X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/559c7f10ed871fab06296b9297b9273240599eee..7ff1117359e03ce00638e9ee1daad537321e75d6:/utf8.h diff --git a/utf8.h b/utf8.h index 5aac904..8504207 100644 --- a/utf8.h +++ b/utf8.h @@ -1,5 +1,11 @@ /* utf8.h * + * This file contains definitions for use with the UTF-8 encoding. It + * actually also works with the variant UTF-8 encoding called UTF-EBCDIC, and + * hides almost all of the differences between these from the caller. In other + * words, someone should #include this file, and if the code is being compiled + * on an EBCDIC platform, things should mostly just work. + * * Copyright (C) 2000, 2001, 2002, 2005, 2006, 2007, 2009, * 2010, 2011 by Larry Wall and others * @@ -20,6 +26,9 @@ # define USE_UTF8_IN_NAMES (PL_hints & HINT_UTF8) #endif +#include "regcharclass.h" +#include "unicode_constants.h" + /* For to_utf8_fold_flags, q.v. */ #define FOLD_FLAGS_LOCALE 0x1 #define FOLD_FLAGS_FULL 0x2 @@ -30,6 +39,13 @@ #define _CORE_SWASH_INIT_RETURN_IF_UNDEF 0x2 #define _CORE_SWASH_INIT_ACCEPT_INVLIST 0x4 +#define uvchr_to_utf8(a,b) uvchr_to_utf8_flags(a,b,0) +#define uvchr_to_utf8_flags(d,uv,flags) \ + uvoffuni_to_utf8_flags(d,NATIVE_TO_UNI(uv),flags) +#define utf8_to_uvchr_buf(s, e, lenp) \ + utf8n_to_uvchr(s, (U8*)(e) - (U8*)(s), lenp, \ + ckWARN_d(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY) + #define to_uni_fold(c, p, lenp) _to_uni_fold_flags(c, p, lenp, FOLD_FLAGS_FULL) #define to_utf8_fold(c, p, lenp) _to_utf8_fold_flags(c, p, lenp, \ FOLD_FLAGS_FULL, NULL) @@ -38,7 +54,6 @@ #define to_utf8_title(a,b,c) _to_utf8_title_flags(a,b,c,0, NULL) /* 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) #define foldEQ_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2) \ @@ -96,28 +111,23 @@ EXTCONST unsigned char PL_utf8skip[]; END_EXTERN_C -#include "regcharclass.h" -#include "unicode_constants.h" +/* Native character to/from iso-8859-1. Are the identity functions on ASCII + * platforms */ +#define NATIVE_TO_LATIN1(ch) (ch) +#define LATIN1_TO_NATIVE(ch) (ch) + +/* I8 is an intermediate version of UTF-8 used only in UTF-EBCDIC. We thus + * consider it to be identical to UTF-8 on ASCII platforms. Strictly speaking + * UTF-8 and UTF-EBCDIC are two different things, but we often conflate them + * because they are 8-bit encodings that serve the same purpose in Perl, and + * rarely do we need to distinguish them. The term "NATIVE_UTF8" applies to + * whichever one is applicable on the current platform */ +#define NATIVE_UTF8_TO_I8(ch) (ch) +#define I8_TO_NATIVE_UTF8(ch) (ch) -/* 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 NATIVE_TO_I8(ch) NATIVE_TO_UTF(ch) /* a clearer synonym */ -#define UTF_TO_NATIVE(ch) (ch) -#define I8_TO_NATIVE(ch) UTF_TO_NATIVE(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 valid_utf8_to_uvchr valid_utf8_to_uvuni -#define uvchr_to_utf8 uvuni_to_utf8 /* @@ -130,7 +140,7 @@ END_EXTERN_C 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 +++++++ utf16 surrogates, not legal utf8 +++++++ + U+D800..U+DFFF ED A0..BF 80..BF (surrogates) 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 @@ -138,13 +148,13 @@ END_EXTERN_C Below are non-Unicode code points U+110000..U+13FFFF F4 90..BF 80..BF 80..BF U+110000..U+1FFFFF F5..F7 80..BF 80..BF 80..BF - U+200000: F8.. * 88..BF 80..BF 80..BF 80..BF + U+200000..: F8.. * 88..BF 80..BF 80..BF 80..BF Note the gaps before several of the byte entries above marked by '*'. These 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). +(and that is what Perl does). The non-shortest ones are called 'overlongs'. */ @@ -166,39 +176,68 @@ Perl's extended UTF-8 means we can have start bytes up to FF. */ +/* Is the representation of the Unicode code point 'c' the same regardless of + * being encoded in UTF-8 or not? */ #define UNI_IS_INVARIANT(c) (((UV)c) < 0x80) + +/* Is the UTF8-encoded byte 'c' part of a variant sequence in UTF-8? This is + * the inverse of UTF8_IS_INVARIANT */ +#define UTF8_IS_CONTINUED(c) (((U8)c) & 0x80) + +/* Is the byte 'c' the first byte of a multi-byte UTF8-8 encoded sequence? + * This doesn't catch invariants (they are single-byte). It also excludes the + * illegal overlong sequences that begin with C0 and C1. */ #define UTF8_IS_START(c) (((U8)c) >= 0xc2) + +/* Is the byte 'c' part of a multi-byte UTF8-8 encoded sequence, and not the + * first byte thereof? */ #define UTF8_IS_CONTINUATION(c) ((((U8)c) & 0xC0) == 0x80) -#define UTF8_IS_CONTINUED(c) (((U8)c) & 0x80) -/* Use UTF8_IS_NEXT_CHAR_DOWNGRADEABLE() instead if the input isn't known to - * be well-formed. Masking with 0xfe allows low bit to be 0 or 1; thus this - * matches 0xc[23]. */ +/* Is the UTF8-encoded byte 'c' the first byte of a two byte sequence? Use + * UTF8_IS_NEXT_CHAR_DOWNGRADEABLE() instead if the input isn't known to + * be well-formed. Masking with 0xfe allows the low bit to be 0 or 1; thus + * this matches 0xc[23]. */ #define UTF8_IS_DOWNGRADEABLE_START(c) (((U8)(c) & 0xfe) == 0xc2) +/* Is the UTF8-encoded byte 'c' the first byte of a sequence of bytes that + * represent a code point > 255? */ #define UTF8_IS_ABOVE_LATIN1(c) ((U8)(c) >= 0xc4) -#define UTF_START_MARK(len) (((len) > 7) ? 0xFF : (0xFE << (7-(len)))) +/* This defines the 1-bits that are to be in the first byte of a multi-byte + * UTF-8 encoded character that give the number of bytes that comprise the + * character. 'len' is the number of bytes in the multi-byte sequence. */ +#define UTF_START_MARK(len) (((len) > 7) ? 0xFF : (0xFF & (0xFE << (7-(len))))) /* Masks out the initial one bits in a start byte, leaving the real data ones. - * Doesn't work on an invariant byte */ + * Doesn't work on an invariant byte. 'len' is the number of bytes in the + * multi-byte sequence that comprises the character. */ #define UTF_START_MASK(len) (((len) >= 7) ? 0x00 : (0x1F >> ((len)-2))) +/* This defines the bits that are to be in the continuation bytes of a multi-byte + * UTF-8 encoded character that indicate it is a continuation byte. */ #define UTF_CONTINUATION_MARK 0x80 + +/* This is the number of low-order bits a continuation byte in a UTF-8 encoded + * sequence contributes to the specification of the code point. In the bit + * maps above, you see that the first 2 bits are a constant '10', leaving 6 of + * real information */ #define UTF_ACCUMULATION_SHIFT 6 /* 2**UTF_ACCUMULATION_SHIFT - 1 */ #define UTF_CONTINUATION_MASK ((U8)0x3f) -/* This sets the UTF_CONTINUATION_MASK in the upper bits of a word. If a value - * is anded with it, and the result is non-zero, then using the original value - * in UTF8_ACCUMULATE will overflow, shifting bits off the left */ +/* If a value is anded with this, and the result is non-zero, then using the + * original value in UTF8_ACCUMULATE will overflow, shifting bits off the left + * */ #define UTF_ACCUMULATION_OVERFLOW_MASK \ (((UV) UTF_CONTINUATION_MASK) << ((sizeof(UV) * CHARBITS) \ - UTF_ACCUMULATION_SHIFT)) -#ifdef HAS_QUAD -#define UNISKIP(uv) ( (uv) < 0x80 ? 1 : \ +#if UVSIZE >= 8 +# define UTF8_QUAD_MAX UINT64_C(0x1000000000) + +/* Input is a true Unicode (not-native) code point */ +#define OFFUNISKIP(uv) ( (uv) < 0x80 ? 1 : \ (uv) < 0x800 ? 2 : \ (uv) < 0x10000 ? 3 : \ (uv) < 0x200000 ? 4 : \ @@ -207,7 +246,7 @@ Perl's extended UTF-8 means we can have start bytes up to FF. (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 : \ +#define OFFUNISKIP(uv) ( (uv) < 0x80 ? 1 : \ (uv) < 0x800 ? 2 : \ (uv) < 0x10000 ? 3 : \ (uv) < 0x200000 ? 4 : \ @@ -232,6 +271,8 @@ Perl's extended UTF-8 means we can have start bytes up to FF. #error UTF8_MAXBYTES must be at least 12 #endif +#define MAX_UTF8_TWO_BYTE 0x7FF + #define UTF8_MAXBYTES_CASE UTF8_MAXBYTES #endif /* EBCDIC vs ASCII */ @@ -241,47 +282,129 @@ Perl's extended UTF-8 means we can have start bytes up to FF. * 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)) +/* In domain restricted to ASCII, these may make more sense to the reader than + * the ones with Latin1 in the name */ +#define NATIVE_TO_ASCII(ch) NATIVE_TO_LATIN1(ch) +#define ASCII_TO_NATIVE(ch) LATIN1_TO_NATIVE(ch) + +/* More or less misleadingly-named defines, retained for back compat */ +#define NATIVE_TO_UTF(ch) NATIVE_UTF8_TO_I8(ch) +#define NATIVE_TO_I8(ch) NATIVE_UTF8_TO_I8(ch) +#define UTF_TO_NATIVE(ch) I8_TO_NATIVE_UTF8(ch) +#define I8_TO_NATIVE(ch) I8_TO_NATIVE_UTF8(ch) +#define NATIVE8_TO_UNI(ch) NATIVE_TO_LATIN1(ch) + +/* Adds a UTF8 continuation byte 'new' of information to a running total code + * point 'old' of all the continuation bytes so far. This is designed to be + * used in a loop to convert from UTF-8 to the code point represented. Note + * that this is asymmetric on EBCDIC platforms, in that the 'new' parameter is + * the UTF-EBCDIC byte, whereas the 'old' parameter is a Unicode (not EBCDIC) + * code point in process of being generated */ +#define UTF8_ACCUMULATE(old, new) (((old) << UTF_ACCUMULATION_SHIFT) \ + | ((NATIVE_UTF8_TO_I8((U8)new)) \ + & UTF_CONTINUATION_MASK)) /* This works in the face of malformed UTF-8. */ #define UTF8_IS_NEXT_CHAR_DOWNGRADEABLE(s, e) (UTF8_IS_DOWNGRADEABLE_START(*s) \ && ( (e) - (s) > 1) \ && UTF8_IS_CONTINUATION(*((s)+1))) -/* Convert a two (not one) byte utf8 character to a unicode code point value. +/* Number of bytes a code point occupies in UTF-8. */ +#define NATIVE_SKIP(uv) OFFUNISKIP(NATIVE_TO_UNI(uv)) + +/* Most code which says UNISKIP is really thinking in terms of native code + * points (0-255) plus all those beyond. This is an imprecise term, but having + * it means existing code continues to work. For precision, use NATIVE_SKIP + * and OFFUNISKIP */ +#define UNISKIP(uv) NATIVE_SKIP(uv) + +/* Convert a two (not one) byte utf8 character to a native code point value. * Needs just one iteration of accumulate. Should not be used unless it is * known that the two bytes are legal: 1) two-byte start, and 2) continuation. * Note that the result can be larger than 255 if the input character is not * downgradable */ -#define TWO_BYTE_UTF8_TO_UNI(HI, LO) \ - UTF8_ACCUMULATE((NATIVE_TO_UTF(HI) & UTF_START_MASK(2)), \ - NATIVE_TO_UTF(LO)) +#define TWO_BYTE_UTF8_TO_NATIVE(HI, LO) \ + UNI_TO_NATIVE(UTF8_ACCUMULATE((NATIVE_UTF8_TO_I8(HI) & UTF_START_MASK(2)), \ + (LO))) + +/* Should never be used, and be deprecated */ +#define TWO_BYTE_UTF8_TO_UNI(HI, LO) NATIVE_TO_UNI(TWO_BYTE_UTF8_TO_NATIVE(HI, LO)) +/* How many bytes in the UTF-8 encoded character whose first (perhaps only) + * byte is pointed to by 's' */ #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)) +/* Is the byte 'c' the same character when encoded in UTF-8 as when not. This + * works on both UTF-8 encoded strings and non-encoded, as it returns TRUE in + * each for the exact same set of bit patterns. (And it works on any byte in a + * UTF-8 encoded string) */ +#define UTF8_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_UTF8_TO_I8(c)) -#define MAX_PORTABLE_UTF8_TWO_BYTE 0x3FF /* constrained by EBCDIC */ +/* Like the above, but its name implies a non-UTF8 input, and is implemented + * differently (for no particular reason) */ +#define NATIVE_BYTE_IS_INVARIANT(c) UNI_IS_INVARIANT(NATIVE_TO_LATIN1(c)) -/* 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) NATIVE_TO_I8(((c) \ - >> UTF_ACCUMULATION_SHIFT) | (0xFF & UTF_START_MARK(2))) -#define UTF8_TWO_BYTE_LO_nocast(c) NATIVE_TO_I8(((c) & UTF_CONTINUATION_MASK) \ - | UTF_CONTINUATION_MARK) +/* Like the above, but accepts any UV as input */ +#define UVCHR_IS_INVARIANT(uv) UNI_IS_INVARIANT(NATIVE_TO_UNI(uv)) -#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))) +#define MAX_PORTABLE_UTF8_TWO_BYTE 0x3FF /* constrained by EBCDIC */ + +/* The macros in the next 4 sets are used to generate the two utf8 or utfebcdic + * bytes from an ordinal that is known to fit into exactly two (not one) bytes; + * it must be less than 0x3FF to work across both encodings. */ + +/* These two are helper macros for the other three sets, and should not be used + * directly anywhere else. 'translate_function' is either NATIVE_TO_LATIN1 + * (which works for code points up to 0xFF) or NATIVE_TO_UNI which works for any + * code point */ +#define __BASE_TWO_BYTE_HI(c, translate_function) \ + I8_TO_NATIVE_UTF8((translate_function(c) >> UTF_ACCUMULATION_SHIFT) \ + | UTF_START_MARK(2)) +#define __BASE_TWO_BYTE_LO(c, translate_function) \ + I8_TO_NATIVE_UTF8((translate_function(c) & UTF_CONTINUATION_MASK) \ + | UTF_CONTINUATION_MARK) + +/* This is another helper macro to avoid preprocessor issues, expanding to an + * assert followed by a comma under DEBUGGING (hence the comma operator). If + * we didn't do this, we would get a comma with nothing before it when not + * DEBUGGING */ +#ifdef DEBUGGING +# define __ASSERT_(statement) assert(statement), +#else +# define __ASSERT_(statement) +#endif -/* 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)) +/* The next two macros should not be used. They were designed to be usable as + * the case label of a switch statement, but this doesn't work for EBCDIC. Use + * regen/unicode_constants.pl instead */ +#define UTF8_TWO_BYTE_HI_nocast(c) __BASE_TWO_BYTE_HI(c, NATIVE_TO_UNI) +#define UTF8_TWO_BYTE_LO_nocast(c) __BASE_TWO_BYTE_LO(c, NATIVE_TO_UNI) + +/* The next two macros are used when the source should be a single byte + * character; checked for under DEBUGGING */ +#define UTF8_EIGHT_BIT_HI(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + ((U8) __BASE_TWO_BYTE_HI(c, NATIVE_TO_LATIN1))) +#define UTF8_EIGHT_BIT_LO(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + ((U8) __BASE_TWO_BYTE_LO(c, NATIVE_TO_LATIN1))) + +/* These final two macros in the series are used when the source can be any + * code point whose UTF-8 is known to occupy 2 bytes; they are less efficient + * than the EIGHT_BIT versions on EBCDIC platforms. We use the logical '~' + * operator instead of "<=" to avoid getting compiler warnings. + * MAX_PORTABLE_UTF8_TWO_BYTE should be exactly all one bits in the lower few + * places, so the ~ works */ +#define UTF8_TWO_BYTE_HI(c) \ + (__ASSERT_((sizeof(c) == 1) \ + || !(((WIDEST_UTYPE)(c)) & ~MAX_PORTABLE_UTF8_TWO_BYTE)) \ + ((U8) __BASE_TWO_BYTE_HI(c, NATIVE_TO_LATIN1))) +#define UTF8_TWO_BYTE_LO(c) \ + (__ASSERT_((sizeof(c) == 1) \ + || !(((WIDEST_UTYPE)(c)) & ~MAX_PORTABLE_UTF8_TWO_BYTE)) \ + ((U8) __BASE_TWO_BYTE_LO(c, NATIVE_TO_LATIN1))) + +/* This is illegal in any well-formed UTF-8 in both EBCDIC and ASCII + * as it is only in overlongs. */ +#define ILLEGAL_UTF8_BYTE I8_TO_NATIVE_UTF8(0xC1) /* * 'UTF' is whether or not p is encoded in UTF8. The names 'foo_lazy_if' stem @@ -293,12 +416,10 @@ Perl's extended UTF-8 means we can have start bytes up to FF. #define isIDFIRST_lazy_if(p,UTF) ((IN_BYTES || !UTF ) \ ? isIDFIRST(*(p)) \ : isIDFIRST_utf8((const U8*)p)) -#define isALNUM_lazy_if(p,UTF) ((IN_BYTES || (!UTF )) \ - ? 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 isWORDCHAR_lazy_if(p,UTF) ((IN_BYTES || (!UTF )) \ + ? isWORDCHAR(*(p)) \ + : isWORDCHAR_utf8((const U8*)p)) +#define isALNUM_lazy_if(p,UTF) isWORDCHAR_lazy_if(p,UTF) #define UTF8_MAXLEN UTF8_MAXBYTES @@ -375,11 +496,12 @@ Perl's extended UTF-8 means we can have start bytes up to FF. * U+110001: \xF4\x90\x80\x81 \xF9\xA2\xA0\xA0\xA1 */ #ifdef EBCDIC /* Both versions assume well-formed UTF8 */ -# define UTF8_IS_SUPER(s) (NATIVE_TO_I8(*(s)) >= 0xF9 \ - && (NATIVE_TO_I8(*(s)) > 0xF9) || (NATIVE_TO_I8(*((s)) + 1 >= 0xA2))) +# define UTF8_IS_SUPER(s) (NATIVE_UTF8_TO_I8(* (U8*) (s)) >= 0xF9 \ + && (NATIVE_UTF8_TO_I8(* (U8*) (s)) > 0xF9 \ + || (NATIVE_UTF8_TO_I8(* ((U8*) (s) + 1)) >= 0xA2))) #else -# define UTF8_IS_SUPER(s) (*(s) >= 0xF4 \ - && (*(s) > 0xF4 || (*((s) + 1) >= 0x90))) +# define UTF8_IS_SUPER(s) (*(U8*) (s) >= 0xF4 \ + && (*(U8*) (s) > 0xF4 || (*((U8*) (s) + 1) >= 0x90))) #endif /* These are now machine generated, and the 'given' clause is no longer @@ -430,10 +552,6 @@ Perl's extended UTF-8 means we can have start bytes up to FF. #define UNICODE_IS_SUPER(c) ((c) > PERL_UNICODE_MAX) #define UNICODE_IS_FE_FF(c) ((c) > 0x7FFFFFFF) -#ifdef HAS_QUAD -# define UTF8_QUAD_MAX UINT64_C(0x1000000000) -#endif - #define LATIN_SMALL_LETTER_SHARP_S LATIN_SMALL_LETTER_SHARP_S_NATIVE #define LATIN_SMALL_LETTER_Y_WITH_DIAERESIS \ LATIN_SMALL_LETTER_Y_WITH_DIAERESIS_NATIVE @@ -446,12 +564,15 @@ Perl's extended UTF-8 means we can have start bytes up to FF. #define UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA 0x03C2 #define UNICODE_GREEK_SMALL_LETTER_SIGMA 0x03C3 #define GREEK_SMALL_LETTER_MU 0x03BC -#define GREEK_CAPITAL_LETTER_MU 0x039C /* Upper and title case of MICRON */ -#define LATIN_CAPITAL_LETTER_Y_WITH_DIAERESIS 0x0178 /* Also is title case */ -#define LATIN_CAPITAL_LETTER_SHARP_S 0x1E9E -#define LATIN_SMALL_LETTER_LONG_S 0x017F -#define KELVIN_SIGN 0x212A -#define ANGSTROM_SIGN 0x212B +#define GREEK_CAPITAL_LETTER_MU 0x039C /* Upper and title case + of MICRON */ +#define LATIN_CAPITAL_LETTER_Y_WITH_DIAERESIS 0x0178 /* Also is title case */ +#define LATIN_CAPITAL_LETTER_SHARP_S 0x1E9E +#define LATIN_SMALL_LETTER_LONG_S 0x017F +#define LATIN_SMALL_LIGATURE_LONG_S_T 0xFB05 +#define LATIN_SMALL_LIGATURE_ST 0xFB06 +#define KELVIN_SIGN 0x212A +#define ANGSTROM_SIGN 0x212B #define UNI_DISPLAY_ISPRINT 0x0001 #define UNI_DISPLAY_BACKSLASH 0x0002 @@ -463,8 +584,8 @@ Perl's extended UTF-8 means we can have start bytes up to FF. (ANYOF_NONBITMAP(node)) && \ (ANYOF_FLAGS(node) & ANYOF_LOC_NONBITMAP_FOLD) && \ ((end) > (input) + 1) && \ - toLOWER((input)[0]) == 's' && \ - toLOWER((input)[1]) == 's') + toFOLD((input)[0]) == 's' && \ + toFOLD((input)[1]) == 's') #define SHARP_S_SKIP 2 /* If you want to exclude surrogates, and beyond legal Unicode, see the blame