1 /*- 2 * Copyright 2014 Garrett D'Amore <garrett@damore.org> 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 char end_fmt[1]; 105 106 static int myargc; 107 static char **myargv; 108 static char **gargv; 109 static char **maxargv; 110 111 int 112 main(int argc, char *argv[]) 113 { 114 size_t len; 115 int end, rval; 116 char *format, *fmt, *start; 117 #ifndef SHELL 118 int ch; 119 120 (void) setlocale(LC_ALL, ""); 121 #endif 122 123 #ifdef SHELL 124 nextopt(""); 125 argc -= argptr - argv; 126 argv = argptr; 127 #else 128 while ((ch = getopt(argc, argv, "")) != -1) 129 switch (ch) { 130 case '?': 131 default: 132 usage(); 133 return (1); 134 } 135 argc -= optind; 136 argv += optind; 137 #endif 138 139 if (argc < 1) { 140 usage(); 141 return (1); 142 } 143 144 #ifdef SHELL 145 INTOFF; 146 #endif 147 /* 148 * Basic algorithm is to scan the format string for conversion 149 * specifications -- once one is found, find out if the field 150 * width or precision is a '*'; if it is, gather up value. Note, 151 * format strings are reused as necessary to use up the provided 152 * arguments, arguments of zero/null string are provided to use 153 * up the format string. 154 */ 155 fmt = format = *argv; 156 escape(fmt, 1, &len); /* backslash interpretation */ 157 rval = end = 0; 158 gargv = ++argv; 159 160 for (;;) { 161 maxargv = gargv; 162 163 myargv = gargv; 164 for (myargc = 0; gargv[myargc]; myargc++) 165 /* nop */; 166 start = fmt; 167 while (fmt < format + len) { 168 if (fmt[0] == '%') { 169 fwrite(start, 1, fmt - start, stdout); 170 if (fmt[1] == '%') { 171 /* %% prints a % */ 172 putchar('%'); 173 fmt += 2; 174 } else { 175 fmt = printf_doformat(fmt, &rval); 176 if (fmt == NULL || fmt == end_fmt) { 177 #ifdef SHELL 178 INTON; 179 #endif 180 return (fmt == NULL ? 1 : rval); 181 } 182 end = 0; 183 } 184 start = fmt; 185 } else 186 fmt++; 187 if (gargv > maxargv) 188 maxargv = gargv; 189 } 190 gargv = maxargv; 191 192 if (end == 1) { 193 warnx("missing format character"); 194 #ifdef SHELL 195 INTON; 196 #endif 197 return (1); 198 } 199 fwrite(start, 1, fmt - start, stdout); 200 if (!*gargv) { 201 #ifdef SHELL 202 INTON; 203 #endif 204 return (rval); 205 } 206 /* Restart at the beginning of the format string. */ 207 fmt = format; 208 end = 1; 209 } 210 /* NOTREACHED */ 211 } 212 213 214 static char * 215 printf_doformat(char *fmt, int *rval) 216 { 217 static const char skip1[] = "#'-+ 0"; 218 int fieldwidth, haveprec, havewidth, mod_ldbl, precision; 219 char convch, nextch; 220 char start[strlen(fmt) + 1]; 221 char **fargv; 222 char *dptr; 223 int l; 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 fputs(p, stdout); 375 free(p); 376 if (getout) 377 return (end_fmt); 378 break; 379 } 380 case 'c': { 381 char p; 382 383 p = getchr(); 384 PF(start, p); 385 break; 386 } 387 case 's': { 388 const char *p; 389 390 p = getstr(); 391 PF(start, p); 392 break; 393 } 394 case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': { 395 char *f; 396 intmax_t val; 397 uintmax_t uval; 398 int signedconv; 399 400 signedconv = (convch == 'd' || convch == 'i'); 401 if ((f = mknum(start, convch)) == NULL) 402 return (NULL); 403 if (getnum(&val, &uval, signedconv)) 404 *rval = 1; 405 if (signedconv) 406 PF(f, val); 407 else 408 PF(f, uval); 409 break; 410 } 411 case 'e': case 'E': 412 case 'f': case 'F': 413 case 'g': case 'G': 414 case 'a': case 'A': { 415 long double p; 416 417 if (getfloating(&p, mod_ldbl)) 418 *rval = 1; 419 if (mod_ldbl) 420 PF(start, p); 421 else 422 PF(start, (double)p); 423 break; 424 } 425 default: 426 warnx("illegal format character %c", convch); 427 return (NULL); 428 } 429 *fmt = nextch; 430 /* return the gargv to the next element */ 431 return (fmt); 432 } 433 434 static char * 435 mknum(char *str, char ch) 436 { 437 static char *copy; 438 static size_t copy_size; 439 char *newcopy; 440 size_t len, newlen; 441 442 len = strlen(str) + 2; 443 if (len > copy_size) { 444 newlen = ((len + 1023) >> 10) << 10; 445 if ((newcopy = realloc(copy, newlen)) == NULL) 446 { 447 warnx("%s", strerror(ENOMEM)); 448 return (NULL); 449 } 450 copy = newcopy; 451 copy_size = newlen; 452 } 453 454 memmove(copy, str, len - 3); 455 copy[len - 3] = 'j'; 456 copy[len - 2] = ch; 457 copy[len - 1] = '\0'; 458 return (copy); 459 } 460 461 static int 462 escape(char *fmt, int percent, size_t *len) 463 { 464 char *save, *store, c; 465 int value; 466 467 for (save = store = fmt; ((c = *fmt) != 0); ++fmt, ++store) { 468 if (c != '\\') { 469 *store = c; 470 continue; 471 } 472 switch (*++fmt) { 473 case '\0': /* EOS, user error */ 474 *store = '\\'; 475 *++store = '\0'; 476 *len = store - save; 477 return (0); 478 case '\\': /* backslash */ 479 case '\'': /* single quote */ 480 *store = *fmt; 481 break; 482 case 'a': /* bell/alert */ 483 *store = '\a'; 484 break; 485 case 'b': /* backspace */ 486 *store = '\b'; 487 break; 488 case 'c': 489 if (!percent) { 490 *store = '\0'; 491 *len = store - save; 492 return (1); 493 } 494 *store = 'c'; 495 break; 496 case 'f': /* form-feed */ 497 *store = '\f'; 498 break; 499 case 'n': /* newline */ 500 *store = '\n'; 501 break; 502 case 'r': /* carriage-return */ 503 *store = '\r'; 504 break; 505 case 't': /* horizontal tab */ 506 *store = '\t'; 507 break; 508 case 'v': /* vertical tab */ 509 *store = '\v'; 510 break; 511 /* octal constant */ 512 case '0': case '1': case '2': case '3': 513 case '4': case '5': case '6': case '7': 514 c = (!percent && *fmt == '0') ? 4 : 3; 515 for (value = 0; 516 c-- && *fmt >= '0' && *fmt <= '7'; ++fmt) { 517 value <<= 3; 518 value += *fmt - '0'; 519 } 520 --fmt; 521 if (percent && value == '%') { 522 *store++ = '%'; 523 *store = '%'; 524 } else 525 *store = (char)value; 526 break; 527 default: 528 *store = *fmt; 529 break; 530 } 531 } 532 *store = '\0'; 533 *len = store - save; 534 return (0); 535 } 536 537 static int 538 getchr(void) 539 { 540 if (!*gargv) 541 return ('\0'); 542 return ((int)**gargv++); 543 } 544 545 static const char * 546 getstr(void) 547 { 548 if (!*gargv) 549 return (""); 550 return (*gargv++); 551 } 552 553 static int 554 getint(int *ip) 555 { 556 intmax_t val; 557 uintmax_t uval; 558 int rval; 559 560 if (getnum(&val, &uval, 1)) 561 return (1); 562 rval = 0; 563 if (val < INT_MIN || val > INT_MAX) { 564 warnx("%s: %s", *gargv, strerror(ERANGE)); 565 rval = 1; 566 } 567 *ip = (int)val; 568 return (rval); 569 } 570 571 static int 572 getnum(intmax_t *ip, uintmax_t *uip, int signedconv) 573 { 574 char *ep; 575 int rval; 576 577 if (!*gargv) { 578 *ip = *uip = 0; 579 return (0); 580 } 581 if (**gargv == '"' || **gargv == '\'') { 582 if (signedconv) 583 *ip = asciicode(); 584 else 585 *uip = asciicode(); 586 return (0); 587 } 588 rval = 0; 589 errno = 0; 590 if (signedconv) 591 *ip = strtoimax(*gargv, &ep, 0); 592 else 593 *uip = strtoumax(*gargv, &ep, 0); 594 if (ep == *gargv) { 595 warnx("%s: expected numeric value", *gargv); 596 rval = 1; 597 } 598 else if (*ep != '\0') { 599 warnx("%s: not completely converted", *gargv); 600 rval = 1; 601 } 602 if (errno == ERANGE) { 603 warnx("%s: %s", *gargv, strerror(ERANGE)); 604 rval = 1; 605 } 606 ++gargv; 607 return (rval); 608 } 609 610 static int 611 getfloating(long double *dp, int mod_ldbl) 612 { 613 char *ep; 614 int rval; 615 616 if (!*gargv) { 617 *dp = 0.0; 618 return (0); 619 } 620 if (**gargv == '"' || **gargv == '\'') { 621 *dp = asciicode(); 622 return (0); 623 } 624 rval = 0; 625 errno = 0; 626 if (mod_ldbl) 627 *dp = strtold(*gargv, &ep); 628 else 629 *dp = strtod(*gargv, &ep); 630 if (ep == *gargv) { 631 warnx("%s: expected numeric value", *gargv); 632 rval = 1; 633 } else if (*ep != '\0') { 634 warnx("%s: not completely converted", *gargv); 635 rval = 1; 636 } 637 if (errno == ERANGE) { 638 warnx("%s: %s", *gargv, strerror(ERANGE)); 639 rval = 1; 640 } 641 ++gargv; 642 return (rval); 643 } 644 645 static int 646 asciicode(void) 647 { 648 int ch; 649 wchar_t wch; 650 mbstate_t mbs; 651 652 ch = (unsigned char)**gargv; 653 if (ch == '\'' || ch == '"') { 654 memset(&mbs, 0, sizeof(mbs)); 655 switch (mbrtowc(&wch, *gargv + 1, MB_LEN_MAX, &mbs)) { 656 case (size_t)-2: 657 case (size_t)-1: 658 wch = (unsigned char)gargv[0][1]; 659 break; 660 case 0: 661 wch = 0; 662 break; 663 } 664 ch = wch; 665 } 666 ++gargv; 667 return (ch); 668 } 669 670 static void 671 usage(void) 672 { 673 (void)fprintf(stderr, "usage: printf format [arguments ...]\n"); 674 } 675