xref: /openbsd-src/lib/libc/gen/unvis.c (revision a28daedfc357b214be5c701aa8ba8adb29a7f1c2)
1 /*	$OpenBSD: unvis.c,v 1.12 2005/08/08 08:05:34 espie Exp $ */
2 /*-
3  * Copyright (c) 1989, 1993
4  *	The Regents of the University of California.  All rights reserved.
5  *
6  * Redistribution and use in source and binary forms, with or without
7  * modification, are permitted provided that the following conditions
8  * are met:
9  * 1. Redistributions of source code must retain the above copyright
10  *    notice, this list of conditions and the following disclaimer.
11  * 2. Redistributions in binary form must reproduce the above copyright
12  *    notice, this list of conditions and the following disclaimer in the
13  *    documentation and/or other materials provided with the distribution.
14  * 3. Neither the name of the University nor the names of its contributors
15  *    may be used to endorse or promote products derived from this software
16  *    without specific prior written permission.
17  *
18  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
19  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
20  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
21  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
22  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
23  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
24  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
25  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
26  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
27  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
28  * SUCH DAMAGE.
29  */
30 
31 #include <sys/types.h>
32 #include <ctype.h>
33 #include <vis.h>
34 
35 /*
36  * decode driven by state machine
37  */
38 #define	S_GROUND	0	/* haven't seen escape char */
39 #define	S_START		1	/* start decoding special sequence */
40 #define	S_META		2	/* metachar started (M) */
41 #define	S_META1		3	/* metachar more, regular char (-) */
42 #define	S_CTRL		4	/* control char started (^) */
43 #define	S_OCTAL2	5	/* octal digit 2 */
44 #define	S_OCTAL3	6	/* octal digit 3 */
45 
46 #define	isoctal(c)	(((u_char)(c)) >= '0' && ((u_char)(c)) <= '7')
47 
48 /*
49  * unvis - decode characters previously encoded by vis
50  */
51 int
52 unvis(char *cp, char c, int *astate, int flag)
53 {
54 
55 	if (flag & UNVIS_END) {
56 		if (*astate == S_OCTAL2 || *astate == S_OCTAL3) {
57 			*astate = S_GROUND;
58 			return (UNVIS_VALID);
59 		}
60 		return (*astate == S_GROUND ? UNVIS_NOCHAR : UNVIS_SYNBAD);
61 	}
62 
63 	switch (*astate) {
64 
65 	case S_GROUND:
66 		*cp = 0;
67 		if (c == '\\') {
68 			*astate = S_START;
69 			return (0);
70 		}
71 		*cp = c;
72 		return (UNVIS_VALID);
73 
74 	case S_START:
75 		switch(c) {
76 		case '\\':
77 			*cp = c;
78 			*astate = S_GROUND;
79 			return (UNVIS_VALID);
80 		case '0': case '1': case '2': case '3':
81 		case '4': case '5': case '6': case '7':
82 			*cp = (c - '0');
83 			*astate = S_OCTAL2;
84 			return (0);
85 		case 'M':
86 			*cp = (char) 0200;
87 			*astate = S_META;
88 			return (0);
89 		case '^':
90 			*astate = S_CTRL;
91 			return (0);
92 		case 'n':
93 			*cp = '\n';
94 			*astate = S_GROUND;
95 			return (UNVIS_VALID);
96 		case 'r':
97 			*cp = '\r';
98 			*astate = S_GROUND;
99 			return (UNVIS_VALID);
100 		case 'b':
101 			*cp = '\b';
102 			*astate = S_GROUND;
103 			return (UNVIS_VALID);
104 		case 'a':
105 			*cp = '\007';
106 			*astate = S_GROUND;
107 			return (UNVIS_VALID);
108 		case 'v':
109 			*cp = '\v';
110 			*astate = S_GROUND;
111 			return (UNVIS_VALID);
112 		case 't':
113 			*cp = '\t';
114 			*astate = S_GROUND;
115 			return (UNVIS_VALID);
116 		case 'f':
117 			*cp = '\f';
118 			*astate = S_GROUND;
119 			return (UNVIS_VALID);
120 		case 's':
121 			*cp = ' ';
122 			*astate = S_GROUND;
123 			return (UNVIS_VALID);
124 		case 'E':
125 			*cp = '\033';
126 			*astate = S_GROUND;
127 			return (UNVIS_VALID);
128 		case '\n':
129 			/*
130 			 * hidden newline
131 			 */
132 			*astate = S_GROUND;
133 			return (UNVIS_NOCHAR);
134 		case '$':
135 			/*
136 			 * hidden marker
137 			 */
138 			*astate = S_GROUND;
139 			return (UNVIS_NOCHAR);
140 		}
141 		*astate = S_GROUND;
142 		return (UNVIS_SYNBAD);
143 
144 	case S_META:
145 		if (c == '-')
146 			*astate = S_META1;
147 		else if (c == '^')
148 			*astate = S_CTRL;
149 		else {
150 			*astate = S_GROUND;
151 			return (UNVIS_SYNBAD);
152 		}
153 		return (0);
154 
155 	case S_META1:
156 		*astate = S_GROUND;
157 		*cp |= c;
158 		return (UNVIS_VALID);
159 
160 	case S_CTRL:
161 		if (c == '?')
162 			*cp |= 0177;
163 		else
164 			*cp |= c & 037;
165 		*astate = S_GROUND;
166 		return (UNVIS_VALID);
167 
168 	case S_OCTAL2:	/* second possible octal digit */
169 		if (isoctal(c)) {
170 			/*
171 			 * yes - and maybe a third
172 			 */
173 			*cp = (*cp << 3) + (c - '0');
174 			*astate = S_OCTAL3;
175 			return (0);
176 		}
177 		/*
178 		 * no - done with current sequence, push back passed char
179 		 */
180 		*astate = S_GROUND;
181 		return (UNVIS_VALIDPUSH);
182 
183 	case S_OCTAL3:	/* third possible octal digit */
184 		*astate = S_GROUND;
185 		if (isoctal(c)) {
186 			*cp = (*cp << 3) + (c - '0');
187 			return (UNVIS_VALID);
188 		}
189 		/*
190 		 * we were done, push back passed char
191 		 */
192 		return (UNVIS_VALIDPUSH);
193 
194 	default:
195 		/*
196 		 * decoder in unknown state - (probably uninitialized)
197 		 */
198 		*astate = S_GROUND;
199 		return (UNVIS_SYNBAD);
200 	}
201 }
202 
203 /*
204  * strunvis - decode src into dst
205  *
206  *	Number of chars decoded into dst is returned, -1 on error.
207  *	Dst is null terminated.
208  */
209 
210 int
211 strunvis(char *dst, const char *src)
212 {
213 	char c;
214 	char *start = dst;
215 	int state = 0;
216 
217 	while ((c = *src++)) {
218 	again:
219 		switch (unvis(dst, c, &state, 0)) {
220 		case UNVIS_VALID:
221 			dst++;
222 			break;
223 		case UNVIS_VALIDPUSH:
224 			dst++;
225 			goto again;
226 		case 0:
227 		case UNVIS_NOCHAR:
228 			break;
229 		default:
230 			*dst = '\0';
231 			return (-1);
232 		}
233 	}
234 	if (unvis(dst, c, &state, UNVIS_END) == UNVIS_VALID)
235 		dst++;
236 	*dst = '\0';
237 	return (dst - start);
238 }
239 
240 ssize_t
241 strnunvis(char *dst, const char *src, size_t sz)
242 {
243 	char c, p;
244 	char *start = dst, *end = dst + sz - 1;
245 	int state = 0;
246 
247 	if (sz > 0)
248 		*end = '\0';
249 	while ((c = *src++)) {
250 	again:
251 		switch (unvis(&p, c, &state, 0)) {
252 		case UNVIS_VALID:
253 			if (dst < end)
254 				*dst = p;
255 			dst++;
256 			break;
257 		case UNVIS_VALIDPUSH:
258 			if (dst < end)
259 				*dst = p;
260 			dst++;
261 			goto again;
262 		case 0:
263 		case UNVIS_NOCHAR:
264 			break;
265 		default:
266 			if (dst <= end)
267 				*dst = '\0';
268 			return (-1);
269 		}
270 	}
271 	if (unvis(&p, c, &state, UNVIS_END) == UNVIS_VALID) {
272 		if (dst < end)
273 			*dst = p;
274 		dst++;
275 	}
276 	if (dst <= end)
277 		*dst = '\0';
278 	return (dst - start);
279 }
280 
281