1 /* 2 * sh.char.h: Table for spotting special characters quickly 3 * Makes for very obscure but efficient coding. 4 */ 5 /*- 6 * Copyright (c) 1980, 1991 The Regents of the University of California. 7 * All rights reserved. 8 * 9 * Redistribution and use in source and binary forms, with or without 10 * modification, are permitted provided that the following conditions 11 * are met: 12 * 1. Redistributions of source code must retain the above copyright 13 * notice, this list of conditions and the following disclaimer. 14 * 2. Redistributions in binary form must reproduce the above copyright 15 * notice, this list of conditions and the following disclaimer in the 16 * documentation and/or other materials provided with the distribution. 17 * 3. Neither the name of the University nor the names of its contributors 18 * may be used to endorse or promote products derived from this software 19 * without specific prior written permission. 20 * 21 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 24 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 31 * SUCH DAMAGE. 32 */ 33 #ifndef _h_sh_char 34 #define _h_sh_char 35 #if defined(NeXT) && defined(NLS) 36 # include <appkit/NXCType.h> 37 #else 38 # include <ctype.h> 39 # ifdef WIDE_STRINGS 40 # ifdef HAVE_WCTYPE_H 41 # include <wctype.h> 42 # else 43 # include <wchar.h> 44 # endif 45 # endif 46 #endif 47 48 typedef unsigned char tcshuc; 49 #ifdef _MINIX 50 # undef _SP 51 #endif /* _MINIX */ 52 extern unsigned short _cmap[]; 53 #if defined(DSPMBYTE) 54 # define CHECK_MBYTEVAR STRdspmbyte 55 #endif 56 extern unsigned short _cmap_c[]; 57 extern unsigned short _cmap_mbyte[]; 58 extern short _enable_mbdisp; 59 extern unsigned short _mbmap[]; 60 extern unsigned short _mbmap_euc[]; 61 extern unsigned short _mbmap_sjis[]; 62 extern unsigned short _mbmap_big5[]; 63 extern unsigned short _mbmap_utf8[]; 64 /* VARIABLE Check str */ 65 /* same compiler require #define even not define DSPMBYTE */ 66 #undef _MB1 67 #define _MB1 0x0001 68 #undef _MB2 69 #define _MB2 0x0002 70 71 #ifndef NLS 72 extern tcshuc _cmap_lower[], _cmap_upper[]; 73 74 #endif 75 76 #ifndef __QNXNTO__ 77 #undef _QF 78 #define _QF 0x0001 /* '" (Forward quotes) */ 79 #undef _QB 80 #define _QB 0x0002 /* ` (Backquote) */ 81 #undef _SP 82 #define _SP 0x0004 /* space and tab */ 83 #else 84 #undef _XD 85 #define _XD 0x0001 /* As in <ctype.h> */ 86 #undef _UP 87 #define _UP 0x0002 /* As in <ctype.h> */ 88 #undef _SP 89 #define _SP 0x0004 /* As in <ctype.h> */ 90 #endif 91 #undef _NL 92 #define _NL 0x0008 /* \n */ 93 #undef _META 94 #define _META 0x0010 /* lex meta characters, sp #'`";&<>()|\t\n */ 95 #undef _GLOB 96 #define _GLOB 0x0020 /* glob characters, *?{[` */ 97 #undef _ESC 98 #define _ESC 0x0040 /* \ */ 99 #undef _DOL 100 #define _DOL 0x0080 /* $ */ 101 #undef _DIG 102 #define _DIG 0x0100 /* 0-9 */ 103 #undef _LET 104 #define _LET 0x0200 /* a-z, A-Z, _, or locale-specific */ 105 #ifndef __QNXNTO__ 106 #undef _UP 107 #define _UP 0x0400 /* A-Z, or locale-specific */ 108 #else 109 #undef _QF 110 #define _QF 0x0400 /* '" (Forward quotes) */ 111 #endif 112 #undef _DOW 113 #define _DOW 0x0800 /* a-z, or locale-specific */ 114 #ifndef __QNXNTO__ 115 #undef _XD 116 #define _XD 0x1000 /* 0-9, a-f, A-F */ 117 #else 118 #undef _QB 119 #define _QB 0x1000 /* 0-9, a-f, A-F */ 120 #endif 121 #undef _CMD 122 #define _CMD 0x2000 /* lex end of command chars, ;&(|` */ 123 #undef _CTR 124 #define _CTR 0x4000 /* control */ 125 #undef _PUN 126 #define _PUN 0x8000 /* punctuation */ 127 128 #ifdef IS_ASCII 129 # define ASC(ch) (ch) 130 # define CTL_ESC(ch) (ch) 131 #else 132 # ifdef _OSD_POSIX 133 /* "BS2000 OSD" is a POSIX on a main frame using a EBCDIC char set */ 134 # include <ascii_ebcdic.h> 135 # else 136 /* "OS/390 USS" is a POSIX on a main frame using an IBM1047 char set */ 137 # endif 138 extern unsigned short _toascii[256]; 139 extern unsigned short _toebcdic[256]; 140 141 /* mainly for comparisons if (ASC(ch)=='\177')... */ 142 # define ASC(ch) _toascii[(tcshuc)(ch)] 143 144 /* Literal escapes ('\010') must be mapped to EBCDIC, 145 * for C-Escapes ('\b'), the compiler already does it. 146 */ 147 # define CTL_ESC(ch) _toebcdic[(tcshuc)(ch)] 148 #endif /*IS_ASCII*/ 149 150 #ifdef WIDE_STRINGS 151 # define cmap(c, bits) \ 152 (((c) & QUOTE) || (unsigned int)(c) >= 0x0080 ? 0 : \ 153 (_cmap[(tcshuc)ASC(c)] & (bits))) 154 #elif defined(SHORT_STRINGS) && defined(KANJI) 155 # define cmap(c, bits) \ 156 ((((c) & QUOTE) || ((tcshuc)(ASC(c) & 0x80) && adrof(STRnokanji))) ? \ 157 0 : (_cmap[(tcshuc)ASC(c)] & (bits))) 158 #else /* SHORT_STRINGS && KANJI */ 159 # define cmap(c, bits) \ 160 (((c) & QUOTE) ? 0 : (_cmap[(tcshuc)ASC(c)] & (bits))) 161 #endif /* SHORT_STRINGS && KANJI */ 162 163 #define isglob(c) cmap((c), _GLOB) 164 #define isspc(c) cmap((c), _SP) 165 #define ismeta(c) cmap((c), _META) 166 #define iscmdmeta(c) cmap((c), _CMD) 167 #ifdef WIDE_STRINGS 168 #define letter(c) (((c) & QUOTE) ? 0 : \ 169 (iswalpha((c)) || (c) == '_')) 170 #define alnum(c) (((c) & QUOTE) ? 0 : \ 171 (iswalnum((c)) || (c) == '_')) 172 #else 173 #define letter(c) (((Char)(c) & QUOTE) ? 0 : \ 174 ((isalpha((c)) && !(cmap((c), _PUN))) \ 175 || (c) == '_')) 176 #define alnum(c) (((Char)(c) & QUOTE) ? 0 : \ 177 ((isalnum((c)) && !(cmap((c), _PUN))) \ 178 || (c) == '_')) 179 180 #endif 181 182 #if defined(DSPMBYTE) 183 # define IsmbyteU(c) (Ismbyte1((Char)(c))||(Ismbyte2((Char)(c))&&((c)&0200))) 184 #endif 185 186 #ifdef NLS 187 # ifdef WIDE_STRINGS 188 # define Isspace(c) (((c) & QUOTE) ? 0 : iswspace(c)) 189 # define Isdigit(c) (((c) & QUOTE) ? 0 : iswdigit(c)) 190 # define Isalpha(c) (((c) & QUOTE) ? 0 : iswalpha(c)) 191 # define Islower(c) (((c) & QUOTE) ? 0 : iswlower(c)) 192 # define Isupper(c) (((c) & QUOTE) ? 0 : iswupper(c)) 193 # define Tolower(c) (((c) & QUOTE) ? 0 : (wchar_t)towlower(c)) 194 # define Toupper(c) (((c) & QUOTE) ? 0 : (wchar_t)towupper(c)) 195 # define Isxdigit(c) (((c) & QUOTE) ? 0 : iswxdigit(c)) 196 # define Isalnum(c) (((c) & QUOTE) ? 0 : iswalnum(c)) 197 # define Iscntrl(c) (((c) & QUOTE) ? 0 : iswcntrl(c)) 198 # define Isprint(c) (((c) & QUOTE) ? 0 : iswprint(c)) 199 # define Ispunct(c) (((c) & QUOTE) ? 0 : iswpunct(c)) 200 # elif defined (NeXT) 201 # define Isspace(c) (((Char)(c) & QUOTE) ? 0 : NXIsSpace((unsigned) (c))) 202 # define Isdigit(c) (((Char)(c) & QUOTE) ? 0 : NXIsDigit((unsigned) (c))) 203 # define Isalpha(c) (((Char)(c) & QUOTE) ? 0 : NXIsAlpha((unsigned) (c))) 204 # define Islower(c) (((Char)(c) & QUOTE) ? 0 : NXIsLower((unsigned) (c))) 205 # define Isupper(c) (((Char)(c) & QUOTE) ? 0 : NXIsUpper((unsigned) (c))) 206 # define Tolower(c) (((Char)(c) & QUOTE) ? 0 : NXToLower((unsigned) (c))) 207 # define Toupper(c) (((Char)(c) & QUOTE) ? 0 : NXToUpper((unsigned) (c))) 208 # define Isxdigit(c) (((Char)(c) & QUOTE) ? 0 : NXIsXDigit((unsigned) (c))) 209 #if defined(DSPMBYTE) 210 # define IscntrlM(c) (((Char)(c) & QUOTE) ? 0 : NXIsCntrl((unsigned) (c))) 211 # define Iscntrl(c) ( (IscntrlM(c)) && !(_enable_mbdisp&&(IsmbyteU((c)))) ) 212 # define IsprintM(c) (((Char)(c) & QUOTE) ? 0 : NXIsPrint((unsigned) (c))) 213 # define Isprint(c) ( (IsprintM(c)) || (_enable_mbdisp&&(IsmbyteU((c)))) ) 214 #else 215 # define Isalnum(c) (((Char)(c) & QUOTE) ? 0 : NXIsAlNum((unsigned) (c))) 216 # define Iscntrl(c) (((Char)(c) & QUOTE) ? 0 : NXIsCntrl((unsigned) (c))) 217 # define Isprint(c) (((Char)(c) & QUOTE) ? 0 : NXIsPrint((unsigned) (c))) 218 #endif /* !defined(DSPMBYTE) */ 219 # define Ispunct(c) (((Char)(c) & QUOTE) ? 0 : NXIsPunct((unsigned) (c))) 220 # else /* !NeXT */ 221 # ifndef WINNT_NATIVE 222 # define Isspace(c) (((Char)(c) & QUOTE) ? 0 : isspace((tcshuc) (c))) 223 # define Isdigit(c) (((Char)(c) & QUOTE) ? 0 : isdigit((tcshuc) (c))) 224 # define Isalpha(c) (((Char)(c) & QUOTE) ? 0 : isalpha((tcshuc) (c))) 225 # define Islower(c) (((Char)(c) & QUOTE) ? 0 : islower((tcshuc) (c))) 226 # define Isupper(c) (((Char)(c) & QUOTE) ? 0 : isupper((tcshuc) (c))) 227 # define Tolower(c) (((Char)(c) & QUOTE) ? 0 : tolower((tcshuc) (c))) 228 # define Toupper(c) (((Char)(c) & QUOTE) ? 0 : toupper((tcshuc) (c))) 229 # define Isxdigit(c) (((Char)(c) & QUOTE) ? 0 : isxdigit((tcshuc) (c))) 230 # define Isalnum(c) (((Char)(c) & QUOTE) ? 0 : isalnum((tcshuc) (c))) 231 #if defined(DSPMBYTE) 232 # define IscntrlM(c) (((Char)(c) & QUOTE) ? 0 : iscntrl((tcshuc) (c))) 233 # define Iscntrl(c) ( (IscntrlM(c)) && !(_enable_mbdisp&&(IsmbyteU((c)))) ) 234 #else 235 # define Iscntrl(c) (((Char)(c) & QUOTE) ? 0 : iscntrl((tcshuc) (c))) 236 #endif /* !defined(DSPMBYTE) */ 237 # if SOLARIS2 == 24 238 /* 239 * From <casper@fwi.uva.nl> Casper Dik: 240 * In Solaris 2.4, isprint('\t') returns true after setlocal(LC_ALL,""). 241 * This breaks commandline editing when you include tabs. 242 * (This is in the en_US locale). 243 */ 244 #if defined(DSPMBYTE) 245 # define IsprintM(c) (((Char)(c) & QUOTE) ? 0 : \ 246 (isprint((tcshuc) (c)) && (c) != '\t')) 247 #else 248 # define Isprint(c) (((Char)(c) & QUOTE) ? 0 : \ 249 (isprint((tcshuc) (c)) && (c) != '\t')) 250 #endif /* !defined(DSPMBYTE) */ 251 # else 252 #if defined(DSPMBYTE) 253 # define IsprintM(c) (((Char)(c) & QUOTE) ? 0 : isprint((tcshuc) (c))) 254 #else 255 # define Isprint(c) (((Char)(c) & QUOTE) ? 0 : isprint((tcshuc) (c))) 256 #endif /* !defined(DSPMBYTE) */ 257 # endif /* SOLARIS2 == 24 */ 258 #if defined(DSPMBYTE) 259 # define Isprint(c) ( (IsprintM(c)) || (_enable_mbdisp&&(IsmbyteU((c)))) ) 260 #endif /* !defined(DSPMBYTE) */ 261 # define Ispunct(c) (((Char)(c) & QUOTE) ? 0 : ispunct((tcshuc) (c))) 262 # else /* WINNT_NATIVE */ 263 # define Isspace(c) (((Char)(c) & QUOTE) ? 0 : isspace( oem_it((tcshuc)(c)))) 264 # define Isdigit(c) (((Char)(c) & QUOTE) ? 0 : isdigit( oem_it((tcshuc)(c)))) 265 # define Isalpha(c) (((Char)(c) & QUOTE) ? 0 : isalpha( oem_it((tcshuc)(c)))) 266 # define Islower(c) (((Char)(c) & QUOTE) ? 0 : islower( oem_it((tcshuc)(c)))) 267 # define Isupper(c) (((Char)(c) & QUOTE) ? 0 : isupper( oem_it((tcshuc)(c)))) 268 # define Tolower(c) (((Char)(c) & QUOTE) ? 0 : tolower( oem_it((tcshuc)(c)))) 269 # define Toupper(c) (((Char)(c) & QUOTE) ? 0 : toupper( oem_it((tcshuc)(c)))) 270 # define Isxdigit(c)(((Char)(c) & QUOTE) ? 0 : isxdigit(oem_it((tcshuc)(c)))) 271 # define Isalnum(c) (((Char)(c) & QUOTE) ? 0 : isalnum( oem_it((tcshuc)(c)))) 272 # define Ispunct(c) (((Char)(c) & QUOTE) ? 0 : ispunct( oem_it((tcshuc)(c)))) 273 #if defined(DSPMBYTE) 274 # define IscntrlM(c) (((Char)(c) & QUOTE) ? 0 : iscntrl( oem_it((tcshuc)(c)))) 275 # define Iscntrl(c) ( (IscntrlM(c)) && !(_enable_mbdisp&&(IsmbyteU((c)))) ) 276 # define IsprintM(c) (((Char)(c) & QUOTE) ? 0 : isprint( oem_it((tcshuc)(c)))) 277 # define Isprint(c) ( (IsprintM(c)) || (_enable_mbdisp&&(IsmbyteU((c)))) ) 278 #else 279 # define Iscntrl(c) (((Char)(c) & QUOTE) ? 0 : iscntrl( oem_it((tcshuc)(c)))) 280 # define Isprint(c) (((Char)(c) & QUOTE) ? 0 : isprint( oem_it((tcshuc)(c)))) 281 #endif /* !defined(DSPMBYTE) */ 282 # endif /* WINNT_NATIVE */ 283 # endif /* !NeXT */ 284 #else /* !NLS */ 285 # define Isspace(c) cmap((c), _SP|_NL) 286 # define Isdigit(c) cmap((c), _DIG) 287 # define Isalpha(c) (cmap((c),_LET) && !(((c) & META) && AsciiOnly)) 288 # define Islower(c) (cmap((c),_DOW) && !(((c) & META) && AsciiOnly)) 289 # define Isupper(c) (cmap((c), _UP) && !(((c) & META) && AsciiOnly)) 290 # define Tolower(c) (_cmap_lower[ASC(c)]) 291 # define Toupper(c) (_cmap_upper[ASC(c)]) 292 # define Isxdigit(c) cmap((c), _XD) 293 # define Isalnum(c) (cmap((c), _DIG|_LET) && !(((Char)(c) & META) && AsciiOnly)) 294 #if defined(DSPMBYTE) 295 # define IscntrlM(c) (cmap((c),_CTR) && !(((c) & META) && AsciiOnly)) 296 # define Iscntrl(c) ( (IscntrlM(c)) && !(_enable_mbdisp&&(IsmbyteU((c)))) ) 297 # define IsprintM(c) (!cmap((c),_CTR) && !(((c) & META) && AsciiOnly)) 298 # define Isprint(c) ( (IsprintM(c)) || (_enable_mbdisp&&(IsmbyteU((c)))) ) 299 #else 300 # define Iscntrl(c) (cmap((c),_CTR) && !(((c) & META) && AsciiOnly)) 301 # define Isprint(c) (!cmap((c),_CTR) && !(((c) & META) && AsciiOnly)) 302 #endif /* !defined(DSPMBYTE) */ 303 # define Ispunct(c) (cmap((c),_PUN) && !(((c) & META) && AsciiOnly)) 304 305 #endif /* !NLS */ 306 307 #if defined(DSPMBYTE) 308 # define Ismbyte1(c) ((_mbmap[(c) & 0377] & _MB1) ? 1 : 0) 309 # define Ismbyte2(c) ((_mbmap[(c) & 0377] & _MB2) ? 1 : 0) 310 #endif 311 312 #endif /* _h_sh_char */ 313