lib/qmark.c

   1 /* qmark.c -- quote 'dangerous' filenames
   2
   3    Copyright (C) 2005 Free Software Foundation, Inc.
   4    Derived from courutils' ls.c:
   5    Copyright (C) 85, 88, 90, 91, 1995-2005 Free Software Foundation, Inc.
   6
   7    This program is free software; you can redistribute it and/or modify
   8    it under the terms of the GNU General Public License as published by
   9    the Free Software Foundation; either version 2, or (at your option)
  10    any later version.
  11
  12    This program is distributed in the hope that it will be useful,
  13    but WITHOUT ANY WARRANTY; without even the implied warranty of
  14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15    GNU General Public License for more details.
  16
  17    You should have received a copy of the GNU General Public License
  18    along with this program; if not, write to the Free Software
  19    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301,
  20    USA.
  21 */
  22
  23 #ifdef HAVE_CONFIG_H
  24 #include <config.h>
  25 #endif
  26
  27 # include <stddef.h>
  28 # include <stdlib.h>
  29 #include <ctype.h>
  30
  31 #if HAVE_STRING_H || STDC_HEADERS
  32 #include <string.h>
  33 #else
  34 #include <strings.h>
  35 #endif
  36
  37
  38 /* Get mbstate_t, mbrtowc(), mbsinit(), wcwidth().  */
  39 #if HAVE_WCHAR_H
  40 # include <wchar.h>
  41 #endif
  42
  43 #include "printquoted.h"
  44
  45
  46 /*
  47    This comment, IN_CTYPE_DOMAIN and ISPRINT were borrowed from
  48    coreutils at Sun Jun  5 21:17:40 2005 UTC.
  49
  50    Jim Meyering writes:
  51
  52    "... Some ctype macros are valid only for character codes that
  53    isascii says are ASCII (SGI's IRIX-4.0.5 is one such system --when
  54    using /bin/cc or gcc but without giving an ansi option).  So, all
  55    ctype uses should be through macros like ISPRINT...  If
  56    STDC_HEADERS is defined, then autoconf has verified that the ctype
  57    macros don't need to be guarded with references to isascii. ...
  58    Defining isascii to 1 should let any compiler worth its salt
  59    eliminate the && through constant folding."
  60
  61    Bruno Haible adds:
  62
  63    "... Furthermore, isupper(c) etc. have an undefined result if c is
  64    outside the range -1 <= c <= 255. One is tempted to write isupper(c)
  65    with c being of type `char', but this is wrong if c is an 8-bit
  66    character >= 128 which gets sign-extended to a negative value.
  67    The macro ISUPPER protects against this as well."  */
  68
  69
  70
  71
  72 /* ISPRINT is defined in <sys/euc.h> on at least Solaris2.6 systems.  */
  73 #undef ISPRINT
  74 #define ISPRINT(c) (IN_CTYPE_DOMAIN (c) && isprint (c))
  75
  76 #if STDC_HEADERS || (!defined (isascii) && !HAVE_ISASCII)
  77 # define IN_CTYPE_DOMAIN(c) 1
  78 #else
  79 # define IN_CTYPE_DOMAIN(c) isascii(c)
  80 #endif
  81
  82
  83
  84
  85
  86 /* Convert a possibly-signed character to an unsigned character.  This is
  87  * a bit safer than casting to unsigned char, since it catches some type
  88  * errors that the cast doesn't.
  89  *
  90  * This code taken from coreutils' system.h header at
  91  * Sun Jun  5 21:05:21 2005 UTC.
  92  */
  93 static inline unsigned char to_uchar (char ch)
  94 {
  95   return ch;
  96 }
  97
  98
  99
 100 static size_t
 101 unibyte_qmark_chars(char *buf, size_t len)
 102 {
 103   char *p = buf;
 104   char const *plimit = buf + len;
 105
 106   while (p < plimit)
 107     {
 108       if (! ISPRINT (to_uchar (*p)))
 109         *p = '?';
 110       p++;
 111     }
 112   return len;
 113 }
 114
 115
 116 #if HAVE_MBRTOWC
 117 static size_t
 118 multibyte_qmark_chars(char *buf, size_t len)
 119 {
 120   if (MB_CUR_MAX <= 1)
 121     {
 122       return unibyte_qmark_chars(buf, len);
 123     }
 124   else
 125     {
 126       char const *p = buf;
 127       char const *plimit = buf + len;
 128       char *q = buf;
 129
 130       while (p < plimit)
 131         switch (*p)
 132           {
 133           case ' ': case '!': case '"': case '#': case '%':
 134           case '&': case '\'': case '(': case ')': case '*':
 135           case '+': case ',': case '-': case '.': case '/':
 136           case '0': case '1': case '2': case '3': case '4':
 137           case '5': case '6': case '7': case '8': case '9':
 138           case ':': case ';': case '<': case '=': case '>':
 139           case '?':
 140           case 'A': case 'B': case 'C': case 'D': case 'E':
 141           case 'F': case 'G': case 'H': case 'I': case 'J':
 142           case 'K': case 'L': case 'M': case 'N': case 'O':
 143           case 'P': case 'Q': case 'R': case 'S': case 'T':
 144           case 'U': case 'V': case 'W': case 'X': case 'Y':
 145           case 'Z':
 146           case '[': case '\\': case ']': case '^': case '_':
 147           case 'a': case 'b': case 'c': case 'd': case 'e':
 148           case 'f': case 'g': case 'h': case 'i': case 'j':
 149           case 'k': case 'l': case 'm': case 'n': case 'o':
 150           case 'p': case 'q': case 'r': case 's': case 't':
 151           case 'u': case 'v': case 'w': case 'x': case 'y':
 152           case 'z': case '{': case '|': case '}': case '~':
 153             /* These characters are printable ASCII characters.  */
 154             *q++ = *p++;
 155             break;
 156           default:
 157             /* If we have a multibyte sequence, copy it until we
 158                reach its end, replacing each non-printable multibyte
 159                character with a single question mark.  */
 160             {
 161               mbstate_t mbstate;
 162               memset (&mbstate, 0, sizeof mbstate);
 163               do
 164                 {
 165                   wchar_t wc;
 166                   size_t bytes;
 167                   int w;
 168
 169                   bytes = mbrtowc (&wc, p, plimit - p, &mbstate);
 170
 171                   if (bytes == (size_t) -1)
 172                     {
 173                       /* An invalid multibyte sequence was
 174                          encountered.  Skip one input byte, and
 175                          put a question mark.  */
 176                       p++;
 177                       *q++ = '?';
 178                       break;
 179                     }
 180
 181                   if (bytes == (size_t) -2)
 182                     {
 183                       /* An incomplete multibyte character
 184                          at the end.  Replace it entirely with
 185                          a question mark.  */
 186                       p = plimit;
 187                       *q++ = '?';
 188                       break;
 189                     }
 190
 191                   if (bytes == 0)
 192                     /* A null wide character was encountered.  */
 193                     bytes = 1;
 194
 195                   w = wcwidth (wc);
 196                   if (w >= 0)
 197                     {
 198                       /* A printable multibyte character.
 199                          Keep it.  */
 200                       for (; bytes > 0; --bytes)
 201                         *q++ = *p++;
 202                     }
 203                   else
 204                     {
 205                       /* An unprintable multibyte character.
 206                          Replace it entirely with a question
 207                          mark.  */
 208                       p += bytes;
 209                       *q++ = '?';
 210                     }
 211                 }
 212               while (! mbsinit (&mbstate));
 213             }
 214             break;
 215           }
 216
 217       /* The buffer may have shrunk.  */
 218       len = q - buf;
 219       return len;
 220     }
 221 }
 222 #endif
 223
 224
 225 /* Scan BUF, replacing any dangerous-looking characters with question
 226  * marks.  This code is taken from the ls.c file in coreutils as at
 227  * Sun Jun  5 20:51:54 2005 UTC.
 228  *
 229  * This function may shrink the buffer.   Either way, the new length
 230  * is returned.
 231  */
 232 size_t
 233 qmark_chars(char *buf, size_t len)
 234 {
 235 #if HAVE_MBRTOWC
 236   return multibyte_qmark_chars(buf, len);
 237 #else
 238   return unibyte_qmark_chars(buf, len);
 239 #endif
 240 }
 241