src/printf.c

   1 /* printf - format and print data
   2    Copyright (C) 1990-2012 Free Software Foundation, Inc.
   3
   4    This program is free software: you can redistribute it and/or modify
   5    it under the terms of the GNU General Public License as published by
   6    the Free Software Foundation, either version 3 of the License, or
   7    (at your option) any later version.
   8
   9    This program is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12    GNU General Public License for more details.
  13
  14    You should have received a copy of the GNU General Public License
  15    along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
  16
  17 /* Usage: printf format [argument...]
  18
  19    A front end to the printf function that lets it be used from the shell.
  20
  21    Backslash escapes:
  22
  23    \" = double quote
  24    \\ = backslash
  25    \a = alert (bell)
  26    \b = backspace
  27    \c = produce no further output
  28    \e = escape
  29    \f = form feed
  30    \n = new line
  31    \r = carriage return
  32    \t = horizontal tab
  33    \v = vertical tab
  34    \ooo = octal number (ooo is 1 to 3 digits)
  35    \xhh = hexadecimal number (hhh is 1 to 2 digits)
  36    \uhhhh = 16-bit Unicode character (hhhh is 4 digits)
  37    \Uhhhhhhhh = 32-bit Unicode character (hhhhhhhh is 8 digits)
  38
  39    Additional directive:
  40
  41    %b = print an argument string, interpreting backslash escapes,
  42      except that octal escapes are of the form \0 or \0ooo.
  43
  44    The 'format' argument is re-used as many times as necessary
  45    to convert all of the given arguments.
  46
  47    David MacKenzie <djm@gnu.ai.mit.edu> */
  48
  49 #include <config.h>
  50 #include <stdio.h>
  51 #include <sys/types.h>
  52
  53 #include "system.h"
  54 #include "c-strtod.h"
  55 #include "error.h"
  56 #include "quote.h"
  57 #include "unicodeio.h"
  58 #include "xprintf.h"
  59
  60 /* The official name of this program (e.g., no 'g' prefix).  */
  61 #define PROGRAM_NAME "printf"
  62
  63 #define AUTHORS proper_name ("David MacKenzie")
  64
  65 #define isodigit(c) ((c) >= '0' && (c) <= '7')
  66 #define hextobin(c) ((c) >= 'a' && (c) <= 'f' ? (c) - 'a' + 10 : \
  67                      (c) >= 'A' && (c) <= 'F' ? (c) - 'A' + 10 : (c) - '0')
  68 #define octtobin(c) ((c) - '0')
  69
  70 /* The value to return to the calling program.  */
  71 static int exit_status;
  72
  73 /* True if the POSIXLY_CORRECT environment variable is set.  */
  74 static bool posixly_correct;
  75
  76 /* This message appears in N_() here rather than just in _() below because
  77    the sole use would have been in a #define.  */
  78 static char const *const cfcc_msg =
  79  N_("warning: %s: character(s) following character constant have been ignored");
  80
  81 void
  82 usage (int status)
  83 {
  84   if (status != EXIT_SUCCESS)
  85     emit_try_help ();
  86   else
  87     {
  88       printf (_("\
  89 Usage: %s FORMAT [ARGUMENT]...\n\
  90   or:  %s OPTION\n\
  91 "),
  92               program_name, program_name);
  93       fputs (_("\
  94 Print ARGUMENT(s) according to FORMAT, or execute according to OPTION:\n\
  95 \n\
  96 "), stdout);
  97       fputs (HELP_OPTION_DESCRIPTION, stdout);
  98       fputs (VERSION_OPTION_DESCRIPTION, stdout);
  99       fputs (_("\
 100 \n\
 101 FORMAT controls the output as in C printf.  Interpreted sequences are:\n\
 102 \n\
 103   \\\"      double quote\n\
 104 "), stdout);
 105       fputs (_("\
 106   \\\\      backslash\n\
 107   \\a      alert (BEL)\n\
 108   \\b      backspace\n\
 109   \\c      produce no further output\n\
 110   \\e      escape\n\
 111   \\f      form feed\n\
 112   \\n      new line\n\
 113   \\r      carriage return\n\
 114   \\t      horizontal tab\n\
 115   \\v      vertical tab\n\
 116 "), stdout);
 117       fputs (_("\
 118   \\NNN    byte with octal value NNN (1 to 3 digits)\n\
 119   \\xHH    byte with hexadecimal value HH (1 to 2 digits)\n\
 120   \\uHHHH  Unicode (ISO/IEC 10646) character with hex value HHHH (4 digits)\n\
 121   \\UHHHHHHHH  Unicode character with hex value HHHHHHHH (8 digits)\n\
 122 "), stdout);
 123       fputs (_("\
 124   %%      a single %\n\
 125   %b      ARGUMENT as a string with '\\' escapes interpreted,\n\
 126           except that octal escapes are of the form \\0 or \\0NNN\n\
 127 \n\
 128 and all C format specifications ending with one of diouxXfeEgGcs, with\n\
 129 ARGUMENTs converted to proper type first.  Variable widths are handled.\n\
 130 "), stdout);
 131       printf (USAGE_BUILTIN_WARNING, PROGRAM_NAME);
 132       emit_ancillary_info ();
 133     }
 134   exit (status);
 135 }
 136
 137 static void
 138 verify_numeric (const char *s, const char *end)
 139 {
 140   if (errno)
 141     {
 142       error (0, errno, "%s", s);
 143       exit_status = EXIT_FAILURE;
 144     }
 145   else if (*end)
 146     {
 147       if (s == end)
 148         error (0, 0, _("%s: expected a numeric value"), s);
 149       else
 150         error (0, 0, _("%s: value not completely converted"), s);
 151       exit_status = EXIT_FAILURE;
 152     }
 153 }
 154
 155 #define STRTOX(TYPE, FUNC_NAME, LIB_FUNC_EXPR)                           \
 156 static TYPE                                                              \
 157 FUNC_NAME (char const *s)                                                \
 158 {                                                                        \
 159   char *end;                                                             \
 160   TYPE val;                                                              \
 161                                                                          \
 162   if ((*s == '\"' || *s == '\'') && *(s + 1))                            \
 163     {                                                                    \
 164       unsigned char ch = *++s;                                           \
 165       val = ch;                                                          \
 166       /* If POSIXLY_CORRECT is not set, then give a warning that there   \
 167          are characters following the character constant and that GNU    \
 168          printf is ignoring those characters.  If POSIXLY_CORRECT *is*   \
 169          set, then don't give the warning.  */                           \
 170       if (*++s != 0 && !posixly_correct)                                 \
 171         error (0, 0, _(cfcc_msg), s);                                    \
 172     }                                                                    \
 173   else                                                                   \
 174     {                                                                    \
 175       errno = 0;                                                         \
 176       val = (LIB_FUNC_EXPR);                                             \
 177       verify_numeric (s, end);                                           \
 178     }                                                                    \
 179   return val;                                                            \
 180 }                                                                        \
 181
 182 STRTOX (intmax_t,    vstrtoimax, strtoimax (s, &end, 0))
 183 STRTOX (uintmax_t,   vstrtoumax, strtoumax (s, &end, 0))
 184 STRTOX (long double, vstrtold,   c_strtold (s, &end))
 185
 186 /* Output a single-character \ escape.  */
 187
 188 static void
 189 print_esc_char (char c)
 190 {
 191   switch (c)
 192     {
 193     case 'a':                   /* Alert. */
 194       putchar ('\a');
 195       break;
 196     case 'b':                   /* Backspace. */
 197       putchar ('\b');
 198       break;
 199     case 'c':                   /* Cancel the rest of the output. */
 200       exit (EXIT_SUCCESS);
 201       break;
 202     case 'e':                   /* Escape. */
 203       putchar ('\x1B');
 204       break;
 205     case 'f':                   /* Form feed. */
 206       putchar ('\f');
 207       break;
 208     case 'n':                   /* New line. */
 209       putchar ('\n');
 210       break;
 211     case 'r':                   /* Carriage return. */
 212       putchar ('\r');
 213       break;
 214     case 't':                   /* Horizontal tab. */
 215       putchar ('\t');
 216       break;
 217     case 'v':                   /* Vertical tab. */
 218       putchar ('\v');
 219       break;
 220     default:
 221       putchar (c);
 222       break;
 223     }
 224 }
 225
 226 /* Print a \ escape sequence starting at ESCSTART.
 227    Return the number of characters in the escape sequence
 228    besides the backslash.
 229    If OCTAL_0 is nonzero, octal escapes are of the form \0ooo, where o
 230    is an octal digit; otherwise they are of the form \ooo.  */
 231
 232 static int
 233 print_esc (const char *escstart, bool octal_0)
 234 {
 235   const char *p = escstart + 1;
 236   int esc_value = 0;            /* Value of \nnn escape. */
 237   int esc_length;               /* Length of \nnn escape. */
 238
 239   if (*p == 'x')
 240     {
 241       /* A hexadecimal \xhh escape sequence must have 1 or 2 hex. digits.  */
 242       for (esc_length = 0, ++p;
 243            esc_length < 2 && isxdigit (to_uchar (*p));
 244            ++esc_length, ++p)
 245         esc_value = esc_value * 16 + hextobin (*p);
 246       if (esc_length == 0)
 247         error (EXIT_FAILURE, 0, _("missing hexadecimal number in escape"));
 248       putchar (esc_value);
 249     }
 250   else if (isodigit (*p))
 251     {
 252       /* Parse \0ooo (if octal_0 && *p == '0') or \ooo (otherwise).
 253          Allow \ooo if octal_0 && *p != '0'; this is an undocumented
 254          extension to POSIX that is compatible with Bash 2.05b.  */
 255       for (esc_length = 0, p += octal_0 && *p == '0';
 256            esc_length < 3 && isodigit (*p);
 257            ++esc_length, ++p)
 258         esc_value = esc_value * 8 + octtobin (*p);
 259       putchar (esc_value);
 260     }
 261   else if (*p && strchr ("\"\\abcefnrtv", *p))
 262     print_esc_char (*p++);
 263   else if (*p == 'u' || *p == 'U')
 264     {
 265       char esc_char = *p;
 266       unsigned int uni_value;
 267
 268       uni_value = 0;
 269       for (esc_length = (esc_char == 'u' ? 4 : 8), ++p;
 270            esc_length > 0;
 271            --esc_length, ++p)
 272         {
 273           if (! isxdigit (to_uchar (*p)))
 274             error (EXIT_FAILURE, 0, _("missing hexadecimal number in escape"));
 275           uni_value = uni_value * 16 + hextobin (*p);
 276         }
 277
 278       /* A universal character name shall not specify a character short
 279          identifier in the range 00000000 through 00000020, 0000007F through
 280          0000009F, or 0000D800 through 0000DFFF inclusive. A universal
 281          character name shall not designate a character in the required
 282          character set.  */
 283       if ((uni_value <= 0x9f
 284            && uni_value != 0x24 && uni_value != 0x40 && uni_value != 0x60)
 285           || (uni_value >= 0xd800 && uni_value <= 0xdfff))
 286         error (EXIT_FAILURE, 0, _("invalid universal character name \\%c%0*x"),
 287                esc_char, (esc_char == 'u' ? 4 : 8), uni_value);
 288
 289       print_unicode_char (stdout, uni_value, 0);
 290     }
 291   else
 292     {
 293       putchar ('\\');
 294       if (*p)
 295         {
 296           putchar (*p);
 297           p++;
 298         }
 299     }
 300   return p - escstart - 1;
 301 }
 302
 303 /* Print string STR, evaluating \ escapes. */
 304
 305 static void
 306 print_esc_string (const char *str)
 307 {
 308   for (; *str; str++)
 309     if (*str == '\\')
 310       str += print_esc (str, true);
 311     else
 312       putchar (*str);
 313 }
 314
 315 /* Evaluate a printf conversion specification.  START is the start of
 316    the directive, LENGTH is its length, and CONVERSION specifies the
 317    type of conversion.  LENGTH does not include any length modifier or
 318    the conversion specifier itself.  FIELD_WIDTH and PRECISION are the
 319    field width and precision for '*' values, if HAVE_FIELD_WIDTH and
 320    HAVE_PRECISION are true, respectively.  ARGUMENT is the argument to
 321    be formatted.  */
 322
 323 static void
 324 print_direc (const char *start, size_t length, char conversion,
 325              bool have_field_width, int field_width,
 326              bool have_precision, int precision,
 327              char const *argument)
 328 {
 329   char *p;              /* Null-terminated copy of % directive. */
 330
 331   /* Create a null-terminated copy of the % directive, with an
 332      intmax_t-wide length modifier substituted for any existing
 333      integer length modifier.  */
 334   {
 335     char *q;
 336     char const *length_modifier;
 337     size_t length_modifier_len;
 338
 339     switch (conversion)
 340       {
 341       case 'd': case 'i': case 'o': case 'u': case 'x': case 'X':
 342         length_modifier = PRIdMAX;
 343         length_modifier_len = sizeof PRIdMAX - 2;
 344         break;
 345
 346       case 'a': case 'e': case 'f': case 'g':
 347       case 'A': case 'E': case 'F': case 'G':
 348         length_modifier = "L";
 349         length_modifier_len = 1;
 350         break;
 351
 352       default:
 353         length_modifier = start;  /* Any valid pointer will do.  */
 354         length_modifier_len = 0;
 355         break;
 356       }
 357
 358     p = xmalloc (length + length_modifier_len + 2);
 359     q = mempcpy (p, start, length);
 360     q = mempcpy (q, length_modifier, length_modifier_len);
 361     *q++ = conversion;
 362     *q = '\0';
 363   }
 364
 365   switch (conversion)
 366     {
 367     case 'd':
 368     case 'i':
 369       {
 370         intmax_t arg = vstrtoimax (argument);
 371         if (!have_field_width)
 372           {
 373             if (!have_precision)
 374               xprintf (p, arg);
 375             else
 376               xprintf (p, precision, arg);
 377           }
 378         else
 379           {
 380             if (!have_precision)
 381               xprintf (p, field_width, arg);
 382             else
 383               xprintf (p, field_width, precision, arg);
 384           }
 385       }
 386       break;
 387
 388     case 'o':
 389     case 'u':
 390     case 'x':
 391     case 'X':
 392       {
 393         uintmax_t arg = vstrtoumax (argument);
 394         if (!have_field_width)
 395           {
 396             if (!have_precision)
 397               xprintf (p, arg);
 398             else
 399               xprintf (p, precision, arg);
 400           }
 401         else
 402           {
 403             if (!have_precision)
 404               xprintf (p, field_width, arg);
 405             else
 406               xprintf (p, field_width, precision, arg);
 407           }
 408       }
 409       break;
 410
 411     case 'a':
 412     case 'A':
 413     case 'e':
 414     case 'E':
 415     case 'f':
 416     case 'F':
 417     case 'g':
 418     case 'G':
 419       {
 420         long double arg = vstrtold (argument);
 421         if (!have_field_width)
 422           {
 423             if (!have_precision)
 424               xprintf (p, arg);
 425             else
 426               xprintf (p, precision, arg);
 427           }
 428         else
 429           {
 430             if (!have_precision)
 431               xprintf (p, field_width, arg);
 432             else
 433               xprintf (p, field_width, precision, arg);
 434           }
 435       }
 436       break;
 437
 438     case 'c':
 439       if (!have_field_width)
 440         xprintf (p, *argument);
 441       else
 442         xprintf (p, field_width, *argument);
 443       break;
 444
 445     case 's':
 446       if (!have_field_width)
 447         {
 448           if (!have_precision)
 449             xprintf (p, argument);
 450           else
 451             xprintf (p, precision, argument);
 452         }
 453       else
 454         {
 455           if (!have_precision)
 456             xprintf (p, field_width, argument);
 457           else
 458             xprintf (p, field_width, precision, argument);
 459         }
 460       break;
 461     }
 462
 463   free (p);
 464 }
 465
 466 /* Print the text in FORMAT, using ARGV (with ARGC elements) for
 467    arguments to any '%' directives.
 468    Return the number of elements of ARGV used.  */
 469
 470 static int
 471 print_formatted (const char *format, int argc, char **argv)
 472 {
 473   int save_argc = argc;         /* Preserve original value.  */
 474   const char *f;                /* Pointer into 'format'.  */
 475   const char *direc_start;      /* Start of % directive.  */
 476   size_t direc_length;          /* Length of % directive.  */
 477   bool have_field_width;        /* True if FIELD_WIDTH is valid.  */
 478   int field_width = 0;          /* Arg to first '*'.  */
 479   bool have_precision;          /* True if PRECISION is valid.  */
 480   int precision = 0;            /* Arg to second '*'.  */
 481   char ok[UCHAR_MAX + 1];       /* ok['x'] is true if %x is allowed.  */
 482
 483   for (f = format; *f; ++f)
 484     {
 485       switch (*f)
 486         {
 487         case '%':
 488           direc_start = f++;
 489           direc_length = 1;
 490           have_field_width = have_precision = false;
 491           if (*f == '%')
 492             {
 493               putchar ('%');
 494               break;
 495             }
 496           if (*f == 'b')
 497             {
 498               /* FIXME: Field width and precision are not supported
 499                  for %b, even though POSIX requires it.  */
 500               if (argc > 0)
 501                 {
 502                   print_esc_string (*argv);
 503                   ++argv;
 504                   --argc;
 505                 }
 506               break;
 507             }
 508
 509           memset (ok, 0, sizeof ok);
 510           ok['a'] = ok['A'] = ok['c'] = ok['d'] = ok['e'] = ok['E'] =
 511             ok['f'] = ok['F'] = ok['g'] = ok['G'] = ok['i'] = ok['o'] =
 512             ok['s'] = ok['u'] = ok['x'] = ok['X'] = 1;
 513
 514           for (;; f++, direc_length++)
 515             switch (*f)
 516               {
 517 #if (__GLIBC__ == 2 && 2 <= __GLIBC_MINOR__) || 3 <= __GLIBC__
 518               case 'I':
 519 #endif
 520               case '\'':
 521                 ok['a'] = ok['A'] = ok['c'] = ok['e'] = ok['E'] =
 522                   ok['o'] = ok['s'] = ok['x'] = ok['X'] = 0;
 523                 break;
 524               case '-': case '+': case ' ':
 525                 break;
 526               case '#':
 527                 ok['c'] = ok['d'] = ok['i'] = ok['s'] = ok['u'] = 0;
 528                 break;
 529               case '0':
 530                 ok['c'] = ok['s'] = 0;
 531                 break;
 532               default:
 533                 goto no_more_flag_characters;
 534               }
 535         no_more_flag_characters:
 536
 537           if (*f == '*')
 538             {
 539               ++f;
 540               ++direc_length;
 541               if (argc > 0)
 542                 {
 543                   intmax_t width = vstrtoimax (*argv);
 544                   if (INT_MIN <= width && width <= INT_MAX)
 545                     field_width = width;
 546                   else
 547                     error (EXIT_FAILURE, 0, _("invalid field width: %s"),
 548                            *argv);
 549                   ++argv;
 550                   --argc;
 551                 }
 552               else
 553                 field_width = 0;
 554               have_field_width = true;
 555             }
 556           else
 557             while (ISDIGIT (*f))
 558               {
 559                 ++f;
 560                 ++direc_length;
 561               }
 562           if (*f == '.')
 563             {
 564               ++f;
 565               ++direc_length;
 566               ok['c'] = 0;
 567               if (*f == '*')
 568                 {
 569                   ++f;
 570                   ++direc_length;
 571                   if (argc > 0)
 572                     {
 573                       intmax_t prec = vstrtoimax (*argv);
 574                       if (prec < 0)
 575                         {
 576                           /* A negative precision is taken as if the
 577                              precision were omitted, so -1 is safe
 578                              here even if prec < INT_MIN.  */
 579                           precision = -1;
 580                         }
 581                       else if (INT_MAX < prec)
 582                         error (EXIT_FAILURE, 0, _("invalid precision: %s"),
 583                                *argv);
 584                       else
 585                         precision = prec;
 586                       ++argv;
 587                       --argc;
 588                     }
 589                   else
 590                     precision = 0;
 591                   have_precision = true;
 592                 }
 593               else
 594                 while (ISDIGIT (*f))
 595                   {
 596                     ++f;
 597                     ++direc_length;
 598                   }
 599             }
 600
 601           while (*f == 'l' || *f == 'L' || *f == 'h'
 602                  || *f == 'j' || *f == 't' || *f == 'z')
 603             ++f;
 604
 605           {
 606             unsigned char conversion = *f;
 607             if (! ok[conversion])
 608               error (EXIT_FAILURE, 0,
 609                      _("%.*s: invalid conversion specification"),
 610                      (int) (f + 1 - direc_start), direc_start);
 611           }
 612
 613           print_direc (direc_start, direc_length, *f,
 614                        have_field_width, field_width,
 615                        have_precision, precision,
 616                        (argc <= 0 ? "" : (argc--, *argv++)));
 617           break;
 618
 619         case '\\':
 620           f += print_esc (f, false);
 621           break;
 622
 623         default:
 624           putchar (*f);
 625         }
 626     }
 627
 628   return save_argc - argc;
 629 }
 630
 631 int
 632 main (int argc, char **argv)
 633 {
 634   char *format;
 635   int args_used;
 636
 637   initialize_main (&argc, &argv);
 638   set_program_name (argv[0]);
 639   setlocale (LC_ALL, "");
 640   bindtextdomain (PACKAGE, LOCALEDIR);
 641   textdomain (PACKAGE);
 642
 643   atexit (close_stdout);
 644
 645   exit_status = EXIT_SUCCESS;
 646
 647   posixly_correct = (getenv ("POSIXLY_CORRECT") != NULL);
 648
 649   /* We directly parse options, rather than use parse_long_options, in
 650      order to avoid accepting abbreviations.  */
 651   if (argc == 2)
 652     {
 653       if (STREQ (argv[1], "--help"))
 654         usage (EXIT_SUCCESS);
 655
 656       if (STREQ (argv[1], "--version"))
 657         {
 658           version_etc (stdout, PROGRAM_NAME, PACKAGE_NAME, Version, AUTHORS,
 659                        (char *) NULL);
 660           exit (EXIT_SUCCESS);
 661         }
 662     }
 663
 664   /* The above handles --help and --version.
 665      Since there is no other invocation of getopt, handle '--' here.  */
 666   if (1 < argc && STREQ (argv[1], "--"))
 667     {
 668       --argc;
 669       ++argv;
 670     }
 671
 672   if (argc <= 1)
 673     {
 674       error (0, 0, _("missing operand"));
 675       usage (EXIT_FAILURE);
 676     }
 677
 678   format = argv[1];
 679   argc -= 2;
 680   argv += 2;
 681
 682   do
 683     {
 684       args_used = print_formatted (format, argc, argv);
 685       argc -= args_used;
 686       argv += args_used;
 687     }
 688   while (args_used > 0 && argc > 0);
 689
 690   if (argc > 0)
 691     error (0, 0,
 692            _("warning: ignoring excess arguments, starting with %s"),
 693            quote (argv[0]));
 694
 695   exit (exit_status);
 696 }