usr.bin/printf/printf.c

   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 "options.h"
  61 #endif
  62
  63 #define PF(f, func) do {                                                \
  64         if (havewidth)                                                  \
  65                 if (haveprec)                                           \
  66                         printf(f, fieldwidth, precision, func); \
  67                 else                                                    \
  68                         printf(f, fieldwidth, func);                    \
  69         else if (haveprec)                                              \
  70                 printf(f, precision, func);                             \
  71         else                                                            \
  72                 printf(f, func);                                        \
  73 } while (0)
  74
  75 static int       asciicode(void);
  76 static char     *printf_doformat(char *, int *);
  77 static int       escape(char *, int, size_t *);
  78 static int       getchr(void);
  79 static int       getfloating(long double *, int);
  80 static int       getint(int *);
  81 static int       getnum(intmax_t *, uintmax_t *, int);
  82 static const char
  83                 *getstr(void);
  84 static char     *mknum(char *, char);
  85 static void      usage(void);
  86
  87 static const char digits[] = "0123456789";
  88
  89 static char end_fmt[1];
  90
  91 static int  myargc;
  92 static char **myargv;
  93 static char **gargv;
  94 static char **maxargv;
  95
  96 int
  97 main(int argc, char *argv[])
  98 {
  99         size_t len;
 100         int end, rval;
 101         char *format, *fmt, *start;
 102
 103 #ifndef SHELL
 104         setlocale(LC_ALL, "");
 105 #endif
 106
 107         /*
 108          * We may not use getopt(3) because calling
 109          * "printf -f%s oo" may not result in an invalid
 110          * option error.
 111          * However common usage and other implementations seem
 112          * to indicate that we need to allow -- as a discardable
 113          * option separator.
 114          */
 115         if (argc > 1 && strcmp(argv[1], "--") == 0) {
 116                 argc--;
 117                 argv++;
 118         }
 119
 120         if (argc < 2) {
 121                 usage();
 122                 return (1);
 123         }
 124
 125         argv++;
 126
 127 #ifdef SHELL
 128         INTOFF;
 129 #endif
 130         /*
 131          * Basic algorithm is to scan the format string for conversion
 132          * specifications -- once one is found, find out if the field
 133          * width or precision is a '*'; if it is, gather up value.  Note,
 134          * format strings are reused as necessary to use up the provided
 135          * arguments, arguments of zero/null string are provided to use
 136          * up the format string.
 137          */
 138         fmt = format = *argv;
 139         escape(fmt, 1, &len);           /* backslash interpretation */
 140         rval = end = 0;
 141         gargv = ++argv;
 142
 143         for (;;) {
 144                 maxargv = gargv;
 145
 146                 myargv = gargv;
 147                 for (myargc = 0; gargv[myargc]; myargc++)
 148                         /* nop */;
 149                 start = fmt;
 150                 while (fmt < format + len) {
 151                         if (fmt[0] == '%') {
 152                                 fwrite(start, 1, fmt - start, stdout);
 153                                 if (fmt[1] == '%') {
 154                                         /* %% prints a % */
 155                                         putchar('%');
 156                                         fmt += 2;
 157                                 } else {
 158                                         fmt = printf_doformat(fmt, &rval);
 159                                         if (fmt == NULL || fmt == end_fmt) {
 160 #ifdef SHELL
 161                                                 INTON;
 162 #endif
 163                                                 return (fmt == NULL ? 1 : rval);
 164                                         }
 165                                         end = 0;
 166                                 }
 167                                 start = fmt;
 168                         } else
 169                                 fmt++;
 170                         if (gargv > maxargv)
 171                                 maxargv = gargv;
 172                 }
 173                 gargv = maxargv;
 174
 175                 if (end == 1) {
 176                         warnx("missing format character");
 177 #ifdef SHELL
 178                         INTON;
 179 #endif
 180                         return (1);
 181                 }
 182                 fwrite(start, 1, fmt - start, stdout);
 183                 if (!*gargv) {
 184 #ifdef SHELL
 185                         INTON;
 186 #endif
 187                         return (rval);
 188                 }
 189                 /* Restart at the beginning of the format string. */
 190                 fmt = format;
 191                 end = 1;
 192         }
 193         /* NOTREACHED */
 194 }
 195
 196
 197 static char *
 198 printf_doformat(char *fmt, int *rval)
 199 {
 200         static const char skip1[] = "#'-+ 0";
 201         int fieldwidth, haveprec, havewidth, mod_ldbl, precision;
 202         char convch, nextch;
 203         char start[strlen(fmt) + 1];
 204         char **fargv;
 205         char *dptr;
 206         int l;
 207
 208         dptr = start;
 209         *dptr++ = '%';
 210         *dptr = 0;
 211
 212         fmt++;
 213
 214         /* look for "n$" field index specifier */
 215         l = strspn(fmt, digits);
 216         if ((l > 0) && (fmt[l] == '$')) {
 217                 int idx = atoi(fmt);
 218                 if (idx <= myargc) {
 219                         gargv = &myargv[idx - 1];
 220                 } else {
 221                         gargv = &myargv[myargc];
 222                 }
 223                 if (gargv > maxargv)
 224                         maxargv = gargv;
 225                 fmt += l + 1;
 226
 227                 /* save format argument */
 228                 fargv = gargv;
 229         } else {
 230                 fargv = NULL;
 231         }
 232
 233         /* skip to field width */
 234         while (*fmt && strchr(skip1, *fmt) != NULL) {
 235                 *dptr++ = *fmt++;
 236                 *dptr = 0;
 237         }
 238
 239         if (*fmt == '*') {
 240
 241                 fmt++;
 242                 l = strspn(fmt, digits);
 243                 if ((l > 0) && (fmt[l] == '$')) {
 244                         int idx = atoi(fmt);
 245                         if (fargv == NULL) {
 246                                 warnx("incomplete use of n$");
 247                                 return (NULL);
 248                         }
 249                         if (idx <= myargc) {
 250                                 gargv = &myargv[idx - 1];
 251                         } else {
 252                                 gargv = &myargv[myargc];
 253                         }
 254                         fmt += l + 1;
 255                 } else if (fargv != NULL) {
 256                         warnx("incomplete use of n$");
 257                         return (NULL);
 258                 }
 259
 260                 if (getint(&fieldwidth))
 261                         return (NULL);
 262                 if (gargv > maxargv)
 263                         maxargv = gargv;
 264                 havewidth = 1;
 265
 266                 *dptr++ = '*';
 267                 *dptr = 0;
 268         } else {
 269                 havewidth = 0;
 270
 271                 /* skip to possible '.', get following precision */
 272                 while (isdigit(*fmt)) {
 273                         *dptr++ = *fmt++;
 274                         *dptr = 0;
 275                 }
 276         }
 277
 278         if (*fmt == '.') {
 279                 /* precision present? */
 280                 fmt++;
 281                 *dptr++ = '.';
 282
 283                 if (*fmt == '*') {
 284
 285                         fmt++;
 286                         l = strspn(fmt, digits);
 287                         if ((l > 0) && (fmt[l] == '$')) {
 288                                 int idx = atoi(fmt);
 289                                 if (fargv == NULL) {
 290                                         warnx("incomplete use of n$");
 291                                         return (NULL);
 292                                 }
 293                                 if (idx <= myargc) {
 294                                         gargv = &myargv[idx - 1];
 295                                 } else {
 296                                         gargv = &myargv[myargc];
 297                                 }
 298                                 fmt += l + 1;
 299                         } else if (fargv != NULL) {
 300                                 warnx("incomplete use of n$");
 301                                 return (NULL);
 302                         }
 303
 304                         if (getint(&precision))
 305                                 return (NULL);
 306                         if (gargv > maxargv)
 307                                 maxargv = gargv;
 308                         haveprec = 1;
 309                         *dptr++ = '*';
 310                         *dptr = 0;
 311                 } else {
 312                         haveprec = 0;
 313
 314                         /* skip to conversion char */
 315                         while (isdigit(*fmt)) {
 316                                 *dptr++ = *fmt++;
 317                                 *dptr = 0;
 318                         }
 319                 }
 320         } else
 321                 haveprec = 0;
 322         if (!*fmt) {
 323                 warnx("missing format character");
 324                 return (NULL);
 325         }
 326         *dptr++ = *fmt;
 327         *dptr = 0;
 328
 329         /*
 330          * Look for a length modifier.  POSIX doesn't have these, so
 331          * we only support them for floating-point conversions, which
 332          * are extensions.  This is useful because the L modifier can
 333          * be used to gain extra range and precision, while omitting
 334          * it is more likely to produce consistent results on different
 335          * architectures.  This is not so important for integers
 336          * because overflow is the only bad thing that can happen to
 337          * them, but consider the command  printf %a 1.1
 338          */
 339         if (*fmt == 'L') {
 340                 mod_ldbl = 1;
 341                 fmt++;
 342                 if (!strchr("aAeEfFgG", *fmt)) {
 343                         warnx("bad modifier L for %%%c", *fmt);
 344                         return (NULL);
 345                 }
 346         } else {
 347                 mod_ldbl = 0;
 348         }
 349
 350         /* save the current arg offset, and set to the format arg */
 351         if (fargv != NULL) {
 352                 gargv = fargv;
 353         }
 354
 355         convch = *fmt;
 356         nextch = *++fmt;
 357
 358         *fmt = '\0';
 359         switch (convch) {
 360         case 'b': {
 361                 size_t len;
 362                 char *p;
 363                 int getout;
 364
 365                 /* Convert "b" to "s" for output. */
 366                 start[strlen(start) - 1] = 's';
 367                 if ((p = strdup(getstr())) == NULL) {
 368                         warnx("%s", strerror(ENOMEM));
 369                         return (NULL);
 370                 }
 371                 getout = escape(p, 0, &len);
 372                 PF(start, p);
 373                 /* Restore format for next loop. */
 374
 375                 free(p);
 376                 if (getout)
 377                         return (end_fmt);
 378                 break;
 379         }
 380         case 'c': {
 381                 char p;
 382
 383                 p = getchr();
 384                 if (p != '\0')
 385                         PF(start, p);
 386                 break;
 387         }
 388         case 's': {
 389                 const char *p;
 390
 391                 p = getstr();
 392                 PF(start, p);
 393                 break;
 394         }
 395         case 'd': case 'i': case 'o': case 'u': case 'x': case 'X': {
 396                 char *f;
 397                 intmax_t val;
 398                 uintmax_t uval;
 399                 int signedconv;
 400
 401                 signedconv = (convch == 'd' || convch == 'i');
 402                 if ((f = mknum(start, convch)) == NULL)
 403                         return (NULL);
 404                 if (getnum(&val, &uval, signedconv))
 405                         *rval = 1;
 406                 if (signedconv)
 407                         PF(f, val);
 408                 else
 409                         PF(f, uval);
 410                 break;
 411         }
 412         case 'e': case 'E':
 413         case 'f': case 'F':
 414         case 'g': case 'G':
 415         case 'a': case 'A': {
 416                 long double p;
 417
 418                 if (getfloating(&p, mod_ldbl))
 419                         *rval = 1;
 420                 if (mod_ldbl)
 421                         PF(start, p);
 422                 else
 423                         PF(start, (double)p);
 424                 break;
 425         }
 426         default:
 427                 warnx("illegal format character %c", convch);
 428                 return (NULL);
 429         }
 430         *fmt = nextch;
 431         /* return the gargv to the next element */
 432         return (fmt);
 433 }
 434
 435 static char *
 436 mknum(char *str, char ch)
 437 {
 438         static char *copy;
 439         static size_t copy_size;
 440         char *newcopy;
 441         size_t len, newlen;
 442
 443         len = strlen(str) + 2;
 444         if (len > copy_size) {
 445                 newlen = ((len + 1023) >> 10) << 10;
 446                 if ((newcopy = realloc(copy, newlen)) == NULL) {
 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         fprintf(stderr, "usage: printf format [arguments ...]\n");
 674 }