[p5sagit/p5-mst-13.2.git] / utf8.h

/*    utf8.h
 *
 *    Copyright (C) 2000, 2001, 2002, 2005, 2006, 2007, 2009 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

/* Source backward compatibility. */
#define uvuni_to_utf8(d, uv)		uvuni_to_utf8_flags(d, uv, 0)
#define is_utf8_string_loc(s, len, ep)	is_utf8_string_loclen(s, len, ep, 0)

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

#include "utfebcdic.h"

#else	/* ! EBCDIC */
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

/* 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 utf8n_to_uvchr utf8n_to_uvuni
#define uvchr_to_utf8  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)
/* Note that C0 and C1 are invalid in legal UTF8, so the lower bound of the
 * below might ought to be C2 */
#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)

#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

#endif /* EBCDIC vs ASCII */

/* Rest of these are attributes of Unicode and perl's internals rather than the
 * encoding, or happen to be the same in both ASCII and EBCDIC (at least at
 * this level; the macros that some of these call may have different
 * definitions in the two encodings */

#define NATIVE8_TO_UNI(ch)     NATIVE_TO_ASCII(ch)	/* a clearer synonym */

#define UTF8_ACCUMULATE(old, new)	(((old) << UTF_ACCUMULATION_SHIFT) | (((U8)new) & UTF_CONTINUATION_MASK))

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

#define UTF8_IS_INVARIANT(c)		UNI_IS_INVARIANT(NATIVE_TO_UTF(c))
#define NATIVE_IS_INVARIANT(c)		UNI_IS_INVARIANT(NATIVE8_TO_UNI(c))

#define MAX_PORTABLE_UTF8_TWO_BYTE 0x3FF    /* constrained by EBCDIC */

/* The macros in the next sets are used to generate the two utf8 or utfebcdic
 * bytes from an ordinal that is known to fit into two bytes; it must be less
 * than 0x3FF to work across both encodings. */
/* Nocast allows these to be used in the case label of a switch statement */
#define UTF8_TWO_BYTE_HI_nocast(c)	UTF_TO_NATIVE(((c)>>UTF_ACCUMULATION_SHIFT)|UTF_START_MARK(2))
#define UTF8_TWO_BYTE_LO_nocast(c)	UTF_TO_NATIVE(((c)&UTF_CONTINUATION_MASK)|UTF_CONTINUATION_MARK)

#define UTF8_TWO_BYTE_HI(c)	((U8) (UTF8_TWO_BYTE_HI_nocast(c)))
#define UTF8_TWO_BYTE_LO(c)	((U8) (UTF8_TWO_BYTE_LO_nocast(c)))

/* This name is used when the source is a single byte */
#define UTF8_EIGHT_BIT_HI(c)	UTF8_TWO_BYTE_HI((U8)(c))
#define UTF8_EIGHT_BIT_LO(c)	UTF8_TWO_BYTE_LO((U8)(c))

/*
 * 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 or UTFEBCDIC
 * character.  Otherwise we risk loading in the heavy-duty swash_init and
 * swash_fetch routines unnecessarily.
 */
#define isIDFIRST_lazy_if(p,c) ((IN_BYTES || (!c || ! UTF8_IS_START(*((const U8*)p)))) \
				? isIDFIRST(*(p)) \
				: isIDFIRST_utf8((const U8*)p))
#define isALNUM_lazy_if(p,c)   ((IN_BYTES || (!c || ! UTF8_IS_START(*((const U8*)p)))) \
				? isALNUM(*(p)) \
				: isALNUM_utf8((const U8*)p))

#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

/* 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 UnicodeData.txt, CaseFolding.txt, and
 * SpecialCasing.txt. */
#define UTF8_MAXBYTES_CASE	6

#define IN_BYTES (CopHINTS_get(PL_curcop) & HINT_BYTES)
#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)
#define IN_UNI_8_BIT (CopHINTS_get(PL_curcop) & HINT_UNI_8_BIT && ! IN_LOCALE_RUNTIME && ! IN_BYTES)

#define UTF8_ALLOW_EMPTY		0x0001
#define UTF8_ALLOW_CONTINUATION		0x0002
#define UTF8_ALLOW_NON_CONTINUATION	0x0004
#define UTF8_ALLOW_FE_FF		0x0008 /* Allow above 0x7fffFFFF */
#define UTF8_ALLOW_SHORT		0x0010
#define UTF8_ALLOW_SURROGATE		0x0020
#define UTF8_ALLOW_FFFF			0x0040 /* Allow UNICODE_ILLEGAL */
#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 UTF8_ALLOW_DEFAULT		(ckWARN(WARN_UTF8) ? 0 : \
					 UTF8_ALLOW_ANYUV)

#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 5.1 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 U+FFF[EF], U+1FFF[EF], ... */
#define UNICODE_ALLOW_SUPER	0x0008	/* Allow past 0x10FFFF */
#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_GREEK_CAPITAL_LETTER_SIGMA	0x03A3
#define UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA	0x03C2
#define UNICODE_GREEK_SMALL_LETTER_SIGMA	0x03C3

#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)

#ifndef EBCDIC
#   define LATIN_SMALL_LETTER_SHARP_S	0x00DF
#   define LATIN_SMALL_LETTER_Y_WITH_DIAERESIS 0x00FF
#   define MICRO_SIGN 0x00B5
#endif

#define ANYOF_FOLD_SHARP_S(node, input, end)	\
	(ANYOF_BITMAP_TEST(node, 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')
#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 */

/*
 * Local variables:
 * c-indentation-style: bsd
 * c-basic-offset: 4
 * indent-tabs-mode: t
 * End:
 *
 * ex: set ts=8 sts=4 sw=4 noet:
 */
Commit	Line	Data
a0ed51b3	1	/* utf8.h
a0ed51b3	2	*
83706693	3	* Copyright (C) 2000, 2001, 2002, 2005, 2006, 2007, 2009 by Larry Wall and others
a0ed51b3	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	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
fd7cb289	19	/* Source backward compatibility. */
	20	#define uvuni_to_utf8(d, uv) uvuni_to_utf8_flags(d, uv, 0)
	21	#define is_utf8_string_loc(s, len, ep) is_utf8_string_loclen(s, len, ep, 0)
	22
1d72bdf6	23	#ifdef EBCDIC
	24	/* The equivalent of these macros but implementing UTF-EBCDIC
	25	are in the following header file:
	26	*/
	27
	28	#include "utfebcdic.h"
fd7cb289	29
d06134e5	30	#else /* ! EBCDIC */
73c4f7a1	31	START_EXTERN_C
73c4f7a1	32
a0ed51b3	33	#ifdef DOINIT
6f06b55f	34	EXTCONST unsigned char PL_utf8skip[] = {
a0ed51b3	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, /* ascii */
	36	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 */
	37	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 */
	38	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 */
	39	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 */
	40	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 */
	41	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	42	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	43	7,13, /* Perl extended (not UTF-8). Up to 72bit allowed (64-bit + reserved). */
a0ed51b3	44	};
a0ed51b3	45	#else
6f06b55f	46	EXTCONST unsigned char PL_utf8skip[];
a0ed51b3	47	#endif
a0ed51b3	48
73c4f7a1	49	END_EXTERN_C
7e2040f0	50
1d72bdf6	51	/* Native character to iso-8859-1 */
	52	#define NATIVE_TO_ASCII(ch) (ch)
	53	#define ASCII_TO_NATIVE(ch) (ch)
	54	/* Transform after encoding */
	55	#define NATIVE_TO_UTF(ch) (ch)
	56	#define UTF_TO_NATIVE(ch) (ch)
	57	/* Transforms in wide UV chars */
	58	#define UNI_TO_NATIVE(ch) (ch)
	59	#define NATIVE_TO_UNI(ch) (ch)
	60	/* Transforms in invariant space */
	61	#define NATIVE_TO_NEED(enc,ch) (ch)
	62	#define ASCII_TO_NEED(enc,ch) (ch)
d7578b48	63
d06134e5	64	/* As there are no translations, avoid the function wrapper */
1754c1a1	65	#define utf8n_to_uvchr utf8n_to_uvuni
1754c1a1	66	#define uvchr_to_utf8 uvuni_to_utf8
2b9d42f0	67
877d9f0d	68	/*
9041c2e3	69
8c007b5a	70	The following table is from Unicode 3.2.
877d9f0d	71
	72	Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
	73
375122d7	74	U+0000..U+007F 00..7F
	75	U+0080..U+07FF C2..DF 80..BF
	76	U+0800..U+0FFF E0 A0..BF 80..BF
	77	U+1000..U+CFFF E1..EC 80..BF 80..BF
	78	U+D000..U+D7FF ED 80..9F 80..BF
8c007b5a	79	U+D800..U+DFFF ***** ill-formed *****
375122d7	80	U+E000..U+FFFF EE..EF 80..BF 80..BF
877d9f0d	81	U+10000..U+3FFFF F0 90..BF 80..BF 80..BF
	82	U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
	83	U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
	84
8c007b5a	85	Note the A0..BF in U+0800..U+0FFF, the 80..9F in U+D000...U+D7FF,
8c007b5a	86	the 90..BF in U+10000..U+3FFFF, and the 80...8F in U+100000..U+10FFFF.
37361303	87	The "gaps" are caused by legal UTF-8 avoiding non-shortest encodings:
	88	it is technically possible to UTF-8-encode a single code point in different
	89	ways, but that is explicitly forbidden, and the shortest possible encoding
	90	should always be used (and that is what Perl does).
8c007b5a	91
877d9f0d	92	*/
877d9f0d	93
8c007b5a	94	/*
	95	Another way to look at it, as bits:
	96
	97	Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
	98
	99	0aaaaaaa 0aaaaaaa
	100	00000bbbbbaaaaaa 110bbbbb 10aaaaaa
	101	ccccbbbbbbaaaaaa 1110cccc 10bbbbbb 10aaaaaa
	102	00000dddccccccbbbbbbaaaaaa 11110ddd 10cccccc 10bbbbbb 10aaaaaa
	103
	104	As you can see, the continuation bytes all begin with C<10>, and the
	105	leading bits of the start byte tell how many bytes the are in the
	106	encoded character.
	107
	108	*/
	109
	110
c4d5f83a	111	#define UNI_IS_INVARIANT(c) (((UV)c) < 0x80)
d06134e5	112	/* Note that C0 and C1 are invalid in legal UTF8, so the lower bound of the
d06134e5	113	* below might ought to be C2 */
c512ce4f	114	#define UTF8_IS_START(c) (((U8)c) >= 0xc0 && (((U8)c) <= 0xfd))
	115	#define UTF8_IS_CONTINUATION(c) (((U8)c) >= 0x80 && (((U8)c) <= 0xbf))
	116	#define UTF8_IS_CONTINUED(c) (((U8)c) & 0x80)
db42d148	117	#define UTF8_IS_DOWNGRADEABLE_START(c) (((U8)c & 0xfc) == 0xc0)
8850bf83	118
22901f30	119	#define UTF_START_MARK(len) (((len) > 7) ? 0xFF : (0xFE << (7-(len))))
22901f30	120	#define UTF_START_MASK(len) (((len) >= 7) ? 0x00 : (0x1F >> ((len)-2)))
1d72bdf6	121
	122	#define UTF_CONTINUATION_MARK 0x80
	123	#define UTF_ACCUMULATION_SHIFT 6
	124	#define UTF_CONTINUATION_MASK ((U8)0x3f)
c512ce4f	125
1d68d6cd	126	#ifdef HAS_QUAD
5bbb0b5a	127	#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd	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	134	#else
1d68d6cd	135	/* No, I'm not even going to TRY putting #ifdef inside a #define */
5bbb0b5a	136	#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \
1d68d6cd	137	(uv) < 0x800 ? 2 : \
	138	(uv) < 0x10000 ? 3 : \
	139	(uv) < 0x200000 ? 4 : \
	140	(uv) < 0x4000000 ? 5 : \
	141	(uv) < 0x80000000 ? 6 : 7 )
	142	#endif
	143
d06134e5	144	#endif /* EBCDIC vs ASCII */
	145
	146	/* Rest of these are attributes of Unicode and perl's internals rather than the
	147	* encoding, or happen to be the same in both ASCII and EBCDIC (at least at
	148	* this level; the macros that some of these call may have different
	149	* definitions in the two encodings */
	150
	151	#define NATIVE8_TO_UNI(ch) NATIVE_TO_ASCII(ch) /* a clearer synonym */
	152
	153	#define UTF8_ACCUMULATE(old, new) (((old) << UTF_ACCUMULATION_SHIFT) \| (((U8)new) & UTF_CONTINUATION_MASK))
	154
	155	#define UTF8SKIP(s) PL_utf8skip[(const U8)(s)]
	156
	157	#define UTF8_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_TO_UTF(c))
	158	#define NATIVE_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE8_TO_UNI(c))
	159
	160	#define MAX_PORTABLE_UTF8_TWO_BYTE 0x3FF /* constrained by EBCDIC */
	161
	162	/* The macros in the next sets are used to generate the two utf8 or utfebcdic
	163	* bytes from an ordinal that is known to fit into two bytes; it must be less
	164	* than 0x3FF to work across both encodings. */
	165	/* Nocast allows these to be used in the case label of a switch statement */
	166	#define UTF8_TWO_BYTE_HI_nocast(c) UTF_TO_NATIVE(((c)>>UTF_ACCUMULATION_SHIFT)\|UTF_START_MARK(2))
	167	#define UTF8_TWO_BYTE_LO_nocast(c) UTF_TO_NATIVE(((c)&UTF_CONTINUATION_MASK)\|UTF_CONTINUATION_MARK)
	168
	169	#define UTF8_TWO_BYTE_HI(c) ((U8) (UTF8_TWO_BYTE_HI_nocast(c)))
	170	#define UTF8_TWO_BYTE_LO(c) ((U8) (UTF8_TWO_BYTE_LO_nocast(c)))
	171
	172	/* This name is used when the source is a single byte */
	173	#define UTF8_EIGHT_BIT_HI(c) UTF8_TWO_BYTE_HI((U8)(c))
	174	#define UTF8_EIGHT_BIT_LO(c) UTF8_TWO_BYTE_LO((U8)(c))
	175
7e2040f0	176	/*
7e2040f0	177	* Note: we try to be careful never to call the isXXX_utf8() functions
d06134e5	178	* unless we're pretty sure we've seen the beginning of a UTF-8 or UTFEBCDIC
	179	* character. Otherwise we risk loading in the heavy-duty swash_init and
	180	* swash_fetch routines unnecessarily.
7e2040f0	181	*/
d06134e5	182	#define isIDFIRST_lazy_if(p,c) ((IN_BYTES \|\| (!c \|\| ! UTF8_IS_START(((const U8)p)))) \
7e2040f0	183	? isIDFIRST(*(p)) \
e1ec3a88	184	: isIDFIRST_utf8((const U8*)p))
d06134e5	185	#define isALNUM_lazy_if(p,c) ((IN_BYTES \|\| (!c \|\| ! UTF8_IS_START(((const U8)p)))) \
7e2040f0	186	? isALNUM(*(p)) \
e1ec3a88	187	: isALNUM_utf8((const U8*)p))
1d72bdf6	188
7e2040f0	189	#define isIDFIRST_lazy(p) isIDFIRST_lazy_if(p,1)
7e2040f0	190	#define isALNUM_lazy(p) isALNUM_lazy_if(p,1)
3bd709b1	191
89ebb4a3	192	#define UTF8_MAXBYTES 13
	193	/* How wide can a single UTF-8 encoded character become in bytes.
	194	* NOTE: Strictly speaking Perl's UTF-8 should not be called UTF-8
	195	* since UTF-8 is an encoding of Unicode and given Unicode's current
	196	* upper limit only four bytes is possible. Perl thinks of UTF-8
	197	* as a way to encode non-negative integers in a binary format. */
	198	#define UTF8_MAXLEN UTF8_MAXBYTES
	199
89ebb4a3	200	/* The maximum number of UTF-8 bytes a single Unicode character can
	201	* uppercase/lowercase/fold into; this number depends on the Unicode
	202	* version. An example of maximal expansion is the U+03B0 which
	203	* uppercases to U+03C5 U+0308 U+0301. The Unicode databases that
d06134e5	204	* tell these things are UnicodeData.txt, CaseFolding.txt, and
89ebb4a3	205	* SpecialCasing.txt. */
89ebb4a3	206	#define UTF8_MAXBYTES_CASE 6
3bd709b1	207
a98fe34d	208	#define IN_BYTES (CopHINTS_get(PL_curcop) & HINT_BYTES)
0064a8a9	209	#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)
00f254e2	210	#define IN_UNI_8_BIT (CopHINTS_get(PL_curcop) & HINT_UNI_8_BIT && ! IN_LOCALE_RUNTIME && ! IN_BYTES)
1d72bdf6	211
	212	#define UTF8_ALLOW_EMPTY 0x0001
	213	#define UTF8_ALLOW_CONTINUATION 0x0002
	214	#define UTF8_ALLOW_NON_CONTINUATION 0x0004
9f7f3913	215	#define UTF8_ALLOW_FE_FF 0x0008 /* Allow above 0x7fffFFFF */
1d72bdf6	216	#define UTF8_ALLOW_SHORT 0x0010
1d72bdf6	217	#define UTF8_ALLOW_SURROGATE 0x0020
9f7f3913	218	#define UTF8_ALLOW_FFFF 0x0040 /* Allow UNICODE_ILLEGAL */
c867b360	219	#define UTF8_ALLOW_LONG 0x0080
1d72bdf6	220	#define UTF8_ALLOW_ANYUV (UTF8_ALLOW_EMPTY\|UTF8_ALLOW_FE_FF\|\
61843245	221	UTF8_ALLOW_SURROGATE\|UTF8_ALLOW_FFFF)
c867b360	222	#define UTF8_ALLOW_ANY 0x00FF
e83d50c9	223	#define UTF8_CHECK_ONLY 0x0200
9f7f3913	224	#define UTF8_ALLOW_DEFAULT (ckWARN(WARN_UTF8) ? 0 : \
9f7f3913	225	UTF8_ALLOW_ANYUV)
1d72bdf6	226
c867b360	227	#define UNICODE_SURROGATE_FIRST 0xD800
	228	#define UNICODE_SURROGATE_LAST 0xDFFF
	229	#define UNICODE_REPLACEMENT 0xFFFD
	230	#define UNICODE_BYTE_ORDER_MARK 0xFEFF
	231	#define UNICODE_ILLEGAL 0xFFFF
1d72bdf6	232
b851fbc1	233	/* Though our UTF-8 encoding can go beyond this,
b3ab6785	234	* let's be conservative and do as Unicode 5.1 says. */
b851fbc1	235	#define PERL_UNICODE_MAX 0x10FFFF
	236
	237	#define UNICODE_ALLOW_SURROGATE 0x0001 /* Allow UTF-16 surrogates (EVIL) */
	238	#define UNICODE_ALLOW_FDD0 0x0002 /* Allow the U+FDD0...U+FDEF */
9f7f3913	239	#define UNICODE_ALLOW_FFFF 0x0004 /* Allow U+FFF[EF], U+1FFF[EF], ... */
9f7f3913	240	#define UNICODE_ALLOW_SUPER 0x0008 /* Allow past 0x10FFFF */
c867b360	241	#define UNICODE_ALLOW_ANY 0x000F
b851fbc1	242
1d72bdf6	243	#define UNICODE_IS_SURROGATE(c) ((c) >= UNICODE_SURROGATE_FIRST && \
1d72bdf6	244	(c) <= UNICODE_SURROGATE_LAST)
a10ec373	245	#define UNICODE_IS_REPLACEMENT(c) ((c) == UNICODE_REPLACEMENT)
872c91ae	246	#define UNICODE_IS_BYTE_ORDER_MARK(c) ((c) == UNICODE_BYTE_ORDER_MARK)
1d72bdf6	247	#define UNICODE_IS_ILLEGAL(c) ((c) == UNICODE_ILLEGAL)
1d72bdf6	248
f067b878	249	#ifdef HAS_QUAD
	250	# define UTF8_QUAD_MAX UINT64_C(0x1000000000)
	251	#endif
3bd709b1	252
1d72bdf6	253	#define UTF8_IS_ASCII(c) UTF8_IS_INVARIANT(c)
09091399	254
	255	#define UNICODE_GREEK_CAPITAL_LETTER_SIGMA 0x03A3
	256	#define UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA 0x03C2
	257	#define UNICODE_GREEK_SMALL_LETTER_SIGMA 0x03C3
	258
9e55ce06	259	#define UNI_DISPLAY_ISPRINT 0x0001
c728cb41	260	#define UNI_DISPLAY_BACKSLASH 0x0002
	261	#define UNI_DISPLAY_QQ (UNI_DISPLAY_ISPRINT\|UNI_DISPLAY_BACKSLASH)
	262	#define UNI_DISPLAY_REGEX (UNI_DISPLAY_ISPRINT\|UNI_DISPLAY_BACKSLASH)
9e55ce06	263
5cd46e1f	264	#ifndef EBCDIC
	265	# define LATIN_SMALL_LETTER_SHARP_S 0x00DF
	266	# define LATIN_SMALL_LETTER_Y_WITH_DIAERESIS 0x00FF
	267	# define MICRO_SIGN 0x00B5
	268	#endif
	269
	270	#define ANYOF_FOLD_SHARP_S(node, input, end) \
	271	(ANYOF_BITMAP_TEST(node, LATIN_SMALL_LETTER_SHARP_S) && \
07b6858f	272	(ANYOF_FLAGS(node) & ANYOF_UNICODE) && \
	273	(ANYOF_FLAGS(node) & ANYOF_FOLD) && \
	274	((end) > (input) + 1) && \
	275	toLOWER((input)[0]) == 's' && \
	276	toLOWER((input)[1]) == 's')
ebc501f0	277	#define SHARP_S_SKIP 2
3b0fc154	278
77263263	279	#ifdef EBCDIC
	280	/* IS_UTF8_CHAR() is not ported to EBCDIC */
	281	#else
3b0fc154	282	#define IS_UTF8_CHAR_1(p) \
	283	((p)[0] <= 0x7F)
	284	#define IS_UTF8_CHAR_2(p) \
	285	((p)[0] >= 0xC2 && (p)[0] <= 0xDF && \
	286	(p)[1] >= 0x80 && (p)[1] <= 0xBF)
	287	#define IS_UTF8_CHAR_3a(p) \
	288	((p)[0] == 0xE0 && \
	289	(p)[1] >= 0xA0 && (p)[1] <= 0xBF && \
	290	(p)[2] >= 0x80 && (p)[2] <= 0xBF)
	291	#define IS_UTF8_CHAR_3b(p) \
	292	((p)[0] >= 0xE1 && (p)[0] <= 0xEC && \
	293	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	294	(p)[2] >= 0x80 && (p)[2] <= 0xBF)
	295	#define IS_UTF8_CHAR_3c(p) \
	296	((p)[0] == 0xED && \
	297	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	298	(p)[2] >= 0x80 && (p)[2] <= 0xBF)
	299	/* In IS_UTF8_CHAR_3c(p) one could use
	300	* (p)[1] >= 0x80 && (p)[1] <= 0x9F
	301	* if one wanted to exclude surrogates. */
	302	#define IS_UTF8_CHAR_3d(p) \
	303	((p)[0] >= 0xEE && (p)[0] <= 0xEF && \
	304	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	305	(p)[2] >= 0x80 && (p)[2] <= 0xBF)
	306	#define IS_UTF8_CHAR_4a(p) \
	307	((p)[0] == 0xF0 && \
	308	(p)[1] >= 0x90 && (p)[1] <= 0xBF && \
	309	(p)[2] >= 0x80 && (p)[2] <= 0xBF && \
	310	(p)[3] >= 0x80 && (p)[3] <= 0xBF)
	311	#define IS_UTF8_CHAR_4b(p) \
	312	((p)[0] >= 0xF1 && (p)[0] <= 0xF3 && \
	313	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	314	(p)[2] >= 0x80 && (p)[2] <= 0xBF && \
	315	(p)[3] >= 0x80 && (p)[3] <= 0xBF)
	316	/* In IS_UTF8_CHAR_4c(p) one could use
	317	* (p)[0] == 0xF4
	318	* if one wanted to stop at the Unicode limit U+10FFFF.
	319	* The 0xF7 allows us to go to 0x1fffff (0x200000 would
	320	* require five bytes). Not doing any further code points
	321	* since that is not needed (and that would not be strict
	322	* UTF-8, anyway). The "slow path" in Perl_is_utf8_char()
	323	* will take care of the "extended UTF-8". */
	324	#define IS_UTF8_CHAR_4c(p) \
	325	((p)[0] == 0xF4 && (p)[0] <= 0xF7 && \
	326	(p)[1] >= 0x80 && (p)[1] <= 0xBF && \
	327	(p)[2] >= 0x80 && (p)[2] <= 0xBF && \
	328	(p)[3] >= 0x80 && (p)[3] <= 0xBF)
	329
	330	#define IS_UTF8_CHAR_3(p) \
	331	(IS_UTF8_CHAR_3a(p) \|\| \
	332	IS_UTF8_CHAR_3b(p) \|\| \
	333	IS_UTF8_CHAR_3c(p) \|\| \
	334	IS_UTF8_CHAR_3d(p))
	335	#define IS_UTF8_CHAR_4(p) \
	336	(IS_UTF8_CHAR_4a(p) \|\| \
	337	IS_UTF8_CHAR_4b(p) \|\| \
	338	IS_UTF8_CHAR_4c(p))
	339
	340	/* IS_UTF8_CHAR(p) is strictly speaking wrong (not UTF-8) because it
	341	* (1) allows UTF-8 encoded UTF-16 surrogates
	342	* (2) it allows code points past U+10FFFF.
	343	* The Perl_is_utf8_char() full "slow" code will handle the Perl
	344	* "extended UTF-8". */
	345	#define IS_UTF8_CHAR(p, n) \
346	((n) == 1 ? IS_UTF8_CHAR_1(p) : \
347	(n) == 2 ? IS_UTF8_CHAR_2(p) : \
348	(n) == 3 ? IS_UTF8_CHAR_3(p) : \
349	(n) == 4 ? IS_UTF8_CHAR_4(p) : 0)
350
768c67ee	351	#define IS_UTF8_CHAR_FAST(n) ((n) <= 4)
768c67ee	352
77263263	353	#endif /* IS_UTF8_CHAR() for UTF-8 */
e9a8c099	354
	355	/*
	356	* Local variables:
	357	* c-indentation-style: bsd
	358	* c-basic-offset: 4
	359	* indent-tabs-mode: t
	360	* End:
	361	*
	362	* ex: set ts=8 sts=4 sw=4 noet:
	363	*/