gecko-dev/intl/uconv/ucvcn/nsGBKToUnicode.cpp

347 строки
11 KiB
C++

/* -*- Mode: C; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* ***** BEGIN LICENSE BLOCK *****
* Version: MPL 1.1/GPL 2.0/LGPL 2.1
*
* The contents of this file are subject to the Mozilla Public License Version
* 1.1 (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
* http://www.mozilla.org/MPL/
*
* Software distributed under the License is distributed on an "AS IS" basis,
* WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
* for the specific language governing rights and limitations under the
* License.
*
* The Original Code is mozilla.org code.
*
* The Initial Developer of the Original Code is
* Netscape Communications Corporation.
* Portions created by the Initial Developer are Copyright (C) 1998
* the Initial Developer. All Rights Reserved.
*
* Contributor(s):
*
* Alternatively, the contents of this file may be used under the terms of
* either of the GNU General Public License Version 2 or later (the "GPL"),
* or the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
* in which case the provisions of the GPL or the LGPL are applicable instead
* of those above. If you wish to allow use of your version of this file only
* under the terms of either the GPL or the LGPL, and not to allow others to
* use your version of this file under the terms of the MPL, indicate your
* decision by deleting the provisions above and replace them with the notice
* and other provisions required by the GPL or the LGPL. If you do not delete
* the provisions above, a recipient may use your version of this file under
* the terms of any one of the MPL, the GPL or the LGPL.
*
* ***** END LICENSE BLOCK ***** */
/**
* A character set converter from GBK to Unicode.
*
*
* @created 07/Sept/1999
* @author Yueheng Xu, Yueheng.Xu@intel.com
*/
#include "nsGBKToUnicode.h"
#include "nsUCvCnDll.h"
#include "gbku.h"
//------------------------------------------------------------
// nsGBKUnique2BytesToUnicode
//------------------------------------------------------------
class nsGBKUnique2BytesToUnicode : public nsTableDecoderSupport
{
public:
nsGBKUnique2BytesToUnicode();
virtual ~nsGBKUnique2BytesToUnicode()
{ }
protected:
};
static const PRUint16 g_utGBKUnique2Bytes[] = {
#include "gbkuniq2b.ut"
};
nsGBKUnique2BytesToUnicode::nsGBKUnique2BytesToUnicode()
: nsTableDecoderSupport(u2BytesCharset, nsnull,
(uMappingTable*) &g_utGBKUnique2Bytes, 1)
{
}
//------------------------------------------------------------
// nsGB18030Unique2BytesToUnicode
//------------------------------------------------------------
class nsGB18030Unique2BytesToUnicode : public nsTableDecoderSupport
{
public:
nsGB18030Unique2BytesToUnicode();
virtual ~nsGB18030Unique2BytesToUnicode()
{ }
protected:
};
static const PRUint16 g_utGB18030Unique2Bytes[] = {
#include "gb18030uniq2b.ut"
};
nsGB18030Unique2BytesToUnicode::nsGB18030Unique2BytesToUnicode()
: nsTableDecoderSupport(u2BytesCharset, nsnull,
(uMappingTable*) &g_utGB18030Unique2Bytes, 1)
{
}
//------------------------------------------------------------
// nsGB18030Unique4BytesToUnicode
//------------------------------------------------------------
class nsGB18030Unique4BytesToUnicode : public nsTableDecoderSupport
{
public:
nsGB18030Unique4BytesToUnicode();
virtual ~nsGB18030Unique4BytesToUnicode()
{ }
protected:
};
static const PRUint16 g_utGB18030Unique4Bytes[] = {
#include "gb180304bytes.ut"
};
nsGB18030Unique4BytesToUnicode::nsGB18030Unique4BytesToUnicode()
: nsTableDecoderSupport(u4BytesGB18030Charset, nsnull,
(uMappingTable*) &g_utGB18030Unique4Bytes, 1)
{
}
//----------------------------------------------------------------------
// Class nsGBKToUnicode [implementation]
//----------------------------------------------------------------------
// Subclassing of nsTablesDecoderSupport class [implementation]
#define LEGAL_GBK_MULTIBYTE_FIRST_BYTE(c) \
(UINT8_IN_RANGE(0x81, (c), 0xFE))
#define FIRST_BYTE_IS_SURROGATE(c) \
(UINT8_IN_RANGE(0x90, (c), 0xFE))
#define LEGAL_GBK_2BYTE_SECOND_BYTE(c) \
(UINT8_IN_RANGE(0x40, (c), 0x7E)|| UINT8_IN_RANGE(0x80, (c), 0xFE))
#define LEGAL_GBK_4BYTE_SECOND_BYTE(c) \
(UINT8_IN_RANGE(0x30, (c), 0x39))
#define LEGAL_GBK_4BYTE_THIRD_BYTE(c) \
(UINT8_IN_RANGE(0x81, (c), 0xFE))
#define LEGAL_GBK_4BYTE_FORTH_BYTE(c) \
(UINT8_IN_RANGE(0x30, (c), 0x39))
NS_IMETHODIMP nsGBKToUnicode::ConvertNoBuff(const char* aSrc,
PRInt32 * aSrcLength,
PRUnichar *aDest,
PRInt32 * aDestLength)
{
PRInt32 i=0;
PRInt32 iSrcLength = (*aSrcLength);
PRInt32 iDestlen = 0;
nsresult rv=NS_OK;
*aSrcLength = 0;
for (i=0;i<iSrcLength;i++)
{
if ( iDestlen >= (*aDestLength) )
{
rv = NS_OK_UDEC_MOREOUTPUT;
break;
}
// The valid range for the 1st byte is [0x81,0xFE]
if(LEGAL_GBK_MULTIBYTE_FIRST_BYTE(*aSrc))
{
if(i+1 >= iSrcLength)
{
rv = NS_OK_UDEC_MOREINPUT;
break;
}
// To make sure, the second byte has to be checked as well.
// In GBK, the second byte range is [0x40,0x7E] and [0x80,0XFE]
if(LEGAL_GBK_2BYTE_SECOND_BYTE(aSrc[1]))
{
// Valid GBK code
*aDest = mUtil.GBKCharToUnicode(aSrc[0], aSrc[1]);
if(UCS2_NO_MAPPING == *aDest)
{
// We cannot map in the common mapping, let's call the
// delegate 2 byte decoder to decode the gbk or gb18030 unique
// 2 byte mapping
if(! TryExtensionDecoder(aSrc, aDest))
{
*aDest = UCS2_NO_MAPPING;
}
}
aSrc += 2;
i++;
}
else if (LEGAL_GBK_4BYTE_SECOND_BYTE(aSrc[1]))
{
// from the first 2 bytes, it looks like a 4 byte GB18030
if(i+3 >= iSrcLength) // make sure we got 4 bytes
{
rv = NS_OK_UDEC_MOREINPUT;
break;
}
// 4 bytes patten
// [0x81-0xfe][0x30-0x39][0x81-0xfe][0x30-0x39]
// preset the
if (LEGAL_GBK_4BYTE_THIRD_BYTE(aSrc[2]) &&
LEGAL_GBK_4BYTE_FORTH_BYTE(aSrc[3]))
{
if ( ! FIRST_BYTE_IS_SURROGATE(aSrc[0]))
{
// let's call the delegated 4 byte gb18030 converter to convert it
if(! Try4BytesDecoder(aSrc, aDest))
*aDest = UCS2_NO_MAPPING;
} else {
// let's try supplement mapping
NS_ASSERTION(( (iDestlen+1) <= (*aDestLength) ), "no enouth output memory");
if ( (iDestlen+1) <= (*aDestLength) )
{
if(DecodeToSurrogate(aSrc, aDest))
{
// surrogte two PRUnichar
iDestlen++;
aDest++;
} else {
*aDest = UCS2_NO_MAPPING;
}
} else {
*aDest = UCS2_NO_MAPPING;
}
}
} else {
*aDest = UCS2_NO_MAPPING;
}
aSrc += 4;
i+=3;
}
else if ((PRUint8) aSrc[0] == (PRUint8)0xA0 )
{
// stand-alone (not followed by a valid second byte) 0xA0 !
// treat it as valid a la Netscape 4.x
*aDest = CAST_CHAR_TO_UNICHAR(*aSrc);
aSrc++;
} else {
// Invalid GBK code point (second byte should be 0x40 or higher)
*aDest = UCS2_NO_MAPPING;
aSrc++;
}
} else {
if(IS_ASCII(*aSrc))
{
// The source is an ASCII
*aDest = CAST_CHAR_TO_UNICHAR(*aSrc);
aSrc++;
} else {
if(IS_GBK_EURO(*aSrc)) {
*aDest = UCS2_EURO;
} else {
*aDest = UCS2_NO_MAPPING;
}
aSrc++;
}
}
iDestlen++;
aDest++;
*aSrcLength = i+1;
}
*aDestLength = iDestlen;
return rv;
}
void nsGBKToUnicode::CreateExtensionDecoder()
{
mExtensionDecoder = new nsGBKUnique2BytesToUnicode();
}
void nsGBKToUnicode::Create4BytesDecoder()
{
m4BytesDecoder = nsnull;
}
void nsGB18030ToUnicode::CreateExtensionDecoder()
{
mExtensionDecoder = new nsGB18030Unique2BytesToUnicode();
}
void nsGB18030ToUnicode::Create4BytesDecoder()
{
m4BytesDecoder = new nsGB18030Unique4BytesToUnicode();
}
PRBool nsGB18030ToUnicode::DecodeToSurrogate(const char* aSrc, PRUnichar* aOut)
{
NS_ASSERTION(FIRST_BYTE_IS_SURROGATE(aSrc[0]), "illegal first byte");
NS_ASSERTION(LEGAL_GBK_4BYTE_SECOND_BYTE(aSrc[1]), "illegal second byte");
NS_ASSERTION(LEGAL_GBK_4BYTE_THIRD_BYTE(aSrc[2]), "illegal third byte");
NS_ASSERTION(LEGAL_GBK_4BYTE_FORTH_BYTE(aSrc[3]), "illegal forth byte");
if(! FIRST_BYTE_IS_SURROGATE(aSrc[0]))
return PR_FALSE;
if(! LEGAL_GBK_4BYTE_SECOND_BYTE(aSrc[1]))
return PR_FALSE;
if(! LEGAL_GBK_4BYTE_THIRD_BYTE(aSrc[2]))
return PR_FALSE;
if(! LEGAL_GBK_4BYTE_FORTH_BYTE(aSrc[3]))
return PR_FALSE;
PRUint8 a1 = (PRUint8) aSrc[0];
PRUint8 a2 = (PRUint8) aSrc[1];
PRUint8 a3 = (PRUint8) aSrc[2];
PRUint8 a4 = (PRUint8) aSrc[3];
a1 -= (PRUint8)0x90;
a2 -= (PRUint8)0x30;
a3 -= (PRUint8)0x81;
a4 -= (PRUint8)0x30;
PRUint32 idx = (((a1 * 10 + a2 ) * 126 + a3) * 10) + a4;
*aOut++ = 0xD800 | (0x000003FF & (idx >> 10));
*aOut = 0xDC00 | (0x000003FF & idx);
return PR_TRUE;
}
PRBool nsGBKToUnicode::TryExtensionDecoder(const char* aSrc, PRUnichar* aOut)
{
if(!mExtensionDecoder)
CreateExtensionDecoder();
NS_ASSERTION(mExtensionDecoder, "cannot creqte 2 bytes unique converter");
if(mExtensionDecoder)
{
nsresult res = mExtensionDecoder->Reset();
NS_ASSERTION(NS_SUCCEEDED(res), "2 bytes unique conversoin reset failed");
PRInt32 len = 2;
PRInt32 dstlen = 1;
res = mExtensionDecoder->Convert(aSrc,&len, aOut, &dstlen);
NS_ASSERTION(NS_FAILED(res) || ((len==2) && (dstlen == 1)),
"some strange conversion result");
// if we failed, we then just use the 0xfffd
// therefore, we ignore the res here.
if(NS_SUCCEEDED(res))
return PR_TRUE;
}
return PR_FALSE;
}
PRBool nsGBKToUnicode::DecodeToSurrogate(const char* aSrc, PRUnichar* aOut)
{
return PR_FALSE;
}
PRBool nsGBKToUnicode::Try4BytesDecoder(const char* aSrc, PRUnichar* aOut)
{
if(!m4BytesDecoder)
Create4BytesDecoder();
if(m4BytesDecoder)
{
nsresult res = m4BytesDecoder->Reset();
NS_ASSERTION(NS_SUCCEEDED(res), "4 bytes unique conversoin reset failed");
PRInt32 len = 4;
PRInt32 dstlen = 1;
res = m4BytesDecoder->Convert(aSrc,&len, aOut, &dstlen);
NS_ASSERTION(NS_FAILED(res) || ((len==4) && (dstlen == 1)),
"some strange conversion result");
// if we failed, we then just use the 0xfffd
// therefore, we ignore the res here.
if(NS_SUCCEEDED(res))
return PR_TRUE;
}
return PR_FALSE;
}