xref: /titanic_44/usr/src/cmd/tr/str.c (revision c10c16dec587a0662068f6e2991c29ed3a9db943)
1 /*
2  * Copyright (c) 1991, 1993
3  *	The Regents of the University of California.  All rights reserved.
4  *
5  * Redistribution and use in source and binary forms, with or without
6  * modification, are permitted provided that the following conditions
7  * are met:
8  * 1. Redistributions of source code must retain the above copyright
9  *    notice, this list of conditions and the following disclaimer.
10  * 2. Redistributions in binary form must reproduce the above copyright
11  *    notice, this list of conditions and the following disclaimer in the
12  *    documentation and/or other materials provided with the distribution.
13  * 3. All advertising materials mentioning features or use of this software
14  *    must display the following acknowledgement:
15  *	This product includes software developed by the University of
16  *	California, Berkeley and its contributors.
17  * 4. 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 
34 #include <sys/types.h>
35 
36 #include <ctype.h>
37 #include <err.h>
38 #include <errno.h>
39 #include <stddef.h>
40 #include <stdio.h>
41 #include <stdlib.h>
42 #include <string.h>
43 #include <strings.h>
44 #include <wchar.h>
45 #include <wctype.h>
46 
47 #include "extern.h"
48 
49 static int	backslash(STR *, int *);
50 static int	bracket(STR *);
51 static void	genclass(STR *);
52 static void	genequiv(STR *);
53 static int	genrange(STR *, int);
54 static void	genseq(STR *);
55 
56 wint_t
57 next(s)
58 	STR *s;
59 {
60 	int is_octal;
61 	wint_t ch;
62 	wchar_t wch;
63 	size_t clen;
64 
65 	switch (s->state) {
66 	case EOS:
67 		return (0);
68 	case INFINITE:
69 		return (1);
70 	case NORMAL:
71 		switch (*s->str) {
72 		case '\0':
73 			s->state = EOS;
74 			return (0);
75 		case '\\':
76 			s->lastch = backslash(s, &is_octal);
77 			break;
78 		case '[':
79 			if (bracket(s))
80 				return (next(s));
81 			/* FALLTHROUGH */
82 		default:
83 			clen = mbrtowc(&wch, s->str, MB_LEN_MAX, NULL);
84 			if (clen == (size_t)-1 || clen == (size_t)-2 ||
85 			    clen == 0) {
86 				(void) fprintf(stderr, "Illegal seqeunce.\n");
87 				exit(1);
88 			}
89 			is_octal = 0;
90 			s->lastch = wch;
91 			s->str += clen;
92 			break;
93 		}
94 
95 		/* We can start a range at any time. */
96 		if (s->str[0] == '-' && genrange(s, is_octal))
97 			return (next(s));
98 		return (1);
99 	case RANGE:
100 		if (s->cnt-- == 0) {
101 			s->state = NORMAL;
102 			return (next(s));
103 		}
104 		++s->lastch;
105 		return (1);
106 	case SEQUENCE:
107 		if (s->cnt-- == 0) {
108 			s->state = NORMAL;
109 			return (next(s));
110 		}
111 		return (1);
112 	case CCLASS:
113 	case CCLASS_UPPER:
114 	case CCLASS_LOWER:
115 		s->cnt++;
116 		ch = nextwctype(s->lastch, s->cclass);
117 		if (ch == -1) {
118 			s->state = NORMAL;
119 			return (next(s));
120 		}
121 		s->lastch = ch;
122 		return (1);
123 	case SET:
124 		if ((ch = s->set[s->cnt++]) == OOBCH) {
125 			s->state = NORMAL;
126 			return (next(s));
127 		}
128 		s->lastch = ch;
129 		return (1);
130 	default:
131 		return (0);
132 	}
133 	/* NOTREACHED */
134 }
135 
136 static int
137 bracket(s)
138 	STR *s;
139 {
140 	char *p;
141 
142 	switch (s->str[1]) {
143 	case ':':				/* "[:class:]" */
144 		if ((p = strchr(s->str + 2, ']')) == NULL)
145 			return (0);
146 		if (*(p - 1) != ':' || p - s->str < 4)
147 			goto repeat;
148 		*(p - 1) = '\0';
149 		s->str += 2;
150 		genclass(s);
151 		s->str = p + 1;
152 		return (1);
153 	case '=':				/* "[=equiv=]" */
154 		if ((p = strchr(s->str + 3, ']')) == NULL)
155 			return (0);
156 		if (*(p - 1) != '=' || p - s->str < 4)
157 			goto repeat;
158 		s->str += 2;
159 		genequiv(s);
160 		return (1);
161 	default:				/* "[\###*n]" or "[#*n]" */
162 	repeat:
163 		if ((p = strpbrk(s->str + 2, "*]")) == NULL)
164 			return (0);
165 		if (p[0] != '*' || index(p, ']') == NULL)
166 			return (0);
167 		s->str += 1;
168 		genseq(s);
169 		return (1);
170 	}
171 	/* NOTREACHED */
172 }
173 
174 static void
175 genclass(s)
176 	STR *s;
177 {
178 
179 	if ((s->cclass = wctype(s->str)) == 0)
180 		errx(1, "unknown class %s", s->str);
181 	s->cnt = 0;
182 	s->lastch = -1;		/* incremented before check in next() */
183 	if (strcmp(s->str, "upper") == 0)
184 		s->state = CCLASS_UPPER;
185 	else if (strcmp(s->str, "lower") == 0)
186 		s->state = CCLASS_LOWER;
187 	else
188 		s->state = CCLASS;
189 }
190 
191 static void
192 genequiv(s)
193 	STR *s;
194 {
195 	int i, p, pri;
196 	char src[2], dst[3];
197 	size_t clen;
198 	wchar_t wc;
199 
200 	if (*s->str == '\\') {
201 		s->equiv[0] = backslash(s, NULL);
202 		if (*s->str != '=')
203 			errx(1, "misplaced equivalence equals sign");
204 		s->str += 2;
205 	} else {
206 		clen = mbrtowc(&wc, s->str, MB_LEN_MAX, NULL);
207 		if (clen == (size_t)-1 || clen == (size_t)-2 || clen == 0) {
208 			errno = EILSEQ;
209 			err(1, NULL);
210 		}
211 		s->equiv[0] = wc;
212 		if (s->str[clen] != '=')
213 			errx(1, "misplaced equivalence equals sign");
214 		s->str += clen + 2;
215 	}
216 
217 	/*
218 	 * Calculate the set of all characters in the same equivalence class
219 	 * as the specified character (they will have the same primary
220 	 * collation weights).
221 	 * XXX Knows too much about how strxfrm() is implemented. Assumes
222 	 * it fills the string with primary collation weight bytes. Only one-
223 	 * to-one mappings are supported.
224 	 * XXX Equivalence classes not supported in multibyte locales.
225 	 */
226 	src[0] = (char)s->equiv[0];
227 	src[1] = '\0';
228 	if (MB_CUR_MAX == 1 && strxfrm(dst, src, sizeof (dst)) == 1) {
229 		pri = (unsigned char)*dst;
230 		for (p = 1, i = 1; i < NCHARS_SB; i++) {
231 			*src = i;
232 			if (strxfrm(dst, src, sizeof (dst)) == 1 && pri &&
233 			    pri == (unsigned char)*dst)
234 				s->equiv[p++] = i;
235 		}
236 		s->equiv[p] = OOBCH;
237 	}
238 
239 	s->cnt = 0;
240 	s->state = SET;
241 	s->set = s->equiv;
242 }
243 
244 static int
245 genrange(STR *s, int was_octal)
246 {
247 	int stopval, octal;
248 	char *savestart;
249 	int n, cnt, *p;
250 	size_t clen;
251 	wchar_t wc;
252 
253 	octal = 0;
254 	savestart = s->str;
255 	if (*++s->str == '\\')
256 		stopval = backslash(s, &octal);
257 	else {
258 		clen = mbrtowc(&wc, s->str, MB_LEN_MAX, NULL);
259 		if (clen == (size_t)-1 || clen == (size_t)-2) {
260 			errno = EILSEQ;
261 			err(1, NULL);
262 		}
263 		stopval = wc;
264 		s->str += clen;
265 	}
266 	/*
267 	 * XXX Characters are not ordered according to collating sequence in
268 	 * multibyte locales.
269 	 */
270 	if (octal || was_octal || MB_CUR_MAX > 1) {
271 		if (stopval < s->lastch) {
272 			s->str = savestart;
273 			return (0);
274 		}
275 		s->cnt = stopval - s->lastch + 1;
276 		s->state = RANGE;
277 		--s->lastch;
278 		return (1);
279 	}
280 	if (charcoll((const void *)&stopval, (const void *)&(s->lastch)) < 0) {
281 		s->str = savestart;
282 		return (0);
283 	}
284 	p = malloc((NCHARS_SB + 1) * sizeof (int));
285 	if ((s->set = (void *)p) == NULL)
286 		err(1, "genrange() malloc");
287 	for (cnt = 0; cnt < NCHARS_SB; cnt++)
288 		if (charcoll((const void *)&cnt, (const void *)&(s->lastch)) >=
289 		    0 &&
290 		    charcoll((const void *)&cnt, (const void *)&stopval) <= 0)
291 			*p++ = cnt;
292 	*p = OOBCH;
293 	n = (int *)p - (int *)s->set;
294 
295 	s->cnt = 0;
296 	s->state = SET;
297 	if (n > 1)
298 		qsort(s->set, n, sizeof (*(s->set)), charcoll);
299 	return (1);
300 }
301 
302 static void
303 genseq(s)
304 	STR *s;
305 {
306 	char *ep;
307 	wchar_t wc;
308 	size_t clen;
309 
310 	if (s->which == STRING1)
311 		errx(1, "sequences only valid in string2");
312 
313 	if (*s->str == '\\')
314 		s->lastch = backslash(s, NULL);
315 	else {
316 		clen = mbrtowc(&wc, s->str, MB_LEN_MAX, NULL);
317 		if (clen == (size_t)-1 || clen == (size_t)-2) {
318 			errno = EILSEQ;
319 			err(1, NULL);
320 		}
321 		s->lastch = wc;
322 		s->str += clen;
323 	}
324 	if (*s->str != '*')
325 		errx(1, "misplaced sequence asterisk");
326 
327 	switch (*++s->str) {
328 	case '\\':
329 		s->cnt = backslash(s, NULL);
330 		break;
331 	case ']':
332 		s->cnt = 0;
333 		++s->str;
334 		break;
335 	default:
336 		if (isdigit((uchar_t)*s->str)) {
337 			s->cnt = strtol(s->str, &ep, 0);
338 			if (*ep == ']') {
339 				s->str = ep + 1;
340 				break;
341 			}
342 		}
343 		errx(1, "illegal sequence count");
344 		/* NOTREACHED */
345 	}
346 
347 	s->state = s->cnt ? SEQUENCE : INFINITE;
348 }
349 
350 /*
351  * Translate \??? into a character.  Up to 3 octal digits, if no digits either
352  * an escape code or a literal character.
353  */
354 static int
355 backslash(STR *s, int *is_octal)
356 {
357 	int ch, cnt, val;
358 
359 	if (is_octal != NULL)
360 		*is_octal = 0;
361 	for (cnt = val = 0; ; ) {
362 		ch = (uchar_t)*++s->str;
363 		if (!isdigit(ch) || ch > '7')
364 			break;
365 		val = val * 8 + ch - '0';
366 		if (++cnt == 3) {
367 			++s->str;
368 			break;
369 		}
370 	}
371 	if (cnt) {
372 		if (is_octal != NULL)
373 			*is_octal = 1;
374 		return (val);
375 	}
376 	if (ch != '\0')
377 		++s->str;
378 	switch (ch) {
379 	case 'a':			/* escape characters */
380 		return ('\7');
381 	case 'b':
382 		return ('\b');
383 	case 'f':
384 		return ('\f');
385 	case 'n':
386 		return ('\n');
387 	case 'r':
388 		return ('\r');
389 	case 't':
390 		return ('\t');
391 	case 'v':
392 		return ('\13');
393 	case '\0':			/*  \" -> \ */
394 		s->state = EOS;
395 		return ('\\');
396 	default:			/* \x" -> x */
397 		return (ch);
398 	}
399 }
400