1 /*- 2 * Copyright 2014 Garrett D'Amore <[email protected]> 3 * Copyright 2010 Nexenta Systems, Inc. All rights reserved. 4 * Copyright (c) 1989, 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 * 4. 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 * Important: This file is used both as a standalone program /usr/bin/printf 33 * and as a builtin for /bin/sh (#define SHELL). 34 */ 35 36 #ifndef SHELL 37 #ifndef lint 38 static char const copyright[] = 39 "@(#) Copyright (c) 1989, 1993\n\ 40 The Regents of the University of California. All rights reserved.\n"; 41 #endif /* not lint */ 42 #endif 43 44 #ifndef lint 45 #if 0 46 static char const sccsid[] = "@(#)printf.c 8.1 (Berkeley) 7/20/93"; 47 #endif 48 static const char rcsid[] = 49 "$FreeBSD$"; 50 #endif /* not lint */ 51 52 #include <sys/types.h> 53 54 #include <ctype.h> 55 #include <err.h> 56 #include <errno.h> 57 #include <inttypes.h> 58 #include <limits.h> 59 #include <locale.h> 60 #include <stdio.h> 61 #include <stdlib.h> 62 #include <string.h> 63 #include <unistd.h> 64 #include <wchar.h> 65 66 #ifdef SHELL 67 #define main printfcmd 68 #include "bltin/bltin.h" 69 #include "error.h" 70 #include "options.h" 71 #endif 72 73 #define PF(f, func) do { \ 74 char *b = NULL; \ 75 if (havewidth) \ 76 if (haveprec) \ 77 (void)asprintf(&b, f, fieldwidth, precision, func); \ 78 else \ 79 (void)asprintf(&b, f, fieldwidth, func); \ 80 else if (haveprec) \ 81 (void)asprintf(&b, f, precision, func); \ 82 else \ 83 (void)asprintf(&b, f, func); \ 84 if (b) { \ 85 (void)fputs(b, stdout); \ 86 free(b); \ 87 } \ 88 } while (0) 89 90 static int asciicode(void); 91 static char *printf_doformat(char *, int *); 92 static int escape(char *, int, size_t *); 93 static int getchr(void); 94 static int getfloating(long double *, int); 95 static int getint(int *); 96 static int getnum(intmax_t *, uintmax_t *, int); 97 static const char 98 *getstr(void); 99 static char *mknum(char *, char); 100 static void usage(void); 101 102 static const char digits[] = "0123456789"; 103 104 static int myargc; 105 static char **myargv; 106 static char **gargv; 107 static char **maxargv; 108 109 int 110 main(int argc, char *argv[]) 111 { 112 size_t len; 113 int chopped, end, rval; 114 char *format, *fmt, *start; 115 #ifndef SHELL 116 int ch; 117 118 (void) setlocale(LC_ALL, ""); 119 #endif 120 121 #ifdef SHELL 122 nextopt(""); 123 argc -= argptr - argv; 124 argv = argptr; 125 #else 126 while ((ch = getopt(argc, argv, "")) != -1) 127 switch (ch) { 128 case '?': 129 default: 130 usage(); 131 return (1); 132 } 133 argc -= optind; 134 argv += optind; 135 #endif 136 137 if (argc < 1) { 138 usage(); 139 return (1); 140 } 141 142 #ifdef SHELL 143 INTOFF; 144 #endif 145 /* 146 * Basic algorithm is to scan the format string for conversion 147 * specifications -- once one is found, find out if the field 148 * width or precision is a '*'; if it is, gather up value. Note, 149 * format strings are reused as necessary to use up the provided 150 * arguments, arguments of zero/null string are provided to use 151 * up the format string. 152 */ 153 fmt = format = *argv; 154 chopped = escape(fmt, 1, &len); /* backslash interpretation */ 155 rval = end = 0; 156 gargv = ++argv; 157 158 for (;;) { 159 maxargv = gargv; 160 161 myargv = gargv; 162 for (myargc = 0; gargv[myargc]; myargc++) 163 /* nop */; 164 start = fmt; 165 while (fmt < format + len) { 166 if (fmt[0] == '%') { 167 fwrite(start, 1, fmt - start, stdout); 168 if (fmt[1] == '%') { 169 /* %% prints a % */ 170 putchar('%'); 171 fmt += 2; 172 } else { 173 fmt = printf_doformat(fmt, &rval); 174 if (fmt == NULL) { 175 #ifdef SHELL 176 INTON; 177 #endif 178 return (1); 179 } 180 end = 0; 181 } 182 start = fmt; 183 } else 184 fmt++; 185 if (gargv > maxargv) 186 maxargv = gargv; 187 } 188 gargv = maxargv; 189 190 if (end == 1) { 191 warnx("missing format character"); 192 #ifdef SHELL 193 INTON; 194 #endif 195 return (1); 196 } 197 fwrite(start, 1, fmt - start, stdout); 198 if (chopped || !*gargv) { 199 #ifdef SHELL 200 INTON; 201 #endif 202 return (rval); 203 } 204 /* Restart at the beginning of the format string. */ 205 fmt = format; 206 end = 1; 207 } 208 /* NOTREACHED */ 209 } 210 211 212 static char * 213 printf_doformat(char *fmt, int *rval) 214 { 215 static const char skip1[] = "#'-+ 0"; 216 int fieldwidth, haveprec, havewidth, mod_ldbl, precision; 217 char convch, nextch; 218 char *start; 219 char **fargv; 220 char *dptr; 221 int l; 222 223 start = alloca(strlen(fmt) + 1); 224 225 dptr = start; 226 *dptr++ = '%'; 227 *dptr = 0; 228 229 fmt++; 230 231 /* look for "n$" field index specifier */ 232 l = strspn(fmt, digits); 233 if ((l > 0) && (fmt[l] == '$')) { 234 int idx = atoi(fmt); 235 if (idx <= myargc) { 236 gargv = &myargv[idx - 1]; 237 } else { 238 gargv = &myargv[myargc]; 239 } 240 if (gargv > maxargv) 241 maxargv = gargv; 242 fmt += l + 1; 243 244 /* save format argument */ 245 fargv = gargv; 246 } else { 247 fargv = NULL; 248 } 249 250 /* skip to field width */ 251 while (strchr(skip1, *fmt) != NULL) { 252 *dptr++ = *fmt++; 253 *dptr = 0; 254 } 255 256 if (*fmt == '*') { 257 258 fmt++; 259 l = strspn(fmt, digits); 260 if ((l > 0) && (fmt[l] == '$')) { 261 int idx = atoi(fmt); 262 if (idx <= myargc) { 263 gargv = &myargv[idx - 1]; 264 } else { 265 gargv = &myargv[myargc]; 266 } 267 fmt += l + 1; 268 } 269 270 if (getint(&fieldwidth)) 271 return (NULL); 272 if (gargv > maxargv) 273 maxargv = gargv; 274 havewidth = 1; 275 276 *dptr++ = '*'; 277 *dptr = 0; 278 } else { 279 havewidth = 0; 280 281 /* skip to possible '.', get following precision */ 282 while (isdigit(*fmt)) { 283 *dptr++ = *fmt++; 284 *dptr = 0; 285 } 286 } 287 288 if (*fmt == '.') { 289 /* precision present? */ 290 fmt++; 291 *dptr++ = '.'; 292 293 if (*fmt == '*') { 294 295 fmt++; 296 l = strspn(fmt, digits); 297 if ((l > 0) && (fmt[l] == '$')) { 298 int idx = atoi(fmt); 299 if (idx <= myargc) { 300 gargv = &myargv[idx - 1]; 301 } else { 302 gargv = &myargv[myargc]; 303 } 304 fmt += l + 1; 305 } 306 307 if (getint(&precision)) 308 return (NULL); 309 if (gargv > maxargv) 310 maxargv = gargv; 311 haveprec = 1; 312 *dptr++ = '*'; 313 *dptr = 0; 314 } else { 315 haveprec = 0; 316 317 /* skip to conversion char */ 318 while (isdigit(*fmt)) { 319 *dptr++ = *fmt++; 320 *dptr = 0; 321 } 322 } 323 } else 324 haveprec = 0; 325 if (!*fmt) { 326 warnx("missing format character"); 327 return (NULL); 328 } 329 *dptr++ = *fmt; 330 *dptr = 0; 331 332 /* 333 * Look for a length modifier. POSIX doesn't have these, so 334 * we only support them for floating-point conversions, which 335 * are extensions. This is useful because the L modifier can 336 * be used to gain extra range and precision, while omitting 337 * it is more likely to produce consistent results on different 338 * architectures. This is not so important for integers 339 * because overflow is the only bad thing that can happen to 340 * them, but consider the command printf %a 1.1 341 */ 342 if (*fmt == 'L') { 343 mod_ldbl = 1; 344 fmt++; 345 if (!strchr("aAeEfFgG", *fmt)) { 346 warnx("bad modifier L for %%%c", *fmt); 347 return (NULL); 348 } 349 } else { 350 mod_ldbl = 0; 351 } 352 353 /* save the current arg offset, and set to the format arg */ 354 if (fargv != NULL) { 355 gargv = fargv; 356 } 357 358 convch = *fmt; 359 nextch = *++fmt; 360 361 *fmt = '\0'; 362 switch (convch) { 363 case 'b': { 364 size_t len; 365 char *p; 366 int getout; 367 368 p = strdup(getstr()); 369 if (p == NULL) { 370 warnx("%s", strerror(ENOMEM)); 371 return (NULL); 372 } 373 getout = escape(p, 0, &len); 374 *(fmt - 1) = 's'; 375 PF(start, p); 376 *(fmt - 1) = 'b'; 377 free(p); 378 if (getout) 379 return (fmt); 380 break; 381 } 382 case 'c': { 383 char p; 384 385 p = getchr(); 386 PF(start, p); 387 break; 388 } 389 case 's': { 390 const char *p; 391 392 p = getstr(); 393 PF(start, p); 394 break; 395 } 396 case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': { 397 char *f; 398 intmax_t val; 399 uintmax_t uval; 400 int signedconv; 401 402 signedconv = (convch == 'd' || convch == 'i'); 403 if ((f = mknum(start, convch)) == NULL) 404 return (NULL); 405 if (getnum(&val, &uval, signedconv)) 406 *rval = 1; 407 if (signedconv) 408 PF(f, val); 409 else 410 PF(f, uval); 411 break; 412 } 413 case 'e': case 'E': 414 case 'f': case 'F': 415 case 'g': case 'G': 416 case 'a': case 'A': { 417 long double p; 418 419 if (getfloating(&p, mod_ldbl)) 420 *rval = 1; 421 if (mod_ldbl) 422 PF(start, p); 423 else 424 PF(start, (double)p); 425 break; 426 } 427 default: 428 warnx("illegal format character %c", convch); 429 return (NULL); 430 } 431 *fmt = nextch; 432 /* return the gargv to the next element */ 433 return (fmt); 434 } 435 436 static char * 437 mknum(char *str, char ch) 438 { 439 static char *copy; 440 static size_t copy_size; 441 char *newcopy; 442 size_t len, newlen; 443 444 len = strlen(str) + 2; 445 if (len > copy_size) { 446 newlen = ((len + 1023) >> 10) << 10; 447 if ((newcopy = realloc(copy, newlen)) == NULL) 448 { 449 warnx("%s", strerror(ENOMEM)); 450 return (NULL); 451 } 452 copy = newcopy; 453 copy_size = newlen; 454 } 455 456 memmove(copy, str, len - 3); 457 copy[len - 3] = 'j'; 458 copy[len - 2] = ch; 459 copy[len - 1] = '\0'; 460 return (copy); 461 } 462 463 static int 464 escape(char *fmt, int percent, size_t *len) 465 { 466 char *save, *store, c; 467 int value; 468 469 for (save = store = fmt; ((c = *fmt) != 0); ++fmt, ++store) { 470 if (c != '\\') { 471 *store = c; 472 continue; 473 } 474 switch (*++fmt) { 475 case '\0': /* EOS, user error */ 476 *store = '\\'; 477 *++store = '\0'; 478 *len = store - save; 479 return (0); 480 case '\\': /* backslash */ 481 case '\'': /* single quote */ 482 *store = *fmt; 483 break; 484 case 'a': /* bell/alert */ 485 *store = '\a'; 486 break; 487 case 'b': /* backspace */ 488 *store = '\b'; 489 break; 490 case 'c': 491 *store = '\0'; 492 *len = store - save; 493 return (1); 494 case 'f': /* form-feed */ 495 *store = '\f'; 496 break; 497 case 'n': /* newline */ 498 *store = '\n'; 499 break; 500 case 'r': /* carriage-return */ 501 *store = '\r'; 502 break; 503 case 't': /* horizontal tab */ 504 *store = '\t'; 505 break; 506 case 'v': /* vertical tab */ 507 *store = '\v'; 508 break; 509 /* octal constant */ 510 case '0': case '1': case '2': case '3': 511 case '4': case '5': case '6': case '7': 512 c = (!percent && *fmt == '0') ? 4 : 3; 513 for (value = 0; 514 c-- && *fmt >= '0' && *fmt <= '7'; ++fmt) { 515 value <<= 3; 516 value += *fmt - '0'; 517 } 518 --fmt; 519 if (percent && value == '%') { 520 *store++ = '%'; 521 *store = '%'; 522 } else 523 *store = (char)value; 524 break; 525 default: 526 *store = *fmt; 527 break; 528 } 529 } 530 *store = '\0'; 531 *len = store - save; 532 return (0); 533 } 534 535 static int 536 getchr(void) 537 { 538 if (!*gargv) 539 return ('\0'); 540 return ((int)**gargv++); 541 } 542 543 static const char * 544 getstr(void) 545 { 546 if (!*gargv) 547 return (""); 548 return (*gargv++); 549 } 550 551 static int 552 getint(int *ip) 553 { 554 intmax_t val; 555 uintmax_t uval; 556 int rval; 557 558 if (getnum(&val, &uval, 1)) 559 return (1); 560 rval = 0; 561 if (val < INT_MIN || val > INT_MAX) { 562 warnx("%s: %s", *gargv, strerror(ERANGE)); 563 rval = 1; 564 } 565 *ip = (int)val; 566 return (rval); 567 } 568 569 static int 570 getnum(intmax_t *ip, uintmax_t *uip, int signedconv) 571 { 572 char *ep; 573 int rval; 574 575 if (!*gargv) { 576 *ip = 0; 577 return (0); 578 } 579 if (**gargv == '"' || **gargv == '\'') { 580 if (signedconv) 581 *ip = asciicode(); 582 else 583 *uip = asciicode(); 584 return (0); 585 } 586 rval = 0; 587 errno = 0; 588 if (signedconv) 589 *ip = strtoimax(*gargv, &ep, 0); 590 else 591 *uip = strtoumax(*gargv, &ep, 0); 592 if (ep == *gargv) { 593 warnx("%s: expected numeric value", *gargv); 594 rval = 1; 595 } 596 else if (*ep != '\0') { 597 warnx("%s: not completely converted", *gargv); 598 rval = 1; 599 } 600 if (errno == ERANGE) { 601 warnx("%s: %s", *gargv, strerror(ERANGE)); 602 rval = 1; 603 } 604 ++gargv; 605 return (rval); 606 } 607 608 static int 609 getfloating(long double *dp, int mod_ldbl) 610 { 611 char *ep; 612 int rval; 613 614 if (!*gargv) { 615 *dp = 0.0; 616 return (0); 617 } 618 if (**gargv == '"' || **gargv == '\'') { 619 *dp = asciicode(); 620 return (0); 621 } 622 rval = 0; 623 errno = 0; 624 if (mod_ldbl) 625 *dp = strtold(*gargv, &ep); 626 else 627 *dp = strtod(*gargv, &ep); 628 if (ep == *gargv) { 629 warnx("%s: expected numeric value", *gargv); 630 rval = 1; 631 } else if (*ep != '\0') { 632 warnx("%s: not completely converted", *gargv); 633 rval = 1; 634 } 635 if (errno == ERANGE) { 636 warnx("%s: %s", *gargv, strerror(ERANGE)); 637 rval = 1; 638 } 639 ++gargv; 640 return (rval); 641 } 642 643 static int 644 asciicode(void) 645 { 646 int ch; 647 wchar_t wch; 648 mbstate_t mbs; 649 650 ch = (unsigned char)**gargv; 651 if (ch == '\'' || ch == '"') { 652 memset(&mbs, 0, sizeof(mbs)); 653 switch (mbrtowc(&wch, *gargv + 1, MB_LEN_MAX, &mbs)) { 654 case (size_t)-2: 655 case (size_t)-1: 656 wch = (unsigned char)gargv[0][1]; 657 break; 658 case 0: 659 wch = 0; 660 break; 661 } 662 ch = wch; 663 } 664 ++gargv; 665 return (ch); 666 } 667 668 static void 669 usage(void) 670 { 671 (void)fprintf(stderr, "usage: printf format [arguments ...]\n"); 672 } 673