tests/tst_punycode.c

   1 /* tst_punycode.c --- Self tests for punycode.
   2  * Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007  Simon Josefsson
   3  *
   4  * This file is part of GNU Libidn.
   5  *
   6  * GNU Libidn is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2 of the License, or
   9  * (at your option) any later version.
  10  *
  11  * GNU Libidn is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License
  17  * along with GNU Libidn; if not, write to the Free Software
  18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
  19  *
  20  */
  21
  22 #ifdef HAVE_CONFIG_H
  23 # include "config.h"
  24 #endif
  25
  26 #include <stdio.h>
  27 #include <stdlib.h>
  28 #include <stdarg.h>
  29 #include <string.h>
  30
  31 #include <punycode.h>
  32
  33 #include "utils.h"
  34
  35 struct punycode
  36 {
  37   const char *name;
  38   size_t inlen;
  39   uint32_t in[100];
  40   const char *out;
  41   int rc;
  42 };
  43
  44 const struct punycode punycode[] = {
  45   {
  46    "(A) Arabic (Egyptian)", 17,
  47    {
  48     0x0644, 0x064A, 0x0647, 0x0645, 0x0627, 0x0628, 0x062A, 0x0643,
  49     0x0644, 0x0645, 0x0648, 0x0634, 0x0639, 0x0631, 0x0628, 0x064A,
  50     0x061F}, "egbpdaj6bu4bxfgehfvwxn", PUNYCODE_SUCCESS},
  51   {
  52    "(B) Chinese (simplified)", 9,
  53    {
  54     0x4ED6, 0x4EEC, 0x4E3A, 0x4EC0, 0x4E48, 0x4E0D, 0x8BF4, 0x4E2D,
  55     0x6587}, "ihqwcrb4cv8a8dqg056pqjye", PUNYCODE_SUCCESS},
  56   {
  57    "(C) Chinese (traditional)", 9,
  58    {
  59     0x4ED6, 0x5011, 0x7232, 0x4EC0, 0x9EBD, 0x4E0D, 0x8AAA, 0x4E2D,
  60     0x6587}, "ihqwctvzc91f659drss3x8bo0yb", PUNYCODE_SUCCESS},
  61   {
  62    "(D) Czech: Pro<ccaron>prost<ecaron>nemluv<iacute><ccaron>esky", 22,
  63    {
  64     0x0050, 0x0072, 0x006F, 0x010D, 0x0070, 0x0072, 0x006F, 0x0073,
  65     0x0074, 0x011B, 0x006E, 0x0065, 0x006D, 0x006C, 0x0075, 0x0076,
  66     0x00ED, 0x010D, 0x0065, 0x0073, 0x006B, 0x0079},
  67    "Proprostnemluvesky-uyb24dma41a", PUNYCODE_SUCCESS},
  68   {
  69    "(E) Hebrew:", 22,
  70    {
  71     0x05DC, 0x05DE, 0x05D4, 0x05D4, 0x05DD, 0x05E4, 0x05E9, 0x05D5,
  72     0x05D8, 0x05DC, 0x05D0, 0x05DE, 0x05D3, 0x05D1, 0x05E8, 0x05D9,
  73     0x05DD, 0x05E2, 0x05D1, 0x05E8, 0x05D9, 0x05EA},
  74    "4dbcagdahymbxekheh6e0a7fei0b", PUNYCODE_SUCCESS},
  75   {
  76    "(F) Hindi (Devanagari):", 30,
  77    {
  78     0x092F, 0x0939, 0x0932, 0x094B, 0x0917, 0x0939, 0x093F, 0x0928,
  79     0x094D, 0x0926, 0x0940, 0x0915, 0x094D, 0x092F, 0x094B, 0x0902,
  80     0x0928, 0x0939, 0x0940, 0x0902, 0x092C, 0x094B, 0x0932, 0x0938,
  81     0x0915, 0x0924, 0x0947, 0x0939, 0x0948, 0x0902},
  82    "i1baa7eci9glrd9b2ae1bj0hfcgg6iyaf8o0a1dig0cd", PUNYCODE_SUCCESS},
  83   {
  84    "(G) Japanese (kanji and hiragana):", 18,
  85    {
  86     0x306A, 0x305C, 0x307F, 0x3093, 0x306A, 0x65E5, 0x672C, 0x8A9E,
  87     0x3092, 0x8A71, 0x3057, 0x3066, 0x304F, 0x308C, 0x306A, 0x3044,
  88     0x306E, 0x304B},
  89    "n8jok5ay5dzabd5bym9f0cm5685rrjetr6pdxa", PUNYCODE_SUCCESS},
  90   {
  91    "(H) Korean (Hangul syllables):", 24,
  92    {
  93     0xC138, 0xACC4, 0xC758, 0xBAA8, 0xB4E0, 0xC0AC, 0xB78C, 0xB4E4,
  94     0xC774, 0xD55C, 0xAD6D, 0xC5B4, 0xB97C, 0xC774, 0xD574, 0xD55C,
  95     0xB2E4, 0xBA74, 0xC5BC, 0xB9C8, 0xB098, 0xC88B, 0xC744, 0xAE4C},
  96    "989aomsvi5e83db1d2a355cv1e0vak1dwrv93d5xbh15a0dt30a5jpsd879ccm6fea98c",
  97    PUNYCODE_SUCCESS},
  98   {
  99    "(I) Russian (Cyrillic):", 28,
 100    {
 101     0x043F, 0x043E, 0x0447, 0x0435, 0x043C, 0x0443, 0x0436, 0x0435,
 102     0x043E, 0x043D, 0x0438, 0x043D, 0x0435, 0x0433, 0x043E, 0x0432,
 103     0x043E, 0x0440, 0x044F, 0x0442, 0x043F, 0x043E, 0x0440, 0x0443,
 104     0x0441, 0x0441, 0x043A, 0x0438},
 105    "b1abfaaepdrnnbgefbadotcwatmq2g4l", PUNYCODE_SUCCESS},
 106   {
 107    "(J) Spanish: Porqu<eacute>nopuedensimplementehablarenEspa<ntilde>ol", 40,
 108    {
 109     0x0050, 0x006F, 0x0072, 0x0071, 0x0075, 0x00E9, 0x006E, 0x006F,
 110     0x0070, 0x0075, 0x0065, 0x0064, 0x0065, 0x006E, 0x0073, 0x0069,
 111     0x006D, 0x0070, 0x006C, 0x0065, 0x006D, 0x0065, 0x006E, 0x0074,
 112     0x0065, 0x0068, 0x0061, 0x0062, 0x006C, 0x0061, 0x0072, 0x0065,
 113     0x006E, 0x0045, 0x0073, 0x0070, 0x0061, 0x00F1, 0x006F, 0x006C},
 114    "PorqunopuedensimplementehablarenEspaol-fmd56a", PUNYCODE_SUCCESS},
 115   {
 116    "(K) Vietnamese:", 31,
 117    {
 118     0x0054, 0x1EA1, 0x0069, 0x0073, 0x0061, 0x006F, 0x0068, 0x1ECD,
 119     0x006B, 0x0068, 0x00F4, 0x006E, 0x0067, 0x0074, 0x0068, 0x1EC3,
 120     0x0063, 0x0068, 0x1EC9, 0x006E, 0x00F3, 0x0069, 0x0074, 0x0069,
 121     0x1EBF, 0x006E, 0x0067, 0x0056, 0x0069, 0x1EC7, 0x0074},
 122    "TisaohkhngthchnitingVit-kjcr8268qyxafd2f1b9g", PUNYCODE_SUCCESS},
 123   {
 124    "(L) 3<nen>B<gumi><kinpachi><sensei>", 8,
 125    {
 126     0x0033, 0x5E74, 0x0042, 0x7D44, 0x91D1, 0x516B, 0x5148, 0x751F},
 127    "3B-ww4c5e180e575a65lsy2b", PUNYCODE_SUCCESS},
 128   {
 129    "(M) <amuro><namie>-with-SUPER-MONKEYS", 24,
 130    {
 131     0x5B89, 0x5BA4, 0x5948, 0x7F8E, 0x6075, 0x002D, 0x0077, 0x0069,
 132     0x0074, 0x0068, 0x002D, 0x0053, 0x0055, 0x0050, 0x0045, 0x0052,
 133     0x002D, 0x004D, 0x004F, 0x004E, 0x004B, 0x0045, 0x0059, 0x0053},
 134    "-with-SUPER-MONKEYS-pc58ag80a8qai00g7n9n", PUNYCODE_SUCCESS},
 135   {
 136    "(N) Hello-Another-Way-<sorezore><no><basho>", 25,
 137    {
 138     0x0048, 0x0065, 0x006C, 0x006C, 0x006F, 0x002D, 0x0041, 0x006E,
 139     0x006F, 0x0074, 0x0068, 0x0065, 0x0072, 0x002D, 0x0057, 0x0061,
 140     0x0079, 0x002D, 0x305D, 0x308C, 0x305E, 0x308C, 0x306E, 0x5834,
 141     0x6240},
 142    "Hello-Another-Way--fc4qua05auwb3674vfr0b", PUNYCODE_SUCCESS},
 143   {
 144    "(O) <hitotsu><yane><no><shita>2", 8,
 145    {
 146     0x3072, 0x3068, 0x3064, 0x5C4B, 0x6839, 0x306E, 0x4E0B, 0x0032},
 147    "2-u9tlzr9756bt3uc0v", PUNYCODE_SUCCESS},
 148   {
 149    "(P) Maji<de>Koi<suru>5<byou><mae>", 13,
 150    {
 151     0x004D, 0x0061, 0x006A, 0x0069, 0x3067, 0x004B, 0x006F, 0x0069,
 152     0x3059, 0x308B, 0x0035, 0x79D2, 0x524D},
 153    "MajiKoi5-783gue6qz075azm5e", PUNYCODE_SUCCESS},
 154   {
 155    "(Q) <pafii>de<runba>", 9,
 156    {
 157     0x30D1, 0x30D5, 0x30A3, 0x30FC, 0x0064, 0x0065, 0x30EB, 0x30F3, 0x30D0},
 158    "de-jg4avhby1noc0d", PUNYCODE_SUCCESS},
 159   {
 160    "(R) <sono><supiido><de>", 7,
 161    {
 162     0x305D, 0x306E, 0x30B9, 0x30D4, 0x30FC, 0x30C9, 0x3067},
 163    "d9juau41awczczp", PUNYCODE_SUCCESS},
 164   {
 165    "(S) -> $1.00 <-", 11,
 166    {
 167     0x002D, 0x003E, 0x0020, 0x0024, 0x0031, 0x002E, 0x0030, 0x0030,
 168     0x0020, 0x003C, 0x002D}, "-> $1.00 <--", PUNYCODE_SUCCESS}
 169 };
 170
 171 void
 172 doit (void)
 173 {
 174   char *p;
 175   uint32_t *q;
 176   int rc;
 177   size_t i, outlen;
 178
 179   p = malloc (sizeof (*p) * BUFSIZ);
 180   if (p == NULL)
 181     fail ("malloc() returned NULL\n");
 182
 183   q = malloc (sizeof (*q) * BUFSIZ);
 184   if (q == NULL)
 185     fail ("malloc() returned NULL\n");
 186
 187   for (i = 0; i < sizeof (punycode) / sizeof (punycode[0]); i++)
 188     {
 189       if (debug)
 190         printf ("PUNYCODE entry %d: %s\n", i, punycode[i].name);
 191
 192       if (debug)
 193         {
 194           printf ("in:\n");
 195           ucs4print (punycode[i].in, punycode[i].inlen);
 196         }
 197
 198       outlen = BUFSIZ;
 199       rc = punycode_encode (punycode[i].inlen, punycode[i].in,
 200                             NULL, &outlen, p);
 201       if (rc != punycode[i].rc)
 202         {
 203           fail ("punycode_encode() entry %d failed: %d\n", i, rc);
 204           if (debug)
 205             printf ("FATAL\n");
 206           continue;
 207         }
 208
 209       if (rc == PUNYCODE_SUCCESS)
 210         p[outlen] = '\0';
 211
 212       if (debug && rc == PUNYCODE_SUCCESS)
 213         {
 214           printf ("computed out: %s\n", p);
 215           printf ("expected out: %s\n", punycode[i].out);
 216         }
 217       else if (debug)
 218         printf ("returned %d expected %d\n", rc, punycode[i].rc);
 219
 220       if (rc == PUNYCODE_SUCCESS)
 221         {
 222           if (strlen (punycode[i].out) != strlen (p) ||
 223               memcmp (punycode[i].out, p, strlen (p)) != 0)
 224             {
 225               fail ("punycode() entry %d failed\n", i);
 226               if (debug)
 227                 printf ("ERROR\n");
 228             }
 229           else if (debug)
 230             printf ("OK\n\n");
 231         }
 232       else if (debug)
 233         printf ("OK\n\n");
 234
 235       if (debug)
 236         {
 237           printf ("in: %s\n", punycode[i].out);
 238         }
 239
 240       outlen = BUFSIZ;
 241       rc = punycode_decode (strlen (punycode[i].out), punycode[i].out,
 242                             &outlen, q, NULL);
 243       if (rc != punycode[i].rc)
 244         {
 245           fail ("punycode() entry %d failed: %d\n", i, rc);
 246           if (debug)
 247             printf ("FATAL\n");
 248           continue;
 249         }
 250
 251       if (debug && rc == PUNYCODE_SUCCESS)
 252         {
 253           printf ("computed out:\n");
 254           ucs4print (q, outlen);
 255           printf ("expected out:\n");
 256           ucs4print (punycode[i].in, punycode[i].inlen);
 257         }
 258       else if (debug)
 259         printf ("returned %d expected %d\n", rc, punycode[i].rc);
 260
 261       if (rc == PUNYCODE_SUCCESS)
 262         {
 263           if (punycode[i].inlen != outlen ||
 264               memcmp (punycode[i].in, q, outlen) != 0)
 265             {
 266               fail ("punycode_decode() entry %d failed\n", i);
 267               if (debug)
 268                 printf ("ERROR\n");
 269             }
 270           else if (debug)
 271             printf ("OK\n\n");
 272         }
 273       else if (debug)
 274         printf ("OK\n\n");
 275     }
 276
 277   free (q);
 278   free (p);
 279 }