1/* $NetBSD: db_lex.c,v 1.22 2011/05/26 15:34:14 joerg Exp $ */
2
3/*
4 * Mach Operating System
5 * Copyright (c) 1991,1990 Carnegie Mellon University
6 * All Rights Reserved.
7 *
8 * Permission to use, copy, modify and distribute this software and its
9 * documentation is hereby granted, provided that both the copyright
10 * notice and this permission notice appear in all copies of the
11 * software, derivative works or modified versions, and any portions
12 * thereof, and that both notices appear in supporting documentation.
13 *
14 * CARNEGIE MELLON ALLOWS FREE USE OF THIS SOFTWARE IN ITS "AS IS"
15 * CONDITION. CARNEGIE MELLON DISCLAIMS ANY LIABILITY OF ANY KIND FOR
16 * ANY DAMAGES WHATSOEVER RESULTING FROM THE USE OF THIS SOFTWARE.
17 *
18 * Carnegie Mellon requests users of this software to return to
19 *
20 * Software Distribution Coordinator or Software.Distribution@CS.CMU.EDU
21 * School of Computer Science
22 * Carnegie Mellon University
23 * Pittsburgh PA 15213-3890
24 *
25 * any improvements or extensions that they make and grant Carnegie the
26 * rights to redistribute these changes.
27 *
28 * Author: David B. Golub, Carnegie Mellon University
29 * Date: 7/90
30 */
31
32/*
33 * Lexical analyzer.
34 */
35
36#include <sys/cdefs.h>
37__KERNEL_RCSID(0, "$NetBSD: db_lex.c,v 1.22 2011/05/26 15:34:14 joerg Exp $");
38
39#include <sys/param.h>
40#include <sys/systm.h>
41
42#include <ddb/ddb.h>
43
44db_expr_t db_tok_number;
45char db_tok_string[TOK_STRING_SIZE];
46
47static char db_line[DB_LINE_MAXLEN];
48static const char *db_lp;
49static const char *db_endlp;
50
51static int db_look_char = 0;
52static int db_look_token = 0;
53
54static void db_flush_line(void);
55static int db_read_char(void);
56static void db_unread_char(int);
57static int db_lex(void);
58
59int
60db_read_line(void)
61{
62 int i;
63
64 i = db_readline(db_line, sizeof(db_line));
65 if (i == 0)
66 return (0); /* EOI */
67 db_set_line(db_line, db_line + i);
68 return (i);
69}
70
71void
72db_set_line(const char *sp, const char *ep)
73{
74
75 db_lp = sp;
76 db_endlp = ep;
77}
78
79static void
80db_flush_line(void)
81{
82
83 db_lp = db_line;
84 db_endlp = db_line;
85}
86
87static int
88db_read_char(void)
89{
90 int c;
91
92 if (db_look_char != 0) {
93 c = db_look_char;
94 db_look_char = 0;
95 }
96 else if (db_lp >= db_endlp)
97 c = -1;
98 else
99 c = *db_lp++;
100 return (c);
101}
102
103static void
104db_unread_char(int c)
105{
106
107 db_look_char = c;
108}
109
110void
111db_unread_token(int t)
112{
113
114 db_look_token = t;
115}
116
117int
118db_read_token(void)
119{
120 int t;
121
122 if (db_look_token) {
123 t = db_look_token;
124 db_look_token = 0;
125 }
126 else
127 t = db_lex();
128 return (t);
129}
130
131int db_radix = 16;
132
133/*
134 * Convert the number to a string in the current radix.
135 * This replaces the non-standard %n printf() format.
136 */
137
138char *
139db_num_to_str(db_expr_t val)
140{
141
142 /*
143 * 2 chars for "0x", 1 for a sign ("-")
144 * up to 21 chars for a 64-bit number:
145 * % echo 2^64 | bc | wc -c
146 * 21
147 * and 1 char for a terminal NUL
148 * 2+1+21+1 => 25
149 */
150 static char buf[25];
151
152 if (db_radix == 16)
153 snprintf(buf, sizeof(buf), "%" DDB_EXPR_FMT "x", val);
154 else if (db_radix == 8)
155 snprintf(buf, sizeof(buf), "%" DDB_EXPR_FMT "o", val);
156 else
157 snprintf(buf, sizeof(buf), "%" DDB_EXPR_FMT "u", val);
158
159 return (buf);
160}
161
162void
163db_flush_lex(void)
164{
165
166 db_flush_line();
167 db_look_char = 0;
168 db_look_token = 0;
169}
170
171static int
172db_lex(void)
173{
174 int c;
175
176 c = db_read_char();
177 while (c <= ' ' || c > '~') {
178 if (c == '\n' || c == -1)
179 return (tEOL);
180 c = db_read_char();
181 }
182
183 if (c >= '0' && c <= '9') {
184 /* number */
185 db_expr_t r, digit = 0;
186
187 if (c > '0')
188 r = db_radix;
189 else {
190 c = db_read_char();
191 if (c == 'O' || c == 'o')
192 r = 8;
193 else if (c == 'T' || c == 't')
194 r = 10;
195 else if (c == 'X' || c == 'x')
196 r = 16;
197 else {
198 r = db_radix;
199 db_unread_char(c);
200 }
201 c = db_read_char();
202 }
203 db_tok_number = 0;
204 for (;;) {
205 if (c >= '0' && c <= ((r == 8) ? '7' : '9'))
206 digit = c - '0';
207 else if (r == 16 && ((c >= 'A' && c <= 'F') ||
208 (c >= 'a' && c <= 'f'))) {
209 if (c >= 'a')
210 digit = c - 'a' + 10;
211 else if (c >= 'A')
212 digit = c - 'A' + 10;
213 }
214 else
215 break;
216 db_tok_number = db_tok_number * r + digit;
217 c = db_read_char();
218 }
219 if ((c >= '0' && c <= '9') ||
220 (c >= 'A' && c <= 'Z') ||
221 (c >= 'a' && c <= 'z') ||
222 (c == '_')) {
223 db_error("Bad character in number\n");
224 /*NOTREACHED*/
225 }
226 db_unread_char(c);
227 return (tNUMBER);
228 }
229 if ((c >= 'A' && c <= 'Z') ||
230 (c >= 'a' && c <= 'z') ||
231 c == '_' || c == '\\') {
232 /* string */
233 char *cp;
234
235 cp = db_tok_string;
236 if (c == '\\') {
237 c = db_read_char();
238 if (c == '\n' || c == -1) {
239 db_error("Bad escape\n");
240 /*NOTREACHED*/
241 }
242 }
243 *cp++ = c;
244 while (1) {
245 c = db_read_char();
246 if ((c >= 'A' && c <= 'Z') ||
247 (c >= 'a' && c <= 'z') ||
248 (c >= '0' && c <= '9') ||
249 c == '_' || c == '\\' || c == ':') {
250 if (c == '\\') {
251 c = db_read_char();
252 if (c == '\n' || c == -1) {
253 db_error("Bad escape\n");
254 /*NOTREACHED*/
255 }
256 }
257 *cp++ = c;
258 if (cp == db_tok_string+sizeof(db_tok_string)) {
259 db_error("String too long\n");
260 /*NOTREACHED*/
261 }
262 continue;
263 } else {
264 *cp = '\0';
265 break;
266 }
267 }
268 db_unread_char(c);
269 return (tIDENT);
270 }
271
272 switch (c) {
273 case '+':
274 return (tPLUS);
275 case '-':
276 return (tMINUS);
277 case '.':
278 c = db_read_char();
279 if (c == '.')
280 return (tDOTDOT);
281 db_unread_char(c);
282 return (tDOT);
283 case '*':
284 return (tSTAR);
285 case '/':
286 return (tSLASH);
287 case '=':
288 return (tEQ);
289 case '%':
290 return (tPCT);
291 case '#':
292 return (tHASH);
293 case '(':
294 return (tLPAREN);
295 case ')':
296 return (tRPAREN);
297 case ',':
298 return (tCOMMA);
299 case '"':
300 return (tDITTO);
301 case '$':
302 return (tDOLLAR);
303 case '!':
304 return (tEXCL);
305 case '<':
306 c = db_read_char();
307 if (c == '<')
308 return (tSHIFT_L);
309 db_unread_char(c);
310 break;
311 case '>':
312 c = db_read_char();
313 if (c == '>')
314 return (tSHIFT_R);
315 db_unread_char(c);
316 break;
317 case -1:
318 return (tEOF);
319 }
320 db_printf("Bad character\n");
321 db_flush_lex();
322 return (tEOF);
323}
324
325/*
326 * Utility routine - discard tokens through end-of-line.
327 */
328void
329db_skip_to_eol(void)
330{
331 int t;
332
333 do {
334 t = db_read_token();
335 } while (t != tEOL);
336}
337
338void
339db_error(const char *s)
340{
341
342 if (s)
343 db_printf("%s", s);
344 db_flush_lex();
345 longjmp(db_recover);
346}
347