1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/LibOVR/Src/Kernel/OVR_UTF8Util.cpp	Wed Jan 14 06:51:16 2015 +0200
     1.3 @@ -0,0 +1,556 @@
     1.4 +/**************************************************************************
     1.5 +
     1.6 +Filename    :   OVR_UTF8Util.cpp
     1.7 +Content     :   UTF8 Unicode character encoding/decoding support
     1.8 +Created     :   September 19, 2012
     1.9 +Notes       : 
    1.10 +Notes       :   Much useful info at "UTF-8 and Unicode FAQ"
    1.11 +                http://www.cl.cam.ac.uk/~mgk25/unicode.html
    1.12 +
    1.13 +Copyright   :   Copyright 2014 Oculus VR, LLC All Rights reserved.
    1.14 +
    1.15 +Licensed under the Oculus VR Rift SDK License Version 3.2 (the "License"); 
    1.16 +you may not use the Oculus VR Rift SDK except in compliance with the License, 
    1.17 +which is provided at the time of installation or download, or which 
    1.18 +otherwise accompanies this software in either electronic or hard copy form.
    1.19 +
    1.20 +You may obtain a copy of the License at
    1.21 +
    1.22 +http://www.oculusvr.com/licenses/LICENSE-3.2 
    1.23 +
    1.24 +Unless required by applicable law or agreed to in writing, the Oculus VR SDK 
    1.25 +distributed under the License is distributed on an "AS IS" BASIS,
    1.26 +WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    1.27 +See the License for the specific language governing permissions and
    1.28 +limitations under the License.
    1.29 +
    1.30 +************************************************************************************/
    1.31 +
    1.32 +#include "OVR_UTF8Util.h"
    1.33 +
    1.34 +namespace OVR { namespace UTF8Util {
    1.35 +
    1.36 +intptr_t OVR_STDCALL GetLength(const char* buf, intptr_t buflen)
    1.37 +{
    1.38 +    const char* p = buf;
    1.39 +    intptr_t length = 0;
    1.40 +
    1.41 +    if (buflen != -1)
    1.42 +    {
    1.43 +        while (p - buf < buflen)
    1.44 +        {
    1.45 +            // We should be able to have ASStrings with 0 in the middle.
    1.46 +            UTF8Util::DecodeNextChar_Advance0(&p);
    1.47 +            length++;
    1.48 +        }
    1.49 +    }
    1.50 +    else
    1.51 +    {
    1.52 +        while (UTF8Util::DecodeNextChar_Advance0(&p))
    1.53 +            length++;
    1.54 +    }
    1.55 +    
    1.56 +    return length;
    1.57 +}
    1.58 +
    1.59 +uint32_t OVR_STDCALL GetCharAt(intptr_t index, const char* putf8str, intptr_t length)
    1.60 +{
    1.61 +    const char* buf = putf8str;
    1.62 +    uint32_t  c = 0;
    1.63 +
    1.64 +    if (length != -1)
    1.65 +    {
    1.66 +        while (buf - putf8str < length)
    1.67 +        {           
    1.68 +            c = UTF8Util::DecodeNextChar_Advance0(&buf);
    1.69 +            if (index == 0)
    1.70 +                return c;
    1.71 +            index--;
    1.72 +        }
    1.73 +
    1.74 +        return c;
    1.75 +    }
    1.76 +
    1.77 +    do 
    1.78 +    {
    1.79 +        c = UTF8Util::DecodeNextChar_Advance0(&buf);
    1.80 +        index--;
    1.81 +
    1.82 +        if (c == 0)
    1.83 +        {
    1.84 +            // We've hit the end of the string; don't go further.
    1.85 +            OVR_ASSERT(index == 0);
    1.86 +            return c;
    1.87 +        }
    1.88 +    } while (index >= 0);
    1.89 +
    1.90 +    return c;
    1.91 +}
    1.92 +
    1.93 +intptr_t OVR_STDCALL GetByteIndex(intptr_t index, const char *putf8str, intptr_t length)
    1.94 +{
    1.95 +    const char* buf = putf8str;
    1.96 +
    1.97 +    if (length != -1)
    1.98 +    {
    1.99 +        while ((buf - putf8str) < length && index > 0)
   1.100 +        {
   1.101 +            UTF8Util::DecodeNextChar_Advance0(&buf);
   1.102 +            index--;
   1.103 +        }
   1.104 +
   1.105 +        return buf-putf8str;
   1.106 +    }
   1.107 +
   1.108 +    while (index > 0) 
   1.109 +    {
   1.110 +        uint32_t c = UTF8Util::DecodeNextChar_Advance0(&buf);
   1.111 +        index--;
   1.112 +
   1.113 +        if (c == 0)
   1.114 +            return buf-putf8str;
   1.115 +    };
   1.116 +
   1.117 +    return buf-putf8str;
   1.118 +}
   1.119 +
   1.120 +int OVR_STDCALL GetEncodeCharSize(uint32_t ucs_character)
   1.121 +{
   1.122 +    if (ucs_character <= 0x7F)
   1.123 +        return 1;
   1.124 +    else if (ucs_character <= 0x7FF)
   1.125 +        return 2;
   1.126 +    else if (ucs_character <= 0xFFFF)
   1.127 +        return 3;
   1.128 +    else if (ucs_character <= 0x1FFFFF)
   1.129 +        return 4;
   1.130 +    else if (ucs_character <= 0x3FFFFFF)
   1.131 +        return 5;
   1.132 +    else if (ucs_character <= 0x7FFFFFFF)
   1.133 +        return 6;
   1.134 +    else
   1.135 +        return 0;
   1.136 +}
   1.137 +
   1.138 +uint32_t OVR_STDCALL DecodeNextChar_Advance0(const char** putf8Buffer)
   1.139 +{
   1.140 +    uint32_t  uc;
   1.141 +    char    c;
   1.142 +    
   1.143 +    // Security considerations:
   1.144 +    //
   1.145 +    // Changed, this is now only the case for DecodeNextChar:
   1.146 +    //  - If we hit a zero byte, we want to return 0 without stepping
   1.147 +    //    the buffer pointer past the 0. th
   1.148 +    //
   1.149 +    // If we hit an "overlong sequence"; i.e. a character encoded
   1.150 +    // in a longer multibyte string than is necessary, then we
   1.151 +    // need to discard the character.  This is so attackers can't
   1.152 +    // disguise dangerous characters or character sequences --
   1.153 +    // there is only one valid encoding for each character.
   1.154 +    //
   1.155 +    // If we decode characters { 0xD800 .. 0xDFFF } or { 0xFFFE,
   1.156 +    // 0xFFFF } then we ignore them; they are not valid in UTF-8.
   1.157 +    
   1.158 +    // This isn't actually an invalid character; it's a valid char that
   1.159 +    // looks like an inverted question mark.
   1.160 +#define INVALID_CHAR 0x0FFFD
   1.161 +    
   1.162 +#define FIRST_BYTE(mask, shift)     \
   1.163 +    uc = (c & (mask)) << (shift);
   1.164 +    
   1.165 +#define NEXT_BYTE(shift) \
   1.166 +    c = **putf8Buffer;   \
   1.167 +    if (c == 0) return 0; /* end of buffer, do not advance */   \
   1.168 +    if ((c & 0xC0) != 0x80) return INVALID_CHAR; /* standard check */  \
   1.169 +    (*putf8Buffer)++;    \
   1.170 +    uc |= (c & 0x3F) << shift;
   1.171 +    
   1.172 +    c = **putf8Buffer;
   1.173 +    (*putf8Buffer)++;
   1.174 +    if (c == 0)
   1.175 +        return 0;   // End of buffer.
   1.176 +    
   1.177 +    if ((c & 0x80) == 0) return (uint32_t) c; // Conventional 7-bit ASCII.
   1.178 +    
   1.179 +    // Multi-byte sequences.
   1.180 +    if ((c & 0xE0) == 0xC0)
   1.181 +    {
   1.182 +        // Two-byte sequence.
   1.183 +        FIRST_BYTE(0x1F, 6);
   1.184 +        NEXT_BYTE(0);
   1.185 +        if (uc < 0x80) return INVALID_CHAR;  // overlong
   1.186 +        return uc;
   1.187 +    }
   1.188 +    else if ((c & 0xF0) == 0xE0)
   1.189 +    {
   1.190 +        // Three-byte sequence.
   1.191 +        FIRST_BYTE(0x0F, 12);
   1.192 +        NEXT_BYTE(6);
   1.193 +        NEXT_BYTE(0);
   1.194 +        if (uc < 0x800) return INVALID_CHAR; // overlong
   1.195 +        // Not valid ISO 10646, but Flash requires these to work
   1.196 +        // see AS3 test e15_5_3_2_3 for String.fromCharCode().charCodeAt(0)
   1.197 +        // if (uc >= 0x0D800 && uc <= 0x0DFFF) return INVALID_CHAR;
   1.198 +        // if (uc == 0x0FFFE || uc == 0x0FFFF) return INVALID_CHAR; // not valid ISO 10646
   1.199 +        return uc;
   1.200 +    }
   1.201 +    else if ((c & 0xF8) == 0xF0)
   1.202 +    {
   1.203 +        // Four-byte sequence.
   1.204 +        FIRST_BYTE(0x07, 18);
   1.205 +        NEXT_BYTE(12);
   1.206 +        NEXT_BYTE(6);
   1.207 +        NEXT_BYTE(0);
   1.208 +        if (uc < 0x010000) return INVALID_CHAR;  // overlong
   1.209 +        return uc;
   1.210 +    }
   1.211 +    else if ((c & 0xFC) == 0xF8)
   1.212 +    {
   1.213 +        // Five-byte sequence.
   1.214 +        FIRST_BYTE(0x03, 24);
   1.215 +        NEXT_BYTE(18);
   1.216 +        NEXT_BYTE(12);
   1.217 +        NEXT_BYTE(6);
   1.218 +        NEXT_BYTE(0);
   1.219 +        if (uc < 0x0200000) return INVALID_CHAR; // overlong
   1.220 +        return uc;
   1.221 +    }
   1.222 +    else if ((c & 0xFE) == 0xFC)
   1.223 +    {
   1.224 +        // Six-byte sequence.
   1.225 +        FIRST_BYTE(0x01, 30);
   1.226 +        NEXT_BYTE(24);
   1.227 +        NEXT_BYTE(18);
   1.228 +        NEXT_BYTE(12);
   1.229 +        NEXT_BYTE(6);
   1.230 +        NEXT_BYTE(0);
   1.231 +        if (uc < 0x04000000) return INVALID_CHAR;    // overlong
   1.232 +        return uc;
   1.233 +    }
   1.234 +    else
   1.235 +    {
   1.236 +        // Invalid.
   1.237 +        return INVALID_CHAR;
   1.238 +    }
   1.239 +}
   1.240 +
   1.241 +
   1.242 +void OVR_STDCALL EncodeChar(char* pbuffer, intptr_t* pindex, uint32_t ucs_character)
   1.243 +{
   1.244 +    if (ucs_character <= 0x7F)
   1.245 +    {
   1.246 +        // Plain single-byte ASCII.
   1.247 +        pbuffer[(*pindex)++] = (char) ucs_character;
   1.248 +    }
   1.249 +    else if (ucs_character <= 0x7FF)
   1.250 +    {
   1.251 +        // Two bytes.
   1.252 +        pbuffer[(*pindex)++] = 0xC0 | (char)(ucs_character >> 6);
   1.253 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 0) & 0x3F);
   1.254 +    }
   1.255 +    else if (ucs_character <= 0xFFFF)
   1.256 +    {
   1.257 +        // Three bytes.
   1.258 +        pbuffer[(*pindex)++] = 0xE0 | (char)(ucs_character >> 12);
   1.259 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 6) & 0x3F);
   1.260 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 0) & 0x3F);
   1.261 +    }
   1.262 +    else if (ucs_character <= 0x1FFFFF)
   1.263 +    {
   1.264 +        // Four bytes.
   1.265 +        pbuffer[(*pindex)++] = 0xF0 | (char)(ucs_character >> 18);
   1.266 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 12) & 0x3F);
   1.267 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 6) & 0x3F);
   1.268 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 0) & 0x3F);
   1.269 +    }
   1.270 +    else if (ucs_character <= 0x3FFFFFF)
   1.271 +    {
   1.272 +        // Five bytes.
   1.273 +        pbuffer[(*pindex)++] = 0xF8 | (char)(ucs_character >> 24);
   1.274 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 18) & 0x3F);
   1.275 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 12) & 0x3F);
   1.276 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 6) & 0x3F);
   1.277 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 0) & 0x3F);
   1.278 +    }
   1.279 +    else if (ucs_character <= 0x7FFFFFFF)
   1.280 +    {
   1.281 +        // Six bytes.
   1.282 +        pbuffer[(*pindex)++] = 0xFC | (char)(ucs_character >> 30);
   1.283 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 24) & 0x3F);
   1.284 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 18) & 0x3F);
   1.285 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 12) & 0x3F);
   1.286 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 6) & 0x3F);
   1.287 +        pbuffer[(*pindex)++] = 0x80 | (char)((ucs_character >> 0) & 0x3F);
   1.288 +    }
   1.289 +    else
   1.290 +    {
   1.291 +        // Invalid char; don't encode anything.
   1.292 +    }
   1.293 +}
   1.294 +
   1.295 +intptr_t OVR_STDCALL GetEncodeStringSize(const wchar_t* pchar, intptr_t length)
   1.296 +{
   1.297 +    intptr_t len = 0;
   1.298 +    if (length != -1)
   1.299 +        for (int i = 0; i < length; i++)
   1.300 +        {
   1.301 +            len += GetEncodeCharSize(pchar[i]);
   1.302 +        }
   1.303 +    else
   1.304 +        for (int i = 0;; i++)
   1.305 +        {
   1.306 +            if (pchar[i] == 0)
   1.307 +                return len;
   1.308 +            len += GetEncodeCharSize(pchar[i]);
   1.309 +        }
   1.310 +    return len;
   1.311 +}
   1.312 +
   1.313 +void OVR_STDCALL EncodeString(char *pbuff, const wchar_t* pchar, intptr_t length)
   1.314 +{
   1.315 +    intptr_t ofs = 0;
   1.316 +    if (length != -1)
   1.317 +    {
   1.318 +        for (int i = 0; i < length; i++)
   1.319 +        {            
   1.320 +            EncodeChar(pbuff, &ofs, pchar[i]);
   1.321 +        }
   1.322 +    }
   1.323 +    else
   1.324 +    {
   1.325 +        for (int i = 0;; i++)
   1.326 +        {
   1.327 +            if (pchar[i] == 0)
   1.328 +                break;
   1.329 +            EncodeChar(pbuff, &ofs, pchar[i]);
   1.330 +        }
   1.331 +    }
   1.332 +    pbuff[ofs] = 0;
   1.333 +}
   1.334 +
   1.335 +size_t OVR_STDCALL DecodeString(wchar_t *pbuff, const char* putf8str, intptr_t bytesLen)
   1.336 +{
   1.337 +    wchar_t *pbegin = pbuff;
   1.338 +    if (bytesLen == -1)
   1.339 +    {
   1.340 +        while (1)
   1.341 +        {
   1.342 +            uint32_t ch = DecodeNextChar_Advance0(&putf8str);
   1.343 +            if (ch == 0)
   1.344 +                break;
   1.345 +            else if (ch >= 0xFFFF)
   1.346 +                ch = 0xFFFD;
   1.347 +            *pbuff++ = wchar_t(ch);
   1.348 +        }
   1.349 +    }
   1.350 +    else
   1.351 +    {
   1.352 +        const char* p = putf8str;
   1.353 +        while ((p - putf8str) < bytesLen)
   1.354 +        {
   1.355 +            uint32_t ch = DecodeNextChar_Advance0(&p);
   1.356 +            if (ch >= 0xFFFF)
   1.357 +                ch = 0xFFFD;
   1.358 +            *pbuff++ = wchar_t(ch);
   1.359 +        }
   1.360 +    }
   1.361 +
   1.362 +    *pbuff = 0;
   1.363 +    return pbuff - pbegin;
   1.364 +}
   1.365 +
   1.366 +
   1.367 +#ifdef UTF8_UNIT_TEST
   1.368 +
   1.369 +// Compile this test case with something like:
   1.370 +//
   1.371 +// gcc utf8.cpp -g -I.. -DUTF8_UNIT_TEST -lstdc++ -o utf8_test
   1.372 +//
   1.373 +//    or
   1.374 +//
   1.375 +// cl utf8.cpp -Zi -Od -DUTF8_UNIT_TEST -I..
   1.376 +//
   1.377 +// If possible, try running the test program with the first arg
   1.378 +// pointing at the file:
   1.379 +//
   1.380 +// http://www.cl.cam.ac.uk/~mgk25/ucs/examples/UTF-8-test.txt
   1.381 +// 
   1.382 +// and examine the results by eye to make sure they are acceptable to
   1.383 +// you.
   1.384 +
   1.385 +
   1.386 +#include "base/utility.h"
   1.387 +#include <stdio.h>
   1.388 +
   1.389 +
   1.390 +bool    check_equal(const char* utf8_in, const uint32_t* ucs_in)
   1.391 +{
   1.392 +    for (;;)
   1.393 +    {
   1.394 +        uint32_t  next_ucs = *ucs_in++;
   1.395 +        uint32_t  next_ucs_from_utf8 = utf8::decode_next_unicode_character(&utf8_in);
   1.396 +        if (next_ucs != next_ucs_from_utf8)
   1.397 +        {
   1.398 +            return false;
   1.399 +        }
   1.400 +        if (next_ucs == 0)
   1.401 +        {
   1.402 +            OVR_ASSERT(next_ucs_from_utf8 == 0);
   1.403 +            break;
   1.404 +        }
   1.405 +    }
   1.406 +    
   1.407 +    return true;
   1.408 +}
   1.409 +
   1.410 +
   1.411 +void    log_ascii(const char* line)
   1.412 +{
   1.413 +    for (;;)
   1.414 +    {
   1.415 +        unsigned char   c = (unsigned char) *line++;
   1.416 +        if (c == 0)
   1.417 +        {
   1.418 +            // End of line.
   1.419 +            return;
   1.420 +        }
   1.421 +        else if (c != '\n'
   1.422 +            && (c < 32 || c > 127))
   1.423 +        {
   1.424 +            // Non-printable as plain ASCII.
   1.425 +            printf("<0x%02X>", (int) c);
   1.426 +        }
   1.427 +        else
   1.428 +        {
   1.429 +            printf("%c", c);
   1.430 +        }
   1.431 +    }
   1.432 +}
   1.433 +
   1.434 +
   1.435 +void    log_ucs(const uint32_t* line)
   1.436 +{
   1.437 +    for (;;)
   1.438 +    {
   1.439 +        uint32_t  uc = *line++;
   1.440 +        if (uc == 0)
   1.441 +        {
   1.442 +            // End of line.
   1.443 +            return;
   1.444 +        }
   1.445 +        else if (uc != '\n'
   1.446 +            && (uc < 32 || uc > 127))
   1.447 +        {
   1.448 +            // Non-printable as plain ASCII.
   1.449 +            printf("<U-%04X>", uc);
   1.450 +        }
   1.451 +        else
   1.452 +        {
   1.453 +            printf("%c", (char) uc);
   1.454 +        }
   1.455 +    }
   1.456 +}
   1.457 +
   1.458 +
   1.459 +// Simple canned test.
   1.460 +int main(int argc, const char* argv[])
   1.461 +{
   1.462 +    {
   1.463 +        const char* test8 = "Ignacio Castaño";
   1.464 +        const uint32_t  test32[] =
   1.465 +        {
   1.466 +            0x49, 0x67, 0x6E, 0x61, 0x63,
   1.467 +                0x69, 0x6F, 0x20, 0x43, 0x61,
   1.468 +                0x73, 0x74, 0x61, 0xF1, 0x6F,
   1.469 +                0x00
   1.470 +        };
   1.471 +        
   1.472 +        OVR_ASSERT(check_equal(test8, test32));
   1.473 +    }
   1.474 +        
   1.475 +        // If user passed an arg, try reading the file as UTF-8 encoded text.
   1.476 +        if (argc > 1)
   1.477 +        {
   1.478 +            const char* filename = argv[1];
   1.479 +            FILE*   fp = fopen(filename, "rb");
   1.480 +            if (fp == NULL)
   1.481 +            {
   1.482 +                printf("Can't open file '%s'\n", filename);
   1.483 +                return 1;
   1.484 +            }
   1.485 +            
   1.486 +            // Read lines from the file, encode/decode them, and highlight discrepancies.
   1.487 +            const int LINE_SIZE = 200;  // max line size
   1.488 +            char    line_buffer_utf8[LINE_SIZE];
   1.489 +            char    reencoded_utf8[6 * LINE_SIZE];
   1.490 +            uint32_t  line_buffer_ucs[LINE_SIZE];
   1.491 +            
   1.492 +            int byte_counter = 0;
   1.493 +            for (;;)
   1.494 +            {
   1.495 +                int c = fgetc(fp);
   1.496 +                if (c == EOF)
   1.497 +                {
   1.498 +                    // Done.
   1.499 +                    break;
   1.500 +                }
   1.501 +                line_buffer_utf8[byte_counter++] = c;
   1.502 +                if (c == '\n' || byte_counter >= LINE_SIZE - 2)
   1.503 +                {
   1.504 +                    // End of line.  Process the line.
   1.505 +                    line_buffer_utf8[byte_counter++] = 0;   // terminate.
   1.506 +                    
   1.507 +                    // Decode into UCS.
   1.508 +                    const char* p = line_buffer_utf8;
   1.509 +                    uint32_t* q = line_buffer_ucs;
   1.510 +                    for (;;)
   1.511 +                    {
   1.512 +                        uint32_t  uc = UTF8Util::DecodeNextChar(&p);
   1.513 +                        *q++ = uc;
   1.514 +                        
   1.515 +                        OVR_ASSERT(q < line_buffer_ucs + LINE_SIZE);
   1.516 +                        OVR_ASSERT(p < line_buffer_utf8 + LINE_SIZE);
   1.517 +                        
   1.518 +                        if (uc == 0) break;
   1.519 +                    }
   1.520 +                    
   1.521 +                    // Encode back into UTF-8.
   1.522 +                    q = line_buffer_ucs;
   1.523 +                    int index = 0;
   1.524 +                    for (;;)
   1.525 +                    {
   1.526 +                        uint32_t  uc = *q++;
   1.527 +                        OVR_ASSERT(index < LINE_SIZE * 6 - 6);
   1.528 +                        int last_index = index;
   1.529 +                        UTF8Util::EncodeChar(reencoded_utf8, &index, uc);
   1.530 +                        OVR_ASSERT(index <= last_index + 6);
   1.531 +                        if (uc == 0) break;
   1.532 +                    }
   1.533 +                    
   1.534 +                    // This can be useful for debugging.
   1.535 +#if 0
   1.536 +                    // Show the UCS and the re-encoded UTF-8.
   1.537 +                    log_ucs(line_buffer_ucs);
   1.538 +                    log_ascii(reencoded_utf8);
   1.539 +#endif // 0
   1.540 +                    
   1.541 +                    OVR_ASSERT(check_equal(line_buffer_utf8, line_buffer_ucs));
   1.542 +                    OVR_ASSERT(check_equal(reencoded_utf8, line_buffer_ucs));
   1.543 +                    
   1.544 +                    // Start next line.
   1.545 +                    byte_counter = 0;
   1.546 +                }
   1.547 +            }
   1.548 +            
   1.549 +            fclose(fp);
   1.550 +        }
   1.551 +        
   1.552 +        return 0;
   1.553 +}
   1.554 +
   1.555 +
   1.556 +#endif // UTF8_UNIT_TEST
   1.557 +
   1.558 +}} // namespace UTF8Util::OVR
   1.559 +