xref: /openbsd-src/gnu/usr.bin/perl/handy.h (revision 2b0358df1d88d06ef4139321dd05bd5e05d91eaf)
1 /*    handy.h
2  *
3  *    Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1999,
4  *    2000, 2001, 2002, 2004, 2005, 2006, 2007, by Larry Wall and others
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  *
9  */
10 
11 #if !defined(__STDC__)
12 #ifdef NULL
13 #undef NULL
14 #endif
15 #ifndef I286
16 #  define NULL 0
17 #else
18 #  define NULL 0L
19 #endif
20 #endif
21 
22 #define Null(type) ((type)NULL)
23 
24 /*
25 =head1 Handy Values
26 
27 =for apidoc AmU||Nullch
28 Null character pointer.
29 
30 =for apidoc AmU||Nullsv
31 Null SV pointer.
32 
33 =cut
34 */
35 
36 #define Nullch Null(char*)
37 #define Nullfp Null(PerlIO*)
38 #define Nullsv Null(SV*)
39 
40 #ifdef TRUE
41 #undef TRUE
42 #endif
43 #ifdef FALSE
44 #undef FALSE
45 #endif
46 #define TRUE (1)
47 #define FALSE (0)
48 
49 
50 /* XXX Configure ought to have a test for a boolean type, if I can
51    just figure out all the headers such a test needs.
52    Andy Dougherty	August 1996
53 */
54 /* bool is built-in for g++-2.6.3 and later, which might be used
55    for extensions.  <_G_config.h> defines _G_HAVE_BOOL, but we can't
56    be sure _G_config.h will be included before this file.  _G_config.h
57    also defines _G_HAVE_BOOL for both gcc and g++, but only g++
58    actually has bool.  Hence, _G_HAVE_BOOL is pretty useless for us.
59    g++ can be identified by __GNUG__.
60    Andy Dougherty	February 2000
61 */
62 #ifdef __GNUG__		/* GNU g++ has bool built-in */
63 #  ifndef HAS_BOOL
64 #    define HAS_BOOL 1
65 #  endif
66 #endif
67 
68 /* The NeXT dynamic loader headers will not build with the bool macro
69    So declare them now to clear confusion.
70 */
71 #if defined(NeXT) || defined(__NeXT__)
72 # undef FALSE
73 # undef TRUE
74   typedef enum bool { FALSE = 0, TRUE = 1 } bool;
75 # define ENUM_BOOL 1
76 # ifndef HAS_BOOL
77 #  define HAS_BOOL 1
78 # endif /* !HAS_BOOL */
79 #endif /* NeXT || __NeXT__ */
80 
81 #ifndef HAS_BOOL
82 # if defined(UTS) || defined(VMS)
83 #  define bool int
84 # else
85 #  define bool char
86 # endif
87 # define HAS_BOOL 1
88 #endif
89 
90 /* Try to figure out __func__ or __FUNCTION__ equivalent, if any.
91  * XXX Should really be a Configure probe, with HAS__FUNCTION__
92  *     and FUNCTION__ as results.
93  * XXX Similarly, a Configure probe for __FILE__ and __LINE__ is needed. */
94 #if (defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L) || (defined(__SUNPRO_C)) /* C99 or close enough. */
95 #  define FUNCTION__ __func__
96 #else
97 #  if (defined(_MSC_VER) && _MSC_VER < 1300) || /* Pre-MSVC 7.0 has neither __func__ nor __FUNCTION and no good workarounds, either. */ \
98       (defined(__DECC_VER)) /* Tru64 or VMS, and strict C89 being used, but not modern enough cc (in Tur64, -c99 not known, only -std1). */
99 #    define FUNCTION__ ""
100 #  else
101 #    define FUNCTION__ __FUNCTION__ /* Common extension. */
102 #  endif
103 #endif
104 
105 /* XXX A note on the perl source internal type system.  The
106    original intent was that I32 be *exactly* 32 bits.
107 
108    Currently, we only guarantee that I32 is *at least* 32 bits.
109    Specifically, if int is 64 bits, then so is I32.  (This is the case
110    for the Cray.)  This has the advantage of meshing nicely with
111    standard library calls (where we pass an I32 and the library is
112    expecting an int), but the disadvantage that an I32 is not 32 bits.
113    Andy Dougherty	August 1996
114 
115    There is no guarantee that there is *any* integral type with
116    exactly 32 bits.  It is perfectly legal for a system to have
117    sizeof(short) == sizeof(int) == sizeof(long) == 8.
118 
119    Similarly, there is no guarantee that I16 and U16 have exactly 16
120    bits.
121 
122    For dealing with issues that may arise from various 32/64-bit
123    systems, we will ask Configure to check out
124 
125 	SHORTSIZE == sizeof(short)
126 	INTSIZE == sizeof(int)
127 	LONGSIZE == sizeof(long)
128 	LONGLONGSIZE == sizeof(long long) (if HAS_LONG_LONG)
129 	PTRSIZE == sizeof(void *)
130 	DOUBLESIZE == sizeof(double)
131 	LONG_DOUBLESIZE == sizeof(long double) (if HAS_LONG_DOUBLE).
132 
133 */
134 
135 #ifdef I_INTTYPES /* e.g. Linux has int64_t without <inttypes.h> */
136 #   include <inttypes.h>
137 #   ifdef INT32_MIN_BROKEN
138 #       undef  INT32_MIN
139 #       define INT32_MIN (-2147483647-1)
140 #   endif
141 #   ifdef INT64_MIN_BROKEN
142 #       undef  INT64_MIN
143 #       define INT64_MIN (-9223372036854775807LL-1)
144 #   endif
145 #endif
146 
147 typedef I8TYPE I8;
148 typedef U8TYPE U8;
149 typedef I16TYPE I16;
150 typedef U16TYPE U16;
151 typedef I32TYPE I32;
152 typedef U32TYPE U32;
153 #ifdef PERL_CORE
154 #   ifdef HAS_QUAD
155 typedef I64TYPE I64;
156 typedef U64TYPE U64;
157 #   endif
158 #endif /* PERL_CORE */
159 
160 #if defined(HAS_QUAD) && defined(USE_64_BIT_INT)
161 #   ifndef UINT64_C /* usually from <inttypes.h> */
162 #       if defined(HAS_LONG_LONG) && QUADKIND == QUAD_IS_LONG_LONG
163 #           define INT64_C(c)	CAT2(c,LL)
164 #           define UINT64_C(c)	CAT2(c,ULL)
165 #       else
166 #           if LONGSIZE == 8 && QUADKIND == QUAD_IS_LONG
167 #               define INT64_C(c)	CAT2(c,L)
168 #               define UINT64_C(c)	CAT2(c,UL)
169 #           else
170 #               define INT64_C(c)	((I64TYPE)(c))
171 #               define UINT64_C(c)	((U64TYPE)(c))
172 #           endif
173 #       endif
174 #   endif
175 #endif
176 
177 /* HMB H.Merijn Brand - a placeholder for preparing Configure patches */
178 #if defined(HAS_MALLOC_SIZE) && defined(LOCALTIME_R_NEEDS_TZSET) && defined(HAS_PSEUDOFORK)
179 /* Not (yet) used at top level, but mention them for metaconfig */
180 #endif
181 
182 /* Mention I8SIZE, U8SIZE, I16SIZE, U16SIZE, I32SIZE, U32SIZE,
183    I64SIZE, and U64SIZE here so that metaconfig pulls them in. */
184 
185 #if defined(UINT8_MAX) && defined(INT16_MAX) && defined(INT32_MAX)
186 
187 /* I8_MAX and I8_MIN constants are not defined, as I8 is an ambiguous type.
188    Please search CHAR_MAX in perl.h for further details. */
189 #define U8_MAX UINT8_MAX
190 #define U8_MIN UINT8_MIN
191 
192 #define I16_MAX INT16_MAX
193 #define I16_MIN INT16_MIN
194 #define U16_MAX UINT16_MAX
195 #define U16_MIN UINT16_MIN
196 
197 #define I32_MAX INT32_MAX
198 #define I32_MIN INT32_MIN
199 #ifndef UINT32_MAX_BROKEN /* e.g. HP-UX with gcc messes this up */
200 #  define U32_MAX UINT32_MAX
201 #else
202 #  define U32_MAX 4294967295U
203 #endif
204 #define U32_MIN UINT32_MIN
205 
206 #else
207 
208 /* I8_MAX and I8_MIN constants are not defined, as I8 is an ambiguous type.
209    Please search CHAR_MAX in perl.h for further details. */
210 #define U8_MAX PERL_UCHAR_MAX
211 #define U8_MIN PERL_UCHAR_MIN
212 
213 #define I16_MAX PERL_SHORT_MAX
214 #define I16_MIN PERL_SHORT_MIN
215 #define U16_MAX PERL_USHORT_MAX
216 #define U16_MIN PERL_USHORT_MIN
217 
218 #if LONGSIZE > 4
219 # define I32_MAX PERL_INT_MAX
220 # define I32_MIN PERL_INT_MIN
221 # define U32_MAX PERL_UINT_MAX
222 # define U32_MIN PERL_UINT_MIN
223 #else
224 # define I32_MAX PERL_LONG_MAX
225 # define I32_MIN PERL_LONG_MIN
226 # define U32_MAX PERL_ULONG_MAX
227 # define U32_MIN PERL_ULONG_MIN
228 #endif
229 
230 #endif
231 
232 /* log(2) is pretty close to  0.30103, just in case anyone is grepping for it */
233 #define BIT_DIGITS(N)   (((N)*146)/485 + 1)  /* log2(10) =~ 146/485 */
234 #define TYPE_DIGITS(T)  BIT_DIGITS(sizeof(T) * 8)
235 #define TYPE_CHARS(T)   (TYPE_DIGITS(T) + 2) /* sign, NUL */
236 
237 #define Ctl(ch) ((ch) & 037)
238 
239 /*
240 =head1 SV-Body Allocation
241 
242 =for apidoc Ama|SV*|newSVpvs|const char* s
243 Like C<newSVpvn>, but takes a literal string instead of a string/length pair.
244 
245 =for apidoc Ama|SV*|newSVpvs_share|const char* s
246 Like C<newSVpvn_share>, but takes a literal string instead of a string/length
247 pair and omits the hash parameter.
248 
249 =for apidoc Am|void|sv_catpvs|SV* sv|const char* s
250 Like C<sv_catpvn>, but takes a literal string instead of a string/length pair.
251 
252 =for apidoc Am|void|sv_setpvs|SV* sv|const char* s
253 Like C<sv_setpvn>, but takes a literal string instead of a string/length pair.
254 
255 =head1 Memory Management
256 
257 =for apidoc Ama|char*|savepvs|const char* s
258 Like C<savepvn>, but takes a literal string instead of a string/length pair.
259 
260 =head1 GV Functions
261 
262 =for apidoc Am|HV*|gv_stashpvs|const char* name|I32 create
263 Like C<gv_stashpvn>, but takes a literal string instead of a string/length pair.
264 
265 =head1 Hash Manipulation Functions
266 
267 =for apidoc Am|SV**|hv_fetchs|HV* tb|const char* key|I32 lval
268 Like C<hv_fetch>, but takes a literal string instead of a string/length pair.
269 
270 =for apidoc Am|SV**|hv_stores|HV* tb|const char* key|NULLOK SV* val
271 Like C<hv_store>, but takes a literal string instead of a string/length pair
272 and omits the hash parameter.
273 
274 =cut
275 */
276 
277 /* concatenating with "" ensures that only literal strings are accepted as argument */
278 #define STR_WITH_LEN(s)  (s ""), (sizeof(s)-1)
279 
280 /* note that STR_WITH_LEN() can't be used as argument to macros or functions that
281  * under some configurations might be macros, which means that it requires the full
282  * Perl_xxx(aTHX_ ...) form for any API calls where it's used.
283  */
284 
285 /* STR_WITH_LEN() shortcuts */
286 #define newSVpvs(str) Perl_newSVpvn(aTHX_ STR_WITH_LEN(str))
287 #define newSVpvs_share(str) Perl_newSVpvn_share(aTHX_ STR_WITH_LEN(str), 0)
288 #define sv_catpvs(sv, str) Perl_sv_catpvn_flags(aTHX_ sv, STR_WITH_LEN(str), SV_GMAGIC)
289 #define sv_setpvs(sv, str) Perl_sv_setpvn(aTHX_ sv, STR_WITH_LEN(str))
290 #define savepvs(str) Perl_savepvn(aTHX_ STR_WITH_LEN(str))
291 #define gv_stashpvs(str, create) Perl_gv_stashpvn(aTHX_ STR_WITH_LEN(str), create)
292 #define gv_fetchpvs(namebeg, add, sv_type) Perl_gv_fetchpvn_flags(aTHX_ STR_WITH_LEN(namebeg), add, sv_type)
293 #define hv_fetchs(hv,key,lval)						\
294   ((SV **)Perl_hv_common(aTHX_ (hv), NULL, STR_WITH_LEN(key), 0,	\
295 			 (lval) ? (HV_FETCH_JUST_SV | HV_FETCH_LVALUE)	\
296 			 : HV_FETCH_JUST_SV, NULL, 0))
297 
298 #define hv_stores(hv,key,val)						\
299   ((SV **)Perl_hv_common(aTHX_ (hv), NULL, STR_WITH_LEN(key), 0,	\
300 			 (HV_FETCH_ISSTORE|HV_FETCH_JUST_SV), (val), 0))
301 
302 
303 /*
304 =head1 Miscellaneous Functions
305 
306 =for apidoc Am|bool|strNE|char* s1|char* s2
307 Test two strings to see if they are different.  Returns true or
308 false.
309 
310 =for apidoc Am|bool|strEQ|char* s1|char* s2
311 Test two strings to see if they are equal.  Returns true or false.
312 
313 =for apidoc Am|bool|strLT|char* s1|char* s2
314 Test two strings to see if the first, C<s1>, is less than the second,
315 C<s2>.  Returns true or false.
316 
317 =for apidoc Am|bool|strLE|char* s1|char* s2
318 Test two strings to see if the first, C<s1>, is less than or equal to the
319 second, C<s2>.  Returns true or false.
320 
321 =for apidoc Am|bool|strGT|char* s1|char* s2
322 Test two strings to see if the first, C<s1>, is greater than the second,
323 C<s2>.  Returns true or false.
324 
325 =for apidoc Am|bool|strGE|char* s1|char* s2
326 Test two strings to see if the first, C<s1>, is greater than or equal to
327 the second, C<s2>.  Returns true or false.
328 
329 =for apidoc Am|bool|strnNE|char* s1|char* s2|STRLEN len
330 Test two strings to see if they are different.  The C<len> parameter
331 indicates the number of bytes to compare.  Returns true or false. (A
332 wrapper for C<strncmp>).
333 
334 =for apidoc Am|bool|strnEQ|char* s1|char* s2|STRLEN len
335 Test two strings to see if they are equal.  The C<len> parameter indicates
336 the number of bytes to compare.  Returns true or false. (A wrapper for
337 C<strncmp>).
338 
339 =cut
340 */
341 
342 #define strNE(s1,s2) (strcmp(s1,s2))
343 #define strEQ(s1,s2) (!strcmp(s1,s2))
344 #define strLT(s1,s2) (strcmp(s1,s2) < 0)
345 #define strLE(s1,s2) (strcmp(s1,s2) <= 0)
346 #define strGT(s1,s2) (strcmp(s1,s2) > 0)
347 #define strGE(s1,s2) (strcmp(s1,s2) >= 0)
348 #define strnNE(s1,s2,l) (strncmp(s1,s2,l))
349 #define strnEQ(s1,s2,l) (!strncmp(s1,s2,l))
350 
351 #ifdef HAS_MEMCMP
352 #  define memNE(s1,s2,l) (memcmp(s1,s2,l))
353 #  define memEQ(s1,s2,l) (!memcmp(s1,s2,l))
354 #else
355 #  define memNE(s1,s2,l) (bcmp(s1,s2,l))
356 #  define memEQ(s1,s2,l) (!bcmp(s1,s2,l))
357 #endif
358 
359 #define memEQs(s1, l, s2) \
360 	(sizeof(s2)-1 == l && memEQ(s1, (s2 ""), (sizeof(s2)-1)))
361 #define memNEs(s1, l, s2) !memEQs(s1, l, s2)
362 
363 /*
364  * Character classes.
365  *
366  * Unfortunately, the introduction of locales means that we
367  * can't trust isupper(), etc. to tell the truth.  And when
368  * it comes to /\w+/ with tainting enabled, we *must* be able
369  * to trust our character classes.
370  *
371  * Therefore, the default tests in the text of Perl will be
372  * independent of locale.  Any code that wants to depend on
373  * the current locale will use the tests that begin with "lc".
374  */
375 
376 #ifdef HAS_SETLOCALE  /* XXX Is there a better test for this? */
377 #  ifndef CTYPE256
378 #    define CTYPE256
379 #  endif
380 #endif
381 
382 /*
383 
384 =head1 Character classes
385 
386 =for apidoc Am|bool|isALNUM|char ch
387 Returns a boolean indicating whether the C C<char> is an ASCII alphanumeric
388 character (including underscore) or digit.
389 
390 =for apidoc Am|bool|isALPHA|char ch
391 Returns a boolean indicating whether the C C<char> is an ASCII alphabetic
392 character.
393 
394 =for apidoc Am|bool|isSPACE|char ch
395 Returns a boolean indicating whether the C C<char> is whitespace.
396 
397 =for apidoc Am|bool|isDIGIT|char ch
398 Returns a boolean indicating whether the C C<char> is an ASCII
399 digit.
400 
401 =for apidoc Am|bool|isUPPER|char ch
402 Returns a boolean indicating whether the C C<char> is an uppercase
403 character.
404 
405 =for apidoc Am|bool|isLOWER|char ch
406 Returns a boolean indicating whether the C C<char> is a lowercase
407 character.
408 
409 =for apidoc Am|char|toUPPER|char ch
410 Converts the specified character to uppercase.
411 
412 =for apidoc Am|char|toLOWER|char ch
413 Converts the specified character to lowercase.
414 
415 =cut
416 */
417 
418 #define isALNUM(c)	(isALPHA(c) || isDIGIT(c) || (c) == '_')
419 #define isIDFIRST(c)	(isALPHA(c) || (c) == '_')
420 #define isALPHA(c)	(isUPPER(c) || isLOWER(c))
421 #define isSPACE(c) \
422 	((c) == ' ' || (c) == '\t' || (c) == '\n' || (c) =='\r' || (c) == '\f')
423 #define isPSXSPC(c)	(isSPACE(c) || (c) == '\v')
424 #define isBLANK(c)	((c) == ' ' || (c) == '\t')
425 #define isDIGIT(c)	((c) >= '0' && (c) <= '9')
426 #ifdef EBCDIC
427     /* In EBCDIC we do not do locales: therefore() isupper() is fine. */
428 #   define isUPPER(c)	isupper(c)
429 #   define isLOWER(c)	islower(c)
430 #   define isALNUMC(c)	isalnum(c)
431 #   define isASCII(c)	isascii(c)
432 #   define isCNTRL(c)	iscntrl(c)
433 #   define isGRAPH(c)	isgraph(c)
434 #   define isPRINT(c)	isprint(c)
435 #   define isPUNCT(c)	ispunct(c)
436 #   define isXDIGIT(c)	isxdigit(c)
437 #   define toUPPER(c)	toupper(c)
438 #   define toLOWER(c)	tolower(c)
439 #else
440 #   define isUPPER(c)	((c) >= 'A' && (c) <= 'Z')
441 #   define isLOWER(c)	((c) >= 'a' && (c) <= 'z')
442 #   define isALNUMC(c)	(isALPHA(c) || isDIGIT(c))
443 #   define isASCII(c)	((c) <= 127)
444 #   define isCNTRL(c)	((c) < ' ' || (c) == 127)
445 #   define isGRAPH(c)	(isALNUM(c) || isPUNCT(c))
446 #   define isPRINT(c)	(((c) >= 32 && (c) < 127))
447 #   define isPUNCT(c)	(((c) >= 33 && (c) <= 47) || ((c) >= 58 && (c) <= 64)  || ((c) >= 91 && (c) <= 96) || ((c) >= 123 && (c) <= 126))
448 #   define isXDIGIT(c)  (isDIGIT(c) || ((c) >= 'a' && (c) <= 'f') || ((c) >= 'A' && (c) <= 'F'))
449 #   define toUPPER(c)	(isLOWER(c) ? (c) - ('a' - 'A') : (c))
450 #   define toLOWER(c)	(isUPPER(c) ? (c) + ('a' - 'A') : (c))
451 #endif
452 
453 #ifdef USE_NEXT_CTYPE
454 
455 #  define isALNUM_LC(c) \
456 	(NXIsAlNum((unsigned int)(c)) || (char)(c) == '_')
457 #  define isIDFIRST_LC(c) \
458 	(NXIsAlpha((unsigned int)(c)) || (char)(c) == '_')
459 #  define isALPHA_LC(c)		NXIsAlpha((unsigned int)(c))
460 #  define isSPACE_LC(c)		NXIsSpace((unsigned int)(c))
461 #  define isDIGIT_LC(c)		NXIsDigit((unsigned int)(c))
462 #  define isUPPER_LC(c)		NXIsUpper((unsigned int)(c))
463 #  define isLOWER_LC(c)		NXIsLower((unsigned int)(c))
464 #  define isALNUMC_LC(c)	NXIsAlNum((unsigned int)(c))
465 #  define isCNTRL_LC(c)		NXIsCntrl((unsigned int)(c))
466 #  define isGRAPH_LC(c)		NXIsGraph((unsigned int)(c))
467 #  define isPRINT_LC(c)		NXIsPrint((unsigned int)(c))
468 #  define isPUNCT_LC(c)		NXIsPunct((unsigned int)(c))
469 #  define toUPPER_LC(c)		NXToUpper((unsigned int)(c))
470 #  define toLOWER_LC(c)		NXToLower((unsigned int)(c))
471 
472 #else /* !USE_NEXT_CTYPE */
473 
474 #  if defined(CTYPE256) || (!defined(isascii) && !defined(HAS_ISASCII))
475 
476 #    define isALNUM_LC(c)   (isalnum((unsigned char)(c)) || (char)(c) == '_')
477 #    define isIDFIRST_LC(c) (isalpha((unsigned char)(c)) || (char)(c) == '_')
478 #    define isALPHA_LC(c)	isalpha((unsigned char)(c))
479 #    define isSPACE_LC(c)	isspace((unsigned char)(c))
480 #    define isDIGIT_LC(c)	isdigit((unsigned char)(c))
481 #    define isUPPER_LC(c)	isupper((unsigned char)(c))
482 #    define isLOWER_LC(c)	islower((unsigned char)(c))
483 #    define isALNUMC_LC(c)	isalnum((unsigned char)(c))
484 #    define isCNTRL_LC(c)	iscntrl((unsigned char)(c))
485 #    define isGRAPH_LC(c)	isgraph((unsigned char)(c))
486 #    define isPRINT_LC(c)	isprint((unsigned char)(c))
487 #    define isPUNCT_LC(c)	ispunct((unsigned char)(c))
488 #    define toUPPER_LC(c)	toupper((unsigned char)(c))
489 #    define toLOWER_LC(c)	tolower((unsigned char)(c))
490 
491 #  else
492 
493 #    define isALNUM_LC(c)	(isascii(c) && (isalnum(c) || (c) == '_'))
494 #    define isIDFIRST_LC(c)	(isascii(c) && (isalpha(c) || (c) == '_'))
495 #    define isALPHA_LC(c)	(isascii(c) && isalpha(c))
496 #    define isSPACE_LC(c)	(isascii(c) && isspace(c))
497 #    define isDIGIT_LC(c)	(isascii(c) && isdigit(c))
498 #    define isUPPER_LC(c)	(isascii(c) && isupper(c))
499 #    define isLOWER_LC(c)	(isascii(c) && islower(c))
500 #    define isALNUMC_LC(c)	(isascii(c) && isalnum(c))
501 #    define isCNTRL_LC(c)	(isascii(c) && iscntrl(c))
502 #    define isGRAPH_LC(c)	(isascii(c) && isgraph(c))
503 #    define isPRINT_LC(c)	(isascii(c) && isprint(c))
504 #    define isPUNCT_LC(c)	(isascii(c) && ispunct(c))
505 #    define toUPPER_LC(c)	toupper(c)
506 #    define toLOWER_LC(c)	tolower(c)
507 
508 #  endif
509 #endif /* USE_NEXT_CTYPE */
510 
511 #define isPSXSPC_LC(c)		(isSPACE_LC(c) || (c) == '\v')
512 #define isBLANK_LC(c)		isBLANK(c) /* could be wrong */
513 
514 #define isALNUM_uni(c)		is_uni_alnum(c)
515 #define isIDFIRST_uni(c)	is_uni_idfirst(c)
516 #define isALPHA_uni(c)		is_uni_alpha(c)
517 #define isSPACE_uni(c)		is_uni_space(c)
518 #define isDIGIT_uni(c)		is_uni_digit(c)
519 #define isUPPER_uni(c)		is_uni_upper(c)
520 #define isLOWER_uni(c)		is_uni_lower(c)
521 #define isALNUMC_uni(c)		is_uni_alnumc(c)
522 #define isASCII_uni(c)		is_uni_ascii(c)
523 #define isCNTRL_uni(c)		is_uni_cntrl(c)
524 #define isGRAPH_uni(c)		is_uni_graph(c)
525 #define isPRINT_uni(c)		is_uni_print(c)
526 #define isPUNCT_uni(c)		is_uni_punct(c)
527 #define isXDIGIT_uni(c)		is_uni_xdigit(c)
528 #define toUPPER_uni(c,s,l)	to_uni_upper(c,s,l)
529 #define toTITLE_uni(c,s,l)	to_uni_title(c,s,l)
530 #define toLOWER_uni(c,s,l)	to_uni_lower(c,s,l)
531 #define toFOLD_uni(c,s,l)	to_uni_fold(c,s,l)
532 
533 #define isPSXSPC_uni(c)		(isSPACE_uni(c) ||(c) == '\f')
534 #define isBLANK_uni(c)		isBLANK(c) /* could be wrong */
535 
536 #define isALNUM_LC_uvchr(c)	(c < 256 ? isALNUM_LC(c) : is_uni_alnum_lc(c))
537 #define isIDFIRST_LC_uvchr(c)	(c < 256 ? isIDFIRST_LC(c) : is_uni_idfirst_lc(c))
538 #define isALPHA_LC_uvchr(c)	(c < 256 ? isALPHA_LC(c) : is_uni_alpha_lc(c))
539 #define isSPACE_LC_uvchr(c)	(c < 256 ? isSPACE_LC(c) : is_uni_space_lc(c))
540 #define isDIGIT_LC_uvchr(c)	(c < 256 ? isDIGIT_LC(c) : is_uni_digit_lc(c))
541 #define isUPPER_LC_uvchr(c)	(c < 256 ? isUPPER_LC(c) : is_uni_upper_lc(c))
542 #define isLOWER_LC_uvchr(c)	(c < 256 ? isLOWER_LC(c) : is_uni_lower_lc(c))
543 #define isALNUMC_LC_uvchr(c)	(c < 256 ? isALNUMC_LC(c) : is_uni_alnumc_lc(c))
544 #define isCNTRL_LC_uvchr(c)	(c < 256 ? isCNTRL_LC(c) : is_uni_cntrl_lc(c))
545 #define isGRAPH_LC_uvchr(c)	(c < 256 ? isGRAPH_LC(c) : is_uni_graph_lc(c))
546 #define isPRINT_LC_uvchr(c)	(c < 256 ? isPRINT_LC(c) : is_uni_print_lc(c))
547 #define isPUNCT_LC_uvchr(c)	(c < 256 ? isPUNCT_LC(c) : is_uni_punct_lc(c))
548 
549 #define isPSXSPC_LC_uni(c)	(isSPACE_LC_uni(c) ||(c) == '\f')
550 #define isBLANK_LC_uni(c)	isBLANK(c) /* could be wrong */
551 
552 #define isALNUM_utf8(p)		is_utf8_alnum(p)
553 /* The ID_Start of Unicode is quite limiting: it assumes a L-class
554  * character (meaning that you cannot have, say, a CJK character).
555  * Instead, let's allow ID_Continue but not digits. */
556 #define isIDFIRST_utf8(p)	(is_utf8_idcont(p) && !is_utf8_digit(p))
557 #define isALPHA_utf8(p)		is_utf8_alpha(p)
558 #define isSPACE_utf8(p)		is_utf8_space(p)
559 #define isDIGIT_utf8(p)		is_utf8_digit(p)
560 #define isUPPER_utf8(p)		is_utf8_upper(p)
561 #define isLOWER_utf8(p)		is_utf8_lower(p)
562 #define isALNUMC_utf8(p)	is_utf8_alnumc(p)
563 #define isASCII_utf8(p)		is_utf8_ascii(p)
564 #define isCNTRL_utf8(p)		is_utf8_cntrl(p)
565 #define isGRAPH_utf8(p)		is_utf8_graph(p)
566 #define isPRINT_utf8(p)		is_utf8_print(p)
567 #define isPUNCT_utf8(p)		is_utf8_punct(p)
568 #define isXDIGIT_utf8(p)	is_utf8_xdigit(p)
569 #define toUPPER_utf8(p,s,l)	to_utf8_upper(p,s,l)
570 #define toTITLE_utf8(p,s,l)	to_utf8_title(p,s,l)
571 #define toLOWER_utf8(p,s,l)	to_utf8_lower(p,s,l)
572 
573 #define isPSXSPC_utf8(c)	(isSPACE_utf8(c) ||(c) == '\f')
574 #define isBLANK_utf8(c)		isBLANK(c) /* could be wrong */
575 
576 #define isALNUM_LC_utf8(p)	isALNUM_LC_uvchr(utf8_to_uvchr(p,  0))
577 #define isIDFIRST_LC_utf8(p)	isIDFIRST_LC_uvchr(utf8_to_uvchr(p,  0))
578 #define isALPHA_LC_utf8(p)	isALPHA_LC_uvchr(utf8_to_uvchr(p,  0))
579 #define isSPACE_LC_utf8(p)	isSPACE_LC_uvchr(utf8_to_uvchr(p,  0))
580 #define isDIGIT_LC_utf8(p)	isDIGIT_LC_uvchr(utf8_to_uvchr(p,  0))
581 #define isUPPER_LC_utf8(p)	isUPPER_LC_uvchr(utf8_to_uvchr(p,  0))
582 #define isLOWER_LC_utf8(p)	isLOWER_LC_uvchr(utf8_to_uvchr(p,  0))
583 #define isALNUMC_LC_utf8(p)	isALNUMC_LC_uvchr(utf8_to_uvchr(p,  0))
584 #define isCNTRL_LC_utf8(p)	isCNTRL_LC_uvchr(utf8_to_uvchr(p,  0))
585 #define isGRAPH_LC_utf8(p)	isGRAPH_LC_uvchr(utf8_to_uvchr(p,  0))
586 #define isPRINT_LC_utf8(p)	isPRINT_LC_uvchr(utf8_to_uvchr(p,  0))
587 #define isPUNCT_LC_utf8(p)	isPUNCT_LC_uvchr(utf8_to_uvchr(p,  0))
588 
589 #define isPSXSPC_LC_utf8(c)	(isSPACE_LC_utf8(c) ||(c) == '\f')
590 #define isBLANK_LC_utf8(c)	isBLANK(c) /* could be wrong */
591 
592 #ifdef EBCDIC
593 #  ifdef PERL_IMPLICIT_CONTEXT
594 #    define toCTRL(c)     Perl_ebcdic_control(aTHX_ c)
595 #  else
596 #    define toCTRL        Perl_ebcdic_control
597 #  endif
598 #else
599   /* This conversion works both ways, strangely enough. */
600 #  define toCTRL(c)    (toUPPER(c) ^ 64)
601 #endif
602 
603 /* Line numbers are unsigned, 32 bits. */
604 typedef U32 line_t;
605 #define NOLINE ((line_t) 4294967295UL)
606 
607 
608 /*
609 =head1 Memory Management
610 
611 =for apidoc Am|void|Newx|void* ptr|int nitems|type
612 The XSUB-writer's interface to the C C<malloc> function.
613 
614 In 5.9.3, Newx() and friends replace the older New() API, and drops
615 the first parameter, I<x>, a debug aid which allowed callers to identify
616 themselves.  This aid has been superseded by a new build option,
617 PERL_MEM_LOG (see L<perlhack/PERL_MEM_LOG>).  The older API is still
618 there for use in XS modules supporting older perls.
619 
620 =for apidoc Am|void|Newxc|void* ptr|int nitems|type|cast
621 The XSUB-writer's interface to the C C<malloc> function, with
622 cast.  See also C<Newx>.
623 
624 =for apidoc Am|void|Newxz|void* ptr|int nitems|type
625 The XSUB-writer's interface to the C C<malloc> function.  The allocated
626 memory is zeroed with C<memzero>.  See also C<Newx>.
627 
628 =for apidoc Am|void|Renew|void* ptr|int nitems|type
629 The XSUB-writer's interface to the C C<realloc> function.
630 
631 =for apidoc Am|void|Renewc|void* ptr|int nitems|type|cast
632 The XSUB-writer's interface to the C C<realloc> function, with
633 cast.
634 
635 =for apidoc Am|void|Safefree|void* ptr
636 The XSUB-writer's interface to the C C<free> function.
637 
638 =for apidoc Am|void|Move|void* src|void* dest|int nitems|type
639 The XSUB-writer's interface to the C C<memmove> function.  The C<src> is the
640 source, C<dest> is the destination, C<nitems> is the number of items, and C<type> is
641 the type.  Can do overlapping moves.  See also C<Copy>.
642 
643 =for apidoc Am|void *|MoveD|void* src|void* dest|int nitems|type
644 Like C<Move> but returns dest. Useful for encouraging compilers to tail-call
645 optimise.
646 
647 =for apidoc Am|void|Copy|void* src|void* dest|int nitems|type
648 The XSUB-writer's interface to the C C<memcpy> function.  The C<src> is the
649 source, C<dest> is the destination, C<nitems> is the number of items, and C<type> is
650 the type.  May fail on overlapping copies.  See also C<Move>.
651 
652 =for apidoc Am|void *|CopyD|void* src|void* dest|int nitems|type
653 
654 Like C<Copy> but returns dest. Useful for encouraging compilers to tail-call
655 optimise.
656 
657 =for apidoc Am|void|Zero|void* dest|int nitems|type
658 
659 The XSUB-writer's interface to the C C<memzero> function.  The C<dest> is the
660 destination, C<nitems> is the number of items, and C<type> is the type.
661 
662 =for apidoc Am|void *|ZeroD|void* dest|int nitems|type
663 
664 Like C<Zero> but returns dest. Useful for encouraging compilers to tail-call
665 optimise.
666 
667 =for apidoc Am|void|StructCopy|type src|type dest|type
668 This is an architecture-independent macro to copy one structure to another.
669 
670 =for apidoc Am|void|PoisonWith|void* dest|int nitems|type|U8 byte
671 
672 Fill up memory with a byte pattern (a byte repeated over and over
673 again) that hopefully catches attempts to access uninitialized memory.
674 
675 =for apidoc Am|void|PoisonNew|void* dest|int nitems|type
676 
677 PoisonWith(0xAB) for catching access to allocated but uninitialized memory.
678 
679 =for apidoc Am|void|PoisonFree|void* dest|int nitems|type
680 
681 PoisonWith(0xEF) for catching access to freed memory.
682 
683 =for apidoc Am|void|Poison|void* dest|int nitems|type
684 
685 PoisonWith(0xEF) for catching access to freed memory.
686 
687 =cut */
688 
689 /* Maintained for backwards-compatibility only. Use newSV() instead. */
690 #ifndef PERL_CORE
691 #define NEWSV(x,len)	newSV(len)
692 #endif
693 
694 #define MEM_SIZE_MAX ((MEM_SIZE)~0)
695 
696 /* The +0.0 in MEM_WRAP_CHECK_ is an attempt to foil
697  * overly eager compilers that will bleat about e.g.
698  * (U16)n > (size_t)~0/sizeof(U16) always being false. */
699 #ifdef PERL_MALLOC_WRAP
700 #define MEM_WRAP_CHECK(n,t) MEM_WRAP_CHECK_1(n,t,PL_memory_wrap)
701 #define MEM_WRAP_CHECK_1(n,t,a) \
702 	(void)(sizeof(t) > 1 && ((MEM_SIZE)(n)+0.0) > MEM_SIZE_MAX/sizeof(t) && (Perl_croak_nocontext(a),0))
703 #define MEM_WRAP_CHECK_(n,t) MEM_WRAP_CHECK(n,t),
704 
705 #define PERL_STRLEN_ROUNDUP(n) ((void)(((n) > MEM_SIZE_MAX - 2 * PERL_STRLEN_ROUNDUP_QUANTUM) ? (Perl_croak_nocontext(PL_memory_wrap),0):0),((n-1+PERL_STRLEN_ROUNDUP_QUANTUM)&~((MEM_SIZE)PERL_STRLEN_ROUNDUP_QUANTUM-1)))
706 
707 #else
708 
709 #define MEM_WRAP_CHECK(n,t)
710 #define MEM_WRAP_CHECK_1(n,t,a)
711 #define MEM_WRAP_CHECK_2(n,t,a,b)
712 #define MEM_WRAP_CHECK_(n,t)
713 
714 #define PERL_STRLEN_ROUNDUP(n) (((n-1+PERL_STRLEN_ROUNDUP_QUANTUM)&~((MEM_SIZE)PERL_STRLEN_ROUNDUP_QUANTUM-1)))
715 
716 #endif
717 
718 #ifdef PERL_MEM_LOG
719 /*
720  * If PERL_MEM_LOG is defined, all Newx()s, Renew()s, and Safefree()s
721  * go through functions, which are handy for debugging breakpoints, but
722  * which more importantly get the immediate calling environment (file and
723  * line number, and C function name if available) passed in.  This info can
724  * then be used for logging the calls, for which one gets a sample
725  * implementation if PERL_MEM_LOG_STDERR is defined.
726  *
727  * Known problems:
728  * - all memory allocs do not get logged, only those
729  *   that go through Newx() and derivatives (while all
730  *  Safefrees do get logged)
731  * - __FILE__ and __LINE__ do not work everywhere
732  * - __func__ or __FUNCTION__ even less so
733  * - I think more goes on after the perlio frees but
734  *   the thing is that STDERR gets closed (as do all
735  *   the file descriptors)
736  * - no deeper calling stack than the caller of the Newx()
737  *   or the kind, but do I look like a C reflection/introspection
738  *   utility to you?
739  * - the function prototypes for the logging functions
740  *   probably should maybe be somewhere else than handy.h
741  * - one could consider inlining (macrofying) the logging
742  *   for speed, but I am too lazy
743  * - one could imagine recording the allocations in a hash,
744  *   (keyed by the allocation address?), and maintain that
745  *   through reallocs and frees, but how to do that without
746  *   any News() happening...?
747  */
748 
749 Malloc_t Perl_mem_log_alloc(const UV n, const UV typesize, const char *typename, Malloc_t newalloc, const char *filename, const int linenumber, const char *funcname);
750 
751 Malloc_t Perl_mem_log_realloc(const UV n, const UV typesize, const char *typename, Malloc_t oldalloc, Malloc_t newalloc, const char *filename, const int linenumber, const char *funcname);
752 
753 Malloc_t Perl_mem_log_free(Malloc_t oldalloc, const char *filename, const int linenumber, const char *funcname);
754 
755 #endif
756 
757 #ifdef PERL_MEM_LOG
758 #define MEM_LOG_ALLOC(n,t,a)     Perl_mem_log_alloc(n,sizeof(t),STRINGIFY(t),a,__FILE__,__LINE__,FUNCTION__)
759 #define MEM_LOG_REALLOC(n,t,v,a) Perl_mem_log_realloc(n,sizeof(t),STRINGIFY(t),v,a,__FILE__,__LINE__,FUNCTION__)
760 #define MEM_LOG_FREE(a)          Perl_mem_log_free(a,__FILE__,__LINE__,FUNCTION__)
761 #endif
762 
763 #ifndef MEM_LOG_ALLOC
764 #define MEM_LOG_ALLOC(n,t,a)     (a)
765 #endif
766 #ifndef MEM_LOG_REALLOC
767 #define MEM_LOG_REALLOC(n,t,v,a) (a)
768 #endif
769 #ifndef MEM_LOG_FREE
770 #define MEM_LOG_FREE(a)          (a)
771 #endif
772 
773 #define Newx(v,n,t)	(v = (MEM_WRAP_CHECK_(n,t) MEM_LOG_ALLOC(n,t,(t*)safemalloc((MEM_SIZE)((n)*sizeof(t))))))
774 #define Newxc(v,n,t,c)	(v = (MEM_WRAP_CHECK_(n,t) MEM_LOG_ALLOC(n,t,(c*)safemalloc((MEM_SIZE)((n)*sizeof(t))))))
775 #define Newxz(v,n,t)	(v = (MEM_WRAP_CHECK_(n,t) MEM_LOG_ALLOC(n,t,(t*)safecalloc((n),sizeof(t)))))
776 
777 #ifndef PERL_CORE
778 /* pre 5.9.x compatibility */
779 #define New(x,v,n,t)	Newx(v,n,t)
780 #define Newc(x,v,n,t,c)	Newxc(v,n,t,c)
781 #define Newz(x,v,n,t)	Newxz(v,n,t)
782 #endif
783 
784 #define Renew(v,n,t) \
785 	  (v = (MEM_WRAP_CHECK_(n,t) MEM_LOG_REALLOC(n,t,v,(t*)saferealloc((Malloc_t)(v),(MEM_SIZE)((n)*sizeof(t))))))
786 #define Renewc(v,n,t,c) \
787 	  (v = (MEM_WRAP_CHECK_(n,t) MEM_LOG_REALLOC(n,t,v,(c*)saferealloc((Malloc_t)(v),(MEM_SIZE)((n)*sizeof(t))))))
788 
789 #ifdef PERL_POISON
790 #define Safefree(d) \
791   ((d) ? (void)(safefree(MEM_LOG_FREE((Malloc_t)(d))), Poison(&(d), 1, Malloc_t)) : (void) 0)
792 #else
793 #define Safefree(d)	safefree(MEM_LOG_FREE((Malloc_t)(d)))
794 #endif
795 
796 #define Move(s,d,n,t)	(MEM_WRAP_CHECK_(n,t) (void)memmove((char*)(d),(const char*)(s), (n) * sizeof(t)))
797 #define Copy(s,d,n,t)	(MEM_WRAP_CHECK_(n,t) (void)memcpy((char*)(d),(const char*)(s), (n) * sizeof(t)))
798 #define Zero(d,n,t)	(MEM_WRAP_CHECK_(n,t) (void)memzero((char*)(d), (n) * sizeof(t)))
799 
800 #define MoveD(s,d,n,t)	(MEM_WRAP_CHECK_(n,t) memmove((char*)(d),(const char*)(s), (n) * sizeof(t)))
801 #define CopyD(s,d,n,t)	(MEM_WRAP_CHECK_(n,t) memcpy((char*)(d),(const char*)(s), (n) * sizeof(t)))
802 #ifdef HAS_MEMSET
803 #define ZeroD(d,n,t)	(MEM_WRAP_CHECK_(n,t) memzero((char*)(d), (n) * sizeof(t)))
804 #else
805 /* Using bzero(), which returns void.  */
806 #define ZeroD(d,n,t)	(MEM_WRAP_CHECK_(n,t) memzero((char*)(d), (n) * sizeof(t)),d)
807 #endif
808 
809 #define PoisonWith(d,n,t,b)	(MEM_WRAP_CHECK_(n,t) (void)memset((char*)(d), (U8)(b), (n) * sizeof(t)))
810 #define PoisonNew(d,n,t)	PoisonWith(d,n,t,0xAB)
811 #define PoisonFree(d,n,t)	PoisonWith(d,n,t,0xEF)
812 #define Poison(d,n,t)		PoisonFree(d,n,t)
813 
814 #ifdef USE_STRUCT_COPY
815 #define StructCopy(s,d,t) (*((t*)(d)) = *((t*)(s)))
816 #else
817 #define StructCopy(s,d,t) Copy(s,d,1,t)
818 #endif
819 
820 #define C_ARRAY_LENGTH(a)	(sizeof(a)/sizeof((a)[0]))
821 
822 #ifdef NEED_VA_COPY
823 # ifdef va_copy
824 #  define Perl_va_copy(s, d) va_copy(d, s)
825 # else
826 #  if defined(__va_copy)
827 #   define Perl_va_copy(s, d) __va_copy(d, s)
828 #  else
829 #   define Perl_va_copy(s, d) Copy(s, d, 1, va_list)
830 #  endif
831 # endif
832 #endif
833 
834 /* convenience debug macros */
835 #ifdef USE_ITHREADS
836 #define pTHX_FORMAT  "Perl interpreter: 0x%p"
837 #define pTHX__FORMAT ", Perl interpreter: 0x%p"
838 #define pTHX_VALUE_   (void *)my_perl,
839 #define pTHX_VALUE    (void *)my_perl
840 #define pTHX__VALUE_ ,(void *)my_perl,
841 #define pTHX__VALUE  ,(void *)my_perl
842 #else
843 #define pTHX_FORMAT
844 #define pTHX__FORMAT
845 #define pTHX_VALUE_
846 #define pTHX_VALUE
847 #define pTHX__VALUE_
848 #define pTHX__VALUE
849 #endif /* USE_ITHREADS */
850 
851