gl/lib/mbsalign.c

   1 /* Align/Truncate a string in a given screen width
   2    Copyright (C) 2009-2012 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 /* Written by Pádraig Brady.  */
  18
  19 #include <config.h>
  20 #include "mbsalign.h"
  21
  22 #include <stdlib.h>
  23 #include <string.h>
  24 #include <stdio.h>
  25 #include <stdint.h>
  26 #include <stdbool.h>
  27 #include <limits.h>
  28 #include <wchar.h>
  29 #include <wctype.h>
  30
  31 #ifndef MIN
  32 # define MIN(a, b) ((a) < (b) ? (a) : (b))
  33 #endif
  34
  35 /* Replace non printable chars.
  36    Note \t and \n etc. are non printable.
  37    Return 1 if replacement made, 0 otherwise.  */
  38
  39 static bool
  40 wc_ensure_printable (wchar_t *wchars)
  41 {
  42   bool replaced = false;
  43   wchar_t *wc = wchars;
  44   while (*wc)
  45     {
  46       if (!iswprint ((wint_t) *wc))
  47         {
  48           *wc = 0xFFFD; /* L'\uFFFD' (replacement char) */
  49           replaced = true;
  50         }
  51       wc++;
  52     }
  53   return replaced;
  54 }
  55
  56 /* Truncate wchar string to width cells.
  57  * Returns number of cells used.  */
  58
  59 static size_t
  60 wc_truncate (wchar_t *wc, size_t width)
  61 {
  62   size_t cells = 0;
  63   int next_cells = 0;
  64
  65   while (*wc)
  66     {
  67       next_cells = wcwidth (*wc);
  68       if (next_cells == -1) /* non printable */
  69         {
  70           *wc = 0xFFFD; /* L'\uFFFD' (replacement char) */
  71           next_cells = 1;
  72         }
  73       if (cells + next_cells > width)
  74         break;
  75       cells += next_cells;
  76       wc++;
  77     }
  78   *wc = L'\0';
  79   return cells;
  80 }
  81
  82 /* Write N_SPACES space characters to DEST while ensuring
  83    nothing is written beyond DEST_END. A terminating NUL
  84    is always added to DEST.
  85    A pointer to the terminating NUL is returned.  */
  86
  87 static char*
  88 mbs_align_pad (char *dest, const char* dest_end, size_t n_spaces)
  89 {
  90   /* FIXME: Should we pad with "figure space" (\u2007)
  91      if non ascii data present?  */
  92   while (n_spaces-- && (dest < dest_end))
  93     *dest++ = ' ';
  94   *dest = '\0';
  95   return dest;
  96 }
  97
  98 /* Align a string, SRC, in a field of *WIDTH columns, handling multi-byte
  99    characters; write the result into the DEST_SIZE-byte buffer, DEST.
 100    ALIGNMENT specifies whether to left- or right-justify or to center.
 101    If SRC requires more than *WIDTH columns, truncate it to fit.
 102    When centering, the number of trailing spaces may be one less than the
 103    number of leading spaces.
 104    Return the length in bytes required for the final result, not counting
 105    the trailing NUL.  A return value of DEST_SIZE or larger means there
 106    wasn't enough space.  DEST will be NUL terminated in any case.
 107    Return SIZE_MAX upon error (invalid multi-byte sequence in SRC,
 108    or malloc failure), unless MBA_UNIBYTE_FALLBACK is specified.
 109    Update *WIDTH to indicate how many columns were used before padding.  */
 110
 111 size_t
 112 mbsalign (const char *src, char *dest, size_t dest_size,
 113           size_t *width, mbs_align_t align, int flags)
 114 {
 115   size_t ret = SIZE_MAX;
 116   size_t src_size = strlen (src) + 1;
 117   char *newstr = NULL;
 118   wchar_t *str_wc = NULL;
 119   const char *str_to_print = src;
 120   size_t n_cols = src_size - 1;
 121   size_t n_used_bytes = n_cols; /* Not including NUL */
 122   size_t n_spaces = 0;
 123   bool conversion = false;
 124   bool wc_enabled = false;
 125
 126   /* In multi-byte locales convert to wide characters
 127      to allow easy truncation. Also determine number
 128      of screen columns used.  */
 129   if (MB_CUR_MAX > 1)
 130     {
 131       size_t src_chars = mbstowcs (NULL, src, 0);
 132       if (src_chars == SIZE_MAX)
 133         {
 134           if (flags & MBA_UNIBYTE_FALLBACK)
 135             goto mbsalign_unibyte;
 136           else
 137             goto mbsalign_cleanup;
 138         }
 139       src_chars += 1; /* make space for NUL */
 140       str_wc = malloc (src_chars * sizeof (wchar_t));
 141       if (str_wc == NULL)
 142         {
 143           if (flags & MBA_UNIBYTE_FALLBACK)
 144             goto mbsalign_unibyte;
 145           else
 146             goto mbsalign_cleanup;
 147         }
 148       if (mbstowcs (str_wc, src, src_chars) != 0)
 149         {
 150           str_wc[src_chars - 1] = L'\0';
 151           wc_enabled = true;
 152           conversion = wc_ensure_printable (str_wc);
 153           n_cols = wcswidth (str_wc, src_chars);
 154         }
 155     }
 156
 157   /* If we transformed or need to truncate the source string
 158      then create a modified copy of it.  */
 159   if (wc_enabled && (conversion || (n_cols > *width)))
 160     {
 161         if (conversion)
 162           {
 163              /* May have increased the size by converting
 164                 \t to \uFFFD for example.  */
 165             src_size = wcstombs (NULL, str_wc, 0) + 1;
 166           }
 167         newstr = malloc (src_size);
 168         if (newstr == NULL)
 169         {
 170           if (flags & MBA_UNIBYTE_FALLBACK)
 171             goto mbsalign_unibyte;
 172           else
 173             goto mbsalign_cleanup;
 174         }
 175         str_to_print = newstr;
 176         n_cols = wc_truncate (str_wc, *width);
 177         n_used_bytes = wcstombs (newstr, str_wc, src_size);
 178     }
 179
 180 mbsalign_unibyte:
 181
 182   if (n_cols > *width) /* Unibyte truncation required.  */
 183     {
 184       n_cols = *width;
 185       n_used_bytes = n_cols;
 186     }
 187
 188   if (*width > n_cols) /* Padding required.  */
 189     n_spaces = *width - n_cols;
 190
 191   /* indicate to caller how many cells needed (not including padding).  */
 192   *width = n_cols;
 193
 194   /* indicate to caller how many bytes needed (not including NUL).  */
 195   ret = n_used_bytes + (n_spaces * 1);
 196
 197   /* Write as much NUL terminated output to DEST as possible.  */
 198   if (dest_size != 0)
 199     {
 200       size_t start_spaces, end_spaces, space_left;
 201       char *dest_end = dest + dest_size - 1;
 202
 203       switch (align)
 204         {
 205         case MBS_ALIGN_LEFT:
 206           start_spaces = 0;
 207           end_spaces = n_spaces;
 208           break;
 209         case MBS_ALIGN_RIGHT:
 210           start_spaces = n_spaces;
 211           end_spaces = 0;
 212           break;
 213         case MBS_ALIGN_CENTER:
 214         default:
 215           start_spaces = n_spaces / 2 + n_spaces % 2;
 216           end_spaces = n_spaces / 2;
 217           break;
 218         }
 219
 220       dest = mbs_align_pad (dest, dest_end, start_spaces);
 221       space_left = dest_end - dest;
 222       dest = mempcpy (dest, str_to_print, MIN (n_used_bytes, space_left));
 223       mbs_align_pad (dest, dest_end, end_spaces);
 224     }
 225
 226 mbsalign_cleanup:
 227
 228   free (str_wc);
 229   free (newstr);
 230
 231   return ret;
 232 }
 233
 234 /* A wrapper around mbsalign() to dynamically allocate the
 235    minimum amount of memory to store the result.
 236    Return NULL on failure.  */
 237
 238 char *
 239 ambsalign (const char *src, size_t *width, mbs_align_t align, int flags)
 240 {
 241   size_t orig_width = *width;
 242   size_t size = *width;         /* Start with enough for unibyte mode.  */
 243   size_t req = size;
 244   char *buf = NULL;
 245
 246   while (req >= size)
 247     {
 248       char *nbuf;
 249       size = req + 1;           /* Space for NUL.  */
 250       nbuf = realloc (buf, size);
 251       if (nbuf == NULL)
 252         {
 253           free (buf);
 254           buf = NULL;
 255           break;
 256         }
 257       buf = nbuf;
 258       *width = orig_width;
 259       req = mbsalign (src, buf, size, width, align, flags);
 260       if (req == SIZE_MAX)
 261         {
 262           free (buf);
 263           buf = NULL;
 264           break;
 265         }
 266     }
 267
 268   return buf;
 269 }