/*

* Copyright (c) 1997-2009 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: 

*

*/

#include <e32std.h>

#include <e32base.h>

#include <utf.h>

const TUint KNotInBase64Alphabet=KMaxTUint;

enum TPanic

	{

	EPanicBad6BitNumber=1,

	EPanicBadUtf7Pointers1,

	EPanicBadUtf7Pointers2,

	EPanicBadUtf7Pointers3,

	EPanicBadUtf7Pointers4,

	EPanicBadUtf7Pointers5,

	EPanicBadUtf7Pointers6,

	EPanicBadUtf7Pointers7,

	EPanicBadUtf7Pointers8,

	EPanicBadUtf7Pointers9,

	EPanicBadUtf7Pointers10,

	EPanicBadUtf7Pointers11,

	EPanicNotInBase64Block,

	EPanicBadUnicodePointers1,

	EPanicBadUnicodePointers2,

	EPanicBadUnicodePointers3,

	EPanicBadUnicodePointers4,

	EPanicBadUnicodePointers5,

	EPanicBadUnicodePointers6,

	EPanicBadUnicodePointers7,

	EPanicBadUnicodePointers8,

	EPanicBadUnicodePointers9,

	EPanicBadUnicodePointers10,

	EPanicBadBitBufferState1,

	EPanicBadBitBufferState2,

	EPanicBadBitBufferState3,

	EPanicBadBitBufferState4,

	EPanicBadBitBufferState5,

	EPanicBadBitBufferState6,

	EPanicBadBitBufferState7,

	EPanicBadBitBufferState8,

	EPanicBadBitBufferState9,

	EPanicBadBitBufferState10,

	EPanicBadBitBufferState11,

	EPanicBadBitBufferState12,

	EPanicBadBitBufferState13,

	EPanicBadBitBufferState14,

	EPanicBadBitBufferState15,

	EPanicBadBitBufferState16,

	EPanicBadBitBufferState17,

	EPanicUnexpectedNumberOfLoopIterations,

	EPanicInitialEscapeCharacterButNoBase64,

	EPanicBase64SequenceDoesNotFallOnUnicodeCharacterBoundary,

	EPanicBadUtf8Pointers1,

	EPanicBadUtf8Pointers2,

	EPanicBadUtf8Pointers3,

	EPanicBadUtf8Pointers4,

	EPanicBadUtf8Pointers5,

	EPanicBadUtf8Pointers6,

	EPanicBadUtf8Pointers7,

	EPanicOutOfSyncUtf7Byte1,

	EPanicOutOfSyncUtf7Byte2,

	EPanicOutOfSyncBase64Decoding

	};

_LIT(KLitPanicText, "CHARCONV-UTF");

LOCAL_C void Panic(TPanic aPanic)

	{

	User::Panic(KLitPanicText, aPanic);

	}

inline TUint EscapeCharacterForStartingBase64Block(TBool aIsImapUtf7) {return aIsImapUtf7? '&': '+';}

LOCAL_C TUint Base64Decoding(TUint aMemberOfBase64Alphabet, TBool aIsImapUtf7)

	{

	if ((aMemberOfBase64Alphabet>='A') && (aMemberOfBase64Alphabet<='Z'))

		{

		return aMemberOfBase64Alphabet-'A';

		}

	if ((aMemberOfBase64Alphabet>='a') && (aMemberOfBase64Alphabet<='z'))

		{

		return aMemberOfBase64Alphabet-('a'-26);

		}

	if ((aMemberOfBase64Alphabet>='0') && (aMemberOfBase64Alphabet<='9'))

		{

		return aMemberOfBase64Alphabet+((26*2)-'0');

		}

	if (aMemberOfBase64Alphabet=='+')

		{

		return 62;

		}

	if (aMemberOfBase64Alphabet==STATIC_CAST(TUint, aIsImapUtf7? ',': '/'))

		{

		return 63;

		}

	return KNotInBase64Alphabet;

	}

LOCAL_C TUint Base64Encoding(TUint a6BitNumber, TBool aIsImapUtf7)

	{

	__ASSERT_DEBUG(a6BitNumber<64, Panic(EPanicBad6BitNumber));

	if ((a6BitNumber==63) && aIsImapUtf7)

		{

		return ',';

		}

	static const TUint8 base64Alphabet[64]={'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '+', '/'};

	return base64Alphabet[a6BitNumber];

	}

LOCAL_C TUint8* PointerToEscapeCharacterStartingBase64Block(TUint8* aPointerToUtf7Byte, const TUint8* aPointerToFirstUtf7Byte, TBool aIsImapUtf7)

	{

	__ASSERT_DEBUG(aPointerToUtf7Byte>=aPointerToFirstUtf7Byte, Panic(EPanicBadUtf7Pointers1));

	TUint8* pointerToCandidateEscapeCharacter=NULL;

	FOREVER

		{

		const TUint utf7Byte=*aPointerToUtf7Byte;

		if (utf7Byte==EscapeCharacterForStartingBase64Block(aIsImapUtf7))

			{

			pointerToCandidateEscapeCharacter=aPointerToUtf7Byte;

			}

		else if (Base64Decoding(utf7Byte, aIsImapUtf7)==KNotInBase64Alphabet)

			{

			break;

			}

		__ASSERT_DEBUG(aPointerToUtf7Byte>=aPointerToFirstUtf7Byte, Panic(EPanicBadUtf7Pointers2));

		if (aPointerToUtf7Byte<=aPointerToFirstUtf7Byte)

			{

			break;

			}

		--aPointerToUtf7Byte;

		}

	__ASSERT_DEBUG(pointerToCandidateEscapeCharacter!=NULL, Panic(EPanicNotInBase64Block));

	return pointerToCandidateEscapeCharacter;

	}

LOCAL_C TBool EncodeInUtf7Directly(TUint aUnicodeCharacter, TBool aIsImapUtf7, TBool aEncodeOptionalDirectCharactersInBase64)

	{

	if (aIsImapUtf7)

		{

		return (aUnicodeCharacter>=0x0020) && (aUnicodeCharacter<=0x007e);

		}

	if ((aUnicodeCharacter>=0x0021) && (aUnicodeCharacter<=0x007d))

		{

		if (aEncodeOptionalDirectCharactersInBase64)

			{

			return (((aUnicodeCharacter>=0x0041) && (aUnicodeCharacter<=0x005a)) ||

					((aUnicodeCharacter>=0x0061) && (aUnicodeCharacter<=0x007a)) ||

					((aUnicodeCharacter>=0x0027) && (aUnicodeCharacter<=0x0029)) ||

					((aUnicodeCharacter>=0x002b) && (aUnicodeCharacter<=0x003a)) ||

					(aUnicodeCharacter==0x003f));

			}

		return aUnicodeCharacter!=0x005c;

		}

	return (aUnicodeCharacter==0x0020) || (aUnicodeCharacter==0x0009) || (aUnicodeCharacter==0x000d) || (aUnicodeCharacter==0x000a);

	}

inline TBool BitBufferContainsNonZeroBits(TUint aBitBuffer, TInt aNumberOfBitsInBuffer)

	{

	return (aBitBuffer&((1<<aNumberOfBitsInBuffer)-1))!=0;

	}

/**  Converts Unicode text into UTF-7 encoding. The fucntion leaves with 

KErrCorrupt if the input string is corrupt.

@param aUnicode A UCS-2 encoded input string.

@param aEncodeOptionalDirectCharactersInBase64  If ETrue then 

characters from UTF-7 set O (optional direct characters) are encoded in 

Modified Base64. If EFalse the characters are encoded directly, 

as their ASCII equivalents.

@return A descriptor containing the UTF-7 encoded output string. */

EXPORT_C HBufC8* CnvUtfConverter::ConvertFromUnicodeToUtf7L(

										const TDesC16& aUnicode, 

										TBool aEncodeOptionalDirectCharactersInBase64)

	{

	// If aUnicode is  Null string, return an empty HBufC

	if (aUnicode.Length() == 0)

		{

		HBufC8* hBuf8 = HBufC8::NewL(1);

		return hBuf8;

		}

	// Otherwise, convert and store result in a buffer, reallocating that buffer if needed.

	TInt length = aUnicode.Length();

	const TInt bufsize = 100;

	TPtrC16 unicode (aUnicode);

	TBuf8<bufsize> buf;

	HBufC8* hBuf8 = HBufC8::NewLC(length);

	TPtr8 utf7 = hBuf8->Des();

	FOREVER

		{

		TInt unconverted = ConvertFromUnicodeToUtf7(buf, unicode, aEncodeOptionalDirectCharactersInBase64);

		if( unconverted == EErrorIllFormedInput || unconverted < 0)

			User::Leave(KErrCorrupt);

		if (utf7.Length() + buf.Length() > utf7.MaxLength())

			{

			// Reallocate the hBuf8

			hBuf8 = hBuf8->ReAllocL(utf7.Length() + buf.Length());

			CleanupStack::Pop();

			CleanupStack::PushL(hBuf8);

			utf7.Set(hBuf8->Des());

			}

		utf7.Append(buf);

		if (unconverted ==0) 

			break;

		unicode.Set(unicode.Right(unconverted));

		}

	CleanupStack::Pop();

	return hBuf8;

	}

/** Converts Unicode text into UTF-7 encoding.

@param aUtf7 On return, contains the UTF-7 encoded output string.

@param aUnicode A UCS-2 encoded input string.

@param aEncodeOptionalDirectCharactersInBase64 If ETrue then characters from 

UTF-7 set O (optional direct characters) are encoded in Modified Base64. If 

EFalse the characters are encoded directly, as their ASCII equivalents.

@return The number of unconverted characters left at the end of the input 

descriptor, or one of the error values defined in TError. */

EXPORT_C TInt CnvUtfConverter::ConvertFromUnicodeToUtf7(

										TDes8& aUtf7, 

										const TDesC16& aUnicode, 

										TBool aEncodeOptionalDirectCharactersInBase64)

	{

	return ConvertFromUnicodeToUtf7(aUtf7, aUnicode, EFalse, aEncodeOptionalDirectCharactersInBase64);

	}

TInt CnvUtfConverter::ConvertFromUnicodeToUtf7(TDes8& aUtf7, 

											   const TDesC16& aUnicode, 

											   TBool aIsImapUtf7, 

											   TBool aEncodeOptionalDirectCharactersInBase64)

	{

	if (aUnicode.Length()==0)

		{

		aUtf7.SetLength(0);

		return 0;

		}

	if (aUtf7.MaxLength()==0)

		{

		return aUnicode.Length();

		}

	const TUint escapeCharacterForStartingBase64Block=EscapeCharacterForStartingBase64Block(aIsImapUtf7);

	TUint8* pointerToPreviousUtf7Byte=CONST_CAST(TUint8*, aUtf7.Ptr()-1);

	const TUint8* const pointerToLastUtf7Byte=pointerToPreviousUtf7Byte+aUtf7.MaxLength();

	const TUint16* pointerToPreviousUnicodeCharacter=aUnicode.Ptr()-1;

	const TUint16* const pointerToLastUnicodeCharacter=pointerToPreviousUnicodeCharacter+aUnicode.Length();

	const TUint KIsInBase64Block=0x80000000u;

	TUint bitBuffer=0;

	TInt numberOfBitsInBuffer=0;

	FOREVER

		{

		__ASSERT_DEBUG(pointerToPreviousUtf7Byte<=pointerToLastUtf7Byte, Panic(EPanicBadUtf7Pointers3));

		__ASSERT_DEBUG(pointerToPreviousUnicodeCharacter<=pointerToLastUnicodeCharacter, Panic(EPanicBadUnicodePointers1));

		TUint currentUnicodeCharacter=(pointerToPreviousUnicodeCharacter==pointerToLastUnicodeCharacter)? 0: *(pointerToPreviousUnicodeCharacter+1);

		if ((pointerToPreviousUnicodeCharacter==pointerToLastUnicodeCharacter) || EncodeInUtf7Directly(currentUnicodeCharacter, aIsImapUtf7, aEncodeOptionalDirectCharactersInBase64))

			{

			__ASSERT_DEBUG((bitBuffer&KIsInBase64Block) || (numberOfBitsInBuffer==0), Panic(EPanicBadBitBufferState1));

			__ASSERT_DEBUG((numberOfBitsInBuffer==0) || (numberOfBitsInBuffer==2) || (numberOfBitsInBuffer==4), Panic(EPanicBadBitBufferState2));

			if (bitBuffer&KIsInBase64Block)

				{

				if (numberOfBitsInBuffer!=0)

					{

					if (pointerToLastUtf7Byte-pointerToPreviousUtf7Byte<2) // make sure there is enough space for the trailing '-' as well as the remains of the bitBuffer as the KIsInBase64Block flag is about to turned off, thus the trailing '-' may never get written

						{

						break;

						}

					++pointerToPreviousUtf7Byte;

					*pointerToPreviousUtf7Byte=STATIC_CAST(TUint8, Base64Encoding((bitBuffer<<(6-numberOfBitsInBuffer))&0x3f, aIsImapUtf7));

					}

				else

					{

					if (pointerToPreviousUtf7Byte==pointerToLastUtf7Byte)

						{

						break;

						}

					}

				++pointerToPreviousUtf7Byte;

				*pointerToPreviousUtf7Byte='-';

				bitBuffer=0;

				numberOfBitsInBuffer=0;

				}

			__ASSERT_DEBUG(pointerToPreviousUnicodeCharacter<=pointerToLastUnicodeCharacter, Panic(EPanicBadUnicodePointers2));

			if (pointerToPreviousUnicodeCharacter>=pointerToLastUnicodeCharacter)

				{

				break;

				}

			__ASSERT_DEBUG(pointerToPreviousUtf7Byte<=pointerToLastUtf7Byte, Panic(EPanicBadUtf7Pointers4));

			if (pointerToLastUtf7Byte-pointerToPreviousUtf7Byte<((currentUnicodeCharacter==escapeCharacterForStartingBase64Block)? 2: 1))

				{

				break;

				}

			++pointerToPreviousUtf7Byte;

			*pointerToPreviousUtf7Byte=STATIC_CAST(TUint8, currentUnicodeCharacter);

			++pointerToPreviousUnicodeCharacter;

			if (currentUnicodeCharacter==escapeCharacterForStartingBase64Block)

				{

				++pointerToPreviousUtf7Byte;

				*pointerToPreviousUtf7Byte='-';

				}

			}

		else

			{

			{

			TInt numberOfUtf7BytesRequired=(numberOfBitsInBuffer+16)/6; // "(numberOfBitsInBuffer+16)/6" is the number of iterations that will happen in the while loop below

			if (~bitBuffer&KIsInBase64Block)

				{

				++numberOfUtf7BytesRequired; // for the initial escapeCharacterForStartingBase64Block

				}

			if (pointerToLastUtf7Byte-pointerToPreviousUtf7Byte<numberOfUtf7BytesRequired)

				{

				break;

				}

			}

			if (~bitBuffer&KIsInBase64Block)

				{

				__ASSERT_DEBUG(pointerToPreviousUtf7Byte<pointerToLastUtf7Byte, Panic(EPanicBadUtf7Pointers5));

				++pointerToPreviousUtf7Byte;

				*pointerToPreviousUtf7Byte=STATIC_CAST(TUint8, escapeCharacterForStartingBase64Block);

				}

			bitBuffer<<=16;

			bitBuffer|=currentUnicodeCharacter;

			numberOfBitsInBuffer+=16;

			++pointerToPreviousUnicodeCharacter;

			__ASSERT_DEBUG(numberOfBitsInBuffer<=20, Panic(EPanicBadBitBufferState3));

			while (numberOfBitsInBuffer>=6)

				{

				numberOfBitsInBuffer-=6;

				__ASSERT_DEBUG(pointerToPreviousUtf7Byte<pointerToLastUtf7Byte, Panic(EPanicBadUtf7Pointers6));

				++pointerToPreviousUtf7Byte;

				*pointerToPreviousUtf7Byte=STATIC_CAST(TUint8, Base64Encoding((bitBuffer>>numberOfBitsInBuffer)&0x3f, aIsImapUtf7));

				}

			bitBuffer&=((1<<numberOfBitsInBuffer)-1); // zero all the consumed bits - not strictly necessary but it leaves the buffer in a cleaner state

			bitBuffer|=KIsInBase64Block;

			}

		}

	__ASSERT_DEBUG((bitBuffer&KIsInBase64Block) || (numberOfBitsInBuffer==0), Panic(EPanicBadBitBufferState4));

	__ASSERT_DEBUG((numberOfBitsInBuffer==0) || (numberOfBitsInBuffer==2) || (numberOfBitsInBuffer==4), Panic(EPanicBadBitBufferState5));

	if (bitBuffer&KIsInBase64Block)

		{

#if defined(_DEBUG)

		TInt numberOfLoopIterations=1;

#endif

		FOREVER // there should never be more than 2 iterations of this loop - the first "if" should always succeed the second time if it doesn't succeed the first time

			{

			__ASSERT_DEBUG(pointerToPreviousUtf7Byte<=pointerToLastUtf7Byte, Panic(EPanicBadUtf7Pointers7));

			__ASSERT_DEBUG((numberOfBitsInBuffer==0) || (numberOfBitsInBuffer==2) || (numberOfBitsInBuffer==4), Panic(EPanicBadBitBufferState6));

			__ASSERT_DEBUG(numberOfLoopIterations<=2, Panic(EPanicUnexpectedNumberOfLoopIterations));

#if defined(_DEBUG)

			++numberOfLoopIterations;

#endif

			if (pointerToLastUtf7Byte-pointerToPreviousUtf7Byte>=((numberOfBitsInBuffer==0)? 1: 2)) // if there's room to finish off the base-64 sequence by (i) flushing the bit-buffer and (ii) appending the trailing '-'

				{

				if (numberOfBitsInBuffer!=0)

					{

					__ASSERT_DEBUG(pointerToPreviousUtf7Byte<pointerToLastUtf7Byte, Panic(EPanicBadUtf7Pointers8));

					++pointerToPreviousUtf7Byte;

					*pointerToPreviousUtf7Byte=STATIC_CAST(TUint8, Base64Encoding((bitBuffer<<(6-numberOfBitsInBuffer))&0x3f, aIsImapUtf7));

					}

				__ASSERT_DEBUG(pointerToPreviousUtf7Byte<pointerToLastUtf7Byte, Panic(EPanicBadUtf7Pointers9));

				++pointerToPreviousUtf7Byte;

				*pointerToPreviousUtf7Byte='-';

				break;

				}

			// it is now necessary to move back pointerToPreviousUtf7Byte so that the base-64 sequence can be terminated - note it must be terminated on a Unicode character boundary hence the reason why pointerToPreviousUnicodeCharacter may be moved back too

			TUint8* pointerToEscapeCharacterStartingBase64Block=PointerToEscapeCharacterStartingBase64Block(pointerToPreviousUtf7Byte, aUtf7.Ptr(), aIsImapUtf7);

			const TInt oldNumberOfBase64Characters=pointerToPreviousUtf7Byte-pointerToEscapeCharacterStartingBase64Block;

			__ASSERT_DEBUG(oldNumberOfBase64Characters>0, Panic(EPanicInitialEscapeCharacterButNoBase64));

			__ASSERT_DEBUG(((oldNumberOfBase64Characters*6)+numberOfBitsInBuffer)%16==0, Panic(EPanicBase64SequenceDoesNotFallOnUnicodeCharacterBoundary));

			pointerToPreviousUnicodeCharacter-=((oldNumberOfBase64Characters*6)+numberOfBitsInBuffer)/16; // move back pointerToPreviousUnicodeCharacter to before the equivalent of the base-64 sequence

			pointerToPreviousUtf7Byte=pointerToEscapeCharacterStartingBase64Block;

			__ASSERT_DEBUG(*pointerToPreviousUtf7Byte==escapeCharacterForStartingBase64Block, Panic(EPanicBadUtf7Pointers10));

			if (oldNumberOfBase64Characters<4) // if the new base-64 sequence will be so short that it won't even be able to contain the UTF-7 encoding of a single Unicode character

				{

				--pointerToPreviousUtf7Byte; // move back pointerToPreviousUtf7Byte to before the escapeCharacterForStartingBase64Block

				break;

				}

			const TInt newNumberOfUnicodeCharacters=((oldNumberOfBase64Characters-1)*3)/8;

			pointerToPreviousUnicodeCharacter+=newNumberOfUnicodeCharacters;

			pointerToPreviousUtf7Byte+=((newNumberOfUnicodeCharacters*8)+2)/3;

			const TInt numberOfBitsToBeZeroedInLastBase64Character=(newNumberOfUnicodeCharacters%3)*2;

			if (numberOfBitsToBeZeroedInLastBase64Character!=0)

				{

				*pointerToPreviousUtf7Byte=STATIC_CAST(TUint8, Base64Encoding(Base64Decoding(*pointerToPreviousUtf7Byte, aIsImapUtf7)&0x3f&~((1<<numberOfBitsToBeZeroedInLastBase64Character)-1), aIsImapUtf7));

				}

			bitBuffer=KIsInBase64Block;

			numberOfBitsInBuffer=0;

			}

		}

	aUtf7.SetLength((pointerToPreviousUtf7Byte-aUtf7.Ptr())+1);

	return pointerToLastUnicodeCharacter-pointerToPreviousUnicodeCharacter;

	}

/** Converts Unicode text into UTF-8 encoding.

@param aUtf8 On return, contains the UTF-8 encoded output string.

@param aUnicode The Unicode-encoded input string.

@return The number of unconverted characters left at the end of the input 

descriptor, or one of the error values defined in TError. */

EXPORT_C TInt CnvUtfConverter::ConvertFromUnicodeToUtf8(TDes8& aUtf8, const TDesC16& aUnicode)

	{

	return ConvertFromUnicodeToUtf8(aUtf8, aUnicode, EFalse);

	}

/**  Converts Unicode text into UTF-8 encoding.

The variant of UTF-8 used internally by Java differs slightly from

standard UTF-8. The TBool argument controls the UTF-8

variant generated by this function. This function leaves with a 

KErrCorrupt if the input string is corrupt. 

@param aUnicode A UCS-2 encoded input string.

@return A pointer to an HBufC8 containing the converted UTF8. */	

EXPORT_C HBufC8* CnvUtfConverter::ConvertFromUnicodeToUtf8L(const TDesC16& aUnicode)

 	{

	// If aUnicode is  Null string, return an empty HBufC

	if (aUnicode.Length() == 0)

		{

		HBufC8* hBuf8 = HBufC8::NewL(1);

		return hBuf8;

		}

	// Otherwise, convert and store result in a buffer, reallocating that buffer if needed.

	const TInt length = aUnicode.Length();

	const TInt bufsize = 100;

	TPtrC16 unicode (aUnicode);

	TBuf8<bufsize> buf;

	HBufC8* hBuf8 = HBufC8::NewLC(length);

	TPtr8 utf8 = hBuf8->Des();

	FOREVER

		{

		TInt unconverted = ConvertFromUnicodeToUtf8(buf, unicode);

		if( unconverted == EErrorIllFormedInput || unconverted < 0)

			User::Leave(KErrCorrupt);

		if (utf8.Length() + buf.Length() > utf8.MaxLength())

			{

			// Reallocate the hBuf8

			hBuf8 = hBuf8->ReAllocL(utf8.Length() + buf.Length());

			CleanupStack::Pop();

			CleanupStack::PushL(hBuf8);

			utf8.Set(hBuf8->Des());

			}

		utf8.Append(buf);

		if (unconverted ==0) 

			break;

		unicode.Set(unicode.Right(unconverted));

		}

	CleanupStack::Pop();

	return hBuf8;

	}

/** Converts Unicode text into UTF-8 encoding. 

Surrogate pairs can be input which will result in a valid 4 byte UTF-8 value.

The variant of UTF-8 used internally by Java differs slightly from standard 

UTF-8. The TBool argument controls the UTF-8 variant generated by this function.

@param aUtf8 On return, contains the UTF-8 encoded output string.

@param aUnicode A UCS-2 encoded input string.

@param aGenerateJavaConformantUtf8 EFalse for orthodox UTF-8. ETrue for Java 

UTF-8. The default is EFalse.

@return The number of unconverted characters left at the end of the input descriptor, 

or one of the error values defined in TError. */

TInt CnvUtfConverter::ConvertFromUnicodeToUtf8(TDes8& aUtf8, 

											   const TDesC16& aUnicode, 

											   TBool aGenerateJavaConformantUtf8)

	{

	if (aUnicode.Length() == 0)

		{

		aUtf8.SetLength(0);

		return 0;

		}

	if (aUtf8.MaxLength() == 0)

		{

		return aUnicode.Length();

		}

	TUint8* pUtf8 = CONST_CAST(TUint8*, aUtf8.Ptr());

	const TUint8* pointerToLastUtf8Byte = pUtf8 + (aUtf8.MaxLength() - 1);

	TBool inputIsTruncated = EFalse;

	const TUint16* pUnicode = aUnicode.Ptr();

	const TUint16* pointerToLastUnicodeCharacter = pUnicode + (aUnicode.Length() - 1);

	FOREVER

		{

		__ASSERT_DEBUG(pUtf8 <= pointerToLastUtf8Byte, Panic(EPanicBadUtf8Pointers1));

		__ASSERT_DEBUG(pUnicode <= pointerToLastUnicodeCharacter, Panic(EPanicBadUnicodePointers3));

		if (pUnicode[0] < 0x80)

			{

			// ascii - 1 byte

			// internally java is different since the \x0000 character is 

			// translated into \xC0 \x80.

			if ((aGenerateJavaConformantUtf8) && (pUnicode[0] == 0x0000))

				{

				if (pUtf8 == pointerToLastUtf8Byte)

					{

					pUtf8--;