xref: /illumos-gate/usr/src/lib/libdtrace/common/dt_string.c (revision 1007fd6fd24227460e77ce89f5ca85641a85a576)
1 /*
2  * CDDL HEADER START
3  *
4  * The contents of this file are subject to the terms of the
5  * Common Development and Distribution License (the "License").
6  * You may not use this file except in compliance with the License.
7  *
8  * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
9  * or http://www.opensolaris.org/os/licensing.
10  * See the License for the specific language governing permissions
11  * and limitations under the License.
12  *
13  * When distributing Covered Code, include this CDDL HEADER in each
14  * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
15  * If applicable, add the following below this CDDL HEADER, with the
16  * fields enclosed by brackets "[]" replaced with your own identifying
17  * information: Portions Copyright [yyyy] [name of copyright owner]
18  *
19  * CDDL HEADER END
20  */
21 
22 /*
23  * Copyright (c) 2003, 2010, Oracle and/or its affiliates. All rights reserved.
24  */
25 
26 #include <strings.h>
27 #include <stdlib.h>
28 #include <errno.h>
29 #include <ctype.h>
30 
31 #include <dt_string.h>
32 #include <dt_impl.h>
33 
34 /*
35  * Create a copy of string s, but only duplicate the first n bytes.
36  */
37 char *
38 strndup(const char *s, size_t n)
39 {
40 	char *s2 = malloc(n + 1);
41 
42 	if (s2 == NULL)
43 		longjmp(yypcb->pcb_jmpbuf, EDT_NOMEM);
44 
45 	(void) strncpy(s2, s, n);
46 	s2[n] = '\0';
47 	return (s2);
48 }
49 
50 /*
51  * Transform string s inline, converting each embedded C escape sequence string
52  * to the corresponding character.  For example, the substring "\n" is replaced
53  * by an inline '\n' character.  The length of the resulting string is returned.
54  */
55 size_t
56 stresc2chr(char *s)
57 {
58 	char *p, *q, c;
59 	int esc = 0;
60 	int x;
61 
62 	for (p = q = s; (c = *p) != '\0'; p++) {
63 		if (esc) {
64 			switch (c) {
65 			case '0':
66 			case '1':
67 			case '2':
68 			case '3':
69 			case '4':
70 			case '5':
71 			case '6':
72 			case '7':
73 				c -= '0';
74 				p++;
75 
76 				if (*p >= '0' && *p <= '7') {
77 					c = c * 8 + *p++ - '0';
78 
79 					if (*p >= '0' && *p <= '7')
80 						c = c * 8 + *p - '0';
81 					else
82 						p--;
83 				} else
84 					p--;
85 
86 				*q++ = c;
87 				break;
88 
89 			case 'a':
90 				*q++ = '\a';
91 				break;
92 			case 'b':
93 				*q++ = '\b';
94 				break;
95 			case 'f':
96 				*q++ = '\f';
97 				break;
98 			case 'n':
99 				*q++ = '\n';
100 				break;
101 			case 'r':
102 				*q++ = '\r';
103 				break;
104 			case 't':
105 				*q++ = '\t';
106 				break;
107 			case 'v':
108 				*q++ = '\v';
109 				break;
110 
111 			case 'x':
112 				for (x = 0; (c = *++p) != '\0'; ) {
113 					if (c >= '0' && c <= '9')
114 						x = x * 16 + c - '0';
115 					else if (c >= 'a' && c <= 'f')
116 						x = x * 16 + c - 'a' + 10;
117 					else if (c >= 'A' && c <= 'F')
118 						x = x * 16 + c - 'A' + 10;
119 					else
120 						break;
121 				}
122 				*q++ = (char)x;
123 				p--;
124 				break;
125 
126 			case '"':
127 			case '\\':
128 				*q++ = c;
129 				break;
130 			default:
131 				*q++ = '\\';
132 				*q++ = c;
133 			}
134 
135 			esc = 0;
136 
137 		} else {
138 			if ((esc = c == '\\') == 0)
139 				*q++ = c;
140 		}
141 	}
142 
143 	*q = '\0';
144 	return ((size_t)(q - s));
145 }
146 
147 /*
148  * Create a copy of string s in which certain unprintable or special characters
149  * have been converted to the string representation of their C escape sequence.
150  * For example, the newline character is expanded to the string "\n".
151  */
152 char *
153 strchr2esc(const char *s, size_t n)
154 {
155 	const char *p;
156 	char *q, *s2, c;
157 	size_t addl = 0;
158 
159 	for (p = s; p < s + n; p++) {
160 		switch (c = *p) {
161 		case '\0':
162 		case '\a':
163 		case '\b':
164 		case '\f':
165 		case '\n':
166 		case '\r':
167 		case '\t':
168 		case '\v':
169 		case '"':
170 		case '\\':
171 			addl++;		/* 1 add'l char needed to follow \ */
172 			break;
173 		case ' ':
174 			break;
175 		default:
176 			if (c < '!' || c > '~')
177 				addl += 3; /* 3 add'l chars following \ */
178 		}
179 	}
180 
181 	if ((s2 = malloc(n + addl + 1)) == NULL)
182 		return (NULL);
183 
184 	for (p = s, q = s2; p < s + n; p++) {
185 		switch (c = *p) {
186 		case '\0':
187 			*q++ = '\\';
188 			*q++ = '0';
189 			break;
190 		case '\a':
191 			*q++ = '\\';
192 			*q++ = 'a';
193 			break;
194 		case '\b':
195 			*q++ = '\\';
196 			*q++ = 'b';
197 			break;
198 		case '\f':
199 			*q++ = '\\';
200 			*q++ = 'f';
201 			break;
202 		case '\n':
203 			*q++ = '\\';
204 			*q++ = 'n';
205 			break;
206 		case '\r':
207 			*q++ = '\\';
208 			*q++ = 'r';
209 			break;
210 		case '\t':
211 			*q++ = '\\';
212 			*q++ = 't';
213 			break;
214 		case '\v':
215 			*q++ = '\\';
216 			*q++ = 'v';
217 			break;
218 		case '"':
219 			*q++ = '\\';
220 			*q++ = '"';
221 			break;
222 		case '\\':
223 			*q++ = '\\';
224 			*q++ = '\\';
225 			break;
226 		case ' ':
227 			*q++ = c;
228 			break;
229 		default:
230 			if (c < '!' || c > '~') {
231 				*q++ = '\\';
232 				*q++ = ((c >> 6) & 3) + '0';
233 				*q++ = ((c >> 3) & 7) + '0';
234 				*q++ = (c & 7) + '0';
235 			} else
236 				*q++ = c;
237 		}
238 
239 		if (c == '\0')
240 			break; /* don't continue past \0 even if p < s + n */
241 	}
242 
243 	*q = '\0';
244 	return (s2);
245 }
246 
247 /*
248  * Return the basename (name after final /) of the given string.  We use
249  * strbasename rather than basename to avoid conflicting with libgen.h's
250  * non-const function prototype.
251  */
252 const char *
253 strbasename(const char *s)
254 {
255 	const char *p = strrchr(s, '/');
256 
257 	if (p == NULL)
258 		return (s);
259 
260 	return (++p);
261 }
262 
263 /*
264  * This function tests a string against the regular expression used for idents
265  * and integers in the D lexer, and should match the superset of RGX_IDENT and
266  * RGX_INT in dt_lex.l.  If an invalid character is found, the function returns
267  * a pointer to it.  Otherwise NULL is returned for a valid string.
268  */
269 const char *
270 strbadidnum(const char *s)
271 {
272 	char *p;
273 	int c;
274 
275 	if (*s == '\0')
276 		return (s);
277 
278 	errno = 0;
279 	(void) strtoull(s, &p, 0);
280 
281 	if (errno == 0 && *p == '\0')
282 		return (NULL); /* matches RGX_INT */
283 
284 	while ((c = *s++) != '\0') {
285 		if (isalnum(c) == 0 && c != '_' && c != '`')
286 			return (s - 1);
287 	}
288 
289 	return (NULL); /* matches RGX_IDENT */
290 }
291 
292 /*
293  * Determine whether the string contains a glob matching pattern or is just a
294  * simple string.  See gmatch(3GEN) and sh(1) for the glob syntax definition.
295  */
296 int
297 strisglob(const char *s)
298 {
299 	char c;
300 
301 	while ((c = *s++) != '\0') {
302 		if (c == '[' || c == '?' || c == '*' || c == '\\')
303 			return (1);
304 	}
305 
306 	return (0);
307 }
308 
309 /*
310  * Hyphenate a string in-place by converting any instances of "__" to "-",
311  * which we use for probe names to improve readability, and return the string.
312  */
313 char *
314 strhyphenate(char *s)
315 {
316 	char *p, *q;
317 
318 	for (p = s, q = p + strlen(p); p < q; p++) {
319 		if (p[0] == '_' && p[1] == '_') {
320 			p[0] = '-';
321 			bcopy(p + 2, p + 1, (size_t)(q - p) - 1);
322 		}
323 	}
324 
325 	return (s);
326 }
327