glib/tests/utf8-misc.c

   1 /* Unit tests for utilities
   2  * Copyright (C) 2010 Red Hat, Inc.
   3  *
   4  * This work is provided "as is"; redistribution and modification
   5  * in whole or in part, in any medium, physical or electronic is
   6  * permitted without restriction.
   7  *
   8  * This work is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
  11  *
  12  * In no event shall the authors or contributors be liable for any
  13  * direct, indirect, incidental, special, exemplary, or consequential
  14  * damages (including, but not limited to, procurement of substitute
  15  * goods or services; loss of use, data, or profits; or business
  16  * interruption) however caused and on any theory of liability, whether
  17  * in contract, strict liability, or tort (including negligence or
  18  * otherwise) arising in any way out of the use of this software, even
  19  * if advised of the possibility of such damage.
  20  *
  21  * Author: Matthias Clasen
  22  */
  23
  24 #include "glib.h"
  25
  26 static void
  27 test_utf8_strlen (void)
  28 {
  29   const gchar *string = "\xe2\x82\xa0gh\xe2\x82\xa4jl";
  30
  31   g_assert_cmpint (g_utf8_strlen (string, -1), ==, 6);
  32   g_assert_cmpint (g_utf8_strlen (string, 0), ==, 0);
  33   g_assert_cmpint (g_utf8_strlen (string, 1), ==, 0);
  34   g_assert_cmpint (g_utf8_strlen (string, 2), ==, 0);
  35   g_assert_cmpint (g_utf8_strlen (string, 3), ==, 1);
  36   g_assert_cmpint (g_utf8_strlen (string, 4), ==, 2);
  37   g_assert_cmpint (g_utf8_strlen (string, 5), ==, 3);
  38   g_assert_cmpint (g_utf8_strlen (string, 6), ==, 3);
  39   g_assert_cmpint (g_utf8_strlen (string, 7), ==, 3);
  40   g_assert_cmpint (g_utf8_strlen (string, 8), ==, 4);
  41   g_assert_cmpint (g_utf8_strlen (string, 9), ==, 5);
  42   g_assert_cmpint (g_utf8_strlen (string, 10), ==, 6);
  43 }
  44
  45 static void
  46 test_utf8_strncpy (void)
  47 {
  48   const gchar *string = "\xe2\x82\xa0gh\xe2\x82\xa4jl";
  49   gchar dest[20];
  50
  51   g_utf8_strncpy (dest, string, 0);
  52   g_assert_cmpstr (dest, ==, "");
  53
  54   g_utf8_strncpy (dest, string, 1);
  55   g_assert_cmpstr (dest, ==, "\xe2\x82\xa0");
  56
  57   g_utf8_strncpy (dest, string, 2);
  58   g_assert_cmpstr (dest, ==, "\xe2\x82\xa0g");
  59
  60   g_utf8_strncpy (dest, string, 3);
  61   g_assert_cmpstr (dest, ==, "\xe2\x82\xa0gh");
  62
  63   g_utf8_strncpy (dest, string, 4);
  64   g_assert_cmpstr (dest, ==, "\xe2\x82\xa0gh\xe2\x82\xa4");
  65
  66   g_utf8_strncpy (dest, string, 5);
  67   g_assert_cmpstr (dest, ==, "\xe2\x82\xa0gh\xe2\x82\xa4j");
  68
  69   g_utf8_strncpy (dest, string, 6);
  70   g_assert_cmpstr (dest, ==, "\xe2\x82\xa0gh\xe2\x82\xa4jl");
  71
  72   g_utf8_strncpy (dest, string, 20);
  73   g_assert_cmpstr (dest, ==, "\xe2\x82\xa0gh\xe2\x82\xa4jl");
  74 }
  75
  76 static void
  77 test_utf8_strrchr (void)
  78 {
  79   const gchar *string = "\xe2\x82\xa0gh\xe2\x82\xa4jl\xe2\x82\xa4jl";
  80
  81   g_assert (g_utf8_strrchr (string, -1, 'j') == string + 13);
  82   g_assert (g_utf8_strrchr (string, -1, 8356) == string + 10);
  83   g_assert (g_utf8_strrchr (string, 9, 8356) == string + 5);
  84   g_assert (g_utf8_strrchr (string, 3, 'j') == NULL);
  85   g_assert (g_utf8_strrchr (string, -1, 'x') == NULL);
  86 }
  87
  88 static void
  89 test_utf8_reverse (void)
  90 {
  91   gchar *r;
  92
  93   r = g_utf8_strreverse ("abcdef", -1);
  94   g_assert_cmpstr (r, ==, "fedcba");
  95   g_free (r);
  96
  97   r = g_utf8_strreverse ("abcdef", 4);
  98   g_assert_cmpstr (r, ==, "dcba");
  99   g_free (r);
 100
 101   /* U+0B0B Oriya Letter Vocalic R
 102    * U+10900 Phoenician Letter Alf
 103    * U+0041 Latin Capital Letter A
 104    * U+1EB6 Latin Capital Letter A With Breve And Dot Below
 105    */
 106   r = g_utf8_strreverse ("\340\254\213\360\220\244\200\101\341\272\266", -1);
 107   g_assert_cmpstr (r, ==, "\341\272\266\101\360\220\244\200\340\254\213");
 108   g_free (r);
 109 }
 110
 111 static void
 112 test_utf8_substring (void)
 113 {
 114   gchar *r;
 115
 116   r = g_utf8_substring ("abcd", 1, 3);
 117   g_assert_cmpstr (r, ==, "bc");
 118   g_free (r);
 119
 120   r = g_utf8_substring ("abcd", 0, 4);
 121   g_assert_cmpstr (r, ==, "abcd");
 122   g_free (r);
 123
 124   r = g_utf8_substring ("abcd", 2, 2);
 125   g_assert_cmpstr (r, ==, "");
 126   g_free (r);
 127
 128   r = g_utf8_substring ("abc\xe2\x82\xa0gh\xe2\x82\xa4", 2, 5);
 129   g_assert_cmpstr (r, ==, "c\xe2\x82\xa0g");
 130   g_free (r);
 131 }
 132
 133 static void
 134 test_utf8_make_valid (void)
 135 {
 136   gchar *r;
 137
 138   /* valid UTF8 */
 139   r = g_utf8_make_valid ("\xe2\x82\xa0gh\xe2\x82\xa4jl", -1);
 140   g_assert_cmpstr (r, ==, "\xe2\x82\xa0gh\xe2\x82\xa4jl");
 141   g_free (r);
 142
 143   /* invalid UTF8 */
 144   r = g_utf8_make_valid ("\xe2\x82\xa0gh\xe2\xffjl", -1);
 145   g_assert_cmpstr (r, ==, "\xe2\x82\xa0gh\xef\xbf\xbd\xef\xbf\xbdjl");
 146   g_free (r);
 147
 148   /* invalid UTF8 without nul terminator followed by something unfortunate */
 149   r = g_utf8_make_valid ("Bj\xc3\xb8", 3);
 150   g_assert_cmpstr (r, ==, "Bj\xef\xbf\xbd");
 151   g_free (r);
 152
 153   /* invalid UTF8 with embedded nul */
 154   r = g_utf8_make_valid ("\xe2\x82\xa0gh\xe2\x00jl", 9);
 155   g_assert_cmpstr (r, ==, "\xe2\x82\xa0gh\xef\xbf\xbd\xef\xbf\xbdjl");
 156   g_free (r);
 157 }
 158
 159 int
 160 main (int   argc,
 161       char *argv[])
 162 {
 163   g_test_init (&argc, &argv, NULL);
 164
 165   g_test_add_func ("/utf8/strlen", test_utf8_strlen);
 166   g_test_add_func ("/utf8/strncpy", test_utf8_strncpy);
 167   g_test_add_func ("/utf8/strrchr", test_utf8_strrchr);
 168   g_test_add_func ("/utf8/reverse", test_utf8_reverse);
 169   g_test_add_func ("/utf8/substring", test_utf8_substring);
 170   g_test_add_func ("/utf8/make-valid", test_utf8_make_valid);
 171
 172   return g_test_run();
 173 }