mbswidth.c

   1 /* Determine the number of screen columns needed for a string.
   2    Copyright (C) 2000-2002 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 Bruno Haible <haible@clisp.cons.org>.  */
  19 /*
  20  * Adapted for abook by JH <jheinonen@users.sourceforge.net>
  21  */
  22
  23 #ifdef HAVE_CONFIG_H
  24 # include <config.h>
  25 #endif
  26
  27 /* Specification.  */
  28 #include "mbswidth.h"
  29
  30 #ifdef HANDLE_MULTIBYTE /* for abook */
  31
  32 /* Get MB_CUR_MAX.  */
  33 #include <stdlib.h>
  34
  35 #include <string.h>
  36
  37 /* Get isprint().  */
  38 #include <ctype.h>
  39
  40 /* Get mbstate_t, mbrtowc(), mbsinit(), wcwidth().  */
  41 #if HAVE_WCHAR_H
  42 # include <wchar.h>
  43 #endif
  44
  45 /* Get iswprint(), iswcntrl().  */
  46 #if HAVE_WCTYPE_H
  47 # include <wctype.h>
  48 #endif
  49 #if !defined iswprint && !HAVE_ISWPRINT
  50 # define iswprint(wc) 1
  51 #endif
  52 #if !defined iswcntrl && !HAVE_ISWCNTRL
  53 # define iswcntrl(wc) 0
  54 #endif
  55
  56 #ifndef mbsinit
  57 # if !HAVE_MBSINIT
  58 #  define mbsinit(ps) 1
  59 # endif
  60 #endif
  61
  62 #ifndef HAVE_DECL_WCWIDTH
  63 #       warn "this configure-time declaration test was not run"
  64 #endif
  65 #if !HAVE_DECL_WCWIDTH
  66 int wcwidth ();
  67 #endif
  68
  69 #ifndef wcwidth
  70 # if !HAVE_WCWIDTH
  71 /* wcwidth doesn't exist, so assume all printable characters have
  72    width 1.  */
  73 #  define wcwidth(wc) ((wc) == 0 ? 0 : iswprint (wc) ? 1 : -1)
  74 # endif
  75 #endif
  76
  77 /* Get ISPRINT.  */
  78 #if defined (STDC_HEADERS) || (!defined (isascii) && !defined (HAVE_ISASCII))
  79 # define IN_CTYPE_DOMAIN(c) 1
  80 #else
  81 # define IN_CTYPE_DOMAIN(c) isascii(c)
  82 #endif
  83 /* Undefine to protect against the definition in wctype.h of solaris2.6.   */
  84 #undef ISPRINT
  85 #define ISPRINT(c) (IN_CTYPE_DOMAIN (c) && isprint (c))
  86 #undef ISCNTRL
  87 #define ISCNTRL(c) (IN_CTYPE_DOMAIN (c) && iscntrl (c))
  88
  89 /* Returns the number of columns needed to represent the multibyte
  90    character string pointed to by STRING.  If a non-printable character
  91    occurs, and MBSW_REJECT_UNPRINTABLE is specified, -1 is returned.
  92    With flags = MBSW_REJECT_INVALID | MBSW_REJECT_UNPRINTABLE, this is
  93    the multibyte analogon of the wcswidth function.  */
  94 int
  95 mbswidth (const char *string, int flags)
  96 {
  97   return mbsnwidth (string, strlen (string), flags);
  98 }
  99
 100 /* Returns the number of columns needed to represent the multibyte
 101    character string pointed to by STRING of length NBYTES.  If a
 102    non-printable character occurs, and MBSW_REJECT_UNPRINTABLE is
 103    specified, -1 is returned.  */
 104 int
 105 mbsnwidth (const char *string, size_t nbytes, int flags)
 106 {
 107   const char *p = string;
 108   const char *plimit = p + nbytes;
 109   int width;
 110
 111   width = 0;
 112 #if HAVE_MBRTOWC
 113   if (MB_CUR_MAX > 1)
 114     {
 115       while (p < plimit)
 116         switch (*p)
 117           {
 118             case ' ': case '!': case '"': case '#': case '%':
 119             case '&': case '\'': case '(': case ')': case '*':
 120             case '+': case ',': case '-': case '.': case '/':
 121             case '0': case '1': case '2': case '3': case '4':
 122             case '5': case '6': case '7': case '8': case '9':
 123             case ':': case ';': case '<': case '=': case '>':
 124             case '?':
 125             case 'A': case 'B': case 'C': case 'D': case 'E':
 126             case 'F': case 'G': case 'H': case 'I': case 'J':
 127             case 'K': case 'L': case 'M': case 'N': case 'O':
 128             case 'P': case 'Q': case 'R': case 'S': case 'T':
 129             case 'U': case 'V': case 'W': case 'X': case 'Y':
 130             case 'Z':
 131             case '[': case '\\': case ']': case '^': case '_':
 132             case 'a': case 'b': case 'c': case 'd': case 'e':
 133             case 'f': case 'g': case 'h': case 'i': case 'j':
 134             case 'k': case 'l': case 'm': case 'n': case 'o':
 135             case 'p': case 'q': case 'r': case 's': case 't':
 136             case 'u': case 'v': case 'w': case 'x': case 'y':
 137             case 'z': case '{': case '|': case '}': case '~':
 138               /* These characters are printable ASCII characters.  */
 139               p++;
 140               width++;
 141               break;
 142             default:
 143               /* If we have a multibyte sequence, scan it up to its end.  */
 144               {
 145                 mbstate_t mbstate;
 146                 memset (&mbstate, 0, sizeof mbstate);
 147                 do
 148                   {
 149                     wchar_t wc;
 150                     size_t bytes;
 151                     int w;
 152
 153                     bytes = mbrtowc (&wc, p, plimit - p, &mbstate);
 154
 155                     if (bytes == (size_t) -1)
 156                       /* An invalid multibyte sequence was encountered.  */
 157                       {
 158                         if (!(flags & MBSW_REJECT_INVALID))
 159                           {
 160                             p++;
 161                             width++;
 162                             break;
 163                           }
 164                         else
 165                           return -1;
 166                       }
 167
 168                     if (bytes == (size_t) -2)
 169                       /* An incomplete multibyte character at the end.  */
 170                       {
 171                         if (!(flags & MBSW_REJECT_INVALID))
 172                           {
 173                             p = plimit;
 174                             width++;
 175                             break;
 176                           }
 177                         else
 178                           return -1;
 179                       }
 180
 181                     if (bytes == 0)
 182                       /* A null wide character was encountered.  */
 183                       bytes = 1;
 184
 185                     w = wcwidth (wc);
 186                     if (w >= 0)
 187                       /* A printable multibyte character.  */
 188                       width += w;
 189                     else
 190                       /* An unprintable multibyte character.  */
 191                       if (!(flags & MBSW_REJECT_UNPRINTABLE))
 192                         width += (iswcntrl (wc) ? 0 : 1);
 193                       else
 194                         return -1;
 195
 196                     p += bytes;
 197                   }
 198                 while (! mbsinit (&mbstate));
 199               }
 200               break;
 201           }
 202       return width;
 203     }
 204 #endif
 205
 206   while (p < plimit)
 207     {
 208       unsigned char c = (unsigned char) *p++;
 209
 210       if (ISPRINT (c))
 211         width++;
 212       else if (!(flags & MBSW_REJECT_UNPRINTABLE))
 213         width += (ISCNTRL (c) ? 0 : 1);
 214       else
 215         return -1;
 216     }
 217   return width;
 218 }
 219
 220 int
 221 mbsnbytes (const char *string, size_t nbytes, int maxwidth, int flags)
 222 {
 223   const char *p = string, *old_p = string;
 224   const char *plimit = p + nbytes;
 225   int width;
 226
 227   width = 0;
 228 #if HAVE_MBRTOWC
 229   if (MB_CUR_MAX > 1)
 230     {
 231       while (p < plimit && width < maxwidth)
 232         {
 233         old_p = p;
 234         switch (*p)
 235           {
 236             case ' ': case '!': case '"': case '#': case '%':
 237             case '&': case '\'': case '(': case ')': case '*':
 238             case '+': case ',': case '-': case '.': case '/':
 239             case '0': case '1': case '2': case '3': case '4':
 240             case '5': case '6': case '7': case '8': case '9':
 241             case ':': case ';': case '<': case '=': case '>':
 242             case '?':
 243             case 'A': case 'B': case 'C': case 'D': case 'E':
 244             case 'F': case 'G': case 'H': case 'I': case 'J':
 245             case 'K': case 'L': case 'M': case 'N': case 'O':
 246             case 'P': case 'Q': case 'R': case 'S': case 'T':
 247             case 'U': case 'V': case 'W': case 'X': case 'Y':
 248             case 'Z':
 249             case '[': case '\\': case ']': case '^': case '_':
 250             case 'a': case 'b': case 'c': case 'd': case 'e':
 251             case 'f': case 'g': case 'h': case 'i': case 'j':
 252             case 'k': case 'l': case 'm': case 'n': case 'o':
 253             case 'p': case 'q': case 'r': case 's': case 't':
 254             case 'u': case 'v': case 'w': case 'x': case 'y':
 255             case 'z': case '{': case '|': case '}': case '~':
 256               /* These characters are printable ASCII characters.  */
 257               p++;
 258               width++;
 259               break;
 260             default:
 261               /* If we have a multibyte sequence, scan it up to its end.  */
 262               {
 263                 mbstate_t mbstate;
 264                 memset (&mbstate, 0, sizeof mbstate);
 265                 do
 266                   {
 267                     wchar_t wc;
 268                     size_t bytes;
 269                     int w;
 270
 271                     bytes = mbrtowc (&wc, p, plimit - p, &mbstate);
 272
 273                     if (bytes == (size_t) -1)
 274                       /* An invalid multibyte sequence was encountered.  */
 275                       {
 276                         if (!(flags & MBSW_REJECT_INVALID))
 277                           {
 278                             p++;
 279                             width++;
 280                             break;
 281                           }
 282                         else
 283                           return -1;
 284                       }
 285
 286                     if (bytes == (size_t) -2)
 287                       /* An incomplete multibyte character at the end.  */
 288                       {
 289                         if (!(flags & MBSW_REJECT_INVALID))
 290                           {
 291                             p = plimit;
 292                             width++;
 293                             break;
 294                           }
 295                         else
 296                           return -1;
 297                       }
 298
 299                     if (bytes == 0)
 300                       /* A null wide character was encountered.  */
 301                       bytes = 1;
 302
 303                     w = wcwidth (wc);
 304                     if (w >= 0)
 305                       /* A printable multibyte character.  */
 306                       width += w;
 307                     else
 308                       /* An unprintable multibyte character.  */
 309                       if (!(flags & MBSW_REJECT_UNPRINTABLE))
 310                         width += (iswcntrl (wc) ? 0 : 1);
 311                       else
 312                         return -1;
 313
 314                     p += bytes;
 315                   }
 316                 while (! mbsinit (&mbstate));
 317               }
 318               break;
 319           }
 320       }
 321
 322       return (width > maxwidth) ? (old_p - string) : (p - string);
 323     }
 324 #endif
 325
 326   return maxwidth;
 327 }
 328
 329 #else /* HANDLE_MULTIBYTE */
 330
 331 #include <string.h>
 332 int
 333 mbswidth (const char *string, int flags)
 334 {
 335   return strlen(string);
 336 }
 337 #endif /* HANDLE_MULTIBYTE */
 338