This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
regcomp.h: Make some #defines sequential
[perl5.git] / handy.h
CommitLineData
a0d0e21e 1/* handy.h
a687059c 2 *
1129b882 3 * Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1999, 2000,
da5d8dbb 4 * 2001, 2002, 2004, 2005, 2006, 2007, 2008, 2012 by Larry Wall and others
a687059c 5 *
6e21c824
LW
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.
8d063cd8 8 *
8d063cd8
LW
9 */
10
85e6fe83 11#if !defined(__STDC__)
378cc40b
LW
12#ifdef NULL
13#undef NULL
14#endif
a687059c 15# define NULL 0
85e6fe83
LW
16#endif
17
24792b8d
NC
18#ifndef PERL_CORE
19# define Null(type) ((type)NULL)
954c1994
GS
20
21/*
ccfc67b7 22=head1 Handy Values
954c1994 23
3609ea0d 24=for apidoc AmU||Nullch
24792b8d 25Null character pointer. (No longer available when C<PERL_CORE> is defined.)
2307c6d0 26
954c1994 27=for apidoc AmU||Nullsv
24792b8d 28Null SV pointer. (No longer available when C<PERL_CORE> is defined.)
954c1994
GS
29
30=cut
31*/
32
24792b8d
NC
33# define Nullch Null(char*)
34# define Nullfp Null(PerlIO*)
35# define Nullsv Null(SV*)
36#endif
8d063cd8 37
641d3f0b
PP
38#ifdef TRUE
39#undef TRUE
40#endif
41#ifdef FALSE
42#undef FALSE
43#endif
44#define TRUE (1)
45#define FALSE (0)
46
cf3f0ffb
DM
47/* The MUTABLE_*() macros cast pointers to the types shown, in such a way
48 * (compiler permitting) that casting away const-ness will give a warning;
49 * e.g.:
50 *
51 * const SV *sv = ...;
52 * AV *av1 = (AV*)sv; <== BAD: the const has been silently cast away
53 * AV *av2 = MUTABLE_AV(sv); <== GOOD: it may warn
54 */
55
b1bc3f34 56#if defined(__GNUC__) && !defined(PERL_GCC_BRACE_GROUPS_FORBIDDEN)
f065df04 57# define MUTABLE_PTR(p) ({ void *_p = (p); _p; })
b1bc3f34
NC
58#else
59# define MUTABLE_PTR(p) ((void *) (p))
60#endif
61
a062e10d 62#define MUTABLE_AV(p) ((AV *)MUTABLE_PTR(p))
ea726b52 63#define MUTABLE_CV(p) ((CV *)MUTABLE_PTR(p))
159b6efe 64#define MUTABLE_GV(p) ((GV *)MUTABLE_PTR(p))
dbebbdb4 65#define MUTABLE_HV(p) ((HV *)MUTABLE_PTR(p))
a45c7426 66#define MUTABLE_IO(p) ((IO *)MUTABLE_PTR(p))
b1bc3f34 67#define MUTABLE_SV(p) ((SV *)MUTABLE_PTR(p))
27d4fb96 68
bd31be4b
NC
69#ifdef I_STDBOOL
70# include <stdbool.h>
71# ifndef HAS_BOOL
72# define HAS_BOOL 1
73# endif
74#endif
75
8e84507e 76/* bool is built-in for g++-2.6.3 and later, which might be used
c1d22f6b
GS
77 for extensions. <_G_config.h> defines _G_HAVE_BOOL, but we can't
78 be sure _G_config.h will be included before this file. _G_config.h
8e84507e 79 also defines _G_HAVE_BOOL for both gcc and g++, but only g++
c1d22f6b
GS
80 actually has bool. Hence, _G_HAVE_BOOL is pretty useless for us.
81 g++ can be identified by __GNUG__.
82 Andy Dougherty February 2000
5d94fbed 83*/
3609ea0d 84#ifdef __GNUG__ /* GNU g++ has bool built-in */
5d94fbed 85# ifndef HAS_BOOL
c1d22f6b 86# define HAS_BOOL 1
5d94fbed 87# endif
5d94fbed
AD
88#endif
89
641d3f0b
PP
90/* The NeXT dynamic loader headers will not build with the bool macro
91 So declare them now to clear confusion.
92*/
8f1f23e8 93#if defined(NeXT) || defined(__NeXT__)
641d3f0b
PP
94# undef FALSE
95# undef TRUE
96 typedef enum bool { FALSE = 0, TRUE = 1 } bool;
97# define ENUM_BOOL 1
98# ifndef HAS_BOOL
99# define HAS_BOOL 1
100# endif /* !HAS_BOOL */
8f1f23e8 101#endif /* NeXT || __NeXT__ */
641d3f0b 102
5d94fbed 103#ifndef HAS_BOOL
cc3315ba 104# if defined(VMS)
5d94fbed
AD
105# define bool int
106# else
107# define bool char
108# endif
c1d22f6b 109# define HAS_BOOL 1
a687059c 110#endif
0d3e774c 111
f2338a2e
DM
112/* a simple (bool) cast may not do the right thing: if bool is defined
113 * as char for example, then the cast from int is implementation-defined
18f5643b 114 * (bool)!!(cbool) in a ternary triggers a bug in xlc on AIX
f2338a2e
DM
115 */
116
18f5643b 117#define cBOOL(cbool) ((cbool) ? (bool)1 : (bool)0)
f2338a2e 118
46c6c7e2 119/* Try to figure out __func__ or __FUNCTION__ equivalent, if any.
e352bcff
JH
120 * XXX Should really be a Configure probe, with HAS__FUNCTION__
121 * and FUNCTION__ as results.
122 * XXX Similarly, a Configure probe for __FILE__ and __LINE__ is needed. */
46c6c7e2
JH
123#if (defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L) || (defined(__SUNPRO_C)) /* C99 or close enough. */
124# define FUNCTION__ __func__
125#else
70764c00 126# if (defined(_MSC_VER) && _MSC_VER < 1300) || /* MSVC6 has neither __func__ nor __FUNCTION and no good workarounds, either. */ \
be73ebb0 127 (defined(__DECC_VER)) /* Tru64 or VMS, and strict C89 being used, but not modern enough cc (in Tur64, -c99 not known, only -std1). */
46c6c7e2
JH
128# define FUNCTION__ ""
129# else
130# define FUNCTION__ __FUNCTION__ /* Common extension. */
131# endif
132#endif
133
27d4fb96
PP
134/* XXX A note on the perl source internal type system. The
135 original intent was that I32 be *exactly* 32 bits.
136
137 Currently, we only guarantee that I32 is *at least* 32 bits.
138 Specifically, if int is 64 bits, then so is I32. (This is the case
139 for the Cray.) This has the advantage of meshing nicely with
140 standard library calls (where we pass an I32 and the library is
141 expecting an int), but the disadvantage that an I32 is not 32 bits.
142 Andy Dougherty August 1996
24fef2a7 143
dc45a647
MB
144 There is no guarantee that there is *any* integral type with
145 exactly 32 bits. It is perfectly legal for a system to have
146 sizeof(short) == sizeof(int) == sizeof(long) == 8.
693762b4 147
dc45a647
MB
148 Similarly, there is no guarantee that I16 and U16 have exactly 16
149 bits.
693762b4 150
8e84507e
NIS
151 For dealing with issues that may arise from various 32/64-bit
152 systems, we will ask Configure to check out
8175356b 153
3609ea0d
JH
154 SHORTSIZE == sizeof(short)
155 INTSIZE == sizeof(int)
156 LONGSIZE == sizeof(long)
dc45a647 157 LONGLONGSIZE == sizeof(long long) (if HAS_LONG_LONG)
3609ea0d 158 PTRSIZE == sizeof(void *)
dc45a647
MB
159 DOUBLESIZE == sizeof(double)
160 LONG_DOUBLESIZE == sizeof(long double) (if HAS_LONG_DOUBLE).
8175356b 161
27d4fb96
PP
162*/
163
69512466
JH
164#ifdef I_INTTYPES /* e.g. Linux has int64_t without <inttypes.h> */
165# include <inttypes.h>
dd0eed91
JH
166# ifdef INT32_MIN_BROKEN
167# undef INT32_MIN
168# define INT32_MIN (-2147483647-1)
169# endif
170# ifdef INT64_MIN_BROKEN
171# undef INT64_MIN
172# define INT64_MIN (-9223372036854775807LL-1)
173# endif
69512466
JH
174#endif
175
8175356b
JH
176typedef I8TYPE I8;
177typedef U8TYPE U8;
178typedef I16TYPE I16;
179typedef U16TYPE U16;
180typedef I32TYPE I32;
181typedef U32TYPE U32;
6b8eaf93
JH
182#ifdef PERL_CORE
183# ifdef HAS_QUAD
8175356b
JH
184typedef I64TYPE I64;
185typedef U64TYPE U64;
6b8eaf93
JH
186# endif
187#endif /* PERL_CORE */
8175356b 188
69512466 189#if defined(HAS_QUAD) && defined(USE_64_BIT_INT)
80524f33
JL
190# if defined(HAS_LONG_LONG) && QUADKIND == QUAD_IS_LONG_LONG
191# define PeRl_INT64_C(c) CAT2(c,LL)
192# define PeRl_UINT64_C(c) CAT2(c,ULL)
193# else
52865553
SH
194# if QUADKIND == QUAD_IS___INT64
195# define PeRl_INT64_C(c) CAT2(c,I64)
196# define PeRl_UINT64_C(c) CAT2(c,UI64)
69512466 197# else
52865553
SH
198# if LONGSIZE == 8 && QUADKIND == QUAD_IS_LONG
199# define PeRl_INT64_C(c) CAT2(c,L)
200# define PeRl_UINT64_C(c) CAT2(c,UL)
69512466 201# else
80524f33
JL
202# define PeRl_INT64_C(c) ((I64TYPE)(c))
203# define PeRl_UINT64_C(c) ((U64TYPE)(c))
69512466
JH
204# endif
205# endif
e8c95190 206# endif
80524f33
JL
207# ifndef UINT64_C
208# define UINT64_C(c) PeRl_UINT64_C(c)
209# endif
210# ifndef INT64_C
211# define INT64_C(c) PeRl_INT64_C(c)
212# endif
e8c95190 213#endif
dc750f2e 214
d8668976 215#if defined(UINT8_MAX) && defined(INT16_MAX) && defined(INT32_MAX)
5ff3f7a4 216
5ff3f7a4
GS
217/* I8_MAX and I8_MIN constants are not defined, as I8 is an ambiguous type.
218 Please search CHAR_MAX in perl.h for further details. */
219#define U8_MAX UINT8_MAX
220#define U8_MIN UINT8_MIN
221
5ff3f7a4
GS
222#define I16_MAX INT16_MAX
223#define I16_MIN INT16_MIN
224#define U16_MAX UINT16_MAX
225#define U16_MIN UINT16_MIN
226
5ff3f7a4
GS
227#define I32_MAX INT32_MAX
228#define I32_MIN INT32_MIN
0e983133
GS
229#ifndef UINT32_MAX_BROKEN /* e.g. HP-UX with gcc messes this up */
230# define U32_MAX UINT32_MAX
231#else
232# define U32_MAX 4294967295U
233#endif
5ff3f7a4
GS
234#define U32_MIN UINT32_MIN
235
236#else
237
5c9fa16e
KA
238/* I8_MAX and I8_MIN constants are not defined, as I8 is an ambiguous type.
239 Please search CHAR_MAX in perl.h for further details. */
27d4fb96
PP
240#define U8_MAX PERL_UCHAR_MAX
241#define U8_MIN PERL_UCHAR_MIN
79072805 242
27d4fb96
PP
243#define I16_MAX PERL_SHORT_MAX
244#define I16_MIN PERL_SHORT_MIN
245#define U16_MAX PERL_USHORT_MAX
246#define U16_MIN PERL_USHORT_MIN
79072805 247
c4f23d77 248#if LONGSIZE > 4
27d4fb96
PP
249# define I32_MAX PERL_INT_MAX
250# define I32_MIN PERL_INT_MIN
251# define U32_MAX PERL_UINT_MAX
252# define U32_MIN PERL_UINT_MIN
79072805 253#else
27d4fb96
PP
254# define I32_MAX PERL_LONG_MAX
255# define I32_MIN PERL_LONG_MIN
256# define U32_MAX PERL_ULONG_MAX
257# define U32_MIN PERL_ULONG_MIN
79072805
LW
258#endif
259
5ff3f7a4
GS
260#endif
261
58a9a5d5 262/* log(2) is pretty close to 0.30103, just in case anyone is grepping for it */
fc36a67e
PP
263#define BIT_DIGITS(N) (((N)*146)/485 + 1) /* log2(10) =~ 146/485 */
264#define TYPE_DIGITS(T) BIT_DIGITS(sizeof(T) * 8)
265#define TYPE_CHARS(T) (TYPE_DIGITS(T) + 2) /* sign, NUL */
266
ff68c719 267#define Ctl(ch) ((ch) & 037)
8d063cd8 268
3fe05580
MHM
269/*
270=head1 SV-Body Allocation
271
272=for apidoc Ama|SV*|newSVpvs|const char* s
273Like C<newSVpvn>, but takes a literal string instead of a string/length pair.
274
84bafc02
NC
275=for apidoc Ama|SV*|newSVpvs_flags|const char* s|U32 flags
276Like C<newSVpvn_flags>, but takes a literal string instead of a string/length
277pair.
278
3fe05580
MHM
279=for apidoc Ama|SV*|newSVpvs_share|const char* s
280Like C<newSVpvn_share>, but takes a literal string instead of a string/length
281pair and omits the hash parameter.
282
9dcc53ea
Z
283=for apidoc Am|void|sv_catpvs_flags|SV* sv|const char* s|I32 flags
284Like C<sv_catpvn_flags>, but takes a literal string instead of a
285string/length pair.
286
287=for apidoc Am|void|sv_catpvs_nomg|SV* sv|const char* s
288Like C<sv_catpvn_nomg>, but takes a literal string instead of a
289string/length pair.
290
7b4e7469 291=for apidoc Am|void|sv_catpvs|SV* sv|const char* s
3fe05580
MHM
292Like C<sv_catpvn>, but takes a literal string instead of a string/length pair.
293
9dcc53ea
Z
294=for apidoc Am|void|sv_catpvs_mg|SV* sv|const char* s
295Like C<sv_catpvn_mg>, but takes a literal string instead of a
296string/length pair.
297
7b4e7469 298=for apidoc Am|void|sv_setpvs|SV* sv|const char* s
3fe05580
MHM
299Like C<sv_setpvn>, but takes a literal string instead of a string/length pair.
300
9dcc53ea
Z
301=for apidoc Am|void|sv_setpvs_mg|SV* sv|const char* s
302Like C<sv_setpvn_mg>, but takes a literal string instead of a
303string/length pair.
304
305=for apidoc Am|SV *|sv_setref_pvs|const char* s
306Like C<sv_setref_pvn>, but takes a literal string instead of a
307string/length pair.
308
3fe05580
MHM
309=head1 Memory Management
310
311=for apidoc Ama|char*|savepvs|const char* s
312Like C<savepvn>, but takes a literal string instead of a string/length pair.
313
9dcc53ea
Z
314=for apidoc Ama|char*|savesharedpvs|const char* s
315A version of C<savepvs()> which allocates the duplicate string in memory
316which is shared between threads.
317
3fe05580
MHM
318=head1 GV Functions
319
320=for apidoc Am|HV*|gv_stashpvs|const char* name|I32 create
321Like C<gv_stashpvn>, but takes a literal string instead of a string/length pair.
322
323=head1 Hash Manipulation Functions
324
325=for apidoc Am|SV**|hv_fetchs|HV* tb|const char* key|I32 lval
326Like C<hv_fetch>, but takes a literal string instead of a string/length pair.
327
328=for apidoc Am|SV**|hv_stores|HV* tb|const char* key|NULLOK SV* val
329Like C<hv_store>, but takes a literal string instead of a string/length pair
330and omits the hash parameter.
331
510966aa
Z
332=head1 Lexer interface
333
334=for apidoc Amx|void|lex_stuff_pvs|const char *pv|U32 flags
335
336Like L</lex_stuff_pvn>, but takes a literal string instead of a
337string/length pair.
338
3fe05580
MHM
339=cut
340*/
341
65177ae0 342/* concatenating with "" ensures that only literal strings are accepted as argument */
26bfd924 343#define STR_WITH_LEN(s) ("" s ""), (sizeof(s)-1)
2efa8cc7 344
ba3a79e7
GA
345/* note that STR_WITH_LEN() can't be used as argument to macros or functions that
346 * under some configurations might be macros, which means that it requires the full
347 * Perl_xxx(aTHX_ ...) form for any API calls where it's used.
348 */
349
350/* STR_WITH_LEN() shortcuts */
351#define newSVpvs(str) Perl_newSVpvn(aTHX_ STR_WITH_LEN(str))
84bafc02
NC
352#define newSVpvs_flags(str,flags) \
353 Perl_newSVpvn_flags(aTHX_ STR_WITH_LEN(str), flags)
ba3a79e7 354#define newSVpvs_share(str) Perl_newSVpvn_share(aTHX_ STR_WITH_LEN(str), 0)
9dcc53ea
Z
355#define sv_catpvs_flags(sv, str, flags) \
356 Perl_sv_catpvn_flags(aTHX_ sv, STR_WITH_LEN(str), flags)
357#define sv_catpvs_nomg(sv, str) \
358 Perl_sv_catpvn_flags(aTHX_ sv, STR_WITH_LEN(str), 0)
359#define sv_catpvs(sv, str) \
360 Perl_sv_catpvn_flags(aTHX_ sv, STR_WITH_LEN(str), SV_GMAGIC)
361#define sv_catpvs_mg(sv, str) \
362 Perl_sv_catpvn_flags(aTHX_ sv, STR_WITH_LEN(str), SV_GMAGIC|SV_SMAGIC)
3fe05580 363#define sv_setpvs(sv, str) Perl_sv_setpvn(aTHX_ sv, STR_WITH_LEN(str))
9dcc53ea
Z
364#define sv_setpvs_mg(sv, str) Perl_sv_setpvn_mg(aTHX_ sv, STR_WITH_LEN(str))
365#define sv_setref_pvs(rv, classname, str) \
366 Perl_sv_setref_pvn(aTHX_ rv, classname, STR_WITH_LEN(str))
ba3a79e7 367#define savepvs(str) Perl_savepvn(aTHX_ STR_WITH_LEN(str))
9dcc53ea
Z
368#define savesharedpvs(str) Perl_savesharedpvn(aTHX_ STR_WITH_LEN(str))
369#define gv_stashpvs(str, create) \
370 Perl_gv_stashpvn(aTHX_ STR_WITH_LEN(str), create)
371#define gv_fetchpvs(namebeg, add, sv_type) \
372 Perl_gv_fetchpvn_flags(aTHX_ STR_WITH_LEN(namebeg), add, sv_type)
373#define gv_fetchpvn(namebeg, len, add, sv_type) \
374 Perl_gv_fetchpvn_flags(aTHX_ namebeg, len, add, sv_type)
375#define sv_catxmlpvs(dsv, str, utf8) \
376 Perl_sv_catxmlpvn(aTHX_ dsv, STR_WITH_LEN(str), utf8)
a038e571
NC
377#define hv_fetchs(hv,key,lval) \
378 ((SV **)Perl_hv_common(aTHX_ (hv), NULL, STR_WITH_LEN(key), 0, \
379 (lval) ? (HV_FETCH_JUST_SV | HV_FETCH_LVALUE) \
380 : HV_FETCH_JUST_SV, NULL, 0))
381
382#define hv_stores(hv,key,val) \
383 ((SV **)Perl_hv_common(aTHX_ (hv), NULL, STR_WITH_LEN(key), 0, \
384 (HV_FETCH_ISSTORE|HV_FETCH_JUST_SV), (val), 0))
ba3a79e7 385
510966aa
Z
386#define lex_stuff_pvs(pv,flags) Perl_lex_stuff_pvn(aTHX_ STR_WITH_LEN(pv), flags)
387
b96d8cd9
NC
388#define get_cvs(str, flags) \
389 Perl_get_cvn_flags(aTHX_ STR_WITH_LEN(str), (flags))
5c1737d1 390
954c1994 391/*
ccfc67b7
JH
392=head1 Miscellaneous Functions
393
954c1994
GS
394=for apidoc Am|bool|strNE|char* s1|char* s2
395Test two strings to see if they are different. Returns true or
396false.
397
398=for apidoc Am|bool|strEQ|char* s1|char* s2
399Test two strings to see if they are equal. Returns true or false.
400
401=for apidoc Am|bool|strLT|char* s1|char* s2
402Test two strings to see if the first, C<s1>, is less than the second,
403C<s2>. Returns true or false.
404
405=for apidoc Am|bool|strLE|char* s1|char* s2
406Test two strings to see if the first, C<s1>, is less than or equal to the
407second, C<s2>. Returns true or false.
408
409=for apidoc Am|bool|strGT|char* s1|char* s2
410Test two strings to see if the first, C<s1>, is greater than the second,
411C<s2>. Returns true or false.
412
413=for apidoc Am|bool|strGE|char* s1|char* s2
414Test two strings to see if the first, C<s1>, is greater than or equal to
415the second, C<s2>. Returns true or false.
416
417=for apidoc Am|bool|strnNE|char* s1|char* s2|STRLEN len
418Test two strings to see if they are different. The C<len> parameter
419indicates the number of bytes to compare. Returns true or false. (A
420wrapper for C<strncmp>).
421
422=for apidoc Am|bool|strnEQ|char* s1|char* s2|STRLEN len
423Test two strings to see if they are equal. The C<len> parameter indicates
424the number of bytes to compare. Returns true or false. (A wrapper for
425C<strncmp>).
426
427=cut
428*/
429
8d063cd8
LW
430#define strNE(s1,s2) (strcmp(s1,s2))
431#define strEQ(s1,s2) (!strcmp(s1,s2))
432#define strLT(s1,s2) (strcmp(s1,s2) < 0)
433#define strLE(s1,s2) (strcmp(s1,s2) <= 0)
434#define strGT(s1,s2) (strcmp(s1,s2) > 0)
435#define strGE(s1,s2) (strcmp(s1,s2) >= 0)
436#define strnNE(s1,s2,l) (strncmp(s1,s2,l))
437#define strnEQ(s1,s2,l) (!strncmp(s1,s2,l))
378cc40b 438
36477c24
PP
439#ifdef HAS_MEMCMP
440# define memNE(s1,s2,l) (memcmp(s1,s2,l))
441# define memEQ(s1,s2,l) (!memcmp(s1,s2,l))
442#else
443# define memNE(s1,s2,l) (bcmp(s1,s2,l))
444# define memEQ(s1,s2,l) (!bcmp(s1,s2,l))
445#endif
446
568a785a 447#define memEQs(s1, l, s2) \
8e6497b6 448 (sizeof(s2)-1 == l && memEQ(s1, ("" s2 ""), (sizeof(s2)-1)))
568a785a
NC
449#define memNEs(s1, l, s2) !memEQs(s1, l, s2)
450
bbce6d69
PP
451/*
452 * Character classes.
453 *
454 * Unfortunately, the introduction of locales means that we
455 * can't trust isupper(), etc. to tell the truth. And when
456 * it comes to /\w+/ with tainting enabled, we *must* be able
457 * to trust our character classes.
458 *
459 * Therefore, the default tests in the text of Perl will be
460 * independent of locale. Any code that wants to depend on
461 * the current locale will use the tests that begin with "lc".
462 */
463
2304df62
AD
464#ifdef HAS_SETLOCALE /* XXX Is there a better test for this? */
465# ifndef CTYPE256
466# define CTYPE256
467# endif
468#endif
469
954c1994 470/*
ccfc67b7
JH
471
472=head1 Character classes
8a58bdcf
KW
473There are three variants for all the functions in this section. The base ones
474operate using the character set of the platform Perl is running on. The ones
475with an C<_A> suffix operate on the ASCII character set, and the ones with an
476C<_L1> suffix operate on the full Latin1 character set. All are unaffected by
181ab5fd 477locale and by C<use bytes>.
8a58bdcf
KW
478
479For ASCII platforms, the base function with no suffix and the one with the
480C<_A> suffix are identical. The function with the C<_L1> suffix imposes the
481Latin-1 character set onto the platform. That is, the code points that are
482ASCII are unaffected, since ASCII is a subset of Latin-1. But the non-ASCII
483code points are treated as if they are Latin-1 characters. For example,
484C<isSPACE_L1()> will return true when called with the code point 0xA0, which is
485the Latin-1 NO-BREAK SPACE.
486
487For EBCDIC platforms, the base function with no suffix and the one with the
488C<_L1> suffix should be identical, since, as of this writing, the EBCDIC code
489pages that Perl knows about all are equivalent to Latin-1. The function that
490ends in an C<_A> suffix will not return true unless the specified character also
491has an ASCII equivalent.
ccfc67b7 492
954c1994 493=for apidoc Am|bool|isALPHA|char ch
2787a470 494Returns a boolean indicating whether the specified character is an
06dc89af 495alphabetic character in the platform's native character set.
8a58bdcf
KW
496See the L<top of this section|/Character classes> for an explanation of variants
497C<isALPHA_A> and C<isALPHA_L1>.
498
499=for apidoc Am|bool|isASCII|char ch
500Returns a boolean indicating whether the specified character is one of the 128
501characters in the ASCII character set. On non-ASCII platforms, it is if this
502character corresponds to an ASCII character. Variants C<isASCII_A()> and
503C<isASCII_L1()> are identical to C<isASCII()>.
954c1994 504
954c1994 505=for apidoc Am|bool|isDIGIT|char ch
2787a470 506Returns a boolean indicating whether the specified character is a
06dc89af 507digit in the platform's native character set.
8a58bdcf 508Variants C<isDIGIT_A> and C<isDIGIT_L1> are identical to C<isDIGIT>.
954c1994 509
0c82b6df 510=for apidoc Am|bool|isLOWER|char ch
2787a470 511Returns a boolean indicating whether the specified character is a
0c82b6df 512lowercase character in the platform's native character set.
8a58bdcf
KW
513See the L<top of this section|/Character classes> for an explanation of variants
514C<isLOWER_A> and C<isLOWER_L1>.
0c82b6df 515
c99e91e9 516=for apidoc Am|bool|isOCTAL|char ch
2787a470 517Returns a boolean indicating whether the specified character is an
06dc89af 518octal digit, [0-7] in the platform's native character set.
8a58bdcf 519Variants C<isOCTAL_A> and C<isOCTAL_L1> are identical to C<isOCTAL>.
c99e91e9 520
0c82b6df 521=for apidoc Am|bool|isSPACE|char ch
2787a470 522Returns a boolean indicating whether the specified character is a
8a58bdcf
KW
523whitespace character in the platform's native character set. This is the same
524as what C<\s> matches in a regular expression.
525See the L<top of this section|/Character classes> for an explanation of variants
526C<isSPACE_A> and C<isSPACE_L1>.
0c82b6df 527
954c1994 528=for apidoc Am|bool|isUPPER|char ch
2787a470 529Returns a boolean indicating whether the specified character is an
06dc89af 530uppercase character in the platform's native character set.
8a58bdcf
KW
531See the L<top of this section|/Character classes> for an explanation of variants
532C<isUPPER_A> and C<isUPPER_L1>.
954c1994 533
8eea39dd
KW
534=for apidoc Am|bool|isWORDCHAR|char ch
535Returns a boolean indicating whether the specified character is a
536character that is any of: alphabetic, numeric, or an underscore. This is the
537same as what C<\w> matches in a regular expression.
538C<isALNUM()> is a synonym provided for backward compatibility. Note that it
539does not have the standard C language meaning of alphanumeric, since it matches
540an underscore and the standard meaning does not.
8a58bdcf
KW
541See the L<top of this section|/Character classes> for an explanation of variants
542C<isWORDCHAR_A> and C<isWORDCHAR_L1>.
543
544=for apidoc Am|bool|isXDIGIT|char ch
545Returns a boolean indicating whether the specified character is a hexadecimal
546digit, [0-9A-Fa-f]. Variants C<isXDIGIT_A()> and C<isXDIGIT_L1()> are
547identical to C<isXDIGIT()>.
8eea39dd 548
95a59cab
YO
549=for apidoc Am|U8|READ_XDIGIT|char str*
550Returns the value of a hex digit and advances the string pointer.
551Behaviour is only well defined when isXDIGIT(*str) is true.
552
e7c1e6c1
KW
553=head1 Character case changing
554
954c1994 555=for apidoc Am|char|toUPPER|char ch
06dc89af
KW
556Converts the specified character to uppercase in the platform's native
557character set, if possible; otherwise returns the input character itself.
954c1994
GS
558
559=for apidoc Am|char|toLOWER|char ch
06dc89af
KW
560Converts the specified character to lowercase in the platform's native
561character set, if possible; otherwise returns the input character itself.
954c1994
GS
562
563=cut
353c9b6f 564
8a58bdcf 565Note that these macros are repeated in Devel::PPPort, so should also be
62fa66b6
KW
566patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc
567
954c1994
GS
568*/
569
7c062697
KW
570/* Specify the widest unsigned type on the platform. Use U64TYPE because U64
571 * is known only in the perl core, and this macro can be called from outside
572 * that */
573#ifdef HAS_QUAD
574# define WIDEST_UTYPE U64TYPE
575#else
576# define WIDEST_UTYPE U32
577#endif
578
3912bc88
KW
579/* FITS_IN_8_BITS(c) returns true if c doesn't have a bit set other than in
580 * the lower 8. It is designed to be hopefully bomb-proof, making sure that no
581 * bits of information are lost even on a 64-bit machine, but to get the
582 * compiler to optimize it out if possible. This is because Configure makes
583 * sure that the machine has an 8-bit byte, so if c is stored in a byte, the
584 * sizeof() guarantees that this evaluates to a constant true at compile time.
220c71bf
KW
585 */
586#define FITS_IN_8_BITS(c) ((sizeof(c) == 1) || !(((WIDEST_UTYPE)(c)) & ~0xFF))
cf301eb7 587
41f43cc2 588#ifdef EBCDIC
ff7eba81 589# define isASCII(c) (FITS_IN_8_BITS(c) && (NATIVE_TO_UNI((U8) (c)) < 128))
41f43cc2
KW
590#else
591# define isASCII(c) ((WIDEST_UTYPE)(c) < 128)
592#endif
593
ec79cfbb 594#define isASCII_A(c) isASCII(c)
3509bb53 595#define isASCII_L1(c) isASCII(c)
6515cb76 596
c2da0b36
KW
597/* The lower 3 bits in both the ASCII and EBCDIC representations of '0' are 0,
598 * and the 8 possible permutations of those bits exactly comprise the 8 octal
599 * digits */
3d33b3e3 600#define isOCTAL_A(c) cBOOL(FITS_IN_8_BITS(c) && (0xF8 & (c)) == '0')
c2da0b36 601
ec79cfbb 602/* ASCII range only */
9fb1bf9d 603#ifdef H_PERL /* If have access to perl.h, lookup in its table */
f4cdb42c 604
66c17564
KW
605/* Character class numbers. For internal core Perl use only. These are used
606 * in PL_charclass[] and the ones up through the one that corresponds to
607 * <_HIGHEST_REGCOMP_DOT_H_SYNC> are used by regcomp.h. These use names used
608 * in l1_char_class_tab.h but their actual definitions are here. If that has a
609 * name not used here, it won't compile. */
f4cdb42c
KW
610# define _CC_WORDCHAR 0
611# define _CC_SPACE 1
612# define _CC_DIGIT 2
613# define _CC_ALNUMC 3
614# define _CC_ALPHA 4
615# define _CC_ASCII 5
616# define _CC_CNTRL 6
617# define _CC_GRAPH 7
618# define _CC_LOWER 8
619# define _CC_PRINT 9
620# define _CC_PUNCT 10
621# define _CC_UPPER 11
622# define _CC_XDIGIT 12
623# define _CC_PSXSPC 13
624# define _CC_BLANK 14
08ada29e
KW
625# define _HIGHEST_REGCOMP_DOT_H_SYNC _CC_BLANK
626
f4cdb42c
KW
627# define _CC_IDFIRST 15
628# define _CC_CHARNAME_CONT 16
629# define _CC_NONLATIN1_FOLD 17
630# define _CC_QUOTEMETA 18
b72a36d4 631# define _CC_NON_FINAL_FOLD 19
37ede926
KW
632# define _CC_IS_IN_SOME_FOLD 20
633/* Unused: 21-31
f4cdb42c
KW
634 * If more bits are needed, one could add a second word for non-64bit
635 * QUAD_IS_INT systems, using some #ifdefs to distinguish between having a 2nd
37ede926
KW
636 * word or not. The IS_IN_SOME_FOLD bit is the most easily expendable, as it
637 * is used only for optimization (as of this writing), and differs in the
638 * Latin1 range from the ALPHA bit only in two relatively unimportant
639 * characters: the masculine and feminine ordinal indicators */
96ac0975
NC
640
641# ifdef DOINIT
642EXTCONST U32 PL_charclass[] = {
643# include "l1_char_class_tab.h"
644};
645
646# else /* ! DOINIT */
647EXTCONST U32 PL_charclass[];
648# endif
649
265c1f46 650 /* The 1U keeps Solaris from griping when shifting sets the uppermost bit */
430b7c70
KW
651# define _CC_mask(classnum) (1U << (classnum))
652# define _generic_isCC(c, classnum) cBOOL(FITS_IN_8_BITS(c) \
653 && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_mask(classnum)))
4eeeb416 654
f4cdb42c
KW
655 /* The mask for the _A versions of the macros; it just adds in the bit for
656 * ASCII. */
657# define _CC_mask_A(classnum) (_CC_mask(classnum) | _CC_mask(_CC_ASCII))
658
659 /* The _A version makes sure that both the desired bit and the ASCII bit
660 * are present */
661# define _generic_isCC_A(c, classnum) (FITS_IN_8_BITS(c) \
662 && ((PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_mask_A(classnum)) \
663 == _CC_mask_A(classnum)))
664
665# define isALNUMC_A(c) _generic_isCC_A(c, _CC_ALNUMC)
666# define isALPHA_A(c) _generic_isCC_A(c, _CC_ALPHA)
667# define isBLANK_A(c) _generic_isCC_A(c, _CC_BLANK)
668# define isCNTRL_A(c) _generic_isCC_A(c, _CC_CNTRL)
669# define isDIGIT_A(c) _generic_isCC(c, _CC_DIGIT)
670# define isGRAPH_A(c) _generic_isCC_A(c, _CC_GRAPH)
671# define isLOWER_A(c) _generic_isCC_A(c, _CC_LOWER)
672# define isPRINT_A(c) _generic_isCC_A(c, _CC_PRINT)
673# define isPSXSPC_A(c) _generic_isCC_A(c, _CC_PSXSPC)
674# define isPUNCT_A(c) _generic_isCC_A(c, _CC_PUNCT)
675# define isSPACE_A(c) _generic_isCC_A(c, _CC_SPACE)
676# define isUPPER_A(c) _generic_isCC_A(c, _CC_UPPER)
677# define isWORDCHAR_A(c) _generic_isCC_A(c, _CC_WORDCHAR)
678# define isXDIGIT_A(c) _generic_isCC(c, _CC_XDIGIT)
679# define isIDFIRST_A(c) _generic_isCC_A(c, ( _CC_IDFIRST))
680
fb9bfbf7
KW
681 /* Either participates in a fold with a character above 255, or is a
682 * multi-char fold */
430b7c70
KW
683# define _HAS_NONLATIN1_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(c) ((! cBOOL(FITS_IN_8_BITS(c))) || (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_mask(_CC_NONLATIN1_FOLD)))
684
4eeeb416 685# define _isQUOTEMETA(c) _generic_isCC(c, _CC_QUOTEMETA)
26faadbd
KW
686# define _IS_NON_FINAL_FOLD_ONLY_FOR_USE_BY_REGCOMP_DOT_C(c) \
687 _generic_isCC(c, _CC_NON_FINAL_FOLD)
37ede926
KW
688# define _IS_IN_SOME_FOLD_ONLY_FOR_USE_BY_REGCOMP_DOT_C(c) \
689 _generic_isCC(c, _CC_IS_IN_SOME_FOLD)
9fb1bf9d 690#else /* No perl.h. */
9fb1bf9d
KW
691# ifdef EBCDIC
692# define isALNUMC_A(c) (isASCII(c) && isALNUMC(c))
693# define isALPHA_A(c) (isASCII(c) && isALPHA(c))
694# define isBLANK_A(c) (isASCII(c) && isBLANK(c))
695# define isCNTRL_A(c) (isASCII(c) && isCNTRL(c))
696# define isDIGIT_A(c) (isASCII(c) && isDIGIT(c))
697# define isGRAPH_A(c) (isASCII(c) && isGRAPH(c))
698# define isIDFIRST_A(c) (isASCII(c) && isIDFIRST(c))
699# define isLOWER_A(c) (isASCII(c) && isLOWER(c))
700# define isPRINT_A(c) (isASCII(c) && isPRINT(c))
701# define isPSXSPC_A(c) (isASCII(c) && isPSXSPC(c))
702# define isPUNCT_A(c) (isASCII(c) && isPUNCT(c))
703# define isSPACE_A(c) (isASCII(c) && isSPACE(c))
704# define isUPPER_A(c) (isASCII(c) && isUPPER(c))
705# define isWORDCHAR_A(c) (isASCII(c) && isWORDCHAR(c))
706# define isXDIGIT_A(c) (isASCII(c) && isXDIGIT(c))
707# else /* ASCII platform, no perl.h */
708# define isALNUMC_A(c) (isALPHA_A(c) || isDIGIT_A(c))
709# define isALPHA_A(c) (isUPPER_A(c) || isLOWER_A(c))
710# define isBLANK_A(c) ((c) == ' ' || (c) == '\t')
ff7eba81 711# define isCNTRL_A(c) (FITS_IN_8_BITS(c) && ((U8) (c) < ' ' || (c) == 127))
2181da29 712# define isDIGIT_A(c) ((c) <= '9' && (c) >= '0')
9fb1bf9d
KW
713# define isGRAPH_A(c) (isWORDCHAR_A(c) || isPUNCT_A(c))
714# define isIDFIRST_A(c) (isALPHA_A(c) || (c) == '_')
715# define isLOWER_A(c) ((c) >= 'a' && (c) <= 'z')
716# define isPRINT_A(c) (((c) >= 32 && (c) < 127))
717# define isPSXSPC_A(c) (isSPACE_A(c) || (c) == '\v')
718# define isPUNCT_A(c) (((c) >= 33 && (c) <= 47) || ((c) >= 58 && (c) <= 64) || ((c) >= 91 && (c) <= 96) || ((c) >= 123 && (c) <= 126))
719# define isSPACE_A(c) ((c) == ' ' || (c) == '\t' || (c) == '\n' || (c) =='\r' || (c) == '\f')
2181da29 720# define isUPPER_A(c) ((c) <= 'Z' && (c) >= 'A')
9fb1bf9d 721# define isWORDCHAR_A(c) (isALPHA_A(c) || isDIGIT_A(c) || (c) == '_')
2181da29 722# define isXDIGIT_A(c) (isDIGIT_A(c) || ((c) >= 'a' && (c) <= 'f') || ((c) <= 'F' && (c) >= 'A'))
9fb1bf9d
KW
723# endif
724#endif /* ASCII range definitions */
ec79cfbb
KW
725
726/* Latin1 definitions */
8a58bdcf 727#ifdef H_PERL
f4cdb42c
KW
728# define isALNUMC_L1(c) _generic_isCC(c, _CC_ALNUMC)
729# define isALPHA_L1(c) _generic_isCC(c, _CC_ALPHA)
730# define isBLANK_L1(c) _generic_isCC(c, _CC_BLANK)
e5dcd934 731
8a58bdcf 732/* continuation character for legal NAME in \N{NAME} */
4eeeb416 733# define isCHARNAME_CONT(c) _generic_isCC(c, _CC_CHARNAME_CONT)
e5dcd934 734
f4cdb42c
KW
735# define isCNTRL_L1(c) _generic_isCC(c, _CC_CNTRL)
736# define isGRAPH_L1(c) _generic_isCC(c, _CC_GRAPH)
737# define isLOWER_L1(c) _generic_isCC(c, _CC_LOWER)
738# define isPRINT_L1(c) _generic_isCC(c, _CC_PRINT)
739# define isPSXSPC_L1(c) _generic_isCC(c, _CC_PSXSPC)
740# define isPUNCT_L1(c) _generic_isCC(c, _CC_PUNCT)
741# define isSPACE_L1(c) _generic_isCC(c, _CC_SPACE)
742# define isUPPER_L1(c) _generic_isCC(c, _CC_UPPER)
743# define isWORDCHAR_L1(c) _generic_isCC(c, _CC_WORDCHAR)
744# define isIDFIRST_L1(c) _generic_isCC(c, _CC_IDFIRST)
8a58bdcf
KW
745#else /* No access to perl.h. Only a few provided here, just in case needed
746 * for backwards compatibility */
747 /* ALPHAU includes Unicode semantics for latin1 characters. It has an extra
748 * >= AA test to speed up ASCII-only tests at the expense of the others */
749# define isALPHA_L1(c) (isALPHA(c) || (NATIVE_TO_UNI((U8) c) >= 0xAA \
750 && ((NATIVE_TO_UNI((U8) c) >= 0xC0 \
751 && NATIVE_TO_UNI((U8) c) != 0xD7 && NATIVE_TO_UNI((U8) c) != 0xF7) \
752 || NATIVE_TO_UNI((U8) c) == 0xAA \
753 || NATIVE_TO_UNI((U8) c) == 0xB5 \
754 || NATIVE_TO_UNI((U8) c) == 0xBA)))
755# define isCHARNAME_CONT(c) (isALNUM_L1(c) || (c) == ' ' || (c) == '-' || (c) == '(' || (c) == ')' || (c) == ':' || NATIVE_TO_UNI((U8) c) == 0xA0)
756#endif
757
758/* Macros for backwards compatibility and for completeness when the ASCII and
759 * Latin1 values are identical */
ec79cfbb
KW
760#define isALNUM(c) isWORDCHAR(c)
761#define isALNUMU(c) isWORDCHAR_L1(c)
8a58bdcf
KW
762#define isALPHAU(c) isALPHA_L1(c)
763#define isDIGIT_L1(c) isDIGIT_A(c)
764#define isOCTAL(c) isOCTAL_A(c)
765#define isOCTAL_L1(c) isOCTAL_A(c)
766#define isXDIGIT_L1(c) isXDIGIT_A(c)
767
768/* Macros that differ between EBCDIC and ASCII. Where C89 defines a function,
769 * that is used in the EBCDIC form, because in EBCDIC we do not do locales:
770 * therefore can use native functions. For those where C89 doesn't define a
771 * function, use our function, assuming that the EBCDIC code page is isomorphic
772 * with Latin1, which the three currently recognized by Perl are. Some libc's
773 * have an isblank(), but it's not guaranteed. */
9d116dd7 774#ifdef EBCDIC
b8c5462f 775# define isALNUMC(c) isalnum(c)
271b5cb3 776# define isALPHA(c) isalpha(c)
0bc70f7f 777# define isBLANK(c) ((c) == ' ' || (c) == '\t' || NATIVE_TO_UNI(c) == 0xA0)
b8c5462f 778# define isCNTRL(c) iscntrl(c)
9916188f 779# define isDIGIT(c) isdigit(c)
b8c5462f 780# define isGRAPH(c) isgraph(c)
8a58bdcf 781# define isIDFIRST(c) (isALPHA(c) || (c) == '_')
7d79683e 782# define isLOWER(c) islower(c)
9d116dd7 783# define isPRINT(c) isprint(c)
56e4b611 784# define isPSXSPC(c) isspace(c)
b8c5462f 785# define isPUNCT(c) ispunct(c)
174b5e76 786# define isSPACE(c) (isPSXSPC(c) && (c) != '\v')
7d79683e 787# define isUPPER(c) isupper(c)
b8c5462f 788# define isXDIGIT(c) isxdigit(c)
8a58bdcf 789# define isWORDCHAR(c) (isalnum(c) || (c) == '_')
9d116dd7 790# define toLOWER(c) tolower(c)
7d79683e 791# define toUPPER(c) toupper(c)
ec79cfbb 792#else /* Not EBCDIC: ASCII-only matching */
e58b1dde 793# define isALNUMC(c) isALNUMC_A(c) /* Mnemonic: "C's alnum" = alpha + digit */
271b5cb3 794# define isALPHA(c) isALPHA_A(c)
ec79cfbb
KW
795# define isBLANK(c) isBLANK_A(c)
796# define isCNTRL(c) isCNTRL_A(c)
797# define isDIGIT(c) isDIGIT_A(c)
798# define isGRAPH(c) isGRAPH_A(c)
8a58bdcf 799# define isIDFIRST(c) isIDFIRST_A(c)
ec79cfbb
KW
800# define isLOWER(c) isLOWER_A(c)
801# define isPRINT(c) isPRINT_A(c)
802# define isPSXSPC(c) isPSXSPC_A(c)
803# define isPUNCT(c) isPUNCT_A(c)
804# define isSPACE(c) isSPACE_A(c)
805# define isUPPER(c) isUPPER_A(c)
8a58bdcf 806# define isWORDCHAR(c) isWORDCHAR_A(c)
ec79cfbb 807# define isXDIGIT(c) isXDIGIT_A(c)
00f254e2 808
8a58bdcf
KW
809 /* ASCII casing. These could also be written as
810 #define toLOWER(c) (isASCII(c) ? toLOWER_LATIN1(c) : (c))
811 #define toUPPER(c) (isASCII(c) ? toUPPER_LATIN1_MOD(c) : (c))
812 which uses table lookup and mask instead of subtraction. (This would
813 work because the _MOD does not apply in the ASCII range) */
9d116dd7 814# define toLOWER(c) (isUPPER(c) ? (c) + ('a' - 'A') : (c))
7d79683e 815# define toUPPER(c) (isLOWER(c) ? (c) - ('a' - 'A') : (c))
9d116dd7 816#endif
bbce6d69 817
1a0901db
KW
818
819/* Use table lookup for speed; return error character for input
820 * out-of-range */
821#define toLOWER_LATIN1(c) (FITS_IN_8_BITS(c) \
822 ? UNI_TO_NATIVE(PL_latin1_lc[ \
823 NATIVE_TO_UNI( (U8) (c)) ]) \
824 : UNICODE_REPLACEMENT)
825/* Modified uc. Is correct uc except for three non-ascii chars which are
826 * all mapped to one of them, and these need special handling; error
827 * character for input out-of-range */
828#define toUPPER_LATIN1_MOD(c) (FITS_IN_8_BITS(c) \
829 ? UNI_TO_NATIVE(PL_mod_latin1_uc[ \
830 NATIVE_TO_UNI( (U8) (c)) ]) \
831 : UNICODE_REPLACEMENT)
832
bbce6d69
PP
833#ifdef USE_NEXT_CTYPE
834
835# define isALNUM_LC(c) \
37bd1396 836 (NXIsAlNum((unsigned int)(c)) || (char)(c) == '_')
ff68c719
PP
837# define isIDFIRST_LC(c) \
838 (NXIsAlpha((unsigned int)(c)) || (char)(c) == '_')
839# define isALPHA_LC(c) NXIsAlpha((unsigned int)(c))
07315176 840# define isASCII_LC(c) isASCII((unsigned int)(c))
67addccf 841# define isBLANK_LC(c) isBLANK((unsigned int)(c))
ff68c719
PP
842# define isSPACE_LC(c) NXIsSpace((unsigned int)(c))
843# define isDIGIT_LC(c) NXIsDigit((unsigned int)(c))
844# define isUPPER_LC(c) NXIsUpper((unsigned int)(c))
845# define isLOWER_LC(c) NXIsLower((unsigned int)(c))
37bd1396 846# define isALNUMC_LC(c) NXIsAlNum((unsigned int)(c))
b8c5462f
JH
847# define isCNTRL_LC(c) NXIsCntrl((unsigned int)(c))
848# define isGRAPH_LC(c) NXIsGraph((unsigned int)(c))
ff68c719 849# define isPRINT_LC(c) NXIsPrint((unsigned int)(c))
b8c5462f 850# define isPUNCT_LC(c) NXIsPunct((unsigned int)(c))
ff68c719
PP
851# define toUPPER_LC(c) NXToUpper((unsigned int)(c))
852# define toLOWER_LC(c) NXToLower((unsigned int)(c))
bbce6d69
PP
853
854#else /* !USE_NEXT_CTYPE */
b8c5462f 855
bbce6d69
PP
856# if defined(CTYPE256) || (!defined(isascii) && !defined(HAS_ISASCII))
857
851131d3
KW
858/* Note that the foo_LC() macros in this case generally are defined only on
859 * code points 0-256, and give undefined, unwarned results if called with
860 * values outside that range */
861
b8c5462f 862# define isALNUM_LC(c) (isalnum((unsigned char)(c)) || (char)(c) == '_')
ff68c719 863# define isIDFIRST_LC(c) (isalpha((unsigned char)(c)) || (char)(c) == '_')
bbce6d69 864# define isALPHA_LC(c) isalpha((unsigned char)(c))
07315176
KW
865# ifdef HAS_ISASCII
866# define isASCII_LC(c) isascii((unsigned char)(c))
867# else
868# define isASCII_LC(c) isASCII((unsigned char)(c))
869# endif
67addccf
KW
870# ifdef HAS_ISBLANK
871# define isBLANK_LC(c) isblank((unsigned char)(c))
872# else
873# define isBLANK_LC(c) isBLANK((unsigned char)(c))
874# endif
bbce6d69
PP
875# define isSPACE_LC(c) isspace((unsigned char)(c))
876# define isDIGIT_LC(c) isdigit((unsigned char)(c))
877# define isUPPER_LC(c) isupper((unsigned char)(c))
878# define isLOWER_LC(c) islower((unsigned char)(c))
b8c5462f
JH
879# define isALNUMC_LC(c) isalnum((unsigned char)(c))
880# define isCNTRL_LC(c) iscntrl((unsigned char)(c))
881# define isGRAPH_LC(c) isgraph((unsigned char)(c))
bbce6d69 882# define isPRINT_LC(c) isprint((unsigned char)(c))
b8c5462f 883# define isPUNCT_LC(c) ispunct((unsigned char)(c))
bbce6d69
PP
884# define toUPPER_LC(c) toupper((unsigned char)(c))
885# define toLOWER_LC(c) tolower((unsigned char)(c))
886
887# else
888
3609ea0d 889# define isALNUM_LC(c) (isascii(c) && (isalnum(c) || (c) == '_'))
bbce6d69
PP
890# define isIDFIRST_LC(c) (isascii(c) && (isalpha(c) || (c) == '_'))
891# define isALPHA_LC(c) (isascii(c) && isalpha(c))
07315176 892# define isASCII_LC(c) isascii(c)
67addccf
KW
893# ifdef HAS_ISBLANK
894# define isBLANK_LC(c) (isascii(c) && isblank(c))
895# else
896# define isBLANK_LC(c) isBLANK(c)
897# endif
bbce6d69
PP
898# define isSPACE_LC(c) (isascii(c) && isspace(c))
899# define isDIGIT_LC(c) (isascii(c) && isdigit(c))
900# define isUPPER_LC(c) (isascii(c) && isupper(c))
901# define isLOWER_LC(c) (isascii(c) && islower(c))
b8c5462f
JH
902# define isALNUMC_LC(c) (isascii(c) && isalnum(c))
903# define isCNTRL_LC(c) (isascii(c) && iscntrl(c))
904# define isGRAPH_LC(c) (isascii(c) && isgraph(c))
bbce6d69 905# define isPRINT_LC(c) (isascii(c) && isprint(c))
b8c5462f 906# define isPUNCT_LC(c) (isascii(c) && ispunct(c))
bbce6d69
PP
907# define toUPPER_LC(c) toupper(c)
908# define toLOWER_LC(c) tolower(c)
909
910# endif
a0d0e21e 911#endif /* USE_NEXT_CTYPE */
55204971 912
aaa51d5e 913#define isPSXSPC_LC(c) (isSPACE_LC(c) || (c) == '\v')
aaa51d5e 914
66c17564
KW
915/* For internal core Perl use only. If the input is Latin1, use the Latin1
916 * macro; otherwise use the function. Won't compile if 'c' isn't unsigned, as
917 * won't match function prototype. The macros do bounds checking, so have
918 * duplicate checks here, so could create versions of the macros that don't,
919 * but experiments show that gcc optimizes them out anyway. */
920
921/* Note that all ignore 'use bytes' */
922#define _generic_uni(classnum, function, c) ((c) < 256 \
aea0c08c
KW
923 ? _generic_isCC(c, classnum) \
924 : function(c))
eb0052d1 925
66c17564 926#define isWORDCHAR_uni(c) _generic_uni(_CC_WORDCHAR, is_uni_alnum, c)
1545c4e9 927#define isALNUM_uni(c) isWORDCHAR_uni(c)
66c17564
KW
928#define isBLANK_uni(c) _generic_uni(_CC_BLANK, is_uni_blank, c)
929#define isIDFIRST_uni(c) _generic_uni(_CC_IDFIRST, is_uni_idfirst, c)
930#define isALPHA_uni(c) _generic_uni(_CC_ALPHA, is_uni_alpha, c)
931#define isSPACE_uni(c) _generic_uni(_CC_SPACE, is_uni_space, c)
932#define isDIGIT_uni(c) _generic_uni(_CC_DIGIT, is_uni_digit, c)
933#define isUPPER_uni(c) _generic_uni(_CC_UPPER, is_uni_upper, c)
934#define isLOWER_uni(c) _generic_uni(_CC_LOWER, is_uni_lower, c)
aea0c08c
KW
935#define isASCII_uni(c) isASCII(c)
936
7b952154 937/* All controls are in Latin1 */
aea0c08c
KW
938#define isCNTRL_uni(c) isCNTRL_L1(c)
939
66c17564
KW
940#define isGRAPH_uni(c) _generic_uni(_CC_GRAPH, is_uni_graph, c)
941#define isPRINT_uni(c) _generic_uni(_CC_PRINT, is_uni_print, c)
942#define isPUNCT_uni(c) _generic_uni(_CC_PUNCT, is_uni_punct, c)
943#define isXDIGIT_uni(c) _generic_uni(_CC_XDIGIT, is_uni_xdigit, c)
b18192e9
KW
944
945/* Posix and regular space differ only in U+000B, which is in Latin1 */
66c17564 946#define isPSXSPC_uni(c) _generic_uni(_CC_PSXSPC, is_uni_space, c)
b18192e9 947
a2a2844f
JH
948#define toUPPER_uni(c,s,l) to_uni_upper(c,s,l)
949#define toTITLE_uni(c,s,l) to_uni_title(c,s,l)
950#define toLOWER_uni(c,s,l) to_uni_lower(c,s,l)
b8d68ded 951#define toFOLD_uni(c,s,l) to_uni_fold(c,s,l)
a0ed51b3 952
9041c2e3
NIS
953#define isALNUM_LC_uvchr(c) (c < 256 ? isALNUM_LC(c) : is_uni_alnum_lc(c))
954#define isIDFIRST_LC_uvchr(c) (c < 256 ? isIDFIRST_LC(c) : is_uni_idfirst_lc(c))
955#define isALPHA_LC_uvchr(c) (c < 256 ? isALPHA_LC(c) : is_uni_alpha_lc(c))
956#define isSPACE_LC_uvchr(c) (c < 256 ? isSPACE_LC(c) : is_uni_space_lc(c))
957#define isDIGIT_LC_uvchr(c) (c < 256 ? isDIGIT_LC(c) : is_uni_digit_lc(c))
958#define isUPPER_LC_uvchr(c) (c < 256 ? isUPPER_LC(c) : is_uni_upper_lc(c))
959#define isLOWER_LC_uvchr(c) (c < 256 ? isLOWER_LC(c) : is_uni_lower_lc(c))
9041c2e3
NIS
960#define isCNTRL_LC_uvchr(c) (c < 256 ? isCNTRL_LC(c) : is_uni_cntrl_lc(c))
961#define isGRAPH_LC_uvchr(c) (c < 256 ? isGRAPH_LC(c) : is_uni_graph_lc(c))
962#define isPRINT_LC_uvchr(c) (c < 256 ? isPRINT_LC(c) : is_uni_print_lc(c))
963#define isPUNCT_LC_uvchr(c) (c < 256 ? isPUNCT_LC(c) : is_uni_punct_lc(c))
a0ed51b3 964
aaa51d5e
JF
965#define isPSXSPC_LC_uni(c) (isSPACE_LC_uni(c) ||(c) == '\f')
966#define isBLANK_LC_uni(c) isBLANK(c) /* could be wrong */
967
66c17564
KW
968/* For internal core Perl use only. If the input is in the Latin1 range, use
969 * the Latin1 macro, after converting from utf8 if necessary; otherwise use the
970 * function. This relies on the fact that ASCII characters have the same
971 * representation whether utf8 or not. Note that all assume that the utf8 has
972 * been validated, and ignore 'use bytes' */
973#define _generic_utf8(classnum, function, p) (UTF8_IS_INVARIANT(*(p)) \
aea0c08c 974 ? _generic_isCC(*(p), classnum) \
7d7a6efc 975 : (UTF8_IS_DOWNGRADEABLE_START(*(p))) \
aea0c08c
KW
976 ? _generic_isCC( \
977 TWO_BYTE_UTF8_TO_UNI(*(p), \
978 *((p)+1 )), \
979 classnum) \
980 : (function)(aTHX_ p))
c8362b00 981
66c17564 982#define isWORDCHAR_utf8(p) _generic_utf8(_CC_WORDCHAR, Perl_is_utf8_alnum, p)
1545c4e9 983#define isALNUM_utf8(p) isWORDCHAR_utf8(p) /* back compat */
e5dcd934 984
c11ff943
KW
985/* To prevent S_scan_word in toke.c from hanging, we have to make sure that
986 * IDFIRST is an alnum. See
c8362b00 987 * http://rt.perl.org/rt3/Ticket/Display.html?id=74022 for more detail than you
9479a769
KW
988 * ever wanted to know about. XXX It is unclear if this should extend to
989 * isIDFIRST_uni() which it hasn't so far. (In the ASCII range, there isn't a
990 * difference.) This used to be not the XID version, but we decided to go with
991 * the more modern Unicode definition */
66c17564 992#define isIDFIRST_utf8(p) _generic_utf8(_CC_IDFIRST, Perl__is_utf8__perl_idstart, p)
e5dcd934 993
66c17564
KW
994#define isIDCONT_utf8(p) _generic_utf8(_CC_WORDCHAR, Perl_is_utf8_xidcont, p)
995#define isALPHA_utf8(p) _generic_utf8(_CC_ALPHA, Perl_is_utf8_alpha, p)
996#define isBLANK_utf8(p) _generic_utf8(_CC_BLANK, Perl_is_utf8_blank, p)
997#define isSPACE_utf8(p) _generic_utf8(_CC_SPACE, Perl_is_utf8_space, p)
998#define isDIGIT_utf8(p) _generic_utf8(_CC_DIGIT, Perl_is_utf8_digit, p)
999#define isUPPER_utf8(p) _generic_utf8(_CC_UPPER, Perl_is_utf8_upper, p)
1000#define isLOWER_utf8(p) _generic_utf8(_CC_LOWER, Perl_is_utf8_lower, p)
aea0c08c 1001
141d8bad 1002/* Because ASCII is invariant under utf8, the non-utf8 macro works */
bce3d7a4 1003#define isASCII_utf8(p) isASCII(*p)
aea0c08c 1004
66c17564
KW
1005#define isCNTRL_utf8(p) _generic_utf8(_CC_CNTRL, Perl_is_utf8_cntrl, p)
1006#define isGRAPH_utf8(p) _generic_utf8(_CC_GRAPH, Perl_is_utf8_graph, p)
1007#define isPRINT_utf8(p) _generic_utf8(_CC_PRINT, Perl_is_utf8_print, p)
1008#define isPUNCT_utf8(p) _generic_utf8(_CC_PUNCT, Perl_is_utf8_punct, p)
1009#define isXDIGIT_utf8(p) _generic_utf8(_CC_XDIGIT, Perl_is_utf8_xdigit, p)
a2a2844f
JH
1010#define toUPPER_utf8(p,s,l) to_utf8_upper(p,s,l)
1011#define toTITLE_utf8(p,s,l) to_utf8_title(p,s,l)
1012#define toLOWER_utf8(p,s,l) to_utf8_lower(p,s,l)
a0ed51b3 1013
c8362b00
KW
1014/* Posix and regular space differ only in U+000B, which is in ASCII (and hence
1015 * Latin1 */
66c17564 1016#define isPSXSPC_utf8(p) _generic_utf8(_CC_PSXSPC, Perl_is_utf8_space, p)
2e8adce6 1017
4b88fb76
KW
1018#define isALNUM_LC_utf8(p) isALNUM_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1019#define isIDFIRST_LC_utf8(p) isIDFIRST_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1020#define isALPHA_LC_utf8(p) isALPHA_LC_uvchr(valid_utf8_to_uvchr(p, 0))
bdd8600f 1021#define isBLANK_LC_utf8(p) isBLANK_LC_uvchr(valid_utf8_to_uvchr(p, 0))
4b88fb76
KW
1022#define isSPACE_LC_utf8(p) isSPACE_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1023#define isDIGIT_LC_utf8(p) isDIGIT_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1024#define isUPPER_LC_utf8(p) isUPPER_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1025#define isLOWER_LC_utf8(p) isLOWER_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1026#define isALNUMC_LC_utf8(p) isALNUMC_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1027#define isCNTRL_LC_utf8(p) isCNTRL_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1028#define isGRAPH_LC_utf8(p) isGRAPH_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1029#define isPRINT_LC_utf8(p) isPRINT_LC_uvchr(valid_utf8_to_uvchr(p, 0))
1030#define isPUNCT_LC_utf8(p) isPUNCT_LC_uvchr(valid_utf8_to_uvchr(p, 0))
a0ed51b3 1031
aaa51d5e 1032#define isPSXSPC_LC_utf8(c) (isSPACE_LC_utf8(c) ||(c) == '\f')
aaa51d5e 1033
6ae9f32a 1034/* This conversion works both ways, strangely enough. On EBCDIC platforms,
e5dcd934
KW
1035 * CTRL-@ is 0, CTRL-A is 1, etc, just like on ASCII, except that they don't
1036 * necessarily mean the same characters, e.g. CTRL-D is 4 on both systems, but
1037 * that is EOT on ASCII; ST on EBCDIC */
6ae9f32a 1038# define toCTRL(c) (toUPPER(NATIVE_TO_UNI(c)) ^ 64)
bbce6d69 1039
dea28490
JJ
1040/* Line numbers are unsigned, 32 bits. */
1041typedef U32 line_t;
e5dcd934 1042#define NOLINE ((line_t) 4294967295UL) /* = FFFFFFFF */
378cc40b 1043
91152fc1
DG
1044/* Helpful alias for version prescan */
1045#define is_LAX_VERSION(a,b) \
1046 (a != Perl_prescan_version(aTHX_ a, FALSE, b, NULL, NULL, NULL, NULL))
1047
1048#define is_STRICT_VERSION(a,b) \
1049 (a != Perl_prescan_version(aTHX_ a, TRUE, b, NULL, NULL, NULL, NULL))
1050
1051#define BADVERSION(a,b,c) \
1052 if (b) { \
1053 *b = c; \
1054 } \
1055 return a;
8c52afec 1056
95a59cab
YO
1057#define READ_XDIGIT(s) (isALPHA(*(s)) ? ((*(s)++ + 9) & 0xf) : (*(s)++ & 0xf))
1058
8e84507e 1059/*
ccfc67b7
JH
1060=head1 Memory Management
1061
a02a5408 1062=for apidoc Am|void|Newx|void* ptr|int nitems|type
954c1994
GS
1063The XSUB-writer's interface to the C C<malloc> function.
1064
c5008215
JC
1065In 5.9.3, Newx() and friends replace the older New() API, and drops
1066the first parameter, I<x>, a debug aid which allowed callers to identify
37b8b4c9 1067themselves. This aid has been superseded by a new build option,
d10b4965 1068PERL_MEM_LOG (see L<perlhacktips/PERL_MEM_LOG>). The older API is still
c5008215
JC
1069there for use in XS modules supporting older perls.
1070
a02a5408 1071=for apidoc Am|void|Newxc|void* ptr|int nitems|type|cast
954c1994 1072The XSUB-writer's interface to the C C<malloc> function, with
c5008215 1073cast. See also C<Newx>.
954c1994 1074
a02a5408 1075=for apidoc Am|void|Newxz|void* ptr|int nitems|type
954c1994 1076The XSUB-writer's interface to the C C<malloc> function. The allocated
c5008215 1077memory is zeroed with C<memzero>. See also C<Newx>.
a02a5408 1078
954c1994
GS
1079=for apidoc Am|void|Renew|void* ptr|int nitems|type
1080The XSUB-writer's interface to the C C<realloc> function.
1081
1082=for apidoc Am|void|Renewc|void* ptr|int nitems|type|cast
1083The XSUB-writer's interface to the C C<realloc> function, with
1084cast.
1085
49b8b560 1086=for apidoc Am|void|Safefree|void* ptr
954c1994
GS
1087The XSUB-writer's interface to the C C<free> function.
1088
1089=for apidoc Am|void|Move|void* src|void* dest|int nitems|type
1090The XSUB-writer's interface to the C C<memmove> function. The C<src> is the
1091source, C<dest> is the destination, C<nitems> is the number of items, and C<type> is
1092the type. Can do overlapping moves. See also C<Copy>.
1093
e90e2364
NC
1094=for apidoc Am|void *|MoveD|void* src|void* dest|int nitems|type
1095Like C<Move> but returns dest. Useful for encouraging compilers to tail-call
1096optimise.
1097
954c1994
GS
1098=for apidoc Am|void|Copy|void* src|void* dest|int nitems|type
1099The XSUB-writer's interface to the C C<memcpy> function. The C<src> is the
1100source, C<dest> is the destination, C<nitems> is the number of items, and C<type> is
1101the type. May fail on overlapping copies. See also C<Move>.
1102
e90e2364
NC
1103=for apidoc Am|void *|CopyD|void* src|void* dest|int nitems|type
1104
1105Like C<Copy> but returns dest. Useful for encouraging compilers to tail-call
1106optimise.
1107
954c1994
GS
1108=for apidoc Am|void|Zero|void* dest|int nitems|type
1109
1110The XSUB-writer's interface to the C C<memzero> function. The C<dest> is the
1111destination, C<nitems> is the number of items, and C<type> is the type.
1112
e90e2364
NC
1113=for apidoc Am|void *|ZeroD|void* dest|int nitems|type
1114
1115Like C<Zero> but returns dest. Useful for encouraging compilers to tail-call
1116optimise.
1117
da5d8dbb 1118=for apidoc Am|void|StructCopy|type *src|type *dest|type
4375e838 1119This is an architecture-independent macro to copy one structure to another.
954c1994 1120
7e337ee0
JH
1121=for apidoc Am|void|PoisonWith|void* dest|int nitems|type|U8 byte
1122
1123Fill up memory with a byte pattern (a byte repeated over and over
1124again) that hopefully catches attempts to access uninitialized memory.
1125
1126=for apidoc Am|void|PoisonNew|void* dest|int nitems|type
1127
1128PoisonWith(0xAB) for catching access to allocated but uninitialized memory.
1129
1c12ffb4 1130=for apidoc Am|void|PoisonFree|void* dest|int nitems|type
7e337ee0
JH
1131
1132PoisonWith(0xEF) for catching access to freed memory.
1133
9965345d
JH
1134=for apidoc Am|void|Poison|void* dest|int nitems|type
1135
7e337ee0 1136PoisonWith(0xEF) for catching access to freed memory.
9965345d
JH
1137
1138=cut */
954c1994 1139
561b68a9
SH
1140/* Maintained for backwards-compatibility only. Use newSV() instead. */
1141#ifndef PERL_CORE
ff06c60c 1142#define NEWSV(x,len) newSV(len)
561b68a9 1143#endif
ff06c60c 1144
19a94d75
JH
1145#define MEM_SIZE_MAX ((MEM_SIZE)~0)
1146
2720f793
JH
1147/* The +0.0 in MEM_WRAP_CHECK_ is an attempt to foil
1148 * overly eager compilers that will bleat about e.g.
1149 * (U16)n > (size_t)~0/sizeof(U16) always being false. */
27d5b266 1150#ifdef PERL_MALLOC_WRAP
f019c49e
DD
1151#define MEM_WRAP_CHECK(n,t) \
1152 (void)(sizeof(t) > 1 && ((MEM_SIZE)(n)+0.0) > MEM_SIZE_MAX/sizeof(t) && (croak_memory_wrap(),0))
27d5b266 1153#define MEM_WRAP_CHECK_1(n,t,a) \
f1f66076 1154 (void)(sizeof(t) > 1 && ((MEM_SIZE)(n)+0.0) > MEM_SIZE_MAX/sizeof(t) && (Perl_croak_nocontext("%s",(a)),0))
8b44ba4c 1155#define MEM_WRAP_CHECK_(n,t) MEM_WRAP_CHECK(n,t),
27d5b266 1156
f019c49e 1157#define PERL_STRLEN_ROUNDUP(n) ((void)(((n) > MEM_SIZE_MAX - 2 * PERL_STRLEN_ROUNDUP_QUANTUM) ? (croak_memory_wrap(),0):0),((n-1+PERL_STRLEN_ROUNDUP_QUANTUM)&~((MEM_SIZE)PERL_STRLEN_ROUNDUP_QUANTUM-1)))
27d5b266
JH
1158#else
1159
410319be
NC
1160#define MEM_WRAP_CHECK(n,t)
1161#define MEM_WRAP_CHECK_1(n,t,a)
1162#define MEM_WRAP_CHECK_2(n,t,a,b)
8b44ba4c
NC
1163#define MEM_WRAP_CHECK_(n,t)
1164
6fff79ce 1165#define PERL_STRLEN_ROUNDUP(n) (((n-1+PERL_STRLEN_ROUNDUP_QUANTUM)&~((MEM_SIZE)PERL_STRLEN_ROUNDUP_QUANTUM-1)))
27d5b266 1166
1936d2a7 1167#endif
8b44ba4c 1168
fe4f188c 1169#ifdef PERL_MEM_LOG
46c6c7e2 1170/*
9f653bb5 1171 * If PERL_MEM_LOG is defined, all Newx()s, Renew()s, and Safefree()s
46c6c7e2
JH
1172 * go through functions, which are handy for debugging breakpoints, but
1173 * which more importantly get the immediate calling environment (file and
e352bcff
JH
1174 * line number, and C function name if available) passed in. This info can
1175 * then be used for logging the calls, for which one gets a sample
73d1d973 1176 * implementation unless -DPERL_MEM_LOG_NOIMPL is also defined.
3609ea0d 1177 *
46c6c7e2 1178 * Known problems:
94e892a6 1179 * - not all memory allocs get logged, only those
46c6c7e2 1180 * that go through Newx() and derivatives (while all
94e892a6 1181 * Safefrees do get logged)
46c6c7e2
JH
1182 * - __FILE__ and __LINE__ do not work everywhere
1183 * - __func__ or __FUNCTION__ even less so
1184 * - I think more goes on after the perlio frees but
1185 * the thing is that STDERR gets closed (as do all
1186 * the file descriptors)
1187 * - no deeper calling stack than the caller of the Newx()
1188 * or the kind, but do I look like a C reflection/introspection
1189 * utility to you?
1190 * - the function prototypes for the logging functions
1191 * probably should maybe be somewhere else than handy.h
1192 * - one could consider inlining (macrofying) the logging
1193 * for speed, but I am too lazy
1194 * - one could imagine recording the allocations in a hash,
1195 * (keyed by the allocation address?), and maintain that
1196 * through reallocs and frees, but how to do that without
1197 * any News() happening...?
73d1d973 1198 * - lots of -Ddefines to get useful/controllable output
b953482e 1199 * - lots of ENV reads
46c6c7e2
JH
1200 */
1201
12754f92 1202PERL_EXPORT_C Malloc_t Perl_mem_log_alloc(const UV n, const UV typesize, const char *type_name, Malloc_t newalloc, const char *filename, const int linenumber, const char *funcname);
46c6c7e2 1203
12754f92 1204PERL_EXPORT_C Malloc_t Perl_mem_log_realloc(const UV n, const UV typesize, const char *type_name, Malloc_t oldalloc, Malloc_t newalloc, const char *filename, const int linenumber, const char *funcname);
46c6c7e2 1205
12754f92 1206PERL_EXPORT_C Malloc_t Perl_mem_log_free(Malloc_t oldalloc, const char *filename, const int linenumber, const char *funcname);
46c6c7e2 1207
0b0ab801 1208# ifdef PERL_CORE
73d1d973 1209# ifndef PERL_MEM_LOG_NOIMPL
0b0ab801
MHM
1210enum mem_log_type {
1211 MLT_ALLOC,
1212 MLT_REALLOC,
d7a2c63c
MHM
1213 MLT_FREE,
1214 MLT_NEW_SV,
1215 MLT_DEL_SV
0b0ab801
MHM
1216};
1217# endif
12754f92 1218# if defined(PERL_IN_SV_C) /* those are only used in sv.c */
d7a2c63c
MHM
1219void Perl_mem_log_new_sv(const SV *sv, const char *filename, const int linenumber, const char *funcname);
1220void Perl_mem_log_del_sv(const SV *sv, const char *filename, const int linenumber, const char *funcname);
12754f92 1221# endif
0b0ab801
MHM
1222# endif
1223
fe4f188c
JH
1224#endif
1225
1226#ifdef PERL_MEM_LOG
d1401ee9
MHM
1227#define MEM_LOG_ALLOC(n,t,a) Perl_mem_log_alloc(n,sizeof(t),STRINGIFY(t),a,__FILE__,__LINE__,FUNCTION__)
1228#define MEM_LOG_REALLOC(n,t,v,a) Perl_mem_log_realloc(n,sizeof(t),STRINGIFY(t),v,a,__FILE__,__LINE__,FUNCTION__)
46c6c7e2 1229#define MEM_LOG_FREE(a) Perl_mem_log_free(a,__FILE__,__LINE__,FUNCTION__)
fe4f188c
JH
1230#endif
1231
1232#ifndef MEM_LOG_ALLOC
1233#define MEM_LOG_ALLOC(n,t,a) (a)
1234#endif
1235#ifndef MEM_LOG_REALLOC
1236#define MEM_LOG_REALLOC(n,t,v,a) (a)
1237#endif
1238#ifndef MEM_LOG_FREE
1239#define MEM_LOG_FREE(a) (a)
1240#endif
1241
d1401ee9
MHM
1242#define Newx(v,n,t) (v = (MEM_WRAP_CHECK_(n,t) (t*)MEM_LOG_ALLOC(n,t,safemalloc((MEM_SIZE)((n)*sizeof(t))))))
1243#define Newxc(v,n,t,c) (v = (MEM_WRAP_CHECK_(n,t) (c*)MEM_LOG_ALLOC(n,t,safemalloc((MEM_SIZE)((n)*sizeof(t))))))
1244#define Newxz(v,n,t) (v = (MEM_WRAP_CHECK_(n,t) (t*)MEM_LOG_ALLOC(n,t,safecalloc((n),sizeof(t)))))
a6f6820f
NC
1245
1246#ifndef PERL_CORE
a02a5408
JC
1247/* pre 5.9.x compatibility */
1248#define New(x,v,n,t) Newx(v,n,t)
1249#define Newc(x,v,n,t,c) Newxc(v,n,t,c)
4541904d 1250#define Newz(x,v,n,t) Newxz(v,n,t)
a6f6820f 1251#endif
a02a5408 1252
ff68c719 1253#define Renew(v,n,t) \
d1401ee9 1254 (v = (MEM_WRAP_CHECK_(n,t) (t*)MEM_LOG_REALLOC(n,t,v,saferealloc((Malloc_t)(v),(MEM_SIZE)((n)*sizeof(t))))))
ff68c719 1255#define Renewc(v,n,t,c) \
d1401ee9 1256 (v = (MEM_WRAP_CHECK_(n,t) (c*)MEM_LOG_REALLOC(n,t,v,saferealloc((Malloc_t)(v),(MEM_SIZE)((n)*sizeof(t))))))
94010e71
NC
1257
1258#ifdef PERL_POISON
1259#define Safefree(d) \
06c0cc96 1260 ((d) ? (void)(safefree(MEM_LOG_FREE((Malloc_t)(d))), Poison(&(d), 1, Malloc_t)) : (void) 0)
94010e71 1261#else
fe4f188c 1262#define Safefree(d) safefree(MEM_LOG_FREE((Malloc_t)(d)))
94010e71 1263#endif
55497cff 1264
8b44ba4c
NC
1265#define Move(s,d,n,t) (MEM_WRAP_CHECK_(n,t) (void)memmove((char*)(d),(const char*)(s), (n) * sizeof(t)))
1266#define Copy(s,d,n,t) (MEM_WRAP_CHECK_(n,t) (void)memcpy((char*)(d),(const char*)(s), (n) * sizeof(t)))
1267#define Zero(d,n,t) (MEM_WRAP_CHECK_(n,t) (void)memzero((char*)(d), (n) * sizeof(t)))
55497cff 1268
8b44ba4c
NC
1269#define MoveD(s,d,n,t) (MEM_WRAP_CHECK_(n,t) memmove((char*)(d),(const char*)(s), (n) * sizeof(t)))
1270#define CopyD(s,d,n,t) (MEM_WRAP_CHECK_(n,t) memcpy((char*)(d),(const char*)(s), (n) * sizeof(t)))
e90e2364 1271#ifdef HAS_MEMSET
8b44ba4c 1272#define ZeroD(d,n,t) (MEM_WRAP_CHECK_(n,t) memzero((char*)(d), (n) * sizeof(t)))
e90e2364 1273#else
8b44ba4c
NC
1274/* Using bzero(), which returns void. */
1275#define ZeroD(d,n,t) (MEM_WRAP_CHECK_(n,t) memzero((char*)(d), (n) * sizeof(t)),d)
e90e2364
NC
1276#endif
1277
7e337ee0
JH
1278#define PoisonWith(d,n,t,b) (MEM_WRAP_CHECK_(n,t) (void)memset((char*)(d), (U8)(b), (n) * sizeof(t)))
1279#define PoisonNew(d,n,t) PoisonWith(d,n,t,0xAB)
1280#define PoisonFree(d,n,t) PoisonWith(d,n,t,0xEF)
1281#define Poison(d,n,t) PoisonFree(d,n,t)
27d5b266 1282
2304df62 1283#ifdef USE_STRUCT_COPY
ff68c719 1284#define StructCopy(s,d,t) (*((t*)(d)) = *((t*)(s)))
bee1dbe2
LW
1285#else
1286#define StructCopy(s,d,t) Copy(s,d,1,t)
1287#endif
2cc61e15 1288
622913ab 1289#define C_ARRAY_LENGTH(a) (sizeof(a)/sizeof((a)[0]))
cd431fde 1290#define C_ARRAY_END(a) (a) + (sizeof(a)/sizeof((a)[0]))
622913ab 1291
2cc61e15
DD
1292#ifdef NEED_VA_COPY
1293# ifdef va_copy
1294# define Perl_va_copy(s, d) va_copy(d, s)
2cc61e15 1295# else
a1866d1b
JH
1296# if defined(__va_copy)
1297# define Perl_va_copy(s, d) __va_copy(d, s)
1298# else
1299# define Perl_va_copy(s, d) Copy(s, d, 1, va_list)
1300# endif
2cc61e15
DD
1301# endif
1302#endif
1303
472d47bc
SB
1304/* convenience debug macros */
1305#ifdef USE_ITHREADS
1306#define pTHX_FORMAT "Perl interpreter: 0x%p"
1307#define pTHX__FORMAT ", Perl interpreter: 0x%p"
f54cb97a
AL
1308#define pTHX_VALUE_ (void *)my_perl,
1309#define pTHX_VALUE (void *)my_perl
1310#define pTHX__VALUE_ ,(void *)my_perl,
1311#define pTHX__VALUE ,(void *)my_perl
472d47bc 1312#else
3609ea0d 1313#define pTHX_FORMAT
472d47bc 1314#define pTHX__FORMAT
3609ea0d 1315#define pTHX_VALUE_
472d47bc 1316#define pTHX_VALUE
3609ea0d 1317#define pTHX__VALUE_
472d47bc
SB
1318#define pTHX__VALUE
1319#endif /* USE_ITHREADS */
3609ea0d 1320
2acdbac1
NC
1321/* Perl_deprecate was not part of the public API, and did not have a deprecate()
1322 shortcut macro defined without -DPERL_CORE. Neither codesearch.google.com nor
1323 CPAN::Unpack show any users outside the core. */
1324#ifdef PERL_CORE
d1d15184 1325# define deprecate(s) Perl_ck_warner_d(aTHX_ packWARN(WARN_DEPRECATED), "Use of " s " is deprecated")
2acdbac1
NC
1326#endif
1327
e9a8c099
MHM
1328/*
1329 * Local variables:
1330 * c-indentation-style: bsd
1331 * c-basic-offset: 4
14d04a33 1332 * indent-tabs-mode: nil
e9a8c099
MHM
1333 * End:
1334 *
14d04a33 1335 * ex: set ts=8 sts=4 sw=4 et:
e9a8c099 1336 */