StdLib: Fix several problems where characters were not being correctly converted...

[mirror_edk2.git] / StdLib / LibC / Locale / multibyte_Utf8.c

/** @file\r
  Copyright (c) 2012, Intel Corporation. All rights reserved.<BR>\r
  This program and the accompanying materials\r
  are licensed and made available under the terms and conditions of the BSD License\r
  which accompanies this distribution.  The full text of the license may be found at\r
  http://opensource.org/licenses/bsd-license.php\r
\r
  THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" BASIS,\r
  WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED.\r
**/\r
#include  <assert.h>\r
#include  <string.h>\r
#include  <errno.h>\r
#include  <stdlib.h>\r
#include  <wchar.h>\r
#include  <sys/types.h>\r
\r
typedef      int      ch_UCS4;\r
\r
static  mbstate_t     LocalConvState = {0};\r
\r
/** Map a UTF-8 encoded prefix byte to a sequence length.\r
    Zero means illegal prefix, but valid surrogate if < 0xC0.\r
    One indicates an ASCII-7 equivalent character.\r
    Two, three, and four are the first byte for 2, 3, and 4 byte sequences, respectively.\r
    See RFC 3629 for details.\r
\r
  TABLE ENCODING:\r
    Low Nibble decodes the first byte into the number of bytes in the sequence.\r
      A value of zero indicates an invalid byte.\r
    The High Nibble encodes a bit mask to be used to match against the high nibble of the second byte.\r
\r
    example:\r
      SequenceLength = code[c0] & 0x0F;\r
      Mask           = 0x80 | code[c0];\r
\r
      Surrogate bytes are valid if: code[cX] & Mask > 0x80;\r
\r
*/\r
static\r
UINT8 utf8_code_length[256] = {\r
  0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, /* 00-0F */\r
  0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01,\r
  0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01,\r
  0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01,\r
  0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01,\r
  0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01,\r
  0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01,\r
  0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, /* 70-7F */\r
  0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, 0x90, /* 80-8F */\r
  0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, 0xA0, /* 90-9F */\r
  0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, /* A0-AF */\r
  0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, /* B0-BF */\r
  0x00, 0x00, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, /* C0-C1 + C2-CF */\r
  0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, 0x72, /* D0-DF */\r
  0x43, 0x73, 0x73, 0x73, 0x73, 0x73, 0x73, 0x73, 0x73, 0x73, 0x73, 0x73, 0x73, 0x33, 0x73, 0x73, /* E0-EF */\r
  0x64, 0x74, 0x74, 0x74, 0x14, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00  /* F0-F4 + F5-FF */\r
};\r
\r
/** Process one byte of a multibyte character.\r
\r
    @param[in]      ch    One byte of a multibyte character.\r
    @param[in,out]  ps    Pointer to a conversion state object.\r
\r
    @retval   -2      ch is an incomplete but potentially valid character.\r
    @retval   -1      ch is not valid in this context.\r
    @retval   1:4     The length, in bytes, of the character ch just completed.\r
**/\r
static\r
int\r
ProcessOneByte(unsigned char ch, mbstate_t *ps)\r
{\r
  UINT32    Mask;\r
  UINT32    Length;\r
  int       RetVal = 0;\r
\r
  if(ps->A > 3) {\r
    // We are in an invalid state\r
    ps->A = 0;    // Initial State\r
  }\r
  ps->C[ps->A] = ch;  // Save the current character\r
  Mask = utf8_code_length[ch];\r
\r
  if(ps->A == 0) {    // Initial State.  First byte of sequence.\r
    ps->E   = Mask | 0x80;\r
    Length  = Mask & 0xF;\r
    switch(Length) {\r
      case 0:                       // State 0, Code 0\r
        errno = EILSEQ;\r
        RetVal = -1;\r
        ps->E = 1;        // Consume this character\r
        break;\r
      case 1:                       // State 0, Code 1\r
        // ASCII-7 Character\r
        ps->B = ps->D[0] = ch;\r
        RetVal = 1;\r
        break;\r
      default:                      // State 0, Code 2, 3, 4\r
        ps->A = 1;    // Next state is State-1\r
        RetVal = -2;  // Incomplete but potentially valid character\r
        break;\r
    }\r
  }\r
  else {\r
    // We are in state 1, 2, or 3 and processing a surrogate byte\r
    Length  = ps->E & 0xF;\r
    if((Mask & ps->E) > 0x80) {\r
      // This byte is valid\r
      switch(ps->A) {   // Process based upon our current state\r
        case 1:             // Second byte of the sequence.\r
          if(Length == 2) {         // State 1, Code 2\r
            Length = ((ps->C[0] & 0x1f) << 6) + (ps->C[1] & 0x3f);\r
            assert ((Length > 0x007F) && (Length <= 0x07FF));\r
            ps->B = ps->D[0] = (UINT16)Length;\r
            ps->A = 0;      // Next state is State-0\r
            RetVal = 2;\r
          }\r
          else {    // This isn't the last character, get more.  State 1, Code 3 or 4\r
            ps->A = 2;\r
            RetVal = -2;\r
          }\r
          break;\r
        case 2:             // Third byte of the sequence\r
          if(Length == 3) {\r
            Length = ((ps->C[0] & 0x0f) << 12) + ((ps->C[1] & 0x3f) << 6) + (ps->C[2] & 0x3f);\r
            assert ((Length > 0x07FF) && (Length <= 0xFFFF));\r
            ps->B = ps->D[0] = (UINT16)Length;\r
            ps->A = 0;      // Next state is State-0\r
            RetVal = 3;\r
          }\r
          else {\r
            ps->A = 3;\r
            RetVal = -2;\r
          }\r
          break;\r
        case 3:             // Fourth byte of the sequence\r
          if(Length == 4) {\r
            Length = ((ps->C[0] & 0x7) << 18) + ((ps->C[1] & 0x3f) << 12) +\r
                     ((ps->C[2] & 0x3f) << 6) + (ps->C[3] & 0x3f);\r
            ps->B = Length;\r
            assert ((Length > 0xFFFF) && (Length <= 0x10ffff));\r
\r
            /*  compute and append the two surrogates: */\r
\r
            /*  translate from 10000..10FFFF to 0..FFFF */\r
            Length -= 0x10000;\r
\r
            /*  high surrogate = top 10 bits added to D800 */\r
            ps->D[0] = (UINT16)(0xD800 + (Length >> 10));\r
\r
            /*  low surrogate = bottom 10 bits added to DC00 */\r
            ps->D[1] = (UINT16)(0xDC00 + (Length & 0x03FF));\r
            ps->A = 0;      // Next state is State-0\r
            RetVal = 4;\r
          }\r
          else {\r
            errno = EILSEQ;\r
            ps->A = 0;\r
            RetVal = -1;\r
            ps->E = 4;      // Can't happen, but consume this character anyway\r
          }\r
          break;\r
      }\r
    }\r
    else {                // Invalid surrogate character\r
      errno = EILSEQ;\r
      ps->A = 0;          // Next is State-0\r
      RetVal = -1;\r
      ps->E = 0;            // Don't Consume, it may be an initial byte\r
    }\r
  }\r
  return RetVal;\r
}\r
\r
/** Convert one Multibyte sequence.\r
\r
    @param[out]   Dest      Pointer to output location, or NULL\r
    @param[in]    Src       Multibyte Source (UTF8)\r
    @param[in]    Len       Max Number of bytes to convert\r
    @param[in]    pS        Pointer to State struct., or NULL\r
\r
    @retval   -2      Bytes processed comprise an incomplete, but potentially valid, character.\r
    @retval   -1      An encoding error was encountered.  ps->E indicates the number of bytes consumed.\r
    @retval   0       Either Src is NULL or it points to a NUL character.\r
    @retval   1:N     N bytes were consumed producing a valid wide character.\r
**/\r
int\r
DecodeOneStateful(\r
  wchar_t    *Dest,       // Pointer to output location, or NULL\r
  const char *Src,        // Multibyte Source (UTF8)\r
  ssize_t     Len,        // Max Number of bytes to convert\r
  mbstate_t  *pS          // Pointer to State struct., or NULL\r
  )\r
{\r
  const char   *SrcEnd;\r
  int           NumConv;\r
  unsigned char ch;\r
\r
  if((Src == NULL) || (*Src == '\0')) {\r
    return 0;\r
  }\r
  if(pS == NULL) {\r
    pS = &LocalConvState;\r
  }\r
  SrcEnd  = Src + Len;\r
  NumConv = 0;\r
  while(Src < SrcEnd) {\r
    ch = (unsigned char)*Src++;\r
    NumConv = ProcessOneByte(ch, pS);\r
    if(NumConv != -2)\r
      break;\r
  }\r
  if((NumConv > 0) && (Dest != NULL)) {\r
    Dest[0] = pS->D[0];\r
    if(NumConv == 4) {\r
      Dest[1] = pS->D[1];\r
    }\r
  }\r
  return NumConv;\r
}\r
\r
/*  Determine the number of bytes needed to represent a Wide character\r
    as a MBCS character.\r
\r
    A single wide character may convert into a one, two, three, or four byte\r
    narrow (MBCS or UTF-8) character.  The number of MBCS bytes can be determined\r
    as follows.\r
\r
    If WCS char      < 0x00000080      One Byte\r
    Else if WCS char < 0x0000D800      Two Bytes\r
    Else                               Three Bytes\r
\r
    Since UEFI only supports the Unicode Base Multilingual Plane (BMP),\r
    Four-byte characters are not supported.\r
\r
    @param[in]    InCh      Wide character to test.\r
\r
    @retval     -1      Improperly formed character\r
    @retval      0      InCh is 0x0000\r
    @retval     >0      Number of bytes needed for the MBCS character\r
*/\r
int\r
EFIAPI\r
OneWcToMcLen(const wchar_t InCh)\r
{\r
  ssize_t   NumBytes;\r
\r
  if(InCh == 0) {             //    Is this a NUL, 0x0000 ?\r
    NumBytes = 0;\r
  }\r
  else if(InCh < 0x0080) {    //    Is this a 1-byte character?\r
    NumBytes = 1;\r
  }\r
  else if(InCh < 0x0800) {    //    Is this a 2-byte character?\r
    NumBytes = 2;\r
  }\r
  else if((InCh >= 0xD800) && (InCh < 0xE000)) {    //    Is this a surrogate?\r
    NumBytes = -1;\r
  }\r
  else {\r
    NumBytes = 3;             //    Otherwise, it must be a 3-byte character.\r
  }\r
  return (int)NumBytes;      // Return extimate of required bytes.\r
}\r
\r
/*  Determine the number of bytes needed to represent a Wide character string\r
    as a MBCS string of given maximum length.  Will optionally return the number\r
    of wide characters that would be consumed.\r
\r
    A single wide character may convert into a one, two, three, or four byte\r
    narrow (MBCS or UTF-8) character.  The number of MBCS bytes can be determined\r
    as follows.\r
\r
    If WCS char      < 0x00000080      One Byte\r
    Else if WCS char < 0x00000800      Two Bytes\r
    Else if WCS char < 0x00010000      Three Bytes\r
    Else                               Four Bytes\r
\r
    Since UEFI only supports the Unicode Base Multilingual Plane (BMP),\r
    Four-byte characters should not be encountered.\r
\r
    @param[in]    Src       Pointer to a wide character string.\r
    @param[in]    Limit     Maximum number of bytes the converted string may occupy.\r
    @param[out]   NumChar   Pointer to where to store the number of wide characters, or NULL.\r
\r
    @return     The number of bytes required to convert Src to MBCS,\r
                not including the terminating NUL.  If NumChar is not NULL, the number\r
                of characters represented by the return value will be written to\r
                where it points.\r
*/\r
size_t\r
EFIAPI\r
EstimateWtoM(const wchar_t * Src, size_t Limit, size_t *NumChar)\r
{\r
  ssize_t    Estimate;\r
  size_t    CharCount;\r
  ssize_t   NumBytes;\r
  wchar_t   EChar;\r
\r
  Estimate  = 0;\r
  CharCount = 0;\r
  EChar = *Src++;               // Get the initial character and point to next\r
  while(((NumBytes = OneWcToMcLen(EChar)) > 0)  &&\r
        ((size_t)(Estimate + NumBytes) < Limit))\r
  {                             // Until one of the source characters is NUL\r
    ++CharCount;                //    Count this character.\r
    Estimate += NumBytes;       //    Count the Bytes for this character\r
    EChar = *Src++;             //    Get the next source character and point to the next.\r
  }\r
  if(NumChar != NULL) {\r
    *NumChar = CharCount;\r
  }\r
  return (size_t)Estimate;      // Return esimate of required bytes.\r
}\r
\r
/*  Determine the number of characters in a MBCS string.\r
    MBCS characters are one to four bytes long.  By examining the first byte\r
    of a MBCS character, one can determine the number of bytes comprising the\r
    character.\r
\r
    0x00 - 0x7F     One\r
    0xC0 - 0xDF     Two\r
    0xE0 - 0xEF     Three\r
    0xF0 - 0xF7     Four\r
\r
    Since UEFI only supports the Unicode Base Multilingual Plane (BMP),\r
    Four-byte characters should not be encountered.\r
\r
    @param[in]    Src     The string to examine\r
\r
    @return   The number of characters represented by the MBCS string.\r
**/\r
size_t\r
EFIAPI\r
CountMbcsChars(const char *Src)\r
{\r
  size_t      Count;\r
  char        EChar;\r
\r
  Count = 0;\r
  EChar = *Src++;\r
  while(EChar != 0) {\r
    if(EChar < 0x80) {\r
      ++Count;\r
    }\r
    else if(EChar < 0xE0) {\r
      Count += 2;\r
      ++Src;\r
    }\r
    else if(EChar < 0xF0) {\r
      Count += 3;\r
      Src += 2;\r
    }\r
    else {\r
      // Ill-formed character\r
      break;\r
    }\r
  }\r
  return Count;\r
}\r
\r
/** Convert a wide character (UTF16) into a multibyte character (UTF8)\r
\r
    Converts a wide character into a corresponding multibyte character that\r
    begins in the conversion state described by the object pointed to by ps.\r
    If dst is not a null pointer, the converted character is then stored into\r
    the array pointed to by dst.\r
\r
    It is the caller's responsibility to ensure that Dest is large enough to\r
    hold the resulting MBCS sequence.\r
\r
    @param  s       Pointer to the wide-character string to convert\r
    @param  Dest    Pointer to the buffer in which to place the converted sequence, or NULL.\r
\r
    @retval   -1    An error occurred.  The error reason is in errno.\r
    @retval   >=0   The number of bytes stored into Dest.\r
**/\r
ssize_t\r
EncodeUtf8(char *Dest, wchar_t ch)\r
{\r
  char       *p;              /* next free byte in build buffer */\r
  int         NumInBuff;      // number of bytes in Buff\r
  char        Buff[4];        // Buffer into which each character is built\r
\r
    p = Buff;\r
\r
  NumInBuff = 0;\r
  if (ch < 0x80) {\r
    /* Encode ASCII -- One Byte */\r
    *p++ = (char) ch;\r
    NumInBuff = 1;\r
  }\r
  else if (ch < 0x0800) {\r
    /* Encode Latin-1 -- Two Byte */\r
    *p++ = (char)(0xc0 | (ch >> 6));\r
    *p++ = (char)(0x80 | (ch & 0x3f));\r
    NumInBuff = 2;\r
  }\r
  else {\r
      /* Encode UCS2 Unicode ordinals -- Three Byte */\r
    /* Special case: check for surrogate -- Shouldn't happen in UEFI */\r
    if (0xD800 <= ch && ch < 0xE000) {\r
      errno = EILSEQ;\r
      return -1;\r
      }\r
    else {\r
      *p++ = (char)(0xe0 | (ch >> 12));\r
      *p++ = (char)(0x80 | ((ch >> 6) & 0x3f));\r
      *p++ = (char)(0x80 | (ch & 0x3f));\r
      NumInBuff = 3;\r
    }\r
  }\r
  /*  At this point, Buff holds the converted character which is NumInBuff bytes long.\r
      NumInBuff is the value 1, 2, 3, or 4\r
  */\r
  if(Dest != NULL) {        // Save character if Dest is not NULL\r
    memcpy(Dest, Buff, NumInBuff);\r
\r
    if(ch != 0) {\r
      // Terminate the destination string.\r
      Dest[NumInBuff] = '\0';\r
    }\r
    else {\r
      NumInBuff = 0;\r
    }\r
  }\r
  return NumInBuff;             // Tell the caller\r
}\r
\r
// ########################  Narrow to Wide Conversions #######################\r
\r
/** If ps is not a null pointer, the mbsinit function determines whether the\r
    pointed-to mbstate_t object describes an initial conversion state.\r
\r
    @param[in]  ps    Pointer to the conversion state object to test.\r
\r
    @return     The mbsinit function returns nonzero if ps is a null pointer\r
                or if the pointed-to object describes an initial conversion\r
                state; otherwise, it returns zero.\r
\r
    Declared in: wchar.h\r
**/\r
int\r
mbsinit(const mbstate_t *ps)\r
{\r
  if((ps == NULL) || (ps->A == 0)) {\r
    return 1;\r
  }\r
  return 0;\r
}\r
\r
/** The mbrlen function is equivalent to the call:<BR>\r
@verbatim\r
    mbrtowc(NULL, s, n, ps != NULL ? ps : &internal)\r
@endverbatim\r
    where internal is the mbstate_t object for the mbrlen function, except that\r
    the expression designated by ps is evaluated only once.\r
\r
    @param[in]  s     Pointer to a multibyte character sequence.\r
    @param[in]  n     Maximum number of bytes to examine.\r
    @param[in]  pS    Pointer to the conversion state object.\r
\r
    @retval   0       The next n or fewer characters complete a NUL.\r
    @retval   1..n    The number of bytes that complete the multibyte character.\r
    @retval   -2      The next n bytes contribute to an incomplete (but potentially valid) multibyte character.\r
    @retval   -1      An encoding error occurred.\r
\r
    Declared in: wchar.h\r
**/\r
size_t\r
mbrlen(\r
  const char *s,\r
  size_t n,\r
  mbstate_t *pS\r
  )\r
{\r
  return mbrtowc(NULL, s, n, pS);\r
}\r
\r
/** Determine the number of bytes comprising a multibyte character.\r
\r
  If S is not a null pointer, the mblen function determines the number of bytes\r
  contained in the multibyte character pointed to by S. Except that the\r
  conversion state of the mbtowc function is not affected, it is equivalent to\r
    mbtowc((wchar_t *)0, S, N);\r
\r
  @param[in]  S   NULL to query whether multibyte characters have\r
                  state-dependent encodings.  Otherwise, points to a\r
                  multibyte character.\r
  @param[in]  N   The maximum number of bytes in a multibyte character.\r
\r
  @return   If S is a null pointer, the mblen function returns a nonzero or\r
            zero value, if multibyte character encodings, respectively, do\r
            or do not have state-dependent encodings. If S is not a null\r
            pointer, the mblen function either returns 0 (if S points to the\r
            null character), or returns the number of bytes that are contained\r
            in the multibyte character (if the next N or fewer bytes form a\r
            valid multibyte character), or returns -1 (if they do not form a\r
            valid multibyte character).\r
\r
    Declared in: stdlib.h\r
**/\r
int\r
mblen(\r
  const char *s,\r
  size_t n\r
  )\r
{\r
  return (int)mbrlen(s, n, NULL);\r
}\r
\r
/**\r
If S is a null pointer, the mbrtowc function is equivalent to the call:<BR>\r
@verbatim\r
        mbrtowc(NULL, "", 1, ps)\r
@endverbatim\r
\r
In this case, the values of the parameters pwc and n are ignored.\r
\r
If S is not a null pointer, the mbrtowc function inspects at most n bytes beginning with\r
the byte pointed to by S to determine the number of bytes needed to complete the next\r
multibyte character (including any shift sequences). If the function determines that the\r
next multibyte character is complete and valid, it determines the value of the\r
corresponding wide character and then, if pwc is not a null pointer, stores that value in\r
the object pointed to by pwc. If the corresponding wide character is the null wide\r
character, the resulting state described is the initial conversion state.\r
\r
    @param[out]   pwc   Pointer to where the resulting wide character is to be stored.\r
    @param[in]     s    Pointer to a multibyte character "string".\r
    @param[in]     n    The maximum number of bytes to inspect.\r
    @param[in]     ps   Pointer to a conversion state object.\r
\r
    @retval   0             if the next n or fewer bytes complete the multibyte\r
                            character that corresponds to the null wide\r
                            character (which is the value stored).\r
    @retval   between_1_and_n_inclusive   if the next n or fewer bytes complete\r
                            a valid multibyte character (which is the value\r
                            stored); the value returned is the number of bytes\r
                            that complete the multibyte character.\r
    @retval   (size_t)(-2)  if the next n bytes contribute to an incomplete\r
                            (but potentially valid) multibyte character, and\r
                            all n bytes have been processed (no value is stored).\r
    @retval   (size_t)(-1)  if an encoding error occurs, in which case the next\r
                            n or fewer bytes do not contribute to a complete and\r
                            valid multibyte character (no value is stored); the\r
                            value of the macro EILSEQ is stored in errno, and\r
                            the conversion state is unspecified.\r
\r
    Declared in: wchar.h\r
**/\r
size_t\r
mbrtowc(\r
  wchar_t *pwc,\r
  const char *s,\r
  size_t n,\r
  mbstate_t *ps\r
  )\r
{\r
  int     RetVal;\r
\r
  RetVal = DecodeOneStateful(pwc, s, (ssize_t)n, ps);\r
  return (size_t)RetVal;\r
}\r
\r
/** Convert a multibyte character into a wide character.\r
\r
    If S is not a null pointer, the mbtowc function inspects at most N bytes\r
    beginning with the byte pointed to by S to determine the number of bytes\r
    needed to complete the next multibyte character (including any shift\r
    sequences). If the function determines that the next multibyte character\r
    is complete and valid, it determines the value of the corresponding wide\r
    character and then, if Pwc is not a null pointer, stores that value in\r
    the object pointed to by Pwc. If the corresponding wide character is the\r
    null wide character, the function is left in the initial conversion state.\r
\r
    @param[out]   Pwc Pointer to a wide-character object to receive the converted character.\r
    @param[in]    S   Pointer to a multibyte character to convert.\r
    @param[in]    N   Maximum number of bytes in a multibyte character.\r
\r
    @return   If S is a null pointer, the mbtowc function returns a nonzero or\r
              zero value, if multibyte character encodings, respectively, do\r
              or do not have state-dependent encodings. If S is not a null\r
              pointer, the mbtowc function either returns 0 (if S points to\r
              the null character), or returns the number of bytes that are\r
              contained in the converted multibyte character (if the next N or\r
              fewer bytes form a valid multibyte character), or returns -1\r
              (if they do not form a valid multibyte character).\r
\r
              In no case will the value returned be greater than N or the value\r
              of the MB_CUR_MAX macro.\r
\r
    Declared in: stdlib.h\r
**/\r
int\r
mbtowc(\r
  wchar_t *pwc,\r
  const char *s,\r
  size_t n\r
  )\r
{\r
  return (int)mbrtowc(pwc, s, n, NULL);\r
}\r
\r
/**\r
The mbsrtowcs function converts a sequence of multibyte characters that begins in the\r
conversion state described by the object pointed to by ps, from the array indirectly\r
pointed to by src into a sequence of corresponding wide characters. If dst is not a null\r
pointer, the converted characters are stored into the array pointed to by dst. Conversion\r
continues up to and including a terminating null character, which is also stored.\r
Conversion stops earlier in two cases: when a sequence of bytes is encountered that does\r
not form a valid multibyte character, or (if dst is not a null pointer) when len wide\r
characters have been stored into the array pointed to by dst. Each conversion takes\r
place as if by a call to the mbrtowc function.\r
\r
If dst is not a null pointer, the pointer object pointed to by src is assigned either a null\r
pointer (if conversion stopped due to reaching a terminating null character) or the address\r
just past the last multibyte character converted (if any). If conversion stopped due to\r
reaching a terminating null character and if dst is not a null pointer, the resulting state\r
described is the initial conversion state.\r
\r
    @param[out]   dst   Pointer to where the resulting wide character sequence is stored.\r
    @param[in]    src   Pointer to a pointer to the multibyte character sequence to convert.\r
    @param[in]    len   Maximum number of wide characters to be stored into dst.\r
    @param[in]    ps    Pointer to a conversion state object.\r
\r
    @return   If the input conversion encounters a sequence of bytes that do\r
              not form a valid multibyte character, an encoding error occurs:\r
              the mbsrtowcs function stores the value of the macro EILSEQ in\r
              errno and returns (size_t)(-1); the conversion state is\r
              unspecified. Otherwise, it returns the number of multibyte\r
              characters successfully converted, not including the terminating\r
              null character (if any).\r
\r
    Declared in: wchar.h\r
**/\r
size_t\r
mbsrtowcs(\r
  wchar_t      *dst,\r
  const char  **src,\r
  size_t        len,\r
  mbstate_t    *ps\r
  )\r
{\r
  int           x;\r
  size_t        RetVal = 0;\r
  const char   *MySrc;\r
\r
  if((src == NULL) || (*src == NULL) || (**src == '\0')) {\r
    return 0;\r
  }\r
\r
  MySrc = *src;\r
  for(x = 1 ; (len != 0) && (x > 0); --len) {\r
    x = DecodeOneStateful(dst, MySrc, MB_LEN_MAX, ps);\r
    switch(x) {\r
      case -2:    // Incomplete character\r
      case -1:    // Encoding error\r
        RetVal = (size_t)x;\r
        break;\r
      case 0:     // Encountered NUL character: done.\r
        if(dst != NULL) {\r
          *dst = 0;\r
          *src = NULL;\r
        }\r
        break;\r
      default:    // Successfully decoded a character, continue with next\r
        MySrc += x;\r
        if(dst != NULL) {\r
          ++dst;\r
          if(x == 4) {\r
            ++dst;\r
          }\r
          *src = MySrc;\r
        }\r
        ++RetVal;\r
        break;\r
    }\r
  }\r
  return RetVal;\r
}\r
\r
/** Convert a multibyte character string into a wide-character string.\r
\r
    The mbstowcs function converts a sequence of multibyte characters that\r
    begins in the initial shift state from the array pointed to by Src into\r
    a sequence of corresponding wide characters and stores not more than limit\r
    wide characters into the array pointed to by Dest.  No multibyte\r
    characters that follow a null character (which is converted into a null\r
    wide character) will be examined or converted. Each multibyte character\r
    is converted as if by a call to the mbtowc function, except that the\r
    conversion state of the mbtowc function is not affected.\r
\r
    No more than Limit elements will be modified in the array pointed to by Dest.\r
    If copying takes place between objects that overlap,\r
    the behavior is undefined.\r
\r
    @param[out]   Dest    Pointer to the array to receive the converted string.\r
    @param[in]    Src     Pointer to the string to be converted.\r
    @param[in]    Limit   Maximum number of elements to be written to Dest.\r
\r
    @return   If an invalid multibyte character is encountered, the mbstowcs\r
              function returns (size_t)(-1). Otherwise, the mbstowcs function\r
              returns the number of array elements modified, not including a\r
              terminating null wide character, if any.\r
\r
    Declared in: stdlib.h\r
**/\r
size_t\r
mbstowcs(\r
  wchar_t *Dest,\r
  const char *Src,\r
  size_t Limit\r
  )\r
{\r
\r
  /* Dest may be NULL */\r
  /* Src may be NULL */\r
\r
  return mbsrtowcs(Dest, &Src, Limit, NULL);\r
}\r
\r
/** The btowc function determines whether C constitutes a valid single-byte\r
    character in the initial shift state.\r
\r
    @param[in]    C   A narrow character to test or convert to wide.\r
\r
    @return   The btowc function returns WEOF if c has the value EOF or if\r
              (unsigned char)C does not constitute a valid single-byte\r
              character in the initial shift state. Otherwise, it returns the\r
              wide character representation of that character.\r
\r
    Declared in: wchar.h\r
**/\r
wint_t\r
btowc(int c)\r
{\r
  int       x;\r
  wchar_t   Dest;\r
  wint_t    RetVal = WEOF;\r
\r
  if (c == EOF)\r
    return WEOF;\r
  x = DecodeOneStateful(&Dest, (const char *)&c, 1, NULL);\r
  if(x == 0) {\r
    RetVal = 0;\r
  }\r
  else if(x == 1) {\r
    RetVal = (wint_t)Dest;\r
  }\r
  return RetVal;\r
}\r
\r
// ########################  Wide to Narrow Conversions #######################\r
\r
/**\r
If S is a null pointer, the wcrtomb function is equivalent to the call:<BR>\r
@verbatim\r
        wcrtomb(buf, L'\0', ps)\r
@endverbatim\r
where buf is an internal buffer.\r
\r
If S is not a null pointer, the wcrtomb function determines the number of bytes needed\r
to represent the multibyte character that corresponds to the wide character given by wc\r
(including any shift sequences), and stores the multibyte character representation in the\r
array whose first element is pointed to by S. At most MB_CUR_MAX bytes are stored. If\r
wc is a null wide character, a null byte is stored, preceded by any shift sequence needed\r
to restore the initial shift state; the resulting state described is the initial conversion state.\r
\r
    @param[out]     Dest    Pointer to the location in which to store the resulting\r
                            multibyte character.  Otherwise, NULL to reset the\r
                            conversion state.\r
    @param[in]      wchar   The wide character to convert.\r
    @param[in,out]  pS      Pointer to a conversion state object, or NULL.\r
\r
    @return   The wcrtomb function returns the number of bytes stored in the\r
              array object (including any shift sequences). When wc is not a\r
              valid wide character, an encoding error occurs: the function\r
              stores the value of the macro EILSEQ in errno and\r
              returns (size_t)(-1); the conversion state is unspecified.\r
\r
    Declared in: wchar.h\r
**/\r
size_t\r
wcrtomb(\r
  char *Dest,\r
  wchar_t wchar,\r
  mbstate_t *pS\r
  )\r
{\r
  size_t    RetVal;\r
\r
  /* Dest may be NULL */\r
  if (Dest == NULL) {\r
    RetVal = 1;\r
  }\r
  else {\r
    if (wchar == L'\0') {\r
      *Dest = '\0';\r
      RetVal = 1;\r
    }\r
    else {\r
      RetVal = EncodeUtf8(Dest, wchar);\r
    }\r
  }\r
  if(pS == NULL) {\r
    pS = &LocalConvState;\r
  }\r
  pS->A = 0;      // Set ps to the initial conversion state\r
\r
  return RetVal;\r
}\r
\r
/** Convert a wide character into a multibyte character.\r
\r
    The wctomb function determines the number of bytes needed to represent the\r
    multibyte character corresponding to the wide character given by WC\r
    (including any shift sequences), and stores the multibyte character\r
    representation in the array whose first element is pointed to by S (if S is\r
    not a null pointer). At most MB_CUR_MAX characters are stored. If WC is a\r
    null wide character, a null byte is stored, preceded by any shift sequence\r
    needed to restore the initial shift state, and the function is left in the\r
    initial conversion state.\r
\r
    @param[out]   S   Pointer to the object to receive the converted multibyte character.\r
    @param[in]    WC  Wide character to be converted.\r
\r
    @return   If S is a null pointer, the wctomb function returns a nonzero or\r
              zero value, if multibyte character encodings, respectively, do or\r
              do not have state-dependent encodings. If S is not a null pointer,\r
              the wctomb function returns -1 if the value of WC does not\r
              correspond to a valid multibyte character, or returns the number\r
              of bytes that are contained in the multibyte character\r
              corresponding to the value of WC.\r
\r
              In no case will the value returned be greater than the value of\r
              the MB_CUR_MAX macro.\r
\r
    Declared in: stdlib.h\r
**/\r
int\r
wctomb(\r
  char *s,\r
  wchar_t wchar\r
  )\r
{\r
  /*\r
    If s is NULL just return whether MB Characters have state\r
    dependent encodings -- they don't.\r
  */\r
  if (s == NULL)\r
    return 0;\r
\r
  return (int)wcrtomb(s, wchar, NULL);\r
}\r
\r
/** The wcsrtombs function converts a sequence of wide characters from the array\r
    indirectly pointed to by Dest into a sequence of corresponding multibyte\r
    characters that begins in the conversion state described by the object\r
    pointed to by ps.\r
\r
    If Dest is not a null pointer, the converted characters are stored into the\r
    array pointed to by Dest.  Conversion continues up to and including a\r
    terminating null wide character, which is also stored. Conversion stops\r
    earlier in two cases: when a wide character is reached that does not\r
    correspond to a valid multibyte character, or (if Dest is not a null\r
    pointer) when the next multibyte character would exceed the limit of Limit\r
    total bytes to be stored into the array pointed to by Dest. Each conversion\r
    takes place as if by a call to the wcrtomb function.)\r
\r
    If Dest is not a null pointer, the pointer object pointed to by Src is\r
    assigned either a null pointer (if conversion stopped due to reaching\r
    a terminating null wide character) or the address just past the last wide\r
    character converted (if any). If conversion stopped due to reaching a\r
    terminating null wide character, the resulting state described is the\r
    initial conversion state.\r
\r
    @param[in]      Dest\r
    @param[in,out]  Src\r
    @param[in]      Limit   Max number of bytes to store in Dest.\r
    @param[in,out]  ps\r
\r
    @return     If conversion stops because a wide character is reached that\r
                does not correspond to a valid multibyte character, an\r
                encoding error occurs: the wcsrtombs function stores the\r
                value of the macro EILSEQ in errno and returns (size_t)(-1);\r
                the conversion state is unspecified. Otherwise, it returns\r
                the number of bytes in the resulting multibyte character\r
                sequence, not including the terminating null character (if any).\r
\r
    Declared in: wchar.h\r
**/\r
size_t\r
wcsrtombs(\r
  char           *Dest,\r
  const wchar_t **Src,\r
  size_t          Limit,\r
  mbstate_t      *ps\r
)\r
{\r
  size_t  NumStored;\r
  ssize_t MaxBytes;\r
  int     count;\r
  wchar_t InCh;\r
\r
  NumStored = 0;\r
  MaxBytes  = (ssize_t)Limit;\r
\r
  /* Dest may be NULL */\r
  /* Src may be NULL */\r
  /* ps appears to be unused */\r
\r
  if (Src == NULL || *Src == NULL)\r
    return (0);\r
\r
  if (Dest == NULL) {\r
    NumStored = EstimateWtoM(*Src, MaxBytes, NULL);\r
  }\r
  else {\r
    while (OneWcToMcLen(InCh = *(*Src)++) <= MaxBytes) {\r
      if(InCh == 0) {\r
        *Src = NULL;\r
        break;\r
      }\r
      count = (int)wcrtomb(Dest, InCh, NULL);\r
      if(count >= 0) {\r
        Dest += count;\r
        MaxBytes -= count;\r
        NumStored += count;\r
      }\r
      else {\r
        NumStored = (size_t)(-1);\r
      }\r
    }\r
  }\r
\r
\r
  return NumStored;\r
}\r
\r
/** Convert a wide-character string into a multibyte character string.\r
\r
    The wcstombs function converts a sequence of wide characters from the\r
    array pointed to by Src into a sequence of corresponding multibyte\r
    characters that begins in the initial shift state, and stores these\r
    multibyte characters into the array pointed to by Dest, stopping if a\r
    multibyte character would exceed the limit of Limit total bytes or if a\r
    null character is stored. Each wide character is converted as if by\r
    a call to the wctomb function, except that the conversion state of\r
    the wctomb function is not affected.\r
\r
    No more than Limit bytes will be modified in the array pointed to by Dest.\r
    If copying takes place between objects that overlap,\r
    the behavior is undefined.\r
\r
    @param[out]   Dest    Pointer to the array to receive the converted string.\r
    @param[in]    Src     Pointer to the string to be converted.\r
    @param[in]    Limit   Maximum number of elements to be written to Dest.\r
\r
    @return   If a wide character is encountered that does not correspond to a\r
              valid multibyte character, the wcstombs function returns\r
              (size_t)(-1). Otherwise, the wcstombs function returns the number\r
              of bytes modified, not including a terminating null character,\r
              if any.\r
\r
    Declared in: stdlib.h\r
**/\r
size_t\r
wcstombs(\r
  char           *Dest,\r
  const wchar_t  *Src,\r
  size_t          Limit\r
)\r
{\r
  /* Dest may be NULL */\r
  return wcsrtombs(Dest, &Src, Limit, NULL);\r
}\r
\r
/** The wctob function determines whether C corresponds to a member of the extended\r
    character set whose multibyte character representation is a single byte when in the initial\r
    shift state.\r
\r
    wctob needs to be consistent with wcrtomb.\r
    If wcrtomb says that a character is representable in 1 byte,\r
    then wctob needs to also represent the character as 1 byte.\r
\r
    @return     The wctob function returns EOF if C does not correspond to a multibyte\r
                character with length one in the initial shift state. Otherwise, it\r
                returns the single-byte representation of that character as an\r
                unsigned char converted to an int.\r
\r
    Declared in: wchar.h\r
**/\r
int\r
wctob(wint_t c)\r
{\r
  int   RetVal;\r
\r
  RetVal = EOF;\r
  if(c == 0) {\r
    RetVal = 0;\r
  }\r
  else if (OneWcToMcLen((const wchar_t)c) == 1) {\r
    RetVal = (int)(c & 0xFF);\r
  }\r
  return RetVal;\r
}\r