xref: /freebsd/usr.sbin/jail/jaillex.l (revision eb69d1f144a6fcc765d1b9d44a5ae8082353e70b)
1 %{
2 /*-
3  * SPDX-License-Identifier: BSD-2-Clause-FreeBSD
4  *
5  * Copyright (c) 2011 James Gritton
6  * 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.
16  *
17  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27  * SUCH DAMAGE.
28  */
29 
30 #include <sys/cdefs.h>
31 __FBSDID("$FreeBSD$");
32 
33 #include <err.h>
34 #include <stddef.h>
35 #include <stdlib.h>
36 #include <string.h>
37 
38 #include "jailp.h"
39 #include "y.tab.h"
40 
41 extern int yynerrs;
42 
43 static ssize_t text2lval(size_t triml, size_t trimr, int tovar);
44 
45 static int instr;
46 static int lineno = 1;
47 
48 #define YY_DECL int yylex(void)
49 %}
50 
51 %option noinput
52 %option nounput
53 
54 %start _ DQ
55 
56 %%
57 
58 			/* Whitespace or equivalent */
59 <_>[ \t]+		instr = 0;
60 <_>#.*			;
61 <_>\/\/.*		;
62 <_>\/\*([^*]|(\*+([^*\/])))*\*+\/ {
63 				const char *s;
64 
65 				for (s = yytext; s < yytext + yyleng; s++)
66 					if (*s == '\n')
67 						lineno++;
68 				instr = 0;
69 			}
70 <_>\n			{
71 				lineno++;
72 				instr = 0;
73 			}
74 
75 			/* Reserved tokens */
76 <_>\+=			{
77 				instr = 0;
78 				return PLEQ;
79 			}
80 <_>[,;={}]		{
81 				instr = 0;
82 				return yytext[0];
83 			}
84 
85 			/* Atomic (unquoted) strings */
86 <_,DQ>[A-Za-z0-9_!%&()\-.:<>?@\[\]^`|~]+ |
87 <_,DQ>\\(.|\n|[0-7]{1,3}|x[0-9A-Fa-f]{1,2}) |
88 <_,DQ>[$*+/\\]		{
89 				(void)text2lval(0, 0, 0);
90 				return instr ? STR1 : (instr = 1, STR);
91 			}
92 
93 			/* Single and double quoted strings */
94 <_>'([^\'\\]|\\(.|\n))*' {
95 				(void)text2lval(1, 1, 0);
96 				return instr ? STR1 : (instr = 1, STR);
97 			}
98 <_>\"([^"\\]|\\(.|\n))*\" |
99 <DQ>[^\"$\\]([^"\\]|\\(.|\n))*\" {
100 				size_t skip;
101 				ssize_t atvar;
102 
103 				skip = yytext[0] == '"' ? 1 : 0;
104 				atvar = text2lval(skip, 1, 1);
105 				if (atvar < 0)
106 					BEGIN _;
107 				else {
108 					/*
109 					 * The string has a variable inside it.
110 					 * Go into DQ mode to get the variable
111 					 * and then the rest of the string.
112 					 */
113 					BEGIN DQ;
114 					yyless(atvar);
115 				}
116 				return instr ? STR1 : (instr = 1, STR);
117 			}
118 <DQ>\"			BEGIN _;
119 
120 			/* Variables, single-word or bracketed */
121 <_,DQ>$[A-Za-z_][A-Za-z_0-9]* {
122 				(void)text2lval(1, 0, 0);
123 				return instr ? VAR1 : (instr = 1, VAR);
124 			}
125 <_>$\{([^\n{}]|\\(.|\n))*\} |
126 <DQ>$\{([^\n\"{}]|\\(.|\n))*\} {
127 				(void)text2lval(2, 1, 0);
128 				return instr ? VAR1 : (instr = 1, VAR);
129 			}
130 
131 			/* Partially formed bits worth complaining about */
132 <_>\/\*([^*]|(\*+([^*\/])))*\** {
133 				warnx("%s line %d: unterminated comment",
134 				    cfname, lineno);
135 				yynerrs++;
136 			}
137 <_>'([^\n'\\]|\\.)*	|
138 <_>\"([^\n\"\\]|\\.)*	{
139 				warnx("%s line %d: unterminated string",
140 				    cfname, lineno);
141 				yynerrs++;
142 			}
143 <_>$\{([^\n{}]|\\.)*	|
144 <DQ>$\{([^\n\"{}]|\\.)*	{
145 				warnx("%s line %d: unterminated variable",
146 				    cfname, lineno);
147 				yynerrs++;
148 			}
149 
150 			/* A hack because "<0>" rules aren't allowed */
151 <_>.			return yytext[0];
152 .|\n			{
153 				BEGIN _;
154 				yyless(0);
155 			}
156 
157 %%
158 
159 void
160 yyerror(const char *s)
161 {
162 	if (!yytext)
163 		warnx("%s line %d: %s", cfname, lineno, s);
164 	else if (!yytext[0])
165 		warnx("%s: unexpected EOF", cfname);
166 	else
167 		warnx("%s line %d: %s: %s", cfname, lineno, yytext, s);
168 }
169 
170 /*
171  * Copy string from yytext to yylval, handling backslash escapes,
172  * and optionally stopping at the beginning of a variable.
173  */
174 static ssize_t
175 text2lval(size_t triml, size_t trimr, int tovar)
176 {
177 	char *d;
178 	const char *s, *se;
179 
180 	yylval.cs = d = emalloc(yyleng - trimr - triml + 1);
181 	se = yytext + (yyleng - trimr);
182 	for (s = yytext + triml; s < se; s++, d++) {
183 		if (*s != '\\') {
184 			if (tovar && *s == '$') {
185 				*d = '\0';
186 				return s - yytext;
187 			}
188 			if (*s == '\n')
189 				lineno++;
190 			*d = *s;
191 			continue;
192 		}
193 		s++;
194 		if (*s >= '0' && *s <= '7') {
195 			*d = *s - '0';
196 			if (s + 1 < se && s[1] >= '0' && s[1] <= '7') {
197 				*d = 010 * *d + (*++s - '0');
198 				if (s + 1 < se && s[1] >= '0' && s[1] <= '7')
199 					*d = 010 * *d + (*++s - '0');
200 			}
201 			continue;
202 		}
203 		switch (*s) {
204 		case 'a':	*d = '\a';	break;
205 		case 'b':	*d = '\b';	break;
206 		case 'f':	*d = '\f';	break;
207 		case 'n':	*d = '\n';	break;
208 		case 'r':	*d = '\r';	break;
209 		case 't':	*d = '\t';	break;
210 		case 'v':	*d = '\v';	break;
211 		case '\n':	d--; lineno++;	break;
212 		default:	*d = *s;	break;
213 		case 'x':
214 			*d = 0;
215 			if (s + 1 >= se)
216 				break;
217 			if (s[1] >= '0' && s[1] <= '9')
218 				*d = *++s - '0';
219 			else if (s[1] >= 'A' && s[1] <= 'F')
220 				*d = *++s + (0xA - 'A');
221 			else if (s[1] >= 'a' && s[1] <= 'f')
222 				*d = *++s + (0xa - 'a');
223 			else
224 				break;
225 			if (s + 1 >= se)
226 				break;
227 			if (s[1] >= '0' && s[1] <= '9')
228 				*d = *d * 0x10 + (*++s - '0');
229 			else if (s[1] >= 'A' && s[1] <= 'F')
230 				*d = *d * 0x10 + (*++s + (0xA - 'A'));
231 			else if (s[1] >= 'a' && s[1] <= 'f')
232 				*d = *d * 0x10 + (*++s + (0xa - 'a'));
233 		}
234 	}
235 	*d = '\0';
236 	return -1;
237 }
238