xref: /illumos-gate/usr/src/cmd/tr/str.c (revision 87c723434df4be9c0f7ef119567cb6e2da776a36)
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 				errx(1, "illegal sequence");
87 			is_octal = 0;
88 			s->lastch = wch;
89 			s->str += clen;
90 			break;
91 		}
92 
93 		/* We can start a range at any time. */
94 		if (s->str[0] == '-' && genrange(s, is_octal))
95 			return (next(s));
96 		return (1);
97 	case RANGE:
98 		if (s->cnt-- == 0) {
99 			s->state = NORMAL;
100 			return (next(s));
101 		}
102 		++s->lastch;
103 		return (1);
104 	case SEQUENCE:
105 		if (s->cnt-- == 0) {
106 			s->state = NORMAL;
107 			return (next(s));
108 		}
109 		return (1);
110 	case CCLASS:
111 	case CCLASS_UPPER:
112 	case CCLASS_LOWER:
113 		s->cnt++;
114 		ch = nextwctype(s->lastch, s->cclass);
115 		if (ch == -1) {
116 			s->state = NORMAL;
117 			return (next(s));
118 		}
119 		s->lastch = ch;
120 		return (1);
121 	case SET:
122 		if ((ch = s->set[s->cnt++]) == OOBCH) {
123 			s->state = NORMAL;
124 			return (next(s));
125 		}
126 		s->lastch = ch;
127 		return (1);
128 	default:
129 		return (0);
130 	}
131 	/* NOTREACHED */
132 }
133 
134 static int
135 bracket(s)
136 	STR *s;
137 {
138 	char *p;
139 
140 	switch (s->str[1]) {
141 	case ':':				/* "[:class:]" */
142 		if ((p = strchr(s->str + 2, ']')) == NULL)
143 			return (0);
144 		if (*(p - 1) != ':' || p - s->str < 4)
145 			goto repeat;
146 		*(p - 1) = '\0';
147 		s->str += 2;
148 		genclass(s);
149 		s->str = p + 1;
150 		return (1);
151 	case '=':				/* "[=equiv=]" */
152 		if ((p = strchr(s->str + 3, ']')) == NULL)
153 			return (0);
154 		if (*(p - 1) != '=' || p - s->str < 4)
155 			goto repeat;
156 		s->str += 2;
157 		genequiv(s);
158 		return (1);
159 	default:				/* "[\###*n]" or "[#*n]" */
160 	repeat:
161 		if ((p = strpbrk(s->str + 2, "*]")) == NULL)
162 			return (0);
163 		if (p[0] != '*' || index(p, ']') == NULL)
164 			return (0);
165 		s->str += 1;
166 		genseq(s);
167 		return (1);
168 	}
169 	/* NOTREACHED */
170 }
171 
172 static void
173 genclass(s)
174 	STR *s;
175 {
176 
177 	if ((s->cclass = wctype(s->str)) == 0)
178 		errx(1, "unknown class %s", s->str);
179 	s->cnt = 0;
180 	s->lastch = -1;		/* incremented before check in next() */
181 	if (strcmp(s->str, "upper") == 0)
182 		s->state = CCLASS_UPPER;
183 	else if (strcmp(s->str, "lower") == 0)
184 		s->state = CCLASS_LOWER;
185 	else
186 		s->state = CCLASS;
187 }
188 
189 static void
190 genequiv(s)
191 	STR *s;
192 {
193 	int i, p, pri;
194 	char src[2], dst[3];
195 	size_t clen;
196 	wchar_t wc;
197 
198 	if (*s->str == '\\') {
199 		s->equiv[0] = backslash(s, NULL);
200 		if (*s->str != '=')
201 			errx(1, "misplaced equivalence equals sign");
202 		s->str += 2;
203 	} else {
204 		clen = mbrtowc(&wc, s->str, MB_LEN_MAX, NULL);
205 		if (clen == (size_t)-1 || clen == (size_t)-2 || clen == 0) {
206 			errno = EILSEQ;
207 			err(1, NULL);
208 		}
209 		s->equiv[0] = wc;
210 		if (s->str[clen] != '=')
211 			errx(1, "misplaced equivalence equals sign");
212 		s->str += clen + 2;
213 	}
214 
215 	/*
216 	 * Calculate the set of all characters in the same equivalence class
217 	 * as the specified character (they will have the same primary
218 	 * collation weights).
219 	 * XXX Knows too much about how strxfrm() is implemented. Assumes
220 	 * it fills the string with primary collation weight bytes. Only one-
221 	 * to-one mappings are supported.
222 	 * XXX Equivalence classes not supported in multibyte locales.
223 	 */
224 	src[0] = (char)s->equiv[0];
225 	src[1] = '\0';
226 	if (MB_CUR_MAX == 1 && strxfrm(dst, src, sizeof (dst)) == 1) {
227 		pri = (unsigned char)*dst;
228 		for (p = 1, i = 1; i < NCHARS_SB; i++) {
229 			*src = i;
230 			if (strxfrm(dst, src, sizeof (dst)) == 1 && pri &&
231 			    pri == (unsigned char)*dst)
232 				s->equiv[p++] = i;
233 		}
234 		s->equiv[p] = OOBCH;
235 	}
236 
237 	s->cnt = 0;
238 	s->state = SET;
239 	s->set = s->equiv;
240 }
241 
242 static int
243 genrange(STR *s, int was_octal)
244 {
245 	int stopval, octal;
246 	char *savestart;
247 	int n, cnt, *p;
248 	size_t clen;
249 	wchar_t wc;
250 
251 	octal = 0;
252 	savestart = s->str;
253 	if (*++s->str == '\\')
254 		stopval = backslash(s, &octal);
255 	else {
256 		clen = mbrtowc(&wc, s->str, MB_LEN_MAX, NULL);
257 		if (clen == (size_t)-1 || clen == (size_t)-2) {
258 			errno = EILSEQ;
259 			err(1, NULL);
260 		}
261 		stopval = wc;
262 		s->str += clen;
263 	}
264 	/*
265 	 * XXX Characters are not ordered according to collating sequence in
266 	 * multibyte locales.
267 	 */
268 	if (octal || was_octal || MB_CUR_MAX > 1) {
269 		if (stopval < s->lastch) {
270 			s->str = savestart;
271 			return (0);
272 		}
273 		s->cnt = stopval - s->lastch + 1;
274 		s->state = RANGE;
275 		--s->lastch;
276 		return (1);
277 	}
278 	if (charcoll((const void *)&stopval, (const void *)&(s->lastch)) < 0) {
279 		s->str = savestart;
280 		return (0);
281 	}
282 	p = malloc((NCHARS_SB + 1) * sizeof (int));
283 	if ((s->set = (void *)p) == NULL)
284 		err(1, "genrange() malloc");
285 	for (cnt = 0; cnt < NCHARS_SB; cnt++)
286 		if (charcoll((const void *)&cnt, (const void *)&(s->lastch)) >=
287 		    0 &&
288 		    charcoll((const void *)&cnt, (const void *)&stopval) <= 0)
289 			*p++ = cnt;
290 	*p = OOBCH;
291 	n = (int *)p - (int *)s->set;
292 
293 	s->cnt = 0;
294 	s->state = SET;
295 	if (n > 1)
296 		qsort(s->set, n, sizeof (*(s->set)), charcoll);
297 	return (1);
298 }
299 
300 static void
301 genseq(s)
302 	STR *s;
303 {
304 	char *ep;
305 	wchar_t wc;
306 	size_t clen;
307 
308 	if (s->which == STRING1)
309 		errx(1, "sequences only valid in string2");
310 
311 	if (*s->str == '\\')
312 		s->lastch = backslash(s, NULL);
313 	else {
314 		clen = mbrtowc(&wc, s->str, MB_LEN_MAX, NULL);
315 		if (clen == (size_t)-1 || clen == (size_t)-2) {
316 			errno = EILSEQ;
317 			err(1, NULL);
318 		}
319 		s->lastch = wc;
320 		s->str += clen;
321 	}
322 	if (*s->str != '*')
323 		errx(1, "misplaced sequence asterisk");
324 
325 	switch (*++s->str) {
326 	case '\\':
327 		s->cnt = backslash(s, NULL);
328 		break;
329 	case ']':
330 		s->cnt = 0;
331 		++s->str;
332 		break;
333 	default:
334 		if (isdigit((uchar_t)*s->str)) {
335 			s->cnt = strtol(s->str, &ep, 0);
336 			if (*ep == ']') {
337 				s->str = ep + 1;
338 				break;
339 			}
340 		}
341 		errx(1, "illegal sequence count");
342 		/* NOTREACHED */
343 	}
344 
345 	s->state = s->cnt ? SEQUENCE : INFINITE;
346 }
347 
348 /*
349  * Translate \??? into a character.  Up to 3 octal digits, if no digits either
350  * an escape code or a literal character.
351  */
352 static int
353 backslash(STR *s, int *is_octal)
354 {
355 	int ch, cnt, val;
356 
357 	if (is_octal != NULL)
358 		*is_octal = 0;
359 	for (cnt = val = 0; ; ) {
360 		ch = (uchar_t)*++s->str;
361 		if (!isdigit(ch) || ch > '7')
362 			break;
363 		val = val * 8 + ch - '0';
364 		if (++cnt == 3) {
365 			++s->str;
366 			break;
367 		}
368 	}
369 	if (cnt) {
370 		if (is_octal != NULL)
371 			*is_octal = 1;
372 		return (val);
373 	}
374 	if (ch != '\0')
375 		++s->str;
376 	switch (ch) {
377 	case 'a':			/* escape characters */
378 		return ('\7');
379 	case 'b':
380 		return ('\b');
381 	case 'f':
382 		return ('\f');
383 	case 'n':
384 		return ('\n');
385 	case 'r':
386 		return ('\r');
387 	case 't':
388 		return ('\t');
389 	case 'v':
390 		return ('\13');
391 	case '\0':			/*  \" -> \ */
392 		s->state = EOS;
393 		return ('\\');
394 	default:			/* \x" -> x */
395 		return (ch);
396 	}
397 }
398