1 /*- 2 * SPDX-License-Identifier: BSD-3-Clause 3 * 4 * Copyright 2018 Staysail Systems, Inc. <info@staysail.tech> 5 * Copyright 2014 Garrett D'Amore <garrett@damore.org> 6 * Copyright 2010 Nexenta Systems, Inc. All rights reserved. 7 * Copyright (c) 1989, 1993 8 * The Regents of the University of California. All rights reserved. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 3. Neither the name of the University nor the names of its contributors 19 * may be used to endorse or promote products derived from this software 20 * without specific prior written permission. 21 * 22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 32 * SUCH DAMAGE. 33 * 34 * @(#) Copyright (c) 1989, 1993 The Regents of the University of California. All rights reserved. 35 * @(#)printf.c 8.1 (Berkeley) 7/20/93 36 * $FreeBSD: head/usr.bin/printf/printf.c 337618 2018-08-11 11:13:34Z jilles $ 37 */ 38 /* 39 * Important: This file is used both as a standalone program /usr/bin/printf 40 * and as a builtin for /bin/sh (#define SHELL). 41 */ 42 43 #include <sys/types.h> 44 45 #include <ctype.h> 46 #include <err.h> 47 #include <errno.h> 48 #include <inttypes.h> 49 #include <limits.h> 50 #include <locale.h> 51 #include <stdio.h> 52 #include <stdlib.h> 53 #include <string.h> 54 #include <unistd.h> 55 #include <wchar.h> 56 57 #ifdef SHELL 58 #define main printfcmd 59 #include "bltin/bltin.h" 60 #include "error.h" 61 #include "options.h" 62 #endif 63 64 #define PF(f, func) do { \ 65 if (havewidth) \ 66 if (haveprec) \ 67 printf(f, fieldwidth, precision, func); \ 68 else \ 69 printf(f, fieldwidth, func); \ 70 else if (haveprec) \ 71 printf(f, precision, func); \ 72 else \ 73 printf(f, func); \ 74 } while (0) 75 76 static int asciicode(void); 77 static char *printf_doformat(char *, int *); 78 static int escape(char *, int, size_t *); 79 static int getchr(void); 80 static int getfloating(long double *, int); 81 static int getint(int *); 82 static int getnum(intmax_t *, uintmax_t *, int); 83 static const char 84 *getstr(void); 85 static char *mknum(char *, char); 86 static void usage(void); 87 88 static const char digits[] = "0123456789"; 89 90 static char end_fmt[1]; 91 92 static int myargc; 93 static char **myargv; 94 static char **gargv; 95 static char **maxargv; 96 97 int 98 main(int argc, char *argv[]) 99 { 100 size_t len; 101 int end, rval; 102 char *format, *fmt, *start; 103 104 #ifndef SHELL 105 setlocale(LC_ALL, ""); 106 #endif 107 108 /* 109 * We may not use getopt(3) because calling 110 * "printf -f%s oo" may not result in an invalid 111 * option error. 112 * However common usage and other implementations seem 113 * to indicate that we need to allow -- as a discardable 114 * option separator. 115 */ 116 if (argc > 1 && strcmp(argv[1], "--") == 0) { 117 argc--; 118 argv++; 119 } 120 121 if (argc < 2) { 122 usage(); 123 return (1); 124 } 125 126 argv++; 127 128 #ifdef SHELL 129 INTOFF; 130 #endif 131 /* 132 * Basic algorithm is to scan the format string for conversion 133 * specifications -- once one is found, find out if the field 134 * width or precision is a '*'; if it is, gather up value. Note, 135 * format strings are reused as necessary to use up the provided 136 * arguments, arguments of zero/null string are provided to use 137 * up the format string. 138 */ 139 fmt = format = *argv; 140 escape(fmt, 1, &len); /* backslash interpretation */ 141 rval = end = 0; 142 gargv = ++argv; 143 144 for (;;) { 145 maxargv = gargv; 146 147 myargv = gargv; 148 for (myargc = 0; gargv[myargc]; myargc++) 149 /* nop */; 150 start = fmt; 151 while (fmt < format + len) { 152 if (fmt[0] == '%') { 153 fwrite(start, 1, fmt - start, stdout); 154 if (fmt[1] == '%') { 155 /* %% prints a % */ 156 putchar('%'); 157 fmt += 2; 158 } else { 159 fmt = printf_doformat(fmt, &rval); 160 if (fmt == NULL || fmt == end_fmt) { 161 #ifdef SHELL 162 INTON; 163 #endif 164 return (fmt == NULL ? 1 : rval); 165 } 166 end = 0; 167 } 168 start = fmt; 169 } else 170 fmt++; 171 if (gargv > maxargv) 172 maxargv = gargv; 173 } 174 gargv = maxargv; 175 176 if (end == 1) { 177 warnx("missing format character"); 178 #ifdef SHELL 179 INTON; 180 #endif 181 return (1); 182 } 183 fwrite(start, 1, fmt - start, stdout); 184 if (!*gargv) { 185 #ifdef SHELL 186 INTON; 187 #endif 188 return (rval); 189 } 190 /* Restart at the beginning of the format string. */ 191 fmt = format; 192 end = 1; 193 } 194 /* NOTREACHED */ 195 } 196 197 198 static char * 199 printf_doformat(char *fmt, int *rval) 200 { 201 static const char skip1[] = "#'-+ 0"; 202 int fieldwidth, haveprec, havewidth, mod_ldbl, precision; 203 char convch, nextch; 204 char start[strlen(fmt) + 1]; 205 char **fargv; 206 char *dptr; 207 int l; 208 209 dptr = start; 210 *dptr++ = '%'; 211 *dptr = 0; 212 213 fmt++; 214 215 /* look for "n$" field index specifier */ 216 l = strspn(fmt, digits); 217 if ((l > 0) && (fmt[l] == '$')) { 218 int idx = atoi(fmt); 219 if (idx <= myargc) { 220 gargv = &myargv[idx - 1]; 221 } else { 222 gargv = &myargv[myargc]; 223 } 224 if (gargv > maxargv) 225 maxargv = gargv; 226 fmt += l + 1; 227 228 /* save format argument */ 229 fargv = gargv; 230 } else { 231 fargv = NULL; 232 } 233 234 /* skip to field width */ 235 while (*fmt && strchr(skip1, *fmt) != NULL) { 236 *dptr++ = *fmt++; 237 *dptr = 0; 238 } 239 240 if (*fmt == '*') { 241 242 fmt++; 243 l = strspn(fmt, digits); 244 if ((l > 0) && (fmt[l] == '$')) { 245 int idx = atoi(fmt); 246 if (fargv == NULL) { 247 warnx("incomplete use of n$"); 248 return (NULL); 249 } 250 if (idx <= myargc) { 251 gargv = &myargv[idx - 1]; 252 } else { 253 gargv = &myargv[myargc]; 254 } 255 fmt += l + 1; 256 } else if (fargv != NULL) { 257 warnx("incomplete use of n$"); 258 return (NULL); 259 } 260 261 if (getint(&fieldwidth)) 262 return (NULL); 263 if (gargv > maxargv) 264 maxargv = gargv; 265 havewidth = 1; 266 267 *dptr++ = '*'; 268 *dptr = 0; 269 } else { 270 havewidth = 0; 271 272 /* skip to possible '.', get following precision */ 273 while (isdigit(*fmt)) { 274 *dptr++ = *fmt++; 275 *dptr = 0; 276 } 277 } 278 279 if (*fmt == '.') { 280 /* precision present? */ 281 fmt++; 282 *dptr++ = '.'; 283 284 if (*fmt == '*') { 285 286 fmt++; 287 l = strspn(fmt, digits); 288 if ((l > 0) && (fmt[l] == '$')) { 289 int idx = atoi(fmt); 290 if (fargv == NULL) { 291 warnx("incomplete use of n$"); 292 return (NULL); 293 } 294 if (idx <= myargc) { 295 gargv = &myargv[idx - 1]; 296 } else { 297 gargv = &myargv[myargc]; 298 } 299 fmt += l + 1; 300 } else if (fargv != NULL) { 301 warnx("incomplete use of n$"); 302 return (NULL); 303 } 304 305 if (getint(&precision)) 306 return (NULL); 307 if (gargv > maxargv) 308 maxargv = gargv; 309 haveprec = 1; 310 *dptr++ = '*'; 311 *dptr = 0; 312 } else { 313 haveprec = 0; 314 315 /* skip to conversion char */ 316 while (isdigit(*fmt)) { 317 *dptr++ = *fmt++; 318 *dptr = 0; 319 } 320 } 321 } else 322 haveprec = 0; 323 if (!*fmt) { 324 warnx("missing format character"); 325 return (NULL); 326 } 327 *dptr++ = *fmt; 328 *dptr = 0; 329 330 /* 331 * Look for a length modifier. POSIX doesn't have these, so 332 * we only support them for floating-point conversions, which 333 * are extensions. This is useful because the L modifier can 334 * be used to gain extra range and precision, while omitting 335 * it is more likely to produce consistent results on different 336 * architectures. This is not so important for integers 337 * because overflow is the only bad thing that can happen to 338 * them, but consider the command printf %a 1.1 339 */ 340 if (*fmt == 'L') { 341 mod_ldbl = 1; 342 fmt++; 343 if (!strchr("aAeEfFgG", *fmt)) { 344 warnx("bad modifier L for %%%c", *fmt); 345 return (NULL); 346 } 347 } else { 348 mod_ldbl = 0; 349 } 350 351 /* save the current arg offset, and set to the format arg */ 352 if (fargv != NULL) { 353 gargv = fargv; 354 } 355 356 convch = *fmt; 357 nextch = *++fmt; 358 359 *fmt = '\0'; 360 switch (convch) { 361 case 'b': { 362 size_t len; 363 char *p; 364 int getout; 365 366 /* Convert "b" to "s" for output. */ 367 start[strlen(start) - 1] = 's'; 368 if ((p = strdup(getstr())) == NULL) { 369 warnx("%s", strerror(ENOMEM)); 370 return (NULL); 371 } 372 getout = escape(p, 0, &len); 373 PF(start, p); 374 /* Restore format for next loop. */ 375 376 free(p); 377 if (getout) 378 return (end_fmt); 379 break; 380 } 381 case 'c': { 382 char p; 383 384 p = getchr(); 385 if (p != '\0') 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 warnx("%s", strerror(ENOMEM)); 449 return (NULL); 450 } 451 copy = newcopy; 452 copy_size = newlen; 453 } 454 455 memmove(copy, str, len - 3); 456 copy[len - 3] = 'j'; 457 copy[len - 2] = ch; 458 copy[len - 1] = '\0'; 459 return (copy); 460 } 461 462 static int 463 escape(char *fmt, int percent, size_t *len) 464 { 465 char *save, *store, c; 466 int value; 467 468 for (save = store = fmt; ((c = *fmt) != 0); ++fmt, ++store) { 469 if (c != '\\') { 470 *store = c; 471 continue; 472 } 473 switch (*++fmt) { 474 case '\0': /* EOS, user error */ 475 *store = '\\'; 476 *++store = '\0'; 477 *len = store - save; 478 return (0); 479 case '\\': /* backslash */ 480 case '\'': /* single quote */ 481 *store = *fmt; 482 break; 483 case 'a': /* bell/alert */ 484 *store = '\a'; 485 break; 486 case 'b': /* backspace */ 487 *store = '\b'; 488 break; 489 case 'c': 490 if (!percent) { 491 *store = '\0'; 492 *len = store - save; 493 return (1); 494 } 495 *store = 'c'; 496 break; 497 case 'f': /* form-feed */ 498 *store = '\f'; 499 break; 500 case 'n': /* newline */ 501 *store = '\n'; 502 break; 503 case 'r': /* carriage-return */ 504 *store = '\r'; 505 break; 506 case 't': /* horizontal tab */ 507 *store = '\t'; 508 break; 509 case 'v': /* vertical tab */ 510 *store = '\v'; 511 break; 512 /* octal constant */ 513 case '0': case '1': case '2': case '3': 514 case '4': case '5': case '6': case '7': 515 c = (!percent && *fmt == '0') ? 4 : 3; 516 for (value = 0; 517 c-- && *fmt >= '0' && *fmt <= '7'; ++fmt) { 518 value <<= 3; 519 value += *fmt - '0'; 520 } 521 --fmt; 522 if (percent && value == '%') { 523 *store++ = '%'; 524 *store = '%'; 525 } else 526 *store = (char)value; 527 break; 528 default: 529 *store = *fmt; 530 break; 531 } 532 } 533 *store = '\0'; 534 *len = store - save; 535 return (0); 536 } 537 538 static int 539 getchr(void) 540 { 541 if (!*gargv) 542 return ('\0'); 543 return ((int)**gargv++); 544 } 545 546 static const char * 547 getstr(void) 548 { 549 if (!*gargv) 550 return (""); 551 return (*gargv++); 552 } 553 554 static int 555 getint(int *ip) 556 { 557 intmax_t val; 558 uintmax_t uval; 559 int rval; 560 561 if (getnum(&val, &uval, 1)) 562 return (1); 563 rval = 0; 564 if (val < INT_MIN || val > INT_MAX) { 565 warnx("%s: %s", *gargv, strerror(ERANGE)); 566 rval = 1; 567 } 568 *ip = (int)val; 569 return (rval); 570 } 571 572 static int 573 getnum(intmax_t *ip, uintmax_t *uip, int signedconv) 574 { 575 char *ep; 576 int rval; 577 578 if (!*gargv) { 579 *ip = *uip = 0; 580 return (0); 581 } 582 if (**gargv == '"' || **gargv == '\'') { 583 if (signedconv) 584 *ip = asciicode(); 585 else 586 *uip = asciicode(); 587 return (0); 588 } 589 rval = 0; 590 errno = 0; 591 if (signedconv) 592 *ip = strtoimax(*gargv, &ep, 0); 593 else 594 *uip = strtoumax(*gargv, &ep, 0); 595 if (ep == *gargv) { 596 warnx("%s: expected numeric value", *gargv); 597 rval = 1; 598 } 599 else if (*ep != '\0') { 600 warnx("%s: not completely converted", *gargv); 601 rval = 1; 602 } 603 if (errno == ERANGE) { 604 warnx("%s: %s", *gargv, strerror(ERANGE)); 605 rval = 1; 606 } 607 ++gargv; 608 return (rval); 609 } 610 611 static int 612 getfloating(long double *dp, int mod_ldbl) 613 { 614 char *ep; 615 int rval; 616 617 if (!*gargv) { 618 *dp = 0.0; 619 return (0); 620 } 621 if (**gargv == '"' || **gargv == '\'') { 622 *dp = asciicode(); 623 return (0); 624 } 625 rval = 0; 626 errno = 0; 627 if (mod_ldbl) 628 *dp = strtold(*gargv, &ep); 629 else 630 *dp = strtod(*gargv, &ep); 631 if (ep == *gargv) { 632 warnx("%s: expected numeric value", *gargv); 633 rval = 1; 634 } else if (*ep != '\0') { 635 warnx("%s: not completely converted", *gargv); 636 rval = 1; 637 } 638 if (errno == ERANGE) { 639 warnx("%s: %s", *gargv, strerror(ERANGE)); 640 rval = 1; 641 } 642 ++gargv; 643 return (rval); 644 } 645 646 static int 647 asciicode(void) 648 { 649 int ch; 650 wchar_t wch; 651 mbstate_t mbs; 652 653 ch = (unsigned char)**gargv; 654 if (ch == '\'' || ch == '"') { 655 memset(&mbs, 0, sizeof(mbs)); 656 switch (mbrtowc(&wch, *gargv + 1, MB_LEN_MAX, &mbs)) { 657 case (size_t)-2: 658 case (size_t)-1: 659 wch = (unsigned char)gargv[0][1]; 660 break; 661 case 0: 662 wch = 0; 663 break; 664 } 665 ch = wch; 666 } 667 ++gargv; 668 return (ch); 669 } 670 671 static void 672 usage(void) 673 { 674 fprintf(stderr, "usage: printf format [arguments ...]\n"); 675 } 676