END
46#else
47#define ASSERT_UTF8_CACHE(cache) NOOP
48#endif
49
f8c7b90f 50#ifdef PERL_OLD_COPY_ON_WRITE
765f542d 51#define SV_COW_NEXT_SV(sv) INT2PTR(SV *,SvUVX(sv))
607fa7f2 52#define SV_COW_NEXT_SV_SET(current,next) SvUV_set(current, PTR2UV(next))
b5ccf5f2 53/* This is a pessimistic view. Scalar must be purely a read-write PV to copy-
765f542d 54 on-write. */
765f542d 55#endif
645c22ef
DM
56
57/* ============================================================================
58
59=head1 Allocation and deallocation of SVs.
60
5e045b90
AMS
61An SV (or AV, HV, etc.) is allocated in two parts: the head (struct sv,
62av, hv...) contains type and reference count information, as well as a
63pointer to the body (struct xrv, xpv, xpviv...), which contains fields
64specific to each type.
65
93e68bfb
JC
66In all but the most memory-paranoid configuations (ex: PURIFY), this
67allocation is done using arenas, which by default are approximately 4K
68chunks of memory parcelled up into N heads or bodies (of same size).
69Sv-bodies are allocated by their sv-type, guaranteeing size
70consistency needed to allocate safely from arrays.
71
72The first slot in each arena is reserved, and is used to hold a link
73to the next arena. In the case of heads, the unused first slot also
74contains some flags and a note of the number of slots. Snaked through
75each arena chain is a linked list of free items; when this becomes
76empty, an extra arena is allocated and divided up into N items which
77are threaded into the free list.
645c22ef
DM
78
79The following global variables are associated with arenas:
80
81 PL_sv_arenaroot pointer to list of SV arenas
82 PL_sv_root pointer to list of free SV structures
83
93e68bfb
JC
84 PL_body_arenaroots[] array of pointers to list of arenas, 1 per svtype
85 PL_body_roots[] array of pointers to list of free bodies of svtype
86 arrays are indexed by the svtype needed
645c22ef 87
93e68bfb
JC
88Note that some of the larger and more rarely used body types (eg
89xpvio) are not allocated using arenas, but are instead just
90malloc()/free()ed as required.
91
92In addition, a few SV heads are not allocated from an arena, but are
93instead directly created as static or auto variables, eg PL_sv_undef.
94The size of arenas can be changed from the default by setting
95PERL_ARENA_SIZE appropriately at compile time.
645c22ef
DM
96
97The SV arena serves the secondary purpose of allowing still-live SVs
98to be located and destroyed during final cleanup.
99
100At the lowest level, the macros new_SV() and del_SV() grab and free
101an SV head. (If debugging with -DD, del_SV() calls the function S_del_sv()
102to return the SV to the free list with error checking.) new_SV() calls
103more_sv() / sv_add_arena() to add an extra arena if the free list is empty.
104SVs in the free list have their SvTYPE field set to all ones.
105
106Similarly, there are macros new_XIV()/del_XIV(), new_XNV()/del_XNV() etc
107that allocate and return individual body types. Normally these are mapped
ff276b08
RG
108to the arena-manipulating functions new_xiv()/del_xiv() etc, but may be
109instead mapped directly to malloc()/free() if PURIFY is defined. The
645c22ef
DM
110new/del functions remove from, or add to, the appropriate PL_foo_root
111list, and call more_xiv() etc to add a new arena if the list is empty.
112
ff276b08 113At the time of very final cleanup, sv_free_arenas() is called from
645c22ef 114perl_destruct() to physically free all the arenas allocated since the
6a93a7e5 115start of the interpreter.
645c22ef
DM
116
117Manipulation of any of the PL_*root pointers is protected by enclosing
118LOCK_SV_MUTEX; ... UNLOCK_SV_MUTEX calls which should Do the Right Thing
119if threads are enabled.
120
121The function visit() scans the SV arenas list, and calls a specified
122function for each SV it finds which is still live - ie which has an SvTYPE
123other than all 1's, and a non-zero SvREFCNT. visit() is used by the
124following functions (specified as [function that calls visit()] / [function
125called by visit() for each SV]):
126
127 sv_report_used() / do_report_used()
128 dump all remaining SVs (debugging aid)
129
130 sv_clean_objs() / do_clean_objs(),do_clean_named_objs()
131 Attempt to free all objects pointed to by RVs,
132 and, unless DISABLE_DESTRUCTOR_KLUDGE is defined,
133 try to do the same for all objects indirectly
134 referenced by typeglobs too. Called once from
135 perl_destruct(), prior to calling sv_clean_all()
136 below.
137
138 sv_clean_all() / do_clean_all()
139 SvREFCNT_dec(sv) each remaining SV, possibly
140 triggering an sv_free(). It also sets the
141 SVf_BREAK flag on the SV to indicate that the
142 refcnt has been artificially lowered, and thus
143 stopping sv_free() from giving spurious warnings
144 about SVs which unexpectedly have a refcnt
145 of zero. called repeatedly from perl_destruct()
146 until there are no SVs left.
147
93e68bfb 148=head2 Arena allocator API Summary
645c22ef
DM
149
150Private API to rest of sv.c
151
152 new_SV(), del_SV(),
153
154 new_XIV(), del_XIV(),
155 new_XNV(), del_XNV(),
156 etc
157
158Public API:
159
8cf8f3d1 160 sv_report_used(), sv_clean_objs(), sv_clean_all(), sv_free_arenas()
645c22ef
DM
161
162
163=cut
164
165============================================================================ */
166
167
51371543 168
4561caa4
CS
169/*
170 * "A time to plant, and a time to uproot what was planted..."
171 */
172
77354fb4
NC
173/*
174 * nice_chunk and nice_chunk size need to be set
175 * and queried under the protection of sv_mutex
176 */
177void
178Perl_offer_nice_chunk(pTHX_ void *chunk, U32 chunk_size)
179{
180 void *new_chunk;
181 U32 new_chunk_size;
182 LOCK_SV_MUTEX;
183 new_chunk = (void *)(chunk);
184 new_chunk_size = (chunk_size);
185 if (new_chunk_size > PL_nice_chunk_size) {
186 Safefree(PL_nice_chunk);
187 PL_nice_chunk = (char *) new_chunk;
188 PL_nice_chunk_size = new_chunk_size;
189 } else {
190 Safefree(chunk);
191 }
192 UNLOCK_SV_MUTEX;
193}
cac9b346 194
fd0854ff 195#ifdef DEBUG_LEAKING_SCALARS
22162ca8 196# define FREE_SV_DEBUG_FILE(sv) Safefree((sv)->sv_debug_file)
fd0854ff
DM
197#else
198# define FREE_SV_DEBUG_FILE(sv)
199#endif
200
48614a46
NC
201#ifdef PERL_POISON
202# define SvARENA_CHAIN(sv) ((sv)->sv_u.svu_rv)
203/* Whilst I'd love to do this, it seems that things like to check on
204 unreferenced scalars
205# define POSION_SV_HEAD(sv) Poison(sv, 1, struct STRUCT_SV)
206*/
207# define POSION_SV_HEAD(sv) Poison(&SvANY(sv), 1, void *), \
208 Poison(&SvREFCNT(sv), 1, U32)
209#else
210# define SvARENA_CHAIN(sv) SvANY(sv)
211# define POSION_SV_HEAD(sv)
212#endif
213
053fc874
GS
214#define plant_SV(p) \
215 STMT_START { \
fd0854ff 216 FREE_SV_DEBUG_FILE(p); \
48614a46
NC
217 POSION_SV_HEAD(p); \
218 SvARENA_CHAIN(p) = (void *)PL_sv_root; \
053fc874
GS
219 SvFLAGS(p) = SVTYPEMASK; \
220 PL_sv_root = (p); \
221 --PL_sv_count; \
222 } STMT_END
a0d0e21e 223
fba3b22e 224/* sv_mutex must be held while calling uproot_SV() */
053fc874
GS
225#define uproot_SV(p) \
226 STMT_START { \
227 (p) = PL_sv_root; \
48614a46 228 PL_sv_root = (SV*)SvARENA_CHAIN(p); \
053fc874
GS
229 ++PL_sv_count; \
230 } STMT_END
231
645c22ef 232
cac9b346
NC
233/* make some more SVs by adding another arena */
234
235/* sv_mutex must be held while calling more_sv() */
236STATIC SV*
237S_more_sv(pTHX)
238{
239 SV* sv;
240
241 if (PL_nice_chunk) {
242 sv_add_arena(PL_nice_chunk, PL_nice_chunk_size, 0);
243 PL_nice_chunk = Nullch;
244 PL_nice_chunk_size = 0;
245 }
246 else {
247 char *chunk; /* must use New here to match call to */
a02a5408 248 Newx(chunk,PERL_ARENA_SIZE,char); /* Safefree() in sv_free_arenas() */
2e7ed132 249 sv_add_arena(chunk, PERL_ARENA_SIZE, 0);
cac9b346
NC
250 }
251 uproot_SV(sv);
252 return sv;
253}
254
645c22ef
DM
255/* new_SV(): return a new, empty SV head */
256
eba0f806
DM
257#ifdef DEBUG_LEAKING_SCALARS
258/* provide a real function for a debugger to play with */
259STATIC SV*
260S_new_SV(pTHX)
261{
262 SV* sv;
263
264 LOCK_SV_MUTEX;
265 if (PL_sv_root)
266 uproot_SV(sv);
267 else
cac9b346 268 sv = S_more_sv(aTHX);
eba0f806
DM
269 UNLOCK_SV_MUTEX;
270 SvANY(sv) = 0;
271 SvREFCNT(sv) = 1;
272 SvFLAGS(sv) = 0;
fd0854ff
DM
273 sv->sv_debug_optype = PL_op ? PL_op->op_type : 0;
274 sv->sv_debug_line = (U16) ((PL_copline == NOLINE) ?
275 (PL_curcop ? CopLINE(PL_curcop) : 0) : PL_copline);
276 sv->sv_debug_inpad = 0;
277 sv->sv_debug_cloned = 0;
fd0854ff 278 sv->sv_debug_file = PL_curcop ? savepv(CopFILE(PL_curcop)): NULL;
fd0854ff 279
eba0f806
DM
280 return sv;
281}
282# define new_SV(p) (p)=S_new_SV(aTHX)
283
284#else
285# define new_SV(p) \
053fc874
GS
286 STMT_START { \
287 LOCK_SV_MUTEX; \
288 if (PL_sv_root) \
289 uproot_SV(p); \
290 else \
cac9b346 291 (p) = S_more_sv(aTHX); \
053fc874
GS
292 UNLOCK_SV_MUTEX; \
293 SvANY(p) = 0; \
294 SvREFCNT(p) = 1; \
295 SvFLAGS(p) = 0; \
296 } STMT_END
eba0f806 297#endif
463ee0b2 298
645c22ef
DM
299
300/* del_SV(): return an empty SV head to the free list */
301
a0d0e21e 302#ifdef DEBUGGING
4561caa4 303
053fc874
GS
304#define del_SV(p) \
305 STMT_START { \
306 LOCK_SV_MUTEX; \
aea4f609 307 if (DEBUG_D_TEST) \
053fc874
GS
308 del_sv(p); \
309 else \
310 plant_SV(p); \
311 UNLOCK_SV_MUTEX; \
312 } STMT_END
a0d0e21e 313
76e3520e 314STATIC void
cea2e8a9 315S_del_sv(pTHX_ SV *p)
463ee0b2 316{
aea4f609 317 if (DEBUG_D_TEST) {
4633a7c4 318 SV* sva;
a3b680e6 319 bool ok = 0;
3280af22 320 for (sva = PL_sv_arenaroot; sva; sva = (SV *) SvANY(sva)) {
53c1dcc0
AL
321 const SV * const sv = sva + 1;
322 const SV * const svend = &sva[SvREFCNT(sva)];
c0ff570e 323 if (p >= sv && p < svend) {
a0d0e21e 324 ok = 1;
c0ff570e
NC
325 break;
326 }
a0d0e21e
LW
327 }
328 if (!ok) {
0453d815 329 if (ckWARN_d(WARN_INTERNAL))
9014280d 330 Perl_warner(aTHX_ packWARN(WARN_INTERNAL),
472d47bc
SB
331 "Attempt to free non-arena SV: 0x%"UVxf
332 pTHX__FORMAT, PTR2UV(p) pTHX__VALUE);
a0d0e21e
LW
333 return;
334 }
335 }
4561caa4 336 plant_SV(p);
463ee0b2 337}
a0d0e21e 338
4561caa4
CS
339#else /* ! DEBUGGING */
340
341#define del_SV(p) plant_SV(p)
342
343#endif /* DEBUGGING */
463ee0b2 344
645c22ef
DM
345
346/*
ccfc67b7
JH
347=head1 SV Manipulation Functions
348
645c22ef
DM
349=for apidoc sv_add_arena
350
351Given a chunk of memory, link it to the head of the list of arenas,
352and split it into a list of free SVs.
353
354=cut
355*/
356
4633a7c4 357void
864dbfa3 358Perl_sv_add_arena(pTHX_ char *ptr, U32 size, U32 flags)
463ee0b2 359{
0bd48802 360 SV* const sva = (SV*)ptr;
463ee0b2
LW
361 register SV* sv;
362 register SV* svend;
4633a7c4
LW
363
364 /* The first SV in an arena isn't an SV. */
3280af22 365 SvANY(sva) = (void *) PL_sv_arenaroot; /* ptr to next arena */
4633a7c4
LW
366 SvREFCNT(sva) = size / sizeof(SV); /* number of SV slots */
367 SvFLAGS(sva) = flags; /* FAKE if not to be freed */
368
3280af22
NIS
369 PL_sv_arenaroot = sva;
370 PL_sv_root = sva + 1;
4633a7c4
LW
371
372 svend = &sva[SvREFCNT(sva) - 1];
373 sv = sva + 1;
463ee0b2 374 while (sv < svend) {
48614a46 375 SvARENA_CHAIN(sv) = (void *)(SV*)(sv + 1);
03e36789 376#ifdef DEBUGGING
978b032e 377 SvREFCNT(sv) = 0;
03e36789
NC
378#endif
379 /* Must always set typemask because it's awlays checked in on cleanup
380 when the arenas are walked looking for objects. */
8990e307 381 SvFLAGS(sv) = SVTYPEMASK;
463ee0b2
LW
382 sv++;
383 }
48614a46 384 SvARENA_CHAIN(sv) = 0;
03e36789
NC
385#ifdef DEBUGGING
386 SvREFCNT(sv) = 0;
387#endif
4633a7c4
LW
388 SvFLAGS(sv) = SVTYPEMASK;
389}
390
055972dc
DM
391/* visit(): call the named function for each non-free SV in the arenas
392 * whose flags field matches the flags/mask args. */
645c22ef 393
5226ed68 394STATIC I32
055972dc 395S_visit(pTHX_ SVFUNC_t f, U32 flags, U32 mask)
8990e307 396{
4633a7c4 397 SV* sva;
5226ed68 398 I32 visited = 0;
8990e307 399
3280af22 400 for (sva = PL_sv_arenaroot; sva; sva = (SV*)SvANY(sva)) {
53c1dcc0 401 register const SV * const svend = &sva[SvREFCNT(sva)];
a3b680e6 402 register SV* sv;
4561caa4 403 for (sv = sva + 1; sv < svend; ++sv) {
055972dc
DM
404 if (SvTYPE(sv) != SVTYPEMASK
405 && (sv->sv_flags & mask) == flags
406 && SvREFCNT(sv))
407 {
acfe0abc 408 (FCALL)(aTHX_ sv);
5226ed68
JH
409 ++visited;
410 }
8990e307
LW
411 }
412 }
5226ed68 413 return visited;
8990e307
LW
414}
415
758a08c3
JH
416#ifdef DEBUGGING
417
645c22ef
DM
418/* called by sv_report_used() for each live SV */
419
420static void
acfe0abc 421do_report_used(pTHX_ SV *sv)
645c22ef
DM
422{
423 if (SvTYPE(sv) != SVTYPEMASK) {
424 PerlIO_printf(Perl_debug_log, "****\n");
425 sv_dump(sv);
426 }
427}
758a08c3 428#endif
645c22ef
DM
429
430/*
431=for apidoc sv_report_used
432
433Dump the contents of all SVs not yet freed. (Debugging aid).
434
435=cut
436*/
437
8990e307 438void
864dbfa3 439Perl_sv_report_used(pTHX)
4561caa4 440{
ff270d3a 441#ifdef DEBUGGING
055972dc 442 visit(do_report_used, 0, 0);
ff270d3a 443#endif
4561caa4
CS
444}
445
645c22ef
DM
446/* called by sv_clean_objs() for each live SV */
447
448static void
e15faf7d 449do_clean_objs(pTHX_ SV *ref)
645c22ef 450{
823a54a3
AL
451 if (SvROK(ref)) {
452 SV * const target = SvRV(ref);
453 if (SvOBJECT(target)) {
454 DEBUG_D((PerlIO_printf(Perl_debug_log, "Cleaning object ref:\n "), sv_dump(ref)));
455 if (SvWEAKREF(ref)) {
456 sv_del_backref(target, ref);
457 SvWEAKREF_off(ref);
458 SvRV_set(ref, NULL);
459 } else {
460 SvROK_off(ref);
461 SvRV_set(ref, NULL);
462 SvREFCNT_dec(target);
463 }
645c22ef
DM
464 }
465 }
466
467 /* XXX Might want to check arrays, etc. */
468}
469
470/* called by sv_clean_objs() for each live SV */
471
472#ifndef DISABLE_DESTRUCTOR_KLUDGE
473static void
acfe0abc 474do_clean_named_objs(pTHX_ SV *sv)
645c22ef
DM
475{
476 if (SvTYPE(sv) == SVt_PVGV && GvGP(sv)) {
c69033f2
NC
477 if ((
478#ifdef PERL_DONT_CREATE_GVSV
479 GvSV(sv) &&
480#endif
481 SvOBJECT(GvSV(sv))) ||
645c22ef
DM
482 (GvAV(sv) && SvOBJECT(GvAV(sv))) ||
483 (GvHV(sv) && SvOBJECT(GvHV(sv))) ||
484 (GvIO(sv) && SvOBJECT(GvIO(sv))) ||
485 (GvCV(sv) && SvOBJECT(GvCV(sv))) )
486 {
487 DEBUG_D((PerlIO_printf(Perl_debug_log, "Cleaning named glob object:\n "), sv_dump(sv)));
ec5f3c78 488 SvFLAGS(sv) |= SVf_BREAK;
645c22ef
DM
489 SvREFCNT_dec(sv);
490 }
491 }
492}
493#endif
494
495/*
496=for apidoc sv_clean_objs
497
498Attempt to destroy all objects not yet freed
499
500=cut
501*/
502
4561caa4 503void
864dbfa3 504Perl_sv_clean_objs(pTHX)
4561caa4 505{
3280af22 506 PL_in_clean_objs = TRUE;
055972dc 507 visit(do_clean_objs, SVf_ROK, SVf_ROK);
4561caa4 508#ifndef DISABLE_DESTRUCTOR_KLUDGE
2d0f3c12 509 /* some barnacles may yet remain, clinging to typeglobs */
055972dc 510 visit(do_clean_named_objs, SVt_PVGV, SVTYPEMASK);
4561caa4 511#endif
3280af22 512 PL_in_clean_objs = FALSE;
4561caa4
CS
513}
514
645c22ef
DM
515/* called by sv_clean_all() for each live SV */
516
517static void
acfe0abc 518do_clean_all(pTHX_ SV *sv)
645c22ef
DM
519{
520 DEBUG_D((PerlIO_printf(Perl_debug_log, "Cleaning loops: SV at 0x%"UVxf"\n", PTR2UV(sv)) ));
521 SvFLAGS(sv) |= SVf_BREAK;
0e705b3b 522 if (PL_comppad == (AV*)sv) {
7d49f689 523 PL_comppad = NULL;
0e705b3b
DM
524 PL_curpad = Null(SV**);
525 }
645c22ef
DM
526 SvREFCNT_dec(sv);
527}
528
529/*
530=for apidoc sv_clean_all
531
532Decrement the refcnt of each remaining SV, possibly triggering a
533cleanup. This function may have to be called multiple times to free
ff276b08 534SVs which are in complex self-referential hierarchies.
645c22ef
DM
535
536=cut
537*/
538
5226ed68 539I32
864dbfa3 540Perl_sv_clean_all(pTHX)
8990e307 541{
5226ed68 542 I32 cleaned;
3280af22 543 PL_in_clean_all = TRUE;
055972dc 544 cleaned = visit(do_clean_all, 0,0);
3280af22 545 PL_in_clean_all = FALSE;
5226ed68 546 return cleaned;
8990e307 547}
463ee0b2 548
7cfef17e
NC
549static void
550S_free_arena(pTHX_ void **root) {
551 while (root) {
1b6737cc 552 void ** const next = *(void **)root;
7cfef17e
NC
553 Safefree(root);
554 root = next;
555 }
556}
557
645c22ef
DM
558/*
559=for apidoc sv_free_arenas
560
561Deallocate the memory used by all arenas. Note that all the individual SV
562heads and bodies within the arenas must already have been freed.
563
564=cut
565*/
7cfef17e
NC
566#define free_arena(name) \
567 STMT_START { \
568 S_free_arena(aTHX_ (void**) PL_ ## name ## _arenaroot); \
569 PL_ ## name ## _arenaroot = 0; \
570 PL_ ## name ## _root = 0; \
571 } STMT_END
572
4633a7c4 573void
864dbfa3 574Perl_sv_free_arenas(pTHX)
4633a7c4
LW
575{
576 SV* sva;
577 SV* svanext;
93e68bfb 578 int i;
4633a7c4
LW
579
580 /* Free arenas here, but be careful about fake ones. (We assume
581 contiguity of the fake ones with the corresponding real ones.) */
582
3280af22 583 for (sva = PL_sv_arenaroot; sva; sva = svanext) {
4633a7c4
LW
584 svanext = (SV*) SvANY(sva);
585 while (svanext && SvFAKE(svanext))
586 svanext = (SV*) SvANY(svanext);
587
588 if (!SvFAKE(sva))
1df70142 589 Safefree(sva);
4633a7c4 590 }
93e68bfb
JC
591
592 for (i=0; i<SVt_LAST; i++) {
593 S_free_arena(aTHX_ (void**) PL_body_arenaroots[i]);
594 PL_body_arenaroots[i] = 0;
595 PL_body_roots[i] = 0;
596 }
597
43c5f42d 598 Safefree(PL_nice_chunk);
3280af22
NIS
599 PL_nice_chunk = Nullch;
600 PL_nice_chunk_size = 0;
601 PL_sv_arenaroot = 0;
602 PL_sv_root = 0;
4633a7c4
LW
603}
604
bd81e77b
NC
605/*
606 Here are mid-level routines that manage the allocation of bodies out
607 of the various arenas. There are 5 kinds of arenas:
29489e7c 608
bd81e77b
NC
609 1. SV-head arenas, which are discussed and handled above
610 2. regular body arenas
611 3. arenas for reduced-size bodies
612 4. Hash-Entry arenas
613 5. pte arenas (thread related)
29489e7c 614
bd81e77b
NC
615 Arena types 2 & 3 are chained by body-type off an array of
616 arena-root pointers, which is indexed by svtype. Some of the
617 larger/less used body types are malloced singly, since a large
618 unused block of them is wasteful. Also, several svtypes dont have
619 bodies; the data fits into the sv-head itself. The arena-root
620 pointer thus has a few unused root-pointers (which may be hijacked
621 later for arena types 4,5)
29489e7c 622
bd81e77b
NC
623 3 differs from 2 as an optimization; some body types have several
624 unused fields in the front of the structure (which are kept in-place
625 for consistency). These bodies can be allocated in smaller chunks,
626 because the leading fields arent accessed. Pointers to such bodies
627 are decremented to point at the unused 'ghost' memory, knowing that
628 the pointers are used with offsets to the real memory.
29489e7c 629
bd81e77b
NC
630 HE, HEK arenas are managed separately, with separate code, but may
631 be merge-able later..
632
633 PTE arenas are not sv-bodies, but they share these mid-level
634 mechanics, so are considered here. The new mid-level mechanics rely
635 on the sv_type of the body being allocated, so we just reserve one
636 of the unused body-slots for PTEs, then use it in those (2) PTE
637 contexts below (line ~10k)
638*/
639
640STATIC void *
641S_more_bodies (pTHX_ size_t size, svtype sv_type)
29489e7c 642{
00b6aa41
AL
643 void ** const arena_root = &PL_body_arenaroots[sv_type];
644 void ** const root = &PL_body_roots[sv_type];
bd81e77b
NC
645 char *start;
646 const char *end;
647 const size_t count = PERL_ARENA_SIZE / size;
29489e7c 648
bd81e77b
NC
649 Newx(start, count*size, char);
650 *((void **) start) = *arena_root;
651 *arena_root = (void *)start;
29489e7c 652
bd81e77b 653 end = start + (count-1) * size;
29489e7c 654
bd81e77b
NC
655 /* The initial slot is used to link the arenas together, so it isn't to be
656 linked into the list of ready-to-use bodies. */
29489e7c 657
bd81e77b 658 start += size;
29489e7c 659
bd81e77b 660 *root = (void *)start;
29489e7c 661
bd81e77b
NC
662 while (start < end) {
663 char * const next = start + size;
664 *(void**) start = (void *)next;
665 start = next;
29489e7c 666 }
bd81e77b
NC
667 *(void **)start = 0;
668
669 return *root;
29489e7c
DM
670}
671
bd81e77b 672/* grab a new thing from the free list, allocating more if necessary */
29489e7c 673
bd81e77b 674/* 1st, the inline version */
29489e7c 675
bd81e77b
NC
676#define new_body_inline(xpv, size, sv_type) \
677 STMT_START { \
00b6aa41 678 void ** const r3wt = &PL_body_roots[sv_type]; \
bd81e77b
NC
679 LOCK_SV_MUTEX; \
680 xpv = *((void **)(r3wt)) \
681 ? *((void **)(r3wt)) : S_more_bodies(aTHX_ size, sv_type); \
682 *(r3wt) = *(void**)(xpv); \
683 UNLOCK_SV_MUTEX; \
684 } STMT_END
29489e7c 685
bd81e77b 686/* now use the inline version in the proper function */
29489e7c 687
bd81e77b 688#ifndef PURIFY
9393da09 689
bd81e77b
NC
690/* This isn't being used with -DPURIFY, so don't declare it. Otherwise
691 compilers issue warnings. */
9393da09 692
bd81e77b
NC
693STATIC void *
694S_new_body(pTHX_ size_t size, svtype sv_type)
695{
696 void *xpv;
697 new_body_inline(xpv, size, sv_type);
698 return xpv;
699}
9393da09 700
bd81e77b 701#endif
53c1dcc0 702
bd81e77b 703/* return a thing to the free list */
29489e7c 704
bd81e77b
NC
705#define del_body(thing, root) \
706 STMT_START { \
00b6aa41 707 void ** const thing_copy = (void **)thing;\
bd81e77b
NC
708 LOCK_SV_MUTEX; \
709 *thing_copy = *root; \
710 *root = (void*)thing_copy; \
711 UNLOCK_SV_MUTEX; \
712 } STMT_END
29489e7c 713
bd81e77b
NC
714/*
715 Revisiting type 3 arenas, there are 4 body-types which have some
716 members that are never accessed. They are XPV, XPVIV, XPVAV,
717 XPVHV, which have corresponding types: xpv_allocated,
718 xpviv_allocated, xpvav_allocated, xpvhv_allocated,
29489e7c 719
bd81e77b
NC
720 For these types, the arenas are carved up into *_allocated size
721 chunks, we thus avoid wasted memory for those unaccessed members.
722 When bodies are allocated, we adjust the pointer back in memory by
723 the size of the bit not allocated, so it's as if we allocated the
724 full structure. (But things will all go boom if you write to the
725 part that is "not there", because you'll be overwriting the last
726 members of the preceding structure in memory.)
29489e7c 727
bd81e77b
NC
728 We calculate the correction using the STRUCT_OFFSET macro. For example, if
729 xpv_allocated is the same structure as XPV then the two OFFSETs sum to zero,
730 and the pointer is unchanged. If the allocated structure is smaller (no
731 initial NV actually allocated) then the net effect is to subtract the size
732 of the NV from the pointer, to return a new pointer as if an initial NV were
733 actually allocated.
29489e7c 734
bd81e77b
NC
735 This is the same trick as was used for NV and IV bodies. Ironically it
736 doesn't need to be used for NV bodies any more, because NV is now at the
737 start of the structure. IV bodies don't need it either, because they are
738 no longer allocated. */
29489e7c 739
bd81e77b
NC
740/* The following 2 arrays hide the above details in a pair of
741 lookup-tables, allowing us to be body-type agnostic.
29489e7c 742
bd81e77b
NC
743 size maps svtype to its body's allocated size.
744 offset maps svtype to the body-pointer adjustment needed
29489e7c 745
bd81e77b
NC
746 NB: elements in latter are 0 or <0, and are added during
747 allocation, and subtracted during deallocation. It may be clearer
748 to invert the values, and call it shrinkage_by_svtype.
29489e7c
DM
749*/
750
bd81e77b
NC
751struct body_details {
752 size_t size; /* Size to allocate */
753 size_t copy; /* Size of structure to copy (may be shorter) */
754 size_t offset;
755 bool cant_upgrade; /* Can upgrade this type */
756 bool zero_nv; /* zero the NV when upgrading from this */
757 bool arena; /* Allocated from an arena */
758};
29489e7c 759
bd81e77b
NC
760#define HADNV FALSE
761#define NONV TRUE
29489e7c 762
bd81e77b
NC
763#ifdef PURIFY
764/* With -DPURFIY we allocate everything directly, and don't use arenas.
765 This seems a rather elegant way to simplify some of the code below. */
766#define HASARENA FALSE
767#else
768#define HASARENA TRUE
769#endif
770#define NOARENA FALSE
29489e7c 771
bd81e77b 772/* A macro to work out the offset needed to subtract from a pointer to (say)
29489e7c 773
bd81e77b
NC
774typedef struct {
775 STRLEN xpv_cur;
776 STRLEN xpv_len;
777} xpv_allocated;
29489e7c 778
bd81e77b 779to make its members accessible via a pointer to (say)
29489e7c 780
bd81e77b
NC
781struct xpv {
782 NV xnv_nv;
783 STRLEN xpv_cur;
784 STRLEN xpv_len;
785};
29489e7c 786
bd81e77b 787*/
29489e7c 788
bd81e77b
NC
789#define relative_STRUCT_OFFSET(longer, shorter, member) \
790 (STRUCT_OFFSET(shorter, member) - STRUCT_OFFSET(longer, member))
29489e7c 791
bd81e77b
NC
792/* Calculate the length to copy. Specifically work out the length less any
793 final padding the compiler needed to add. See the comment in sv_upgrade
794 for why copying the padding proved to be a bug. */
29489e7c 795
bd81e77b
NC
796#define copy_length(type, last_member) \
797 STRUCT_OFFSET(type, last_member) \
798 + sizeof (((type*)SvANY((SV*)0))->last_member)
29489e7c 799
bd81e77b
NC
800static const struct body_details bodies_by_type[] = {
801 {0, 0, 0, FALSE, NONV, NOARENA},
802 /* IVs are in the head, so the allocation size is 0 */
803 {0, sizeof(IV), STRUCT_OFFSET(XPVIV, xiv_iv), FALSE, NONV, NOARENA},
804 /* 8 bytes on most ILP32 with IEEE doubles */
805 {sizeof(NV), sizeof(NV), 0, FALSE, HADNV, HASARENA},
806 /* RVs are in the head now */
807 /* However, this slot is overloaded and used by the pte */
808 {0, 0, 0, FALSE, NONV, NOARENA},
809 /* 8 bytes on most ILP32 with IEEE doubles */
810 {sizeof(xpv_allocated),
811 copy_length(XPV, xpv_len)
d41c018a
NC
812 - relative_STRUCT_OFFSET(xpv_allocated, XPV, xpv_cur),
813 + relative_STRUCT_OFFSET(xpv_allocated, XPV, xpv_cur),
bd81e77b
NC
814 FALSE, NONV, HASARENA},
815 /* 12 */
816 {sizeof(xpviv_allocated),
817 copy_length(XPVIV, xiv_u)
d41c018a
NC
818 - relative_STRUCT_OFFSET(xpviv_allocated, XPVIV, xpv_cur),
819 + relative_STRUCT_OFFSET(xpviv_allocated, XPVIV, xpv_cur),
bd81e77b
NC
820 FALSE, NONV, HASARENA},
821 /* 20 */
822 {sizeof(XPVNV), copy_length(XPVNV, xiv_u), 0, FALSE, HADNV, HASARENA},
823 /* 28 */
824 {sizeof(XPVMG), copy_length(XPVMG, xmg_stash), 0, FALSE, HADNV, HASARENA},
825 /* 36 */
826 {sizeof(XPVBM), sizeof(XPVBM), 0, TRUE, HADNV, HASARENA},
827 /* 48 */
828 {sizeof(XPVGV), sizeof(XPVGV), 0, TRUE, HADNV, HASARENA},
829 /* 64 */
830 {sizeof(XPVLV), sizeof(XPVLV), 0, TRUE, HADNV, HASARENA},
831 /* 20 */
832 {sizeof(xpvav_allocated),
833 copy_length(XPVAV, xmg_stash)
d41c018a
NC
834 - relative_STRUCT_OFFSET(xpvav_allocated, XPVAV, xav_fill),
835 + relative_STRUCT_OFFSET(xpvav_allocated, XPVAV, xav_fill),
bd81e77b
NC
836 TRUE, HADNV, HASARENA},
837 /* 20 */
838 {sizeof(xpvhv_allocated),
839 copy_length(XPVHV, xmg_stash)
d41c018a
NC
840 - relative_STRUCT_OFFSET(xpvhv_allocated, XPVHV, xhv_fill),
841 + relative_STRUCT_OFFSET(xpvhv_allocated, XPVHV, xhv_fill),
bd81e77b
NC
842 TRUE, HADNV, HASARENA},
843 /* 76 */
844 {sizeof(XPVCV), sizeof(XPVCV), 0, TRUE, HADNV, HASARENA},
845 /* 80 */
846 {sizeof(XPVFM), sizeof(XPVFM), 0, TRUE, HADNV, NOARENA},
847 /* 84 */
848 {sizeof(XPVIO), sizeof(XPVIO), 0, TRUE, HADNV, NOARENA}
849};
29489e7c 850
bd81e77b
NC
851#define new_body_type(sv_type) \
852 (void *)((char *)S_new_body(aTHX_ bodies_by_type[sv_type].size, sv_type)\
853 - bodies_by_type[sv_type].offset)
29489e7c 854
bd81e77b
NC
855#define del_body_type(p, sv_type) \
856 del_body(p, &PL_body_roots[sv_type])
29489e7c 857
29489e7c 858
bd81e77b
NC
859#define new_body_allocated(sv_type) \
860 (void *)((char *)S_new_body(aTHX_ bodies_by_type[sv_type].size, sv_type)\
861 - bodies_by_type[sv_type].offset)
29489e7c 862
bd81e77b
NC
863#define del_body_allocated(p, sv_type) \
864 del_body(p + bodies_by_type[sv_type].offset, &PL_body_roots[sv_type])
29489e7c 865
29489e7c 866
bd81e77b
NC
867#define my_safemalloc(s) (void*)safemalloc(s)
868#define my_safecalloc(s) (void*)safecalloc(s, 1)
869#define my_safefree(p) safefree((char*)p)
29489e7c 870
bd81e77b 871#ifdef PURIFY
29489e7c 872
bd81e77b
NC
873#define new_XNV() my_safemalloc(sizeof(XPVNV))
874#define del_XNV(p) my_safefree(p)
29489e7c 875
bd81e77b
NC
876#define new_XPVNV() my_safemalloc(sizeof(XPVNV))
877#define del_XPVNV(p) my_safefree(p)
29489e7c 878
bd81e77b
NC
879#define new_XPVAV() my_safemalloc(sizeof(XPVAV))
880#define del_XPVAV(p) my_safefree(p)
29489e7c 881
bd81e77b
NC
882#define new_XPVHV() my_safemalloc(sizeof(XPVHV))
883#define del_XPVHV(p) my_safefree(p)
29489e7c 884
bd81e77b
NC
885#define new_XPVMG() my_safemalloc(sizeof(XPVMG))
886#define del_XPVMG(p) my_safefree(p)
29489e7c 887
bd81e77b
NC
888#define new_XPVGV() my_safemalloc(sizeof(XPVGV))
889#define del_XPVGV(p) my_safefree(p)
29489e7c 890
bd81e77b 891#else /* !PURIFY */
29489e7c 892
bd81e77b
NC
893#define new_XNV() new_body_type(SVt_NV)
894#define del_XNV(p) del_body_type(p, SVt_NV)
29489e7c 895
bd81e77b
NC
896#define new_XPVNV() new_body_type(SVt_PVNV)
897#define del_XPVNV(p) del_body_type(p, SVt_PVNV)
29489e7c 898
bd81e77b
NC
899#define new_XPVAV() new_body_allocated(SVt_PVAV)
900#define del_XPVAV(p) del_body_allocated(p, SVt_PVAV)
645c22ef 901
bd81e77b
NC
902#define new_XPVHV() new_body_allocated(SVt_PVHV)
903#define del_XPVHV(p) del_body_allocated(p, SVt_PVHV)
645c22ef 904
bd81e77b
NC
905#define new_XPVMG() new_body_type(SVt_PVMG)
906#define del_XPVMG(p) del_body_type(p, SVt_PVMG)
645c22ef 907
bd81e77b
NC
908#define new_XPVGV() new_body_type(SVt_PVGV)
909#define del_XPVGV(p) del_body_type(p, SVt_PVGV)
1d7c1841 910
bd81e77b 911#endif /* PURIFY */
93e68bfb 912
bd81e77b 913/* no arena for you! */
93e68bfb 914
bd81e77b
NC
915#define new_NOARENA(details) \
916 my_safemalloc((details)->size + (details)->offset)
917#define new_NOARENAZ(details) \
918 my_safecalloc((details)->size + (details)->offset)
93e68bfb 919
bd81e77b
NC
920/*
921=for apidoc sv_upgrade
93e68bfb 922
bd81e77b
NC
923Upgrade an SV to a more complex form. Generally adds a new body type to the
924SV, then copies across as much information as possible from the old body.
925You generally want to use the C<SvUPGRADE> macro wrapper. See also C<svtype>.
93e68bfb 926
bd81e77b 927=cut
93e68bfb 928*/
93e68bfb 929
bd81e77b
NC
930void
931Perl_sv_upgrade(pTHX_ register SV *sv, U32 new_type)
cac9b346 932{
bd81e77b
NC
933 void* old_body;
934 void* new_body;
935 const U32 old_type = SvTYPE(sv);
936 const struct body_details *const old_type_details
937 = bodies_by_type + old_type;
938 const struct body_details *new_type_details = bodies_by_type + new_type;
cac9b346 939
bd81e77b
NC
940 if (new_type != SVt_PV && SvIsCOW(sv)) {
941 sv_force_normal_flags(sv, 0);
942 }
cac9b346 943
bd81e77b
NC
944 if (old_type == new_type)
945 return;
cac9b346 946
bd81e77b
NC
947 if (old_type > new_type)
948 Perl_croak(aTHX_ "sv_upgrade from type %d down to type %d",
949 (int)old_type, (int)new_type);
cac9b346 950
cac9b346 951
bd81e77b 952 old_body = SvANY(sv);
de042e1d 953
bd81e77b
NC
954 /* Copying structures onto other structures that have been neatly zeroed
955 has a subtle gotcha. Consider XPVMG
cac9b346 956
bd81e77b
NC
957 +------+------+------+------+------+-------+-------+
958 | NV | CUR | LEN | IV | MAGIC | STASH |
959 +------+------+------+------+------+-------+-------+
960 0 4 8 12 16 20 24 28
645c22ef 961
bd81e77b
NC
962 where NVs are aligned to 8 bytes, so that sizeof that structure is
963 actually 32 bytes long, with 4 bytes of padding at the end:
08742458 964
bd81e77b
NC
965 +------+------+------+------+------+-------+-------+------+
966 | NV | CUR | LEN | IV | MAGIC | STASH | ??? |
967 +------+------+------+------+------+-------+-------+------+
968 0 4 8 12 16 20 24 28 32
08742458 969
bd81e77b 970 so what happens if you allocate memory for this structure:
30f9da9e 971
bd81e77b
NC
972 +------+------+------+------+------+-------+-------+------+------+...
973 | NV | CUR | LEN | IV | MAGIC | STASH | GP | NAME |
974 +------+------+------+------+------+-------+-------+------+------+...
975 0 4 8 12 16 20 24 28 32 36
bfc44f79 976
bd81e77b
NC
977 zero it, then copy sizeof(XPVMG) bytes on top of it? Not quite what you
978 expect, because you copy the area marked ??? onto GP. Now, ??? may have
979 started out as zero once, but it's quite possible that it isn't. So now,
980 rather than a nicely zeroed GP, you have it pointing somewhere random.
981 Bugs ensue.
bfc44f79 982
bd81e77b
NC
983 (In fact, GP ends up pointing at a previous GP structure, because the
984 principle cause of the padding in XPVMG getting garbage is a copy of
985 sizeof(XPVMG) bytes from a XPVGV structure in sv_unglob)
30f9da9e 986
bd81e77b
NC
987 So we are careful and work out the size of used parts of all the
988 structures. */
bfc44f79 989
bd81e77b
NC
990 switch (old_type) {
991 case SVt_NULL:
992 break;
993 case SVt_IV:
994 if (new_type < SVt_PVIV) {
995 new_type = (new_type == SVt_NV)
996 ? SVt_PVNV : SVt_PVIV;
997 new_type_details = bodies_by_type + new_type;
998 }
999 break;
1000 case SVt_NV:
1001 if (new_type < SVt_PVNV) {
1002 new_type = SVt_PVNV;
1003 new_type_details = bodies_by_type + new_type;
1004 }
1005 break;
1006 case SVt_RV:
1007 break;
1008 case SVt_PV:
1009 assert(new_type > SVt_PV);
1010 assert(SVt_IV < SVt_PV);
1011 assert(SVt_NV < SVt_PV);
1012 break;
1013 case SVt_PVIV:
1014 break;
1015 case SVt_PVNV:
1016 break;
1017 case SVt_PVMG:
1018 /* Because the XPVMG of PL_mess_sv isn't allocated from the arena,
1019 there's no way that it can be safely upgraded, because perl.c
1020 expects to Safefree(SvANY(PL_mess_sv)) */
1021 assert(sv != PL_mess_sv);
1022 /* This flag bit is used to mean other things in other scalar types.
1023 Given that it only has meaning inside the pad, it shouldn't be set
1024 on anything that can get upgraded. */
1025 assert((SvFLAGS(sv) & SVpad_TYPED) == 0);
1026 break;
1027 default:
1028 if (old_type_details->cant_upgrade)
1029 Perl_croak(aTHX_ "Can't upgrade that kind of scalar");
1030 }
645c22ef 1031
bd81e77b
NC
1032 SvFLAGS(sv) &= ~SVTYPEMASK;
1033 SvFLAGS(sv) |= new_type;
932e9ff9 1034
bd81e77b
NC
1035 switch (new_type) {
1036 case SVt_NULL:
1037 Perl_croak(aTHX_ "Can't upgrade to undef");
1038 case SVt_IV:
1039 assert(old_type == SVt_NULL);
1040 SvANY(sv) = (XPVIV*)((char*)&(sv->sv_u.svu_iv) - STRUCT_OFFSET(XPVIV, xiv_iv));
1041 SvIV_set(sv, 0);
1042 return;
1043 case SVt_NV:
1044 assert(old_type == SVt_NULL);
1045 SvANY(sv) = new_XNV();
1046 SvNV_set(sv, 0);
1047 return;
1048 case SVt_RV:
1049 assert(old_type == SVt_NULL);
1050 SvANY(sv) = &sv->sv_u.svu_rv;
1051 SvRV_set(sv, 0);
1052 return;
1053 case SVt_PVHV:
1054 SvANY(sv) = new_XPVHV();
1055 HvFILL(sv) = 0;
1056 HvMAX(sv) = 0;
1057 HvTOTALKEYS(sv) = 0;
645c22ef 1058
bd81e77b 1059 goto hv_av_common;
aeb18a1e 1060
bd81e77b
NC
1061 case SVt_PVAV:
1062 SvANY(sv) = new_XPVAV();
1063 AvMAX(sv) = -1;
1064 AvFILLp(sv) = -1;
1065 AvALLOC(sv) = 0;
1066 AvREAL_only(sv);
aeb18a1e 1067
bd81e77b
NC
1068 hv_av_common:
1069 /* SVt_NULL isn't the only thing upgraded to AV or HV.
1070 The target created by newSVrv also is, and it can have magic.
1071 However, it never has SvPVX set.
1072 */
1073 if (old_type >= SVt_RV) {
1074 assert(SvPVX_const(sv) == 0);
1075 }
aeb18a1e 1076
bd81e77b
NC
1077 /* Could put this in the else clause below, as PVMG must have SvPVX
1078 0 already (the assertion above) */
1079 SvPV_set(sv, (char*)0);
93e68bfb 1080
bd81e77b
NC
1081 if (old_type >= SVt_PVMG) {
1082 SvMAGIC_set(sv, ((XPVMG*)old_body)->xmg_magic);
1083 SvSTASH_set(sv, ((XPVMG*)old_body)->xmg_stash);
1084 } else {
1085 SvMAGIC_set(sv, 0);
1086 SvSTASH_set(sv, 0);
1087 }
1088 break;
93e68bfb 1089
93e68bfb 1090
bd81e77b
NC
1091 case SVt_PVIV:
1092 /* XXX Is this still needed? Was it ever needed? Surely as there is
1093 no route from NV to PVIV, NOK can never be true */
1094 assert(!SvNOKp(sv));
1095 assert(!SvNOK(sv));
1096 case SVt_PVIO:
1097 case SVt_PVFM:
1098 case SVt_PVBM:
1099 case SVt_PVGV:
1100 case SVt_PVCV:
1101 case SVt_PVLV:
1102 case SVt_PVMG:
1103 case SVt_PVNV:
1104 case SVt_PV:
93e68bfb 1105
bd81e77b
NC
1106 assert(new_type_details->size);
1107 /* We always allocated the full length item with PURIFY. To do this
1108 we fake things so that arena is false for all 16 types.. */
1109 if(new_type_details->arena) {
1110 /* This points to the start of the allocated area. */
1111 new_body_inline(new_body, new_type_details->size, new_type);
1112 Zero(new_body, new_type_details->size, char);
1113 new_body = ((char *)new_body) - new_type_details->offset;
1114 } else {
1115 new_body = new_NOARENAZ(new_type_details);
1116 }
1117 SvANY(sv) = new_body;
5e2fc214 1118
bd81e77b
NC
1119 if (old_type_details->copy) {
1120 Copy((char *)old_body + old_type_details->offset,
1121 (char *)new_body + old_type_details->offset,
1122 old_type_details->copy, char);
1123 }
1124
1125#ifndef NV_ZERO_IS_ALLBITS_ZERO
1126 /* If NV 0.0 is store as all bits 0 then Zero() already creates a correct
1127 0.0 for us. */
1128 if (old_type_details->zero_nv)
1129 SvNV_set(sv, 0);
82048762 1130#endif
5e2fc214 1131
bd81e77b
NC
1132 if (new_type == SVt_PVIO)
1133 IoPAGE_LEN(sv) = 60;
1134 if (old_type < SVt_RV)
1135 SvPV_set(sv, 0);
1136 break;
1137 default:
1138 Perl_croak(aTHX_ "panic: sv_upgrade to unknown type %lu", new_type);
1139 }
73171d91 1140
bd81e77b
NC
1141 if (old_type_details->size) {
1142 /* If the old body had an allocated size, then we need to free it. */
1143#ifdef PURIFY
1144 my_safefree(old_body);
1145#else
1146 del_body((void*)((char*)old_body + old_type_details->offset),
1147 &PL_body_roots[old_type]);
1148#endif
1149 }
1150}
73171d91 1151
bd81e77b
NC
1152/*
1153=for apidoc sv_backoff
73171d91 1154
bd81e77b
NC
1155Remove any string offset. You should normally use the C<SvOOK_off> macro
1156wrapper instead.
73171d91 1157
bd81e77b 1158=cut
73171d91
NC
1159*/
1160
bd81e77b
NC
1161int
1162Perl_sv_backoff(pTHX_ register SV *sv)
1163{
1164 assert(SvOOK(sv));
1165 assert(SvTYPE(sv) != SVt_PVHV);
1166 assert(SvTYPE(sv) != SVt_PVAV);
1167 if (SvIVX(sv)) {
1168 const char * const s = SvPVX_const(sv);
1169 SvLEN_set(sv, SvLEN(sv) + SvIVX(sv));
1170 SvPV_set(sv, SvPVX(sv) - SvIVX(sv));
1171 SvIV_set(sv, 0);
1172 Move(s, SvPVX(sv), SvCUR(sv)+1, char);
1173 }
1174 SvFLAGS(sv) &= ~SVf_OOK;
1175 return 0;
1176}
73171d91 1177
bd81e77b
NC
1178/*
1179=for apidoc sv_grow
73171d91 1180
bd81e77b
NC
1181Expands the character buffer in the SV. If necessary, uses C<sv_unref> and
1182upgrades the SV to C<SVt_PV>. Returns a pointer to the character buffer.
1183Use the C<SvGROW> wrapper instead.
93e68bfb 1184
bd81e77b
NC
1185=cut
1186*/
93e68bfb 1187
bd81e77b
NC
1188char *
1189Perl_sv_grow(pTHX_ register SV *sv, register STRLEN newlen)
1190{
1191 register char *s;
93e68bfb 1192
bd81e77b
NC
1193#ifdef HAS_64K_LIMIT
1194 if (newlen >= 0x10000) {
1195 PerlIO_printf(Perl_debug_log,
1196 "Allocation too large: %"UVxf"\n", (UV)newlen);
1197 my_exit(1);
1198 }
1199#endif /* HAS_64K_LIMIT */
1200 if (SvROK(sv))
1201 sv_unref(sv);
1202 if (SvTYPE(sv) < SVt_PV) {
1203 sv_upgrade(sv, SVt_PV);
1204 s = SvPVX_mutable(sv);
1205 }
1206 else if (SvOOK(sv)) { /* pv is offset? */
1207 sv_backoff(sv);
1208 s = SvPVX_mutable(sv);
1209 if (newlen > SvLEN(sv))
1210 newlen += 10 * (newlen - SvCUR(sv)); /* avoid copy each time */
1211#ifdef HAS_64K_LIMIT
1212 if (newlen >= 0x10000)
1213 newlen = 0xFFFF;
1214#endif
1215 }
1216 else
1217 s = SvPVX_mutable(sv);
aeb18a1e 1218
bd81e77b
NC
1219 if (newlen > SvLEN(sv)) { /* need more room? */
1220 newlen = PERL_STRLEN_ROUNDUP(newlen);
1221 if (SvLEN(sv) && s) {
1222#ifdef MYMALLOC
1223 const STRLEN l = malloced_size((void*)SvPVX_const(sv));
1224 if (newlen <= l) {
1225 SvLEN_set(sv, l);
1226 return s;
1227 } else
1228#endif
1229 s = saferealloc(s, newlen);
1230 }
1231 else {
1232 s = safemalloc(newlen);
1233 if (SvPVX_const(sv) && SvCUR(sv)) {
1234 Move(SvPVX_const(sv), s, (newlen < SvCUR(sv)) ? newlen : SvCUR(sv), char);
1235 }
1236 }
1237 SvPV_set(sv, s);
1238 SvLEN_set(sv, newlen);
1239 }
1240 return s;
1241}
aeb18a1e 1242
bd81e77b
NC
1243/*
1244=for apidoc sv_setiv
932e9ff9 1245
bd81e77b
NC
1246Copies an integer into the given SV, upgrading first if necessary.
1247Does not handle 'set' magic. See also C<sv_setiv_mg>.
463ee0b2 1248
bd81e77b
NC
1249=cut
1250*/
463ee0b2 1251
bd81e77b
NC
1252void
1253Perl_sv_setiv(pTHX_ register SV *sv, IV i)
1254{
1255 SV_CHECK_THINKFIRST_COW_DROP(sv);
1256 switch (SvTYPE(sv)) {
1257 case SVt_NULL:
1258 sv_upgrade(sv, SVt_IV);
1259 break;
1260 case SVt_NV:
1261 sv_upgrade(sv, SVt_PVNV);
1262 break;
1263 case SVt_RV:
1264 case SVt_PV:
1265 sv_upgrade(sv, SVt_PVIV);
1266 break;
463ee0b2 1267
bd81e77b
NC
1268 case SVt_PVGV:
1269 case SVt_PVAV:
1270 case SVt_PVHV:
1271 case SVt_PVCV:
1272 case SVt_PVFM:
1273 case SVt_PVIO:
1274 Perl_croak(aTHX_ "Can't coerce %s to integer in %s", sv_reftype(sv,0),
1275 OP_DESC(PL_op));
1276 }
1277 (void)SvIOK_only(sv); /* validate number */
1278 SvIV_set(sv, i);
1279 SvTAINT(sv);
1280}
932e9ff9 1281
bd81e77b
NC
1282/*
1283=for apidoc sv_setiv_mg
d33b2eba 1284
bd81e77b 1285Like C<sv_setiv>, but also handles 'set' magic.
1c846c1f 1286
bd81e77b
NC
1287=cut
1288*/
d33b2eba 1289
bd81e77b
NC
1290void
1291Perl_sv_setiv_mg(pTHX_ register SV *sv, IV i)
1292{
1293 sv_setiv(sv,i);
1294 SvSETMAGIC(sv);
1295}
727879eb 1296
bd81e77b
NC
1297/*
1298=for apidoc sv_setuv
d33b2eba 1299
bd81e77b
NC
1300Copies an unsigned integer into the given SV, upgrading first if necessary.
1301Does not handle 'set' magic. See also C<sv_setuv_mg>.
9b94d1dd 1302
bd81e77b
NC
1303=cut
1304*/
d33b2eba 1305
bd81e77b
NC
1306void
1307Perl_sv_setuv(pTHX_ register SV *sv, UV u)
1308{
1309 /* With these two if statements:
1310 u=1.49 s=0.52 cu=72.49 cs=10.64 scripts=270 tests=20865
d33b2eba 1311
bd81e77b
NC
1312 without
1313 u=1.35 s=0.47 cu=73.45 cs=11.43 scripts=270 tests=20865
1c846c1f 1314
bd81e77b
NC
1315 If you wish to remove them, please benchmark to see what the effect is
1316 */
1317 if (u <= (UV)IV_MAX) {
1318 sv_setiv(sv, (IV)u);
1319 return;
1320 }
1321 sv_setiv(sv, 0);
1322 SvIsUV_on(sv);
1323 SvUV_set(sv, u);
1324}
d33b2eba 1325
bd81e77b
NC
1326/*
1327=for apidoc sv_setuv_mg
727879eb 1328
bd81e77b 1329Like C<sv_setuv>, but also handles 'set' magic.
9b94d1dd 1330
bd81e77b
NC
1331=cut
1332*/
5e2fc214 1333
bd81e77b
NC
1334void
1335Perl_sv_setuv_mg(pTHX_ register SV *sv, UV u)
1336{
1337 sv_setiv(sv, 0);
1338 SvIsUV_on(sv);
1339 sv_setuv(sv,u);
1340 SvSETMAGIC(sv);
1341}
5e2fc214 1342
954c1994 1343/*
bd81e77b 1344=for apidoc sv_setnv
954c1994 1345
bd81e77b
NC
1346Copies a double into the given SV, upgrading first if necessary.
1347Does not handle 'set' magic. See also C<sv_setnv_mg>.
954c1994
GS
1348
1349=cut
1350*/
1351
63f97190 1352void
bd81e77b 1353Perl_sv_setnv(pTHX_ register SV *sv, NV num)
79072805 1354{
bd81e77b
NC
1355 SV_CHECK_THINKFIRST_COW_DROP(sv);
1356 switch (SvTYPE(sv)) {
79072805 1357 case SVt_NULL:
79072805 1358 case SVt_IV:
bd81e77b 1359 sv_upgrade(sv, SVt_NV);
79072805 1360 break;
ed6116ce 1361 case SVt_RV:
79072805 1362 case SVt_PV:
79072805 1363 case SVt_PVIV:
bd81e77b 1364 sv_upgrade(sv, SVt_PVNV);
79072805 1365 break;
bd4b1eb5 1366
bd4b1eb5 1367 case SVt_PVGV:
bd81e77b
NC
1368 case SVt_PVAV:
1369 case SVt_PVHV:
79072805 1370 case SVt_PVCV:
bd81e77b
NC
1371 case SVt_PVFM:
1372 case SVt_PVIO:
1373 Perl_croak(aTHX_ "Can't coerce %s to number in %s", sv_reftype(sv,0),
1374 OP_NAME(PL_op));
2068cd4d 1375 }
bd81e77b
NC
1376 SvNV_set(sv, num);
1377 (void)SvNOK_only(sv); /* validate number */
1378 SvTAINT(sv);
79072805
LW
1379}
1380
645c22ef 1381/*
bd81e77b 1382=for apidoc sv_setnv_mg
645c22ef 1383
bd81e77b 1384Like C<sv_setnv>, but also handles 'set' magic.
645c22ef
DM
1385
1386=cut
1387*/
1388
bd81e77b
NC
1389void
1390Perl_sv_setnv_mg(pTHX_ register SV *sv, NV num)
79072805 1391{
bd81e77b
NC
1392 sv_setnv(sv,num);
1393 SvSETMAGIC(sv);
79072805
LW
1394}
1395
bd81e77b
NC
1396/* Print an "isn't numeric" warning, using a cleaned-up,
1397 * printable version of the offending string
1398 */
954c1994 1399
bd81e77b
NC
1400STATIC void
1401S_not_a_number(pTHX_ SV *sv)
79072805 1402{
bd81e77b
NC
1403 SV *dsv;
1404 char tmpbuf[64];
1405 const char *pv;
94463019
JH
1406
1407 if (DO_UTF8(sv)) {
d0043bd1 1408 dsv = sv_2mortal(newSVpvn("", 0));
94463019
JH
1409 pv = sv_uni_display(dsv, sv, 10, 0);
1410 } else {
1411 char *d = tmpbuf;
551405c4 1412 const char * const limit = tmpbuf + sizeof(tmpbuf) - 8;
94463019
JH
1413 /* each *s can expand to 4 chars + "...\0",
1414 i.e. need room for 8 chars */
ecdeb87c 1415
00b6aa41
AL
1416 const char *s = SvPVX_const(sv);
1417 const char * const end = s + SvCUR(sv);
1418 for ( ; s < end && d < limit; s++ ) {
94463019
JH
1419 int ch = *s & 0xFF;
1420 if (ch & 128 && !isPRINT_LC(ch)) {
1421 *d++ = 'M';
1422 *d++ = '-';
1423 ch &= 127;
1424 }
1425 if (ch == '\n') {
1426 *d++ = '\\';
1427 *d++ = 'n';
1428 }
1429 else if (ch == '\r') {
1430 *d++ = '\\';
1431 *d++ = 'r';
1432 }
1433 else if (ch == '\f') {
1434 *d++ = '\\';
1435 *d++ = 'f';
1436 }
1437 else if (ch == '\\') {
1438 *d++ = '\\';
1439 *d++ = '\\';
1440 }
1441 else if (ch == '\0') {
1442 *d++ = '\\';
1443 *d++ = '0';
1444 }
1445 else if (isPRINT_LC(ch))
1446 *d++ = ch;
1447 else {
1448 *d++ = '^';
1449 *d++ = toCTRL(ch);
1450 }
1451 }
1452 if (s < end) {
1453 *d++ = '.';
1454 *d++ = '.';
1455 *d++ = '.';
1456 }
1457 *d = '\0';
1458 pv = tmpbuf;
a0d0e21e 1459 }
a0d0e21e 1460
533c011a 1461 if (PL_op)
9014280d 1462 Perl_warner(aTHX_ packWARN(WARN_NUMERIC),
94463019
JH
1463 "Argument \"%s\" isn't numeric in %s", pv,
1464 OP_DESC(PL_op));
a0d0e21e 1465 else
9014280d 1466 Perl_warner(aTHX_ packWARN(WARN_NUMERIC),
94463019 1467 "Argument \"%s\" isn't numeric", pv);
a0d0e21e
LW
1468}
1469
c2988b20
NC
1470/*
1471=for apidoc looks_like_number
1472
645c22ef
DM
1473Test if the content of an SV looks like a number (or is a number).
1474C<Inf> and C<Infinity> are treated as numbers (so will not issue a
1475non-numeric warning), even if your atof() doesn't grok them.
c2988b20
NC
1476
1477=cut
1478*/
1479
1480I32
1481Perl_looks_like_number(pTHX_ SV *sv)
1482{
a3b680e6 1483 register const char *sbegin;
c2988b20
NC
1484 STRLEN len;
1485
1486 if (SvPOK(sv)) {
3f7c398e 1487 sbegin = SvPVX_const(sv);
c2988b20
NC
1488 len = SvCUR(sv);
1489 }
1490 else if (SvPOKp(sv))
83003860 1491 sbegin = SvPV_const(sv, len);
c2988b20 1492 else
e0ab1c0e 1493 return SvFLAGS(sv) & (SVf_NOK|SVp_NOK|SVf_IOK|SVp_IOK);
c2988b20
NC
1494 return grok_number(sbegin, len, NULL);
1495}
25da4f38
IZ
1496
1497/* Actually, ISO C leaves conversion of UV to IV undefined, but
1498 until proven guilty, assume that things are not that bad... */
1499
645c22ef
DM
1500/*
1501 NV_PRESERVES_UV:
1502
1503 As 64 bit platforms often have an NV that doesn't preserve all bits of
28e5dec8
JH
1504 an IV (an assumption perl has been based on to date) it becomes necessary
1505 to remove the assumption that the NV always carries enough precision to
1506 recreate the IV whenever needed, and that the NV is the canonical form.
1507 Instead, IV/UV and NV need to be given equal rights. So as to not lose
645c22ef 1508 precision as a side effect of conversion (which would lead to insanity
28e5dec8
JH
1509 and the dragon(s) in t/op/numconvert.t getting very angry) the intent is
1510 1) to distinguish between IV/UV/NV slots that have cached a valid
1511 conversion where precision was lost and IV/UV/NV slots that have a
1512 valid conversion which has lost no precision
645c22ef 1513 2) to ensure that if a numeric conversion to one form is requested that
28e5dec8
JH
1514 would lose precision, the precise conversion (or differently
1515 imprecise conversion) is also performed and cached, to prevent
1516 requests for different numeric formats on the same SV causing
1517 lossy conversion chains. (lossless conversion chains are perfectly
1518 acceptable (still))
1519
1520
1521 flags are used:
1522 SvIOKp is true if the IV slot contains a valid value
1523 SvIOK is true only if the IV value is accurate (UV if SvIOK_UV true)
1524 SvNOKp is true if the NV slot contains a valid value
1525 SvNOK is true only if the NV value is accurate
1526
1527 so
645c22ef 1528 while converting from PV to NV, check to see if converting that NV to an
28e5dec8
JH
1529 IV(or UV) would lose accuracy over a direct conversion from PV to
1530 IV(or UV). If it would, cache both conversions, return NV, but mark
1531 SV as IOK NOKp (ie not NOK).
1532
645c22ef 1533 While converting from PV to IV, check to see if converting that IV to an
28e5dec8
JH
1534 NV would lose accuracy over a direct conversion from PV to NV. If it
1535 would, cache both conversions, flag similarly.
1536
1537 Before, the SV value "3.2" could become NV=3.2 IV=3 NOK, IOK quite
1538 correctly because if IV & NV were set NV *always* overruled.
645c22ef
DM
1539 Now, "3.2" will become NV=3.2 IV=3 NOK, IOKp, because the flag's meaning
1540 changes - now IV and NV together means that the two are interchangeable:
28e5dec8 1541 SvIVX == (IV) SvNVX && SvNVX == (NV) SvIVX;
d460ef45 1542
645c22ef
DM
1543 The benefit of this is that operations such as pp_add know that if
1544 SvIOK is true for both left and right operands, then integer addition
1545 can be used instead of floating point (for cases where the result won't
1546 overflow). Before, floating point was always used, which could lead to
28e5dec8
JH
1547 loss of precision compared with integer addition.
1548
1549 * making IV and NV equal status should make maths accurate on 64 bit
1550 platforms
1551 * may speed up maths somewhat if pp_add and friends start to use
645c22ef 1552 integers when possible instead of fp. (Hopefully the overhead in
28e5dec8
JH
1553 looking for SvIOK and checking for overflow will not outweigh the
1554 fp to integer speedup)
1555 * will slow down integer operations (callers of SvIV) on "inaccurate"
1556 values, as the change from SvIOK to SvIOKp will cause a call into
1557 sv_2iv each time rather than a macro access direct to the IV slot
1558 * should speed up number->string conversion on integers as IV is
645c22ef 1559 favoured when IV and NV are equally accurate
28e5dec8
JH
1560
1561 ####################################################################
645c22ef
DM
1562 You had better be using SvIOK_notUV if you want an IV for arithmetic:
1563 SvIOK is true if (IV or UV), so you might be getting (IV)SvUV.
1564 On the other hand, SvUOK is true iff UV.
28e5dec8
JH
1565 ####################################################################
1566
645c22ef 1567 Your mileage will vary depending your CPU's relative fp to integer
28e5dec8
JH
1568 performance ratio.
1569*/
1570
1571#ifndef NV_PRESERVES_UV
645c22ef
DM
1572# define IS_NUMBER_UNDERFLOW_IV 1
1573# define IS_NUMBER_UNDERFLOW_UV 2
1574# define IS_NUMBER_IV_AND_UV 2
1575# define IS_NUMBER_OVERFLOW_IV 4
1576# define IS_NUMBER_OVERFLOW_UV 5
1577
1578/* sv_2iuv_non_preserve(): private routine for use by sv_2iv() and sv_2uv() */
28e5dec8
JH
1579
1580/* For sv_2nv these three cases are "SvNOK and don't bother casting" */
1581STATIC int
645c22ef 1582S_sv_2iuv_non_preserve(pTHX_ register SV *sv, I32 numtype)
28e5dec8 1583{
3f7c398e 1584 DEBUG_c(PerlIO_printf(Perl_debug_log,"sv_2iuv_non '%s', IV=0x%"UVxf" NV=%"NVgf" inttype=%"UVXf"\n", SvPVX_const(sv), SvIVX(sv), SvNVX(sv), (UV)numtype));
28e5dec8
JH
1585 if (SvNVX(sv) < (NV)IV_MIN) {
1586 (void)SvIOKp_on(sv);
1587 (void)SvNOK_on(sv);
45977657 1588 SvIV_set(sv, IV_MIN);
28e5dec8
JH
1589 return IS_NUMBER_UNDERFLOW_IV;
1590 }
1591 if (SvNVX(sv) > (NV)UV_MAX) {
1592 (void)SvIOKp_on(sv);
1593 (void)SvNOK_on(sv);
1594 SvIsUV_on(sv);
607fa7f2 1595 SvUV_set(sv, UV_MAX);
28e5dec8
JH
1596 return IS_NUMBER_OVERFLOW_UV;
1597 }
c2988b20
NC
1598 (void)SvIOKp_on(sv);
1599 (void)SvNOK_on(sv);
1600 /* Can't use strtol etc to convert this string. (See truth table in
1601 sv_2iv */
1602 if (SvNVX(sv) <= (UV)IV_MAX) {
45977657 1603 SvIV_set(sv, I_V(SvNVX(sv)));
c2988b20
NC
1604 if ((NV)(SvIVX(sv)) == SvNVX(sv)) {
1605 SvIOK_on(sv); /* Integer is precise. NOK, IOK */
1606 } else {
1607 /* Integer is imprecise. NOK, IOKp */
1608 }
1609 return SvNVX(sv) < 0 ? IS_NUMBER_UNDERFLOW_UV : IS_NUMBER_IV_AND_UV;
1610 }
1611 SvIsUV_on(sv);
607fa7f2 1612 SvUV_set(sv, U_V(SvNVX(sv)));
c2988b20
NC
1613 if ((NV)(SvUVX(sv)) == SvNVX(sv)) {
1614 if (SvUVX(sv) == UV_MAX) {
1615 /* As we know that NVs don't preserve UVs, UV_MAX cannot
1616 possibly be preserved by NV. Hence, it must be overflow.
1617 NOK, IOKp */
1618 return IS_NUMBER_OVERFLOW_UV;
1619 }
1620 SvIOK_on(sv); /* Integer is precise. NOK, UOK */
1621 } else {
1622 /* Integer is imprecise. NOK, IOKp */
28e5dec8 1623 }
c2988b20 1624 return IS_NUMBER_OVERFLOW_IV;
28e5dec8 1625}
645c22ef
DM
1626#endif /* !NV_PRESERVES_UV*/
1627
af359546
NC
1628STATIC bool
1629S_sv_2iuv_common(pTHX_ SV *sv) {
1630 if (SvNOKp(sv)) {
28e5dec8
JH
1631 /* erm. not sure. *should* never get NOKp (without NOK) from sv_2nv
1632 * without also getting a cached IV/UV from it at the same time
1633 * (ie PV->NV conversion should detect loss of accuracy and cache
af359546
NC
1634 * IV or UV at same time to avoid this. */
1635 /* IV-over-UV optimisation - choose to cache IV if possible */
25da4f38
IZ
1636
1637 if (SvTYPE(sv) == SVt_NV)
1638 sv_upgrade(sv, SVt_PVNV);
1639
28e5dec8
JH
1640 (void)SvIOKp_on(sv); /* Must do this first, to clear any SvOOK */
1641 /* < not <= as for NV doesn't preserve UV, ((NV)IV_MAX+1) will almost
1642 certainly cast into the IV range at IV_MAX, whereas the correct
1643 answer is the UV IV_MAX +1. Hence < ensures that dodgy boundary
1644 cases go to UV */
1645 if (SvNVX(sv) < (NV)IV_MAX + 0.5) {
45977657 1646 SvIV_set(sv, I_V(SvNVX(sv)));
28e5dec8
JH
1647 if (SvNVX(sv) == (NV) SvIVX(sv)
1648#ifndef NV_PRESERVES_UV
1649 && (((UV)1 << NV_PRESERVES_UV_BITS) >
1650 (UV)(SvIVX(sv) > 0 ? SvIVX(sv) : -SvIVX(sv)))
1651 /* Don't flag it as "accurately an integer" if the number
1652 came from a (by definition imprecise) NV operation, and
1653 we're outside the range of NV integer precision */
1654#endif
1655 ) {
1656 SvIOK_on(sv); /* Can this go wrong with rounding? NWC */
1657 DEBUG_c(PerlIO_printf(Perl_debug_log,
7234c960 1658 "0x%"UVxf" iv(%"NVgf" => %"IVdf") (precise)\n",
28e5dec8
JH
1659 PTR2UV(sv),
1660 SvNVX(sv),
1661 SvIVX(sv)));
1662
1663 } else {
1664 /* IV not precise. No need to convert from PV, as NV
1665 conversion would already have cached IV if it detected
1666 that PV->IV would be better than PV->NV->IV
1667 flags already correct - don't set public IOK. */
1668 DEBUG_c(PerlIO_printf(Perl_debug_log,
7234c960 1669 "0x%"UVxf" iv(%"NVgf" => %"IVdf") (imprecise)\n",
28e5dec8
JH
1670 PTR2UV(sv),
1671 SvNVX(sv),
1672 SvIVX(sv)));
1673 }
1674 /* Can the above go wrong if SvIVX == IV_MIN and SvNVX < IV_MIN,
1675 but the cast (NV)IV_MIN rounds to a the value less (more
1676 negative) than IV_MIN which happens to be equal to SvNVX ??
1677 Analogous to 0xFFFFFFFFFFFFFFFF rounding up to NV (2**64) and
1678 NV rounding back to 0xFFFFFFFFFFFFFFFF, so UVX == UV(NVX) and
1679 (NV)UVX == NVX are both true, but the values differ. :-(
1680 Hopefully for 2s complement IV_MIN is something like
1681 0x8000000000000000 which will be exact. NWC */
d460ef45 1682 }
25da4f38 1683 else {
607fa7f2 1684 SvUV_set(sv, U_V(SvNVX(sv)));
28e5dec8
JH
1685 if (
1686 (SvNVX(sv) == (NV) SvUVX(sv))
1687#ifndef NV_PRESERVES_UV
1688 /* Make sure it's not 0xFFFFFFFFFFFFFFFF */
1689 /*&& (SvUVX(sv) != UV_MAX) irrelevant with code below */
1690 && (((UV)1 << NV_PRESERVES_UV_BITS) > SvUVX(sv))
1691 /* Don't flag it as "accurately an integer" if the number
1692 came from a (by definition imprecise) NV operation, and
1693 we're outside the range of NV integer precision */
1694#endif
1695 )
1696 SvIOK_on(sv);
25da4f38 1697 SvIsUV_on(sv);
1c846c1f 1698 DEBUG_c(PerlIO_printf(Perl_debug_log,
57def98f 1699 "0x%"UVxf" 2iv(%"UVuf" => %"IVdf") (as unsigned)\n",
56431972 1700 PTR2UV(sv),
57def98f
JH
1701 SvUVX(sv),
1702 SvUVX(sv)));
25da4f38 1703 }
748a9306
LW
1704 }
1705 else if (SvPOKp(sv) && SvLEN(sv)) {
c2988b20 1706 UV value;
504618e9 1707 const int numtype = grok_number(SvPVX_const(sv), SvCUR(sv), &value);
af359546 1708 /* We want to avoid a possible problem when we cache an IV/ a UV which
25da4f38 1709 may be later translated to an NV, and the resulting NV is not
c2988b20
NC
1710 the same as the direct translation of the initial string
1711 (eg 123.456 can shortcut to the IV 123 with atol(), but we must
1712 be careful to ensure that the value with the .456 is around if the
1713 NV value is requested in the future).
1c846c1f 1714
af359546 1715 This means that if we cache such an IV/a UV, we need to cache the
25da4f38 1716 NV as well. Moreover, we trade speed for space, and do not
28e5dec8 1717 cache the NV if we are sure it's not needed.
25da4f38 1718 */
16b7a9a4 1719
c2988b20
NC
1720 /* SVt_PVNV is one higher than SVt_PVIV, hence this order */
1721 if ((numtype & (IS_NUMBER_IN_UV | IS_NUMBER_NOT_INT))
1722 == IS_NUMBER_IN_UV) {
5e045b90 1723 /* It's definitely an integer, only upgrade to PVIV */
28e5dec8
JH
1724 if (SvTYPE(sv) < SVt_PVIV)
1725 sv_upgrade(sv, SVt_PVIV);
f7bbb42a 1726 (void)SvIOK_on(sv);
c2988b20
NC
1727 } else if (SvTYPE(sv) < SVt_PVNV)
1728 sv_upgrade(sv, SVt_PVNV);
28e5dec8 1729
c2988b20
NC
1730 /* If NV preserves UV then we only use the UV value if we know that
1731 we aren't going to call atof() below. If NVs don't preserve UVs
1732 then the value returned may have more precision than atof() will
1733 return, even though value isn't perfectly accurate. */
1734 if ((numtype & (IS_NUMBER_IN_UV
1735#ifdef NV_PRESERVES_UV
1736 | IS_NUMBER_NOT_INT
1737#endif
1738 )) == IS_NUMBER_IN_UV) {
1739 /* This won't turn off the public IOK flag if it was set above */
1740 (void)SvIOKp_on(sv);
1741
1742 if (!(numtype & IS_NUMBER_NEG)) {
1743 /* positive */;
1744 if (value <= (UV)IV_MAX) {
45977657 1745 SvIV_set(sv, (IV)value);
c2988b20 1746 } else {
af359546 1747 /* it didn't overflow, and it was positive. */
607fa7f2 1748 SvUV_set(sv, value);
c2988b20
NC
1749 SvIsUV_on(sv);
1750 }
1751 } else {
1752 /* 2s complement assumption */
1753 if (value <= (UV)IV_MIN) {
45977657 1754 SvIV_set(sv, -(IV)value);
c2988b20
NC
1755 } else {
1756 /* Too negative for an IV. This is a double upgrade, but
d1be9408 1757 I'm assuming it will be rare. */
c2988b20
NC
1758 if (SvTYPE(sv) < SVt_PVNV)
1759 sv_upgrade(sv, SVt_PVNV);
1760 SvNOK_on(sv);
1761 SvIOK_off(sv);
1762 SvIOKp_on(sv);
9d6ce603 1763 SvNV_set(sv, -(NV)value);
45977657 1764 SvIV_set(sv, IV_MIN);
c2988b20
NC
1765 }
1766 }
1767 }
1768 /* For !NV_PRESERVES_UV and IS_NUMBER_IN_UV and IS_NUMBER_NOT_INT we
1769 will be in the previous block to set the IV slot, and the next
1770 block to set the NV slot. So no else here. */
1771
1772 if ((numtype & (IS_NUMBER_IN_UV | IS_NUMBER_NOT_INT))
1773 != IS_NUMBER_IN_UV) {
1774 /* It wasn't an (integer that doesn't overflow the UV). */
3f7c398e 1775 SvNV_set(sv, Atof(SvPVX_const(sv)));
28e5dec8 1776
c2988b20
NC
1777 if (! numtype && ckWARN(WARN_NUMERIC))
1778 not_a_number(sv);
28e5dec8 1779
65202027 1780#if defined(USE_LONG_DOUBLE)
c2988b20
NC
1781 DEBUG_c(PerlIO_printf(Perl_debug_log, "0x%"UVxf" 2iv(%" PERL_PRIgldbl ")\n",
1782 PTR2UV(sv), SvNVX(sv)));
65202027 1783#else
1779d84d 1784 DEBUG_c(PerlIO_printf(Perl_debug_log, "0x%"UVxf" 2iv(%"NVgf")\n",
c2988b20 1785 PTR2UV(sv), SvNVX(sv)));
65202027 1786#endif
28e5dec8 1787
28e5dec8 1788#ifdef NV_PRESERVES_UV
af359546
NC
1789 (void)SvIOKp_on(sv);
1790 (void)SvNOK_on(sv);
1791 if (SvNVX(sv) < (NV)IV_MAX + 0.5) {
1792 SvIV_set(sv, I_V(SvNVX(sv)));
1793 if ((NV)(SvIVX(sv)) == SvNVX(sv)) {
1794 SvIOK_on(sv);
1795 } else {
1796 /* Integer is imprecise. NOK, IOKp */
1797 }
1798 /* UV will not work better than IV */
1799 } else {
1800 if (SvNVX(sv) > (NV)UV_MAX) {
1801 SvIsUV_on(sv);
1802 /* Integer is inaccurate. NOK, IOKp, is UV */
1803 SvUV_set(sv, UV_MAX);
af359546
NC
1804 } else {
1805 SvUV_set(sv, U_V(SvNVX(sv)));
1806 /* 0xFFFFFFFFFFFFFFFF not an issue in here, NVs
1807 NV preservse UV so can do correct comparison. */
1808 if ((NV)(SvUVX(sv)) == SvNVX(sv)) {
1809 SvIOK_on(sv);
af359546
NC
1810 } else {
1811 /* Integer is imprecise. NOK, IOKp, is UV */
af359546
NC
1812 }
1813 }
4b0c9573 1814 SvIsUV_on(sv);
af359546 1815 }
28e5dec8 1816#else /* NV_PRESERVES_UV */
c2988b20
NC
1817 if ((numtype & (IS_NUMBER_IN_UV | IS_NUMBER_NOT_INT))
1818 == (IS_NUMBER_IN_UV | IS_NUMBER_NOT_INT)) {
af359546 1819 /* The IV/UV slot will have been set from value returned by
c2988b20
NC
1820 grok_number above. The NV slot has just been set using
1821 Atof. */
560b0c46 1822 SvNOK_on(sv);
c2988b20
NC
1823 assert (SvIOKp(sv));
1824 } else {
1825 if (((UV)1 << NV_PRESERVES_UV_BITS) >
1826 U_V(SvNVX(sv) > 0 ? SvNVX(sv) : -SvNVX(sv))) {
1827 /* Small enough to preserve all bits. */
1828 (void)SvIOKp_on(sv);
1829 SvNOK_on(sv);
45977657 1830 SvIV_set(sv, I_V(SvNVX(sv)));
c2988b20
NC
1831 if ((NV)(SvIVX(sv)) == SvNVX(sv))
1832 SvIOK_on(sv);
1833 /* Assumption: first non-preserved integer is < IV_MAX,
1834 this NV is in the preserved range, therefore: */
1835 if (!(U_V(SvNVX(sv) > 0 ? SvNVX(sv) : -SvNVX(sv))
1836 < (UV)IV_MAX)) {
32fdb065 1837 Perl_croak(aTHX_ "sv_2iv assumed (U_V(fabs((double)SvNVX(sv))) < (UV)IV_MAX) but SvNVX(sv)=%"NVgf" U_V is 0x%"UVxf", IV_MAX is 0x%"UVxf"\n", SvNVX(sv), U_V(SvNVX(sv)), (UV)IV_MAX);
c2988b20
NC
1838 }
1839 } else {
1840 /* IN_UV NOT_INT
1841 0 0 already failed to read UV.
1842 0 1 already failed to read UV.
1843 1 0 you won't get here in this case. IV/UV
1844 slot set, public IOK, Atof() unneeded.
1845 1 1 already read UV.
1846 so there's no point in sv_2iuv_non_preserve() attempting
1847 to use atol, strtol, strtoul etc. */
40a17c4c 1848 sv_2iuv_non_preserve (sv, numtype);
c2988b20
NC
1849 }
1850 }
28e5dec8 1851#endif /* NV_PRESERVES_UV */
25da4f38 1852 }
af359546
NC
1853 }
1854 else {
1855 if (!(SvFLAGS(sv) & SVs_PADTMP)) {
1856 if (!PL_localizing && ckWARN(WARN_UNINITIALIZED))
1857 report_uninit(sv);
1858 }
25da4f38
IZ
1859 if (SvTYPE(sv) < SVt_IV)
1860 /* Typically the caller expects that sv_any is not NULL now. */
1861 sv_upgrade(sv, SVt_IV);
af359546
NC
1862 /* Return 0 from the caller. */
1863 return TRUE;
1864 }
1865 return FALSE;
1866}
1867
1868/*
1869=for apidoc sv_2iv_flags
1870
1871Return the integer value of an SV, doing any necessary string
1872conversion. If flags includes SV_GMAGIC, does an mg_get() first.
1873Normally used via the C<SvIV(sv)> and C<SvIVx(sv)> macros.
1874
1875=cut
1876*/
1877
1878IV
1879Perl_sv_2iv_flags(pTHX_ register SV *sv, I32 flags)
1880{
1881 if (!sv)
a0d0e21e 1882 return 0;
af359546
NC
1883 if (SvGMAGICAL(sv)) {
1884 if (flags & SV_GMAGIC)
1885 mg_get(sv);
1886 if (SvIOKp(sv))
1887 return SvIVX(sv);
1888 if (SvNOKp(sv)) {
1889 return I_V(SvNVX(sv));
1890 }
71c558c3
NC
1891 if (SvPOKp(sv) && SvLEN(sv)) {
1892 UV value;
1893 const int numtype
1894 = grok_number(SvPVX_const(sv), SvCUR(sv), &value);
1895
1896 if ((numtype & (IS_NUMBER_IN_UV | IS_NUMBER_NOT_INT))
1897 == IS_NUMBER_IN_UV) {
1898 /* It's definitely an integer */
1899 if (numtype & IS_NUMBER_NEG) {
1900 if (value < (UV)IV_MIN)
1901 return -(IV)value;
1902 } else {
1903 if (value < (UV)IV_MAX)
1904 return (IV)value;
1905 }
1906 }
1907 if (!numtype) {
1908 if (ckWARN(WARN_NUMERIC))
1909 not_a_number(sv);
1910 }
1911 return I_V(Atof(SvPVX_const(sv)));
1912 }
1c7ff15e
NC
1913 if (SvROK(sv)) {
1914 goto return_rok;
af359546 1915 }
1c7ff15e
NC
1916 assert(SvTYPE(sv) >= SVt_PVMG);
1917 /* This falls through to the report_uninit inside S_sv_2iuv_common. */
4cb1ec55 1918 } else if (SvTHINKFIRST(sv)) {
af359546 1919 if (SvROK(sv)) {
1c7ff15e 1920 return_rok:
af359546
NC
1921 if (SvAMAGIC(sv)) {
1922 SV * const tmpstr=AMG_CALLun(sv,numer);
1923 if (tmpstr && (!SvROK(tmpstr) || (SvRV(tmpstr) != SvRV(sv)))) {
1924 return SvIV(tmpstr);
1925 }
1926 }
1927 return PTR2IV(SvRV(sv));
1928 }
1929 if (SvIsCOW(sv)) {
1930 sv_force_normal_flags(sv, 0);
1931 }
1932 if (SvREADONLY(sv) && !SvOK(sv)) {
1933 if (ckWARN(WARN_UNINITIALIZED))
1934 report_uninit(sv);
1935 return 0;
1936 }
1937 }
1938 if (!SvIOKp(sv)) {
1939 if (S_sv_2iuv_common(aTHX_ sv))
1940 return 0;
79072805 1941 }
1d7c1841
GS
1942 DEBUG_c(PerlIO_printf(Perl_debug_log, "0x%"UVxf" 2iv(%"IVdf")\n",
1943 PTR2UV(sv),SvIVX(sv)));
25da4f38 1944 return SvIsUV(sv) ? (IV)SvUVX(sv) : SvIVX(sv);
79072805
LW
1945}
1946
645c22ef 1947/*
891f9566 1948=for apidoc sv_2uv_flags
645c22ef
DM
1949
1950Return the unsigned integer value of an SV, doing any necessary string
891f9566
YST
1951conversion. If flags includes SV_GMAGIC, does an mg_get() first.
1952Normally used via the C<SvUV(sv)> and C<SvUVx(sv)> macros.
645c22ef
DM
1953
1954=cut
1955*/
1956
ff68c719 1957UV
891f9566 1958Perl_sv_2uv_flags(pTHX_ register SV *sv, I32 flags)
ff68c719 1959{
1960 if (!sv)
1961 return 0;
1962 if (SvGMAGICAL(sv)) {
891f9566
YST
1963 if (flags & SV_GMAGIC)
1964 mg_get(sv);
ff68c719 1965 if (SvIOKp(sv))
1966 return SvUVX(sv);
1967 if (SvNOKp(sv))
1968 return U_V(SvNVX(sv));
71c558c3
NC
1969 if (SvPOKp(sv) && SvLEN(sv)) {
1970 UV value;
1971 const int numtype
1972 = grok_number(SvPVX_const(sv), SvCUR(sv), &value);
1973
1974 if ((numtype & (IS_NUMBER_IN_UV | IS_NUMBER_NOT_INT))
1975 == IS_NUMBER_IN_UV) {
1976 /* It's definitely an integer */
1977 if (!(numtype & IS_NUMBER_NEG))
1978 return value;
1979 }
1980 if (!numtype) {
1981 if (ckWARN(WARN_NUMERIC))
1982 not_a_number(sv);
1983 }
1984 return U_V(Atof(SvPVX_const(sv)));
1985 }
1c7ff15e
NC
1986 if (SvROK(sv)) {
1987 goto return_rok;
3fe9a6f1 1988 }
1c7ff15e
NC
1989 assert(SvTYPE(sv) >= SVt_PVMG);
1990 /* This falls through to the report_uninit inside S_sv_2iuv_common. */
4cb1ec55 1991 } else if (SvTHINKFIRST(sv)) {
ff68c719 1992 if (SvROK(sv)) {
1c7ff15e 1993 return_rok:
deb46114
NC
1994 if (SvAMAGIC(sv)) {
1995 SV *const tmpstr = AMG_CALLun(sv,numer);
1996 if (tmpstr && (!SvROK(tmpstr) || (SvRV(tmpstr) != SvRV(sv)))) {
1997 return SvUV(tmpstr);
1998 }
1999 }
2000 return PTR2UV(SvRV(sv));
ff68c719 2001 }
765f542d
NC
2002 if (SvIsCOW(sv)) {
2003 sv_force_normal_flags(sv, 0);
8a818333 2004 }
0336b60e 2005 if (SvREADONLY(sv) && !SvOK(sv)) {
0336b60e 2006 if (ckWARN(WARN_UNINITIALIZED))
29489e7c 2007 report_uninit(sv);
ff68c719 2008 return 0;
2009 }
2010 }
af359546
NC
2011 if (!SvIOKp(sv)) {
2012 if (S_sv_2iuv_common(aTHX_ sv))
2013 return 0;
ff68c719 2014 }
25da4f38 2015
1d7c1841
GS
2016 DEBUG_c(PerlIO_printf(Perl_debug_log, "0x%"UVxf" 2uv(%"UVuf")\n",
2017 PTR2UV(sv),SvUVX(sv)));
25da4f38 2018 return SvIsUV(sv) ? SvUVX(sv) : (UV)SvIVX(sv);
ff68c719 2019}
2020
645c22ef
DM
2021/*
2022=for apidoc sv_2nv
2023
2024Return the num value of an SV, doing any necessary string or integer
2025conversion, magic etc. Normally used via the C<SvNV(sv)> and C<SvNVx(sv)>
2026macros.
2027
2028=cut
2029*/
2030
65202027 2031NV
864dbfa3 2032Perl_sv_2nv(pTHX_ register SV *sv)
79072805
LW
2033{
2034 if (!sv)
2035 return 0.0;
8990e307 2036 if (SvGMAGICAL(sv)) {
463ee0b2
LW
2037 mg_get(sv);
2038 if (SvNOKp(sv))
2039 return SvNVX(sv);
a0d0e21e 2040 if (SvPOKp(sv) && SvLEN(sv)) {
041457d9 2041 if (!SvIOKp(sv) && ckWARN(WARN_NUMERIC) &&
504618e9 2042 !grok_number(SvPVX_const(sv), SvCUR(sv), NULL))
a0d0e21e 2043 not_a_number(sv);
3f7c398e 2044 return Atof(SvPVX_const(sv));
a0d0e21e 2045 }
25da4f38 2046 if (SvIOKp(sv)) {
1c846c1f 2047 if (SvIsUV(sv))
65202027 2048 return (NV)SvUVX(sv);
25da4f38 2049 else
65202027 2050 return (NV)SvIVX(sv);
47a72cb8
NC
2051 }
2052 if (SvROK(sv)) {
2053 goto return_rok;
2054 }
2055 assert(SvTYPE(sv) >= SVt_PVMG);
2056 /* This falls through to the report_uninit near the end of the
2057 function. */
2058 } else if (SvTHINKFIRST(sv)) {
a0d0e21e 2059 if (SvROK(sv)) {
47a72cb8 2060 return_rok:
deb46114
NC
2061 if (SvAMAGIC(sv)) {
2062 SV *const tmpstr = AMG_CALLun(sv,numer);
2063 if (tmpstr && (!SvROK(tmpstr) || (SvRV(tmpstr) != SvRV(sv)))) {
2064 return SvNV(tmpstr);
2065 }
2066 }
2067 return PTR2NV(SvRV(sv));
a0d0e21e 2068 }
765f542d
NC
2069 if (SvIsCOW(sv)) {
2070 sv_force_normal_flags(sv, 0);
8a818333 2071 }
0336b60e 2072 if (SvREADONLY(sv) && !SvOK(sv)) {
599cee73 2073 if (ckWARN(WARN_UNINITIALIZED))
29489e7c 2074 report_uninit(sv);
ed6116ce
LW
2075 return 0.0;
2076 }
79072805
LW
2077 }
2078 if (SvTYPE(sv) < SVt_NV) {
7e25a7e9
NC
2079 /* The logic to use SVt_PVNV if necessary is in sv_upgrade. */
2080 sv_upgrade(sv, SVt_NV);
906f284f 2081#ifdef USE_LONG_DOUBLE
097ee67d 2082 DEBUG_c({
f93f4e46 2083 STORE_NUMERIC_LOCAL_SET_STANDARD();
1d7c1841
GS
2084 PerlIO_printf(Perl_debug_log,
2085 "0x%"UVxf" num(%" PERL_PRIgldbl ")\n",
2086 PTR2UV(sv), SvNVX(sv));
572bbb43
GS
2087 RESTORE_NUMERIC_LOCAL();
2088 });
65202027 2089#else
572bbb43 2090 DEBUG_c({
f93f4e46 2091 STORE_NUMERIC_LOCAL_SET_STANDARD();
1779d84d 2092 PerlIO_printf(Perl_debug_log, "0x%"UVxf" num(%"NVgf")\n",
1d7c1841 2093 PTR2UV(sv), SvNVX(sv));
097ee67d
JH
2094 RESTORE_NUMERIC_LOCAL();
2095 });
572bbb43 2096#endif
79072805
LW
2097 }
2098 else if (SvTYPE(sv) < SVt_PVNV)
2099 sv_upgrade(sv, SVt_PVNV);
59d8ce62
NC
2100 if (SvNOKp(sv)) {
2101 return SvNVX(sv);
61604483 2102 }
59d8ce62 2103 if (SvIOKp(sv)) {
9d6ce603 2104 SvNV_set(sv, SvIsUV(sv) ? (NV)SvUVX(sv) : (NV)SvIVX(sv));
28e5dec8
JH
2105#ifdef NV_PRESERVES_UV
2106 SvNOK_on(sv);
2107#else
2108 /* Only set the public NV OK flag if this NV preserves the IV */
2109 /* Check it's not 0xFFFFFFFFFFFFFFFF */
2110 if (SvIsUV(sv) ? ((SvUVX(sv) != UV_MAX)&&(SvUVX(sv) == U_V(SvNVX(sv))))
2111 : (SvIVX(sv) == I_V(SvNVX(sv))))
2112 SvNOK_on(sv);
2113 else
2114 SvNOKp_on(sv);
2115#endif
93a17b20 2116 }
748a9306 2117 else if (SvPOKp(sv) && SvLEN(sv)) {
c2988b20 2118 UV value;
3f7c398e 2119 const int numtype = grok_number(SvPVX_const(sv), SvCUR(sv), &value);
041457d9 2120 if (!SvIOKp(sv) && !numtype && ckWARN(WARN_NUMERIC))
a0d0e21e 2121 not_a_number(sv);
28e5dec8 2122#ifdef NV_PRESERVES_UV
c2988b20
NC
2123 if ((numtype & (IS_NUMBER_IN_UV | IS_NUMBER_NOT_INT))
2124 == IS_NUMBER_IN_UV) {
5e045b90 2125 /* It's definitely an integer */
9d6ce603 2126 SvNV_set(sv, (numtype & IS_NUMBER_NEG) ? -(NV)value : (NV)value);
c2988b20 2127 } else
3f7c398e 2128 SvNV_set(sv, Atof(SvPVX_const(sv)));
28e5dec8
JH
2129 SvNOK_on(sv);
2130#else
3f7c398e 2131 SvNV_set(sv, Atof(SvPVX_const(sv)));
28e5dec8
JH
2132 /* Only set the public NV OK flag if this NV preserves the value in
2133 the PV at least as well as an IV/UV would.
2134 Not sure how to do this 100% reliably. */
2135 /* if that shift count is out of range then Configure's test is
2136 wonky. We shouldn't be in here with NV_PRESERVES_UV_BITS ==
2137 UV_BITS */
2138 if (((UV)1 << NV_PRESERVES_UV_BITS) >
c2988b20 2139 U_V(SvNVX(sv) > 0 ? SvNVX(sv) : -SvNVX(sv))) {
28e5dec8 2140 SvNOK_on(sv); /* Definitely small enough to preserve all bits */
c2988b20
NC
2141 } else if (!(numtype & IS_NUMBER_IN_UV)) {
2142 /* Can't use strtol etc to convert this string, so don't try.
2143 sv_2iv and sv_2uv will use the NV to convert, not the PV. */
2144 SvNOK_on(sv);
2145 } else {
2146 /* value has been set. It may not be precise. */
2147 if ((numtype & IS_NUMBER_NEG) && (value > (UV)IV_MIN)) {
2148 /* 2s complement assumption for (UV)IV_MIN */
2149 SvNOK_on(sv); /* Integer is too negative. */
2150 } else {
2151 SvNOKp_on(sv);
2152 SvIOKp_on(sv);
6fa402ec 2153
c2988b20 2154 if (numtype & IS_NUMBER_NEG) {
45977657 2155 SvIV_set(sv, -(IV)value);
c2988b20 2156 } else if (value <= (UV)IV_MAX) {
45977657 2157 SvIV_set(sv, (IV)value);
c2988b20 2158 } else {
607fa7f2 2159 SvUV_set(sv, value);
c2988b20
NC
2160 SvIsUV_on(sv);
2161 }
2162
2163 if (numtype & IS_NUMBER_NOT_INT) {
2164 /* I believe that even if the original PV had decimals,
2165 they are lost beyond the limit of the FP precision.
2166 However, neither is canonical, so both only get p
2167 flags. NWC, 2000/11/25 */
2168 /* Both already have p flags, so do nothing */
2169 } else {
66a1b24b 2170 const NV nv = SvNVX(sv);
c2988b20
NC
2171 if (SvNVX(sv) < (NV)IV_MAX + 0.5) {
2172 if (SvIVX(sv) == I_V(nv)) {
2173 SvNOK_on(sv);
c2988b20 2174 } else {
c2988b20
NC
2175 /* It had no "." so it must be integer. */
2176 }
00b6aa41 2177 SvIOK_on(sv);
c2988b20
NC
2178 } else {
2179 /* between IV_MAX and NV(UV_MAX).
2180 Could be slightly > UV_MAX */
6fa402ec 2181
c2988b20
NC
2182 if (numtype & IS_NUMBER_NOT_INT) {
2183 /* UV and NV both imprecise. */
2184 } else {
66a1b24b 2185 const UV nv_as_uv = U_V(nv);
c2988b20
NC
2186
2187 if (value == nv_as_uv && SvUVX(sv) != UV_MAX) {
2188 SvNOK_on(sv);
c2988b20 2189 }
00b6aa41 2190 SvIOK_on(sv);
c2988b20
NC
2191 }
2192 }
2193 }
2194 }
2195 }
28e5dec8 2196#endif /* NV_PRESERVES_UV */
93a17b20 2197 }
79072805 2198 else {
041457d9 2199 if (!PL_localizing && !(SvFLAGS(sv) & SVs_PADTMP) && ckWARN(WARN_UNINITIALIZED))
29489e7c 2200 report_uninit(sv);
7e25a7e9
NC
2201 assert (SvTYPE(sv) >= SVt_NV);
2202 /* Typically the caller expects that sv_any is not NULL now. */
2203 /* XXX Ilya implies that this is a bug in callers that assume this
2204 and ideally should be fixed. */
a0d0e21e 2205 return 0.0;
79072805 2206 }
572bbb43 2207#if defined(USE_LONG_DOUBLE)
097ee67d 2208 DEBUG_c({
f93f4e46 2209 STORE_NUMERIC_LOCAL_SET_STANDARD();
1d7c1841
GS
2210 PerlIO_printf(Perl_debug_log, "0x%"UVxf" 2nv(%" PERL_PRIgldbl ")\n",
2211 PTR2UV(sv), SvNVX(sv));
572bbb43
GS
2212 RESTORE_NUMERIC_LOCAL();
2213 });
65202027 2214#else
572bbb43 2215 DEBUG_c({
f93f4e46 2216 STORE_NUMERIC_LOCAL_SET_STANDARD();
1779d84d 2217 PerlIO_printf(Perl_debug_log, "0x%"UVxf" 1nv(%"NVgf")\n",
1d7c1841 2218 PTR2UV(sv), SvNVX(sv));
097ee67d
JH
2219 RESTORE_NUMERIC_LOCAL();
2220 });
572bbb43 2221#endif
463ee0b2 2222 return SvNVX(sv);
79072805
LW
2223}
2224
645c22ef
DM
2225/* uiv_2buf(): private routine for use by sv_2pv_flags(): print an IV or
2226 * UV as a string towards the end of buf, and return pointers to start and
2227 * end of it.
2228 *
2229 * We assume that buf is at least TYPE_CHARS(UV) long.
2230 */
2231
864dbfa3 2232static char *
aec46f14 2233S_uiv_2buf(char *buf, IV iv, UV uv, int is_uv, char **peob)
25da4f38 2234{
25da4f38 2235 char *ptr = buf + TYPE_CHARS(UV);
823a54a3 2236 char * const ebuf = ptr;
25da4f38 2237 int sign;
25da4f38
IZ
2238
2239 if (is_uv)
2240 sign = 0;
2241 else if (iv >= 0) {
2242 uv = iv;
2243 sign = 0;
2244 } else {
2245 uv = -iv;
2246 sign = 1;
2247 }
2248 do {
eb160463 2249 *--ptr = '0' + (char)(uv % 10);
25da4f38
IZ
2250 } while (uv /= 10);
2251 if (sign)
2252 *--ptr = '-';
2253 *peob = ebuf;
2254 return ptr;
2255}
2256
9af30d34
NC
2257/* stringify_regexp(): private routine for use by sv_2pv_flags(): converts
2258 * a regexp to its stringified form.
2259 */
2260
2261static char *
2262S_stringify_regexp(pTHX_ SV *sv, MAGIC *mg, STRLEN *lp) {
00b6aa41 2263 const regexp * const re = (regexp *)mg->mg_obj;
9af30d34
NC
2264
2265 if (!mg->mg_ptr) {
2266 const char *fptr = "msix";
2267 char reflags[6];
2268 char ch;
2269 int left = 0;
2270 int right = 4;
00b6aa41 2271 bool need_newline = 0;
9af30d34
NC
2272 U16 reganch = (U16)((re->reganch & PMf_COMPILETIME) >> 12);
2273
2274 while((ch = *fptr++)) {
2275 if(reganch & 1) {
2276 reflags[left++] = ch;
2277 }
2278 else {
2279 reflags[right--] = ch;
2280 }
2281 reganch >>= 1;
2282 }
2283 if(left != 4) {
2284 reflags[left] = '-';
2285 left = 5;
2286 }
2287
2288 mg->mg_len = re->prelen + 4 + left;
2289 /*
2290 * If /x was used, we have to worry about a regex ending with a
2291 * comment later being embedded within another regex. If so, we don't
2292 * want this regex's "commentization" to leak out to the right part of
2293 * the enclosing regex, we must cap it with a newline.
2294 *
2295 * So, if /x was used, we scan backwards from the end of the regex. If
2296 * we find a '#' before we find a newline, we need to add a newline
2297 * ourself. If we find a '\n' first (or if we don't find '#' or '\n'),
2298 * we don't need to add anything. -jfriedl
2299 */
2300 if (PMf_EXTENDED & re->reganch) {
2301 const char *endptr = re->precomp + re->prelen;
2302 while (endptr >= re->precomp) {
2303 const char c = *(endptr--);
2304 if (c == '\n')
2305 break; /* don't need another */
2306 if (c == '#') {
2307 /* we end while in a comment, so we need a newline */
2308 mg->mg_len++; /* save space for it */
2309 need_newline = 1; /* note to add it */
2310 break;
2311 }
2312 }
2313 }
2314
2315 Newx(mg->mg_ptr, mg->mg_len + 1 + left, char);
2316 mg->mg_ptr[0] = '(';
2317 mg->mg_ptr[1] = '?';
2318 Copy(reflags, mg->mg_ptr+2, left, char);
2319 *(mg->mg_ptr+left+2) = ':';
2320 Copy(re->precomp, mg->mg_ptr+3+left, re->prelen, char);
2321 if (need_newline)
2322 mg->mg_ptr[mg->mg_len - 2] = '\n';
2323 mg->mg_ptr[mg->mg_len - 1] = ')';
2324 mg->mg_ptr[mg->mg_len] = 0;
2325 }
2326 PL_reginterp_cnt += re->program[0].next_off;
2327
2328 if (re->reganch & ROPT_UTF8)
2329 SvUTF8_on(sv);
2330 else
2331 SvUTF8_off(sv);
2332 if (lp)
2333 *lp = mg->mg_len;
2334 return mg->mg_ptr;
2335}
2336
645c22ef
DM
2337/*
2338=for apidoc sv_2pv_flags
2339
ff276b08 2340Returns a pointer to the string value of an SV, and sets *lp to its length.
645c22ef
DM
2341If flags includes SV_GMAGIC, does an mg_get() first. Coerces sv to a string
2342if necessary.
2343Normally invoked via the C<SvPV_flags> macro. C<sv_2pv()> and C<sv_2pv_nomg>
2344usually end up here too.
2345
2346=cut
2347*/
2348
8d6d96c1
HS
2349char *
2350Perl_sv_2pv_flags(pTHX_ register SV *sv, STRLEN *lp, I32 flags)
2351{
79072805 2352 register char *s;
79072805 2353
463ee0b2 2354 if (!sv) {
cdb061a3
NC
2355 if (lp)
2356 *lp = 0;
73d840c0 2357 return (char *)"";
463ee0b2 2358 }
8990e307 2359 if (SvGMAGICAL(sv)) {
8d6d96c1
HS
2360 if (flags & SV_GMAGIC)
2361 mg_get(sv);
463ee0b2 2362 if (SvPOKp(sv)) {
cdb061a3
NC
2363 if (lp)
2364 *lp = SvCUR(sv);
10516c54
NC
2365 if (flags & SV_MUTABLE_RETURN)
2366 return SvPVX_mutable(sv);
4d84ee25
NC
2367 if (flags & SV_CONST_RETURN)
2368 return (char *)SvPVX_const(sv);
463ee0b2
LW
2369 return SvPVX(sv);
2370 }
75dfc8ec
NC
2371 if (SvIOKp(sv) || SvNOKp(sv)) {
2372 char tbuf[64]; /* Must fit sprintf/Gconvert of longest IV/NV */
75dfc8ec
NC
2373 STRLEN len;
2374
2375 if (SvIOKp(sv)) {
e8ada2d0
NC
2376 len = SvIsUV(sv) ? my_sprintf(tbuf,"%"UVuf, (UV)SvUVX(sv))
2377 : my_sprintf(tbuf,"%"IVdf, (IV)SvIVX(sv));
75dfc8ec 2378 } else {
e8ada2d0
NC
2379 Gconvert(SvNVX(sv), NV_DIG, 0, tbuf);
2380 len = strlen(tbuf);
75dfc8ec
NC
2381 }
2382 if (SvROK(sv)) { /* XXX Skip this when sv_pvn_force calls */
2383 /* Sneaky stuff here */
00b6aa41 2384 SV * const tsv = newSVpvn(tbuf, len);
75dfc8ec
NC
2385
2386 sv_2mortal(tsv);
2387 if (lp)
2388 *lp = SvCUR(tsv);
2389 return SvPVX(tsv);
2390 }
2391 else {
2392 dVAR;
2393
2394#ifdef FIXNEGATIVEZERO
e8ada2d0
NC
2395 if (len == 2 && tbuf[0] == '-' && tbuf[1] == '0') {
2396 tbuf[0] = '0';
2397 tbuf[1] = 0;
75dfc8ec
NC
2398 len = 1;
2399 }
2400#endif
2401 SvUPGRADE(sv, SVt_PV);
2402 if (lp)
2403 *lp = len;
2404 s = SvGROW_mutable(sv, len + 1);
2405 SvCUR_set(sv, len);
2406 SvPOKp_on(sv);
e8ada2d0 2407 return memcpy(s, tbuf, len + 1);
75dfc8ec 2408 }
463ee0b2 2409 }
1c7ff15e
NC
2410 if (SvROK(sv)) {
2411 goto return_rok;
2412 }
2413 assert(SvTYPE(sv) >= SVt_PVMG);
2414 /* This falls through to the report_uninit near the end of the
2415 function. */
2416 } else if (SvTHINKFIRST(sv)) {
ed6116ce 2417 if (SvROK(sv)) {
1c7ff15e 2418 return_rok:
deb46114
NC
2419 if (SvAMAGIC(sv)) {
2420 SV *const tmpstr = AMG_CALLun(sv,string);
2421 if (tmpstr && (!SvROK(tmpstr) || (SvRV(tmpstr) != SvRV(sv)))) {
2422 /* Unwrap this: */
2423 /* char *pv = lp ? SvPV(tmpstr, *lp) : SvPV_nolen(tmpstr);
2424 */
2425
2426 char *pv;
2427 if ((SvFLAGS(tmpstr) & (SVf_POK)) == SVf_POK) {
2428 if (flags & SV_CONST_RETURN) {
2429 pv = (char *) SvPVX_const(tmpstr);
2430 } else {
2431 pv = (flags & SV_MUTABLE_RETURN)
2432 ? SvPVX_mutable(tmpstr) : SvPVX(tmpstr);
2433 }
2434 if (lp)
2435 *lp = SvCUR(tmpstr);
50adf7d2 2436 } else {
deb46114 2437 pv = sv_2pv_flags(tmpstr, lp, flags);
50adf7d2 2438 }
deb46114
NC
2439 if (SvUTF8(tmpstr))
2440 SvUTF8_on(sv);
2441 else
2442 SvUTF8_off(sv);
2443 return pv;
50adf7d2 2444 }
deb46114
NC
2445 }
2446 {
75dfc8ec 2447 SV *tsv;
f9277f47 2448 MAGIC *mg;
d8eae41e
NC
2449 const SV *const referent = (SV*)SvRV(sv);
2450
2451 if (!referent) {
042dae7a
NC
2452 tsv = sv_2mortal(newSVpvn("NULLREF", 7));
2453 } else if (SvTYPE(referent) == SVt_PVMG
2454 && ((SvFLAGS(referent) &
2455 (SVs_OBJECT|SVf_OK|SVs_GMG|SVs_SMG|SVs_RMG))
2456 == (SVs_OBJECT|SVs_SMG))
2457 && (mg = mg_find(referent, PERL_MAGIC_qr))) {
c445ea15 2458 return stringify_regexp(sv, mg, lp);
d8eae41e
NC
2459 } else {
2460 const char *const typestr = sv_reftype(referent, 0);
2461
2462 tsv = sv_newmortal();
2463 if (SvOBJECT(referent)) {
2464 const char *const name = HvNAME_get(SvSTASH(referent));
2465 Perl_sv_setpvf(aTHX_ tsv, "%s=%s(0x%"UVxf")",
2466 name ? name : "__ANON__" , typestr,
2467 PTR2UV(referent));
2468 }
2469 else
2470 Perl_sv_setpvf(aTHX_ tsv, "%s(0x%"UVxf")", typestr,
2471 PTR2UV(referent));
c080367d 2472 }
042dae7a
NC
2473 if (lp)
2474 *lp = SvCUR(tsv);
2475 return SvPVX(tsv);
463ee0b2 2476 }
79072805 2477 }
0336b60e 2478 if (SvREADONLY(sv) && !SvOK(sv)) {
0336b60e 2479 if (ckWARN(WARN_UNINITIALIZED))
29489e7c 2480 report_uninit(sv);
cdb061a3
NC
2481 if (lp)
2482 *lp = 0;
73d840c0 2483 return (char *)"";
79072805 2484 }
79072805 2485 }
28e5dec8
JH
2486 if (SvIOK(sv) || ((SvIOKp(sv) && !SvNOKp(sv)))) {
2487 /* I'm assuming that if both IV and NV are equally valid then
2488 converting the IV is going to be more efficient */
e1ec3a88
AL
2489 const U32 isIOK = SvIOK(sv);
2490 const U32 isUIOK = SvIsUV(sv);
28e5dec8
JH
2491 char buf[TYPE_CHARS(UV)];
2492 char *ebuf, *ptr;
2493
2494 if (SvTYPE(sv) < SVt_PVIV)
2495 sv_upgrade(sv, SVt_PVIV);
4ea1d550 2496 ptr = uiv_2buf(buf, SvIVX(sv), SvUVX(sv), isUIOK, &ebuf);
5902b6a9
NC
2497 /* inlined from sv_setpvn */
2498 SvGROW_mutable(sv, (STRLEN)(ebuf - ptr + 1));
4d84ee25 2499 Move(ptr,SvPVX_mutable(sv),ebuf - ptr,char);
28e5dec8
JH
2500 SvCUR_set(sv, ebuf - ptr);
2501 s = SvEND(sv);
2502 *s = '\0';
2503 if (isIOK)
2504 SvIOK_on(sv);
2505 else
2506 SvIOKp_on(sv);
2507 if (isUIOK)
2508 SvIsUV_on(sv);
2509 }
2510 else if (SvNOKp(sv)) {
c81271c3 2511 const int olderrno = errno;
79072805
LW
2512 if (SvTYPE(sv) < SVt_PVNV)
2513 sv_upgrade(sv, SVt_PVNV);
1c846c1f 2514 /* The +20 is pure guesswork. Configure test needed. --jhi */
5902b6a9 2515 s = SvGROW_mutable(sv, NV_DIG + 20);
c81271c3 2516 /* some Xenix systems wipe out errno here */
79072805 2517#ifdef apollo
463ee0b2 2518 if (SvNVX(sv) == 0.0)
79072805
LW
2519 (void)strcpy(s,"0");
2520 else
2521#endif /*apollo*/
bbce6d69 2522 {
2d4389e4 2523 Gconvert(SvNVX(sv), NV_DIG, 0, s);
bbce6d69 2524 }
79072805 2525 errno = olderrno;
a0d0e21e
LW
2526#ifdef FIXNEGATIVEZERO
2527 if (*s == '-' && s[1] == '0' && !s[2])
2528 strcpy(s,"0");
2529#endif
79072805
LW
2530 while (*s) s++;
2531#ifdef hcx
2532 if (s[-1] == '.')
46fc3d4c 2533 *--s = '\0';
79072805
LW
2534#endif
2535 }
79072805 2536 else {
041457d9 2537 if (!PL_localizing && !(SvFLAGS(sv) & SVs_PADTMP) && ckWARN(WARN_UNINITIALIZED))
29489e7c 2538 report_uninit(sv);
cdb061a3 2539 if (lp)
00b6aa41 2540 *lp = 0;
25da4f38
IZ
2541 if (SvTYPE(sv) < SVt_PV)
2542 /* Typically the caller expects that sv_any is not NULL now. */
2543 sv_upgrade(sv, SVt_PV);
73d840c0 2544 return (char *)"";
79072805 2545 }
cdb061a3 2546 {
823a54a3 2547 const STRLEN len = s - SvPVX_const(sv);
cdb061a3
NC
2548 if (lp)
2549 *lp = len;
2550 SvCUR_set(sv, len);
2551 }
79072805 2552 SvPOK_on(sv);
1d7c1841 2553 DEBUG_c(PerlIO_printf(Perl_debug_log, "0x%"UVxf" 2pv(%s)\n",
3f7c398e 2554 PTR2UV(sv),SvPVX_const(sv)));
4d84ee25
NC
2555 if (flags & SV_CONST_RETURN)
2556 return (char *)SvPVX_const(sv);
10516c54
NC
2557 if (flags & SV_MUTABLE_RETURN)
2558 return SvPVX_mutable(sv);
463ee0b2
LW
2559 return SvPVX(sv);
2560}
2561
645c22ef 2562/*
6050d10e
JP
2563=for apidoc sv_copypv
2564
2565Copies a stringified representation of the source SV into the
2566destination SV. Automatically performs any necessary mg_get and
54f0641b 2567coercion of numeric values into strings. Guaranteed to preserve
6050d10e 2568UTF-8 flag even from overloaded objects. Similar in nature to
54f0641b
NIS
2569sv_2pv[_flags] but operates directly on an SV instead of just the
2570string. Mostly uses sv_2pv_flags to do its work, except when that
6050d10e
JP
2571would lose the UTF-8'ness of the PV.
2572
2573=cut
2574*/
2575
2576void
2577Perl_sv_copypv(pTHX_ SV *dsv, register SV *ssv)
2578{
446eaa42 2579 STRLEN len;
53c1dcc0 2580 const char * const s = SvPV_const(ssv,len);
cb50f42d 2581 sv_setpvn(dsv,s,len);
446eaa42 2582 if (SvUTF8(ssv))
cb50f42d 2583 SvUTF8_on(dsv);
446eaa42 2584 else
cb50f42d 2585 SvUTF8_off(dsv);
6050d10e
JP
2586}
2587
2588/*
645c22ef
DM
2589=for apidoc sv_2pvbyte
2590
2591Return a pointer to the byte-encoded representation of the SV, and set *lp
1e54db1a 2592to its length. May cause the SV to be downgraded from UTF-8 as a
645c22ef
DM
2593side-effect.
2594
2595Usually accessed via the C<SvPVbyte> macro.
2596
2597=cut
2598*/
2599
7340a771
GS
2600char *
2601Perl_sv_2pvbyte(pTHX_ register SV *sv, STRLEN *lp)
2602{
0875d2fe 2603 sv_utf8_downgrade(sv,0);
97972285 2604 return lp ? SvPV(sv,*lp) : SvPV_nolen(sv);
7340a771
GS
2605}
2606
645c22ef 2607/*
035cbb0e
RGS
2608=for apidoc sv_2pvutf8
2609
2610Return a pointer to the UTF-8-encoded representation of the SV, and set *lp
2611to its length. May cause the SV to be upgraded to UTF-8 as a side-effect.
2612
2613Usually accessed via the C<SvPVutf8> macro.
2614
2615=cut
2616*/
645c22ef 2617
7340a771
GS
2618char *
2619Perl_sv_2pvutf8(pTHX_ register SV *sv, STRLEN *lp)
2620{
035cbb0e
RGS
2621 sv_utf8_upgrade(sv);
2622 return lp ? SvPV(sv,*lp) : SvPV_nolen(sv);
7340a771 2623}
1c846c1f 2624
7ee2227d 2625
645c22ef
DM
2626/*
2627=for apidoc sv_2bool
2628
2629This function is only called on magical items, and is only used by
8cf8f3d1 2630sv_true() or its macro equivalent.
645c22ef
DM
2631
2632=cut
2633*/
2634
463ee0b2 2635bool
864dbfa3 2636Perl_sv_2bool(pTHX_ register SV *sv)
463ee0b2 2637{
5b295bef 2638 SvGETMAGIC(sv);
463ee0b2 2639
a0d0e21e
LW
2640 if (!SvOK(sv))
2641 return 0;
2642 if (SvROK(sv)) {
fabdb6c0
AL
2643 if (SvAMAGIC(sv)) {
2644 SV * const tmpsv = AMG_CALLun(sv,bool_);
2645 if (tmpsv && (!SvROK(tmpsv) || (SvRV(tmpsv) != SvRV(sv))))
2646 return (bool)SvTRUE(tmpsv);
2647 }
2648 return SvRV(sv) != 0;
a0d0e21e 2649 }
463ee0b2 2650 if (SvPOKp(sv)) {
53c1dcc0
AL
2651 register XPV* const Xpvtmp = (XPV*)SvANY(sv);
2652 if (Xpvtmp &&
339049b0 2653 (*sv->sv_u.svu_pv > '0' ||
11343788 2654 Xpvtmp->xpv_cur > 1 ||
339049b0 2655 (Xpvtmp->xpv_cur && *sv->sv_u.svu_pv != '0')))
463ee0b2
LW
2656 return 1;
2657 else
2658 return 0;
2659 }
2660 else {
2661 if (SvIOKp(sv))
2662 return SvIVX(sv) != 0;
2663 else {
2664 if (SvNOKp(sv))
2665 return SvNVX(sv) != 0.0;
2666 else
2667 return FALSE;
2668 }
2669 }
79072805
LW
2670}
2671
c461cf8f
JH
2672/*
2673=for apidoc sv_utf8_upgrade
2674
78ea37eb 2675Converts the PV of an SV to its UTF-8-encoded form.
645c22ef 2676Forces the SV to string form if it is not already.
4411f3b6
NIS
2677Always sets the SvUTF8 flag to avoid future validity checks even
2678if all the bytes have hibit clear.
c461cf8f 2679
13a6c0e0
JH
2680This is not as a general purpose byte encoding to Unicode interface:
2681use the Encode extension for that.
2682
8d6d96c1
HS
2683=for apidoc sv_utf8_upgrade_flags
2684
78ea37eb 2685Converts the PV of an SV to its UTF-8-encoded form.
645c22ef 2686Forces the SV to string form if it is not already.
8d6d96c1
HS
2687Always sets the SvUTF8 flag to avoid future validity checks even
2688if all the bytes have hibit clear. If C<flags> has C<SV_GMAGIC> bit set,
2689will C<mg_get> on C<sv> if appropriate, else not. C<sv_utf8_upgrade> and
2690C<sv_utf8_upgrade_nomg> are implemented in terms of this function.
2691
13a6c0e0
JH
2692This is not as a general purpose byte encoding to Unicode interface:
2693use the Encode extension for that.
2694
8d6d96c1
HS
2695=cut
2696*/
2697
2698STRLEN
2699Perl_sv_utf8_upgrade_flags(pTHX_ register SV *sv, I32 flags)
2700{
808c356f
RGS
2701 if (sv == &PL_sv_undef)
2702 return 0;
e0e62c2a
NIS
2703 if (!SvPOK(sv)) {
2704 STRLEN len = 0;
d52b7888
NC
2705 if (SvREADONLY(sv) && (SvPOKp(sv) || SvIOKp(sv) || SvNOKp(sv))) {
2706 (void) sv_2pv_flags(sv,&len, flags);
2707 if (SvUTF8(sv))
2708 return len;
2709 } else {
2710 (void) SvPV_force(sv,len);
2711 }
e0e62c2a 2712 }
4411f3b6 2713
f5cee72b 2714 if (SvUTF8(sv)) {
5fec3b1d 2715 return SvCUR(sv);
f5cee72b 2716 }
5fec3b1d 2717
765f542d
NC
2718 if (SvIsCOW(sv)) {
2719 sv_force_normal_flags(sv, 0);
db42d148
NIS
2720 }
2721
88632417 2722 if (PL_encoding && !(flags & SV_UTF8_NO_ENCODING))
799ef3cb 2723 sv_recode_to_utf8(sv, PL_encoding);
9f4817db 2724 else { /* Assume Latin-1/EBCDIC */
c4e7c712
NC
2725 /* This function could be much more efficient if we
2726 * had a FLAG in SVs to signal if there are any hibit
2727 * chars in the PV. Given that there isn't such a flag
2728 * make the loop as fast as possible. */
00b6aa41 2729 const U8 * const s = (U8 *) SvPVX_const(sv);
c4420975 2730 const U8 * const e = (U8 *) SvEND(sv);
93524f2b 2731 const U8 *t = s;
c4e7c712
NC
2732
2733 while (t < e) {
53c1dcc0 2734 const U8 ch = *t++;
00b6aa41
AL
2735 /* Check for hi bit */
2736 if (!NATIVE_IS_INVARIANT(ch)) {
2737 STRLEN len = SvCUR(sv) + 1; /* Plus the \0 */
2738 U8 * const recoded = bytes_to_utf8((U8*)s, &len);
2739
2740 SvPV_free(sv); /* No longer using what was there before. */
2741 SvPV_set(sv, (char*)recoded);
2742 SvCUR_set(sv, len - 1);
2743 SvLEN_set(sv, len); /* No longer know the real size. */
c4e7c712 2744 break;
00b6aa41 2745 }
c4e7c712
NC
2746 }
2747 /* Mark as UTF-8 even if no hibit - saves scanning loop */
2748 SvUTF8_on(sv);
560a288e 2749 }
4411f3b6 2750 return SvCUR(sv);
560a288e
GS
2751}
2752
c461cf8f
JH
2753/*
2754=for apidoc sv_utf8_downgrade
2755
78ea37eb
TS
2756Attempts to convert the PV of an SV from characters to bytes.
2757If the PV contains a character beyond byte, this conversion will fail;
2758in this case, either returns false or, if C<fail_ok> is not
c461cf8f
JH
2759true, croaks.
2760
13a6c0e0
JH
2761This is not as a general purpose Unicode to byte encoding interface:
2762use the Encode extension for that.
2763
c461cf8f
JH
2764=cut
2765*/
2766
560a288e
GS
2767bool
2768Perl_sv_utf8_downgrade(pTHX_ register SV* sv, bool fail_ok)
2769{
78ea37eb 2770 if (SvPOKp(sv) && SvUTF8(sv)) {
fa301091 2771 if (SvCUR(sv)) {
03cfe0ae 2772 U8 *s;
652088fc 2773 STRLEN len;
fa301091 2774
765f542d
NC
2775 if (SvIsCOW(sv)) {
2776 sv_force_normal_flags(sv, 0);
2777 }
03cfe0ae
NIS
2778 s = (U8 *) SvPV(sv, len);
2779 if (!utf8_to_bytes(s, &len)) {
fa301091
JH
2780 if (fail_ok)
2781 return FALSE;
2782 else {
2783 if (PL_op)
2784 Perl_croak(aTHX_ "Wide character in %s",
53e06cf0 2785 OP_DESC(PL_op));
fa301091
JH
2786 else
2787 Perl_croak(aTHX_ "Wide character");
2788 }
4b3603a4 2789 }
b162af07 2790 SvCUR_set(sv, len);
67e989fb 2791 }
560a288e 2792 }
ffebcc3e 2793 SvUTF8_off(sv);
560a288e
GS
2794 return TRUE;
2795}
2796
c461cf8f
JH
2797/*
2798=for apidoc sv_utf8_encode
2799
78ea37eb
TS
2800Converts the PV of an SV to UTF-8, but then turns the C<SvUTF8>
2801flag off so that it looks like octets again.
c461cf8f
JH
2802
2803=cut
2804*/
2805
560a288e
GS
2806void
2807Perl_sv_utf8_encode(pTHX_ register SV *sv)
2808{
4411f3b6 2809 (void) sv_utf8_upgrade(sv);
4c94c214
NC
2810 if (SvIsCOW(sv)) {
2811 sv_force_normal_flags(sv, 0);
2812 }
2813 if (SvREADONLY(sv)) {
2814 Perl_croak(aTHX_ PL_no_modify);
2815 }
560a288e
GS
2816 SvUTF8_off(sv);
2817}
2818
4411f3b6
NIS
2819/*
2820=for apidoc sv_utf8_decode
2821
78ea37eb
TS
2822If the PV of the SV is an octet sequence in UTF-8
2823and contains a multiple-byte character, the C<SvUTF8> flag is turned on
2824so that it looks like a character. If the PV contains only single-byte
2825characters, the C<SvUTF8> flag stays being off.
2826Scans PV for validity and returns false if the PV is invalid UTF-8.
4411f3b6
NIS
2827
2828=cut
2829*/
2830
560a288e
GS
2831bool
2832Perl_sv_utf8_decode(pTHX_ register SV *sv)
2833{
78ea37eb 2834 if (SvPOKp(sv)) {
93524f2b
NC
2835 const U8 *c;
2836 const U8 *e;
9cbac4c7 2837
645c22ef
DM
2838 /* The octets may have got themselves encoded - get them back as
2839 * bytes
2840 */
2841 if (!sv_utf8_downgrade(sv, TRUE))
560a288e
GS
2842 return FALSE;
2843
2844 /* it is actually just a matter of turning the utf8 flag on, but
2845 * we want to make sure everything inside is valid utf8 first.
2846 */
93524f2b 2847 c = (const U8 *) SvPVX_const(sv);
63cd0674 2848 if (!is_utf8_string(c, SvCUR(sv)+1))
67e989fb 2849 return FALSE;
93524f2b 2850 e = (const U8 *) SvEND(sv);
511c2ff0 2851 while (c < e) {
b64e5050 2852 const U8 ch = *c++;
c4d5f83a 2853 if (!UTF8_IS_INVARIANT(ch)) {
67e989fb
JH
2854 SvUTF8_on(sv);
2855 break;
2856 }
560a288e 2857 }
560a288e
GS
2858 }
2859 return TRUE;
2860}
2861
954c1994
GS
2862/*
2863=for apidoc sv_setsv
2864
645c22ef
DM
2865Copies the contents of the source SV C<ssv> into the destination SV
2866C<dsv>. The source SV may be destroyed if it is mortal, so don't use this
2867function if the source SV needs to be reused. Does not handle 'set' magic.
2868Loosely speaking, it performs a copy-by-value, obliterating any previous
2869content of the destination.
2870
2871You probably want to use one of the assortment of wrappers, such as
2872C<SvSetSV>, C<SvSetSV_nosteal>, C<SvSetMagicSV> and
2873C<SvSetMagicSV_nosteal>.
2874
8d6d96c1
HS
2875=for apidoc sv_setsv_flags
2876
645c22ef
DM
2877Copies the contents of the source SV C<ssv> into the destination SV
2878C<dsv>. The source SV may be destroyed if it is mortal, so don't use this
2879function if the source SV needs to be reused. Does not handle 'set' magic.
2880Loosely speaking, it performs a copy-by-value, obliterating any previous
2881content of the destination.
2882If the C<flags> parameter has the C<SV_GMAGIC> bit set, will C<mg_get> on
5fcdf167
NC
2883C<ssv> if appropriate, else not. If the C<flags> parameter has the
2884C<NOSTEAL> bit set then the buffers of temps will not be stolen. <sv_setsv>
2885and C<sv_setsv_nomg> are implemented in terms of this function.
645c22ef
DM
2886
2887You probably want to use one of the assortment of wrappers, such as
2888C<SvSetSV>, C<SvSetSV_nosteal>, C<SvSetMagicSV> and
2889C<SvSetMagicSV_nosteal>.
2890
2891This is the primary function for copying scalars, and most other
2892copy-ish functions and macros use this underneath.
8d6d96c1
HS
2893
2894=cut
2895*/
2896
2897void
2898Perl_sv_setsv_flags(pTHX_ SV *dstr, register SV *sstr, I32 flags)
2899{
8990e307
LW
2900 register U32 sflags;
2901 register int dtype;
2902 register int stype;
463ee0b2 2903
79072805
LW
2904 if (sstr == dstr)
2905 return;
765f542d 2906 SV_CHECK_THINKFIRST_COW_DROP(dstr);
79072805 2907 if (!sstr)
3280af22 2908 sstr = &PL_sv_undef;
8990e307
LW
2909 stype = SvTYPE(sstr);
2910 dtype = SvTYPE(dstr);
79072805 2911
a0d0e21e 2912 SvAMAGIC_off(dstr);
7a5fa8a2 2913 if ( SvVOK(dstr) )
ece467f9
JP
2914 {
2915 /* need to nuke the magic */
2916 mg_free(dstr);
2917 SvRMAGICAL_off(dstr);
2918 }
9e7bc3e8 2919
463ee0b2 2920 /* There's a lot of redundancy below but we're going for speed here */
79072805 2921
8990e307 2922 switch (stype) {
79072805 2923 case SVt_NULL:
aece5585 2924 undef_sstr:
20408e3c
GS
2925 if (dtype != SVt_PVGV) {
2926 (void)SvOK_off(dstr);
2927 return;
2928 }
2929 break;
463ee0b2 2930 case SVt_IV:
aece5585
GA
2931 if (SvIOK(sstr)) {
2932 switch (dtype) {
2933 case SVt_NULL:
8990e307 2934 sv_upgrade(dstr, SVt_IV);
aece5585
GA
2935 break;
2936 case SVt_NV:
8990e307 2937 sv_upgrade(dstr, SVt_PVNV);
aece5585
GA
2938 break;
2939 case SVt_RV:
2940 case SVt_PV:
a0d0e21e 2941 sv_upgrade(dstr, SVt_PVIV);
aece5585
GA
2942 break;
2943 }
2944 (void)SvIOK_only(dstr);
45977657 2945 SvIV_set(dstr, SvIVX(sstr));
25da4f38
IZ
2946 if (SvIsUV(sstr))
2947 SvIsUV_on(dstr);
27c9684d
AP
2948 if (SvTAINTED(sstr))
2949 SvTAINT(dstr);
aece5585 2950 return;
8990e307 2951 }
aece5585
GA
2952 goto undef_sstr;
2953
463ee0b2 2954 case SVt_NV:
aece5585
GA
2955 if (SvNOK(sstr)) {
2956 switch (dtype) {
2957 case SVt_NULL:
2958 case SVt_IV:
8990e307 2959 sv_upgrade(dstr, SVt_NV);
aece5585
GA
2960 break;
2961 case SVt_RV:
2962 case SVt_PV:
2963 case SVt_PVIV:
a0d0e21e 2964 sv_upgrade(dstr, SVt_PVNV);
aece5585
GA
2965 break;
2966 }
9d6ce603 2967 SvNV_set(dstr, SvNVX(sstr));
aece5585 2968 (void)SvNOK_only(dstr);
27c9684d
AP
2969 if (SvTAINTED(sstr))
2970 SvTAINT(dstr);
aece5585 2971 return;
8990e307 2972 }
aece5585
GA
2973 goto undef_sstr;
2974
ed6116ce 2975 case SVt_RV:
8990e307 2976 if (dtype < SVt_RV)
ed6116ce 2977 sv_upgrade(dstr, SVt_RV);
c07a80fd 2978 else if (dtype == SVt_PVGV &&
23bb1b96 2979 SvROK(sstr) && SvTYPE(SvRV(sstr)) == SVt_PVGV) {
c07a80fd 2980 sstr = SvRV(sstr);
a5f75d66 2981 if (sstr == dstr) {
1d7c1841
GS
2982 if (GvIMPORTED(dstr) != GVf_IMPORTED
2983 && CopSTASH_ne(PL_curcop, GvSTASH(dstr)))
2984 {
a5f75d66 2985 GvIMPORTED_on(dstr);
1d7c1841 2986 }
a5f75d66
AD
2987 GvMULTI_on(dstr);
2988 return;
2989 }
c07a80fd 2990 goto glob_assign;
2991 }
ed6116ce 2992 break;
fc36a67e 2993 case SVt_PVFM:
f8c7b90f 2994#ifdef PERL_OLD_COPY_ON_WRITE
d89fc664
NC
2995 if ((SvFLAGS(sstr) & CAN_COW_MASK) == CAN_COW_FLAGS) {
2996 if (dtype < SVt_PVIV)
2997 sv_upgrade(dstr, SVt_PVIV);
2998 break;
2999 }
3000 /* Fall through */
3001#endif
3002 case SVt_PV:
8990e307 3003 if (dtype < SVt_PV)
463ee0b2 3004 sv_upgrade(dstr, SVt_PV);
463ee0b2
LW
3005 break;
3006 case SVt_PVIV:
8990e307 3007 if (dtype < SVt_PVIV)
463ee0b2 3008 sv_upgrade(dstr, SVt_PVIV);
463ee0b2
LW
3009 break;
3010 case SVt_PVNV:
8990e307 3011 if (dtype < SVt_PVNV)
463ee0b2 3012 sv_upgrade(dstr, SVt_PVNV);
463ee0b2 3013 break;
4633a7c4
LW
3014 case SVt_PVAV:
3015 case SVt_PVHV:
3016 case SVt_PVCV:
4633a7c4 3017 case SVt_PVIO:
a3b680e6
AL
3018 {
3019 const char * const type = sv_reftype(sstr,0);
533c011a 3020 if (PL_op)
a3b680e6 3021 Perl_croak(aTHX_ "Bizarre copy of %s in %s", type, OP_NAME(PL_op));
4633a7c4 3022 else
a3b680e6
AL
3023 Perl_croak(aTHX_ "Bizarre copy of %s", type);
3024 }
4633a7c4
LW
3025 break;
3026
79072805 3027 case SVt_PVGV:
8990e307 3028 if (dtype <= SVt_PVGV) {
c07a80fd 3029 glob_assign:
a5f75d66 3030 if (dtype != SVt_PVGV) {
a3b680e6
AL
3031 const char * const name = GvNAME(sstr);
3032 const STRLEN len = GvNAMELEN(sstr);
b76195c2
DM
3033 /* don't upgrade SVt_PVLV: it can hold a glob */
3034 if (dtype != SVt_PVLV)
3035 sv_upgrade(dstr, SVt_PVGV);
14befaf4 3036 sv_magic(dstr, dstr, PERL_MAGIC_glob, Nullch, 0);
e15faf7d
NC
3037 GvSTASH(dstr) = GvSTASH(sstr);
3038 if (GvSTASH(dstr))
3039 Perl_sv_add_backref(aTHX_ (SV*)GvSTASH(dstr), dstr);
a0d0e21e
LW
3040 GvNAME(dstr) = savepvn(name, len);
3041 GvNAMELEN(dstr) = len;
3042 SvFAKE_on(dstr); /* can coerce to non-glob */
3043 }
5bd07a3d 3044
7fb37951
AMS
3045#ifdef GV_UNIQUE_CHECK
3046 if (GvUNIQUE((GV*)dstr)) {
5bd07a3d
DM
3047 Perl_croak(aTHX_ PL_no_modify);
3048 }
3049#endif
3050
a0d0e21e 3051 (void)SvOK_off(dstr);
a5f75d66 3052 GvINTRO_off(dstr); /* one-shot flag */
1edc1566 3053 gp_free((GV*)dstr);
79072805 3054 GvGP(dstr) = gp_ref(GvGP(sstr));
27c9684d
AP
3055 if (SvTAINTED(sstr))
3056 SvTAINT(dstr);
1d7c1841
GS
3057 if (GvIMPORTED(dstr) != GVf_IMPORTED
3058 && CopSTASH_ne(PL_curcop, GvSTASH(dstr)))
3059 {
a5f75d66 3060 GvIMPORTED_on(dstr);
1d7c1841 3061 }
a5f75d66 3062 GvMULTI_on(dstr);
79072805
LW
3063 return;
3064 }
3065 /* FALL THROUGH */
3066
3067 default:
8d6d96c1 3068 if (SvGMAGICAL(sstr) && (flags & SV_GMAGIC)) {
973f89ab 3069 mg_get(sstr);
eb160463 3070 if ((int)SvTYPE(sstr) != stype) {
973f89ab
CS
3071 stype = SvTYPE(sstr);
3072 if (stype == SVt_PVGV && dtype <= SVt_PVGV)
3073 goto glob_assign;
3074 }
3075 }
ded42b9f 3076 if (stype == SVt_PVLV)
862a34c6 3077 SvUPGRADE(dstr, SVt_PVNV);
ded42b9f 3078 else
862a34c6 3079 SvUPGRADE(dstr, (U32)stype);
79072805
LW
3080 }
3081
8990e307
LW
3082 sflags = SvFLAGS(sstr);
3083
3084 if (sflags & SVf_ROK) {
3085 if (dtype >= SVt_PV) {
3086 if (dtype == SVt_PVGV) {
823a54a3 3087 SV * const sref = SvREFCNT_inc(SvRV(sstr));
8990e307 3088 SV *dref = 0;
a3b680e6 3089 const int intro = GvINTRO(dstr);
a0d0e21e 3090
7fb37951
AMS
3091#ifdef GV_UNIQUE_CHECK
3092 if (GvUNIQUE((GV*)dstr)) {
5bd07a3d
DM
3093 Perl_croak(aTHX_ PL_no_modify);
3094 }
3095#endif
3096
a0d0e21e 3097 if (intro) {
a5f75d66 3098 GvINTRO_off(dstr); /* one-shot flag */
1d7c1841 3099 GvLINE(dstr) = CopLINE(PL_curcop);
1edc1566 3100 GvEGV(dstr) = (GV*)dstr;
a0d0e21e 3101 }
a5f75d66 3102 GvMULTI_on(dstr);
8990e307
LW
3103 switch (SvTYPE(sref)) {
3104 case SVt_PVAV:
a0d0e21e 3105 if (intro)
890ed176 3106 SAVEGENERICSV(GvAV(dstr));
a0d0e21e
LW
3107 else
3108 dref = (SV*)GvAV(dstr);
8990e307 3109 GvAV(dstr) = (AV*)sref;
39bac7f7 3110 if (!GvIMPORTED_AV(dstr)
1d7c1841
GS
3111 && CopSTASH_ne(PL_curcop, GvSTASH(dstr)))
3112 {
a5f75d66 3113 GvIMPORTED_AV_on(dstr);
1d7c1841 3114 }
8990e307
LW
3115 break;
3116 case SVt_PVHV:
a0d0e21e 3117 if (intro)
890ed176 3118 SAVEGENERICSV(GvHV(dstr));
a0d0e21e
LW
3119 else
3120 dref = (SV*)GvHV(dstr);
8990e307 3121 GvHV(dstr) = (HV*)sref;
39bac7f7 3122 if (!GvIMPORTED_HV(dstr)
1d7c1841
GS
3123 && CopSTASH_ne(PL_curcop, GvSTASH(dstr)))
3124 {
a5f75d66 3125 GvIMPORTED_HV_on(dstr);
1d7c1841 3126 }
8990e307
LW
3127 break;
3128 case SVt_PVCV:
8ebc5c01 3129 if (intro) {
3130 if (GvCVGEN(dstr) && GvCV(dstr) != (CV*)sref) {
3131 SvREFCNT_dec(GvCV(dstr));
3132 GvCV(dstr) = Nullcv;
68dc0745 3133 GvCVGEN(dstr) = 0; /* Switch off cacheness. */
3280af22 3134 PL_sub_generation++;
8ebc5c01 3135 }
890ed176 3136 SAVEGENERICSV(GvCV(dstr));
8ebc5c01 3137 }
68dc0745 3138 else
3139 dref = (SV*)GvCV(dstr);
3140 if (GvCV(dstr) != (CV*)sref) {
823a54a3 3141 CV* const cv = GvCV(dstr);
4633a7c4 3142 if (cv) {
68dc0745 3143 if (!GvCVGEN((GV*)dstr) &&
3144 (CvROOT(cv) || CvXSUB(cv)))
3145 {
beab0874
JT
3146 /* Redefining a sub - warning is mandatory if
3147 it was a const and its value changed. */
2111d928
NC
3148 if (CvCONST(cv) && CvCONST((CV*)sref)
3149 && cv_const_sv(cv)
3150 == cv_const_sv((CV*)sref)) {
3151 /* They are 2 constant subroutines
3152 generated from the same constant.
3153 This probably means that they are
3154 really the "same" proxy subroutine
3155 instantiated in 2 places. Most likely
3156 this is when a constant is exported
3157 twice. Don't warn. */
3158 }
3159 else if (ckWARN(WARN_REDEFINE)
beab0874
JT
3160 || (CvCONST(cv)
3161 && (!CvCONST((CV*)sref)
3162 || sv_cmp(cv_const_sv(cv),
3163 cv_const_sv((CV*)sref)))))
3164 {
9014280d 3165 Perl_warner(aTHX_ packWARN(WARN_REDEFINE),
beab0874 3166 CvCONST(cv)
910764e6
RGS
3167 ? "Constant subroutine %s::%s redefined"
3168 : "Subroutine %s::%s redefined",
bfcb3514 3169 HvNAME_get(GvSTASH((GV*)dstr)),
beab0874
JT
3170 GvENAME((GV*)dstr));
3171 }
9607fc9c 3172 }
fb24441d
RGS
3173 if (!intro)
3174 cv_ckproto(cv, (GV*)dstr,
93524f2b
NC
3175 SvPOK(sref)
3176 ? SvPVX_const(sref) : Nullch);
4633a7c4 3177 }
a5f75d66 3178 GvCV(dstr) = (CV*)sref;
7a4c00b4 3179 GvCVGEN(dstr) = 0; /* Switch off cacheness. */
a5f75d66 3180 GvASSUMECV_on(dstr);
3280af22 3181 PL_sub_generation++;
a5f75d66 3182 }
39bac7f7 3183 if (!GvIMPORTED_CV(dstr)
1d7c1841
GS
3184 && CopSTASH_ne(PL_curcop, GvSTASH(dstr)))
3185 {
a5f75d66 3186 GvIMPORTED_CV_on(dstr);
1d7c1841 3187 }
8990e307 3188 break;
91bba347
LW
3189 case SVt_PVIO:
3190 if (intro)
890ed176 3191 SAVEGENERICSV(GvIOp(dstr));
91bba347
LW
3192 else
3193 dref = (SV*)GvIOp(dstr);
3194 GvIOp(dstr) = (IO*)sref;
3195 break;
f4d13ee9
JH
3196 case SVt_PVFM:
3197 if (intro)
890ed176 3198 SAVEGENERICSV(GvFORM(dstr));
f4d13ee9
JH
3199 else
3200 dref = (SV*)GvFORM(dstr);
3201 GvFORM(dstr) = (CV*)sref;
3202 break;
8990e307 3203 default:
a0d0e21e 3204 if (intro)
890ed176 3205 SAVEGENERICSV(GvSV(dstr));
a0d0e21e
LW
3206 else
3207 dref = (SV*)GvSV(dstr);
8990e307 3208 GvSV(dstr) = sref;
39bac7f7 3209 if (!GvIMPORTED_SV(dstr)
1d7c1841
GS
3210 && CopSTASH_ne(PL_curcop, GvSTASH(dstr)))
3211 {
a5f75d66 3212 GvIMPORTED_SV_on(dstr);
1d7c1841 3213 }
8990e307
LW
3214 break;
3215 }
3216 if (dref)
3217 SvREFCNT_dec(dref);
27c9684d
AP
3218 if (SvTAINTED(sstr))
3219 SvTAINT(dstr);
8990e307
LW
3220 return;
3221 }
3f7c398e 3222 if (SvPVX_const(dstr)) {
8bd4d4c5 3223 SvPV_free(dstr);
b162af07
SP
3224 SvLEN_set(dstr, 0);
3225 SvCUR_set(dstr, 0);
a0d0e21e 3226 }
8990e307 3227 }
a0d0e21e 3228 (void)SvOK_off(dstr);
b162af07 3229 SvRV_set(dstr, SvREFCNT_inc(SvRV(sstr)));
ed6116ce 3230 SvROK_on(dstr);
8990e307 3231 if (sflags & SVp_NOK) {
3332b3c1
JH
3232 SvNOKp_on(dstr);
3233 /* Only set the public OK flag if the source has public OK. */
3234 if (sflags & SVf_NOK)
3235 SvFLAGS(dstr) |= SVf_NOK;
9d6ce603 3236 SvNV_set(dstr, SvNVX(sstr));
ed6116ce 3237 }
8990e307 3238 if (sflags & SVp_IOK) {
3332b3c1
JH
3239 (void)SvIOKp_on(dstr);
3240 if (sflags & SVf_IOK)
3241 SvFLAGS(dstr) |= SVf_IOK;
2b1c7e3e 3242 if (sflags & SVf_IVisUV)
25da4f38 3243 SvIsUV_on(dstr);
45977657 3244 SvIV_set(dstr, SvIVX(sstr));
ed6116ce 3245 }
a0d0e21e
LW
3246 if (SvAMAGIC(sstr)) {
3247 SvAMAGIC_on(dstr);
3248 }
ed6116ce 3249 }
8990e307 3250 else if (sflags & SVp_POK) {
765f542d 3251 bool isSwipe = 0;
79072805
LW
3252
3253 /*
3254 * Check to see if we can just swipe the string. If so, it's a
3255 * possible small lose on short strings, but a big win on long ones.
3f7c398e
SP
3256 * It might even be a win on short strings if SvPVX_const(dstr)
3257 * has to be allocated and SvPVX_const(sstr) has to be freed.
79072805
LW
3258 */
3259
120fac95
NC
3260 /* Whichever path we take through the next code, we want this true,
3261 and doing it now facilitates the COW check. */
3262 (void)SvPOK_only(dstr);
3263
765f542d 3264 if (
b8f9541a
NC
3265 /* We're not already COW */
3266 ((sflags & (SVf_FAKE | SVf_READONLY)) != (SVf_FAKE | SVf_READONLY)
f8c7b90f 3267#ifndef PERL_OLD_COPY_ON_WRITE
b8f9541a
NC
3268 /* or we are, but dstr isn't a suitable target. */
3269 || (SvFLAGS(dstr) & CAN_COW_MASK) != CAN_COW_FLAGS
3270#endif
3271 )
765f542d 3272 &&
765f542d
NC
3273 !(isSwipe =
3274 (sflags & SVs_TEMP) && /* slated for free anyway? */
3275 !(sflags & SVf_OOK) && /* and not involved in OOK hack? */
5fcdf167
NC
3276 (!(flags & SV_NOSTEAL)) &&
3277 /* and we're allowed to steal temps */
765f542d
NC
3278 SvREFCNT(sstr) == 1 && /* and no other references to it? */
3279 SvLEN(sstr) && /* and really is a string */
645c22ef 3280 /* and won't be needed again, potentially */
765f542d 3281 !(PL_op && PL_op->op_type == OP_AASSIGN))
f8c7b90f 3282#ifdef PERL_OLD_COPY_ON_WRITE
765f542d 3283 && !((sflags & CAN_COW_MASK) == CAN_COW_FLAGS
120fac95 3284 && (SvFLAGS(dstr) & CAN_COW_MASK) == CAN_COW_FLAGS
765f542d
NC
3285 && SvTYPE(sstr) >= SVt_PVIV)
3286#endif
3287 ) {
3288 /* Failed the swipe test, and it's not a shared hash key either.
3289 Have to copy the string. */
3290 STRLEN len = SvCUR(sstr);
3291 SvGROW(dstr, len + 1); /* inlined from sv_setpvn */
3f7c398e 3292 Move(SvPVX_const(sstr),SvPVX(dstr),len,char);
765f542d
NC
3293 SvCUR_set(dstr, len);
3294 *SvEND(dstr) = '\0';
765f542d 3295 } else {
f8c7b90f 3296 /* If PERL_OLD_COPY_ON_WRITE is not defined, then isSwipe will always
765f542d 3297 be true in here. */
765f542d
NC
3298 /* Either it's a shared hash key, or it's suitable for
3299 copy-on-write or we can swipe the string. */
46187eeb 3300 if (DEBUG_C_TEST) {
ed252734 3301 PerlIO_printf(Perl_debug_log, "Copy on write: sstr --> dstr\n");
e419cbc5
NC
3302 sv_dump(sstr);
3303 sv_dump(dstr);
46187eeb 3304 }
f8c7b90f 3305#ifdef PERL_OLD_COPY_ON_WRITE
765f542d
NC
3306 if (!isSwipe) {
3307 /* I believe I should acquire a global SV mutex if
3308 it's a COW sv (not a shared hash key) to stop
3309 it going un copy-on-write.
3310 If the source SV has gone un copy on write between up there
3311 and down here, then (assert() that) it is of the correct
3312 form to make it copy on write again */
3313 if ((sflags & (SVf_FAKE | SVf_READONLY))
3314 != (SVf_FAKE | SVf_READONLY)) {
3315 SvREADONLY_on(sstr);
3316 SvFAKE_on(sstr);
3317 /* Make the source SV into a loop of 1.
3318 (about to become 2) */
a29f6d03 3319 SV_COW_NEXT_SV_SET(sstr, sstr);
765f542d
NC
3320 }
3321 }
3322#endif
3323 /* Initial code is common. */
94010e71
NC
3324 if (SvPVX_const(dstr)) { /* we know that dtype >= SVt_PV */
3325 SvPV_free(dstr);
79072805 3326 }
765f542d 3327
765f542d
NC
3328 if (!isSwipe) {
3329 /* making another shared SV. */
3330 STRLEN cur = SvCUR(sstr);
3331 STRLEN len = SvLEN(sstr);
f8c7b90f 3332#ifdef PERL_OLD_COPY_ON_WRITE
765f542d 3333 if (len) {
b8f9541a 3334 assert (SvTYPE(dstr) >= SVt_PVIV);
765f542d
NC
3335 /* SvIsCOW_normal */
3336 /* splice us in between source and next-after-source. */
a29f6d03
NC
3337 SV_COW_NEXT_SV_SET(dstr, SV_COW_NEXT_SV(sstr));
3338 SV_COW_NEXT_SV_SET(sstr, dstr);
940132f3 3339 SvPV_set(dstr, SvPVX_mutable(sstr));
a604c751
NC
3340 } else
3341#endif
3342 {
765f542d 3343 /* SvIsCOW_shared_hash */
46187eeb
NC
3344 DEBUG_C(PerlIO_printf(Perl_debug_log,
3345 "Copy on write: Sharing hash\n"));
b8f9541a 3346
bdd68bc3 3347 assert (SvTYPE(dstr) >= SVt_PV);
765f542d 3348 SvPV_set(dstr,
d1db91c6 3349 HEK_KEY(share_hek_hek(SvSHARED_HEK_FROM_PV(SvPVX_const(sstr)))));
bdd68bc3 3350 }
87a1ef3d
SP
3351 SvLEN_set(dstr, len);
3352 SvCUR_set(dstr, cur);
765f542d
NC
3353 SvREADONLY_on(dstr);
3354 SvFAKE_on(dstr);
3355 /* Relesase a global SV mutex. */
3356 }
3357 else
765f542d 3358 { /* Passes the swipe test. */
78d1e721 3359 SvPV_set(dstr, SvPVX_mutable(sstr));
765f542d
NC
3360 SvLEN_set(dstr, SvLEN(sstr));
3361 SvCUR_set(dstr, SvCUR(sstr));
3362
3363 SvTEMP_off(dstr);
3364 (void)SvOK_off(sstr); /* NOTE: nukes most SvFLAGS on sstr */
3365 SvPV_set(sstr, Nullch);
3366 SvLEN_set(sstr, 0);
3367 SvCUR_set(sstr, 0);
3368 SvTEMP_off(sstr);
3369 }
3370 }
9aa983d2 3371 if (sflags & SVf_UTF8)
a7cb1f99 3372 SvUTF8_on(dstr);
8990e307 3373 if (sflags & SVp_NOK) {
3332b3c1
JH
3374 SvNOKp_on(dstr);
3375 if (sflags & SVf_NOK)
3376 SvFLAGS(dstr) |= SVf_NOK;
9d6ce603 3377 SvNV_set(dstr, SvNVX(sstr));
79072805 3378 }
8990e307 3379 if (sflags & SVp_IOK) {
3332b3c1
JH
3380 (void)SvIOKp_on(dstr);
3381 if (sflags & SVf_IOK)
3382 SvFLAGS(dstr) |= SVf_IOK;
2b1c7e3e 3383 if (sflags & SVf_IVisUV)
25da4f38 3384 SvIsUV_on(dstr);
45977657 3385 SvIV_set(dstr, SvIVX(sstr));
79072805 3386 }
92f0c265 3387 if (SvVOK(sstr)) {
00b6aa41 3388 const MAGIC * const smg = mg_find(sstr,PERL_MAGIC_vstring);
ece467f9
JP
3389 sv_magic(dstr, NULL, PERL_MAGIC_vstring,
3390 smg->mg_ptr, smg->mg_len);
439cb1c4 3391 SvRMAGICAL_on(dstr);
7a5fa8a2 3392 }
79072805 3393 }
8990e307 3394 else if (sflags & SVp_IOK) {
3332b3c1
JH
3395 if (sflags & SVf_IOK)
3396 (void)SvIOK_only(dstr);
3397 else {
9cbac4c7
DM
3398 (void)SvOK_off(dstr);
3399 (void)SvIOKp_on(dstr);
3332b3c1
JH
3400 }
3401 /* XXXX Do we want to set IsUV for IV(ROK)? Be extra safe... */
2b1c7e3e 3402 if (sflags & SVf_IVisUV)
25da4f38 3403 SvIsUV_on(dstr);
45977657 3404 SvIV_set(dstr, SvIVX(sstr));
3332b3c1
JH
3405 if (sflags & SVp_NOK) {
3406 if (sflags & SVf_NOK)
3407 (void)SvNOK_on(dstr);
3408 else
3409 (void)SvNOKp_on(dstr);
9d6ce603 3410 SvNV_set(dstr, SvNVX(sstr));
3332b3c1
JH
3411 }
3412 }
3413 else if (sflags & SVp_NOK) {
3414 if (sflags & SVf_NOK)
3415 (void)SvNOK_only(dstr);
3416 else {
9cbac4c7 3417 (void)SvOK_off(dstr);
3332b3c1
JH
3418 SvNOKp_on(dstr);
3419 }
9d6ce603 3420 SvNV_set(dstr, SvNVX(sstr));
79072805
LW
3421 }
3422 else {
20408e3c 3423 if (dtype == SVt_PVGV) {
e476b1b5 3424 if (ckWARN(WARN_MISC))
9014280d 3425 Perl_warner(aTHX_ packWARN(WARN_MISC), "Undefined value assigned to typeglob");
20408e3c
GS
3426 }
3427 else
3428 (void)SvOK_off(dstr);
a0d0e21e 3429 }
27c9684d
AP
3430 if (SvTAINTED(sstr))
3431 SvTAINT(dstr);
79072805
LW
3432}
3433
954c1994
GS
3434/*
3435=for apidoc sv_setsv_mg
3436
3437Like C<sv_setsv>, but also handles 'set' magic.
3438
3439=cut
3440*/
3441
79072805 3442void
864dbfa3 3443Perl_sv_setsv_mg(pTHX_ SV *dstr, register SV *sstr)
ef50df4b
GS
3444{
3445 sv_setsv(dstr,sstr);
3446 SvSETMAGIC(dstr);
3447}
3448
f8c7b90f 3449#ifdef PERL_OLD_COPY_ON_WRITE
ed252734
NC
3450SV *
3451Perl_sv_setsv_cow(pTHX_ SV *dstr, SV *sstr)
3452{
3453 STRLEN cur = SvCUR(sstr);
3454 STRLEN len = SvLEN(sstr);
3455 register char *new_pv;
3456
3457 if (DEBUG_C_TEST) {
3458 PerlIO_printf(Perl_debug_log, "Fast copy on write: %p -> %p\n",
3459 sstr, dstr);
3460 sv_dump(sstr);
3461 if (dstr)
3462 sv_dump(dstr);
3463 }
3464
3465 if (dstr) {
3466 if (SvTHINKFIRST(dstr))
3467 sv_force_normal_flags(dstr, SV_COW_DROP_PV);
3f7c398e
SP
3468 else if (SvPVX_const(dstr))
3469 Safefree(SvPVX_const(dstr));
ed252734
NC
3470 }
3471 else
3472 new_SV(dstr);
862a34c6 3473 SvUPGRADE(dstr, SVt_PVIV);
ed252734
NC
3474
3475 assert (SvPOK(sstr));
3476 assert (SvPOKp(sstr));
3477 assert (!SvIOK(sstr));
3478 assert (!SvIOKp(sstr));
3479 assert (!SvNOK(sstr));
3480 assert (!SvNOKp(sstr));
3481
3482 if (SvIsCOW(sstr)) {
3483
3484 if (SvLEN(sstr) == 0) {
3485 /* source is a COW shared hash key. */
ed252734
NC
3486 DEBUG_C(PerlIO_printf(Perl_debug_log,
3487 "Fast copy on write: Sharing hash\n"));
d1db91c6 3488 new_pv = HEK_KEY(share_hek_hek(SvSHARED_HEK_FROM_PV(SvPVX_const(sstr))));
ed252734
NC
3489 goto common_exit;
3490 }
3491 SV_COW_NEXT_SV_SET(dstr, SV_COW_NEXT_SV(sstr));
3492 } else {
3493 assert ((SvFLAGS(sstr) & CAN_COW_MASK) == CAN_COW_FLAGS);
862a34c6 3494 SvUPGRADE(sstr, SVt_PVIV);
ed252734
NC
3495 SvREADONLY_on(sstr);
3496 SvFAKE_on(sstr);
3497 DEBUG_C(PerlIO_printf(Perl_debug_log,
3498 "Fast copy on write: Converting sstr to COW\n"));
3499 SV_COW_NEXT_SV_SET(dstr, sstr);
3500 }
3501 SV_COW_NEXT_SV_SET(sstr, dstr);
940132f3 3502 new_pv = SvPVX_mutable(sstr);
ed252734
NC
3503
3504 common_exit:
3505 SvPV_set(dstr, new_pv);
3506 SvFLAGS(dstr) = (SVt_PVIV|SVf_POK|SVp_POK|SVf_FAKE|SVf_READONLY);
3507 if (SvUTF8(sstr))
3508 SvUTF8_on(dstr);
87a1ef3d
SP
3509 SvLEN_set(dstr, len);
3510 SvCUR_set(dstr, cur);
ed252734
NC
3511 if (DEBUG_C_TEST) {
3512 sv_dump(dstr);
3513 }
3514 return dstr;
3515}
3516#endif
3517
954c1994
GS
3518/*
3519=for apidoc sv_setpvn
3520
3521Copies a string into an SV. The C<len> parameter indicates the number of
9e09f5f2
MHM
3522bytes to be copied. If the C<ptr> argument is NULL the SV will become
3523undefined. Does not handle 'set' magic. See C<sv_setpvn_mg>.
954c1994
GS
3524
3525=cut
3526*/
3527
ef50df4b 3528void
864dbfa3 3529Perl_sv_setpvn(pTHX_ register SV *sv, register const char *ptr, register STRLEN len)
79072805 3530{
c6f8c383 3531 register char *dptr;
22c522df 3532
765f542d 3533 SV_CHECK_THINKFIRST_COW_DROP(sv);
463ee0b2 3534 if (!ptr) {
a0d0e21e 3535 (void)SvOK_off(sv);
463ee0b2
LW
3536 return;
3537 }
22c522df
JH
3538 else {
3539 /* len is STRLEN which is unsigned, need to copy to signed */
a3b680e6 3540 const IV iv = len;
9c5ffd7c
JH
3541 if (iv < 0)
3542 Perl_croak(aTHX_ "panic: sv_setpvn called with negative strlen");
22c522df 3543 }
862a34c6 3544 SvUPGRADE(sv, SVt_PV);
c6f8c383 3545
5902b6a9 3546 dptr = SvGROW(sv, len + 1);
c6f8c383
GA
3547 Move(ptr,dptr,len,char);
3548 dptr[len] = '\0';
79072805 3549 SvCUR_set(sv, len);
1aa99e6b 3550 (void)SvPOK_only_UTF8(sv); /* validate pointer */
463ee0b2 3551 SvTAINT(sv);
79072805
LW
3552}
3553
954c1994
GS
3554/*
3555=for apidoc sv_setpvn_mg
3556
3557Like C<sv_setpvn>, but also handles 'set' magic.
3558
3559=cut
3560*/
3561
79072805 3562void
864dbfa3 3563Perl_sv_setpvn_mg(pTHX_ register SV *sv, register const char *ptr, register STRLEN len)
ef50df4b
GS
3564{
3565 sv_setpvn(sv,ptr,len);
3566 SvSETMAGIC(sv);
3567}
3568
954c1994
GS
3569/*
3570=for apidoc sv_setpv
3571
3572Copies a string into an SV. The string must be null-terminated. Does not
3573handle 'set' magic. See C<sv_setpv_mg>.
3574
3575=cut
3576*/
3577
ef50df4b 3578void
864dbfa3 3579Perl_sv_setpv(pTHX_ register SV *sv, register const char *ptr)
79072805
LW
3580{
3581 register STRLEN len;
3582
765f542d 3583 SV_CHECK_THINKFIRST_COW_DROP(sv);
463ee0b2 3584 if (!ptr) {
a0d0e21e 3585 (void)SvOK_off(sv);
463ee0b2
LW
3586 return;
3587 }
79072805 3588 len = strlen(ptr);
862a34c6 3589 SvUPGRADE(sv, SVt_PV);
c6f8c383 3590
79072805 3591 SvGROW(sv, len + 1);
463ee0b2 3592 Move(ptr,SvPVX(sv),len+1,char);
79072805 3593 SvCUR_set(sv, len);
1aa99e6b 3594 (void)SvPOK_only_UTF8(sv); /* validate pointer */
463ee0b2
LW
3595 SvTAINT(sv);
3596}
3597
954c1994
GS
3598/*
3599=for apidoc sv_setpv_mg
3600
3601Like C<sv_setpv>, but also handles 'set' magic.
3602
3603=cut
3604*/
3605
463ee0b2 3606void
864dbfa3 3607Perl_sv_setpv_mg(pTHX_ register SV *sv, register const char *ptr)
ef50df4b
GS
3608{
3609 sv_setpv(sv,ptr);
3610 SvSETMAGIC(sv);
3611}
3612
954c1994
GS
3613/*
3614=for apidoc sv_usepvn
3615
3616Tells an SV to use C<ptr> to find its string value. Normally the string is
1c846c1f 3617stored inside the SV but sv_usepvn allows the SV to use an outside string.
954c1994
GS
3618The C<ptr> should point to memory that was allocated by C<malloc>. The
3619string length, C<len>, must be supplied. This function will realloc the
3620memory pointed to by C<ptr>, so that pointer should not be freed or used by
3621the programmer after giving it to sv_usepvn. Does not handle 'set' magic.
3622See C<sv_usepvn_mg>.
3623
3624=cut
3625*/
3626
ef50df4b 3627void
864dbfa3 3628Perl_sv_usepvn(pTHX_ register SV *sv, register char *ptr, register STRLEN len)
463ee0b2 3629{
1936d2a7 3630 STRLEN allocate;
765f542d 3631 SV_CHECK_THINKFIRST_COW_DROP(sv);
862a34c6 3632 SvUPGRADE(sv, SVt_PV);
463ee0b2 3633 if (!ptr) {
a0d0e21e 3634 (void)SvOK_off(sv);
463ee0b2
LW
3635 return;
3636 }
3f7c398e 3637 if (SvPVX_const(sv))
8bd4d4c5 3638 SvPV_free(sv);
1936d2a7
NC
3639
3640 allocate = PERL_STRLEN_ROUNDUP(len + 1);
7a9b70e9 3641 ptr = saferealloc (ptr, allocate);
f880fe2f 3642 SvPV_set(sv, ptr);
463ee0b2 3643 SvCUR_set(sv, len);
1936d2a7 3644 SvLEN_set(sv, allocate);
463ee0b2 3645 *SvEND(sv) = '\0';
1aa99e6b 3646 (void)SvPOK_only_UTF8(sv); /* validate pointer */
463ee0b2 3647 SvTAINT(sv);
79072805
LW
3648}
3649
954c1994
GS
3650/*
3651=for apidoc sv_usepvn_mg
3652
3653Like C<sv_usepvn>, but also handles 'set' magic.
3654
3655=cut
3656*/
3657
ef50df4b 3658void
864dbfa3 3659Perl_sv_usepvn_mg(pTHX_ register SV *sv, register char *ptr, register STRLEN len)
ef50df4b 3660{
51c1089b 3661 sv_usepvn(sv,ptr,len);
ef50df4b
GS
3662 SvSETMAGIC(sv);
3663}
3664
f8c7b90f 3665#ifdef PERL_OLD_COPY_ON_WRITE
765f542d
NC
3666/* Need to do this *after* making the SV normal, as we need the buffer
3667 pointer to remain valid until after we've copied it. If we let go too early,
3668 another thread could invalidate it by unsharing last of the same hash key
3669 (which it can do by means other than releasing copy-on-write Svs)
3670 or by changing the other copy-on-write SVs in the loop. */
3671STATIC void
bdd68bc3 3672S_sv_release_COW(pTHX_ register SV *sv, const char *pvx, STRLEN len, SV *after)
765f542d
NC
3673{
3674 if (len) { /* this SV was SvIsCOW_normal(sv) */
3675 /* we need to find the SV pointing to us. */
aec46f14 3676 SV * const current = SV_COW_NEXT_SV(after);
7a5fa8a2 3677
765f542d
NC
3678 if (current == sv) {
3679 /* The SV we point to points back to us (there were only two of us
3680 in the loop.)
3681 Hence other SV is no longer copy on write either. */
3682 SvFAKE_off(after);
3683 SvREADONLY_off(after);
3684 } else {
3685 /* We need to follow the pointers around the loop. */
3686 SV *next;
3687 while ((next = SV_COW_NEXT_SV(current)) != sv) {
3688 assert (next);
3689 current = next;
3690 /* don't loop forever if the structure is bust, and we have
3691 a pointer into a closed loop. */
3692 assert (current != after);
3f7c398e 3693 assert (SvPVX_const(current) == pvx);
765f542d
NC
3694 }
3695 /* Make the SV before us point to the SV after us. */
a29f6d03 3696 SV_COW_NEXT_SV_SET(current, after);
765f542d
NC
3697 }
3698 } else {
bdd68bc3 3699 unshare_hek(SvSHARED_HEK_FROM_PV(pvx));
765f542d
NC
3700 }
3701}
3702
3703int
3704Perl_sv_release_IVX(pTHX_ register SV *sv)
3705{
3706 if (SvIsCOW(sv))
3707 sv_force_normal_flags(sv, 0);
0c34ef67
MHM
3708 SvOOK_off(sv);
3709 return 0;
765f542d
NC
3710}
3711#endif
645c22ef
DM
3712/*
3713=for apidoc sv_force_normal_flags
3714
3715Undo various types of fakery on an SV: if the PV is a shared string, make
3716a private copy; if we're a ref, stop refing; if we're a glob, downgrade to
765f542d
NC
3717an xpvmg; if we're a copy-on-write scalar, this is the on-write time when
3718we do the copy, and is also used locally. If C<SV_COW_DROP_PV> is set
3719then a copy-on-write scalar drops its PV buffer (if any) and becomes
3720SvPOK_off rather than making a copy. (Used where this scalar is about to be
d3050d9d 3721set to some other value.) In addition, the C<flags> parameter gets passed to
765f542d
NC
3722C<sv_unref_flags()> when unrefing. C<sv_force_normal> calls this function
3723with flags set to 0.
645c22ef
DM
3724
3725=cut
3726*/
3727
6fc92669 3728void
840a7b70 3729Perl_sv_force_normal_flags(pTHX_ register SV *sv, U32 flags)
0f15f207 3730{
f8c7b90f 3731#ifdef PERL_OLD_COPY_ON_WRITE
765f542d
NC
3732 if (SvREADONLY(sv)) {
3733 /* At this point I believe I should acquire a global SV mutex. */
3734 if (SvFAKE(sv)) {
b64e5050 3735 const char * const pvx = SvPVX_const(sv);
a28509cc
AL
3736 const STRLEN len = SvLEN(sv);
3737 const STRLEN cur = SvCUR(sv);
a28509cc 3738 SV * const next = SV_COW_NEXT_SV(sv); /* next COW sv in the loop. */
46187eeb
NC
3739 if (DEBUG_C_TEST) {
3740 PerlIO_printf(Perl_debug_log,
3741 "Copy on write: Force normal %ld\n",
3742 (long) flags);
e419cbc5 3743 sv_dump(sv);
46187eeb 3744 }
765f542d
NC
3745 SvFAKE_off(sv);
3746 SvREADONLY_off(sv);
9f653bb5 3747 /* This SV doesn't own the buffer, so need to Newx() a new one: */
f880fe2f 3748 SvPV_set(sv, (char*)0);
87a1ef3d 3749 SvLEN_set(sv, 0);
765f542d
NC
3750 if (flags & SV_COW_DROP_PV) {
3751 /* OK, so we don't need to copy our buffer. */
3752 SvPOK_off(sv);
3753 } else {
3754 SvGROW(sv, cur + 1);
3755 Move(pvx,SvPVX(sv),cur,char);
87a1ef3d 3756 SvCUR_set(sv, cur);
765f542d
NC
3757 *SvEND(sv) = '\0';
3758 }
bdd68bc3 3759 sv_release_COW(sv, pvx, len, next);
46187eeb 3760 if (DEBUG_C_TEST) {
e419cbc5 3761 sv_dump(sv);
46187eeb 3762 }
765f542d 3763 }
923e4eb5 3764 else if (IN_PERL_RUNTIME)
765f542d
NC
3765 Perl_croak(aTHX_ PL_no_modify);
3766 /* At this point I believe that I can drop the global SV mutex. */
3767 }
3768#else
2213622d 3769 if (SvREADONLY(sv)) {
1c846c1f 3770 if (SvFAKE(sv)) {
b64e5050 3771 const char * const pvx = SvPVX_const(sv);
66a1b24b 3772 const STRLEN len = SvCUR(sv);
10bcdfd6
NC
3773 SvFAKE_off(sv);
3774 SvREADONLY_off(sv);
66a1b24b
AL
3775 SvPV_set(sv, Nullch);
3776 SvLEN_set(sv, 0);
1c846c1f 3777 SvGROW(sv, len + 1);
706aa1c9 3778 Move(pvx,SvPVX(sv),len,char);
1c846c1f 3779 *SvEND(sv) = '\0';
bdd68bc3 3780 unshare_hek(SvSHARED_HEK_FROM_PV(pvx));
1c846c1f 3781 }
923e4eb5 3782 else if (IN_PERL_RUNTIME)
cea2e8a9 3783 Perl_croak(aTHX_ PL_no_modify);
0f15f207 3784 }
765f542d 3785#endif
2213622d 3786 if (SvROK(sv))
840a7b70 3787 sv_unref_flags(sv, flags);
6fc92669
GS
3788 else if (SvFAKE(sv) && SvTYPE(sv) == SVt_PVGV)
3789 sv_unglob(sv);
0f15f207 3790}
1c846c1f 3791
645c22ef 3792/*
954c1994
GS
3793=for apidoc sv_chop
3794
1c846c1f 3795Efficient removal of characters from the beginning of the string buffer.
954c1994
GS
3796SvPOK(sv) must be true and the C<ptr> must be a pointer to somewhere inside
3797the string buffer. The C<ptr> becomes the first character of the adjusted
645c22ef 3798string. Uses the "OOK hack".
3f7c398e 3799Beware: after this function returns, C<ptr> and SvPVX_const(sv) may no longer
31869a79 3800refer to the same chunk of data.
954c1994
GS
3801
3802=cut
3803*/
3804
79072805 3805void
f54cb97a 3806Perl_sv_chop(pTHX_ register SV *sv, register const char *ptr)
79072805
LW
3807{
3808 register STRLEN delta;
a0d0e21e 3809 if (!ptr || !SvPOKp(sv))
79072805 3810 return;
3f7c398e 3811 delta = ptr - SvPVX_const(sv);
2213622d 3812 SV_CHECK_THINKFIRST(sv);
79072805
LW
3813 if (SvTYPE(sv) < SVt_PVIV)
3814 sv_upgrade(sv,SVt_PVIV);
3815
3816 if (!SvOOK(sv)) {
50483b2c 3817 if (!SvLEN(sv)) { /* make copy of shared string */
3f7c398e 3818 const char *pvx = SvPVX_const(sv);
a28509cc 3819 const STRLEN len = SvCUR(sv);
50483b2c 3820 SvGROW(sv, len + 1);
706aa1c9 3821 Move(pvx,SvPVX(sv),len,char);
50483b2c
JD
3822 *SvEND(sv) = '\0';
3823 }
45977657 3824 SvIV_set(sv, 0);
a4bfb290
AB
3825 /* Same SvOOK_on but SvOOK_on does a SvIOK_off
3826 and we do that anyway inside the SvNIOK_off
3827 */
7a5fa8a2 3828 SvFLAGS(sv) |= SVf_OOK;
79072805 3829 }
a4bfb290 3830 SvNIOK_off(sv);
b162af07
SP
3831 SvLEN_set(sv, SvLEN(sv) - delta);
3832 SvCUR_set(sv, SvCUR(sv) - delta);
f880fe2f 3833 SvPV_set(sv, SvPVX(sv) + delta);
45977657 3834 SvIV_set(sv, SvIVX(sv) + delta);
79072805
LW
3835}
3836
954c1994
GS
3837/*
3838=for apidoc sv_catpvn
3839
3840Concatenates the string onto the end of the string which is in the SV. The
1e54db1a
JH
3841C<len> indicates number of bytes to copy. If the SV has the UTF-8
3842status set, then the bytes appended should be valid UTF-8.
d5ce4a7c 3843Handles 'get' magic, but not 'set' magic. See C<sv_catpvn_mg>.
954c1994 3844
8d6d96c1
HS
3845=for apidoc sv_catpvn_flags
3846
3847Concatenates the string onto the end of the string which is in the SV. The
1e54db1a
JH
3848C<len> indicates number of bytes to copy. If the SV has the UTF-8
3849status set, then the bytes appended should be valid UTF-8.
8d6d96c1
HS
3850If C<flags> has C<SV_GMAGIC> bit set, will C<mg_get> on C<dsv> if
3851appropriate, else not. C<sv_catpvn> and C<sv_catpvn_nomg> are implemented
3852in terms of this function.
3853
3854=cut
3855*/
3856
3857void
3858Perl_sv_catpvn_flags(pTHX_ register SV *dsv, register const char *sstr, register STRLEN slen, I32 flags)
3859{
3860 STRLEN dlen;
fabdb6c0 3861 const char * const dstr = SvPV_force_flags(dsv, dlen, flags);
8d6d96c1 3862
8d6d96c1
HS
3863 SvGROW(dsv, dlen + slen + 1);
3864 if (sstr == dstr)
3f7c398e 3865 sstr = SvPVX_const(dsv);
8d6d96c1 3866 Move(sstr, SvPVX(dsv) + dlen, slen, char);
b162af07 3867 SvCUR_set(dsv, SvCUR(dsv) + slen);
8d6d96c1
HS
3868 *SvEND(dsv) = '\0';
3869 (void)SvPOK_only_UTF8(dsv); /* validate pointer */
3870 SvTAINT(dsv);
bddd5118
NC
3871 if (flags & SV_SMAGIC)
3872 SvSETMAGIC(dsv);
79072805
LW
3873}
3874
954c1994 3875/*
954c1994
GS
3876=for apidoc sv_catsv
3877
13e8c8e3
JH
3878Concatenates the string from SV C<ssv> onto the end of the string in
3879SV C<dsv>. Modifies C<dsv> but not C<ssv>. Handles 'get' magic, but
3880not 'set' magic. See C<sv_catsv_mg>.
954c1994 3881
8d6d96c1
HS
3882=for apidoc sv_catsv_flags
3883
3884Concatenates the string from SV C<ssv> onto the end of the string in
3885SV C<dsv>. Modifies C<dsv> but not C<ssv>. If C<flags> has C<SV_GMAGIC>
3886bit set, will C<mg_get> on the SVs if appropriate, else not. C<sv_catsv>
3887and C<sv_catsv_nomg> are implemented in terms of this function.
3888
3889=cut */
3890
ef50df4b 3891void
8d6d96c1 3892Perl_sv_catsv_flags(pTHX_ SV *dsv, register SV *ssv, I32 flags)
79072805 3893{
bddd5118 3894 if (ssv) {
00b6aa41
AL
3895 STRLEN slen;
3896 const char *spv = SvPV_const(ssv, slen);
3897 if (spv) {
bddd5118
NC
3898 /* sutf8 and dutf8 were type bool, but under USE_ITHREADS,
3899 gcc version 2.95.2 20000220 (Debian GNU/Linux) for
3900 Linux xxx 2.2.17 on sparc64 with gcc -O2, we erroneously
3901 get dutf8 = 0x20000000, (i.e. SVf_UTF8) even though
3902 dsv->sv_flags doesn't have that bit set.
4fd84b44 3903 Andy Dougherty 12 Oct 2001
bddd5118
NC
3904 */
3905 const I32 sutf8 = DO_UTF8(ssv);
3906 I32 dutf8;
13e8c8e3 3907
bddd5118
NC
3908 if (SvGMAGICAL(dsv) && (flags & SV_GMAGIC))
3909 mg_get(dsv);
3910 dutf8 = DO_UTF8(dsv);
8d6d96c1 3911
bddd5118
NC
3912 if (dutf8 != sutf8) {
3913 if (dutf8) {
3914 /* Not modifying source SV, so taking a temporary copy. */
00b6aa41 3915 SV* const csv = sv_2mortal(newSVpvn(spv, slen));
13e8c8e3 3916
bddd5118
NC
3917 sv_utf8_upgrade(csv);
3918 spv = SvPV_const(csv, slen);
3919 }
3920 else
3921 sv_utf8_upgrade_nomg(dsv);
13e8c8e3 3922 }
bddd5118 3923 sv_catpvn_nomg(dsv, spv, slen);
e84ff256 3924 }
560a288e 3925 }
bddd5118
NC
3926 if (flags & SV_SMAGIC)
3927 SvSETMAGIC(dsv);
79072805
LW
3928}
3929
954c1994 3930/*
954c1994
GS
3931=for apidoc sv_catpv
3932
3933Concatenates the string onto the end of the string which is in the SV.
1e54db1a
JH
3934If the SV has the UTF-8 status set, then the bytes appended should be
3935valid UTF-8. Handles 'get' magic, but not 'set' magic. See C<sv_catpv_mg>.
954c1994 3936
d5ce4a7c 3937=cut */
954c1994 3938
ef50df4b 3939void
0c981600 3940Perl_sv_catpv(pTHX_ register SV *sv, register const char *ptr)
79072805
LW
3941{
3942 register STRLEN len;
463ee0b2 3943 STRLEN tlen;
748a9306 3944 char *junk;
79072805 3945
0c981600 3946 if (!ptr)
79072805 3947 return;
748a9306 3948 junk = SvPV_force(sv, tlen);
0c981600 3949 len = strlen(ptr);
463ee0b2 3950 SvGROW(sv, tlen + len + 1);
0c981600 3951 if (ptr == junk)
3f7c398e 3952 ptr = SvPVX_const(sv);
0c981600 3953 Move(ptr,SvPVX(sv)+tlen,len+1,char);
b162af07 3954 SvCUR_set(sv, SvCUR(sv) + len);
d41ff1b8 3955 (void)SvPOK_only_UTF8(sv); /* validate pointer */
463ee0b2 3956 SvTAINT(sv);
79072805
LW
3957}
3958
954c1994
GS
3959/*
3960=for apidoc sv_catpv_mg
3961
3962Like C<sv_catpv>, but also handles 'set' magic.
3963
3964=cut
3965*/
3966
ef50df4b 3967void
0c981600 3968Perl_sv_catpv_mg(pTHX_ register SV *sv, register const char *ptr)
ef50df4b 3969{
0c981600 3970 sv_catpv(sv,ptr);
ef50df4b
GS
3971 SvSETMAGIC(sv);
3972}
3973
645c22ef
DM
3974/*
3975=for apidoc newSV
3976
3977Create a new null SV, or if len > 0, create a new empty SVt_PV type SV
3978with an initial PV allocation of len+1. Normally accessed via the C<NEWSV>
3979macro.
3980
3981=cut
3982*/
3983
79072805 3984SV *
864dbfa3 3985Perl_newSV(pTHX_ STRLEN len)
79072805
LW
3986{
3987 register SV *sv;
1c846c1f 3988
4561caa4 3989 new_SV(sv);
79072805
LW
3990 if (len) {
3991 sv_upgrade(sv, SVt_PV);
3992 SvGROW(sv, len + 1);
3993 }
3994 return sv;
3995}
954c1994 3996/*
92110913 3997=for apidoc sv_magicext
954c1994 3998
68795e93 3999Adds magic to an SV, upgrading it if necessary. Applies the
2d8d5d5a 4000supplied vtable and returns a pointer to the magic added.
92110913 4001
2d8d5d5a
SH
4002Note that C<sv_magicext> will allow things that C<sv_magic> will not.
4003In particular, you can add magic to SvREADONLY SVs, and add more than
4004one instance of the same 'how'.
645c22ef 4005
2d8d5d5a
SH
4006If C<namlen> is greater than zero then a C<savepvn> I<copy> of C<name> is
4007stored, if C<namlen> is zero then C<name> is stored as-is and - as another
4008special case - if C<(name && namlen == HEf_SVKEY)> then C<name> is assumed
4009to contain an C<SV*> and is stored as-is with its REFCNT incremented.
92110913 4010
2d8d5d5a 4011(This is now used as a subroutine by C<sv_magic>.)
954c1994
GS
4012
4013=cut
4014*/
92110913 4015MAGIC *
e1ec3a88 4016Perl_sv_magicext(pTHX_ SV* sv, SV* obj, int how, const MGVTBL *vtable,
92110913 4017 const char* name, I32 namlen)
79072805
LW
4018{
4019 MAGIC* mg;
68795e93 4020
92110913 4021 if (SvTYPE(sv) < SVt_PVMG) {
862a34c6 4022 SvUPGRADE(sv, SVt_PVMG);
463ee0b2 4023 }
a02a5408 4024 Newxz(mg, 1, MAGIC);
79072805 4025 mg->mg_moremagic = SvMAGIC(sv);
b162af07 4026 SvMAGIC_set(sv, mg);
75f9d97a 4027
05f95b08
SB
4028 /* Sometimes a magic contains a reference loop, where the sv and
4029 object refer to each other. To prevent a reference loop that
4030 would prevent such objects being freed, we look for such loops
4031 and if we find one we avoid incrementing the object refcount.
87f0b213
JH
4032
4033 Note we cannot do this to avoid self-tie loops as intervening RV must
b5ccf5f2 4034 have its REFCNT incremented to keep it in existence.
87f0b213
JH
4035
4036 */
14befaf4
DM
4037 if (!obj || obj == sv ||
4038 how == PERL_MAGIC_arylen ||
4039 how == PERL_MAGIC_qr ||
8d2f4536 4040 how == PERL_MAGIC_symtab ||
75f9d97a
JH
4041 (SvTYPE(obj) == SVt_PVGV &&
4042 (GvSV(obj) == sv || GvHV(obj) == (HV*)sv || GvAV(obj) == (AV*)sv ||
4043 GvCV(obj) == (CV*)sv || GvIOp(obj) == (IO*)sv ||
2628be26 4044 GvFORM(obj) == (CV*)sv)))
75f9d97a 4045 {
8990e307 4046 mg->mg_obj = obj;
75f9d97a 4047 }
85e6fe83 4048 else {
8990e307 4049 mg->mg_obj = SvREFCNT_inc(obj);
85e6fe83
LW
4050 mg->mg_flags |= MGf_REFCOUNTED;
4051 }
b5ccf5f2
YST
4052
4053 /* Normal self-ties simply pass a null object, and instead of
4054 using mg_obj directly, use the SvTIED_obj macro to produce a
4055 new RV as needed. For glob "self-ties", we are tieing the PVIO
4056 with an RV obj pointing to the glob containing the PVIO. In
4057 this case, to avoid a reference loop, we need to weaken the
4058 reference.
4059 */
4060
4061 if (how == PERL_MAGIC_tiedscalar && SvTYPE(sv) == SVt_PVIO &&
4062 obj && SvROK(obj) && GvIO(SvRV(obj)) == (IO*)sv)
4063 {
4064 sv_rvweaken(obj);
4065 }
4066
79072805 4067 mg->mg_type = how;
565764a8 4068 mg->mg_len = namlen;
9cbac4c7 4069 if (name) {
92110913 4070 if (namlen > 0)
1edc1566 4071 mg->mg_ptr = savepvn(name, namlen);
c6ee37c5 4072 else if (namlen == HEf_SVKEY)
1edc1566 4073 mg->mg_ptr = (char*)SvREFCNT_inc((SV*)name);
68795e93 4074 else
92110913 4075 mg->mg_ptr = (char *) name;
9cbac4c7 4076 }
92110913 4077 mg->mg_virtual = vtable;
68795e93 4078
92110913
NIS
4079 mg_magical(sv);
4080 if (SvGMAGICAL(sv))
4081 SvFLAGS(sv) &= ~(SVf_IOK|SVf_NOK|SVf_POK);
4082 return mg;
4083}
4084
4085/*
4086=for apidoc sv_magic
1c846c1f 4087
92110913
NIS
4088Adds magic to an SV. First upgrades C<sv> to type C<SVt_PVMG> if necessary,
4089then adds a new magic item of type C<how> to the head of the magic list.
4090
2d8d5d5a
SH
4091See C<sv_magicext> (which C<sv_magic> now calls) for a description of the
4092handling of the C<name> and C<namlen> arguments.
4093
4509d3fb
SB
4094You need to use C<sv_magicext> to add magic to SvREADONLY SVs and also
4095to add more than one instance of the same 'how'.
4096
92110913
NIS
4097=cut
4098*/
4099
4100void
4101Perl_sv_magic(pTHX_ register SV *sv, SV *obj, int how, const char *name, I32 namlen)
68795e93 4102{
aec46f14 4103 const MGVTBL *vtable;
92110913 4104 MAGIC* mg;
92110913 4105
f8c7b90f 4106#ifdef PERL_OLD_COPY_ON_WRITE
765f542d
NC
4107 if (SvIsCOW(sv))
4108 sv_force_normal_flags(sv, 0);
4109#endif
92110913 4110 if (SvREADONLY(sv)) {
d8084ca5
DM
4111 if (
4112 /* its okay to attach magic to shared strings; the subsequent
4113 * upgrade to PVMG will unshare the string */
4114 !(SvFAKE(sv) && SvTYPE(sv) < SVt_PVMG)
4115
4116 && IN_PERL_RUNTIME
92110913
NIS
4117 && how != PERL_MAGIC_regex_global
4118 && how != PERL_MAGIC_bm
4119 && how != PERL_MAGIC_fm
4120 && how != PERL_MAGIC_sv
e6469971 4121 && how != PERL_MAGIC_backref
92110913
NIS
4122 )
4123 {
4124 Perl_croak(aTHX_ PL_no_modify);
4125 }
4126 }
4127 if (SvMAGICAL(sv) || (how == PERL_MAGIC_taint && SvTYPE(sv) >= SVt_PVMG)) {
4128 if (SvMAGIC(sv) && (mg = mg_find(sv, how))) {
68795e93
NIS
4129 /* sv_magic() refuses to add a magic of the same 'how' as an
4130 existing one
92110913
NIS
4131 */
4132 if (how == PERL_MAGIC_taint)
4133 mg->mg_len |= 1;
4134 return;
4135 }
4136 }
68795e93 4137
79072805 4138 switch (how) {
14befaf4 4139 case PERL_MAGIC_sv:
92110913 4140 vtable = &PL_vtbl_sv;
79072805 4141 break;
14befaf4 4142 case PERL_MAGIC_overload:
92110913 4143 vtable = &PL_vtbl_amagic;
a0d0e21e 4144 break;
14befaf4 4145 case PERL_MAGIC_overload_elem:
92110913 4146 vtable = &PL_vtbl_amagicelem;
a0d0e21e 4147 break;
14befaf4 4148 case PERL_MAGIC_overload_table:
92110913 4149 vtable = &PL_vtbl_ovrld;
a0d0e21e 4150 break;
14befaf4 4151 case PERL_MAGIC_bm:
92110913 4152 vtable = &PL_vtbl_bm;
79072805 4153 break;
14befaf4 4154 case PERL_MAGIC_regdata:
92110913 4155 vtable = &PL_vtbl_regdata;
6cef1e77 4156 break;
14befaf4 4157 case PERL_MAGIC_regdatum:
92110913 4158 vtable = &PL_vtbl_regdatum;
6cef1e77 4159 break;
14befaf4 4160 case PERL_MAGIC_env:
92110913 4161 vtable = &PL_vtbl_env;
79072805 4162 break;
14befaf4 4163 case PERL_MAGIC_fm:
92110913 4164 vtable = &PL_vtbl_fm;
55497cff 4165 break;
14befaf4 4166 case PERL_MAGIC_envelem:
92110913 4167 vtable = &PL_vtbl_envelem;
79072805 4168 break;
14befaf4 4169 case PERL_MAGIC_regex_global:
92110913 4170 vtable = &PL_vtbl_mglob;
93a17b20 4171 break;
14befaf4 4172 case PERL_MAGIC_isa:
92110913 4173 vtable = &PL_vtbl_isa;
463ee0b2 4174 break;
14befaf4 4175 case PERL_MAGIC_isaelem:
92110913 4176 vtable = &PL_vtbl_isaelem;
463ee0b2 4177 break;
14befaf4 4178 case PERL_MAGIC_nkeys:
92110913 4179 vtable = &PL_vtbl_nkeys;
16660edb 4180 break;
14befaf4 4181 case PERL_MAGIC_dbfile:
aec46f14 4182 vtable = NULL;
93a17b20 4183 break;
14befaf4 4184 case PERL_MAGIC_dbline:
92110913 4185 vtable = &PL_vtbl_dbline;
79072805 4186 break;
36477c24 4187#ifdef USE_LOCALE_COLLATE
14befaf4 4188 case PERL_MAGIC_collxfrm:
92110913 4189 vtable = &PL_vtbl_collxfrm;
bbce6d69 4190 break;
36477c24 4191#endif /* USE_LOCALE_COLLATE */
14befaf4 4192 case PERL_MAGIC_tied:
92110913 4193 vtable = &PL_vtbl_pack;
463ee0b2 4194 break;
14befaf4
DM
4195 case PERL_MAGIC_tiedelem:
4196 case PERL_MAGIC_tiedscalar:
92110913 4197 vtable = &PL_vtbl_packelem;
463ee0b2 4198 break;
14befaf4 4199 case PERL_MAGIC_qr:
92110913 4200 vtable = &PL_vtbl_regexp;
c277df42 4201 break;
14befaf4 4202 case PERL_MAGIC_sig:
92110913 4203 vtable = &PL_vtbl_sig;
79072805 4204 break;
14befaf4 4205 case PERL_MAGIC_sigelem:
92110913 4206 vtable = &PL_vtbl_sigelem;
79072805 4207 break;
14befaf4 4208 case PERL_MAGIC_taint:
92110913 4209 vtable = &PL_vtbl_taint;
463ee0b2 4210 break;
14befaf4 4211 case PERL_MAGIC_uvar:
92110913 4212 vtable = &PL_vtbl_uvar;
79072805 4213 break;
14befaf4 4214 case PERL_MAGIC_vec:
92110913 4215 vtable = &PL_vtbl_vec;
79072805 4216 break;
a3874608 4217 case PERL_MAGIC_arylen_p:
bfcb3514 4218 case PERL_MAGIC_rhash:
8d2f4536 4219 case PERL_MAGIC_symtab:
ece467f9 4220 case PERL_MAGIC_vstring:
aec46f14 4221 vtable = NULL;
ece467f9 4222 break;
7e8c5dac
HS
4223 case PERL_MAGIC_utf8:
4224 vtable = &PL_vtbl_utf8;
4225 break;
14befaf4 4226 case PERL_MAGIC_substr:
92110913 4227 vtable = &PL_vtbl_substr;
79072805 4228 break;
14befaf4 4229 case PERL_MAGIC_defelem:
92110913 4230 vtable = &PL_vtbl_defelem;
5f05dabc 4231 break;
14befaf4 4232 case PERL_MAGIC_glob:
92110913 4233 vtable = &PL_vtbl_glob;
79072805 4234 break;
14befaf4 4235 case PERL_MAGIC_arylen:
92110913 4236 vtable = &PL_vtbl_arylen;
79072805 4237 break;
14befaf4 4238 case PERL_MAGIC_pos:
92110913 4239 vtable = &PL_vtbl_pos;
a0d0e21e 4240 break;
14befaf4 4241 case PERL_MAGIC_backref:
92110913 4242 vtable = &PL_vtbl_backref;
810b8aa5 4243 break;
14befaf4
DM
4244 case PERL_MAGIC_ext:
4245 /* Reserved for use by extensions not perl internals. */
4633a7c4
LW
4246 /* Useful for attaching extension internal data to perl vars. */
4247 /* Note that multiple extensions may clash if magical scalars */
4248 /* etc holding private data from one are passed to another. */
aec46f14 4249 vtable = NULL;
a0d0e21e 4250 break;
79072805 4251 default:
14befaf4 4252 Perl_croak(aTHX_ "Don't know how to handle magic of type \\%o", how);
463ee0b2 4253 }
68795e93 4254
92110913 4255 /* Rest of work is done else where */
aec46f14 4256 mg = sv_magicext(sv,obj,how,vtable,name,namlen);
68795e93 4257
92110913
NIS
4258 switch (how) {
4259 case PERL_MAGIC_taint:
4260 mg->mg_len = 1;
4261 break;
4262 case PERL_MAGIC_ext:
4263 case PERL_MAGIC_dbfile:
4264 SvRMAGICAL_on(sv);
4265 break;
4266 }
463ee0b2
LW
4267}
4268
c461cf8f
JH
4269/*
4270=for apidoc sv_unmagic
4271
645c22ef 4272Removes all magic of type C<type> from an SV.
c461cf8f
JH
4273
4274=cut
4275*/
4276
463ee0b2 4277int
864dbfa3 4278Perl_sv_unmagic(pTHX_ SV *sv, int type)
463ee0b2
LW
4279{
4280 MAGIC* mg;
4281 MAGIC** mgp;
91bba347 4282 if (SvTYPE(sv) < SVt_PVMG || !SvMAGIC(sv))
463ee0b2
LW
4283 return 0;
4284 mgp = &SvMAGIC(sv);
4285 for (mg = *mgp; mg; mg = *mgp) {
4286 if (mg->mg_type == type) {
e1ec3a88 4287 const MGVTBL* const vtbl = mg->mg_virtual;
463ee0b2 4288 *mgp = mg->mg_moremagic;
1d7c1841 4289 if (vtbl && vtbl->svt_free)
fc0dc3b3 4290 CALL_FPTR(vtbl->svt_free)(aTHX_ sv, mg);
14befaf4 4291 if (mg->mg_ptr && mg->mg_type != PERL_MAGIC_regex_global) {
92110913 4292 if (mg->mg_len > 0)
1edc1566 4293 Safefree(mg->mg_ptr);
565764a8 4294 else if (mg->mg_len == HEf_SVKEY)
1edc1566 4295 SvREFCNT_dec((SV*)mg->mg_ptr);
7e8c5dac
HS
4296 else if (mg->mg_type == PERL_MAGIC_utf8 && mg->mg_ptr)
4297 Safefree(mg->mg_ptr);
9cbac4c7 4298 }
a0d0e21e
LW
4299 if (mg->mg_flags & MGf_REFCOUNTED)
4300 SvREFCNT_dec(mg->mg_obj);
463ee0b2
LW
4301 Safefree(mg);
4302 }
4303 else
4304 mgp = &mg->mg_moremagic;
79072805 4305 }
91bba347 4306 if (!SvMAGIC(sv)) {
463ee0b2 4307 SvMAGICAL_off(sv);
06759ea0 4308 SvFLAGS(sv) |= (SvFLAGS(sv) & (SVp_NOK|SVp_POK)) >> PRIVSHIFT;
463ee0b2
LW
4309 }
4310
4311 return 0;
79072805
LW
4312}
4313
c461cf8f
JH
4314/*
4315=for apidoc sv_rvweaken
4316
645c22ef
DM
4317Weaken a reference: set the C<SvWEAKREF> flag on this RV; give the
4318referred-to SV C<PERL_MAGIC_backref> magic if it hasn't already; and
4319push a back-reference to this RV onto the array of backreferences
4320associated with that magic.
c461cf8f
JH
4321
4322=cut
4323*/
4324
810b8aa5 4325SV *
864dbfa3 4326Perl_sv_rvweaken(pTHX_ SV *sv)
810b8aa5
GS
4327{
4328 SV *tsv;
4329 if (!SvOK(sv)) /* let undefs pass */
4330 return sv;
4331 if (!SvROK(sv))
cea2e8a9 4332 Perl_croak(aTHX_ "Can't weaken a nonreference");
810b8aa5 4333 else if (SvWEAKREF(sv)) {
810b8aa5 4334 if (ckWARN(WARN_MISC))
9014280d 4335 Perl_warner(aTHX_ packWARN(WARN_MISC), "Reference is already weak");
810b8aa5
GS
4336 return sv;
4337 }
4338 tsv = SvRV(sv);
e15faf7d 4339 Perl_sv_add_backref(aTHX_ tsv, sv);
810b8aa5 4340 SvWEAKREF_on(sv);
1c846c1f 4341 SvREFCNT_dec(tsv);
810b8aa5
GS
4342 return sv;
4343}
4344
645c22ef
DM
4345/* Give tsv backref magic if it hasn't already got it, then push a
4346 * back-reference to sv onto the array associated with the backref magic.
4347 */
4348
e15faf7d
NC
4349void
4350Perl_sv_add_backref(pTHX_ SV *tsv, SV *sv)
810b8aa5
GS
4351{
4352 AV *av;
4353 MAGIC *mg;
14befaf4 4354 if (SvMAGICAL(tsv) && (mg = mg_find(tsv, PERL_MAGIC_backref)))
810b8aa5
GS
4355 av = (AV*)mg->mg_obj;
4356 else {
4357 av = newAV();
e3d99884 4358 AvREAL_off(av);
14befaf4 4359 sv_magic(tsv, (SV*)av, PERL_MAGIC_backref, NULL, 0);
d99b02a1
DM
4360 /* av now has a refcnt of 2, which avoids it getting freed
4361 * before us during global cleanup. The extra ref is removed
4362 * by magic_killbackrefs() when tsv is being freed */
810b8aa5 4363 }
d91d49e8 4364 if (AvFILLp(av) >= AvMAX(av)) {
d91d49e8
MM
4365 av_extend(av, AvFILLp(av)+1);
4366 }
4367 AvARRAY(av)[++AvFILLp(av)] = sv; /* av_push() */
810b8aa5
GS
4368}
4369
645c22ef
DM
4370/* delete a back-reference to ourselves from the backref magic associated
4371 * with the SV we point to.
4372 */
4373
1c846c1f 4374STATIC void
e15faf7d 4375S_sv_del_backref(pTHX_ SV *tsv, SV *sv)
810b8aa5
GS
4376{
4377 AV *av;
4378 SV **svp;
4379 I32 i;
c04a4dfe 4380 MAGIC *mg = NULL;
e15faf7d
NC
4381 if (!SvMAGICAL(tsv) || !(mg = mg_find(tsv, PERL_MAGIC_backref))) {
4382 if (PL_in_clean_all)
4383 return;
4384 }
14befaf4 4385 if (!SvMAGICAL(tsv) || !(mg = mg_find(tsv, PERL_MAGIC_backref)))
cea2e8a9 4386 Perl_croak(aTHX_ "panic: del_backref");
810b8aa5
GS
4387 av = (AV *)mg->mg_obj;
4388 svp = AvARRAY(av);
6a76db8b
NC
4389 /* We shouldn't be in here more than once, but for paranoia reasons lets
4390 not assume this. */
4391 for (i = AvFILLp(av); i >= 0; i--) {
4392 if (svp[i] == sv) {
4393 const SSize_t fill = AvFILLp(av);
4394 if (i != fill) {
4395 /* We weren't the last entry.
4396 An unordered list has this property that you can take the
4397 last element off the end to fill the hole, and it's still
4398 an unordered list :-)
4399 */
4400 svp[i] = svp[fill];
4401 }
4402 svp[fill] = Nullsv;
4403 AvFILLp(av) = fill - 1;
4404 }
4405 }
810b8aa5
GS
4406}
4407
954c1994
GS
4408/*
4409=for apidoc sv_insert
4410
4411Inserts a string at the specified offset/length within the SV. Similar to
4412the Perl substr() function.
4413
4414=cut
4415*/
4416
79072805 4417void
e1ec3a88 4418Perl_sv_insert(pTHX_ SV *bigstr, STRLEN offset, STRLEN len, const char *little, STRLEN littlelen)
79072805
LW
4419{
4420 register char *big;
4421 register char *mid;
4422 register char *midend;
4423 register char *bigend;
4424 register I32 i;
6ff81951 4425 STRLEN curlen;
1c846c1f 4426
79072805 4427
8990e307 4428 if (!bigstr)
cea2e8a9 4429 Perl_croak(aTHX_ "Can't modify non-existent substring");
6ff81951 4430 SvPV_force(bigstr, curlen);
60fa28ff 4431 (void)SvPOK_only_UTF8(bigstr);
6ff81951
GS
4432 if (offset + len > curlen) {
4433 SvGROW(bigstr, offset+len+1);
93524f2b 4434 Zero(SvPVX(bigstr)+curlen, offset+len-curlen, char);
6ff81951
GS
4435 SvCUR_set(bigstr, offset+len);
4436 }
79072805 4437
69b47968 4438 SvTAINT(bigstr);
79072805
LW
4439 i = littlelen - len;
4440 if (i > 0) { /* string might grow */
a0d0e21e 4441 big = SvGROW(bigstr, SvCUR(bigstr) + i + 1);
79072805
LW
4442 mid = big + offset + len;
4443 midend = bigend = big + SvCUR(bigstr);
4444 bigend += i;
4445 *bigend = '\0';
4446 while (midend > mid) /* shove everything down */
4447 *--bigend = *--midend;
4448 Move(little,big+offset,littlelen,char);
b162af07 4449 SvCUR_set(bigstr, SvCUR(bigstr) + i);
79072805
LW
4450 SvSETMAGIC(bigstr);
4451 return;
4452 }
4453 else if (i == 0) {
463ee0b2 4454 Move(little,SvPVX(bigstr)+offset,len,char);
79072805
LW
4455 SvSETMAGIC(bigstr);
4456 return;
4457 }
4458
463ee0b2 4459 big = SvPVX(bigstr);
79072805
LW
4460 mid = big + offset;
4461 midend = mid + len;
4462 bigend = big + SvCUR(bigstr);
4463
4464 if (midend > bigend)
cea2e8a9 4465 Perl_croak(aTHX_ "panic: sv_insert");
79072805
LW
4466
4467 if (mid - big > bigend - midend) { /* faster to shorten from end */
4468 if (littlelen) {
4469 Move(little, mid, littlelen,char);
4470 mid += littlelen;
4471 }
4472 i = bigend - midend;
4473 if (i > 0) {
4474 Move(midend, mid, i,char);
4475 mid += i;
4476 }
4477 *mid = '\0';
4478 SvCUR_set(bigstr, mid - big);
4479 }
155aba94 4480 else if ((i = mid - big)) { /* faster from front */
79072805
LW
4481 midend -= littlelen;
4482 mid = midend;
4483 sv_chop(bigstr,midend-i);
4484 big += i;
4485 while (i--)
4486 *--midend = *--big;
4487 if (littlelen)
4488 Move(little, mid, littlelen,char);
4489 }
4490 else if (littlelen) {
4491 midend -= littlelen;
4492 sv_chop(bigstr,midend);
4493 Move(little,midend,littlelen,char);
4494 }
4495 else {
4496 sv_chop(bigstr,midend);
4497 }
4498 SvSETMAGIC(bigstr);
4499}
4500
c461cf8f
JH
4501/*
4502=for apidoc sv_replace
4503
4504Make the first argument a copy of the second, then delete the original.
645c22ef
DM
4505The target SV physically takes over ownership of the body of the source SV
4506and inherits its flags; however, the target keeps any magic it owns,
4507and any magic in the source is discarded.
ff276b08 4508Note that this is a rather specialist SV copying operation; most of the
645c22ef 4509time you'll want to use C<sv_setsv> or one of its many macro front-ends.
c461cf8f
JH
4510
4511=cut
4512*/
79072805
LW
4513
4514void
864dbfa3 4515Perl_sv_replace(pTHX_ register SV *sv, register SV *nsv)
79072805 4516{
a3b680e6 4517 const U32 refcnt = SvREFCNT(sv);
765f542d 4518 SV_CHECK_THINKFIRST_COW_DROP(sv);
30e5c352 4519 if (SvREFCNT(nsv) != 1) {
7437becc 4520 Perl_croak(aTHX_ "panic: reference miscount on nsv in sv_replace() (%"
30e5c352
NC
4521 UVuf " != 1)", (UV) SvREFCNT(nsv));
4522 }
93a17b20 4523 if (SvMAGICAL(sv)) {
a0d0e21e
LW
4524 if (SvMAGICAL(nsv))
4525 mg_free(nsv);
4526 else
4527 sv_upgrade(nsv, SVt_PVMG);
b162af07 4528 SvMAGIC_set(nsv, SvMAGIC(sv));
a0d0e21e 4529 SvFLAGS(nsv) |= SvMAGICAL(sv);
93a17b20 4530 SvMAGICAL_off(sv);
b162af07 4531 SvMAGIC_set(sv, NULL);
93a17b20 4532 }
79072805
LW
4533 SvREFCNT(sv) = 0;
4534 sv_clear(sv);
477f5d66 4535 assert(!SvREFCNT(sv));
fd0854ff
DM
4536#ifdef DEBUG_LEAKING_SCALARS
4537 sv->sv_flags = nsv->sv_flags;
4538 sv->sv_any = nsv->sv_any;
4539 sv->sv_refcnt = nsv->sv_refcnt;
f34d0642 4540 sv->sv_u = nsv->sv_u;
fd0854ff 4541#else
79072805 4542 StructCopy(nsv,sv,SV);
fd0854ff 4543#endif
7b2c381c
NC
4544 /* Currently could join these into one piece of pointer arithmetic, but
4545 it would be unclear. */
4546 if(SvTYPE(sv) == SVt_IV)
4547 SvANY(sv)
339049b0 4548 = (XPVIV*)((char*)&(sv->sv_u.svu_iv) - STRUCT_OFFSET(XPVIV, xiv_iv));
7b2c381c 4549 else if (SvTYPE(sv) == SVt_RV) {
339049b0 4550 SvANY(sv) = &sv->sv_u.svu_rv;
7b2c381c
NC
4551 }
4552
fd0854ff 4553
f8c7b90f 4554#ifdef PERL_OLD_COPY_ON_WRITE
d3d0e6f1
NC
4555 if (SvIsCOW_normal(nsv)) {
4556 /* We need to follow the pointers around the loop to make the
4557 previous SV point to sv, rather than nsv. */
4558 SV *next;
4559 SV *current = nsv;
4560 while ((next = SV_COW_NEXT_SV(current)) != nsv) {
4561 assert(next);
4562 current = next;
3f7c398e 4563 assert(SvPVX_const(current) == SvPVX_const(nsv));
d3d0e6f1
NC
4564 }
4565 /* Make the SV before us point to the SV after us. */
4566 if (DEBUG_C_TEST) {
4567 PerlIO_printf(Perl_debug_log, "previous is\n");
4568 sv_dump(current);
a29f6d03
NC
4569 PerlIO_printf(Perl_debug_log,
4570 "move it from 0x%"UVxf" to 0x%"UVxf"\n",
d3d0e6f1
NC
4571 (UV) SV_COW_NEXT_SV(current), (UV) sv);
4572 }
a29f6d03 4573 SV_COW_NEXT_SV_SET(current, sv);
d3d0e6f1
NC
4574 }
4575#endif
79072805 4576 SvREFCNT(sv) = refcnt;
1edc1566 4577 SvFLAGS(nsv) |= SVTYPEMASK; /* Mark as freed */
39cf41c2 4578 SvREFCNT(nsv) = 0;
463ee0b2 4579 del_SV(nsv);
79072805
LW
4580}
4581
c461cf8f
JH
4582/*
4583=for apidoc sv_clear
4584
645c22ef
DM
4585Clear an SV: call any destructors, free up any memory used by the body,
4586and free the body itself. The SV's head is I<not> freed, although
4587its type is set to all 1's so that it won't inadvertently be assumed
4588to be live during global destruction etc.
4589This function should only be called when REFCNT is zero. Most of the time
4590you'll want to call C<sv_free()> (or its macro wrapper C<SvREFCNT_dec>)
4591instead.
c461cf8f
JH
4592
4593=cut
4594*/
4595
79072805 4596void
864dbfa3 4597Perl_sv_clear(pTHX_ register SV *sv)
79072805 4598{
27da23d5 4599 dVAR;
82bb6deb 4600 const U32 type = SvTYPE(sv);
8edfc514
NC
4601 const struct body_details *const sv_type_details
4602 = bodies_by_type + type;
82bb6deb 4603
79072805
LW
4604 assert(sv);
4605 assert(SvREFCNT(sv) == 0);
4606
82bb6deb
NC
4607 if (type <= SVt_IV)
4608 return;
4609
ed6116ce 4610 if (SvOBJECT(sv)) {
3280af22 4611 if (PL_defstash) { /* Still have a symbol table? */
39644a26 4612 dSP;
893645bd 4613 HV* stash;
d460ef45 4614 do {
b464bac0 4615 CV* destructor;
4e8e7886 4616 stash = SvSTASH(sv);
32251b26 4617 destructor = StashHANDLER(stash,DESTROY);
4e8e7886 4618 if (destructor) {
1b6737cc 4619 SV* const tmpref = newRV(sv);
5cc433a6 4620 SvREADONLY_on(tmpref); /* DESTROY() could be naughty */
4e8e7886 4621 ENTER;
e788e7d3 4622 PUSHSTACKi(PERLSI_DESTROY);
4e8e7886
GS
4623 EXTEND(SP, 2);
4624 PUSHMARK(SP);
5cc433a6 4625 PUSHs(tmpref);
4e8e7886 4626 PUTBACK;
44389ee9 4627 call_sv((SV*)destructor, G_DISCARD|G_EVAL|G_KEEPERR|G_VOID);
7a5fa8a2
NIS
4628
4629
d3acc0f7 4630 POPSTACK;
3095d977 4631 SPAGAIN;
4e8e7886 4632 LEAVE;
5cc433a6
AB
4633 if(SvREFCNT(tmpref) < 2) {
4634 /* tmpref is not kept alive! */
4635 SvREFCNT(sv)--;
b162af07 4636 SvRV_set(tmpref, NULL);
5cc433a6
AB
4637 SvROK_off(tmpref);
4638 }
4639 SvREFCNT_dec(tmpref);
4e8e7886
GS
4640 }
4641 } while (SvOBJECT(sv) && SvSTASH(sv) != stash);
8ebc5c01 4642
6f44e0a4
JP
4643
4644 if (SvREFCNT(sv)) {
4645 if (PL_in_clean_objs)
cea2e8a9 4646 Perl_croak(aTHX_ "DESTROY created new reference to dead object '%s'",
bfcb3514 4647 HvNAME_get(stash));
6f44e0a4
JP
4648 /* DESTROY gave object new lease on life */
4649 return;
4650 }
a0d0e21e 4651 }
4e8e7886 4652
a0d0e21e 4653 if (SvOBJECT(sv)) {
4e8e7886 4654 SvREFCNT_dec(SvSTASH(sv)); /* possibly of changed persuasion */
a0d0e21e 4655 SvOBJECT_off(sv); /* Curse the object. */
82bb6deb 4656 if (type != SVt_PVIO)
3280af22 4657 --PL_sv_objcount; /* XXX Might want something more general */
a0d0e21e 4658 }
463ee0b2 4659 }
82bb6deb 4660 if (type >= SVt_PVMG) {
524189f1
JH
4661 if (SvMAGIC(sv))
4662 mg_free(sv);
82bb6deb 4663 if (type == SVt_PVMG && SvFLAGS(sv) & SVpad_TYPED)
524189f1
JH
4664 SvREFCNT_dec(SvSTASH(sv));
4665 }
82bb6deb 4666 switch (type) {
8990e307 4667 case SVt_PVIO:
df0bd2f4
GS
4668 if (IoIFP(sv) &&
4669 IoIFP(sv) != PerlIO_stdin() &&
5f05dabc 4670 IoIFP(sv) != PerlIO_stdout() &&
4671 IoIFP(sv) != PerlIO_stderr())
93578b34 4672 {
f2b5be74 4673 io_close((IO*)sv, FALSE);
93578b34 4674 }
1d7c1841 4675 if (IoDIRP(sv) && !(IoFLAGS(sv) & IOf_FAKE_DIRP))
1236053a 4676 PerlDir_close(IoDIRP(sv));
1d7c1841 4677 IoDIRP(sv) = (DIR*)NULL;
8990e307
LW
4678 Safefree(IoTOP_NAME(sv));
4679 Safefree(IoFMT_NAME(sv));
4680 Safefree(IoBOTTOM_NAME(sv));
82bb6deb 4681 goto freescalar;
79072805 4682 case SVt_PVBM:
a0d0e21e 4683 goto freescalar;
79072805 4684 case SVt_PVCV:
748a9306 4685 case SVt_PVFM:
85e6fe83 4686 cv_undef((CV*)sv);
a0d0e21e 4687 goto freescalar;
79072805 4688 case SVt_PVHV:
85e6fe83 4689 hv_undef((HV*)sv);
a0d0e21e 4690 break;
79072805 4691 case SVt_PVAV:
85e6fe83 4692 av_undef((AV*)sv);
a0d0e21e 4693 break;
02270b4e 4694 case SVt_PVLV:
dd28f7bb
DM
4695 if (LvTYPE(sv) == 'T') { /* for tie: return HE to pool */
4696 SvREFCNT_dec(HeKEY_sv((HE*)LvTARG(sv)));
4697 HeNEXT((HE*)LvTARG(sv)) = PL_hv_fetch_ent_mh;
4698 PL_hv_fetch_ent_mh = (HE*)LvTARG(sv);
4699 }
4700 else if (LvTYPE(sv) != 't') /* unless tie: unrefcnted fake SV** */
4701 SvREFCNT_dec(LvTARG(sv));
02270b4e 4702 goto freescalar;
a0d0e21e 4703 case SVt_PVGV:
1edc1566 4704 gp_free((GV*)sv);
a0d0e21e 4705 Safefree(GvNAME(sv));
893645bd
NC
4706 /* If we're in a stash, we don't own a reference to it. However it does
4707 have a back reference to us, which needs to be cleared. */
4708 if (GvSTASH(sv))
4709 sv_del_backref((SV*)GvSTASH(sv), sv);
79072805 4710 case SVt_PVMG:
79072805
LW
4711 case SVt_PVNV:
4712 case SVt_PVIV:
a0d0e21e 4713 freescalar:
5228ca4e
NC
4714 /* Don't bother with SvOOK_off(sv); as we're only going to free it. */
4715 if (SvOOK(sv)) {
93524f2b 4716 SvPV_set(sv, SvPVX_mutable(sv) - SvIVX(sv));
5228ca4e
NC
4717 /* Don't even bother with turning off the OOK flag. */
4718 }
79072805 4719 case SVt_PV:
a0d0e21e 4720 case SVt_RV:
810b8aa5 4721 if (SvROK(sv)) {
e15faf7d 4722 SV *target = SvRV(sv);
810b8aa5 4723 if (SvWEAKREF(sv))
e15faf7d 4724 sv_del_backref(target, sv);
810b8aa5 4725 else
e15faf7d 4726 SvREFCNT_dec(target);
810b8aa5 4727 }
f8c7b90f 4728#ifdef PERL_OLD_COPY_ON_WRITE
3f7c398e 4729 else if (SvPVX_const(sv)) {
765f542d
NC
4730 if (SvIsCOW(sv)) {
4731 /* I believe I need to grab the global SV mutex here and
4732 then recheck the COW status. */
46187eeb
NC
4733 if (DEBUG_C_TEST) {
4734 PerlIO_printf(Perl_debug_log, "Copy on write: clear\n");
e419cbc5 4735 sv_dump(sv);
46187eeb 4736 }
bdd68bc3
NC
4737 sv_release_COW(sv, SvPVX_const(sv), SvLEN(sv),
4738 SV_COW_NEXT_SV(sv));
765f542d
NC
4739 /* And drop it here. */
4740 SvFAKE_off(sv);
4741 } else if (SvLEN(sv)) {
3f7c398e 4742 Safefree(SvPVX_const(sv));
765f542d
NC
4743 }
4744 }
4745#else
3f7c398e 4746 else if (SvPVX_const(sv) && SvLEN(sv))
94010e71 4747 Safefree(SvPVX_mutable(sv));
3f7c398e 4748 else if (SvPVX_const(sv) && SvREADONLY(sv) && SvFAKE(sv)) {
bdd68bc3 4749 unshare_hek(SvSHARED_HEK_FROM_PV(SvPVX_const(sv)));
1c846c1f
NIS
4750 SvFAKE_off(sv);
4751 }
765f542d 4752#endif
79072805
LW
4753 break;
4754 case SVt_NV:
79072805
LW
4755 break;
4756 }
4757
893645bd
NC
4758 SvFLAGS(sv) &= SVf_BREAK;
4759 SvFLAGS(sv) |= SVTYPEMASK;
4760
8edfc514 4761 if (sv_type_details->arena) {
b9502f15 4762 del_body(((char *)SvANY(sv) + sv_type_details->offset),
8edfc514
NC
4763 &PL_body_roots[type]);
4764 }
4765 else if (sv_type_details->size) {
4766 my_safefree(SvANY(sv));
4767 }
79072805
LW
4768}
4769
645c22ef
DM
4770/*
4771=for apidoc sv_newref
4772
4773Increment an SV's reference count. Use the C<SvREFCNT_inc()> wrapper
4774instead.
4775
4776=cut
4777*/
4778
79072805 4779SV *
864dbfa3 4780Perl_sv_newref(pTHX_ SV *sv)
79072805 4781{
463ee0b2 4782 if (sv)
4db098f4 4783 (SvREFCNT(sv))++;
79072805
LW
4784 return sv;
4785}
4786
c461cf8f
JH
4787/*
4788=for apidoc sv_free
4789
645c22ef
DM
4790Decrement an SV's reference count, and if it drops to zero, call
4791C<sv_clear> to invoke destructors and free up any memory used by
4792the body; finally, deallocate the SV's head itself.
4793Normally called via a wrapper macro C<SvREFCNT_dec>.
c461cf8f
JH
4794
4795=cut
4796*/
4797
79072805 4798void
864dbfa3 4799Perl_sv_free(pTHX_ SV *sv)
79072805 4800{
27da23d5 4801 dVAR;
79072805
LW
4802 if (!sv)
4803 return;
a0d0e21e
LW
4804 if (SvREFCNT(sv) == 0) {
4805 if (SvFLAGS(sv) & SVf_BREAK)
645c22ef
DM
4806 /* this SV's refcnt has been artificially decremented to
4807 * trigger cleanup */
a0d0e21e 4808 return;
3280af22 4809 if (PL_in_clean_all) /* All is fair */
1edc1566 4810 return;
d689ffdd
JP
4811 if (SvREADONLY(sv) && SvIMMORTAL(sv)) {
4812 /* make sure SvREFCNT(sv)==0 happens very seldom */
4813 SvREFCNT(sv) = (~(U32)0)/2;
4814 return;
4815 }
41e4abd8 4816 if (ckWARN_d(WARN_INTERNAL)) {
d5dede04 4817 Perl_warner(aTHX_ packWARN(WARN_INTERNAL),
472d47bc
SB
4818 "Attempt to free unreferenced scalar: SV 0x%"UVxf
4819 pTHX__FORMAT, PTR2UV(sv) pTHX__VALUE);
41e4abd8
NC
4820#ifdef DEBUG_LEAKING_SCALARS_FORK_DUMP
4821 Perl_dump_sv_child(aTHX_ sv);
4822#endif
4823 }
79072805
LW
4824 return;
4825 }
4db098f4 4826 if (--(SvREFCNT(sv)) > 0)
8990e307 4827 return;
8c4d3c90
NC
4828 Perl_sv_free2(aTHX_ sv);
4829}
4830
4831void
4832Perl_sv_free2(pTHX_ SV *sv)
4833{
27da23d5 4834 dVAR;
463ee0b2
LW
4835#ifdef DEBUGGING
4836 if (SvTEMP(sv)) {
0453d815 4837 if (ckWARN_d(WARN_DEBUGGING))
9014280d 4838 Perl_warner(aTHX_ packWARN(WARN_DEBUGGING),
472d47bc
SB
4839 "Attempt to free temp prematurely: SV 0x%"UVxf
4840 pTHX__FORMAT, PTR2UV(sv) pTHX__VALUE);
79072805 4841 return;
79072805 4842 }
463ee0b2 4843#endif
d689ffdd
JP
4844 if (SvREADONLY(sv) && SvIMMORTAL(sv)) {
4845 /* make sure SvREFCNT(sv)==0 happens very seldom */
4846 SvREFCNT(sv) = (~(U32)0)/2;
4847 return;
4848 }
79072805 4849 sv_clear(sv);
477f5d66
CS
4850 if (! SvREFCNT(sv))
4851 del_SV(sv);
79072805
LW
4852}
4853
954c1994
GS
4854/*
4855=for apidoc sv_len
4856
645c22ef
DM
4857Returns the length of the string in the SV. Handles magic and type
4858coercion. See also C<SvCUR>, which gives raw access to the xpv_cur slot.
954c1994
GS
4859
4860=cut
4861*/
4862
79072805 4863STRLEN
864dbfa3 4864Perl_sv_len(pTHX_ register SV *sv)
79072805 4865{
463ee0b2 4866 STRLEN len;
79072805
LW
4867
4868 if (!sv)
4869 return 0;
4870
8990e307 4871 if (SvGMAGICAL(sv))
565764a8 4872 len = mg_length(sv);
8990e307 4873 else
4d84ee25 4874 (void)SvPV_const(sv, len);
463ee0b2 4875 return len;
79072805
LW
4876}
4877
c461cf8f
JH
4878/*
4879=for apidoc sv_len_utf8
4880
4881Returns the number of characters in the string in an SV, counting wide
1e54db1a 4882UTF-8 bytes as a single character. Handles magic and type coercion.
c461cf8f
JH
4883
4884=cut
4885*/
4886
7e8c5dac
HS
4887/*
4888 * The length is cached in PERL_UTF8_magic, in the mg_len field. Also the
4889 * mg_ptr is used, by sv_pos_u2b(), see the comments of S_utf8_mg_pos_init().
4890 * (Note that the mg_len is not the length of the mg_ptr field.)
7a5fa8a2 4891 *
7e8c5dac
HS
4892 */
4893
a0ed51b3 4894STRLEN
864dbfa3 4895Perl_sv_len_utf8(pTHX_ register SV *sv)
a0ed51b3 4896{
a0ed51b3
LW
4897 if (!sv)
4898 return 0;
4899
a0ed51b3 4900 if (SvGMAGICAL(sv))
b76347f2 4901 return mg_length(sv);
a0ed51b3 4902 else
b76347f2 4903 {
7e8c5dac 4904 STRLEN len, ulen;
e62f0680 4905 const U8 *s = (U8*)SvPV_const(sv, len);
7e8c5dac
HS
4906 MAGIC *mg = SvMAGICAL(sv) ? mg_find(sv, PERL_MAGIC_utf8) : 0;
4907
e23c8137 4908 if (mg && mg->mg_len != -1 && (mg->mg_len > 0 || len == 0)) {
7e8c5dac 4909 ulen = mg->mg_len;
e23c8137
JH
4910#ifdef PERL_UTF8_CACHE_ASSERT
4911 assert(ulen == Perl_utf8_length(aTHX_ s, s + len));
4912#endif
4913 }
7e8c5dac
HS
4914 else {
4915 ulen = Perl_utf8_length(aTHX_ s, s + len);
4916 if (!mg && !SvREADONLY(sv)) {
4917 sv_magic(sv, 0, PERL_MAGIC_utf8, 0, 0);
4918 mg = mg_find(sv, PERL_MAGIC_utf8);
4919 assert(mg);
4920 }
4921 if (mg)
4922 mg->mg_len = ulen;
4923 }
4924 return ulen;
4925 }
4926}
4927
4928/* S_utf8_mg_pos_init() is used to initialize the mg_ptr field of
4929 * a PERL_UTF8_magic. The mg_ptr is used to store the mapping
4930 * between UTF-8 and byte offsets. There are two (substr offset and substr
4931 * length, the i offset, PERL_MAGIC_UTF8_CACHESIZE) times two (UTF-8 offset
4932 * and byte offset) cache positions.
4933 *
4934 * The mg_len field is used by sv_len_utf8(), see its comments.
4935 * Note that the mg_len is not the length of the mg_ptr field.
4936 *
4937 */
4938STATIC bool
245d4a47
NC
4939S_utf8_mg_pos_init(pTHX_ SV *sv, MAGIC **mgp, STRLEN **cachep, I32 i,
4940 I32 offsetp, const U8 *s, const U8 *start)
7e8c5dac 4941{
7a5fa8a2 4942 bool found = FALSE;
7e8c5dac
HS
4943
4944 if (SvMAGICAL(sv) && !SvREADONLY(sv)) {
8f78557a 4945 if (!*mgp)
27da23d5 4946 *mgp = sv_magicext(sv, 0, PERL_MAGIC_utf8, (MGVTBL*)&PL_vtbl_utf8, 0, 0);
7e8c5dac 4947 assert(*mgp);
b76347f2 4948
7e8c5dac
HS
4949 if ((*mgp)->mg_ptr)
4950 *cachep = (STRLEN *) (*mgp)->mg_ptr;
4951 else {
a02a5408 4952 Newxz(*cachep, PERL_MAGIC_UTF8_CACHESIZE * 2, STRLEN);
7e8c5dac
HS
4953 (*mgp)->mg_ptr = (char *) *cachep;
4954 }
4955 assert(*cachep);
4956
a3b680e6 4957 (*cachep)[i] = offsetp;
7e8c5dac
HS
4958 (*cachep)[i+1] = s - start;
4959 found = TRUE;
a0ed51b3 4960 }
7e8c5dac
HS
4961
4962 return found;
a0ed51b3
LW
4963}
4964
645c22ef 4965/*
7e8c5dac
HS
4966 * S_utf8_mg_pos() is used to query and update mg_ptr field of
4967 * a PERL_UTF8_magic. The mg_ptr is used to store the mapping
4968 * between UTF-8 and byte offsets. See also the comments of
4969 * S_utf8_mg_pos_init().
4970 *
4971 */
4972STATIC bool
245d4a47 4973S_utf8_mg_pos(pTHX_ SV *sv, MAGIC **mgp, STRLEN **cachep, I32 i, I32 *offsetp, I32 uoff, const U8 **sp, const U8 *start, const U8 *send)
7e8c5dac
HS
4974{
4975 bool found = FALSE;
4976
4977 if (SvMAGICAL(sv) && !SvREADONLY(sv)) {
4978 if (!*mgp)
4979 *mgp = mg_find(sv, PERL_MAGIC_utf8);
4980 if (*mgp && (*mgp)->mg_ptr) {
4981 *cachep = (STRLEN *) (*mgp)->mg_ptr;
e23c8137 4982 ASSERT_UTF8_CACHE(*cachep);
667208dd 4983 if ((*cachep)[i] == (STRLEN)uoff) /* An exact match. */
7a5fa8a2 4984 found = TRUE;
7e8c5dac
HS
4985 else { /* We will skip to the right spot. */
4986 STRLEN forw = 0;
4987 STRLEN backw = 0;
a3b680e6 4988 const U8* p = NULL;
7e8c5dac
HS
4989
4990 /* The assumption is that going backward is half
4991 * the speed of going forward (that's where the
4992 * 2 * backw in the below comes from). (The real
4993 * figure of course depends on the UTF-8 data.) */
4994
667208dd 4995 if ((*cachep)[i] > (STRLEN)uoff) {
7e8c5dac 4996 forw = uoff;
667208dd 4997 backw = (*cachep)[i] - (STRLEN)uoff;
7e8c5dac
HS
4998
4999 if (forw < 2 * backw)
5000 p = start;
5001 else
5002 p = start + (*cachep)[i+1];
5003 }
5004 /* Try this only for the substr offset (i == 0),
5005 * not for the substr length (i == 2). */
5006 else if (i == 0) { /* (*cachep)[i] < uoff */
a3b680e6 5007 const STRLEN ulen = sv_len_utf8(sv);
7e8c5dac 5008
667208dd
JH
5009 if ((STRLEN)uoff < ulen) {
5010 forw = (STRLEN)uoff - (*cachep)[i];
5011 backw = ulen - (STRLEN)uoff;
7e8c5dac
HS
5012
5013 if (forw < 2 * backw)
5014 p = start + (*cachep)[i+1];
5015 else
5016 p = send;
5017 }
5018
5019 /* If the string is not long enough for uoff,
5020 * we could extend it, but not at this low a level. */
5021 }
5022
5023 if (p) {
5024 if (forw < 2 * backw) {
5025 while (forw--)
5026 p += UTF8SKIP(p);
5027 }
5028 else {
5029 while (backw--) {
5030 p--;
5031 while (UTF8_IS_CONTINUATION(*p))
5032 p--;
5033 }
5034 }
5035
5036 /* Update the cache. */
667208dd 5037 (*cachep)[i] = (STRLEN)uoff;
7e8c5dac 5038 (*cachep)[i+1] = p - start;
8f78557a
AE
5039
5040 /* Drop the stale "length" cache */
5041 if (i == 0) {
5042 (*cachep)[2] = 0;
5043 (*cachep)[3] = 0;
5044 }
7a5fa8a2 5045
7e8c5dac
HS
5046 found = TRUE;
5047 }
5048 }
5049 if (found) { /* Setup the return values. */
5050 *offsetp = (*cachep)[i+1];
5051 *sp = start + *offsetp;
5052 if (*sp >= send) {
5053 *sp = send;
5054 *offsetp = send - start;
5055 }
5056 else if (*sp < start) {
5057 *sp = start;
5058 *offsetp = 0;
5059 }
5060 }
5061 }
e23c8137
JH
5062#ifdef PERL_UTF8_CACHE_ASSERT
5063 if (found) {
5064 U8 *s = start;
5065 I32 n = uoff;
5066
5067 while (n-- && s < send)
5068 s += UTF8SKIP(s);
5069
5070 if (i == 0) {
5071 assert(*offsetp == s - start);
5072 assert((*cachep)[0] == (STRLEN)uoff);
5073 assert((*cachep)[1] == *offsetp);
5074 }
5075 ASSERT_UTF8_CACHE(*cachep);
5076 }
5077#endif
7e8c5dac 5078 }
e23c8137 5079
7e8c5dac
HS
5080 return found;
5081}
7a5fa8a2 5082
7e8c5dac 5083/*
645c22ef
DM
5084=for apidoc sv_pos_u2b
5085
1e54db1a 5086Converts the value pointed to by offsetp from a count of UTF-8 chars from
645c22ef
DM
5087the start of the string, to a count of the equivalent number of bytes; if
5088lenp is non-zero, it does the same to lenp, but this time starting from
5089the offset, rather than from the start of the string. Handles magic and
5090type coercion.
5091
5092=cut
5093*/
5094
7e8c5dac
HS
5095/*
5096 * sv_pos_u2b() uses, like sv_pos_b2u(), the mg_ptr of the potential
5097 * PERL_UTF8_magic of the sv to store the mapping between UTF-8 and
5098 * byte offsets. See also the comments of S_utf8_mg_pos().
5099 *
5100 */
5101
a0ed51b3 5102void
864dbfa3 5103Perl_sv_pos_u2b(pTHX_ register SV *sv, I32* offsetp, I32* lenp)
a0ed51b3 5104{
245d4a47 5105 const U8 *start;
a0ed51b3
LW
5106 STRLEN len;
5107
5108 if (!sv)
5109 return;
5110
245d4a47 5111 start = (U8*)SvPV_const(sv, len);
7e8c5dac 5112 if (len) {
b464bac0
AL
5113 STRLEN boffset = 0;
5114 STRLEN *cache = 0;
245d4a47
NC
5115 const U8 *s = start;
5116 I32 uoffset = *offsetp;
9d4ba2ae 5117 const U8 * const send = s + len;
245d4a47
NC
5118 MAGIC *mg = 0;
5119 bool found = FALSE;
7e8c5dac 5120
bdf77a2a 5121 if (utf8_mg_pos(sv, &mg, &cache, 0, offsetp, *offsetp, &s, start, send))
7e8c5dac
HS
5122 found = TRUE;
5123 if (!found && uoffset > 0) {
5124 while (s < send && uoffset--)
5125 s += UTF8SKIP(s);
5126 if (s >= send)
5127 s = send;
a3b680e6 5128 if (utf8_mg_pos_init(sv, &mg, &cache, 0, *offsetp, s, start))
7e8c5dac
HS
5129 boffset = cache[1];
5130 *offsetp = s - start;
5131 }
5132 if (lenp) {
5133 found = FALSE;
5134 start = s;
ec062429 5135 if (utf8_mg_pos(sv, &mg, &cache, 2, lenp, *lenp, &s, start, send)) {
7e8c5dac
HS
5136 *lenp -= boffset;
5137 found = TRUE;
5138 }
5139 if (!found && *lenp > 0) {
5140 I32 ulen = *lenp;
5141 if (ulen > 0)
5142 while (s < send && ulen--)
5143 s += UTF8SKIP(s);
5144 if (s >= send)
5145 s = send;
a3b680e6 5146 utf8_mg_pos_init(sv, &mg, &cache, 2, *lenp, s, start);
7e8c5dac
HS
5147 }
5148 *lenp = s - start;
5149 }
e23c8137 5150 ASSERT_UTF8_CACHE(cache);
7e8c5dac
HS
5151 }
5152 else {
5153 *offsetp = 0;
5154 if (lenp)
5155 *lenp = 0;
a0ed51b3 5156 }
e23c8137 5157
a0ed51b3
LW
5158 return;
5159}
5160
645c22ef
DM
5161/*
5162=for apidoc sv_pos_b2u
5163
5164Converts the value pointed to by offsetp from a count of bytes from the
1e54db1a 5165start of the string, to a count of the equivalent number of UTF-8 chars.
645c22ef
DM
5166Handles magic and type coercion.
5167
5168=cut
5169*/
5170
7e8c5dac
HS
5171/*
5172 * sv_pos_b2u() uses, like sv_pos_u2b(), the mg_ptr of the potential
5173 * PERL_UTF8_magic of the sv to store the mapping between UTF-8 and
5174 * byte offsets. See also the comments of S_utf8_mg_pos().
5175 *
5176 */
5177
a0ed51b3 5178void
7e8c5dac 5179Perl_sv_pos_b2u(pTHX_ register SV* sv, I32* offsetp)
a0ed51b3 5180{
83003860 5181 const U8* s;
a0ed51b3
LW
5182 STRLEN len;
5183
5184 if (!sv)
5185 return;
5186
83003860 5187 s = (const U8*)SvPV_const(sv, len);
eb160463 5188 if ((I32)len < *offsetp)
a0dbb045 5189 Perl_croak(aTHX_ "panic: sv_pos_b2u: bad byte offset");
7e8c5dac 5190 else {
83003860 5191 const U8* send = s + *offsetp;
7e8c5dac
HS
5192 MAGIC* mg = NULL;
5193 STRLEN *cache = NULL;
5194
5195 len = 0;
5196
5197 if (SvMAGICAL(sv) && !SvREADONLY(sv)) {
5198 mg = mg_find(sv, PERL_MAGIC_utf8);
5199 if (mg && mg->mg_ptr) {
5200 cache = (STRLEN *) mg->mg_ptr;
c5661c80 5201 if (cache[1] == (STRLEN)*offsetp) {
7e8c5dac
HS
5202 /* An exact match. */
5203 *offsetp = cache[0];
5204
5205 return;
5206 }
c5661c80 5207 else if (cache[1] < (STRLEN)*offsetp) {
7e8c5dac
HS
5208 /* We already know part of the way. */
5209 len = cache[0];
5210 s += cache[1];
7a5fa8a2 5211 /* Let the below loop do the rest. */
7e8c5dac
HS
5212 }
5213 else { /* cache[1] > *offsetp */
5214 /* We already know all of the way, now we may
5215 * be able to walk back. The same assumption
5216 * is made as in S_utf8_mg_pos(), namely that
5217 * walking backward is twice slower than
5218 * walking forward. */
9d4ba2ae 5219 const STRLEN forw = *offsetp;
7e8c5dac
HS
5220 STRLEN backw = cache[1] - *offsetp;
5221
5222 if (!(forw < 2 * backw)) {
83003860 5223 const U8 *p = s + cache[1];
7e8c5dac 5224 STRLEN ubackw = 0;
7a5fa8a2 5225
a5b510f2
AE
5226 cache[1] -= backw;
5227
7e8c5dac
HS
5228 while (backw--) {
5229 p--;
0aeb64d0 5230 while (UTF8_IS_CONTINUATION(*p)) {
7e8c5dac 5231 p--;
0aeb64d0
JH
5232 backw--;
5233 }
7e8c5dac
HS
5234 ubackw++;
5235 }
5236
5237 cache[0] -= ubackw;
0aeb64d0 5238 *offsetp = cache[0];
a67d7df9
TS
5239
5240 /* Drop the stale "length" cache */
5241 cache[2] = 0;
5242 cache[3] = 0;
5243
0aeb64d0 5244 return;
7e8c5dac
HS
5245 }
5246 }
5247 }
e23c8137 5248 ASSERT_UTF8_CACHE(cache);
a0dbb045 5249 }
7e8c5dac
HS
5250
5251 while (s < send) {
5252 STRLEN n = 1;
5253
5254 /* Call utf8n_to_uvchr() to validate the sequence
5255 * (unless a simple non-UTF character) */
5256 if (!UTF8_IS_INVARIANT(*s))
5257 utf8n_to_uvchr(s, UTF8SKIP(s), &n, 0);
5258 if (n > 0) {
5259 s += n;
5260 len++;
5261 }
5262 else
5263 break;
5264 }
5265
5266 if (!SvREADONLY(sv)) {
5267 if (!mg) {
5268 sv_magic(sv, 0, PERL_MAGIC_utf8, 0, 0);
5269 mg = mg_find(sv, PERL_MAGIC_utf8);
5270 }
5271 assert(mg);
5272
5273 if (!mg->mg_ptr) {
a02a5408 5274 Newxz(cache, PERL_MAGIC_UTF8_CACHESIZE * 2, STRLEN);
7e8c5dac
HS
5275 mg->mg_ptr = (char *) cache;
5276 }
5277 assert(cache);
5278
5279 cache[0] = len;
5280 cache[1] = *offsetp;
a67d7df9
TS
5281 /* Drop the stale "length" cache */
5282 cache[2] = 0;
5283 cache[3] = 0;
7e8c5dac
HS
5284 }
5285
5286 *offsetp = len;
a0ed51b3 5287 }
a0ed51b3
LW
5288 return;
5289}
5290
954c1994
GS
5291/*
5292=for apidoc sv_eq
5293
5294Returns a boolean indicating whether the strings in the two SVs are
645c22ef
DM
5295identical. Is UTF-8 and 'use bytes' aware, handles get magic, and will
5296coerce its args to strings if necessary.
954c1994
GS
5297
5298=cut
5299*/
5300
79072805 5301I32
e01b9e88 5302Perl_sv_eq(pTHX_ register SV *sv1, register SV *sv2)
79072805 5303{
e1ec3a88 5304 const char *pv1;
463ee0b2 5305 STRLEN cur1;
e1ec3a88 5306 const char *pv2;
463ee0b2 5307 STRLEN cur2;
e01b9e88 5308 I32 eq = 0;
553e1bcc
AT
5309 char *tpv = Nullch;
5310 SV* svrecode = Nullsv;
79072805 5311
e01b9e88 5312 if (!sv1) {
79072805
LW
5313 pv1 = "";
5314 cur1 = 0;
5315 }
463ee0b2 5316 else
4d84ee25 5317 pv1 = SvPV_const(sv1, cur1);
79072805 5318
e01b9e88
SC
5319 if (!sv2){
5320 pv2 = "";
5321 cur2 = 0;
92d29cee 5322 }
e01b9e88 5323 else
4d84ee25 5324 pv2 = SvPV_const(sv2, cur2);
79072805 5325
cf48d248 5326 if (cur1 && cur2 && SvUTF8(sv1) != SvUTF8(sv2) && !IN_BYTES) {
799ef3cb
JH
5327 /* Differing utf8ness.
5328 * Do not UTF8size the comparands as a side-effect. */
5329 if (PL_encoding) {
5330 if (SvUTF8(sv1)) {
553e1bcc
AT
5331 svrecode = newSVpvn(pv2, cur2);
5332 sv_recode_to_utf8(svrecode, PL_encoding);
93524f2b 5333 pv2 = SvPV_const(svrecode, cur2);
799ef3cb
JH
5334 }
5335 else {
553e1bcc
AT
5336 svrecode = newSVpvn(pv1, cur1);
5337 sv_recode_to_utf8(svrecode, PL_encoding);
93524f2b 5338 pv1 = SvPV_const(svrecode, cur1);
799ef3cb
JH
5339 }
5340 /* Now both are in UTF-8. */
0a1bd7ac
DM
5341 if (cur1 != cur2) {
5342 SvREFCNT_dec(svrecode);
799ef3cb 5343 return FALSE;
0a1bd7ac 5344 }
799ef3cb
JH
5345 }
5346 else {
5347 bool is_utf8 = TRUE;
5348
5349 if (SvUTF8(sv1)) {
5350 /* sv1 is the UTF-8 one,
5351 * if is equal it must be downgrade-able */
9d4ba2ae 5352 char * const pv = (char*)bytes_from_utf8((const U8*)pv1,
799ef3cb
JH
5353 &cur1, &is_utf8);
5354 if (pv != pv1)
553e1bcc 5355 pv1 = tpv = pv;
799ef3cb
JH
5356 }
5357 else {
5358 /* sv2 is the UTF-8 one,
5359 * if is equal it must be downgrade-able */
9d4ba2ae 5360 char * const pv = (char *)bytes_from_utf8((const U8*)pv2,
799ef3cb
JH
5361 &cur2, &is_utf8);
5362 if (pv != pv2)
553e1bcc 5363 pv2 = tpv = pv;
799ef3cb
JH
5364 }
5365 if (is_utf8) {
5366 /* Downgrade not possible - cannot be eq */
bf694877 5367 assert (tpv == 0);
799ef3cb
JH
5368 return FALSE;
5369 }
5370 }
cf48d248
JH
5371 }
5372
5373 if (cur1 == cur2)
765f542d 5374 eq = (pv1 == pv2) || memEQ(pv1, pv2, cur1);
e01b9e88 5375
553e1bcc
AT
5376 if (svrecode)
5377 SvREFCNT_dec(svrecode);
799ef3cb 5378
553e1bcc
AT
5379 if (tpv)
5380 Safefree(tpv);
cf48d248 5381
e01b9e88 5382 return eq;
79072805
LW
5383}
5384
954c1994
GS
5385/*
5386=for apidoc sv_cmp
5387
5388Compares the strings in two SVs. Returns -1, 0, or 1 indicating whether the
5389string in C<sv1> is less than, equal to, or greater than the string in
645c22ef
DM
5390C<sv2>. Is UTF-8 and 'use bytes' aware, handles get magic, and will
5391coerce its args to strings if necessary. See also C<sv_cmp_locale>.
954c1994
GS
5392
5393=cut
5394*/
5395
79072805 5396I32
e01b9e88 5397Perl_sv_cmp(pTHX_ register SV *sv1, register SV *sv2)
79072805 5398{
560a288e 5399 STRLEN cur1, cur2;
e1ec3a88
AL
5400 const char *pv1, *pv2;
5401 char *tpv = Nullch;
cf48d248 5402 I32 cmp;
553e1bcc 5403 SV *svrecode = Nullsv;
560a288e 5404
e01b9e88
SC
5405 if (!sv1) {
5406 pv1 = "";
560a288e
GS
5407 cur1 = 0;
5408 }
e01b9e88 5409 else
4d84ee25 5410 pv1 = SvPV_const(sv1, cur1);
560a288e 5411
553e1bcc 5412 if (!sv2) {
e01b9e88 5413 pv2 = "";
560a288e
GS
5414 cur2 = 0;
5415 }
e01b9e88 5416 else
4d84ee25 5417 pv2 = SvPV_const(sv2, cur2);
79072805 5418
cf48d248 5419 if (cur1 && cur2 && SvUTF8(sv1) != SvUTF8(sv2) && !IN_BYTES) {
799ef3cb
JH
5420 /* Differing utf8ness.
5421 * Do not UTF8size the comparands as a side-effect. */
cf48d248 5422 if (SvUTF8(sv1)) {
799ef3cb 5423 if (PL_encoding) {
553e1bcc
AT
5424 svrecode = newSVpvn(pv2, cur2);
5425 sv_recode_to_utf8(svrecode, PL_encoding);
93524f2b 5426 pv2 = SvPV_const(svrecode, cur2);
799ef3cb
JH
5427 }
5428 else {
e1ec3a88 5429 pv2 = tpv = (char*)bytes_to_utf8((const U8*)pv2, &cur2);
799ef3cb 5430 }
cf48d248
JH
5431 }
5432 else {
799ef3cb 5433 if (PL_encoding) {
553e1bcc
AT
5434 svrecode = newSVpvn(pv1, cur1);
5435 sv_recode_to_utf8(svrecode, PL_encoding);
93524f2b 5436 pv1 = SvPV_const(svrecode, cur1);
799ef3cb
JH
5437 }
5438 else {
e1ec3a88 5439 pv1 = tpv = (char*)bytes_to_utf8((const U8*)pv1, &cur1);
799ef3cb 5440 }
cf48d248
JH
5441 }
5442 }
5443
e01b9e88 5444 if (!cur1) {
cf48d248 5445 cmp = cur2 ? -1 : 0;
e01b9e88 5446 } else if (!cur2) {
cf48d248
JH
5447 cmp = 1;
5448 } else {
e1ec3a88 5449 const I32 retval = memcmp((const void*)pv1, (const void*)pv2, cur1 < cur2 ? cur1 : cur2);
e01b9e88
SC
5450
5451 if (retval) {
cf48d248 5452 cmp = retval < 0 ? -1 : 1;
e01b9e88 5453 } else if (cur1 == cur2) {
cf48d248
JH
5454 cmp = 0;
5455 } else {
5456 cmp = cur1 < cur2 ? -1 : 1;
e01b9e88 5457 }
cf48d248 5458 }
16660edb 5459
553e1bcc
AT
5460 if (svrecode)
5461 SvREFCNT_dec(svrecode);
799ef3cb 5462
553e1bcc
AT
5463 if (tpv)
5464 Safefree(tpv);
cf48d248
JH
5465
5466 return cmp;
bbce6d69 5467}
16660edb 5468
c461cf8f
JH
5469/*
5470=for apidoc sv_cmp_locale
5471
645c22ef
DM
5472Compares the strings in two SVs in a locale-aware manner. Is UTF-8 and
5473'use bytes' aware, handles get magic, and will coerce its args to strings
5474if necessary. See also C<sv_cmp_locale>. See also C<sv_cmp>.
c461cf8f
JH
5475
5476=cut
5477*/
5478
bbce6d69 5479I32
864dbfa3 5480Perl_sv_cmp_locale(pTHX_ register SV *sv1, register SV *sv2)
bbce6d69 5481{
36477c24 5482#ifdef USE_LOCALE_COLLATE
16660edb 5483
bbce6d69 5484 char *pv1, *pv2;
5485 STRLEN len1, len2;
5486 I32 retval;
16660edb 5487
3280af22 5488 if (PL_collation_standard)
bbce6d69 5489 goto raw_compare;
16660edb 5490
bbce6d69 5491 len1 = 0;
8ac85365 5492 pv1 = sv1 ? sv_collxfrm(sv1, &len1) : (char *) NULL;
bbce6d69 5493 len2 = 0;
8ac85365 5494 pv2 = sv2 ? sv_collxfrm(sv2, &len2) : (char *) NULL;
16660edb 5495
bbce6d69 5496 if (!pv1 || !len1) {
5497 if (pv2 && len2)
5498 return -1;
5499 else
5500 goto raw_compare;
5501 }
5502 else {
5503 if (!pv2 || !len2)
5504 return 1;
5505 }
16660edb 5506
bbce6d69 5507 retval = memcmp((void*)pv1, (void*)pv2, len1 < len2 ? len1 : len2);
16660edb 5508
bbce6d69 5509 if (retval)
16660edb 5510 return retval < 0 ? -1 : 1;
5511
bbce6d69 5512 /*
5513 * When the result of collation is equality, that doesn't mean
5514 * that there are no differences -- some locales exclude some
5515 * characters from consideration. So to avoid false equalities,
5516 * we use the raw string as a tiebreaker.
5517 */
16660edb 5518
bbce6d69 5519 raw_compare:
5520 /* FALL THROUGH */
16660edb 5521
36477c24 5522#endif /* USE_LOCALE_COLLATE */
16660edb 5523
bbce6d69 5524 return sv_cmp(sv1, sv2);
5525}
79072805 5526
645c22ef 5527
36477c24 5528#ifdef USE_LOCALE_COLLATE
645c22ef 5529
7a4c00b4 5530/*
645c22ef
DM
5531=for apidoc sv_collxfrm
5532
5533Add Collate Transform magic to an SV if it doesn't already have it.
5534
5535Any scalar variable may carry PERL_MAGIC_collxfrm magic that contains the
5536scalar data of the variable, but transformed to such a format that a normal
5537memory comparison can be used to compare the data according to the locale
5538settings.
5539
5540=cut
5541*/
5542
bbce6d69 5543char *
864dbfa3 5544Perl_sv_collxfrm(pTHX_ SV *sv, STRLEN *nxp)
bbce6d69 5545{
7a4c00b4 5546 MAGIC *mg;
16660edb 5547
14befaf4 5548 mg = SvMAGICAL(sv) ? mg_find(sv, PERL_MAGIC_collxfrm) : (MAGIC *) NULL;
3280af22 5549 if (!mg || !mg->mg_ptr || *(U32*)mg->mg_ptr != PL_collation_ix) {
93524f2b
NC
5550 const char *s;
5551 char *xf;
bbce6d69 5552 STRLEN len, xlen;
5553
7a4c00b4 5554 if (mg)
5555 Safefree(mg->mg_ptr);
93524f2b 5556 s = SvPV_const(sv, len);
bbce6d69 5557 if ((xf = mem_collxfrm(s, len, &xlen))) {
ff0cee69 5558 if (SvREADONLY(sv)) {
5559 SAVEFREEPV(xf);
5560 *nxp = xlen;
3280af22 5561 return xf + sizeof(PL_collation_ix);
ff0cee69 5562 }
7a4c00b4 5563 if (! mg) {
14befaf4
DM
5564 sv_magic(sv, 0, PERL_MAGIC_collxfrm, 0, 0);
5565 mg = mg_find(sv, PERL_MAGIC_collxfrm);
7a4c00b4 5566 assert(mg);
bbce6d69 5567 }
7a4c00b4 5568 mg->mg_ptr = xf;
565764a8 5569 mg->mg_len = xlen;
7a4c00b4 5570 }
5571 else {
ff0cee69 5572 if (mg) {
5573 mg->mg_ptr = NULL;
565764a8 5574 mg->mg_len = -1;
ff0cee69 5575 }
bbce6d69 5576 }
5577 }
7a4c00b4 5578 if (mg && mg->mg_ptr) {
565764a8 5579 *nxp = mg->mg_len;
3280af22 5580 return mg->mg_ptr + sizeof(PL_collation_ix);
bbce6d69 5581 }
5582 else {
5583 *nxp = 0;
5584 return NULL;
16660edb 5585 }
79072805
LW
5586}
5587
36477c24 5588#endif /* USE_LOCALE_COLLATE */
bbce6d69 5589
c461cf8f
JH
5590/*
5591=for apidoc sv_gets
5592
5593Get a line from the filehandle and store it into the SV, optionally
5594appending to the currently-stored string.
5595
5596=cut
5597*/
5598
79072805 5599char *
864dbfa3 5600Perl_sv_gets(pTHX_ register SV *sv, register PerlIO *fp, I32 append)
79072805 5601{
e1ec3a88 5602 const char *rsptr;
c07a80fd 5603 STRLEN rslen;
5604 register STDCHAR rslast;
5605 register STDCHAR *bp;
5606 register I32 cnt;
9c5ffd7c 5607 I32 i = 0;
8bfdd7d9 5608 I32 rspara = 0;
e311fd51 5609 I32 recsize;
c07a80fd 5610
bc44a8a2
NC
5611 if (SvTHINKFIRST(sv))
5612 sv_force_normal_flags(sv, append ? 0 : SV_COW_DROP_PV);
765f542d
NC
5613 /* XXX. If you make this PVIV, then copy on write can copy scalars read
5614 from <>.
5615 However, perlbench says it's slower, because the existing swipe code
5616 is faster than copy on write.
5617 Swings and roundabouts. */
862a34c6 5618 SvUPGRADE(sv, SVt_PV);
99491443 5619
ff68c719 5620 SvSCREAM_off(sv);
efd8b2ba
AE
5621
5622 if (append) {
5623 if (PerlIO_isutf8(fp)) {
5624 if (!SvUTF8(sv)) {
5625 sv_utf8_upgrade_nomg(sv);
5626 sv_pos_u2b(sv,&append,0);
5627 }
5628 } else if (SvUTF8(sv)) {
1b6737cc 5629 SV * const tsv = NEWSV(0,0);
efd8b2ba
AE
5630 sv_gets(tsv, fp, 0);
5631 sv_utf8_upgrade_nomg(tsv);
5632 SvCUR_set(sv,append);
5633 sv_catsv(sv,tsv);
5634 sv_free(tsv);
5635 goto return_string_or_null;
5636 }
5637 }
5638
5639 SvPOK_only(sv);
5640 if (PerlIO_isutf8(fp))
5641 SvUTF8_on(sv);
c07a80fd 5642
923e4eb5 5643 if (IN_PERL_COMPILETIME) {
8bfdd7d9
HS
5644 /* we always read code in line mode */
5645 rsptr = "\n";
5646 rslen = 1;
5647 }
5648 else if (RsSNARF(PL_rs)) {
7a5fa8a2
NIS
5649 /* If it is a regular disk file use size from stat() as estimate
5650 of amount we are going to read - may result in malloc-ing
5651 more memory than we realy need if layers bellow reduce
e468d35b
NIS
5652 size we read (e.g. CRLF or a gzip layer)
5653 */
e311fd51 5654 Stat_t st;
e468d35b 5655 if (!PerlLIO_fstat(PerlIO_fileno(fp), &st) && S_ISREG(st.st_mode)) {
f54cb97a 5656 const Off_t offset = PerlIO_tell(fp);
58f1856e 5657 if (offset != (Off_t) -1 && st.st_size + append > offset) {
e468d35b
NIS
5658 (void) SvGROW(sv, (STRLEN)((st.st_size - offset) + append + 1));
5659 }
5660 }
c07a80fd 5661 rsptr = NULL;
5662 rslen = 0;
5663 }
3280af22 5664 else if (RsRECORD(PL_rs)) {
e311fd51 5665 I32 bytesread;
5b2b9c68
HM
5666 char *buffer;
5667
5668 /* Grab the size of the record we're getting */
3280af22 5669 recsize = SvIV(SvRV(PL_rs));
e311fd51 5670 buffer = SvGROW(sv, (STRLEN)(recsize + append + 1)) + append;
5b2b9c68
HM
5671 /* Go yank in */
5672#ifdef VMS
5673 /* VMS wants read instead of fread, because fread doesn't respect */
5674 /* RMS record boundaries. This is not necessarily a good thing to be */
e468d35b
NIS
5675 /* doing, but we've got no other real choice - except avoid stdio
5676 as implementation - perhaps write a :vms layer ?
5677 */
5b2b9c68
HM
5678 bytesread = PerlLIO_read(PerlIO_fileno(fp), buffer, recsize);
5679#else
5680 bytesread = PerlIO_read(fp, buffer, recsize);
5681#endif
27e6ca2d
AE
5682 if (bytesread < 0)
5683 bytesread = 0;
e311fd51 5684 SvCUR_set(sv, bytesread += append);
e670df4e 5685 buffer[bytesread] = '\0';
efd8b2ba 5686 goto return_string_or_null;
5b2b9c68 5687 }
3280af22 5688 else if (RsPARA(PL_rs)) {
c07a80fd 5689 rsptr = "\n\n";
5690 rslen = 2;
8bfdd7d9 5691 rspara = 1;
c07a80fd 5692 }
7d59b7e4
NIS
5693 else {
5694 /* Get $/ i.e. PL_rs into same encoding as stream wants */
5695 if (PerlIO_isutf8(fp)) {
5696 rsptr = SvPVutf8(PL_rs, rslen);
5697 }
5698 else {
5699 if (SvUTF8(PL_rs)) {
5700 if (!sv_utf8_downgrade(PL_rs, TRUE)) {
5701 Perl_croak(aTHX_ "Wide character in $/");
5702 }
5703 }
93524f2b 5704 rsptr = SvPV_const(PL_rs, rslen);
7d59b7e4
NIS
5705 }
5706 }
5707
c07a80fd 5708 rslast = rslen ? rsptr[rslen - 1] : '\0';
5709
8bfdd7d9 5710 if (rspara) { /* have to do this both before and after */
79072805 5711 do { /* to make sure file boundaries work right */
760ac839 5712 if (PerlIO_eof(fp))
a0d0e21e 5713 return 0;
760ac839 5714 i = PerlIO_getc(fp);
79072805 5715 if (i != '\n') {
a0d0e21e
LW
5716 if (i == -1)
5717 return 0;
760ac839 5718 PerlIO_ungetc(fp,i);
79072805
LW
5719 break;
5720 }
5721 } while (i != EOF);
5722 }
c07a80fd 5723
760ac839
LW
5724 /* See if we know enough about I/O mechanism to cheat it ! */
5725
5726 /* This used to be #ifdef test - it is made run-time test for ease
1c846c1f 5727 of abstracting out stdio interface. One call should be cheap
760ac839
LW
5728 enough here - and may even be a macro allowing compile
5729 time optimization.
5730 */
5731
5732 if (PerlIO_fast_gets(fp)) {
5733
5734 /*
5735 * We're going to steal some values from the stdio struct
5736 * and put EVERYTHING in the innermost loop into registers.
5737 */
5738 register STDCHAR *ptr;
5739 STRLEN bpx;
5740 I32 shortbuffered;
5741
16660edb 5742#if defined(VMS) && defined(PERLIO_IS_STDIO)
5743 /* An ungetc()d char is handled separately from the regular
5744 * buffer, so we getc() it back out and stuff it in the buffer.
5745 */
5746 i = PerlIO_getc(fp);
5747 if (i == EOF) return 0;
5748 *(--((*fp)->_ptr)) = (unsigned char) i;
5749 (*fp)->_cnt++;
5750#endif
c07a80fd 5751
c2960299 5752 /* Here is some breathtakingly efficient cheating */
c07a80fd 5753
a20bf0c3 5754 cnt = PerlIO_get_cnt(fp); /* get count into register */
e468d35b 5755 /* make sure we have the room */
7a5fa8a2 5756 if ((I32)(SvLEN(sv) - append) <= cnt + 1) {
e468d35b 5757 /* Not room for all of it
7a5fa8a2 5758 if we are looking for a separator and room for some
e468d35b
NIS
5759 */
5760 if (rslen && cnt > 80 && (I32)SvLEN(sv) > append) {
7a5fa8a2 5761 /* just process what we have room for */
79072805
LW
5762 shortbuffered = cnt - SvLEN(sv) + append + 1;
5763 cnt -= shortbuffered;
5764 }
5765 else {
5766 shortbuffered = 0;
bbce6d69 5767 /* remember that cnt can be negative */
eb160463 5768 SvGROW(sv, (STRLEN)(append + (cnt <= 0 ? 2 : (cnt + 1))));
79072805
LW
5769 }
5770 }
7a5fa8a2 5771 else
79072805 5772 shortbuffered = 0;
3f7c398e 5773 bp = (STDCHAR*)SvPVX_const(sv) + append; /* move these two too to registers */
a20bf0c3 5774 ptr = (STDCHAR*)PerlIO_get_ptr(fp);
16660edb 5775 DEBUG_P(PerlIO_printf(Perl_debug_log,
1d7c1841 5776 "Screamer: entering, ptr=%"UVuf", cnt=%ld\n",PTR2UV(ptr),(long)cnt));
16660edb 5777 DEBUG_P(PerlIO_printf(Perl_debug_log,
ba7abf9d 5778 "Screamer: entering: PerlIO * thinks ptr=%"UVuf", cnt=%ld, base=%"UVuf"\n",
1c846c1f 5779 PTR2UV(PerlIO_get_ptr(fp)), (long)PerlIO_get_cnt(fp),
1d7c1841 5780 PTR2UV(PerlIO_has_base(fp) ? PerlIO_get_base(fp) : 0)));
79072805
LW
5781 for (;;) {
5782 screamer:
93a17b20 5783 if (cnt > 0) {
c07a80fd 5784 if (rslen) {
760ac839
LW
5785 while (cnt > 0) { /* this | eat */
5786 cnt--;
c07a80fd 5787 if ((*bp++ = *ptr++) == rslast) /* really | dust */
5788 goto thats_all_folks; /* screams | sed :-) */
5789 }
5790 }
5791 else {
1c846c1f
NIS
5792 Copy(ptr, bp, cnt, char); /* this | eat */
5793 bp += cnt; /* screams | dust */
c07a80fd 5794 ptr += cnt; /* louder | sed :-) */
a5f75d66 5795 cnt = 0;
93a17b20 5796 }
79072805
LW
5797 }
5798
748a9306 5799 if (shortbuffered) { /* oh well, must extend */
79072805
LW
5800 cnt = shortbuffered;
5801 shortbuffered = 0;
3f7c398e 5802 bpx = bp - (STDCHAR*)SvPVX_const(sv); /* box up before relocation */
79072805
LW
5803 SvCUR_set(sv, bpx);
5804 SvGROW(sv, SvLEN(sv) + append + cnt + 2);
3f7c398e 5805 bp = (STDCHAR*)SvPVX_const(sv) + bpx; /* unbox after relocation */
79072805
LW
5806 continue;
5807 }
5808
16660edb 5809 DEBUG_P(PerlIO_printf(Perl_debug_log,
1d7c1841
GS
5810 "Screamer: going to getc, ptr=%"UVuf", cnt=%ld\n",
5811 PTR2UV(ptr),(long)cnt));
cc00df79 5812 PerlIO_set_ptrcnt(fp, (STDCHAR*)ptr, cnt); /* deregisterize cnt and ptr */
ba7abf9d 5813#if 0
16660edb 5814 DEBUG_P(PerlIO_printf(Perl_debug_log,
1d7c1841 5815 "Screamer: pre: FILE * thinks ptr=%"UVuf", cnt=%ld, base=%"UVuf"\n",
1c846c1f 5816 PTR2UV(PerlIO_get_ptr(fp)), (long)PerlIO_get_cnt(fp),
1d7c1841 5817 PTR2UV(PerlIO_has_base (fp) ? PerlIO_get_base(fp) : 0)));
ba7abf9d 5818#endif
1c846c1f 5819 /* This used to call 'filbuf' in stdio form, but as that behaves like
774d564b 5820 getc when cnt <= 0 we use PerlIO_getc here to avoid introducing
5821 another abstraction. */
760ac839 5822 i = PerlIO_getc(fp); /* get more characters */
ba7abf9d 5823#if 0
16660edb 5824 DEBUG_P(PerlIO_printf(Perl_debug_log,
1d7c1841 5825 "Screamer: post: FILE * thinks ptr=%"UVuf", cnt=%ld, base=%"UVuf"\n",
1c846c1f 5826 PTR2UV(PerlIO_get_ptr(fp)), (long)PerlIO_get_cnt(fp),
1d7c1841 5827 PTR2UV(PerlIO_has_base (fp) ? PerlIO_get_base(fp) : 0)));
ba7abf9d 5828#endif
a20bf0c3
JH
5829 cnt = PerlIO_get_cnt(fp);
5830 ptr = (STDCHAR*)PerlIO_get_ptr(fp); /* reregisterize cnt and ptr */
16660edb 5831 DEBUG_P(PerlIO_printf(Perl_debug_log,
1d7c1841 5832 "Screamer: after getc, ptr=%"UVuf", cnt=%ld\n",PTR2UV(ptr),(long)cnt));
79072805 5833
748a9306
LW
5834 if (i == EOF) /* all done for ever? */
5835 goto thats_really_all_folks;
5836
3f7c398e 5837 bpx = bp - (STDCHAR*)SvPVX_const(sv); /* box up before relocation */
79072805
LW
5838 SvCUR_set(sv, bpx);
5839 SvGROW(sv, bpx + cnt + 2);
3f7c398e 5840 bp = (STDCHAR*)SvPVX_const(sv) + bpx; /* unbox after relocation */
c07a80fd 5841
eb160463 5842 *bp++ = (STDCHAR)i; /* store character from PerlIO_getc */
79072805 5843
c07a80fd 5844 if (rslen && (STDCHAR)i == rslast) /* all done for now? */
79072805 5845 goto thats_all_folks;
79072805
LW
5846 }
5847
5848thats_all_folks:
3f7c398e 5849 if ((rslen > 1 && (STRLEN)(bp - (STDCHAR*)SvPVX_const(sv)) < rslen) ||
36477c24 5850 memNE((char*)bp - rslen, rsptr, rslen))
760ac839 5851 goto screamer; /* go back to the fray */
79072805
LW
5852thats_really_all_folks:
5853 if (shortbuffered)
5854 cnt += shortbuffered;
16660edb 5855 DEBUG_P(PerlIO_printf(Perl_debug_log,
1d7c1841 5856 "Screamer: quitting, ptr=%"UVuf", cnt=%ld\n",PTR2UV(ptr),(long)cnt));
cc00df79 5857 PerlIO_set_ptrcnt(fp, (STDCHAR*)ptr, cnt); /* put these back or we're in trouble */
16660edb 5858 DEBUG_P(PerlIO_printf(Perl_debug_log,
1d7c1841 5859 "Screamer: end: FILE * thinks ptr=%"UVuf", cnt=%ld, base=%"UVuf"\n",
1c846c1f 5860 PTR2UV(PerlIO_get_ptr(fp)), (long)PerlIO_get_cnt(fp),
1d7c1841 5861 PTR2UV(PerlIO_has_base (fp) ? PerlIO_get_base(fp) : 0)));
79072805 5862 *bp = '\0';
3f7c398e 5863 SvCUR_set(sv, bp - (STDCHAR*)SvPVX_const(sv)); /* set length */
16660edb 5864 DEBUG_P(PerlIO_printf(Perl_debug_log,
fb73857a 5865 "Screamer: done, len=%ld, string=|%.*s|\n",
3f7c398e 5866 (long)SvCUR(sv),(int)SvCUR(sv),SvPVX_const(sv)));
760ac839
LW
5867 }
5868 else
79072805 5869 {
6edd2cd5 5870 /*The big, slow, and stupid way. */
27da23d5 5871#ifdef USE_HEAP_INSTEAD_OF_STACK /* Even slower way. */
6edd2cd5 5872 STDCHAR *buf = 0;
a02a5408 5873 Newx(buf, 8192, STDCHAR);
6edd2cd5 5874 assert(buf);
4d2c4e07 5875#else
6edd2cd5 5876 STDCHAR buf[8192];
4d2c4e07 5877#endif
79072805 5878
760ac839 5879screamer2:
c07a80fd 5880 if (rslen) {
00b6aa41 5881 register const STDCHAR * const bpe = buf + sizeof(buf);
760ac839 5882 bp = buf;
eb160463 5883 while ((i = PerlIO_getc(fp)) != EOF && (*bp++ = (STDCHAR)i) != rslast && bp < bpe)
760ac839
LW
5884 ; /* keep reading */
5885 cnt = bp - buf;
c07a80fd 5886 }
5887 else {
760ac839 5888 cnt = PerlIO_read(fp,(char*)buf, sizeof(buf));
16660edb 5889 /* Accomodate broken VAXC compiler, which applies U8 cast to
5890 * both args of ?: operator, causing EOF to change into 255
5891 */
37be0adf 5892 if (cnt > 0)
cbe9e203
JH
5893 i = (U8)buf[cnt - 1];
5894 else
37be0adf 5895 i = EOF;
c07a80fd 5896 }
79072805 5897
cbe9e203
JH
5898 if (cnt < 0)
5899 cnt = 0; /* we do need to re-set the sv even when cnt <= 0 */
5900 if (append)
5901 sv_catpvn(sv, (char *) buf, cnt);
5902 else
5903 sv_setpvn(sv, (char *) buf, cnt);
c07a80fd 5904
5905 if (i != EOF && /* joy */
5906 (!rslen ||
5907 SvCUR(sv) < rslen ||
3f7c398e 5908 memNE(SvPVX_const(sv) + SvCUR(sv) - rslen, rsptr, rslen)))
79072805
LW
5909 {
5910 append = -1;
63e4d877
CS
5911 /*
5912 * If we're reading from a TTY and we get a short read,
5913 * indicating that the user hit his EOF character, we need
5914 * to notice it now, because if we try to read from the TTY
5915 * again, the EOF condition will disappear.
5916 *
5917 * The comparison of cnt to sizeof(buf) is an optimization
5918 * that prevents unnecessary calls to feof().
5919 *
5920 * - jik 9/25/96
5921 */
5922 if (!(cnt < sizeof(buf) && PerlIO_eof(fp)))
5923 goto screamer2;
79072805 5924 }
6edd2cd5 5925
27da23d5 5926#ifdef USE_HEAP_INSTEAD_OF_STACK
6edd2cd5
JH
5927 Safefree(buf);
5928#endif
79072805
LW
5929 }
5930
8bfdd7d9 5931 if (rspara) { /* have to do this both before and after */
c07a80fd 5932 while (i != EOF) { /* to make sure file boundaries work right */
760ac839 5933 i = PerlIO_getc(fp);
79072805 5934 if (i != '\n') {
760ac839 5935 PerlIO_ungetc(fp,i);
79072805
LW
5936 break;
5937 }
5938 }
5939 }
c07a80fd 5940
efd8b2ba 5941return_string_or_null:
c07a80fd 5942 return (SvCUR(sv) - append) ? SvPVX(sv) : Nullch;
79072805
LW
5943}
5944
954c1994
GS
5945/*
5946=for apidoc sv_inc
5947
645c22ef
DM
5948Auto-increment of the value in the SV, doing string to numeric conversion
5949if necessary. Handles 'get' magic.
954c1994
GS
5950
5951=cut
5952*/
5953
79072805 5954void
864dbfa3 5955Perl_sv_inc(pTHX_ register SV *sv)
79072805
LW
5956{
5957 register char *d;
463ee0b2 5958 int flags;
79072805
LW
5959
5960 if (!sv)
5961 return;
5b295bef 5962 SvGETMAGIC(sv);
ed6116ce 5963 if (SvTHINKFIRST(sv)) {
765f542d
NC
5964 if (SvIsCOW(sv))
5965 sv_force_normal_flags(sv, 0);
0f15f207 5966 if (SvREADONLY(sv)) {
923e4eb5 5967 if (IN_PERL_RUNTIME)
cea2e8a9 5968 Perl_croak(aTHX_ PL_no_modify);
0f15f207 5969 }
a0d0e21e 5970 if (SvROK(sv)) {
b5be31e9 5971 IV i;
9e7bc3e8
JD
5972 if (SvAMAGIC(sv) && AMG_CALLun(sv,inc))
5973 return;
56431972 5974 i = PTR2IV(SvRV(sv));
b5be31e9
SM
5975 sv_unref(sv);
5976 sv_setiv(sv, i);
a0d0e21e 5977 }
ed6116ce 5978 }
8990e307 5979 flags = SvFLAGS(sv);
28e5dec8
JH
5980 if ((flags & (SVp_NOK|SVp_IOK)) == SVp_NOK) {
5981 /* It's (privately or publicly) a float, but not tested as an
5982 integer, so test it to see. */
d460ef45 5983 (void) SvIV(sv);
28e5dec8
JH
5984 flags = SvFLAGS(sv);
5985 }
5986 if ((flags & SVf_IOK) || ((flags & (SVp_IOK | SVp_NOK)) == SVp_IOK)) {
5987 /* It's publicly an integer, or privately an integer-not-float */
59d8ce62 5988#ifdef PERL_PRESERVE_IVUV
28e5dec8 5989 oops_its_int:
59d8ce62 5990#endif
25da4f38
IZ
5991 if (SvIsUV(sv)) {
5992 if (SvUVX(sv) == UV_MAX)
a1e868e7 5993 sv_setnv(sv, UV_MAX_P1);
25da4f38
IZ
5994 else
5995 (void)SvIOK_only_UV(sv);
607fa7f2 5996 SvUV_set(sv, SvUVX(sv) + 1);
25da4f38
IZ
5997 } else {
5998 if (SvIVX(sv) == IV_MAX)
28e5dec8 5999 sv_setuv(sv, (UV)IV_MAX + 1);
25da4f38
IZ
6000 else {
6001 (void)SvIOK_only(sv);
45977657 6002 SvIV_set(sv, SvIVX(sv) + 1);
1c846c1f 6003 }
55497cff 6004 }
79072805
LW
6005 return;
6006 }
28e5dec8
JH
6007 if (flags & SVp_NOK) {
6008 (void)SvNOK_only(sv);
9d6ce603 6009 SvNV_set(sv, SvNVX(sv) + 1.0);
28e5dec8
JH
6010 return;
6011 }
6012
3f7c398e 6013 if (!(flags & SVp_POK) || !*SvPVX_const(sv)) {
28e5dec8 6014 if ((flags & SVTYPEMASK) < SVt_PVIV)
f5282e15 6015 sv_upgrade(sv, ((flags & SVTYPEMASK) > SVt_IV ? SVt_PVIV : SVt_IV));
28e5dec8 6016 (void)SvIOK_only(sv);
45977657 6017 SvIV_set(sv, 1);
79072805
LW
6018 return;
6019 }
463ee0b2 6020 d = SvPVX(sv);
79072805
LW
6021 while (isALPHA(*d)) d++;
6022 while (isDIGIT(*d)) d++;
6023 if (*d) {
28e5dec8 6024#ifdef PERL_PRESERVE_IVUV
d1be9408 6025 /* Got to punt this as an integer if needs be, but we don't issue
28e5dec8
JH
6026 warnings. Probably ought to make the sv_iv_please() that does
6027 the conversion if possible, and silently. */
504618e9 6028 const int numtype = grok_number(SvPVX_const(sv), SvCUR(sv), NULL);
28e5dec8
JH
6029 if (numtype && !(numtype & IS_NUMBER_INFINITY)) {
6030 /* Need to try really hard to see if it's an integer.
6031 9.22337203685478e+18 is an integer.
6032 but "9.22337203685478e+18" + 0 is UV=9223372036854779904
6033 so $a="9.22337203685478e+18"; $a+0; $a++
6034 needs to be the same as $a="9.22337203685478e+18"; $a++
6035 or we go insane. */
d460ef45 6036
28e5dec8
JH
6037 (void) sv_2iv(sv);
6038 if (SvIOK(sv))
6039 goto oops_its_int;
6040
6041 /* sv_2iv *should* have made this an NV */
6042 if (flags & SVp_NOK) {
6043 (void)SvNOK_only(sv);
9d6ce603 6044 SvNV_set(sv, SvNVX(sv) + 1.0);
28e5dec8
JH
6045 return;
6046 }
6047 /* I don't think we can get here. Maybe I should assert this
6048 And if we do get here I suspect that sv_setnv will croak. NWC
6049 Fall through. */
6050#if defined(USE_LONG_DOUBLE)
6051 DEBUG_c(PerlIO_printf(Perl_debug_log,"sv_inc punt failed to convert '%s' to IOK or NOKp, UV=0x%"UVxf" NV=%"PERL_PRIgldbl"\n",
3f7c398e 6052 SvPVX_const(sv), SvIVX(sv), SvNVX(sv)));
28e5dec8 6053#else
1779d84d 6054 DEBUG_c(PerlIO_printf(Perl_debug_log,"sv_inc punt failed to convert '%s' to IOK or NOKp, UV=0x%"UVxf" NV=%"NVgf"\n",
3f7c398e 6055 SvPVX_const(sv), SvIVX(sv), SvNVX(sv)));
28e5dec8
JH
6056#endif
6057 }
6058#endif /* PERL_PRESERVE_IVUV */
3f7c398e 6059 sv_setnv(sv,Atof(SvPVX_const(sv)) + 1.0);
79072805
LW
6060 return;
6061 }
6062 d--;
3f7c398e 6063 while (d >= SvPVX_const(sv)) {
79072805
LW
6064 if (isDIGIT(*d)) {
6065 if (++*d <= '9')
6066 return;
6067 *(d--) = '0';
6068 }
6069 else {
9d116dd7
JH
6070#ifdef EBCDIC
6071 /* MKS: The original code here died if letters weren't consecutive.
6072 * at least it didn't have to worry about non-C locales. The
6073 * new code assumes that ('z'-'a')==('Z'-'A'), letters are
1c846c1f 6074 * arranged in order (although not consecutively) and that only
9d116dd7
JH
6075 * [A-Za-z] are accepted by isALPHA in the C locale.
6076 */
6077 if (*d != 'z' && *d != 'Z') {
6078 do { ++*d; } while (!isALPHA(*d));
6079 return;
6080 }
6081 *(d--) -= 'z' - 'a';
6082#else
79072805
LW
6083 ++*d;
6084 if (isALPHA(*d))
6085 return;
6086 *(d--) -= 'z' - 'a' + 1;
9d116dd7 6087#endif
79072805
LW
6088 }
6089 }
6090 /* oh,oh, the number grew */
6091 SvGROW(sv, SvCUR(sv) + 2);
b162af07 6092 SvCUR_set(sv, SvCUR(sv) + 1);
3f7c398e 6093 for (d = SvPVX(sv) + SvCUR(sv); d > SvPVX_const(sv); d--)
79072805
LW
6094 *d = d[-1];
6095 if (isDIGIT(d[1]))
6096 *d = '1';
6097 else
6098 *d = d[1];
6099}
6100
954c1994
GS
6101/*
6102=for apidoc sv_dec
6103
645c22ef
DM
6104Auto-decrement of the value in the SV, doing string to numeric conversion
6105if necessary. Handles 'get' magic.
954c1994
GS
6106
6107=cut
6108*/
6109
79072805 6110void
864dbfa3 6111Perl_sv_dec(pTHX_ register SV *sv)
79072805 6112{
463ee0b2
LW
6113 int flags;
6114
79072805
LW
6115 if (!sv)
6116 return;
5b295bef 6117 SvGETMAGIC(sv);
ed6116ce 6118 if (SvTHINKFIRST(sv)) {
765f542d
NC
6119 if (SvIsCOW(sv))
6120 sv_force_normal_flags(sv, 0);
0f15f207 6121 if (SvREADONLY(sv)) {
923e4eb5 6122 if (IN_PERL_RUNTIME)
cea2e8a9 6123 Perl_croak(aTHX_ PL_no_modify);
0f15f207 6124 }
a0d0e21e 6125 if (SvROK(sv)) {
b5be31e9 6126 IV i;
9e7bc3e8
JD
6127 if (SvAMAGIC(sv) && AMG_CALLun(sv,dec))
6128 return;
56431972 6129 i = PTR2IV(SvRV(sv));
b5be31e9
SM
6130 sv_unref(sv);
6131 sv_setiv(sv, i);
a0d0e21e 6132 }
ed6116ce 6133 }
28e5dec8
JH
6134 /* Unlike sv_inc we don't have to worry about string-never-numbers
6135 and keeping them magic. But we mustn't warn on punting */
8990e307 6136 flags = SvFLAGS(sv);
28e5dec8
JH
6137 if ((flags & SVf_IOK) || ((flags & (SVp_IOK | SVp_NOK)) == SVp_IOK)) {
6138 /* It's publicly an integer, or privately an integer-not-float */
59d8ce62 6139#ifdef PERL_PRESERVE_IVUV
28e5dec8 6140 oops_its_int:
59d8ce62 6141#endif
25da4f38
IZ
6142 if (SvIsUV(sv)) {
6143 if (SvUVX(sv) == 0) {
6144 (void)SvIOK_only(sv);
45977657 6145 SvIV_set(sv, -1);
25da4f38
IZ
6146 }
6147 else {
6148 (void)SvIOK_only_UV(sv);
f4eee32f 6149 SvUV_set(sv, SvUVX(sv) - 1);
1c846c1f 6150 }
25da4f38
IZ
6151 } else {
6152 if (SvIVX(sv) == IV_MIN)
65202027 6153 sv_setnv(sv, (NV)IV_MIN - 1.0);
25da4f38
IZ
6154 else {
6155 (void)SvIOK_only(sv);
45977657 6156 SvIV_set(sv, SvIVX(sv) - 1);
1c846c1f 6157 }
55497cff 6158 }
6159 return;
6160 }
28e5dec8 6161 if (flags & SVp_NOK) {
9d6ce603 6162 SvNV_set(sv, SvNVX(sv) - 1.0);
28e5dec8
JH
6163 (void)SvNOK_only(sv);
6164 return;
6165 }
8990e307 6166 if (!(flags & SVp_POK)) {
ef088171
NC
6167 if ((flags & SVTYPEMASK) < SVt_PVIV)
6168 sv_upgrade(sv, ((flags & SVTYPEMASK) > SVt_IV) ? SVt_PVIV : SVt_IV);
6169 SvIV_set(sv, -1);
6170 (void)SvIOK_only(sv);
79072805
LW
6171 return;
6172 }
28e5dec8
JH
6173#ifdef PERL_PRESERVE_IVUV
6174 {
504618e9 6175 const int numtype = grok_number(SvPVX_const(sv), SvCUR(sv), NULL);
28e5dec8
JH
6176 if (numtype && !(numtype & IS_NUMBER_INFINITY)) {
6177 /* Need to try really hard to see if it's an integer.
6178 9.22337203685478e+18 is an integer.
6179 but "9.22337203685478e+18" + 0 is UV=9223372036854779904
6180 so $a="9.22337203685478e+18"; $a+0; $a--
6181 needs to be the same as $a="9.22337203685478e+18"; $a--
6182 or we go insane. */
d460ef45 6183
28e5dec8
JH
6184 (void) sv_2iv(sv);
6185 if (SvIOK(sv))
6186 goto oops_its_int;
6187
6188 /* sv_2iv *should* have made this an NV */
6189 if (flags & SVp_NOK) {
6190 (void)SvNOK_only(sv);
9d6ce603 6191 SvNV_set(sv, SvNVX(sv) - 1.0);
28e5dec8
JH
6192 return;
6193 }
6194 /* I don't think we can get here. Maybe I should assert this
6195 And if we do get here I suspect that sv_setnv will croak. NWC
6196 Fall through. */
6197#if defined(USE_LONG_DOUBLE)
6198 DEBUG_c(PerlIO_printf(Perl_debug_log,"sv_dec punt failed to convert '%s' to IOK or NOKp, UV=0x%"UVxf" NV=%"PERL_PRIgldbl"\n",
3f7c398e 6199 SvPVX_const(sv), SvIVX(sv), SvNVX(sv)));
28e5dec8 6200#else
1779d84d 6201 DEBUG_c(PerlIO_printf(Perl_debug_log,"sv_dec punt failed to convert '%s' to IOK or NOKp, UV=0x%"UVxf" NV=%"NVgf"\n",
3f7c398e 6202 SvPVX_const(sv), SvIVX(sv), SvNVX(sv)));
28e5dec8
JH
6203#endif
6204 }
6205 }
6206#endif /* PERL_PRESERVE_IVUV */
3f7c398e 6207 sv_setnv(sv,Atof(SvPVX_const(sv)) - 1.0); /* punt */
79072805
LW
6208}
6209
954c1994
GS
6210/*
6211=for apidoc sv_mortalcopy
6212
645c22ef 6213Creates a new SV which is a copy of the original SV (using C<sv_setsv>).
d4236ebc
DM
6214The new SV is marked as mortal. It will be destroyed "soon", either by an
6215explicit call to FREETMPS, or by an implicit call at places such as
6216statement boundaries. See also C<sv_newmortal> and C<sv_2mortal>.
954c1994
GS
6217
6218=cut
6219*/
6220
79072805
LW
6221/* Make a string that will exist for the duration of the expression
6222 * evaluation. Actually, it may have to last longer than that, but
6223 * hopefully we won't free it until it has been assigned to a
6224 * permanent location. */
6225
6226SV *
864dbfa3 6227Perl_sv_mortalcopy(pTHX_ SV *oldstr)
79072805 6228{
463ee0b2 6229 register SV *sv;
b881518d 6230
4561caa4 6231 new_SV(sv);
79072805 6232 sv_setsv(sv,oldstr);
677b06e3
GS
6233 EXTEND_MORTAL(1);
6234 PL_tmps_stack[++PL_tmps_ix] = sv;
8990e307
LW
6235 SvTEMP_on(sv);
6236 return sv;
6237}
6238
954c1994
GS
6239/*
6240=for apidoc sv_newmortal
6241
645c22ef 6242Creates a new null SV which is mortal. The reference count of the SV is
d4236ebc
DM
6243set to 1. It will be destroyed "soon", either by an explicit call to
6244FREETMPS, or by an implicit call at places such as statement boundaries.
6245See also C<sv_mortalcopy> and C<sv_2mortal>.
954c1994
GS
6246
6247=cut
6248*/
6249
8990e307 6250SV *
864dbfa3 6251Perl_sv_newmortal(pTHX)
8990e307
LW
6252{
6253 register SV *sv;
6254
4561caa4 6255 new_SV(sv);
8990e307 6256 SvFLAGS(sv) = SVs_TEMP;
677b06e3
GS
6257 EXTEND_MORTAL(1);
6258 PL_tmps_stack[++PL_tmps_ix] = sv;
79072805
LW
6259 return sv;
6260}
6261
954c1994
GS
6262/*
6263=for apidoc sv_2mortal
6264
d4236ebc
DM
6265Marks an existing SV as mortal. The SV will be destroyed "soon", either
6266by an explicit call to FREETMPS, or by an implicit call at places such as
37d2ac18
NC
6267statement boundaries. SvTEMP() is turned on which means that the SV's
6268string buffer can be "stolen" if this SV is copied. See also C<sv_newmortal>
6269and C<sv_mortalcopy>.
954c1994
GS
6270
6271=cut
6272*/
6273
79072805 6274SV *
864dbfa3 6275Perl_sv_2mortal(pTHX_ register SV *sv)
79072805 6276{
27da23d5 6277 dVAR;
79072805
LW
6278 if (!sv)
6279 return sv;
d689ffdd 6280 if (SvREADONLY(sv) && SvIMMORTAL(sv))
11162842 6281 return sv;
677b06e3
GS
6282 EXTEND_MORTAL(1);
6283 PL_tmps_stack[++PL_tmps_ix] = sv;
8990e307 6284 SvTEMP_on(sv);
79072805
LW
6285 return sv;
6286}
6287
954c1994
GS
6288/*
6289=for apidoc newSVpv
6290
6291Creates a new SV and copies a string into it. The reference count for the
6292SV is set to 1. If C<len> is zero, Perl will compute the length using
6293strlen(). For efficiency, consider using C<newSVpvn> instead.
6294
6295=cut
6296*/
6297
79072805 6298SV *
864dbfa3 6299Perl_newSVpv(pTHX_ const char *s, STRLEN len)
79072805 6300{
463ee0b2 6301 register SV *sv;
79072805 6302
4561caa4 6303 new_SV(sv);
616d8c9c 6304 sv_setpvn(sv,s,len ? len : strlen(s));
79072805
LW
6305 return sv;
6306}
6307
954c1994
GS
6308/*
6309=for apidoc newSVpvn
6310
6311Creates a new SV and copies a string into it. The reference count for the
1c846c1f 6312SV is set to 1. Note that if C<len> is zero, Perl will create a zero length
954c1994 6313string. You are responsible for ensuring that the source string is at least
9e09f5f2 6314C<len> bytes long. If the C<s> argument is NULL the new SV will be undefined.
954c1994
GS
6315
6316=cut
6317*/
6318
9da1e3b5 6319SV *
864dbfa3 6320Perl_newSVpvn(pTHX_ const char *s, STRLEN len)
9da1e3b5
MUN
6321{
6322 register SV *sv;
6323
6324 new_SV(sv);
9da1e3b5
MUN
6325 sv_setpvn(sv,s,len);
6326 return sv;
6327}
6328
bd08039b
NC
6329
6330/*
926f8064 6331=for apidoc newSVhek
bd08039b
NC
6332
6333Creates a new SV from the hash key structure. It will generate scalars that
5aaec2b4
NC
6334point to the shared string table where possible. Returns a new (undefined)
6335SV if the hek is NULL.
bd08039b
NC
6336
6337=cut
6338*/
6339
6340SV *
c1b02ed8 6341Perl_newSVhek(pTHX_ const HEK *hek)
bd08039b 6342{
5aaec2b4
NC
6343 if (!hek) {
6344 SV *sv;
6345
6346 new_SV(sv);
6347 return sv;
6348 }
6349
bd08039b
NC
6350 if (HEK_LEN(hek) == HEf_SVKEY) {
6351 return newSVsv(*(SV**)HEK_KEY(hek));
6352 } else {
6353 const int flags = HEK_FLAGS(hek);
6354 if (flags & HVhek_WASUTF8) {
6355 /* Trouble :-)
6356 Andreas would like keys he put in as utf8 to come back as utf8
6357 */
6358 STRLEN utf8_len = HEK_LEN(hek);
b64e5050
AL
6359 const U8 *as_utf8 = bytes_to_utf8 ((U8*)HEK_KEY(hek), &utf8_len);
6360 SV * const sv = newSVpvn ((const char*)as_utf8, utf8_len);
bd08039b
NC
6361
6362 SvUTF8_on (sv);
6363 Safefree (as_utf8); /* bytes_to_utf8() allocates a new string */
6364 return sv;
6365 } else if (flags & HVhek_REHASH) {
6366 /* We don't have a pointer to the hv, so we have to replicate the
6367 flag into every HEK. This hv is using custom a hasing
6368 algorithm. Hence we can't return a shared string scalar, as
6369 that would contain the (wrong) hash value, and might get passed
6370 into an hv routine with a regular hash */
6371
b64e5050 6372 SV * const sv = newSVpvn (HEK_KEY(hek), HEK_LEN(hek));
bd08039b
NC
6373 if (HEK_UTF8(hek))
6374 SvUTF8_on (sv);
6375 return sv;
6376 }
6377 /* This will be overwhelminly the most common case. */
6378 return newSVpvn_share(HEK_KEY(hek),
6379 (HEK_UTF8(hek) ? -HEK_LEN(hek) : HEK_LEN(hek)),
6380 HEK_HASH(hek));
6381 }
6382}
6383
1c846c1f
NIS
6384/*
6385=for apidoc newSVpvn_share
6386
3f7c398e 6387Creates a new SV with its SvPVX_const pointing to a shared string in the string
645c22ef
DM
6388table. If the string does not already exist in the table, it is created
6389first. Turns on READONLY and FAKE. The string's hash is stored in the UV
6390slot of the SV; if the C<hash> parameter is non-zero, that value is used;
6391otherwise the hash is computed. The idea here is that as the string table
3f7c398e 6392is used for shared hash keys these strings will have SvPVX_const == HeKEY and
645c22ef 6393hash lookup will avoid string compare.
1c846c1f
NIS
6394
6395=cut
6396*/
6397
6398SV *
c3654f1a 6399Perl_newSVpvn_share(pTHX_ const char *src, I32 len, U32 hash)
1c846c1f
NIS
6400{
6401 register SV *sv;
c3654f1a
IH
6402 bool is_utf8 = FALSE;
6403 if (len < 0) {
77caf834 6404 STRLEN tmplen = -len;
c3654f1a 6405 is_utf8 = TRUE;
75a54232 6406 /* See the note in hv.c:hv_fetch() --jhi */
e1ec3a88 6407 src = (char*)bytes_from_utf8((const U8*)src, &tmplen, &is_utf8);
75a54232
JH
6408 len = tmplen;
6409 }
1c846c1f 6410 if (!hash)
5afd6d42 6411 PERL_HASH(hash, src, len);
1c846c1f 6412 new_SV(sv);
bdd68bc3 6413 sv_upgrade(sv, SVt_PV);
f880fe2f 6414 SvPV_set(sv, sharepvn(src, is_utf8?-len:len, hash));
b162af07 6415 SvCUR_set(sv, len);
b162af07 6416 SvLEN_set(sv, 0);
1c846c1f
NIS
6417 SvREADONLY_on(sv);
6418 SvFAKE_on(sv);
6419 SvPOK_on(sv);
c3654f1a
IH
6420 if (is_utf8)
6421 SvUTF8_on(sv);
1c846c1f
NIS
6422 return sv;
6423}
6424
645c22ef 6425
cea2e8a9 6426#if defined(PERL_IMPLICIT_CONTEXT)
645c22ef
DM
6427
6428/* pTHX_ magic can't cope with varargs, so this is a no-context
6429 * version of the main function, (which may itself be aliased to us).
6430 * Don't access this version directly.
6431 */
6432
46fc3d4c 6433SV *
cea2e8a9 6434Perl_newSVpvf_nocontext(const char* pat, ...)
46fc3d4c 6435{
cea2e8a9 6436 dTHX;
46fc3d4c 6437 register SV *sv;
6438 va_list args;
46fc3d4c 6439 va_start(args, pat);
c5be433b 6440 sv = vnewSVpvf(pat, &args);
46fc3d4c 6441 va_end(args);
6442 return sv;
6443}
cea2e8a9 6444#endif
46fc3d4c 6445
954c1994
GS
6446/*
6447=for apidoc newSVpvf
6448
645c22ef 6449Creates a new SV and initializes it with the string formatted like
954c1994
GS
6450C<sprintf>.
6451
6452=cut
6453*/
6454
cea2e8a9
GS
6455SV *
6456Perl_newSVpvf(pTHX_ const char* pat, ...)
6457{
6458 register SV *sv;
6459 va_list args;
cea2e8a9 6460 va_start(args, pat);
c5be433b 6461 sv = vnewSVpvf(pat, &args);
cea2e8a9
GS
6462 va_end(args);
6463 return sv;
6464}
46fc3d4c 6465
645c22ef
DM
6466/* backend for newSVpvf() and newSVpvf_nocontext() */
6467
79072805 6468SV *
c5be433b
GS
6469Perl_vnewSVpvf(pTHX_ const char* pat, va_list* args)
6470{
6471 register SV *sv;
6472 new_SV(sv);
6473 sv_vsetpvfn(sv, pat, strlen(pat), args, Null(SV**), 0, Null(bool*));
6474 return sv;
6475}
6476
954c1994
GS
6477/*
6478=for apidoc newSVnv
6479
6480Creates a new SV and copies a floating point value into it.
6481The reference count for the SV is set to 1.
6482
6483=cut
6484*/
6485
c5be433b 6486SV *
65202027 6487Perl_newSVnv(pTHX_ NV n)
79072805 6488{
463ee0b2 6489 register SV *sv;
79072805 6490
4561caa4 6491 new_SV(sv);
79072805
LW
6492 sv_setnv(sv,n);
6493 return sv;
6494}
6495
954c1994
GS
6496/*
6497=for apidoc newSViv
6498
6499Creates a new SV and copies an integer into it. The reference count for the
6500SV is set to 1.
6501
6502=cut
6503*/
6504
79072805 6505SV *
864dbfa3 6506Perl_newSViv(pTHX_ IV i)
79072805 6507{
463ee0b2 6508 register SV *sv;
79072805 6509
4561caa4 6510 new_SV(sv);
79072805
LW
6511 sv_setiv(sv,i);
6512 return sv;
6513}
6514
954c1994 6515/*
1a3327fb
JH
6516=for apidoc newSVuv
6517
6518Creates a new SV and copies an unsigned integer into it.
6519The reference count for the SV is set to 1.
6520
6521=cut
6522*/
6523
6524SV *
6525Perl_newSVuv(pTHX_ UV u)
6526{
6527 register SV *sv;
6528
6529 new_SV(sv);
6530 sv_setuv(sv,u);
6531 return sv;
6532}
6533
6534/*
954c1994
GS
6535=for apidoc newRV_noinc
6536
6537Creates an RV wrapper for an SV. The reference count for the original
6538SV is B<not> incremented.
6539
6540=cut
6541*/
6542
2304df62 6543SV *
864dbfa3 6544Perl_newRV_noinc(pTHX_ SV *tmpRef)
2304df62
AD
6545{
6546 register SV *sv;
6547
4561caa4 6548 new_SV(sv);
2304df62 6549 sv_upgrade(sv, SVt_RV);
76e3520e 6550 SvTEMP_off(tmpRef);
b162af07 6551 SvRV_set(sv, tmpRef);
2304df62 6552 SvROK_on(sv);
2304df62
AD
6553 return sv;
6554}
6555
ff276b08 6556/* newRV_inc is the official function name to use now.
645c22ef
DM
6557 * newRV_inc is in fact #defined to newRV in sv.h
6558 */
6559
5f05dabc 6560SV *
864dbfa3 6561Perl_newRV(pTHX_ SV *tmpRef)
5f05dabc 6562{
5f6447b6 6563 return newRV_noinc(SvREFCNT_inc(tmpRef));
5f05dabc 6564}
5f05dabc 6565
954c1994
GS
6566/*
6567=for apidoc newSVsv
6568
6569Creates a new SV which is an exact duplicate of the original SV.
645c22ef 6570(Uses C<sv_setsv>).
954c1994
GS
6571
6572=cut
6573*/
6574
79072805 6575SV *
864dbfa3 6576Perl_newSVsv(pTHX_ register SV *old)
79072805 6577{
463ee0b2 6578 register SV *sv;
79072805
LW
6579
6580 if (!old)
6581 return Nullsv;
8990e307 6582 if (SvTYPE(old) == SVTYPEMASK) {
0453d815 6583 if (ckWARN_d(WARN_INTERNAL))
9014280d 6584 Perl_warner(aTHX_ packWARN(WARN_INTERNAL), "semi-panic: attempt to dup freed string");
79072805
LW
6585 return Nullsv;
6586 }
4561caa4 6587 new_SV(sv);
e90aabeb
NC
6588 /* SV_GMAGIC is the default for sv_setv()
6589 SV_NOSTEAL prevents TEMP buffers being, well, stolen, and saves games
6590 with SvTEMP_off and SvTEMP_on round a call to sv_setsv. */
6591 sv_setsv_flags(sv, old, SV_GMAGIC | SV_NOSTEAL);
463ee0b2 6592 return sv;
79072805
LW
6593}
6594
645c22ef
DM
6595/*
6596=for apidoc sv_reset
6597
6598Underlying implementation for the C<reset> Perl function.
6599Note that the perl-level function is vaguely deprecated.
6600
6601=cut
6602*/
6603
79072805 6604void
e1ec3a88 6605Perl_sv_reset(pTHX_ register const char *s, HV *stash)
79072805 6606{
27da23d5 6607 dVAR;
4802d5d7 6608 char todo[PERL_UCHAR_MAX+1];
79072805 6609
49d8d3a1
MB
6610 if (!stash)
6611 return;
6612
79072805 6613 if (!*s) { /* reset ?? searches */
aec46f14 6614 MAGIC * const mg = mg_find((SV *)stash, PERL_MAGIC_symtab);
8d2f4536
NC
6615 if (mg) {
6616 PMOP *pm = (PMOP *) mg->mg_obj;
6617 while (pm) {
6618 pm->op_pmdynflags &= ~PMdf_USED;
6619 pm = pm->op_pmnext;
6620 }
79072805
LW
6621 }
6622 return;
6623 }
6624
6625 /* reset variables */
6626
6627 if (!HvARRAY(stash))
6628 return;
463ee0b2
LW
6629
6630 Zero(todo, 256, char);
79072805 6631 while (*s) {
b464bac0
AL
6632 I32 max;
6633 I32 i = (unsigned char)*s;
79072805
LW
6634 if (s[1] == '-') {
6635 s += 2;
6636 }
4802d5d7 6637 max = (unsigned char)*s++;
79072805 6638 for ( ; i <= max; i++) {
463ee0b2
LW
6639 todo[i] = 1;
6640 }
a0d0e21e 6641 for (i = 0; i <= (I32) HvMAX(stash); i++) {
b464bac0 6642 HE *entry;
79072805 6643 for (entry = HvARRAY(stash)[i];
9e35f4b3
GS
6644 entry;
6645 entry = HeNEXT(entry))
6646 {
b464bac0
AL
6647 register GV *gv;
6648 register SV *sv;
6649
1edc1566 6650 if (!todo[(U8)*HeKEY(entry)])
463ee0b2 6651 continue;
1edc1566 6652 gv = (GV*)HeVAL(entry);
79072805 6653 sv = GvSV(gv);
e203899d
NC
6654 if (sv) {
6655 if (SvTHINKFIRST(sv)) {
6656 if (!SvREADONLY(sv) && SvROK(sv))
6657 sv_unref(sv);
6658 /* XXX Is this continue a bug? Why should THINKFIRST
6659 exempt us from resetting arrays and hashes? */
6660 continue;
6661 }
6662 SvOK_off(sv);
6663 if (SvTYPE(sv) >= SVt_PV) {
6664 SvCUR_set(sv, 0);
6665 if (SvPVX_const(sv) != Nullch)
6666 *SvPVX(sv) = '\0';
6667 SvTAINT(sv);
6668 }
79072805
LW
6669 }
6670 if (GvAV(gv)) {
6671 av_clear(GvAV(gv));
6672 }
bfcb3514 6673 if (GvHV(gv) && !HvNAME_get(GvHV(gv))) {
b0269e46
AB
6674#if defined(VMS)
6675 Perl_die(aTHX_ "Can't reset %%ENV on this system");
6676#else /* ! VMS */
463ee0b2 6677 hv_clear(GvHV(gv));
b0269e46
AB
6678# if defined(USE_ENVIRON_ARRAY)
6679 if (gv == PL_envgv)
6680 my_clearenv();
6681# endif /* USE_ENVIRON_ARRAY */
6682#endif /* VMS */
79072805
LW
6683 }
6684 }
6685 }
6686 }
6687}
6688
645c22ef
DM
6689/*
6690=for apidoc sv_2io
6691
6692Using various gambits, try to get an IO from an SV: the IO slot if its a
6693GV; or the recursive result if we're an RV; or the IO slot of the symbol
6694named after the PV if we're a string.
6695
6696=cut
6697*/
6698
46fc3d4c 6699IO*
864dbfa3 6700Perl_sv_2io(pTHX_ SV *sv)
46fc3d4c 6701{
6702 IO* io;
6703 GV* gv;
6704
6705 switch (SvTYPE(sv)) {
6706 case SVt_PVIO:
6707 io = (IO*)sv;
6708 break;
6709 case SVt_PVGV:
6710 gv = (GV*)sv;
6711 io = GvIO(gv);
6712 if (!io)
cea2e8a9 6713 Perl_croak(aTHX_ "Bad filehandle: %s", GvNAME(gv));
46fc3d4c 6714 break;
6715 default:
6716 if (!SvOK(sv))
cea2e8a9 6717 Perl_croak(aTHX_ PL_no_usym, "filehandle");
46fc3d4c 6718 if (SvROK(sv))
6719 return sv_2io(SvRV(sv));
f776e3cd 6720 gv = gv_fetchsv(sv, 0, SVt_PVIO);
46fc3d4c 6721 if (gv)
6722 io = GvIO(gv);
6723 else
6724 io = 0;
6725 if (!io)
35c1215d 6726 Perl_croak(aTHX_ "Bad filehandle: %"SVf, sv);
46fc3d4c 6727 break;
6728 }
6729 return io;
6730}
6731
645c22ef
DM
6732/*
6733=for apidoc sv_2cv
6734
6735Using various gambits, try to get a CV from an SV; in addition, try if
6736possible to set C<*st> and C<*gvp> to the stash and GV associated with it.
f2c0649b 6737The flags in C<lref> are passed to sv_fetchsv.
645c22ef
DM
6738
6739=cut
6740*/
6741
79072805 6742CV *
864dbfa3 6743Perl_sv_2cv(pTHX_ SV *sv, HV **st, GV **gvp, I32 lref)
79072805 6744{
27da23d5 6745 dVAR;
c04a4dfe
JH
6746 GV *gv = Nullgv;
6747 CV *cv = Nullcv;
79072805
LW
6748
6749 if (!sv)
ef58ba18 6750 return *st = NULL, *gvp = Nullgv, Nullcv;
79072805 6751 switch (SvTYPE(sv)) {
79072805
LW
6752 case SVt_PVCV:
6753 *st = CvSTASH(sv);
6754 *gvp = Nullgv;
6755 return (CV*)sv;
6756 case SVt_PVHV:
6757 case SVt_PVAV:
ef58ba18 6758 *st = NULL;
79072805
LW
6759 *gvp = Nullgv;
6760 return Nullcv;
8990e307
LW
6761 case SVt_PVGV:
6762 gv = (GV*)sv;
a0d0e21e 6763 *gvp = gv;
8990e307
LW
6764 *st = GvESTASH(gv);
6765 goto fix_gv;
6766
79072805 6767 default:
5b295bef 6768 SvGETMAGIC(sv);
a0d0e21e 6769 if (SvROK(sv)) {
823a54a3 6770 SV * const *sp = &sv; /* Used in tryAMAGICunDEREF macro. */
f5284f61
IZ
6771 tryAMAGICunDEREF(to_cv);
6772
62f274bf
GS
6773 sv = SvRV(sv);
6774 if (SvTYPE(sv) == SVt_PVCV) {
6775 cv = (CV*)sv;
6776 *gvp = Nullgv;
6777 *st = CvSTASH(cv);
6778 return cv;
6779 }
6780 else if(isGV(sv))
6781 gv = (GV*)sv;
6782 else
cea2e8a9 6783 Perl_croak(aTHX_ "Not a subroutine reference");
a0d0e21e 6784 }
62f274bf 6785 else if (isGV(sv))
79072805
LW
6786 gv = (GV*)sv;
6787 else
7a5fd60d 6788 gv = gv_fetchsv(sv, lref, SVt_PVCV);
79072805 6789 *gvp = gv;
ef58ba18
NC
6790 if (!gv) {
6791 *st = NULL;
79072805 6792 return Nullcv;
ef58ba18 6793 }
e26df76a
NC
6794 /* Some flags to gv_fetchsv mean don't really create the GV */
6795 if (SvTYPE(gv) != SVt_PVGV) {
6796 *st = NULL;
6797 return NULL;
6798 }
79072805 6799 *st = GvESTASH(gv);
8990e307 6800 fix_gv:
8ebc5c01 6801 if (lref && !GvCVu(gv)) {
4633a7c4 6802 SV *tmpsv;
748a9306 6803 ENTER;
4633a7c4 6804 tmpsv = NEWSV(704,0);
16660edb 6805 gv_efullname3(tmpsv, gv, Nullch);
f6ec51f7
GS
6806 /* XXX this is probably not what they think they're getting.
6807 * It has the same effect as "sub name;", i.e. just a forward
6808 * declaration! */
774d564b 6809 newSUB(start_subparse(FALSE, 0),
4633a7c4
LW
6810 newSVOP(OP_CONST, 0, tmpsv),
6811 Nullop,
8990e307 6812 Nullop);
748a9306 6813 LEAVE;
8ebc5c01 6814 if (!GvCVu(gv))
35c1215d
NC
6815 Perl_croak(aTHX_ "Unable to create sub named \"%"SVf"\"",
6816 sv);
8990e307 6817 }
8ebc5c01 6818 return GvCVu(gv);
79072805
LW
6819 }
6820}
6821
c461cf8f
JH
6822/*
6823=for apidoc sv_true
6824
6825Returns true if the SV has a true value by Perl's rules.
645c22ef
DM
6826Use the C<SvTRUE> macro instead, which may call C<sv_true()> or may
6827instead use an in-line version.
c461cf8f
JH
6828
6829=cut
6830*/
6831
79072805 6832I32
864dbfa3 6833Perl_sv_true(pTHX_ register SV *sv)
79072805 6834{
8990e307
LW
6835 if (!sv)
6836 return 0;
79072805 6837 if (SvPOK(sv)) {
823a54a3
AL
6838 register const XPV* const tXpv = (XPV*)SvANY(sv);
6839 if (tXpv &&
c2f1de04 6840 (tXpv->xpv_cur > 1 ||
339049b0 6841 (tXpv->xpv_cur && *sv->sv_u.svu_pv != '0')))
79072805
LW
6842 return 1;
6843 else
6844 return 0;
6845 }
6846 else {
6847 if (SvIOK(sv))
463ee0b2 6848 return SvIVX(sv) != 0;
79072805
LW
6849 else {
6850 if (SvNOK(sv))
463ee0b2 6851 return SvNVX(sv) != 0.0;
79072805 6852 else
463ee0b2 6853 return sv_2bool(sv);
79072805
LW
6854 }
6855 }
6856}
79072805 6857
645c22ef 6858/*
c461cf8f
JH
6859=for apidoc sv_pvn_force
6860
6861Get a sensible string out of the SV somehow.
645c22ef
DM
6862A private implementation of the C<SvPV_force> macro for compilers which
6863can't cope with complex macro expressions. Always use the macro instead.
c461cf8f 6864
8d6d96c1
HS
6865=for apidoc sv_pvn_force_flags
6866
6867Get a sensible string out of the SV somehow.
6868If C<flags> has C<SV_GMAGIC> bit set, will C<mg_get> on C<sv> if
6869appropriate, else not. C<sv_pvn_force> and C<sv_pvn_force_nomg> are
6870implemented in terms of this function.
645c22ef
DM
6871You normally want to use the various wrapper macros instead: see
6872C<SvPV_force> and C<SvPV_force_nomg>
8d6d96c1
HS
6873
6874=cut
6875*/
6876
6877char *
6878Perl_sv_pvn_force_flags(pTHX_ SV *sv, STRLEN *lp, I32 flags)
6879{
a0d0e21e 6880
6fc92669 6881 if (SvTHINKFIRST(sv) && !SvROK(sv))
765f542d 6882 sv_force_normal_flags(sv, 0);
1c846c1f 6883
a0d0e21e 6884 if (SvPOK(sv)) {
13c5b33c
NC
6885 if (lp)
6886 *lp = SvCUR(sv);
a0d0e21e
LW
6887 }
6888 else {
a3b680e6 6889 char *s;
13c5b33c
NC
6890 STRLEN len;
6891
4d84ee25 6892 if (SvREADONLY(sv) && !(flags & SV_MUTABLE_RETURN)) {
b64e5050 6893 const char * const ref = sv_reftype(sv,0);
4d84ee25
NC
6894 if (PL_op)
6895 Perl_croak(aTHX_ "Can't coerce readonly %s to string in %s",
b64e5050 6896 ref, OP_NAME(PL_op));
4d84ee25 6897 else
b64e5050 6898 Perl_croak(aTHX_ "Can't coerce readonly %s to string", ref);
4d84ee25 6899 }
b64e5050 6900 if (SvTYPE(sv) > SVt_PVLV && SvTYPE(sv) != SVt_PVFM)
cea2e8a9 6901 Perl_croak(aTHX_ "Can't coerce %s to string in %s", sv_reftype(sv,0),
53e06cf0 6902 OP_NAME(PL_op));
b64e5050 6903 s = sv_2pv_flags(sv, &len, flags);
13c5b33c
NC
6904 if (lp)
6905 *lp = len;
6906
3f7c398e 6907 if (s != SvPVX_const(sv)) { /* Almost, but not quite, sv_setpvn() */
a0d0e21e
LW
6908 if (SvROK(sv))
6909 sv_unref(sv);
862a34c6 6910 SvUPGRADE(sv, SVt_PV); /* Never FALSE */
a0d0e21e 6911 SvGROW(sv, len + 1);
706aa1c9 6912 Move(s,SvPVX(sv),len,char);
a0d0e21e
LW
6913 SvCUR_set(sv, len);
6914 *SvEND(sv) = '\0';
6915 }
6916 if (!SvPOK(sv)) {
6917 SvPOK_on(sv); /* validate pointer */
6918 SvTAINT(sv);
1d7c1841 6919 DEBUG_c(PerlIO_printf(Perl_debug_log, "0x%"UVxf" 2pv(%s)\n",
3f7c398e 6920 PTR2UV(sv),SvPVX_const(sv)));
a0d0e21e
LW
6921 }
6922 }
4d84ee25 6923 return SvPVX_mutable(sv);
a0d0e21e
LW
6924}
6925
645c22ef 6926/*
645c22ef
DM
6927=for apidoc sv_pvbyten_force
6928
0feed65a 6929The backend for the C<SvPVbytex_force> macro. Always use the macro instead.
645c22ef
DM
6930
6931=cut
6932*/
6933
7340a771
GS
6934char *
6935Perl_sv_pvbyten_force(pTHX_ SV *sv, STRLEN *lp)
6936{
46ec2f14 6937 sv_pvn_force(sv,lp);
ffebcc3e 6938 sv_utf8_downgrade(sv,0);
46ec2f14
TS
6939 *lp = SvCUR(sv);
6940 return SvPVX(sv);
7340a771
GS
6941}
6942
645c22ef 6943/*
c461cf8f
JH
6944=for apidoc sv_pvutf8n_force
6945
0feed65a 6946The backend for the C<SvPVutf8x_force> macro. Always use the macro instead.
c461cf8f
JH
6947
6948=cut
6949*/
6950
7340a771
GS
6951char *
6952Perl_sv_pvutf8n_force(pTHX_ SV *sv, STRLEN *lp)
6953{
46ec2f14 6954 sv_pvn_force(sv,lp);
560a288e 6955 sv_utf8_upgrade(sv);
46ec2f14
TS
6956 *lp = SvCUR(sv);
6957 return SvPVX(sv);
7340a771
GS
6958}
6959
c461cf8f
JH
6960/*
6961=for apidoc sv_reftype
6962
6963Returns a string describing what the SV is a reference to.
6964
6965=cut
6966*/
6967
1cb0ed9b 6968char *
bfed75c6 6969Perl_sv_reftype(pTHX_ const SV *sv, int ob)
a0d0e21e 6970{
07409e01
NC
6971 /* The fact that I don't need to downcast to char * everywhere, only in ?:
6972 inside return suggests a const propagation bug in g++. */
c86bf373 6973 if (ob && SvOBJECT(sv)) {
1b6737cc 6974 char * const name = HvNAME_get(SvSTASH(sv));
07409e01 6975 return name ? name : (char *) "__ANON__";
c86bf373 6976 }
a0d0e21e
LW
6977 else {
6978 switch (SvTYPE(sv)) {
6979 case SVt_NULL:
6980 case SVt_IV:
6981 case SVt_NV:
6982 case SVt_RV:
6983 case SVt_PV:
6984 case SVt_PVIV:
6985 case SVt_PVNV:
6986 case SVt_PVMG:
6987 case SVt_PVBM:
1cb0ed9b 6988 if (SvVOK(sv))
439cb1c4 6989 return "VSTRING";
a0d0e21e
LW
6990 if (SvROK(sv))
6991 return "REF";
6992 else
6993 return "SCALAR";
1cb0ed9b 6994
07409e01 6995 case SVt_PVLV: return (char *) (SvROK(sv) ? "REF"
be65207d
DM
6996 /* tied lvalues should appear to be
6997 * scalars for backwards compatitbility */
6998 : (LvTYPE(sv) == 't' || LvTYPE(sv) == 'T')
07409e01 6999 ? "SCALAR" : "LVALUE");
a0d0e21e
LW
7000 case SVt_PVAV: return "ARRAY";
7001 case SVt_PVHV: return "HASH";
7002 case SVt_PVCV: return "CODE";
7003 case SVt_PVGV: return "GLOB";
1d2dff63 7004 case SVt_PVFM: return "FORMAT";
27f9d8f3 7005 case SVt_PVIO: return "IO";
a0d0e21e
LW
7006 default: return "UNKNOWN";
7007 }
7008 }
7009}
7010
954c1994
GS
7011/*
7012=for apidoc sv_isobject
7013
7014Returns a boolean indicating whether the SV is an RV pointing to a blessed
7015object. If the SV is not an RV, or if the object is not blessed, then this
7016will return false.
7017
7018=cut
7019*/
7020
463ee0b2 7021int
864dbfa3 7022Perl_sv_isobject(pTHX_ SV *sv)
85e6fe83 7023{
68dc0745 7024 if (!sv)
7025 return 0;
5b295bef 7026 SvGETMAGIC(sv);
85e6fe83
LW
7027 if (!SvROK(sv))
7028 return 0;
7029 sv = (SV*)SvRV(sv);
7030 if (!SvOBJECT(sv))
7031 return 0;
7032 return 1;
7033}
7034
954c1994
GS
7035/*
7036=for apidoc sv_isa
7037
7038Returns a boolean indicating whether the SV is blessed into the specified
7039class. This does not check for subtypes; use C<sv_derived_from> to verify
7040an inheritance relationship.
7041
7042=cut
7043*/
7044
85e6fe83 7045int
864dbfa3 7046Perl_sv_isa(pTHX_ SV *sv, const char *name)
463ee0b2 7047{
bfcb3514 7048 const char *hvname;
68dc0745 7049 if (!sv)
7050 return 0;
5b295bef 7051 SvGETMAGIC(sv);
ed6116ce 7052 if (!SvROK(sv))
463ee0b2 7053 return 0;
ed6116ce
LW
7054 sv = (SV*)SvRV(sv);
7055 if (!SvOBJECT(sv))
463ee0b2 7056 return 0;
bfcb3514
NC
7057 hvname = HvNAME_get(SvSTASH(sv));
7058 if (!hvname)
e27ad1f2 7059 return 0;
463ee0b2 7060
bfcb3514 7061 return strEQ(hvname, name);
463ee0b2
LW
7062}
7063
954c1994
GS
7064/*
7065=for apidoc newSVrv
7066
7067Creates a new SV for the RV, C<rv>, to point to. If C<rv> is not an RV then
7068it will be upgraded to one. If C<classname> is non-null then the new SV will
7069be blessed in the specified package. The new SV is returned and its
7070reference count is 1.
7071
7072=cut
7073*/
7074
463ee0b2 7075SV*
864dbfa3 7076Perl_newSVrv(pTHX_ SV *rv, const char *classname)
463ee0b2 7077{
463ee0b2
LW
7078 SV *sv;
7079
4561caa4 7080 new_SV(sv);
51cf62d8 7081
765f542d 7082 SV_CHECK_THINKFIRST_COW_DROP(rv);
51cf62d8 7083 SvAMAGIC_off(rv);
51cf62d8 7084
0199fce9 7085 if (SvTYPE(rv) >= SVt_PVMG) {
a3b680e6 7086 const U32 refcnt = SvREFCNT(rv);
0199fce9
JD
7087 SvREFCNT(rv) = 0;
7088 sv_clear(rv);
7089 SvFLAGS(rv) = 0;
7090 SvREFCNT(rv) = refcnt;
7091 }
7092
51cf62d8 7093 if (SvTYPE(rv) < SVt_RV)
0199fce9
JD
7094 sv_upgrade(rv, SVt_RV);
7095 else if (SvTYPE(rv) > SVt_RV) {
8bd4d4c5 7096 SvPV_free(rv);
0199fce9
JD
7097 SvCUR_set(rv, 0);
7098 SvLEN_set(rv, 0);
7099 }
51cf62d8 7100
0c34ef67 7101 SvOK_off(rv);
b162af07 7102 SvRV_set(rv, sv);
ed6116ce 7103 SvROK_on(rv);
463ee0b2 7104
a0d0e21e 7105 if (classname) {
1b6737cc 7106 HV* const stash = gv_stashpv(classname, TRUE);
a0d0e21e
LW
7107 (void)sv_bless(rv, stash);
7108 }
7109 return sv;
7110}
7111
954c1994
GS
7112/*
7113=for apidoc sv_setref_pv
7114
7115Copies a pointer into a new SV, optionally blessing the SV. The C<rv>
7116argument will be upgraded to an RV. That RV will be modified to point to
7117the new SV. If the C<pv> argument is NULL then C<PL_sv_undef> will be placed
7118into the SV. The C<classname> argument indicates the package for the
7119blessing. Set C<classname> to C<Nullch> to avoid the blessing. The new SV
d34c2299 7120will have a reference count of 1, and the RV will be returned.
954c1994
GS
7121
7122Do not use with other Perl types such as HV, AV, SV, CV, because those
7123objects will become corrupted by the pointer copy process.
7124
7125Note that C<sv_setref_pvn> copies the string while this copies the pointer.
7126
7127=cut
7128*/
7129
a0d0e21e 7130SV*
864dbfa3 7131Perl_sv_setref_pv(pTHX_ SV *rv, const char *classname, void *pv)
a0d0e21e 7132{
189b2af5 7133 if (!pv) {
3280af22 7134 sv_setsv(rv, &PL_sv_undef);
189b2af5
GS
7135 SvSETMAGIC(rv);
7136 }
a0d0e21e 7137 else
56431972 7138 sv_setiv(newSVrv(rv,classname), PTR2IV(pv));
a0d0e21e
LW
7139 return rv;
7140}
7141
954c1994
GS
7142/*
7143=for apidoc sv_setref_iv
7144
7145Copies an integer into a new SV, optionally blessing the SV. The C<rv>
7146argument will be upgraded to an RV. That RV will be modified to point to
7147the new SV. The C<classname> argument indicates the package for the
7148blessing. Set C<classname> to C<Nullch> to avoid the blessing. The new SV
d34c2299 7149will have a reference count of 1, and the RV will be returned.
954c1994
GS
7150
7151=cut
7152*/
7153
a0d0e21e 7154SV*
864dbfa3 7155Perl_sv_setref_iv(pTHX_ SV *rv, const char *classname, IV iv)
a0d0e21e
LW
7156{
7157 sv_setiv(newSVrv(rv,classname), iv);
7158 return rv;
7159}
7160
954c1994 7161/*
e1c57cef
JH
7162=for apidoc sv_setref_uv
7163
7164Copies an unsigned integer into a new SV, optionally blessing the SV. The C<rv>
7165argument will be upgraded to an RV. That RV will be modified to point to
7166the new SV. The C<classname> argument indicates the package for the
7167blessing. Set C<classname> to C<Nullch> to avoid the blessing. The new SV
d34c2299 7168will have a reference count of 1, and the RV will be returned.
e1c57cef
JH
7169
7170=cut
7171*/
7172
7173SV*
7174Perl_sv_setref_uv(pTHX_ SV *rv, const char *classname, UV uv)
7175{
7176 sv_setuv(newSVrv(rv,classname), uv);
7177 return rv;
7178}
7179
7180/*
954c1994
GS
7181=for apidoc sv_setref_nv
7182
7183Copies a double into a new SV, optionally blessing the SV. The C<rv>
7184argument will be upgraded to an RV. That RV will be modified to point to
7185the new SV. The C<classname> argument indicates the package for the
7186blessing. Set C<classname> to C<Nullch> to avoid the blessing. The new SV
d34c2299 7187will have a reference count of 1, and the RV will be returned.
954c1994
GS
7188
7189=cut
7190*/
7191
a0d0e21e 7192SV*
65202027 7193Perl_sv_setref_nv(pTHX_ SV *rv, const char *classname, NV nv)
a0d0e21e
LW
7194{
7195 sv_setnv(newSVrv(rv,classname), nv);
7196 return rv;
7197}
463ee0b2 7198
954c1994
GS
7199/*
7200=for apidoc sv_setref_pvn
7201
7202Copies a string into a new SV, optionally blessing the SV. The length of the
7203string must be specified with C<n>. The C<rv> argument will be upgraded to
7204an RV. That RV will be modified to point to the new SV. The C<classname>
7205argument indicates the package for the blessing. Set C<classname> to
7a5fa8a2 7206C<Nullch> to avoid the blessing. The new SV will have a reference count
d34c2299 7207of 1, and the RV will be returned.
954c1994
GS
7208
7209Note that C<sv_setref_pv> copies the pointer while this copies the string.
7210
7211=cut
7212*/
7213
a0d0e21e 7214SV*
1b6737cc 7215Perl_sv_setref_pvn(pTHX_ SV *rv, const char *classname, const char *pv, STRLEN n)
a0d0e21e
LW
7216{
7217 sv_setpvn(newSVrv(rv,classname), pv, n);
463ee0b2
LW
7218 return rv;
7219}
7220
954c1994
GS
7221/*
7222=for apidoc sv_bless
7223
7224Blesses an SV into a specified package. The SV must be an RV. The package
7225must be designated by its stash (see C<gv_stashpv()>). The reference count
7226of the SV is unaffected.
7227
7228=cut
7229*/
7230
a0d0e21e 7231SV*
864dbfa3 7232Perl_sv_bless(pTHX_ SV *sv, HV *stash)
a0d0e21e 7233{
76e3520e 7234 SV *tmpRef;
a0d0e21e 7235 if (!SvROK(sv))
cea2e8a9 7236 Perl_croak(aTHX_ "Can't bless non-reference value");
76e3520e
GS
7237 tmpRef = SvRV(sv);
7238 if (SvFLAGS(tmpRef) & (SVs_OBJECT|SVf_READONLY)) {
7239 if (SvREADONLY(tmpRef))
cea2e8a9 7240 Perl_croak(aTHX_ PL_no_modify);
76e3520e
GS
7241 if (SvOBJECT(tmpRef)) {
7242 if (SvTYPE(tmpRef) != SVt_PVIO)
3280af22 7243 --PL_sv_objcount;
76e3520e 7244 SvREFCNT_dec(SvSTASH(tmpRef));
2e3febc6 7245 }
a0d0e21e 7246 }
76e3520e
GS
7247 SvOBJECT_on(tmpRef);
7248 if (SvTYPE(tmpRef) != SVt_PVIO)
3280af22 7249 ++PL_sv_objcount;
862a34c6 7250 SvUPGRADE(tmpRef, SVt_PVMG);
b162af07 7251 SvSTASH_set(tmpRef, (HV*)SvREFCNT_inc(stash));
a0d0e21e 7252
2e3febc6
CS
7253 if (Gv_AMG(stash))
7254 SvAMAGIC_on(sv);
7255 else
7256 SvAMAGIC_off(sv);
a0d0e21e 7257
1edbfb88
AB
7258 if(SvSMAGICAL(tmpRef))
7259 if(mg_find(tmpRef, PERL_MAGIC_ext) || mg_find(tmpRef, PERL_MAGIC_uvar))
7260 mg_set(tmpRef);
7261
7262
ecdeb87c 7263
a0d0e21e
LW
7264 return sv;
7265}
7266
645c22ef 7267/* Downgrades a PVGV to a PVMG.
645c22ef
DM
7268 */
7269
76e3520e 7270STATIC void
cea2e8a9 7271S_sv_unglob(pTHX_ SV *sv)
a0d0e21e 7272{
850fabdf
GS
7273 void *xpvmg;
7274
a0d0e21e
LW
7275 assert(SvTYPE(sv) == SVt_PVGV);
7276 SvFAKE_off(sv);
7277 if (GvGP(sv))
1edc1566 7278 gp_free((GV*)sv);
e826b3c7 7279 if (GvSTASH(sv)) {
e15faf7d 7280 sv_del_backref((SV*)GvSTASH(sv), sv);
5c284bb0 7281 GvSTASH(sv) = NULL;
e826b3c7 7282 }
14befaf4 7283 sv_unmagic(sv, PERL_MAGIC_glob);
a0d0e21e 7284 Safefree(GvNAME(sv));
a5f75d66 7285 GvMULTI_off(sv);
850fabdf
GS
7286
7287 /* need to keep SvANY(sv) in the right arena */
7288 xpvmg = new_XPVMG();
7289 StructCopy(SvANY(sv), xpvmg, XPVMG);
7290 del_XPVGV(SvANY(sv));
7291 SvANY(sv) = xpvmg;
7292
a0d0e21e
LW
7293 SvFLAGS(sv) &= ~SVTYPEMASK;
7294 SvFLAGS(sv) |= SVt_PVMG;
7295}
7296
954c1994 7297/*
840a7b70 7298=for apidoc sv_unref_flags
954c1994
GS
7299
7300Unsets the RV status of the SV, and decrements the reference count of
7301whatever was being referenced by the RV. This can almost be thought of
840a7b70
IZ
7302as a reversal of C<newSVrv>. The C<cflags> argument can contain
7303C<SV_IMMEDIATE_UNREF> to force the reference count to be decremented
7304(otherwise the decrementing is conditional on the reference count being
7305different from one or the reference being a readonly SV).
7889fe52 7306See C<SvROK_off>.
954c1994
GS
7307
7308=cut
7309*/
7310
ed6116ce 7311void
e15faf7d 7312Perl_sv_unref_flags(pTHX_ SV *ref, U32 flags)
ed6116ce 7313{
b64e5050 7314 SV* const target = SvRV(ref);
810b8aa5 7315
e15faf7d
NC
7316 if (SvWEAKREF(ref)) {
7317 sv_del_backref(target, ref);
7318 SvWEAKREF_off(ref);
7319 SvRV_set(ref, NULL);
810b8aa5
GS
7320 return;
7321 }
e15faf7d
NC
7322 SvRV_set(ref, NULL);
7323 SvROK_off(ref);
7324 /* You can't have a || SvREADONLY(target) here, as $a = $$a, where $a was
04ca4930 7325 assigned to as BEGIN {$a = \"Foo"} will fail. */
e15faf7d
NC
7326 if (SvREFCNT(target) != 1 || (flags & SV_IMMEDIATE_UNREF))
7327 SvREFCNT_dec(target);
840a7b70 7328 else /* XXX Hack, but hard to make $a=$a->[1] work otherwise */
e15faf7d 7329 sv_2mortal(target); /* Schedule for freeing later */
ed6116ce 7330}
8990e307 7331
840a7b70 7332/*
645c22ef
DM
7333=for apidoc sv_untaint
7334
7335Untaint an SV. Use C<SvTAINTED_off> instead.
7336=cut
7337*/
7338
bbce6d69 7339void
864dbfa3 7340Perl_sv_untaint(pTHX_ SV *sv)
bbce6d69 7341{
13f57bf8 7342 if (SvTYPE(sv) >= SVt_PVMG && SvMAGIC(sv)) {
b64e5050 7343 MAGIC * const mg = mg_find(sv, PERL_MAGIC_taint);
36477c24 7344 if (mg)
565764a8 7345 mg->mg_len &= ~1;
36477c24 7346 }
bbce6d69 7347}
7348
645c22ef
DM
7349/*
7350=for apidoc sv_tainted
7351
7352Test an SV for taintedness. Use C<SvTAINTED> instead.
7353=cut
7354*/
7355
bbce6d69 7356bool
864dbfa3 7357Perl_sv_tainted(pTHX_ SV *sv)
bbce6d69 7358{
13f57bf8 7359 if (SvTYPE(sv) >= SVt_PVMG && SvMAGIC(sv)) {
823a54a3 7360 const MAGIC * const mg = mg_find(sv, PERL_MAGIC_taint);
2ddb8a4f 7361 if (mg && (mg->mg_len & 1) )
36477c24 7362 return TRUE;
7363 }
7364 return FALSE;
bbce6d69 7365}
7366
09540bc3
JH
7367/*
7368=for apidoc sv_setpviv
7369
7370Copies an integer into the given SV, also updating its string value.
7371Does not handle 'set' magic. See C<sv_setpviv_mg>.
7372
7373=cut
7374*/
7375
7376void
7377Perl_sv_setpviv(pTHX_ SV *sv, IV iv)
7378{
7379 char buf[TYPE_CHARS(UV)];
7380 char *ebuf;
b64e5050 7381 char * const ptr = uiv_2buf(buf, iv, 0, 0, &ebuf);
09540bc3
JH
7382
7383 sv_setpvn(sv, ptr, ebuf - ptr);
7384}
7385
7386/*
7387=for apidoc sv_setpviv_mg
7388
7389Like C<sv_setpviv>, but also handles 'set' magic.
7390
7391=cut
7392*/
7393
7394void
7395Perl_sv_setpviv_mg(pTHX_ SV *sv, IV iv)
7396{
df7eb254 7397 sv_setpviv(sv, iv);
09540bc3
JH
7398 SvSETMAGIC(sv);
7399}
7400
cea2e8a9 7401#if defined(PERL_IMPLICIT_CONTEXT)
645c22ef
DM
7402
7403/* pTHX_ magic can't cope with varargs, so this is a no-context
7404 * version of the main function, (which may itself be aliased to us).
7405 * Don't access this version directly.
7406 */
7407
cea2e8a9
GS
7408void
7409Perl_sv_setpvf_nocontext(SV *sv, const char* pat, ...)
7410{
7411 dTHX;
7412 va_list args;
7413 va_start(args, pat);
c5be433b 7414 sv_vsetpvf(sv, pat, &args);
cea2e8a9
GS
7415 va_end(args);
7416}
7417
645c22ef
DM
7418/* pTHX_ magic can't cope with varargs, so this is a no-context
7419 * version of the main function, (which may itself be aliased to us).
7420 * Don't access this version directly.
7421 */
cea2e8a9
GS
7422
7423void
7424Perl_sv_setpvf_mg_nocontext(SV *sv, const char* pat, ...)
7425{
7426 dTHX;
7427 va_list args;
7428 va_start(args, pat);
c5be433b 7429 sv_vsetpvf_mg(sv, pat, &args);
cea2e8a9 7430 va_end(args);
cea2e8a9
GS
7431}
7432#endif
7433
954c1994
GS
7434/*
7435=for apidoc sv_setpvf
7436
bffc3d17
SH
7437Works like C<sv_catpvf> but copies the text into the SV instead of
7438appending it. Does not handle 'set' magic. See C<sv_setpvf_mg>.
954c1994
GS
7439
7440=cut
7441*/
7442
46fc3d4c 7443void
864dbfa3 7444Perl_sv_setpvf(pTHX_ SV *sv, const char* pat, ...)
46fc3d4c 7445{
7446 va_list args;
46fc3d4c 7447 va_start(args, pat);
c5be433b 7448 sv_vsetpvf(sv, pat, &args);
46fc3d4c 7449 va_end(args);
7450}
7451
bffc3d17
SH
7452/*
7453=for apidoc sv_vsetpvf
7454
7455Works like C<sv_vcatpvf> but copies the text into the SV instead of
7456appending it. Does not handle 'set' magic. See C<sv_vsetpvf_mg>.
7457
7458Usually used via its frontend C<sv_setpvf>.
7459
7460=cut
7461*/
645c22ef 7462
c5be433b
GS
7463void
7464Perl_sv_vsetpvf(pTHX_ SV *sv, const char* pat, va_list* args)
7465{
7466 sv_vsetpvfn(sv, pat, strlen(pat), args, Null(SV**), 0, Null(bool*));
7467}
ef50df4b 7468
954c1994
GS
7469/*
7470=for apidoc sv_setpvf_mg
7471
7472Like C<sv_setpvf>, but also handles 'set' magic.
7473
7474=cut
7475*/
7476
ef50df4b 7477void
864dbfa3 7478Perl_sv_setpvf_mg(pTHX_ SV *sv, const char* pat, ...)
ef50df4b
GS
7479{
7480 va_list args;
ef50df4b 7481 va_start(args, pat);
c5be433b 7482 sv_vsetpvf_mg(sv, pat, &args);
ef50df4b 7483 va_end(args);
c5be433b
GS
7484}
7485
bffc3d17
SH
7486/*
7487=for apidoc sv_vsetpvf_mg
7488
7489Like C<sv_vsetpvf>, but also handles 'set' magic.
7490
7491Usually used via its frontend C<sv_setpvf_mg>.
7492
7493=cut
7494*/
645c22ef 7495
c5be433b
GS
7496void
7497Perl_sv_vsetpvf_mg(pTHX_ SV *sv, const char* pat, va_list* args)
7498{
7499 sv_vsetpvfn(sv, pat, strlen(pat), args, Null(SV**), 0, Null(bool*));
ef50df4b
GS
7500 SvSETMAGIC(sv);
7501}
7502
cea2e8a9 7503#if defined(PERL_IMPLICIT_CONTEXT)
645c22ef
DM
7504
7505/* pTHX_ magic can't cope with varargs, so this is a no-context
7506 * version of the main function, (which may itself be aliased to us).
7507 * Don't access this version directly.
7508 */
7509
cea2e8a9
GS
7510void
7511Perl_sv_catpvf_nocontext(SV *sv, const char* pat, ...)
7512{
7513 dTHX;
7514 va_list args;
7515 va_start(args, pat);
c5be433b 7516 sv_vcatpvf(sv, pat, &args);
cea2e8a9
GS
7517 va_end(args);
7518}
7519
645c22ef
DM
7520/* pTHX_ magic can't cope with varargs, so this is a no-context
7521 * version of the main function, (which may itself be aliased to us).
7522 * Don't access this version directly.
7523 */
7524
cea2e8a9
GS
7525void
7526Perl_sv_catpvf_mg_nocontext(SV *sv, const char* pat, ...)
7527{
7528 dTHX;
7529 va_list args;
7530 va_start(args, pat);
c5be433b 7531 sv_vcatpvf_mg(sv, pat, &args);
cea2e8a9 7532 va_end(args);
cea2e8a9
GS
7533}
7534#endif
7535
954c1994
GS
7536/*
7537=for apidoc sv_catpvf
7538
d5ce4a7c
GA
7539Processes its arguments like C<sprintf> and appends the formatted
7540output to an SV. If the appended data contains "wide" characters
7541(including, but not limited to, SVs with a UTF-8 PV formatted with %s,
7542and characters >255 formatted with %c), the original SV might get
bffc3d17 7543upgraded to UTF-8. Handles 'get' magic, but not 'set' magic. See
cdd94ca7
NC
7544C<sv_catpvf_mg>. If the original SV was UTF-8, the pattern should be
7545valid UTF-8; if the original SV was bytes, the pattern should be too.
954c1994 7546
d5ce4a7c 7547=cut */
954c1994 7548
46fc3d4c 7549void
864dbfa3 7550Perl_sv_catpvf(pTHX_ SV *sv, const char* pat, ...)
46fc3d4c 7551{
7552 va_list args;
46fc3d4c 7553 va_start(args, pat);
c5be433b 7554 sv_vcatpvf(sv, pat, &args);
46fc3d4c 7555 va_end(args);
7556}
7557
bffc3d17
SH
7558/*
7559=for apidoc sv_vcatpvf
7560
7561Processes its arguments like C<vsprintf> and appends the formatted output
7562to an SV. Does not handle 'set' magic. See C<sv_vcatpvf_mg>.
7563
7564Usually used via its frontend C<sv_catpvf>.
7565
7566=cut
7567*/
645c22ef 7568
ef50df4b 7569void
c5be433b
GS
7570Perl_sv_vcatpvf(pTHX_ SV *sv, const char* pat, va_list* args)
7571{
7572 sv_vcatpvfn(sv, pat, strlen(pat), args, Null(SV**), 0, Null(bool*));
7573}
7574
954c1994
GS
7575/*
7576=for apidoc sv_catpvf_mg
7577
7578Like C<sv_catpvf>, but also handles 'set' magic.
7579
7580=cut
7581*/
7582
c5be433b 7583void
864dbfa3 7584Perl_sv_catpvf_mg(pTHX_ SV *sv, const char* pat, ...)
ef50df4b
GS
7585{
7586 va_list args;
ef50df4b 7587 va_start(args, pat);
c5be433b 7588 sv_vcatpvf_mg(sv, pat, &args);
ef50df4b 7589 va_end(args);
c5be433b
GS
7590}
7591
bffc3d17
SH
7592/*
7593=for apidoc sv_vcatpvf_mg
7594
7595Like C<sv_vcatpvf>, but also handles 'set' magic.
7596
7597Usually used via its frontend C<sv_catpvf_mg>.
7598
7599=cut
7600*/
645c22ef 7601
c5be433b
GS
7602void
7603Perl_sv_vcatpvf_mg(pTHX_ SV *sv, const char* pat, va_list* args)
7604{
7605 sv_vcatpvfn(sv, pat, strlen(pat), args, Null(SV**), 0, Null(bool*));
ef50df4b
GS
7606 SvSETMAGIC(sv);
7607}
7608
954c1994
GS
7609/*
7610=for apidoc sv_vsetpvfn
7611
bffc3d17 7612Works like C<sv_vcatpvfn> but copies the text into the SV instead of
954c1994
GS
7613appending it.
7614
bffc3d17 7615Usually used via one of its frontends C<sv_vsetpvf> and C<sv_vsetpvf_mg>.
645c22ef 7616
954c1994
GS
7617=cut
7618*/
7619
46fc3d4c 7620void
7d5ea4e7 7621Perl_sv_vsetpvfn(pTHX_ SV *sv, const char *pat, STRLEN patlen, va_list *args, SV **svargs, I32 svmax, bool *maybe_tainted)
46fc3d4c 7622{
7623 sv_setpvn(sv, "", 0);
7d5ea4e7 7624 sv_vcatpvfn(sv, pat, patlen, args, svargs, svmax, maybe_tainted);
46fc3d4c 7625}
7626
2d00ba3b 7627STATIC I32
9dd79c3f 7628S_expect_number(pTHX_ char** pattern)
211dfcf1
HS
7629{
7630 I32 var = 0;
7631 switch (**pattern) {
7632 case '1': case '2': case '3':
7633 case '4': case '5': case '6':
7634 case '7': case '8': case '9':
2fba7546
GA
7635 var = *(*pattern)++ - '0';
7636 while (isDIGIT(**pattern)) {
7637 I32 tmp = var * 10 + (*(*pattern)++ - '0');
7638 if (tmp < var)
7639 Perl_croak(aTHX_ "Integer overflow in format string for %s", (PL_op ? OP_NAME(PL_op) : "sv_vcatpvfn"));
7640 var = tmp;
7641 }
211dfcf1
HS
7642 }
7643 return var;
7644}
211dfcf1 7645
c445ea15
AL
7646STATIC char *
7647S_F0convert(NV nv, char *endbuf, STRLEN *len)
4151a5fe 7648{
a3b680e6 7649 const int neg = nv < 0;
4151a5fe 7650 UV uv;
4151a5fe
IZ
7651
7652 if (neg)
7653 nv = -nv;
7654 if (nv < UV_MAX) {
b464bac0 7655 char *p = endbuf;
4151a5fe 7656 nv += 0.5;
028f8eaa 7657 uv = (UV)nv;
4151a5fe
IZ
7658 if (uv & 1 && uv == nv)
7659 uv--; /* Round to even */
7660 do {
a3b680e6 7661 const unsigned dig = uv % 10;
4151a5fe
IZ
7662 *--p = '0' + dig;
7663 } while (uv /= 10);
7664 if (neg)
7665 *--p = '-';
7666 *len = endbuf - p;
7667 return p;
7668 }
7669 return Nullch;
7670}
7671
7672
954c1994
GS
7673/*
7674=for apidoc sv_vcatpvfn
7675
7676Processes its arguments like C<vsprintf> and appends the formatted output
7677to an SV. Uses an array of SVs if the C style variable argument list is
7678missing (NULL). When running with taint checks enabled, indicates via
7679C<maybe_tainted> if results are untrustworthy (often due to the use of
7680locales).
7681
bffc3d17 7682Usually used via one of its frontends C<sv_vcatpvf> and C<sv_vcatpvf_mg>.
645c22ef 7683
954c1994
GS
7684=cut
7685*/
7686
8896765a
RB
7687
7688#define VECTORIZE_ARGS vecsv = va_arg(*args, SV*);\
7689 vecstr = (U8*)SvPV_const(vecsv,veclen);\
7690 vec_utf8 = DO_UTF8(vecsv);
7691
1ef29b0e
RGS
7692/* XXX maybe_tainted is never assigned to, so the doc above is lying. */
7693
46fc3d4c 7694void
7d5ea4e7 7695Perl_sv_vcatpvfn(pTHX_ SV *sv, const char *pat, STRLEN patlen, va_list *args, SV **svargs, I32 svmax, bool *maybe_tainted)
46fc3d4c 7696{
7697 char *p;
7698 char *q;
a3b680e6 7699 const char *patend;
fc36a67e 7700 STRLEN origlen;
46fc3d4c 7701 I32 svix = 0;
27da23d5 7702 static const char nullstr[] = "(null)";
9c5ffd7c 7703 SV *argsv = Nullsv;
b464bac0
AL
7704 bool has_utf8 = DO_UTF8(sv); /* has the result utf8? */
7705 const bool pat_utf8 = has_utf8; /* the pattern is in utf8? */
db79b45b 7706 SV *nsv = Nullsv;
4151a5fe
IZ
7707 /* Times 4: a decimal digit takes more than 3 binary digits.
7708 * NV_DIG: mantissa takes than many decimal digits.
7709 * Plus 32: Playing safe. */
7710 char ebuf[IV_DIG * 4 + NV_DIG + 32];
7711 /* large enough for "%#.#f" --chip */
7712 /* what about long double NVs? --jhi */
db79b45b 7713
53c1dcc0
AL
7714 PERL_UNUSED_ARG(maybe_tainted);
7715
46fc3d4c 7716 /* no matter what, this is a string now */
fc36a67e 7717 (void)SvPV_force(sv, origlen);
46fc3d4c 7718
8896765a 7719 /* special-case "", "%s", and "%-p" (SVf - see below) */
46fc3d4c 7720 if (patlen == 0)
7721 return;
0dbb1585 7722 if (patlen == 2 && pat[0] == '%' && pat[1] == 's') {
2d03de9c
AL
7723 if (args) {
7724 const char * const s = va_arg(*args, char*);
7725 sv_catpv(sv, s ? s : nullstr);
7726 }
7727 else if (svix < svmax) {
7728 sv_catsv(sv, *svargs);
2d03de9c
AL
7729 }
7730 return;
0dbb1585 7731 }
8896765a
RB
7732 if (args && patlen == 3 && pat[0] == '%' &&
7733 pat[1] == '-' && pat[2] == 'p') {
7734 argsv = va_arg(*args, SV*);
7735 sv_catsv(sv, argsv);
8896765a 7736 return;
46fc3d4c 7737 }
7738
1d917b39 7739#ifndef USE_LONG_DOUBLE
4151a5fe 7740 /* special-case "%.<number>[gf]" */
7af36d83 7741 if ( !args && patlen <= 5 && pat[0] == '%' && pat[1] == '.'
4151a5fe
IZ
7742 && (pat[patlen-1] == 'g' || pat[patlen-1] == 'f') ) {
7743 unsigned digits = 0;
7744 const char *pp;
7745
7746 pp = pat + 2;
7747 while (*pp >= '0' && *pp <= '9')
7748 digits = 10 * digits + (*pp++ - '0');
028f8eaa 7749 if (pp - pat == (int)patlen - 1) {
4151a5fe
IZ
7750 NV nv;
7751
7af36d83 7752 if (svix < svmax)
4151a5fe
IZ
7753 nv = SvNV(*svargs);
7754 else
7755 return;
7756 if (*pp == 'g') {
2873255c
NC
7757 /* Add check for digits != 0 because it seems that some
7758 gconverts are buggy in this case, and we don't yet have
7759 a Configure test for this. */
7760 if (digits && digits < sizeof(ebuf) - NV_DIG - 10) {
7761 /* 0, point, slack */
2e59c212 7762 Gconvert(nv, (int)digits, 0, ebuf);
4151a5fe
IZ
7763 sv_catpv(sv, ebuf);
7764 if (*ebuf) /* May return an empty string for digits==0 */
7765 return;
7766 }
7767 } else if (!digits) {
7768 STRLEN l;
7769
7770 if ((p = F0convert(nv, ebuf + sizeof ebuf, &l))) {
7771 sv_catpvn(sv, p, l);
7772 return;
7773 }
7774 }
7775 }
7776 }
1d917b39 7777#endif /* !USE_LONG_DOUBLE */
4151a5fe 7778
2cf2cfc6 7779 if (!args && svix < svmax && DO_UTF8(*svargs))
205f51d8 7780 has_utf8 = TRUE;
2cf2cfc6 7781
46fc3d4c 7782 patend = (char*)pat + patlen;
7783 for (p = (char*)pat; p < patend; p = q) {
7784 bool alt = FALSE;
7785 bool left = FALSE;
b22c7a20 7786 bool vectorize = FALSE;
211dfcf1 7787 bool vectorarg = FALSE;
2cf2cfc6 7788 bool vec_utf8 = FALSE;
46fc3d4c 7789 char fill = ' ';
7790 char plus = 0;
7791 char intsize = 0;
7792 STRLEN width = 0;
fc36a67e 7793 STRLEN zeros = 0;
46fc3d4c 7794 bool has_precis = FALSE;
7795 STRLEN precis = 0;
c445ea15 7796 const I32 osvix = svix;
2cf2cfc6 7797 bool is_utf8 = FALSE; /* is this item utf8? */
20f6aaab
AS
7798#ifdef HAS_LDBL_SPRINTF_BUG
7799 /* This is to try to fix a bug with irix/nonstop-ux/powerux and
205f51d8 7800 with sfio - Allen <allens@cpan.org> */
20f6aaab
AS
7801 bool fix_ldbl_sprintf_bug = FALSE;
7802#endif
205f51d8 7803
46fc3d4c 7804 char esignbuf[4];
89ebb4a3 7805 U8 utf8buf[UTF8_MAXBYTES+1];
46fc3d4c 7806 STRLEN esignlen = 0;
7807
4d84ee25 7808 const char *eptr = Nullch;
fc36a67e 7809 STRLEN elen = 0;
81f715da 7810 SV *vecsv = Nullsv;
245d4a47 7811 const U8 *vecstr = Null(U8*);
b22c7a20 7812 STRLEN veclen = 0;
934abaf1 7813 char c = 0;
46fc3d4c 7814 int i;
9c5ffd7c 7815 unsigned base = 0;
8c8eb53c
RB
7816 IV iv = 0;
7817 UV uv = 0;
9e5b023a
JH
7818 /* we need a long double target in case HAS_LONG_DOUBLE but
7819 not USE_LONG_DOUBLE
7820 */
35fff930 7821#if defined(HAS_LONG_DOUBLE) && LONG_DOUBLESIZE > DOUBLESIZE
9e5b023a
JH
7822 long double nv;
7823#else
65202027 7824 NV nv;
9e5b023a 7825#endif
46fc3d4c 7826 STRLEN have;
7827 STRLEN need;
7828 STRLEN gap;
7af36d83 7829 const char *dotstr = ".";
b22c7a20 7830 STRLEN dotstrlen = 1;
211dfcf1 7831 I32 efix = 0; /* explicit format parameter index */
eb3fce90 7832 I32 ewix = 0; /* explicit width index */
211dfcf1
HS
7833 I32 epix = 0; /* explicit precision index */
7834 I32 evix = 0; /* explicit vector index */
eb3fce90 7835 bool asterisk = FALSE;
46fc3d4c 7836
211dfcf1 7837 /* echo everything up to the next format specification */
46fc3d4c 7838 for (q = p; q < patend && *q != '%'; ++q) ;
7839 if (q > p) {
db79b45b
JH
7840 if (has_utf8 && !pat_utf8)
7841 sv_catpvn_utf8_upgrade(sv, p, q - p, nsv);
7842 else
7843 sv_catpvn(sv, p, q - p);
46fc3d4c 7844 p = q;
7845 }
7846 if (q++ >= patend)
7847 break;
7848
211dfcf1
HS
7849/*
7850 We allow format specification elements in this order:
7851 \d+\$ explicit format parameter index
7852 [-+ 0#]+ flags
a472f209 7853 v|\*(\d+\$)?v vector with optional (optionally specified) arg
f3583277 7854 0 flag (as above): repeated to allow "v02"
211dfcf1
HS
7855 \d+|\*(\d+\$)? width using optional (optionally specified) arg
7856 \.(\d*|\*(\d+\$)?) precision using optional (optionally specified) arg
7857 [hlqLV] size
8896765a
RB
7858 [%bcdefginopsuxDFOUX] format (mandatory)
7859*/
7860
7861 if (args) {
7862/*
7863 As of perl5.9.3, printf format checking is on by default.
7864 Internally, perl uses %p formats to provide an escape to
7865 some extended formatting. This block deals with those
7866 extensions: if it does not match, (char*)q is reset and
7867 the normal format processing code is used.
7868
7869 Currently defined extensions are:
7870 %p include pointer address (standard)
7871 %-p (SVf) include an SV (previously %_)
7872 %-<num>p include an SV with precision <num>
7873 %1p (VDf) include a v-string (as %vd)
7874 %<num>p reserved for future extensions
7875
7876 Robin Barker 2005-07-14
211dfcf1 7877*/
8896765a
RB
7878 char* r = q;
7879 bool sv = FALSE;
7880 STRLEN n = 0;
7881 if (*q == '-')
7882 sv = *q++;
c445ea15 7883 n = expect_number(&q);
8896765a
RB
7884 if (*q++ == 'p') {
7885 if (sv) { /* SVf */
7886 if (n) {
7887 precis = n;
7888 has_precis = TRUE;
7889 }
7890 argsv = va_arg(*args, SV*);
7891 eptr = SvPVx_const(argsv, elen);
7892 if (DO_UTF8(argsv))
7893 is_utf8 = TRUE;
7894 goto string;
7895 }
7896#if vdNUMBER
7897 else if (n == vdNUMBER) { /* VDf */
7898 vectorize = TRUE;
7899 VECTORIZE_ARGS
7900 goto format_vd;
7901 }
7902#endif
7903 else if (n) {
7904 if (ckWARN_d(WARN_INTERNAL))
7905 Perl_warner(aTHX_ packWARN(WARN_INTERNAL),
7906 "internal %%<num>p might conflict with future printf extensions");
7907 }
7908 }
7909 q = r;
7910 }
7911
c445ea15 7912 if ( (width = expect_number(&q)) ) {
211dfcf1
HS
7913 if (*q == '$') {
7914 ++q;
7915 efix = width;
7916 } else {
7917 goto gotwidth;
7918 }
7919 }
7920
fc36a67e 7921 /* FLAGS */
7922
46fc3d4c 7923 while (*q) {
7924 switch (*q) {
7925 case ' ':
7926 case '+':
7927 plus = *q++;
7928 continue;
7929
7930 case '-':
7931 left = TRUE;
7932 q++;
7933 continue;
7934
7935 case '0':
7936 fill = *q++;
7937 continue;
7938
7939 case '#':
7940 alt = TRUE;
7941 q++;
7942 continue;
7943
fc36a67e 7944 default:
7945 break;
7946 }
7947 break;
7948 }
46fc3d4c 7949
211dfcf1 7950 tryasterisk:
eb3fce90 7951 if (*q == '*') {
211dfcf1 7952 q++;
c445ea15 7953 if ( (ewix = expect_number(&q)) )
211dfcf1
HS
7954 if (*q++ != '$')
7955 goto unknown;
eb3fce90 7956 asterisk = TRUE;
211dfcf1
HS
7957 }
7958 if (*q == 'v') {
eb3fce90 7959 q++;
211dfcf1
HS
7960 if (vectorize)
7961 goto unknown;
9cbac4c7 7962 if ((vectorarg = asterisk)) {
211dfcf1
HS
7963 evix = ewix;
7964 ewix = 0;
7965 asterisk = FALSE;
7966 }
7967 vectorize = TRUE;
7968 goto tryasterisk;
eb3fce90
JH
7969 }
7970
211dfcf1 7971 if (!asterisk)
858a90f9 7972 {
7a5fa8a2 7973 if( *q == '0' )
f3583277 7974 fill = *q++;
c445ea15 7975 width = expect_number(&q);
858a90f9 7976 }
211dfcf1
HS
7977
7978 if (vectorize) {
7979 if (vectorarg) {
7980 if (args)
7981 vecsv = va_arg(*args, SV*);
7ad96abb
NC
7982 else if (evix) {
7983 vecsv = (evix > 0 && evix <= svmax)
7984 ? svargs[evix-1] : &PL_sv_undef;
7985 } else {
7986 vecsv = svix < svmax ? svargs[svix++] : &PL_sv_undef;
7987 }
245d4a47 7988 dotstr = SvPV_const(vecsv, dotstrlen);
640283f5
NC
7989 /* Keep the DO_UTF8 test *after* the SvPV call, else things go
7990 bad with tied or overloaded values that return UTF8. */
211dfcf1 7991 if (DO_UTF8(vecsv))
2cf2cfc6 7992 is_utf8 = TRUE;
640283f5
NC
7993 else if (has_utf8) {
7994 vecsv = sv_mortalcopy(vecsv);
7995 sv_utf8_upgrade(vecsv);
7996 dotstr = SvPV_const(vecsv, dotstrlen);
7997 is_utf8 = TRUE;
7998 }
211dfcf1
HS
7999 }
8000 if (args) {
8896765a 8001 VECTORIZE_ARGS
eb3fce90 8002 }
7ad96abb 8003 else if (efix ? (efix > 0 && efix <= svmax) : svix < svmax) {
211dfcf1 8004 vecsv = svargs[efix ? efix-1 : svix++];
245d4a47 8005 vecstr = (U8*)SvPV_const(vecsv,veclen);
2cf2cfc6 8006 vec_utf8 = DO_UTF8(vecsv);
96b8f7ce
JP
8007
8008 /* if this is a version object, we need to convert
8009 * back into v-string notation and then let the
8010 * vectorize happen normally
d7aa5382 8011 */
96b8f7ce
JP
8012 if (sv_derived_from(vecsv, "version")) {
8013 char *version = savesvpv(vecsv);
8014 vecsv = sv_newmortal();
8015 /* scan_vstring is expected to be called during
8016 * tokenization, so we need to fake up the end
8017 * of the buffer for it
8018 */
8019 PL_bufend = version + veclen;
8020 scan_vstring(version, vecsv);
8021 vecstr = (U8*)SvPV_const(vecsv, veclen);
8022 vec_utf8 = DO_UTF8(vecsv);
8023 Safefree(version);
d7aa5382 8024 }
211dfcf1
HS
8025 }
8026 else {
8027 vecstr = (U8*)"";
8028 veclen = 0;
8029 }
eb3fce90 8030 }
fc36a67e 8031
eb3fce90 8032 if (asterisk) {
fc36a67e 8033 if (args)
8034 i = va_arg(*args, int);
8035 else
eb3fce90
JH
8036 i = (ewix ? ewix <= svmax : svix < svmax) ?
8037 SvIVx(svargs[ewix ? ewix-1 : svix++]) : 0;
fc36a67e 8038 left |= (i < 0);
8039 width = (i < 0) ? -i : i;
fc36a67e 8040 }
211dfcf1 8041 gotwidth:
fc36a67e 8042
8043 /* PRECISION */
46fc3d4c 8044
fc36a67e 8045 if (*q == '.') {
8046 q++;
8047 if (*q == '*') {
211dfcf1 8048 q++;
c445ea15 8049 if ( ((epix = expect_number(&q))) && (*q++ != '$') )
7b8dd722
HS
8050 goto unknown;
8051 /* XXX: todo, support specified precision parameter */
8052 if (epix)
211dfcf1 8053 goto unknown;
46fc3d4c 8054 if (args)
8055 i = va_arg(*args, int);
8056 else
eb3fce90
JH
8057 i = (ewix ? ewix <= svmax : svix < svmax)
8058 ? SvIVx(svargs[ewix ? ewix-1 : svix++]) : 0;
fc36a67e 8059 precis = (i < 0) ? 0 : i;
fc36a67e 8060 }
8061 else {
8062 precis = 0;
8063 while (isDIGIT(*q))
8064 precis = precis * 10 + (*q++ - '0');
8065 }
8066 has_precis = TRUE;
8067 }
46fc3d4c 8068
fc36a67e 8069 /* SIZE */
46fc3d4c 8070
fc36a67e 8071 switch (*q) {
c623ac67
GS
8072#ifdef WIN32
8073 case 'I': /* Ix, I32x, and I64x */
8074# ifdef WIN64
8075 if (q[1] == '6' && q[2] == '4') {
8076 q += 3;
8077 intsize = 'q';
8078 break;
8079 }
8080# endif
8081 if (q[1] == '3' && q[2] == '2') {
8082 q += 3;
8083 break;
8084 }
8085# ifdef WIN64
8086 intsize = 'q';
8087# endif
8088 q++;
8089 break;
8090#endif
9e5b023a 8091#if defined(HAS_QUAD) || defined(HAS_LONG_DOUBLE)
6f9bb7fd 8092 case 'L': /* Ld */
e5c81feb 8093 /* FALL THROUGH */
e5c81feb 8094#ifdef HAS_QUAD
6f9bb7fd 8095 case 'q': /* qd */
9e5b023a 8096#endif
6f9bb7fd
GS
8097 intsize = 'q';
8098 q++;
8099 break;
8100#endif
fc36a67e 8101 case 'l':
9e5b023a 8102#if defined(HAS_QUAD) || defined(HAS_LONG_DOUBLE)
205f51d8 8103 if (*(q + 1) == 'l') { /* lld, llf */
fc36a67e 8104 intsize = 'q';
8105 q += 2;
46fc3d4c 8106 break;
cf2093f6 8107 }
fc36a67e 8108#endif
6f9bb7fd 8109 /* FALL THROUGH */
fc36a67e 8110 case 'h':
cf2093f6 8111 /* FALL THROUGH */
fc36a67e 8112 case 'V':
8113 intsize = *q++;
46fc3d4c 8114 break;
8115 }
8116
fc36a67e 8117 /* CONVERSION */
8118
211dfcf1
HS
8119 if (*q == '%') {
8120 eptr = q++;
8121 elen = 1;
26372e71
GA
8122 if (vectorize) {
8123 c = '%';
8124 goto unknown;
8125 }
211dfcf1
HS
8126 goto string;
8127 }
8128
26372e71 8129 if (!vectorize && !args) {
86c51f8b
NC
8130 if (efix) {
8131 const I32 i = efix-1;
8132 argsv = (i >= 0 && i < svmax) ? svargs[i] : &PL_sv_undef;
8133 } else {
8134 argsv = (svix >= 0 && svix < svmax)
8135 ? svargs[svix++] : &PL_sv_undef;
8136 }
863811b2 8137 }
211dfcf1 8138
46fc3d4c 8139 switch (c = *q++) {
8140
8141 /* STRINGS */
8142
46fc3d4c 8143 case 'c':
26372e71
GA
8144 if (vectorize)
8145 goto unknown;
8146 uv = (args) ? va_arg(*args, int) : SvIVx(argsv);
1bd104fb
JH
8147 if ((uv > 255 ||
8148 (!UNI_IS_INVARIANT(uv) && SvUTF8(sv)))
0064a8a9 8149 && !IN_BYTES) {
dfe13c55 8150 eptr = (char*)utf8buf;
9041c2e3 8151 elen = uvchr_to_utf8((U8*)eptr, uv) - utf8buf;
2cf2cfc6 8152 is_utf8 = TRUE;
7e2040f0
GS
8153 }
8154 else {
8155 c = (char)uv;
8156 eptr = &c;
8157 elen = 1;
a0ed51b3 8158 }
46fc3d4c 8159 goto string;
8160
46fc3d4c 8161 case 's':
26372e71
GA
8162 if (vectorize)
8163 goto unknown;
8164 if (args) {
fc36a67e 8165 eptr = va_arg(*args, char*);
c635e13b 8166 if (eptr)
1d7c1841
GS
8167#ifdef MACOS_TRADITIONAL
8168 /* On MacOS, %#s format is used for Pascal strings */
8169 if (alt)
8170 elen = *eptr++;
8171 else
8172#endif
c635e13b 8173 elen = strlen(eptr);
8174 else {
27da23d5 8175 eptr = (char *)nullstr;
c635e13b 8176 elen = sizeof nullstr - 1;
8177 }
46fc3d4c 8178 }
211dfcf1 8179 else {
4d84ee25 8180 eptr = SvPVx_const(argsv, elen);
7e2040f0 8181 if (DO_UTF8(argsv)) {
a0ed51b3
LW
8182 if (has_precis && precis < elen) {
8183 I32 p = precis;
7e2040f0 8184 sv_pos_u2b(argsv, &p, 0); /* sticks at end */
a0ed51b3
LW
8185 precis = p;
8186 }
8187 if (width) { /* fudge width (can't fudge elen) */
7e2040f0 8188 width += elen - sv_len_utf8(argsv);
a0ed51b3 8189 }
2cf2cfc6 8190 is_utf8 = TRUE;
a0ed51b3
LW
8191 }
8192 }
fc36a67e 8193
46fc3d4c 8194 string:
8195 if (has_precis && elen > precis)
8196 elen = precis;
8197 break;
8198
8199 /* INTEGERS */
8200
fc36a67e 8201 case 'p':
be75b157 8202 if (alt || vectorize)
c2e66d9e 8203 goto unknown;
211dfcf1 8204 uv = PTR2UV(args ? va_arg(*args, void*) : argsv);
fc36a67e 8205 base = 16;
8206 goto integer;
8207
46fc3d4c 8208 case 'D':
29fe7a80 8209#ifdef IV_IS_QUAD
22f3ae8c 8210 intsize = 'q';
29fe7a80 8211#else
46fc3d4c 8212 intsize = 'l';
29fe7a80 8213#endif
46fc3d4c 8214 /* FALL THROUGH */
8215 case 'd':
8216 case 'i':
8896765a
RB
8217#if vdNUMBER
8218 format_vd:
8219#endif
b22c7a20 8220 if (vectorize) {
ba210ebe 8221 STRLEN ulen;
211dfcf1
HS
8222 if (!veclen)
8223 continue;
2cf2cfc6
A
8224 if (vec_utf8)
8225 uv = utf8n_to_uvchr(vecstr, veclen, &ulen,
8226 UTF8_ALLOW_ANYUV);
b22c7a20 8227 else {
e83d50c9 8228 uv = *vecstr;
b22c7a20
GS
8229 ulen = 1;
8230 }
8231 vecstr += ulen;
8232 veclen -= ulen;
e83d50c9
JP
8233 if (plus)
8234 esignbuf[esignlen++] = plus;
b22c7a20
GS
8235 }
8236 else if (args) {
46fc3d4c 8237 switch (intsize) {
8238 case 'h': iv = (short)va_arg(*args, int); break;
46fc3d4c 8239 case 'l': iv = va_arg(*args, long); break;
fc36a67e 8240 case 'V': iv = va_arg(*args, IV); break;
b10c0dba 8241 default: iv = va_arg(*args, int); break;
cf2093f6
JH
8242#ifdef HAS_QUAD
8243 case 'q': iv = va_arg(*args, Quad_t); break;
8244#endif
46fc3d4c 8245 }
8246 }
8247 else {
b10c0dba 8248 IV tiv = SvIVx(argsv); /* work around GCC bug #13488 */
46fc3d4c 8249 switch (intsize) {
b10c0dba
MHM
8250 case 'h': iv = (short)tiv; break;
8251 case 'l': iv = (long)tiv; break;
8252 case 'V':
8253 default: iv = tiv; break;
cf2093f6 8254#ifdef HAS_QUAD
b10c0dba 8255 case 'q': iv = (Quad_t)tiv; break;
cf2093f6 8256#endif
46fc3d4c 8257 }
8258 }
e83d50c9
JP
8259 if ( !vectorize ) /* we already set uv above */
8260 {
8261 if (iv >= 0) {
8262 uv = iv;
8263 if (plus)
8264 esignbuf[esignlen++] = plus;
8265 }
8266 else {
8267 uv = -iv;
8268 esignbuf[esignlen++] = '-';
8269 }
46fc3d4c 8270 }
8271 base = 10;
8272 goto integer;
8273
fc36a67e 8274 case 'U':
29fe7a80 8275#ifdef IV_IS_QUAD
22f3ae8c 8276 intsize = 'q';
29fe7a80 8277#else
fc36a67e 8278 intsize = 'l';
29fe7a80 8279#endif
fc36a67e 8280 /* FALL THROUGH */
8281 case 'u':
8282 base = 10;
8283 goto uns_integer;
8284
4f19785b
WSI
8285 case 'b':
8286 base = 2;
8287 goto uns_integer;
8288
46fc3d4c 8289 case 'O':
29fe7a80 8290#ifdef IV_IS_QUAD
22f3ae8c 8291 intsize = 'q';
29fe7a80 8292#else
46fc3d4c 8293 intsize = 'l';
29fe7a80 8294#endif
46fc3d4c 8295 /* FALL THROUGH */
8296 case 'o':
8297 base = 8;
8298 goto uns_integer;
8299
8300 case 'X':
46fc3d4c 8301 case 'x':
8302 base = 16;
46fc3d4c 8303
8304 uns_integer:
b22c7a20 8305 if (vectorize) {
ba210ebe 8306 STRLEN ulen;
b22c7a20 8307 vector:
211dfcf1
HS
8308 if (!veclen)
8309 continue;
2cf2cfc6
A
8310 if (vec_utf8)
8311 uv = utf8n_to_uvchr(vecstr, veclen, &ulen,
8312 UTF8_ALLOW_ANYUV);
b22c7a20 8313 else {
a05b299f 8314 uv = *vecstr;
b22c7a20
GS
8315 ulen = 1;
8316 }
8317 vecstr += ulen;
8318 veclen -= ulen;
8319 }
8320 else if (args) {
46fc3d4c 8321 switch (intsize) {
8322 case 'h': uv = (unsigned short)va_arg(*args, unsigned); break;
46fc3d4c 8323 case 'l': uv = va_arg(*args, unsigned long); break;
fc36a67e 8324 case 'V': uv = va_arg(*args, UV); break;
b10c0dba 8325 default: uv = va_arg(*args, unsigned); break;
cf2093f6 8326#ifdef HAS_QUAD
9e3321a5 8327 case 'q': uv = va_arg(*args, Uquad_t); break;
cf2093f6 8328#endif
46fc3d4c 8329 }
8330 }
8331 else {
b10c0dba 8332 UV tuv = SvUVx(argsv); /* work around GCC bug #13488 */
46fc3d4c 8333 switch (intsize) {
b10c0dba
MHM
8334 case 'h': uv = (unsigned short)tuv; break;
8335 case 'l': uv = (unsigned long)tuv; break;
8336 case 'V':
8337 default: uv = tuv; break;
cf2093f6 8338#ifdef HAS_QUAD
b10c0dba 8339 case 'q': uv = (Uquad_t)tuv; break;
cf2093f6 8340#endif
46fc3d4c 8341 }
8342 }
8343
8344 integer:
4d84ee25
NC
8345 {
8346 char *ptr = ebuf + sizeof ebuf;
8347 switch (base) {
8348 unsigned dig;
8349 case 16:
8350 if (!uv)
8351 alt = FALSE;
8352 p = (char*)((c == 'X')
8353 ? "0123456789ABCDEF" : "0123456789abcdef");
8354 do {
8355 dig = uv & 15;
8356 *--ptr = p[dig];
8357 } while (uv >>= 4);
8358 if (alt) {
8359 esignbuf[esignlen++] = '0';
8360 esignbuf[esignlen++] = c; /* 'x' or 'X' */
8361 }
8362 break;
8363 case 8:
8364 do {
8365 dig = uv & 7;
8366 *--ptr = '0' + dig;
8367 } while (uv >>= 3);
8368 if (alt && *ptr != '0')
8369 *--ptr = '0';
8370 break;
8371 case 2:
ed2b91d2
GA
8372 if (!uv)
8373 alt = FALSE;
4d84ee25
NC
8374 do {
8375 dig = uv & 1;
8376 *--ptr = '0' + dig;
8377 } while (uv >>= 1);
8378 if (alt) {
8379 esignbuf[esignlen++] = '0';
8380 esignbuf[esignlen++] = 'b';
8381 }
8382 break;
8383 default: /* it had better be ten or less */
8384 do {
8385 dig = uv % base;
8386 *--ptr = '0' + dig;
8387 } while (uv /= base);
8388 break;
46fc3d4c 8389 }
4d84ee25
NC
8390 elen = (ebuf + sizeof ebuf) - ptr;
8391 eptr = ptr;
8392 if (has_precis) {
8393 if (precis > elen)
8394 zeros = precis - elen;
8395 else if (precis == 0 && elen == 1 && *eptr == '0')
8396 elen = 0;
eda88b6d 8397 }
c10ed8b9 8398 }
46fc3d4c 8399 break;
8400
8401 /* FLOATING POINT */
8402
fc36a67e 8403 case 'F':
8404 c = 'f'; /* maybe %F isn't supported here */
8405 /* FALL THROUGH */
46fc3d4c 8406 case 'e': case 'E':
fc36a67e 8407 case 'f':
46fc3d4c 8408 case 'g': case 'G':
26372e71
GA
8409 if (vectorize)
8410 goto unknown;
46fc3d4c 8411
8412 /* This is evil, but floating point is even more evil */
8413
9e5b023a
JH
8414 /* for SV-style calling, we can only get NV
8415 for C-style calling, we assume %f is double;
8416 for simplicity we allow any of %Lf, %llf, %qf for long double
8417 */
8418 switch (intsize) {
8419 case 'V':
8420#if defined(USE_LONG_DOUBLE)
8421 intsize = 'q';
8422#endif
8423 break;
8a2e3f14 8424/* [perl #20339] - we should accept and ignore %lf rather than die */
00e17364
HS
8425 case 'l':
8426 /* FALL THROUGH */
9e5b023a
JH
8427 default:
8428#if defined(USE_LONG_DOUBLE)
8429 intsize = args ? 0 : 'q';
8430#endif
8431 break;
8432 case 'q':
8433#if defined(HAS_LONG_DOUBLE)
8434 break;
8435#else
8436 /* FALL THROUGH */
8437#endif
8438 case 'h':
9e5b023a
JH
8439 goto unknown;
8440 }
8441
8442 /* now we need (long double) if intsize == 'q', else (double) */
26372e71 8443 nv = (args) ?
35fff930
JH
8444#if LONG_DOUBLESIZE > DOUBLESIZE
8445 intsize == 'q' ?
205f51d8
AS
8446 va_arg(*args, long double) :
8447 va_arg(*args, double)
35fff930 8448#else
205f51d8 8449 va_arg(*args, double)
35fff930 8450#endif
9e5b023a 8451 : SvNVx(argsv);
fc36a67e 8452
8453 need = 0;
8454 if (c != 'e' && c != 'E') {
8455 i = PERL_INT_MIN;
9e5b023a
JH
8456 /* FIXME: if HAS_LONG_DOUBLE but not USE_LONG_DOUBLE this
8457 will cast our (long double) to (double) */
73b309ea 8458 (void)Perl_frexp(nv, &i);
fc36a67e 8459 if (i == PERL_INT_MIN)
cea2e8a9 8460 Perl_die(aTHX_ "panic: frexp");
c635e13b 8461 if (i > 0)
fc36a67e 8462 need = BIT_DIGITS(i);
8463 }
8464 need += has_precis ? precis : 6; /* known default */
20f6aaab 8465
fc36a67e 8466 if (need < width)
8467 need = width;
8468
20f6aaab
AS
8469#ifdef HAS_LDBL_SPRINTF_BUG
8470 /* This is to try to fix a bug with irix/nonstop-ux/powerux and
205f51d8
AS
8471 with sfio - Allen <allens@cpan.org> */
8472
8473# ifdef DBL_MAX
8474# define MY_DBL_MAX DBL_MAX
8475# else /* XXX guessing! HUGE_VAL may be defined as infinity, so not using */
8476# if DOUBLESIZE >= 8
8477# define MY_DBL_MAX 1.7976931348623157E+308L
8478# else
8479# define MY_DBL_MAX 3.40282347E+38L
8480# endif
8481# endif
8482
8483# ifdef HAS_LDBL_SPRINTF_BUG_LESS1 /* only between -1L & 1L - Allen */
8484# define MY_DBL_MAX_BUG 1L
20f6aaab 8485# else
205f51d8 8486# define MY_DBL_MAX_BUG MY_DBL_MAX
20f6aaab 8487# endif
20f6aaab 8488
205f51d8
AS
8489# ifdef DBL_MIN
8490# define MY_DBL_MIN DBL_MIN
8491# else /* XXX guessing! -Allen */
8492# if DOUBLESIZE >= 8
8493# define MY_DBL_MIN 2.2250738585072014E-308L
8494# else
8495# define MY_DBL_MIN 1.17549435E-38L
8496# endif
8497# endif
20f6aaab 8498
205f51d8
AS
8499 if ((intsize == 'q') && (c == 'f') &&
8500 ((nv < MY_DBL_MAX_BUG) && (nv > -MY_DBL_MAX_BUG)) &&
8501 (need < DBL_DIG)) {
8502 /* it's going to be short enough that
8503 * long double precision is not needed */
8504
8505 if ((nv <= 0L) && (nv >= -0L))
8506 fix_ldbl_sprintf_bug = TRUE; /* 0 is 0 - easiest */
8507 else {
8508 /* would use Perl_fp_class as a double-check but not
8509 * functional on IRIX - see perl.h comments */
8510
8511 if ((nv >= MY_DBL_MIN) || (nv <= -MY_DBL_MIN)) {
8512 /* It's within the range that a double can represent */
8513#if defined(DBL_MAX) && !defined(DBL_MIN)
8514 if ((nv >= ((long double)1/DBL_MAX)) ||
8515 (nv <= (-(long double)1/DBL_MAX)))
20f6aaab 8516#endif
205f51d8 8517 fix_ldbl_sprintf_bug = TRUE;
20f6aaab 8518 }
205f51d8
AS
8519 }
8520 if (fix_ldbl_sprintf_bug == TRUE) {
8521 double temp;
8522
8523 intsize = 0;
8524 temp = (double)nv;
8525 nv = (NV)temp;
8526 }
20f6aaab 8527 }
205f51d8
AS
8528
8529# undef MY_DBL_MAX
8530# undef MY_DBL_MAX_BUG
8531# undef MY_DBL_MIN
8532
20f6aaab
AS
8533#endif /* HAS_LDBL_SPRINTF_BUG */
8534
46fc3d4c 8535 need += 20; /* fudge factor */
80252599
GS
8536 if (PL_efloatsize < need) {
8537 Safefree(PL_efloatbuf);
8538 PL_efloatsize = need + 20; /* more fudge */
a02a5408 8539 Newx(PL_efloatbuf, PL_efloatsize, char);
7d5ea4e7 8540 PL_efloatbuf[0] = '\0';
46fc3d4c 8541 }
8542
4151a5fe
IZ
8543 if ( !(width || left || plus || alt) && fill != '0'
8544 && has_precis && intsize != 'q' ) { /* Shortcuts */
2873255c
NC
8545 /* See earlier comment about buggy Gconvert when digits,
8546 aka precis is 0 */
8547 if ( c == 'g' && precis) {
2e59c212 8548 Gconvert((NV)nv, (int)precis, 0, PL_efloatbuf);
4150c189
NC
8549 /* May return an empty string for digits==0 */
8550 if (*PL_efloatbuf) {
8551 elen = strlen(PL_efloatbuf);
4151a5fe 8552 goto float_converted;
4150c189 8553 }
4151a5fe
IZ
8554 } else if ( c == 'f' && !precis) {
8555 if ((eptr = F0convert(nv, ebuf + sizeof ebuf, &elen)))
8556 break;
8557 }
8558 }
4d84ee25
NC
8559 {
8560 char *ptr = ebuf + sizeof ebuf;
8561 *--ptr = '\0';
8562 *--ptr = c;
8563 /* FIXME: what to do if HAS_LONG_DOUBLE but not PERL_PRIfldbl? */
9e5b023a 8564#if defined(HAS_LONG_DOUBLE) && defined(PERL_PRIfldbl)
4d84ee25
NC
8565 if (intsize == 'q') {
8566 /* Copy the one or more characters in a long double
8567 * format before the 'base' ([efgEFG]) character to
8568 * the format string. */
8569 static char const prifldbl[] = PERL_PRIfldbl;
8570 char const *p = prifldbl + sizeof(prifldbl) - 3;
8571 while (p >= prifldbl) { *--ptr = *p--; }
8572 }
65202027 8573#endif
4d84ee25
NC
8574 if (has_precis) {
8575 base = precis;
8576 do { *--ptr = '0' + (base % 10); } while (base /= 10);
8577 *--ptr = '.';
8578 }
8579 if (width) {
8580 base = width;
8581 do { *--ptr = '0' + (base % 10); } while (base /= 10);
8582 }
8583 if (fill == '0')
8584 *--ptr = fill;
8585 if (left)
8586 *--ptr = '-';
8587 if (plus)
8588 *--ptr = plus;
8589 if (alt)
8590 *--ptr = '#';
8591 *--ptr = '%';
8592
8593 /* No taint. Otherwise we are in the strange situation
8594 * where printf() taints but print($float) doesn't.
8595 * --jhi */
9e5b023a 8596#if defined(HAS_LONG_DOUBLE)
4150c189
NC
8597 elen = ((intsize == 'q')
8598 ? my_sprintf(PL_efloatbuf, ptr, nv)
8599 : my_sprintf(PL_efloatbuf, ptr, (double)nv));
9e5b023a 8600#else
4150c189 8601 elen = my_sprintf(PL_efloatbuf, ptr, nv);
9e5b023a 8602#endif
4d84ee25 8603 }
4151a5fe 8604 float_converted:
80252599 8605 eptr = PL_efloatbuf;
46fc3d4c 8606 break;
8607
fc36a67e 8608 /* SPECIAL */
8609
8610 case 'n':
26372e71
GA
8611 if (vectorize)
8612 goto unknown;
fc36a67e 8613 i = SvCUR(sv) - origlen;
26372e71 8614 if (args) {
c635e13b 8615 switch (intsize) {
8616 case 'h': *(va_arg(*args, short*)) = i; break;
8617 default: *(va_arg(*args, int*)) = i; break;
8618 case 'l': *(va_arg(*args, long*)) = i; break;
8619 case 'V': *(va_arg(*args, IV*)) = i; break;
cf2093f6
JH
8620#ifdef HAS_QUAD
8621 case 'q': *(va_arg(*args, Quad_t*)) = i; break;
8622#endif
c635e13b 8623 }
fc36a67e 8624 }
9dd79c3f 8625 else
211dfcf1 8626 sv_setuv_mg(argsv, (UV)i);
fc36a67e 8627 continue; /* not "break" */
8628
8629 /* UNKNOWN */
8630
46fc3d4c 8631 default:
fc36a67e 8632 unknown:
041457d9
DM
8633 if (!args
8634 && (PL_op->op_type == OP_PRTF || PL_op->op_type == OP_SPRINTF)
8635 && ckWARN(WARN_PRINTF))
8636 {
c4420975 8637 SV * const msg = sv_newmortal();
35c1215d
NC
8638 Perl_sv_setpvf(aTHX_ msg, "Invalid conversion in %sprintf: ",
8639 (PL_op->op_type == OP_PRTF) ? "" : "s");
0f4b6630 8640 if (c) {
0f4b6630 8641 if (isPRINT(c))
1c846c1f 8642 Perl_sv_catpvf(aTHX_ msg,
0f4b6630
JH
8643 "\"%%%c\"", c & 0xFF);
8644 else
8645 Perl_sv_catpvf(aTHX_ msg,
57def98f 8646 "\"%%\\%03"UVof"\"",
0f4b6630 8647 (UV)c & 0xFF);
0f4b6630 8648 } else
c635e13b 8649 sv_catpv(msg, "end of string");
9014280d 8650 Perl_warner(aTHX_ packWARN(WARN_PRINTF), "%"SVf, msg); /* yes, this is reentrant */
c635e13b 8651 }
fb73857a 8652
8653 /* output mangled stuff ... */
8654 if (c == '\0')
8655 --q;
46fc3d4c 8656 eptr = p;
8657 elen = q - p;
fb73857a 8658
8659 /* ... right here, because formatting flags should not apply */
8660 SvGROW(sv, SvCUR(sv) + elen + 1);
8661 p = SvEND(sv);
4459522c 8662 Copy(eptr, p, elen, char);
fb73857a 8663 p += elen;
8664 *p = '\0';
3f7c398e 8665 SvCUR_set(sv, p - SvPVX_const(sv));
58e33a90 8666 svix = osvix;
fb73857a 8667 continue; /* not "break" */
46fc3d4c 8668 }
8669
6c94ec8b
HS
8670 /* calculate width before utf8_upgrade changes it */
8671 have = esignlen + zeros + elen;
ed2b91d2
GA
8672 if (have < zeros)
8673 Perl_croak_nocontext(PL_memory_wrap);
6c94ec8b 8674
d2876be5
JH
8675 if (is_utf8 != has_utf8) {
8676 if (is_utf8) {
8677 if (SvCUR(sv))
8678 sv_utf8_upgrade(sv);
8679 }
8680 else {
53c1dcc0 8681 SV * const nsv = sv_2mortal(newSVpvn(eptr, elen));
d2876be5 8682 sv_utf8_upgrade(nsv);
93524f2b 8683 eptr = SvPVX_const(nsv);
d2876be5
JH
8684 elen = SvCUR(nsv);
8685 }
8686 SvGROW(sv, SvCUR(sv) + elen + 1);
8687 p = SvEND(sv);
8688 *p = '\0';
8689 }
6af65485 8690
46fc3d4c 8691 need = (have > width ? have : width);
8692 gap = need - have;
8693
d2641cbd
PC
8694 if (need >= (((STRLEN)~0) - SvCUR(sv) - dotstrlen - 1))
8695 Perl_croak_nocontext(PL_memory_wrap);
b22c7a20 8696 SvGROW(sv, SvCUR(sv) + need + dotstrlen + 1);
46fc3d4c 8697 p = SvEND(sv);
8698 if (esignlen && fill == '0') {
53c1dcc0 8699 int i;
eb160463 8700 for (i = 0; i < (int)esignlen; i++)
46fc3d4c 8701 *p++ = esignbuf[i];
8702 }
8703 if (gap && !left) {
8704 memset(p, fill, gap);
8705 p += gap;
8706 }
8707 if (esignlen && fill != '0') {
53c1dcc0 8708 int i;
eb160463 8709 for (i = 0; i < (int)esignlen; i++)
46fc3d4c 8710 *p++ = esignbuf[i];
8711 }
fc36a67e 8712 if (zeros) {
53c1dcc0 8713 int i;
fc36a67e 8714 for (i = zeros; i; i--)
8715 *p++ = '0';
8716 }
46fc3d4c 8717 if (elen) {
4459522c 8718 Copy(eptr, p, elen, char);
46fc3d4c 8719 p += elen;
8720 }
8721 if (gap && left) {
8722 memset(p, ' ', gap);
8723 p += gap;
8724 }
b22c7a20
GS
8725 if (vectorize) {
8726 if (veclen) {
4459522c 8727 Copy(dotstr, p, dotstrlen, char);
b22c7a20
GS
8728 p += dotstrlen;
8729 }
8730 else
8731 vectorize = FALSE; /* done iterating over vecstr */
8732 }
2cf2cfc6
A
8733 if (is_utf8)
8734 has_utf8 = TRUE;
8735 if (has_utf8)
7e2040f0 8736 SvUTF8_on(sv);
46fc3d4c 8737 *p = '\0';
3f7c398e 8738 SvCUR_set(sv, p - SvPVX_const(sv));
b22c7a20
GS
8739 if (vectorize) {
8740 esignlen = 0;
8741 goto vector;
8742 }
46fc3d4c 8743 }
8744}
51371543 8745
645c22ef
DM
8746/* =========================================================================
8747
8748=head1 Cloning an interpreter
8749
8750All the macros and functions in this section are for the private use of
8751the main function, perl_clone().
8752
8753The foo_dup() functions make an exact copy of an existing foo thinngy.
8754During the course of a cloning, a hash table is used to map old addresses
8755to new addresses. The table is created and manipulated with the
8756ptr_table_* functions.
8757
8758=cut
8759
8760============================================================================*/
8761
8762
1d7c1841
GS
8763#if defined(USE_ITHREADS)
8764
1d7c1841
GS
8765#ifndef GpREFCNT_inc
8766# define GpREFCNT_inc(gp) ((gp) ? (++(gp)->gp_refcnt, (gp)) : (GP*)NULL)
8767#endif
8768
8769
d2d73c3e
AB
8770#define sv_dup_inc(s,t) SvREFCNT_inc(sv_dup(s,t))
8771#define av_dup(s,t) (AV*)sv_dup((SV*)s,t)
8772#define av_dup_inc(s,t) (AV*)SvREFCNT_inc(sv_dup((SV*)s,t))
8773#define hv_dup(s,t) (HV*)sv_dup((SV*)s,t)
8774#define hv_dup_inc(s,t) (HV*)SvREFCNT_inc(sv_dup((SV*)s,t))
8775#define cv_dup(s,t) (CV*)sv_dup((SV*)s,t)
8776#define cv_dup_inc(s,t) (CV*)SvREFCNT_inc(sv_dup((SV*)s,t))
8777#define io_dup(s,t) (IO*)sv_dup((SV*)s,t)
8778#define io_dup_inc(s,t) (IO*)SvREFCNT_inc(sv_dup((SV*)s,t))
8779#define gv_dup(s,t) (GV*)sv_dup((SV*)s,t)
8780#define gv_dup_inc(s,t) (GV*)SvREFCNT_inc(sv_dup((SV*)s,t))
1d7c1841
GS
8781#define SAVEPV(p) (p ? savepv(p) : Nullch)
8782#define SAVEPVN(p,n) (p ? savepvn(p,n) : Nullch)
8cf8f3d1 8783
d2d73c3e 8784
d2f185dc
AMS
8785/* Duplicate a regexp. Required reading: pregcomp() and pregfree() in
8786 regcomp.c. AMS 20010712 */
645c22ef 8787
1d7c1841 8788REGEXP *
53c1dcc0 8789Perl_re_dup(pTHX_ const REGEXP *r, CLONE_PARAMS *param)
1d7c1841 8790{
27da23d5 8791 dVAR;
d2f185dc
AMS
8792 REGEXP *ret;
8793 int i, len, npar;
8794 struct reg_substr_datum *s;
8795
8796 if (!r)
8797 return (REGEXP *)NULL;
8798
8799 if ((ret = (REGEXP *)ptr_table_fetch(PL_ptr_table, r)))
8800 return ret;
8801
8802 len = r->offsets[0];
8803 npar = r->nparens+1;
8804
a02a5408 8805 Newxc(ret, sizeof(regexp) + (len+1)*sizeof(regnode), char, regexp);
d2f185dc
AMS
8806 Copy(r->program, ret->program, len+1, regnode);
8807
a02a5408 8808 Newx(ret->startp, npar, I32);
d2f185dc 8809 Copy(r->startp, ret->startp, npar, I32);
a02a5408 8810 Newx(ret->endp, npar, I32);
d2f185dc
AMS
8811 Copy(r->startp, ret->startp, npar, I32);
8812
a02a5408 8813 Newx(ret->substrs, 1, struct reg_substr_data);
d2f185dc
AMS
8814 for (s = ret->substrs->data, i = 0; i < 3; i++, s++) {
8815 s->min_offset = r->substrs->data[i].min_offset;
8816 s->max_offset = r->substrs->data[i].max_offset;
8817 s->substr = sv_dup_inc(r->substrs->data[i].substr, param);
33b8afdf 8818 s->utf8_substr = sv_dup_inc(r->substrs->data[i].utf8_substr, param);
d2f185dc
AMS
8819 }
8820
70612e96 8821 ret->regstclass = NULL;
d2f185dc
AMS
8822 if (r->data) {
8823 struct reg_data *d;
e1ec3a88 8824 const int count = r->data->count;
53c1dcc0 8825 int i;
d2f185dc 8826
a02a5408 8827 Newxc(d, sizeof(struct reg_data) + count*sizeof(void *),
d2f185dc 8828 char, struct reg_data);
a02a5408 8829 Newx(d->what, count, U8);
d2f185dc
AMS
8830
8831 d->count = count;
8832 for (i = 0; i < count; i++) {
8833 d->what[i] = r->data->what[i];
8834 switch (d->what[i]) {
a3621e74
YO
8835 /* legal options are one of: sfpont
8836 see also regcomp.h and pregfree() */
d2f185dc
AMS
8837 case 's':
8838 d->data[i] = sv_dup_inc((SV *)r->data->data[i], param);
8839 break;
8840 case 'p':
8841 d->data[i] = av_dup_inc((AV *)r->data->data[i], param);
8842 break;
8843 case 'f':
8844 /* This is cheating. */
a02a5408 8845 Newx(d->data[i], 1, struct regnode_charclass_class);
d2f185dc
AMS
8846 StructCopy(r->data->data[i], d->data[i],
8847 struct regnode_charclass_class);
70612e96 8848 ret->regstclass = (regnode*)d->data[i];
d2f185dc
AMS
8849 break;
8850 case 'o':
33773810
AMS
8851 /* Compiled op trees are readonly, and can thus be
8852 shared without duplication. */
b34c0dd4 8853 OP_REFCNT_LOCK;
9b978d73 8854 d->data[i] = (void*)OpREFCNT_inc((OP*)r->data->data[i]);
b34c0dd4 8855 OP_REFCNT_UNLOCK;
9b978d73 8856 break;
d2f185dc
AMS
8857 case 'n':
8858 d->data[i] = r->data->data[i];
8859 break;
a3621e74
YO
8860 case 't':
8861 d->data[i] = r->data->data[i];
8862 OP_REFCNT_LOCK;
8863 ((reg_trie_data*)d->data[i])->refcount++;
8864 OP_REFCNT_UNLOCK;
8865 break;
8866 default:
8867 Perl_croak(aTHX_ "panic: re_dup unknown data code '%c'", r->data->what[i]);
d2f185dc
AMS
8868 }
8869 }
8870
8871 ret->data = d;
8872 }
8873 else
8874 ret->data = NULL;
8875
a02a5408 8876 Newx(ret->offsets, 2*len+1, U32);
d2f185dc
AMS
8877 Copy(r->offsets, ret->offsets, 2*len+1, U32);
8878
e01c5899 8879 ret->precomp = SAVEPVN(r->precomp, r->prelen);
d2f185dc
AMS
8880 ret->refcnt = r->refcnt;
8881 ret->minlen = r->minlen;
8882 ret->prelen = r->prelen;
8883 ret->nparens = r->nparens;
8884 ret->lastparen = r->lastparen;
8885 ret->lastcloseparen = r->lastcloseparen;
8886 ret->reganch = r->reganch;
8887
70612e96
RG
8888 ret->sublen = r->sublen;
8889
8890 if (RX_MATCH_COPIED(ret))
e01c5899 8891 ret->subbeg = SAVEPVN(r->subbeg, r->sublen);
70612e96
RG
8892 else
8893 ret->subbeg = Nullch;
f8c7b90f 8894#ifdef PERL_OLD_COPY_ON_WRITE
9a26048b
NC
8895 ret->saved_copy = Nullsv;
8896#endif
70612e96 8897
d2f185dc
AMS
8898 ptr_table_store(PL_ptr_table, r, ret);
8899 return ret;
1d7c1841
GS
8900}
8901
d2d73c3e 8902/* duplicate a file handle */
645c22ef 8903
1d7c1841 8904PerlIO *
a8fc9800 8905Perl_fp_dup(pTHX_ PerlIO *fp, char type, CLONE_PARAMS *param)
1d7c1841
GS
8906{
8907 PerlIO *ret;
53c1dcc0
AL
8908
8909 PERL_UNUSED_ARG(type);
73d840c0 8910
1d7c1841
GS
8911 if (!fp)
8912 return (PerlIO*)NULL;
8913
8914 /* look for it in the table first */
8915 ret = (PerlIO*)ptr_table_fetch(PL_ptr_table, fp);
8916 if (ret)
8917 return ret;
8918
8919 /* create anew and remember what it is */
ecdeb87c 8920 ret = PerlIO_fdupopen(aTHX_ fp, param, PERLIO_DUP_CLONE);
1d7c1841
GS
8921 ptr_table_store(PL_ptr_table, fp, ret);
8922 return ret;
8923}
8924
645c22ef
DM
8925/* duplicate a directory handle */
8926
1d7c1841
GS
8927DIR *
8928Perl_dirp_dup(pTHX_ DIR *dp)
8929{
8930 if (!dp)
8931 return (DIR*)NULL;
8932 /* XXX TODO */
8933 return dp;
8934}
8935
ff276b08 8936/* duplicate a typeglob */
645c22ef 8937
1d7c1841 8938GP *
a8fc9800 8939Perl_gp_dup(pTHX_ GP *gp, CLONE_PARAMS* param)
1d7c1841
GS
8940{
8941 GP *ret;
8942 if (!gp)
8943 return (GP*)NULL;
8944 /* look for it in the table first */
8945 ret = (GP*)ptr_table_fetch(PL_ptr_table, gp);
8946 if (ret)
8947 return ret;
8948
8949 /* create anew and remember what it is */
a02a5408 8950 Newxz(ret, 1, GP);
1d7c1841
GS
8951 ptr_table_store(PL_ptr_table, gp, ret);
8952
8953 /* clone */
8954 ret->gp_refcnt = 0; /* must be before any other dups! */
d2d73c3e
AB
8955 ret->gp_sv = sv_dup_inc(gp->gp_sv, param);
8956 ret->gp_io = io_dup_inc(gp->gp_io, param);
8957 ret->gp_form = cv_dup_inc(gp->gp_form, param);
8958 ret->gp_av = av_dup_inc(gp->gp_av, param);
8959 ret->gp_hv = hv_dup_inc(gp->gp_hv, param);
8960 ret->gp_egv = gv_dup(gp->gp_egv, param);/* GvEGV is not refcounted */
8961 ret->gp_cv = cv_dup_inc(gp->gp_cv, param);
1d7c1841 8962 ret->gp_cvgen = gp->gp_cvgen;
1d7c1841
GS
8963 ret->gp_line = gp->gp_line;
8964 ret->gp_file = gp->gp_file; /* points to COP.cop_file */
8965 return ret;
8966}
8967
645c22ef
DM
8968/* duplicate a chain of magic */
8969
1d7c1841 8970MAGIC *
a8fc9800 8971Perl_mg_dup(pTHX_ MAGIC *mg, CLONE_PARAMS* param)
1d7c1841 8972{
cb359b41
JH
8973 MAGIC *mgprev = (MAGIC*)NULL;
8974 MAGIC *mgret;
1d7c1841
GS
8975 if (!mg)
8976 return (MAGIC*)NULL;
8977 /* look for it in the table first */
8978 mgret = (MAGIC*)ptr_table_fetch(PL_ptr_table, mg);
8979 if (mgret)
8980 return mgret;
8981
8982 for (; mg; mg = mg->mg_moremagic) {
8983 MAGIC *nmg;
a02a5408 8984 Newxz(nmg, 1, MAGIC);
cb359b41 8985 if (mgprev)
1d7c1841 8986 mgprev->mg_moremagic = nmg;
cb359b41
JH
8987 else
8988 mgret = nmg;
1d7c1841
GS
8989 nmg->mg_virtual = mg->mg_virtual; /* XXX copy dynamic vtable? */
8990 nmg->mg_private = mg->mg_private;
8991 nmg->mg_type = mg->mg_type;
8992 nmg->mg_flags = mg->mg_flags;
14befaf4 8993 if (mg->mg_type == PERL_MAGIC_qr) {
d2f185dc 8994 nmg->mg_obj = (SV*)re_dup((REGEXP*)mg->mg_obj, param);
1d7c1841 8995 }
05bd4103 8996 else if(mg->mg_type == PERL_MAGIC_backref) {
7fc63493 8997 const AV * const av = (AV*) mg->mg_obj;
fdc9a813
AE
8998 SV **svp;
8999 I32 i;
7fc63493 9000 (void)SvREFCNT_inc(nmg->mg_obj = (SV*)newAV());
e3d99884 9001 AvREAL_off((AV*)nmg->mg_obj);
fdc9a813
AE
9002 svp = AvARRAY(av);
9003 for (i = AvFILLp(av); i >= 0; i--) {
3a81978b 9004 if (!svp[i]) continue;
fdc9a813
AE
9005 av_push((AV*)nmg->mg_obj,sv_dup(svp[i],param));
9006 }
05bd4103 9007 }
8d2f4536
NC
9008 else if (mg->mg_type == PERL_MAGIC_symtab) {
9009 nmg->mg_obj = mg->mg_obj;
9010 }
1d7c1841
GS
9011 else {
9012 nmg->mg_obj = (mg->mg_flags & MGf_REFCOUNTED)
d2d73c3e
AB
9013 ? sv_dup_inc(mg->mg_obj, param)
9014 : sv_dup(mg->mg_obj, param);
1d7c1841
GS
9015 }
9016 nmg->mg_len = mg->mg_len;
9017 nmg->mg_ptr = mg->mg_ptr; /* XXX random ptr? */
14befaf4 9018 if (mg->mg_ptr && mg->mg_type != PERL_MAGIC_regex_global) {
68795e93 9019 if (mg->mg_len > 0) {
1d7c1841 9020 nmg->mg_ptr = SAVEPVN(mg->mg_ptr, mg->mg_len);
14befaf4
DM
9021 if (mg->mg_type == PERL_MAGIC_overload_table &&
9022 AMT_AMAGIC((AMT*)mg->mg_ptr))
9023 {
c445ea15 9024 const AMT * const amtp = (AMT*)mg->mg_ptr;
0bcc34c2 9025 AMT * const namtp = (AMT*)nmg->mg_ptr;
1d7c1841
GS
9026 I32 i;
9027 for (i = 1; i < NofAMmeth; i++) {
d2d73c3e 9028 namtp->table[i] = cv_dup_inc(amtp->table[i], param);
1d7c1841
GS
9029 }
9030 }
9031 }
9032 else if (mg->mg_len == HEf_SVKEY)
d2d73c3e 9033 nmg->mg_ptr = (char*)sv_dup_inc((SV*)mg->mg_ptr, param);
1d7c1841 9034 }
68795e93
NIS
9035 if ((mg->mg_flags & MGf_DUP) && mg->mg_virtual && mg->mg_virtual->svt_dup) {
9036 CALL_FPTR(nmg->mg_virtual->svt_dup)(aTHX_ nmg, param);
9037 }
1d7c1841
GS
9038 mgprev = nmg;
9039 }
9040 return mgret;
9041}
9042
645c22ef
DM
9043/* create a new pointer-mapping table */
9044
1d7c1841
GS
9045PTR_TBL_t *
9046Perl_ptr_table_new(pTHX)
9047{
9048 PTR_TBL_t *tbl;
a02a5408 9049 Newxz(tbl, 1, PTR_TBL_t);
1d7c1841
GS
9050 tbl->tbl_max = 511;
9051 tbl->tbl_items = 0;
a02a5408 9052 Newxz(tbl->tbl_ary, tbl->tbl_max + 1, PTR_TBL_ENT_t*);
1d7c1841
GS
9053 return tbl;
9054}
9055
7119fd33
NC
9056#define PTR_TABLE_HASH(ptr) \
9057 ((PTR2UV(ptr) >> 3) ^ (PTR2UV(ptr) >> (3 + 7)) ^ (PTR2UV(ptr) >> (3 + 17)))
134ca3d6 9058
93e68bfb
JC
9059/*
9060 we use the PTE_SVSLOT 'reservation' made above, both here (in the
9061 following define) and at call to new_body_inline made below in
9062 Perl_ptr_table_store()
9063 */
9064
9065#define del_pte(p) del_body_type(p, PTE_SVSLOT)
32e691d0 9066
645c22ef
DM
9067/* map an existing pointer using a table */
9068
7bf61b54 9069STATIC PTR_TBL_ENT_t *
e4cd1874 9070S_ptr_table_find(pTHX_ PTR_TBL_t *tbl, const void *sv) {
1d7c1841 9071 PTR_TBL_ENT_t *tblent;
4373e329 9072 const UV hash = PTR_TABLE_HASH(sv);
1d7c1841
GS
9073 assert(tbl);
9074 tblent = tbl->tbl_ary[hash & tbl->tbl_max];
9075 for (; tblent; tblent = tblent->next) {
9076 if (tblent->oldval == sv)
7bf61b54 9077 return tblent;
1d7c1841 9078 }
7bf61b54
NC
9079 return 0;
9080}
9081
9082void *
9083Perl_ptr_table_fetch(pTHX_ PTR_TBL_t *tbl, const void *sv)
9084{
e4cd1874 9085 PTR_TBL_ENT_t const *const tblent = S_ptr_table_find(aTHX_ tbl, sv);
7bf61b54 9086 return tblent ? tblent->newval : (void *) 0;
1d7c1841
GS
9087}
9088
645c22ef
DM
9089/* add a new entry to a pointer-mapping table */
9090
1d7c1841 9091void
44f8325f 9092Perl_ptr_table_store(pTHX_ PTR_TBL_t *tbl, const void *oldsv, void *newsv)
1d7c1841 9093{
e4cd1874 9094 PTR_TBL_ENT_t *tblent = S_ptr_table_find(aTHX_ tbl, oldsv);
1d7c1841 9095
7bf61b54
NC
9096 if (tblent) {
9097 tblent->newval = newsv;
9098 } else {
9099 const UV entry = PTR_TABLE_HASH(oldsv) & tbl->tbl_max;
9100
9101 new_body_inline(tblent, sizeof(struct ptr_tbl_ent), PTE_SVSLOT);
9102 tblent->oldval = oldsv;
9103 tblent->newval = newsv;
9104 tblent->next = tbl->tbl_ary[entry];
9105 tbl->tbl_ary[entry] = tblent;
9106 tbl->tbl_items++;
9107 if (tblent->next && tbl->tbl_items > tbl->tbl_max)
9108 ptr_table_split(tbl);
1d7c1841 9109 }
1d7c1841
GS
9110}
9111
645c22ef
DM
9112/* double the hash bucket size of an existing ptr table */
9113
1d7c1841
GS
9114void
9115Perl_ptr_table_split(pTHX_ PTR_TBL_t *tbl)
9116{
9117 PTR_TBL_ENT_t **ary = tbl->tbl_ary;
4373e329 9118 const UV oldsize = tbl->tbl_max + 1;
1d7c1841
GS
9119 UV newsize = oldsize * 2;
9120 UV i;
9121
9122 Renew(ary, newsize, PTR_TBL_ENT_t*);
9123 Zero(&ary[oldsize], newsize-oldsize, PTR_TBL_ENT_t*);
9124 tbl->tbl_max = --newsize;
9125 tbl->tbl_ary = ary;
9126 for (i=0; i < oldsize; i++, ary++) {
9127 PTR_TBL_ENT_t **curentp, **entp, *ent;
9128 if (!*ary)
9129 continue;
9130 curentp = ary + oldsize;
9131 for (entp = ary, ent = *ary; ent; ent = *entp) {
134ca3d6 9132 if ((newsize & PTR_TABLE_HASH(ent->oldval)) != i) {
1d7c1841
GS
9133 *entp = ent->next;
9134 ent->next = *curentp;
9135 *curentp = ent;
9136 continue;
9137 }
9138 else
9139 entp = &ent->next;
9140 }
9141 }
9142}
9143
645c22ef
DM
9144/* remove all the entries from a ptr table */
9145
a0739874
DM
9146void
9147Perl_ptr_table_clear(pTHX_ PTR_TBL_t *tbl)
9148{
d5cefff9 9149 if (tbl && tbl->tbl_items) {
c445ea15 9150 register PTR_TBL_ENT_t * const * const array = tbl->tbl_ary;
d5cefff9 9151 UV riter = tbl->tbl_max;
a0739874 9152
d5cefff9
NC
9153 do {
9154 PTR_TBL_ENT_t *entry = array[riter];
ab1e7f95 9155
d5cefff9 9156 while (entry) {
00b6aa41 9157 PTR_TBL_ENT_t * const oentry = entry;
d5cefff9
NC
9158 entry = entry->next;
9159 del_pte(oentry);
9160 }
9161 } while (riter--);
a0739874 9162
d5cefff9
NC
9163 tbl->tbl_items = 0;
9164 }
a0739874
DM
9165}
9166
645c22ef
DM
9167/* clear and free a ptr table */
9168
a0739874
DM
9169void
9170Perl_ptr_table_free(pTHX_ PTR_TBL_t *tbl)
9171{
9172 if (!tbl) {
9173 return;
9174 }
9175 ptr_table_clear(tbl);
9176 Safefree(tbl->tbl_ary);
9177 Safefree(tbl);
9178}
9179
5bd07a3d 9180
83841fad
NIS
9181void
9182Perl_rvpv_dup(pTHX_ SV *dstr, SV *sstr, CLONE_PARAMS* param)
9183{
9184 if (SvROK(sstr)) {
b162af07
SP
9185 SvRV_set(dstr, SvWEAKREF(sstr)
9186 ? sv_dup(SvRV(sstr), param)
9187 : sv_dup_inc(SvRV(sstr), param));
f880fe2f 9188
83841fad 9189 }
3f7c398e 9190 else if (SvPVX_const(sstr)) {
83841fad
NIS
9191 /* Has something there */
9192 if (SvLEN(sstr)) {
68795e93 9193 /* Normal PV - clone whole allocated space */
3f7c398e 9194 SvPV_set(dstr, SAVEPVN(SvPVX_const(sstr), SvLEN(sstr)-1));
d3d0e6f1
NC
9195 if (SvREADONLY(sstr) && SvFAKE(sstr)) {
9196 /* Not that normal - actually sstr is copy on write.
9197 But we are a true, independant SV, so: */
9198 SvREADONLY_off(dstr);
9199 SvFAKE_off(dstr);
9200 }
68795e93 9201 }
83841fad
NIS
9202 else {
9203 /* Special case - not normally malloced for some reason */
ef10be65
NC
9204 if ((SvREADONLY(sstr) && SvFAKE(sstr))) {
9205 /* A "shared" PV - clone it as "shared" PV */
9206 SvPV_set(dstr,
9207 HEK_KEY(hek_dup(SvSHARED_HEK_FROM_PV(SvPVX_const(sstr)),
9208 param)));
83841fad
NIS
9209 }
9210 else {
9211 /* Some other special case - random pointer */
f880fe2f 9212 SvPV_set(dstr, SvPVX(sstr));
d3d0e6f1 9213 }
83841fad
NIS
9214 }
9215 }
9216 else {
9217 /* Copy the Null */
f880fe2f 9218 if (SvTYPE(dstr) == SVt_RV)
b162af07 9219 SvRV_set(dstr, NULL);
f880fe2f
SP
9220 else
9221 SvPV_set(dstr, 0);
83841fad
NIS
9222 }
9223}
9224
662fb8b2
NC
9225/* duplicate an SV of any type (including AV, HV etc) */
9226
1d7c1841 9227SV *
a8fc9800 9228Perl_sv_dup(pTHX_ SV *sstr, CLONE_PARAMS* param)
1d7c1841 9229{
27da23d5 9230 dVAR;
1d7c1841
GS
9231 SV *dstr;
9232
9233 if (!sstr || SvTYPE(sstr) == SVTYPEMASK)
9234 return Nullsv;
9235 /* look for it in the table first */
9236 dstr = (SV*)ptr_table_fetch(PL_ptr_table, sstr);
9237 if (dstr)
9238 return dstr;
9239
0405e91e
AB
9240 if(param->flags & CLONEf_JOIN_IN) {
9241 /** We are joining here so we don't want do clone
9242 something that is bad **/
bfcb3514 9243 const char *hvname;
0405e91e
AB
9244
9245 if(SvTYPE(sstr) == SVt_PVHV &&
bfcb3514 9246 (hvname = HvNAME_get(sstr))) {
0405e91e 9247 /** don't clone stashes if they already exist **/
551405c4 9248 return (SV*)gv_stashpv(hvname,0);
0405e91e
AB
9249 }
9250 }
9251
1d7c1841
GS
9252 /* create anew and remember what it is */
9253 new_SV(dstr);
fd0854ff
DM
9254
9255#ifdef DEBUG_LEAKING_SCALARS
9256 dstr->sv_debug_optype = sstr->sv_debug_optype;
9257 dstr->sv_debug_line = sstr->sv_debug_line;
9258 dstr->sv_debug_inpad = sstr->sv_debug_inpad;
9259 dstr->sv_debug_cloned = 1;
9260# ifdef NETWARE
9261 dstr->sv_debug_file = savepv(sstr->sv_debug_file);
9262# else
9263 dstr->sv_debug_file = savesharedpv(sstr->sv_debug_file);
9264# endif
9265#endif
9266
1d7c1841
GS
9267 ptr_table_store(PL_ptr_table, sstr, dstr);
9268
9269 /* clone */
9270 SvFLAGS(dstr) = SvFLAGS(sstr);
9271 SvFLAGS(dstr) &= ~SVf_OOK; /* don't propagate OOK hack */
9272 SvREFCNT(dstr) = 0; /* must be before any other dups! */
9273
9274#ifdef DEBUGGING
3f7c398e 9275 if (SvANY(sstr) && PL_watch_pvx && SvPVX_const(sstr) == PL_watch_pvx)
1d7c1841 9276 PerlIO_printf(Perl_debug_log, "watch at %p hit, found string \"%s\"\n",
3f7c398e 9277 PL_watch_pvx, SvPVX_const(sstr));
1d7c1841
GS
9278#endif
9279
9660f481
DM
9280 /* don't clone objects whose class has asked us not to */
9281 if (SvOBJECT(sstr) && ! (SvFLAGS(SvSTASH(sstr)) & SVphv_CLONEABLE)) {
9282 SvFLAGS(dstr) &= ~SVTYPEMASK;
9283 SvOBJECT_off(dstr);
9284 return dstr;
9285 }
9286
1d7c1841
GS
9287 switch (SvTYPE(sstr)) {
9288 case SVt_NULL:
9289 SvANY(dstr) = NULL;
9290 break;
9291 case SVt_IV:
339049b0 9292 SvANY(dstr) = (XPVIV*)((char*)&(dstr->sv_u.svu_iv) - STRUCT_OFFSET(XPVIV, xiv_iv));
45977657 9293 SvIV_set(dstr, SvIVX(sstr));
1d7c1841
GS
9294 break;
9295 case SVt_NV:
9296 SvANY(dstr) = new_XNV();
9d6ce603 9297 SvNV_set(dstr, SvNVX(sstr));
1d7c1841
GS
9298 break;
9299 case SVt_RV:
339049b0 9300 SvANY(dstr) = &(dstr->sv_u.svu_rv);
83841fad 9301 Perl_rvpv_dup(aTHX_ dstr, sstr, param);
1d7c1841 9302 break;
662fb8b2
NC
9303 default:
9304 {
9305 /* These are all the types that need complex bodies allocating. */
662fb8b2 9306 void *new_body;
2bcc16b3
NC
9307 const svtype sv_type = SvTYPE(sstr);
9308 const struct body_details *const sv_type_details
9309 = bodies_by_type + sv_type;
662fb8b2 9310
93e68bfb 9311 switch (sv_type) {
662fb8b2
NC
9312 default:
9313 Perl_croak(aTHX_ "Bizarre SvTYPE [%" IVdf "]",
9314 (IV)SvTYPE(sstr));
9315 break;
9316
662fb8b2
NC
9317 case SVt_PVGV:
9318 if (GvUNIQUE((GV*)sstr)) {
93e68bfb 9319 /* Do sharing here, and fall through */
662fb8b2 9320 }
c22188b4
NC
9321 case SVt_PVIO:
9322 case SVt_PVFM:
9323 case SVt_PVHV:
9324 case SVt_PVAV:
93e68bfb 9325 case SVt_PVBM:
662fb8b2 9326 case SVt_PVCV:
662fb8b2 9327 case SVt_PVLV:
662fb8b2 9328 case SVt_PVMG:
662fb8b2 9329 case SVt_PVNV:
662fb8b2 9330 case SVt_PVIV:
662fb8b2 9331 case SVt_PV:
3043b442 9332 assert(sv_type_details->size);
c22188b4 9333 if (sv_type_details->arena) {
3043b442 9334 new_body_inline(new_body, sv_type_details->size, sv_type);
c22188b4 9335 new_body
b9502f15 9336 = (void*)((char*)new_body - sv_type_details->offset);
c22188b4
NC
9337 } else {
9338 new_body = new_NOARENA(sv_type_details);
9339 }
1d7c1841 9340 }
662fb8b2
NC
9341 assert(new_body);
9342 SvANY(dstr) = new_body;
9343
2bcc16b3 9344#ifndef PURIFY
b9502f15
NC
9345 Copy(((char*)SvANY(sstr)) + sv_type_details->offset,
9346 ((char*)SvANY(dstr)) + sv_type_details->offset,
f32993d6 9347 sv_type_details->copy, char);
2bcc16b3
NC
9348#else
9349 Copy(((char*)SvANY(sstr)),
9350 ((char*)SvANY(dstr)),
b9502f15 9351 sv_type_details->size + sv_type_details->offset, char);
2bcc16b3 9352#endif
662fb8b2 9353
f32993d6 9354 if (sv_type != SVt_PVAV && sv_type != SVt_PVHV)
662fb8b2
NC
9355 Perl_rvpv_dup(aTHX_ dstr, sstr, param);
9356
9357 /* The Copy above means that all the source (unduplicated) pointers
9358 are now in the destination. We can check the flags and the
9359 pointers in either, but it's possible that there's less cache
9360 missing by always going for the destination.
9361 FIXME - instrument and check that assumption */
f32993d6 9362 if (sv_type >= SVt_PVMG) {
662fb8b2
NC
9363 if (SvMAGIC(dstr))
9364 SvMAGIC_set(dstr, mg_dup(SvMAGIC(dstr), param));
9365 if (SvSTASH(dstr))
9366 SvSTASH_set(dstr, hv_dup_inc(SvSTASH(dstr), param));
1d7c1841 9367 }
662fb8b2 9368
f32993d6
NC
9369 /* The cast silences a GCC warning about unhandled types. */
9370 switch ((int)sv_type) {
662fb8b2
NC
9371 case SVt_PV:
9372 break;
9373 case SVt_PVIV:
9374 break;
9375 case SVt_PVNV:
9376 break;
9377 case SVt_PVMG:
9378 break;
9379 case SVt_PVBM:
9380 break;
9381 case SVt_PVLV:
9382 /* XXX LvTARGOFF sometimes holds PMOP* when DEBUGGING */
9383 if (LvTYPE(dstr) == 't') /* for tie: unrefcnted fake (SV**) */
9384 LvTARG(dstr) = dstr;
9385 else if (LvTYPE(dstr) == 'T') /* for tie: fake HE */
9386 LvTARG(dstr) = (SV*)he_dup((HE*)LvTARG(dstr), 0, param);
9387 else
9388 LvTARG(dstr) = sv_dup_inc(LvTARG(dstr), param);
9389 break;
9390 case SVt_PVGV:
9391 GvNAME(dstr) = SAVEPVN(GvNAME(dstr), GvNAMELEN(dstr));
e15faf7d
NC
9392 GvSTASH(dstr) = hv_dup(GvSTASH(dstr), param);
9393 /* Don't call sv_add_backref here as it's going to be created
9394 as part of the magic cloning of the symbol table. */
662fb8b2
NC
9395 GvGP(dstr) = gp_dup(GvGP(dstr), param);
9396 (void)GpREFCNT_inc(GvGP(dstr));
9397 break;
9398 case SVt_PVIO:
9399 IoIFP(dstr) = fp_dup(IoIFP(dstr), IoTYPE(dstr), param);
9400 if (IoOFP(dstr) == IoIFP(sstr))
9401 IoOFP(dstr) = IoIFP(dstr);
9402 else
9403 IoOFP(dstr) = fp_dup(IoOFP(dstr), IoTYPE(dstr), param);
9404 /* PL_rsfp_filters entries have fake IoDIRP() */
9405 if (IoDIRP(dstr) && !(IoFLAGS(dstr) & IOf_FAKE_DIRP))
9406 IoDIRP(dstr) = dirp_dup(IoDIRP(dstr));
9407 if(IoFLAGS(dstr) & IOf_FAKE_DIRP) {
9408 /* I have no idea why fake dirp (rsfps)
9409 should be treated differently but otherwise
9410 we end up with leaks -- sky*/
9411 IoTOP_GV(dstr) = gv_dup_inc(IoTOP_GV(dstr), param);
9412 IoFMT_GV(dstr) = gv_dup_inc(IoFMT_GV(dstr), param);
9413 IoBOTTOM_GV(dstr) = gv_dup_inc(IoBOTTOM_GV(dstr), param);
9414 } else {
9415 IoTOP_GV(dstr) = gv_dup(IoTOP_GV(dstr), param);
9416 IoFMT_GV(dstr) = gv_dup(IoFMT_GV(dstr), param);
9417 IoBOTTOM_GV(dstr) = gv_dup(IoBOTTOM_GV(dstr), param);
9418 }
9419 IoTOP_NAME(dstr) = SAVEPV(IoTOP_NAME(dstr));
9420 IoFMT_NAME(dstr) = SAVEPV(IoFMT_NAME(dstr));
9421 IoBOTTOM_NAME(dstr) = SAVEPV(IoBOTTOM_NAME(dstr));
9422 break;
9423 case SVt_PVAV:
9424 if (AvARRAY((AV*)sstr)) {
9425 SV **dst_ary, **src_ary;
9426 SSize_t items = AvFILLp((AV*)sstr) + 1;
9427
9428 src_ary = AvARRAY((AV*)sstr);
a02a5408 9429 Newxz(dst_ary, AvMAX((AV*)sstr)+1, SV*);
662fb8b2
NC
9430 ptr_table_store(PL_ptr_table, src_ary, dst_ary);
9431 SvPV_set(dstr, (char*)dst_ary);
9432 AvALLOC((AV*)dstr) = dst_ary;
9433 if (AvREAL((AV*)sstr)) {
9434 while (items-- > 0)
9435 *dst_ary++ = sv_dup_inc(*src_ary++, param);
9436 }
9437 else {
9438 while (items-- > 0)
9439 *dst_ary++ = sv_dup(*src_ary++, param);
9440 }
9441 items = AvMAX((AV*)sstr) - AvFILLp((AV*)sstr);
9442 while (items-- > 0) {
9443 *dst_ary++ = &PL_sv_undef;
9444 }
bfcb3514 9445 }
662fb8b2
NC
9446 else {
9447 SvPV_set(dstr, Nullch);
9448 AvALLOC((AV*)dstr) = (SV**)NULL;
b79f7545 9449 }
662fb8b2
NC
9450 break;
9451 case SVt_PVHV:
9452 {
9453 HEK *hvname = 0;
9454
9455 if (HvARRAY((HV*)sstr)) {
9456 STRLEN i = 0;
9457 const bool sharekeys = !!HvSHAREKEYS(sstr);
9458 XPVHV * const dxhv = (XPVHV*)SvANY(dstr);
9459 XPVHV * const sxhv = (XPVHV*)SvANY(sstr);
9460 char *darray;
a02a5408 9461 Newx(darray, PERL_HV_ARRAY_ALLOC_BYTES(dxhv->xhv_max+1)
662fb8b2
NC
9462 + (SvOOK(sstr) ? sizeof(struct xpvhv_aux) : 0),
9463 char);
9464 HvARRAY(dstr) = (HE**)darray;
9465 while (i <= sxhv->xhv_max) {
5c4138a0 9466 const HE *source = HvARRAY(sstr)[i];
662fb8b2
NC
9467 HvARRAY(dstr)[i] = source
9468 ? he_dup(source, sharekeys, param) : 0;
9469 ++i;
9470 }
9471 if (SvOOK(sstr)) {
00b6aa41
AL
9472 struct xpvhv_aux * const saux = HvAUX(sstr);
9473 struct xpvhv_aux * const daux = HvAUX(dstr);
662fb8b2
NC
9474 /* This flag isn't copied. */
9475 /* SvOOK_on(hv) attacks the IV flags. */
9476 SvFLAGS(dstr) |= SVf_OOK;
9477
9478 hvname = saux->xhv_name;
dd690478
NC
9479 daux->xhv_name
9480 = hvname ? hek_dup(hvname, param) : hvname;
662fb8b2
NC
9481
9482 daux->xhv_riter = saux->xhv_riter;
9483 daux->xhv_eiter = saux->xhv_eiter
dd690478
NC
9484 ? he_dup(saux->xhv_eiter,
9485 (bool)!!HvSHAREKEYS(sstr), param) : 0;
662fb8b2
NC
9486 }
9487 }
9488 else {
9489 SvPV_set(dstr, Nullch);
9490 }
9491 /* Record stashes for possible cloning in Perl_clone(). */
9492 if(hvname)
9493 av_push(param->stashes, dstr);
9494 }
9495 break;
9496 case SVt_PVFM:
9497 case SVt_PVCV:
9498 /* NOTE: not refcounted */
9499 CvSTASH(dstr) = hv_dup(CvSTASH(dstr), param);
9500 OP_REFCNT_LOCK;
9501 CvROOT(dstr) = OpREFCNT_inc(CvROOT(dstr));
9502 OP_REFCNT_UNLOCK;
9503 if (CvCONST(dstr)) {
9504 CvXSUBANY(dstr).any_ptr = GvUNIQUE(CvGV(dstr)) ?
9505 SvREFCNT_inc(CvXSUBANY(dstr).any_ptr) :
9506 sv_dup_inc((SV *)CvXSUBANY(dstr).any_ptr, param);
9507 }
9508 /* don't dup if copying back - CvGV isn't refcounted, so the
9509 * duped GV may never be freed. A bit of a hack! DAPM */
9510 CvGV(dstr) = (param->flags & CLONEf_JOIN_IN) ?
9511 Nullgv : gv_dup(CvGV(dstr), param) ;
9512 if (!(param->flags & CLONEf_COPY_STACKS)) {
9513 CvDEPTH(dstr) = 0;
9514 }
9515 PAD_DUP(CvPADLIST(dstr), CvPADLIST(sstr), param);
9516 CvOUTSIDE(dstr) =
9517 CvWEAKOUTSIDE(sstr)
9518 ? cv_dup( CvOUTSIDE(dstr), param)
9519 : cv_dup_inc(CvOUTSIDE(dstr), param);
9520 if (!CvXSUB(dstr))
9521 CvFILE(dstr) = SAVEPV(CvFILE(dstr));
9522 break;
bfcb3514 9523 }
1d7c1841 9524 }
1d7c1841
GS
9525 }
9526
9527 if (SvOBJECT(dstr) && SvTYPE(dstr) != SVt_PVIO)
9528 ++PL_sv_objcount;
9529
9530 return dstr;
d2d73c3e 9531 }
1d7c1841 9532
645c22ef
DM
9533/* duplicate a context */
9534
1d7c1841 9535PERL_CONTEXT *
a8fc9800 9536Perl_cx_dup(pTHX_ PERL_CONTEXT *cxs, I32 ix, I32 max, CLONE_PARAMS* param)
1d7c1841
GS
9537{
9538 PERL_CONTEXT *ncxs;
9539
9540 if (!cxs)
9541 return (PERL_CONTEXT*)NULL;
9542
9543 /* look for it in the table first */
9544 ncxs = (PERL_CONTEXT*)ptr_table_fetch(PL_ptr_table, cxs);
9545 if (ncxs)
9546 return ncxs;
9547
9548 /* create anew and remember what it is */
a02a5408 9549 Newxz(ncxs, max + 1, PERL_CONTEXT);
1d7c1841
GS
9550 ptr_table_store(PL_ptr_table, cxs, ncxs);
9551
9552 while (ix >= 0) {
c445ea15
AL
9553 PERL_CONTEXT * const cx = &cxs[ix];
9554 PERL_CONTEXT * const ncx = &ncxs[ix];
1d7c1841
GS
9555 ncx->cx_type = cx->cx_type;
9556 if (CxTYPE(cx) == CXt_SUBST) {
9557 Perl_croak(aTHX_ "Cloning substitution context is unimplemented");
9558 }
9559 else {
9560 ncx->blk_oldsp = cx->blk_oldsp;
9561 ncx->blk_oldcop = cx->blk_oldcop;
1d7c1841
GS
9562 ncx->blk_oldmarksp = cx->blk_oldmarksp;
9563 ncx->blk_oldscopesp = cx->blk_oldscopesp;
9564 ncx->blk_oldpm = cx->blk_oldpm;
9565 ncx->blk_gimme = cx->blk_gimme;
9566 switch (CxTYPE(cx)) {
9567 case CXt_SUB:
9568 ncx->blk_sub.cv = (cx->blk_sub.olddepth == 0
d2d73c3e
AB
9569 ? cv_dup_inc(cx->blk_sub.cv, param)
9570 : cv_dup(cx->blk_sub.cv,param));
1d7c1841 9571 ncx->blk_sub.argarray = (cx->blk_sub.hasargs
d2d73c3e 9572 ? av_dup_inc(cx->blk_sub.argarray, param)
7d49f689 9573 : NULL);
d2d73c3e 9574 ncx->blk_sub.savearray = av_dup_inc(cx->blk_sub.savearray, param);
1d7c1841
GS
9575 ncx->blk_sub.olddepth = cx->blk_sub.olddepth;
9576 ncx->blk_sub.hasargs = cx->blk_sub.hasargs;
9577 ncx->blk_sub.lval = cx->blk_sub.lval;
f39bc417 9578 ncx->blk_sub.retop = cx->blk_sub.retop;
1d7c1841
GS
9579 break;
9580 case CXt_EVAL:
9581 ncx->blk_eval.old_in_eval = cx->blk_eval.old_in_eval;
9582 ncx->blk_eval.old_op_type = cx->blk_eval.old_op_type;
b47cad08 9583 ncx->blk_eval.old_namesv = sv_dup_inc(cx->blk_eval.old_namesv, param);
1d7c1841 9584 ncx->blk_eval.old_eval_root = cx->blk_eval.old_eval_root;
d2d73c3e 9585 ncx->blk_eval.cur_text = sv_dup(cx->blk_eval.cur_text, param);
f39bc417 9586 ncx->blk_eval.retop = cx->blk_eval.retop;
1d7c1841
GS
9587 break;
9588 case CXt_LOOP:
9589 ncx->blk_loop.label = cx->blk_loop.label;
9590 ncx->blk_loop.resetsp = cx->blk_loop.resetsp;
9591 ncx->blk_loop.redo_op = cx->blk_loop.redo_op;
9592 ncx->blk_loop.next_op = cx->blk_loop.next_op;
9593 ncx->blk_loop.last_op = cx->blk_loop.last_op;
9594 ncx->blk_loop.iterdata = (CxPADLOOP(cx)
9595 ? cx->blk_loop.iterdata
d2d73c3e 9596 : gv_dup((GV*)cx->blk_loop.iterdata, param));
f3548bdc
DM
9597 ncx->blk_loop.oldcomppad
9598 = (PAD*)ptr_table_fetch(PL_ptr_table,
9599 cx->blk_loop.oldcomppad);
d2d73c3e
AB
9600 ncx->blk_loop.itersave = sv_dup_inc(cx->blk_loop.itersave, param);
9601 ncx->blk_loop.iterlval = sv_dup_inc(cx->blk_loop.iterlval, param);
9602 ncx->blk_loop.iterary = av_dup_inc(cx->blk_loop.iterary, param);
1d7c1841
GS
9603 ncx->blk_loop.iterix = cx->blk_loop.iterix;
9604 ncx->blk_loop.itermax = cx->blk_loop.itermax;
9605 break;
9606 case CXt_FORMAT:
d2d73c3e
AB
9607 ncx->blk_sub.cv = cv_dup(cx->blk_sub.cv, param);
9608 ncx->blk_sub.gv = gv_dup(cx->blk_sub.gv, param);
9609 ncx->blk_sub.dfoutgv = gv_dup_inc(cx->blk_sub.dfoutgv, param);
1d7c1841 9610 ncx->blk_sub.hasargs = cx->blk_sub.hasargs;
f39bc417 9611 ncx->blk_sub.retop = cx->blk_sub.retop;
1d7c1841
GS
9612 break;
9613 case CXt_BLOCK:
9614 case CXt_NULL:
9615 break;
9616 }
9617 }
9618 --ix;
9619 }
9620 return ncxs;
9621}
9622
645c22ef
DM
9623/* duplicate a stack info structure */
9624
1d7c1841 9625PERL_SI *
a8fc9800 9626Perl_si_dup(pTHX_ PERL_SI *si, CLONE_PARAMS* param)
1d7c1841
GS
9627{
9628 PERL_SI *nsi;
9629
9630 if (!si)
9631 return (PERL_SI*)NULL;
9632
9633 /* look for it in the table first */
9634 nsi = (PERL_SI*)ptr_table_fetch(PL_ptr_table, si);
9635 if (nsi)
9636 return nsi;
9637
9638 /* create anew and remember what it is */
a02a5408 9639 Newxz(nsi, 1, PERL_SI);
1d7c1841
GS
9640 ptr_table_store(PL_ptr_table, si, nsi);
9641
d2d73c3e 9642 nsi->si_stack = av_dup_inc(si->si_stack, param);
1d7c1841
GS
9643 nsi->si_cxix = si->si_cxix;
9644 nsi->si_cxmax = si->si_cxmax;
d2d73c3e 9645 nsi->si_cxstack = cx_dup(si->si_cxstack, si->si_cxix, si->si_cxmax, param);
1d7c1841 9646 nsi->si_type = si->si_type;
d2d73c3e
AB
9647 nsi->si_prev = si_dup(si->si_prev, param);
9648 nsi->si_next = si_dup(si->si_next, param);
1d7c1841
GS
9649 nsi->si_markoff = si->si_markoff;
9650
9651 return nsi;
9652}
9653
9654#define POPINT(ss,ix) ((ss)[--(ix)].any_i32)
9655#define TOPINT(ss,ix) ((ss)[ix].any_i32)
9656#define POPLONG(ss,ix) ((ss)[--(ix)].any_long)
9657#define TOPLONG(ss,ix) ((ss)[ix].any_long)
9658#define POPIV(ss,ix) ((ss)[--(ix)].any_iv)
9659#define TOPIV(ss,ix) ((ss)[ix].any_iv)
38d8b13e
HS
9660#define POPBOOL(ss,ix) ((ss)[--(ix)].any_bool)
9661#define TOPBOOL(ss,ix) ((ss)[ix].any_bool)
1d7c1841
GS
9662#define POPPTR(ss,ix) ((ss)[--(ix)].any_ptr)
9663#define TOPPTR(ss,ix) ((ss)[ix].any_ptr)
9664#define POPDPTR(ss,ix) ((ss)[--(ix)].any_dptr)
9665#define TOPDPTR(ss,ix) ((ss)[ix].any_dptr)
9666#define POPDXPTR(ss,ix) ((ss)[--(ix)].any_dxptr)
9667#define TOPDXPTR(ss,ix) ((ss)[ix].any_dxptr)
9668
9669/* XXXXX todo */
9670#define pv_dup_inc(p) SAVEPV(p)
9671#define pv_dup(p) SAVEPV(p)
9672#define svp_dup_inc(p,pp) any_dup(p,pp)
9673
645c22ef
DM
9674/* map any object to the new equivent - either something in the
9675 * ptr table, or something in the interpreter structure
9676 */
9677
1d7c1841 9678void *
53c1dcc0 9679Perl_any_dup(pTHX_ void *v, const PerlInterpreter *proto_perl)
1d7c1841
GS
9680{
9681 void *ret;
9682
9683 if (!v)
9684 return (void*)NULL;
9685
9686 /* look for it in the table first */
9687 ret = ptr_table_fetch(PL_ptr_table, v);
9688 if (ret)
9689 return ret;
9690
9691 /* see if it is part of the interpreter structure */
9692 if (v >= (void*)proto_perl && v < (void*)(proto_perl+1))
acfe0abc 9693 ret = (void*)(((char*)aTHX) + (((char*)v) - (char*)proto_perl));
05ec9bb3 9694 else {
1d7c1841 9695 ret = v;
05ec9bb3 9696 }
1d7c1841
GS
9697
9698 return ret;
9699}
9700
645c22ef
DM
9701/* duplicate the save stack */
9702
1d7c1841 9703ANY *
a8fc9800 9704Perl_ss_dup(pTHX_ PerlInterpreter *proto_perl, CLONE_PARAMS* param)
1d7c1841 9705{
53c1dcc0
AL
9706 ANY * const ss = proto_perl->Tsavestack;
9707 const I32 max = proto_perl->Tsavestack_max;
9708 I32 ix = proto_perl->Tsavestack_ix;
1d7c1841
GS
9709 ANY *nss;
9710 SV *sv;
9711 GV *gv;
9712 AV *av;
9713 HV *hv;
9714 void* ptr;
9715 int intval;
9716 long longval;
9717 GP *gp;
9718 IV iv;
c4e33207 9719 char *c = NULL;
1d7c1841 9720 void (*dptr) (void*);
acfe0abc 9721 void (*dxptr) (pTHX_ void*);
1d7c1841 9722
a02a5408 9723 Newxz(nss, max, ANY);
1d7c1841
GS
9724
9725 while (ix > 0) {
b464bac0 9726 I32 i = POPINT(ss,ix);
1d7c1841
GS
9727 TOPINT(nss,ix) = i;
9728 switch (i) {
9729 case SAVEt_ITEM: /* normal string */
9730 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9731 TOPPTR(nss,ix) = sv_dup_inc(sv, param);
1d7c1841 9732 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9733 TOPPTR(nss,ix) = sv_dup_inc(sv, param);
1d7c1841
GS
9734 break;
9735 case SAVEt_SV: /* scalar reference */
9736 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9737 TOPPTR(nss,ix) = sv_dup_inc(sv, param);
1d7c1841 9738 gv = (GV*)POPPTR(ss,ix);
d2d73c3e 9739 TOPPTR(nss,ix) = gv_dup_inc(gv, param);
1d7c1841 9740 break;
f4dd75d9
GS
9741 case SAVEt_GENERIC_PVREF: /* generic char* */
9742 c = (char*)POPPTR(ss,ix);
9743 TOPPTR(nss,ix) = pv_dup(c);
9744 ptr = POPPTR(ss,ix);
9745 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9746 break;
05ec9bb3
NIS
9747 case SAVEt_SHARED_PVREF: /* char* in shared space */
9748 c = (char*)POPPTR(ss,ix);
9749 TOPPTR(nss,ix) = savesharedpv(c);
9750 ptr = POPPTR(ss,ix);
9751 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9752 break;
1d7c1841
GS
9753 case SAVEt_GENERIC_SVREF: /* generic sv */
9754 case SAVEt_SVREF: /* scalar reference */
9755 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9756 TOPPTR(nss,ix) = sv_dup_inc(sv, param);
1d7c1841
GS
9757 ptr = POPPTR(ss,ix);
9758 TOPPTR(nss,ix) = svp_dup_inc((SV**)ptr, proto_perl);/* XXXXX */
9759 break;
9760 case SAVEt_AV: /* array reference */
9761 av = (AV*)POPPTR(ss,ix);
d2d73c3e 9762 TOPPTR(nss,ix) = av_dup_inc(av, param);
1d7c1841 9763 gv = (GV*)POPPTR(ss,ix);
d2d73c3e 9764 TOPPTR(nss,ix) = gv_dup(gv, param);
1d7c1841
GS
9765 break;
9766 case SAVEt_HV: /* hash reference */
9767 hv = (HV*)POPPTR(ss,ix);
d2d73c3e 9768 TOPPTR(nss,ix) = hv_dup_inc(hv, param);
1d7c1841 9769 gv = (GV*)POPPTR(ss,ix);
d2d73c3e 9770 TOPPTR(nss,ix) = gv_dup(gv, param);
1d7c1841
GS
9771 break;
9772 case SAVEt_INT: /* int reference */
9773 ptr = POPPTR(ss,ix);
9774 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9775 intval = (int)POPINT(ss,ix);
9776 TOPINT(nss,ix) = intval;
9777 break;
9778 case SAVEt_LONG: /* long reference */
9779 ptr = POPPTR(ss,ix);
9780 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9781 longval = (long)POPLONG(ss,ix);
9782 TOPLONG(nss,ix) = longval;
9783 break;
9784 case SAVEt_I32: /* I32 reference */
9785 case SAVEt_I16: /* I16 reference */
9786 case SAVEt_I8: /* I8 reference */
9787 ptr = POPPTR(ss,ix);
9788 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9789 i = POPINT(ss,ix);
9790 TOPINT(nss,ix) = i;
9791 break;
9792 case SAVEt_IV: /* IV reference */
9793 ptr = POPPTR(ss,ix);
9794 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9795 iv = POPIV(ss,ix);
9796 TOPIV(nss,ix) = iv;
9797 break;
9798 case SAVEt_SPTR: /* SV* reference */
9799 ptr = POPPTR(ss,ix);
9800 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9801 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9802 TOPPTR(nss,ix) = sv_dup(sv, param);
1d7c1841
GS
9803 break;
9804 case SAVEt_VPTR: /* random* reference */
9805 ptr = POPPTR(ss,ix);
9806 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9807 ptr = POPPTR(ss,ix);
9808 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9809 break;
9810 case SAVEt_PPTR: /* char* reference */
9811 ptr = POPPTR(ss,ix);
9812 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9813 c = (char*)POPPTR(ss,ix);
9814 TOPPTR(nss,ix) = pv_dup(c);
9815 break;
9816 case SAVEt_HPTR: /* HV* reference */
9817 ptr = POPPTR(ss,ix);
9818 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9819 hv = (HV*)POPPTR(ss,ix);
d2d73c3e 9820 TOPPTR(nss,ix) = hv_dup(hv, param);
1d7c1841
GS
9821 break;
9822 case SAVEt_APTR: /* AV* reference */
9823 ptr = POPPTR(ss,ix);
9824 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9825 av = (AV*)POPPTR(ss,ix);
d2d73c3e 9826 TOPPTR(nss,ix) = av_dup(av, param);
1d7c1841
GS
9827 break;
9828 case SAVEt_NSTAB:
9829 gv = (GV*)POPPTR(ss,ix);
d2d73c3e 9830 TOPPTR(nss,ix) = gv_dup(gv, param);
1d7c1841
GS
9831 break;
9832 case SAVEt_GP: /* scalar reference */
9833 gp = (GP*)POPPTR(ss,ix);
d2d73c3e 9834 TOPPTR(nss,ix) = gp = gp_dup(gp, param);
1d7c1841
GS
9835 (void)GpREFCNT_inc(gp);
9836 gv = (GV*)POPPTR(ss,ix);
2ed3c8fc 9837 TOPPTR(nss,ix) = gv_dup_inc(gv, param);
1d7c1841
GS
9838 c = (char*)POPPTR(ss,ix);
9839 TOPPTR(nss,ix) = pv_dup(c);
9840 iv = POPIV(ss,ix);
9841 TOPIV(nss,ix) = iv;
9842 iv = POPIV(ss,ix);
9843 TOPIV(nss,ix) = iv;
9844 break;
9845 case SAVEt_FREESV:
26d9b02f 9846 case SAVEt_MORTALIZESV:
1d7c1841 9847 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9848 TOPPTR(nss,ix) = sv_dup_inc(sv, param);
1d7c1841
GS
9849 break;
9850 case SAVEt_FREEOP:
9851 ptr = POPPTR(ss,ix);
9852 if (ptr && (((OP*)ptr)->op_private & OPpREFCOUNTED)) {
9853 /* these are assumed to be refcounted properly */
53c1dcc0 9854 OP *o;
1d7c1841
GS
9855 switch (((OP*)ptr)->op_type) {
9856 case OP_LEAVESUB:
9857 case OP_LEAVESUBLV:
9858 case OP_LEAVEEVAL:
9859 case OP_LEAVE:
9860 case OP_SCOPE:
9861 case OP_LEAVEWRITE:
e977893f
GS
9862 TOPPTR(nss,ix) = ptr;
9863 o = (OP*)ptr;
9864 OpREFCNT_inc(o);
1d7c1841
GS
9865 break;
9866 default:
9867 TOPPTR(nss,ix) = Nullop;
9868 break;
9869 }
9870 }
9871 else
9872 TOPPTR(nss,ix) = Nullop;
9873 break;
9874 case SAVEt_FREEPV:
9875 c = (char*)POPPTR(ss,ix);
9876 TOPPTR(nss,ix) = pv_dup_inc(c);
9877 break;
9878 case SAVEt_CLEARSV:
9879 longval = POPLONG(ss,ix);
9880 TOPLONG(nss,ix) = longval;
9881 break;
9882 case SAVEt_DELETE:
9883 hv = (HV*)POPPTR(ss,ix);
d2d73c3e 9884 TOPPTR(nss,ix) = hv_dup_inc(hv, param);
1d7c1841
GS
9885 c = (char*)POPPTR(ss,ix);
9886 TOPPTR(nss,ix) = pv_dup_inc(c);
9887 i = POPINT(ss,ix);
9888 TOPINT(nss,ix) = i;
9889 break;
9890 case SAVEt_DESTRUCTOR:
9891 ptr = POPPTR(ss,ix);
9892 TOPPTR(nss,ix) = any_dup(ptr, proto_perl); /* XXX quite arbitrary */
9893 dptr = POPDPTR(ss,ix);
8141890a
JH
9894 TOPDPTR(nss,ix) = DPTR2FPTR(void (*)(void*),
9895 any_dup(FPTR2DPTR(void *, dptr),
9896 proto_perl));
1d7c1841
GS
9897 break;
9898 case SAVEt_DESTRUCTOR_X:
9899 ptr = POPPTR(ss,ix);
9900 TOPPTR(nss,ix) = any_dup(ptr, proto_perl); /* XXX quite arbitrary */
9901 dxptr = POPDXPTR(ss,ix);
8141890a
JH
9902 TOPDXPTR(nss,ix) = DPTR2FPTR(void (*)(pTHX_ void*),
9903 any_dup(FPTR2DPTR(void *, dxptr),
9904 proto_perl));
1d7c1841
GS
9905 break;
9906 case SAVEt_REGCONTEXT:
9907 case SAVEt_ALLOC:
9908 i = POPINT(ss,ix);
9909 TOPINT(nss,ix) = i;
9910 ix -= i;
9911 break;
9912 case SAVEt_STACK_POS: /* Position on Perl stack */
9913 i = POPINT(ss,ix);
9914 TOPINT(nss,ix) = i;
9915 break;
9916 case SAVEt_AELEM: /* array element */
9917 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9918 TOPPTR(nss,ix) = sv_dup_inc(sv, param);
1d7c1841
GS
9919 i = POPINT(ss,ix);
9920 TOPINT(nss,ix) = i;
9921 av = (AV*)POPPTR(ss,ix);
d2d73c3e 9922 TOPPTR(nss,ix) = av_dup_inc(av, param);
1d7c1841
GS
9923 break;
9924 case SAVEt_HELEM: /* hash element */
9925 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9926 TOPPTR(nss,ix) = sv_dup_inc(sv, param);
1d7c1841 9927 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9928 TOPPTR(nss,ix) = sv_dup_inc(sv, param);
1d7c1841 9929 hv = (HV*)POPPTR(ss,ix);
d2d73c3e 9930 TOPPTR(nss,ix) = hv_dup_inc(hv, param);
1d7c1841
GS
9931 break;
9932 case SAVEt_OP:
9933 ptr = POPPTR(ss,ix);
9934 TOPPTR(nss,ix) = ptr;
9935 break;
9936 case SAVEt_HINTS:
9937 i = POPINT(ss,ix);
9938 TOPINT(nss,ix) = i;
9939 break;
c4410b1b
GS
9940 case SAVEt_COMPPAD:
9941 av = (AV*)POPPTR(ss,ix);
58ed4fbe 9942 TOPPTR(nss,ix) = av_dup(av, param);
c4410b1b 9943 break;
c3564e5c
GS
9944 case SAVEt_PADSV:
9945 longval = (long)POPLONG(ss,ix);
9946 TOPLONG(nss,ix) = longval;
9947 ptr = POPPTR(ss,ix);
9948 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
9949 sv = (SV*)POPPTR(ss,ix);
d2d73c3e 9950 TOPPTR(nss,ix) = sv_dup(sv, param);
c3564e5c 9951 break;
a1bb4754 9952 case SAVEt_BOOL:
38d8b13e 9953 ptr = POPPTR(ss,ix);
b9609c01 9954 TOPPTR(nss,ix) = any_dup(ptr, proto_perl);
38d8b13e 9955 longval = (long)POPBOOL(ss,ix);
b9609c01 9956 TOPBOOL(nss,ix) = (bool)longval;
a1bb4754 9957 break;
8bd2680e
MHM
9958 case SAVEt_SET_SVFLAGS:
9959 i = POPINT(ss,ix);
9960 TOPINT(nss,ix) = i;
9961 i = POPINT(ss,ix);
9962 TOPINT(nss,ix) = i;
9963 sv = (SV*)POPPTR(ss,ix);
9964 TOPPTR(nss,ix) = sv_dup(sv, param);
9965 break;
1d7c1841
GS
9966 default:
9967 Perl_croak(aTHX_ "panic: ss_dup inconsistency");
9968 }
9969 }
9970
bd81e77b
NC
9971 return nss;
9972}
9973
9974
9975/* if sv is a stash, call $class->CLONE_SKIP(), and set the SVphv_CLONEABLE
9976 * flag to the result. This is done for each stash before cloning starts,
9977 * so we know which stashes want their objects cloned */
9978
9979static void
9980do_mark_cloneable_stash(pTHX_ SV *sv)
9981{
9982 const HEK * const hvname = HvNAME_HEK((HV*)sv);
9983 if (hvname) {
9984 GV* const cloner = gv_fetchmethod_autoload((HV*)sv, "CLONE_SKIP", 0);
9985 SvFLAGS(sv) |= SVphv_CLONEABLE; /* clone objects by default */
9986 if (cloner && GvCV(cloner)) {
9987 dSP;
9988 UV status;
9989
9990 ENTER;
9991 SAVETMPS;
9992 PUSHMARK(SP);
9993 XPUSHs(sv_2mortal(newSVhek(hvname)));
9994 PUTBACK;
9995 call_sv((SV*)GvCV(cloner), G_SCALAR);
9996 SPAGAIN;
9997 status = POPu;
9998 PUTBACK;
9999 FREETMPS;
10000 LEAVE;
10001 if (status)
10002 SvFLAGS(sv) &= ~SVphv_CLONEABLE;
10003 }
10004 }
10005}
10006
10007
10008
10009/*
10010=for apidoc perl_clone
10011
10012Create and return a new interpreter by cloning the current one.
10013
10014perl_clone takes these flags as parameters:
10015
10016CLONEf_COPY_STACKS - is used to, well, copy the stacks also,
10017without it we only clone the data and zero the stacks,
10018with it we copy the stacks and the new perl interpreter is
10019ready to run at the exact same point as the previous one.
10020The pseudo-fork code uses COPY_STACKS while the
10021threads->new doesn't.
10022
10023CLONEf_KEEP_PTR_TABLE
10024perl_clone keeps a ptr_table with the pointer of the old
10025variable as a key and the new variable as a value,
10026this allows it to check if something has been cloned and not
10027clone it again but rather just use the value and increase the
10028refcount. If KEEP_PTR_TABLE is not set then perl_clone will kill
10029the ptr_table using the function
10030C<ptr_table_free(PL_ptr_table); PL_ptr_table = NULL;>,
10031reason to keep it around is if you want to dup some of your own
10032variable who are outside the graph perl scans, example of this
10033code is in threads.xs create
10034
10035CLONEf_CLONE_HOST
10036This is a win32 thing, it is ignored on unix, it tells perls
10037win32host code (which is c++) to clone itself, this is needed on
10038win32 if you want to run two threads at the same time,
10039if you just want to do some stuff in a separate perl interpreter
10040and then throw it away and return to the original one,
10041you don't need to do anything.
10042
10043=cut
10044*/
10045
10046/* XXX the above needs expanding by someone who actually understands it ! */
10047EXTERN_C PerlInterpreter *
10048perl_clone_host(PerlInterpreter* proto_perl, UV flags);
10049
10050PerlInterpreter *
10051perl_clone(PerlInterpreter *proto_perl, UV flags)
10052{
10053 dVAR;
10054#ifdef PERL_IMPLICIT_SYS
10055
10056 /* perlhost.h so we need to call into it
10057 to clone the host, CPerlHost should have a c interface, sky */
10058
10059 if (flags & CLONEf_CLONE_HOST) {
10060 return perl_clone_host(proto_perl,flags);
10061 }
10062 return perl_clone_using(proto_perl, flags,
10063 proto_perl->IMem,
10064 proto_perl->IMemShared,
10065 proto_perl->IMemParse,
10066 proto_perl->IEnv,
10067 proto_perl->IStdIO,
10068 proto_perl->ILIO,
10069 proto_perl->IDir,
10070 proto_perl->ISock,
10071 proto_perl->IProc);
10072}
10073
10074PerlInterpreter *
10075perl_clone_using(PerlInterpreter *proto_perl, UV flags,
10076 struct IPerlMem* ipM, struct IPerlMem* ipMS,
10077 struct IPerlMem* ipMP, struct IPerlEnv* ipE,
10078 struct IPerlStdIO* ipStd, struct IPerlLIO* ipLIO,
10079 struct IPerlDir* ipD, struct IPerlSock* ipS,
10080 struct IPerlProc* ipP)
10081{
10082 /* XXX many of the string copies here can be optimized if they're
10083 * constants; they need to be allocated as common memory and just
10084 * their pointers copied. */
10085
10086 IV i;
10087 CLONE_PARAMS clone_params;
10088 CLONE_PARAMS* param = &clone_params;
10089
10090 PerlInterpreter *my_perl = (PerlInterpreter*)(*ipM->pMalloc)(ipM, sizeof(PerlInterpreter));
10091 /* for each stash, determine whether its objects should be cloned */
10092 S_visit(proto_perl, do_mark_cloneable_stash, SVt_PVHV, SVTYPEMASK);
10093 PERL_SET_THX(my_perl);
10094
10095# ifdef DEBUGGING
10096 Poison(my_perl, 1, PerlInterpreter);
10097 PL_op = Nullop;
10098 PL_curcop = (COP *)Nullop;
10099 PL_markstack = 0;
10100 PL_scopestack = 0;
10101 PL_savestack = 0;
10102 PL_savestack_ix = 0;
10103 PL_savestack_max = -1;
10104 PL_sig_pending = 0;
10105 Zero(&PL_debug_pad, 1, struct perl_debug_pad);
10106# else /* !DEBUGGING */
10107 Zero(my_perl, 1, PerlInterpreter);
10108# endif /* DEBUGGING */
10109
10110 /* host pointers */
10111 PL_Mem = ipM;
10112 PL_MemShared = ipMS;
10113 PL_MemParse = ipMP;
10114 PL_Env = ipE;
10115 PL_StdIO = ipStd;
10116 PL_LIO = ipLIO;
10117 PL_Dir = ipD;
10118 PL_Sock = ipS;
10119 PL_Proc = ipP;
10120#else /* !PERL_IMPLICIT_SYS */
10121 IV i;
10122 CLONE_PARAMS clone_params;
10123 CLONE_PARAMS* param = &clone_params;
10124 PerlInterpreter *my_perl = (PerlInterpreter*)PerlMem_malloc(sizeof(PerlInterpreter));
10125 /* for each stash, determine whether its objects should be cloned */
10126 S_visit(proto_perl, do_mark_cloneable_stash, SVt_PVHV, SVTYPEMASK);
10127 PERL_SET_THX(my_perl);
10128
10129# ifdef DEBUGGING
10130 Poison(my_perl, 1, PerlInterpreter);
10131 PL_op = Nullop;
10132 PL_curcop = (COP *)Nullop;
10133 PL_markstack = 0;
10134 PL_scopestack = 0;
10135 PL_savestack = 0;
10136 PL_savestack_ix = 0;
10137 PL_savestack_max = -1;
10138 PL_sig_pending = 0;
10139 Zero(&PL_debug_pad, 1, struct perl_debug_pad);
10140# else /* !DEBUGGING */
10141 Zero(my_perl, 1, PerlInterpreter);
10142# endif /* DEBUGGING */
10143#endif /* PERL_IMPLICIT_SYS */
10144 param->flags = flags;
10145 param->proto_perl = proto_perl;
10146
10147 Zero(&PL_body_arenaroots, 1, PL_body_arenaroots);
10148 Zero(&PL_body_roots, 1, PL_body_roots);
10149
10150 PL_nice_chunk = NULL;
10151 PL_nice_chunk_size = 0;
10152 PL_sv_count = 0;
10153 PL_sv_objcount = 0;
10154 PL_sv_root = Nullsv;
10155 PL_sv_arenaroot = Nullsv;
10156
10157 PL_debug = proto_perl->Idebug;
10158
10159 PL_hash_seed = proto_perl->Ihash_seed;
10160 PL_rehash_seed = proto_perl->Irehash_seed;
10161
10162#ifdef USE_REENTRANT_API
10163 /* XXX: things like -Dm will segfault here in perlio, but doing
10164 * PERL_SET_CONTEXT(proto_perl);
10165 * breaks too many other things
10166 */
10167 Perl_reentrant_init(aTHX);
10168#endif
10169
10170 /* create SV map for pointer relocation */
10171 PL_ptr_table = ptr_table_new();
10172
10173 /* initialize these special pointers as early as possible */
10174 SvANY(&PL_sv_undef) = NULL;
10175 SvREFCNT(&PL_sv_undef) = (~(U32)0)/2;
10176 SvFLAGS(&PL_sv_undef) = SVf_READONLY|SVt_NULL;
10177 ptr_table_store(PL_ptr_table, &proto_perl->Isv_undef, &PL_sv_undef);
10178
10179 SvANY(&PL_sv_no) = new_XPVNV();
10180 SvREFCNT(&PL_sv_no) = (~(U32)0)/2;
10181 SvFLAGS(&PL_sv_no) = SVp_IOK|SVf_IOK|SVp_NOK|SVf_NOK
10182 |SVp_POK|SVf_POK|SVf_READONLY|SVt_PVNV;
10183 SvPV_set(&PL_sv_no, SAVEPVN(PL_No, 0));
10184 SvCUR_set(&PL_sv_no, 0);
10185 SvLEN_set(&PL_sv_no, 1);
10186 SvIV_set(&PL_sv_no, 0);
10187 SvNV_set(&PL_sv_no, 0);
10188 ptr_table_store(PL_ptr_table, &proto_perl->Isv_no, &PL_sv_no);
10189
10190 SvANY(&PL_sv_yes) = new_XPVNV();
10191 SvREFCNT(&PL_sv_yes) = (~(U32)0)/2;
10192 SvFLAGS(&PL_sv_yes) = SVp_IOK|SVf_IOK|SVp_NOK|SVf_NOK
10193 |SVp_POK|SVf_POK|SVf_READONLY|SVt_PVNV;
10194 SvPV_set(&PL_sv_yes, SAVEPVN(PL_Yes, 1));
10195 SvCUR_set(&PL_sv_yes, 1);
10196 SvLEN_set(&PL_sv_yes, 2);
10197 SvIV_set(&PL_sv_yes, 1);
10198 SvNV_set(&PL_sv_yes, 1);
10199 ptr_table_store(PL_ptr_table, &proto_perl->Isv_yes, &PL_sv_yes);
10200
10201 /* create (a non-shared!) shared string table */
10202 PL_strtab = newHV();
10203 HvSHAREKEYS_off(PL_strtab);
10204 hv_ksplit(PL_strtab, HvTOTALKEYS(proto_perl->Istrtab));
10205 ptr_table_store(PL_ptr_table, proto_perl->Istrtab, PL_strtab);
10206
10207 PL_compiling = proto_perl->Icompiling;
10208
10209 /* These two PVs will be free'd special way so must set them same way op.c does */
10210 PL_compiling.cop_stashpv = savesharedpv(PL_compiling.cop_stashpv);
10211 ptr_table_store(PL_ptr_table, proto_perl->Icompiling.cop_stashpv, PL_compiling.cop_stashpv);
10212
10213 PL_compiling.cop_file = savesharedpv(PL_compiling.cop_file);
10214 ptr_table_store(PL_ptr_table, proto_perl->Icompiling.cop_file, PL_compiling.cop_file);
10215
10216 ptr_table_store(PL_ptr_table, &proto_perl->Icompiling, &PL_compiling);
10217 if (!specialWARN(PL_compiling.cop_warnings))
10218 PL_compiling.cop_warnings = sv_dup_inc(PL_compiling.cop_warnings, param);
10219 if (!specialCopIO(PL_compiling.cop_io))
10220 PL_compiling.cop_io = sv_dup_inc(PL_compiling.cop_io, param);
10221 PL_curcop = (COP*)any_dup(proto_perl->Tcurcop, proto_perl);
10222
10223 /* pseudo environmental stuff */
10224 PL_origargc = proto_perl->Iorigargc;
10225 PL_origargv = proto_perl->Iorigargv;
10226
10227 param->stashes = newAV(); /* Setup array of objects to call clone on */
10228
10229 /* Set tainting stuff before PerlIO_debug can possibly get called */
10230 PL_tainting = proto_perl->Itainting;
10231 PL_taint_warn = proto_perl->Itaint_warn;
10232
10233#ifdef PERLIO_LAYERS
10234 /* Clone PerlIO tables as soon as we can handle general xx_dup() */
10235 PerlIO_clone(aTHX_ proto_perl, param);
10236#endif
10237
10238 PL_envgv = gv_dup(proto_perl->Ienvgv, param);
10239 PL_incgv = gv_dup(proto_perl->Iincgv, param);
10240 PL_hintgv = gv_dup(proto_perl->Ihintgv, param);
10241 PL_origfilename = SAVEPV(proto_perl->Iorigfilename);
10242 PL_diehook = sv_dup_inc(proto_perl->Idiehook, param);
10243 PL_warnhook = sv_dup_inc(proto_perl->Iwarnhook, param);
10244
10245 /* switches */
10246 PL_minus_c = proto_perl->Iminus_c;
10247 PL_patchlevel = sv_dup_inc(proto_perl->Ipatchlevel, param);
10248 PL_localpatches = proto_perl->Ilocalpatches;
10249 PL_splitstr = proto_perl->Isplitstr;
10250 PL_preprocess = proto_perl->Ipreprocess;
10251 PL_minus_n = proto_perl->Iminus_n;
10252 PL_minus_p = proto_perl->Iminus_p;
10253 PL_minus_l = proto_perl->Iminus_l;
10254 PL_minus_a = proto_perl->Iminus_a;
bc9b29db 10255 PL_minus_E = proto_perl->Iminus_E;
bd81e77b
NC
10256 PL_minus_F = proto_perl->Iminus_F;
10257 PL_doswitches = proto_perl->Idoswitches;
10258 PL_dowarn = proto_perl->Idowarn;
10259 PL_doextract = proto_perl->Idoextract;
10260 PL_sawampersand = proto_perl->Isawampersand;
10261 PL_unsafe = proto_perl->Iunsafe;
10262 PL_inplace = SAVEPV(proto_perl->Iinplace);
10263 PL_e_script = sv_dup_inc(proto_perl->Ie_script, param);
10264 PL_perldb = proto_perl->Iperldb;
10265 PL_perl_destruct_level = proto_perl->Iperl_destruct_level;
10266 PL_exit_flags = proto_perl->Iexit_flags;
10267
10268 /* magical thingies */
10269 /* XXX time(&PL_basetime) when asked for? */
10270 PL_basetime = proto_perl->Ibasetime;
10271 PL_formfeed = sv_dup(proto_perl->Iformfeed, param);
10272
10273 PL_maxsysfd = proto_perl->Imaxsysfd;
10274 PL_multiline = proto_perl->Imultiline;
10275 PL_statusvalue = proto_perl->Istatusvalue;
10276#ifdef VMS
10277 PL_statusvalue_vms = proto_perl->Istatusvalue_vms;
10278#else
10279 PL_statusvalue_posix = proto_perl->Istatusvalue_posix;
10280#endif
10281 PL_encoding = sv_dup(proto_perl->Iencoding, param);
10282
10283 sv_setpvn(PERL_DEBUG_PAD(0), "", 0); /* For regex debugging. */
10284 sv_setpvn(PERL_DEBUG_PAD(1), "", 0); /* ext/re needs these */
10285 sv_setpvn(PERL_DEBUG_PAD(2), "", 0); /* even without DEBUGGING. */
10286
10287 /* Clone the regex array */
10288 PL_regex_padav = newAV();
10289 {
10290 const I32 len = av_len((AV*)proto_perl->Iregex_padav);
10291 SV** const regexen = AvARRAY((AV*)proto_perl->Iregex_padav);
10292 IV i;
10293 av_push(PL_regex_padav,
10294 sv_dup_inc(regexen[0],param));
10295 for(i = 1; i <= len; i++) {
10296 if(SvREPADTMP(regexen[i])) {
10297 av_push(PL_regex_padav, sv_dup_inc(regexen[i], param));
10298 } else {
10299 av_push(PL_regex_padav,
10300 SvREFCNT_inc(
10301 newSViv(PTR2IV(re_dup(INT2PTR(REGEXP *,
10302 SvIVX(regexen[i])), param)))
10303 ));
10304 }
10305 }
10306 }
10307 PL_regex_pad = AvARRAY(PL_regex_padav);
10308
10309 /* shortcuts to various I/O objects */
10310 PL_stdingv = gv_dup(proto_perl->Istdingv, param);
10311 PL_stderrgv = gv_dup(proto_perl->Istderrgv, param);
10312 PL_defgv = gv_dup(proto_perl->Idefgv, param);
10313 PL_argvgv = gv_dup(proto_perl->Iargvgv, param);
10314 PL_argvoutgv = gv_dup(proto_perl->Iargvoutgv, param);
10315 PL_argvout_stack = av_dup_inc(proto_perl->Iargvout_stack, param);
1d7c1841 10316
bd81e77b
NC
10317 /* shortcuts to regexp stuff */
10318 PL_replgv = gv_dup(proto_perl->Ireplgv, param);
9660f481 10319
bd81e77b
NC
10320 /* shortcuts to misc objects */
10321 PL_errgv = gv_dup(proto_perl->Ierrgv, param);
9660f481 10322
bd81e77b
NC
10323 /* shortcuts to debugging objects */
10324 PL_DBgv = gv_dup(proto_perl->IDBgv, param);
10325 PL_DBline = gv_dup(proto_perl->IDBline, param);
10326 PL_DBsub = gv_dup(proto_perl->IDBsub, param);
10327 PL_DBsingle = sv_dup(proto_perl->IDBsingle, param);
10328 PL_DBtrace = sv_dup(proto_perl->IDBtrace, param);
10329 PL_DBsignal = sv_dup(proto_perl->IDBsignal, param);
10330 PL_DBassertion = sv_dup(proto_perl->IDBassertion, param);
10331 PL_lineary = av_dup(proto_perl->Ilineary, param);
10332 PL_dbargs = av_dup(proto_perl->Idbargs, param);
9660f481 10333
bd81e77b
NC
10334 /* symbol tables */
10335 PL_defstash = hv_dup_inc(proto_perl->Tdefstash, param);
10336 PL_curstash = hv_dup(proto_perl->Tcurstash, param);
10337 PL_debstash = hv_dup(proto_perl->Idebstash, param);
10338 PL_globalstash = hv_dup(proto_perl->Iglobalstash, param);
10339 PL_curstname = sv_dup_inc(proto_perl->Icurstname, param);
10340
10341 PL_beginav = av_dup_inc(proto_perl->Ibeginav, param);
10342 PL_beginav_save = av_dup_inc(proto_perl->Ibeginav_save, param);
10343 PL_checkav_save = av_dup_inc(proto_perl->Icheckav_save, param);
10344 PL_endav = av_dup_inc(proto_perl->Iendav, param);
10345 PL_checkav = av_dup_inc(proto_perl->Icheckav, param);
10346 PL_initav = av_dup_inc(proto_perl->Iinitav, param);
10347
10348 PL_sub_generation = proto_perl->Isub_generation;
10349
10350 /* funky return mechanisms */
10351 PL_forkprocess = proto_perl->Iforkprocess;
10352
10353 /* subprocess state */
10354 PL_fdpid = av_dup_inc(proto_perl->Ifdpid, param);
10355
10356 /* internal state */
10357 PL_maxo = proto_perl->Imaxo;
10358 if (proto_perl->Iop_mask)
10359 PL_op_mask = SAVEPVN(proto_perl->Iop_mask, PL_maxo);
10360 else
10361 PL_op_mask = Nullch;
10362 /* PL_asserting = proto_perl->Iasserting; */
10363
10364 /* current interpreter roots */
10365 PL_main_cv = cv_dup_inc(proto_perl->Imain_cv, param);
10366 PL_main_root = OpREFCNT_inc(proto_perl->Imain_root);
10367 PL_main_start = proto_perl->Imain_start;
10368 PL_eval_root = proto_perl->Ieval_root;
10369 PL_eval_start = proto_perl->Ieval_start;
10370
10371 /* runtime control stuff */
10372 PL_curcopdb = (COP*)any_dup(proto_perl->Icurcopdb, proto_perl);
10373 PL_copline = proto_perl->Icopline;
10374
10375 PL_filemode = proto_perl->Ifilemode;
10376 PL_lastfd = proto_perl->Ilastfd;
10377 PL_oldname = proto_perl->Ioldname; /* XXX not quite right */
10378 PL_Argv = NULL;
10379 PL_Cmd = Nullch;
10380 PL_gensym = proto_perl->Igensym;
10381 PL_preambled = proto_perl->Ipreambled;
10382 PL_preambleav = av_dup_inc(proto_perl->Ipreambleav, param);
10383 PL_laststatval = proto_perl->Ilaststatval;
10384 PL_laststype = proto_perl->Ilaststype;
10385 PL_mess_sv = Nullsv;
10386
10387 PL_ors_sv = sv_dup_inc(proto_perl->Iors_sv, param);
10388
10389 /* interpreter atexit processing */
10390 PL_exitlistlen = proto_perl->Iexitlistlen;
10391 if (PL_exitlistlen) {
10392 Newx(PL_exitlist, PL_exitlistlen, PerlExitListEntry);
10393 Copy(proto_perl->Iexitlist, PL_exitlist, PL_exitlistlen, PerlExitListEntry);
9660f481 10394 }
bd81e77b
NC
10395 else
10396 PL_exitlist = (PerlExitListEntry*)NULL;
f16dd614
DM
10397
10398 PL_my_cxt_size = proto_perl->Imy_cxt_size;
10399 if (PL_my_cxt_size) {
10400 Newx(PL_my_cxt_list, PL_my_cxt_size, void *);
10401 Copy(proto_perl->Imy_cxt_list, PL_my_cxt_list, PL_my_cxt_size, void *);
10402 }
10403 else
10404 PL_my_cxt_list = (void**)NULL;
bd81e77b
NC
10405 PL_modglobal = hv_dup_inc(proto_perl->Imodglobal, param);
10406 PL_custom_op_names = hv_dup_inc(proto_perl->Icustom_op_names,param);
10407 PL_custom_op_descs = hv_dup_inc(proto_perl->Icustom_op_descs,param);
10408
10409 PL_profiledata = NULL;
10410 PL_rsfp = fp_dup(proto_perl->Irsfp, '<', param);
10411 /* PL_rsfp_filters entries have fake IoDIRP() */
10412 PL_rsfp_filters = av_dup_inc(proto_perl->Irsfp_filters, param);
9660f481 10413
bd81e77b 10414 PL_compcv = cv_dup(proto_perl->Icompcv, param);
9660f481 10415
bd81e77b 10416 PAD_CLONE_VARS(proto_perl, param);
9660f481 10417
bd81e77b
NC
10418#ifdef HAVE_INTERP_INTERN
10419 sys_intern_dup(&proto_perl->Isys_intern, &PL_sys_intern);
10420#endif
645c22ef 10421
bd81e77b
NC
10422 /* more statics moved here */
10423 PL_generation = proto_perl->Igeneration;
10424 PL_DBcv = cv_dup(proto_perl->IDBcv, param);
645c22ef 10425
bd81e77b
NC
10426 PL_in_clean_objs = proto_perl->Iin_clean_objs;
10427 PL_in_clean_all = proto_perl->Iin_clean_all;
6a78b4db 10428
bd81e77b
NC
10429 PL_uid = proto_perl->Iuid;
10430 PL_euid = proto_perl->Ieuid;
10431 PL_gid = proto_perl->Igid;
10432 PL_egid = proto_perl->Iegid;
10433 PL_nomemok = proto_perl->Inomemok;
10434 PL_an = proto_perl->Ian;
10435 PL_evalseq = proto_perl->Ievalseq;
10436 PL_origenviron = proto_perl->Iorigenviron; /* XXX not quite right */
10437 PL_origalen = proto_perl->Iorigalen;
10438#ifdef PERL_USES_PL_PIDSTATUS
10439 PL_pidstatus = newHV(); /* XXX flag for cloning? */
10440#endif
10441 PL_osname = SAVEPV(proto_perl->Iosname);
10442 PL_sighandlerp = proto_perl->Isighandlerp;
6a78b4db 10443
bd81e77b 10444 PL_runops = proto_perl->Irunops;
6a78b4db 10445
bd81e77b 10446 Copy(proto_perl->Itokenbuf, PL_tokenbuf, 256, char);
6a78b4db 10447
bd81e77b
NC
10448#ifdef CSH
10449 PL_cshlen = proto_perl->Icshlen;
10450 PL_cshname = proto_perl->Icshname; /* XXX never deallocated */
10451#endif
645c22ef 10452
bd81e77b
NC
10453 PL_lex_state = proto_perl->Ilex_state;
10454 PL_lex_defer = proto_perl->Ilex_defer;
10455 PL_lex_expect = proto_perl->Ilex_expect;
10456 PL_lex_formbrack = proto_perl->Ilex_formbrack;
10457 PL_lex_dojoin = proto_perl->Ilex_dojoin;
10458 PL_lex_starts = proto_perl->Ilex_starts;
10459 PL_lex_stuff = sv_dup_inc(proto_perl->Ilex_stuff, param);
10460 PL_lex_repl = sv_dup_inc(proto_perl->Ilex_repl, param);
10461 PL_lex_op = proto_perl->Ilex_op;
10462 PL_lex_inpat = proto_perl->Ilex_inpat;
10463 PL_lex_inwhat = proto_perl->Ilex_inwhat;
10464 PL_lex_brackets = proto_perl->Ilex_brackets;
10465 i = (PL_lex_brackets < 120 ? 120 : PL_lex_brackets);
10466 PL_lex_brackstack = SAVEPVN(proto_perl->Ilex_brackstack,i);
10467 PL_lex_casemods = proto_perl->Ilex_casemods;
10468 i = (PL_lex_casemods < 12 ? 12 : PL_lex_casemods);
10469 PL_lex_casestack = SAVEPVN(proto_perl->Ilex_casestack,i);
645c22ef 10470
bd81e77b
NC
10471 Copy(proto_perl->Inextval, PL_nextval, 5, YYSTYPE);
10472 Copy(proto_perl->Inexttype, PL_nexttype, 5, I32);
10473 PL_nexttoke = proto_perl->Inexttoke;
c43294b8 10474
bd81e77b
NC
10475 /* XXX This is probably masking the deeper issue of why
10476 * SvANY(proto_perl->Ilinestr) can be NULL at this point. For test case:
10477 * http://archive.develooper.com/perl5-porters%40perl.org/msg83298.html
10478 * (A little debugging with a watchpoint on it may help.)
10479 */
10480 if (SvANY(proto_perl->Ilinestr)) {
10481 PL_linestr = sv_dup_inc(proto_perl->Ilinestr, param);
10482 i = proto_perl->Ibufptr - SvPVX_const(proto_perl->Ilinestr);
10483 PL_bufptr = SvPVX(PL_linestr) + (i < 0 ? 0 : i);
10484 i = proto_perl->Ioldbufptr - SvPVX_const(proto_perl->Ilinestr);
10485 PL_oldbufptr = SvPVX(PL_linestr) + (i < 0 ? 0 : i);
10486 i = proto_perl->Ioldoldbufptr - SvPVX_const(proto_perl->Ilinestr);
10487 PL_oldoldbufptr = SvPVX(PL_linestr) + (i < 0 ? 0 : i);
10488 i = proto_perl->Ilinestart - SvPVX_const(proto_perl->Ilinestr);
10489 PL_linestart = SvPVX(PL_linestr) + (i < 0 ? 0 : i);
10490 }
10491 else {
10492 PL_linestr = NEWSV(65,79);
10493 sv_upgrade(PL_linestr,SVt_PVIV);
10494 sv_setpvn(PL_linestr,"",0);
10495 PL_bufptr = PL_oldbufptr = PL_oldoldbufptr = PL_linestart = SvPVX(PL_linestr);
10496 }
10497 PL_bufend = SvPVX(PL_linestr) + SvCUR(PL_linestr);
10498 PL_pending_ident = proto_perl->Ipending_ident;
10499 PL_sublex_info = proto_perl->Isublex_info; /* XXX not quite right */
10500
10501 PL_expect = proto_perl->Iexpect;
10502
10503 PL_multi_start = proto_perl->Imulti_start;
10504 PL_multi_end = proto_perl->Imulti_end;
10505 PL_multi_open = proto_perl->Imulti_open;
10506 PL_multi_close = proto_perl->Imulti_close;
10507
10508 PL_error_count = proto_perl->Ierror_count;
10509 PL_subline = proto_perl->Isubline;
10510 PL_subname = sv_dup_inc(proto_perl->Isubname, param);
c43294b8 10511
bd81e77b
NC
10512 /* XXX See comment on SvANY(proto_perl->Ilinestr) above */
10513 if (SvANY(proto_perl->Ilinestr)) {
10514 i = proto_perl->Ilast_uni - SvPVX_const(proto_perl->Ilinestr);
10515 PL_last_uni = SvPVX(PL_linestr) + (i < 0 ? 0 : i);
10516 i = proto_perl->Ilast_lop - SvPVX_const(proto_perl->Ilinestr);
10517 PL_last_lop = SvPVX(PL_linestr) + (i < 0 ? 0 : i);
10518 PL_last_lop_op = proto_perl->Ilast_lop_op;
10519 }
10520 else {
10521 PL_last_uni = SvPVX(PL_linestr);
10522 PL_last_lop = SvPVX(PL_linestr);
10523 PL_last_lop_op = 0;
10524 }
10525 PL_in_my = proto_perl->Iin_my;
10526 PL_in_my_stash = hv_dup(proto_perl->Iin_my_stash, param);
10527#ifdef FCRYPT
10528 PL_cryptseen = proto_perl->Icryptseen;
10529#endif
1d7c1841 10530
bd81e77b 10531 PL_hints = proto_perl->Ihints;
1d7c1841 10532
bd81e77b 10533 PL_amagic_generation = proto_perl->Iamagic_generation;
d2d73c3e 10534
bd81e77b
NC
10535#ifdef USE_LOCALE_COLLATE
10536 PL_collation_ix = proto_perl->Icollation_ix;
10537 PL_collation_name = SAVEPV(proto_perl->Icollation_name);
10538 PL_collation_standard = proto_perl->Icollation_standard;
10539 PL_collxfrm_base = proto_perl->Icollxfrm_base;
10540 PL_collxfrm_mult = proto_perl->Icollxfrm_mult;
10541#endif /* USE_LOCALE_COLLATE */
1d7c1841 10542
bd81e77b
NC
10543#ifdef USE_LOCALE_NUMERIC
10544 PL_numeric_name = SAVEPV(proto_perl->Inumeric_name);
10545 PL_numeric_standard = proto_perl->Inumeric_standard;
10546 PL_numeric_local = proto_perl->Inumeric_local;
10547 PL_numeric_radix_sv = sv_dup_inc(proto_perl->Inumeric_radix_sv, param);
10548#endif /* !USE_LOCALE_NUMERIC */
1d7c1841 10549
bd81e77b
NC
10550 /* utf8 character classes */
10551 PL_utf8_alnum = sv_dup_inc(proto_perl->Iutf8_alnum, param);
10552 PL_utf8_alnumc = sv_dup_inc(proto_perl->Iutf8_alnumc, param);
10553 PL_utf8_ascii = sv_dup_inc(proto_perl->Iutf8_ascii, param);
10554 PL_utf8_alpha = sv_dup_inc(proto_perl->Iutf8_alpha, param);
10555 PL_utf8_space = sv_dup_inc(proto_perl->Iutf8_space, param);
10556 PL_utf8_cntrl = sv_dup_inc(proto_perl->Iutf8_cntrl, param);
10557 PL_utf8_graph = sv_dup_inc(proto_perl->Iutf8_graph, param);
10558 PL_utf8_digit = sv_dup_inc(proto_perl->Iutf8_digit, param);
10559 PL_utf8_upper = sv_dup_inc(proto_perl->Iutf8_upper, param);
10560 PL_utf8_lower = sv_dup_inc(proto_perl->Iutf8_lower, param);
10561 PL_utf8_print = sv_dup_inc(proto_perl->Iutf8_print, param);
10562 PL_utf8_punct = sv_dup_inc(proto_perl->Iutf8_punct, param);
10563 PL_utf8_xdigit = sv_dup_inc(proto_perl->Iutf8_xdigit, param);
10564 PL_utf8_mark = sv_dup_inc(proto_perl->Iutf8_mark, param);
10565 PL_utf8_toupper = sv_dup_inc(proto_perl->Iutf8_toupper, param);
10566 PL_utf8_totitle = sv_dup_inc(proto_perl->Iutf8_totitle, param);
10567 PL_utf8_tolower = sv_dup_inc(proto_perl->Iutf8_tolower, param);
10568 PL_utf8_tofold = sv_dup_inc(proto_perl->Iutf8_tofold, param);
10569 PL_utf8_idstart = sv_dup_inc(proto_perl->Iutf8_idstart, param);
10570 PL_utf8_idcont = sv_dup_inc(proto_perl->Iutf8_idcont, param);
1d7c1841 10571
bd81e77b
NC
10572 /* Did the locale setup indicate UTF-8? */
10573 PL_utf8locale = proto_perl->Iutf8locale;
10574 /* Unicode features (see perlrun/-C) */
10575 PL_unicode = proto_perl->Iunicode;
1d7c1841 10576
bd81e77b
NC
10577 /* Pre-5.8 signals control */
10578 PL_signals = proto_perl->Isignals;
1d7c1841 10579
bd81e77b
NC
10580 /* times() ticks per second */
10581 PL_clocktick = proto_perl->Iclocktick;
1d7c1841 10582
bd81e77b
NC
10583 /* Recursion stopper for PerlIO_find_layer */
10584 PL_in_load_module = proto_perl->Iin_load_module;
8df990a8 10585
bd81e77b
NC
10586 /* sort() routine */
10587 PL_sort_RealCmp = proto_perl->Isort_RealCmp;
e5dd39fc 10588
bd81e77b
NC
10589 /* Not really needed/useful since the reenrant_retint is "volatile",
10590 * but do it for consistency's sake. */
10591 PL_reentrant_retint = proto_perl->Ireentrant_retint;
1d7c1841 10592
bd81e77b
NC
10593 /* Hooks to shared SVs and locks. */
10594 PL_sharehook = proto_perl->Isharehook;
10595 PL_lockhook = proto_perl->Ilockhook;
10596 PL_unlockhook = proto_perl->Iunlockhook;
10597 PL_threadhook = proto_perl->Ithreadhook;
1d7c1841 10598
bd81e77b
NC
10599 PL_runops_std = proto_perl->Irunops_std;
10600 PL_runops_dbg = proto_perl->Irunops_dbg;
1d7c1841 10601
bd81e77b
NC
10602#ifdef THREADS_HAVE_PIDS
10603 PL_ppid = proto_perl->Ippid;
10604#endif
1d7c1841 10605
bd81e77b 10606 /* swatch cache */
5c284bb0 10607 PL_last_swash_hv = NULL; /* reinits on demand */
bd81e77b
NC
10608 PL_last_swash_klen = 0;
10609 PL_last_swash_key[0]= '\0';
10610 PL_last_swash_tmps = (U8*)NULL;
10611 PL_last_swash_slen = 0;
1d7c1841 10612
bd81e77b
NC
10613 PL_glob_index = proto_perl->Iglob_index;
10614 PL_srand_called = proto_perl->Isrand_called;
10615 PL_uudmap['M'] = 0; /* reinits on demand */
10616 PL_bitcount = Nullch; /* reinits on demand */
05ec9bb3 10617
bd81e77b
NC
10618 if (proto_perl->Ipsig_pend) {
10619 Newxz(PL_psig_pend, SIG_SIZE, int);
10620 }
10621 else {
10622 PL_psig_pend = (int*)NULL;
10623 }
05ec9bb3 10624
bd81e77b
NC
10625 if (proto_perl->Ipsig_ptr) {
10626 Newxz(PL_psig_ptr, SIG_SIZE, SV*);
10627 Newxz(PL_psig_name, SIG_SIZE, SV*);
10628 for (i = 1; i < SIG_SIZE; i++) {
10629 PL_psig_ptr[i] = sv_dup_inc(proto_perl->Ipsig_ptr[i], param);
10630 PL_psig_name[i] = sv_dup_inc(proto_perl->Ipsig_name[i], param);
10631 }
10632 }
10633 else {
10634 PL_psig_ptr = (SV**)NULL;
10635 PL_psig_name = (SV**)NULL;
10636 }
05ec9bb3 10637
bd81e77b 10638 /* thrdvar.h stuff */
1d7c1841 10639
bd81e77b
NC
10640 if (flags & CLONEf_COPY_STACKS) {
10641 /* next allocation will be PL_tmps_stack[PL_tmps_ix+1] */
10642 PL_tmps_ix = proto_perl->Ttmps_ix;
10643 PL_tmps_max = proto_perl->Ttmps_max;
10644 PL_tmps_floor = proto_perl->Ttmps_floor;
10645 Newxz(PL_tmps_stack, PL_tmps_max, SV*);
10646 i = 0;
10647 while (i <= PL_tmps_ix) {
10648 PL_tmps_stack[i] = sv_dup_inc(proto_perl->Ttmps_stack[i], param);
10649 ++i;
10650 }
d2d73c3e 10651
bd81e77b
NC
10652 /* next PUSHMARK() sets *(PL_markstack_ptr+1) */
10653 i = proto_perl->Tmarkstack_max - proto_perl->Tmarkstack;
10654 Newxz(PL_markstack, i, I32);
10655 PL_markstack_max = PL_markstack + (proto_perl->Tmarkstack_max
10656 - proto_perl->Tmarkstack);
10657 PL_markstack_ptr = PL_markstack + (proto_perl->Tmarkstack_ptr
10658 - proto_perl->Tmarkstack);
10659 Copy(proto_perl->Tmarkstack, PL_markstack,
10660 PL_markstack_ptr - PL_markstack + 1, I32);
d2d73c3e 10661
bd81e77b
NC
10662 /* next push_scope()/ENTER sets PL_scopestack[PL_scopestack_ix]
10663 * NOTE: unlike the others! */
10664 PL_scopestack_ix = proto_perl->Tscopestack_ix;
10665 PL_scopestack_max = proto_perl->Tscopestack_max;
10666 Newxz(PL_scopestack, PL_scopestack_max, I32);
10667 Copy(proto_perl->Tscopestack, PL_scopestack, PL_scopestack_ix, I32);
d419787a 10668
bd81e77b
NC
10669 /* NOTE: si_dup() looks at PL_markstack */
10670 PL_curstackinfo = si_dup(proto_perl->Tcurstackinfo, param);
d2d73c3e 10671
bd81e77b
NC
10672 /* PL_curstack = PL_curstackinfo->si_stack; */
10673 PL_curstack = av_dup(proto_perl->Tcurstack, param);
10674 PL_mainstack = av_dup(proto_perl->Tmainstack, param);
1d7c1841 10675
bd81e77b
NC
10676 /* next PUSHs() etc. set *(PL_stack_sp+1) */
10677 PL_stack_base = AvARRAY(PL_curstack);
10678 PL_stack_sp = PL_stack_base + (proto_perl->Tstack_sp
10679 - proto_perl->Tstack_base);
10680 PL_stack_max = PL_stack_base + AvMAX(PL_curstack);
1d7c1841 10681
bd81e77b
NC
10682 /* next SSPUSHFOO() sets PL_savestack[PL_savestack_ix]
10683 * NOTE: unlike the others! */
10684 PL_savestack_ix = proto_perl->Tsavestack_ix;
10685 PL_savestack_max = proto_perl->Tsavestack_max;
10686 /*Newxz(PL_savestack, PL_savestack_max, ANY);*/
10687 PL_savestack = ss_dup(proto_perl, param);
10688 }
10689 else {
10690 init_stacks();
10691 ENTER; /* perl_destruct() wants to LEAVE; */
10692 }
1d7c1841 10693
bd81e77b
NC
10694 PL_start_env = proto_perl->Tstart_env; /* XXXXXX */
10695 PL_top_env = &PL_start_env;
1d7c1841 10696
bd81e77b 10697 PL_op = proto_perl->Top;
4a4c6fe3 10698
bd81e77b
NC
10699 PL_Sv = Nullsv;
10700 PL_Xpv = (XPV*)NULL;
10701 PL_na = proto_perl->Tna;
1fcf4c12 10702
bd81e77b
NC
10703 PL_statbuf = proto_perl->Tstatbuf;
10704 PL_statcache = proto_perl->Tstatcache;
10705 PL_statgv = gv_dup(proto_perl->Tstatgv, param);
10706 PL_statname = sv_dup_inc(proto_perl->Tstatname, param);
10707#ifdef HAS_TIMES
10708 PL_timesbuf = proto_perl->Ttimesbuf;
10709#endif
1d7c1841 10710
bd81e77b
NC
10711 PL_tainted = proto_perl->Ttainted;
10712 PL_curpm = proto_perl->Tcurpm; /* XXX No PMOP ref count */
10713 PL_rs = sv_dup_inc(proto_perl->Trs, param);
10714 PL_last_in_gv = gv_dup(proto_perl->Tlast_in_gv, param);
10715 PL_ofs_sv = sv_dup_inc(proto_perl->Tofs_sv, param);
10716 PL_defoutgv = gv_dup_inc(proto_perl->Tdefoutgv, param);
10717 PL_chopset = proto_perl->Tchopset; /* XXX never deallocated */
10718 PL_toptarget = sv_dup_inc(proto_perl->Ttoptarget, param);
10719 PL_bodytarget = sv_dup_inc(proto_perl->Tbodytarget, param);
10720 PL_formtarget = sv_dup(proto_perl->Tformtarget, param);
1d7c1841 10721
bd81e77b
NC
10722 PL_restartop = proto_perl->Trestartop;
10723 PL_in_eval = proto_perl->Tin_eval;
10724 PL_delaymagic = proto_perl->Tdelaymagic;
10725 PL_dirty = proto_perl->Tdirty;
10726 PL_localizing = proto_perl->Tlocalizing;
1d7c1841 10727
bd81e77b
NC
10728 PL_errors = sv_dup_inc(proto_perl->Terrors, param);
10729 PL_hv_fetch_ent_mh = Nullhe;
10730 PL_modcount = proto_perl->Tmodcount;
10731 PL_lastgotoprobe = Nullop;
10732 PL_dumpindent = proto_perl->Tdumpindent;
1d7c1841 10733
bd81e77b
NC
10734 PL_sortcop = (OP*)any_dup(proto_perl->Tsortcop, proto_perl);
10735 PL_sortstash = hv_dup(proto_perl->Tsortstash, param);
10736 PL_firstgv = gv_dup(proto_perl->Tfirstgv, param);
10737 PL_secondgv = gv_dup(proto_perl->Tsecondgv, param);
10738 PL_efloatbuf = Nullch; /* reinits on demand */
10739 PL_efloatsize = 0; /* reinits on demand */
d2d73c3e 10740
bd81e77b 10741 /* regex stuff */
1d7c1841 10742
bd81e77b
NC
10743 PL_screamfirst = NULL;
10744 PL_screamnext = NULL;
10745 PL_maxscream = -1; /* reinits on demand */
10746 PL_lastscream = Nullsv;
1d7c1841 10747
bd81e77b
NC
10748 PL_watchaddr = NULL;
10749 PL_watchok = Nullch;
1d7c1841 10750
bd81e77b
NC
10751 PL_regdummy = proto_perl->Tregdummy;
10752 PL_regprecomp = Nullch;
10753 PL_regnpar = 0;
10754 PL_regsize = 0;
10755 PL_colorset = 0; /* reinits PL_colors[] */
10756 /*PL_colors[6] = {0,0,0,0,0,0};*/
10757 PL_reginput = Nullch;
10758 PL_regbol = Nullch;
10759 PL_regeol = Nullch;
10760 PL_regstartp = (I32*)NULL;
10761 PL_regendp = (I32*)NULL;
10762 PL_reglastparen = (U32*)NULL;
10763 PL_reglastcloseparen = (U32*)NULL;
10764 PL_regtill = Nullch;
10765 PL_reg_start_tmp = (char**)NULL;
10766 PL_reg_start_tmpl = 0;
10767 PL_regdata = (struct reg_data*)NULL;
10768 PL_bostr = Nullch;
10769 PL_reg_flags = 0;
10770 PL_reg_eval_set = 0;
10771 PL_regnarrate = 0;
10772 PL_regprogram = (regnode*)NULL;
10773 PL_regindent = 0;
10774 PL_regcc = (CURCUR*)NULL;
10775 PL_reg_call_cc = (struct re_cc_state*)NULL;
10776 PL_reg_re = (regexp*)NULL;
10777 PL_reg_ganch = Nullch;
10778 PL_reg_sv = Nullsv;
10779 PL_reg_match_utf8 = FALSE;
10780 PL_reg_magic = (MAGIC*)NULL;
10781 PL_reg_oldpos = 0;
10782 PL_reg_oldcurpm = (PMOP*)NULL;
10783 PL_reg_curpm = (PMOP*)NULL;
10784 PL_reg_oldsaved = Nullch;
10785 PL_reg_oldsavedlen = 0;
10786#ifdef PERL_OLD_COPY_ON_WRITE
10787 PL_nrs = Nullsv;
10788#endif
10789 PL_reg_maxiter = 0;
10790 PL_reg_leftiter = 0;
10791 PL_reg_poscache = Nullch;
10792 PL_reg_poscache_size= 0;
1d7c1841 10793
bd81e77b
NC
10794 /* RE engine - function pointers */
10795 PL_regcompp = proto_perl->Tregcompp;
10796 PL_regexecp = proto_perl->Tregexecp;
10797 PL_regint_start = proto_perl->Tregint_start;
10798 PL_regint_string = proto_perl->Tregint_string;
10799 PL_regfree = proto_perl->Tregfree;
1d7c1841 10800
bd81e77b
NC
10801 PL_reginterp_cnt = 0;
10802 PL_reg_starttry = 0;
1d7c1841 10803
bd81e77b
NC
10804 /* Pluggable optimizer */
10805 PL_peepp = proto_perl->Tpeepp;
1d7c1841 10806
bd81e77b 10807 PL_stashcache = newHV();
1d7c1841 10808
bd81e77b
NC
10809 if (!(flags & CLONEf_KEEP_PTR_TABLE)) {
10810 ptr_table_free(PL_ptr_table);
10811 PL_ptr_table = NULL;
10812 }
1d7c1841 10813
bd81e77b
NC
10814 /* Call the ->CLONE method, if it exists, for each of the stashes
10815 identified by sv_dup() above.
10816 */
10817 while(av_len(param->stashes) != -1) {
10818 HV* const stash = (HV*) av_shift(param->stashes);
10819 GV* const cloner = gv_fetchmethod_autoload(stash, "CLONE", 0);
10820 if (cloner && GvCV(cloner)) {
10821 dSP;
10822 ENTER;
10823 SAVETMPS;
10824 PUSHMARK(SP);
10825 XPUSHs(sv_2mortal(newSVhek(HvNAME_HEK(stash))));
10826 PUTBACK;
10827 call_sv((SV*)GvCV(cloner), G_DISCARD);
10828 FREETMPS;
10829 LEAVE;
10830 }
1d7c1841 10831 }
1d7c1841 10832
bd81e77b 10833 SvREFCNT_dec(param->stashes);
1d7c1841 10834
bd81e77b
NC
10835 /* orphaned? eg threads->new inside BEGIN or use */
10836 if (PL_compcv && ! SvREFCNT(PL_compcv)) {
10837 (void)SvREFCNT_inc(PL_compcv);
10838 SAVEFREESV(PL_compcv);
10839 }
dd2155a4 10840
bd81e77b
NC
10841 return my_perl;
10842}
1d7c1841 10843
bd81e77b 10844#endif /* USE_ITHREADS */
1d7c1841 10845
bd81e77b
NC
10846/*
10847=head1 Unicode Support
1d7c1841 10848
bd81e77b 10849=for apidoc sv_recode_to_utf8
1d7c1841 10850
bd81e77b
NC
10851The encoding is assumed to be an Encode object, on entry the PV
10852of the sv is assumed to be octets in that encoding, and the sv
10853will be converted into Unicode (and UTF-8).
1d7c1841 10854
bd81e77b
NC
10855If the sv already is UTF-8 (or if it is not POK), or if the encoding
10856is not a reference, nothing is done to the sv. If the encoding is not
10857an C<Encode::XS> Encoding object, bad things will happen.
10858(See F<lib/encoding.pm> and L<Encode>).
1d7c1841 10859
bd81e77b 10860The PV of the sv is returned.
1d7c1841 10861
bd81e77b 10862=cut */
1d7c1841 10863
bd81e77b
NC
10864char *
10865Perl_sv_recode_to_utf8(pTHX_ SV *sv, SV *encoding)
10866{
10867 dVAR;
10868 if (SvPOK(sv) && !SvUTF8(sv) && !IN_BYTES && SvROK(encoding)) {
10869 SV *uni;
10870 STRLEN len;
10871 const char *s;
10872 dSP;
10873 ENTER;
10874 SAVETMPS;
10875 save_re_context();
10876 PUSHMARK(sp);
10877 EXTEND(SP, 3);
10878 XPUSHs(encoding);
10879 XPUSHs(sv);
10880/*
10881 NI-S 2002/07/09
10882 Passing sv_yes is wrong - it needs to be or'ed set of constants
10883 for Encode::XS, while UTf-8 decode (currently) assumes a true value means
10884 remove converted chars from source.
1d7c1841 10885
bd81e77b 10886 Both will default the value - let them.
1d7c1841 10887
bd81e77b
NC
10888 XPUSHs(&PL_sv_yes);
10889*/
10890 PUTBACK;
10891 call_method("decode", G_SCALAR);
10892 SPAGAIN;
10893 uni = POPs;
10894 PUTBACK;
10895 s = SvPV_const(uni, len);
10896 if (s != SvPVX_const(sv)) {
10897 SvGROW(sv, len + 1);
10898 Move(s, SvPVX(sv), len + 1, char);
10899 SvCUR_set(sv, len);
10900 }
10901 FREETMPS;
10902 LEAVE;
10903 SvUTF8_on(sv);
10904 return SvPVX(sv);
389edf32 10905 }
bd81e77b
NC
10906 return SvPOKp(sv) ? SvPVX(sv) : NULL;
10907}
1d7c1841 10908
bd81e77b
NC
10909/*
10910=for apidoc sv_cat_decode
1d7c1841 10911
bd81e77b
NC
10912The encoding is assumed to be an Encode object, the PV of the ssv is
10913assumed to be octets in that encoding and decoding the input starts
10914from the position which (PV + *offset) pointed to. The dsv will be
10915concatenated the decoded UTF-8 string from ssv. Decoding will terminate
10916when the string tstr appears in decoding output or the input ends on
10917the PV of the ssv. The value which the offset points will be modified
10918to the last input position on the ssv.
1d7c1841 10919
bd81e77b 10920Returns TRUE if the terminator was found, else returns FALSE.
1d7c1841 10921
bd81e77b
NC
10922=cut */
10923
10924bool
10925Perl_sv_cat_decode(pTHX_ SV *dsv, SV *encoding,
10926 SV *ssv, int *offset, char *tstr, int tlen)
10927{
10928 dVAR;
10929 bool ret = FALSE;
10930 if (SvPOK(ssv) && SvPOK(dsv) && SvROK(encoding) && offset) {
10931 SV *offsv;
10932 dSP;
10933 ENTER;
10934 SAVETMPS;
10935 save_re_context();
10936 PUSHMARK(sp);
10937 EXTEND(SP, 6);
10938 XPUSHs(encoding);
10939 XPUSHs(dsv);
10940 XPUSHs(ssv);
10941 XPUSHs(offsv = sv_2mortal(newSViv(*offset)));
10942 XPUSHs(sv_2mortal(newSVpvn(tstr, tlen)));
10943 PUTBACK;
10944 call_method("cat_decode", G_SCALAR);
10945 SPAGAIN;
10946 ret = SvTRUE(TOPs);
10947 *offset = SvIV(offsv);
10948 PUTBACK;
10949 FREETMPS;
10950 LEAVE;
389edf32 10951 }
bd81e77b
NC
10952 else
10953 Perl_croak(aTHX_ "Invalid argument to sv_cat_decode");
10954 return ret;
1d7c1841 10955
bd81e77b 10956}
1d7c1841 10957
bd81e77b
NC
10958/* ---------------------------------------------------------------------
10959 *
10960 * support functions for report_uninit()
10961 */
1d7c1841 10962
bd81e77b
NC
10963/* the maxiumum size of array or hash where we will scan looking
10964 * for the undefined element that triggered the warning */
1d7c1841 10965
bd81e77b 10966#define FUV_MAX_SEARCH_SIZE 1000
1d7c1841 10967
bd81e77b
NC
10968/* Look for an entry in the hash whose value has the same SV as val;
10969 * If so, return a mortal copy of the key. */
1d7c1841 10970
bd81e77b
NC
10971STATIC SV*
10972S_find_hash_subscript(pTHX_ HV *hv, SV* val)
10973{
10974 dVAR;
10975 register HE **array;
10976 I32 i;
6c3182a5 10977
bd81e77b
NC
10978 if (!hv || SvMAGICAL(hv) || !HvARRAY(hv) ||
10979 (HvTOTALKEYS(hv) > FUV_MAX_SEARCH_SIZE))
10980 return Nullsv;
6c3182a5 10981
bd81e77b 10982 array = HvARRAY(hv);
6c3182a5 10983
bd81e77b
NC
10984 for (i=HvMAX(hv); i>0; i--) {
10985 register HE *entry;
10986 for (entry = array[i]; entry; entry = HeNEXT(entry)) {
10987 if (HeVAL(entry) != val)
10988 continue;
10989 if ( HeVAL(entry) == &PL_sv_undef ||
10990 HeVAL(entry) == &PL_sv_placeholder)
10991 continue;
10992 if (!HeKEY(entry))
10993 return Nullsv;
10994 if (HeKLEN(entry) == HEf_SVKEY)
10995 return sv_mortalcopy(HeKEY_sv(entry));
10996 return sv_2mortal(newSVpvn(HeKEY(entry), HeKLEN(entry)));
10997 }
10998 }
10999 return Nullsv;
11000}
6c3182a5 11001
bd81e77b
NC
11002/* Look for an entry in the array whose value has the same SV as val;
11003 * If so, return the index, otherwise return -1. */
6c3182a5 11004
bd81e77b
NC
11005STATIC I32
11006S_find_array_subscript(pTHX_ AV *av, SV* val)
11007{
11008 SV** svp;
11009 I32 i;
11010 if (!av || SvMAGICAL(av) || !AvARRAY(av) ||
11011 (AvFILLp(av) > FUV_MAX_SEARCH_SIZE))
11012 return -1;
57c6e6d2 11013
bd81e77b
NC
11014 svp = AvARRAY(av);
11015 for (i=AvFILLp(av); i>=0; i--) {
11016 if (svp[i] == val && svp[i] != &PL_sv_undef)
11017 return i;
11018 }
11019 return -1;
11020}
15a5279a 11021
bd81e77b
NC
11022/* S_varname(): return the name of a variable, optionally with a subscript.
11023 * If gv is non-zero, use the name of that global, along with gvtype (one
11024 * of "$", "@", "%"); otherwise use the name of the lexical at pad offset
11025 * targ. Depending on the value of the subscript_type flag, return:
11026 */
bce260cd 11027
bd81e77b
NC
11028#define FUV_SUBSCRIPT_NONE 1 /* "@foo" */
11029#define FUV_SUBSCRIPT_ARRAY 2 /* "$foo[aindex]" */
11030#define FUV_SUBSCRIPT_HASH 3 /* "$foo{keyname}" */
11031#define FUV_SUBSCRIPT_WITHIN 4 /* "within @foo" */
bce260cd 11032
bd81e77b
NC
11033STATIC SV*
11034S_varname(pTHX_ GV *gv, const char gvtype, PADOFFSET targ,
11035 SV* keyname, I32 aindex, int subscript_type)
11036{
1d7c1841 11037
bd81e77b
NC
11038 SV * const name = sv_newmortal();
11039 if (gv) {
11040 char buffer[2];
11041 buffer[0] = gvtype;
11042 buffer[1] = 0;
1d7c1841 11043
bd81e77b 11044 /* as gv_fullname4(), but add literal '^' for $^FOO names */
66fe0623 11045
bd81e77b 11046 gv_fullname4(name, gv, buffer, 0);
1d7c1841 11047
bd81e77b
NC
11048 if ((unsigned int)SvPVX(name)[1] <= 26) {
11049 buffer[0] = '^';
11050 buffer[1] = SvPVX(name)[1] + 'A' - 1;
1d7c1841 11051
bd81e77b
NC
11052 /* Swap the 1 unprintable control character for the 2 byte pretty
11053 version - ie substr($name, 1, 1) = $buffer; */
11054 sv_insert(name, 1, 1, buffer, 2);
1d7c1841 11055 }
bd81e77b
NC
11056 }
11057 else {
11058 U32 unused;
11059 CV * const cv = find_runcv(&unused);
11060 SV *sv;
11061 AV *av;
1d7c1841 11062
bd81e77b
NC
11063 if (!cv || !CvPADLIST(cv))
11064 return Nullsv;
11065 av = (AV*)(*av_fetch(CvPADLIST(cv), 0, FALSE));
11066 sv = *av_fetch(av, targ, FALSE);
11067 /* SvLEN in a pad name is not to be trusted */
11068 sv_setpv(name, SvPV_nolen_const(sv));
11069 }
1d7c1841 11070
bd81e77b
NC
11071 if (subscript_type == FUV_SUBSCRIPT_HASH) {
11072 SV * const sv = NEWSV(0,0);
11073 *SvPVX(name) = '$';
11074 Perl_sv_catpvf(aTHX_ name, "{%s}",
11075 pv_display(sv,SvPVX_const(keyname), SvCUR(keyname), 0, 32));
11076 SvREFCNT_dec(sv);
11077 }
11078 else if (subscript_type == FUV_SUBSCRIPT_ARRAY) {
11079 *SvPVX(name) = '$';
11080 Perl_sv_catpvf(aTHX_ name, "[%"IVdf"]", (IV)aindex);
11081 }
11082 else if (subscript_type == FUV_SUBSCRIPT_WITHIN)
11083 sv_insert(name, 0, 0, "within ", 7);
1d7c1841 11084
bd81e77b
NC
11085 return name;
11086}
1d7c1841 11087
1d7c1841 11088
bd81e77b
NC
11089/*
11090=for apidoc find_uninit_var
1d7c1841 11091
bd81e77b
NC
11092Find the name of the undefined variable (if any) that caused the operator o
11093to issue a "Use of uninitialized value" warning.
11094If match is true, only return a name if it's value matches uninit_sv.
11095So roughly speaking, if a unary operator (such as OP_COS) generates a
11096warning, then following the direct child of the op may yield an
11097OP_PADSV or OP_GV that gives the name of the undefined variable. On the
11098other hand, with OP_ADD there are two branches to follow, so we only print
11099the variable name if we get an exact match.
1d7c1841 11100
bd81e77b 11101The name is returned as a mortal SV.
1d7c1841 11102
bd81e77b
NC
11103Assumes that PL_op is the op that originally triggered the error, and that
11104PL_comppad/PL_curpad points to the currently executing pad.
1d7c1841 11105
bd81e77b
NC
11106=cut
11107*/
1d7c1841 11108
bd81e77b
NC
11109STATIC SV *
11110S_find_uninit_var(pTHX_ OP* obase, SV* uninit_sv, bool match)
11111{
11112 dVAR;
11113 SV *sv;
11114 AV *av;
11115 GV *gv;
11116 OP *o, *o2, *kid;
1d7c1841 11117
bd81e77b
NC
11118 if (!obase || (match && (!uninit_sv || uninit_sv == &PL_sv_undef ||
11119 uninit_sv == &PL_sv_placeholder)))
11120 return Nullsv;
1d7c1841 11121
bd81e77b 11122 switch (obase->op_type) {
1d7c1841 11123
bd81e77b
NC
11124 case OP_RV2AV:
11125 case OP_RV2HV:
11126 case OP_PADAV:
11127 case OP_PADHV:
11128 {
11129 const bool pad = (obase->op_type == OP_PADAV || obase->op_type == OP_PADHV);
11130 const bool hash = (obase->op_type == OP_PADHV || obase->op_type == OP_RV2HV);
11131 I32 index = 0;
11132 SV *keysv = Nullsv;
11133 int subscript_type = FUV_SUBSCRIPT_WITHIN;
1d7c1841 11134
bd81e77b
NC
11135 if (pad) { /* @lex, %lex */
11136 sv = PAD_SVl(obase->op_targ);
11137 gv = Nullgv;
11138 }
11139 else {
11140 if (cUNOPx(obase)->op_first->op_type == OP_GV) {
11141 /* @global, %global */
11142 gv = cGVOPx_gv(cUNOPx(obase)->op_first);
11143 if (!gv)
11144 break;
11145 sv = hash ? (SV*)GvHV(gv): (SV*)GvAV(gv);
11146 }
11147 else /* @{expr}, %{expr} */
11148 return find_uninit_var(cUNOPx(obase)->op_first,
11149 uninit_sv, match);
11150 }
1d7c1841 11151
bd81e77b
NC
11152 /* attempt to find a match within the aggregate */
11153 if (hash) {
11154 keysv = S_find_hash_subscript(aTHX_ (HV*)sv, uninit_sv);
11155 if (keysv)
11156 subscript_type = FUV_SUBSCRIPT_HASH;
11157 }
11158 else {
11159 index = S_find_array_subscript(aTHX_ (AV*)sv, uninit_sv);
11160 if (index >= 0)
11161 subscript_type = FUV_SUBSCRIPT_ARRAY;
11162 }
1d7c1841 11163
bd81e77b
NC
11164 if (match && subscript_type == FUV_SUBSCRIPT_WITHIN)
11165 break;
1d7c1841 11166
bd81e77b
NC
11167 return varname(gv, hash ? '%' : '@', obase->op_targ,
11168 keysv, index, subscript_type);
11169 }
1d7c1841 11170
bd81e77b
NC
11171 case OP_PADSV:
11172 if (match && PAD_SVl(obase->op_targ) != uninit_sv)
11173 break;
11174 return varname(Nullgv, '$', obase->op_targ,
11175 Nullsv, 0, FUV_SUBSCRIPT_NONE);
1d7c1841 11176
bd81e77b
NC
11177 case OP_GVSV:
11178 gv = cGVOPx_gv(obase);
11179 if (!gv || (match && GvSV(gv) != uninit_sv))
11180 break;
11181 return varname(gv, '$', 0, Nullsv, 0, FUV_SUBSCRIPT_NONE);
1d7c1841 11182
bd81e77b
NC
11183 case OP_AELEMFAST:
11184 if (obase->op_flags & OPf_SPECIAL) { /* lexical array */
11185 if (match) {
11186 SV **svp;
11187 av = (AV*)PAD_SV(obase->op_targ);
11188 if (!av || SvRMAGICAL(av))
11189 break;
11190 svp = av_fetch(av, (I32)obase->op_private, FALSE);
11191 if (!svp || *svp != uninit_sv)
11192 break;
11193 }
11194 return varname(Nullgv, '$', obase->op_targ,
11195 Nullsv, (I32)obase->op_private, FUV_SUBSCRIPT_ARRAY);
11196 }
11197 else {
11198 gv = cGVOPx_gv(obase);
11199 if (!gv)
11200 break;
11201 if (match) {
11202 SV **svp;
11203 av = GvAV(gv);
11204 if (!av || SvRMAGICAL(av))
11205 break;
11206 svp = av_fetch(av, (I32)obase->op_private, FALSE);
11207 if (!svp || *svp != uninit_sv)
11208 break;
11209 }
11210 return varname(gv, '$', 0,
11211 Nullsv, (I32)obase->op_private, FUV_SUBSCRIPT_ARRAY);
11212 }
11213 break;
1d7c1841 11214
bd81e77b
NC
11215 case OP_EXISTS:
11216 o = cUNOPx(obase)->op_first;
11217 if (!o || o->op_type != OP_NULL ||
11218 ! (o->op_targ == OP_AELEM || o->op_targ == OP_HELEM))
11219 break;
11220 return find_uninit_var(cBINOPo->op_last, uninit_sv, match);
a2efc822 11221
bd81e77b
NC
11222 case OP_AELEM:
11223 case OP_HELEM:
11224 if (PL_op == obase)
11225 /* $a[uninit_expr] or $h{uninit_expr} */
11226 return find_uninit_var(cBINOPx(obase)->op_last, uninit_sv, match);
081fc587 11227
bd81e77b
NC
11228 gv = Nullgv;
11229 o = cBINOPx(obase)->op_first;
11230 kid = cBINOPx(obase)->op_last;
8cf8f3d1 11231
bd81e77b
NC
11232 /* get the av or hv, and optionally the gv */
11233 sv = Nullsv;
11234 if (o->op_type == OP_PADAV || o->op_type == OP_PADHV) {
11235 sv = PAD_SV(o->op_targ);
11236 }
11237 else if ((o->op_type == OP_RV2AV || o->op_type == OP_RV2HV)
11238 && cUNOPo->op_first->op_type == OP_GV)
11239 {
11240 gv = cGVOPx_gv(cUNOPo->op_first);
11241 if (!gv)
11242 break;
11243 sv = o->op_type == OP_RV2HV ? (SV*)GvHV(gv) : (SV*)GvAV(gv);
11244 }
11245 if (!sv)
11246 break;
11247
11248 if (kid && kid->op_type == OP_CONST && SvOK(cSVOPx_sv(kid))) {
11249 /* index is constant */
11250 if (match) {
11251 if (SvMAGICAL(sv))
11252 break;
11253 if (obase->op_type == OP_HELEM) {
11254 HE* he = hv_fetch_ent((HV*)sv, cSVOPx_sv(kid), 0, 0);
11255 if (!he || HeVAL(he) != uninit_sv)
11256 break;
11257 }
11258 else {
00b6aa41 11259 SV * const * const svp = av_fetch((AV*)sv, SvIV(cSVOPx_sv(kid)), FALSE);
bd81e77b
NC
11260 if (!svp || *svp != uninit_sv)
11261 break;
11262 }
11263 }
11264 if (obase->op_type == OP_HELEM)
11265 return varname(gv, '%', o->op_targ,
11266 cSVOPx_sv(kid), 0, FUV_SUBSCRIPT_HASH);
11267 else
11268 return varname(gv, '@', o->op_targ, Nullsv,
11269 SvIV(cSVOPx_sv(kid)), FUV_SUBSCRIPT_ARRAY);
bd81e77b
NC
11270 }
11271 else {
11272 /* index is an expression;
11273 * attempt to find a match within the aggregate */
11274 if (obase->op_type == OP_HELEM) {
11275 SV * const keysv = S_find_hash_subscript(aTHX_ (HV*)sv, uninit_sv);
11276 if (keysv)
11277 return varname(gv, '%', o->op_targ,
11278 keysv, 0, FUV_SUBSCRIPT_HASH);
11279 }
11280 else {
11281 const I32 index = S_find_array_subscript(aTHX_ (AV*)sv, uninit_sv);
11282 if (index >= 0)
11283 return varname(gv, '@', o->op_targ,
11284 Nullsv, index, FUV_SUBSCRIPT_ARRAY);
11285 }
11286 if (match)
11287 break;
11288 return varname(gv,
11289 (o->op_type == OP_PADAV || o->op_type == OP_RV2AV)
11290 ? '@' : '%',
11291 o->op_targ, Nullsv, 0, FUV_SUBSCRIPT_WITHIN);
f284b03f 11292 }
a0739874 11293
bd81e77b 11294 break;
dc507217 11295
bd81e77b
NC
11296 case OP_AASSIGN:
11297 /* only examine RHS */
11298 return find_uninit_var(cBINOPx(obase)->op_first, uninit_sv, match);
6d26897e 11299
bd81e77b
NC
11300 case OP_OPEN:
11301 o = cUNOPx(obase)->op_first;
11302 if (o->op_type == OP_PUSHMARK)
11303 o = o->op_sibling;
1d7c1841 11304
bd81e77b
NC
11305 if (!o->op_sibling) {
11306 /* one-arg version of open is highly magical */
a0ae6670 11307
bd81e77b
NC
11308 if (o->op_type == OP_GV) { /* open FOO; */
11309 gv = cGVOPx_gv(o);
11310 if (match && GvSV(gv) != uninit_sv)
11311 break;
11312 return varname(gv, '$', 0,
11313 Nullsv, 0, FUV_SUBSCRIPT_NONE);
11314 }
11315 /* other possibilities not handled are:
11316 * open $x; or open my $x; should return '${*$x}'
11317 * open expr; should return '$'.expr ideally
11318 */
11319 break;
11320 }
11321 goto do_op;
ccfc67b7 11322
bd81e77b
NC
11323 /* ops where $_ may be an implicit arg */
11324 case OP_TRANS:
11325 case OP_SUBST:
11326 case OP_MATCH:
11327 if ( !(obase->op_flags & OPf_STACKED)) {
11328 if (uninit_sv == ((obase->op_private & OPpTARGET_MY)
11329 ? PAD_SVl(obase->op_targ)
11330 : DEFSV))
11331 {
11332 sv = sv_newmortal();
11333 sv_setpvn(sv, "$_", 2);
11334 return sv;
11335 }
11336 }
11337 goto do_op;
9f4817db 11338
bd81e77b
NC
11339 case OP_PRTF:
11340 case OP_PRINT:
11341 /* skip filehandle as it can't produce 'undef' warning */
11342 o = cUNOPx(obase)->op_first;
11343 if ((obase->op_flags & OPf_STACKED) && o->op_type == OP_PUSHMARK)
11344 o = o->op_sibling->op_sibling;
11345 goto do_op2;
9f4817db 11346
9f4817db 11347
bd81e77b
NC
11348 case OP_RV2SV:
11349 case OP_CUSTOM:
11350 case OP_ENTERSUB:
11351 match = 1; /* XS or custom code could trigger random warnings */
11352 goto do_op;
9f4817db 11353
bd81e77b
NC
11354 case OP_SCHOMP:
11355 case OP_CHOMP:
11356 if (SvROK(PL_rs) && uninit_sv == SvRV(PL_rs))
11357 return sv_2mortal(newSVpvn("${$/}", 5));
11358 /* FALL THROUGH */
5d170f3a 11359
bd81e77b
NC
11360 default:
11361 do_op:
11362 if (!(obase->op_flags & OPf_KIDS))
11363 break;
11364 o = cUNOPx(obase)->op_first;
11365
11366 do_op2:
11367 if (!o)
11368 break;
f9893866 11369
bd81e77b
NC
11370 /* if all except one arg are constant, or have no side-effects,
11371 * or are optimized away, then it's unambiguous */
11372 o2 = Nullop;
11373 for (kid=o; kid; kid = kid->op_sibling) {
11374 if (kid &&
11375 ( (kid->op_type == OP_CONST && SvOK(cSVOPx_sv(kid)))
11376 || (kid->op_type == OP_NULL && ! (kid->op_flags & OPf_KIDS))
11377 || (kid->op_type == OP_PUSHMARK)
11378 )
11379 )
11380 continue;
11381 if (o2) { /* more than one found */
11382 o2 = Nullop;
11383 break;
11384 }
11385 o2 = kid;
11386 }
11387 if (o2)
11388 return find_uninit_var(o2, uninit_sv, match);
7a5fa8a2 11389
bd81e77b
NC
11390 /* scan all args */
11391 while (o) {
11392 sv = find_uninit_var(o, uninit_sv, 1);
11393 if (sv)
11394 return sv;
11395 o = o->op_sibling;
d0063567 11396 }
bd81e77b 11397 break;
f9893866 11398 }
bd81e77b 11399 return Nullsv;
9f4817db
JH
11400}
11401
220e2d4e 11402
bd81e77b
NC
11403/*
11404=for apidoc report_uninit
68795e93 11405
bd81e77b 11406Print appropriate "Use of uninitialized variable" warning
220e2d4e 11407
bd81e77b
NC
11408=cut
11409*/
220e2d4e 11410
bd81e77b
NC
11411void
11412Perl_report_uninit(pTHX_ SV* uninit_sv)
220e2d4e 11413{
bd81e77b
NC
11414 if (PL_op) {
11415 SV* varname = Nullsv;
11416 if (uninit_sv) {
11417 varname = find_uninit_var(PL_op, uninit_sv,0);
11418 if (varname)
11419 sv_insert(varname, 0, 0, " ", 1);
11420 }
11421 Perl_warner(aTHX_ packWARN(WARN_UNINITIALIZED), PL_warn_uninit,
11422 varname ? SvPV_nolen_const(varname) : "",
11423 " in ", OP_DESC(PL_op));
220e2d4e 11424 }
a73e8557 11425 else
bd81e77b
NC
11426 Perl_warner(aTHX_ packWARN(WARN_UNINITIALIZED), PL_warn_uninit,
11427 "", "", "");
220e2d4e 11428}
f9893866 11429
241d1a3b
NC
11430/*
11431 * Local variables:
11432 * c-indentation-style: bsd
11433 * c-basic-offset: 4
11434 * indent-tabs-mode: t
11435 * End:
11436 *
37442d52
RGS
11437 * ex: set ts=8 sts=4 sw=4 noet:
11438 */