tests/uninorm/test-u32-nfd.c

   1 /* Test of canonical decomposition of UTF-32 strings.
   2    Copyright (C) 2009-2018 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>, 2009.  */
  18
  19 #include <config.h>
  20
  21 #if GNULIB_TEST_UNINORM_U32_NORMALIZE
  22
  23 #include "uninorm.h"
  24
  25 #include <signal.h>
  26 #include <stdlib.h>
  27 #include <unistd.h>
  28
  29 #include "unistr.h"
  30 #include "macros.h"
  31
  32 static int
  33 check (const uint32_t *input, size_t input_length,
  34        const uint32_t *expected, size_t expected_length)
  35 {
  36   size_t length;
  37   uint32_t *result;
  38
  39   /* Test return conventions with resultbuf == NULL.  */
  40   result = u32_normalize (UNINORM_NFD, input, input_length, NULL, &length);
  41   if (!(result != NULL))
  42     return 1;
  43   if (!(length == expected_length))
  44     return 2;
  45   if (!(u32_cmp (result, expected, expected_length) == 0))
  46     return 3;
  47   free (result);
  48
  49   /* Test return conventions with resultbuf too small.  */
  50   if (expected_length > 0)
  51     {
  52       uint32_t *preallocated;
  53
  54       length = expected_length - 1;
  55       preallocated = (uint32_t *) malloc (length * sizeof (uint32_t));
  56       result = u32_normalize (UNINORM_NFD, input, input_length, preallocated, &length);
  57       if (!(result != NULL))
  58         return 4;
  59       if (!(result != preallocated))
  60         return 5;
  61       if (!(length == expected_length))
  62         return 6;
  63       if (!(u32_cmp (result, expected, expected_length) == 0))
  64         return 7;
  65       free (result);
  66       free (preallocated);
  67     }
  68
  69   /* Test return conventions with resultbuf large enough.  */
  70   {
  71     uint32_t *preallocated;
  72
  73     length = expected_length;
  74     preallocated = (uint32_t *) malloc (length * sizeof (uint32_t));
  75     result = u32_normalize (UNINORM_NFD, input, input_length, preallocated, &length);
  76     if (!(result != NULL))
  77       return 8;
  78     if (!(preallocated == NULL || result == preallocated))
  79       return 9;
  80     if (!(length == expected_length))
  81       return 10;
  82     if (!(u32_cmp (result, expected, expected_length) == 0))
  83       return 11;
  84     free (preallocated);
  85   }
  86
  87   return 0;
  88 }
  89
  90 void
  91 test_u32_nfd (void)
  92 {
  93   { /* Empty string.  */
  94     ASSERT (check (NULL, 0, NULL, 0) == 0);
  95   }
  96   { /* SPACE */
  97     static const uint32_t input[]    = { 0x0020 };
  98     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
  99   }
 100
 101   { /* LATIN CAPITAL LETTER A WITH DIAERESIS */
 102     static const uint32_t input[]    = { 0x00C4 };
 103     static const uint32_t expected[] = { 0x0041, 0x0308 };
 104     ASSERT (check (input, SIZEOF (input), expected, SIZEOF (expected)) == 0);
 105   }
 106
 107   { /* LATIN CAPITAL LETTER A WITH DIAERESIS AND MACRON */
 108     static const uint32_t input[]    = { 0x01DE };
 109     static const uint32_t expected[] = { 0x0041, 0x0308, 0x0304 };
 110     ASSERT (check (input, SIZEOF (input), expected, SIZEOF (expected)) == 0);
 111   }
 112
 113   { /* GREEK DIALYTIKA AND PERISPOMENI */
 114     static const uint32_t input[]    = { 0x1FC1 };
 115     static const uint32_t expected[] = { 0x00A8, 0x0342 };
 116     ASSERT (check (input, SIZEOF (input), expected, SIZEOF (expected)) == 0);
 117   }
 118
 119   { /* SCRIPT SMALL L */
 120     static const uint32_t input[]    = { 0x2113 };
 121     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 122   }
 123
 124   { /* NO-BREAK SPACE */
 125     static const uint32_t input[]    = { 0x00A0 };
 126     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 127   }
 128
 129   { /* ARABIC LETTER VEH INITIAL FORM */
 130     static const uint32_t input[]    = { 0xFB6C };
 131     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 132   }
 133
 134   { /* ARABIC LETTER VEH MEDIAL FORM */
 135     static const uint32_t input[]    = { 0xFB6D };
 136     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 137   }
 138
 139   { /* ARABIC LETTER VEH FINAL FORM */
 140     static const uint32_t input[]    = { 0xFB6B };
 141     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 142   }
 143
 144   { /* ARABIC LETTER VEH ISOLATED FORM */
 145     static const uint32_t input[]    = { 0xFB6A };
 146     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 147   }
 148
 149   { /* CIRCLED NUMBER FIFTEEN */
 150     static const uint32_t input[]    = { 0x246E };
 151     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 152   }
 153
 154   { /* TRADE MARK SIGN */
 155     static const uint32_t input[]    = { 0x2122 };
 156     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 157   }
 158
 159   { /* LATIN SUBSCRIPT SMALL LETTER I */
 160     static const uint32_t input[]    = { 0x1D62 };
 161     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 162   }
 163
 164   { /* PRESENTATION FORM FOR VERTICAL LEFT PARENTHESIS */
 165     static const uint32_t input[]    = { 0xFE35 };
 166     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 167   }
 168
 169   { /* FULLWIDTH LATIN CAPITAL LETTER A */
 170     static const uint32_t input[]    = { 0xFF21 };
 171     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 172   }
 173
 174   { /* HALFWIDTH IDEOGRAPHIC COMMA */
 175     static const uint32_t input[]    = { 0xFF64 };
 176     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 177   }
 178
 179   { /* SMALL IDEOGRAPHIC COMMA */
 180     static const uint32_t input[]    = { 0xFE51 };
 181     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 182   }
 183
 184   { /* SQUARE MHZ */
 185     static const uint32_t input[]    = { 0x3392 };
 186     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 187   }
 188
 189   { /* VULGAR FRACTION THREE EIGHTHS */
 190     static const uint32_t input[]    = { 0x215C };
 191     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 192   }
 193
 194   { /* MICRO SIGN */
 195     static const uint32_t input[]    = { 0x00B5 };
 196     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 197   }
 198
 199   { /* ARABIC LIGATURE SALLALLAHOU ALAYHE WASALLAM */
 200     static const uint32_t input[]    = { 0xFDFA };
 201     ASSERT (check (input, SIZEOF (input), input, SIZEOF (input)) == 0);
 202   }
 203
 204   { /* HANGUL SYLLABLE GEUL */
 205     static const uint32_t input[]    = { 0xAE00 };
 206     static const uint32_t expected[] = { 0x1100, 0x1173, 0x11AF };
 207     ASSERT (check (input, SIZEOF (input), expected, SIZEOF (expected)) == 0);
 208   }
 209
 210   { /* HANGUL SYLLABLE GEU */
 211     static const uint32_t input[]    = { 0xADF8 };
 212     static const uint32_t expected[] = { 0x1100, 0x1173 };
 213     ASSERT (check (input, SIZEOF (input), expected, SIZEOF (expected)) == 0);
 214   }
 215
 216   { /* "Grüß Gott. Здравствуйте! x=(-b±sqrt(b²-4ac))/(2a)  日本語,中文,한글" */
 217     static const uint32_t input[] =
 218       { 'G', 'r', 0x00FC, 0x00DF, ' ', 'G', 'o', 't', 't', '.', ' ',
 219         0x0417, 0x0434, 0x0440, 0x0430, 0x0432, 0x0441, 0x0442, 0x0432, 0x0443,
 220         0x0439, 0x0442, 0x0435, '!', ' ',
 221         'x', '=', '(', '-', 'b', 0x00B1, 's', 'q', 'r', 't', '(', 'b', 0x00B2,
 222         '-', '4', 'a', 'c', ')', ')', '/', '(', '2', 'a', ')', ' ', ' ',
 223         0x65E5, 0x672C, 0x8A9E, ',', 0x4E2D, 0x6587, ',', 0xD55C, 0xAE00, '\n'
 224       };
 225     static const uint32_t expected[] =
 226       { 'G', 'r', 0x0075, 0x0308, 0x00DF, ' ', 'G', 'o', 't', 't', '.', ' ',
 227         0x0417, 0x0434, 0x0440, 0x0430, 0x0432, 0x0441, 0x0442, 0x0432, 0x0443,
 228         0x0438, 0x0306, 0x0442, 0x0435, '!', ' ',
 229         'x', '=', '(', '-', 'b', 0x00B1, 's', 'q', 'r', 't', '(', 'b', 0x00B2,
 230         '-', '4', 'a', 'c', ')', ')', '/', '(', '2', 'a', ')', ' ', ' ',
 231         0x65E5, 0x672C, 0x8A9E, ',', 0x4E2D, 0x6587, ',',
 232         0x1112, 0x1161, 0x11AB, 0x1100, 0x1173, 0x11AF, '\n'
 233       };
 234     ASSERT (check (input, SIZEOF (input), expected, SIZEOF (expected)) == 0);
 235   }
 236
 237 #if HAVE_DECL_ALARM
 238   /* Declare failure if test takes too long, by using default abort
 239      caused by SIGALRM.  */
 240   signal (SIGALRM, SIG_DFL);
 241   alarm (50);
 242 #endif
 243
 244   /* Check that the sorting is not O(n²) but O(n log n).  */
 245   {
 246     int pass;
 247     for (pass = 0; pass < 3; pass++)
 248       {
 249         size_t repeat = 1;
 250         size_t m = 100000;
 251         uint32_t *input = (uint32_t *) malloc (2 * m * sizeof (uint32_t));
 252         if (input != NULL)
 253           {
 254             uint32_t *expected = input + m;
 255             size_t m1 = m / 2;
 256             size_t m2 = (m - 1) / 2;
 257             /* NB: m1 + m2 == m - 1.  */
 258             uint32_t *p;
 259             size_t i;
 260
 261             input[0] = 0x0041;
 262             p = input + 1;
 263             switch (pass)
 264               {
 265               case 0:
 266                 for (i = 0; i < m1; i++)
 267                   *p++ = 0x0319;
 268                 for (i = 0; i < m2; i++)
 269                   *p++ = 0x0300;
 270                 break;
 271
 272               case 1:
 273                 for (i = 0; i < m2; i++)
 274                   *p++ = 0x0300;
 275                 for (i = 0; i < m1; i++)
 276                   *p++ = 0x0319;
 277                 break;
 278
 279               case 2:
 280                 for (i = 0; i < m2; i++)
 281                   {
 282                     *p++ = 0x0319;
 283                     *p++ = 0x0300;
 284                   }
 285                 for (; i < m1; i++)
 286                   *p++ = 0x0319;
 287                 break;
 288
 289               default:
 290                 abort ();
 291               }
 292
 293             expected[0] = 0x0041;
 294             p = expected + 1;
 295             for (i = 0; i < m1; i++)
 296               *p++ = 0x0319;
 297             for (i = 0; i < m2; i++)
 298               *p++ = 0x0300;
 299
 300             for (; repeat > 0; repeat--)
 301               ASSERT (check (input, m, expected, m) == 0);
 302
 303             free (input);
 304           }
 305       }
 306   }
 307 }
 308
 309 #else
 310
 311 void
 312 test_u32_nfd (void)
 313 {
 314 }
 315
 316 #endif