str.c revision 118412 
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/cdefs.h>
35
36__FBSDID("$FreeBSD: head/usr.bin/tr/str.c 118412 2003-08-04 04:20:04Z ache $");
37
38#ifndef lint
39static const char sccsid[] = "@(#)str.c	8.2 (Berkeley) 4/28/95";
40#endif
41
42#include <sys/cdefs.h>
43#include <sys/types.h>
44
45#include <ctype.h>
46#include <err.h>
47#include <stddef.h>
48#include <stdio.h>
49#include <stdlib.h>
50#include <string.h>
51
52#include "extern.h"
53
54static int      backslash(STR *, int *);
55static int	bracket(STR *);
56static int	c_class(const void *, const void *);
57static void	genclass(STR *);
58static void	genequiv(STR *);
59static int      genrange(STR *, int);
60static void	genseq(STR *);
61
62int
63next(s)
64	STR *s;
65{
66	int ch, is_octal;
67
68	switch (s->state) {
69	case EOS:
70		return (0);
71	case INFINITE:
72		return (1);
73	case NORMAL:
74		switch (ch = (u_char)*s->str) {
75		case '\0':
76			s->state = EOS;
77			return (0);
78		case '\\':
79			s->lastch = backslash(s, &is_octal);
80			break;
81		case '[':
82			if (bracket(s))
83				return (next(s));
84			/* FALLTHROUGH */
85		default:
86			is_octal = 0;
87			++s->str;
88			s->lastch = ch;
89			break;
90		}
91
92		/* We can start a range at any time. */
93		if (s->str[0] == '-' && genrange(s, is_octal))
94			return (next(s));
95		return (1);
96	case SEQUENCE:
97		if (s->cnt-- == 0) {
98			s->state = NORMAL;
99			return (next(s));
100		}
101		return (1);
102	case SET:
103	case SET_UPPER:
104	case SET_LOWER:
105		if ((ch = s->set[s->cnt++]) == OOBCH) {
106			s->state = NORMAL;
107			return (next(s));
108		}
109		s->lastch = ch;
110		return (1);
111	default:
112		return (0);
113	}
114	/* NOTREACHED */
115}
116
117static int
118bracket(s)
119	STR *s;
120{
121	char *p;
122
123	switch (s->str[1]) {
124	case ':':				/* "[:class:]" */
125		if ((p = strchr(s->str + 2, ']')) == NULL)
126			return (0);
127		if (*(p - 1) != ':' || p - s->str < 4)
128			goto repeat;
129		*(p - 1) = '\0';
130		s->str += 2;
131		genclass(s);
132		s->str = p + 1;
133		return (1);
134	case '=':				/* "[=equiv=]" */
135		if ((p = strchr(s->str + 2, ']')) == NULL)
136			return (0);
137		if (*(p - 1) != '=' || p - s->str < 4)
138			goto repeat;
139		s->str += 2;
140		genequiv(s);
141		return (1);
142	default:				/* "[\###*n]" or "[#*n]" */
143	repeat:
144		if ((p = strpbrk(s->str + 2, "*]")) == NULL)
145			return (0);
146		if (p[0] != '*' || index(p, ']') == NULL)
147			return (0);
148		s->str += 1;
149		genseq(s);
150		return (1);
151	}
152	/* NOTREACHED */
153}
154
155typedef struct {
156	const char *name;
157	int (*func)(int);
158	int *set;
159} CLASS;
160
161static CLASS classes[] = {
162#undef isalnum
163	{ "alnum",  isalnum,  NULL },
164#undef isalpha
165	{ "alpha",  isalpha,  NULL },
166#undef isblank
167	{ "blank",  isblank,  NULL },
168#undef iscntrl
169	{ "cntrl",  iscntrl,  NULL },
170#undef isdigit
171	{ "digit",  isdigit,  NULL },
172#undef isgraph
173	{ "graph",  isgraph,  NULL },
174#undef islower
175	{ "lower",  islower,  NULL },
176#undef isprint
177	{ "print",  isprint,  NULL },
178#undef ispunct
179	{ "punct",  ispunct,  NULL },
180#undef isspace
181	{ "space",  isspace,  NULL },
182#undef isupper
183	{ "upper",  isupper,  NULL },
184#undef isxdigit
185	{ "xdigit", isxdigit, NULL },
186};
187
188static void
189genclass(s)
190	STR *s;
191{
192	int cnt, (*func)(int);
193	CLASS *cp, tmp;
194	int *p, n;
195
196	tmp.name = s->str;
197	if ((cp = (CLASS *)bsearch(&tmp, classes, sizeof(classes) /
198	    sizeof(CLASS), sizeof(CLASS), c_class)) == NULL)
199		errx(1, "unknown class %s", s->str);
200
201	if ((cp->set = p = malloc((NCHARS + 1) * sizeof(int))) == NULL)
202		err(1, "genclass() malloc");
203	bzero(p, (NCHARS + 1) * sizeof(int));
204	for (cnt = 0, func = cp->func; cnt < NCHARS; ++cnt)
205		if ((func)(cnt))
206			*p++ = cnt;
207	*p = OOBCH;
208	n = p - cp->set;
209
210	s->cnt = 0;
211	s->set = cp->set;
212	if (strcmp(s->str, "upper") == 0)
213		s->state = SET_UPPER;
214	else if (strcmp(s->str, "lower") == 0) {
215		s->state = SET_LOWER;
216	} else
217		s->state = SET;
218	if ((s->state == SET_LOWER || s->state == SET_UPPER) && n > 1)
219		mergesort(s->set, n, sizeof(*(s->set)), charcoll);
220}
221
222static int
223c_class(a, b)
224	const void *a, *b;
225{
226	return (strcmp(((const CLASS *)a)->name, ((const CLASS *)b)->name));
227}
228
229static void
230genequiv(s)
231	STR *s;
232{
233	int i, p, pri;
234	char src[2], dst[3];
235
236	if (*s->str == '\\') {
237		s->equiv[0] = backslash(s, NULL);
238		if (*s->str != '=')
239			errx(1, "misplaced equivalence equals sign");
240		s->str += 2;
241	} else {
242		s->equiv[0] = s->str[0];
243		if (s->str[1] != '=')
244			errx(1, "misplaced equivalence equals sign");
245		s->str += 3;
246	}
247
248	/*
249	 * Calculate the set of all characters in the same equivalence class
250	 * as the specified character (they will have the same primary
251	 * collation weights).
252	 * XXX Knows too much about how strxfrm() is implemented. Assumes
253	 * it fills the string with primary collation weight bytes. Only one-
254	 * to-one mappings are supported.
255	 */
256	src[0] = s->equiv[0];
257	src[1] = '\0';
258	if (strxfrm(dst, src, sizeof(dst)) == 1) {
259		pri = (unsigned char)*dst;
260		for (p = 1, i = 1; i < NCHARS; i++) {
261			*src = i;
262			if (strxfrm(dst, src, sizeof(dst)) == 1 && pri &&
263			    pri == (unsigned char)*dst)
264				s->equiv[p++] = i;
265		}
266		s->equiv[p] = OOBCH;
267	}
268
269	s->cnt = 0;
270	s->state = SET;
271	s->set = s->equiv;
272}
273
274static int
275genrange(STR *s, int was_octal)
276{
277	int stopval, octal;
278	char *savestart;
279	int n, cnt, *p;
280
281	octal = 0;
282	savestart = s->str;
283	stopval = *++s->str == '\\' ? backslash(s, &octal) : (u_char)*s->str++;
284	if (!octal)
285		octal = was_octal;
286
287	if ((octal && stopval < s->lastch) ||
288	    (!octal &&
289	     charcoll((const void *)&stopval, (const void *)&(s->lastch)) < 0)) {
290		s->str = savestart;
291		return (0);
292	}
293	if ((s->set = p = malloc((NCHARS + 1) * sizeof(int))) == NULL)
294		err(1, "genrange() malloc");
295	bzero(p, (NCHARS + 1) * sizeof(int));
296	if (octal) {
297		for (cnt = s->lastch; cnt <= stopval; cnt++)
298			*p++ = cnt;
299	} else {
300		for (cnt = 0; cnt < NCHARS; cnt++)
301			if (charcoll((const void *)&cnt, (const void *)&(s->lastch)) >= 0 &&
302			    charcoll((const void *)&cnt, (const void *)&stopval) <= 0)
303				*p++ = cnt;
304	}
305	*p = OOBCH;
306	n = p - s->set;
307
308	s->cnt = 0;
309	s->state = SET;
310	if (!octal && n > 1)
311		mergesort(s->set, n, sizeof(*(s->set)), charcoll);
312	return (1);
313}
314
315static void
316genseq(s)
317	STR *s;
318{
319	char *ep;
320
321	if (s->which == STRING1)
322		errx(1, "sequences only valid in string2");
323
324	if (*s->str == '\\')
325		s->lastch = backslash(s, NULL);
326	else
327		s->lastch = *s->str++;
328	if (*s->str != '*')
329		errx(1, "misplaced sequence asterisk");
330
331	switch (*++s->str) {
332	case '\\':
333		s->cnt = backslash(s, NULL);
334		break;
335	case ']':
336		s->cnt = 0;
337		++s->str;
338		break;
339	default:
340		if (isdigit((u_char)*s->str)) {
341			s->cnt = strtol(s->str, &ep, 0);
342			if (*ep == ']') {
343				s->str = ep + 1;
344				break;
345			}
346		}
347		errx(1, "illegal sequence count");
348		/* NOTREACHED */
349	}
350
351	s->state = s->cnt ? SEQUENCE : INFINITE;
352}
353
354/*
355 * Translate \??? into a character.  Up to 3 octal digits, if no digits either
356 * an escape code or a literal character.
357 */
358static int
359backslash(STR *s, int *is_octal)
360{
361	int ch, cnt, val;
362
363	if (is_octal != NULL)
364		*is_octal = 0;
365	for (cnt = val = 0;;) {
366		ch = (u_char)*++s->str;
367		if (!isdigit(ch))
368			break;
369		val = val * 8 + ch - '0';
370		if (++cnt == 3) {
371			++s->str;
372			break;
373		}
374	}
375	if (cnt) {
376		if (is_octal != NULL)
377			*is_octal = 1;
378		return (val);
379	}
380	if (ch != '\0')
381		++s->str;
382	switch (ch) {
383		case 'a':			/* escape characters */
384			return ('\7');
385		case 'b':
386			return ('\b');
387		case 'f':
388			return ('\f');
389		case 'n':
390			return ('\n');
391		case 'r':
392			return ('\r');
393		case 't':
394			return ('\t');
395		case 'v':
396			return ('\13');
397		case '\0':			/*  \" -> \ */
398			s->state = EOS;
399			return ('\\');
400		default:			/* \x" -> x */
401			return (ch);
402	}
403}
404