printf/printf.c

   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: head/usr.bin/printf/printf.c 279503 2015-03-01 21:46:55Z jilles $";
  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 "options.h"
  70 #endif
  71
  72 #define PF(f, func) do {                                                \
  73         char *b = NULL;                                                 \
  74         if (havewidth)                                                  \
  75                 if (haveprec)                                           \
  76                         (void)asprintf(&b, f, fieldwidth, precision, func); \
  77                 else                                                    \
  78                         (void)asprintf(&b, f, fieldwidth, func);        \
  79         else if (haveprec)                                              \
  80                 (void)asprintf(&b, f, precision, func);                 \
  81         else                                                            \
  82                 (void)asprintf(&b, f, func);                            \
  83         if (b) {                                                        \
  84                 (void)fputs(b, stdout);                                 \
  85                 free(b);                                                \
  86         }                                                               \
  87 } while (0)
  88
  89 static int       asciicode(void);
  90 static char     *printf_doformat(char *, int *);
  91 static int       escape(char *, int, size_t *);
  92 static int       getchr(void);
  93 static int       getfloating(long double *, int);
  94 static int       getint(int *);
  95 static int       getnum(intmax_t *, uintmax_t *, int);
  96 static const char
  97                 *getstr(void);
  98 static char     *mknum(char *, char);
  99 static void      usage(void);
 100
 101 static const char digits[] = "0123456789";
 102
 103 static char end_fmt[1];
 104
 105 static int  myargc;
 106 static char **myargv;
 107 static char **gargv;
 108 static char **maxargv;
 109
 110 int
 111 main(int argc, char *argv[])
 112 {
 113         size_t len;
 114         int end, rval;
 115         char *format, *fmt, *start;
 116 #ifndef SHELL
 117         int ch;
 118
 119         (void) setlocale(LC_ALL, "");
 120 #endif
 121
 122 #ifdef SHELL
 123         nextopt("");
 124         argc -= argptr - argv;
 125         argv = argptr;
 126 #else
 127         while ((ch = getopt(argc, argv, "")) != -1)
 128                 switch (ch) {
 129                 case '?':
 130                 default:
 131                         usage();
 132                         return (1);
 133                 }
 134         argc -= optind;
 135         argv += optind;
 136 #endif
 137
 138         if (argc < 1) {
 139                 usage();
 140                 return (1);
 141         }
 142
 143 #ifdef SHELL
 144         INTOFF;
 145 #endif
 146         /*
 147          * Basic algorithm is to scan the format string for conversion
 148          * specifications -- once one is found, find out if the field
 149          * width or precision is a '*'; if it is, gather up value.  Note,
 150          * format strings are reused as necessary to use up the provided
 151          * arguments, arguments of zero/null string are provided to use
 152          * up the format string.
 153          */
 154         fmt = format = *argv;
 155         escape(fmt, 1, &len);           /* backslash interpretation */
 156         rval = end = 0;
 157         gargv = ++argv;
 158
 159         for (;;) {
 160                 maxargv = gargv;
 161
 162                 myargv = gargv;
 163                 for (myargc = 0; gargv[myargc]; myargc++)
 164                         /* nop */;
 165                 start = fmt;
 166                 while (fmt < format + len) {
 167                         if (fmt[0] == '%') {
 168                                 fwrite(start, 1, fmt - start, stdout);
 169                                 if (fmt[1] == '%') {
 170                                         /* %% prints a % */
 171                                         putchar('%');
 172                                         fmt += 2;
 173                                 } else {
 174                                         fmt = printf_doformat(fmt, &rval);
 175                                         if (fmt == NULL || fmt == end_fmt) {
 176 #ifdef SHELL
 177                                                 INTON;
 178 #endif
 179                                                 return (fmt == NULL ? 1 : rval);
 180                                         }
 181                                         end = 0;
 182                                 }
 183                                 start = fmt;
 184                         } else
 185                                 fmt++;
 186                         if (gargv > maxargv)
 187                                 maxargv = gargv;
 188                 }
 189                 gargv = maxargv;
 190
 191                 if (end == 1) {
 192                         warnx("missing format character");
 193 #ifdef SHELL
 194                         INTON;
 195 #endif
 196                         return (1);
 197                 }
 198                 fwrite(start, 1, fmt - start, stdout);
 199                 if (!*gargv) {
 200 #ifdef SHELL
 201                         INTON;
 202 #endif
 203                         return (rval);
 204                 }
 205                 /* Restart at the beginning of the format string. */
 206                 fmt = format;
 207                 end = 1;
 208         }
 209         /* NOTREACHED */
 210 }
 211
 212
 213 static char *
 214 printf_doformat(char *fmt, int *rval)
 215 {
 216         static const char skip1[] = "#'-+ 0";
 217         int fieldwidth, haveprec, havewidth, mod_ldbl, precision;
 218         char convch, nextch;
 219         char start[strlen(fmt) + 1];
 220         char **fargv;
 221         char *dptr;
 222         int l;
 223
 224         dptr = start;
 225         *dptr++ = '%';
 226         *dptr = 0;
 227
 228         fmt++;
 229
 230         /* look for "n$" field index specifier */
 231         l = strspn(fmt, digits);
 232         if ((l > 0) && (fmt[l] == '$')) {
 233                 int idx = atoi(fmt);
 234                 if (idx <= myargc) {
 235                         gargv = &myargv[idx - 1];
 236                 } else {
 237                         gargv = &myargv[myargc];
 238                 }
 239                 if (gargv > maxargv)
 240                         maxargv = gargv;
 241                 fmt += l + 1;
 242
 243                 /* save format argument */
 244                 fargv = gargv;
 245         } else {
 246                 fargv = NULL;
 247         }
 248
 249         /* skip to field width */
 250         while (*fmt && strchr(skip1, *fmt) != NULL) {
 251                 *dptr++ = *fmt++;
 252                 *dptr = 0;
 253         }
 254
 255         if (*fmt == '*') {
 256
 257                 fmt++;
 258                 l = strspn(fmt, digits);
 259                 if ((l > 0) && (fmt[l] == '$')) {
 260                         int idx = atoi(fmt);
 261                         if (fargv == NULL) {
 262                                 warnx("incomplete use of n$");
 263                                 return (NULL);
 264                         }
 265                         if (idx <= myargc) {
 266                                 gargv = &myargv[idx - 1];
 267                         } else {
 268                                 gargv = &myargv[myargc];
 269                         }
 270                         fmt += l + 1;
 271                 } else if (fargv != NULL) {
 272                         warnx("incomplete use of n$");
 273                         return (NULL);
 274                 }
 275
 276                 if (getint(&fieldwidth))
 277                         return (NULL);
 278                 if (gargv > maxargv)
 279                         maxargv = gargv;
 280                 havewidth = 1;
 281
 282                 *dptr++ = '*';
 283                 *dptr = 0;
 284         } else {
 285                 havewidth = 0;
 286
 287                 /* skip to possible '.', get following precision */
 288                 while (isdigit(*fmt)) {
 289                         *dptr++ = *fmt++;
 290                         *dptr = 0;
 291                 }
 292         }
 293
 294         if (*fmt == '.') {
 295                 /* precision present? */
 296                 fmt++;
 297                 *dptr++ = '.';
 298
 299                 if (*fmt == '*') {
 300
 301                         fmt++;
 302                         l = strspn(fmt, digits);
 303                         if ((l > 0) && (fmt[l] == '$')) {
 304                                 int idx = atoi(fmt);
 305                                 if (fargv == NULL) {
 306                                         warnx("incomplete use of n$");
 307                                         return (NULL);
 308                                 }
 309                                 if (idx <= myargc) {
 310                                         gargv = &myargv[idx - 1];
 311                                 } else {
 312                                         gargv = &myargv[myargc];
 313                                 }
 314                                 fmt += l + 1;
 315                         } else if (fargv != NULL) {
 316                                 warnx("incomplete use of n$");
 317                                 return (NULL);
 318                         }
 319
 320                         if (getint(&precision))
 321                                 return (NULL);
 322                         if (gargv > maxargv)
 323                                 maxargv = gargv;
 324                         haveprec = 1;
 325                         *dptr++ = '*';
 326                         *dptr = 0;
 327                 } else {
 328                         haveprec = 0;
 329
 330                         /* skip to conversion char */
 331                         while (isdigit(*fmt)) {
 332                                 *dptr++ = *fmt++;
 333                                 *dptr = 0;
 334                         }
 335                 }
 336         } else
 337                 haveprec = 0;
 338         if (!*fmt) {
 339                 warnx("missing format character");
 340                 return (NULL);
 341         }
 342         *dptr++ = *fmt;
 343         *dptr = 0;
 344
 345         /*
 346          * Look for a length modifier.  POSIX doesn't have these, so
 347          * we only support them for floating-point conversions, which
 348          * are extensions.  This is useful because the L modifier can
 349          * be used to gain extra range and precision, while omitting
 350          * it is more likely to produce consistent results on different
 351          * architectures.  This is not so important for integers
 352          * because overflow is the only bad thing that can happen to
 353          * them, but consider the command  printf %a 1.1
 354          */
 355         if (*fmt == 'L') {
 356                 mod_ldbl = 1;
 357                 fmt++;
 358                 if (!strchr("aAeEfFgG", *fmt)) {
 359                         warnx("bad modifier L for %%%c", *fmt);
 360                         return (NULL);
 361                 }
 362         } else {
 363                 mod_ldbl = 0;
 364         }
 365
 366         /* save the current arg offset, and set to the format arg */
 367         if (fargv != NULL) {
 368                 gargv = fargv;
 369         }
 370
 371         convch = *fmt;
 372         nextch = *++fmt;
 373
 374         *fmt = '\0';
 375         switch (convch) {
 376         case 'b': {
 377                 size_t len;
 378                 char *p;
 379                 int getout;
 380
 381                 p = strdup(getstr());
 382                 if (p == NULL) {
 383                         warnx("%s", strerror(ENOMEM));
 384                         return (NULL);
 385                 }
 386                 getout = escape(p, 0, &len);
 387                 fputs(p, stdout);
 388                 free(p);
 389                 if (getout)
 390                         return (end_fmt);
 391                 break;
 392         }
 393         case 'c': {
 394                 char p;
 395
 396                 p = getchr();
 397                 PF(start, p);
 398                 break;
 399         }
 400         case 's': {
 401                 const char *p;
 402
 403                 p = getstr();
 404                 PF(start, p);
 405                 break;
 406         }
 407         case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': {
 408                 char *f;
 409                 intmax_t val;
 410                 uintmax_t uval;
 411                 int signedconv;
 412
 413                 signedconv = (convch == 'd' || convch == 'i');
 414                 if ((f = mknum(start, convch)) == NULL)
 415                         return (NULL);
 416                 if (getnum(&val, &uval, signedconv))
 417                         *rval = 1;
 418                 if (signedconv)
 419                         PF(f, val);
 420                 else
 421                         PF(f, uval);
 422                 break;
 423         }
 424         case 'e': case 'E':
 425         case 'f': case 'F':
 426         case 'g': case 'G':
 427         case 'a': case 'A': {
 428                 long double p;
 429
 430                 if (getfloating(&p, mod_ldbl))
 431                         *rval = 1;
 432                 if (mod_ldbl)
 433                         PF(start, p);
 434                 else
 435                         PF(start, (double)p);
 436                 break;
 437         }
 438         default:
 439                 warnx("illegal format character %c", convch);
 440                 return (NULL);
 441         }
 442         *fmt = nextch;
 443         /* return the gargv to the next element */
 444         return (fmt);
 445 }
 446
 447 static char *
 448 mknum(char *str, char ch)
 449 {
 450         static char *copy;
 451         static size_t copy_size;
 452         char *newcopy;
 453         size_t len, newlen;
 454
 455         len = strlen(str) + 2;
 456         if (len > copy_size) {
 457                 newlen = ((len + 1023) >> 10) << 10;
 458                 if ((newcopy = realloc(copy, newlen)) == NULL) {
 459                         warnx("%s", strerror(ENOMEM));
 460                         return (NULL);
 461                 }
 462                 copy = newcopy;
 463                 copy_size = newlen;
 464         }
 465
 466         memmove(copy, str, len - 3);
 467         copy[len - 3] = 'j';
 468         copy[len - 2] = ch;
 469         copy[len - 1] = '\0';
 470         return (copy);
 471 }
 472
 473 static int
 474 escape(char *fmt, int percent, size_t *len)
 475 {
 476         char *save, *store, c;
 477         int value;
 478
 479         for (save = store = fmt; ((c = *fmt) != 0); ++fmt, ++store) {
 480                 if (c != '\\') {
 481                         *store = c;
 482                         continue;
 483                 }
 484                 switch (*++fmt) {
 485                 case '\0':              /* EOS, user error */
 486                         *store = '\\';
 487                         *++store = '\0';
 488                         *len = store - save;
 489                         return (0);
 490                 case '\\':              /* backslash */
 491                 case '\'':              /* single quote */
 492                         *store = *fmt;
 493                         break;
 494                 case 'a':               /* bell/alert */
 495                         *store = '\a';
 496                         break;
 497                 case 'b':               /* backspace */
 498                         *store = '\b';
 499                         break;
 500                 case 'c':
 501                         if (!percent) {
 502                                 *store = '\0';
 503                                 *len = store - save;
 504                                 return (1);
 505                         }
 506                         *store = 'c';
 507                         break;
 508                 case 'f':               /* form-feed */
 509                         *store = '\f';
 510                         break;
 511                 case 'n':               /* newline */
 512                         *store = '\n';
 513                         break;
 514                 case 'r':               /* carriage-return */
 515                         *store = '\r';
 516                         break;
 517                 case 't':               /* horizontal tab */
 518                         *store = '\t';
 519                         break;
 520                 case 'v':               /* vertical tab */
 521                         *store = '\v';
 522                         break;
 523                                         /* octal constant */
 524                 case '0': case '1': case '2': case '3':
 525                 case '4': case '5': case '6': case '7':
 526                         c = (!percent && *fmt == '0') ? 4 : 3;
 527                         for (value = 0;
 528                             c-- && *fmt >= '0' && *fmt <= '7'; ++fmt) {
 529                                 value <<= 3;
 530                                 value += *fmt - '0';
 531                         }
 532                         --fmt;
 533                         if (percent && value == '%') {
 534                                 *store++ = '%';
 535                                 *store = '%';
 536                         } else
 537                                 *store = (char)value;
 538                         break;
 539                 default:
 540                         *store = *fmt;
 541                         break;
 542                 }
 543         }
 544         *store = '\0';
 545         *len = store - save;
 546         return (0);
 547 }
 548
 549 static int
 550 getchr(void)
 551 {
 552         if (!*gargv)
 553                 return ('\0');
 554         return ((int)**gargv++);
 555 }
 556
 557 static const char *
 558 getstr(void)
 559 {
 560         if (!*gargv)
 561                 return ("");
 562         return (*gargv++);
 563 }
 564
 565 static int
 566 getint(int *ip)
 567 {
 568         intmax_t val;
 569         uintmax_t uval;
 570         int rval;
 571
 572         if (getnum(&val, &uval, 1))
 573                 return (1);
 574         rval = 0;
 575         if (val < INT_MIN || val > INT_MAX) {
 576                 warnx("%s: %s", *gargv, strerror(ERANGE));
 577                 rval = 1;
 578         }
 579         *ip = (int)val;
 580         return (rval);
 581 }
 582
 583 static int
 584 getnum(intmax_t *ip, uintmax_t *uip, int signedconv)
 585 {
 586         char *ep;
 587         int rval;
 588
 589         if (!*gargv) {
 590                 *ip = *uip = 0;
 591                 return (0);
 592         }
 593         if (**gargv == '"' || **gargv == '\'') {
 594                 if (signedconv)
 595                         *ip = asciicode();
 596                 else
 597                         *uip = asciicode();
 598                 return (0);
 599         }
 600         rval = 0;
 601         errno = 0;
 602         if (signedconv)
 603                 *ip = strtoimax(*gargv, &ep, 0);
 604         else
 605                 *uip = strtoumax(*gargv, &ep, 0);
 606         if (ep == *gargv) {
 607                 warnx("%s: expected numeric value", *gargv);
 608                 rval = 1;
 609         }
 610         else if (*ep != '\0') {
 611                 warnx("%s: not completely converted", *gargv);
 612                 rval = 1;
 613         }
 614         if (errno == ERANGE) {
 615                 warnx("%s: %s", *gargv, strerror(ERANGE));
 616                 rval = 1;
 617         }
 618         ++gargv;
 619         return (rval);
 620 }
 621
 622 static int
 623 getfloating(long double *dp, int mod_ldbl)
 624 {
 625         char *ep;
 626         int rval;
 627
 628         if (!*gargv) {
 629                 *dp = 0.0;
 630                 return (0);
 631         }
 632         if (**gargv == '"' || **gargv == '\'') {
 633                 *dp = asciicode();
 634                 return (0);
 635         }
 636         rval = 0;
 637         errno = 0;
 638         if (mod_ldbl)
 639                 *dp = strtold(*gargv, &ep);
 640         else
 641                 *dp = strtod(*gargv, &ep);
 642         if (ep == *gargv) {
 643                 warnx("%s: expected numeric value", *gargv);
 644                 rval = 1;
 645         } else if (*ep != '\0') {
 646                 warnx("%s: not completely converted", *gargv);
 647                 rval = 1;
 648         }
 649         if (errno == ERANGE) {
 650                 warnx("%s: %s", *gargv, strerror(ERANGE));
 651                 rval = 1;
 652         }
 653         ++gargv;
 654         return (rval);
 655 }
 656
 657 static int
 658 asciicode(void)
 659 {
 660         int ch;
 661         wchar_t wch;
 662         mbstate_t mbs;
 663
 664         ch = (unsigned char)**gargv;
 665         if (ch == '\'' || ch == '"') {
 666                 memset(&mbs, 0, sizeof(mbs));
 667                 switch (mbrtowc(&wch, *gargv + 1, MB_LEN_MAX, &mbs)) {
 668                 case (size_t)-2:
 669                 case (size_t)-1:
 670                         wch = (unsigned char)gargv[0][1];
 671                         break;
 672                 case 0:
 673                         wch = 0;
 674                         break;
 675                 }
 676                 ch = wch;
 677         }
 678         ++gargv;
 679         return (ch);
 680 }
 681
 682 static void
 683 usage(void)
 684 {
 685         (void)fprintf(stderr, "usage: printf format [arguments ...]\n");
 686 }