xref: /freebsd/sys/ddb/db_lex.c (revision 5ca8e32633c4ffbbcd6762e5888b6a4ba0708c6c)
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/param.h>
37 #include <sys/libkern.h>
38 #include <sys/lock.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(int);
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(void)
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(void)
92 {
93 	db_lp = db_line;
94 	db_endlp = db_line;
95 }
96 
97 static int
98 db_read_char(void)
99 {
100 	int	c;
101 
102 	if (db_lp >= db_endlp)
103 	    c = -1;
104 	else
105 	    c = *db_lp++;
106 	return (c);
107 }
108 
109 static void
110 db_unread_char(int c)
111 {
112 
113 	if (c == -1) {
114 		/* Unread EOL at EOL is okay. */
115 		if (db_lp < db_endlp)
116 			db_error("db_unread_char(-1) before end of line\n");
117 	} else {
118 		if (db_lp > db_line) {
119 			db_lp--;
120 			if (*db_lp != c)
121 				db_error("db_unread_char() wrong char\n");
122 		} else {
123 			db_error("db_unread_char() at beginning of line\n");
124 		}
125 	}
126 }
127 
128 static int	db_look_token = 0;
129 
130 void
131 db_unread_token(int t)
132 {
133 	db_look_token = t;
134 }
135 
136 int
137 db_read_token_flags(int flags)
138 {
139 	int	t;
140 
141 	MPASS((flags & ~(DRT_VALID_FLAGS_MASK)) == 0);
142 
143 	if (db_look_token) {
144 	    t = db_look_token;
145 	    db_look_token = 0;
146 	}
147 	else
148 	    t = db_lex(flags);
149 	return (t);
150 }
151 
152 db_expr_t	db_tok_number;
153 char	db_tok_string[TOK_STRING_SIZE];
154 
155 db_expr_t	db_radix = 16;
156 
157 void
158 db_flush_lex(void)
159 {
160 	db_flush_line();
161 	db_look_token = 0;
162 }
163 
164 static int
165 db_lex(int flags)
166 {
167 	int	c, n, radix_mode;
168 	bool	lex_wspace, lex_hex_numbers;
169 
170 	switch (flags & DRT_RADIX_MASK) {
171 	case DRT_DEFAULT_RADIX:
172 		radix_mode = -1;
173 		break;
174 	case DRT_OCTAL:
175 		radix_mode = 8;
176 		break;
177 	case DRT_DECIMAL:
178 		radix_mode = 10;
179 		break;
180 	case DRT_HEXADECIMAL:
181 		radix_mode = 16;
182 		break;
183 	}
184 
185 	lex_wspace = ((flags & DRT_WSPACE) != 0);
186 	lex_hex_numbers = ((flags & DRT_HEX) != 0);
187 
188 	c = db_read_char();
189 	for (n = 0; c <= ' ' || c > '~'; n++) {
190 	    if (c == '\n' || c == -1)
191 		return (tEOL);
192 	    c = db_read_char();
193 	}
194 	if (lex_wspace && n != 0) {
195 	    db_unread_char(c);
196 	    return (tWSPACE);
197 	}
198 
199 	if ((c >= '0' && c <= '9') ||
200 	   (lex_hex_numbers &&
201 	   ((c >= 'a' && c <= 'f') ||
202 	   (c >= 'A' && c <= 'F')))) {
203 	    /* number */
204 	    int	r, digit = 0;
205 
206 	    if (radix_mode != -1)
207 		r = radix_mode;
208 	    else if (c != '0')
209 		r = db_radix;
210 	    else {
211 		c = db_read_char();
212 		if (c == 'O' || c == 'o')
213 		    r = 8;
214 		else if (c == 'T' || c == 't')
215 		    r = 10;
216 		else if (c == 'X' || c == 'x')
217 		    r = 16;
218 		else {
219 		    r = db_radix;
220 		    db_unread_char(c);
221 		}
222 		c = db_read_char();
223 	    }
224 	    db_tok_number = 0;
225 	    for (;;) {
226 		if (c >= '0' && c <= ((r == 8) ? '7' : '9'))
227 		    digit = c - '0';
228 		else if (r == 16 && ((c >= 'A' && c <= 'F') ||
229 				     (c >= 'a' && c <= 'f'))) {
230 		    if (c >= 'a')
231 			digit = c - 'a' + 10;
232 		    else if (c >= 'A')
233 			digit = c - 'A' + 10;
234 		}
235 		else
236 		    break;
237 		db_tok_number = db_tok_number * r + digit;
238 		c = db_read_char();
239 	    }
240 	    if ((c >= '0' && c <= '9') ||
241 		(c >= 'A' && c <= 'Z') ||
242 		(c >= 'a' && c <= 'z') ||
243 		(c == '_'))
244 	    {
245 		db_error("Bad character in number\n");
246 		db_flush_lex();
247 		return (tEOF);
248 	    }
249 	    db_unread_char(c);
250 	    return (tNUMBER);
251 	}
252 	if ((c >= 'A' && c <= 'Z') ||
253 	    (c >= 'a' && c <= 'z') ||
254 	    c == '_' || c == '\\')
255 	{
256 	    /* string */
257 	    char *cp;
258 
259 	    cp = db_tok_string;
260 	    if (c == '\\') {
261 		c = db_read_char();
262 		if (c == '\n' || c == -1)
263 		    db_error("Bad escape\n");
264 	    }
265 	    *cp++ = c;
266 	    while (1) {
267 		c = db_read_char();
268 		if ((c >= 'A' && c <= 'Z') ||
269 		    (c >= 'a' && c <= 'z') ||
270 		    (c >= '0' && c <= '9') ||
271 		    c == '_' || c == '\\' || c == ':' || c == '.')
272 		{
273 		    if (c == '\\') {
274 			c = db_read_char();
275 			if (c == '\n' || c == -1)
276 			    db_error("Bad escape\n");
277 		    }
278 		    *cp++ = c;
279 		    if (cp == db_tok_string+sizeof(db_tok_string)) {
280 			db_error("String too long\n");
281 			db_flush_lex();
282 			return (tEOF);
283 		    }
284 		    continue;
285 		}
286 		else {
287 		    *cp = '\0';
288 		    break;
289 		}
290 	    }
291 	    db_unread_char(c);
292 	    return (tIDENT);
293 	}
294 
295 	switch (c) {
296 	    case '+':
297 		return (tPLUS);
298 	    case '-':
299 		return (tMINUS);
300 	    case '.':
301 		c = db_read_char();
302 		if (c == '.')
303 		    return (tDOTDOT);
304 		db_unread_char(c);
305 		return (tDOT);
306 	    case '*':
307 		return (tSTAR);
308 	    case '/':
309 		return (tSLASH);
310 	    case '=':
311 		c = db_read_char();
312 		if (c == '=')
313 		    return (tLOG_EQ);
314 		db_unread_char(c);
315 		return (tEQ);
316 	    case '%':
317 		return (tPCT);
318 	    case '#':
319 		return (tHASH);
320 	    case '(':
321 		return (tLPAREN);
322 	    case ')':
323 		return (tRPAREN);
324 	    case ',':
325 		return (tCOMMA);
326 	    case '"':
327 		return (tDITTO);
328 	    case '$':
329 		return (tDOLLAR);
330 	    case '!':
331 		c = db_read_char();
332 		if (c == '='){
333 			return (tLOG_NOT_EQ);
334 		}
335 		db_unread_char(c);
336 		return (tEXCL);
337 	    case ':':
338 		c = db_read_char();
339 		if (c == ':')
340 			return (tCOLONCOLON);
341 		db_unread_char(c);
342 		return (tCOLON);
343 	    case ';':
344 		return (tSEMI);
345 	    case '&':
346 		c = db_read_char();
347 		if (c == '&')
348 		    return (tLOG_AND);
349 		db_unread_char(c);
350 		return (tBIT_AND);
351 	    case '|':
352 		c = db_read_char();
353 		if (c == '|')
354 		    return (tLOG_OR);
355 		db_unread_char(c);
356 		return (tBIT_OR);
357 	    case '<':
358 		c = db_read_char();
359 		if (c == '<')
360 		    return (tSHIFT_L);
361 		if (c == '=')
362 		    return (tLESS_EQ);
363 		db_unread_char(c);
364 		return (tLESS);
365 	    case '>':
366 		c = db_read_char();
367 		if (c == '>')
368 		    return (tSHIFT_R);
369 		if (c == '=')
370 		    return (tGREATER_EQ);
371 		db_unread_char(c);
372 		return (tGREATER);
373 	    case '?':
374 		return (tQUESTION);
375 	    case '~':
376 		return (tBIT_NOT);
377 	    case -1:
378 		return (tEOF);
379 	}
380 	db_printf("Bad character\n");
381 	db_flush_lex();
382 	return (tEOF);
383 }
384