xref: /freebsd/sys/ddb/db_lex.c (revision d0b2dbfa0ecf2bbc9709efc5e20baf8e4b44bbbf)
1 /*-
2  * SPDX-License-Identifier: MIT-CMU
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
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 /*
29  *	Author: David B. Golub, Carnegie Mellon University
30  *	Date:	7/90
31  */
32 /*
33  * Lexical analyzer.
34  */
35 
36 #include <sys/cdefs.h>
37 #include <sys/param.h>
38 #include <sys/libkern.h>
39 #include <sys/lock.h>
40 
41 #include <ddb/ddb.h>
42 #include <ddb/db_lex.h>
43 
44 static char	db_line[DB_MAXLINE];
45 static char *	db_lp, *db_endlp;
46 
47 static int	db_lex(int);
48 static void 	db_flush_line(void);
49 static int 	db_read_char(void);
50 static void 	db_unread_char(int);
51 
52 int
53 db_read_line(void)
54 {
55 	int	i;
56 
57 	i = db_readline(db_line, sizeof(db_line));
58 	if (i == 0)
59 	    return (0);	/* EOI */
60 	db_lp = db_line;
61 	db_endlp = db_lp + i;
62 	return (i);
63 }
64 
65 /*
66  * Simulate a line of input into DDB.
67  */
68 void
69 db_inject_line(const char *command)
70 {
71 
72 	strlcpy(db_line, command, sizeof(db_line));
73 	db_lp = db_line;
74 	db_endlp = db_lp + strlen(command);
75 }
76 
77 /*
78  * In rare cases, we may want to pull the remainder of the line input
79  * verbatim, rather than lexing it.  For example, when assigning literal
80  * values associated with scripts.  In that case, return a static pointer to
81  * the current location in the input buffer.  The caller must be aware that
82  * the contents are not stable if other lex/input calls are made.
83  */
84 char *
85 db_get_line(void)
86 {
87 
88 	return (db_lp);
89 }
90 
91 static void
92 db_flush_line(void)
93 {
94 	db_lp = db_line;
95 	db_endlp = db_line;
96 }
97 
98 static int
99 db_read_char(void)
100 {
101 	int	c;
102 
103 	if (db_lp >= db_endlp)
104 	    c = -1;
105 	else
106 	    c = *db_lp++;
107 	return (c);
108 }
109 
110 static void
111 db_unread_char(int c)
112 {
113 
114 	if (c == -1) {
115 		/* Unread EOL at EOL is okay. */
116 		if (db_lp < db_endlp)
117 			db_error("db_unread_char(-1) before end of line\n");
118 	} else {
119 		if (db_lp > db_line) {
120 			db_lp--;
121 			if (*db_lp != c)
122 				db_error("db_unread_char() wrong char\n");
123 		} else {
124 			db_error("db_unread_char() at beginning of line\n");
125 		}
126 	}
127 }
128 
129 static int	db_look_token = 0;
130 
131 void
132 db_unread_token(int t)
133 {
134 	db_look_token = t;
135 }
136 
137 int
138 db_read_token_flags(int flags)
139 {
140 	int	t;
141 
142 	MPASS((flags & ~(DRT_VALID_FLAGS_MASK)) == 0);
143 
144 	if (db_look_token) {
145 	    t = db_look_token;
146 	    db_look_token = 0;
147 	}
148 	else
149 	    t = db_lex(flags);
150 	return (t);
151 }
152 
153 db_expr_t	db_tok_number;
154 char	db_tok_string[TOK_STRING_SIZE];
155 
156 db_expr_t	db_radix = 16;
157 
158 void
159 db_flush_lex(void)
160 {
161 	db_flush_line();
162 	db_look_token = 0;
163 }
164 
165 static int
166 db_lex(int flags)
167 {
168 	int	c, n, radix_mode;
169 	bool	lex_wspace, lex_hex_numbers;
170 
171 	switch (flags & DRT_RADIX_MASK) {
172 	case DRT_DEFAULT_RADIX:
173 		radix_mode = -1;
174 		break;
175 	case DRT_OCTAL:
176 		radix_mode = 8;
177 		break;
178 	case DRT_DECIMAL:
179 		radix_mode = 10;
180 		break;
181 	case DRT_HEXADECIMAL:
182 		radix_mode = 16;
183 		break;
184 	}
185 
186 	lex_wspace = ((flags & DRT_WSPACE) != 0);
187 	lex_hex_numbers = ((flags & DRT_HEX) != 0);
188 
189 	c = db_read_char();
190 	for (n = 0; c <= ' ' || c > '~'; n++) {
191 	    if (c == '\n' || c == -1)
192 		return (tEOL);
193 	    c = db_read_char();
194 	}
195 	if (lex_wspace && n != 0) {
196 	    db_unread_char(c);
197 	    return (tWSPACE);
198 	}
199 
200 	if ((c >= '0' && c <= '9') ||
201 	   (lex_hex_numbers &&
202 	   ((c >= 'a' && c <= 'f') ||
203 	   (c >= 'A' && c <= 'F')))) {
204 	    /* number */
205 	    int	r, digit = 0;
206 
207 	    if (radix_mode != -1)
208 		r = radix_mode;
209 	    else if (c != '0')
210 		r = db_radix;
211 	    else {
212 		c = db_read_char();
213 		if (c == 'O' || c == 'o')
214 		    r = 8;
215 		else if (c == 'T' || c == 't')
216 		    r = 10;
217 		else if (c == 'X' || c == 'x')
218 		    r = 16;
219 		else {
220 		    r = db_radix;
221 		    db_unread_char(c);
222 		}
223 		c = db_read_char();
224 	    }
225 	    db_tok_number = 0;
226 	    for (;;) {
227 		if (c >= '0' && c <= ((r == 8) ? '7' : '9'))
228 		    digit = c - '0';
229 		else if (r == 16 && ((c >= 'A' && c <= 'F') ||
230 				     (c >= 'a' && c <= 'f'))) {
231 		    if (c >= 'a')
232 			digit = c - 'a' + 10;
233 		    else if (c >= 'A')
234 			digit = c - 'A' + 10;
235 		}
236 		else
237 		    break;
238 		db_tok_number = db_tok_number * r + digit;
239 		c = db_read_char();
240 	    }
241 	    if ((c >= '0' && c <= '9') ||
242 		(c >= 'A' && c <= 'Z') ||
243 		(c >= 'a' && c <= 'z') ||
244 		(c == '_'))
245 	    {
246 		db_error("Bad character in number\n");
247 		db_flush_lex();
248 		return (tEOF);
249 	    }
250 	    db_unread_char(c);
251 	    return (tNUMBER);
252 	}
253 	if ((c >= 'A' && c <= 'Z') ||
254 	    (c >= 'a' && c <= 'z') ||
255 	    c == '_' || c == '\\')
256 	{
257 	    /* string */
258 	    char *cp;
259 
260 	    cp = db_tok_string;
261 	    if (c == '\\') {
262 		c = db_read_char();
263 		if (c == '\n' || c == -1)
264 		    db_error("Bad escape\n");
265 	    }
266 	    *cp++ = c;
267 	    while (1) {
268 		c = db_read_char();
269 		if ((c >= 'A' && c <= 'Z') ||
270 		    (c >= 'a' && c <= 'z') ||
271 		    (c >= '0' && c <= '9') ||
272 		    c == '_' || c == '\\' || c == ':' || c == '.')
273 		{
274 		    if (c == '\\') {
275 			c = db_read_char();
276 			if (c == '\n' || c == -1)
277 			    db_error("Bad escape\n");
278 		    }
279 		    *cp++ = c;
280 		    if (cp == db_tok_string+sizeof(db_tok_string)) {
281 			db_error("String too long\n");
282 			db_flush_lex();
283 			return (tEOF);
284 		    }
285 		    continue;
286 		}
287 		else {
288 		    *cp = '\0';
289 		    break;
290 		}
291 	    }
292 	    db_unread_char(c);
293 	    return (tIDENT);
294 	}
295 
296 	switch (c) {
297 	    case '+':
298 		return (tPLUS);
299 	    case '-':
300 		return (tMINUS);
301 	    case '.':
302 		c = db_read_char();
303 		if (c == '.')
304 		    return (tDOTDOT);
305 		db_unread_char(c);
306 		return (tDOT);
307 	    case '*':
308 		return (tSTAR);
309 	    case '/':
310 		return (tSLASH);
311 	    case '=':
312 		c = db_read_char();
313 		if (c == '=')
314 		    return (tLOG_EQ);
315 		db_unread_char(c);
316 		return (tEQ);
317 	    case '%':
318 		return (tPCT);
319 	    case '#':
320 		return (tHASH);
321 	    case '(':
322 		return (tLPAREN);
323 	    case ')':
324 		return (tRPAREN);
325 	    case ',':
326 		return (tCOMMA);
327 	    case '"':
328 		return (tDITTO);
329 	    case '$':
330 		return (tDOLLAR);
331 	    case '!':
332 		c = db_read_char();
333 		if (c == '='){
334 			return (tLOG_NOT_EQ);
335 		}
336 		db_unread_char(c);
337 		return (tEXCL);
338 	    case ':':
339 		c = db_read_char();
340 		if (c == ':')
341 			return (tCOLONCOLON);
342 		db_unread_char(c);
343 		return (tCOLON);
344 	    case ';':
345 		return (tSEMI);
346 	    case '&':
347 		c = db_read_char();
348 		if (c == '&')
349 		    return (tLOG_AND);
350 		db_unread_char(c);
351 		return (tBIT_AND);
352 	    case '|':
353 		c = db_read_char();
354 		if (c == '|')
355 		    return (tLOG_OR);
356 		db_unread_char(c);
357 		return (tBIT_OR);
358 	    case '<':
359 		c = db_read_char();
360 		if (c == '<')
361 		    return (tSHIFT_L);
362 		if (c == '=')
363 		    return (tLESS_EQ);
364 		db_unread_char(c);
365 		return (tLESS);
366 	    case '>':
367 		c = db_read_char();
368 		if (c == '>')
369 		    return (tSHIFT_R);
370 		if (c == '=')
371 		    return (tGREATER_EQ);
372 		db_unread_char(c);
373 		return (tGREATER);
374 	    case '?':
375 		return (tQUESTION);
376 	    case '~':
377 		return (tBIT_NOT);
378 	    case -1:
379 		return (tEOF);
380 	}
381 	db_printf("Bad character\n");
382 	db_flush_lex();
383 	return (tEOF);
384 }
385