Annotation of src/usr.bin/tr/str.c, Revision 1.10
1.10 ! djm 1: /* $OpenBSD: str.c,v 1.9 2004/09/15 22:12:19 deraadt Exp $ */
1.1 deraadt 2: /* $NetBSD: str.c,v 1.7 1995/08/31 22:13:47 jtc Exp $ */
3:
4: /*-
5: * Copyright (c) 1991, 1993
6: * The Regents of the University of California. All rights reserved.
7: *
8: * Redistribution and use in source and binary forms, with or without
9: * modification, are permitted provided that the following conditions
10: * are met:
11: * 1. Redistributions of source code must retain the above copyright
12: * notice, this list of conditions and the following disclaimer.
13: * 2. Redistributions in binary form must reproduce the above copyright
14: * notice, this list of conditions and the following disclaimer in the
15: * documentation and/or other materials provided with the distribution.
1.8 millert 16: * 3. Neither the name of the University nor the names of its contributors
1.1 deraadt 17: * may be used to endorse or promote products derived from this software
18: * without specific prior written permission.
19: *
20: * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
21: * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22: * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23: * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
24: * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
30: * SUCH DAMAGE.
31: */
32:
33: #ifndef lint
34: #if 0
35: static char sccsid[] = "@(#)str.c 8.2 (Berkeley) 4/28/95";
36: #endif
1.10 ! djm 37: static char rcsid[] = "$OpenBSD: str.c,v 1.9 2004/09/15 22:12:19 deraadt Exp $";
1.1 deraadt 38: #endif /* not lint */
39:
40: #include <sys/cdefs.h>
41: #include <sys/types.h>
42:
43: #include <errno.h>
44: #include <stddef.h>
45: #include <stdio.h>
46: #include <stdlib.h>
47: #include <string.h>
48: #include <ctype.h>
1.4 mickey 49: #include <err.h>
1.1 deraadt 50:
51: #include "extern.h"
52:
1.7 millert 53: static int backslash(STR *);
54: static int bracket(STR *);
55: static int c_class(const void *, const void *);
56: static void genclass(STR *);
57: static void genequiv(STR *);
58: static int genrange(STR *);
59: static void genseq(STR *);
1.1 deraadt 60:
61: int
62: next(s)
1.6 mpech 63: STR *s;
1.1 deraadt 64: {
1.6 mpech 65: int ch;
1.1 deraadt 66:
67: switch (s->state) {
68: case EOS:
69: return (0);
70: case INFINITE:
71: return (1);
72: case NORMAL:
73: switch (ch = *s->str) {
74: case '\0':
75: s->state = EOS;
76: return (0);
77: case '\\':
78: s->lastch = backslash(s);
79: break;
80: case '[':
81: if (bracket(s))
82: return (next(s));
83: /* FALLTHROUGH */
84: default:
85: ++s->str;
86: s->lastch = ch;
87: break;
88: }
89:
90: /* We can start a range at any time. */
91: if (s->str[0] == '-' && genrange(s))
92: return (next(s));
93: return (1);
94: case RANGE:
95: if (s->cnt-- == 0) {
96: s->state = NORMAL;
97: return (next(s));
98: }
99: ++s->lastch;
100: return (1);
101: case SEQUENCE:
102: if (s->cnt-- == 0) {
103: s->state = NORMAL;
104: return (next(s));
105: }
106: return (1);
107: case SET:
108: if ((s->lastch = s->set[s->cnt++]) == OOBCH) {
109: s->state = NORMAL;
110: return (next(s));
111: }
112: return (1);
1.4 mickey 113: default:
114: return 0;
1.1 deraadt 115: }
116: /* NOTREACHED */
117: }
118:
119: static int
120: bracket(s)
1.6 mpech 121: STR *s;
1.1 deraadt 122: {
1.6 mpech 123: char *p;
1.1 deraadt 124:
125: switch (s->str[1]) {
126: case ':': /* "[:class:]" */
1.9 deraadt 127: if ((p = strstr((char *)s->str + 2, ":]")) == NULL)
1.1 deraadt 128: return (0);
129: *p = '\0';
130: s->str += 2;
131: genclass(s);
1.9 deraadt 132: s->str = (unsigned char *)p + 2;
1.1 deraadt 133: return (1);
134: case '=': /* "[=equiv=]" */
1.9 deraadt 135: if ((p = strstr((char *)s->str + 2, "=]")) == NULL)
1.1 deraadt 136: return (0);
137: s->str += 2;
138: genequiv(s);
139: return (1);
140: default: /* "[\###*n]" or "[#*n]" */
1.9 deraadt 141: if ((p = strpbrk((char *)s->str + 2, "*]")) == NULL)
1.1 deraadt 142: return (0);
1.3 millert 143: if (p[0] != '*' || strchr(p, ']') == NULL)
1.1 deraadt 144: return (0);
145: s->str += 1;
146: genseq(s);
147: return (1);
148: }
149: /* NOTREACHED */
150: }
151:
152: typedef struct {
153: char *name;
1.7 millert 154: int (*func)(int);
1.1 deraadt 155: int *set;
156: } CLASS;
157:
158: static CLASS classes[] = {
159: { "alnum", isalnum, },
160: { "alpha", isalpha, },
161: { "blank", isblank, },
162: { "cntrl", iscntrl, },
163: { "digit", isdigit, },
164: { "graph", isgraph, },
165: { "lower", islower, },
1.5 deraadt 166: { "print", isprint, },
1.1 deraadt 167: { "punct", ispunct, },
168: { "space", isspace, },
169: { "upper", isupper, },
170: { "xdigit", isxdigit, },
171: };
172:
173: static void
174: genclass(s)
175: STR *s;
176: {
1.7 millert 177: int cnt, (*func)(int);
1.1 deraadt 178: CLASS *cp, tmp;
179: int *p;
180:
1.9 deraadt 181: tmp.name = (char *)s->str;
1.1 deraadt 182: if ((cp = (CLASS *)bsearch(&tmp, classes, sizeof(classes) /
183: sizeof(CLASS), sizeof(CLASS), c_class)) == NULL)
1.4 mickey 184: errx(1, "unknown class %s", s->str);
1.1 deraadt 185:
1.10 ! djm 186: if ((cp->set = p = calloc(NCHARS + 1, sizeof(int))) == NULL)
1.4 mickey 187: errx(1, "no memory for a class");
1.1 deraadt 188: for (cnt = 0, func = cp->func; cnt < NCHARS; ++cnt)
189: if ((func)(cnt))
190: *p++ = cnt;
191: *p = OOBCH;
192:
193: s->cnt = 0;
194: s->state = SET;
195: s->set = cp->set;
196: }
197:
198: static int
199: c_class(a, b)
200: const void *a, *b;
201: {
202: return (strcmp(((CLASS *)a)->name, ((CLASS *)b)->name));
203: }
204:
205: /*
206: * English doesn't have any equivalence classes, so for now
207: * we just syntax check and grab the character.
208: */
209: static void
210: genequiv(s)
211: STR *s;
212: {
213: if (*s->str == '\\') {
214: s->equiv[0] = backslash(s);
215: if (*s->str != '=')
1.4 mickey 216: errx(1, "misplaced equivalence equals sign");
1.1 deraadt 217: } else {
218: s->equiv[0] = s->str[0];
219: if (s->str[1] != '=')
1.4 mickey 220: errx(1, "misplaced equivalence equals sign");
1.1 deraadt 221: }
222: s->str += 2;
223: s->cnt = 0;
224: s->state = SET;
225: s->set = s->equiv;
226: }
227:
228: static int
229: genrange(s)
230: STR *s;
231: {
232: int stopval;
1.9 deraadt 233: unsigned char *savestart;
1.1 deraadt 234:
235: savestart = s->str;
236: stopval = *++s->str == '\\' ? backslash(s) : *s->str++;
237: if (stopval < (u_char)s->lastch) {
238: s->str = savestart;
239: return (0);
240: }
241: s->cnt = stopval - s->lastch + 1;
242: s->state = RANGE;
243: --s->lastch;
244: return (1);
245: }
246:
247: static void
248: genseq(s)
249: STR *s;
250: {
251: char *ep;
252:
253: if (s->which == STRING1)
1.4 mickey 254: errx(1, "sequences only valid in string2");
1.1 deraadt 255:
256: if (*s->str == '\\')
257: s->lastch = backslash(s);
258: else
259: s->lastch = *s->str++;
260: if (*s->str != '*')
1.4 mickey 261: errx(1, "misplaced sequence asterisk");
1.1 deraadt 262:
263: switch (*++s->str) {
264: case '\\':
265: s->cnt = backslash(s);
266: break;
267: case ']':
268: s->cnt = 0;
269: ++s->str;
270: break;
271: default:
272: if (isdigit(*s->str)) {
1.9 deraadt 273: s->cnt = strtol((char *)s->str, &ep, 0);
1.1 deraadt 274: if (*ep == ']') {
1.9 deraadt 275: s->str = (unsigned char *)ep + 1;
1.1 deraadt 276: break;
277: }
278: }
1.4 mickey 279: errx(1, "illegal sequence count");
1.1 deraadt 280: /* NOTREACHED */
281: }
282:
283: s->state = s->cnt ? SEQUENCE : INFINITE;
284: }
285:
286: /*
287: * Translate \??? into a character. Up to 3 octal digits, if no digits either
288: * an escape code or a literal character.
289: */
290: static int
291: backslash(s)
1.6 mpech 292: STR *s;
1.1 deraadt 293: {
1.6 mpech 294: int ch, cnt, val;
1.1 deraadt 295:
296: for (cnt = val = 0;;) {
297: ch = *++s->str;
298: if (!isascii(ch) || !isdigit(ch))
299: break;
300: val = val * 8 + ch - '0';
301: if (++cnt == 3) {
302: ++s->str;
303: break;
304: }
305: }
306: if (cnt)
307: return (val);
308: if (ch != '\0')
309: ++s->str;
310: switch (ch) {
311: case 'a': /* escape characters */
312: return ('\7');
313: case 'b':
314: return ('\b');
315: case 'f':
316: return ('\f');
317: case 'n':
318: return ('\n');
319: case 'r':
320: return ('\r');
321: case 't':
322: return ('\t');
323: case 'v':
324: return ('\13');
325: case '\0': /* \" -> \ */
326: s->state = EOS;
327: return ('\\');
328: default: /* \x" -> x */
329: return (ch);
330: }
331: }