[perl5.git] / utf8.h

/*    utf8.h
 *
 *    Copyright (c) 1998-2001, Larry Wall
 *
 *    You may distribute under the terms of either the GNU General Public
 *    License or the Artistic License, as specified in the README file.
 *
 */

/* Use UTF-8 as the default script encoding?
 * Turning this on will break scripts having non-UTF8 binary
 * data (such as Latin-1) in string literals. */
#ifdef USE_UTF8_SCRIPTS
#    define USE_UTF8_IN_NAMES (!IN_BYTES)
#else
#    define USE_UTF8_IN_NAMES (PL_hints & HINT_UTF8)
#endif

#ifdef EBCDIC
/* The equivalent of these macros but implementing UTF-EBCDIC
   are in the following header file:
 */

#include "utfebcdic.h"
#else
START_EXTERN_C

#ifdef DOINIT
EXTCONST unsigned char PL_utf8skip[] = {
1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus */
1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus */
2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, /* scripts */
3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,5,5,5,5,6,6,	 /* cjk etc. */
7,13, /* Perl extended (not UTF-8).  Up to 72bit allowed (64-bit + reserved). */
};
#else
EXTCONST unsigned char PL_utf8skip[];
#endif

END_EXTERN_C
#define UTF8SKIP(s) PL_utf8skip[*(U8*)s]

/* Native character to iso-8859-1 */
#define NATIVE_TO_ASCII(ch)      (ch)
#define ASCII_TO_NATIVE(ch)      (ch)
/* Transform after encoding */
#define NATIVE_TO_UTF(ch)        (ch)
#define UTF_TO_NATIVE(ch)        (ch)
/* Transforms in wide UV chars */
#define UNI_TO_NATIVE(ch)        (ch)
#define NATIVE_TO_UNI(ch)        (ch)
/* Transforms in invariant space */
#define NATIVE_TO_NEED(enc,ch)   (ch)
#define ASCII_TO_NEED(enc,ch)    (ch)

/* As there are no translations avoid the function wrapper */
#define Perl_utf8n_to_uvchr Perl_utf8n_to_uvuni
#define Perl_uvchr_to_utf8  Perl_uvuni_to_utf8

/*

 The following table is from Unicode 3.1.

 Code Points		1st Byte  2nd Byte  3rd Byte  4th Byte

   U+0000..U+007F	00..7F   
   U+0080..U+07FF	C2..DF    80..BF   
   U+0800..U+0FFF	E0        A0..BF    80..BF  
   U+1000..U+FFFF	E1..EF    80..BF    80..BF  
  U+10000..U+3FFFF	F0        90..BF    80..BF    80..BF
  U+40000..U+FFFFF	F1..F3    80..BF    80..BF    80..BF
 U+100000..U+10FFFF	F4        80..8F    80..BF    80..BF

 */

#define UNI_IS_INVARIANT(c)		(((UV)c) <  0x80)
#define UTF8_IS_INVARIANT(c)		UNI_IS_INVARIANT(NATIVE_TO_UTF(c))
#define NATIVE_IS_INVARIANT(c)		UNI_IS_INVARIANT(NATIVE_TO_ASCII(c))
#define UTF8_IS_START(c)		(((U8)c) >= 0xc0 && (((U8)c) <= 0xfd))
#define UTF8_IS_CONTINUATION(c)		(((U8)c) >= 0x80 && (((U8)c) <= 0xbf))
#define UTF8_IS_CONTINUED(c) 		(((U8)c) &  0x80)
#define UTF8_IS_DOWNGRADEABLE_START(c)	(((U8)c & 0xfc) == 0xc0)

#define UTF_START_MARK(len) ((len >  7) ? 0xFF : (0xFE << (7-len)))
#define UTF_START_MASK(len) ((len >= 7) ? 0x00 : (0x1F >> (len-2)))

#define UTF_CONTINUATION_MARK		0x80
#define UTF_ACCUMULATION_SHIFT		6
#define UTF_CONTINUATION_MASK		((U8)0x3f)
#define UTF8_ACCUMULATE(old, new)	(((old) << UTF_ACCUMULATION_SHIFT) | (((U8)new) & UTF_CONTINUATION_MASK))

#define UTF8_EIGHT_BIT_HI(c)	((((U8)(c))>>UTF_ACCUMULATION_SHIFT)|UTF_START_MARK(2))
#define UTF8_EIGHT_BIT_LO(c)	(((((U8)(c)))&UTF_CONTINUATION_MASK)|UTF_CONTINUATION_MARK)

#ifdef HAS_QUAD
#define UNISKIP(uv) ( (uv) < 0x80           ? 1 : \
		      (uv) < 0x800          ? 2 : \
		      (uv) < 0x10000        ? 3 : \
		      (uv) < 0x200000       ? 4 : \
		      (uv) < 0x4000000      ? 5 : \
		      (uv) < 0x80000000     ? 6 : \
                      (uv) < UTF8_QUAD_MAX ? 7 : 13 )
#else
/* No, I'm not even going to *TRY* putting #ifdef inside a #define */
#define UNISKIP(uv) ( (uv) < 0x80           ? 1 : \
		      (uv) < 0x800          ? 2 : \
		      (uv) < 0x10000        ? 3 : \
		      (uv) < 0x200000       ? 4 : \
		      (uv) < 0x4000000      ? 5 : \
		      (uv) < 0x80000000     ? 6 : 7 )
#endif

/*
 * Note: we try to be careful never to call the isXXX_utf8() functions
 * unless we're pretty sure we've seen the beginning of a UTF-8 character
 * (that is, the two high bits are set).  Otherwise we risk loading in the
 * heavy-duty SWASHINIT and SWASHGET routines unnecessarily.
 */
#define isIDFIRST_lazy_if(p,c) ((IN_BYTES || (!c || (*((U8*)p) < 0xc0))) \
				? isIDFIRST(*(p)) \
				: isIDFIRST_utf8((U8*)p))
#define isALNUM_lazy_if(p,c)   ((IN_BYTES || (!c || (*((U8*)p) < 0xc0))) \
				? isALNUM(*(p)) \
				: isALNUM_utf8((U8*)p))


#endif /* EBCDIC vs ASCII */

/* Rest of these are attributes of Unicode and perl's internals rather than the encoding */

#define isIDFIRST_lazy(p)	isIDFIRST_lazy_if(p,1)
#define isALNUM_lazy(p)		isALNUM_lazy_if(p,1)

#define UTF8_MAXLEN 13 /* how wide can a single UTF8 encoded character become */

#define IN_BYTES (PL_curcop->op_private & HINT_BYTES)
#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)

#define UTF8_ALLOW_EMPTY		0x0001
#define UTF8_ALLOW_CONTINUATION		0x0002
#define UTF8_ALLOW_NON_CONTINUATION	0x0004
#define UTF8_ALLOW_FE_FF		0x0008
#define UTF8_ALLOW_SHORT		0x0010
#define UTF8_ALLOW_SURROGATE		0x0020
#define UTF8_ALLOW_BOM			0x0040
#define UTF8_ALLOW_FFFF			0x0080
#define UTF8_ALLOW_LONG			0x0100
#define UTF8_ALLOW_ANYUV		(UTF8_ALLOW_EMPTY|UTF8_ALLOW_FE_FF|\
					 UTF8_ALLOW_SURROGATE|UTF8_ALLOW_BOM|\
					 UTF8_ALLOW_FFFF|UTF8_ALLOW_LONG)
#define UTF8_ALLOW_ANY			0x00ff
#define UTF8_CHECK_ONLY			0x0200

#define UNICODE_SURROGATE_FIRST		0xd800
#define UNICODE_SURROGATE_LAST		0xdfff
#define UNICODE_REPLACEMENT		0xfffd
#define UNICODE_BYTER_ORDER_MARK	0xfffe
#define UNICODE_ILLEGAL			0xffff

#define UNICODE_IS_SURROGATE(c)		((c) >= UNICODE_SURROGATE_FIRST && \
					 (c) <= UNICODE_SURROGATE_LAST)
#define UNICODE_IS_REPLACEMENT(c)	((c) == UNICODE_REPLACEMENT)
#define UNICODE_IS_BYTE_ORDER_MARK(c)	((c) == UNICODE_BYTER_ORDER_MARK)
#define UNICODE_IS_ILLEGAL(c)		((c) == UNICODE_ILLEGAL)

#ifdef HAS_QUAD
#    define UTF8_QUAD_MAX	UINT64_C(0x1000000000)
#endif

#define UTF8_IS_ASCII(c) UTF8_IS_INVARIANT(c)
Commit	Line	Data
a0ed51b3 LW	1	/* utf8.h
a0ed51b3 LW	2	*
bc89e66f	3	* Copyright (c) 1998-2001, Larry Wall
a0ed51b3 LW	4	*
	5	* You may distribute under the terms of either the GNU General Public
	6	* License or the Artistic License, as specified in the README file.
	7	*
	8	*/
	9
39e02b42 JH	10	/* Use UTF-8 as the default script encoding?
	11	* Turning this on will break scripts having non-UTF8 binary
	12	* data (such as Latin-1) in string literals. */
	13	#ifdef USE_UTF8_SCRIPTS
	14	# define USE_UTF8_IN_NAMES (!IN_BYTES)
	15	#else
	16	# define USE_UTF8_IN_NAMES (PL_hints & HINT_UTF8)
	17	#endif
	18
1d72bdf6 NIS	19	#ifdef EBCDIC
	20	/* The equivalent of these macros but implementing UTF-EBCDIC
	21	are in the following header file:
	22	*/
	23
	24	#include "utfebcdic.h"
	25	#else
73c4f7a1 GS	26	START_EXTERN_C
73c4f7a1 GS	27
a0ed51b3	28	#ifdef DOINIT
6f06b55f	29	EXTCONST unsigned char PL_utf8skip[] = {
a0ed51b3 LW	30	1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
	31	1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
	32	1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
	33	1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
	34	1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus */
	35	1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus */
	36	2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, /* scripts */
3c77ea2b GS	37	3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,5,5,5,5,6,6, /* cjk etc. */
3c77ea2b GS	38	7,13, /* Perl extended (not UTF-8). Up to 72bit allowed (64-bit + reserved). */
a0ed51b3 LW	39	};
a0ed51b3 LW	40	#else
6f06b55f	41	EXTCONST unsigned char PL_utf8skip[];
a0ed51b3 LW	42	#endif
a0ed51b3 LW	43
73c4f7a1	44	END_EXTERN_C
6f06b55f	45	#define UTF8SKIP(s) PL_utf8skip[(U8)s]
7e2040f0	46
1d72bdf6 NIS	47	/* Native character to iso-8859-1 */
	48	#define NATIVE_TO_ASCII(ch) (ch)
	49	#define ASCII_TO_NATIVE(ch) (ch)
	50	/* Transform after encoding */
	51	#define NATIVE_TO_UTF(ch) (ch)
	52	#define UTF_TO_NATIVE(ch) (ch)
	53	/* Transforms in wide UV chars */
	54	#define UNI_TO_NATIVE(ch) (ch)
	55	#define NATIVE_TO_UNI(ch) (ch)
	56	/* Transforms in invariant space */
	57	#define NATIVE_TO_NEED(enc,ch) (ch)
	58	#define ASCII_TO_NEED(enc,ch) (ch)
d7578b48	59
2b9d42f0 NIS	60	/* As there are no translations avoid the function wrapper */
	61	#define Perl_utf8n_to_uvchr Perl_utf8n_to_uvuni
	62	#define Perl_uvchr_to_utf8 Perl_uvuni_to_utf8
	63
877d9f0d	64	/*
9041c2e3	65
877d9f0d JH	66	The following table is from Unicode 3.1.
	67
	68	Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
	69
	70	U+0000..U+007F 00..7F
	71	U+0080..U+07FF C2..DF 80..BF
	72	U+0800..U+0FFF E0 A0..BF 80..BF
	73	U+1000..U+FFFF E1..EF 80..BF 80..BF
	74	U+10000..U+3FFFF F0 90..BF 80..BF 80..BF
	75	U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
	76	U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
	77
	78	*/
	79
c4d5f83a NIS	80	#define UNI_IS_INVARIANT(c) (((UV)c) < 0x80)
	81	#define UTF8_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_TO_UTF(c))
	82	#define NATIVE_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_TO_ASCII(c))
c512ce4f JH	83	#define UTF8_IS_START(c) (((U8)c) >= 0xc0 && (((U8)c) <= 0xfd))
	84	#define UTF8_IS_CONTINUATION(c) (((U8)c) >= 0x80 && (((U8)c) <= 0xbf))
	85	#define UTF8_IS_CONTINUED(c) (((U8)c) & 0x80)
db42d148	86	#define UTF8_IS_DOWNGRADEABLE_START(c) (((U8)c & 0xfc) == 0xc0)
8850bf83	87
1d72bdf6 NIS	88	#define UTF_START_MARK(len) ((len > 7) ? 0xFF : (0xFE << (7-len)))
	89	#define UTF_START_MASK(len) ((len >= 7) ? 0x00 : (0x1F >> (len-2)))
	90
	91	#define UTF_CONTINUATION_MARK 0x80
	92	#define UTF_ACCUMULATION_SHIFT 6
	93	#define UTF_CONTINUATION_MASK ((U8)0x3f)
cb359b41	94	#define UTF8_ACCUMULATE(old, new) (((old) << UTF_ACCUMULATION_SHIFT) \| (((U8)new) & UTF_CONTINUATION_MASK))
421a8bf2	95
1d72bdf6 NIS	96	#define UTF8_EIGHT_BIT_HI(c) ((((U8)(c))>>UTF_ACCUMULATION_SHIFT)\|UTF_START_MARK(2))
1d72bdf6 NIS	97	#define UTF8_EIGHT_BIT_LO(c) (((((U8)(c)))&UTF_CONTINUATION_MASK)\|UTF_CONTINUATION_MARK)
c512ce4f	98
1d68d6cd	99	#ifdef HAS_QUAD
5bbb0b5a	100	#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd SC	101	(uv) < 0x800 ? 2 : \
	102	(uv) < 0x10000 ? 3 : \
	103	(uv) < 0x200000 ? 4 : \
	104	(uv) < 0x4000000 ? 5 : \
	105	(uv) < 0x80000000 ? 6 : \
9041c2e3	106	(uv) < UTF8_QUAD_MAX ? 7 : 13 )
1d68d6cd SC	107	#else
1d68d6cd SC	108	/* No, I'm not even going to TRY putting #ifdef inside a #define */
5bbb0b5a	109	#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd SC	110	(uv) < 0x800 ? 2 : \
	111	(uv) < 0x10000 ? 3 : \
	112	(uv) < 0x200000 ? 4 : \
	113	(uv) < 0x4000000 ? 5 : \
	114	(uv) < 0x80000000 ? 6 : 7 )
	115	#endif
	116
7e2040f0 GS	117	/*
	118	* Note: we try to be careful never to call the isXXX_utf8() functions
	119	* unless we're pretty sure we've seen the beginning of a UTF-8 character
	120	* (that is, the two high bits are set). Otherwise we risk loading in the
	121	* heavy-duty SWASHINIT and SWASHGET routines unnecessarily.
	122	*/
0064a8a9	123	#define isIDFIRST_lazy_if(p,c) ((IN_BYTES \|\| (!c \|\| (((U8)p) < 0xc0))) \
7e2040f0 GS	124	? isIDFIRST(*(p)) \
7e2040f0 GS	125	: isIDFIRST_utf8((U8*)p))
0064a8a9	126	#define isALNUM_lazy_if(p,c) ((IN_BYTES \|\| (!c \|\| (((U8)p) < 0xc0))) \
7e2040f0 GS	127	? isALNUM(*(p)) \
7e2040f0 GS	128	: isALNUM_utf8((U8*)p))
1d72bdf6 NIS	129
	130
	131	#endif /* EBCDIC vs ASCII */
	132
	133	/* Rest of these are attributes of Unicode and perl's internals rather than the encoding */
	134
7e2040f0 GS	135	#define isIDFIRST_lazy(p) isIDFIRST_lazy_if(p,1)
7e2040f0 GS	136	#define isALNUM_lazy(p) isALNUM_lazy_if(p,1)
3bd709b1	137
1d72bdf6	138	#define UTF8_MAXLEN 13 /* how wide can a single UTF8 encoded character become */
3bd709b1	139
0064a8a9 JH	140	#define IN_BYTES (PL_curcop->op_private & HINT_BYTES)
0064a8a9 JH	141	#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)
1d72bdf6 NIS	142
	143	#define UTF8_ALLOW_EMPTY 0x0001
	144	#define UTF8_ALLOW_CONTINUATION 0x0002
	145	#define UTF8_ALLOW_NON_CONTINUATION 0x0004
	146	#define UTF8_ALLOW_FE_FF 0x0008
	147	#define UTF8_ALLOW_SHORT 0x0010
	148	#define UTF8_ALLOW_SURROGATE 0x0020
	149	#define UTF8_ALLOW_BOM 0x0040
	150	#define UTF8_ALLOW_FFFF 0x0080
	151	#define UTF8_ALLOW_LONG 0x0100
	152	#define UTF8_ALLOW_ANYUV (UTF8_ALLOW_EMPTY\|UTF8_ALLOW_FE_FF\|\
	153	UTF8_ALLOW_SURROGATE\|UTF8_ALLOW_BOM\|\
	154	UTF8_ALLOW_FFFF\|UTF8_ALLOW_LONG)
	155	#define UTF8_ALLOW_ANY 0x00ff
e83d50c9	156	#define UTF8_CHECK_ONLY 0x0200
1d72bdf6 NIS	157
	158	#define UNICODE_SURROGATE_FIRST 0xd800
	159	#define UNICODE_SURROGATE_LAST 0xdfff
	160	#define UNICODE_REPLACEMENT 0xfffd
	161	#define UNICODE_BYTER_ORDER_MARK 0xfffe
	162	#define UNICODE_ILLEGAL 0xffff
	163
	164	#define UNICODE_IS_SURROGATE(c) ((c) >= UNICODE_SURROGATE_FIRST && \
	165	(c) <= UNICODE_SURROGATE_LAST)
a10ec373	166	#define UNICODE_IS_REPLACEMENT(c) ((c) == UNICODE_REPLACEMENT)
1d72bdf6 NIS	167	#define UNICODE_IS_BYTE_ORDER_MARK(c) ((c) == UNICODE_BYTER_ORDER_MARK)
	168	#define UNICODE_IS_ILLEGAL(c) ((c) == UNICODE_ILLEGAL)
	169
f067b878 NA	170	#ifdef HAS_QUAD
	171	# define UTF8_QUAD_MAX UINT64_C(0x1000000000)
	172	#endif
3bd709b1	173
1d72bdf6	174	#define UTF8_IS_ASCII(c) UTF8_IS_INVARIANT(c)