usr.bin/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  * 3. 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  * @(#) Copyright (c) 1989, 1993 The Regents of the University of California.  All rights reserved.
  32  * @(#)printf.c 8.1 (Berkeley) 7/20/93
  33  * $FreeBSD: head/usr.bin/printf/printf.c 279503 2015-03-01 21:46:55Z jilles $
  34  */
  35 /*
  36  * Important: This file is used both as a standalone program /usr/bin/printf
  37  * and as a builtin for /bin/sh (#define SHELL).
  38  */
  39
  40 #include <sys/types.h>
  41
  42 #include <ctype.h>
  43 #include <err.h>
  44 #include <errno.h>
  45 #include <inttypes.h>
  46 #include <limits.h>
  47 #include <locale.h>
  48 #include <stdio.h>
  49 #include <stdlib.h>
  50 #include <string.h>
  51 #include <unistd.h>
  52 #include <wchar.h>
  53
  54 #ifdef SHELL
  55 #define main printfcmd
  56 #include "bltin/bltin.h"
  57 #include "error.h"
  58 #endif
  59
  60 #define PF(f, func) do {                                                \
  61         char *b = NULL;                                                 \
  62         if (havewidth)                                                  \
  63                 if (haveprec)                                           \
  64                         asprintf(&b, f, fieldwidth, precision, func);   \
  65                 else                                                    \
  66                         asprintf(&b, f, fieldwidth, func);              \
  67         else if (haveprec)                                              \
  68                 asprintf(&b, f, precision, func);                       \
  69         else                                                            \
  70                 asprintf(&b, f, func);                                  \
  71         if (b) {                                                        \
  72                 fputs(b, stdout);                                       \
  73                 free(b);                                                \
  74         }                                                               \
  75 } while (0)
  76
  77 static int       asciicode(void);
  78 static char     *printf_doformat(char *, int *);
  79 static int       escape(char *, int, size_t *);
  80 static int       getchr(void);
  81 static int       getfloating(long double *, int);
  82 static int       getint(int *);
  83 static int       getnum(intmax_t *, uintmax_t *, int);
  84 static const char
  85                 *getstr(void);
  86 static char     *mknum(char *, char);
  87 static void      usage(void);
  88
  89 static const char digits[] = "0123456789";
  90
  91 static char end_fmt[1];
  92
  93 static int  myargc;
  94 static char **myargv;
  95 static char **gargv;
  96 static char **maxargv;
  97
  98 int
  99 main(int argc, char *argv[])
 100 {
 101         size_t len;
 102         int end, rval;
 103         char *format, *fmt, *start;
 104
 105 #ifndef SHELL
 106         setlocale(LC_ALL, "");
 107 #endif
 108
 109         /*
 110          * We may not use getopt(3) because calling
 111          * "printf -f%s oo" may not result in an invalid
 112          * option error.
 113          * However common usage and other implementations seem
 114          * to indicate that we need to allow -- as a discardable
 115          * option separator.
 116          */
 117         if (argc > 1 && strcmp(argv[1], "--") == 0) {
 118                 argc--;
 119                 argv++;
 120         }
 121
 122         if (argc < 2) {
 123                 usage();
 124                 return (1);
 125         }
 126
 127         argv++;
 128
 129 #ifdef SHELL
 130         INTOFF;
 131 #endif
 132         /*
 133          * Basic algorithm is to scan the format string for conversion
 134          * specifications -- once one is found, find out if the field
 135          * width or precision is a '*'; if it is, gather up value.  Note,
 136          * format strings are reused as necessary to use up the provided
 137          * arguments, arguments of zero/null string are provided to use
 138          * up the format string.
 139          */
 140         fmt = format = *argv;
 141         escape(fmt, 1, &len);           /* backslash interpretation */
 142         rval = end = 0;
 143         gargv = ++argv;
 144
 145         for (;;) {
 146                 maxargv = gargv;
 147
 148                 myargv = gargv;
 149                 for (myargc = 0; gargv[myargc]; myargc++)
 150                         /* nop */;
 151                 start = fmt;
 152                 while (fmt < format + len) {
 153                         if (fmt[0] == '%') {
 154                                 fwrite(start, 1, fmt - start, stdout);
 155                                 if (fmt[1] == '%') {
 156                                         /* %% prints a % */
 157                                         putchar('%');
 158                                         fmt += 2;
 159                                 } else {
 160                                         fmt = printf_doformat(fmt, &rval);
 161                                         if (fmt == NULL || fmt == end_fmt) {
 162 #ifdef SHELL
 163                                                 INTON;
 164 #endif
 165                                                 return (fmt == NULL ? 1 : rval);
 166                                         }
 167                                         end = 0;
 168                                 }
 169                                 start = fmt;
 170                         } else
 171                                 fmt++;
 172                         if (gargv > maxargv)
 173                                 maxargv = gargv;
 174                 }
 175                 gargv = maxargv;
 176
 177                 if (end == 1) {
 178                         warnx("missing format character");
 179 #ifdef SHELL
 180                         INTON;
 181 #endif
 182                         return (1);
 183                 }
 184                 fwrite(start, 1, fmt - start, stdout);
 185                 if (!*gargv) {
 186 #ifdef SHELL
 187                         INTON;
 188 #endif
 189                         return (rval);
 190                 }
 191                 /* Restart at the beginning of the format string. */
 192                 fmt = format;
 193                 end = 1;
 194         }
 195         /* NOTREACHED */
 196 }
 197
 198
 199 static char *
 200 printf_doformat(char *fmt, int *rval)
 201 {
 202         static const char skip1[] = "#'-+ 0";
 203         int fieldwidth, haveprec, havewidth, mod_ldbl, precision;
 204         char convch, nextch;
 205         char start[strlen(fmt) + 1];
 206         char **fargv;
 207         char *dptr;
 208         int l;
 209
 210         dptr = start;
 211         *dptr++ = '%';
 212         *dptr = 0;
 213
 214         fmt++;
 215
 216         /* look for "n$" field index specifier */
 217         l = strspn(fmt, digits);
 218         if ((l > 0) && (fmt[l] == '$')) {
 219                 int idx = atoi(fmt);
 220                 if (idx <= myargc) {
 221                         gargv = &myargv[idx - 1];
 222                 } else {
 223                         gargv = &myargv[myargc];
 224                 }
 225                 if (gargv > maxargv)
 226                         maxargv = gargv;
 227                 fmt += l + 1;
 228
 229                 /* save format argument */
 230                 fargv = gargv;
 231         } else {
 232                 fargv = NULL;
 233         }
 234
 235         /* skip to field width */
 236         while (*fmt && strchr(skip1, *fmt) != NULL) {
 237                 *dptr++ = *fmt++;
 238                 *dptr = 0;
 239         }
 240
 241         if (*fmt == '*') {
 242
 243                 fmt++;
 244                 l = strspn(fmt, digits);
 245                 if ((l > 0) && (fmt[l] == '$')) {
 246                         int idx = atoi(fmt);
 247                         if (fargv == NULL) {
 248                                 warnx("incomplete use of n$");
 249                                 return (NULL);
 250                         }
 251                         if (idx <= myargc) {
 252                                 gargv = &myargv[idx - 1];
 253                         } else {
 254                                 gargv = &myargv[myargc];
 255                         }
 256                         fmt += l + 1;
 257                 } else if (fargv != NULL) {
 258                         warnx("incomplete use of n$");
 259                         return (NULL);
 260                 }
 261
 262                 if (getint(&fieldwidth))
 263                         return (NULL);
 264                 if (gargv > maxargv)
 265                         maxargv = gargv;
 266                 havewidth = 1;
 267
 268                 *dptr++ = '*';
 269                 *dptr = 0;
 270         } else {
 271                 havewidth = 0;
 272
 273                 /* skip to possible '.', get following precision */
 274                 while (isdigit(*fmt)) {
 275                         *dptr++ = *fmt++;
 276                         *dptr = 0;
 277                 }
 278         }
 279
 280         if (*fmt == '.') {
 281                 /* precision present? */
 282                 fmt++;
 283                 *dptr++ = '.';
 284
 285                 if (*fmt == '*') {
 286
 287                         fmt++;
 288                         l = strspn(fmt, digits);
 289                         if ((l > 0) && (fmt[l] == '$')) {
 290                                 int idx = atoi(fmt);
 291                                 if (fargv == NULL) {
 292                                         warnx("incomplete use of n$");
 293                                         return (NULL);
 294                                 }
 295                                 if (idx <= myargc) {
 296                                         gargv = &myargv[idx - 1];
 297                                 } else {
 298                                         gargv = &myargv[myargc];
 299                                 }
 300                                 fmt += l + 1;
 301                         } else if (fargv != NULL) {
 302                                 warnx("incomplete use of n$");
 303                                 return (NULL);
 304                         }
 305
 306                         if (getint(&precision))
 307                                 return (NULL);
 308                         if (gargv > maxargv)
 309                                 maxargv = gargv;
 310                         haveprec = 1;
 311                         *dptr++ = '*';
 312                         *dptr = 0;
 313                 } else {
 314                         haveprec = 0;
 315
 316                         /* skip to conversion char */
 317                         while (isdigit(*fmt)) {
 318                                 *dptr++ = *fmt++;
 319                                 *dptr = 0;
 320                         }
 321                 }
 322         } else
 323                 haveprec = 0;
 324         if (!*fmt) {
 325                 warnx("missing format character");
 326                 return (NULL);
 327         }
 328         *dptr++ = *fmt;
 329         *dptr = 0;
 330
 331         /*
 332          * Look for a length modifier.  POSIX doesn't have these, so
 333          * we only support them for floating-point conversions, which
 334          * are extensions.  This is useful because the L modifier can
 335          * be used to gain extra range and precision, while omitting
 336          * it is more likely to produce consistent results on different
 337          * architectures.  This is not so important for integers
 338          * because overflow is the only bad thing that can happen to
 339          * them, but consider the command  printf %a 1.1
 340          */
 341         if (*fmt == 'L') {
 342                 mod_ldbl = 1;
 343                 fmt++;
 344                 if (!strchr("aAeEfFgG", *fmt)) {
 345                         warnx("bad modifier L for %%%c", *fmt);
 346                         return (NULL);
 347                 }
 348         } else {
 349                 mod_ldbl = 0;
 350         }
 351
 352         /* save the current arg offset, and set to the format arg */
 353         if (fargv != NULL) {
 354                 gargv = fargv;
 355         }
 356
 357         convch = *fmt;
 358         nextch = *++fmt;
 359
 360         *fmt = '\0';
 361         switch (convch) {
 362         case 'b': {
 363                 size_t len;
 364                 char *p;
 365                 int getout;
 366
 367                 p = strdup(getstr());
 368                 if (p == NULL) {
 369                         warnx("%s", strerror(ENOMEM));
 370                         return (NULL);
 371                 }
 372                 getout = escape(p, 0, &len);
 373                 fputs(p, stdout);
 374                 free(p);
 375                 if (getout)
 376                         return (end_fmt);
 377                 break;
 378         }
 379         case 'c': {
 380                 char p;
 381
 382                 p = getchr();
 383                 PF(start, p);
 384                 break;
 385         }
 386         case 's': {
 387                 const char *p;
 388
 389                 p = getstr();
 390                 PF(start, p);
 391                 break;
 392         }
 393         case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': {
 394                 char *f;
 395                 intmax_t val;
 396                 uintmax_t uval;
 397                 int signedconv;
 398
 399                 signedconv = (convch == 'd' || convch == 'i');
 400                 if ((f = mknum(start, convch)) == NULL)
 401                         return (NULL);
 402                 if (getnum(&val, &uval, signedconv))
 403                         *rval = 1;
 404                 if (signedconv)
 405                         PF(f, val);
 406                 else
 407                         PF(f, uval);
 408                 break;
 409         }
 410         case 'e': case 'E':
 411         case 'f': case 'F':
 412         case 'g': case 'G':
 413         case 'a': case 'A': {
 414                 long double p;
 415
 416                 if (getfloating(&p, mod_ldbl))
 417                         *rval = 1;
 418                 if (mod_ldbl)
 419                         PF(start, p);
 420                 else
 421                         PF(start, (double)p);
 422                 break;
 423         }
 424         default:
 425                 warnx("illegal format character %c", convch);
 426                 return (NULL);
 427         }
 428         *fmt = nextch;
 429         /* return the gargv to the next element */
 430         return (fmt);
 431 }
 432
 433 static char *
 434 mknum(char *str, char ch)
 435 {
 436         static char *copy;
 437         static size_t copy_size;
 438         char *newcopy;
 439         size_t len, newlen;
 440
 441         len = strlen(str) + 2;
 442         if (len > copy_size) {
 443                 newlen = ((len + 1023) >> 10) << 10;
 444                 if ((newcopy = realloc(copy, newlen)) == NULL) {
 445                         warnx("%s", strerror(ENOMEM));
 446                         return (NULL);
 447                 }
 448                 copy = newcopy;
 449                 copy_size = newlen;
 450         }
 451
 452         memmove(copy, str, len - 3);
 453         copy[len - 3] = 'j';
 454         copy[len - 2] = ch;
 455         copy[len - 1] = '\0';
 456         return (copy);
 457 }
 458
 459 static int
 460 escape(char *fmt, int percent, size_t *len)
 461 {
 462         char *save, *store, c;
 463         int value;
 464
 465         for (save = store = fmt; ((c = *fmt) != 0); ++fmt, ++store) {
 466                 if (c != '\\') {
 467                         *store = c;
 468                         continue;
 469                 }
 470                 switch (*++fmt) {
 471                 case '\0':              /* EOS, user error */
 472                         *store = '\\';
 473                         *++store = '\0';
 474                         *len = store - save;
 475                         return (0);
 476                 case '\\':              /* backslash */
 477                 case '\'':              /* single quote */
 478                         *store = *fmt;
 479                         break;
 480                 case 'a':               /* bell/alert */
 481                         *store = '\a';
 482                         break;
 483                 case 'b':               /* backspace */
 484                         *store = '\b';
 485                         break;
 486                 case 'c':
 487                         if (!percent) {
 488                                 *store = '\0';
 489                                 *len = store - save;
 490                                 return (1);
 491                         }
 492                         *store = 'c';
 493                         break;
 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 = *uip = 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         fprintf(stderr, "usage: printf format [arguments ...]\n");
 672 }