tests/test-mbsrtoc32s.c

   1 /* Test of conversion of string to 32-bit wide string.
   2    Copyright (C) 2008-2021 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 <https://www.gnu.org/licenses/>.  */
  16
  17 /* Written by Bruno Haible <bruno@clisp.org>, 2008.  */
  18
  19 #include <config.h>
  20
  21 #include <uchar.h>
  22
  23 #include "signature.h"
  24 SIGNATURE_CHECK (mbsrtoc32s, size_t,
  25                  (char32_t *, const char **, size_t, mbstate_t *));
  26
  27 #include <locale.h>
  28 #include <stdio.h>
  29 #include <string.h>
  30
  31 #include "macros.h"
  32
  33 int
  34 main (int argc, char *argv[])
  35 {
  36   mbstate_t state;
  37   char32_t wc;
  38   size_t ret;
  39
  40   /* configure should already have checked that the locale is supported.  */
  41   if (setlocale (LC_ALL, "") == NULL)
  42     return 1;
  43
  44   /* Test NUL byte input.  */
  45   {
  46     const char *src;
  47
  48     memset (&state, '\0', sizeof (mbstate_t));
  49
  50     src = "";
  51     ret = mbsrtoc32s (NULL, &src, 0, &state);
  52     ASSERT (ret == 0);
  53     ASSERT (mbsinit (&state));
  54
  55     src = "";
  56     ret = mbsrtoc32s (NULL, &src, 1, &state);
  57     ASSERT (ret == 0);
  58     ASSERT (mbsinit (&state));
  59
  60     wc = (char32_t) 0xBADFACE;
  61     src = "";
  62     ret = mbsrtoc32s (&wc, &src, 0, &state);
  63     ASSERT (ret == 0);
  64     ASSERT (wc == (char32_t) 0xBADFACE);
  65     ASSERT (mbsinit (&state));
  66
  67     wc = (char32_t) 0xBADFACE;
  68     src = "";
  69     ret = mbsrtoc32s (&wc, &src, 1, &state);
  70     ASSERT (ret == 0);
  71     ASSERT (wc == 0);
  72     ASSERT (mbsinit (&state));
  73   }
  74
  75   if (argc > 1)
  76     {
  77       int unlimited;
  78
  79       for (unlimited = 0; unlimited < 2; unlimited++)
  80         {
  81           #define BUFSIZE 10
  82           char32_t buf[BUFSIZE];
  83           const char *src;
  84           mbstate_t temp_state;
  85
  86           {
  87             size_t i;
  88             for (i = 0; i < BUFSIZE; i++)
  89               buf[i] = (char32_t) 0xBADFACE;
  90           }
  91
  92           switch (argv[1][0])
  93             {
  94             case '1':
  95               /* Locale encoding is ISO-8859-1 or ISO-8859-15.  */
  96               {
  97                 char input[] = "B\374\337er"; /* "Büßer" */
  98                 memset (&state, '\0', sizeof (mbstate_t));
  99
 100                 wc = (char32_t) 0xBADFACE;
 101                 ret = mbrtoc32 (&wc, input, 1, &state);
 102                 ASSERT (ret == 1);
 103                 ASSERT (wc == 'B');
 104                 ASSERT (mbsinit (&state));
 105                 input[0] = '\0';
 106
 107                 wc = (char32_t) 0xBADFACE;
 108                 ret = mbrtoc32 (&wc, input + 1, 1, &state);
 109                 ASSERT (ret == 1);
 110                 ASSERT (c32tob (wc) == (unsigned char) '\374');
 111                 ASSERT (mbsinit (&state));
 112                 input[1] = '\0';
 113
 114                 src = input + 2;
 115                 temp_state = state;
 116                 ret = mbsrtoc32s (NULL, &src, unlimited ? BUFSIZE : 1, &temp_state);
 117                 ASSERT (ret == 3);
 118                 ASSERT (src == input + 2);
 119                 ASSERT (mbsinit (&state));
 120
 121                 src = input + 2;
 122                 ret = mbsrtoc32s (buf, &src, unlimited ? BUFSIZE : 1, &state);
 123                 ASSERT (ret == (unlimited ? 3 : 1));
 124                 ASSERT (src == (unlimited ? NULL : input + 3));
 125                 ASSERT (c32tob (buf[0]) == (unsigned char) '\337');
 126                 if (unlimited)
 127                   {
 128                     ASSERT (buf[1] == 'e');
 129                     ASSERT (buf[2] == 'r');
 130                     ASSERT (buf[3] == 0);
 131                     ASSERT (buf[4] == (char32_t) 0xBADFACE);
 132                   }
 133                 else
 134                   ASSERT (buf[1] == (char32_t) 0xBADFACE);
 135                 ASSERT (mbsinit (&state));
 136               }
 137               break;
 138
 139             case '2':
 140               /* Locale encoding is UTF-8.  */
 141               {
 142                 char input[] = "s\303\274\303\237\360\237\230\213!"; /* "süß😋!" */
 143                 memset (&state, '\0', sizeof (mbstate_t));
 144
 145                 wc = (char32_t) 0xBADFACE;
 146                 ret = mbrtoc32 (&wc, input, 1, &state);
 147                 ASSERT (ret == 1);
 148                 ASSERT (wc == 's');
 149                 ASSERT (mbsinit (&state));
 150                 input[0] = '\0';
 151
 152                 wc = (char32_t) 0xBADFACE;
 153                 ret = mbrtoc32 (&wc, input + 1, 1, &state);
 154                 ASSERT (ret == (size_t)(-2));
 155                 ASSERT (wc == (char32_t) 0xBADFACE);
 156                 ASSERT (!mbsinit (&state));
 157                 input[1] = '\0';
 158
 159                 src = input + 2;
 160                 temp_state = state;
 161                 ret = mbsrtoc32s (NULL, &src, unlimited ? BUFSIZE : 2, &temp_state);
 162                 ASSERT (ret == 4);
 163                 ASSERT (src == input + 2);
 164                 ASSERT (!mbsinit (&state));
 165
 166                 src = input + 2;
 167                 ret = mbsrtoc32s (buf, &src, unlimited ? BUFSIZE : 2, &state);
 168                 ASSERT (ret == (unlimited ? 4 : 2));
 169                 ASSERT (src == (unlimited ? NULL : input + 5));
 170                 ASSERT (c32tob (buf[0]) == EOF);
 171                 ASSERT (c32tob (buf[1]) == EOF);
 172                 if (unlimited)
 173                   {
 174                     ASSERT (buf[2] == 0x1F60B); /* expect Unicode encoding */
 175                     ASSERT (buf[3] == '!');
 176                     ASSERT (buf[4] == 0);
 177                     ASSERT (buf[5] == (char32_t) 0xBADFACE);
 178                   }
 179                 else
 180                   ASSERT (buf[2] == (char32_t) 0xBADFACE);
 181                 ASSERT (mbsinit (&state));
 182               }
 183               break;
 184
 185             case '3':
 186               /* Locale encoding is EUC-JP.  */
 187               {
 188                 char input[] = "<\306\374\313\334\270\354>"; /* "<日本語>" */
 189                 memset (&state, '\0', sizeof (mbstate_t));
 190
 191                 wc = (char32_t) 0xBADFACE;
 192                 ret = mbrtoc32 (&wc, input, 1, &state);
 193                 ASSERT (ret == 1);
 194                 ASSERT (wc == '<');
 195                 ASSERT (mbsinit (&state));
 196                 input[0] = '\0';
 197
 198                 wc = (char32_t) 0xBADFACE;
 199                 ret = mbrtoc32 (&wc, input + 1, 2, &state);
 200                 ASSERT (ret == 2);
 201                 ASSERT (c32tob (wc) == EOF);
 202                 ASSERT (mbsinit (&state));
 203                 input[1] = '\0';
 204                 input[2] = '\0';
 205
 206                 wc = (char32_t) 0xBADFACE;
 207                 ret = mbrtoc32 (&wc, input + 3, 1, &state);
 208                 ASSERT (ret == (size_t)(-2));
 209                 ASSERT (wc == (char32_t) 0xBADFACE);
 210                 ASSERT (!mbsinit (&state));
 211                 input[3] = '\0';
 212
 213                 src = input + 4;
 214                 temp_state = state;
 215                 ret = mbsrtoc32s (NULL, &src, unlimited ? BUFSIZE : 2, &temp_state);
 216                 ASSERT (ret == 3);
 217                 ASSERT (src == input + 4);
 218                 ASSERT (!mbsinit (&state));
 219
 220                 src = input + 4;
 221                 ret = mbsrtoc32s (buf, &src, unlimited ? BUFSIZE : 2, &state);
 222                 ASSERT (ret == (unlimited ? 3 : 2));
 223                 ASSERT (src == (unlimited ? NULL : input + 7));
 224                 ASSERT (c32tob (buf[0]) == EOF);
 225                 ASSERT (c32tob (buf[1]) == EOF);
 226                 if (unlimited)
 227                   {
 228                     ASSERT (buf[2] == '>');
 229                     ASSERT (buf[3] == 0);
 230                     ASSERT (buf[4] == (char32_t) 0xBADFACE);
 231                   }
 232                 else
 233                   ASSERT (buf[2] == (char32_t) 0xBADFACE);
 234                 ASSERT (mbsinit (&state));
 235               }
 236               break;
 237
 238             case '4':
 239               /* Locale encoding is GB18030.  */
 240               {
 241                 char input[] = "s\250\271\201\060\211\070\224\071\375\067!"; /* "süß😋!" */
 242                 memset (&state, '\0', sizeof (mbstate_t));
 243
 244                 wc = (char32_t) 0xBADFACE;
 245                 ret = mbrtoc32 (&wc, input, 1, &state);
 246                 ASSERT (ret == 1);
 247                 ASSERT (wc == 's');
 248                 ASSERT (mbsinit (&state));
 249                 input[0] = '\0';
 250
 251                 wc = (char32_t) 0xBADFACE;
 252                 ret = mbrtoc32 (&wc, input + 1, 1, &state);
 253                 ASSERT (ret == (size_t)(-2));
 254                 ASSERT (wc == (char32_t) 0xBADFACE);
 255                 ASSERT (!mbsinit (&state));
 256                 input[1] = '\0';
 257
 258                 src = input + 2;
 259                 temp_state = state;
 260                 ret = mbsrtoc32s (NULL, &src, unlimited ? BUFSIZE : 2, &temp_state);
 261                 ASSERT (ret == 4);
 262                 ASSERT (src == input + 2);
 263                 ASSERT (!mbsinit (&state));
 264
 265                 src = input + 2;
 266                 ret = mbsrtoc32s (buf, &src, unlimited ? BUFSIZE : 2, &state);
 267                 ASSERT (ret == (unlimited ? 4 : 2));
 268                 ASSERT (src == (unlimited ? NULL : input + 7));
 269                 ASSERT (c32tob (buf[0]) == EOF);
 270                 ASSERT (c32tob (buf[1]) == EOF);
 271                 if (unlimited)
 272                   {
 273                     ASSERT (c32tob (buf[2]) == EOF);
 274                     ASSERT (buf[3] == '!');
 275                     ASSERT (buf[4] == 0);
 276                     ASSERT (buf[5] == (char32_t) 0xBADFACE);
 277                   }
 278                 else
 279                   ASSERT (buf[2] == (char32_t) 0xBADFACE);
 280                 ASSERT (mbsinit (&state));
 281               }
 282               break;
 283
 284             default:
 285               return 1;
 286             }
 287         }
 288
 289       return 0;
 290     }
 291
 292   return 1;
 293 }