michael@0: /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
michael@0: /* This Source Code Form is subject to the terms of the Mozilla Public
michael@0:  * License, v. 2.0. If a copy of the MPL was not distributed with this
michael@0:  * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
michael@0: 
michael@0: #include "nsUnicodeToISO2022JP.h"
michael@0: #include "nsUCVJADll.h"
michael@0: #include "nsUnicodeEncodeHelper.h"
michael@0: 
michael@0: //----------------------------------------------------------------------
michael@0: // Global functions and data [declaration]
michael@0: 
michael@0: // Basic mapping from Hankaku to Zenkaku
michael@0: // Nigori and Maru are taken care of outside this basic mapping
michael@0: static const char16_t gBasicMapping[0x40] =
michael@0: {
michael@0: // 0xff60
michael@0: 0xff60,0x3002,0x300c,0x300d,0x3001,0x30fb,0x30f2,0x30a1,
michael@0: // 0xff68
michael@0: 0x30a3,0x30a5,0x30a7,0x30a9,0x30e3,0x30e5,0x30e7,0x30c3,
michael@0: // 0xff70
michael@0: 0x30fc,0x30a2,0x30a4,0x30a6,0x30a8,0x30aa,0x30ab,0x30ad,
michael@0: // 0xff78
michael@0: 0x30af,0x30b1,0x30b3,0x30b5,0x30b7,0x30b9,0x30bb,0x30bd,
michael@0: // 0xff80
michael@0: 0x30bf,0x30c1,0x30c4,0x30c6,0x30c8,0x30ca,0x30cb,0x30cc,
michael@0: // 0xff88
michael@0: 0x30cd,0x30ce,0x30cf,0x30d2,0x30d5,0x30d8,0x30db,0x30de,
michael@0: // 0xff90
michael@0: 0x30df,0x30e0,0x30e1,0x30e2,0x30e4,0x30e6,0x30e8,0x30e9,
michael@0: // 0xff98
michael@0: 0x30ea,0x30eb,0x30ec,0x30ed,0x30ef,0x30f3,0x309b,0x309c
michael@0: };
michael@0: 
michael@0: // Do we need to check for Nigori for the next unicode ?
michael@0: #define NEED_TO_CHECK_NIGORI(u) (((0xff76<=(u))&&((u)<=0xff84))||((0xff8a<=(u))&&((u)<=0xff8e)))
michael@0: 
michael@0: // Do we need to check for Maru for the next unicode ?
michael@0: #define NEED_TO_CHECK_MARU(u) ((0xff8a<=(u))&&((u)<=0xff8e))
michael@0: 
michael@0: // The  unicode is in Katakana Hankaku block
michael@0: #define IS_HANKAKU(u) ((0xff61 <= (u)) && ((u) <= 0xff9f))
michael@0: #define IS_NIGORI(u) (0xff9e == (u))
michael@0: #define IS_MARU(u)   (0xff9f == (u))
michael@0: #define NIGORI_MODIFIER 1
michael@0: #define MARU_MODIFIER   2
michael@0: 
michael@0: static const uint16_t g_ufAsciiMapping [] = {
michael@0:   0x0001, 0x0004, 0x0005, 0x0008, 0x0000, 0x0000, 0x007F, 0x0000
michael@0: };
michael@0: 
michael@0: #define SIZE_OF_ISO2022JP_TABLES 5
michael@0: static const uint16_t * g_ufMappingTables[SIZE_OF_ISO2022JP_TABLES] = {
michael@0:   g_ufAsciiMapping,             // ASCII           ISOREG 6
michael@0:   g_uf0201GLMapping,            // JIS X 0201-1976 ISOREG 14
michael@0:   g_uf0208Mapping,              // JIS X 0208-1983 ISOREG 87
michael@0:   g_uf0208extMapping,           // JIS X 0208 - cp932 ext
michael@0:   g_uf0208Mapping,              // JIS X 0208-1978 ISOREG 42
michael@0: };
michael@0: 
michael@0: static const uScanClassID g_ufScanClassIDs[SIZE_OF_ISO2022JP_TABLES] = {
michael@0:   u1ByteCharset,                // ASCII           ISOREG 6
michael@0:   u1ByteCharset,                // JIS X 0201-1976 ISOREG 14
michael@0:   u2BytesCharset,               // JIS X 0208-1983 ISOREG 87
michael@0:   u2BytesCharset,               // JIS X 0208- cp932 ext
michael@0:   u2BytesCharset,               // JIS X 0208-1978 ISOREG 42
michael@0: };
michael@0: #define JIS_X_208_INDEX 2
michael@0: 
michael@0: //----------------------------------------------------------------------
michael@0: // Class nsUnicodeToISO2022JP [implementation]
michael@0: 
michael@0: // worst case max length: 
michael@0: //  1  2 3  4  5  6  7 8
michael@0: // ESC $ B XX XX ESC ( B
michael@0: nsUnicodeToISO2022JP::nsUnicodeToISO2022JP() 
michael@0: : nsEncoderSupport(8)
michael@0: {
michael@0:   Reset();
michael@0: }
michael@0: 
michael@0: nsUnicodeToISO2022JP::~nsUnicodeToISO2022JP() 
michael@0: {
michael@0: }
michael@0: 
michael@0: nsresult nsUnicodeToISO2022JP::ChangeCharset(int32_t aCharset,
michael@0:                                              char * aDest, 
michael@0:                                              int32_t * aDestLength)
michael@0: {
michael@0:   // both 2 and 3 generate the same escape sequence. 2 is for
michael@0:   // the standard JISx0208 table, and 3 is for theCP932 extensions
michael@0:   // therefore, we treat them as the same one.
michael@0:   if(((2 == aCharset) && ( 3 == mCharset)) ||
michael@0:      ((3 == aCharset) && ( 2 == mCharset)) )
michael@0:   {
michael@0:     mCharset = aCharset;
michael@0:   }
michael@0: 
michael@0:   if(aCharset == mCharset) 
michael@0:   {
michael@0:     *aDestLength = 0;
michael@0:     return NS_OK;
michael@0:   } 
michael@0:   
michael@0:   if (*aDestLength < 3) {
michael@0:     *aDestLength = 0;
michael@0:     return NS_OK_UENC_MOREOUTPUT;
michael@0:   }
michael@0: 
michael@0:   switch (aCharset) {
michael@0:     case 0: // ASCII ISOREG 6
michael@0:       aDest[0] = 0x1b;
michael@0:       aDest[1] = '(';
michael@0:       aDest[2] = 'B';
michael@0:       break;
michael@0:     case 1: // JIS X 0201-1976 ("Roman" set) ISOREG 14
michael@0:       aDest[0] = 0x1b;
michael@0:       aDest[1] = '(';
michael@0:       aDest[2] = 'J';
michael@0:       break;
michael@0:     case 2: // JIS X 0208-1983 ISOREG 87
michael@0:     case 3: // JIS X 0208-1983 
michael@0:             // we currently use this for CP932 ext
michael@0:       aDest[0] = 0x1b;
michael@0:       aDest[1] = '$';
michael@0:       aDest[2] = 'B';
michael@0:       break;
michael@0:     case 4: // JIS X 0201-1978 ISOREG 87- 
michael@0:             // we currently do not have a diff mapping for it.
michael@0:       aDest[0] = 0x1b;
michael@0:       aDest[1] = '$';
michael@0:       aDest[2] = '@';
michael@0:       break;
michael@0:   }
michael@0: 
michael@0:   mCharset = aCharset;
michael@0:   *aDestLength = 3;
michael@0:   return NS_OK;
michael@0: }
michael@0: 
michael@0: nsresult nsUnicodeToISO2022JP::ConvertHankaku(const char16_t * aSrc,
michael@0:                                               int32_t * aSrcLength,
michael@0:                                               char * aDest,
michael@0:                                               int32_t * aDestLength)
michael@0: {
michael@0:   nsresult res = NS_OK;
michael@0: 
michael@0:   const char16_t * src = aSrc;
michael@0:   const char16_t * srcEnd = aSrc + *aSrcLength;
michael@0:   char * dest = aDest;
michael@0:   char * destEnd = aDest + *aDestLength;
michael@0:   char16_t srcChar, tempChar;
michael@0:   int32_t bcr, bcw;
michael@0: 
michael@0:   bcw = destEnd - dest;
michael@0:   res = ChangeCharset(JIS_X_208_INDEX, dest, &bcw);
michael@0:   dest += bcw;
michael@0:   if (res != NS_OK) {
michael@0:     return res;
michael@0:   }
michael@0: 
michael@0:   while (src < srcEnd) {
michael@0:     srcChar = *src;
michael@0:     if (!IS_HANKAKU(srcChar)) {
michael@0:       break;
michael@0:     }
michael@0:     ++src;
michael@0:     tempChar = gBasicMapping[(srcChar) - 0xff60];
michael@0: 
michael@0:     if (src < srcEnd) {
michael@0:       // if the character could take a modifier, and the next char
michael@0:       // is a modifier, modify it and eat one char16_t
michael@0:       if (NEED_TO_CHECK_NIGORI(srcChar) && IS_NIGORI(*src)) {
michael@0:         tempChar += NIGORI_MODIFIER;
michael@0:         ++src;
michael@0:       } else if (NEED_TO_CHECK_MARU(srcChar) && IS_MARU(*src)) {
michael@0:         tempChar += MARU_MODIFIER;
michael@0:         ++src;
michael@0:       }
michael@0:     }
michael@0:     bcr = 1;
michael@0:     bcw = destEnd - dest;
michael@0:     res = nsUnicodeEncodeHelper::ConvertByTable(
michael@0:              &tempChar, &bcr, dest, &bcw, g_ufScanClassIDs[JIS_X_208_INDEX],
michael@0:              nullptr, (uMappingTable *) g_ufMappingTables[JIS_X_208_INDEX]);
michael@0:     dest += bcw;
michael@0:     if (res != NS_OK)
michael@0:       break;
michael@0:   }
michael@0:   *aDestLength = dest - aDest;
michael@0:   *aSrcLength = src - aSrc;
michael@0:   return res;
michael@0: }
michael@0: 
michael@0: //----------------------------------------------------------------------
michael@0: // Subclassing of nsTableEncoderSupport class [implementation]
michael@0: 
michael@0: NS_IMETHODIMP nsUnicodeToISO2022JP::ConvertNoBuffNoErr(
michael@0:                                     const char16_t * aSrc, 
michael@0:                                     int32_t * aSrcLength, 
michael@0:                                     char * aDest, 
michael@0:                                     int32_t * aDestLength)
michael@0: {
michael@0:   nsresult res = NS_OK;
michael@0: 
michael@0:   const char16_t * src = aSrc;
michael@0:   const char16_t * srcEnd = aSrc + *aSrcLength;
michael@0:   char * dest = aDest;
michael@0:   char * destEnd = aDest + *aDestLength;
michael@0:   int32_t bcr, bcw;
michael@0:   int32_t i;
michael@0: 
michael@0:   while (src < srcEnd) {
michael@0:     for (i=0; i< SIZE_OF_ISO2022JP_TABLES ; i++) {
michael@0:       bcr = 1;
michael@0:       bcw = destEnd - dest;
michael@0:       res = nsUnicodeEncodeHelper::ConvertByTable(src, &bcr, dest, &bcw, 
michael@0:                                       g_ufScanClassIDs[i], nullptr,
michael@0:                                       (uMappingTable *) g_ufMappingTables[i]);
michael@0:       if (res != NS_ERROR_UENC_NOMAPPING) break;
michael@0:     }
michael@0: 
michael@0:     if ( i>=  SIZE_OF_ISO2022JP_TABLES) {
michael@0:       if (IS_HANKAKU(*src)) {
michael@0:         bcr = srcEnd - src;
michael@0:         bcw = destEnd - dest;
michael@0:         res = ConvertHankaku(src, &bcr, dest, &bcw);
michael@0:         dest += bcw;
michael@0:         src += bcr;
michael@0:         if (res == NS_OK) continue;
michael@0:       } else {
michael@0:         res = NS_ERROR_UENC_NOMAPPING;
michael@0:         src++;
michael@0:       }
michael@0:     }
michael@0:     if (res != NS_OK) break;
michael@0: 
michael@0:     bcw = destEnd - dest;
michael@0:     res = ChangeCharset(i, dest, &bcw);
michael@0:     dest += bcw;
michael@0:     if (res != NS_OK) break;
michael@0: 
michael@0:     bcr = srcEnd - src;
michael@0:     bcw = destEnd - dest;
michael@0:     res = nsUnicodeEncodeHelper::ConvertByTable(src, &bcr, dest, &bcw, 
michael@0:                                       g_ufScanClassIDs[i], nullptr,
michael@0:                                       (uMappingTable *) g_ufMappingTables[i]);
michael@0:     src += bcr;
michael@0:     dest += bcw;
michael@0: 
michael@0:     if ((res != NS_OK) && (res != NS_ERROR_UENC_NOMAPPING)) break;
michael@0:     if (res == NS_ERROR_UENC_NOMAPPING) src--;
michael@0:   }
michael@0: 
michael@0:   *aSrcLength = src - aSrc;
michael@0:   *aDestLength  = dest - aDest;
michael@0:   return res;
michael@0: }
michael@0: 
michael@0: NS_IMETHODIMP nsUnicodeToISO2022JP::FinishNoBuff(char * aDest, 
michael@0:                                                  int32_t * aDestLength)
michael@0: {
michael@0:   ChangeCharset(0, aDest, aDestLength);
michael@0:   return NS_OK;
michael@0: }
michael@0: 
michael@0: NS_IMETHODIMP nsUnicodeToISO2022JP::Reset()
michael@0: {
michael@0:   mCharset = 0;
michael@0:   return nsEncoderSupport::Reset();
michael@0: }