X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/f180b2926a9378db829862d88921feefe2460d35..e17f68231dd7d8e7771a792525d5dbb108b9a2dc:/utf8.h?ds=sidebyside diff --git a/utf8.h b/utf8.h index 3da10e0..e6c7864 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 @@ -34,11 +34,6 @@ #define FOLD_FLAGS_FULL 0x2 #define FOLD_FLAGS_NOMIX_ASCII 0x4 -/* For _core_swash_init(), internal core use only */ -#define _CORE_SWASH_INIT_USER_DEFINED_PROPERTY 0x1 -#define _CORE_SWASH_INIT_RETURN_IF_UNDEF 0x2 -#define _CORE_SWASH_INIT_ACCEPT_INVLIST 0x4 - /* =head1 Unicode Support L has an introduction to this API. @@ -66,20 +61,30 @@ the string is invariant. #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 uvoffuni_to_utf8_flags(d,uv,flags) \ + uvoffuni_to_utf8_flags_msgs(d, uv, flags, 0) #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) + uvchr_to_utf8_flags_msgs(d,uv,flags, 0) +#define uvchr_to_utf8_flags_msgs(d,uv,flags,msgs) \ + uvoffuni_to_utf8_flags_msgs(d,NATIVE_TO_UNI(uv),flags, msgs) #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) + utf8_to_uvchr_buf_helper((const U8 *) (s), (const U8 *) e, lenp) #define utf8n_to_uvchr(s, len, lenp, flags) \ utf8n_to_uvchr_error(s, len, lenp, flags, 0) +#define utf8n_to_uvchr_error(s, len, lenp, flags, errors) \ + utf8n_to_uvchr_msgs(s, len, lenp, flags, errors, 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) + +#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) @@ -103,11 +108,19 @@ the string is invariant. #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 */ +/* + +=for apidoc AmnU|STRLEN|UTF8_MAXBYTES + +The maximum width of a single UTF-8 encoded character, 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. + +=cut + */ #define UTF8_MAXBYTES 13 #ifdef DOINIT @@ -145,8 +158,51 @@ END_EXTERN_C #define PERL_SMALL_MACRO_BUFFER #endif -/* Native character to/from iso-8859-1. Are the identity functions on ASCII - * platforms */ +/* + +=for apidoc Am|U8|NATIVE_TO_LATIN1|U8 ch + +Returns the Latin-1 (including ASCII and control characters) equivalent of the +input native code point given by C. Thus, C on +EBCDIC platforms returns 65. These each represent the character C<"A"> on +their respective platforms. On ASCII platforms no conversion is needed, so +this macro expands to just its input, adding no time nor space requirements to +the implementation. + +For conversion of code points potentially larger than will fit in a character, +use L. + +=for apidoc Am|U8|LATIN1_TO_NATIVE|U8 ch + +Returns the native equivalent of the input Latin-1 code point (including ASCII +and control characters) given by C. Thus, C on +EBCDIC platforms returns 194. These each represent the character C<"B"> on +their respective platforms. On ASCII platforms no conversion is needed, so +this macro expands to just its input, adding no time nor space requirements to +the implementation. + +For conversion of code points potentially larger than will fit in a character, +use L. + +=for apidoc Am|UV|NATIVE_TO_UNI|UV ch + +Returns the Unicode equivalent of the input native code point given by C. +Thus, C on EBCDIC platforms returns 67. These each +represent the character C<"C"> on their respective platforms. On ASCII +platforms no conversion is needed, so this macro expands to just its input, +adding no time nor space requirements to the implementation. + +=for apidoc Am|UV|UNI_TO_NATIVE|UV ch + +Returns the native equivalent of the input Unicode code point given by C. +Thus, C on EBCDIC platforms returns 196. These each +represent the character C<"D"> on their respective platforms. On ASCII +platforms no conversion is needed, so this macro expands to just its input, +adding no time nor space requirements to the implementation. + +=cut +*/ + #ifdef PERL_SMALL_MACRO_BUFFER #define NATIVE_TO_LATIN1(ch) ((U8)(ch)) #define LATIN1_TO_NATIVE(ch) ((U8)(ch)) @@ -169,7 +225,6 @@ END_EXTERN_C #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) ((UV) (ch)) #define NATIVE_TO_UNI(ch) ((UV) (ch)) @@ -228,15 +283,9 @@ 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. +=cut */ -/* Anything larger than this will overflow the word if it were converted into a UV */ -#if defined(UV_IS_QUAD) -# define HIGHEST_REPRESENTABLE_UTF8 "\xFF\x80\x8F\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF" -#else -# define HIGHEST_REPRESENTABLE_UTF8 "\xFE\x83\xBF\xBF\xBF\xBF\xBF" -#endif - /* Is the representation of the Unicode code point 'cp' the same regardless of * being encoded in UTF-8 or not? */ #define OFFUNI_IS_INVARIANT(cp) isASCII(cp) @@ -261,13 +310,15 @@ C is Unicode if above 255; otherwise is platform-native. /* 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) (((U8)((c) | 0)) & UTF_CONTINUATION_MARK) +#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. The |0 makes sure * this isn't mistakenly called with a ptr argument */ -#define UTF8_IS_START(c) (((U8)((c) | 0)) >= 0xc2) +#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 @@ -275,20 +326,22 @@ C is Unicode if above 255; otherwise is platform-native. /* Is the byte 'c' part of a multi-byte UTF8-8 encoded sequence, and not the * first byte thereof? The |0 makes sure this isn't mistakenly called with a * ptr argument */ -#define UTF8_IS_CONTINUATION(c) \ - ((((U8)((c) | 0)) & UTF_IS_CONTINUATION_MASK) == UTF_CONTINUATION_MARK) +#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]. The |0 makes sure this isn't mistakenly called with a * ptr argument */ -#define UTF8_IS_DOWNGRADEABLE_START(c) ((((U8)((c) | 0)) & 0xfe) == 0xc2) +#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? The |0 makes sure this isn't mistakenly * called with a ptr argument */ -#define UTF8_IS_ABOVE_LATIN1(c) (((U8)((c) | 0)) >= 0xc4) +#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 @@ -304,105 +357,10 @@ C is Unicode if above 255; otherwise is platform-native. * 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) ((U8) c >= 0xED) - -/* 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 +#define isUTF8_POSSIBLY_PROBLEMATIC(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + (U8) c >= 0xED) - 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) #endif /* EBCDIC vs ASCII */ @@ -524,14 +482,16 @@ encoded as UTF-8. C is a native (ASCII or EBCDIC) code point if less than * 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) @@ -566,8 +526,10 @@ 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_(PL_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)))) @@ -585,6 +547,20 @@ only) byte is pointed to by C. #define UTF8SKIP(s) PL_utf8skip[*(const U8*)(s)] #define UTF8_SKIP(s) UTF8SKIP(s) +/* + +=for apidoc Am|STRLEN|UTF8_SAFE_SKIP|char* s|char* e +returns 0 if S= e>>; otherwise returns the number of bytes in the +UTF-8 encoded character whose first byte is pointed to by C. But it never +returns beyond C. On DEBUGGING builds, it asserts that S= e>>. + +=cut + */ +#define UTF8_SAFE_SKIP(s, e) (__ASSERT_((e) >= (s)) \ + ((e) - (s)) <= 0 \ + ? 0 \ + : MIN(((e) - (s)), UTF8_SKIP(s))) + /* 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) @@ -672,31 +648,37 @@ with a ptr argument. * 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) /* @@ -715,12 +697,12 @@ 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 */ @@ -730,7 +712,7 @@ case any call to string overloading updates the internal UTF-8 encoding flag. #define UTF8_ALLOW_CONTINUATION 0x0002 #define UTF8_GOT_CONTINUATION UTF8_ALLOW_CONTINUATION -/* Unexpected continuation byte */ +/* Unexpected non-continuation byte */ #define UTF8_ALLOW_NON_CONTINUATION 0x0004 #define UTF8_GOT_NON_CONTINUATION UTF8_ALLOW_NON_CONTINUATION @@ -738,43 +720,59 @@ case any call to string overloading updates the internal UTF-8 encoding flag. #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_ALLOW_LONG_AND_ITS_VALUE (UTF8_ALLOW_LONG|0x0020) #define UTF8_GOT_LONG UTF8_ALLOW_LONG -/* Currently no way to allow overflow */ -#define UTF8_GOT_OVERFLOW 0x0080 +#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 -#define UTF8_DISALLOW_NONCHAR 0x0400 /* Unicode non-character */ +/* Unicode non-character code points */ +#define UTF8_DISALLOW_NONCHAR 0x0400 #define UTF8_GOT_NONCHAR UTF8_DISALLOW_NONCHAR -#define UTF8_WARN_NONCHAR 0x0800 /* code points */ +#define UTF8_WARN_NONCHAR 0x0800 -#define UTF8_DISALLOW_SUPER 0x1000 /* Super-set of Unicode: code */ +/* 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 /* points above the legal max */ - -/* Code points which never were part of the original UTF-8 standard, which only - * went up to 2 ** 31 - 1. Note that these all overflow a signed 32-bit word, - * The first byte of these code points is FE or FF on ASCII platforms. If the - * first byte is FF, it will overflow a 32-bit word. */ -#define UTF8_DISALLOW_ABOVE_31_BIT 0x4000 -#define UTF8_GOT_ABOVE_31_BIT UTF8_DISALLOW_ABOVE_31_BIT -#define UTF8_WARN_ABOVE_31_BIT 0x8000 - -/* For back compat, these old names are misleading for UTF_EBCDIC */ -#define UTF8_DISALLOW_FE_FF UTF8_DISALLOW_ABOVE_31_BIT -#define UTF8_WARN_FE_FF UTF8_WARN_ABOVE_31_BIT +#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 * just the single non-character 0xFFFF */ #define UTF8_ALLOW_FFFF 0 +#define UTF8_ALLOW_FE_FF 0 #define UTF8_ALLOW_SURROGATE 0 /* C9 refers to Unicode Corrigendum #9: allows but discourages non-chars */ @@ -787,12 +785,20 @@ case any call to string overloading updates the internal UTF-8 encoding flag. #define UTF8_WARN_ILLEGAL_INTERCHANGE \ (UTF8_WARN_ILLEGAL_C9_INTERCHANGE|UTF8_WARN_NONCHAR) -#define UTF8_ALLOW_ANY \ - (~( UTF8_DISALLOW_ILLEGAL_INTERCHANGE|UTF8_DISALLOW_ABOVE_31_BIT \ - |UTF8_WARN_ILLEGAL_INTERCHANGE|UTF8_WARN_ABOVE_31_BIT)) -#define UTF8_ALLOW_ANYUV UTF8_ALLOW_EMPTY -#define UTF8_ALLOW_DEFAULT (ckWARN(WARN_UTF8) ? 0 : \ - UTF8_ALLOW_ANYUV) +/* 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. */ +#define UTF8_ALLOW_ANYUV 0 +#define UTF8_ALLOW_DEFAULT UTF8_ALLOW_ANYUV /* =for apidoc Am|bool|UTF8_IS_SURROGATE|const U8 *s|const U8 *e @@ -810,6 +816,8 @@ point's representation. #define UTF8_IS_REPLACEMENT(s, send) is_REPLACEMENT_utf8_safe(s,send) +#define MAX_LEGAL_CP ((UV)IV_MAX) + /* =for apidoc Am|bool|UTF8_IS_SUPER|const U8 *s|const U8 *e @@ -826,7 +834,7 @@ 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 +Note that it is illegal to have code points that are larger than what can fit in an IV on the current machine. =cut @@ -843,14 +851,14 @@ fit in an IV on the current machine. && ( 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) + ? is_utf8_char_helper(s, s + UTF8SKIP(s), 0) : 0) #else # 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) + ? is_utf8_char_helper(s, s + UTF8SKIP(s), 0) : 0) #endif /* These are now machine generated, and the 'given' clause is no longer @@ -867,6 +875,10 @@ 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. +=for apidoc AmnU|UV|UNICODE_REPLACEMENT + +Evaluates to 0xFFFD, the code point of the Unicode REPLACEMENT CHARACTER + =cut */ #define UTF8_IS_NONCHAR(s, e) \ @@ -881,14 +893,22 @@ point's representation. * 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_ABOVE_31_BIT 0x0008 /* Above 0x7FFF_FFFF */ -#define UNICODE_DISALLOW_SURROGATE 0x0010 -#define UNICODE_DISALLOW_NONCHAR 0x0020 -#define UNICODE_DISALLOW_SUPER 0x0040 -#define UNICODE_DISALLOW_ABOVE_31_BIT 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_GOT_SURROGATE UNICODE_DISALLOW_SURROGATE +#define UNICODE_GOT_NONCHAR UNICODE_DISALLOW_NONCHAR +#define UNICODE_GOT_SUPER UNICODE_DISALLOW_SUPER +#define UNICODE_GOT_PERL_EXTENDED UNICODE_DISALLOW_PERL_EXTENDED + #define UNICODE_WARN_ILLEGAL_C9_INTERCHANGE \ (UNICODE_WARN_SURROGATE|UNICODE_WARN_SUPER) #define UNICODE_WARN_ILLEGAL_INTERCHANGE \ @@ -927,7 +947,6 @@ point's representation. && UNICODE_IS_END_PLANE_NONCHAR_GIVEN_NOT_SUPER(uv))) #define UNICODE_IS_SUPER(uv) ((UV) (uv) > PERL_UNICODE_MAX) -#define UNICODE_IS_ABOVE_31_BIT(uv) ((UV) (uv) > 0x7FFFFFFF) #define LATIN_SMALL_LETTER_SHARP_S LATIN_SMALL_LETTER_SHARP_S_NATIVE #define LATIN_SMALL_LETTER_Y_WITH_DIAERESIS \ @@ -969,129 +988,9 @@ point's representation. #define SHARP_S_SKIP 2 -/* - -=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) - -/* - -=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 -*/ - -#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|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. - -Use 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 -*/ - -#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 bytes_from_utf8(s, lenp, is_utf8p) \ + bytes_from_utf8_loc(s, lenp, is_utf8p, 0) /* @@ -1130,13 +1029,13 @@ L to check entire strings. ? 1 \ : UNLIKELY(((e) - (s)) < UTF8SKIP(s)) \ ? 0 \ - : _is_utf8_char_helper(s, e, flags)) + : 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: