xref: /freebsd/sys/ddb/db_lex.c (revision 641a6cfb86023499caafe26a4d821a0b885cf00b)
1 /*-
2  * Mach Operating System
3  * Copyright (c) 1991,1990 Carnegie Mellon University
4  * All Rights Reserved.
5  *
6  * Permission to use, copy, modify and distribute this software and its
7  * documentation is hereby granted, provided that both the copyright
8  * notice and this permission notice appear in all copies of the
9  * software, derivative works or modified versions, and any portions
10  * thereof, and that both notices appear in supporting documentation.
11  *
12  * CARNEGIE MELLON ALLOWS FREE USE OF THIS SOFTWARE IN ITS
13  * CONDITION.  CARNEGIE MELLON DISCLAIMS ANY LIABILITY OF ANY KIND FOR
14  * ANY DAMAGES WHATSOEVER RESULTING FROM THE USE OF THIS SOFTWARE.
15  *
16  * Carnegie Mellon requests users of this software to return to
17  *
18  *  Software Distribution Coordinator  or  Software.Distribution@CS.CMU.EDU
19  *  School of Computer Science
20  *  Carnegie Mellon University
21  *  Pittsburgh PA 15213-3890
22  *
23  * any improvements or extensions that they make and grant Carnegie the
24  * rights to redistribute these changes.
25  */
26 /*
27  *	Author: David B. Golub, Carnegie Mellon University
28  *	Date:	7/90
29  */
30 /*
31  * Lexical analyzer.
32  */
33 
34 #include <sys/cdefs.h>
35 __FBSDID("$FreeBSD$");
36 
37 #include <sys/param.h>
38 #include <sys/libkern.h>
39 
40 #include <ddb/ddb.h>
41 #include <ddb/db_lex.h>
42 
43 static char	db_line[DB_MAXLINE];
44 static char *	db_lp, *db_endlp;
45 
46 static int	db_lex(void);
47 static void 	db_flush_line(void);
48 static int 	db_read_char(void);
49 static void 	db_unread_char(int);
50 
51 int
52 db_read_line()
53 {
54 	int	i;
55 
56 	i = db_readline(db_line, sizeof(db_line));
57 	if (i == 0)
58 	    return (0);	/* EOI */
59 	db_lp = db_line;
60 	db_endlp = db_lp + i;
61 	return (i);
62 }
63 
64 /*
65  * Simulate a line of input into DDB.
66  */
67 void
68 db_inject_line(const char *command)
69 {
70 
71 	strlcpy(db_line, command, sizeof(db_line));
72 	db_lp = db_line;
73 	db_endlp = db_lp + strlen(command);
74 }
75 
76 /*
77  * In rare cases, we may want to pull the remainder of the line input
78  * verbatim, rather than lexing it.  For example, when assigning literal
79  * values associated with scripts.  In that case, return a static pointer to
80  * the current location in the input buffer.  The caller must be aware that
81  * the contents are not stable if other lex/input calls are made.
82  */
83 char *
84 db_get_line(void)
85 {
86 
87 	return (db_lp);
88 }
89 
90 static void
91 db_flush_line()
92 {
93 	db_lp = db_line;
94 	db_endlp = db_line;
95 }
96 
97 static int	db_look_char = 0;
98 
99 static int
100 db_read_char()
101 {
102 	int	c;
103 
104 	if (db_look_char != 0) {
105 	    c = db_look_char;
106 	    db_look_char = 0;
107 	}
108 	else if (db_lp >= db_endlp)
109 	    c = -1;
110 	else
111 	    c = *db_lp++;
112 	return (c);
113 }
114 
115 static void
116 db_unread_char(c)
117 	int c;
118 {
119 	db_look_char = c;
120 }
121 
122 static int	db_look_token = 0;
123 
124 void
125 db_unread_token(t)
126 	int	t;
127 {
128 	db_look_token = t;
129 }
130 
131 int
132 db_read_token()
133 {
134 	int	t;
135 
136 	if (db_look_token) {
137 	    t = db_look_token;
138 	    db_look_token = 0;
139 	}
140 	else
141 	    t = db_lex();
142 	return (t);
143 }
144 
145 db_expr_t	db_tok_number;
146 char	db_tok_string[TOK_STRING_SIZE];
147 
148 db_expr_t	db_radix = 16;
149 
150 void
151 db_flush_lex()
152 {
153 	db_flush_line();
154 	db_look_char = 0;
155 	db_look_token = 0;
156 }
157 
158 static int
159 db_lex()
160 {
161 	int	c;
162 
163 	c = db_read_char();
164 	while (c <= ' ' || c > '~') {
165 	    if (c == '\n' || c == -1)
166 		return (tEOL);
167 	    c = db_read_char();
168 	}
169 
170 	if (c >= '0' && c <= '9') {
171 	    /* number */
172 	    int	r, digit = 0;
173 
174 	    if (c > '0')
175 		r = db_radix;
176 	    else {
177 		c = db_read_char();
178 		if (c == 'O' || c == 'o')
179 		    r = 8;
180 		else if (c == 'T' || c == 't')
181 		    r = 10;
182 		else if (c == 'X' || c == 'x')
183 		    r = 16;
184 		else {
185 		    r = db_radix;
186 		    db_unread_char(c);
187 		}
188 		c = db_read_char();
189 	    }
190 	    db_tok_number = 0;
191 	    for (;;) {
192 		if (c >= '0' && c <= ((r == 8) ? '7' : '9'))
193 		    digit = c - '0';
194 		else if (r == 16 && ((c >= 'A' && c <= 'F') ||
195 				     (c >= 'a' && c <= 'f'))) {
196 		    if (c >= 'a')
197 			digit = c - 'a' + 10;
198 		    else if (c >= 'A')
199 			digit = c - 'A' + 10;
200 		}
201 		else
202 		    break;
203 		db_tok_number = db_tok_number * r + digit;
204 		c = db_read_char();
205 	    }
206 	    if ((c >= '0' && c <= '9') ||
207 		(c >= 'A' && c <= 'Z') ||
208 		(c >= 'a' && c <= 'z') ||
209 		(c == '_'))
210 	    {
211 		db_error("Bad character in number\n");
212 		db_flush_lex();
213 		return (tEOF);
214 	    }
215 	    db_unread_char(c);
216 	    return (tNUMBER);
217 	}
218 	if ((c >= 'A' && c <= 'Z') ||
219 	    (c >= 'a' && c <= 'z') ||
220 	    c == '_' || c == '\\')
221 	{
222 	    /* string */
223 	    char *cp;
224 
225 	    cp = db_tok_string;
226 	    if (c == '\\') {
227 		c = db_read_char();
228 		if (c == '\n' || c == -1)
229 		    db_error("Bad escape\n");
230 	    }
231 	    *cp++ = c;
232 	    while (1) {
233 		c = db_read_char();
234 		if ((c >= 'A' && c <= 'Z') ||
235 		    (c >= 'a' && c <= 'z') ||
236 		    (c >= '0' && c <= '9') ||
237 		    c == '_' || c == '\\' || c == ':' || c == '.')
238 		{
239 		    if (c == '\\') {
240 			c = db_read_char();
241 			if (c == '\n' || c == -1)
242 			    db_error("Bad escape\n");
243 		    }
244 		    *cp++ = c;
245 		    if (cp == db_tok_string+sizeof(db_tok_string)) {
246 			db_error("String too long\n");
247 			db_flush_lex();
248 			return (tEOF);
249 		    }
250 		    continue;
251 		}
252 		else {
253 		    *cp = '\0';
254 		    break;
255 		}
256 	    }
257 	    db_unread_char(c);
258 	    return (tIDENT);
259 	}
260 
261 	switch (c) {
262 	    case '+':
263 		return (tPLUS);
264 	    case '-':
265 		return (tMINUS);
266 	    case '.':
267 		c = db_read_char();
268 		if (c == '.')
269 		    return (tDOTDOT);
270 		db_unread_char(c);
271 		return (tDOT);
272 	    case '*':
273 		return (tSTAR);
274 	    case '/':
275 		return (tSLASH);
276 	    case '=':
277 		return (tEQ);
278 	    case '%':
279 		return (tPCT);
280 	    case '#':
281 		return (tHASH);
282 	    case '(':
283 		return (tLPAREN);
284 	    case ')':
285 		return (tRPAREN);
286 	    case ',':
287 		return (tCOMMA);
288 	    case '"':
289 		return (tDITTO);
290 	    case '$':
291 		return (tDOLLAR);
292 	    case '!':
293 		return (tEXCL);
294 	    case ';':
295 		return (tSEMI);
296 	    case '<':
297 		c = db_read_char();
298 		if (c == '<')
299 		    return (tSHIFT_L);
300 		db_unread_char(c);
301 		break;
302 	    case '>':
303 		c = db_read_char();
304 		if (c == '>')
305 		    return (tSHIFT_R);
306 		db_unread_char(c);
307 		break;
308 	    case -1:
309 		return (tEOF);
310 	}
311 	db_printf("Bad character\n");
312 	db_flush_lex();
313 	return (tEOF);
314 }
315