lib/uniwbrk/u8-wordbreaks.c

   1 /* Word breaks in UTF-8 strings.
   2    Copyright (C) 2009-2024 Free Software Foundation, Inc.
   3    Written by Bruno Haible <bruno@clisp.org>, 2009.
   4
   5    This file is free software.
   6    It is dual-licensed under "the GNU LGPLv3+ or the GNU GPLv2+".
   7    You can redistribute it and/or modify it under either
   8      - the terms of the GNU Lesser General Public License as published
   9        by the Free Software Foundation, either version 3, or (at your
  10        option) any later version, or
  11      - the terms of the GNU General Public License as published by the
  12        Free Software Foundation; either version 2, or (at your option)
  13        any later version, or
  14      - the same dual license "the GNU LGPLv3+ or the GNU GPLv2+".
  15
  16    This file is distributed in the hope that it will be useful,
  17    but WITHOUT ANY WARRANTY; without even the implied warranty of
  18    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  19    Lesser General Public License and the GNU General Public License
  20    for more details.
  21
  22    You should have received a copy of the GNU Lesser General Public
  23    License and of the GNU General Public License along with this
  24    program.  If not, see <https://www.gnu.org/licenses/>.  */
  25
  26 #include <config.h>
  27
  28 /* Specification.  */
  29 #include "uniwbrk.h"
  30
  31 #include <string.h>
  32
  33 #include "unictype.h"
  34 #include "unistr.h"
  35 #include "uniwbrk/wbrktable.h"
  36
  37 #define FUNC u8_wordbreaks
  38 #define UNIT uint8_t
  39 #define U_MBTOUC_UNSAFE u8_mbtouc_unsafe
  40 #include "u-wordbreaks.h"
  41
  42
  43 #ifdef TEST
  44
  45 #include <stdio.h>
  46 #include <stdlib.h>
  47
  48 /* Read the contents of an input stream, and return it, terminated with a NUL
  49    byte. */
  50 char *
  51 read_file (FILE *stream)
  52 {
  53 #define BUFSIZE 4096
  54   char *buf = NULL;
  55   int alloc = 0;
  56   int size = 0;
  57   int count;
  58
  59   while (! feof (stream))
  60     {
  61       if (size + BUFSIZE > alloc)
  62         {
  63           alloc = alloc + alloc / 2;
  64           if (alloc < size + BUFSIZE)
  65             alloc = size + BUFSIZE;
  66           buf = realloc (buf, alloc);
  67           if (buf == NULL)
  68             {
  69               fprintf (stderr, "out of memory\n");
  70               exit (1);
  71             }
  72         }
  73       count = fread (buf + size, 1, BUFSIZE, stream);
  74       if (count == 0)
  75         {
  76           if (ferror (stream))
  77             {
  78               perror ("fread");
  79               exit (1);
  80             }
  81         }
  82       else
  83         size += count;
  84     }
  85   buf = realloc (buf, size + 1);
  86   if (buf == NULL)
  87     {
  88       fprintf (stderr, "out of memory\n");
  89       exit (1);
  90     }
  91   buf[size] = '\0';
  92   return buf;
  93 #undef BUFSIZE
  94 }
  95
  96 int
  97 main (int argc, char * argv[])
  98 {
  99   if (argc == 1)
 100     {
 101       /* Display all the word breaks in the input string.  */
 102       char *input = read_file (stdin);
 103       int length = strlen (input);
 104       char *breaks = malloc (length);
 105       int i;
 106
 107       u8_wordbreaks ((uint8_t *) input, length, breaks);
 108
 109       for (i = 0; i < length; i++)
 110         {
 111           switch (breaks[i])
 112             {
 113             case 1:
 114               /* U+2027 in UTF-8 encoding */
 115               putc (0xe2, stdout); putc (0x80, stdout); putc (0xa7, stdout);
 116               break;
 117             case 0:
 118               break;
 119             default:
 120               abort ();
 121             }
 122           putc (input[i], stdout);
 123         }
 124
 125       free (breaks);
 126
 127       return 0;
 128     }
 129   else
 130     return 1;
 131 }
 132
 133 #endif /* TEST */