1 /* $OpenBSD: parse.c,v 1.20 2016/03/15 04:19:13 mmcc Exp $ */ 2 /* $NetBSD: parse.c,v 1.12 2001/12/07 13:37:39 bjh21 Exp $ */ 3 4 /* 5 * Copyright (c) 1989, 1993 6 * The Regents of the University of California. All rights reserved. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 3. Neither the name of the University nor the names of its contributors 17 * may be used to endorse or promote products derived from this software 18 * without specific prior written permission. 19 * 20 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 23 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 30 * SUCH DAMAGE. 31 */ 32 33 #include <sys/types.h> 34 #include <sys/file.h> 35 36 #include <ctype.h> 37 #include <err.h> 38 #include <errno.h> 39 #include <fcntl.h> 40 #include <stdio.h> 41 #include <stdlib.h> 42 #include <string.h> 43 44 #include "hexdump.h" 45 46 FU *endfu; /* format at end-of-data */ 47 48 static __dead void badcnt(char *); 49 static __dead void badconv(char *); 50 static __dead void badfmt(const char *); 51 static __dead void badsfmt(void); 52 static void escape(char *); 53 54 void 55 addfile(char *name) 56 { 57 FILE *fp; 58 size_t len; 59 char *buf, *lbuf, *p; 60 61 if ((fp = fopen(name, "r")) == NULL) 62 err(1, "fopen %s", name); 63 64 lbuf = NULL; 65 while ((buf = fgetln(fp, &len))) { 66 if (buf[len - 1] == '\n') 67 buf[len - 1] = '\0'; 68 else { 69 /* EOF without EOL, copy and add the NUL */ 70 if ((lbuf = malloc(len + 1)) == NULL) 71 err(1, NULL); 72 memcpy(lbuf, buf, len); 73 lbuf[len] = '\0'; 74 buf = lbuf; 75 } 76 for (p = buf; isspace((unsigned char)*p); ++p); 77 if (!*p || *p == '#') 78 continue; 79 add(p); 80 } 81 free(lbuf); 82 (void)fclose(fp); 83 } 84 85 void 86 add(const char *fmt) 87 { 88 const char *p; 89 static FS **nextfs; 90 FS *tfs; 91 FU *tfu, **nextfu; 92 const char *savep; 93 94 /* start new linked list of format units */ 95 if ((tfs = calloc(1, sizeof(FS))) == NULL) 96 err(1, NULL); 97 if (!fshead) 98 fshead = tfs; 99 else 100 *nextfs = tfs; 101 nextfs = &tfs->nextfs; 102 nextfu = &tfs->nextfu; 103 104 /* take the format string and break it up into format units */ 105 for (p = fmt;;) { 106 /* skip leading white space */ 107 for (; isspace((unsigned char)*p); ++p); 108 if (!*p) 109 break; 110 111 /* allocate a new format unit and link it in */ 112 if ((tfu = calloc(1, sizeof(FU))) == NULL) 113 err(1, NULL); 114 *nextfu = tfu; 115 nextfu = &tfu->nextfu; 116 tfu->reps = 1; 117 118 /* if leading digit, repetition count */ 119 if (isdigit((unsigned char)*p)) { 120 for (savep = p; isdigit((unsigned char)*p); ++p); 121 if (!isspace((unsigned char)*p) && *p != '/') 122 badfmt(fmt); 123 /* may overwrite either white space or slash */ 124 tfu->reps = atoi(savep); 125 tfu->flags = F_SETREP; 126 /* skip trailing white space */ 127 for (++p; isspace((unsigned char)*p); ++p); 128 } 129 130 /* skip slash and trailing white space */ 131 if (*p == '/') 132 while (isspace((unsigned char)*++p)); 133 134 /* byte count */ 135 if (isdigit((unsigned char)*p)) { 136 for (savep = p; isdigit((unsigned char)*p); ++p); 137 if (!isspace((unsigned char)*p)) 138 badfmt(fmt); 139 tfu->bcnt = atoi(savep); 140 /* skip trailing white space */ 141 for (++p; isspace((unsigned char)*p); ++p); 142 } 143 144 /* format */ 145 if (*p != '"') 146 badfmt(fmt); 147 for (savep = ++p; *p != '"';) 148 if (*p++ == 0) 149 badfmt(fmt); 150 tfu->fmt = strndup(savep, p - savep); 151 if (tfu->fmt == NULL) 152 err(1, NULL); 153 escape(tfu->fmt); 154 p++; 155 } 156 } 157 158 static const char *spec = ".#-+ 0123456789"; 159 160 int 161 size(FS *fs) 162 { 163 FU *fu; 164 int bcnt, cursize; 165 char *fmt; 166 int prec; 167 168 /* figure out the data block size needed for each format unit */ 169 for (cursize = 0, fu = fs->nextfu; fu; fu = fu->nextfu) { 170 if (fu->bcnt) { 171 cursize += fu->bcnt * fu->reps; 172 continue; 173 } 174 for (bcnt = prec = 0, fmt = fu->fmt; *fmt; ++fmt) { 175 if (*fmt != '%') 176 continue; 177 /* 178 * skip any special chars -- save precision in 179 * case it's a %s format. 180 */ 181 while (*++fmt && strchr(spec + 1, *fmt)); 182 if (*fmt == '.' && isdigit((unsigned char)*++fmt)) { 183 prec = atoi(fmt); 184 while (isdigit((unsigned char)*++fmt)); 185 } 186 switch(*fmt) { 187 case 'c': 188 bcnt += 1; 189 break; 190 case 'd': case 'i': case 'o': case 'u': 191 case 'x': case 'X': 192 bcnt += 4; 193 break; 194 case 'e': case 'E': case 'f': case 'g': case 'G': 195 bcnt += 8; 196 break; 197 case 's': 198 bcnt += prec; 199 break; 200 case '_': 201 switch(*++fmt) { 202 case 'c': case 'p': case 'u': 203 bcnt += 1; 204 break; 205 } 206 } 207 } 208 cursize += bcnt * fu->reps; 209 } 210 return (cursize); 211 } 212 213 void 214 rewrite(FS *fs) 215 { 216 enum { NOTOKAY, USEBCNT, USEPREC } sokay; 217 PR *pr, **nextpr; 218 FU *fu; 219 char *p1, *p2; 220 char savech, *fmtp, cs[3]; 221 int nconv, prec; 222 size_t len; 223 224 nextpr = NULL; 225 prec = 0; 226 for (fu = fs->nextfu; fu; fu = fu->nextfu) { 227 /* 228 * Break each format unit into print units; each conversion 229 * character gets its own. 230 */ 231 for (nconv = 0, fmtp = fu->fmt; *fmtp; nextpr = &pr->nextpr) { 232 if ((pr = calloc(1, sizeof(PR))) == NULL) 233 err(1, NULL); 234 if (!fu->nextpr) 235 fu->nextpr = pr; 236 else 237 *nextpr = pr; 238 239 /* Skip preceding text and up to the next % sign. */ 240 for (p1 = fmtp; *p1 && *p1 != '%'; ++p1); 241 242 /* Only text in the string. */ 243 if (!*p1) { 244 pr->fmt = fmtp; 245 pr->flags = F_TEXT; 246 break; 247 } 248 249 /* 250 * Get precision for %s -- if have a byte count, don't 251 * need it. 252 */ 253 if (fu->bcnt) { 254 sokay = USEBCNT; 255 /* Skip to conversion character. */ 256 for (++p1; *p1 && strchr(spec, *p1); ++p1); 257 } else { 258 /* Skip any special chars, field width. */ 259 while (*++p1 && strchr(spec + 1, *p1)); 260 if (*p1 == '.' && 261 isdigit((unsigned char)*++p1)) { 262 sokay = USEPREC; 263 prec = atoi(p1); 264 while (isdigit((unsigned char)*++p1)) 265 continue; 266 } else 267 sokay = NOTOKAY; 268 } 269 270 p2 = *p1 ? p1 + 1 : p1; /* Set end pointer. */ 271 cs[0] = *p1; /* Set conversion string. */ 272 cs[1] = '\0'; 273 274 /* 275 * Figure out the byte count for each conversion; 276 * rewrite the format as necessary, set up blank- 277 * padding for end of data. 278 */ 279 switch(cs[0]) { 280 case 'c': 281 pr->flags = F_CHAR; 282 switch(fu->bcnt) { 283 case 0: case 1: 284 pr->bcnt = 1; 285 break; 286 default: 287 p1[1] = '\0'; 288 badcnt(p1); 289 } 290 break; 291 case 'd': case 'i': 292 case 'o': case 'u': case 'x': case 'X': 293 if (cs[0] == 'd' || cs[0] == 'i') 294 pr->flags = F_INT; 295 else 296 pr->flags = F_UINT; 297 298 cs[2] = '\0'; 299 cs[1] = cs[0]; 300 cs[0] = 'q'; 301 switch(fu->bcnt) { 302 case 0: case 4: 303 pr->bcnt = 4; 304 break; 305 case 1: 306 pr->bcnt = 1; 307 break; 308 case 2: 309 pr->bcnt = 2; 310 break; 311 case 8: 312 pr->bcnt = 8; 313 break; 314 default: 315 p1[1] = '\0'; 316 badcnt(p1); 317 } 318 break; 319 case 'e': case 'E': case 'f': case 'g': case 'G': 320 pr->flags = F_DBL; 321 switch(fu->bcnt) { 322 case 0: case 8: 323 pr->bcnt = 8; 324 break; 325 case 4: 326 pr->bcnt = 4; 327 break; 328 default: 329 p1[1] = '\0'; 330 badcnt(p1); 331 } 332 break; 333 case 's': 334 pr->flags = F_STR; 335 switch(sokay) { 336 case NOTOKAY: 337 badsfmt(); 338 case USEBCNT: 339 pr->bcnt = fu->bcnt; 340 break; 341 case USEPREC: 342 pr->bcnt = prec; 343 break; 344 } 345 break; 346 case '_': 347 ++p2; 348 switch(p1[1]) { 349 case 'A': 350 endfu = fu; 351 fu->flags |= F_IGNORE; 352 /* FALLTHROUGH */ 353 case 'a': 354 pr->flags = F_ADDRESS; 355 ++p2; 356 switch(p1[2]) { 357 case 'd': case 'o': case'x': 358 cs[0] = 'q'; 359 cs[1] = p1[2]; 360 cs[2] = '\0'; 361 break; 362 default: 363 if (p1[2]) 364 p1[3] = '\0'; 365 badconv(p1); 366 } 367 break; 368 case 'c': 369 case 'p': 370 case 'u': 371 if (p1[1] == 'c') { 372 pr->flags = F_C; 373 /* cs[0] = 'c'; set in conv_c */ 374 } else if (p1[1] == 'p') { 375 pr->flags = F_P; 376 cs[0] = 'c'; 377 } else { 378 pr->flags = F_U; 379 /* cs[0] = 'c'; set in conv_u */ 380 } 381 382 switch(fu->bcnt) { 383 case 0: case 1: 384 pr->bcnt = 1; 385 break; 386 default: 387 p1[2] = '\0'; 388 badcnt(p1); 389 } 390 break; 391 default: 392 if (p1[1]) 393 p1[2] = '\0'; 394 badconv(p1); 395 } 396 break; 397 default: 398 if (cs[0]) 399 p1[1] = '\0'; 400 badconv(p1); 401 } 402 403 /* 404 * Copy to PR format string, set conversion character 405 * pointer, update original. 406 */ 407 savech = *p2; 408 p1[0] = '\0'; 409 len = strlen(fmtp) + strlen(cs) + 1; 410 if ((pr->fmt = calloc(1, len)) == NULL) 411 err(1, NULL); 412 snprintf(pr->fmt, len, "%s%s", fmtp, cs); 413 *p2 = savech; 414 pr->cchar = pr->fmt + (p1 - fmtp); 415 fmtp = p2; 416 417 /* Only one conversion character if byte count. */ 418 if (!(pr->flags&F_ADDRESS) && fu->bcnt && nconv++) 419 errx(1, 420 "byte count with multiple conversion characters"); 421 } 422 /* 423 * If format unit byte count not specified, figure it out 424 * so can adjust rep count later. 425 */ 426 if (!fu->bcnt) 427 for (pr = fu->nextpr; pr; pr = pr->nextpr) 428 fu->bcnt += pr->bcnt; 429 } 430 /* 431 * If the format string interprets any data at all, and it's 432 * not the same as the blocksize, and its last format unit 433 * interprets any data at all, and has no iteration count, 434 * repeat it as necessary. 435 * 436 * If, rep count is greater than 1, no trailing whitespace 437 * gets output from the last iteration of the format unit. 438 */ 439 for (fu = fs->nextfu; fu; fu = fu->nextfu) { 440 if (!fu->nextfu && fs->bcnt < blocksize && 441 !(fu->flags&F_SETREP) && fu->bcnt) 442 fu->reps += (blocksize - fs->bcnt) / fu->bcnt; 443 if (fu->reps > 1) { 444 if (!fu->nextpr) 445 break; 446 for (pr = fu->nextpr;; pr = pr->nextpr) 447 if (!pr->nextpr) 448 break; 449 for (p1 = pr->fmt, p2 = NULL; *p1; ++p1) 450 p2 = isspace((unsigned char)*p1) ? p1 : NULL; 451 if (p2) 452 pr->nospace = p2; 453 } 454 } 455 #ifdef DEBUG 456 for (fu = fs->nextfu; fu; fu = fu->nextfu) { 457 (void)printf("fmt:"); 458 for (pr = fu->nextpr; pr; pr = pr->nextpr) 459 (void)printf(" {%s}", pr->fmt); 460 (void)printf("\n"); 461 } 462 #endif 463 } 464 465 static void 466 escape(char *p1) 467 { 468 char *p2; 469 470 /* alphabetic escape sequences have to be done in place */ 471 for (p2 = p1;; ++p1, ++p2) { 472 if (!*p1) { 473 *p2 = *p1; 474 break; 475 } 476 if (*p1 == '\\') { 477 switch(*++p1) { 478 case '\0': 479 *p2++ = '\\'; 480 *p2 = '\0'; 481 return; /* incomplete escape sequence */ 482 case 'a': 483 /* *p2 = '\a'; */ 484 *p2 = '\007'; 485 break; 486 case 'b': 487 *p2 = '\b'; 488 break; 489 case 'f': 490 *p2 = '\f'; 491 break; 492 case 'n': 493 *p2 = '\n'; 494 break; 495 case 'r': 496 *p2 = '\r'; 497 break; 498 case 't': 499 *p2 = '\t'; 500 break; 501 case 'v': 502 *p2 = '\v'; 503 break; 504 default: 505 *p2 = *p1; 506 break; 507 } 508 } else 509 *p2 = *p1; 510 } 511 } 512 513 static __dead void 514 badcnt(char *s) 515 { 516 errx(1, "%s: bad byte count", s); 517 } 518 519 static __dead void 520 badsfmt(void) 521 { 522 errx(1, "%%s: requires a precision or a byte count"); 523 } 524 525 static __dead void 526 badfmt(const char *fmt) 527 { 528 errx(1, "\"%s\": bad format", fmt); 529 } 530 531 static __dead void 532 badconv(char *ch) 533 { 534 errx(1, "%%%s: bad conversion character", ch); 535 } 536