xref: /illumos-gate/usr/src/cmd/sgs/libconv/common/c_literal.c (revision 9ab815e1e50104cb1004a5ccca7a6da582994b57)
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 2010 Sun Microsystems, Inc.  All rights reserved.
24  * Use is subject to license terms.
25  */
26 
27 
28 /*
29  * Translate a string into C literal string constant notation.
30  */
31 
32 #include	<stdio.h>
33 #include	<ctype.h>
34 #include	<_conv.h>
35 #include	<c_literal_msg.h>
36 
37 
38 /*
39  * Convert characters to the form used by the C language to represent
40  * literal strings:
41  *	- Printable characters are shown as themselves
42  *	- Convert special characters to their 2-character escaped forms:
43  *		alert (bell)	\a
44  *		backspace	\b
45  *		formfeed	\f
46  *		newline		\n
47  *		return		\r
48  *		horizontal tab	\t
49  *		vertical tab	\v
50  *		backspace	\\
51  *		single quote	\'
52  *		double quote	\"
53  *	- Display other non-printable characters as 4-character escaped
54  *		octal constants.
55  *
56  * entry:
57  *	buf - Buffer of characters to be processed
58  *	n # of characters in buf to be processed
59  *	outfunc - Function to be called to move output characters.
60  *	uvalue - User value. This argument is passed to outfunc without
61  *		examination. The caller can use it to pass additional
62  *		information required by the callback.
63  *
64  * exit:
65  *	The string has been processed, with the resulting data passed
66  *	to outfunc for processing.
67  */
68 void
69 conv_str_to_c_literal(const char *buf, size_t n,
70     Conv_str_to_c_literal_func_t *outfunc, void *uvalue)
71 {
72 	char	bs_buf[2];	/* For two-character backslash codes */
73 	char	octal_buf[10];	/* For \000 style octal constants */
74 
75 	bs_buf[0] = '\\';
76 	while (n > 0) {
77 		switch (*buf) {
78 		case '\0':
79 			bs_buf[1] = '0';
80 			break;
81 		case '\a':
82 			bs_buf[1] = 'a';
83 			break;
84 		case '\b':
85 			bs_buf[1] = 'b';
86 			break;
87 		case '\f':
88 			bs_buf[1] = 'f';
89 			break;
90 		case '\n':
91 			bs_buf[1] = 'n';
92 			break;
93 		case '\r':
94 			bs_buf[1] = 'r';
95 			break;
96 		case '\t':
97 			bs_buf[1] = 't';
98 			break;
99 		case '\v':
100 			bs_buf[1] = 'v';
101 			break;
102 		case '\\':
103 			bs_buf[1] = '\\';
104 			break;
105 		case '\'':
106 			bs_buf[1] = '\'';
107 			break;
108 		case '"':
109 			bs_buf[1] = '"';
110 			break;
111 		default:
112 			bs_buf[1] = '\0';
113 		}
114 
115 		if (bs_buf[1] != '\0') {
116 			(*outfunc)(bs_buf, 2, uvalue);
117 			buf++;
118 			n--;
119 		} else if (isprint(*buf)) {
120 			/*
121 			 * Output the entire sequence of printable
122 			 * characters in a single shot.
123 			 */
124 			const char	*start = buf;
125 			size_t		outlen = 0;
126 
127 			for (start = buf; (n > 0) && isprint(*buf); buf++, n--)
128 				outlen++;
129 			(*outfunc)(start, outlen, uvalue);
130 		} else {
131 			/* Generic unprintable character: Use octal notation */
132 			(void) snprintf(octal_buf, sizeof (octal_buf),
133 			    MSG_ORIG(MSG_FMT_OCTCONST), (uchar_t)*buf);
134 			(*outfunc)(octal_buf, strlen(octal_buf), uvalue);
135 			buf++;
136 			n--;
137 		}
138 	}
139 }
140 
141 /*
142  * Given the pointer to the character following a '\' character in
143  * a C style literal, return the ASCII character code it represents,
144  * and advance the string pointer to the character following the last
145  * character in the escape sequence.
146  *
147  * entry:
148  *	str - Address of string pointer to first character following
149  *		the backslash.
150  *
151  * exit:
152  *	If the character is not valid, -1 is returned. Otherwise
153  *	it returns the ASCII code for the translated character, and
154  *	*str has been advanced.
155  */
156 int
157 conv_translate_c_esc(char **str)
158 {
159 	char	*s = *str;
160 	int	ch, i;
161 
162 	ch = *s++;
163 	switch (ch) {
164 	case 'a':
165 		ch = '\a';
166 		break;
167 	case 'b':
168 		ch = '\b';
169 		break;
170 	case 'f':
171 		ch = '\f';
172 		break;
173 	case 'n':
174 		ch = '\n';
175 		break;
176 	case 'r':
177 		ch = '\r';
178 		break;
179 	case 't':
180 		ch = '\t';
181 		break;
182 	case 'v':
183 		ch = '\v';
184 		break;
185 
186 	case '0':
187 	case '1':
188 	case '2':
189 	case '3':
190 	case '4':
191 	case '5':
192 	case '6':
193 	case '7':
194 		/* Octal constant: There can be up to 3 digits */
195 		ch -= '0';
196 		for (i = 0; i < 2; i++) {
197 			if ((*s < '0') || (*s > '7'))
198 				break;
199 			ch = (ch << 3) + (*s++ - '0');
200 		}
201 		break;
202 
203 	/*
204 	 * There are some cases where ch already has the desired value.
205 	 * These cases exist simply to remove the special meaning that
206 	 * character would otherwise have. We need to match them to
207 	 * prevent them from falling into the default error case.
208 	 */
209 	case '\\':
210 	case '\'':
211 	case '"':
212 		break;
213 
214 	default:
215 		ch = -1;
216 		break;
217 	}
218 
219 	*str = s;
220 	return (ch);
221 }
222