1/* $Header: /p/tcsh/cvsroot/tcsh/sh.char.h,v 3.35 2008/09/25 14:41:34 christos Exp $ */
2/*
3 * sh.char.h: Table for spotting special characters quickly
4 * 	      Makes for very obscure but efficient coding.
5 */
6/*-
7 * Copyright (c) 1980, 1991 The Regents of the University of California.
8 * All rights reserved.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 *    notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 *    notice, this list of conditions and the following disclaimer in the
17 *    documentation and/or other materials provided with the distribution.
18 * 3. Neither the name of the University nor the names of its contributors
19 *    may be used to endorse or promote products derived from this software
20 *    without specific prior written permission.
21 *
22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 * SUCH DAMAGE.
33 */
34#ifndef _h_sh_char
35#define _h_sh_char
36#if defined(NeXT) && defined(NLS)
37# include <appkit/NXCType.h>
38#else
39# include <ctype.h>
40# ifdef WIDE_STRINGS
41#  ifdef HAVE_WCTYPE_H
42#   include <wctype.h>
43#  else
44#   include <wchar.h>
45#  endif
46# endif
47#endif
48
49typedef unsigned char tcshuc;
50#ifdef _MINIX
51# undef _SP
52#endif /* _MINIX */
53extern unsigned short _cmap[];
54#if defined(DSPMBYTE)
55# define CHECK_MBYTEVAR	STRdspmbyte
56#endif
57extern unsigned short _cmap_c[];
58extern unsigned short _cmap_mbyte[];
59extern short _enable_mbdisp;
60extern unsigned short _mbmap[];
61extern unsigned short _mbmap_euc[];
62extern unsigned short _mbmap_sjis[];
63extern unsigned short _mbmap_big5[];
64extern unsigned short _mbmap_utf8[];
65/* VARIABLE Check str */
66/* same compiler require #define even not define DSPMBYTE */
67#undef	_MB1
68#define _MB1	0x0001
69#undef	_MB2
70#define _MB2	0x0002
71
72#ifndef NLS
73extern tcshuc _cmap_lower[], _cmap_upper[];
74
75#endif
76
77#ifndef __QNXNTO__
78#undef	_QF
79#define	_QF	0x0001		/* '" (Forward quotes) */
80#undef	_QB
81#define	_QB	0x0002		/* ` (Backquote) */
82#undef	_SP
83#define	_SP	0x0004		/* space and tab */
84#else
85#undef	_XD
86#define	_XD	0x0001		/* As in <ctype.h> */
87#undef	_UP
88#define	_UP	0x0002		/* As in <ctype.h> */
89#undef	_SP
90#define	_SP	0x0004		/* As in <ctype.h> */
91#endif
92#undef	_NL
93#define	_NL	0x0008		/* \n */
94#undef	_META
95#define	_META	0x0010		/* lex meta characters, sp #'`";&<>()|\t\n */
96#undef	_GLOB
97#define	_GLOB	0x0020		/* glob characters, *?{[` */
98#undef	_ESC
99#define	_ESC	0x0040		/* \ */
100#undef	_DOL
101#define	_DOL	0x0080		/* $ */
102#undef	_DIG
103#define	_DIG  	0x0100		/* 0-9 */
104#undef	_LET
105#define	_LET  	0x0200		/* a-z, A-Z, _, or locale-specific */
106#ifndef __QNXNTO__
107#undef	_UP
108#define	_UP   	0x0400		/* A-Z, or locale-specific */
109#else
110#undef	_QF
111#define	_QF	0x0400		/* '" (Forward quotes) */
112#endif
113#undef	_DOW
114#define	_DOW  	0x0800		/* a-z, or locale-specific */
115#ifndef __QNXNTO__
116#undef	_XD
117#define	_XD 	0x1000		/* 0-9, a-f, A-F */
118#else
119#undef	_QB
120#define	_QB	0x1000		/* 0-9, a-f, A-F */
121#endif
122#undef	_CMD
123#define	_CMD	0x2000		/* lex end of command chars, ;&(|` */
124#undef	_CTR
125#define _CTR	0x4000		/* control */
126#undef	_PUN
127#define _PUN	0x8000		/* punctuation */
128
129#ifdef IS_ASCII
130# define ASC(ch) (ch)
131# define CTL_ESC(ch) (ch)
132#else
133# ifdef _OSD_POSIX
134/* "BS2000 OSD" is a POSIX on a main frame using a EBCDIC char set */
135#   include <ascii_ebcdic.h>
136# else
137/* "OS/390 USS" is a POSIX on a main frame using an IBM1047 char set */
138# endif
139  extern unsigned short _toascii[256];
140  extern unsigned short _toebcdic[256];
141
142/* mainly for comparisons if (ASC(ch)=='\177')... */
143#  define ASC(ch)     _toascii[(tcshuc)(ch)]
144
145/* Literal escapes ('\010') must be mapped to EBCDIC,
146 * for C-Escapes   ('\b'), the compiler already does it.
147 */
148#  define CTL_ESC(ch) _toebcdic[(tcshuc)(ch)]
149#endif /*IS_ASCII*/
150
151#ifdef WIDE_STRINGS
152# define cmap(c, bits)	\
153	(((c) < 0) ? 0 : \
154	((c) & QUOTE) || (c) >= 0x0080 ? 0 : (_cmap[(tcshuc)ASC(c)] & (bits)))
155#elif defined(SHORT_STRINGS) && defined(KANJI)
156#  define cmap(c, bits)	\
157	(((c) < 0) ? 0 : \
158	(((c) & QUOTE) || ((ASC(c) & 0x80) && adrof(STRnokanji))) ? \
159	0 : (_cmap[(tcshuc)ASC(c)] & (bits)))
160#else /* SHORT_STRINGS && KANJI */
161# define cmap(c, bits)	\
162	(((c) < 0) ? 0 : \
163	((c) & QUOTE) ? 0 : (_cmap[(tcshuc)ASC(c)] & (bits)))
164#endif /* SHORT_STRINGS && KANJI */
165
166#define isglob(c)	cmap((c), _GLOB)
167#define isspc(c)	cmap((c), _SP)
168#define ismeta(c)	cmap((c), _META)
169#define iscmdmeta(c)	cmap((c), _CMD)
170#ifdef WIDE_STRINGS
171#define letter(c)	(((c) & QUOTE) ? 0 :  \
172			 (iswalpha((tcshuc) (c)) || (c) == '_'))
173#define alnum(c)	(((c) & QUOTE) ? 0 :  \
174		         (iswalnum((tcshuc) (c)) || (c) == '_'))
175#else
176#define letter(c)	(((Char)(c) & QUOTE) ? 0 :  \
177			 ((isalpha((tcshuc) (c)) && !(cmap((c), _PUN))) \
178			  || (c) == '_'))
179#define alnum(c)	(((Char)(c) & QUOTE) ? 0 :  \
180		         ((isalnum((tcshuc) (c)) && !(cmap((c), _PUN))) \
181			  || (c) == '_'))
182
183#endif
184
185#if defined(DSPMBYTE)
186# define IsmbyteU(c)	(Ismbyte1((Char)(c))||(Ismbyte2((Char)(c))&&((c)&0200)))
187#endif
188
189#ifdef NLS
190# ifdef WIDE_STRINGS
191#  define Isspace(c)	(((c) & QUOTE) ? 0 : iswspace(c))
192#  define Isdigit(c)	(((c) & QUOTE) ? 0 : iswdigit(c))
193#  define Isalpha(c)	(((c) & QUOTE) ? 0 : iswalpha(c))
194#  define Islower(c)	(((c) & QUOTE) ? 0 : iswlower(c))
195#  define Isupper(c)	(((c) & QUOTE) ? 0 : iswupper(c))
196#  define Tolower(c) 	(((c) & QUOTE) ? 0 : (wchar_t)towlower(c))
197#  define Toupper(c) 	(((c) & QUOTE) ? 0 : (wchar_t)towupper(c))
198#  define Isxdigit(c)	(((c) & QUOTE) ? 0 : iswxdigit(c))
199#  define Isalnum(c)	(((c) & QUOTE) ? 0 : iswalnum(c))
200#  define Iscntrl(c) 	(((c) & QUOTE) ? 0 : iswcntrl(c))
201#  define Isprint(c) 	(((c) & QUOTE) ? 0 : iswprint(c))
202#  define Ispunct(c) 	(((c) & QUOTE) ? 0 : iswpunct(c))
203# elif defined (NeXT)
204#  define Isspace(c)	(((Char)(c) & QUOTE) ? 0 : NXIsSpace((unsigned) (c)))
205#  define Isdigit(c)	(((Char)(c) & QUOTE) ? 0 : NXIsDigit((unsigned) (c)))
206#  define Isalpha(c)	(((Char)(c) & QUOTE) ? 0 : NXIsAlpha((unsigned) (c)))
207#  define Islower(c)	(((Char)(c) & QUOTE) ? 0 : NXIsLower((unsigned) (c)))
208#  define Isupper(c)	(((Char)(c) & QUOTE) ? 0 : NXIsUpper((unsigned) (c)))
209#  define Tolower(c) 	(((Char)(c) & QUOTE) ? 0 : NXToLower((unsigned) (c)))
210#  define Toupper(c) 	(((Char)(c) & QUOTE) ? 0 : NXToUpper((unsigned) (c)))
211#  define Isxdigit(c)	(((Char)(c) & QUOTE) ? 0 : NXIsXDigit((unsigned) (c)))
212#if defined(DSPMBYTE)
213#  define IscntrlM(c) 	(((Char)(c) & QUOTE) ? 0 : NXIsCntrl((unsigned) (c)))
214#  define Iscntrl(c)	( (IscntrlM(c)) && !(_enable_mbdisp&&(IsmbyteU((c)))) )
215#  define IsprintM(c) 	(((Char)(c) & QUOTE) ? 0 : NXIsPrint((unsigned) (c)))
216#  define Isprint(c)	( (IsprintM(c)) || (_enable_mbdisp&&(IsmbyteU((c)))) )
217#else
218#  define Isalnum(c)	(((Char)(c) & QUOTE) ? 0 : NXIsAlNum((unsigned) (c)))
219#  define Iscntrl(c) 	(((Char)(c) & QUOTE) ? 0 : NXIsCntrl((unsigned) (c)))
220#  define Isprint(c) 	(((Char)(c) & QUOTE) ? 0 : NXIsPrint((unsigned) (c)))
221#endif /* !defined(DSPMBYTE) */
222#  define Ispunct(c) 	(((Char)(c) & QUOTE) ? 0 : NXIsPunct((unsigned) (c)))
223# else /* !NeXT */
224#  ifndef WINNT_NATIVE
225#   define Isspace(c)	(((Char)(c) & QUOTE) ? 0 : isspace((tcshuc) (c)))
226#   define Isdigit(c)	(((Char)(c) & QUOTE) ? 0 : isdigit((tcshuc) (c)))
227#   define Isalpha(c)	(((Char)(c) & QUOTE) ? 0 : isalpha((tcshuc) (c)))
228#   define Islower(c)	(((Char)(c) & QUOTE) ? 0 : islower((tcshuc) (c)))
229#   define Isupper(c)	(((Char)(c) & QUOTE) ? 0 : isupper((tcshuc) (c)))
230#   define Tolower(c) 	(((Char)(c) & QUOTE) ? 0 : tolower((tcshuc) (c)))
231#   define Toupper(c) 	(((Char)(c) & QUOTE) ? 0 : toupper((tcshuc) (c)))
232#   define Isxdigit(c)	(((Char)(c) & QUOTE) ? 0 : isxdigit((tcshuc) (c)))
233#   define Isalnum(c)	(((Char)(c) & QUOTE) ? 0 : isalnum((tcshuc) (c)))
234#if defined(DSPMBYTE)
235#   define IscntrlM(c) 	(((Char)(c) & QUOTE) ? 0 : iscntrl((tcshuc) (c)))
236#   define Iscntrl(c)	( (IscntrlM(c)) && !(_enable_mbdisp&&(IsmbyteU((c)))) )
237#else
238#   define Iscntrl(c) 	(((Char)(c) & QUOTE) ? 0 : iscntrl((tcshuc) (c)))
239#endif /* !defined(DSPMBYTE) */
240#   if SOLARIS2 == 24
241    /*
242     * From <casper@fwi.uva.nl> Casper Dik:
243     * In Solaris 2.4, isprint('\t') returns true after setlocal(LC_ALL,"").
244     * This breaks commandline editing when you include tabs.
245     * (This is in the en_US locale).
246     */
247#if defined(DSPMBYTE)
248#    define IsprintM(c) 	(((Char)(c) & QUOTE) ? 0 : \
249				(isprint((tcshuc) (c)) && (c) != '\t'))
250#else
251#    define Isprint(c) 	(((Char)(c) & QUOTE) ? 0 : \
252				(isprint((tcshuc) (c)) && (c) != '\t'))
253#endif /* !defined(DSPMBYTE) */
254#   else
255#if defined(DSPMBYTE)
256#    define IsprintM(c) (((Char)(c) & QUOTE) ? 0 : isprint((tcshuc) (c)))
257#else
258#    define Isprint(c) 	(((Char)(c) & QUOTE) ? 0 : isprint((tcshuc) (c)))
259#endif /* !defined(DSPMBYTE) */
260#   endif /* SOLARIS2 == 24 */
261#if defined(DSPMBYTE)
262#   define Isprint(c)	( (IsprintM(c)) || (_enable_mbdisp&&(IsmbyteU((c)))) )
263#endif /* !defined(DSPMBYTE) */
264#    define Ispunct(c) 	(((Char)(c) & QUOTE) ? 0 : ispunct((tcshuc) (c)))
265#  else /* WINNT_NATIVE */
266#   define Isspace(c) (((Char)(c) & QUOTE) ? 0 : isspace( oem_it((tcshuc)(c))))
267#   define Isdigit(c) (((Char)(c) & QUOTE) ? 0 : isdigit( oem_it((tcshuc)(c))))
268#   define Isalpha(c) (((Char)(c) & QUOTE) ? 0 : isalpha( oem_it((tcshuc)(c))))
269#   define Islower(c) (((Char)(c) & QUOTE) ? 0 : islower( oem_it((tcshuc)(c))))
270#   define Isupper(c) (((Char)(c) & QUOTE) ? 0 : isupper( oem_it((tcshuc)(c))))
271#   define Tolower(c) (((Char)(c) & QUOTE) ? 0 : tolower( oem_it((tcshuc)(c))))
272#   define Toupper(c) (((Char)(c) & QUOTE) ? 0 : toupper( oem_it((tcshuc)(c))))
273#   define Isxdigit(c)(((Char)(c) & QUOTE) ? 0 : isxdigit(oem_it((tcshuc)(c))))
274#   define Isalnum(c) (((Char)(c) & QUOTE) ? 0 : isalnum( oem_it((tcshuc)(c))))
275#   define Ispunct(c) (((Char)(c) & QUOTE) ? 0 : ispunct( oem_it((tcshuc)(c))))
276#if defined(DSPMBYTE)
277#   define IscntrlM(c) (((Char)(c) & QUOTE) ? 0 : iscntrl( oem_it((tcshuc)(c))))
278#   define Iscntrl(c)	( (IscntrlM(c)) && !(_enable_mbdisp&&(IsmbyteU((c)))) )
279#   define IsprintM(c) (((Char)(c) & QUOTE) ? 0 : isprint( oem_it((tcshuc)(c))))
280#   define Isprint(c)	( (IsprintM(c)) || (_enable_mbdisp&&(IsmbyteU((c)))) )
281#else
282#   define Iscntrl(c) (((Char)(c) & QUOTE) ? 0 : iscntrl( oem_it((tcshuc)(c))))
283#   define Isprint(c) (((Char)(c) & QUOTE) ? 0 : isprint( oem_it((tcshuc)(c))))
284#endif /* !defined(DSPMBYTE) */
285#  endif /* WINNT_NATIVE */
286# endif /* !NeXT */
287#else /* !NLS */
288# define Isspace(c)	cmap((c), _SP|_NL)
289# define Isdigit(c)	cmap((c), _DIG)
290# define Isalpha(c)	(cmap((c),_LET) && !(((c) & META) && AsciiOnly))
291# define Islower(c)	(cmap((c),_DOW) && !(((c) & META) && AsciiOnly))
292# define Isupper(c)	(cmap((c), _UP) && !(((c) & META) && AsciiOnly))
293# define Tolower(c)	(_cmap_lower[ASC(c)])
294# define Toupper(c)	(_cmap_upper[ASC(c)])
295# define Isxdigit(c)	cmap((c), _XD)
296# define Isalnum(c)	(cmap((c), _DIG|_LET) && !(((Char)(c) & META) && AsciiOnly))
297#if defined(DSPMBYTE)
298# define IscntrlM(c)	(cmap((c),_CTR) && !(((c) & META) && AsciiOnly))
299# define Iscntrl(c)	( (IscntrlM(c)) && !(_enable_mbdisp&&(IsmbyteU((c)))) )
300# define IsprintM(c)	(!cmap((c),_CTR) && !(((c) & META) && AsciiOnly))
301# define Isprint(c)	( (IsprintM(c)) || (_enable_mbdisp&&(IsmbyteU((c)))) )
302#else
303# define Iscntrl(c)	(cmap((c),_CTR) && !(((c) & META) && AsciiOnly))
304# define Isprint(c)	(!cmap((c),_CTR) && !(((c) & META) && AsciiOnly))
305#endif /* !defined(DSPMBYTE) */
306# define Ispunct(c)	(cmap((c),_PUN) && !(((c) & META) && AsciiOnly))
307
308#endif /* !NLS */
309
310#if defined(DSPMBYTE)
311# define Ismbyte1(c)	((_mbmap[(c) & 0377] & _MB1) ? 1 : 0)
312# define Ismbyte2(c)	((_mbmap[(c) & 0377] & _MB2) ? 1 : 0)
313#endif
314
315#endif /* _h_sh_char */
316