tests/unicase/test-u32-casefold.c

   1 /* Test of casefolding mapping for UTF-32 strings.
   2    Copyright (C) 2009-2020 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 #include "unicase.h"
  22
  23 #include <stdlib.h>
  24
  25 #include "unistr.h"
  26 #include "uninorm.h"
  27 #include "macros.h"
  28
  29 static int
  30 check (const uint32_t *input, size_t input_length,
  31        const char *iso639_language, uninorm_t nf,
  32        const uint32_t *expected, size_t expected_length)
  33 {
  34   size_t length;
  35   uint32_t *result;
  36
  37   /* Test return conventions with resultbuf == NULL.  */
  38   result = u32_casefold (input, input_length, iso639_language, nf, NULL, &length);
  39   if (!(result != NULL))
  40     return 1;
  41   if (!(length == expected_length))
  42     return 2;
  43   if (!(u32_cmp (result, expected, expected_length) == 0))
  44     return 3;
  45   free (result);
  46
  47   /* Test return conventions with resultbuf too small.  */
  48   if (expected_length > 0)
  49     {
  50       uint32_t *preallocated;
  51
  52       length = expected_length - 1;
  53       preallocated = (uint32_t *) malloc (length * sizeof (uint32_t));
  54       result = u32_casefold (input, input_length, iso639_language, nf, preallocated, &length);
  55       if (!(result != NULL))
  56         return 4;
  57       if (!(result != preallocated))
  58         return 5;
  59       if (!(length == expected_length))
  60         return 6;
  61       if (!(u32_cmp (result, expected, expected_length) == 0))
  62         return 7;
  63       free (result);
  64       free (preallocated);
  65     }
  66
  67   /* Test return conventions with resultbuf large enough.  */
  68   {
  69     uint32_t *preallocated;
  70
  71     length = expected_length;
  72     preallocated = (uint32_t *) malloc (length * sizeof (uint32_t));
  73     result = u32_casefold (input, input_length, iso639_language, nf, preallocated, &length);
  74     if (!(result != NULL))
  75       return 8;
  76     if (!(preallocated == NULL || result == preallocated))
  77       return 9;
  78     if (!(length == expected_length))
  79       return 10;
  80     if (!(u32_cmp (result, expected, expected_length) == 0))
  81       return 11;
  82     free (preallocated);
  83   }
  84
  85   return 0;
  86 }
  87
  88 int
  89 main ()
  90 {
  91   { /* Empty string.  */
  92     ASSERT (check (NULL, 0, NULL, NULL, NULL, 0) == 0);
  93     ASSERT (check (NULL, 0, NULL, UNINORM_NFC, NULL, 0) == 0);
  94   }
  95
  96   /* Simple string.  */
  97   { /* "Grüß Gott. Здравствуйте! x=(-b±sqrt(b²-4ac))/(2a)  日本語,中文,한글" */
  98     static const uint32_t input[] =
  99       { 'G', 'r', 0x00FC, 0x00DF, ' ', 'G', 'o', 't', 't', '.', ' ',
 100         0x0417, 0x0434, 0x0440, 0x0430, 0x0432, 0x0441, 0x0442, 0x0432, 0x0443,
 101         0x0439, 0x0442, 0x0435, '!', ' ',
 102         'x', '=', '(', '-', 'b', 0x00B1, 's', 'q', 'r', 't', '(', 'b', 0x00B2,
 103         '-', '4', 'a', 'c', ')', ')', '/', '(', '2', 'a', ')', ' ', ' ',
 104         0x65E5, 0x672C, 0x8A9E, ',', 0x4E2D, 0x6587, ',', 0xD55C, 0xAE00, '\n'
 105       };
 106     static const uint32_t casefolded[] =
 107       { 'g', 'r', 0x00FC, 0x0073, 0x0073, ' ', 'g', 'o', 't', 't', '.', ' ',
 108         0x0437, 0x0434, 0x0440, 0x0430, 0x0432, 0x0441, 0x0442, 0x0432, 0x0443,
 109         0x0439, 0x0442, 0x0435, '!', ' ',
 110         'x', '=', '(', '-', 'b', 0x00B1, 's', 'q', 'r', 't', '(', 'b', 0x00B2,
 111         '-', '4', 'a', 'c', ')', ')', '/', '(', '2', 'a', ')', ' ', ' ',
 112         0x65E5, 0x672C, 0x8A9E, ',', 0x4E2D, 0x6587, ',', 0xD55C, 0xAE00, '\n'
 113       };
 114     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 115   }
 116
 117   /* Case mapping can increase the number of Unicode characters.  */
 118   { /* LATIN SMALL LETTER N PRECEDED BY APOSTROPHE */
 119     static const uint32_t input[]      = { 0x0149 };
 120     static const uint32_t casefolded[] = { 0x02BC, 0x006E };
 121     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 122   }
 123   { /* GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS */
 124     static const uint32_t input[]      = { 0x0390 };
 125     static const uint32_t casefolded[] = { 0x03B9, 0x0308, 0x0301 };
 126     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 127   }
 128
 129   /* Turkish letters i İ ı I */
 130   { /* LATIN CAPITAL LETTER I */
 131     static const uint32_t input[]         = { 0x0049 };
 132     static const uint32_t casefolded[]    = { 0x0069 };
 133     static const uint32_t casefolded_tr[] = { 0x0131 };
 134     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 135     ASSERT (check (input, SIZEOF (input), "tr", NULL, casefolded_tr, SIZEOF (casefolded_tr)) == 0);
 136   }
 137   { /* LATIN SMALL LETTER I */
 138     static const uint32_t input[]      = { 0x0069 };
 139     static const uint32_t casefolded[] = { 0x0069 };
 140     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 141     ASSERT (check (input, SIZEOF (input), "tr", NULL, casefolded, SIZEOF (casefolded)) == 0);
 142   }
 143   { /* LATIN CAPITAL LETTER I WITH DOT ABOVE */
 144     static const uint32_t input[]         = { 0x0130 };
 145     static const uint32_t casefolded[]    = { 0x0069, 0x0307 };
 146     static const uint32_t casefolded_tr[] = { 0x0069 };
 147     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 148     ASSERT (check (input, SIZEOF (input), "tr", NULL, casefolded_tr, SIZEOF (casefolded_tr)) == 0);
 149   }
 150   { /* LATIN SMALL LETTER DOTLESS I */
 151     static const uint32_t input[]      = { 0x0131 };
 152     static const uint32_t casefolded[] = { 0x0131 };
 153     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 154     ASSERT (check (input, SIZEOF (input), "tr", NULL, casefolded, SIZEOF (casefolded)) == 0);
 155   }
 156   { /* "topkapı" */
 157     static const uint32_t input[] =
 158       { 0x0074, 0x006F, 0x0070, 0x006B, 0x0061, 0x0070, 0x0131 };
 159     static const uint32_t casefolded[] =
 160       { 0x0074, 0x006F, 0x0070, 0x006B, 0x0061, 0x0070, 0x0131 };
 161     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 162     ASSERT (check (input, SIZEOF (input), "tr", NULL, casefolded, SIZEOF (casefolded)) == 0);
 163   }
 164
 165   /* Uppercasing can increase the number of Unicode characters.  */
 166   { /* "heiß" */
 167     static const uint32_t input[]      = { 0x0068, 0x0065, 0x0069, 0x00DF };
 168     static const uint32_t casefolded[] = { 0x0068, 0x0065, 0x0069, 0x0073, 0x0073 };
 169     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 170   }
 171
 172   /* Case mappings for some characters can depend on the surrounding characters.  */
 173   { /* "περισσότερες πληροφορίες" */
 174     static const uint32_t input[] =
 175       {
 176         0x03C0, 0x03B5, 0x03C1, 0x03B9, 0x03C3, 0x03C3, 0x03CC, 0x03C4,
 177         0x03B5, 0x03C1, 0x03B5, 0x03C2, 0x0020, 0x03C0, 0x03BB, 0x03B7,
 178         0x03C1, 0x03BF, 0x03C6, 0x03BF, 0x03C1, 0x03AF, 0x03B5, 0x03C2
 179       };
 180     static const uint32_t casefolded[] =
 181       {
 182         0x03C0, 0x03B5, 0x03C1, 0x03B9, 0x03C3, 0x03C3, 0x03CC, 0x03C4,
 183         0x03B5, 0x03C1, 0x03B5, 0x03C3, 0x0020, 0x03C0, 0x03BB, 0x03B7,
 184         0x03C1, 0x03BF, 0x03C6, 0x03BF, 0x03C1, 0x03AF, 0x03B5, 0x03C3
 185       };
 186     ASSERT (check (input, SIZEOF (input), NULL, NULL, casefolded, SIZEOF (casefolded)) == 0);
 187   }
 188
 189   /* Case mapping can require subsequent normalization.  */
 190   { /* LATIN SMALL LETTER J WITH CARON, COMBINING DOT BELOW */
 191     static const uint32_t input[]                 = { 0x01F0, 0x0323 };
 192     static const uint32_t casefolded[]            = { 0x006A, 0x030C, 0x0323 };
 193     static const uint32_t casefolded_decomposed[] = { 0x006A, 0x0323, 0x030C };
 194     static const uint32_t casefolded_normalized[] = { 0x01F0, 0x0323 };
 195     ASSERT (check (input, SIZEOF (input), NULL, NULL,        casefolded, SIZEOF (casefolded)) == 0);
 196     ASSERT (check (input, SIZEOF (input), NULL, UNINORM_NFD, casefolded_decomposed, SIZEOF (casefolded_decomposed)) == 0);
 197     ASSERT (check (input, SIZEOF (input), NULL, UNINORM_NFC, casefolded_normalized, SIZEOF (casefolded_normalized)) == 0);
 198   }
 199
 200   return 0;
 201 }