/*

* Copyright (c) 2002 Nokia Corporation and/or its subsidiary(-ies). 

* All rights reserved.

* This component and the accompanying materials are made available

* under the terms of "Eclipse Public License v1.0"

* which accompanies this distribution, and is available

* at the URL "http://www.eclipse.org/legal/epl-v10.html".

*

* Initial Contributors:

* Nokia Corporation - initial contribution.

*

* Contributors:

*

* Description:    Implements the characterconversion plug-in

*                for ISCII characterset.

*

*/

#include <e32std.h>

#include <charconv.h>

#include <convgeneratedcpp.h>

#include <convutils.h>

//The maximum length of any intermediate buffer allocated for conversion.

const TInt KMaximumLengthOfIntermediateBuffer=5;

//The ISCII ATR code point, used for ISCII script switching mechanism.

const TUint KControlCharacterEscape=0xef;

//The number of Indic scripts supported by the plug-in.

//ISCII in general addresses all the Indic character sets.

const TUint KNumberOfIndicCharactersetsSupported = 1;

//The common reason for panic for all panics raised by the iscii plug-in

_LIT16(KPanicReason,"ISCII Plug-in Panic"); 

//The escape sequence for ISCII (ATR) is 0xEF and immidiate byte following 

//that is the script selection code for Devanagari.

_LIT8(KEscapeSequenceDevanagari,"\xef\x42");

//The sequence for Explicit Halant, in unicode it gets converted to VIRAMA+ZWNJ

_LIT8(KExplicitHalant,"\xe8\xe8");

//For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForExplicitHalant,"\xe8\xfc");

//The sequence for Soft Halant, in unicode it gets converted to VIRAMA+ZWJ

_LIT8(KSoftHalant,"\xe8\xe9");

//For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForSoftHalant,"\xe8\xfd");

//Devanagari character Om

_LIT8(KOm,"\xa1\xe9");

////For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForOm,"\xfe");

//Devanagari character Avagraha

_LIT8(KAvagraha,"\xea\xe9");

//For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForAvagraha,"\xff");

//Devanagari character VOCALIC RR

_LIT8(KVocalicRr,"\xaa\xe9");

//For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForVocalicRr,"\x80");

//Devanagari character VOCALIC LL

_LIT8(KVocalicLl,"\xa7\xe9");

//For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForVocalicLl,"\x81");

//Devanagari character VOCALIC L SIGN

_LIT8(KVocalicLSign,"\xdb\xe9");

//For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForVocalicLSign,"\x82");

//Devanagari character VOCALIC LL SIGN

_LIT8(KVocalicLlSign,"\xdc\xe9");

//For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForVocalicLlSign,"\x83");

//Devanagari character VOCALIC L

_LIT8(KVocalicL,"\xa6\xe9");

//For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForVocalicL,"\x84");

//Devanagari character VOCALIC RR SIGN

_LIT8(KVocalicRrSign,"\xdf\xe9");

//For supportiing ISCII to Unicode conversion of multi byte ISCII sequences, the

//sequence is converted to intermediate unused iscii code point.

_LIT8(KReplacementForVocalicRrSign,"\x85");

//Unconvertible ISCII character

_LIT8(KIsciiUnconvertibleCharacter,"\xeb");

enum TPanic

{

	//The panic raised by ConvertToUnicodeFromModalForeign_Internal() if the input

	//conversion flag is CCnvCharacterSetConverter::EInputConversionFlagStopAtFirstUnconvertibleCharacter

	EPanicBadInputConversionFlags=1,

	//Panic raised if the buffer does not start with the escape sequence 0xEF

	EPanicBadRemainderOfForeign,

	//Panic is raised if the length of the search buffer is greater than the length of the 

	//replacement buffer

	EPanicBadReplacementBuffer,

	//If the offset of start of the escape sequence is not an unsigned number.

	EPanicBadStartOfNextEscapeSequence

};

//The dummy datastructure for the dummy conversion data i.e. used for conversion if the 

//script selection code is not supported.

#define ARRAY_LENGTH(aArray) (sizeof(aArray)/sizeof((aArray)[0]))

LOCAL_D const SCnvConversionData::SOneDirectionData::SRange::UData::SKeyedTable1616::SEntry keyedTable1616_foreignToUnicode_1[]=

	{

		{

		0xa0,

		0xfffd

		}

	};

LOCAL_D const SCnvConversionData::SOneDirectionData::SRange::UData::SKeyedTable1616::SEntry keyedTable1616_unicodeToForeign_1[]=

	{

		{

		0xfffd,

		0xa0

		}

	};

LOCAL_D const SCnvConversionData::SVariableByteData::SRange foreignVariableByteDataRanges[]=

	{

		{

		0x00,

		0xff,

		0,

		0

		}

	};

LOCAL_D const SCnvConversionData::SOneDirectionData::SRange foreignToUnicodeDataRanges[]=

	{

		{

		0x00,

		0x7f,

		SCnvConversionData::SOneDirectionData::SRange::EDirect,

		0,

		0,

			{

			0

			}

		},

		{

		0xa0,

		0xff,

		SCnvConversionData::SOneDirectionData::SRange::EKeyedTable1616,

		0,

		0,

			{

			UData_SKeyedTable1616(keyedTable1616_foreignToUnicode_1)

			}

		}

	};

LOCAL_D const SCnvConversionData::SOneDirectionData::SRange unicodeToForeignDataRanges[]=

	{

		{

		0x0000,

		0x007f,

		SCnvConversionData::SOneDirectionData::SRange::EDirect,

		1,

		0,

			{

			0

			}

		},

		{

		0x00a0,

		0xffff,

		SCnvConversionData::SOneDirectionData::SRange::EKeyedTable1616,

		1,

		0,

			{

			UData_SKeyedTable1616(keyedTable1616_unicodeToForeign_1)

			}

		}

	};

//The dummy conversion data to be used for conversion if the iscii code sequence is not

//Devanagari (i.e. the script selection code is not 0x42 and something else.

//In this case the ISCII characters are converted to unconvertible characters.

LOCAL_D const SCnvConversionData conversionDataDummy=

	{

	SCnvConversionData::EFixedBigEndian,

		{

		ARRAY_LENGTH(foreignVariableByteDataRanges),

		foreignVariableByteDataRanges

		},

		{

		ARRAY_LENGTH(foreignToUnicodeDataRanges),

		foreignToUnicodeDataRanges

		},

		{

		ARRAY_LENGTH(unicodeToForeignDataRanges),

		unicodeToForeignDataRanges

		},

	NULL,

	NULL

	};

#ifdef EKA2

///////////////////////////////////////////////////////////////

// 3.1 Code

// INCLUDES

#include <ecom/implementationproxy.h>

#include <charactersetconverter.h>

/**

* The character conversion plug-in implementation for Iscii.

*

*  @lib ecom.lib

*  @since Series 60 3.1

*/

class CIsciiImplementation : public CCharacterSetConverterPluginInterface

{

    public:

    	//From CCharacterSetConverterPluginInterface

        virtual const TDesC8& ReplacementForUnconvertibleUnicodeCharacters();

	//From CCharacterSetConverterPluginInterface

        virtual TInt ConvertFromUnicode(

            CCnvCharacterSetConverter::TEndianness aDefaultEndiannessOfForeignCharacters, 

            const TDesC8& aReplacementForUnconvertibleUnicodeCharacters, 

            TDes8& aForeign, 

            const TDesC16& aUnicode, 

            CCnvCharacterSetConverter::TArrayOfAscendingIndices& aIndicesOfUnconvertibleCharacters );

	//From CCharacterSetConverterPluginInterface

        virtual TInt ConvertToUnicode(

            CCnvCharacterSetConverter::TEndianness aDefaultEndiannessOfForeignCharacters, 

            TDes16& aUnicode, 

            const TDesC8& aForeign, 

            TInt&, 

            TInt& aNumberOfUnconvertibleCharacters, 

            TInt& aIndexOfFirstByteOfFirstUnconvertibleCharacter );

	//From CCharacterSetConverterPluginInterface

        virtual TBool IsInThisCharacterSetL(

            TBool& aSetToTrue, 

            TInt& aConfidenceLevel, 

            const TDesC8& );

        static CIsciiImplementation* NewL();

        virtual ~CIsciiImplementation();

    private:

        CIsciiImplementation();

};

//Checks if a descriptor starts with another descriptor at the begining.

LOCAL_C TBool IsStartOf(const TDesC8& aEscapeSequence, const TDesC8& aBuffer)

{

	const TInt lengthOfStart=aEscapeSequence.Length();

	return (aBuffer.Length()>=lengthOfStart) && (aBuffer.Left(lengthOfStart)==aEscapeSequence);

}

// -----------------------------------------------------------------------------

// MatchesEscapeSequence()

//If the remainder of the foreign text starts with the passed escapesequence, modifies the remainder of the foreign text

//and  sets the homogeneous run buffer that uses the same conversion data.

//The homogeneous run buffer is the buffer that will use a single conversion data, it doesn't contain the attribute code

// neither it contains the script switching code.

//The aRemainderOfForeign buffer 

// Returns: ETrue: If the sequence contains the escape sequence

//              EFalse: If the sequence does not contain the escape sequence

//

// -----------------------------------------------------------------------------

//

LOCAL_C TBool MatchesEscapeSequence(TInt& aNumberOfForeignBytesConsumed, TPtrC8& aHomogeneousRun,

		TPtrC8& aRemainderOfForeign, const TDesC8& aEscapeSequence)

{

	const TInt lengthOfEscapeSequence=aEscapeSequence.Length();

	if (IsStartOf(aEscapeSequence, aRemainderOfForeign))

		{

		aRemainderOfForeign.Set(aRemainderOfForeign.Mid(lengthOfEscapeSequence));

		const TInt startOfNextEscapeSequence=aRemainderOfForeign.Locate(KControlCharacterEscape);

		if (startOfNextEscapeSequence==KErrNotFound)

			{

			aHomogeneousRun.Set(aRemainderOfForeign);

			aRemainderOfForeign.Set(NULL, 0);

			}

		else

			{

			aHomogeneousRun.Set(aRemainderOfForeign.Left(startOfNextEscapeSequence));

			aRemainderOfForeign.Set(aRemainderOfForeign.Mid(startOfNextEscapeSequence));

			}

		aNumberOfForeignBytesConsumed+=lengthOfEscapeSequence;

		return ETrue;

		}

	return EFalse;

}

// -----------------------------------------------------------------------------

// NextHomogeneousForeignRun()

//Matches the escape sequence of each of the elements of the SState array with the remainder of

//foreign text and if the escape sequence matches with the start of remainder of the foreign text, 

//then the conversion data is set to the conversion data corresponding to the escape sequence

//Also the homogeneous foreign text for conversion with the same escape sequence is set.

// Returns: ETrue: If length of the remainder of foreign buffer is nonzero.

//              EFalse: If length of the remainder of foreign buffer is zero.

//

// -----------------------------------------------------------------------------

//

LOCAL_C TBool NextHomogeneousForeignRun(const SCnvConversionData*& aConversionData, TInt& aNumberOfForeignBytesConsumed, TPtrC8& aHomogeneousRun, TPtrC8& aRemainderOfForeign, const TArray<CnvUtilities::SState>& aArrayOfStates, TUint& aOutputConversionFlags)

{

	TBool returnValue = EFalse;

	TBool foundState = EFalse;

	__ASSERT_DEBUG((aRemainderOfForeign.Length()==0) || (aRemainderOfForeign[0]==KControlCharacterEscape), User::Panic(KPanicReason,EPanicBadRemainderOfForeign));

	if (aRemainderOfForeign.Length()==0)

		{

		return returnValue;

		}

	const TInt numberOfStates=aArrayOfStates.Count();

	TInt i;

	for (i=0; i<numberOfStates; ++i)

		{

		const CnvUtilities::SState& state=aArrayOfStates[i];

		if (MatchesEscapeSequence(aNumberOfForeignBytesConsumed, aHomogeneousRun, aRemainderOfForeign, *state.iEscapeSequence))

			{

			aConversionData=state.iConversionData;

			foundState = ETrue;

			}

		}

	if(!foundState)

		{

		for (i=0; i<numberOfStates; ++i)

			{

			if (IsStartOf(aRemainderOfForeign, *aArrayOfStates[i].iEscapeSequence))

				{

				// aRemainderOfForeign ends with a truncated escape sequence, so ConvertToUnicode cannot convert any more

				aOutputConversionFlags|=CCnvCharacterSetConverter::EOutputConversionFlagInputIsTruncated;

				break;

				}

			}

		MatchesEscapeSequence(aNumberOfForeignBytesConsumed,aHomogeneousRun,aRemainderOfForeign,aRemainderOfForeign.Left(2));

		aConversionData = &conversionDataDummy;	

		returnValue = ETrue;

		}

		if (aHomogeneousRun.Length()>0)

			{

			returnValue = ETrue;

			}

		return returnValue;		

}

// -----------------------------------------------------------------------------

// ConvertFromUnicodeIntermediateBufferInPlace()

//Default implementation for conversion to the intermediate buffer

//It modifies the unicode buffer before it is converted back to iscii.

//The current implementation of iscii plug-in doesn't require any

//modification to the default implementation

// Returns: Nothing

//

// -----------------------------------------------------------------------------

//

LOCAL_C void ConvertFromUnicodeIntermediateBufferInPlace(TInt aStartPositionInDescriptor, TDes8& aDescriptor, TInt& aNumberOfCharactersThatDroppedOut)

{

	CnvUtilities::ConvertFromIntermediateBufferInPlace(aStartPositionInDescriptor, aDescriptor, aNumberOfCharactersThatDroppedOut, KEscapeSequenceDevanagari, 1);

}

// -----------------------------------------------------------------------------

// DoFindAndModifyBuffer()

//Modifies the iscii buffer by replacing the search buffer with the replacement buffer.

//Introduced for handling multibyte iscii sequence.

//Takes the search buffer array and the replacement buffer arrays as input to it and replaces all

//the occurances of the search buffer with the corresponding replace buffer.

// Returns: Nothing

//

// -----------------------------------------------------------------------------

//

LOCAL_C void DoFindAndModifyBuffer(TDes8& aModifyBuffer,const TDesC8& aSearchBuffer,const TDesC8& aReplaceBuffer)

{

	FOREVER

	{

		TInt offset;

		__ASSERT_ALWAYS((aSearchBuffer.Length()>= aReplaceBuffer.Length()),User::Panic(KPanicReason,EPanicBadReplacementBuffer));

		if((offset = aModifyBuffer.Find(aSearchBuffer)) != KErrNotFound)

		{

			TUint8 *pointerToBuffer = const_cast<TUint8*> (aModifyBuffer.Ptr());

			Mem::Copy(pointerToBuffer+offset,aReplaceBuffer.Ptr(),aReplaceBuffer.Length());

			Mem::Copy(pointerToBuffer+offset+aReplaceBuffer.Length(),pointerToBuffer+offset+aSearchBuffer.Length(),aModifyBuffer.Length()-aSearchBuffer.Length()-offset);

			aModifyBuffer.SetLength(aModifyBuffer.Length() - aSearchBuffer.Length() + aReplaceBuffer.Length());

		}

		else

			break;

	}

}

// -----------------------------------------------------------------------------

// FindAndModifyBuffer()

//Modifies the iscii buffer by replacing the search buffer with the replacement buffer.

//Calls DoFindAndModifyBuffer() and supplies the search buffer and replacement buffer.

//Introduced for handling multibyte iscii sequence.

// Returns: Nothing

//

// -----------------------------------------------------------------------------

//

LOCAL_C void FindAndModifyBuffer(TDes8& aModifyBuffer)

{

	RArray<TPtrC8> searchBuffer;

	RArray<TPtrC8> replaceBuffer;

	//If the passed buffer contains the replacement buffer,

	//Then it should not get converted to respective Unicode

	//buffer rather it should get converted to replacement for

	//unconvertible character.

	searchBuffer.Append(KReplacementForExplicitHalant().Right(1));

	searchBuffer.Append(KReplacementForSoftHalant().Right(1));

	searchBuffer.Append(KReplacementForOm().Right(1));

	searchBuffer.Append(KReplacementForAvagraha().Right(1));

	searchBuffer.Append(KReplacementForVocalicRr().Right(1));

	searchBuffer.Append(KReplacementForVocalicLl().Right(1));

	searchBuffer.Append(KReplacementForVocalicLSign().Right(1));

	searchBuffer.Append(KReplacementForVocalicLlSign().Right(1));

	searchBuffer.Append(KReplacementForVocalicL().Right(1));

	searchBuffer.Append(KReplacementForVocalicRrSign().Right(1));

	//All normal search buffers

	searchBuffer.Append(KExplicitHalant().Mid(0));

	searchBuffer.Append(KSoftHalant().Mid(0));

	searchBuffer.Append(KOm().Mid(0));

	searchBuffer.Append(KAvagraha().Mid(0));

	searchBuffer.Append(KVocalicRr().Mid(0));

	searchBuffer.Append(KVocalicLl().Mid(0));

	searchBuffer.Append(KVocalicLSign().Mid(0));

	searchBuffer.Append(KVocalicLlSign().Mid(0));

	searchBuffer.Append(KVocalicL().Mid(0));

	searchBuffer.Append(KVocalicRrSign().Mid(0));

	//The replacement buffer for the odd cases to restrict the 

	//replacement buffers not to get converted to the corresponding 

	//unicode buffer

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	replaceBuffer.Append(KIsciiUnconvertibleCharacter().Mid(0));

	//All normal replace buffers		

	replaceBuffer.Append(KReplacementForExplicitHalant().Mid(0));

	replaceBuffer.Append(KReplacementForSoftHalant().Mid(0));

	replaceBuffer.Append(KReplacementForOm().Mid(0));

	replaceBuffer.Append(KReplacementForAvagraha().Mid(0));

	replaceBuffer.Append(KReplacementForVocalicRr().Mid(0));

	replaceBuffer.Append(KReplacementForVocalicLl().Mid(0));

	replaceBuffer.Append(KReplacementForVocalicLSign().Mid(0));

	replaceBuffer.Append(KReplacementForVocalicLlSign().Mid(0));

	replaceBuffer.Append(KReplacementForVocalicL().Mid(0));

	replaceBuffer.Append(KReplacementForVocalicRrSign().Mid(0));

	for(TInt counter=0;counter<searchBuffer.Count();counter++)

	{

		DoFindAndModifyBuffer(aModifyBuffer,searchBuffer[counter],replaceBuffer[counter]);

	}

	searchBuffer.Reset();

	replaceBuffer.Reset();

}

// -----------------------------------------------------------------------------

// DoNormalizeReturnValue()

//Modifies the return value(Number of bytes did not get converted) according to the modified

//intermediate buffer. Takes the modified intermediate buffer, return value, search and replace

//buffers. Searches for the replace buffer in the intermediate buffer and if found, it looks for

//the actual buffer corresponding to the replace buffer and adds the difference of the lengths

//of the actual to replace buffer to the return value.

// Returns: Nothing

//

// -----------------------------------------------------------------------------

//

LOCAL_C void DoNormalizeReturnValue(TUint& aReturnValue,const TDesC8& aBuffer,RArray<TPtrC8>& anArrayOfSearches,RArray<TPtrC8>& anArrayOfReplaces)

{

	TPtr8 buffer(const_cast<TUint8*>(aBuffer.Ptr()),aBuffer.Length(),aBuffer.Length());

	TUint count = anArrayOfSearches.Count();

	FOREVER

	{

		TBool flag = EFalse;

		for(TUint i=0;i<count;i++)

		{

			TPtrC8 searchBufferForComparison(buffer.Right(anArrayOfReplaces[i].Length()));

			TInt returnCompare = searchBufferForComparison.Compare(anArrayOfReplaces[i]);

			if(returnCompare == 0)

			{

				flag =ETrue;

				aReturnValue += (anArrayOfSearches[i].Length() - anArrayOfReplaces[i].Length());

				buffer=buffer.MidTPtr(0,buffer.Length()-anArrayOfReplaces[i].Length());

				break;

			}

		}

		if(buffer.Length() == 0)

		{