1 /*- 2 * SPDX-License-Identifier: BSD-3-Clause 3 * 4 * Copyright (c) 1980, 1993 5 * The Regents of the University of California. All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 3. Neither the name of the University nor the names of its contributors 16 * may be used to endorse or promote products derived from this software 17 * without specific prior written permission. 18 * 19 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 22 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 29 * SUCH DAMAGE. 30 */ 31 32 33 #ifndef lint 34 static const char copyright[] = 35 "@(#) Copyright (c) 1980, 1993\n\ 36 The Regents of the University of California. All rights reserved.\n"; 37 #endif 38 39 40 #include <sys/types.h> 41 42 #include <ctype.h> 43 #include <err.h> 44 #include <stdio.h> 45 #include <stdlib.h> 46 #include <signal.h> 47 #include <string.h> 48 #include <unistd.h> 49 50 #include "pathnames.h" 51 52 /* 53 * xstr - extract and hash strings in a C program 54 * 55 * Bill Joy UCB 56 * November, 1978 57 */ 58 59 #define ignore(a) ((void) a) 60 61 static off_t tellpt; 62 63 static off_t mesgpt; 64 static char cstrings[] = "strings"; 65 static char *strings = cstrings; 66 67 static int cflg; 68 static int vflg; 69 static int readstd; 70 71 static char lastchr(char *); 72 73 static int fgetNUL(char *, int, FILE *); 74 static int istail(char *, char *); 75 static int octdigit(char); 76 static int xgetc(FILE *); 77 78 static off_t hashit(char *, int); 79 static off_t yankstr(char **); 80 81 static void usage(void) __dead2; 82 83 static void flushsh(void); 84 static void found(int, off_t, char *); 85 static void inithash(void); 86 static void onintr(int); 87 static void process(const char *); 88 static void prstr(char *); 89 static void xsdotc(void); 90 91 int 92 main(int argc, char *argv[]) 93 { 94 int c; 95 int fdesc; 96 97 while ((c = getopt(argc, argv, "-cv")) != -1) 98 switch (c) { 99 case '-': 100 readstd++; 101 break; 102 case 'c': 103 cflg++; 104 break; 105 case 'v': 106 vflg++; 107 break; 108 default: 109 usage(); 110 } 111 argc -= optind; 112 argv += optind; 113 114 if (signal(SIGINT, SIG_IGN) == SIG_DFL) 115 signal(SIGINT, onintr); 116 if (cflg || (argc == 0 && !readstd)) 117 inithash(); 118 else { 119 strings = strdup(_PATH_TMP); 120 if (strings == NULL) 121 err(1, "strdup() failed"); 122 fdesc = mkstemp(strings); 123 if (fdesc == -1) 124 err(1, "Unable to create temporary file"); 125 close(fdesc); 126 } 127 128 while (readstd || argc > 0) { 129 if (freopen("x.c", "w", stdout) == NULL) 130 err(1, "x.c"); 131 if (!readstd && freopen(argv[0], "r", stdin) == NULL) 132 err(2, "%s", argv[0]); 133 process("x.c"); 134 if (readstd == 0) 135 argc--, argv++; 136 else 137 readstd = 0; 138 } 139 flushsh(); 140 if (cflg == 0) 141 xsdotc(); 142 if (strings[0] == '/') 143 ignore(unlink(strings)); 144 exit(0); 145 } 146 147 static void 148 usage(void) 149 { 150 fprintf(stderr, "usage: xstr [-cv] [-] [file ...]\n"); 151 exit (1); 152 } 153 154 static char linebuf[BUFSIZ]; 155 156 static void 157 process(const char *name) 158 { 159 char *cp; 160 int c; 161 int incomm = 0; 162 int ret; 163 164 printf("extern char\txstr[];\n"); 165 for (;;) { 166 if (fgets(linebuf, sizeof linebuf, stdin) == NULL) { 167 if (ferror(stdin)) 168 err(3, "%s", name); 169 break; 170 } 171 if (linebuf[0] == '#') { 172 if (linebuf[1] == ' ' && isdigit(linebuf[2])) 173 printf("#line%s", &linebuf[1]); 174 else 175 printf("%s", linebuf); 176 continue; 177 } 178 for (cp = linebuf; (c = *cp++);) switch (c) { 179 180 case '"': 181 if (incomm) 182 goto def; 183 if ((ret = (int) yankstr(&cp)) == -1) 184 goto out; 185 printf("(&xstr[%d])", ret); 186 break; 187 188 case '\'': 189 if (incomm) 190 goto def; 191 putchar(c); 192 if (*cp) 193 putchar(*cp++); 194 break; 195 196 case '/': 197 if (incomm || *cp != '*') 198 goto def; 199 incomm = 1; 200 cp++; 201 printf("/*"); 202 continue; 203 204 case '*': 205 if (incomm && *cp == '/') { 206 incomm = 0; 207 cp++; 208 printf("*/"); 209 continue; 210 } 211 goto def; 212 213 def: 214 default: 215 putchar(c); 216 break; 217 } 218 } 219 out: 220 if (ferror(stdout)) 221 warn("x.c"), onintr(0); 222 } 223 224 static off_t 225 yankstr(char **cpp) 226 { 227 char *cp = *cpp; 228 int c, ch; 229 char dbuf[BUFSIZ]; 230 char *dp = dbuf; 231 char *tp; 232 static char tmp[] = "b\bt\tr\rn\nf\f\\\\\"\""; 233 234 while ((c = *cp++)) { 235 if (dp == dbuf + sizeof(dbuf) - 3) 236 errx(1, "message too long"); 237 switch (c) { 238 239 case '"': 240 cp++; 241 goto out; 242 243 case '\\': 244 c = *cp++; 245 if (c == 0) 246 break; 247 if (c == '\n') { 248 if (fgets(linebuf, sizeof linebuf, stdin) 249 == NULL) { 250 if (ferror(stdin)) 251 err(3, "x.c"); 252 return(-1); 253 } 254 cp = linebuf; 255 continue; 256 } 257 for (tp = tmp; (ch = *tp++); tp++) 258 if (c == ch) { 259 c = *tp; 260 goto gotc; 261 } 262 if (!octdigit(c)) { 263 *dp++ = '\\'; 264 break; 265 } 266 c -= '0'; 267 if (!octdigit(*cp)) 268 break; 269 c <<= 3, c += *cp++ - '0'; 270 if (!octdigit(*cp)) 271 break; 272 c <<= 3, c += *cp++ - '0'; 273 break; 274 } 275 gotc: 276 *dp++ = c; 277 } 278 out: 279 *cpp = --cp; 280 *dp = 0; 281 return (hashit(dbuf, 1)); 282 } 283 284 static int 285 octdigit(char c) 286 { 287 return (isdigit(c) && c != '8' && c != '9'); 288 } 289 290 static void 291 inithash(void) 292 { 293 char buf[BUFSIZ]; 294 FILE *mesgread = fopen(strings, "r"); 295 296 if (mesgread == NULL) 297 return; 298 for (;;) { 299 mesgpt = tellpt; 300 if (fgetNUL(buf, sizeof buf, mesgread) == 0) 301 break; 302 ignore(hashit(buf, 0)); 303 } 304 ignore(fclose(mesgread)); 305 } 306 307 static int 308 fgetNUL(char *obuf, int rmdr, FILE *file) 309 { 310 int c; 311 char *buf = obuf; 312 313 while (--rmdr > 0 && (c = xgetc(file)) != 0 && c != EOF) 314 *buf++ = c; 315 *buf++ = 0; 316 return ((feof(file) || ferror(file)) ? 0 : 1); 317 } 318 319 static int 320 xgetc(FILE *file) 321 { 322 323 tellpt++; 324 return (getc(file)); 325 } 326 327 #define BUCKETS 128 328 329 static struct hash { 330 off_t hpt; 331 char *hstr; 332 struct hash *hnext; 333 short hnew; 334 } bucket[BUCKETS]; 335 336 static off_t 337 hashit(char *str, int new) 338 { 339 int i; 340 struct hash *hp, *hp0; 341 342 hp = hp0 = &bucket[lastchr(str) & 0177]; 343 while (hp->hnext) { 344 hp = hp->hnext; 345 i = istail(str, hp->hstr); 346 if (i >= 0) 347 return (hp->hpt + i); 348 } 349 if ((hp = (struct hash *) calloc(1, sizeof (*hp))) == NULL) 350 errx(8, "calloc"); 351 hp->hpt = mesgpt; 352 if (!(hp->hstr = strdup(str))) 353 err(1, NULL); 354 mesgpt += strlen(hp->hstr) + 1; 355 hp->hnext = hp0->hnext; 356 hp->hnew = new; 357 hp0->hnext = hp; 358 return (hp->hpt); 359 } 360 361 static void 362 flushsh(void) 363 { 364 int i; 365 struct hash *hp; 366 FILE *mesgwrit; 367 int old = 0, new = 0; 368 369 for (i = 0; i < BUCKETS; i++) 370 for (hp = bucket[i].hnext; hp != NULL; hp = hp->hnext) 371 if (hp->hnew) 372 new++; 373 else 374 old++; 375 if (new == 0 && old != 0) 376 return; 377 mesgwrit = fopen(strings, old ? "r+" : "w"); 378 if (mesgwrit == NULL) 379 err(4, "%s", strings); 380 for (i = 0; i < BUCKETS; i++) 381 for (hp = bucket[i].hnext; hp != NULL; hp = hp->hnext) { 382 found(hp->hnew, hp->hpt, hp->hstr); 383 if (hp->hnew) { 384 fseek(mesgwrit, hp->hpt, 0); 385 ignore(fwrite(hp->hstr, strlen(hp->hstr) + 1, 1, mesgwrit)); 386 if (ferror(mesgwrit)) 387 err(4, "%s", strings); 388 } 389 } 390 if (fclose(mesgwrit) == EOF) 391 err(4, "%s", strings); 392 } 393 394 static void 395 found(int new, off_t off, char *str) 396 { 397 if (vflg == 0) 398 return; 399 if (!new) 400 fprintf(stderr, "found at %d:", (int) off); 401 else 402 fprintf(stderr, "new at %d:", (int) off); 403 prstr(str); 404 fprintf(stderr, "\n"); 405 } 406 407 static void 408 prstr(char *cp) 409 { 410 int c; 411 412 while ((c = (*cp++ & 0377))) 413 if (c < ' ') 414 fprintf(stderr, "^%c", c + '`'); 415 else if (c == 0177) 416 fprintf(stderr, "^?"); 417 else if (c > 0200) 418 fprintf(stderr, "\\%03o", c); 419 else 420 fprintf(stderr, "%c", c); 421 } 422 423 static void 424 xsdotc(void) 425 { 426 FILE *strf = fopen(strings, "r"); 427 FILE *xdotcf; 428 429 if (strf == NULL) 430 err(5, "%s", strings); 431 xdotcf = fopen("xs.c", "w"); 432 if (xdotcf == NULL) 433 err(6, "xs.c"); 434 fprintf(xdotcf, "char\txstr[] = {\n"); 435 for (;;) { 436 int i, c; 437 438 for (i = 0; i < 8; i++) { 439 c = getc(strf); 440 if (ferror(strf)) { 441 warn("%s", strings); 442 onintr(0); 443 } 444 if (feof(strf)) { 445 fprintf(xdotcf, "\n"); 446 goto out; 447 } 448 fprintf(xdotcf, "0x%02x,", c); 449 } 450 fprintf(xdotcf, "\n"); 451 } 452 out: 453 fprintf(xdotcf, "};\n"); 454 ignore(fclose(xdotcf)); 455 ignore(fclose(strf)); 456 } 457 458 static char 459 lastchr(char *cp) 460 { 461 462 while (cp[0] && cp[1]) 463 cp++; 464 return (*cp); 465 } 466 467 static int 468 istail(char *str, char *of) 469 { 470 int d = strlen(of) - strlen(str); 471 472 if (d < 0 || strcmp(&of[d], str) != 0) 473 return (-1); 474 return (d); 475 } 476 477 static void 478 onintr(int dummy __unused) 479 { 480 481 ignore(signal(SIGINT, SIG_IGN)); 482 if (strings[0] == '/') 483 ignore(unlink(strings)); 484 ignore(unlink("x.c")); 485 ignore(unlink("xs.c")); 486 exit(7); 487 } 488