X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/5352a76315308cf38aae6f4e6f20f967d5c88300..90a9496b5607497839aa08a0ddcdfa54c9d47678:/utf8.h diff --git a/utf8.h b/utf8.h index 99545a2..4d2d01b 100644 --- a/utf8.h +++ b/utf8.h @@ -14,8 +14,8 @@ * */ -#ifndef H_UTF8 /* Guard against recursive inclusion */ -#define H_UTF8 1 +#ifndef PERL_UTF8_H_ /* Guard against recursive inclusion */ +#define PERL_UTF8_H_ 1 /* Use UTF-8 as the default script encoding? * Turning this on will break scripts having non-UTF-8 binary @@ -50,14 +50,21 @@ Search for the string "utf8" in this document. =for apidoc is_ascii_string -This is a misleadingly-named synonym for L. +This is a misleadingly-named synonym for L. On ASCII-ish platforms, the name isn't misleading: the ASCII-range characters are exactly the UTF-8 invariants. But EBCDIC machines have more invariants -than just the ASCII characters, so C is preferred. +than just the ASCII characters, so C is preferred. + +=for apidoc is_invariant_string + +This is a somewhat misleadingly-named synonym for L. +C is preferred, as it indicates under what conditions +the string is invariant. =cut */ -#define is_ascii_string(s, len) is_invariant_string(s, len) +#define is_ascii_string(s, len) is_utf8_invariant_string(s, len) +#define is_invariant_string(s, len) is_utf8_invariant_string(s, len) #define uvchr_to_utf8(a,b) uvchr_to_utf8_flags(a,b,0) #define uvchr_to_utf8_flags(d,uv,flags) \ @@ -65,15 +72,19 @@ than just the ASCII characters, so C is preferred. #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 utf8n_to_uvchr(s, len, lenp, flags) \ + utf8n_to_uvchr_error(s, len, lenp, flags, 0) #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) -#define to_utf8_lower(a,b,c) _to_utf8_lower_flags(a,b,c,0) -#define to_utf8_upper(a,b,c) _to_utf8_upper_flags(a,b,c,0) -#define to_utf8_title(a,b,c) _to_utf8_title_flags(a,b,c,0) -/* Source backward compatibility. */ -#define is_utf8_string_loc(s, len, ep) is_utf8_string_loclen(s, len, ep, 0) +#define to_utf8_fold(s, r, lenr) \ + _to_utf8_fold_flags (s, NULL, r, lenr, FOLD_FLAGS_FULL, __FILE__, __LINE__) +#define to_utf8_lower(s, r, lenr) \ + _to_utf8_lower_flags(s, NULL, r ,lenr, 0, __FILE__, __LINE__) +#define to_utf8_upper(s, r, lenr) \ + _to_utf8_upper_flags(s, NULL, r, lenr, 0, __FILE__, __LINE__) +#define to_utf8_title(s, r, lenr) \ + _to_utf8_title_flags(s, NULL, r, lenr ,0, __FILE__, __LINE__) #define foldEQ_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2) \ foldEQ_utf8_flags(s1, pe1, l1, u1, s2, pe2, l2, u2, 0) @@ -97,6 +108,13 @@ than just the ASCII characters, so C is preferred. #else /* ! EBCDIC */ START_EXTERN_C +/* 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 Unicode's upper limit, 0x10FFFF, can be + * expressed with 4 bytes. However, Perl thinks of UTF-8 as a way to encode + * non-negative integers in a binary format, even those above Unicode */ +#define UTF8_MAXBYTES 13 + #ifdef DOINIT EXTCONST unsigned char PL_utf8skip[] = { /* 0x00 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */ @@ -116,8 +134,10 @@ EXTCONST unsigned char PL_utf8skip[] = { /* 0xD0 */ 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, /* U+0400 to U+07FF */ /* 0xE0 */ 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, /* U+0800 to U+FFFF */ /* 0xF0 */ 4,4,4,4,4,4,4,4,5,5,5,5,6,6, /* above BMP to 2**31 - 1 */ -/* 0xFE */ 7,13, /* Perl extended (never was official UTF-8). Up to 72bit - allowed (64-bit + reserved). */ + /* Perl extended (never was official UTF-8). Up to 36 bit */ +/* 0xFE */ 7, + /* More extended, Up to 72 bits (64-bit + reserved) */ +/* 0xFF */ UTF8_MAXBYTES }; #else EXTCONST unsigned char PL_utf8skip[]; @@ -125,10 +145,20 @@ EXTCONST unsigned char PL_utf8skip[]; END_EXTERN_C +#if defined(_MSC_VER) && _MSC_VER < 1400 +/* older MSVC versions have a smallish macro buffer */ +#define PERL_SMALL_MACRO_BUFFER +#endif + /* 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) +#ifdef PERL_SMALL_MACRO_BUFFER +#define NATIVE_TO_LATIN1(ch) ((U8)(ch)) +#define LATIN1_TO_NATIVE(ch) ((U8)(ch)) +#else +#define NATIVE_TO_LATIN1(ch) (__ASSERT_(FITS_IN_8_BITS(ch)) ((U8) (ch))) +#define LATIN1_TO_NATIVE(ch) (__ASSERT_(FITS_IN_8_BITS(ch)) ((U8) (ch))) +#endif /* 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 @@ -136,33 +166,44 @@ END_EXTERN_C * 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) +#ifdef PERL_SMALL_MACRO_BUFFER +#define NATIVE_UTF8_TO_I8(ch) ((U8) (ch)) +#define I8_TO_NATIVE_UTF8(ch) ((U8) (ch)) +#else +#define NATIVE_UTF8_TO_I8(ch) (__ASSERT_(FITS_IN_8_BITS(ch)) ((U8) (ch))) +#define I8_TO_NATIVE_UTF8(ch) (__ASSERT_(FITS_IN_8_BITS(ch)) ((U8) (ch))) +#endif /* Transforms in wide UV chars */ -#define UNI_TO_NATIVE(ch) (ch) -#define NATIVE_TO_UNI(ch) (ch) +#define UNI_TO_NATIVE(ch) ((UV) (ch)) +#define NATIVE_TO_UNI(ch) ((UV) (ch)) /* - The following table is from Unicode 3.2. + The following table is from Unicode 3.2, plus the Perl extensions for above + U+10FFFF - Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte + Code Points 1st Byte 2nd Byte 3rd 4th 5th 6th 7th 8th-13th 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 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 - U+100000..U+10FFFF F4 80..8F 80..BF 80..BF - 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+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 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 + U+100000..U+10FFFF F4 80..8F 80..BF 80..BF + Below are above-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..U+FFFFFF F8 * 88..BF 80..BF 80..BF 80..BF +U+1000000..U+3FFFFFF F9..FB 80..BF 80..BF 80..BF 80..BF +U+4000000..U+3FFFFFFF FC * 84..BF 80..BF 80..BF 80..BF 80..BF +U+40000000..U+7FFFFFFF FD 80..BF 80..BF 80..BF 80..BF 80..BF +U+80000000..U+FFFFFFFFF FE * 82..BF 80..BF 80..BF 80..BF 80..BF 80..BF +U+1000000000.. FF 80..BF 80..BF 80..BF 80..BF 80..BF * 81..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 @@ -186,55 +227,70 @@ As you can see, the continuation bytes all begin with C<10>, and the leading bits of the start byte tell how many bytes there are in the encoded character. -Perl's extended UTF-8 means we can have start bytes up to FF. +Perl's extended UTF-8 means we can have start bytes up through FF, though any +beginning with FF yields a code point that is too large for 32-bit ASCII +platforms. FF signals to use 13 bytes for the encoded character. This breaks +the paradigm that the number of leading bits gives how many total bytes there +are in the character. */ /* Is the representation of the Unicode code point 'cp' the same regardless of * being encoded in UTF-8 or not? */ -#define UNI_IS_INVARIANT(cp) isASCII(cp) +#define OFFUNI_IS_INVARIANT(cp) isASCII(cp) -/* Is the representation of the code point 'cp' the same regardless of - * being encoded in UTF-8 or not? 'cp' is native if < 256; Unicode otherwise - * */ -#define UVCHR_IS_INVARIANT(uv) UNI_IS_INVARIANT(uv) +/* +=for apidoc Am|bool|UVCHR_IS_INVARIANT|UV cp + +Evaluates to 1 if the representation of code point C is the same whether or +not it is encoded in UTF-8; otherwise evaluates to 0. UTF-8 invariant +characters can be copied as-is when converting to/from UTF-8, saving time. +C is Unicode if above 255; otherwise is platform-native. -/* 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) +=cut + */ + +#define UVCHR_IS_INVARIANT(cp) OFFUNI_IS_INVARIANT(cp) + +/* This defines the bits that are to be in the continuation bytes of a multi-byte + * UTF-8 encoded character that mark it is a continuation byte. */ +#define UTF_CONTINUATION_MARK 0x80 + +/* Misleadingly named: is the UTF8-encoded byte 'c' part of a variant sequence + * in UTF-8? This is the inverse of UTF8_IS_INVARIANT. The |0 makes sure this + * isn't mistakenly called with a ptr argument */ +#define UTF8_IS_CONTINUED(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + ((U8)((c) | 0)) & UTF_CONTINUATION_MARK) /* 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) + * illegal overlong sequences that begin with C0 and C1. The |0 makes sure + * this isn't mistakenly called with a ptr argument */ +#define UTF8_IS_START(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + ((U8)((c) | 0)) >= 0xc2) + +/* For use in UTF8_IS_CONTINUATION() below */ +#define UTF_IS_CONTINUATION_MASK 0xC0 /* 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) + * first byte thereof? The |0 makes sure this isn't mistakenly called with a + * ptr argument */ +#define UTF8_IS_CONTINUATION(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + (((U8)((c) | 0)) & UTF_IS_CONTINUATION_MASK) == UTF_CONTINUATION_MARK) /* 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) + * this matches 0xc[23]. The |0 makes sure this isn't mistakenly called with a + * ptr argument */ +#define UTF8_IS_DOWNGRADEABLE_START(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + (((U8)((c) | 0)) & 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) - -/* 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. '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 + * represent a code point > 255? The |0 makes sure this isn't mistakenly + * called with a ptr argument */ +#define UTF8_IS_ABOVE_LATIN1(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + ((U8)((c) | 0)) >= 0xc4) /* 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 @@ -242,61 +298,168 @@ Perl's extended UTF-8 means we can have start bytes up to FF. * real information */ #define UTF_ACCUMULATION_SHIFT 6 -/* 2**UTF_ACCUMULATION_SHIFT - 1 */ -#define UTF_CONTINUATION_MASK ((U8)0x3f) +/* ^? is defined to be DEL on ASCII systems. See the definition of toCTRL() + * for more */ +#define QUESTION_MARK_CTRL DEL_NATIVE -/* 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)) +/* Surrogates, non-character code points and above-Unicode code points are + * problematic in some contexts. This allows code that needs to check for + * those to to quickly exclude the vast majority of code points it will + * encounter */ +#define isUTF8_POSSIBLY_PROBLEMATIC(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + (U8) c >= 0xED) -#if UVSIZE >= 8 -# define UTF8_QUAD_MAX UINT64_C(0x1000000000) +/* A helper macro for isUTF8_CHAR, so use that one instead of this. This was + * generated by regen/regcharclass.pl, and then moved here. Then it was + * hand-edited to add some LIKELY() calls, presuming that malformations are + * unlikely. The lines that generated it were then commented out. This was + * done because it takes on the order of 10 minutes to generate, and is never + * going to change, unless the generated code is improved, and figuring out + * the LIKELYs there would be hard. + * + UTF8_CHAR: Matches legal UTF-8 variant code points up through 0x1FFFFFF -/* Input is a true Unicode (not-native) code point */ -#define OFFUNISKIP(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 OFFUNISKIP(uv) ( (uv) < 0x80 ? 1 : \ - (uv) < 0x800 ? 2 : \ - (uv) < 0x10000 ? 3 : \ - (uv) < 0x200000 ? 4 : \ - (uv) < 0x4000000 ? 5 : \ - (uv) < 0x80000000 ? 6 : 7 ) -#endif + 0x80 - 0x1FFFFF +*/ +/*** GENERATED CODE ***/ +#define is_UTF8_CHAR_utf8_no_length_checks(s) \ +( ( 0xC2 <= ((const U8*)s)[0] && ((const U8*)s)[0] <= 0xDF ) ? \ + ( LIKELY( ( ((const U8*)s)[1] & 0xC0 ) == 0x80 ) ? 2 : 0 ) \ +: ( 0xE0 == ((const U8*)s)[0] ) ? \ + ( LIKELY( ( ( ((const U8*)s)[1] & 0xE0 ) == 0xA0 ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ +: ( 0xE1 <= ((const U8*)s)[0] && ((const U8*)s)[0] <= 0xEF ) ? \ + ( LIKELY( ( ( ((const U8*)s)[1] & 0xC0 ) == 0x80 ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ +: ( 0xF0 == ((const U8*)s)[0] ) ? \ + ( LIKELY( ( ( 0x90 <= ((const U8*)s)[1] && ((const U8*)s)[1] <= 0xBF ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) && ( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 )\ +: ( ( ( ( 0xF1 <= ((const U8*)s)[0] && ((const U8*)s)[0] <= 0xF7 ) && LIKELY( ( ((const U8*)s)[1] & 0xC0 ) == 0x80 ) ) && LIKELY( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) && LIKELY( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 ) + +/* The above macro handles UTF-8 that has this start byte as the maximum */ +#define _IS_UTF8_CHAR_HIGHEST_START_BYTE 0xF7 + +/* A helper macro for isSTRICT_UTF8_CHAR, so use that one instead of this. + * Like is_UTF8_CHAR_utf8_no_length_checks(), this was moved here and LIKELYs + * added manually. + * + STRICT_UTF8_CHAR: Matches legal Unicode UTF-8 variant code points, no + surrrogates nor non-character code points +*/ +/*** GENERATED CODE ***/ +#define is_STRICT_UTF8_CHAR_utf8_no_length_checks(s) \ +( ( 0xC2 <= ((const U8*)s)[0] && ((const U8*)s)[0] <= 0xDF ) ? \ + ( LIKELY( ( ((const U8*)s)[1] & 0xC0 ) == 0x80 ) ? 2 : 0 ) \ +: ( 0xE0 == ((const U8*)s)[0] ) ? \ + ( LIKELY( ( ( ((const U8*)s)[1] & 0xE0 ) == 0xA0 ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ +: ( ( 0xE1 <= ((const U8*)s)[0] && ((const U8*)s)[0] <= 0xEC ) || 0xEE == ((const U8*)s)[0] ) ?\ + ( ( ( ( ((const U8*)s)[1] & 0xC0 ) == 0x80 ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ +: ( 0xED == ((const U8*)s)[0] ) ? \ + ( LIKELY( ( ( ((const U8*)s)[1] & 0xE0 ) == 0x80 ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ +: ( 0xEF == ((const U8*)s)[0] ) ? \ + ( ( ( 0x80 <= ((const U8*)s)[1] && ((const U8*)s)[1] <= 0xB6 ) || ( 0xB8 <= ((const U8*)s)[1] && ((const U8*)s)[1] <= 0xBE ) ) ?\ + ( LIKELY( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ? 3 : 0 ) \ + : ( 0xB7 == ((const U8*)s)[1] ) ? \ + ( LIKELY( ( ((const U8*)s)[2] & 0xF0 ) == 0x80 || ( ((const U8*)s)[2] & 0xF0 ) == 0xB0 ) ? 3 : 0 )\ + : ( ( 0xBF == ((const U8*)s)[1] ) && ( 0x80 <= ((const U8*)s)[2] && ((const U8*)s)[2] <= 0xBD ) ) ? 3 : 0 )\ +: ( 0xF0 == ((const U8*)s)[0] ) ? \ + ( ( ( 0x90 <= ((const U8*)s)[1] && ((const U8*)s)[1] <= 0x9E ) || ( 0xA0 <= ((const U8*)s)[1] && ((const U8*)s)[1] <= 0xAE ) || ( 0xB0 <= ((const U8*)s)[1] && ((const U8*)s)[1] <= 0xBE ) ) ?\ + ( LIKELY( ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) && ( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 )\ + : ( ((const U8*)s)[1] == 0x9F || ( ( ((const U8*)s)[1] & 0xEF ) == 0xAF ) ) ? \ + ( ( 0x80 <= ((const U8*)s)[2] && ((const U8*)s)[2] <= 0xBE ) ? \ + ( LIKELY( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ? 4 : 0 ) \ + : LIKELY( ( 0xBF == ((const U8*)s)[2] ) && ( 0x80 <= ((const U8*)s)[3] && ((const U8*)s)[3] <= 0xBD ) ) ? 4 : 0 )\ + : 0 ) \ +: ( 0xF1 <= ((const U8*)s)[0] && ((const U8*)s)[0] <= 0xF3 ) ? \ + ( ( ( ( ((const U8*)s)[1] & 0xC8 ) == 0x80 ) || ( ( ((const U8*)s)[1] & 0xCC ) == 0x88 ) || ( ( ((const U8*)s)[1] & 0xCE ) == 0x8C ) || ( ( ((const U8*)s)[1] & 0xCF ) == 0x8E ) ) ?\ + ( LIKELY( ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) && ( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 )\ + : ( ( ((const U8*)s)[1] & 0xCF ) == 0x8F ) ? \ + ( ( 0x80 <= ((const U8*)s)[2] && ((const U8*)s)[2] <= 0xBE ) ? \ + ( LIKELY( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ? 4 : 0 ) \ + : LIKELY( ( 0xBF == ((const U8*)s)[2] ) && ( 0x80 <= ((const U8*)s)[3] && ((const U8*)s)[3] <= 0xBD ) ) ? 4 : 0 )\ + : 0 ) \ +: ( 0xF4 == ((const U8*)s)[0] ) ? \ + ( ( 0x80 <= ((const U8*)s)[1] && ((const U8*)s)[1] <= 0x8E ) ? \ + ( LIKELY( ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) && ( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 )\ + : ( 0x8F == ((const U8*)s)[1] ) ? \ + ( ( 0x80 <= ((const U8*)s)[2] && ((const U8*)s)[2] <= 0xBE ) ? \ + ( LIKELY( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ? 4 : 0 ) \ + : LIKELY( ( 0xBF == ((const U8*)s)[2] ) && ( 0x80 <= ((const U8*)s)[3] && ((const U8*)s)[3] <= 0xBD ) ) ? 4 : 0 )\ + : 0 ) \ +: 0 ) + +/* Similarly, + C9_STRICT_UTF8_CHAR: Matches legal Unicode UTF-8 variant code + points, no surrogates + 0x0080 - 0xD7FF + 0xE000 - 0x10FFFF +*/ +/*** GENERATED CODE ***/ +#define is_C9_STRICT_UTF8_CHAR_utf8_no_length_checks(s) \ +( ( 0xC2 <= ((const U8*)s)[0] && ((const U8*)s)[0] <= 0xDF ) ? \ + ( LIKELY( ( ((const U8*)s)[1] & 0xC0 ) == 0x80 ) ? 2 : 0 ) \ +: ( 0xE0 == ((const U8*)s)[0] ) ? \ + ( LIKELY( ( ( ((const U8*)s)[1] & 0xE0 ) == 0xA0 ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ +: ( ( 0xE1 <= ((const U8*)s)[0] && ((const U8*)s)[0] <= 0xEC ) || ( ((const U8*)s)[0] & 0xFE ) == 0xEE ) ?\ + ( LIKELY( ( ( ((const U8*)s)[1] & 0xC0 ) == 0x80 ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ +: ( 0xED == ((const U8*)s)[0] ) ? \ + ( LIKELY( ( ( ((const U8*)s)[1] & 0xE0 ) == 0x80 ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ +: ( 0xF0 == ((const U8*)s)[0] ) ? \ + ( LIKELY( ( ( 0x90 <= ((const U8*)s)[1] && ((const U8*)s)[1] <= 0xBF ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) && ( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 )\ +: ( 0xF1 <= ((const U8*)s)[0] && ((const U8*)s)[0] <= 0xF3 ) ? \ + ( LIKELY( ( ( ( ((const U8*)s)[1] & 0xC0 ) == 0x80 ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) && ( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 )\ +: LIKELY( ( ( ( 0xF4 == ((const U8*)s)[0] ) && ( ( ((const U8*)s)[1] & 0xF0 ) == 0x80 ) ) && ( ( ((const U8*)s)[2] & 0xC0 ) == 0x80 ) ) && ( ( ((const U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 ) + +#define UNICODE_IS_PERL_EXTENDED(uv) UNLIKELY((UV) (uv) > 0x7FFFFFFF) -/* 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 Unicode's upper limit, 0x10FFFF, can be - * expressed with 4 bytes. However, Perl thinks of UTF-8 as a way to encode - * non-negative integers in a binary format, even those above Unicode */ -#define UTF8_MAXBYTES 13 +#endif /* EBCDIC vs ASCII */ -/* The maximum number of UTF-8 bytes a single Unicode character can - * uppercase/lowercase/fold into. Unicode guarantees that the maximum - * expansion is 3 characters. On ASCIIish platforms, the highest Unicode - * character occupies 4 bytes, therefore this number would be 12, but this is - * smaller than the maximum width a single above-Unicode character can occupy, - * so use that instead */ -#if UTF8_MAXBYTES < 12 -#error UTF8_MAXBYTES must be at least 12 +/* 2**UTF_ACCUMULATION_SHIFT - 1 */ +#define UTF_CONTINUATION_MASK ((U8) ((1U << UTF_ACCUMULATION_SHIFT) - 1)) + +/* Internal macro to be used only in this file to aid in constructing other + * publicly accessible macros. + * The number of bytes required to express this uv in UTF-8, for just those + * uv's requiring 2 through 6 bytes, as these are common to all platforms and + * word sizes. The number of bytes needed is given by the number of leading 1 + * bits in the start byte. There are 32 start bytes that have 2 initial 1 bits + * (C0-DF); there are 16 that have 3 initial 1 bits (E0-EF); 8 that have 4 + * initial 1 bits (F0-F8); 4 that have 5 initial 1 bits (F9-FB), and 2 that + * have 6 initial 1 bits (FC-FD). The largest number a string of n bytes can + * represent is (the number of possible start bytes for 'n') + * * (the number of possiblities for each start byte + * The latter in turn is + * 2 ** ( (how many continuation bytes there are) + * * (the number of bits of information each + * continuation byte holds)) + * + * If we were on a platform where we could use a fast find first set bit + * instruction (or count leading zeros instruction) this could be replaced by + * using that to find the log2 of the uv, and divide that by the number of bits + * of information in each continuation byte, adjusting for large cases and how + * much information is in a start byte for that length */ +#define __COMMON_UNI_SKIP(uv) \ + (UV) (uv) < (32 * (1U << ( UTF_ACCUMULATION_SHIFT))) ? 2 : \ + (UV) (uv) < (16 * (1U << (2 * UTF_ACCUMULATION_SHIFT))) ? 3 : \ + (UV) (uv) < ( 8 * (1U << (3 * UTF_ACCUMULATION_SHIFT))) ? 4 : \ + (UV) (uv) < ( 4 * (1U << (4 * UTF_ACCUMULATION_SHIFT))) ? 5 : \ + (UV) (uv) < ( 2 * (1U << (5 * UTF_ACCUMULATION_SHIFT))) ? 6 : + +/* Internal macro to be used only in this file. + * This adds to __COMMON_UNI_SKIP the details at this platform's upper range. + * For any-sized EBCDIC platforms, or 64-bit ASCII ones, we need one more test + * to see if just 7 bytes is needed, or if the maximum is needed. For 32-bit + * ASCII platforms, everything is representable by 7 bytes */ +#if defined(UV_IS_QUAD) || defined(EBCDIC) +# define __BASE_UNI_SKIP(uv) (__COMMON_UNI_SKIP(uv) \ + (UV) (uv) < ((UV) 1U << (6 * UTF_ACCUMULATION_SHIFT)) ? 7 : UTF8_MAXBYTES) +#else +# define __BASE_UNI_SKIP(uv) (__COMMON_UNI_SKIP(uv) 7) #endif -/* ^? is defined to be DEL on ASCII systems. See the definition of toCTRL() - * for more */ -#define QUESTION_MARK_CTRL DEL_NATIVE - -#define MAX_UTF8_TWO_BYTE 0x7FF +/* The next two macros use the base macro defined above, and add in the tests + * at the low-end of the range, for just 1 byte, yielding complete macros, + * publicly accessible. */ -#define UTF8_MAXBYTES_CASE UTF8_MAXBYTES +/* Input is a true Unicode (not-native) code point */ +#define OFFUNISKIP(uv) (OFFUNI_IS_INVARIANT(uv) ? 1 : __BASE_UNI_SKIP(uv)) /* @@ -307,9 +470,30 @@ encoded as UTF-8. C is a native (ASCII or EBCDIC) code point if less than =cut */ -#define UVCHR_SKIP(uv) OFFUNISKIP(uv) +#define UVCHR_SKIP(uv) ( UVCHR_IS_INVARIANT(uv) ? 1 : __BASE_UNI_SKIP(uv)) -#endif /* EBCDIC vs ASCII */ +/* The largest code point representable by two UTF-8 bytes on this platform. + * As explained in the comments for __COMMON_UNI_SKIP, 32 start bytes with + * UTF_ACCUMULATION_SHIFT bits of information each */ +#define MAX_UTF8_TWO_BYTE (32 * (1U << UTF_ACCUMULATION_SHIFT) - 1) + +/* The largest code point representable by two UTF-8 bytes on any platform that + * Perl runs on. This value is constrained by EBCDIC which has 5 bits per + * continuation byte */ +#define MAX_PORTABLE_UTF8_TWO_BYTE (32 * (1U << 5) - 1) + +/* The maximum number of UTF-8 bytes a single Unicode character can + * uppercase/lowercase/fold into. Unicode guarantees that the maximum + * expansion is UTF8_MAX_FOLD_CHAR_EXPAND characters, but any above-Unicode + * code point will fold to itself, so we only have to look at the expansion of + * the maximum Unicode code point. But this number may be less than the space + * occupied by a very large code point under Perl's extended UTF-8. We have to + * make it large enough to fit any single character. (It turns out that ASCII + * and EBCDIC differ in which is larger) */ +#define UTF8_MAXBYTES_CASE \ + (UTF8_MAXBYTES >= (UTF8_MAX_FOLD_CHAR_EXPAND * OFFUNISKIP(0x10FFFF)) \ + ? UTF8_MAXBYTES \ + : (UTF8_MAX_FOLD_CHAR_EXPAND * OFFUNISKIP(0x10FFFF))) /* 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 @@ -328,20 +512,33 @@ encoded as UTF-8. C is a native (ASCII or EBCDIC) code point if less than #define I8_TO_NATIVE(ch) I8_TO_NATIVE_UTF8(ch) #define NATIVE8_TO_UNI(ch) NATIVE_TO_LATIN1(ch) +/* This defines the 1-bits that are to be in the first byte of a multi-byte + * UTF-8 encoded character that mark it as a start byte and 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. '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))) + /* 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) \ +#define UTF8_ACCUMULATE(old, new) (__ASSERT_(FITS_IN_8_BITS(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))) +#define UTF8_IS_NEXT_CHAR_DOWNGRADEABLE(s, e) \ + ( UTF8_IS_DOWNGRADEABLE_START(*(s)) \ + && ( (e) - (s) > 1) \ + && UTF8_IS_CONTINUATION(*((s)+1))) /* Number of bytes a code point occupies in UTF-8. */ #define NATIVE_SKIP(uv) UVCHR_SKIP(uv) @@ -376,8 +573,10 @@ encoded as UTF-8. C is a native (ASCII or EBCDIC) code point if less than * Note that the result can be larger than 255 if the input character is not * downgradable */ #define TWO_BYTE_UTF8_TO_NATIVE(HI, LO) \ - ( __ASSERT_(UTF8SKIP(HI) == 2) \ - __ASSERT_(UTF8_IS_CONTINUATION(LO)) \ + (__ASSERT_(FITS_IN_8_BITS(HI)) \ + __ASSERT_(FITS_IN_8_BITS(LO)) \ + __ASSERT_(PL_utf8skip[HI] == 2) \ + __ASSERT_(UTF8_IS_CONTINUATION(LO)) \ UNI_TO_NATIVE(UTF8_ACCUMULATE((NATIVE_UTF8_TO_I8(HI) & UTF_START_MASK(2)), \ (LO)))) @@ -395,16 +594,36 @@ only) byte is pointed to by C. #define UTF8SKIP(s) PL_utf8skip[*(const U8*)(s)] #define UTF8_SKIP(s) UTF8SKIP(s) -/* 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)) +/* Most code that says 'UNI_' really means the native value for code points up + * through 255 */ +#define UNI_IS_INVARIANT(cp) UVCHR_IS_INVARIANT(cp) -/* Like the above, but its name implies a non-UTF8 input */ -#define NATIVE_BYTE_IS_INVARIANT(c) UVCHR_IS_INVARIANT(c) +/* +=for apidoc Am|bool|UTF8_IS_INVARIANT|char c + +Evaluates to 1 if the byte C represents the same character when encoded in +UTF-8 as when not; otherwise evaluates to 0. UTF-8 invariant characters can be +copied as-is when converting to/from UTF-8, saving time. -#define MAX_PORTABLE_UTF8_TWO_BYTE 0x3FF /* constrained by EBCDIC */ +In spite of the name, this macro gives the correct result if the input string +from which C comes is not encoded in UTF-8. + +See C> for checking if a UV is invariant. + +=cut + +The reason it works on both UTF-8 encoded strings and non-UTF-8 encoded, is +that it returns TRUE in each for the exact same set of bit patterns. It is +valid on a subset of what UVCHR_IS_INVARIANT is valid on, so can just use that; +and the compiler should optimize out anything extraneous given the +implementation of the latter. The |0 makes sure this isn't mistakenly called +with a ptr argument. +*/ +#define UTF8_IS_INVARIANT(c) UVCHR_IS_INVARIANT((c) | 0) + +/* Like the above, but its name implies a non-UTF8 input, which as the comments + * above show, doesn't matter as to its implementation */ +#define NATIVE_BYTE_IS_INVARIANT(c) UVCHR_IS_INVARIANT(c) /* 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; @@ -415,11 +634,13 @@ only) byte is pointed to by C. * (which works for code points up through 0xFF) or NATIVE_TO_UNI which works * for any code point */ #define __BASE_TWO_BYTE_HI(c, translate_function) \ + (__ASSERT_(! UVCHR_IS_INVARIANT(c)) \ I8_TO_NATIVE_UTF8((translate_function(c) >> UTF_ACCUMULATION_SHIFT) \ - | UTF_START_MARK(2)) + | UTF_START_MARK(2))) #define __BASE_TWO_BYTE_LO(c, translate_function) \ + (__ASSERT_(! UVCHR_IS_INVARIANT(c)) \ I8_TO_NATIVE_UTF8((translate_function(c) & UTF_CONTINUATION_MASK) \ - | UTF_CONTINUATION_MARK) + | UTF_CONTINUATION_MARK)) /* 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 @@ -430,9 +651,9 @@ only) byte is pointed to by C. /* 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))) + ( __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))) + (__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 @@ -443,11 +664,11 @@ only) byte is pointed to by C. #define UTF8_TWO_BYTE_HI(c) \ (__ASSERT_((sizeof(c) == 1) \ || !(((WIDEST_UTYPE)(c)) & ~MAX_UTF8_TWO_BYTE)) \ - ((U8) __BASE_TWO_BYTE_HI(c, NATIVE_TO_UNI))) + (__BASE_TWO_BYTE_HI(c, NATIVE_TO_UNI))) #define UTF8_TWO_BYTE_LO(c) \ (__ASSERT_((sizeof(c) == 1) \ || !(((WIDEST_UTYPE)(c)) & ~MAX_UTF8_TWO_BYTE)) \ - ((U8) __BASE_TWO_BYTE_LO(c, NATIVE_TO_UNI))) + (__BASE_TWO_BYTE_LO(c, NATIVE_TO_UNI))) /* This is illegal in any well-formed UTF-8 in both EBCDIC and ASCII * as it is only in overlongs. */ @@ -460,20 +681,37 @@ only) byte is pointed to by C. * beginning of a utf8 character. Now that foo_utf8() determines that itself, * no need to do it again here */ -#define isIDFIRST_lazy_if(p,UTF) ((IN_BYTES || !UTF ) \ - ? isIDFIRST(*(p)) \ - : isIDFIRST_utf8((const U8*)p)) -#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 isIDFIRST_lazy_if(p,UTF) \ + _is_utf8_FOO(_CC_IDFIRST, (const U8 *) p, "isIDFIRST_lazy_if", \ + "isIDFIRST_lazy_if_safe", \ + cBOOL(UTF && ! IN_BYTES), 0, __FILE__,__LINE__) + +#define isIDFIRST_lazy_if_safe(p, e, UTF) \ + ((IN_BYTES || !UTF) \ + ? isIDFIRST(*(p)) \ + : isIDFIRST_utf8_safe(p, e)) + +#define isWORDCHAR_lazy_if(p,UTF) \ + _is_utf8_FOO(_CC_IDFIRST, (const U8 *) p, "isWORDCHAR_lazy_if", \ + "isWORDCHAR_lazy_if_safe", \ + cBOOL(UTF && ! IN_BYTES), 0, __FILE__,__LINE__) + +#define isWORDCHAR_lazy_if_safe(p, e, UTF) \ + ((IN_BYTES || !UTF) \ + ? isWORDCHAR(*(p)) \ + : isWORDCHAR_utf8_safe((U8 *) p, (U8 *) e)) + +#define isALNUM_lazy_if(p,UTF) \ + _is_utf8_FOO(_CC_IDFIRST, (const U8 *) p, "isALNUM_lazy_if", \ + "isWORDCHAR_lazy_if_safe", \ + cBOOL(UTF && ! IN_BYTES), 0, __FILE__,__LINE__) #define UTF8_MAXLEN UTF8_MAXBYTES /* A Unicode character can fold to up to 3 characters */ #define UTF8_MAX_FOLD_CHAR_EXPAND 3 -#define IN_BYTES (CopHINTS_get(PL_curcop) & HINT_BYTES) +#define IN_BYTES UNLIKELY(CopHINTS_get(PL_curcop) & HINT_BYTES) /* @@ -492,45 +730,76 @@ case any call to string overloading updates the internal UTF-8 encoding flag. * Is so within 'feature unicode_strings' or 'locale :not_characters', and not * within 'use bytes'. UTF-8 locales are not tested for here, but perhaps * could be */ -#define IN_UNI_8_BIT \ - (((CopHINTS_get(PL_curcop) & (HINT_UNI_8_BIT)) \ - || (CopHINTS_get(PL_curcop) & HINT_LOCALE_PARTIAL \ - /* -1 below is for :not_characters */ \ - && _is_in_locale_category(FALSE, -1))) \ - && ! IN_BYTES) +#define IN_UNI_8_BIT \ + (( ( (CopHINTS_get(PL_curcop) & HINT_UNI_8_BIT)) \ + || ( CopHINTS_get(PL_curcop) & HINT_LOCALE_PARTIAL \ + /* -1 below is for :not_characters */ \ + && _is_in_locale_category(FALSE, -1))) \ + && (! IN_BYTES)) #define UTF8_ALLOW_EMPTY 0x0001 /* Allow a zero length string */ +#define UTF8_GOT_EMPTY UTF8_ALLOW_EMPTY /* Allow first byte to be a continuation byte */ #define UTF8_ALLOW_CONTINUATION 0x0002 +#define UTF8_GOT_CONTINUATION UTF8_ALLOW_CONTINUATION -/* Allow second... bytes to be non-continuation bytes */ +/* Unexpected continuation byte */ #define UTF8_ALLOW_NON_CONTINUATION 0x0004 +#define UTF8_GOT_NON_CONTINUATION UTF8_ALLOW_NON_CONTINUATION /* expecting more bytes than were available in the string */ #define UTF8_ALLOW_SHORT 0x0008 +#define UTF8_GOT_SHORT UTF8_ALLOW_SHORT -/* Overlong sequence; i.e., the code point can be specified in fewer bytes. */ +/* Overlong sequence; i.e., the code point can be specified in fewer bytes. + * First one will convert the overlong to the REPLACEMENT CHARACTER; second + * will return what the overlong evaluates to */ #define UTF8_ALLOW_LONG 0x0010 - -#define UTF8_DISALLOW_SURROGATE 0x0020 /* Unicode surrogates */ -#define UTF8_WARN_SURROGATE 0x0040 - -#define UTF8_DISALLOW_NONCHAR 0x0080 /* Unicode non-character */ -#define UTF8_WARN_NONCHAR 0x0100 /* code points */ - -#define UTF8_DISALLOW_SUPER 0x0200 /* Super-set of Unicode: code */ -#define UTF8_WARN_SUPER 0x0400 /* points above the legal max */ - -/* Code points which never were part of the original UTF-8 standard, the first - * byte of which is a FE or FF on ASCII platforms. If the first byte is FF, it - * will overflow a 32-bit word. If the first byte is FE, it will overflow a - * signed 32-bit word. */ -#define UTF8_DISALLOW_FE_FF 0x0800 -#define UTF8_WARN_FE_FF 0x1000 - -#define UTF8_CHECK_ONLY 0x2000 +#define UTF8_ALLOW_LONG_AND_ITS_VALUE (UTF8_ALLOW_LONG|0x0020) +#define UTF8_GOT_LONG UTF8_ALLOW_LONG + +#define UTF8_ALLOW_OVERFLOW 0x0080 +#define UTF8_GOT_OVERFLOW UTF8_ALLOW_OVERFLOW + +#define UTF8_DISALLOW_SURROGATE 0x0100 /* Unicode surrogates */ +#define UTF8_GOT_SURROGATE UTF8_DISALLOW_SURROGATE +#define UTF8_WARN_SURROGATE 0x0200 + +/* Unicode non-character code points */ +#define UTF8_DISALLOW_NONCHAR 0x0400 +#define UTF8_GOT_NONCHAR UTF8_DISALLOW_NONCHAR +#define UTF8_WARN_NONCHAR 0x0800 + +/* Super-set of Unicode: code points above the legal max */ +#define UTF8_DISALLOW_SUPER 0x1000 +#define UTF8_GOT_SUPER UTF8_DISALLOW_SUPER +#define UTF8_WARN_SUPER 0x2000 + +/* The original UTF-8 standard did not define UTF-8 with start bytes of 0xFE or + * 0xFF, though UTF-EBCDIC did. This allowed both versions to represent code + * points up to 2 ** 31 - 1. Perl extends UTF-8 so that 0xFE and 0xFF are + * usable on ASCII platforms, and 0xFF means something different than + * UTF-EBCDIC defines. These changes allow code points of 64 bits (actually + * somewhat more) to be represented on both platforms. But these are Perl + * extensions, and not likely to be interchangeable with other languages. Note + * that on ASCII platforms, FE overflows a signed 32-bit word, and FF an + * unsigned one. */ +#define UTF8_DISALLOW_PERL_EXTENDED 0x4000 +#define UTF8_GOT_PERL_EXTENDED UTF8_DISALLOW_PERL_EXTENDED +#define UTF8_WARN_PERL_EXTENDED 0x8000 + +/* For back compat, these old names are misleading for overlongs and + * UTF_EBCDIC. */ +#define UTF8_DISALLOW_ABOVE_31_BIT UTF8_DISALLOW_PERL_EXTENDED +#define UTF8_GOT_ABOVE_31_BIT UTF8_GOT_PERL_EXTENDED +#define UTF8_WARN_ABOVE_31_BIT UTF8_WARN_PERL_EXTENDED +#define UTF8_DISALLOW_FE_FF UTF8_DISALLOW_PERL_EXTENDED +#define UTF8_WARN_FE_FF UTF8_WARN_PERL_EXTENDED + +#define UTF8_CHECK_ONLY 0x10000 +#define _UTF8_NO_CONFIDENCE_IN_CURLEN 0x20000 /* Internal core use only */ /* For backwards source compatibility. They do nothing, as the default now * includes what they used to mean. The first one's meaning was to allow the @@ -538,54 +807,107 @@ case any call to string overloading updates the internal UTF-8 encoding flag. #define UTF8_ALLOW_FFFF 0 #define UTF8_ALLOW_SURROGATE 0 -#define UTF8_DISALLOW_ILLEGAL_INTERCHANGE (UTF8_DISALLOW_SUPER|UTF8_DISALLOW_NONCHAR|UTF8_DISALLOW_SURROGATE|UTF8_DISALLOW_FE_FF) +/* C9 refers to Unicode Corrigendum #9: allows but discourages non-chars */ +#define UTF8_DISALLOW_ILLEGAL_C9_INTERCHANGE \ + (UTF8_DISALLOW_SUPER|UTF8_DISALLOW_SURROGATE) +#define UTF8_WARN_ILLEGAL_C9_INTERCHANGE (UTF8_WARN_SUPER|UTF8_WARN_SURROGATE) + +#define UTF8_DISALLOW_ILLEGAL_INTERCHANGE \ + (UTF8_DISALLOW_ILLEGAL_C9_INTERCHANGE|UTF8_DISALLOW_NONCHAR) #define UTF8_WARN_ILLEGAL_INTERCHANGE \ - (UTF8_WARN_SUPER|UTF8_WARN_NONCHAR|UTF8_WARN_SURROGATE|UTF8_WARN_FE_FF) -#define UTF8_ALLOW_ANY \ - (~(UTF8_DISALLOW_ILLEGAL_INTERCHANGE|UTF8_WARN_ILLEGAL_INTERCHANGE)) -#define UTF8_ALLOW_ANYUV \ - (UTF8_ALLOW_EMPTY \ - & ~(UTF8_DISALLOW_ILLEGAL_INTERCHANGE|UTF8_WARN_ILLEGAL_INTERCHANGE)) -#define UTF8_ALLOW_DEFAULT (ckWARN(WARN_UTF8) ? 0 : \ - UTF8_ALLOW_ANYUV) + (UTF8_WARN_ILLEGAL_C9_INTERCHANGE|UTF8_WARN_NONCHAR) + +/* This is typically used for code that processes UTF-8 input and doesn't want + * to have to deal with any malformations that might be present. All such will + * be safely replaced by the REPLACEMENT CHARACTER, unless other flags + * overriding this are also present. */ +#define UTF8_ALLOW_ANY ( UTF8_ALLOW_CONTINUATION \ + |UTF8_ALLOW_NON_CONTINUATION \ + |UTF8_ALLOW_SHORT \ + |UTF8_ALLOW_LONG \ + |UTF8_ALLOW_OVERFLOW) + +/* Accept any Perl-extended UTF-8 that evaluates to any UV on the platform, but + * not any malformed. This is the default. (Note that UVs above IV_MAX are + * deprecated. */ +#define UTF8_ALLOW_ANYUV 0 +#define UTF8_ALLOW_DEFAULT UTF8_ALLOW_ANYUV -/* Surrogates, non-character code points and above-Unicode code points are - * problematic in some contexts. This allows code that needs to check for - * those to to quickly exclude the vast majority of code points it will - * encounter */ -#ifdef EBCDIC -# define isUTF8_POSSIBLY_PROBLEMATIC(c) (NATIVE_UTF8_TO_I8(c) >= 0xF1) -#else -# define isUTF8_POSSIBLY_PROBLEMATIC(c) ((U8) c >= 0xED) -#endif +/* +=for apidoc Am|bool|UTF8_IS_SURROGATE|const U8 *s|const U8 *e -/* Several of the macros below have a second parameter that is currently - * unused; but could be used in the future to make sure that the input is - * well-formed. */ +Evaluates to non-zero if the first few bytes of the string starting at C and +looking no further than S> are well-formed UTF-8 that represents one +of the Unicode surrogate code points; otherwise it evaluates to 0. If +non-zero, the value gives how many bytes starting at C comprise the code +point's representation. + +=cut + */ +#define UTF8_IS_SURROGATE(s, e) is_SURROGATE_utf8_safe(s, e) -#define UTF8_IS_SURROGATE(s, e) cBOOL(is_SURROGATE_utf8(s)) -#define UTF8_IS_REPLACEMENT(s, send) cBOOL(is_REPLACEMENT_utf8_safe(s,send)) -/* ASCII EBCDIC I8 +#define UTF8_IS_REPLACEMENT(s, send) is_REPLACEMENT_utf8_safe(s,send) + +/* +=for apidoc Am|bool|UTF8_IS_SUPER|const U8 *s|const U8 *e + +Recall that Perl recognizes an extension to UTF-8 that can encode code +points larger than the ones defined by Unicode, which are 0..0x10FFFF. + +This macro evaluates to non-zero if the first few bytes of the string starting +at C and looking no further than S> are from this UTF-8 extension; +otherwise it evaluates to 0. If non-zero, the value gives how many bytes +starting at C comprise the code point's representation. + +0 is returned if the bytes are not well-formed extended UTF-8, or if they +represent a code point that cannot fit in a UV on the current platform. Hence +this macro can give different results when run on a 64-bit word machine than on +one with a 32-bit word size. + +Note that it is deprecated to have code points that are larger than what can +fit in an IV on the current machine. + +=cut + + * ASCII EBCDIC I8 * U+10FFFF: \xF4\x8F\xBF\xBF \xF9\xA1\xBF\xBF\xBF max legal Unicode * U+110000: \xF4\x90\x80\x80 \xF9\xA2\xA0\xA0\xA0 * U+110001: \xF4\x90\x80\x81 \xF9\xA2\xA0\xA0\xA1 - * - * BE AWARE that this test doesn't rule out malformed code points, in - * particular overlongs */ -#ifdef EBCDIC /* Both versions assume well-formed UTF8 */ -# define UTF8_IS_SUPER(s, e) (NATIVE_UTF8_TO_I8(* (U8*) (s)) >= 0xF9 \ - && (NATIVE_UTF8_TO_I8(* (U8*) (s)) > 0xF9 \ - || (NATIVE_UTF8_TO_I8(* ((U8*) (s) + 1)) >= 0xA2))) + */ +#ifdef EBCDIC +# define UTF8_IS_SUPER(s, e) \ + (( LIKELY((e) > (s) + 4) \ + && NATIVE_UTF8_TO_I8(*(s)) >= 0xF9 \ + && ( NATIVE_UTF8_TO_I8(*(s)) > 0xF9 \ + || (NATIVE_UTF8_TO_I8(*((s) + 1)) >= 0xA2)) \ + && LIKELY((s) + UTF8SKIP(s) <= (e))) \ + ? _is_utf8_char_helper(s, s + UTF8SKIP(s), 0) : 0) #else -# define UTF8_IS_SUPER(s, e) (*(U8*) (s) >= 0xF4 \ - && (*(U8*) (s) > 0xF4 || (*((U8*) (s) + 1) >= 0x90))) +# define UTF8_IS_SUPER(s, e) \ + (( LIKELY((e) > (s) + 3) \ + && (*(U8*) (s)) >= 0xF4 \ + && ((*(U8*) (s)) > 0xF4 || (*((U8*) (s) + 1) >= 0x90))\ + && LIKELY((s) + UTF8SKIP(s) <= (e))) \ + ? _is_utf8_char_helper(s, s + UTF8SKIP(s), 0) : 0) #endif /* These are now machine generated, and the 'given' clause is no longer * applicable */ #define UTF8_IS_NONCHAR_GIVEN_THAT_NON_SUPER_AND_GE_PROBLEMATIC(s, e) \ - cBOOL(is_NONCHAR_utf8(s)) + cBOOL(is_NONCHAR_utf8_safe(s,e)) + +/* +=for apidoc Am|bool|UTF8_IS_NONCHAR|const U8 *s|const U8 *e + +Evaluates to non-zero if the first few bytes of the string starting at C and +looking no further than S> are well-formed UTF-8 that represents one +of the Unicode non-character code points; otherwise it evaluates to 0. If +non-zero, the value gives how many bytes starting at C comprise the code +point's representation. + +=cut + */ #define UTF8_IS_NONCHAR(s, e) \ UTF8_IS_NONCHAR_GIVEN_THAT_NON_SUPER_AND_GE_PROBLEMATIC(s, e) @@ -598,37 +920,54 @@ case any call to string overloading updates the internal UTF-8 encoding flag. * let's be conservative and do as Unicode says. */ #define PERL_UNICODE_MAX 0x10FFFF -#define UNICODE_WARN_SURROGATE 0x0001 /* UTF-16 surrogates */ -#define UNICODE_WARN_NONCHAR 0x0002 /* Non-char code points */ -#define UNICODE_WARN_SUPER 0x0004 /* Above 0x10FFFF */ -#define UNICODE_WARN_FE_FF 0x0008 /* Above 0x10FFFF */ -#define UNICODE_DISALLOW_SURROGATE 0x0010 -#define UNICODE_DISALLOW_NONCHAR 0x0020 -#define UNICODE_DISALLOW_SUPER 0x0040 -#define UNICODE_DISALLOW_FE_FF 0x0080 +#define UNICODE_WARN_SURROGATE 0x0001 /* UTF-16 surrogates */ +#define UNICODE_WARN_NONCHAR 0x0002 /* Non-char code points */ +#define UNICODE_WARN_SUPER 0x0004 /* Above 0x10FFFF */ +#define UNICODE_WARN_PERL_EXTENDED 0x0008 /* Above 0x7FFF_FFFF */ +#define UNICODE_WARN_ABOVE_31_BIT UNICODE_WARN_PERL_EXTENDED +#define UNICODE_DISALLOW_SURROGATE 0x0010 +#define UNICODE_DISALLOW_NONCHAR 0x0020 +#define UNICODE_DISALLOW_SUPER 0x0040 +#define UNICODE_DISALLOW_PERL_EXTENDED 0x0080 +#define UNICODE_DISALLOW_ABOVE_31_BIT UNICODE_DISALLOW_PERL_EXTENDED +#define UNICODE_WARN_ILLEGAL_C9_INTERCHANGE \ + (UNICODE_WARN_SURROGATE|UNICODE_WARN_SUPER) #define UNICODE_WARN_ILLEGAL_INTERCHANGE \ - (UNICODE_WARN_SURROGATE|UNICODE_WARN_NONCHAR|UNICODE_WARN_SUPER) + (UNICODE_WARN_ILLEGAL_C9_INTERCHANGE|UNICODE_WARN_NONCHAR) +#define UNICODE_DISALLOW_ILLEGAL_C9_INTERCHANGE \ + (UNICODE_DISALLOW_SURROGATE|UNICODE_DISALLOW_SUPER) #define UNICODE_DISALLOW_ILLEGAL_INTERCHANGE \ - (UNICODE_DISALLOW_SURROGATE|UNICODE_DISALLOW_NONCHAR|UNICODE_DISALLOW_SUPER) + (UNICODE_DISALLOW_ILLEGAL_C9_INTERCHANGE|UNICODE_DISALLOW_NONCHAR) /* For backward source compatibility, as are now the default */ #define UNICODE_ALLOW_SURROGATE 0 #define UNICODE_ALLOW_SUPER 0 #define UNICODE_ALLOW_ANY 0 -#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_NONCHAR(c) ((c >= 0xFDD0 && c <= 0xFDEF) \ - /* The other noncharacters end in FFFE or FFFF, which \ - * the mask below catches both of, but beyond the last \ - * official unicode code point, they aren't \ - * noncharacters, since those aren't Unicode \ - * characters at all */ \ - || ((((c & 0xFFFE) == 0xFFFE)) && ! UNICODE_IS_SUPER(c))) -#define UNICODE_IS_SUPER(c) ((c) > PERL_UNICODE_MAX) -#define UNICODE_IS_FE_FF(c) ((c) > 0x7FFFFFFF) +/* This matches the 2048 code points between UNICODE_SURROGATE_FIRST (0xD800) and + * UNICODE_SURROGATE_LAST (0xDFFF) */ +#define UNICODE_IS_SURROGATE(uv) (((UV) (uv) & (~0xFFFF | 0xF800)) \ + == 0xD800) + +#define UNICODE_IS_REPLACEMENT(uv) ((UV) (uv) == UNICODE_REPLACEMENT) +#define UNICODE_IS_BYTE_ORDER_MARK(uv) ((UV) (uv) == UNICODE_BYTE_ORDER_MARK) + +/* Is 'uv' one of the 32 contiguous-range noncharacters? */ +#define UNICODE_IS_32_CONTIGUOUS_NONCHARS(uv) ((UV) (uv) >= 0xFDD0 \ + && (UV) (uv) <= 0xFDEF) + +/* Is 'uv' one of the 34 plane-ending noncharacters 0xFFFE, 0xFFFF, 0x1FFFE, + * 0x1FFFF, ... 0x10FFFE, 0x10FFFF, given that we know that 'uv' is not above + * the Unicode legal max */ +#define UNICODE_IS_END_PLANE_NONCHAR_GIVEN_NOT_SUPER(uv) \ + (((UV) (uv) & 0xFFFE) == 0xFFFE) + +#define UNICODE_IS_NONCHAR(uv) \ + ( UNICODE_IS_32_CONTIGUOUS_NONCHARS(uv) \ + || ( LIKELY( ! UNICODE_IS_SUPER(uv)) \ + && UNICODE_IS_END_PLANE_NONCHAR_GIVEN_NOT_SUPER(uv))) + +#define UNICODE_IS_SUPER(uv) ((UV) (uv) > PERL_UNICODE_MAX) #define LATIN_SMALL_LETTER_SHARP_S LATIN_SMALL_LETTER_SHARP_S_NATIVE #define LATIN_SMALL_LETTER_Y_WITH_DIAERESIS \ @@ -670,81 +1009,176 @@ case any call to string overloading updates the internal UTF-8 encoding flag. #define SHARP_S_SKIP 2 -/* If you want to exclude surrogates, and beyond legal Unicode, see the blame - * log for earlier versions which gave details for these */ +/* -/* A helper macro for isUTF8_CHAR, so use that one, and not this one. This is - * retained solely for backwards compatibility and may be deprecated and - * removed in a future Perl version. - * - * regen/regcharclass.pl generates is_UTF8_CHAR_utf8() macros for up to these - * number of bytes. So this has to be coordinated with that file */ -#ifdef EBCDIC -# define IS_UTF8_CHAR_FAST(n) ((n) <= 3) -#else -# define IS_UTF8_CHAR_FAST(n) ((n) <= 4) -#endif +=for apidoc Am|STRLEN|isUTF8_CHAR|const U8 *s|const U8 *e + +Evaluates to non-zero if the first few bytes of the string starting at C and +looking no further than S> are well-formed UTF-8, as extended by Perl, +that represents some code point; otherwise it evaluates to 0. If non-zero, the +value gives how many bytes starting at C comprise the code point's +representation. Any bytes remaining before C, but beyond the ones needed to +form the first code point in C, are not examined. + +The code point can be any that will fit in a UV on this machine, using Perl's +extension to official UTF-8 to represent those higher than the Unicode maximum +of 0x10FFFF. That means that this macro is used to efficiently decide if the +next few bytes in C is legal UTF-8 for a single character. + +Use C> to restrict the acceptable code points to those +defined by Unicode to be fully interchangeable across applications; +C> to use the L definition of allowable +code points; and C> for a more customized definition. + +Use C>, C>, and +C> to check entire strings. + +Note that it is deprecated to use code points higher than what will fit in an +IV. This macro does not raise any warnings for such code points, treating them +as valid. + +Note also that a UTF-8 INVARIANT character (i.e. ASCII on non-EBCDIC machines) +is a valid UTF-8 character. + +=cut +*/ + +#define isUTF8_CHAR(s, e) \ + (UNLIKELY((e) <= (s)) \ + ? 0 \ + : (UTF8_IS_INVARIANT(*s)) \ + ? 1 \ + : UNLIKELY(((e) - (s)) < UTF8SKIP(s)) \ + ? 0 \ + : LIKELY(NATIVE_UTF8_TO_I8(*s) <= _IS_UTF8_CHAR_HIGHEST_START_BYTE) \ + ? is_UTF8_CHAR_utf8_no_length_checks(s) \ + : _is_utf8_char_helper(s, e, 0)) + +#define is_utf8_char_buf(buf, buf_end) isUTF8_CHAR(buf, buf_end) +#define bytes_from_utf8(s, lenp, is_utf8p) \ + bytes_from_utf8_loc(s, lenp, is_utf8p, 0) -#ifndef EBCDIC -/* A helper macro for isUTF8_CHAR, so use that one instead of this. This was - * generated by regen/regcharclass.pl, and then moved here. The lines that - * generated it were then commented out. This was done solely because it takes - * on the order of 10 minutes to generate, and is never going to change, unless - * the generated code is improved. - * - * The EBCDIC versions have been cut to not cover all of legal Unicode, - * otherwise they take too long to generate; besides there is a separate one - * for each code page, so they are in regcharclass.h instead of here */ /* - UTF8_CHAR: Matches legal UTF-8 encoded characters from 2 through 4 bytes - 0x80 - 0x1FFFFF +=for apidoc Am|STRLEN|isSTRICT_UTF8_CHAR|const U8 *s|const U8 *e + +Evaluates to non-zero if the first few bytes of the string starting at C and +looking no further than S> are well-formed UTF-8 that represents some +Unicode code point completely acceptable for open interchange between all +applications; otherwise it evaluates to 0. If non-zero, the value gives how +many bytes starting at C comprise the code point's representation. Any +bytes remaining before C, but beyond the ones needed to form the first code +point in C, are not examined. + +The largest acceptable code point is the Unicode maximum 0x10FFFF, and must not +be a surrogate nor a non-character code point. Thus this excludes any code +point from Perl's extended UTF-8. + +This is used to efficiently decide if the next few bytes in C is +legal Unicode-acceptable UTF-8 for a single character. + +Use C> to use the L definition of allowable +code points; C> to check for Perl's extended UTF-8; +and C> for a more customized definition. + +Use C>, C>, and +C> to check entire strings. + +=cut */ -/*** GENERATED CODE ***/ -#define is_UTF8_CHAR_utf8_no_length_checks(s) \ -( ( 0xC2 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xDF ) ? \ - ( ( ( ((U8*)s)[1] & 0xC0 ) == 0x80 ) ? 2 : 0 ) \ -: ( 0xE0 == ((U8*)s)[0] ) ? \ - ( ( ( ( ((U8*)s)[1] & 0xE0 ) == 0xA0 ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ -: ( 0xE1 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xEF ) ? \ - ( ( ( ( ((U8*)s)[1] & 0xC0 ) == 0x80 ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) ? 3 : 0 )\ -: ( 0xF0 == ((U8*)s)[0] ) ? \ - ( ( ( ( 0x90 <= ((U8*)s)[1] && ((U8*)s)[1] <= 0xBF ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) && ( ( ((U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 )\ -: ( ( ( ( 0xF1 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xF7 ) && ( ( ((U8*)s)[1] & 0xC0 ) == 0x80 ) ) && ( ( ((U8*)s)[2] & 0xC0 ) == 0x80 ) ) && ( ( ((U8*)s)[3] & 0xC0 ) == 0x80 ) ) ? 4 : 0 ) -#endif + +#define isSTRICT_UTF8_CHAR(s, e) \ + (UNLIKELY((e) <= (s)) \ + ? 0 \ + : (UTF8_IS_INVARIANT(*s)) \ + ? 1 \ + : UNLIKELY(((e) - (s)) < UTF8SKIP(s)) \ + ? 0 \ + : is_STRICT_UTF8_CHAR_utf8_no_length_checks(s)) /* -=for apidoc Am|STRLEN|isUTF8_CHAR|const U8 *s|const U8 *e +=for apidoc Am|STRLEN|isC9_STRICT_UTF8_CHAR|const U8 *s|const U8 *e + +Evaluates to non-zero if the first few bytes of the string starting at C and +looking no further than S> are well-formed UTF-8 that represents some +Unicode non-surrogate code point; otherwise it evaluates to 0. If non-zero, +the value gives how many bytes starting at C comprise the code point's +representation. Any bytes remaining before C, but beyond the ones needed to +form the first code point in C, are not examined. + +The largest acceptable code point is the Unicode maximum 0x10FFFF. This +differs from C> only in that it accepts non-character +code points. This corresponds to +L. +which said that non-character code points are merely discouraged rather than +completely forbidden in open interchange. See +L. -Returns the number of bytes beginning at C which form a legal UTF-8 (or -UTF-EBCDIC) encoded character, looking no further than S> bytes into -C. Returns 0 if the sequence starting at C through S> is not -well-formed UTF-8. +Use C> to check for Perl's extended UTF-8; and +C> for a more customized definition. -Note that an INVARIANT character (i.e. ASCII on non-EBCDIC -machines) is a valid UTF-8 character. +Use C>, C>, and +C> to check entire strings. =cut */ -#define isUTF8_CHAR(s, e) (UNLIKELY((e) <= (s)) \ - ? 0 \ - : (UTF8_IS_INVARIANT(*s)) \ - ? 1 \ - : UNLIKELY(((e) - (s)) < UTF8SKIP(s)) \ - ? 0 \ - : LIKELY(IS_UTF8_CHAR_FAST(UTF8SKIP(s))) \ - ? is_UTF8_CHAR_utf8_no_length_checks(s) \ - : _is_utf8_char_slow(s, e)) +#define isC9_STRICT_UTF8_CHAR(s, e) \ + (UNLIKELY((e) <= (s)) \ + ? 0 \ + : (UTF8_IS_INVARIANT(*s)) \ + ? 1 \ + : UNLIKELY(((e) - (s)) < UTF8SKIP(s)) \ + ? 0 \ + : is_C9_STRICT_UTF8_CHAR_utf8_no_length_checks(s)) -#define is_utf8_char_buf(buf, buf_end) isUTF8_CHAR(buf, buf_end) +/* + +=for apidoc Am|STRLEN|isUTF8_CHAR_flags|const U8 *s|const U8 *e| const U32 flags + +Evaluates to non-zero if the first few bytes of the string starting at C and +looking no further than S> are well-formed UTF-8, as extended by Perl, +that represents some code point, subject to the restrictions given by C; +otherwise it evaluates to 0. If non-zero, the value gives how many bytes +starting at C comprise the code point's representation. Any bytes remaining +before C, but beyond the ones needed to form the first code point in C, +are not examined. + +If C is 0, this gives the same results as C>; +if C is C, this gives the same results +as C>; +and if C is C, this gives +the same results as C>. +Otherwise C may be any combination of the C> flags +understood by C>, with the same meanings. + +The three alternative macros are for the most commonly needed validations; they +are likely to run somewhat faster than this more general one, as they can be +inlined into your code. + +Use L, L, and +L to check entire strings. + +=cut +*/ + +#define isUTF8_CHAR_flags(s, e, flags) \ + (UNLIKELY((e) <= (s)) \ + ? 0 \ + : (UTF8_IS_INVARIANT(*s)) \ + ? 1 \ + : UNLIKELY(((e) - (s)) < UTF8SKIP(s)) \ + ? 0 \ + : _is_utf8_char_helper(s, e, flags)) /* Do not use; should be deprecated. Use isUTF8_CHAR() instead; this is * retained solely for backwards compatibility */ #define IS_UTF8_CHAR(p, n) (isUTF8_CHAR(p, (p) + (n)) == n) -#endif /* H_UTF8 */ +#endif /* PERL_UTF8_H_ */ /* * ex: set ts=8 sts=4 sw=4 et: