xref: /freebsd/sys/ddb/db_lex.c (revision f7c32ed617858bcd22f8d1b03199099d50125721)
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 __FBSDID("$FreeBSD$");
38 
39 #include <sys/param.h>
40 #include <sys/libkern.h>
41 #include <sys/lock.h>
42 
43 #include <ddb/ddb.h>
44 #include <ddb/db_lex.h>
45 
46 static char	db_line[DB_MAXLINE];
47 static char *	db_lp, *db_endlp;
48 
49 static int	db_lex(int);
50 static void 	db_flush_line(void);
51 static int 	db_read_char(void);
52 static void 	db_unread_char(int);
53 
54 int
55 db_read_line(void)
56 {
57 	int	i;
58 
59 	i = db_readline(db_line, sizeof(db_line));
60 	if (i == 0)
61 	    return (0);	/* EOI */
62 	db_lp = db_line;
63 	db_endlp = db_lp + i;
64 	return (i);
65 }
66 
67 /*
68  * Simulate a line of input into DDB.
69  */
70 void
71 db_inject_line(const char *command)
72 {
73 
74 	strlcpy(db_line, command, sizeof(db_line));
75 	db_lp = db_line;
76 	db_endlp = db_lp + strlen(command);
77 }
78 
79 /*
80  * In rare cases, we may want to pull the remainder of the line input
81  * verbatim, rather than lexing it.  For example, when assigning literal
82  * values associated with scripts.  In that case, return a static pointer to
83  * the current location in the input buffer.  The caller must be aware that
84  * the contents are not stable if other lex/input calls are made.
85  */
86 char *
87 db_get_line(void)
88 {
89 
90 	return (db_lp);
91 }
92 
93 static void
94 db_flush_line()
95 {
96 	db_lp = db_line;
97 	db_endlp = db_line;
98 }
99 
100 static int
101 db_read_char(void)
102 {
103 	int	c;
104 
105 	if (db_lp >= db_endlp)
106 	    c = -1;
107 	else
108 	    c = *db_lp++;
109 	return (c);
110 }
111 
112 static void
113 db_unread_char(int c)
114 {
115 
116 	if (c == -1) {
117 		/* Unread EOL at EOL is okay. */
118 		if (db_lp < db_endlp)
119 			db_error("db_unread_char(-1) before end of line\n");
120 	} else {
121 		if (db_lp > db_line) {
122 			db_lp--;
123 			if (*db_lp != c)
124 				db_error("db_unread_char() wrong char\n");
125 		} else {
126 			db_error("db_unread_char() at beginning of line\n");
127 		}
128 	}
129 }
130 
131 static int	db_look_token = 0;
132 
133 void
134 db_unread_token(int t)
135 {
136 	db_look_token = t;
137 }
138 
139 int
140 db_read_token_flags(int flags)
141 {
142 	int	t;
143 
144 	MPASS((flags & ~(DRT_VALID_FLAGS_MASK)) == 0);
145 
146 	if (db_look_token) {
147 	    t = db_look_token;
148 	    db_look_token = 0;
149 	}
150 	else
151 	    t = db_lex(flags);
152 	return (t);
153 }
154 
155 db_expr_t	db_tok_number;
156 char	db_tok_string[TOK_STRING_SIZE];
157 
158 db_expr_t	db_radix = 16;
159 
160 void
161 db_flush_lex(void)
162 {
163 	db_flush_line();
164 	db_look_token = 0;
165 }
166 
167 static int
168 db_lex(int flags)
169 {
170 	int	c, n, radix_mode;
171 	bool	lex_wspace, lex_hex_numbers;
172 
173 	switch (flags & DRT_RADIX_MASK) {
174 	case DRT_DEFAULT_RADIX:
175 		radix_mode = -1;
176 		break;
177 	case DRT_OCTAL:
178 		radix_mode = 8;
179 		break;
180 	case DRT_DECIMAL:
181 		radix_mode = 10;
182 		break;
183 	case DRT_HEXADECIMAL:
184 		radix_mode = 16;
185 		break;
186 	}
187 
188 	lex_wspace = ((flags & DRT_WSPACE) != 0);
189 	lex_hex_numbers = ((flags & DRT_HEX) != 0);
190 
191 	c = db_read_char();
192 	for (n = 0; c <= ' ' || c > '~'; n++) {
193 	    if (c == '\n' || c == -1)
194 		return (tEOL);
195 	    c = db_read_char();
196 	}
197 	if (lex_wspace && n != 0) {
198 	    db_unread_char(c);
199 	    return (tWSPACE);
200 	}
201 
202 	if ((c >= '0' && c <= '9') ||
203 	   (lex_hex_numbers &&
204 	   ((c >= 'a' && c <= 'f') ||
205 	   (c >= 'A' && c <= 'F')))) {
206 	    /* number */
207 	    int	r, digit = 0;
208 
209 	    if (radix_mode != -1)
210 		r = radix_mode;
211 	    else if (c != '0')
212 		r = db_radix;
213 	    else {
214 		c = db_read_char();
215 		if (c == 'O' || c == 'o')
216 		    r = 8;
217 		else if (c == 'T' || c == 't')
218 		    r = 10;
219 		else if (c == 'X' || c == 'x')
220 		    r = 16;
221 		else {
222 		    r = db_radix;
223 		    db_unread_char(c);
224 		}
225 		c = db_read_char();
226 	    }
227 	    db_tok_number = 0;
228 	    for (;;) {
229 		if (c >= '0' && c <= ((r == 8) ? '7' : '9'))
230 		    digit = c - '0';
231 		else if (r == 16 && ((c >= 'A' && c <= 'F') ||
232 				     (c >= 'a' && c <= 'f'))) {
233 		    if (c >= 'a')
234 			digit = c - 'a' + 10;
235 		    else if (c >= 'A')
236 			digit = c - 'A' + 10;
237 		}
238 		else
239 		    break;
240 		db_tok_number = db_tok_number * r + digit;
241 		c = db_read_char();
242 	    }
243 	    if ((c >= '0' && c <= '9') ||
244 		(c >= 'A' && c <= 'Z') ||
245 		(c >= 'a' && c <= 'z') ||
246 		(c == '_'))
247 	    {
248 		db_error("Bad character in number\n");
249 		db_flush_lex();
250 		return (tEOF);
251 	    }
252 	    db_unread_char(c);
253 	    return (tNUMBER);
254 	}
255 	if ((c >= 'A' && c <= 'Z') ||
256 	    (c >= 'a' && c <= 'z') ||
257 	    c == '_' || c == '\\')
258 	{
259 	    /* string */
260 	    char *cp;
261 
262 	    cp = db_tok_string;
263 	    if (c == '\\') {
264 		c = db_read_char();
265 		if (c == '\n' || c == -1)
266 		    db_error("Bad escape\n");
267 	    }
268 	    *cp++ = c;
269 	    while (1) {
270 		c = db_read_char();
271 		if ((c >= 'A' && c <= 'Z') ||
272 		    (c >= 'a' && c <= 'z') ||
273 		    (c >= '0' && c <= '9') ||
274 		    c == '_' || c == '\\' || c == ':' || c == '.')
275 		{
276 		    if (c == '\\') {
277 			c = db_read_char();
278 			if (c == '\n' || c == -1)
279 			    db_error("Bad escape\n");
280 		    }
281 		    *cp++ = c;
282 		    if (cp == db_tok_string+sizeof(db_tok_string)) {
283 			db_error("String too long\n");
284 			db_flush_lex();
285 			return (tEOF);
286 		    }
287 		    continue;
288 		}
289 		else {
290 		    *cp = '\0';
291 		    break;
292 		}
293 	    }
294 	    db_unread_char(c);
295 	    return (tIDENT);
296 	}
297 
298 	switch (c) {
299 	    case '+':
300 		return (tPLUS);
301 	    case '-':
302 		return (tMINUS);
303 	    case '.':
304 		c = db_read_char();
305 		if (c == '.')
306 		    return (tDOTDOT);
307 		db_unread_char(c);
308 		return (tDOT);
309 	    case '*':
310 		return (tSTAR);
311 	    case '/':
312 		return (tSLASH);
313 	    case '=':
314 		c = db_read_char();
315 		if (c == '=')
316 		    return (tLOG_EQ);
317 		db_unread_char(c);
318 		return (tEQ);
319 	    case '%':
320 		return (tPCT);
321 	    case '#':
322 		return (tHASH);
323 	    case '(':
324 		return (tLPAREN);
325 	    case ')':
326 		return (tRPAREN);
327 	    case ',':
328 		return (tCOMMA);
329 	    case '"':
330 		return (tDITTO);
331 	    case '$':
332 		return (tDOLLAR);
333 	    case '!':
334 		c = db_read_char();
335 		if (c == '='){
336 			return (tLOG_NOT_EQ);
337 		}
338 		db_unread_char(c);
339 		return (tEXCL);
340 	    case ':':
341 		c = db_read_char();
342 		if (c == ':')
343 			return (tCOLONCOLON);
344 		db_unread_char(c);
345 		return (tCOLON);
346 	    case ';':
347 		return (tSEMI);
348 	    case '&':
349 		c = db_read_char();
350 		if (c == '&')
351 		    return (tLOG_AND);
352 		db_unread_char(c);
353 		return (tBIT_AND);
354 	    case '|':
355 		c = db_read_char();
356 		if (c == '|')
357 		    return (tLOG_OR);
358 		db_unread_char(c);
359 		return (tBIT_OR);
360 	    case '<':
361 		c = db_read_char();
362 		if (c == '<')
363 		    return (tSHIFT_L);
364 		if (c == '=')
365 		    return (tLESS_EQ);
366 		db_unread_char(c);
367 		return (tLESS);
368 	    case '>':
369 		c = db_read_char();
370 		if (c == '>')
371 		    return (tSHIFT_R);
372 		if (c == '=')
373 		    return (tGREATER_EQ);
374 		db_unread_char(c);
375 		return (tGREATER);
376 	    case '?':
377 		return (tQUESTION);
378 	    case '~':
379 		return (tBIT_NOT);
380 	    case -1:
381 		return (tEOF);
382 	}
383 	db_printf("Bad character\n");
384 	db_flush_lex();
385 	return (tEOF);
386 }
387