core/ASCOfficePDFReader/UnicodeMap.cpp

#include "stdafx.h"

#include <stdio.h>
#include <string.h>
#include "MemoryUtils.h"
#include "File.h"
#include "StringExt.h"
#include "List.h"
#include "GlobalParams.h"
#include "UnicodeMap.h"

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

#define maxExtCode 16

struct UnicodeMapExt 
{
	Unicode      nUnicode;
	char         arrCode[maxExtCode];
	unsigned int unBytesCount;
};

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

UnicodeMap *UnicodeMap::Parse(StringExt *seEncodingName, GlobalParams *pGlobalParams) 
{
	FILE *pFile = NULL;
	if ( pGlobalParams && !( pFile = pGlobalParams->GetUnicodeMapFile(seEncodingName) ) ) 
	{
		// TO DO: Error "Couldn't find UnicodeMap file for the encoding"
		return NULL;
	}

	UnicodeMap *pMap = new UnicodeMap( seEncodingName->Copy() );

	int nSize = 8;
	pMap->m_pRanges = (UnicodeMapRange *)MemUtilsMallocArray( nSize, sizeof(UnicodeMapRange) );
	int nEMapsSize = 0;

	int nLineNum = 1;
	char sBuffer[256];
	while ( GetLine( sBuffer, sizeof(sBuffer), pFile ) ) 
	{
		char *sToken1, *sToken2, *sToken3;
		if ( ( sToken1 = strtok(sBuffer, " \t\r\n" ) ) && ( sToken2 = strtok(NULL, " \t\r\n" ) ) ) 
		{
			if ( !( sToken3 = strtok(NULL, " \t\r\n" ) ) ) 
			{
				sToken3 = sToken2;
				sToken2 = sToken1;
			}
			int nBytesCount = strlen(sToken3) / 2;
			
			if ( nBytesCount <= 4 ) 
			{
				if ( pMap->m_nLen == nSize ) 
				{
					nSize *= 2;
					pMap->m_pRanges = (UnicodeMapRange *)MemUtilsReallocArray(pMap->m_pRanges, nSize, sizeof(UnicodeMapRange));
				}
				UnicodeMapRange *pRange = &pMap->m_pRanges[pMap->m_nLen];
				sscanf( sToken1, "%x", &pRange->nStart);
				sscanf( sToken2, "%x", &pRange->nEnd);
				sscanf( sToken3, "%x", &pRange->unCode);
				pRange->nBytesCount = nBytesCount;
				++pMap->m_nLen;
			} 
			else if ( sToken2 == sToken1 ) 
			{
				if ( pMap->m_nEMapsLen == nEMapsSize ) 
				{
					nEMapsSize += 16;
					pMap->m_pEMaps = (UnicodeMapExt *)MemUtilsReallocArray(pMap->m_pEMaps, nEMapsSize, sizeof(UnicodeMapExt));
				}
				UnicodeMapExt *pEMap = &pMap->m_pEMaps[pMap->m_nEMapsLen];
				sscanf( sToken1, "%x", &pEMap->nUnicode);
				for ( int nIndex = 0; nIndex < nBytesCount; ++nIndex ) 
				{
					int nValue = 0;
					sscanf( sToken3 + nIndex * 2, "%2x", &nValue);
					pEMap->arrCode[nIndex] = (char)nValue;
				}
				pEMap->unBytesCount = nBytesCount;
				++pMap->m_nEMapsLen;
			} 
			else 
			{
				// TO DO: Error "Bad line in unicodeMap file for encoding"
			}
		} 
		else 
		{
			// TO DO: Error "Bad line in unicodeMap file for encoding"
		}
		++nLineNum;
	}

	fclose(pFile);

	return pMap;
}

UnicodeMap::UnicodeMap(StringExt *seEncodingName) 
{
	m_seEncodingName = seEncodingName;
	m_bUnicodeOut    = FALSE;
	m_eType          = unicodeMapUser;

	m_pRanges = NULL;
	m_nLen    = 0;

	m_pEMaps    = NULL;
	m_nEMapsLen = 0;

	m_nRefCount = 1;

	InitializeCriticalSection( &m_oCS );
}

UnicodeMap::UnicodeMap(char *sEncodingName, BOOL bUnicodeOut, UnicodeMapRange *pRanges, int nLen) 
{
	m_seEncodingName = new StringExt(sEncodingName);
	m_bUnicodeOut    = bUnicodeOut;
	m_eType          = unicodeMapResident;

	m_pRanges = pRanges;
	m_nLen    = nLen;

	m_pEMaps    = NULL;
	m_nEMapsLen = 0;

	m_nRefCount = 1;

	InitializeCriticalSection( &m_oCS );
}

UnicodeMap::UnicodeMap(char *sEncodingName, BOOL bUnicodeOut, UnicodeMapFunc pFunction) 
{
	m_seEncodingName = new StringExt(sEncodingName);
	m_bUnicodeOut    = bUnicodeOut;
	m_eType          = unicodeMapFunc;

	m_pFunction = pFunction;

	m_pEMaps    = NULL;
	m_nEMapsLen = 0;
	
	m_nRefCount = 1;

	InitializeCriticalSection( &m_oCS );
}

UnicodeMap::~UnicodeMap() 
{
	if ( m_seEncodingName )
		delete m_seEncodingName;

	if ( m_eType == unicodeMapUser && m_pRanges ) 
	{
		MemUtilsFree( m_pRanges );
	}

	if ( m_pEMaps ) 
	{
		MemUtilsFree(m_pEMaps);
	}

	DeleteCriticalSection( &m_oCS );
}

void UnicodeMap::AddRef() 
{
	CTemporaryCS *pCS = new CTemporaryCS( &m_oCS );

	++m_nRefCount;

	RELEASEOBJECT( pCS );
}

void UnicodeMap::Release() 
{
	CTemporaryCS *pCS = new CTemporaryCS( &m_oCS );

	BOOL bDone = (--m_nRefCount == 0);

	RELEASEOBJECT( pCS );

	if ( bDone ) 
	{
		delete this;
	}
}

BOOL UnicodeMap::Match(StringExt *seEncodingName) 
{
	return ( !m_seEncodingName->Compare(seEncodingName) );
}

int UnicodeMap::MapUnicode(Unicode nUnicode, char *sBuffer, int nBufferSize) 
{
	if ( m_eType == unicodeMapFunc ) 
	{
		return (*m_pFunction)( nUnicode, sBuffer, nBufferSize);
	}

	int nFirst = 0;
	int nLast = m_nLen;
	if ( nUnicode >= m_pRanges[nFirst].nStart ) 
	{
		// m_pRanges[nFirst].nStart <= nUnicode < m_pRanges[nLast].nStart
		while ( nLast - nFirst > 1 ) 
		{
			int nMiddle = (nFirst + nLast) / 2;
			if ( nUnicode >= m_pRanges[nMiddle].nStart ) 
			{
				nFirst = nMiddle;
			} 
			else if ( nUnicode < m_pRanges[nMiddle].nStart ) 
			{
				nLast = nMiddle;
			}
		}

		if ( nUnicode <= m_pRanges[nFirst].nEnd ) 
		{
			int nLen = m_pRanges[nFirst].nBytesCount;
			if ( nLen > nBufferSize ) 
			{
				return 0;
			}
			unsigned int unCode = m_pRanges[nFirst].unCode + ( nUnicode - m_pRanges[nFirst].nStart );
			for ( int nIndex = nLen - 1; nIndex >= 0; --nIndex ) 
			{
				sBuffer[nIndex] = (char)(unCode & 0xff);
				unCode >>= 8;
			}
			return nLen;
		}
	}

	for ( int nIndex = 0; nIndex < m_nEMapsLen; ++nIndex ) 
	{
		if ( m_pEMaps[nIndex].nUnicode == nUnicode ) 
		{
			int nLen = m_pEMaps[nIndex].unBytesCount;
			for ( int nJ = 0; nJ < nLen; ++nJ ) 
			{
				sBuffer[nJ] = m_pEMaps[nIndex].arrCode[nJ];
			}
			return nLen;
		}
	}

	return 0;
}

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

UnicodeMapCache::UnicodeMapCache() 
{
	for ( int nIndex = 0; nIndex < unicodeMapCacheSize; ++nIndex )
	{
		m_ppCache[nIndex] = NULL;
	}
}

UnicodeMapCache::~UnicodeMapCache() 
{
	for ( int nIndex = 0; nIndex < unicodeMapCacheSize; ++nIndex ) 
	{
		if ( m_ppCache[nIndex] ) 
		{
			m_ppCache[nIndex]->Release();
		}
	}
}

UnicodeMap *UnicodeMapCache::GetUnicodeMap(StringExt *seEncodingName, GlobalParams *pGlobalParams) 
{
	if ( m_ppCache[0] && m_ppCache[0]->Match(seEncodingName) ) 
	{
		m_ppCache[0]->AddRef();
		return m_ppCache[0];
	}
	UnicodeMap *pMap = NULL;
	for ( int nIndex = 1; nIndex < unicodeMapCacheSize; ++nIndex ) 
	{
		if ( m_ppCache[nIndex] && m_ppCache[nIndex]->Match(seEncodingName) ) 
		{
			pMap = m_ppCache[nIndex];
			for ( int nJ = nIndex; nJ >= 1; --nJ ) 
			{
				m_ppCache[nJ] = m_ppCache[nJ - 1];
			}
			m_ppCache[0] = pMap;
			pMap->AddRef();
			return pMap;
		}
	}
	if ( ( pMap = UnicodeMap::Parse(seEncodingName, pGlobalParams) ) ) 
	{
		if ( m_ppCache[unicodeMapCacheSize - 1] ) 
		{
			m_ppCache[unicodeMapCacheSize - 1]->Release();
		}
		for ( int nJ = unicodeMapCacheSize - 1; nJ >= 1; --nJ ) 
		{
			m_ppCache[nJ] = m_ppCache[nJ - 1];
		}
		m_ppCache[0] = pMap;
		pMap->AddRef();
		return pMap;
	}
	return NULL;
}