tests/test-mbrtowc.c

   1 /* Test of conversion of multibyte character to wide character.
   2    Copyright (C) 2008-2018 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 <wchar.h>
  22
  23 #include "signature.h"
  24 SIGNATURE_CHECK (mbrtowc, size_t, (wchar_t *, char const *, size_t,
  25                                    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   wchar_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 zero-length input.  */
  45   {
  46     memset (&state, '\0', sizeof (mbstate_t));
  47     wc = (wchar_t) 0xBADFACE;
  48     ret = mbrtowc (&wc, "x", 0, &state);
  49     ASSERT (ret == (size_t)(-2));
  50     ASSERT (mbsinit (&state));
  51   }
  52
  53   /* Test NUL byte input.  */
  54   {
  55     memset (&state, '\0', sizeof (mbstate_t));
  56     wc = (wchar_t) 0xBADFACE;
  57     ret = mbrtowc (&wc, "", 1, &state);
  58     ASSERT (ret == 0);
  59     ASSERT (wc == 0);
  60     ASSERT (mbsinit (&state));
  61     ret = mbrtowc (NULL, "", 1, &state);
  62     ASSERT (ret == 0);
  63     ASSERT (mbsinit (&state));
  64   }
  65
  66   /* Test single-byte input.  */
  67   {
  68     int c;
  69     char buf[1];
  70
  71     memset (&state, '\0', sizeof (mbstate_t));
  72     for (c = 0; c < 0x100; c++)
  73       switch (c)
  74         {
  75         default:
  76           if (! (c && 1 < argc && argv[1][0] == '5'))
  77             break;
  78           FALLTHROUGH;
  79         case '\t': case '\v': case '\f':
  80         case ' ': case '!': case '"': case '#': case '%':
  81         case '&': case '\'': case '(': case ')': case '*':
  82         case '+': case ',': case '-': case '.': case '/':
  83         case '0': case '1': case '2': case '3': case '4':
  84         case '5': case '6': case '7': case '8': case '9':
  85         case ':': case ';': case '<': case '=': case '>':
  86         case '?':
  87         case 'A': case 'B': case 'C': case 'D': case 'E':
  88         case 'F': case 'G': case 'H': case 'I': case 'J':
  89         case 'K': case 'L': case 'M': case 'N': case 'O':
  90         case 'P': case 'Q': case 'R': case 'S': case 'T':
  91         case 'U': case 'V': case 'W': case 'X': case 'Y':
  92         case 'Z':
  93         case '[': case '\\': case ']': case '^': case '_':
  94         case 'a': case 'b': case 'c': case 'd': case 'e':
  95         case 'f': case 'g': case 'h': case 'i': case 'j':
  96         case 'k': case 'l': case 'm': case 'n': case 'o':
  97         case 'p': case 'q': case 'r': case 's': case 't':
  98         case 'u': case 'v': case 'w': case 'x': case 'y':
  99         case 'z': case '{': case '|': case '}': case '~':
 100           /* c is in the ISO C "basic character set", or argv[1] starts
 101              with '5' so we are testing all nonnull bytes.  */
 102           buf[0] = c;
 103           wc = (wchar_t) 0xBADFACE;
 104           ret = mbrtowc (&wc, buf, 1, &state);
 105           ASSERT (ret == 1);
 106           if (c < 0x80)
 107             /* c is an ASCII character.  */
 108             ASSERT (wc == c);
 109           else
 110             /* argv[1] starts with '5', that is, we are testing the C or POSIX
 111                locale.
 112                On most platforms, the bytes 0x80..0xFF map to U+0080..U+00FF.
 113                But on musl libc, the bytes 0x80..0xFF map to U+DF80..U+DFFF.  */
 114             ASSERT (wc == (btowc (c) == WEOF ? c : btowc (c)));
 115           ASSERT (mbsinit (&state));
 116           ret = mbrtowc (NULL, buf, 1, &state);
 117           ASSERT (ret == 1);
 118           ASSERT (mbsinit (&state));
 119           break;
 120         }
 121   }
 122
 123   /* Test special calling convention, passing a NULL pointer.  */
 124   {
 125     memset (&state, '\0', sizeof (mbstate_t));
 126     wc = (wchar_t) 0xBADFACE;
 127     ret = mbrtowc (&wc, NULL, 5, &state);
 128     ASSERT (ret == 0);
 129     ASSERT (wc == (wchar_t) 0xBADFACE);
 130     ASSERT (mbsinit (&state));
 131   }
 132
 133   if (argc > 1)
 134     switch (argv[1][0])
 135       {
 136       case '1':
 137         /* Locale encoding is ISO-8859-1 or ISO-8859-15.  */
 138         {
 139           char input[] = "B\374\337er"; /* "Büßer" */
 140           memset (&state, '\0', sizeof (mbstate_t));
 141
 142           wc = (wchar_t) 0xBADFACE;
 143           ret = mbrtowc (&wc, input, 1, &state);
 144           ASSERT (ret == 1);
 145           ASSERT (wc == 'B');
 146           ASSERT (mbsinit (&state));
 147           input[0] = '\0';
 148
 149           wc = (wchar_t) 0xBADFACE;
 150           ret = mbrtowc (&wc, input + 1, 1, &state);
 151           ASSERT (ret == 1);
 152           ASSERT (wctob (wc) == (unsigned char) '\374');
 153           ASSERT (mbsinit (&state));
 154           input[1] = '\0';
 155
 156           /* Test support of NULL first argument.  */
 157           ret = mbrtowc (NULL, input + 2, 3, &state);
 158           ASSERT (ret == 1);
 159           ASSERT (mbsinit (&state));
 160
 161           wc = (wchar_t) 0xBADFACE;
 162           ret = mbrtowc (&wc, input + 2, 3, &state);
 163           ASSERT (ret == 1);
 164           ASSERT (wctob (wc) == (unsigned char) '\337');
 165           ASSERT (mbsinit (&state));
 166           input[2] = '\0';
 167
 168           wc = (wchar_t) 0xBADFACE;
 169           ret = mbrtowc (&wc, input + 3, 2, &state);
 170           ASSERT (ret == 1);
 171           ASSERT (wc == 'e');
 172           ASSERT (mbsinit (&state));
 173           input[3] = '\0';
 174
 175           wc = (wchar_t) 0xBADFACE;
 176           ret = mbrtowc (&wc, input + 4, 1, &state);
 177           ASSERT (ret == 1);
 178           ASSERT (wc == 'r');
 179           ASSERT (mbsinit (&state));
 180         }
 181         return 0;
 182
 183       case '2':
 184         /* Locale encoding is UTF-8.  */
 185         {
 186           char input[] = "B\303\274\303\237er"; /* "Büßer" */
 187           memset (&state, '\0', sizeof (mbstate_t));
 188
 189           wc = (wchar_t) 0xBADFACE;
 190           ret = mbrtowc (&wc, input, 1, &state);
 191           ASSERT (ret == 1);
 192           ASSERT (wc == 'B');
 193           ASSERT (mbsinit (&state));
 194           input[0] = '\0';
 195
 196           wc = (wchar_t) 0xBADFACE;
 197           ret = mbrtowc (&wc, input + 1, 1, &state);
 198           ASSERT (ret == (size_t)(-2));
 199           ASSERT (wc == (wchar_t) 0xBADFACE);
 200           ASSERT (!mbsinit (&state));
 201           input[1] = '\0';
 202
 203           wc = (wchar_t) 0xBADFACE;
 204           ret = mbrtowc (&wc, input + 2, 5, &state);
 205           ASSERT (ret == 1);
 206           ASSERT (wctob (wc) == EOF);
 207           ASSERT (mbsinit (&state));
 208           input[2] = '\0';
 209
 210           /* Test support of NULL first argument.  */
 211           ret = mbrtowc (NULL, input + 3, 4, &state);
 212           ASSERT (ret == 2);
 213           ASSERT (mbsinit (&state));
 214
 215           wc = (wchar_t) 0xBADFACE;
 216           ret = mbrtowc (&wc, input + 3, 4, &state);
 217           ASSERT (ret == 2);
 218           ASSERT (wctob (wc) == EOF);
 219           ASSERT (mbsinit (&state));
 220           input[3] = '\0';
 221           input[4] = '\0';
 222
 223           wc = (wchar_t) 0xBADFACE;
 224           ret = mbrtowc (&wc, input + 5, 2, &state);
 225           ASSERT (ret == 1);
 226           ASSERT (wc == 'e');
 227           ASSERT (mbsinit (&state));
 228           input[5] = '\0';
 229
 230           wc = (wchar_t) 0xBADFACE;
 231           ret = mbrtowc (&wc, input + 6, 1, &state);
 232           ASSERT (ret == 1);
 233           ASSERT (wc == 'r');
 234           ASSERT (mbsinit (&state));
 235         }
 236         return 0;
 237
 238       case '3':
 239         /* Locale encoding is EUC-JP.  */
 240         {
 241           char input[] = "<\306\374\313\334\270\354>"; /* "<日本語>" */
 242           memset (&state, '\0', sizeof (mbstate_t));
 243
 244           wc = (wchar_t) 0xBADFACE;
 245           ret = mbrtowc (&wc, input, 1, &state);
 246           ASSERT (ret == 1);
 247           ASSERT (wc == '<');
 248           ASSERT (mbsinit (&state));
 249           input[0] = '\0';
 250
 251           wc = (wchar_t) 0xBADFACE;
 252           ret = mbrtowc (&wc, input + 1, 2, &state);
 253           ASSERT (ret == 2);
 254           ASSERT (wctob (wc) == EOF);
 255           ASSERT (mbsinit (&state));
 256           input[1] = '\0';
 257           input[2] = '\0';
 258
 259           wc = (wchar_t) 0xBADFACE;
 260           ret = mbrtowc (&wc, input + 3, 1, &state);
 261           ASSERT (ret == (size_t)(-2));
 262           ASSERT (wc == (wchar_t) 0xBADFACE);
 263           ASSERT (!mbsinit (&state));
 264           input[3] = '\0';
 265
 266           wc = (wchar_t) 0xBADFACE;
 267           ret = mbrtowc (&wc, input + 4, 4, &state);
 268           ASSERT (ret == 1);
 269           ASSERT (wctob (wc) == EOF);
 270           ASSERT (mbsinit (&state));
 271           input[4] = '\0';
 272
 273           /* Test support of NULL first argument.  */
 274           ret = mbrtowc (NULL, input + 5, 3, &state);
 275           ASSERT (ret == 2);
 276           ASSERT (mbsinit (&state));
 277
 278           wc = (wchar_t) 0xBADFACE;
 279           ret = mbrtowc (&wc, input + 5, 3, &state);
 280           ASSERT (ret == 2);
 281           ASSERT (wctob (wc) == EOF);
 282           ASSERT (mbsinit (&state));
 283           input[5] = '\0';
 284           input[6] = '\0';
 285
 286           wc = (wchar_t) 0xBADFACE;
 287           ret = mbrtowc (&wc, input + 7, 1, &state);
 288           ASSERT (ret == 1);
 289           ASSERT (wc == '>');
 290           ASSERT (mbsinit (&state));
 291         }
 292         return 0;
 293
 294       case '4':
 295         /* Locale encoding is GB18030.  */
 296         {
 297           char input[] = "B\250\271\201\060\211\070er"; /* "Büßer" */
 298           memset (&state, '\0', sizeof (mbstate_t));
 299
 300           wc = (wchar_t) 0xBADFACE;
 301           ret = mbrtowc (&wc, input, 1, &state);
 302           ASSERT (ret == 1);
 303           ASSERT (wc == 'B');
 304           ASSERT (mbsinit (&state));
 305           input[0] = '\0';
 306
 307           wc = (wchar_t) 0xBADFACE;
 308           ret = mbrtowc (&wc, input + 1, 1, &state);
 309           ASSERT (ret == (size_t)(-2));
 310           ASSERT (wc == (wchar_t) 0xBADFACE);
 311           ASSERT (!mbsinit (&state));
 312           input[1] = '\0';
 313
 314           wc = (wchar_t) 0xBADFACE;
 315           ret = mbrtowc (&wc, input + 2, 7, &state);
 316           ASSERT (ret == 1);
 317           ASSERT (wctob (wc) == EOF);
 318           ASSERT (mbsinit (&state));
 319           input[2] = '\0';
 320
 321           /* Test support of NULL first argument.  */
 322           ret = mbrtowc (NULL, input + 3, 6, &state);
 323           ASSERT (ret == 4);
 324           ASSERT (mbsinit (&state));
 325
 326           wc = (wchar_t) 0xBADFACE;
 327           ret = mbrtowc (&wc, input + 3, 6, &state);
 328           ASSERT (ret == 4);
 329           ASSERT (wctob (wc) == EOF);
 330           ASSERT (mbsinit (&state));
 331           input[3] = '\0';
 332           input[4] = '\0';
 333           input[5] = '\0';
 334           input[6] = '\0';
 335
 336           wc = (wchar_t) 0xBADFACE;
 337           ret = mbrtowc (&wc, input + 7, 2, &state);
 338           ASSERT (ret == 1);
 339           ASSERT (wc == 'e');
 340           ASSERT (mbsinit (&state));
 341           input[5] = '\0';
 342
 343           wc = (wchar_t) 0xBADFACE;
 344           ret = mbrtowc (&wc, input + 8, 1, &state);
 345           ASSERT (ret == 1);
 346           ASSERT (wc == 'r');
 347           ASSERT (mbsinit (&state));
 348         }
 349         return 0;
 350
 351       case '5':
 352         /* C locale; tested above.  */
 353         return 0;
 354       }
 355
 356   return 1;
 357 }