diff -r 000000000000 -r 1fb32624e06b textrendering/texthandling/stext/TXTSCAN.CPP
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/textrendering/texthandling/stext/TXTSCAN.CPP	Tue Feb 02 02:02:46 2010 +0200
@@ -0,0 +1,441 @@
+/*
+* Copyright (c) 2003-2009 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 "TXTETEXT.H"
+#include "TXTSTD.H"
+
+
+EXPORT_C TInt CPlainText::CharPosOfParagraph(TInt& aLength,TInt aParaOffset)const
+/** Finds the length and the start position of a paragraph identified by its 
+paragraph number the first paragraph is numbered zero.
+
+Notes:
+
+if aParaOffset is invalid, (equal to or greater than the total number of 
+paragraphs), the function's return value is EScanEndOfData
+
+@param aLength On return contains the length of the specified paragraph. 
+@param aParaOffset The paragraph number. The first paragraph is numbered zero. 
+@return The document position of the first character in the paragraph. */
+	{
+	__TEST_INVARIANT;
+
+	TInt startPos=aLength=0;
+	TUint scanMask=(EScanToUnitStart);
+	aLength=ScanParas(startPos,scanMask);
+	for (TInt offset=1;offset<=aParaOffset;offset++)
+		{
+		if (startPos<=EScanEndOfData)
+			return EScanEndOfData;
+		aLength=ScanParas(startPos,scanMask);
+		}
+	if (startPos==EScanEndOfData)
+		startPos=iByteStore->Size()/sizeof(TText);
+	return startPos-aLength;
+	}
+
+EXPORT_C TInt CPlainText::ParagraphNumberForPos(TInt& aPos)const
+/** Gets the number of the paragraph which contains a document position. 
+Paragraph numbering begins at zero.
+
+@param aPos A document position. Must be valid or a panic occurs. On return, 
+contains the document position of the first character in the paragraph in 
+which it is located. 
+@return The number of the paragraph containing the specified document position.
+The first paragraph is numbered zero. */
+	{
+	__TEST_INVARIANT;
+
+	__ASSERT_ALWAYS(aPos>=0 && aPos<=DocumentLength(),Panic(ECharPosBeyondDocument));
+	TUint scanMask=(EScanBackwards|EScanStayIfBoundary|EScanToUnitStart);
+	ScanParas(aPos,scanMask);
+	// aPos now holds the 1st char of the paragraph.
+	TInt paraOffset=0,currentPos=0;
+	scanMask=EScanToUnitStart;
+	while (currentPos<aPos)
+		{// Move forwards, counting paragraphs until we get to the current position.
+		paraOffset++;
+		ScanParas(currentPos,scanMask);
+		}
+	return paraOffset;
+	}
+
+EXPORT_C void CPlainText::GetWordInfo(TInt aCurrentPos,TInt& aStartPos,TInt& aLength,
+									  TBool aPictureIsDelimiter,TBool aPunctuationIsDelimiter)const
+/** Gets the start position and length of the word that contains the document 
+position specified.
+
+@param aCurrentPos A document position. Must be valid or a panic occurs. 
+@param aStartPos On return, the document position of the first character in 
+the word containing the position aCurrentPos.
+@param aLength On return, the length of the word containing document position 
+aCurrentPos.
+@param aPictureIsDelimiter ETrue if picture characters should be treated as 
+word delimiters, false if not. For example, when navigating text, this might 
+be EFalse, but when spell checking, it might be ETrue. 
+@param aPunctuationIsDelimiter ETrue if punctuation characters should be treated 
+as word delimiters, EFalse if not. */
+	{
+	__TEST_INVARIANT;
+
+	__ASSERT_ALWAYS(aCurrentPos>=0 && aCurrentPos<=DocumentLength(),Panic(ECharPosBeyondDocument));
+	aStartPos=aLength=aCurrentPos;
+	// First find the start of the word.
+	TUint scanMask=(EScanBackwards|
+					EScanStayIfBoundary|
+					EScanToUnitStart);
+	if (aPictureIsDelimiter)
+		scanMask|=EScanPictureIsDelimiter;
+	if (aPunctuationIsDelimiter)
+		scanMask|=EScanPunctuationIsDelimiter;
+	ScanWords(aStartPos,scanMask);
+	// Next find the end of the word.
+	scanMask=(EScanStayIfBoundary|EScanToUnitEnd);
+	if (aPictureIsDelimiter)
+		scanMask|=EScanPictureIsDelimiter;
+	if (aPunctuationIsDelimiter)
+		scanMask|=EScanPunctuationIsDelimiter;
+	ScanWords(aLength,scanMask);
+	aLength-=aStartPos;
+	__ASSERT_DEBUG(aLength+aStartPos<=(DocumentLength()+1),Panic(ECharPosBeyondDocument));
+
+	__TEST_INVARIANT;
+	}
+
+EXPORT_C TInt CPlainText::ToParagraphStart(TInt& aPos) const
+/** Gets the document position of the start of the paragraph containing the 
+specified document position.
+
+@param aPos A document position. Must be valid or a panic occurs. On return, 
+contains the document position of the first character in the paragraph in 
+which it is located.
+@return The number of characters skipped in scanning to the start of the 
+paragraph. */
+	{
+	__TEST_INVARIANT;
+	__ASSERT_ALWAYS(aPos>=0 && aPos<=DocumentLength(),Panic(ECharPosBeyondDocument));
+
+	int skipped = 0;
+	while (aPos > 0)
+		{
+		TPtrC8 p = iByteStore->BackPtr(aPos * sizeof(TText));
+		const TText* start = (TText*)p.Ptr();
+		const TText* end = start + p.Length() / sizeof(TText);
+		while (end > start)
+			{
+			end--;
+			if (*end == EParagraphDelimiter)
+				return skipped;
+			aPos--;
+			skipped++;
+			}
+		}
+	return skipped;
+	}
+
+EXPORT_C TInt CPlainText::WordCount()const
+/** Gets a count of the number of words in the text object.
+
+@return The number of words in the text object. */
+	{return CountUnits(EUnitIsWord);}
+
+EXPORT_C TInt CPlainText::ParagraphCount()const
+/** Gets a count of the number of paragraphs in the text object.
+
+Note: the paragraph delimiter which terminates every text object means this 
+function always returns a count of at least one.
+
+@return The number of paragraphs contained in the text object. */
+	{//return CountUnits(EUnitIsParagraph);}
+	TInt pos=0,unitCount=0;
+	TUint scanMask=(EScanToUnitStart);
+	while (pos!=EScanEndOfData)
+		{
+		unitCount++;
+		ScanParas(pos,scanMask);
+		}
+	return unitCount;
+	}
+
+TInt CPlainText::CountUnits(TUnitOfText aContext)const
+// Returns the number of units in the document, where
+// units are defined by the parameter aContext.
+//
+	{
+	SScanData scanData;
+	TInt pos=0,unitCount=0;
+	TUint scanMask=(EScanToUnitStart|EScanJoinDelimiters|EScanStayIfBoundary);
+	InitScanControl(pos,scanMask,aContext,scanData);
+	ScanUnit(pos,aContext,scanData);
+	// The above code skips over leading white space
+	scanData.scanMask &=~ EScanStayIfBoundary;
+	while(pos!=EScanEndOfData)
+		{
+		unitCount++;
+		ScanUnit(pos,aContext,scanData);
+		}//	Count complete, so return pos to the beginning of the data
+	pos=0;
+	return unitCount;
+	}
+
+
+
+
+EXPORT_C TInt CPlainText::ScanWords(TInt& aPos,TUint& aScanMask)const
+/** Scans from a document position to the beginning or end of a word. The 
+destination is determined by a scan mask. The scan can either be forwards 
+(the default) or backwards, and the destination may be the first or last 
+character in the word containing the position, or the first character in 
+the next word.
+
+Note: If an attempt is made to scan beyond the end of text delimiter, on return, 
+aPos is set to EScanEndOfData  and the function's return value indicates 
+the number of characters skipped in passing the end of text delimiter.
+
+@param aPos The document position from which to scan. Must be valid, or a 
+panic occurs. On return, contains the new document position. 
+@param aScanMask The scan mask. See the enumeration whose values begin with 
+CPlainText::EScanBackwards. 
+@return The number of characters skipped to reach the new document position. */
+	{
+	__ASSERT_ALWAYS(aPos>=0 && aPos<=DocumentLength(),Panic(ECharPosBeyondDocument));
+	SScanData scanData;
+	InitScanControl(aPos,aScanMask,EUnitIsWord,scanData);
+	return ScanUnit(aPos,EUnitIsWord,scanData);
+	}
+
+EXPORT_C TInt CPlainText::ScanParas(TInt& aPos,TUint& aScanMask)const
+/** Scans from a document position to the beginning or end of a paragraph. The 
+destination is determined by a scan mask. The scan can either be forwards 
+(the default) or backwards, and the destination may be the first or last 
+character in the paragraph containing the position, or the first character in 
+the next paragraph. 
+
+Note: if an attempt is made to scan beyond the end of text delimiter, on return, 
+aPos is set to EScanEndOfData ) and the function's return value indicates 
+the number of characters skipped in passing the end of text delimiter.
+
+@param aPos The document position to scan from. Must be valid or a panic occurs. 
+On return, contains the new document position. 
+@param aScanMask The scan mask. See the enumeration whose values begin with 
+CPlainText::EScanBackwards. 
+@return The number of characters skipped to reach the new document position. */
+	{
+	__ASSERT_ALWAYS(aPos>=0 && aPos<=DocumentLength(),Panic(ECharPosBeyondDocument));
+	SScanData scanData;
+	InitScanControl(aPos,aScanMask,EUnitIsParagraph,scanData);
+	return ScanUnit(aPos,EUnitIsParagraph,scanData);
+	}
+
+
+void CPlainText::InitScanControl(TInt& aPos,TUint& aScanMask,TUnitOfText aContext,SScanData& aScanData)const
+// Initialises the control segment of the bitmask
+// used to determine when scan has completed.
+// Calls the initialisation of aScanData.
+//
+	{
+	aScanData.pos=aPos;
+	aScanData.scanMask=0;
+	aScanData.scanMask|=aScanMask;
+	if (aScanData.scanMask & EScanStayIfBoundary)
+		{//Scan one unit the other way first.
+			aScanData.scanMask ^= EScanBackwards;  // Alter scan direction.
+			InitScanData(aScanData.pos,aScanData);				
+			TestForDelimiter(aContext,aScanData);
+			aScanData.scanMask ^= EScanBackwards;  // Restore original scan direction.
+		}
+	InitScanData(aScanData.pos,aScanData);
+	if(!aPos && (aScanData.scanMask&(EScanStayIfBoundary|EScanBackwards))==EScanStayIfBoundary)
+		{//At the start of the buffer and scanning forwards and ScanStayIfBoundary set.
+		aScanData.scanMask &=~aScanData.EInsideUnit;
+		}
+	else 
+		{
+		TestForDelimiter(aContext,aScanData);
+		if (aScanData.scanMask & aScanData.EIsDelimiter)
+			aScanData.scanMask &=~ aScanData.EInsideUnit;
+		else
+ 			aScanData.scanMask |= aScanData.EInsideUnit;
+		}
+	if (aScanData.scanMask & EScanBackwards)
+		{//Set stop flags relative to scan direction
+		if (aScanData.scanMask &  EScanToUnitStart)
+			aScanData.scanMask |= aScanData.EStopEnd;
+		if (aScanData.scanMask &  EScanToUnitEnd)
+			aScanData.scanMask |= aScanData.EStopBegin;
+		}
+	else
+		{//Set stop flags relative to scan direction
+		if (aScanData.scanMask &  EScanToUnitStart)
+			aScanData.scanMask |= aScanData.EStopBegin;
+		if (aScanData.scanMask &  EScanToUnitEnd)
+			aScanData.scanMask |= aScanData.EStopEnd;
+		}
+	}
+
+
+void CPlainText::InitScanData(TInt aPos,SScanData& aScanData)const
+// Initialises the necessary elements of the scan structure, for ScanXxx 
+// methods.  These are used to track scan progress.
+//
+	{
+	aScanData.pos=aPos;
+	aScanData.oldPos=aPos;
+	aScanData.currentSegLen=0;
+	if (aScanData.scanMask & EScanBackwards)
+		{//Scanning backwards.
+		aScanData.delta=-1;
+		aScanData.totalBufLen=aPos;
+		}
+	else
+		{//Scanning forwards.
+		aScanData.delta=+1;
+		aScanData.totalBufLen=((iByteStore->Size()/sizeof(TText))-aScanData.pos);
+		}
+	}	
+
+
+TInt CPlainText::ScanUnit(TInt& aPos,TUnitOfText aContext,SScanData& aScanData)const
+// Scan from position aPos, by one unit.
+// Update aPos to the current position, and return the number
+// of characters skipped or EScanEndOfData if at the end of the buffer.
+//
+	{
+	while (TestForDelimiter(aContext,aScanData))
+		{//More data has been read
+		if (aScanData.scanMask & aScanData.EInsideUnit)
+			{
+			if (aScanData.scanMask & aScanData.EIsDelimiter)
+				{
+				aScanData.scanMask &=~aScanData.EInsideUnit;
+				if (aScanData.scanMask & aScanData.EStopEnd)
+					break;
+				}
+			 }
+		else
+			{
+			if (aScanData.scanMask & aScanData.EIsDelimiter)
+				{
+				if (!(aScanData.scanMask & EScanJoinDelimiters))
+					break;
+				}
+	   		else
+				{
+				aScanData.scanMask |= aScanData.EInsideUnit;
+				if (aScanData.scanMask & aScanData.EStopBegin)
+					break;
+				}
+			}
+		}
+	TInt charsSkipped=(aScanData.scanMask & EScanBackwards)? aPos-aScanData.oldPos : 
+										aScanData.oldPos-aPos;
+	aPos=(aScanData.oldPos<(TInt)(iByteStore->Size()/sizeof(TText)))? aScanData.oldPos : EScanEndOfData;
+	return charsSkipped;
+	}
+
+
+TBool CPlainText::TestForDelimiter(TUnitOfText aContext,SScanData& aScanData) const
+/** Return ETrue if a character was read successfully.
+Set EIsDelimiter flag if the character read
+is a delimiter of the appropriate type.*/	
+    {
+	aScanData.scanMask |= aScanData.EIsDelimiter;
+	aScanData.oldPos = aScanData.pos;
+	TChar c;
+	if (!GetChar(aScanData,c))
+		return FALSE;
+
+	TBool found = FALSE;
+	if (aContext == EUnitIsParagraph)
+		found = c == EParagraphDelimiter;
+	else if (aContext == EUnitIsWord)
+		{
+		if (c == EParagraphDelimiter ||
+			c == ELineBreak ||
+			c == EPageBreak ||
+			c == ESpace ||
+			c == ETabCharacter)
+			found = TRUE;
+		if (!found && (aScanData.scanMask & EScanPictureIsDelimiter))
+			{
+			if (c == EPictureCharacter)
+				found = TRUE;
+			}
+		if (!found)
+			{
+			TChar::TCategory cat = c.GetCategory();
+			if (cat == TChar::EZsCategory) // it's a space
+				found = TRUE;
+			if (!found &&
+				(aScanData.scanMask & EScanPunctuationIsDelimiter) &&
+				(cat & TChar::EPunctuationGroup))
+				{
+				if (c != EHyphenMinus &&
+					c != EApostrophe &&
+					c != EPotentialHyphen &&
+					c != EHyphen && 
+					c != ERightSingleQuote)
+					found = TRUE;
+				}
+			}
+		}
+
+	if (found)
+		aScanData.scanMask |= aScanData.EIsDelimiter;
+	else
+		aScanData.scanMask &= ~aScanData.EIsDelimiter;
+	return TRUE;
+	}
+
+
+TBool CPlainText::GetChar(SScanData& aScanData,TChar& aChar)const
+// If there is one, write the next character to aChar, and return ETrue,
+// otherwise return EFalse.
+//
+	{
+	TPtrC view(_S("a"));  // Dummy value cos no default constructor.
+	if (!aScanData.currentSegLen)	
+		{//First time in, or new segment.
+		if (!aScanData.totalBufLen)
+			{//No data left to read.
+			return EFalse;	
+			}
+		if (aScanData.delta < 0)
+			{//Scanning backwards
+			TPtrC8 tempView=iByteStore->BackPtr(aScanData.pos*sizeof(TText));
+			view.Set((TText*)tempView.Ptr(),tempView.Length()/sizeof(TText));
+			}
+		else
+			{//Scanning forwards
+			TPtrC8 tempView=iByteStore->Ptr(aScanData.pos*sizeof(TText));
+			view.Set((TText*)tempView.Ptr(),tempView.Length()/sizeof(TText));
+			}//Ptr now returns a TDes.
+		aScanData.currentSegLen=(TInt)Min(aScanData.totalBufLen,(TInt)view.Length());
+		aScanData.totalBufLen-=aScanData.currentSegLen;
+		aScanData.buf=(TText*)view.Ptr();
+		if (aScanData.delta < 0)
+			aScanData.buf+=aScanData.currentSegLen-1;
+		}
+	aChar=*(aScanData.buf);  // aChar takes the character just read.
+	aScanData.buf+=aScanData.delta;  // pText increments correctly for both ASCII and Unicode
+	aScanData.pos+=aScanData.delta;
+	aScanData.currentSegLen-=1;  // len holds the number of chars left, for ASCII and Unicode
+	return ETrue;
+	}
+
+