[perl5.git] / utf8.h

/*    utf8.h
 *
 *    Copyright (C) 2000, 2001, 2002, by Larry Wall and others
 *
 *    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-UTF-8 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[*(const 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 3.2.

 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+CFFF       E1..EC    80..BF    80..BF
   U+D000..U+D7FF       ED        80..9F    80..BF
   U+D800..U+DFFF       ******* ill-formed *******
   U+E000..U+FFFF       EE..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

Note the A0..BF in U+0800..U+0FFF, the 80..9F in U+D000...U+D7FF,
the 90..BF in U+10000..U+3FFFF, and the 80...8F in U+100000..U+10FFFF.
The "gaps" are caused by legal UTF-8 avoiding non-shortest encodings:
it is technically possible to UTF-8-encode a single code point in different
ways, but that is explicitly forbidden, and the shortest possible encoding
should always be used (and that is what Perl does).

 */

/*
 Another way to look at it, as bits:

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

                    0aaaaaaa     0aaaaaaa
            00000bbbbbaaaaaa     110bbbbb  10aaaaaa
            ccccbbbbbbaaaaaa     1110cccc  10bbbbbb  10aaaaaa
  00000dddccccccbbbbbbaaaaaa     11110ddd  10cccccc  10bbbbbb  10aaaaaa

As you can see, the continuation bytes all begin with C<10>, and the
leading bits of the start byte tell how many bytes the are in the
encoded character.

*/


#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 || (*((const U8*)p) < 0xc0))) \
				? isIDFIRST(*(p)) \
				: isIDFIRST_utf8((const U8*)p))
#define isALNUM_lazy_if(p,c)   ((IN_BYTES || (!c || (*((const U8*)p) < 0xc0))) \
				? isALNUM(*(p)) \
				: isALNUM_utf8((const 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_MAXBYTES 13
/* How wide can a single UTF-8 encoded character become in bytes.
 * NOTE: Strictly speaking Perl's UTF-8 should not be called UTF-8
 * since UTF-8 is an encoding of Unicode and given Unicode's current
 * upper limit only four bytes is possible.  Perl thinks of UTF-8
 * as a way to encode non-negative integers in a binary format. */
#define UTF8_MAXLEN UTF8_MAXBYTES

#define UTF8_MAXLEN_UCLC 3		/* Obsolete, do not use. */
#define UTF8_MAXLEN_UCLC_MULT 39	/* Obsolete, do not use. */
#define UTF8_MAXLEN_FOLD 3		/* Obsolete, do not use. */
#define UTF8_MAXLEN_FOLD_MULT 39	/* Obsolete, do not use. */

/* The maximum number of UTF-8 bytes a single Unicode character can
 * uppercase/lowercase/fold into; this number depends on the Unicode
 * version.  An example of maximal expansion is the U+03B0 which
 * uppercases to U+03C5 U+0308 U+0301.  The Unicode databases that
 * tell these things are UnicodeDatabase.txt, CaseFolding.txt, and
 * SpecialCasing.txt. */
#define UTF8_MAXBYTES_CASE	6

#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_FFFF			0x0040 /* Allows also FFFE. */
#define UTF8_ALLOW_LONG			0x0080
#define UTF8_ALLOW_ANYUV		(UTF8_ALLOW_EMPTY|UTF8_ALLOW_FE_FF|\
					 UTF8_ALLOW_SURROGATE|UTF8_ALLOW_FFFF)
#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_BYTE_ORDER_MARK		0xFEFF
#define UNICODE_ILLEGAL			0xFFFF

/* Though our UTF-8 encoding can go beyond this,
 * let's be conservative and do as Unicode 3.2 says. */
#define PERL_UNICODE_MAX	0x10FFFF

#define UNICODE_ALLOW_SURROGATE 0x0001	/* Allow UTF-16 surrogates (EVIL) */
#define UNICODE_ALLOW_FDD0	0x0002	/* Allow the U+FDD0...U+FDEF */
#define UNICODE_ALLOW_FFFF	0x0004	/* Allow 0xFFF[EF], 0x1FFF[EF], ... */
#define UNICODE_ALLOW_SUPER	0x0008	/* Allow past 10xFFFF */
#define UNICODE_ALLOW_ANY	0x000F

#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_BYTE_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)

#define UNICODE_LATIN_SMALL_LETTER_SHARP_S	0x00DF
#define UNICODE_GREEK_CAPITAL_LETTER_SIGMA	0x03A3
#define UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA	0x03C2
#define UNICODE_GREEK_SMALL_LETTER_SIGMA	0x03C3

#define EBCDIC_LATIN_SMALL_LETTER_SHARP_S	0x0059

#define UNI_DISPLAY_ISPRINT	0x0001
#define UNI_DISPLAY_BACKSLASH	0x0002
#define UNI_DISPLAY_QQ		(UNI_DISPLAY_ISPRINT|UNI_DISPLAY_BACKSLASH)
#define UNI_DISPLAY_REGEX	(UNI_DISPLAY_ISPRINT|UNI_DISPLAY_BACKSLASH)

#ifdef EBCDIC
#   define ANYOF_FOLD_SHARP_S(node, input, end)	\
	(ANYOF_BITMAP_TEST(node, EBCDIC_LATIN_SMALL_LETTER_SHARP_S) && \
	 (ANYOF_FLAGS(node) & ANYOF_UNICODE) && \
	 (ANYOF_FLAGS(node) & ANYOF_FOLD) && \
	 ((end) > (input) + 1) && \
	 toLOWER((input)[0]) == 's' && \
	 toLOWER((input)[1]) == 's')
#else
#   define ANYOF_FOLD_SHARP_S(node, input, end)	\
	(ANYOF_BITMAP_TEST(node, UNICODE_LATIN_SMALL_LETTER_SHARP_S) && \
	 (ANYOF_FLAGS(node) & ANYOF_UNICODE) && \
	 (ANYOF_FLAGS(node) & ANYOF_FOLD) && \
	 ((end) > (input) + 1) && \
	 toLOWER((input)[0]) == 's' && \
	 toLOWER((input)[1]) == 's')
#endif
#define SHARP_S_SKIP 2

#ifdef EBCDIC
/* IS_UTF8_CHAR() is not ported to EBCDIC */
#else
#define IS_UTF8_CHAR_1(p)	\
	((p)[0] <= 0x7F)
#define IS_UTF8_CHAR_2(p)	\
	((p)[0] >= 0xC2 && (p)[0] <= 0xDF && \
	 (p)[1] >= 0x80 && (p)[1] <= 0xBF)
#define IS_UTF8_CHAR_3a(p)	\
	((p)[0] == 0xE0 && \
	 (p)[1] >= 0xA0 && (p)[1] <= 0xBF && \
	 (p)[2] >= 0x80 && (p)[2] <= 0xBF)
#define IS_UTF8_CHAR_3b(p)	\
	((p)[0] >= 0xE1 && (p)[0] <= 0xEC && \
	 (p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	 (p)[2] >= 0x80 && (p)[2] <= 0xBF)
#define IS_UTF8_CHAR_3c(p)	\
	((p)[0] == 0xED && \
	 (p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	 (p)[2] >= 0x80 && (p)[2] <= 0xBF)
/* In IS_UTF8_CHAR_3c(p) one could use
 * (p)[1] >= 0x80 && (p)[1] <= 0x9F
 * if one wanted to exclude surrogates. */
#define IS_UTF8_CHAR_3d(p)	\
	((p)[0] >= 0xEE && (p)[0] <= 0xEF && \
	 (p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	 (p)[2] >= 0x80 && (p)[2] <= 0xBF)
#define IS_UTF8_CHAR_4a(p)	\
	((p)[0] == 0xF0 && \
	 (p)[1] >= 0x90 && (p)[1] <= 0xBF && \
	 (p)[2] >= 0x80 && (p)[2] <= 0xBF && \
	 (p)[3] >= 0x80 && (p)[3] <= 0xBF)
#define IS_UTF8_CHAR_4b(p)	\
	((p)[0] >= 0xF1 && (p)[0] <= 0xF3 && \
	 (p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	 (p)[2] >= 0x80 && (p)[2] <= 0xBF && \
	 (p)[3] >= 0x80 && (p)[3] <= 0xBF)
/* In IS_UTF8_CHAR_4c(p) one could use
 * (p)[0] == 0xF4
 * if one wanted to stop at the Unicode limit U+10FFFF.
 * The 0xF7 allows us to go to 0x1fffff (0x200000 would
 * require five bytes).  Not doing any further code points
 * since that is not needed (and that would not be strict
 * UTF-8, anyway).  The "slow path" in Perl_is_utf8_char()
 * will take care of the "extended UTF-8". */
#define IS_UTF8_CHAR_4c(p)	\
	((p)[0] == 0xF4 && (p)[0] <= 0xF7 && \
	 (p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	 (p)[2] >= 0x80 && (p)[2] <= 0xBF && \
	 (p)[3] >= 0x80 && (p)[3] <= 0xBF)

#define IS_UTF8_CHAR_3(p)	\
	(IS_UTF8_CHAR_3a(p) || \
	 IS_UTF8_CHAR_3b(p) || \
	 IS_UTF8_CHAR_3c(p) || \
	 IS_UTF8_CHAR_3d(p))
#define IS_UTF8_CHAR_4(p)	\
	(IS_UTF8_CHAR_4a(p) || \
	 IS_UTF8_CHAR_4b(p) || \
	 IS_UTF8_CHAR_4c(p))

/* IS_UTF8_CHAR(p) is strictly speaking wrong (not UTF-8) because it
 * (1) allows UTF-8 encoded UTF-16 surrogates
 * (2) it allows code points past U+10FFFF.
 * The Perl_is_utf8_char() full "slow" code will handle the Perl
 * "extended UTF-8". */
#define IS_UTF8_CHAR(p, n)	\
	((n) == 1 ? IS_UTF8_CHAR_1(p) : \
 	 (n) == 2 ? IS_UTF8_CHAR_2(p) : \
	 (n) == 3 ? IS_UTF8_CHAR_3(p) : \
	 (n) == 4 ? IS_UTF8_CHAR_4(p) : 0)

#define IS_UTF8_CHAR_FAST(n) ((n) <= 4)

#endif /* IS_UTF8_CHAR() for UTF-8 */
Commit	Line	Data
a0ed51b3 LW	1	/* utf8.h
a0ed51b3 LW	2	*
4bb101f2	3	* Copyright (C) 2000, 2001, 2002, by Larry Wall and others
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	10	/* Use UTF-8 as the default script encoding?
1e54db1a	11	* Turning this on will break scripts having non-UTF-8 binary
39e02b42 JH	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
27da23d5	45	#define UTF8SKIP(s) PL_utf8skip[(const 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
8c007b5a	66	The following table is from Unicode 3.2.
877d9f0d JH	67
	68	Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
	69
375122d7 JH	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+CFFF E1..EC 80..BF 80..BF
	74	U+D000..U+D7FF ED 80..9F 80..BF
8c007b5a	75	U+D800..U+DFFF ***** ill-formed *****
375122d7	76	U+E000..U+FFFF EE..EF 80..BF 80..BF
877d9f0d JH	77	U+10000..U+3FFFF F0 90..BF 80..BF 80..BF
	78	U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
	79	U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
	80
8c007b5a JH	81	Note the A0..BF in U+0800..U+0FFF, the 80..9F in U+D000...U+D7FF,
8c007b5a JH	82	the 90..BF in U+10000..U+3FFFF, and the 80...8F in U+100000..U+10FFFF.
37361303 JH	83	The "gaps" are caused by legal UTF-8 avoiding non-shortest encodings:
	84	it is technically possible to UTF-8-encode a single code point in different
	85	ways, but that is explicitly forbidden, and the shortest possible encoding
	86	should always be used (and that is what Perl does).
8c007b5a	87
877d9f0d JH	88	*/
877d9f0d JH	89
8c007b5a JH	90	/*
	91	Another way to look at it, as bits:
	92
	93	Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
	94
	95	0aaaaaaa 0aaaaaaa
	96	00000bbbbbaaaaaa 110bbbbb 10aaaaaa
	97	ccccbbbbbbaaaaaa 1110cccc 10bbbbbb 10aaaaaa
	98	00000dddccccccbbbbbbaaaaaa 11110ddd 10cccccc 10bbbbbb 10aaaaaa
	99
	100	As you can see, the continuation bytes all begin with C<10>, and the
	101	leading bits of the start byte tell how many bytes the are in the
	102	encoded character.
	103
	104	*/
	105
	106
c4d5f83a NIS	107	#define UNI_IS_INVARIANT(c) (((UV)c) < 0x80)
	108	#define UTF8_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_TO_UTF(c))
	109	#define NATIVE_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_TO_ASCII(c))
c512ce4f JH	110	#define UTF8_IS_START(c) (((U8)c) >= 0xc0 && (((U8)c) <= 0xfd))
	111	#define UTF8_IS_CONTINUATION(c) (((U8)c) >= 0x80 && (((U8)c) <= 0xbf))
	112	#define UTF8_IS_CONTINUED(c) (((U8)c) & 0x80)
db42d148	113	#define UTF8_IS_DOWNGRADEABLE_START(c) (((U8)c & 0xfc) == 0xc0)
8850bf83	114
1d72bdf6 NIS	115	#define UTF_START_MARK(len) ((len > 7) ? 0xFF : (0xFE << (7-len)))
	116	#define UTF_START_MASK(len) ((len >= 7) ? 0x00 : (0x1F >> (len-2)))
	117
	118	#define UTF_CONTINUATION_MARK 0x80
	119	#define UTF_ACCUMULATION_SHIFT 6
	120	#define UTF_CONTINUATION_MASK ((U8)0x3f)
cb359b41	121	#define UTF8_ACCUMULATE(old, new) (((old) << UTF_ACCUMULATION_SHIFT) \| (((U8)new) & UTF_CONTINUATION_MASK))
421a8bf2	122
1d72bdf6 NIS	123	#define UTF8_EIGHT_BIT_HI(c) ((((U8)(c))>>UTF_ACCUMULATION_SHIFT)\|UTF_START_MARK(2))
1d72bdf6 NIS	124	#define UTF8_EIGHT_BIT_LO(c) (((((U8)(c)))&UTF_CONTINUATION_MASK)\|UTF_CONTINUATION_MARK)
c512ce4f	125
1d68d6cd	126	#ifdef HAS_QUAD
5bbb0b5a	127	#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd SC	128	(uv) < 0x800 ? 2 : \
	129	(uv) < 0x10000 ? 3 : \
	130	(uv) < 0x200000 ? 4 : \
	131	(uv) < 0x4000000 ? 5 : \
	132	(uv) < 0x80000000 ? 6 : \
9041c2e3	133	(uv) < UTF8_QUAD_MAX ? 7 : 13 )
1d68d6cd SC	134	#else
1d68d6cd SC	135	/* No, I'm not even going to TRY putting #ifdef inside a #define */
5bbb0b5a	136	#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd SC	137	(uv) < 0x800 ? 2 : \
	138	(uv) < 0x10000 ? 3 : \
	139	(uv) < 0x200000 ? 4 : \
	140	(uv) < 0x4000000 ? 5 : \
	141	(uv) < 0x80000000 ? 6 : 7 )
	142	#endif
	143
7e2040f0 GS	144	/*
	145	* Note: we try to be careful never to call the isXXX_utf8() functions
	146	* unless we're pretty sure we've seen the beginning of a UTF-8 character
	147	* (that is, the two high bits are set). Otherwise we risk loading in the
	148	* heavy-duty SWASHINIT and SWASHGET routines unnecessarily.
	149	*/
e1ec3a88	150	#define isIDFIRST_lazy_if(p,c) ((IN_BYTES \|\| (!c \|\| (((const U8)p) < 0xc0))) \
7e2040f0	151	? isIDFIRST(*(p)) \
e1ec3a88 AL	152	: isIDFIRST_utf8((const U8*)p))
e1ec3a88 AL	153	#define isALNUM_lazy_if(p,c) ((IN_BYTES \|\| (!c \|\| (((const U8)p) < 0xc0))) \
7e2040f0	154	? isALNUM(*(p)) \
e1ec3a88	155	: isALNUM_utf8((const U8*)p))
1d72bdf6 NIS	156
	157
	158	#endif /* EBCDIC vs ASCII */
	159
	160	/* Rest of these are attributes of Unicode and perl's internals rather than the encoding */
	161
7e2040f0 GS	162	#define isIDFIRST_lazy(p) isIDFIRST_lazy_if(p,1)
7e2040f0 GS	163	#define isALNUM_lazy(p) isALNUM_lazy_if(p,1)
3bd709b1	164
89ebb4a3 JH	165	#define UTF8_MAXBYTES 13
	166	/* How wide can a single UTF-8 encoded character become in bytes.
	167	* NOTE: Strictly speaking Perl's UTF-8 should not be called UTF-8
	168	* since UTF-8 is an encoding of Unicode and given Unicode's current
	169	* upper limit only four bytes is possible. Perl thinks of UTF-8
	170	* as a way to encode non-negative integers in a binary format. */
	171	#define UTF8_MAXLEN UTF8_MAXBYTES
	172
	173	#define UTF8_MAXLEN_UCLC 3 /* Obsolete, do not use. */
	174	#define UTF8_MAXLEN_UCLC_MULT 39 /* Obsolete, do not use. */
	175	#define UTF8_MAXLEN_FOLD 3 /* Obsolete, do not use. */
	176	#define UTF8_MAXLEN_FOLD_MULT 39 /* Obsolete, do not use. */
	177
	178	/* The maximum number of UTF-8 bytes a single Unicode character can
	179	* uppercase/lowercase/fold into; this number depends on the Unicode
	180	* version. An example of maximal expansion is the U+03B0 which
	181	* uppercases to U+03C5 U+0308 U+0301. The Unicode databases that
	182	* tell these things are UnicodeDatabase.txt, CaseFolding.txt, and
	183	* SpecialCasing.txt. */
	184	#define UTF8_MAXBYTES_CASE 6
3bd709b1	185
0064a8a9 JH	186	#define IN_BYTES (PL_curcop->op_private & HINT_BYTES)
0064a8a9 JH	187	#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)
1d72bdf6 NIS	188
	189	#define UTF8_ALLOW_EMPTY 0x0001
	190	#define UTF8_ALLOW_CONTINUATION 0x0002
	191	#define UTF8_ALLOW_NON_CONTINUATION 0x0004
	192	#define UTF8_ALLOW_FE_FF 0x0008
	193	#define UTF8_ALLOW_SHORT 0x0010
	194	#define UTF8_ALLOW_SURROGATE 0x0020
c867b360 JH	195	#define UTF8_ALLOW_FFFF 0x0040 /* Allows also FFFE. */
c867b360 JH	196	#define UTF8_ALLOW_LONG 0x0080
1d72bdf6	197	#define UTF8_ALLOW_ANYUV (UTF8_ALLOW_EMPTY\|UTF8_ALLOW_FE_FF\|\
61843245	198	UTF8_ALLOW_SURROGATE\|UTF8_ALLOW_FFFF)
c867b360	199	#define UTF8_ALLOW_ANY 0x00FF
e83d50c9	200	#define UTF8_CHECK_ONLY 0x0200
1d72bdf6	201
c867b360 JH	202	#define UNICODE_SURROGATE_FIRST 0xD800
	203	#define UNICODE_SURROGATE_LAST 0xDFFF
	204	#define UNICODE_REPLACEMENT 0xFFFD
	205	#define UNICODE_BYTE_ORDER_MARK 0xFEFF
	206	#define UNICODE_ILLEGAL 0xFFFF
1d72bdf6	207
b851fbc1	208	/* Though our UTF-8 encoding can go beyond this,
872c91ae	209	* let's be conservative and do as Unicode 3.2 says. */
b851fbc1 JH	210	#define PERL_UNICODE_MAX 0x10FFFF
	211
	212	#define UNICODE_ALLOW_SURROGATE 0x0001 /* Allow UTF-16 surrogates (EVIL) */
	213	#define UNICODE_ALLOW_FDD0 0x0002 /* Allow the U+FDD0...U+FDEF */
c867b360 JH	214	#define UNICODE_ALLOW_FFFF 0x0004 /* Allow 0xFFF[EF], 0x1FFF[EF], ... */
	215	#define UNICODE_ALLOW_SUPER 0x0008 /* Allow past 10xFFFF */
	216	#define UNICODE_ALLOW_ANY 0x000F
b851fbc1	217
1d72bdf6 NIS	218	#define UNICODE_IS_SURROGATE(c) ((c) >= UNICODE_SURROGATE_FIRST && \
1d72bdf6 NIS	219	(c) <= UNICODE_SURROGATE_LAST)
a10ec373	220	#define UNICODE_IS_REPLACEMENT(c) ((c) == UNICODE_REPLACEMENT)
872c91ae	221	#define UNICODE_IS_BYTE_ORDER_MARK(c) ((c) == UNICODE_BYTE_ORDER_MARK)
1d72bdf6 NIS	222	#define UNICODE_IS_ILLEGAL(c) ((c) == UNICODE_ILLEGAL)
1d72bdf6 NIS	223
f067b878 NA	224	#ifdef HAS_QUAD
	225	# define UTF8_QUAD_MAX UINT64_C(0x1000000000)
	226	#endif
3bd709b1	227
1d72bdf6	228	#define UTF8_IS_ASCII(c) UTF8_IS_INVARIANT(c)
09091399	229
e0f9d4a8	230	#define UNICODE_LATIN_SMALL_LETTER_SHARP_S 0x00DF
09091399 JH	231	#define UNICODE_GREEK_CAPITAL_LETTER_SIGMA 0x03A3
	232	#define UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA 0x03C2
	233	#define UNICODE_GREEK_SMALL_LETTER_SIGMA 0x03C3
	234
ebc501f0 JH	235	#define EBCDIC_LATIN_SMALL_LETTER_SHARP_S 0x0059
ebc501f0 JH	236
9e55ce06	237	#define UNI_DISPLAY_ISPRINT 0x0001
c728cb41 JH	238	#define UNI_DISPLAY_BACKSLASH 0x0002
	239	#define UNI_DISPLAY_QQ (UNI_DISPLAY_ISPRINT\|UNI_DISPLAY_BACKSLASH)
	240	#define UNI_DISPLAY_REGEX (UNI_DISPLAY_ISPRINT\|UNI_DISPLAY_BACKSLASH)
9e55ce06	241
ebc501f0 JH	242	#ifdef EBCDIC
	243	# define ANYOF_FOLD_SHARP_S(node, input, end) \
	244	(ANYOF_BITMAP_TEST(node, EBCDIC_LATIN_SMALL_LETTER_SHARP_S) && \
	245	(ANYOF_FLAGS(node) & ANYOF_UNICODE) && \
	246	(ANYOF_FLAGS(node) & ANYOF_FOLD) && \
	247	((end) > (input) + 1) && \
	248	toLOWER((input)[0]) == 's' && \
	249	toLOWER((input)[1]) == 's')
	250	#else
	251	# define ANYOF_FOLD_SHARP_S(node, input, end) \
07b6858f JH	252	(ANYOF_BITMAP_TEST(node, UNICODE_LATIN_SMALL_LETTER_SHARP_S) && \
	253	(ANYOF_FLAGS(node) & ANYOF_UNICODE) && \
	254	(ANYOF_FLAGS(node) & ANYOF_FOLD) && \
	255	((end) > (input) + 1) && \
	256	toLOWER((input)[0]) == 's' && \
	257	toLOWER((input)[1]) == 's')
ebc501f0 JH	258	#endif
ebc501f0 JH	259	#define SHARP_S_SKIP 2
3b0fc154	260
77263263 TS	261	#ifdef EBCDIC
	262	/* IS_UTF8_CHAR() is not ported to EBCDIC */
	263	#else
3b0fc154 JH	264	#define IS_UTF8_CHAR_1(p) \
	265	((p)[0] <= 0x7F)
	266	#define IS_UTF8_CHAR_2(p) \
	267	((p)[0] >= 0xC2 && (p)[0] <= 0xDF && \
	268	(p)[1] >= 0x80 && (p)[1] <= 0xBF)
	269	#define IS_UTF8_CHAR_3a(p) \
	270	((p)[0] == 0xE0 && \
	271	(p)[1] >= 0xA0 && (p)[1] <= 0xBF && \
	272	(p)[2] >= 0x80 && (p)[2] <= 0xBF)
	273	#define IS_UTF8_CHAR_3b(p) \
	274	((p)[0] >= 0xE1 && (p)[0] <= 0xEC && \
	275	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	276	(p)[2] >= 0x80 && (p)[2] <= 0xBF)
	277	#define IS_UTF8_CHAR_3c(p) \
	278	((p)[0] == 0xED && \
	279	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	280	(p)[2] >= 0x80 && (p)[2] <= 0xBF)
	281	/* In IS_UTF8_CHAR_3c(p) one could use
	282	* (p)[1] >= 0x80 && (p)[1] <= 0x9F
	283	* if one wanted to exclude surrogates. */
	284	#define IS_UTF8_CHAR_3d(p) \
	285	((p)[0] >= 0xEE && (p)[0] <= 0xEF && \
	286	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	287	(p)[2] >= 0x80 && (p)[2] <= 0xBF)
	288	#define IS_UTF8_CHAR_4a(p) \
	289	((p)[0] == 0xF0 && \
	290	(p)[1] >= 0x90 && (p)[1] <= 0xBF && \
	291	(p)[2] >= 0x80 && (p)[2] <= 0xBF && \
	292	(p)[3] >= 0x80 && (p)[3] <= 0xBF)
	293	#define IS_UTF8_CHAR_4b(p) \
	294	((p)[0] >= 0xF1 && (p)[0] <= 0xF3 && \
	295	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	296	(p)[2] >= 0x80 && (p)[2] <= 0xBF && \
	297	(p)[3] >= 0x80 && (p)[3] <= 0xBF)
	298	/* In IS_UTF8_CHAR_4c(p) one could use
	299	* (p)[0] == 0xF4
	300	* if one wanted to stop at the Unicode limit U+10FFFF.
	301	* The 0xF7 allows us to go to 0x1fffff (0x200000 would
	302	* require five bytes). Not doing any further code points
	303	* since that is not needed (and that would not be strict
	304	* UTF-8, anyway). The "slow path" in Perl_is_utf8_char()
	305	* will take care of the "extended UTF-8". */
	306	#define IS_UTF8_CHAR_4c(p) \
	307	((p)[0] == 0xF4 && (p)[0] <= 0xF7 && \
	308	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	309	(p)[2] >= 0x80 && (p)[2] <= 0xBF && \
	310	(p)[3] >= 0x80 && (p)[3] <= 0xBF)
	311
	312	#define IS_UTF8_CHAR_3(p) \
	313	(IS_UTF8_CHAR_3a(p) \|\| \
	314	IS_UTF8_CHAR_3b(p) \|\| \
	315	IS_UTF8_CHAR_3c(p) \|\| \
	316	IS_UTF8_CHAR_3d(p))
	317	#define IS_UTF8_CHAR_4(p) \
	318	(IS_UTF8_CHAR_4a(p) \|\| \
	319	IS_UTF8_CHAR_4b(p) \|\| \
	320	IS_UTF8_CHAR_4c(p))
	321
	322	/* IS_UTF8_CHAR(p) is strictly speaking wrong (not UTF-8) because it
	323	* (1) allows UTF-8 encoded UTF-16 surrogates
	324	* (2) it allows code points past U+10FFFF.
	325	* The Perl_is_utf8_char() full "slow" code will handle the Perl
	326	* "extended UTF-8". */
	327	#define IS_UTF8_CHAR(p, n) \
328	((n) == 1 ? IS_UTF8_CHAR_1(p) : \
329	(n) == 2 ? IS_UTF8_CHAR_2(p) : \
330	(n) == 3 ? IS_UTF8_CHAR_3(p) : \
331	(n) == 4 ? IS_UTF8_CHAR_4(p) : 0)
332
768c67ee JH	333	#define IS_UTF8_CHAR_FAST(n) ((n) <= 4)
768c67ee JH	334
77263263	335	#endif /* IS_UTF8_CHAR() for UTF-8 */