lib/quotearg.c

   1 /* quotearg.c - quote arguments for output
   2    Copyright (C) 1998, 1999, 2000, 2001, 2002 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 2, or (at your option)
   7    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, write to the Free Software Foundation,
  16    Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
  17
  18 /* Written by Paul Eggert <eggert@twinsun.com> */
  19
  20 #if HAVE_CONFIG_H
  21 # include <config.h>
  22 #endif
  23
  24 #if HAVE_STDDEF_H
  25 # include <stddef.h>  /* For the definition of size_t on windows w/MSVC.  */
  26 #endif
  27 #include <sys/types.h>
  28 #include "quotearg.h"
  29 #include "xalloc.h"
  30
  31 #include <ctype.h>
  32
  33 #if ENABLE_NLS
  34 # include <libintl.h>
  35 # define _(text) gettext (text)
  36 #else
  37 # define _(text) text
  38 #endif
  39 #define N_(text) text
  40
  41 #if HAVE_LIMITS_H
  42 # include <limits.h>
  43 #endif
  44 #ifndef CHAR_BIT
  45 # define CHAR_BIT 8
  46 #endif
  47 #ifndef SIZE_MAX
  48 # define SIZE_MAX ((size_t) -1)
  49 #endif
  50 #ifndef UCHAR_MAX
  51 # define UCHAR_MAX ((unsigned char) -1)
  52 #endif
  53 #ifndef UINT_MAX
  54 # define UINT_MAX ((unsigned int) -1)
  55 #endif
  56
  57 #if HAVE_C_BACKSLASH_A
  58 # define ALERT_CHAR '\a'
  59 #else
  60 # define ALERT_CHAR '\7'
  61 #endif
  62
  63 #if HAVE_STDLIB_H
  64 # include <stdlib.h>
  65 #endif
  66
  67 #if HAVE_STRING_H
  68 # include <string.h>
  69 #endif
  70
  71 #if HAVE_WCHAR_H
  72
  73 /* BSD/OS 4.1 wchar.h requires FILE and struct tm to be declared.  */
  74 # include <stdio.h>
  75 # include <time.h>
  76
  77 # include <wchar.h>
  78 #endif
  79
  80 #if !HAVE_MBRTOWC
  81 /* Disable multibyte processing entirely.  Since MB_CUR_MAX is 1, the
  82    other macros are defined only for documentation and to satisfy C
  83    syntax.  */
  84 # undef MB_CUR_MAX
  85 # define MB_CUR_MAX 1
  86 # define mbrtowc(pwc, s, n, ps) ((*(pwc) = *(s)) != 0)
  87 # define iswprint(wc) ISPRINT ((unsigned char) (wc))
  88 # undef HAVE_MBSINIT
  89 #endif
  90
  91 #if !defined mbsinit && !HAVE_MBSINIT
  92 # define mbsinit(ps) 1
  93 #endif
  94
  95 #ifndef iswprint
  96 # if HAVE_WCTYPE_H
  97 #  include <wctype.h>
  98 # endif
  99 # if !defined iswprint && !HAVE_ISWPRINT
 100 #  define iswprint(wc) 1
 101 # endif
 102 #endif
 103
 104 #define INT_BITS (sizeof (int) * CHAR_BIT)
 105
 106 #if defined (STDC_HEADERS) || (!defined (isascii) && !defined (HAVE_ISASCII))
 107 # define IN_CTYPE_DOMAIN(c) 1
 108 #else
 109 # define IN_CTYPE_DOMAIN(c) isascii(c)
 110 #endif
 111
 112 /* Undefine to protect against the definition in wctype.h of solaris2.6.   */
 113 #undef ISPRINT
 114 #define ISPRINT(c) (IN_CTYPE_DOMAIN (c) && isprint (c))
 115
 116 struct quoting_options
 117 {
 118   /* Basic quoting style.  */
 119   enum quoting_style style;
 120
 121   /* Quote the characters indicated by this bit vector even if the
 122      quoting style would not normally require them to be quoted.  */
 123   int quote_these_too[(UCHAR_MAX / INT_BITS) + 1];
 124 };
 125
 126 /* Names of quoting styles.  */
 127 char const *const quoting_style_args[] =
 128 {
 129   "literal",
 130   "shell",
 131   "shell-always",
 132   "c",
 133   "escape",
 134   "locale",
 135   "clocale",
 136   0
 137 };
 138
 139 /* Correspondences to quoting style names.  */
 140 enum quoting_style const quoting_style_vals[] =
 141 {
 142   literal_quoting_style,
 143   shell_quoting_style,
 144   shell_always_quoting_style,
 145   c_quoting_style,
 146   escape_quoting_style,
 147   locale_quoting_style,
 148   clocale_quoting_style
 149 };
 150
 151 /* The default quoting options.  */
 152 static struct quoting_options default_quoting_options;
 153
 154 /* Allocate a new set of quoting options, with contents initially identical
 155    to O if O is not null, or to the default if O is null.
 156    It is the caller's responsibility to free the result.  */
 157 struct quoting_options *
 158 clone_quoting_options (struct quoting_options *o)
 159 {
 160   struct quoting_options *p
 161     = (struct quoting_options *) xmalloc (sizeof (struct quoting_options));
 162   *p = *(o ? o : &default_quoting_options);
 163   return p;
 164 }
 165
 166 /* Get the value of O's quoting style.  If O is null, use the default.  */
 167 enum quoting_style
 168 get_quoting_style (struct quoting_options *o)
 169 {
 170   return (o ? o : &default_quoting_options)->style;
 171 }
 172
 173 /* In O (or in the default if O is null),
 174    set the value of the quoting style to S.  */
 175 void
 176 set_quoting_style (struct quoting_options *o, enum quoting_style s)
 177 {
 178   (o ? o : &default_quoting_options)->style = s;
 179 }
 180
 181 /* In O (or in the default if O is null),
 182    set the value of the quoting options for character C to I.
 183    Return the old value.  Currently, the only values defined for I are
 184    0 (the default) and 1 (which means to quote the character even if
 185    it would not otherwise be quoted).  */
 186 int
 187 set_char_quoting (struct quoting_options *o, char c, int i)
 188 {
 189   unsigned char uc = c;
 190   int *p = (o ? o : &default_quoting_options)->quote_these_too + uc / INT_BITS;
 191   int shift = uc % INT_BITS;
 192   int r = (*p >> shift) & 1;
 193   *p ^= ((i & 1) ^ r) << shift;
 194   return r;
 195 }
 196
 197 /* MSGID approximates a quotation mark.  Return its translation if it
 198    has one; otherwise, return either it or "\"", depending on S.  */
 199 static char const *
 200 gettext_quote (char const *msgid, enum quoting_style s)
 201 {
 202   char const *translation = _(msgid);
 203   if (translation == msgid && s == clocale_quoting_style)
 204     translation = "\"";
 205   return translation;
 206 }
 207
 208 /* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of
 209    argument ARG (of size ARGSIZE), using QUOTING_STYLE and the
 210    non-quoting-style part of O to control quoting.
 211    Terminate the output with a null character, and return the written
 212    size of the output, not counting the terminating null.
 213    If BUFFERSIZE is too small to store the output string, return the
 214    value that would have been returned had BUFFERSIZE been large enough.
 215    If ARGSIZE is -1, use the string length of the argument for ARGSIZE.
 216
 217    This function acts like quotearg_buffer (BUFFER, BUFFERSIZE, ARG,
 218    ARGSIZE, O), except it uses QUOTING_STYLE instead of the quoting
 219    style specified by O, and O may not be null.  */
 220
 221 static size_t
 222 quotearg_buffer_restyled (char *buffer, size_t buffersize,
 223                           char const *arg, size_t argsize,
 224                           enum quoting_style quoting_style,
 225                           struct quoting_options const *o)
 226 {
 227   size_t i;
 228   size_t len = 0;
 229   char const *quote_string = 0;
 230   size_t quote_string_len = 0;
 231   int backslash_escapes = 0;
 232   int unibyte_locale = MB_CUR_MAX == 1;
 233
 234 #define STORE(c) \
 235     do \
 236       { \
 237         if (len < buffersize) \
 238           buffer[len] = (c); \
 239         len++; \
 240       } \
 241     while (0)
 242
 243   switch (quoting_style)
 244     {
 245     case c_quoting_style:
 246       STORE ('"');
 247       backslash_escapes = 1;
 248       quote_string = "\"";
 249       quote_string_len = 1;
 250       break;
 251
 252     case escape_quoting_style:
 253       backslash_escapes = 1;
 254       break;
 255
 256     case locale_quoting_style:
 257     case clocale_quoting_style:
 258       {
 259         /* Get translations for open and closing quotation marks.
 260
 261            The message catalog should translate "`" to a left
 262            quotation mark suitable for the locale, and similarly for
 263            "'".  If the catalog has no translation,
 264            locale_quoting_style quotes `like this', and
 265            clocale_quoting_style quotes "like this".
 266
 267            For example, an American English Unicode locale should
 268            translate "`" to U+201C (LEFT DOUBLE QUOTATION MARK), and
 269            should translate "'" to U+201D (RIGHT DOUBLE QUOTATION
 270            MARK).  A British English Unicode locale should instead
 271            translate these to U+2018 (LEFT SINGLE QUOTATION MARK) and
 272            U+2019 (RIGHT SINGLE QUOTATION MARK), respectively.  */
 273
 274         char const *left = gettext_quote (N_("`"), quoting_style);
 275         char const *right = gettext_quote (N_("'"), quoting_style);
 276         for (quote_string = left; *quote_string; quote_string++)
 277           STORE (*quote_string);
 278         backslash_escapes = 1;
 279         quote_string = right;
 280         quote_string_len = strlen (quote_string);
 281       }
 282       break;
 283
 284     case shell_always_quoting_style:
 285       STORE ('\'');
 286       quote_string = "'";
 287       quote_string_len = 1;
 288       break;
 289
 290     default:
 291       break;
 292     }
 293
 294   for (i = 0;  ! (argsize == (size_t) -1 ? arg[i] == '\0' : i == argsize);  i++)
 295     {
 296       unsigned char c;
 297       unsigned char esc;
 298
 299       if (backslash_escapes
 300           && quote_string_len
 301           && i + quote_string_len <= argsize
 302           && memcmp (arg + i, quote_string, quote_string_len) == 0)
 303         STORE ('\\');
 304
 305       c = arg[i];
 306       switch (c)
 307         {
 308         case '\0':
 309           if (backslash_escapes)
 310             {
 311               STORE ('\\');
 312               STORE ('0');
 313               STORE ('0');
 314               c = '0';
 315             }
 316           break;
 317
 318         case '?':
 319           switch (quoting_style)
 320             {
 321             case shell_quoting_style:
 322               goto use_shell_always_quoting_style;
 323
 324             case c_quoting_style:
 325               if (i + 2 < argsize && arg[i + 1] == '?')
 326                 switch (arg[i + 2])
 327                   {
 328                   case '!': case '\'':
 329                   case '(': case ')': case '-': case '/':
 330                   case '<': case '=': case '>':
 331                     /* Escape the second '?' in what would otherwise be
 332                        a trigraph.  */
 333                     c = arg[i + 2];
 334                     i += 2;
 335                     STORE ('?');
 336                     STORE ('\\');
 337                     STORE ('?');
 338                     break;
 339                   }
 340               break;
 341
 342             default:
 343               break;
 344             }
 345           break;
 346
 347         case ALERT_CHAR: esc = 'a'; goto c_escape;
 348         case '\b': esc = 'b'; goto c_escape;
 349         case '\f': esc = 'f'; goto c_escape;
 350         case '\n': esc = 'n'; goto c_and_shell_escape;
 351         case '\r': esc = 'r'; goto c_and_shell_escape;
 352         case '\t': esc = 't'; goto c_and_shell_escape;
 353         case '\v': esc = 'v'; goto c_escape;
 354         case '\\': esc = c; goto c_and_shell_escape;
 355
 356         c_and_shell_escape:
 357           if (quoting_style == shell_quoting_style)
 358             goto use_shell_always_quoting_style;
 359         c_escape:
 360           if (backslash_escapes)
 361             {
 362               c = esc;
 363               goto store_escape;
 364             }
 365           break;
 366
 367         case '#': case '~':
 368           if (i != 0)
 369             break;
 370           /* Fall through.  */
 371         case ' ':
 372         case '!': /* special in bash */
 373         case '"': case '$': case '&':
 374         case '(': case ')': case '*': case ';':
 375         case '<': case '>': case '[':
 376         case '^': /* special in old /bin/sh, e.g. SunOS 4.1.4 */
 377         case '`': case '|':
 378           /* A shell special character.  In theory, '$' and '`' could
 379              be the first bytes of multibyte characters, which means
 380              we should check them with mbrtowc, but in practice this
 381              doesn't happen so it's not worth worrying about.  */
 382           if (quoting_style == shell_quoting_style)
 383             goto use_shell_always_quoting_style;
 384           break;
 385
 386         case '\'':
 387           switch (quoting_style)
 388             {
 389             case shell_quoting_style:
 390               goto use_shell_always_quoting_style;
 391
 392             case shell_always_quoting_style:
 393               STORE ('\'');
 394               STORE ('\\');
 395               STORE ('\'');
 396               break;
 397
 398             default:
 399               break;
 400             }
 401           break;
 402
 403         case '%': case '+': case ',': case '-': case '.': case '/':
 404         case '0': case '1': case '2': case '3': case '4': case '5':
 405         case '6': case '7': case '8': case '9': case ':': case '=':
 406         case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
 407         case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
 408         case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
 409         case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
 410         case 'Y': case 'Z': case ']': case '_': case 'a': case 'b':
 411         case 'c': case 'd': case 'e': case 'f': case 'g': case 'h':
 412         case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
 413         case 'o': case 'p': case 'q': case 'r': case 's': case 't':
 414         case 'u': case 'v': case 'w': case 'x': case 'y': case 'z':
 415         case '{': case '}':
 416           /* These characters don't cause problems, no matter what the
 417              quoting style is.  They cannot start multibyte sequences.  */
 418           break;
 419
 420         default:
 421           /* If we have a multibyte sequence, copy it until we reach
 422              its end, find an error, or come back to the initial shift
 423              state.  For C-like styles, if the sequence has
 424              unprintable characters, escape the whole sequence, since
 425              we can't easily escape single characters within it.  */
 426           {
 427             /* Length of multibyte sequence found so far.  */
 428             size_t m;
 429
 430             int printable;
 431
 432             if (unibyte_locale)
 433               {
 434                 m = 1;
 435                 printable = ISPRINT (c);
 436               }
 437             else
 438               {
 439                 mbstate_t mbstate;
 440                 memset (&mbstate, 0, sizeof mbstate);
 441
 442                 m = 0;
 443                 printable = 1;
 444                 if (argsize == (size_t) -1)
 445                   argsize = strlen (arg);
 446
 447                 do
 448                   {
 449                     wchar_t w;
 450                     size_t bytes = mbrtowc (&w, &arg[i + m],
 451                                             argsize - (i + m), &mbstate);
 452                     if (bytes == 0)
 453                       break;
 454                     else if (bytes == (size_t) -1)
 455                       {
 456                         printable = 0;
 457                         break;
 458                       }
 459                     else if (bytes == (size_t) -2)
 460                       {
 461                         printable = 0;
 462                         while (i + m < argsize && arg[i + m])
 463                           m++;
 464                         break;
 465                       }
 466                     else
 467                       {
 468                         if (! iswprint (w))
 469                           printable = 0;
 470                         m += bytes;
 471                       }
 472                   }
 473                 while (! mbsinit (&mbstate));
 474               }
 475
 476             if (1 < m || (backslash_escapes && ! printable))
 477               {
 478                 /* Output a multibyte sequence, or an escaped
 479                    unprintable unibyte character.  */
 480                 size_t ilim = i + m;
 481
 482                 for (;;)
 483                   {
 484                     if (backslash_escapes && ! printable)
 485                       {
 486                         STORE ('\\');
 487                         STORE ('0' + (c >> 6));
 488                         STORE ('0' + ((c >> 3) & 7));
 489                         c = '0' + (c & 7);
 490                       }
 491                     if (ilim <= i + 1)
 492                       break;
 493                     STORE (c);
 494                     c = arg[++i];
 495                   }
 496
 497                 goto store_c;
 498               }
 499           }
 500         }
 501
 502       if (! (backslash_escapes
 503              && o->quote_these_too[c / INT_BITS] & (1 << (c % INT_BITS))))
 504         goto store_c;
 505
 506     store_escape:
 507       STORE ('\\');
 508
 509     store_c:
 510       STORE (c);
 511     }
 512
 513   if (quote_string)
 514     for (; *quote_string; quote_string++)
 515       STORE (*quote_string);
 516
 517   if (len < buffersize)
 518     buffer[len] = '\0';
 519   return len;
 520
 521  use_shell_always_quoting_style:
 522   return quotearg_buffer_restyled (buffer, buffersize, arg, argsize,
 523                                    shell_always_quoting_style, o);
 524 }
 525
 526 /* Place into buffer BUFFER (of size BUFFERSIZE) a quoted version of
 527    argument ARG (of size ARGSIZE), using O to control quoting.
 528    If O is null, use the default.
 529    Terminate the output with a null character, and return the written
 530    size of the output, not counting the terminating null.
 531    If BUFFERSIZE is too small to store the output string, return the
 532    value that would have been returned had BUFFERSIZE been large enough.
 533    If ARGSIZE is -1, use the string length of the argument for ARGSIZE.  */
 534 size_t
 535 quotearg_buffer (char *buffer, size_t buffersize,
 536                  char const *arg, size_t argsize,
 537                  struct quoting_options const *o)
 538 {
 539   struct quoting_options const *p = o ? o : &default_quoting_options;
 540   return quotearg_buffer_restyled (buffer, buffersize, arg, argsize,
 541                                    p->style, p);
 542 }
 543
 544 /* Use storage slot N to return a quoted version of argument ARG.
 545    ARG is of size ARGSIZE, but if that is -1, ARG is a null-terminated string.
 546    OPTIONS specifies the quoting options.
 547    The returned value points to static storage that can be
 548    reused by the next call to this function with the same value of N.
 549    N must be nonnegative.  N is deliberately declared with type "int"
 550    to allow for future extensions (using negative values).  */
 551 static char *
 552 quotearg_n_options (int n, char const *arg, size_t argsize,
 553                     struct quoting_options const *options)
 554 {
 555   /* Preallocate a slot 0 buffer, so that the caller can always quote
 556      one small component of a "memory exhausted" message in slot 0.  */
 557   static char slot0[256];
 558   static unsigned int nslots = 1;
 559   unsigned int n0 = n;
 560   struct slotvec
 561     {
 562       size_t size;
 563       char *val;
 564     };
 565   static struct slotvec slotvec0 = {sizeof slot0, slot0};
 566   static struct slotvec *slotvec = &slotvec0;
 567
 568   if (n < 0)
 569     abort ();
 570
 571   if (nslots <= n0)
 572     {
 573       unsigned int n1 = n0 + 1;
 574       size_t s = n1 * sizeof *slotvec;
 575
 576       if (SIZE_MAX / UINT_MAX <= sizeof *slotvec
 577           && n1 != s / sizeof *slotvec)
 578         xalloc_die ();
 579
 580       if (slotvec == &slotvec0)
 581         {
 582           slotvec = (struct slotvec *) xmalloc (sizeof *slotvec);
 583           *slotvec = slotvec0;
 584         }
 585       slotvec = (struct slotvec *) xrealloc (slotvec, s);
 586       memset (slotvec + nslots, 0, (n1 - nslots) * sizeof *slotvec);
 587       nslots = n1;
 588     }
 589
 590   {
 591     size_t size = slotvec[n].size;
 592     char *val = slotvec[n].val;
 593     size_t qsize = quotearg_buffer (val, size, arg, argsize, options);
 594
 595     if (size <= qsize)
 596       {
 597         slotvec[n].size = size = qsize + 1;
 598         slotvec[n].val = val = xrealloc (val == slot0 ? 0 : val, size);
 599         quotearg_buffer (val, size, arg, argsize, options);
 600       }
 601
 602     return val;
 603   }
 604 }
 605
 606 char *
 607 quotearg_n (int n, char const *arg)
 608 {
 609   return quotearg_n_options (n, arg, (size_t) -1, &default_quoting_options);
 610 }
 611
 612 char *
 613 quotearg (char const *arg)
 614 {
 615   return quotearg_n (0, arg);
 616 }
 617
 618 /* Return quoting options for STYLE, with no extra quoting.  */
 619 static struct quoting_options
 620 quoting_options_from_style (enum quoting_style style)
 621 {
 622   struct quoting_options o;
 623   o.style = style;
 624   memset (o.quote_these_too, 0, sizeof o.quote_these_too);
 625   return o;
 626 }
 627
 628 char *
 629 quotearg_n_style (int n, enum quoting_style s, char const *arg)
 630 {
 631   struct quoting_options const o = quoting_options_from_style (s);
 632   return quotearg_n_options (n, arg, (size_t) -1, &o);
 633 }
 634
 635 char *
 636 quotearg_n_style_mem (int n, enum quoting_style s,
 637                       char const *arg, size_t argsize)
 638 {
 639   struct quoting_options const o = quoting_options_from_style (s);
 640   return quotearg_n_options (n, arg, argsize, &o);
 641 }
 642
 643 char *
 644 quotearg_style (enum quoting_style s, char const *arg)
 645 {
 646   return quotearg_n_style (0, s, arg);
 647 }
 648
 649 char *
 650 quotearg_char (char const *arg, char ch)
 651 {
 652   struct quoting_options options;
 653   options = default_quoting_options;
 654   set_char_quoting (&options, ch, 1);
 655   return quotearg_n_options (0, arg, (size_t) -1, &options);
 656 }
 657
 658 char *
 659 quotearg_colon (char const *arg)
 660 {
 661   return quotearg_char (arg, ':');
 662 }