/*++ Copyright (c) 1991-2000, Microsoft Corporation All rights reserved. Module Name: mbcs.c Abstract: This file contains functions that convert multibyte character strings to wide character strings, convert wide character strings to multibyte character strings, convert a multibyte character string from one code page to a multibyte character string of another code page, and get the DBCS leadbyte ranges for a given code page. APIs found in this file: IsValidCodePage GetACP GetOEMCP GetCPInfo GetCPInfoExW IsDBCSLeadByte IsDBCSLeadByteEx MultiByteToWideChar WideCharToMultiByte Revision History: 05-31-91 JulieB Created. --*/ // // Include Files. // #include "nls.h" // // Forward Declarations. // int GetWCCompSB( PMB_TABLE pMBTbl, LPBYTE pMBStr, LPWSTR pWCStr, LPWSTR pEndWCStr); int GetWCCompMB( PCP_HASH pHashN, PMB_TABLE pMBTbl, LPBYTE pMBStr, LPBYTE pEndMBStr, LPWSTR pWCStr, LPWSTR pEndWCStr, int *pmbIncr); int GetWCCompSBErr( PCP_HASH pHashN, PMB_TABLE pMBTbl, LPBYTE pMBStr, LPWSTR pWCStr, LPWSTR pEndWCStr); int GetWCCompMBErr( PCP_HASH pHashN, PMB_TABLE pMBTbl, LPBYTE pMBStr, LPBYTE pEndMBStr, LPWSTR pWCStr, LPWSTR pEndWCStr, int *pmbIncr); int GetMBNoDefault( PCP_HASH pHashN, LPWSTR pWCStr, LPWSTR pEndWCStr, LPBYTE pMBStr, int cbMultiByte, DWORD dwFlags); int GetMBDefault( PCP_HASH pHashN, LPWSTR pWCStr, LPWSTR pEndWCStr, LPBYTE pMBStr, int cbMultiByte, WORD wDefault, LPBOOL pUsedDef, DWORD dwFlags); int GetMBDefaultComp( PCP_HASH pHashN, LPWSTR pWCStr, LPWSTR pEndWCStr, LPBYTE pMBStr, int cbMultiByte, WORD wDefault, LPBOOL pUsedDef, DWORD dwFlags); int GetMBCompSB( PCP_HASH pHashN, DWORD dwFlags, LPWSTR pWCStr, LPBYTE pMBStr, int mbCount, WORD wDefault, LPBOOL pUsedDef); int GetMBCompMB( PCP_HASH pHashN, DWORD dwFlags, LPWSTR pWCStr, LPBYTE pMBStr, int mbCount, WORD wDefault, LPBOOL pUsedDef, BOOL *fError, BOOL fOnlyOne); UINT GetMacCodePage(void); //-------------------------------------------------------------------------// // INTERNAL MACROS // //-------------------------------------------------------------------------// //////////////////////////////////////////////////////////////////////////// // // CHECK_DBCS_LEAD_BYTE // // Returns the offset to the DBCS table for the given leadbyte character. // If the given character is not a leadbyte, then it returns zero (table // value). // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define CHECK_DBCS_LEAD_BYTE(pDBCSOff, Ch) \ (pDBCSOff ? ((WORD)(pDBCSOff[Ch])) : ((WORD)0)) //////////////////////////////////////////////////////////////////////////// // // CHECK_ERROR_WC_SINGLE // // Checks to see if the default character was used due to an invalid // character. Sets last error and returns 0 characters written if an // invalid character was used. // // NOTE: This macro may return if an error is encountered. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define CHECK_ERROR_WC_SINGLE( pHashN, \ wch, \ Ch ) \ { \ if ( ( (wch == pHashN->pCPInfo->wUniDefaultChar) && \ (Ch != pHashN->pCPInfo->wTransUniDefaultChar) ) || \ ( (wch >= PRIVATE_USE_BEGIN) && (wch <= PRIVATE_USE_END) ) ) \ { \ SetLastError(ERROR_NO_UNICODE_TRANSLATION); \ return (0); \ } \ } //////////////////////////////////////////////////////////////////////////// // // CHECK_ERROR_WC_MULTI // // Checks to see if the default character was used due to an invalid // character. Sets last error and returns 0 characters written if an // invalid character was used. // // NOTE: This macro may return if an error is encountered. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define CHECK_ERROR_WC_MULTI( pHashN, \ wch, \ lead, \ trail ) \ { \ if ((wch == pHashN->pCPInfo->wUniDefaultChar) && \ (MAKEWORD(trail, lead) != pHashN->pCPInfo->wTransUniDefaultChar)) \ { \ SetLastError(ERROR_NO_UNICODE_TRANSLATION); \ return (0); \ } \ } //////////////////////////////////////////////////////////////////////////// // // CHECK_ERROR_WC_MULTI_SPECIAL // // Checks to see if the default character was used due to an invalid // character. Sets it to 0xffff if invalid. // // DEFINED AS A MACRO. // // 08-21-95 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define CHECK_ERROR_WC_MULTI_SPECIAL( pHashN, \ pWCStr, \ lead, \ trail ) \ { \ if ((*pWCStr == pHashN->pCPInfo->wUniDefaultChar) && \ (MAKEWORD(trail, lead) != pHashN->pCPInfo->wTransUniDefaultChar)) \ { \ *pWCStr = 0xffff; \ } \ } //////////////////////////////////////////////////////////////////////////// // // GET_WC_SINGLE // // Fills in pWCStr with the wide character(s) for the corresponding single // byte character from the appropriate translation table. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_WC_SINGLE( pMBTbl, \ pMBStr, \ pWCStr ) \ { \ *pWCStr = pMBTbl[*pMBStr]; \ } //////////////////////////////////////////////////////////////////////////// // // GET_WC_SINGLE_SPECIAL // // Fills in pWCStr with the wide character(s) for the corresponding single // byte character from the appropriate translation table. Also checks for // invalid characters - if invalid, it fills in 0xffff instead. // // DEFINED AS A MACRO. // // 08-21-95 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_WC_SINGLE_SPECIAL( pHashN, \ pMBTbl, \ pMBStr, \ pWCStr ) \ { \ *pWCStr = pMBTbl[*pMBStr]; \ \ if ( ( (*pWCStr == pHashN->pCPInfo->wUniDefaultChar) && \ (*pMBStr != pHashN->pCPInfo->wTransUniDefaultChar) ) || \ ( (*pWCStr >= PRIVATE_USE_BEGIN) && \ (*pWCStr <= PRIVATE_USE_END) ) ) \ { \ *pWCStr = 0xffff; \ } \ } //////////////////////////////////////////////////////////////////////////// // // GET_WC_MULTI // // Fills in pWCStr with the wide character(s) for the corresponding multibyte // character from the appropriate translation table. The number of bytes // used from the pMBStr buffer (single byte or double byte) is stored in // the mbIncr parameter. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_WC_MULTI( pHashN, \ pMBTbl, \ pMBStr, \ pEndMBStr, \ pWCStr, \ pEndWCStr, \ mbIncr ) \ { \ WORD Offset; /* offset to DBCS table for range */ \ \ \ if (Offset = CHECK_DBCS_LEAD_BYTE(pHashN->pDBCSOffsets, *pMBStr)) \ { \ /* \ * DBCS Lead Byte. Make sure there is a trail byte with the \ * lead byte. \ */ \ if (pMBStr + 1 == pEndMBStr) \ { \ /* \ * There is no trail byte with the lead byte. The lead byte \ * is the LAST character in the string. Translate to NULL. \ */ \ *pWCStr = (WCHAR)0; \ mbIncr = 1; \ } \ else if (*(pMBStr + 1) == 0) \ { \ /* \ * There is no trail byte with the lead byte. The lead byte \ * is followed by a NULL. Translate to NULL. \ * \ * Increment by 2 so that the null is not counted twice. \ */ \ *pWCStr = (WCHAR)0; \ mbIncr = 2; \ } \ else \ { \ /* \ * Fill in the wide character translation from the double \ * byte character table. \ */ \ *pWCStr = (pHashN->pDBCSOffsets + Offset)[*(pMBStr + 1)]; \ mbIncr = 2; \ } \ } \ else \ { \ /* \ * Not DBCS Lead Byte. Fill in the wide character translation \ * from the single byte character table. \ */ \ *pWCStr = pMBTbl[*pMBStr]; \ mbIncr = 1; \ } \ } //////////////////////////////////////////////////////////////////////////// // // GET_WC_MULTI_ERR // // Fills in pWCStr with the wide character(s) for the corresponding multibyte // character from the appropriate translation table. The number of bytes // used from the pMBStr buffer (single byte or double byte) is stored in // the mbIncr parameter. // // Once the character has been translated, it checks to be sure the // character was valid. If not, it sets last error and return 0 characters // written. // // NOTE: This macro may return if an error is encountered. // // DEFINED AS A MACRO. // // 09-01-93 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_WC_MULTI_ERR( pHashN, \ pMBTbl, \ pMBStr, \ pEndMBStr, \ pWCStr, \ pEndWCStr, \ mbIncr ) \ { \ WORD Offset; /* offset to DBCS table for range */ \ \ \ if (Offset = CHECK_DBCS_LEAD_BYTE(pHashN->pDBCSOffsets, *pMBStr)) \ { \ /* \ * DBCS Lead Byte. Make sure there is a trail byte with the \ * lead byte. \ */ \ if ((pMBStr + 1 == pEndMBStr) || (*(pMBStr + 1) == 0)) \ { \ /* \ * There is no trail byte with the lead byte. Return error. \ */ \ SetLastError(ERROR_NO_UNICODE_TRANSLATION); \ return (0); \ } \ \ /* \ * Fill in the wide character translation from the double \ * byte character table. \ */ \ *pWCStr = (pHashN->pDBCSOffsets + Offset)[*(pMBStr + 1)]; \ mbIncr = 2; \ \ /* \ * Make sure an invalid character was not translated to \ * the default char. Return an error if invalid. \ */ \ CHECK_ERROR_WC_MULTI( pHashN, \ *pWCStr, \ *pMBStr, \ *(pMBStr + 1) ); \ } \ else \ { \ /* \ * Not DBCS Lead Byte. Fill in the wide character translation \ * from the single byte character table. \ */ \ *pWCStr = pMBTbl[*pMBStr]; \ mbIncr = 1; \ \ /* \ * Make sure an invalid character was not translated to \ * the default char. Return an error if invalid. \ */ \ CHECK_ERROR_WC_SINGLE( pHashN, \ *pWCStr, \ *pMBStr ); \ } \ } //////////////////////////////////////////////////////////////////////////// // // GET_WC_MULTI_ERR_SPECIAL // // Fills in pWCStr with the wide character(s) for the corresponding multibyte // character from the appropriate translation table. The number of bytes // used from the pMBStr buffer (single byte or double byte) is stored in // the mbIncr parameter. // // Once the character has been translated, it checks to be sure the // character was valid. If not, it fills in 0xffff. // // DEFINED AS A MACRO. // // 08-21-95 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_WC_MULTI_ERR_SPECIAL( pHashN, \ pMBTbl, \ pMBStr, \ pEndMBStr, \ pWCStr, \ pEndWCStr, \ mbIncr ) \ { \ WORD Offset; /* offset to DBCS table for range */ \ \ \ if (Offset = CHECK_DBCS_LEAD_BYTE(pHashN->pDBCSOffsets, *pMBStr)) \ { \ /* \ * DBCS Lead Byte. Make sure there is a trail byte with the \ * lead byte. \ */ \ if ((pMBStr + 1 == pEndMBStr) || (*(pMBStr + 1) == 0)) \ { \ /* \ * There is no trail byte with the lead byte. The lead byte \ * is the LAST character in the string. Translate to 0xffff. \ */ \ *pWCStr = (WCHAR)0xffff; \ mbIncr = 1; \ } \ else \ { \ /* \ * Fill in the wide character translation from the double \ * byte character table. \ */ \ *pWCStr = (pHashN->pDBCSOffsets + Offset)[*(pMBStr + 1)]; \ mbIncr = 2; \ \ /* \ * Make sure an invalid character was not translated to \ * the default char. Translate to 0xffff if invalid. \ */ \ CHECK_ERROR_WC_MULTI_SPECIAL( pHashN, \ pWCStr, \ *pMBStr, \ *(pMBStr + 1) ); \ } \ } \ else \ { \ /* \ * Not DBCS Lead Byte. Fill in the wide character translation \ * from the single byte character table. \ * Make sure an invalid character was not translated to \ * the default char. Return an error if invalid. \ */ \ GET_WC_SINGLE_SPECIAL( pHashN, \ pMBTbl, \ pMBStr, \ pWCStr ); \ mbIncr = 1; \ } \ } //////////////////////////////////////////////////////////////////////////// // // COPY_MB_CHAR // // Copies a multibyte character to the given string buffer. If the // high byte of the multibyte word is zero, then it is a single byte // character and the number of characters written (returned) is 1. // Otherwise, it is a double byte character and the number of characters // written (returned) is 2. // // NumByte will be 0 if the buffer is too small for the translation. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define COPY_MB_CHAR( mbChar, \ pMBStr, \ NumByte, \ fOnlyOne ) \ { \ if (HIBYTE(mbChar)) \ { \ /* \ * Make sure there is enough room in the buffer for both bytes. \ */ \ if (fOnlyOne) \ { \ NumByte = 0; \ } \ else \ { \ /* \ * High Byte is NOT zero, so it's a DOUBLE byte char. \ * Return 2 characters written. \ */ \ *pMBStr = HIBYTE(mbChar); \ *(pMBStr + 1) = LOBYTE(mbChar); \ NumByte = 2; \ } \ } \ else \ { \ /* \ * High Byte IS zero, so it's a SINGLE byte char. \ * Return 1 character written. \ */ \ *pMBStr = LOBYTE(mbChar); \ NumByte = 1; \ } \ } //////////////////////////////////////////////////////////////////////////// // // GET_SB // // Fills in pMBStr with the single byte character for the corresponding // wide character from the appropriate translation table. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_SB( pWC, \ wChar, \ pMBStr ) \ { \ *pMBStr = ((BYTE *)(pWC))[wChar]; \ } //////////////////////////////////////////////////////////////////////////// // // GET_MB // // Fills in pMBStr with the multi byte character for the corresponding // wide character from the appropriate translation table. // // mbCount will be 0 if the buffer is too small for the translation. // // Broken Down Version: // -------------------- // mbChar = ((WORD *)(pHashN->pWC))[wChar]; // COPY_MB_CHAR(mbChar, pMBStr, mbCount); // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_MB( pWC, \ wChar, \ pMBStr, \ mbCount, \ fOnlyOne ) \ { \ COPY_MB_CHAR( ((WORD *)(pWC))[wChar], \ pMBStr, \ mbCount, \ fOnlyOne ); \ } //////////////////////////////////////////////////////////////////////////// // // ELIMINATE_BEST_FIT_SB // // Checks to see if a single byte Best Fit character was used. If so, // it replaces it with a single byte default character. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define ELIMINATE_BEST_FIT_SB( pHashN, \ wChar, \ pMBStr ) \ { \ if ((pHashN->pMBTbl)[*pMBStr] != wChar) \ { \ *pMBStr = LOBYTE(pHashN->pCPInfo->wDefaultChar); \ } \ } //////////////////////////////////////////////////////////////////////////// // // ELIMINATE_BEST_FIT_MB // // Checks to see if a multi byte Best Fit character was used. If so, // it replaces it with a multi byte default character. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define ELIMINATE_BEST_FIT_MB( pHashN, \ wChar, \ pMBStr, \ mbCount, \ fOnlyOne ) \ { \ WORD Offset; \ WORD wDefault; \ \ if (((mbCount == 1) && ((pHashN->pMBTbl)[*pMBStr] != wChar)) || \ ((mbCount == 2) && \ (Offset = CHECK_DBCS_LEAD_BYTE(pHashN->pDBCSOffsets, *pMBStr)) && \ (((pHashN->pDBCSOffsets + Offset)[*(pMBStr + 1)]) != wChar))) \ { \ wDefault = pHashN->pCPInfo->wDefaultChar; \ if (HIBYTE(wDefault)) \ { \ if (fOnlyOne) \ { \ mbCount = 0; \ } \ else \ { \ *pMBStr = HIBYTE(wDefault); \ *(pMBStr + 1) = LOBYTE(wDefault); \ mbCount = 2; \ } \ } \ else \ { \ *pMBStr = LOBYTE(wDefault); \ mbCount = 1; \ } \ } \ } //////////////////////////////////////////////////////////////////////////// // // GET_DEFAULT_WORD // // Takes a pointer to a character string (either one or two characters), // and converts it to a WORD value. If the character is not DBCS, then it // zero extends the high byte. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_DEFAULT_WORD(pOff, pDefault) \ (CHECK_DBCS_LEAD_BYTE(pOff, *pDefault) \ ? MAKEWORD(*(pDefault + 1), *pDefault) \ : MAKEWORD(*pDefault, 0)) //////////////////////////////////////////////////////////////////////////// // // DEFAULT_CHAR_CHECK_SB // // Checks to see if the default character is used. If it is, it sets // pUsedDef to TRUE (if non-null). If the user specified a default, then // the user's default character is used. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define DEFAULT_CHAR_CHECK_SB( pHashN, \ wch, \ pMBStr, \ wDefChar, \ pUsedDef ) \ { \ WORD wSysDefChar = pHashN->pCPInfo->wDefaultChar; \ \ \ /* \ * Check for default character being used. \ */ \ if ((*pMBStr == (BYTE)wSysDefChar) && \ (wch != pHashN->pCPInfo->wTransDefaultChar)) \ { \ /* \ * Default was used. Set the pUsedDef parameter to TRUE. \ */ \ *pUsedDef = TRUE; \ \ /* \ * If the user specified a different default character than \ * the system default, use that character instead. \ */ \ if (wSysDefChar != wDefChar) \ { \ *pMBStr = LOBYTE(wDefChar); \ } \ } \ } //////////////////////////////////////////////////////////////////////////// // // DEFAULT_CHAR_CHECK_MB // // Checks to see if the default character is used. If it is, it sets // pUsedDef to TRUE (if non-null). If the user specified a default, then // the user's default character is used. The number of bytes written to // the buffer is returned. // // NumByte will be -1 if the buffer is too small for the translation. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define DEFAULT_CHAR_CHECK_MB( pHashN, \ wch, \ pMBStr, \ wDefChar, \ pUsedDef, \ NumByte, \ fOnlyOne ) \ { \ WORD wSysDefChar = pHashN->pCPInfo->wDefaultChar; \ \ \ /* \ * Set NumByte to zero for return (zero bytes written). \ */ \ NumByte = 0; \ \ /* \ * Check for default character being used. \ */ \ if ((*pMBStr == (BYTE)wSysDefChar) && \ (wch != pHashN->pCPInfo->wTransDefaultChar)) \ { \ /* \ * Default was used. Set the pUsedDef parameter to TRUE. \ */ \ *pUsedDef = TRUE; \ \ /* \ * If the user specified a different default character than \ * the system default, use that character instead. \ */ \ if (wSysDefChar != wDefChar) \ { \ COPY_MB_CHAR( wDefChar, \ pMBStr, \ NumByte, \ fOnlyOne ); \ if (NumByte == 0) \ { \ NumByte = -1; \ } \ } \ } \ } //////////////////////////////////////////////////////////////////////////// // // GET_WC_TRANSLATION_SB // // Gets the 1:1 translation of a given wide character. It fills in the // string pointer with the single byte character. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_WC_TRANSLATION_SB( pHashN, \ wch, \ pMBStr, \ wDefault, \ pUsedDef, \ dwFlags ) \ { \ GET_SB( pHashN->pWC, \ wch, \ pMBStr ); \ if (dwFlags & WC_NO_BEST_FIT_CHARS) \ { \ ELIMINATE_BEST_FIT_SB( pHashN, \ wch, \ pMBStr ); \ } \ DEFAULT_CHAR_CHECK_SB( pHashN, \ wch, \ pMBStr, \ wDefault, \ pUsedDef ); \ } //////////////////////////////////////////////////////////////////////////// // // GET_WC_TRANSLATION_MB // // Gets the 1:1 translation of a given wide character. It fills in the // appropriate number of characters for the multibyte character and then // returns the number of characters written to the multibyte string. // // mbCnt will be 0 if the buffer is too small for the translation. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_WC_TRANSLATION_MB( pHashN, \ wch, \ pMBStr, \ wDefault, \ pUsedDef, \ mbCnt, \ fOnlyOne, \ dwFlags ) \ { \ int mbCnt2; /* number of characters written */ \ \ \ GET_MB( pHashN->pWC, \ wch, \ pMBStr, \ mbCnt, \ fOnlyOne ); \ if (dwFlags & WC_NO_BEST_FIT_CHARS) \ { \ ELIMINATE_BEST_FIT_MB( pHashN, \ wch, \ pMBStr, \ mbCnt, \ fOnlyOne ); \ } \ if (mbCnt) \ { \ DEFAULT_CHAR_CHECK_MB( pHashN, \ wch, \ pMBStr, \ wDefault, \ pUsedDef, \ mbCnt2, \ fOnlyOne ); \ if (mbCnt2 == -1) \ { \ mbCnt = 0; \ } \ else if (mbCnt2) \ { \ mbCnt = mbCnt2; \ } \ } \ } //////////////////////////////////////////////////////////////////////////// // // GET_CP_HASH_NODE // // Sets the code page value (if a special value is passed in) and the // hash node pointer. If the code page value is invalid, the pointer // to the hash node will be set to NULL. // // DEFINED AS A MACRO. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// #define GET_CP_HASH_NODE( CodePage, \ pHashN ) \ { \ PLOC_HASH pHashLoc; \ \ \ /* \ * Check for the ACP, OEMCP, or MACCP. Fill in the appropriate \ * value for the code page if one of these values is given. \ * Otherwise, just get the hash node for the given code page. \ */ \ if (CodePage == gAnsiCodePage) \ { \ pHashN = gpACPHashN; \ } \ else if (CodePage == gOemCodePage) \ { \ pHashN = gpOEMCPHashN; \ } \ else if (CodePage == CP_ACP) \ { \ CodePage = gAnsiCodePage; \ pHashN = gpACPHashN; \ } \ else if (CodePage == CP_OEMCP) \ { \ CodePage = gOemCodePage; \ pHashN = gpOEMCPHashN; \ } \ else if (CodePage == CP_SYMBOL) \ { \ pHashN = NULL; \ } \ else if (CodePage == CP_MACCP) \ { \ CodePage = GetMacCodePage(); \ pHashN = gpMACCPHashN; \ } \ else if (CodePage == CP_THREAD_ACP) \ { \ VALIDATE_LOCALE(NtCurrentTeb()->CurrentLocale, pHashLoc, FALSE); \ if (pHashLoc != NULL) \ { \ CodePage = pHashLoc->pLocaleFixed->DefaultACP; \ } \ if (CodePage == CP_ACP) \ { \ CodePage = gAnsiCodePage; \ pHashN = gpACPHashN; \ } \ else if (CodePage == CP_OEMCP) \ { \ CodePage = gOemCodePage; \ pHashN = gpOEMCPHashN; \ } \ else if (CodePage == CP_MACCP) \ { \ CodePage = GetMacCodePage(); \ pHashN = gpMACCPHashN; \ } \ else \ { \ pHashN = GetCPHashNode(CodePage); \ } \ } \ else \ { \ pHashN = GetCPHashNode(CodePage); \ } \ } //-------------------------------------------------------------------------// // API ROUTINES // //-------------------------------------------------------------------------// //////////////////////////////////////////////////////////////////////////// // // IsValidCodePage // // Checks that the given code page is a valid one. It does so by querying // the registry. If the code page is found, then TRUE is returned. // Otherwise, FALSE is returned. // // 05-31-1991 JulieB Created. // 05-31-2002 ShawnSte Make it not force the loading of the code page //////////////////////////////////////////////////////////////////////////// BOOL WINAPI IsValidCodePage( UINT CodePage) { // // Do not allow special code page values to be valid here. // (CP_ACP, CP_OEMCP, CP_MACCP, CP_THREAD_ACP, CP_SYMBOL are invalid) // // // Do the quick check for the code page value equal to either // the Ansi code page value or the OEM code page value. // if ((CodePage == gAnsiCodePage) || (CodePage == gOemCodePage) || (CodePage == CP_UTF7) || (CodePage == CP_UTF8)) { // // Return success. // return (TRUE); } // // Check for other code page values. // // If a node already exists, then we're OK if (IsCPHashNodeLoaded(CodePage) == TRUE) { // // Return success. // return (TRUE); } // // Hash node doesn't exist. Have to look in the registry. // True if this works, false if it doesn't // return GetCPFileNameFromRegistry(CodePage, NULL, 0); } //////////////////////////////////////////////////////////////////////////// // // GetACP // // Returns the ANSI code page for the system. If the registry value is // not readable, then the chosen default ACP is used (NLS_DEFAULT_ACP). // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// UINT WINAPI GetACP() { // // Return the ACP stored in the cache. // return (gAnsiCodePage); } //////////////////////////////////////////////////////////////////////////// // // SetCPGlobal // // Sets the code page global, used by Setup to force the code page into // the correct value during GUI mode. // // 02-15-99 JimSchm Created. //////////////////////////////////////////////////////////////////////////// UINT WINAPI SetCPGlobal ( IN UINT NewAcp ) { UINT oldVal; oldVal = gAnsiCodePage; // // Sets the ACP global. This is a private exported routine, not an API. // gAnsiCodePage = NewAcp; return oldVal; } //////////////////////////////////////////////////////////////////////////// // // GetOEMCP // // Returns the OEM code page for the system. If the registry value is // not readable, then the chosen default ACP is used (NLS_DEFAULT_OEMCP). // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// UINT WINAPI GetOEMCP() { // // Return the OEMCP stored in the cache. // return (gOemCodePage); } //////////////////////////////////////////////////////////////////////////// // // GetCPInfo // // Returns information about a given code page. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// BOOL WINAPI GetCPInfo( UINT CodePage, LPCPINFO lpCPInfo) { PCP_HASH pHashN; // ptr to CP hash node PCP_TABLE pInfo; // ptr to CP information in file WORD wDefChar; // default character BYTE *pLeadBytes; // ptr to lead byte ranges UINT Ctr; // loop counter // // See if it's a special code page value for UTF translations. // if (CodePage >= NLS_CP_ALGORITHM_RANGE) { return (UTFCPInfo(CodePage, lpCPInfo, FALSE)); } // // Get the code page value and the appropriate hash node. // GET_CP_HASH_NODE(CodePage, pHashN); // // Invalid Parameter Check: // - validate code page - get hash node containing translation tables // - lpCPInfo is NULL // if ( (pHashN == NULL) || ((pHashN->pCPInfo == NULL) && (pHashN->pfnCPProc == NULL)) || (lpCPInfo == NULL) ) { SetLastError(ERROR_INVALID_PARAMETER); return (FALSE); } // // See if the given code page is in the DLL range. // if (pHashN->pfnCPProc) { // // Call the DLL to get the code page information. // return ( (*(pHashN->pfnCPProc))( CodePage, NLS_CP_CPINFO, NULL, 0, NULL, 0, lpCPInfo ) ); } // // Fill in the CPINFO structure with the appropriate information. // pInfo = pHashN->pCPInfo; // // Get the max char size. // lpCPInfo->MaxCharSize = (UINT)((WORD)pInfo->MaxCharSize); // // Get the default character. // wDefChar = pInfo->wDefaultChar; if (HIBYTE(wDefChar)) { (lpCPInfo->DefaultChar)[0] = HIBYTE(wDefChar); (lpCPInfo->DefaultChar)[1] = LOBYTE(wDefChar); } else { (lpCPInfo->DefaultChar)[0] = LOBYTE(wDefChar); (lpCPInfo->DefaultChar)[1] = (BYTE)0; } // // Get the leadbytes. // pLeadBytes = pInfo->LeadByte; for (Ctr = 0; Ctr < MAX_LEADBYTES; Ctr++) { (lpCPInfo->LeadByte)[Ctr] = pLeadBytes[Ctr]; } // // Return success. // return (TRUE); } //////////////////////////////////////////////////////////////////////////// // // GetCPInfoExW // // Returns information about a given code page. // // 11-15-96 JulieB Created. //////////////////////////////////////////////////////////////////////////// BOOL WINAPI GetCPInfoExW( UINT CodePage, DWORD dwFlags, LPCPINFOEXW lpCPInfoEx) { PCP_HASH pHashN; // ptr to CP hash node PCP_TABLE pInfo; // ptr to CP information in file WORD wDefChar; // default character BYTE *pLeadBytes; // ptr to lead byte ranges UINT Ctr; // loop counter // // See if it's a special code page value for UTF translations. // if (CodePage >= NLS_CP_ALGORITHM_RANGE) { if (UTFCPInfo(CodePage, (LPCPINFO)lpCPInfoEx, TRUE)) { if (GetStringTableEntry( CodePage, 0, lpCPInfoEx->CodePageName, MAX_PATH, RC_CODE_PAGE_NAME ) != 0) { return (TRUE); } } return (FALSE); } // // Get the code page value and the appropriate hash node. // GET_CP_HASH_NODE(CodePage, pHashN); // // Invalid Parameter Check: // - validate code page - get hash node containing translation tables // - lpCPInfoEx is NULL // if ( (pHashN == NULL) || ((pHashN->pCPInfo == NULL) && (pHashN->pfnCPProc == NULL)) || (lpCPInfoEx == NULL) ) { SetLastError(ERROR_INVALID_PARAMETER); return (FALSE); } // // Invalid Flags Check: // - flags not 0 // if (dwFlags != 0) { SetLastError(ERROR_INVALID_FLAGS); return (FALSE); } // // See if the given code page is in the DLL range. // if (pHashN->pfnCPProc) { // // Call the DLL to get the code page information. // if (((*(pHashN->pfnCPProc))( CodePage, NLS_CP_CPINFOEX, NULL, 0, NULL, 0, (LPCPINFO)lpCPInfoEx )) == TRUE) { return (TRUE); } else { // // See if the CPINFO will succeed. If so, then add the // default CPINFOEX info to the structure. // if (((*(pHashN->pfnCPProc))( CodePage, NLS_CP_CPINFO, NULL, 0, NULL, 0, (LPCPINFO)lpCPInfoEx )) == TRUE) { // // Fill in the Ex version info. // lpCPInfoEx->UnicodeDefaultChar = L'?'; lpCPInfoEx->CodePage = CodePage; GetStringTableEntry( CodePage, 0, lpCPInfoEx->CodePageName, MAX_PATH, RC_CODE_PAGE_NAME ); SetLastError(NO_ERROR); return (TRUE); } return (FALSE); } } // // Fill in the CPINFO structure with the appropriate information. // pInfo = pHashN->pCPInfo; // // Get the max char size. // lpCPInfoEx->MaxCharSize = (UINT)((WORD)pInfo->MaxCharSize); // // Get the default character. // wDefChar = pInfo->wDefaultChar; if (HIBYTE(wDefChar)) { (lpCPInfoEx->DefaultChar)[0] = HIBYTE(wDefChar); (lpCPInfoEx->DefaultChar)[1] = LOBYTE(wDefChar); } else { (lpCPInfoEx->DefaultChar)[0] = LOBYTE(wDefChar); (lpCPInfoEx->DefaultChar)[1] = (BYTE)0; } // // Get the leadbytes. // pLeadBytes = pInfo->LeadByte; for (Ctr = 0; Ctr < MAX_LEADBYTES; Ctr++) { (lpCPInfoEx->LeadByte)[Ctr] = pLeadBytes[Ctr]; } // // Get the Unicode default character. // lpCPInfoEx->UnicodeDefaultChar = pInfo->wUniDefaultChar; // // Get the code page id. // lpCPInfoEx->CodePage = CodePage; // // Get the code page name. // if (GetStringTableEntry( CodePage, 0, lpCPInfoEx->CodePageName, MAX_PATH, RC_CODE_PAGE_NAME ) == 0) { return (FALSE); } // // Return success. // return (TRUE); } //////////////////////////////////////////////////////////////////////////// // // IsDBCSLeadByte // // Checks to see if a given character is a DBCS lead byte in the ACP. // Returns TRUE if it is, FALSE if it is not. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// BOOL WINAPI IsDBCSLeadByte( BYTE TestChar) { // // Get the hash node for the ACP. // if (gpACPHashN == NULL) { SetLastError(ERROR_FILE_NOT_FOUND); return (FALSE); } // // See if the given character is a DBCS lead byte. // if (CHECK_DBCS_LEAD_BYTE(gpACPHashN->pDBCSOffsets, TestChar)) { // // Return success - IS a DBCS lead byte. // return (TRUE); } // // Return failure - is NOT a DBCS lead byte. // return (FALSE); } //////////////////////////////////////////////////////////////////////////// // // IsDBCSLeadByteEx // // Checks to see if a given character is a DBCS lead byte in the given // code page. Returns TRUE if it is, FALSE if it is not. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// BOOL WINAPI IsDBCSLeadByteEx( UINT CodePage, BYTE TestChar) { PCP_HASH pHashN; // ptr to CP hash node // // See if it's a special code page value for UTF translations. // if (CodePage >= NLS_CP_ALGORITHM_RANGE) { // // Return that it's not a DBCS leadbyte. // return (FALSE); } // // Get the code page value and the appropriate hash node. // GET_CP_HASH_NODE(CodePage, pHashN); // // Invalid Parameter Check: // - validate code page // if (pHashN == NULL) { SetLastError(ERROR_INVALID_PARAMETER); return (FALSE); } // // See if the given character is a DBCS lead byte. // if (CHECK_DBCS_LEAD_BYTE(pHashN->pDBCSOffsets, TestChar)) { // // Return success - IS a DBCS lead byte. // return (TRUE); } // // Return failure - is NOT a DBCS lead byte. // return (FALSE); } //////////////////////////////////////////////////////////////////////////// // // MultiByteToWideChar // // Maps a multibyte character string to its wide character string // counterpart. // // 05-31-91 JulieB Created. // 09-01-93 JulieB Add support for MB_ERR_INVALID_CHARS flag. //////////////////////////////////////////////////////////////////////////// int WINAPI MultiByteToWideChar( UINT CodePage, DWORD dwFlags, LPCSTR lpMultiByteStr, int cbMultiByte, LPWSTR lpWideCharStr, int cchWideChar) { PCP_HASH pHashN; // ptr to CP hash node register LPBYTE pMBStr; // ptr to search through MB string register LPWSTR pWCStr; // ptr to search through WC string LPBYTE pEndMBStr; // ptr to end of MB search string LPWSTR pEndWCStr; // ptr to end of WC string buffer int wcIncr; // amount to increment pWCStr int mbIncr; // amount to increment pMBStr int wcCount = 0; // count of wide chars written int CompSet; // if MB_COMPOSITE flag is set PMB_TABLE pMBTbl; // ptr to correct MB table (MB or GLYPH) int ctr; // loop counter // // See if it's a special code page value for UTF translations. // if (CodePage >= NLS_CP_ALGORITHM_RANGE) { return (UTFToUnicode( CodePage, dwFlags, lpMultiByteStr, cbMultiByte, lpWideCharStr, cchWideChar )); } // // Get the code page value and the appropriate hash node. // GET_CP_HASH_NODE(CodePage, pHashN); // // Invalid Parameter Check: // - length of MB string is 0 // - wide char buffer size is negative // - MB string is NULL // - length of WC string is NOT zero AND // (WC string is NULL OR src and dest pointers equal) // if ( (cbMultiByte == 0) || (cchWideChar < 0) || (lpMultiByteStr == NULL) || ((cchWideChar != 0) && ((lpWideCharStr == NULL) || (lpMultiByteStr == (LPSTR)lpWideCharStr))) ) { SetLastError(ERROR_INVALID_PARAMETER); return (0); } // // If cbMultiByte is -1, then the string is null terminated and we // need to get the length of the string. Add one to the length to // include the null termination. (This will always be at least 1.) // if (cbMultiByte <= -1) { cbMultiByte = strlen(lpMultiByteStr) + 1; } // // Check for valid code page. // if (pHashN == NULL) { // // Special case the CP_SYMBOL code page. // if ((CodePage == CP_SYMBOL) && (dwFlags == 0)) { // // If the caller just wants the size of the buffer needed // to do this translation, return the size of the MB string. // if (cchWideChar == 0) { return (cbMultiByte); } // // Make sure the buffer is large enough. // if (cchWideChar < cbMultiByte) { SetLastError(ERROR_INSUFFICIENT_BUFFER); return (0); } // // Translate SB char xx to Unicode f0xx. // 0x00->0x1f map to 0x0000->0x001f // 0x20->0xff map to 0xf020->0xf0ff // for (ctr = 0; ctr < cbMultiByte; ctr++) { lpWideCharStr[ctr] = ((BYTE)(lpMultiByteStr[ctr]) < 0x20) ? (WCHAR)lpMultiByteStr[ctr] : MAKEWORD(lpMultiByteStr[ctr], 0xf0); } return (cbMultiByte); } else { SetLastError(((CodePage == CP_SYMBOL) && (dwFlags != 0)) ? ERROR_INVALID_FLAGS : ERROR_INVALID_PARAMETER); return (0); } } // // See if the given code page is in the DLL range. // if (pHashN->pfnCPProc) { // // Invalid Flags Check: // - flags not 0 // if (dwFlags != 0) { SetLastError(ERROR_INVALID_FLAGS); return (0); } // // Call the DLL to do the translation. // return ( (*(pHashN->pfnCPProc))( CodePage, NLS_CP_MBTOWC, (LPSTR)lpMultiByteStr, cbMultiByte, (LPWSTR)lpWideCharStr, cchWideChar, NULL ) ); } // // Invalid Flags Check: // - flags other than valid ones // - composite and precomposed both set // if ( (dwFlags & MB_INVALID_FLAG) || ((dwFlags & MB_PRECOMPOSED) && (dwFlags & MB_COMPOSITE)) ) { SetLastError(ERROR_INVALID_FLAGS); return (0); } // // Initialize multibyte character loop pointers. // pMBStr = (LPBYTE)lpMultiByteStr; pEndMBStr = pMBStr + cbMultiByte; CompSet = dwFlags & MB_COMPOSITE; // // Get the correct MB table (MB or GLYPH). // if ((dwFlags & MB_USEGLYPHCHARS) && (pHashN->pGlyphTbl != NULL)) { pMBTbl = pHashN->pGlyphTbl; } else { pMBTbl = pHashN->pMBTbl; } // // If cchWideChar is 0, then we can't use lpWideCharStr. In this // case, we simply want to count the number of characters that would // be written to the buffer. // if (cchWideChar == 0) { WCHAR pTempStr[MAX_COMPOSITE]; // tmp buffer - max for composite // // For each multibyte char, translate it to its corresponding // wide char and increment the wide character count. // pEndWCStr = pTempStr + MAX_COMPOSITE; if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // if (CompSet) { // // Composite flag is set. // if (dwFlags & MB_ERR_INVALID_CHARS) { // // Error check flag is set. // while (pMBStr < pEndMBStr) { if (!(wcIncr = GetWCCompSBErr( pHashN, pMBTbl, pMBStr, pTempStr, pEndWCStr ))) { return (0); } pMBStr++; wcCount += wcIncr; } } else { // // Error check flag is NOT set. // while (pMBStr < pEndMBStr) { wcCount += GetWCCompSB( pMBTbl, pMBStr, pTempStr, pEndWCStr ); pMBStr++; } } } else { // // Composite flag is NOT set. // if (dwFlags & MB_ERR_INVALID_CHARS) { // // Error check flag is set. // wcCount = (int)(pEndMBStr - pMBStr); while (pMBStr < pEndMBStr) { GET_WC_SINGLE( pMBTbl, pMBStr, pTempStr ); CHECK_ERROR_WC_SINGLE( pHashN, *pTempStr, *pMBStr ); pMBStr++; } } else { // // Error check flag is NOT set. // // Just return the size of the MB string, since // it's a 1:1 translation. // wcCount = (int)(pEndMBStr - pMBStr); } } } else { // // Multi Byte Character Code Page. // if (CompSet) { // // Composite flag is set. // if (dwFlags & MB_ERR_INVALID_CHARS) { // // Error check flag is set. // while (pMBStr < pEndMBStr) { if (!(wcIncr = GetWCCompMBErr( pHashN, pMBTbl, pMBStr, pEndMBStr, pTempStr, pEndWCStr, &mbIncr ))) { return (0); } pMBStr += mbIncr; wcCount += wcIncr; } } else { // // Error check flag is NOT set. // while (pMBStr < pEndMBStr) { wcCount += GetWCCompMB( pHashN, pMBTbl, pMBStr, pEndMBStr, pTempStr, pEndWCStr, &mbIncr ); pMBStr += mbIncr; } } } else { // // Composite flag is NOT set. // if (dwFlags & MB_ERR_INVALID_CHARS) { // // Error check flag is set. // while (pMBStr < pEndMBStr) { GET_WC_MULTI_ERR( pHashN, pMBTbl, pMBStr, pEndMBStr, pTempStr, pEndWCStr, mbIncr ); pMBStr += mbIncr; wcCount++; } } else { // // Error check flag is NOT set. // while (pMBStr < pEndMBStr) { GET_WC_MULTI( pHashN, pMBTbl, pMBStr, pEndMBStr, pTempStr, pEndWCStr, mbIncr ); pMBStr += mbIncr; wcCount++; } } } } } else { // // Initialize wide character loop pointers. // pWCStr = lpWideCharStr; pEndWCStr = pWCStr + cchWideChar; // // For each multibyte char, translate it to its corresponding // wide char, store it in lpWideCharStr, and increment the wide // character count. // if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // if (CompSet) { // // Composite flag is set. // if (dwFlags & MB_ERR_INVALID_CHARS) { // // Error check flag is set. // while ((pMBStr < pEndMBStr) && (pWCStr < pEndWCStr)) { if (!(wcIncr = GetWCCompSBErr( pHashN, pMBTbl, pMBStr, pWCStr, pEndWCStr ))) { return (0); } pMBStr++; pWCStr += wcIncr; } wcCount = (int)(pWCStr - lpWideCharStr); } else { // // Error check flag is NOT set. // while ((pMBStr < pEndMBStr) && (pWCStr < pEndWCStr)) { pWCStr += GetWCCompSB( pMBTbl, pMBStr, pWCStr, pEndWCStr ); pMBStr++; } wcCount = (int)(pWCStr - lpWideCharStr); } } else { // // Composite flag is NOT set. // if (dwFlags & MB_ERR_INVALID_CHARS) { // // Error check flag is set. // wcCount = (int)(pEndMBStr - pMBStr); if ((pEndWCStr - pWCStr) < wcCount) { wcCount = (int)(pEndWCStr - pWCStr); } for (ctr = wcCount; ctr > 0; ctr--) { GET_WC_SINGLE( pMBTbl, pMBStr, pWCStr ); CHECK_ERROR_WC_SINGLE( pHashN, *pWCStr, *pMBStr ); pMBStr++; pWCStr++; } } else { // // Error check flag is NOT set. // wcCount = (int)(pEndMBStr - pMBStr); if ((pEndWCStr - pWCStr) < wcCount) { wcCount = (int)(pEndWCStr - pWCStr); } for (ctr = wcCount; ctr > 0; ctr--) { GET_WC_SINGLE( pMBTbl, pMBStr, pWCStr ); pMBStr++; pWCStr++; } } } } else { // // Multi Byte Character Code Page. // if (CompSet) { // // Composite flag is set. // if (dwFlags & MB_ERR_INVALID_CHARS) { // // Error check flag is set. // while ((pMBStr < pEndMBStr) && (pWCStr < pEndWCStr)) { if (!(wcIncr = GetWCCompMBErr( pHashN, pMBTbl, pMBStr, pEndMBStr, pWCStr, pEndWCStr, &mbIncr ))) { return (0); } pMBStr += mbIncr; pWCStr += wcIncr; } wcCount = (int)(pWCStr - lpWideCharStr); } else { // // Error check flag is NOT set. // while ((pMBStr < pEndMBStr) && (pWCStr < pEndWCStr)) { pWCStr += GetWCCompMB( pHashN, pMBTbl, pMBStr, pEndMBStr, pWCStr, pEndWCStr, &mbIncr ); pMBStr += mbIncr; } wcCount = (int)(pWCStr - lpWideCharStr); } } else { // // Composite flag is NOT set. // if (dwFlags & MB_ERR_INVALID_CHARS) { // // Error check flag is set. // while ((pMBStr < pEndMBStr) && (pWCStr < pEndWCStr)) { GET_WC_MULTI_ERR( pHashN, pMBTbl, pMBStr, pEndMBStr, pWCStr, pEndWCStr, mbIncr ); pMBStr += mbIncr; pWCStr++; } wcCount = (int)(pWCStr - lpWideCharStr); } else { // // Error check flag is NOT set. // while ((pMBStr < pEndMBStr) && (pWCStr < pEndWCStr)) { GET_WC_MULTI( pHashN, pMBTbl, pMBStr, pEndMBStr, pWCStr, pEndWCStr, mbIncr ); pMBStr += mbIncr; pWCStr++; } wcCount = (int)(pWCStr - lpWideCharStr); } } } // // Make sure wide character buffer was large enough. // if (pMBStr < pEndMBStr) { SetLastError(ERROR_INSUFFICIENT_BUFFER); return (0); } } // // Return the number of characters written (or that would have // been written) to the buffer. // return (wcCount); } //////////////////////////////////////////////////////////////////////////// // // WideCharToMultiByte // // Maps a wide character string to its multibyte character string // counterpart. // // NOTE: Most significant bit of dwFlags parameter is used by this routine // to indicate that the caller only wants the count of the number of // characters written, not the string (ie. do not back up in buffer). // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// int WINAPI WideCharToMultiByte( UINT CodePage, DWORD dwFlags, LPCWSTR lpWideCharStr, int cchWideChar, LPSTR lpMultiByteStr, int cbMultiByte, LPCSTR lpDefaultChar, LPBOOL lpUsedDefaultChar) { PCP_HASH pHashN; // ptr to CP hash node LPWSTR pWCStr; // ptr to search through WC string LPWSTR pEndWCStr; // ptr to end of WC string buffer WORD wDefault = 0; // default character as a word int IfNoDefault; // if default check is to be made int IfCompositeChk; // if check for composite BOOL TmpUsed; // temp storage for default used int ctr; // loop counter // // See if it's a special code page value for UTF translations. // if (CodePage >= NLS_CP_ALGORITHM_RANGE) { return (UnicodeToUTF( CodePage, dwFlags, lpWideCharStr, cchWideChar, lpMultiByteStr, cbMultiByte, lpDefaultChar, lpUsedDefaultChar )); } // // Get the code page value and the appropriate hash node. // GET_CP_HASH_NODE(CodePage, pHashN); // // Invalid Parameter Check: // - length of WC string is 0 // - multibyte buffer size is negative // - WC string is NULL // - length of WC string is NOT zero AND // (MB string is NULL OR src and dest pointers equal) // if ( (cchWideChar == 0) || (cbMultiByte < 0) || (lpWideCharStr == NULL) || ((cbMultiByte != 0) && ((lpMultiByteStr == NULL) || (lpWideCharStr == (LPWSTR)lpMultiByteStr))) ) { SetLastError(ERROR_INVALID_PARAMETER); return (0); } // // If cchWideChar is -1, then the string is null terminated and we // need to get the length of the string. Add one to the length to // include the null termination. (This will always be at least 1.) // if (cchWideChar <= -1) { cchWideChar = NlsStrLenW(lpWideCharStr) + 1; } // // Check for valid code page. // if (pHashN == NULL) { // // Special case the CP_SYMBOL code page. // if ((CodePage == CP_SYMBOL) && (dwFlags == 0) && (lpDefaultChar == NULL) && (lpUsedDefaultChar == NULL)) { // // If the caller just wants the size of the buffer needed // to do this translation, return the size of the MB string. // if (cbMultiByte == 0) { return (cchWideChar); } // // Make sure the buffer is large enough. // if (cbMultiByte < cchWideChar) { SetLastError(ERROR_INSUFFICIENT_BUFFER); return (0); } // // Translate Unicode char f0xx to SB xx. // 0x0000->0x001f map to 0x00->0x1f // 0xf020->0xf0ff map to 0x20->0xff // for (ctr = 0; ctr < cchWideChar; ctr++) { if ((lpWideCharStr[ctr] >= 0x0020) && ((lpWideCharStr[ctr] < 0xf020) || (lpWideCharStr[ctr] > 0xf0ff))) { SetLastError(ERROR_NO_UNICODE_TRANSLATION); \ return (0); } lpMultiByteStr[ctr] = (BYTE)lpWideCharStr[ctr]; } return (cchWideChar); } else { SetLastError(((CodePage == CP_SYMBOL) && (dwFlags != 0)) ? ERROR_INVALID_FLAGS : ERROR_INVALID_PARAMETER); return (0); } } // // See if the given code page is in the DLL range. // if (pHashN->pfnCPProc) { // // Invalid Parameter Check: // - lpDefaultChar not NULL // - lpUsedDefaultChar not NULL // if ((lpDefaultChar != NULL) || (lpUsedDefaultChar != NULL)) { SetLastError(ERROR_INVALID_PARAMETER); return (0); } // // Invalid Flags Check: // - flags not 0 // if (dwFlags != 0) { SetLastError(ERROR_INVALID_FLAGS); return (0); } // // Call the DLL to do the translation. // return ( (*(pHashN->pfnCPProc))( CodePage, NLS_CP_WCTOMB, (LPSTR)lpMultiByteStr, cbMultiByte, (LPWSTR)lpWideCharStr, cchWideChar, NULL ) ); } // // Invalid Flags Check: // - compositechk flag is not set AND any of comp flags are set // - flags other than valid ones // if ( ((!(IfCompositeChk = (dwFlags & WC_COMPOSITECHECK))) && (dwFlags & WC_COMPCHK_FLAGS)) || (dwFlags & WC_INVALID_FLAG) ) { SetLastError(ERROR_INVALID_FLAGS); return (0); } // // Initialize wide character loop pointers. // pWCStr = (LPWSTR)lpWideCharStr; pEndWCStr = pWCStr + cchWideChar; // // Set the IfNoDefault parameter to TRUE if both lpDefaultChar and // lpUsedDefaultChar are NULL. // IfNoDefault = ((lpDefaultChar == NULL) && (lpUsedDefaultChar == NULL)); // // If the composite check flag is NOT set AND both of the default // parameters (lpDefaultChar and lpUsedDefaultChar) are null, then // do the quick translation. // if (IfNoDefault && !IfCompositeChk) { // // Translate WC string to MB string, ignoring default chars. // return (GetMBNoDefault( pHashN, pWCStr, pEndWCStr, (LPBYTE)lpMultiByteStr, cbMultiByte, dwFlags )); } // // Set the system default character. // wDefault = pHashN->pCPInfo->wDefaultChar; // // See if the default check is needed. // if (!IfNoDefault) { // // If lpDefaultChar is NULL, then use the system default. // Form a word out of the default character. Single byte // characters are zero extended, DBCS characters are as is. // if (lpDefaultChar != NULL) { wDefault = GET_DEFAULT_WORD( pHashN->pDBCSOffsets, (LPBYTE)lpDefaultChar ); } // // If lpUsedDefaultChar is NULL, then it won't be used later // on if a default character is detected. Otherwise, we need // to initialize it. // if (lpUsedDefaultChar == NULL) { lpUsedDefaultChar = &TmpUsed; } *lpUsedDefaultChar = FALSE; // // Check for "composite check" flag. // if (!IfCompositeChk) { // // Translate WC string to MB string, checking for the use of the // default character. // return (GetMBDefault( pHashN, pWCStr, pEndWCStr, (LPBYTE)lpMultiByteStr, cbMultiByte, wDefault, lpUsedDefaultChar, dwFlags )); } else { // // Translate WC string to MB string, checking for the use of the // default character. // return (GetMBDefaultComp( pHashN, pWCStr, pEndWCStr, (LPBYTE)lpMultiByteStr, cbMultiByte, wDefault, lpUsedDefaultChar, dwFlags )); } } else { // // The only case left here is that the Composite check // flag IS set and the default check flag is NOT set. // // Translate WC string to MB string, checking for the use of the // default character. // return (GetMBDefaultComp( pHashN, pWCStr, pEndWCStr, (LPBYTE)lpMultiByteStr, cbMultiByte, wDefault, &TmpUsed, dwFlags )); } } //-------------------------------------------------------------------------// // INTERNAL ROUTINES // //-------------------------------------------------------------------------// //////////////////////////////////////////////////////////////////////////// // // GetWCCompSB // // Fills in pWCStr with the wide character(s) for the corresponding single // byte character from the appropriate translation table and returns the // number of wide characters written. This routine should only be called // when the precomposed forms need to be translated to composite. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// int GetWCCompSB( PMB_TABLE pMBTbl, LPBYTE pMBStr, LPWSTR pWCStr, LPWSTR pEndWCStr) { // // Get the single byte to wide character translation. // GET_WC_SINGLE(pMBTbl, pMBStr, pWCStr); // // Fill in the composite form of the character (if one exists) // and return the number of wide characters written. // return (InsertCompositeForm(pWCStr, pEndWCStr)); } //////////////////////////////////////////////////////////////////////////// // // GetWCCompMB // // Fills in pWCStr with the wide character(s) for the corresponding multibyte // character from the appropriate translation table and returns the number // of wide characters written. The number of bytes used from the pMBStr // buffer (single byte or double byte) is returned in the mbIncr parameter. // This routine should only be called when the precomposed forms need to be // translated to composite. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// int GetWCCompMB( PCP_HASH pHashN, PMB_TABLE pMBTbl, LPBYTE pMBStr, LPBYTE pEndMBStr, LPWSTR pWCStr, LPWSTR pEndWCStr, int *pmbIncr) { // // Get the multibyte to wide char translation. // GET_WC_MULTI( pHashN, pMBTbl, pMBStr, pEndMBStr, pWCStr, pEndWCStr, *pmbIncr ); // // Fill in the composite form of the character (if one exists) // and return the number of wide characters written. // return (InsertCompositeForm(pWCStr, pEndWCStr)); } //////////////////////////////////////////////////////////////////////////// // // GetWCCompSBErr // // Fills in pWCStr with the wide character(s) for the corresponding single // byte character from the appropriate translation table and returns the // number of wide characters written. This routine should only be called // when the precomposed forms need to be translated to composite. // // Checks to be sure an invalid character is not translated to the default // character. If so, it sets last error and returns 0 characters written. // // 09-01-93 JulieB Created. //////////////////////////////////////////////////////////////////////////// int GetWCCompSBErr( PCP_HASH pHashN, PMB_TABLE pMBTbl, LPBYTE pMBStr, LPWSTR pWCStr, LPWSTR pEndWCStr) { // // Get the single byte to wide character translation. // GET_WC_SINGLE(pMBTbl, pMBStr, pWCStr); // // Make sure an invalid character was not translated to the // default char. If it was, set last error and return 0 // characters written. // CHECK_ERROR_WC_SINGLE(pHashN, *pWCStr, *pMBStr); // // Fill in the composite form of the character (if one exists) // and return the number of wide characters written. // return (InsertCompositeForm(pWCStr, pEndWCStr)); } //////////////////////////////////////////////////////////////////////////// // // GetWCCompMBErr // // Fills in pWCStr with the wide character(s) for the corresponding multibyte // character from the appropriate translation table and returns the number // of wide characters written. The number of bytes used from the pMBStr // buffer (single byte or double byte) is returned in the mbIncr parameter. // This routine should only be called when the precomposed forms need to be // translated to composite. // // Checks to be sure an invalid character is not translated to the default // character. If so, it sets last error and returns 0 characters written. // // 09-01-93 JulieB Created. //////////////////////////////////////////////////////////////////////////// int GetWCCompMBErr( PCP_HASH pHashN, PMB_TABLE pMBTbl, LPBYTE pMBStr, LPBYTE pEndMBStr, LPWSTR pWCStr, LPWSTR pEndWCStr, int *pmbIncr) { // // Get the multibyte to wide char translation. // // Make sure an invalid character was not translated to the // default char. If it was, set last error and return 0 // characters written. // GET_WC_MULTI_ERR( pHashN, pMBTbl, pMBStr, pEndMBStr, pWCStr, pEndWCStr, *pmbIncr ); // // Fill in the composite form of the character (if one exists) // and return the number of wide characters written. // return (InsertCompositeForm(pWCStr, pEndWCStr)); } //////////////////////////////////////////////////////////////////////////// // // GetMBNoDefault // // Translates the wide character string to a multibyte string and returns // the number of bytes written. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// int GetMBNoDefault( PCP_HASH pHashN, LPWSTR pWCStr, LPWSTR pEndWCStr, LPBYTE pMBStr, int cbMultiByte, DWORD dwFlags) { int mbIncr; // amount to increment pMBStr int mbCount = 0; // count of multibyte chars written LPBYTE pEndMBStr; // ptr to end of MB string buffer PWC_TABLE pWC = pHashN->pWC; // ptr to WC table int ctr; // loop counter // // If cbMultiByte is 0, then we can't use pMBStr. In this // case, we simply want to count the number of characters that // would be written to the buffer. // if (cbMultiByte == 0) { BYTE pTempStr[2]; // tmp buffer - 2 bytes for DBCS // // For each wide char, translate it to its corresponding multibyte // char and increment the multibyte character count. // if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // // Just return the count of characters - it will be the // same number of characters as the source string. // mbCount = (int)(pEndWCStr - pWCStr); } else { // // Multi Byte Character Code Page. // if (dwFlags & WC_NO_BEST_FIT_CHARS) { while (pWCStr < pEndWCStr) { GET_MB( pWC, *pWCStr, pTempStr, mbIncr, FALSE ); ELIMINATE_BEST_FIT_MB( pHashN, *pWCStr, pTempStr, mbIncr, FALSE ); pWCStr++; mbCount += mbIncr; } } else { while (pWCStr < pEndWCStr) { GET_MB( pWC, *pWCStr, pTempStr, mbIncr, FALSE ); pWCStr++; mbCount += mbIncr; } } } } else { // // Initialize multibyte loop pointers. // pEndMBStr = pMBStr + cbMultiByte; // // For each wide char, translate it to its corresponding // multibyte char, store it in pMBStr, and increment the // multibyte character count. // if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // mbCount = (int)(pEndWCStr - pWCStr); if ((pEndMBStr - pMBStr) < mbCount) { mbCount = (int)(pEndMBStr - pMBStr); } if (dwFlags & WC_NO_BEST_FIT_CHARS) { for (ctr = mbCount; ctr > 0; ctr--) { GET_SB( pWC, *pWCStr, pMBStr ); ELIMINATE_BEST_FIT_SB( pHashN, *pWCStr, pMBStr ); pWCStr++; pMBStr++; } } else { for (ctr = mbCount; ctr > 0; ctr--) { GET_SB( pWC, *pWCStr, pMBStr ); pWCStr++; pMBStr++; } } } else { // // Multi Byte Character Code Page. // if (dwFlags & WC_NO_BEST_FIT_CHARS) { while ((pWCStr < pEndWCStr) && (pMBStr < pEndMBStr)) { GET_MB( pWC, *pWCStr, pMBStr, mbIncr, ((pMBStr + 1) < pEndMBStr) ? FALSE : TRUE ); ELIMINATE_BEST_FIT_MB( pHashN, *pWCStr, pMBStr, mbIncr, ((pMBStr + 1) < pEndMBStr) ? FALSE : TRUE ); if (mbIncr == 0) { // // Not enough space in buffer. // break; } pWCStr++; mbCount += mbIncr; pMBStr += mbIncr; } } else { while ((pWCStr < pEndWCStr) && (pMBStr < pEndMBStr)) { GET_MB( pWC, *pWCStr, pMBStr, mbIncr, ((pMBStr + 1) < pEndMBStr) ? FALSE : TRUE ); if (mbIncr == 0) { // // Not enough space in buffer. // break; } pWCStr++; mbCount += mbIncr; pMBStr += mbIncr; } } } // // Make sure multibyte character buffer was large enough. // if (pWCStr < pEndWCStr) { SetLastError(ERROR_INSUFFICIENT_BUFFER); return (0); } } // // Return the number of characters written (or that would have // been written) to the buffer. // return (mbCount); } //////////////////////////////////////////////////////////////////////////// // // GetMBDefault // // Translates the wide character string to a multibyte string and returns // the number of bytes written. This also checks for the use of the default // character, so the translation is slower. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// int GetMBDefault( PCP_HASH pHashN, LPWSTR pWCStr, LPWSTR pEndWCStr, LPBYTE pMBStr, int cbMultiByte, WORD wDefault, LPBOOL pUsedDef, DWORD dwFlags) { int mbIncr; // amount to increment pMBStr int mbIncr2; // amount to increment pMBStr int mbCount = 0; // count of multibyte chars written LPBYTE pEndMBStr; // ptr to end of MB string buffer PWC_TABLE pWC = pHashN->pWC; // ptr to WC table int ctr; // loop counter // // If cbMultiByte is 0, then we can't use pMBStr. In this // case, we simply want to count the number of characters that // would be written to the buffer. // if (cbMultiByte == 0) { BYTE pTempStr[2]; // tmp buffer - 2 bytes for DBCS // // For each wide char, translate it to its corresponding multibyte // char and increment the multibyte character count. // if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // mbCount = (int)(pEndWCStr - pWCStr); if (dwFlags & WC_NO_BEST_FIT_CHARS) { while (pWCStr < pEndWCStr) { GET_SB( pWC, *pWCStr, pTempStr ); ELIMINATE_BEST_FIT_SB( pHashN, *pWCStr, pTempStr ); DEFAULT_CHAR_CHECK_SB( pHashN, *pWCStr, pTempStr, wDefault, pUsedDef ); pWCStr++; } } else { while (pWCStr < pEndWCStr) { GET_SB( pWC, *pWCStr, pTempStr ); DEFAULT_CHAR_CHECK_SB( pHashN, *pWCStr, pTempStr, wDefault, pUsedDef ); pWCStr++; } } } else { // // Multi Byte Character Code Page. // if (dwFlags & WC_NO_BEST_FIT_CHARS) { while (pWCStr < pEndWCStr) { GET_MB( pWC, *pWCStr, pTempStr, mbIncr, FALSE ); ELIMINATE_BEST_FIT_MB( pHashN, *pWCStr, pTempStr, mbIncr, FALSE ); DEFAULT_CHAR_CHECK_MB( pHashN, *pWCStr, pTempStr, wDefault, pUsedDef, mbIncr2, FALSE ); mbCount += (mbIncr2) ? (mbIncr2) : (mbIncr); pWCStr++; } } else { while (pWCStr < pEndWCStr) { GET_MB( pWC, *pWCStr, pTempStr, mbIncr, FALSE ); DEFAULT_CHAR_CHECK_MB( pHashN, *pWCStr, pTempStr, wDefault, pUsedDef, mbIncr2, FALSE ); mbCount += (mbIncr2) ? (mbIncr2) : (mbIncr); pWCStr++; } } } } else { // // Initialize multibyte loop pointers. // pEndMBStr = pMBStr + cbMultiByte; // // For each wide char, translate it to its corresponding // multibyte char, store it in pMBStr, and increment the // multibyte character count. // if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // mbCount = (int)(pEndWCStr - pWCStr); if ((pEndMBStr - pMBStr) < mbCount) { mbCount = (int)(pEndMBStr - pMBStr); } if (dwFlags & WC_NO_BEST_FIT_CHARS) { for (ctr = mbCount; ctr > 0; ctr--) { GET_SB( pWC, *pWCStr, pMBStr ); ELIMINATE_BEST_FIT_SB( pHashN, *pWCStr, pMBStr ); DEFAULT_CHAR_CHECK_SB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef ); pWCStr++; pMBStr++; } } else { for (ctr = mbCount; ctr > 0; ctr--) { GET_SB( pWC, *pWCStr, pMBStr ); DEFAULT_CHAR_CHECK_SB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef ); pWCStr++; pMBStr++; } } } else { // // Multi Byte Character Code Page. // if (dwFlags & WC_NO_BEST_FIT_CHARS) { while ((pWCStr < pEndWCStr) && (pMBStr < pEndMBStr)) { GET_MB( pWC, *pWCStr, pMBStr, mbIncr, ((pMBStr + 1) < pEndMBStr) ? FALSE : TRUE ); ELIMINATE_BEST_FIT_MB( pHashN, *pWCStr, pMBStr, mbIncr, ((pMBStr + 1) < pEndMBStr) ? FALSE : TRUE ); DEFAULT_CHAR_CHECK_MB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef, mbIncr2, ((pMBStr + 1) < pEndMBStr) ? FALSE : TRUE ); if ((mbIncr == 0) || (mbIncr2 == -1)) { // // Not enough room in buffer. // break; } mbCount += (mbIncr2) ? (mbIncr2) : (mbIncr); pWCStr++; pMBStr += mbIncr; } } else { while ((pWCStr < pEndWCStr) && (pMBStr < pEndMBStr)) { GET_MB( pWC, *pWCStr, pMBStr, mbIncr, ((pMBStr + 1) < pEndMBStr) ? FALSE : TRUE ); DEFAULT_CHAR_CHECK_MB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef, mbIncr2, ((pMBStr + 1) < pEndMBStr) ? FALSE : TRUE ); if ((mbIncr == 0) || (mbIncr2 == -1)) { // // Not enough room in buffer. // break; } mbCount += (mbIncr2) ? (mbIncr2) : (mbIncr); pWCStr++; pMBStr += mbIncr; } } } // // Make sure multibyte character buffer was large enough. // if (pWCStr < pEndWCStr) { SetLastError(ERROR_INSUFFICIENT_BUFFER); return (0); } } // // Return the number of characters written (or that would have // been written) to the buffer. // return (mbCount); } //////////////////////////////////////////////////////////////////////////// // // GetMBDefaultComp // // Translates the wide character string to a multibyte string and returns // the number of bytes written. This also checks for the use of the default // character and tries to convert composite forms to precomposed forms, so // the translation is a lot slower. // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// int GetMBDefaultComp( PCP_HASH pHashN, LPWSTR pWCStr, LPWSTR pEndWCStr, LPBYTE pMBStr, int cbMultiByte, WORD wDefault, LPBOOL pUsedDef, DWORD dwFlags) { int mbIncr; // amount to increment pMBStr int mbCount = 0; // count of multibyte chars written LPBYTE pEndMBStr; // ptr to end of MB string buffer BOOL fError; // if error during MB conversion // // If cbMultiByte is 0, then we can't use pMBStr. In this // case, we simply want to count the number of characters that // would be written to the buffer. // if (cbMultiByte == 0) { BYTE pTempStr[2]; // tmp buffer - 2 bytes for DBCS // // Set most significant bit of flags to indicate to the // GetMBComp routine that it's using a temporary storage // area, so don't back up in the buffer. // SET_MSB(dwFlags); // // For each wide char, translate it to its corresponding multibyte // char and increment the multibyte character count. // if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // while (pWCStr < pEndWCStr) { // // Get the translation. // mbCount += GetMBCompSB( pHashN, dwFlags, pWCStr, pTempStr, mbCount, wDefault, pUsedDef ); pWCStr++; } } else { // // Multi Byte Character Code Page. // while (pWCStr < pEndWCStr) { // // Get the translation. // mbCount += GetMBCompMB( pHashN, dwFlags, pWCStr, pTempStr, mbCount, wDefault, pUsedDef, &fError, FALSE ); pWCStr++; } } } else { // // Initialize multibyte loop pointers. // pEndMBStr = pMBStr + cbMultiByte; // // For each wide char, translate it to its corresponding // multibyte char, store it in pMBStr, and increment the // multibyte character count. // if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // while ((pWCStr < pEndWCStr) && (pMBStr < pEndMBStr)) { // // Get the translation. // mbIncr = GetMBCompSB( pHashN, dwFlags, pWCStr, pMBStr, mbCount, wDefault, pUsedDef ); pWCStr++; mbCount += mbIncr; pMBStr += mbIncr; } } else { // // Multi Byte Character Code Page. // while ((pWCStr < pEndWCStr) && (pMBStr < pEndMBStr)) { // // Get the translation. // mbIncr = GetMBCompMB( pHashN, dwFlags, pWCStr, pMBStr, mbCount, wDefault, pUsedDef, &fError, ((pMBStr + 1) < pEndMBStr) ? FALSE : TRUE ); if (fError) { // // Not enough room in the buffer. // break; } pWCStr++; mbCount += mbIncr; pMBStr += mbIncr; } } // // Make sure multibyte character buffer was large enough. // if (pWCStr < pEndWCStr) { SetLastError(ERROR_INSUFFICIENT_BUFFER); return (0); } } // // Return the number of characters written (or that would have // been written) to the buffer. // return (mbCount); } //////////////////////////////////////////////////////////////////////////// // // GetMBCompSB // // Fills in pMBStr with the byte character(s) for the corresponding wide // character from the appropriate translation table and returns the number // of byte characters written to pMBStr. This routine is only called if // the defaultcheck and compositecheck flags were both set. // // NOTE: Most significant bit of dwFlags parameter is used by this routine // to indicate that the caller only wants the count of the number of // characters written, not the string (ie. do not back up in buffer). // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// int GetMBCompSB( PCP_HASH pHashN, DWORD dwFlags, LPWSTR pWCStr, LPBYTE pMBStr, int mbCount, WORD wDefault, LPBOOL pUsedDef) { WCHAR PreComp; // precomposed wide character if ((pTblPtrs->pDefaultSortkey == NULL) || (!IS_NONSPACE_ONLY(pTblPtrs->pDefaultSortkey, *pWCStr))) { // // Get the 1:1 translation from wide char to single byte. // GET_WC_TRANSLATION_SB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef, dwFlags ); return (1); } else { if (mbCount < 1) { // // Need to handle the nonspace character by itself, since // it is the first character in the string. // if (dwFlags & WC_DISCARDNS) { // // Discard the non-spacing char, so just return with // zero chars written. // return (0); } else if (dwFlags & WC_DEFAULTCHAR) { // // Need to replace the nonspace character with the default // character and return the number of characters written // to the multibyte string. // *pUsedDef = TRUE; *pMBStr = LOBYTE(wDefault); return (1); } else // WC_SEPCHARS - default { // // Get the 1:1 translation from wide char to multibyte // of the non-spacing char and return the number of // characters written to the multibyte string. // GET_WC_TRANSLATION_SB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef, dwFlags ); return (1); } } else if (PreComp = GetPreComposedChar(*pWCStr, *(pWCStr - 1))) { // // Back up in the single byte string and write the // precomposed char. // if (!IS_MSB(dwFlags)) { pMBStr--; } GET_WC_TRANSLATION_SB( pHashN, PreComp, pMBStr, wDefault, pUsedDef, dwFlags ); return (0); } else { if (dwFlags & WC_DISCARDNS) { // // Discard the non-spacing char, so just return with // zero chars written. // return (0); } else if (dwFlags & WC_DEFAULTCHAR) { // // Need to replace the base character with the default // character. Since we've already written the base // translation char in the single byte string, we need to // back up in the single byte string and write the default // char. // if (!IS_MSB(dwFlags)) { pMBStr--; } *pUsedDef = TRUE; *pMBStr = LOBYTE(wDefault); return (0); } else // WC_SEPCHARS - default { // // Get the 1:1 translation from wide char to multibyte // of the non-spacing char and return the number of // characters written to the multibyte string. // GET_WC_TRANSLATION_SB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef, dwFlags ); return (1); } } } } //////////////////////////////////////////////////////////////////////////// // // GetMBCompMB // // Fills in pMBStr with the byte character(s) for the corresponding wide // character from the appropriate translation table and returns the number // of byte characters written to pMBStr. This routine is only called if // the defaultcheck and compositecheck flags were both set. // // If the buffer was too small, the fError flag will be set to TRUE. // // NOTE: Most significant bit of dwFlags parameter is used by this routine // to indicate that the caller only wants the count of the number of // characters written, not the string (ie. do not back up in buffer). // // 05-31-91 JulieB Created. //////////////////////////////////////////////////////////////////////////// int GetMBCompMB( PCP_HASH pHashN, DWORD dwFlags, LPWSTR pWCStr, LPBYTE pMBStr, int mbCount, WORD wDefault, LPBOOL pUsedDef, BOOL *fError, BOOL fOnlyOne) { WCHAR PreComp; // precomposed wide character BYTE pTmpSp[2]; // temp space - 2 bytes for DBCS int nCnt; // number of characters written *fError = FALSE; if ((pTblPtrs->pDefaultSortkey == NULL) || (!IS_NONSPACE_ONLY(pTblPtrs->pDefaultSortkey, *pWCStr))) { // // Get the 1:1 translation from wide char to multibyte. // This also handles DBCS and returns the number of characters // written to the multibyte string. // GET_WC_TRANSLATION_MB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef, nCnt, fOnlyOne, dwFlags ); if (nCnt == 0) { *fError = TRUE; } return (nCnt); } else { if (mbCount < 1) { // // Need to handle the nonspace character by itself, since // it is the first character in the string. // if (dwFlags & WC_DISCARDNS) { // // Discard the non-spacing char, so just return with // zero chars written. // return (0); } else if (dwFlags & WC_DEFAULTCHAR) { // // Need to replace the nonspace character with the default // character and return the number of characters written // to the multibyte string. // *pUsedDef = TRUE; COPY_MB_CHAR( wDefault, pMBStr, nCnt, fOnlyOne ); if (nCnt == 0) { *fError = TRUE; } return (nCnt); } else // WC_SEPCHARS - default { // // Get the 1:1 translation from wide char to multibyte // of the non-spacing char and return the number of // characters written to the multibyte string. // GET_WC_TRANSLATION_MB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef, nCnt, fOnlyOne, dwFlags ); if (nCnt == 0) { *fError = TRUE; } return (nCnt); } } else if (PreComp = GetPreComposedChar(*pWCStr, *(pWCStr - 1))) { // // Get the 1:1 translation from wide char to multibyte // of the precomposed char, back up in the multibyte string, // write the precomposed char, and return the DIFFERENCE of // the number of characters written to the the multibyte // string. // GET_WC_TRANSLATION_MB( pHashN, *(pWCStr - 1), pTmpSp, wDefault, pUsedDef, nCnt, fOnlyOne, dwFlags ); if (nCnt == 0) { *fError = TRUE; return (nCnt); } if (!IS_MSB(dwFlags)) { pMBStr -= nCnt; } GET_WC_TRANSLATION_MB( pHashN, PreComp, pMBStr, wDefault, pUsedDef, mbCount, fOnlyOne, dwFlags ); if (mbCount == 0) { *fError = TRUE; } return (mbCount - nCnt); } else { if (dwFlags & WC_DISCARDNS) { // // Discard the non-spacing char, so just return with // zero chars written. // return (0); } else if (dwFlags & WC_DEFAULTCHAR) { // // Need to replace the base character with the default // character. Since we've already written the base // translation char in the multibyte string, we need to // back up in the multibyte string and return the // DIFFERENCE of the number of characters written // (could be negative). // // // If the previous character written is the default // character, then the base character for this nonspace // character has already been replaced. Simply throw // this character away and return zero chars written. // if (!IS_MSB(dwFlags)) { // // Not using a temporary buffer, so find out if the // previous character translated was the default char. // if ((MAKEWORD(*(pMBStr - 1), 0) == wDefault) || ((mbCount > 1) && (MAKEWORD(*(pMBStr - 1), *(pMBStr - 2)) == wDefault))) { return (0); } } else { // // Using a temporary buffer. The temp buffer is 2 bytes // in length and contains the previous character written. // if ((MAKEWORD(*pMBStr, 0) == wDefault) || ((mbCount > 1) && (MAKEWORD(*pMBStr, *(pMBStr + 1)) == wDefault))) { return (0); } } // // Get the 1:1 translation from wide char to multibyte // of the base char, back up in the multibyte string, // write the default char, and return the DIFFERENCE of // the number of characters written to the the multibyte // string. // GET_WC_TRANSLATION_MB( pHashN, *(pWCStr - 1), pTmpSp, wDefault, pUsedDef, nCnt, fOnlyOne, dwFlags ); if (nCnt == 0) { *fError = TRUE; return (nCnt); } if (!IS_MSB(dwFlags)) { pMBStr -= nCnt; } *pUsedDef = TRUE; COPY_MB_CHAR( wDefault, pMBStr, mbCount, fOnlyOne ); if (mbCount == 0) { *fError = TRUE; } return (mbCount - nCnt); } else // WC_SEPCHARS - default { // // Get the 1:1 translation from wide char to multibyte // of the non-spacing char and return the number of // characters written to the multibyte string. // GET_WC_TRANSLATION_MB( pHashN, *pWCStr, pMBStr, wDefault, pUsedDef, nCnt, fOnlyOne, dwFlags ); if (nCnt == 0) { *fError = TRUE; } return (nCnt); } } } } //////////////////////////////////////////////////////////////////////////// // // GetMacCodePage // // Returns the system default Mac code page. // // 09-22-93 JulieB Created. //////////////////////////////////////////////////////////////////////////// UINT GetMacCodePage() { PKEY_VALUE_FULL_INFORMATION pKeyValueFull; // ptr to query information BYTE pStatic[MAX_KEY_VALUE_FULLINFO]; // ptr to static buffer UNICODE_STRING ObUnicodeStr; // unicode string UINT CodePage; // code page value PCP_HASH pHashN; // ptr to hash node // // See if the Mac code page globals have been initialized yet. // If they have, return the mac code page value. // if (gMacCodePage != 0) { return (gMacCodePage); } // // Make sure code page key is open. // OPEN_CODEPAGE_KEY(NLS_DEFAULT_MACCP); // // Query the registry for the Mac CP value. // CodePage = 0; pKeyValueFull = (PKEY_VALUE_FULL_INFORMATION)pStatic; if ((QueryRegValue( hCodePageKey, NLS_VALUE_MACCP, &pKeyValueFull, MAX_KEY_VALUE_FULLINFO, NULL )) == NO_ERROR) { // // Convert the value to an integer. // RtlInitUnicodeString(&ObUnicodeStr, GET_VALUE_DATA_PTR(pKeyValueFull)); if (RtlUnicodeStringToInteger(&ObUnicodeStr, 10, (PULONG)&CodePage)) { CodePage = 0; } } // // Make sure the CodePage value was set. // if (CodePage == 0) { // // Registry value is corrupt, so use default Mac code page. // CodePage = NLS_DEFAULT_MACCP; } // // Get the hash node for the Mac code page. // pHashN = GetCPHashNode(CodePage); // // Make sure the Mac hash node is valid. // if (pHashN == NULL) { // // Invalid hash node, which means either the registry is // corrupt, or setup failed to install a file. Use the // Ansi code page values. // CodePage = gAnsiCodePage; pHashN = gpACPHashN; } // // Set the final MAC CP values. // RtlEnterCriticalSection(&gcsTblPtrs); if (gMacCodePage == 0) { gpMACCPHashN = pHashN; gMacCodePage = CodePage; } RtlLeaveCriticalSection(&gcsTblPtrs); // // Return the Mac code page value. // return (gMacCodePage); } //////////////////////////////////////////////////////////////////////////// // // SpecialMBToWC // // Maps a multibyte character string to its wide character string // counterpart. // // 08-21-95 JulieB Created. //////////////////////////////////////////////////////////////////////////// int SpecialMBToWC( PCP_HASH pHashN, DWORD dwFlags, LPCSTR lpMultiByteStr, int cbMultiByte, LPWSTR lpWideCharStr, int cchWideChar) { register LPBYTE pMBStr; // ptr to search through MB string register LPWSTR pWCStr; // ptr to search through WC string LPBYTE pEndMBStr; // ptr to end of MB search string LPWSTR pEndWCStr; // ptr to end of WC string buffer int mbIncr; // amount to increment pMBStr int wcCount = 0; // count of wide chars written PMB_TABLE pMBTbl; // ptr to MB table int ctr; // loop counter // // Initialize multibyte character loop pointers. // pMBStr = (LPBYTE)lpMultiByteStr; pEndMBStr = pMBStr + cbMultiByte; // // Get the MB table. // pMBTbl = pHashN->pMBTbl; // // If cchWideChar is 0, then we can't use lpWideCharStr. In this // case, we simply want to count the number of characters that would // be written to the buffer. // if (cchWideChar == 0) { // // For each multibyte char, translate it to its corresponding // wide char and increment the wide character count. // if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // wcCount = (int)(pEndMBStr - pMBStr); } else { // // Multi Byte Character Code Page. // WCHAR pTempStr[MAX_COMPOSITE]; // tmp buffer pEndWCStr = pTempStr + MAX_COMPOSITE; while (pMBStr < pEndMBStr) { GET_WC_MULTI( pHashN, pMBTbl, pMBStr, pEndMBStr, pTempStr, pEndWCStr, mbIncr ); pMBStr += mbIncr; wcCount++; } } } else { // // Initialize wide character loop pointers. // pWCStr = lpWideCharStr; pEndWCStr = pWCStr + cchWideChar; // // For each multibyte char, translate it to its corresponding // wide char, store it in lpWideCharStr, and increment the wide // character count. // if (IS_SBCS_CP(pHashN)) { // // Single Byte Character Code Page. // wcCount = (int)(pEndMBStr - pMBStr); if ((pEndWCStr - pWCStr) < wcCount) { wcCount = (int)(pEndWCStr - pWCStr); } if (dwFlags & MB_INVALID_CHAR_CHECK) { // // Error check flag is set. // for (ctr = wcCount; ctr > 0; ctr--) { GET_WC_SINGLE_SPECIAL( pHashN, pMBTbl, pMBStr, pWCStr ); pMBStr++; pWCStr++; } } else { // // Error check flag is NOT set. // for (ctr = wcCount; ctr > 0; ctr--) { GET_WC_SINGLE( pMBTbl, pMBStr, pWCStr ); pMBStr++; pWCStr++; } } } else { // // Multi Byte Character Code Page. // if (dwFlags & MB_INVALID_CHAR_CHECK) { // // Error check flag is set. // while ((pMBStr < pEndMBStr) && (pWCStr < pEndWCStr)) { GET_WC_MULTI_ERR_SPECIAL( pHashN, pMBTbl, pMBStr, pEndMBStr, pWCStr, pEndWCStr, mbIncr ); pMBStr += mbIncr; pWCStr++; } wcCount = (int)(pWCStr - lpWideCharStr); } else { // // Error check flag is NOT set. // while ((pMBStr < pEndMBStr) && (pWCStr < pEndWCStr)) { GET_WC_MULTI( pHashN, pMBTbl, pMBStr, pEndMBStr, pWCStr, pEndWCStr, mbIncr ); pMBStr += mbIncr; pWCStr++; } wcCount = (int)(pWCStr - lpWideCharStr); } } // // Make sure wide character buffer was large enough. // if (pMBStr < pEndMBStr) { SetLastError(ERROR_INSUFFICIENT_BUFFER); return (0); } } // // Return the number of characters written (or that would have // been written) to the buffer. // return (wcCount); }