glib/tests/utf8-pointer.c

   1 /* GLIB - Library of useful routines for C programming
   2  * Copyright (C) 1995-1997  Peter Mattis, Spencer Kimball and Josh MacDonald
   3  *
   4  * This library is free software; you can redistribute it and/or
   5  * modify it under the terms of the GNU Lesser General Public
   6  * License as published by the Free Software Foundation; either
   7  * version 2.1 of the License, or (at your option) any later version.
   8  *
   9  * This library 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 GNU
  12  * Lesser General Public License for more details.
  13  *
  14  * You should have received a copy of the GNU Lesser General Public
  15  * License along with this library; if not, see <http://www.gnu.org/licenses/>.
  16  */
  17
  18 /*
  19  * Modified by the GLib Team and others 1997-2000.  See the AUTHORS
  20  * file for a list of people on the GLib Team.  See the ChangeLog
  21  * files for a list of changes.  These files are distributed with
  22  * GLib at ftp://ftp.gtk.org/pub/gtk/.
  23  */
  24
  25 #include <string.h>
  26 #include <glib.h>
  27
  28 /* Test conversions between offsets and pointers */
  29
  30 static void test_utf8 (gconstpointer d)
  31 {
  32   gint num_chars;
  33   const gchar **p;
  34   gint i, j;
  35   const gchar *string = d;
  36
  37   g_assert (g_utf8_validate (string, -1, NULL));
  38
  39   num_chars = g_utf8_strlen (string, -1);
  40
  41   p = (const gchar **) g_malloc (num_chars * sizeof (gchar *));
  42
  43   p[0] = string;
  44   for (i = 1; i < num_chars; i++)
  45     p[i] = g_utf8_next_char (p[i-1]);
  46
  47   for (i = 0; i < num_chars; i++)
  48     for (j = 0; j < num_chars; j++)
  49       {
  50         g_assert (g_utf8_offset_to_pointer (p[i], j - i) == p[j]);
  51         g_assert (g_utf8_pointer_to_offset (p[i], p[j]) == j - i);
  52       }
  53
  54   g_free (p);
  55 }
  56
  57 gchar *longline = "asdasdas dsaf asfd as fdasdf asfd asdf as dfas dfasdf a"
  58 "asd fasdf asdf asdf asd fasfd as fdasfd asdf as fdççççççççças ffsd asfd as fdASASASAs As"
  59 "Asfdsf sdfg sdfg dsfg dfg sdfgsdfgsdfgsdfg sdfgsdfg sdfg sdfg sdf gsdfg sdfg sd"
  60 "asd fasdf asdf asdf asd fasfd as fdaèèèèèèè òòòòòòòòòòòòsfd asdf as fdas ffsd asfd as fdASASASAs D"
  61 "Asfdsf sdfg sdfg dsfg dfg sdfgsdfgsdfgsdfg sdfgsdfg sdfgùùùùùùùùùùùùùù sdfg sdf gsdfg sdfg sd"
  62 "asd fasdf asdf asdf asd fasfd as fdasfd asd@@@@@@@f as fdas ffsd asfd as fdASASASAs D "
  63 "Asfdsf sdfg sdfg dsfg dfg sdfgsdfgsdfgsdfg sdfgsdf€€€€€€€€€€€€€€€€€€g sdfg sdfg sdf gsdfg sdfg sd"
  64 "asd fasdf asdf asdf asd fasfd as fdasfd asdf as fdas ffsd asfd as fdASASASAs D"
  65 "Asfdsf sdfg sdfg dsfg dfg sdfgsdfgsdfgsdfg sdfgsdfg sdfg sdfg sdf gsdfg sdfg sd\n\nlalala\n";
  66
  67 static void
  68 test_length (void)
  69 {
  70   g_assert (g_utf8_strlen ("1234", -1) == 4);
  71   g_assert (g_utf8_strlen ("1234", 0) == 0);
  72   g_assert (g_utf8_strlen ("1234", 1) == 1);
  73   g_assert (g_utf8_strlen ("1234", 2) == 2);
  74   g_assert (g_utf8_strlen ("1234", 3) == 3);
  75   g_assert (g_utf8_strlen ("1234", 4) == 4);
  76   g_assert (g_utf8_strlen ("1234", 5) == 4);
  77
  78   g_assert (g_utf8_strlen (longline, -1) == 762);
  79   g_assert (g_utf8_strlen (longline, strlen (longline)) == 762);
  80   g_assert (g_utf8_strlen (longline, 1024) == 762);
  81
  82   g_assert (g_utf8_strlen (NULL, 0) == 0);
  83
  84   g_assert (g_utf8_strlen ("a\340\250\201c", -1) == 3);
  85   g_assert (g_utf8_strlen ("a\340\250\201c", 1) == 1);
  86   g_assert (g_utf8_strlen ("a\340\250\201c", 2) == 1);
  87   g_assert (g_utf8_strlen ("a\340\250\201c", 3) == 1);
  88   g_assert (g_utf8_strlen ("a\340\250\201c", 4) == 2);
  89   g_assert (g_utf8_strlen ("a\340\250\201c", 5) == 3);
  90 }
  91
  92 static void
  93 test_find (void)
  94 {
  95   /* U+0B0B Oriya Letter Vocalic R (\340\254\213)
  96    * U+10900 Phoenician Letter Alf (\360\220\244\200)
  97    * U+0041 Latin Capital Letter A (\101)
  98    * U+1EB6 Latin Capital Letter A With Breve And Dot Below (\341\272\266)
  99    */
 100   const gchar *str = "\340\254\213\360\220\244\200\101\341\272\266\0\101";
 101   const gchar *p = str + strlen (str);
 102   const gchar *q;
 103
 104   q = g_utf8_find_prev_char (str, p);
 105   g_assert (q == str + 8);
 106   q = g_utf8_find_prev_char (str, q);
 107   g_assert (q == str + 7);
 108   q = g_utf8_find_prev_char (str, q);
 109   g_assert (q == str + 3);
 110   q = g_utf8_find_prev_char (str, q);
 111   g_assert (q == str);
 112   q = g_utf8_find_prev_char (str, q);
 113   g_assert (q == NULL);
 114
 115   p = str + 2;
 116   q = g_utf8_find_next_char (p, NULL);
 117   g_assert (q == str + 3);
 118   q = g_utf8_find_next_char (q, NULL);
 119   g_assert (q == str + 7);
 120
 121   q = g_utf8_find_next_char (p, str + 6);
 122   g_assert (q == str + 3);
 123   q = g_utf8_find_next_char (q, str + 6);
 124   g_assert (q == NULL);
 125
 126   q = g_utf8_find_next_char (str, str);
 127   g_assert (q == NULL);
 128
 129   q = g_utf8_find_next_char (str + strlen (str), NULL);
 130   g_assert (q == str + strlen (str) + 1);
 131
 132   /* Check return values when reaching the end of the string,
 133    * with @end set and unset. */
 134   q = g_utf8_find_next_char (str + 10, NULL);
 135   g_assert_nonnull (q);
 136   g_assert (*q == '\0');
 137
 138   q = g_utf8_find_next_char (str + 10, str + 11);
 139   g_assert_null (q);
 140 }
 141
 142 int main (int argc, char *argv[])
 143 {
 144   g_test_init (&argc, &argv, NULL);
 145
 146   g_test_add_data_func ("/utf8/offsets", longline, test_utf8);
 147   g_test_add_func ("/utf8/lengths", test_length);
 148   g_test_add_func ("/utf8/find", test_find);
 149
 150   return g_test_run ();
 151 }