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