xref: /freebsd/usr.bin/tr/tr.c (revision cc16dea626cf2fc80cde667ac4798065108e596c)
1 /*
2  * Copyright (c) 1988, 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  * 4. Neither the name of the University nor the names of its contributors
14  *    may be used to endorse or promote products derived from this software
15  *    without specific prior written permission.
16  *
17  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27  * SUCH DAMAGE.
28  */
29 
30 #include <sys/cdefs.h>
31 
32 __FBSDID("$FreeBSD$");
33 
34 #ifndef lint
35 static const char copyright[] =
36 "@(#) Copyright (c) 1988, 1993\n\
37 	The Regents of the University of California.  All rights reserved.\n";
38 #endif
39 
40 #ifndef lint
41 static const char sccsid[] = "@(#)tr.c	8.2 (Berkeley) 5/4/95";
42 #endif
43 
44 #include <sys/types.h>
45 
46 #include <ctype.h>
47 #include <err.h>
48 #include <limits.h>
49 #include <locale.h>
50 #include <stdint.h>
51 #include <stdio.h>
52 #include <stdlib.h>
53 #include <string.h>
54 #include <unistd.h>
55 #include <wchar.h>
56 #include <wctype.h>
57 
58 #include "cmap.h"
59 #include "cset.h"
60 #include "extern.h"
61 
62 static STR s1 = { STRING1, NORMAL, 0, OOBCH, 0, { 0, OOBCH }, NULL, NULL };
63 static STR s2 = { STRING2, NORMAL, 0, OOBCH, 0, { 0, OOBCH }, NULL, NULL };
64 
65 static struct cset *setup(char *, STR *, int, int);
66 static void usage(void);
67 
68 int
69 main(int argc, char **argv)
70 {
71 	static int carray[NCHARS_SB];
72 	struct cmap *map;
73 	struct cset *delete, *squeeze;
74 	int n, *p;
75 	int Cflag, cflag, dflag, sflag, isstring2;
76 	wint_t ch, cnt, lastch;
77 
78 	(void)setlocale(LC_ALL, "");
79 
80 	Cflag = cflag = dflag = sflag = 0;
81 	while ((ch = getopt(argc, argv, "Ccdsu")) != -1)
82 		switch((char)ch) {
83 		case 'C':
84 			Cflag = 1;
85 			cflag = 0;
86 			break;
87 		case 'c':
88 			cflag = 1;
89 			Cflag = 0;
90 			break;
91 		case 'd':
92 			dflag = 1;
93 			break;
94 		case 's':
95 			sflag = 1;
96 			break;
97 		case 'u':
98 			setbuf(stdout, (char *)NULL);
99 			break;
100 		case '?':
101 		default:
102 			usage();
103 		}
104 	argc -= optind;
105 	argv += optind;
106 
107 	switch(argc) {
108 	case 0:
109 	default:
110 		usage();
111 		/* NOTREACHED */
112 	case 1:
113 		isstring2 = 0;
114 		break;
115 	case 2:
116 		isstring2 = 1;
117 		break;
118 	}
119 
120 	/*
121 	 * tr -ds [-Cc] string1 string2
122 	 * Delete all characters (or complemented characters) in string1.
123 	 * Squeeze all characters in string2.
124 	 */
125 	if (dflag && sflag) {
126 		if (!isstring2)
127 			usage();
128 
129 		delete = setup(argv[0], &s1, cflag, Cflag);
130 		squeeze = setup(argv[1], &s2, 0, 0);
131 
132 		for (lastch = OOBCH; (ch = getwchar()) != WEOF;)
133 			if (!cset_in(delete, ch) &&
134 			    (lastch != ch || !cset_in(squeeze, ch))) {
135 				lastch = ch;
136 				(void)putwchar(ch);
137 			}
138 		if (ferror(stdin))
139 			err(1, NULL);
140 		exit(0);
141 	}
142 
143 	/*
144 	 * tr -d [-Cc] string1
145 	 * Delete all characters (or complemented characters) in string1.
146 	 */
147 	if (dflag) {
148 		if (isstring2)
149 			usage();
150 
151 		delete = setup(argv[0], &s1, cflag, Cflag);
152 
153 		while ((ch = getwchar()) != WEOF)
154 			if (!cset_in(delete, ch))
155 				(void)putwchar(ch);
156 		if (ferror(stdin))
157 			err(1, NULL);
158 		exit(0);
159 	}
160 
161 	/*
162 	 * tr -s [-Cc] string1
163 	 * Squeeze all characters (or complemented characters) in string1.
164 	 */
165 	if (sflag && !isstring2) {
166 		squeeze = setup(argv[0], &s1, cflag, Cflag);
167 
168 		for (lastch = OOBCH; (ch = getwchar()) != WEOF;)
169 			if (lastch != ch || !cset_in(squeeze, ch)) {
170 				lastch = ch;
171 				(void)putwchar(ch);
172 			}
173 		if (ferror(stdin))
174 			err(1, NULL);
175 		exit(0);
176 	}
177 
178 	/*
179 	 * tr [-Ccs] string1 string2
180 	 * Replace all characters (or complemented characters) in string1 with
181 	 * the character in the same position in string2.  If the -s option is
182 	 * specified, squeeze all the characters in string2.
183 	 */
184 	if (!isstring2)
185 		usage();
186 
187 	map = cmap_alloc();
188 	if (map == NULL)
189 		err(1, NULL);
190 	squeeze = cset_alloc();
191 	if (squeeze == NULL)
192 		err(1, NULL);
193 
194 	s1.str = argv[0];
195 
196 	if (Cflag || cflag) {
197 		cmap_default(map, OOBCH);
198 		if ((s2.str = strdup(argv[1])) == NULL)
199 			errx(1, "strdup(argv[1])");
200 	} else
201 		s2.str = argv[1];
202 
203 	if (!next(&s2))
204 		errx(1, "empty string2");
205 
206 	/*
207 	 * For -s result will contain only those characters defined
208 	 * as the second characters in each of the toupper or tolower
209 	 * pairs.
210 	 */
211 
212 	/* If string2 runs out of characters, use the last one specified. */
213 	while (next(&s1)) {
214 	again:
215 		if (s1.state == CCLASS_LOWER &&
216 		    s2.state == CCLASS_UPPER &&
217 		    s1.cnt == 1 && s2.cnt == 1) {
218 			do {
219 				ch = towupper(s1.lastch);
220 				cmap_add(map, s1.lastch, ch);
221 				if (sflag && iswupper(ch))
222 					cset_add(squeeze, ch);
223 				if (!next(&s1))
224 					goto endloop;
225 			} while (s1.state == CCLASS_LOWER && s1.cnt > 1);
226 			/* skip upper set */
227 			do {
228 				if (!next(&s2))
229 					break;
230 			} while (s2.state == CCLASS_UPPER && s2.cnt > 1);
231 			goto again;
232 		} else if (s1.state == CCLASS_UPPER &&
233 			   s2.state == CCLASS_LOWER &&
234 			   s1.cnt == 1 && s2.cnt == 1) {
235 			do {
236 				ch = towlower(s1.lastch);
237 				cmap_add(map, s1.lastch, ch);
238 				if (sflag && iswlower(ch))
239 					cset_add(squeeze, ch);
240 				if (!next(&s1))
241 					goto endloop;
242 			} while (s1.state == CCLASS_UPPER && s1.cnt > 1);
243 			/* skip lower set */
244 			do {
245 				if (!next(&s2))
246 					break;
247 			} while (s2.state == CCLASS_LOWER && s2.cnt > 1);
248 			goto again;
249 		} else {
250 			cmap_add(map, s1.lastch, s2.lastch);
251 			if (sflag)
252 				cset_add(squeeze, s2.lastch);
253 		}
254 		(void)next(&s2);
255 	}
256 endloop:
257 	if (cflag || (Cflag && MB_CUR_MAX > 1)) {
258 		/*
259 		 * This is somewhat tricky: since the character set is
260 		 * potentially huge, we need to avoid allocating a map
261 		 * entry for every character. Our strategy is to set the
262 		 * default mapping to the last character of string #2
263 		 * (= the one that gets automatically repeated), then to
264 		 * add back identity mappings for characters that should
265 		 * remain unchanged. We don't waste space on identity mappings
266 		 * for non-characters with the -C option; those are simulated
267 		 * in the I/O loop.
268 		 */
269 		s2.str = argv[1];
270 		s2.state = NORMAL;
271 		for (cnt = 0; cnt < WINT_MAX; cnt++) {
272 			if (Cflag && !iswrune(cnt))
273 				continue;
274 			if (cmap_lookup(map, cnt) == OOBCH) {
275 				if (next(&s2))
276 					cmap_add(map, cnt, s2.lastch);
277 				if (sflag)
278 					cset_add(squeeze, s2.lastch);
279 			} else
280 				cmap_add(map, cnt, cnt);
281 			if ((s2.state == EOS || s2.state == INFINITE) &&
282 			    cnt >= cmap_max(map))
283 				break;
284 		}
285 		cmap_default(map, s2.lastch);
286 	} else if (Cflag) {
287 		for (p = carray, cnt = 0; cnt < NCHARS_SB; cnt++) {
288 			if (cmap_lookup(map, cnt) == OOBCH && iswrune(cnt))
289 				*p++ = cnt;
290 			else
291 				cmap_add(map, cnt, cnt);
292 		}
293 		n = p - carray;
294 		if (Cflag && n > 1)
295 			(void)mergesort(carray, n, sizeof(*carray), charcoll);
296 
297 		s2.str = argv[1];
298 		s2.state = NORMAL;
299 		for (cnt = 0; cnt < n; cnt++) {
300 			(void)next(&s2);
301 			cmap_add(map, carray[cnt], s2.lastch);
302 			/*
303 			 * Chars taken from s2 can be different this time
304 			 * due to lack of complex upper/lower processing,
305 			 * so fill string2 again to not miss some.
306 			 */
307 			if (sflag)
308 				cset_add(squeeze, s2.lastch);
309 		}
310 	}
311 
312 	cset_cache(squeeze);
313 	cmap_cache(map);
314 
315 	if (sflag)
316 		for (lastch = OOBCH; (ch = getwchar()) != WEOF;) {
317 			if (!Cflag || iswrune(ch))
318 				ch = cmap_lookup(map, ch);
319 			if (lastch != ch || !cset_in(squeeze, ch)) {
320 				lastch = ch;
321 				(void)putwchar(ch);
322 			}
323 		}
324 	else
325 		while ((ch = getwchar()) != WEOF) {
326 			if (!Cflag || iswrune(ch))
327 				ch = cmap_lookup(map, ch);
328 			(void)putwchar(ch);
329 		}
330 	if (ferror(stdin))
331 		err(1, NULL);
332 	exit (0);
333 }
334 
335 static struct cset *
336 setup(char *arg, STR *str, int cflag, int Cflag)
337 {
338 	struct cset *cs;
339 
340 	cs = cset_alloc();
341 	if (cs == NULL)
342 		err(1, NULL);
343 	str->str = arg;
344 	while (next(str))
345 		cset_add(cs, str->lastch);
346 	if (Cflag)
347 		cset_addclass(cs, wctype("rune"), true);
348 	if (cflag || Cflag)
349 		cset_invert(cs);
350 	cset_cache(cs);
351 	return (cs);
352 }
353 
354 int
355 charcoll(const void *a, const void *b)
356 {
357 	static char sa[2], sb[2];
358 
359 	sa[0] = *(const int *)a;
360 	sb[0] = *(const int *)b;
361 	return (strcoll(sa, sb));
362 }
363 
364 static void
365 usage(void)
366 {
367 	(void)fprintf(stderr, "%s\n%s\n%s\n%s\n",
368 		"usage: tr [-Ccsu] string1 string2",
369 		"       tr [-Ccu] -d string1",
370 		"       tr [-Ccu] -s string1",
371 		"       tr [-Ccu] -ds string1 string2");
372 	exit(1);
373 }
374