lib/unilbrk/ulc-possible-linebreaks.c

   1 /* Line breaking of strings.
   2    Copyright (C) 2001-2003, 2006-2020 Free Software Foundation, Inc.
   3    Written by Bruno Haible <bruno@clisp.org>, 2001.
   4
   5    This program is free software: you can redistribute it and/or modify it
   6    under the terms of the GNU Lesser General Public License as published
   7    by the Free Software Foundation; either version 3 of the License, or
   8    (at your option) any later version.
   9
  10    This program is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13    Lesser General Public License for more details.
  14
  15    You should have received a copy of the GNU Lesser General Public License
  16    along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
  17
  18 #include <config.h>
  19
  20 /* Specification.  */
  21 #include "unilbrk.h"
  22
  23 #include <stdlib.h>
  24 #include <string.h>
  25
  26 #include "c-ctype.h"
  27 #include "uniconv.h"
  28 #include "unilbrk/ulc-common.h"
  29
  30 /* Line breaking of a string in an arbitrary encoding.
  31
  32    We convert the input string to Unicode.
  33
  34    The standardized Unicode encodings are UTF-8, UCS-2, UCS-4, UTF-16,
  35    UTF-16BE, UTF-16LE, UTF-7.  UCS-2 supports only characters up to
  36    \U0000FFFF.  UTF-16 and variants support only characters up to
  37    \U0010FFFF.  UTF-7 is way too complex and not supported by glibc-2.1.
  38    UCS-4 specification leaves doubts about endianness and byte order mark.
  39    glibc currently interprets it as big endian without byte order mark,
  40    but this is not backed by an RFC.  So we use UTF-8. It supports
  41    characters up to \U7FFFFFFF and is unambiguously defined.  */
  42
  43 void
  44 ulc_possible_linebreaks (const char *s, size_t n, const char *encoding,
  45                          char *p)
  46 {
  47   if (n > 0)
  48     {
  49       if (is_utf8_encoding (encoding))
  50         u8_possible_linebreaks ((const uint8_t *) s, n, encoding, p);
  51       else
  52         {
  53           /* Convert the string to UTF-8 and build a translation table
  54              from offsets into s to offsets into the translated string.  */
  55           size_t *offsets = (size_t *) malloc (n * sizeof (size_t));
  56
  57           if (offsets != NULL)
  58             {
  59               uint8_t *t;
  60               size_t m;
  61
  62               t = u8_conv_from_encoding (encoding, iconveh_question_mark,
  63                                          s, n, offsets, NULL, &m);
  64               if (t != NULL)
  65                 {
  66                   char *q = (char *) (m > 0 ? malloc (m) : NULL);
  67
  68                   if (m == 0 || q != NULL)
  69                     {
  70                       size_t i;
  71
  72                       /* Determine the possible line breaks of the UTF-8
  73                          string.  */
  74                       u8_possible_linebreaks (t, m, encoding, q);
  75
  76                       /* Translate the result back to the original string.  */
  77                       memset (p, UC_BREAK_PROHIBITED, n);
  78                       for (i = 0; i < n; i++)
  79                         if (offsets[i] != (size_t)(-1))
  80                           p[i] = q[offsets[i]];
  81
  82                       free (q);
  83                       free (t);
  84                       free (offsets);
  85                       return;
  86                     }
  87                   free (t);
  88                 }
  89               free (offsets);
  90             }
  91
  92           /* Impossible to convert.  */
  93 #if C_CTYPE_ASCII
  94           if (is_all_ascii (s, n))
  95             {
  96               /* ASCII is a subset of UTF-8.  */
  97               u8_possible_linebreaks ((const uint8_t *) s, n, encoding, p);
  98               return;
  99             }
 100 #endif
 101           /* We have a non-ASCII string and cannot convert it.
 102              Don't produce line breaks except those already present in the
 103              input string.  All we assume here is that the encoding is
 104              minimally ASCII compatible.  */
 105           {
 106             const char *s_end = s + n;
 107             while (s < s_end)
 108               {
 109                 *p = (*s == '\n' ? UC_BREAK_MANDATORY : UC_BREAK_PROHIBITED);
 110                 s++;
 111                 p++;
 112               }
 113           }
 114         }
 115     }
 116 }
 117
 118
 119 #ifdef TEST
 120
 121 #include <stdio.h>
 122 #include <locale.h>
 123 #include <string.h>
 124
 125 /* Read the contents of an input stream, and return it, terminated with a NUL
 126    byte. */
 127 char *
 128 read_file (FILE *stream)
 129 {
 130 #define BUFSIZE 4096
 131   char *buf = NULL;
 132   int alloc = 0;
 133   int size = 0;
 134   int count;
 135
 136   while (! feof (stream))
 137     {
 138       if (size + BUFSIZE > alloc)
 139         {
 140           alloc = alloc + alloc / 2;
 141           if (alloc < size + BUFSIZE)
 142             alloc = size + BUFSIZE;
 143           buf = realloc (buf, alloc);
 144           if (buf == NULL)
 145             {
 146               fprintf (stderr, "out of memory\n");
 147               exit (1);
 148             }
 149         }
 150       count = fread (buf + size, 1, BUFSIZE, stream);
 151       if (count == 0)
 152         {
 153           if (ferror (stream))
 154             {
 155               perror ("fread");
 156               exit (1);
 157             }
 158         }
 159       else
 160         size += count;
 161     }
 162   buf = realloc (buf, size + 1);
 163   if (buf == NULL)
 164     {
 165       fprintf (stderr, "out of memory\n");
 166       exit (1);
 167     }
 168   buf[size] = '\0';
 169   return buf;
 170 #undef BUFSIZE
 171 }
 172
 173 int
 174 main (int argc, char * argv[])
 175 {
 176   setlocale (LC_CTYPE, "");
 177   if (argc == 1)
 178     {
 179       /* Display all the break opportunities in the input string.  */
 180       char *input = read_file (stdin);
 181       int length = strlen (input);
 182       char *breaks = malloc (length);
 183       int i;
 184
 185       ulc_possible_linebreaks (input, length, locale_charset (), breaks);
 186
 187       for (i = 0; i < length; i++)
 188         {
 189           switch (breaks[i])
 190             {
 191             case UC_BREAK_POSSIBLE:
 192               putc ('|', stdout);
 193               break;
 194             case UC_BREAK_MANDATORY:
 195               break;
 196             case UC_BREAK_PROHIBITED:
 197               break;
 198             default:
 199               abort ();
 200             }
 201           putc (input[i], stdout);
 202         }
 203
 204       free (breaks);
 205
 206       return 0;
 207     }
 208   else
 209     return 1;
 210 }
 211
 212 #endif /* TEST */