diff options
Diffstat (limited to 'Master/tlpkg/tlperl/lib/CORE/regcharclass.h')
-rw-r--r-- | Master/tlpkg/tlperl/lib/CORE/regcharclass.h | 883 |
1 files changed, 640 insertions, 243 deletions
diff --git a/Master/tlpkg/tlperl/lib/CORE/regcharclass.h b/Master/tlpkg/tlperl/lib/CORE/regcharclass.h index 5e233ea730e..e51fe64af40 100644 --- a/Master/tlpkg/tlperl/lib/CORE/regcharclass.h +++ b/Master/tlpkg/tlperl/lib/CORE/regcharclass.h @@ -12,17 +12,15 @@ * Any changes made here will be lost! */ + +#ifndef H_REGCHARCLASS /* Guard against nested #includes */ +#define H_REGCHARCLASS 1 + /* LNBREAK: Line Break: \R "\x0D\x0A" # CRLF - Network (Windows) line ending - 0x0A # LF | LINE FEED - 0x0B # VT | VERTICAL TAB - 0x0C # FF | FORM FEED - 0x0D # CR | CARRIAGE RETURN - 0x85 # NEL | NEXT LINE - 0x2028 # LINE SEPARATOR - 0x2029 # PARAGRAPH SEPARATOR + \p{VertSpace} */ /*** GENERATED CODE ***/ #define is_LNBREAK(s,is_utf8) \ @@ -32,9 +30,7 @@ : ( is_utf8 ) ? \ ( ( 0xC2 == ((U8*)s)[0] ) ? \ ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xE2 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0xA8 == ((U8*)s)[2] || 0xA9 == ((U8*)s)[2] ) ) ? 3 : 0 )\ - : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ : ( 0x85 == ((U8*)s)[0] ) ) /*** GENERATED CODE ***/ @@ -46,9 +42,7 @@ : ( is_utf8 ) ? \ ( ( 0xC2 == ((U8*)s)[0] ) ? \ ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xE2 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0xA8 == ((U8*)s)[2] || 0xA9 == ((U8*)s)[2] ) ) ? 3 : 0 )\ - : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ : ( 0x85 == ((U8*)s)[0] ) ) \ : ((e)-(s) > 1) ? \ ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0C ) ? 1 \ @@ -71,9 +65,7 @@ ( ( 0x0A == ((U8*)s)[1] ) ? 2 : 1 ) \ : ( 0xC2 == ((U8*)s)[0] ) ? \ ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ -: ( 0xE2 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0xA8 == ((U8*)s)[2] || 0xA9 == ((U8*)s)[2] ) ) ? 3 : 0 )\ -: 0 ) +: ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 ) /*** GENERATED CODE ***/ #define is_LNBREAK_utf8_safe(s,e) \ @@ -83,34 +75,30 @@ ( ( 0x0A == ((U8*)s)[1] ) ? 2 : 1 ) \ : ( 0xC2 == ((U8*)s)[0] ) ? \ ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xE2 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0xA8 == ((U8*)s)[2] || 0xA9 == ((U8*)s)[2] ) ) ? 3 : 0 )\ - : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ : ((e)-(s) > 1) ? \ ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0C ) ? 1 \ : ( 0x0D == ((U8*)s)[0] ) ? \ ( ( 0x0A == ((U8*)s)[1] ) ? 2 : 1 ) \ - : ( 0xC2 == ((U8*)s)[0] ) ? \ - ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : 0 ) \ + : ( ( 0xC2 == ((U8*)s)[0] ) && ( 0x85 == ((U8*)s)[1] ) ) ? 2 : 0 ) \ : ((e)-(s) > 0) ? \ ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) \ : 0 ) /*** GENERATED CODE ***/ #define is_LNBREAK_latin1(s) \ -( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0C ) ? 1 \ +( ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0C ) || 0x85 == ((U8*)s)[0] ) ? 1\ : ( 0x0D == ((U8*)s)[0] ) ? \ ( ( 0x0A == ((U8*)s)[1] ) ? 2 : 1 ) \ -: ( 0x85 == ((U8*)s)[0] ) ) +: 0 ) /*** GENERATED CODE ***/ #define is_LNBREAK_latin1_safe(s,e) \ ( ((e)-(s) > 1) ? \ - ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0C ) ? 1 \ + ( ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0C ) || 0x85 == ((U8*)s)[0] ) ? 1\ : ( 0x0D == ((U8*)s)[0] ) ? \ ( ( 0x0A == ((U8*)s)[1] ) ? 2 : 1 ) \ - : ( 0x85 == ((U8*)s)[0] ) ) \ + : 0 ) \ : ((e)-(s) > 0) ? \ ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) || 0x85 == ((U8*)s)[0] )\ : 0 ) @@ -118,25 +106,7 @@ /* HORIZWS: Horizontal Whitespace: \h \H - 0x09 # HT - 0x20 # SPACE - 0xa0 # NBSP - 0x1680 # OGHAM SPACE MARK - 0x180e # MONGOLIAN VOWEL SEPARATOR - 0x2000 # EN QUAD - 0x2001 # EM QUAD - 0x2002 # EN SPACE - 0x2003 # EM SPACE - 0x2004 # THREE-PER-EM SPACE - 0x2005 # FOUR-PER-EM SPACE - 0x2006 # SIX-PER-EM SPACE - 0x2007 # FIGURE SPACE - 0x2008 # PUNCTUATION SPACE - 0x2009 # THIN SPACE - 0x200A # HAIR SPACE - 0x202f # NARROW NO-BREAK SPACE - 0x205f # MEDIUM MATHEMATICAL SPACE - 0x3000 # IDEOGRAPHIC SPACE + \p{HorizSpace} */ /*** GENERATED CODE ***/ #define is_HORIZWS(s,is_utf8) \ @@ -147,18 +117,12 @@ : ( 0xE1 == ((U8*)s)[0] ) ? \ ( ( 0x9A == ((U8*)s)[1] ) ? \ ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ - : ( 0xA0 == ((U8*)s)[1] ) ? \ - ( ( 0x8E == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ : ( 0xE2 == ((U8*)s)[0] ) ? \ ( ( 0x80 == ((U8*)s)[1] ) ? \ ( ( ( 0x80 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x8A ) || 0xAF == ((U8*)s)[2] ) ? 3 : 0 )\ - : ( 0x81 == ((U8*)s)[1] ) ? \ - ( ( 0x9F == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ - : ( 0xE3 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ - : 0 ) \ + : ( ( 0x81 == ((U8*)s)[1] ) && ( 0x9F == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( ( ( 0xE3 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 )\ : ( 0xA0 == ((U8*)s)[0] ) ) /*** GENERATED CODE ***/ @@ -171,18 +135,12 @@ : ( 0xE1 == ((U8*)s)[0] ) ? \ ( ( 0x9A == ((U8*)s)[1] ) ? \ ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ - : ( 0xA0 == ((U8*)s)[1] ) ? \ - ( ( 0x8E == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8E == ((U8*)s)[2] ) ) ? 3 : 0 )\ : ( 0xE2 == ((U8*)s)[0] ) ? \ ( ( 0x80 == ((U8*)s)[1] ) ? \ ( ( ( 0x80 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x8A ) || 0xAF == ((U8*)s)[2] ) ? 3 : 0 )\ - : ( 0x81 == ((U8*)s)[1] ) ? \ - ( ( 0x9F == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ - : ( 0xE3 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 )\ - : 0 ) \ + : ( ( 0x81 == ((U8*)s)[1] ) && ( 0x9F == ((U8*)s)[2] ) ) ? 3 : 0 )\ + : ( ( ( 0xE3 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 )\ : ( 0xA0 == ((U8*)s)[0] ) ) \ : ((e)-(s) > 1) ? \ ( ( 0x09 == ((U8*)s)[0] || 0x20 == ((U8*)s)[0] ) ? 1 \ @@ -204,18 +162,12 @@ : ( 0xE1 == ((U8*)s)[0] ) ? \ ( ( 0x9A == ((U8*)s)[1] ) ? \ ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ - : ( 0xA0 == ((U8*)s)[1] ) ? \ - ( ( 0x8E == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ : ( 0xE2 == ((U8*)s)[0] ) ? \ ( ( 0x80 == ((U8*)s)[1] ) ? \ - ( ( ( 0x80 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x8A ) || 0xAF == ((U8*)s)[2] ) ? 3 : 0 )\ - : ( 0x81 == ((U8*)s)[1] ) ? \ - ( ( 0x9F == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ -: ( 0xE3 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ -: 0 ) + ( ( ( ((U8*)s)[2] <= 0x8A ) || 0xAF == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( ( 0x81 == ((U8*)s)[1] ) && ( 0x9F == ((U8*)s)[2] ) ) ? 3 : 0 ) \ +: ( ( ( 0xE3 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 ) /*** GENERATED CODE ***/ #define is_HORIZWS_utf8_safe(s,e) \ @@ -226,35 +178,53 @@ : ( 0xE1 == ((U8*)s)[0] ) ? \ ( ( 0x9A == ((U8*)s)[1] ) ? \ ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ - : ( 0xA0 == ((U8*)s)[1] ) ? \ - ( ( 0x8E == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ : ( 0xE2 == ((U8*)s)[0] ) ? \ ( ( 0x80 == ((U8*)s)[1] ) ? \ ( ( ( 0x80 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x8A ) || 0xAF == ((U8*)s)[2] ) ? 3 : 0 )\ - : ( 0x81 == ((U8*)s)[1] ) ? \ - ( ( 0x9F == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ - : ( 0xE3 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ - : 0 ) \ + : ( ( 0x81 == ((U8*)s)[1] ) && ( 0x9F == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( ( ( 0xE3 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 )\ : ((e)-(s) > 1) ? \ ( ( 0x09 == ((U8*)s)[0] || 0x20 == ((U8*)s)[0] ) ? 1 \ - : ( 0xC2 == ((U8*)s)[0] ) ? \ - ( ( 0xA0 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : 0 ) \ + : ( ( 0xC2 == ((U8*)s)[0] ) && ( 0xA0 == ((U8*)s)[1] ) ) ? 2 : 0 ) \ : ((e)-(s) > 0) ? \ ( 0x09 == ((U8*)s)[0] || 0x20 == ((U8*)s)[0] ) \ : 0 ) /*** GENERATED CODE ***/ #define is_HORIZWS_latin1(s) \ -( 0x09 == ((U8*)s)[0] || 0x20 == ((U8*)s)[0] || 0xA0 == ((U8*)s)[0] ) +( ((U8*)s)[0] == 0x09 || ( ( ((U8*)s)[0] & 0x7F ) == 0x20 ) ) /*** GENERATED CODE ***/ #define is_HORIZWS_latin1_safe(s,e) \ ( ((e)-(s) > 0) ? \ - ( 0x09 == ((U8*)s)[0] || 0x20 == ((U8*)s)[0] || 0xA0 == ((U8*)s)[0] ) \ + ( ((U8*)s)[0] == 0x09 || ( ( ((U8*)s)[0] & 0x7F ) == 0x20 ) ) \ +: 0 ) + +/*** GENERATED CODE ***/ +#define is_HORIZWS_high(s) \ +( ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0x9A == ((U8*)s)[1] ) ? \ + ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ +: ( 0xE2 == ((U8*)s)[0] ) ? \ + ( ( 0x80 == ((U8*)s)[1] ) ? \ + ( ( ( ((U8*)s)[2] <= 0x8A ) || 0xAF == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( ( 0x81 == ((U8*)s)[1] ) && ( 0x9F == ((U8*)s)[2] ) ) ? 3 : 0 ) \ +: ( ( ( 0xE3 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 ) + +/*** GENERATED CODE ***/ +#define is_HORIZWS_high_safe(s,e) \ +( ((e)-(s) > 2) ? \ + ( ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0x9A == ((U8*)s)[1] ) ? \ + ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0xE2 == ((U8*)s)[0] ) ? \ + ( ( 0x80 == ((U8*)s)[1] ) ? \ + ( ( ( 0x80 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x8A ) || 0xAF == ((U8*)s)[2] ) ? 3 : 0 )\ + : ( ( 0x81 == ((U8*)s)[1] ) && ( 0x9F == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( ( ( 0xE3 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 )\ : 0 ) /*** GENERATED CODE ***/ @@ -266,19 +236,20 @@ ( 0x180E == cp || ( 0x180E < cp && \ ( ( 0x2000 <= cp && cp <= 0x200A ) || ( 0x200A < cp && \ ( 0x202F == cp || ( 0x202F < cp && \ -( 0x205F == cp || ( 0x205F < cp && \ -0x3000 == cp ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) +( 0x205F == cp || 0x3000 == cp ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) + +/*** GENERATED CODE ***/ +#define is_HORIZWS_cp_high(cp) \ +( 0x1680 == cp || ( 0x1680 < cp && \ +( 0x180E == cp || ( 0x180E < cp && \ +( ( 0x2000 <= cp && cp <= 0x200A ) || ( 0x200A < cp && \ +( 0x202F == cp || ( 0x202F < cp && \ +( 0x205F == cp || 0x3000 == cp ) ) ) ) ) ) ) ) ) /* VERTWS: Vertical Whitespace: \v \V - 0x0A # LF - 0x0B # VT - 0x0C # FF - 0x0D # CR - 0x85 # NEL - 0x2028 # LINE SEPARATOR - 0x2029 # PARAGRAPH SEPARATOR + \p{VertSpace} */ /*** GENERATED CODE ***/ #define is_VERTWS(s,is_utf8) \ @@ -286,9 +257,7 @@ : ( is_utf8 ) ? \ ( ( 0xC2 == ((U8*)s)[0] ) ? \ ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xE2 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0xA8 == ((U8*)s)[2] || 0xA9 == ((U8*)s)[2] ) ) ? 3 : 0 )\ - : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ : ( 0x85 == ((U8*)s)[0] ) ) /*** GENERATED CODE ***/ @@ -298,9 +267,7 @@ : ( is_utf8 ) ? \ ( ( 0xC2 == ((U8*)s)[0] ) ? \ ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xE2 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0xA8 == ((U8*)s)[2] || 0xA9 == ((U8*)s)[2] ) ) ? 3 : 0 )\ - : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ : ( 0x85 == ((U8*)s)[0] ) ) \ : ((e)-(s) > 1) ? \ ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) ? 1 \ @@ -319,9 +286,7 @@ ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) ? 1 \ : ( 0xC2 == ((U8*)s)[0] ) ? \ ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ -: ( 0xE2 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0xA8 == ((U8*)s)[2] || 0xA9 == ((U8*)s)[2] ) ) ? 3 : 0 )\ -: 0 ) +: ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 ) /*** GENERATED CODE ***/ #define is_VERTWS_utf8_safe(s,e) \ @@ -329,19 +294,23 @@ ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) ? 1 \ : ( 0xC2 == ((U8*)s)[0] ) ? \ ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xE2 == ((U8*)s)[0] ) ? \ - ( ( ( 0x80 == ((U8*)s)[1] ) && ( 0xA8 == ((U8*)s)[2] || 0xA9 == ((U8*)s)[2] ) ) ? 3 : 0 )\ - : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ : ((e)-(s) > 1) ? \ ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) ? 1 \ - : ( 0xC2 == ((U8*)s)[0] ) ? \ - ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : 0 ) \ + : ( ( 0xC2 == ((U8*)s)[0] ) && ( 0x85 == ((U8*)s)[1] ) ) ? 2 : 0 ) \ : ((e)-(s) > 0) ? \ ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) \ : 0 ) /*** GENERATED CODE ***/ +#define is_VERTWS_high(s) \ +( ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 ) + +/*** GENERATED CODE ***/ +#define is_VERTWS_high_safe(s,e) \ +( ( ( ( ((e)-(s) > 2) && ( 0xE2 == ((U8*)s)[0] ) ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 ) + +/*** GENERATED CODE ***/ #define is_VERTWS_latin1(s) \ ( ( 0x0A <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) || 0x85 == ((U8*)s)[0] ) @@ -355,171 +324,599 @@ #define is_VERTWS_cp(cp) \ ( ( 0x0A <= cp && cp <= 0x0D ) || ( 0x0D < cp && \ ( 0x85 == cp || ( 0x85 < cp && \ -( 0x2028 == cp || ( 0x2028 < cp && \ -0x2029 == cp ) ) ) ) ) ) +( 0x2028 == cp || 0x2029 == cp ) ) ) ) ) + +/*** GENERATED CODE ***/ +#define is_VERTWS_cp_high(cp) \ +( 0x2028 == cp || 0x2029 == cp ) /* - TRICKYFOLD: Problematic fold case letters. When adding to this list, also should add them to regcomp.c and fold_grind.t - - 0x00DF # LATIN SMALL LETTER SHARP S - 0x0390 # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS - 0x03B0 # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND TONOS - 0x1E9E # LATIN CAPITAL LETTER SHARP S, because maps to same as 00DF - 0x1FD3 # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND OXIA; maps same as 0390 - 0x1FE3 # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND OXIA; maps same as 03B0 + XDIGIT: Hexadecimal digits + + \p{XDigit} */ /*** GENERATED CODE ***/ -#define is_TRICKYFOLD(s,is_utf8) \ -( ( is_utf8 ) ? \ - ( ( 0xC3 == ((U8*)s)[0] ) ? \ - ( ( 0x9F == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xCE == ((U8*)s)[0] ) ? \ - ( ( 0x90 == ((U8*)s)[1] || 0xB0 == ((U8*)s)[1] ) ? 2 : 0 ) \ +#define is_XDIGIT_utf8(s) \ +( ( ( 0x30 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x39 ) || ( 0x41 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x46 ) || ( 0x61 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x66 ) ) ? 1\ +: ( 0xEF == ((U8*)s)[0] ) ? \ + ( ( 0xBC == ((U8*)s)[1] ) ? \ + ( ( ( 0x90 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x99 ) || ( 0xA1 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xA6 ) ) ? 3 : 0 )\ + : ( ( 0xBD == ((U8*)s)[1] ) && ( 0x81 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x86 ) ) ? 3 : 0 )\ +: 0 ) + +/*** GENERATED CODE ***/ +#define is_XDIGIT_high(s) \ +( ( 0xEF == ((U8*)s)[0] ) ? \ + ( ( 0xBC == ((U8*)s)[1] ) ? \ + ( ( ( 0x90 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x99 ) || ( 0xA1 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xA6 ) ) ? 3 : 0 )\ + : ( ( 0xBD == ((U8*)s)[1] ) && ( 0x81 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x86 ) ) ? 3 : 0 )\ +: 0 ) + +/*** GENERATED CODE ***/ +#define is_XDIGIT_cp_high(cp) \ +( ( 0xFF10 <= cp && cp <= 0xFF19 ) || ( 0xFF19 < cp && \ +( ( 0xFF21 <= cp && cp <= 0xFF26 ) || ( 0xFF41 <= cp && cp <= 0xFF46 ) ) ) ) + +/* + XPERLSPACE: \p{XPerlSpace} + + \p{XPerlSpace} +*/ +/*** GENERATED CODE ***/ +#define is_XPERLSPACE(s,is_utf8) \ +( ( ( 0x09 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) || 0x20 == ((U8*)s)[0] ) ? 1\ +: ( is_utf8 ) ? \ + ( ( 0xC2 == ((U8*)s)[0] ) ? \ + ( ( 0x85 == ((U8*)s)[1] || 0xA0 == ((U8*)s)[1] ) ? 2 : 0 ) \ : ( 0xE1 == ((U8*)s)[0] ) ? \ - ( ( 0xBA == ((U8*)s)[1] ) ? \ - ( ( 0x9E == ((U8*)s)[2] ) ? 3 : 0 ) \ - : ( 0xBF == ((U8*)s)[1] ) ? \ - ( ( 0x93 == ((U8*)s)[2] || 0xA3 == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ + ( ( 0x9A == ((U8*)s)[1] ) ? \ + ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0xE2 == ((U8*)s)[0] ) ? \ + ( ( 0x80 == ((U8*)s)[1] ) ? \ + ( ( ( 0x80 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x8A ) || ( ((U8*)s)[2] & 0xFE ) == 0xA8 || 0xAF == ((U8*)s)[2] ) ? 3 : 0 )\ + : ( ( 0x81 == ((U8*)s)[1] ) && ( 0x9F == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( ( ( 0xE3 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 )\ +: ( 0x85 == ((U8*)s)[0] || 0xA0 == ((U8*)s)[0] ) ) + +/*** GENERATED CODE ***/ +#define is_XPERLSPACE_utf8(s) \ +( ( ( 0x09 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) || 0x20 == ((U8*)s)[0] ) ? 1\ +: ( 0xC2 == ((U8*)s)[0] ) ? \ + ( ( 0x85 == ((U8*)s)[1] || 0xA0 == ((U8*)s)[1] ) ? 2 : 0 ) \ +: ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0x9A == ((U8*)s)[1] ) ? \ + ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ +: ( 0xE2 == ((U8*)s)[0] ) ? \ + ( ( 0x80 == ((U8*)s)[1] ) ? \ + ( ( ( ((U8*)s)[2] <= 0x8A ) || ( ((U8*)s)[2] & 0xFE ) == 0xA8 || 0xAF == ((U8*)s)[2] ) ? 3 : 0 )\ + : ( ( 0x81 == ((U8*)s)[1] ) && ( 0x9F == ((U8*)s)[2] ) ) ? 3 : 0 ) \ +: ( ( ( 0xE3 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 ) + +/*** GENERATED CODE ***/ +#define is_XPERLSPACE_high(s) \ +( ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0x9A == ((U8*)s)[1] ) ? \ + ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ +: ( 0xE2 == ((U8*)s)[0] ) ? \ + ( ( 0x80 == ((U8*)s)[1] ) ? \ + ( ( ( ((U8*)s)[2] <= 0x8A ) || ( ((U8*)s)[2] & 0xFE ) == 0xA8 || 0xAF == ((U8*)s)[2] ) ? 3 : 0 )\ + : ( ( 0x81 == ((U8*)s)[1] ) && ( 0x9F == ((U8*)s)[2] ) ) ? 3 : 0 ) \ +: ( ( ( 0xE3 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( 0x80 == ((U8*)s)[2] ) ) ? 3 : 0 ) + +/*** GENERATED CODE ***/ +#define is_XPERLSPACE_cp_high(cp) \ +( 0x1680 == cp || ( 0x1680 < cp && \ +( 0x180E == cp || ( 0x180E < cp && \ +( ( 0x2000 <= cp && cp <= 0x200A ) || ( 0x200A < cp && \ +( 0x2028 == cp || ( 0x2028 < cp && \ +( 0x2029 == cp || ( 0x2029 < cp && \ +( 0x202F == cp || ( 0x202F < cp && \ +( 0x205F == cp || 0x3000 == cp ) ) ) ) ) ) ) ) ) ) ) ) ) + +/* + REPLACEMENT: Unicode REPLACEMENT CHARACTER + + 0xFFFD +*/ +/*** GENERATED CODE ***/ +#define is_REPLACEMENT_utf8_safe(s,e) \ +( ( ( ( ((e)-(s) > 2) && ( 0xEF == ((U8*)s)[0] ) ) && ( 0xBF == ((U8*)s)[1] ) ) && ( 0xBD == ((U8*)s)[2] ) ) ? 3 : 0 ) + +/* + NONCHAR: Non character code points + + \p{Nchar} +*/ +/*** GENERATED CODE ***/ +#define is_NONCHAR_utf8(s) \ +( ( 0xEF == ((U8*)s)[0] ) ? \ + ( ( 0xB7 == ((U8*)s)[1] ) ? \ + ( ( 0x90 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xAF ) ? 3 : 0 ) \ + : ( ( 0xBF == ((U8*)s)[1] ) && ( ((U8*)s)[2] >= 0xBE ) ) ? 3 : 0 ) \ +: ( 0xF0 == ((U8*)s)[0] ) ? \ + ( ( ( ( ((U8*)s)[1] == 0x9F || ( ( ((U8*)s)[1] & 0xEF ) == 0xAF ) ) && ( 0xBF == ((U8*)s)[2] ) ) && ( ((U8*)s)[3] >= 0xBE ) ) ? 4 : 0 )\ +: ( 0xF1 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0xF3 ) ? \ + ( ( ( ( ( ((U8*)s)[1] & 0xCF ) == 0x8F ) && ( 0xBF == ((U8*)s)[2] ) ) && ( ((U8*)s)[3] >= 0xBE ) ) ? 4 : 0 )\ +: ( ( ( ( 0xF4 == ((U8*)s)[0] ) && ( 0x8F == ((U8*)s)[1] ) ) && ( 0xBF == ((U8*)s)[2] ) ) && ( ((U8*)s)[3] >= 0xBE ) ) ? 4 : 0 ) + +/* + SURROGATE: Surrogate characters + + \p{Gc=Cs} +*/ +/*** GENERATED CODE ***/ +#define is_SURROGATE_utf8(s) \ +( ( ( 0xED == ((U8*)s)[0] ) && ( ((U8*)s)[1] >= 0xA0 ) ) ? 3 : 0 ) + +/* + GCB_L: Grapheme_Cluster_Break=L + + \p{_X_GCB_L} +*/ +/*** GENERATED CODE ***/ +#define is_GCB_L_utf8(s) \ +( ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0x84 == ((U8*)s)[1] ) ? \ + 3 \ + : ( ( 0x85 == ((U8*)s)[1] ) && ( ((U8*)s)[2] <= 0x9F ) ) ? 3 : 0 ) \ +: ( ( ( 0xEA == ((U8*)s)[0] ) && ( 0xA5 == ((U8*)s)[1] ) ) && ( 0xA0 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xBC ) ) ? 3 : 0 ) + +/* + GCB_LV_LVT_V: Grapheme_Cluster_Break=(LV or LVT or V) + + \p{_X_LV_LVT_V} +*/ +/*** GENERATED CODE ***/ +#define is_GCB_LV_LVT_V_utf8(s) \ +( ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0x85 == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] >= 0xA0 ) ? 3 : 0 ) \ + : ( ( 0x86 == ((U8*)s)[1] ) && ( ((U8*)s)[2] <= 0xA7 ) ) ? 3 : 0 ) \ +: ( 0xEA == ((U8*)s)[0] ) ? \ + ( ( ((U8*)s)[1] >= 0xB0 ) ? \ + 3 \ : 0 ) \ -: ( 0xDF == ((U8*)s)[0] ) ) +: ( 0xEB == ((U8*)s)[0] || 0xEC == ((U8*)s)[0] ) ? \ + 3 \ +: ( 0xED == ((U8*)s)[0] ) ? \ + ( ( ((U8*)s)[1] <= 0x9D ) ? \ + 3 \ + : ( 0x9E == ((U8*)s)[1] ) ? \ + ( ( ( ((U8*)s)[2] <= 0xA3 ) || ( ((U8*)s)[2] >= 0xB0 ) ) ? 3 : 0 ) \ + : ( ( 0x9F == ((U8*)s)[1] ) && ( ((U8*)s)[2] <= 0x86 ) ) ? 3 : 0 ) \ +: 0 ) + +/* + GCB_Prepend: Grapheme_Cluster_Break=Prepend + \p{_X_GCB_Prepend} +*/ /*** GENERATED CODE ***/ -#define is_TRICKYFOLD_safe(s,e,is_utf8) \ -( ((e)-(s) > 2) ? \ - ( ( is_utf8 ) ? \ - ( ( 0xC3 == ((U8*)s)[0] ) ? \ - ( ( 0x9F == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xCE == ((U8*)s)[0] ) ? \ - ( ( 0x90 == ((U8*)s)[1] || 0xB0 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xE1 == ((U8*)s)[0] ) ? \ - ( ( 0xBA == ((U8*)s)[1] ) ? \ - ( ( 0x9E == ((U8*)s)[2] ) ? 3 : 0 ) \ - : ( 0xBF == ((U8*)s)[1] ) ? \ - ( ( 0x93 == ((U8*)s)[2] || 0xA3 == ((U8*)s)[2] ) ? 3 : 0 ) \ - : 0 ) \ - : 0 ) \ - : ( 0xDF == ((U8*)s)[0] ) ) \ -: ((e)-(s) > 1) ? \ - ( ( is_utf8 ) ? \ - ( ( 0xC3 == ((U8*)s)[0] ) ? \ - ( ( 0x9F == ((U8*)s)[1] ) ? 2 : 0 ) \ - : ( 0xCE == ((U8*)s)[0] ) ? \ - ( ( 0x90 == ((U8*)s)[1] || 0xB0 == ((U8*)s)[1] ) ? 2 : 0 ) \ - : 0 ) \ - : ( 0xDF == ((U8*)s)[0] ) ) \ -: ((e)-(s) > 0) ? \ - ( ( !( is_utf8 ) ) ? \ - ( 0xDF == ((U8*)s)[0] ) \ +#define is_GCB_Prepend_utf8(s) \ +( 0 ) + +/* + GCB_RI: Grapheme_Cluster_Break=RI + + \p{_X_RI} +*/ +/*** GENERATED CODE ***/ +#define is_GCB_RI_utf8(s) \ +( ( ( ( ( 0xF0 == ((U8*)s)[0] ) && ( 0x9F == ((U8*)s)[1] ) ) && ( 0x87 == ((U8*)s)[2] ) ) && ( ((U8*)s)[3] >= 0xA6 ) ) ? 4 : 0 ) + +/* + GCB_SPECIAL_BEGIN_START: Grapheme_Cluster_Break=special_begin_starts + + \p{_X_Special_Begin_Start} +*/ +/*** GENERATED CODE ***/ +#define is_GCB_SPECIAL_BEGIN_START_utf8(s) \ +( ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( ( ((U8*)s)[1] & 0xFC ) == 0x84 ) ? \ + 3 \ + : 0 ) \ +: ( 0xEA == ((U8*)s)[0] ) ? \ + ( ( 0xA5 == ((U8*)s)[1] ) ? \ + ( ( 0xA0 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xBC ) ? 3 : 0 ) \ + : ( ((U8*)s)[1] >= 0xB0 ) ? \ + 3 \ + : 0 ) \ +: ( 0xEB == ((U8*)s)[0] || 0xEC == ((U8*)s)[0] ) ? \ + 3 \ +: ( 0xED == ((U8*)s)[0] ) ? \ + ( ( ((U8*)s)[1] <= 0x9D ) ? \ + 3 \ + : ( 0x9E == ((U8*)s)[1] ) ? \ + ( ( ( ((U8*)s)[2] <= 0xA3 ) || ( ((U8*)s)[2] >= 0xB0 ) ) ? 3 : 0 ) \ + : ( ( 0x9F == ((U8*)s)[1] ) && ( ( ((U8*)s)[2] <= 0x86 ) || ( 0x8B <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xBB ) ) ) ? 3 : 0 )\ +: ( ( ( ( 0xF0 == ((U8*)s)[0] ) && ( 0x9F == ((U8*)s)[1] ) ) && ( 0x87 == ((U8*)s)[2] ) ) && ( ((U8*)s)[3] >= 0xA6 ) ) ? 4 : 0 ) + +/* + GCB_T: Grapheme_Cluster_Break=T + + \p{_X_GCB_T} +*/ +/*** GENERATED CODE ***/ +#define is_GCB_T_utf8(s) \ +( ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0x86 == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] >= 0xA8 ) ? 3 : 0 ) \ + : ( 0x87 == ((U8*)s)[1] ) ? \ + 3 \ : 0 ) \ +: ( ( ( 0xED == ((U8*)s)[0] ) && ( 0x9F == ((U8*)s)[1] ) ) && ( 0x8B <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xBB ) ) ? 3 : 0 ) + +/* + GCB_V: Grapheme_Cluster_Break=V + + \p{_X_GCB_V} +*/ +/*** GENERATED CODE ***/ +#define is_GCB_V_utf8(s) \ +( ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0x85 == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] >= 0xA0 ) ? 3 : 0 ) \ + : ( ( 0x86 == ((U8*)s)[1] ) && ( ((U8*)s)[2] <= 0xA7 ) ) ? 3 : 0 ) \ +: ( 0xED == ((U8*)s)[0] ) ? \ + ( ( 0x9E == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] >= 0xB0 ) ? 3 : 0 ) \ + : ( ( 0x9F == ((U8*)s)[1] ) && ( ((U8*)s)[2] <= 0x86 ) ) ? 3 : 0 ) \ : 0 ) +/* + QUOTEMETA: Meta-characters that \Q should quote + + \p{_Perl_Quotemeta} +*/ /*** GENERATED CODE ***/ -#define is_TRICKYFOLD_cp(cp) \ -( 0xDF == cp || ( 0xDF < cp && \ -( 0x390 == cp || ( 0x390 < cp && \ -( 0x3B0 == cp || ( 0x3B0 < cp && \ -( 0x1E9E == cp || ( 0x1E9E < cp && \ -( 0x1FD3 == cp || ( 0x1FD3 < cp && \ -0x1FE3 == cp ) ) ) ) ) ) ) ) ) ) +#define is_QUOTEMETA_high(s) \ +( ( 0xCD == ((U8*)s)[0] ) ? \ + ( ( 0x8F == ((U8*)s)[1] ) ? 2 : 0 ) \ +: ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0x85 == ((U8*)s)[1] ) ? \ + ( ( 0x9F == ((U8*)s)[2] || 0xA0 == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( 0x9A == ((U8*)s)[1] ) ? \ + ( ( 0x80 == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( 0x9E == ((U8*)s)[1] ) ? \ + ( ( ( ((U8*)s)[2] & 0xFE ) == 0xB4 ) ? 3 : 0 ) \ + : ( ( 0xA0 == ((U8*)s)[1] ) && ( 0x8B <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x8E ) ) ? 3 : 0 )\ +: ( 0xE2 == ((U8*)s)[0] ) ? \ + ( ( 0x80 == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] <= 0xBE ) ? 3 : 0 ) \ + : ( 0x81 == ((U8*)s)[1] ) ? \ + ( ( ( 0x81 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0x93 ) || ( 0x95 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xAF ) ) ? 3 : 0 )\ + : ( 0x86 == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] >= 0x90 ) ? 3 : 0 ) \ + : ( ( 0x87 <= ((U8*)s)[1] && ((U8*)s)[1] <= 0x90 ) || ( 0x94 <= ((U8*)s)[1] && ((U8*)s)[1] <= 0x9C ) || ( 0x9F <= ((U8*)s)[1] && ((U8*)s)[1] <= 0xAF ) || ( ((U8*)s)[1] & 0xFE ) == 0xB8 ) ?\ + 3 \ + : ( 0x91 == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] <= 0x9F ) ? 3 : 0 ) \ + : ( 0x9D == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] <= 0xB5 ) ? 3 : 0 ) \ + : ( ( 0x9E == ((U8*)s)[1] ) && ( ((U8*)s)[2] >= 0x94 ) ) ? 3 : 0 ) \ +: ( 0xE3 == ((U8*)s)[0] ) ? \ + ( ( 0x80 == ((U8*)s)[1] ) ? \ + ( ( ( ((U8*)s)[2] <= 0x83 ) || ( 0x88 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xA0 ) || 0xB0 == ((U8*)s)[2] ) ? 3 : 0 )\ + : ( ( 0x85 == ((U8*)s)[1] ) && ( 0xA4 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ +: ( 0xEF == ((U8*)s)[0] ) ? \ + ( ( 0xB4 == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] >= 0xBE ) ? 3 : 0 ) \ + : ( 0xB8 == ((U8*)s)[1] ) ? \ + ( ( ((U8*)s)[2] <= 0x8F ) ? 3 : 0 ) \ + : ( 0xB9 == ((U8*)s)[1] ) ? \ + ( ( 0x85 == ((U8*)s)[2] || 0x86 == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( 0xBB == ((U8*)s)[1] ) ? \ + ( ( 0xBF == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( 0xBE == ((U8*)s)[1] ) ? \ + ( ( 0xA0 == ((U8*)s)[2] ) ? 3 : 0 ) \ + : ( ( 0xBF == ((U8*)s)[1] ) && ( 0xB0 <= ((U8*)s)[2] && ((U8*)s)[2] <= 0xB8 ) ) ? 3 : 0 )\ +: ( 0xF0 == ((U8*)s)[0] ) ? \ + ( ( ( ( 0x9D == ((U8*)s)[1] ) && ( 0x85 == ((U8*)s)[2] ) ) && ( 0xB3 <= ((U8*)s)[3] && ((U8*)s)[3] <= 0xBA ) ) ? 4 : 0 )\ +: ( ( 0xF3 == ((U8*)s)[0] ) && ( 0xA0 == ((U8*)s)[1] ) ) ? 4 : 0 ) + +/* + MULTI_CHAR_FOLD: multi-char strings that are folded to by a single character + do regen/regcharclass_multi_char_folds.pl + ®charclass_multi_char_folds::multi_char_folds(1) +*/ /*** GENERATED CODE ***/ -#define what_TRICKYFOLD(s,is_utf8) \ -( ( is_utf8 ) ? \ - ( ( 0xC3 == ((U8*)s)[0] ) ? \ - ( ( 0x9F == ((U8*)s)[1] ) ? 0xDF : 0 ) \ +#define is_MULTI_CHAR_FOLD_utf8_safe_part0(s,e) \ +( ( 0x61 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCA == ((U8*)s)[1] ) && ( 0xBE == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x66 == ((U8*)s)[0] ) ? \ + ( ( 0x66 == ((U8*)s)[1] ) ? \ + ( ( 0x69 == ((U8*)s)[2] || 0x6C == ((U8*)s)[2] ) ? 3 : 2 ) \ + : ( 0x69 == ((U8*)s)[1] || 0x6C == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( 0x68 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0xB1 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x69 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x87 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x6A == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x8C == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x73 == ((U8*)s)[0] ) ? \ + ( ( 0x73 == ((U8*)s)[1] || 0x74 == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( 0x74 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x88 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x77 == ((U8*)s)[0] || 0x79 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x8A == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0xCA == ((U8*)s)[0] ) ? \ + ( ( ( 0xBC == ((U8*)s)[1] ) && ( 0x6E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ : ( 0xCE == ((U8*)s)[0] ) ? \ - ( ( 0x90 == ((U8*)s)[1] ) ? 0x390 \ - : ( 0xB0 == ((U8*)s)[1] ) ? 0x3B0 : 0 ) \ + ( ( ( ((U8*)s)[1] & 0xFD ) == 0xAC ) ? \ + ( ( ( 0xCE == ((U8*)s)[2] ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xB1 == ((U8*)s)[1] || 0xB7 == ((U8*)s)[1] ) ? \ + ( ( 0xCD == ((U8*)s)[2] ) ? \ + ( ( 0x82 == ((U8*)s)[3] ) ? 4 : 0 ) \ + : ( ( 0xCE == ((U8*)s)[2] ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( ( ( 0xB9 == ((U8*)s)[1] ) && ( 0xCD == ((U8*)s)[2] ) ) && ( 0x82 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xCF == ((U8*)s)[0] ) ? \ + ( ( 0x81 == ((U8*)s)[1] ) ? \ + ( ( ( 0xCC == ((U8*)s)[2] ) && ( 0x93 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0x85 == ((U8*)s)[1] ) ? \ + ( ( 0xCC == ((U8*)s)[2] ) ? \ + ( ( 0x93 == ((U8*)s)[3] ) ? 4 : 0 ) \ + : ( ( 0xCD == ((U8*)s)[2] ) && ( 0x82 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0x89 == ((U8*)s)[1] ) ? \ + ( ( 0xCD == ((U8*)s)[2] ) ? \ + ( ( 0x82 == ((U8*)s)[3] ) ? 4 : 0 ) \ + : ( ( 0xCE == ((U8*)s)[2] ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( ( ( 0x8E == ((U8*)s)[1] ) && ( 0xCE == ((U8*)s)[2] ) ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xD5 == ((U8*)s)[0] ) ? \ + ( ( 0xA5 == ((U8*)s)[1] ) ? \ + ( ( ( 0xD6 == ((U8*)s)[2] ) && ( 0x82 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xB4 == ((U8*)s)[1] ) ? \ + ( ( ( 0xD5 == ((U8*)s)[2] ) && ( ( ( ((U8*)s)[3] & 0xF7 ) == 0xA5 ) || ((U8*)s)[3] == 0xAB || ((U8*)s)[3] == 0xB6 ) ) ? 4 : 0 )\ + : ( ( ( 0xBE == ((U8*)s)[1] ) && ( 0xD5 == ((U8*)s)[2] ) ) && ( 0xB6 == ((U8*)s)[3] ) ) ? 4 : 0 )\ : ( 0xE1 == ((U8*)s)[0] ) ? \ - ( ( 0xBA == ((U8*)s)[1] ) ? \ - ( ( 0x9E == ((U8*)s)[2] ) ? 0x1E9E : 0 ) \ - : ( 0xBF == ((U8*)s)[1] ) ? \ - ( ( 0x93 == ((U8*)s)[2] ) ? 0x1FD3 \ - : ( 0xA3 == ((U8*)s)[2] ) ? 0x1FE3 : 0 ) \ + ( ( 0xBC == ((U8*)s)[1] ) ? \ + ( ( ( ( ( ((U8*)s)[2] & 0xD8 ) == 0x80 ) && ( 0xCE == ((U8*)s)[3] ) ) && ( 0xB9 == ((U8*)s)[4] ) ) ? 5 : 0 )\ + : ( ( ( ( 0xBD == ((U8*)s)[1] ) && ( ( ( ((U8*)s)[2] & 0xF8 ) == 0xA0 ) || ( ( ((U8*)s)[2] & 0xFB ) == 0xB0 ) || ((U8*)s)[2] == 0xBC ) ) && ( 0xCE == ((U8*)s)[3] ) ) && ( 0xB9 == ((U8*)s)[4] ) ) ? 5 : 0 )\ + : 0 ) + + +/*** GENERATED CODE ***/ +#define is_MULTI_CHAR_FOLD_utf8_safe_part1(s,e) \ +( ((e)-(s) > 3) ? \ + ( ( 0x61 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCA == ((U8*)s)[1] ) && ( 0xBE == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x66 == ((U8*)s)[0] ) ? \ + ( ( 0x66 == ((U8*)s)[1] ) ? \ + ( ( 0x69 == ((U8*)s)[2] || 0x6C == ((U8*)s)[2] ) ? 3 : 2 ) \ + : ( 0x69 == ((U8*)s)[1] || 0x6C == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( 0x68 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0xB1 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x69 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x87 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x6A == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x8C == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x73 == ((U8*)s)[0] ) ? \ + ( ( 0x73 == ((U8*)s)[1] || 0x74 == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( 0x74 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x88 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x77 == ((U8*)s)[0] || 0x79 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x8A == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0xCA == ((U8*)s)[0] ) ? \ + ( ( ( 0xBC == ((U8*)s)[1] ) && ( 0x6E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0xCE == ((U8*)s)[0] ) ? \ + ( ( ( ((U8*)s)[1] & 0xFD ) == 0xAC ) ? \ + ( ( ( 0xCE == ((U8*)s)[2] ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xB1 == ((U8*)s)[1] || 0xB7 == ((U8*)s)[1] ) ? \ + ( ( 0xCD == ((U8*)s)[2] ) ? \ + ( ( 0x82 == ((U8*)s)[3] ) ? 4 : 0 ) \ + : ( ( 0xCE == ((U8*)s)[2] ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( ( ( 0xB9 == ((U8*)s)[1] ) && ( 0xCD == ((U8*)s)[2] ) ) && ( 0x82 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xCF == ((U8*)s)[0] ) ? \ + ( ( 0x81 == ((U8*)s)[1] ) ? \ + ( ( ( 0xCC == ((U8*)s)[2] ) && ( 0x93 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0x85 == ((U8*)s)[1] ) ? \ + ( ( 0xCC == ((U8*)s)[2] ) ? \ + ( ( 0x93 == ((U8*)s)[3] ) ? 4 : 0 ) \ + : ( ( 0xCD == ((U8*)s)[2] ) && ( 0x82 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0x89 == ((U8*)s)[1] ) ? \ + ( ( 0xCD == ((U8*)s)[2] ) ? \ + ( ( 0x82 == ((U8*)s)[3] ) ? 4 : 0 ) \ + : ( ( 0xCE == ((U8*)s)[2] ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( ( ( 0x8E == ((U8*)s)[1] ) && ( 0xCE == ((U8*)s)[2] ) ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xD5 == ((U8*)s)[0] ) ? \ + ( ( 0xA5 == ((U8*)s)[1] ) ? \ + ( ( ( 0xD6 == ((U8*)s)[2] ) && ( 0x82 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xB4 == ((U8*)s)[1] ) ? \ + ( ( ( 0xD5 == ((U8*)s)[2] ) && ( ( ( ((U8*)s)[3] & 0xF7 ) == 0xA5 ) || ((U8*)s)[3] == 0xAB || ((U8*)s)[3] == 0xB6 ) ) ? 4 : 0 )\ + : ( ( ( 0xBE == ((U8*)s)[1] ) && ( 0xD5 == ((U8*)s)[2] ) ) && ( 0xB6 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : 0 ) \ +: ((e)-(s) > 2) ? \ + ( ( 0x61 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCA == ((U8*)s)[1] ) && ( 0xBE == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x66 == ((U8*)s)[0] ) ? \ + ( ( 0x66 == ((U8*)s)[1] ) ? \ + ( ( 0x69 == ((U8*)s)[2] || 0x6C == ((U8*)s)[2] ) ? 3 : 2 ) \ + : ( 0x69 == ((U8*)s)[1] || 0x6C == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( 0x68 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0xB1 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x69 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x87 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x6A == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x8C == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x73 == ((U8*)s)[0] ) ? \ + ( ( 0x73 == ((U8*)s)[1] || 0x74 == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( 0x74 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x88 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x77 == ((U8*)s)[0] || 0x79 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x8A == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( ( ( 0xCA == ((U8*)s)[0] ) && ( 0xBC == ((U8*)s)[1] ) ) && ( 0x6E == ((U8*)s)[2] ) ) ? 3 : 0 )\ +: ((e)-(s) > 1) ? \ + ( ( 0x66 == ((U8*)s)[0] ) ? \ + ( ( 0x66 == ((U8*)s)[1] || 0x69 == ((U8*)s)[1] || 0x6C == ((U8*)s)[1] ) ? 2 : 0 )\ + : ( ( 0x73 == ((U8*)s)[0] ) && ( 0x73 == ((U8*)s)[1] || 0x74 == ((U8*)s)[1] ) ) ? 2 : 0 )\ +: 0 ) + + +/*** GENERATED CODE ***/ +#define is_MULTI_CHAR_FOLD_utf8_safe(s,e) \ +( ((e)-(s) > 5) ? \ + ( ( 0x61 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCA == ((U8*)s)[1] ) && ( 0xBE == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x66 == ((U8*)s)[0] ) ? \ + ( ( 0x66 == ((U8*)s)[1] ) ? \ + ( ( 0x69 == ((U8*)s)[2] || 0x6C == ((U8*)s)[2] ) ? 3 : 2 ) \ + : ( 0x69 == ((U8*)s)[1] || 0x6C == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( 0x68 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0xB1 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x69 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x87 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x6A == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x8C == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x73 == ((U8*)s)[0] ) ? \ + ( ( 0x73 == ((U8*)s)[1] || 0x74 == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( 0x74 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x88 == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0x77 == ((U8*)s)[0] || 0x79 == ((U8*)s)[0] ) ? \ + ( ( ( 0xCC == ((U8*)s)[1] ) && ( 0x8A == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0xCA == ((U8*)s)[0] ) ? \ + ( ( ( 0xBC == ((U8*)s)[1] ) && ( 0x6E == ((U8*)s)[2] ) ) ? 3 : 0 ) \ + : ( 0xCE == ((U8*)s)[0] ) ? \ + ( ( ( ((U8*)s)[1] & 0xFD ) == 0xAC ) ? \ + ( ( ( 0xCE == ((U8*)s)[2] ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xB1 == ((U8*)s)[1] || 0xB7 == ((U8*)s)[1] ) ? \ + ( ( 0xCD == ((U8*)s)[2] ) ? \ + ( ( 0x82 == ((U8*)s)[3] ) ? \ + ( ( ( 0xCE == ((U8*)s)[4] ) && ( 0xB9 == ((U8*)s)[5] ) ) ? 6 : 4 )\ + : 0 ) \ + : ( ( 0xCE == ((U8*)s)[2] ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xB9 == ((U8*)s)[1] ) ? \ + ( ( 0xCC == ((U8*)s)[2] ) ? \ + ( ( 0x88 == ((U8*)s)[3] ) ? \ + ( ( 0xCC == ((U8*)s)[4] ) ? \ + ( ( ( ((U8*)s)[5] & 0xFE ) == 0x80 ) ? 6 : 0 ) \ + : ( ( 0xCD == ((U8*)s)[4] ) && ( 0x82 == ((U8*)s)[5] ) ) ? 6 : 0 )\ + : 0 ) \ + : ( ( 0xCD == ((U8*)s)[2] ) && ( 0x82 == ((U8*)s)[3] ) ) ? 4 : 0 )\ : 0 ) \ + : ( 0xCF == ((U8*)s)[0] ) ? \ + ( ( 0x81 == ((U8*)s)[1] ) ? \ + ( ( ( 0xCC == ((U8*)s)[2] ) && ( 0x93 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0x85 == ((U8*)s)[1] ) ? \ + ( ( 0xCC == ((U8*)s)[2] ) ? \ + ( ( 0x88 == ((U8*)s)[3] ) ? \ + ( ( 0xCC == ((U8*)s)[4] ) ? \ + ( ( ( ((U8*)s)[5] & 0xFE ) == 0x80 ) ? 6 : 0 ) \ + : ( ( 0xCD == ((U8*)s)[4] ) && ( 0x82 == ((U8*)s)[5] ) ) ? 6 : 0 )\ + : ( 0x93 == ((U8*)s)[3] ) ? \ + ( ( 0xCC == ((U8*)s)[4] ) ? \ + ( ( ( ((U8*)s)[5] & 0xFE ) == 0x80 ) ? 6 : 4 ) \ + : ( ( 0xCD == ((U8*)s)[4] ) && ( 0x82 == ((U8*)s)[5] ) ) ? 6 : 4 )\ + : 0 ) \ + : ( ( 0xCD == ((U8*)s)[2] ) && ( 0x82 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0x89 == ((U8*)s)[1] ) ? \ + ( ( 0xCD == ((U8*)s)[2] ) ? \ + ( ( 0x82 == ((U8*)s)[3] ) ? \ + ( ( ( 0xCE == ((U8*)s)[4] ) && ( 0xB9 == ((U8*)s)[5] ) ) ? 6 : 4 )\ + : 0 ) \ + : ( ( 0xCE == ((U8*)s)[2] ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( ( ( 0x8E == ((U8*)s)[1] ) && ( 0xCE == ((U8*)s)[2] ) ) && ( 0xB9 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xD5 == ((U8*)s)[0] ) ? \ + ( ( 0xA5 == ((U8*)s)[1] ) ? \ + ( ( ( 0xD6 == ((U8*)s)[2] ) && ( 0x82 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xB4 == ((U8*)s)[1] ) ? \ + ( ( ( 0xD5 == ((U8*)s)[2] ) && ( ( ( ((U8*)s)[3] & 0xF7 ) == 0xA5 ) || ((U8*)s)[3] == 0xAB || ((U8*)s)[3] == 0xB6 ) ) ? 4 : 0 )\ + : ( ( ( 0xBE == ((U8*)s)[1] ) && ( 0xD5 == ((U8*)s)[2] ) ) && ( 0xB6 == ((U8*)s)[3] ) ) ? 4 : 0 )\ + : ( 0xE1 == ((U8*)s)[0] ) ? \ + ( ( 0xBC == ((U8*)s)[1] ) ? \ + ( ( ( ( ( ((U8*)s)[2] & 0xD8 ) == 0x80 ) && ( 0xCE == ((U8*)s)[3] ) ) && ( 0xB9 == ((U8*)s)[4] ) ) ? 5 : 0 )\ + : ( ( ( ( 0xBD == ((U8*)s)[1] ) && ( ( ( ((U8*)s)[2] & 0xF8 ) == 0xA0 ) || ( ( ((U8*)s)[2] & 0xFB ) == 0xB0 ) || ((U8*)s)[2] == 0xBC ) ) && ( 0xCE == ((U8*)s)[3] ) ) && ( 0xB9 == ((U8*)s)[4] ) ) ? 5 : 0 )\ : 0 ) \ -: ( 0xDF == ((U8*)s)[0] ) ? 0xDF : 0 ) +: ((e)-(s) > 4) ? is_MULTI_CHAR_FOLD_utf8_safe_part0(s,e) : is_MULTI_CHAR_FOLD_utf8_safe_part1(s,e) ) + +/* + MULTI_CHAR_FOLD: multi-char strings that are folded to by a single character + ®charclass_multi_char_folds::multi_char_folds(0) +*/ /*** GENERATED CODE ***/ -#define what_TRICKYFOLD_safe(s,e,is_utf8) \ +#define is_MULTI_CHAR_FOLD_latin1_safe(s,e) \ ( ((e)-(s) > 2) ? \ - ( ( is_utf8 ) ? \ - ( ( 0xC3 == ((U8*)s)[0] ) ? \ - ( ( 0x9F == ((U8*)s)[1] ) ? 0xDF : 0 ) \ - : ( 0xCE == ((U8*)s)[0] ) ? \ - ( ( 0x90 == ((U8*)s)[1] ) ? 0x390 \ - : ( 0xB0 == ((U8*)s)[1] ) ? 0x3B0 : 0 ) \ - : ( 0xE1 == ((U8*)s)[0] ) ? \ - ( ( 0xBA == ((U8*)s)[1] ) ? \ - ( ( 0x9E == ((U8*)s)[2] ) ? 0x1E9E : 0 ) \ - : ( 0xBF == ((U8*)s)[1] ) ? \ - ( ( 0x93 == ((U8*)s)[2] ) ? 0x1FD3 \ - : ( 0xA3 == ((U8*)s)[2] ) ? 0x1FE3 : 0 ) \ - : 0 ) \ - : 0 ) \ - : ( 0xDF == ((U8*)s)[0] ) ? 0xDF : 0 ) \ + ( ( ( ((U8*)s)[0] & 0xDF ) == 0x46 ) ? \ + ( ( ( ((U8*)s)[1] & 0xDF ) == 0x46 ) ? \ + ( ( ( ( ((U8*)s)[2] & 0xDF ) == 0x49 ) || ( ( ((U8*)s)[2] & 0xDF ) == 0x4C ) ) ? 3 : 2 )\ + : ( ( ( ((U8*)s)[1] & 0xDF ) == 0x49 ) || ( ( ((U8*)s)[1] & 0xDF ) == 0x4C ) ) ? 2 : 0 )\ + : ( ( ( ((U8*)s)[0] & 0xDF ) == 0x53 ) && ( ( ( ((U8*)s)[1] & 0xDF ) == 0x53 ) || ( ( ((U8*)s)[1] & 0xDF ) == 0x54 ) ) ) ? 2 : 0 )\ : ((e)-(s) > 1) ? \ - ( ( is_utf8 ) ? \ - ( ( 0xC3 == ((U8*)s)[0] ) ? \ - ( ( 0x9F == ((U8*)s)[1] ) ? 0xDF : 0 ) \ - : ( 0xCE == ((U8*)s)[0] ) ? \ - ( ( 0x90 == ((U8*)s)[1] ) ? 0x390 \ - : ( 0xB0 == ((U8*)s)[1] ) ? 0x3B0 : 0 ) \ - : 0 ) \ - : ( 0xDF == ((U8*)s)[0] ) ? 0xDF : 0 ) \ + ( ( ( ((U8*)s)[0] & 0xDF ) == 0x46 ) ? \ + ( ( ( ( ((U8*)s)[1] & 0xDF ) == 0x46 ) || ( ( ((U8*)s)[1] & 0xDF ) == 0x49 ) || ( ( ((U8*)s)[1] & 0xDF ) == 0x4C ) ) ? 2 : 0 )\ + : ( ( ( ((U8*)s)[0] & 0xDF ) == 0x53 ) && ( ( ( ((U8*)s)[1] & 0xDF ) == 0x53 ) || ( ( ((U8*)s)[1] & 0xDF ) == 0x54 ) ) ) ? 2 : 0 )\ +: 0 ) + +/* + PATWS: pattern white space + + \p{PatWS} +*/ +/*** GENERATED CODE ***/ +#define is_PATWS(s,is_utf8) \ +( ( ( 0x09 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) || 0x20 == ((U8*)s)[0] ) ? 1\ +: ( is_utf8 ) ? \ + ( ( 0xC2 == ((U8*)s)[0] ) ? \ + ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0x8E || ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ +: ( 0x85 == ((U8*)s)[0] ) ) + +/*** GENERATED CODE ***/ +#define is_PATWS_safe(s,e,is_utf8) \ +( ((e)-(s) > 2) ? \ + ( ( ( 0x09 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) || 0x20 == ((U8*)s)[0] ) ? 1\ + : ( is_utf8 ) ? \ + ( ( 0xC2 == ((U8*)s)[0] ) ? \ + ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0x8E || ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ + : ( 0x85 == ((U8*)s)[0] ) ) \ +: ((e)-(s) > 1) ? \ + ( ( ( 0x09 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) || 0x20 == ((U8*)s)[0] ) ? 1\ + : ( is_utf8 ) ? \ + ( ( ( 0xC2 == ((U8*)s)[0] ) && ( 0x85 == ((U8*)s)[1] ) ) ? 2 : 0 ) \ + : ( 0x85 == ((U8*)s)[0] ) ) \ : ((e)-(s) > 0) ? \ - ( ( ( !( is_utf8 ) ) && ( 0xDF == ((U8*)s)[0] ) ) ? 0xDF : 0 ) \ + ( ( ( 0x09 <= ((U8*)s)[0] && ((U8*)s)[0] <= 0x0D ) || 0x20 == ((U8*)s)[0] ) ? 1\ + : ( !( is_utf8 ) ) ? \ + ( 0x85 == ((U8*)s)[0] ) \ + : 0 ) \ : 0 ) /*** GENERATED CODE ***/ -#define what_len_TRICKYFOLD(s,is_utf8,len) \ +#define is_PATWS_non_low(s,is_utf8) \ ( ( is_utf8 ) ? \ - ( ( 0xC3 == ((U8*)s)[0] ) ? \ - ( ( 0x9F == ((U8*)s)[1] ) ? len=2, 0xDF : 0 ) \ - : ( 0xCE == ((U8*)s)[0] ) ? \ - ( ( 0x90 == ((U8*)s)[1] ) ? len=2, 0x390 \ - : ( 0xB0 == ((U8*)s)[1] ) ? len=2, 0x3B0 : 0 ) \ - : ( 0xE1 == ((U8*)s)[0] ) ? \ - ( ( 0xBA == ((U8*)s)[1] ) ? \ - ( ( 0x9E == ((U8*)s)[2] ) ? len=3, 0x1E9E : 0 ) \ - : ( 0xBF == ((U8*)s)[1] ) ? \ - ( ( 0x93 == ((U8*)s)[2] ) ? len=3, 0x1FD3 \ - : ( 0xA3 == ((U8*)s)[2] ) ? len=3, 0x1FE3 : 0 ) \ - : 0 ) \ - : 0 ) \ -: ( 0xDF == ((U8*)s)[0] ) ? len=1, 0xDF : 0 ) + ( ( 0xC2 == ((U8*)s)[0] ) ? \ + ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0x8E || ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ +: ( 0x85 == ((U8*)s)[0] ) ) /*** GENERATED CODE ***/ -#define what_len_TRICKYFOLD_safe(s,e,is_utf8,len) \ +#define is_PATWS_non_low_safe(s,e,is_utf8) \ ( ((e)-(s) > 2) ? \ ( ( is_utf8 ) ? \ - ( ( 0xC3 == ((U8*)s)[0] ) ? \ - ( ( 0x9F == ((U8*)s)[1] ) ? len=2, 0xDF : 0 ) \ - : ( 0xCE == ((U8*)s)[0] ) ? \ - ( ( 0x90 == ((U8*)s)[1] ) ? len=2, 0x390 \ - : ( 0xB0 == ((U8*)s)[1] ) ? len=2, 0x3B0 : 0 ) \ - : ( 0xE1 == ((U8*)s)[0] ) ? \ - ( ( 0xBA == ((U8*)s)[1] ) ? \ - ( ( 0x9E == ((U8*)s)[2] ) ? len=3, 0x1E9E : 0 ) \ - : ( 0xBF == ((U8*)s)[1] ) ? \ - ( ( 0x93 == ((U8*)s)[2] ) ? len=3, 0x1FD3 \ - : ( 0xA3 == ((U8*)s)[2] ) ? len=3, 0x1FE3 : 0 ) \ - : 0 ) \ - : 0 ) \ - : ( 0xDF == ((U8*)s)[0] ) ? len=1, 0xDF : 0 ) \ + ( ( 0xC2 == ((U8*)s)[0] ) ? \ + ( ( 0x85 == ((U8*)s)[1] ) ? 2 : 0 ) \ + : ( ( ( 0xE2 == ((U8*)s)[0] ) && ( 0x80 == ((U8*)s)[1] ) ) && ( ( ((U8*)s)[2] & 0xFE ) == 0x8E || ( ((U8*)s)[2] & 0xFE ) == 0xA8 ) ) ? 3 : 0 )\ + : ( 0x85 == ((U8*)s)[0] ) ) \ : ((e)-(s) > 1) ? \ ( ( is_utf8 ) ? \ - ( ( 0xC3 == ((U8*)s)[0] ) ? \ - ( ( 0x9F == ((U8*)s)[1] ) ? len=2, 0xDF : 0 ) \ - : ( 0xCE == ((U8*)s)[0] ) ? \ - ( ( 0x90 == ((U8*)s)[1] ) ? len=2, 0x390 \ - : ( 0xB0 == ((U8*)s)[1] ) ? len=2, 0x3B0 : 0 ) \ - : 0 ) \ - : ( 0xDF == ((U8*)s)[0] ) ? len=1, 0xDF : 0 ) \ -: ((e)-(s) > 0) ? \ - ( ( ( !( is_utf8 ) ) && ( 0xDF == ((U8*)s)[0] ) ) ? len=1, 0xDF : 0 ) \ -: 0 ) + ( ( ( 0xC2 == ((U8*)s)[0] ) && ( 0x85 == ((U8*)s)[1] ) ) ? 2 : 0 ) \ + : ( 0x85 == ((U8*)s)[0] ) ) \ +: ( ((e)-(s) > 0) && ( !( is_utf8 ) ) ) ? ( 0x85 == ((U8*)s)[0] ) : 0 ) + +/*** GENERATED CODE ***/ +#define is_PATWS_cp(cp) \ +( ( 0x09 <= cp && cp <= 0x0D ) || ( 0x0D < cp && \ +( 0x20 == cp || ( 0x20 < cp && \ +( 0x85 == cp || ( 0x85 < cp && \ +( 0x200E == cp || ( 0x200E < cp && \ +( 0x200F == cp || ( 0x200F < cp && \ +( 0x2028 == cp || 0x2029 == cp ) ) ) ) ) ) ) ) ) ) ) + +#endif /* H_REGCHARCLASS */ /* ex: set ro: */ |