/* -----------------------------------------------------------------------------
*
*  npcsv.c
*
*  ANTz - realtime 3D data visualization tools for the real-world, based on NPE.
*
*  ANTz is hosted at http://openantz.com and NPE at http://neuralphysics.org
*
*  Written in 2010-2016 by Shane Saxon - saxon@openantz.com
*
*  Please see main.c for a complete list of additional code contributors.
*
*  To the extent possible under law, the author(s) have dedicated all copyright 
*  and related and neighboring rights to this software to the public domain
*  worldwide. This software is distributed without any warranty.
*
*  Released under the CC0 license, which is GPL compatible.
*
*  You should have received a copy of the CC0 Public Domain Dedication along
*  with this software (license file named LICENSE.txt). If not, see
*  http://creativecommons.org/publicdomain/zero/1.0/
*
* --------------------------------------------------------------------------- */

#include "npcsv.h"


#include "../../os/npos.h"
#include "../npfile.h"

#include "../db/npdb.h"			//zz db2 added to access member elements

#include "../gl/nptags.h"
#include "../npgl.h"
#include "../../npctrl.h"
#include <stdio.h> // lv mac port

//-----------------------------------------------------------------------------
void npInitCSV (void* dataRef)
{
	return;
}

//-----------------------------------------------------------------------------
void npCloseCSV (FILE* filePtr, void* dataRef)
{
	fclose(filePtr);
	
	return;
}

//-----------------------------------------------------------------------------
// mode usage: 3 parameters "r/w/a" followed by optional "b" and optional "+"
// example mode = "rb+"
// r open file for reading
// w open file for overwriting. Create new file if doesn't already exist
// a open file for appending data. 
// b open file as a binary file
// + open file for read and writing
//-----------------------------------------------------------------------------
FILE* npOpenCSV (const char* fileName, const char* mode, void* dataRef)
{
	FILE* file = NULL;

	file = fopen(fileName, mode); 

	if ( file == NULL )
	{
		perror("Failed to open/create file");
		return NULL;
	}

	// add code to add the file ptr to the data->map							debug, zz

	return file;
}


//-----------------------------------------------------------------------------
int npReadCSV (void* readBuffer, int elementSize, int elementCount, 
				 FILE* file, void* dataRef)
{
	int bytes = 0;

	bytes = fread (readBuffer, elementSize, elementCount, file);
	
	if ( bytes == 0 )
		printf("Failed Reading\n");
	else
		printf("Success Reading\n");

	return bytes;
}


//-----------------------------------------------------------------------------
void npTestCSV (void* fileName, int wordSize, int size, void* dataRef)
{
	FILE *file;
	size_t count;
	char *str = (char*)fileName;
	int i = 0;


	char buffer[256];
/*
	// get the Current Working Directory
	npGetCWD (buffer, &size);

	printf("CWD path : ");
	for(i = 0; i < size; i++)
		printf ("%c", buffer[i]);

	printf("\n");
*/
	// open/create and write to the file, "w" overwrites existing files
	file = npOpenCSV (fileName, "r+", dataRef);

	// write the file path in the file as text
	count = npWriteCSV (buffer, 1, strlen(buffer), file, dataRef);

	printf("wrote %d bytes.\n", count); 

	// read the file
//	npFileRewind(file);

	size = npReadCSV (buffer, 1, 256, file, dataRef);


	printf("file contents : \n");
	for(i = 0; i < size; i++)
		printf ("%c", buffer[i]);

//mkdir(strB, S_IRWXU | S_IRWXG | S_IROTH | S_IXOTH);

}

//-----------------------------------------------------------------------------
FILE* npCSVDialog (const char* fileName, int dialogType, void* dataRef)
{	
	
	switch (dialogType)
	{
		case kNPfileDialogNew : ; break;

		case kNPfileDialogOpen : ; break;
		case kNPfileDialogClose : ; break;

		case kNPfileDialogSave : ; break;
		case kNPfileDialogSaveAs : ; break;

		case kNPfileDialogImport : ; break;
		case kNPfileDialogExport : ; break;

		default : break;
	}

	return NULL;
}


//-----------------------------------------------------------------------------
int npWriteCSV (const void* str, int wordSize, int size, FILE* file, void* dataRef)
{
	return fwrite( str, 1, strlen(str), file );
}



void npSaveCSV (FILE* file, void* dataRef)
{
	return;
}


//-----------------------------------------------------------------------------
void npImportCSV (FILE* file, int type, void* dataRef)
{
	return;
}

//-----------------------------------------------------------------------------
void npExportCSV (FILE* file, void* dataRef)
{
	return;
}

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

// if not saved then opens SaveAs dialog before reseting app
void npCloseFileCSV (int size, void* dataRef)
{
	return;
}

//-----------------------------------------------------------------------------
void npCSVGetAttrib (int size, void* dataRef)
{
	return;
}

//-----------------------------------------------------------------------------
void npCSVSetAttrib (int size, void* dataRef)
{
	return;
}

//-----------------------------------------------------------------------------
//passed an open State File and reads it in based on type
void npCSVtoMap (FILE* file, int type, void*dataRef)
{
	size_t count = 0;
	int i = 0;
	int byteCount = 0;
	char* bufferIndex = 0;
	
	pData data = (pData) dataRef;
	pNPnode node = NULL;


	char* buffer = (char*)malloc(kNPmapFileBufferMax);
	if (buffer == NULL)
	{
		printf ("err 4285 - malloc failed, cannot write file");
		return;
	}

	if( file != NULL )   // check to see if file exists
	{
		printf("Reading File...\n");
		count = npFileRead (buffer, 1, kNPmapFileBufferMax, file, dataRef);
		printf("Bytes Read: %d\n", count);

		npFileClose (file, dataRef);	//done reading, close file
		
		if (count > 0)
		{
			printf("Loading File\n");			//add table type handling, debug zz
			npLoadMapFile (buffer, 1, count, dataRef);	//process data
			printf ("Done\n\n");
		}
		else
			printf("err 4297 - zero bytes read\n");
	}
	else
		printf("err 4298 - cannot open file \n");


	free (buffer);

	return;
}

//zz debug, need a table similar to npmap.c generates for node header with ptrs like chmap
//need arrays for sub elements to, color = black, white...

//-----------------------------------------------------------------------------
char* npGetValueStr(char* dest, pNPmapLink item, void* dataRef) //*(int*)mapGlobals[id].mapPtr,
{
	//*(int*)mapGlobals[id].mapPtr,

	
	pNPubyteRGBA	ubyteRGBA = NULL;
	pNPfloatRGBA	floatRGBA = NULL;
	pNPfloatXYZ		floatXYZ = NULL;
	pNPfloatXY		floatXY = NULL;
	pNPintXYZ		intXYZ = NULL;
	pNPintXY		intXY = NULL;

	//zz debug, should this type be here???
	pNPgl			gl = NULL;

	//	_itoa(*(int*)item->mapPtr, dest, 0);	//zz debug optimize

	// sort order is map_path, np_osc, tag match, element match
	//type tags match
	//elements match
	//then elementB
//	switch (item->typeTagA)
//	{
//		case 
	switch (item->typeA)
	{
		//the constant is the functionID, plugins can use number or 'name'
		case kNPcmdFuncPtr : sprintf (dest, "%d", kNPcmdFuncPtr); break;

		case kNPfloat : sprintf (dest, "%f", *(float*)item->mapPtr); break;
		case kNPint : sprintf (dest, "%d", *(int*)item->mapPtr); break;
		case kNPcstrPtr : sprintf (dest, "%s", (char*)item->mapPtr); break;
		

		case kNPubyteRGBA :
			ubyteRGBA = item->mapPtr;
			sprintf(dest, "%d,%d,%d,%d", 
					ubyteRGBA->r,
					ubyteRGBA->g,
					ubyteRGBA->b,
					ubyteRGBA->a
					);
			break;

		case kNPfloatRGBA :
			floatRGBA = item->mapPtr;
			sprintf(dest, "%f,%f,%f,%f",
					floatRGBA->r,
					floatRGBA->g,
					floatRGBA->b,
					floatRGBA->a
					);
			break;

		case kNPfloatXYZ :
			floatXYZ = item->mapPtr;
			sprintf (dest, "%f,%f,%f", floatXYZ->x, floatXYZ->y, floatXYZ->z);
			break;

		case kNPintXYZ :
			intXYZ = item->mapPtr;
			sprintf (dest, "%d,%d,%d", intXYZ->x, intXYZ->y, intXYZ->z);
			break;

		case kNPfloatXY :
			floatXY = item->mapPtr;
			sprintf (dest, "%f,%f", floatXY->x, floatXY->y);
			break;

		case kNPintXY :
			intXY = item->mapPtr;
			sprintf (dest, "%d,%d", intXY->x, intXY->y);
			break;

	//zz debug probably should not be here this is the element not the type_id...
		case kNPpositionXY :
			gl = item->mapPtr;
			sprintf (dest, "%d,%d", gl->position.x, gl->position.y);
			break;

		case kNPwindowSizeXY :
			gl = item->mapPtr;
			sprintf (dest, "%d,%d", gl->windowSize.x, gl->windowSize.y);
			break;

		default : //report err
				strcpy (dest, "22");//"err 9999 - unknown pNPmapType item");
			break;
	}

	
	return  dest;
}


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

	// ISSUES 
	// best not to limit osc str length
	// need to get global ptr when populating list, or servicing it....
	// advantage to both runtime const to ptr conversion and pre-mapped
	// may change data pointer for struct member mapping pointer offsets
	// nodeMemberPtrOffset = (&node->translate - node)
	// (pNPfloatXYZ)node+nodeMemberPtrOffset

	// for example /np/node/42 iiiifff...i 42 1 1 7... 420
	// needs to get node 42 then populate, more of an offset method


//zz debug, this would seem to be a map_path_id not typeA?
//this is currently not called by anything
//npMapIDtoPtr
void* npMapElementIDtoPtr (int elementID, void* dataRef);
//-----------------------------------------------------------------------------
void* npMapElementIDtoPtr (int elementID, void* dataRef)
{
	int i = 0;
	int itemCount = 0;

	pData data = (pData) dataRef;

	pNPmapLink mapType = NULL;


	//get the map and itemCount
	//first check globals
	mapType = data->map.typeMapGlobals;
	itemCount = data->map.globalsCount;

	//scan map using the ID and if found return a pointer to the data
	for (i=0; i < itemCount; i++)
	{
		if (mapType[i].elementID == elementID)
			return mapType[i].mapPtr;
	}

	//then check osc
	mapType = data->map.typeMapOSC;
	itemCount = data->map.oscCount;

	//scan map using the ID and if found return a pointer to the data
	for (i=0; i < itemCount; i++)
	{
		if (mapType[i].elementID == elementID)
			return mapType[i].mapPtr;
	}

	return NULL;
}


char* npGetMapTypeName (int elementID, void* dataRef);
//-----------------------------------------------------------------------------
char* npGetMapTypeName (int elementID, void* dataRef)
{
	int i = 0;
	int itemCount = 0;

	pData data = (pData) dataRef;

	pNPmapLink mapType = NULL;


	//get the map and itemCount
	//first check globals
	mapType = data->map.typeMapGlobals;
	itemCount = data->map.globalsCount;

	//scan map using the ID and if found return a pointer to the data
	for (i=0; i < itemCount; i++)
	{
		if (mapType[i].elementID == elementID)
			return mapType[i].elementA;
	}

	//then check osc
	mapType = data->map.typeMapOSC;
	itemCount = data->map.oscCount;

	//scan map using the ID and if found return a pointer to the data
	for (i=0; i < itemCount; i++)
	{
		if (mapType[i].elementID == elementID)
			return mapType[i].elementA;
	}

	return NULL;
}


int npMapStrToID (const char* str, void* dataRef);
//-----------------------------------------------------------------------------
int npMapStrToID (const char* str, void* dataRef)
{
	int i = 0;
	int result = 0;
	int itemCount = 0;

	pData data = (pData) dataRef;

	pNPmapLink mapType = NULL;
	pNPmapLink mapItem = NULL;

	//first check globals
	mapType = data->map.typeMapGlobals;
	itemCount = data->map.globalsCount;

	//scan elementA lists and return the corresponding constant kNP... = 42
	for (i=0; i < itemCount; i++)
	{
		mapItem = &mapType[i];

		if (strncmp(str, (const char*)mapItem->elementA, kNPnameMax) == 0)
		{
			result = mapItem->elementID;
			break;
		}
	}
		
	//then check OSC map
	mapType = data->map.typeMapOSC;
	itemCount = data->map.oscCount;

	//scan elementA lists and return the corresponding constant kNP... = 42
	for (i=0; i < itemCount; i++)
	{
		mapItem = &mapType[i];

		if (strncmp(str, (const char*)mapItem->elementA, kNPnameMax) == 0)
		{
			result = mapItem->elementID;
			break;
		}
	}

	return result;
}

void* npMapStrToPtr (const char* str, void* dataRef);
//-----------------------------------------------------------------------------
void* npMapStrToPtr (const char* str, void* dataRef)
{
	//quick implementation... not efficient, make direct like others..
//	return nMapIDtoPtr ( npMapStrToID (str, dataRef), dataRef );

	int i = 0;
	int itemCount = 0;

	pData data = (pData) dataRef;

	pNPmapLink mapType = NULL;

	//get the map and itemCount
	mapType = data->map.typeMapGlobals;
	itemCount = data->map.globalsCount;

	//scan map using the ID and if found return a pointer to the data
	for (i=0; i < itemCount; i++)
	{
		if (strncmp(str, (const char*)mapType[i].elementA, kNPnameMax) == 0)
			return mapType[i].mapPtr;
	}

	//get the map and itemCount
	mapType = data->map.typeMapOSC;
	itemCount = data->map.oscCount;

	//scan map using the ID and if found return a pointer to the data
	for (i=0; i < itemCount; i++)
	{
		if (strncmp(str, (const char*)mapType[i].elementA, kNPnameMax) == 0)
			return mapType[i].mapPtr;
	}

	return NULL;
}


//upgrade to HASH method or similar
//-----------------------------------------------------------------------------
void* npMapElementToPtr (const char* str, void* dataRef)
{
	//quick implementation... not efficient, make direct like others..
//	return nMapIDtoPtr ( npMapStrToID (str, dataRef), dataRef );

	int i = 0;
	int itemCount = 0;

	pData data = (pData) dataRef;

	pNPmapLink mapType = NULL;

	//get the map and itemCount
	mapType = data->map.typeMapGlobals;
	itemCount = data->map.globalsCount;

	//scan map using the ID and if found return a pointer to the data
//	printf("\nElement globalsCount: %d\n", itemCount); // temp, lde
//	printf("\nstr : %s\n", str); // temp, lde
	if( !str)
	{
		printf(" err 4426 - npMapElementToPtr str is NULL\n");
		return NULL;
	}

	for( i=1; i < itemCount; i++) // Fixed, lde
	{
		//printf("%d ",i);
		//printf("\n(const char*)mapType[%d].elementA : %s\n", i, (const char*)mapType[i].elementA);
		if (strncmp(str, (const char*)mapType[i].elementA, kNPnameMax) == 0) 
		{
			//printf("\nFound It 5472\n");
			return mapType[i].mapPtr;
		}
	}
	
/*
	for( i=0; i < itemCount; i++ )		//zz debug, why i=1 and <= works
	{									//instead of i=0 and <
		if (strncmp(str, (const char*)mapType[i].elementA, kNPnameMax) == 0) // Fails here, lde @todo
			return mapType[i].mapPtr;
	}	
*/
	
	//get the map and itemCount
	mapType = data->map.typeMapOSC;
	itemCount = data->map.oscCount;
//	printf("itemCount: %i\n", itemCount);
	//scan map using the ID and if found return a pointer to the data
	for (i=0; i < itemCount; i++)
	{
		printf("%s\n", mapType[i].elementA); // lv mac port window bug
		if (strncmp(str, (const char*)mapType[i].elementA, kNPnameMax) == 0)
			return mapType[i].mapPtr;
	}

	return NULL;
}

//-----------------------------------------------------------------------------
char* npMapTypeName (int mapType, void* dataRef)
{
	//quick implementation... not efficient, make direct like others..
//	return nMapIDtoPtr ( npMapStrToID (str, dataRef), dataRef );

	int i = 0;
	int itemCount = 0;

	pData data = (pData) dataRef;

	pNPmapType mapTypeList = NULL;

	//get the map and itemCount
	mapTypeList = data->map.mapTypeList;
	itemCount = data->map.mapTypeCount;

	//scan map using the ID and if found return a pointer to the data
	for( i=0; i < itemCount; i++ )		//zz debug, why i=1 and <= works
	{									//instead of i=0 and <
		if( mapTypeList[i].type == mapType )
			return mapTypeList[i].name;
	}

	//if not found
	npPostMsg( " err 4427 - unknown mapType", kNPmsgErr, data );

	return NULL;
}

//-----------------------------------------------------------------------------
void* npMapAddressToPtr (const char* str, void* dataRef)
{
	//quick implementation... not efficient, make direct like others..
//	return nMapIDtoPtr ( npMapStrToID (str, dataRef), dataRef );

	int i = 0;
	int itemCount = 0;

	pData data = (pData) dataRef;

	pNPmapLink mapType = NULL;

	//get the map and itemCount
	mapType = data->map.typeMapGlobals;
	itemCount = data->map.globalsCount;

	//scan map using the ID and if found return a pointer to the data
	for (i=0; i < itemCount; i++)
	{
		if (strncmp(str, (const char*)mapType[i].elementB, kNPnameMax) == 0)
			return mapType[i].mapPtr;
	}

	//get the map and itemCount
	mapType = data->map.typeMapOSC;
	itemCount = data->map.oscCount;

	//scan map using the ID and if found return a pointer to the data
	for (i=0; i < itemCount; i++)
	{
		if (strncmp(str, (const char*)mapType[i].elementB, kNPnameMax) == 0)
			return mapType[i].mapPtr;
	}

	return NULL;
}


//-----------------------------------------------------------------------------
pNPmapLink npMapAddressToMapTypeItem (const char* str, void* dataRef)
{
	//quick implementation... not efficient, make direct like others..
//	return nMapIDtoPtr ( npMapStrToID (str, dataRef), dataRef );

	int i = 0;
	int count = 0;
	char* mapMatch = NULL;

	pData data = (pData) dataRef;

	pNPmapLink mapType = NULL;

	//get the map and itemCount
	mapType = data->map.typeMapGlobals;
	count = data->map.globalsCount;

	if( !str)
	{
		printf("err 4430 - npMapElementToPtr str is NULL\n");
		return NULL;
	}

	//scan map using the ID and if found return a pointer to the data
	for( i=0; i < count; i++)
	{
		mapMatch = mapType[i].elementB;
		if( !mapMatch)
			continue;
		//printf("elementB: %s\n", mapType[i].elementB);
		if (strncmp(str, (const char*)mapType[i].elementB, kNPnameMax) == 0)
			return &mapType[i];
	}

	//get the map and itemCount
	mapType = data->map.typeMapOSC;
	count = data->map.oscCount;

	//scan map using the ID and if found return a pointer to the data
	for( i=0; i < count; i++)
	{
		
		printf("OSC elementB: %s\n", mapType[i].elementB);
		mapMatch = mapType[i].elementB;
		if( !mapMatch)
			continue;
		if (strncmp(str, (const char*)mapType[i].elementB, kNPnameMax) == 0)
			return &mapType[i];
	}

	printf( "OSC Address: %s\n", str ); //zz osc debug

	return NULL;
}


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


int npMapToCSV (char* csvStr, int mapType, int size, int* index, void* dataRef);
//-----------------------------------------------------------------------------
int npMapToCSV (char* csvStr, int mapType, int size, int* index, void* dataRef)
{
	int id = 1;				//first id should start at 1
	int itemCount = 0;

	int n = 0;				//curs byte count offset
	char* curs = csvStr;

	char str[4096];			//temp memory for value string, optimize this out, //zz debug			

	int i = 0, j = 0;		//node tag sort stuff
	int count = 0;
	void** nodes;
			
    pNPgeolist geolist = NULL; // lv mac port models
    pNPtexmap texmap = NULL; // lv mac port models
    char* rel = NULL; // lv mac port models
    
	pNPmapLink mapList = NULL;
	pNPmapLink mapItem = NULL;

	pNPnode node = NULL;
	pNPtag tag = NULL;

	pData data = (pData) dataRef;

//zz-JJ		itemCount = data->map.globalsCount;

	if ( mapType == kNPmapOSC )
	{
		// write the first row header, includes CRLF line ending
		// ---
		// 'item_id' = record_id for nodes or window_id for GL window context... etc.
		//
		// 'map_id' = maps the data location... file, db table, list ptr
		// map_path connects MySQL, URL, OSC, files...
		// table names such as 'globals', 'node', 'gl', 'track', 'cmd'
		// custom table names... node_tbl_5... or MyCustomTableName...
		//example: 1,node,42,translate_xyz,/np/node/42/translate_xyz
		// ---
		mapList = data->map.typeMapOSC;
		itemCount = data->map.oscCount;

		// write first row header first field as table_name_id based on mapType
	//	includes CRLF line ending
		n += sprintf (curs, "np_osc_id");	//npGetMapName(mapType, dataRef); //zz debug
	}
	else if ( mapType == kNPmapGlobals )
	{
		mapList = data->map.typeMapGlobals;
		itemCount = data->map.globalsCount;
	
		n += sprintf (curs, "np_globals_id");
	}
	else if ( mapType == kNPmapTag )
	{
		//mapList = data->map.typeMapTag;
		itemCount = kNPnodeRootPin;	//data->map.tagsCount;					//zz debug
		
		//allocate a pair of destination buffers for the nodes and tags
		nodes = npMalloc(0, sizeof(pNPnode) * kNPnodeMax, data);			//data->map.nodeCount
		if (!nodes) return 0;

		/// @todo implement a hash table and/or balanced tree //zz
		//populate the sort lists, skip over null nodes		
		for (i = kNPnodeRootPin; i < kNPnodeMax; i++)	//data->map.nodeCount; i++)
			if (data->map.nodeID[i] != NULL)
				nodes[j++] = data->map.nodeID[i];

#define kNPtagWriteLimit 110000			//zz debug temp workaround, need file block procedure
		if (j > kNPtagWriteLimit )
			j = kNPtagWriteLimit;		//zz debug kNPtagWriteLimit

		//sort nodes and tags grouped first by table_id and second by record_id
	//	npSort(nodes, j, sizeof(void*), npCompareNodes, data);

		//zz debug change to np_tag_id,map_path,record_id,title,description
		n += sprintf (curs, "np_tag_id,record_id,table_id,title,description\n");
		//for (id = 1; id <= itemCount; id++)
		for (i=0; i < j; i++)
		{
			node = nodes[i];
		//	node = data->map.node[id];
			if ( node )
			{
				tag = node->tag;
				if( tag && node->recordID ) //zz debug, test w/o '&& node->recordID' to save blank tags
				{	
					n += sprintf ((curs + n), "%d,%lld,%d,\"%s\",\"%s\"\n",	//zz optimize with npitoa...
					id++,
					node->recordID,
					node->tableID,
					tag->title,
					tag->desc
					);
				}
			}
		}

		// cleanup					//zz debug, add handling for &index with file blocks 
		npFree( nodes, data );
		return n;
    } // lv mac port models start
    else if(mapType == kNPmapModels) {
    //printf("kNPmapModels\n");
    n += sprintf (curs, "np_geo_id,np_texture_id,type,center_x,center_y,center_z,rotate_x,rotate_y,rotate_z,scale_x,scale_y,scale_z,object_name,file_name,path\n");
    geolist = &data->io.gl.geolist[0];
    for( i=0; i < 2000; i++ )
    {
        geolist = &data->io.gl.geolist[i];
        if( geolist->geometryId != 0 )
        {
            //rel = npFilePathAbsToRel(geolist->modelPath, dataRef); // lv mac port models
            
            printf(" i : %d \n tex id : %d\n\n", i, geolist->textureId);
            n += sprintf((curs + n), "%d,%d,%d,%f,%f,%f,%f,%f,%f,%f,%f,%f,\"%s\",\"%s\",\"%s\"\n",
                         geolist->geometryId,
                         geolist->textureId,
                         0,
                         geolist->center.x,
                         geolist->center.y,
                         geolist->center.z,
                         geolist->rotate.x,
                         geolist->rotate.y,
                         geolist->rotate.z,
                         geolist->scale.x,
                         geolist->scale.y,
                         geolist->scale.z,
                         geolist->name,
                         geolist->modelFile,
                         //rel
                         //    geolist->modelPath // lv mac port
                         "/usr/global/models"
                         );
            free(rel);
        }
    }
    return n;
}
else if(mapType == kNPmapTextures)
{
    printf("-------------Saving kNPmapTextures------------\n");
    n += sprintf (curs, "np_texture_id,type,file_name,path\n");
    // kNPtexListMax 2000
    for( i = 1; i < 2000; i++ )
    {
        texmap = &data->io.gl.texmap[i];
        
        if(/*texmap->extTexId != 0 && */texmap->filename[0] != '\0')
        {
            rel = npFilePathAbsToRel(texmap->path, dataRef);
            if( (rel[strlen(rel)-1] == '\\') && (rel[strlen(rel)-2] == '\\') )
            {
                /// lv, remove extra slash
                rel[strlen(rel)-1] = '\0';
            }
            
            printf("i %d texture rel path : %s\n", i, rel);
            
            printf("7G tex ext id : %d\n", texmap->extTexId);
            printf("7G tex int id : %d\n", texmap->intTexId);
            
            n += sprintf((curs + n), "%d,%d,\"%s\",\"%s\"\n",
                         texmap->intTexId
                         ,
                         0,
                         texmap->filename,
                         "usr/global/images"
                         );
            
            free(rel);
        }
    }
    
    printf("kNPmapTextures : %s\n", curs);
    return n;
} // lv mac port models stop
	else
	{
		npPostMsg("err 4545 - unknown mapType", kNPmsgErr, data);
		return 0;
	}

	//append rest of the header row
	n += sprintf ((curs + n), ",map_path,item_id,element,type,permisions" );

	n += sprintf ((curs + n), ",name,desc,value\n" );

	for (id = 1; id <= itemCount; id++)
	{
		mapItem = &mapList[id];

		n += sprintf ((curs + n), "%d,\"%s\",%d,\"%s\",\"%s\",%d,\"%s\",\"%s\",\"%s\"\n",	//zz optimize with npitoa...
					id,

					mapItem->mapPathA,		//"node"	//mapItem->baseType	//zz-JJ this will be added to array. mapItem->baseType
					mapItem->itemA,			//list array index or 0 for single elementA
					mapItem->elementA,
					mapItem->typeTagA,
					mapItem->permissions,	//permissions +0 for full duplex, +1 = A to B, -1 = B to A
	
					mapItem->name,
					mapItem->desc,

					npGetValueStr( str, mapItem, data)	//formats parameter values as csv 
				 );
	}
/*	}
	else if ( mapType == kNPmapGlobals )
	{
	//		npMapTypeInit(data);
		mapTypePtr = data->map.typeMapGlobals;
		itemCount = data->map.globalsCount;

		//write the first row header, includes CRLF line ending
		n += sprintf (curs, "np_globals_id,a_map_path,a_item_id,a_element,a_type_id,b_map_path,b_item_id,b_element,b_type_tag,init_value,title,desc\n");
		
		//example: 1,gl,1,clear_rgba,"1.0,0.5,0.5,1.0","background color RGBA"
	
		//loop and write out all globals in the mapType[] list
		//populates the rows using the globals map descriptor array 
		for (id = 1; id <= itemCount; id++)
		{
	//zz-JJ		mapItem = &data->map.typeMapGlobals[id];
			mapItem = &mapTypePtr[id];

			n += sprintf ((curs + n), "%d,%s,%d,\"%s\",%s,%s,%d,\"%s\",%s,\"%s\",\"%s\",\"%s\"\n",	//zz optimize with npitoa...
						id,
						mapItem->mapPathA,	//"node"	//mapItem->baseType	//zz-JJ this will be added to array. mapItem->baseType
						mapItem->itemA,		//list index or 0 for single elementA
						mapItem->elementA,
						mapItem->typeTagA,		//this is a type by 'ID' rather then type tag by 'NAME'

						mapItem->mapPathB,
						mapItem->itemB,
						mapItem->elementB,
						mapItem->typeTagB,	//this one is by name

						npGetValueStr(str, mapItem, data),	//format value str

						"",					//title
						""					//desc
					 );
		}
	}
*/
	return n;
}

int npGetFormat( const char* file, const char* curs, int size, void* dataRef );
//-----------------------------------------------------------------------------
int npGetFormat( const char* file, const char* curs, int size, void* dataRef )
{
	int mapType = 0;
	//if *.csv then check for comma, line returns, etc... make sure it really is a csv...

	return kNPformatCSV;

	if (strncmp(curs, "map_id", 6) == 0)		// native CSV format
		mapType = kNPmapCSV;

	else if (strncmp(curs, "osc_id", 6) == 0)	//3rd party OSC handler
		mapType = kNPformatOSC;
	else if (strncmp(curs, "json_id", 7) == 0)
		mapType = kNPformatJSON;
	else if (strncmp(curs, "snmp_id", 7) == 0)
		mapType = kNPformatSNMP;
	else if (strncmp(curs, "xml_id", 6) == 0)
		mapType = kNPformatXML;
	else if (strncmp(curs, "kml_id", 6) == 0)
		mapType = kNPformatKML;

	//	default : break; //run heuristics on content samples...
	// try and figure out the data type, compare against a list of known types
	// using a folder/path(s) to a bunch of sample data file/types

	return mapType;
}

//non-native (alien) universal format translator
int npMapAuto( int format, const char* curs, int size, void* dataRef );
//------------------------------------------------------------------------------
int npMapAuto( int format, const char* curs, int size, void* dataRef )
{
	int result = 0;

	return result;		//zz debug, add 3rd party support CSV, JSON, DB...
}


//identifies map format based on first field text, native types are np_....._id
int npGetMapType( int format, const char* curs, int size, void* dataRef );
//------------------------------------------------------------------------------
int npGetMapType( int format, const char* curs, int size, void* dataRef )
{
	int mapType = 0;
	char msg[128];

	if( strncmp( curs, "np_", 3 ) != 0 )
		return npMapAuto(format, curs, size, dataRef );		//dispatch non-native format

	//advance cursor to compare next str component
	curs += 3;

	if (strncmp(curs, "_id", 3) == 0)
		mapType = kNPmapNP;				// "np_id" default handler, agnostic
	else if (strncmp(curs, "globals_id", 10) == 0)		//globals map
		mapType = kNPmapGlobals;
	else if (strncmp(curs, "node_id", 7) == 0)			//node map
		mapType = kNPmapNode;						
	else if (strncmp(curs, "tag_id", 6) == 0)			//tag map
		mapType = kNPmapTag;
	else if (strncmp(curs, "map_id", 6) == 0)		//describes other maps... native CSV format
		mapType = kNPmapCSV;
	else if (strncmp(curs, "osc_id", 6) == 0)		//3rd party OSC handler
		mapType = kNPmapOSC;
	else if (strncmp(curs, "json_id", 7) == 0)
		mapType = kNPmapJSON;
	else if (strncmp(curs, "snmp_id", 7) == 0)
		mapType = kNPmapSNMP;
	else if (strncmp(curs, "kml_id", 6) == 0)
		mapType = kNPmapKML;
	else if (strncmp(curs, "xml_id", 6) == 0)
		mapType = kNPmapXML;
	else
	{
		//zz, could attempt other detection methods and/or generic automapper
		sprintf(msg,"err 7822 - unknown native mapType: %.20s...",curs);
		npPostMsg(msg, kNPmsgErr, dataRef);
	}

	return mapType;
}

void* npMapItemPtr (char* mapPath, char* item, char* element, char* typeTag, void* dataRef);
void* npMapItemPtr (char* mapPath, char* item, char* element, char* typeTag, void* dataRef)
{
	return npMapElementToPtr (element, dataRef);
}

void npUpdateGlobals( void* dataRef );
/// update global states at startup and after loading datasets with globals
//-----------------------------------------------------------------------------
void npUpdateGlobals( void* dataRef )
{
	pData data = (pData) dataRef;

	pNPgl gl = &data->io.gl;

	/// Don't sync if this is the startup load, GL context does NOT yet exist!
    if ( data->ctrl.startup ){
        printf("this is startup, not setting window size at this time\n"); // lv mac port window bug
		return;
    }

	printf( "fullscreen: %d  pos_x: %d  pos_y: %d  width: %d  height: %d\n",
			gl->fullscreen, gl->position.x, gl->position.y,
			gl->windowSize.x, gl->windowSize.y );

	/// Toggle fullscreen mode if does not match current state
	//if ( glutGet( GLUT_FULL_SCREEN ) != gl->fullscreen ) // lv mac port
    if( glutGet(0x01FF) != gl->fullscreen)
	{
		gl->fullscreen = 1 - gl->fullscreen;	 // reverse fullscreen flag
		npCtrlCommand( kNPcmdFullscreen, data ); // command flips flag back
		
		return;		// exit to prevent resizing when in fullscreen mode
	}
	
	/// Update window size and position if changed
	if ( glutGet( GLUT_WINDOW_X ) != gl->position.x 
		|| glutGet( GLUT_WINDOW_Y ) != gl->position.y
		|| glutGet( GLUT_WINDOW_WIDTH ) != gl->windowSize.x
		|| glutGet( GLUT_WINDOW_HEIGHT ) != gl->windowSize.y )
	{

		gl->width = gl->windowSize.x;
		gl->height = gl->windowSize.y;

		npGLResizeScene(gl->windowSize.x, gl->windowSize.y);
			
		gl->width = gl->windowSize.x;
		gl->height = gl->windowSize.y;
		glutReshapeWindow (gl->windowSize.x, gl->windowSize.y);
		glutPositionWindow (gl->position.x, gl->position.y);
	}
}

/// @todo add parallel (threaded) file and database parsing routines
// file parsing tends to be CPU intensive, be mindful of performance
// the parsing routines are optimized for native antz structures
// max bandwidth is achieved through large block read/write such as 4MB or 8MB
// currently some systems perform well up to 12MB block size, others slow down
// high IOPS can only be achieved with good flow control and efficient parsing
//-----------------------------------------------------------------------------
int npOpenMapCSV (char* filePath, int mapType, void* dataRef)
{
	int mapFormat = kNPformatNull;		//kNPformatCSV
//	int mapType = kNPmapNull;					//zz debug, reconcile different definitions of mapType

	int result = 0;
	int eof = 0;
	int endPoint = 0;

	FILE* file = NULL;
	size_t count = 0;
	size_t size = 0;
	int i = 0;
	int byteCount = 0;
	char* bufferIndex = 0;
	char* buffer = NULL;		//file read buffer
	char* curs = NULL;

	char*	id			= NULL;	//record ID
	char*	permissions	= NULL;

	char*	mapPathA	= NULL;
	char*	itemA		= NULL;		//typically an integer
	char*	typeTagA	= NULL;
	char*	elementA	= NULL;

	char*	mapPathB	= NULL;
	char*	itemB		= NULL;
	char*	typeTagB	= NULL;
	char*	elementB	= NULL;

	char*	name		= NULL;
	char*	desc		= NULL;
	char*	valueStr	= NULL;

	void*	elementRef	= NULL;
	int		subElement = 0;

	char msg[512] = {'\0'};

	pData data = (pData) dataRef;

	// allocate read buffer
	buffer = npMalloc( 0, kNPfileBlockSize, data); if( !buffer )return 0;

	// open file in read only mode "r"
	file = npFileOpen( filePath, "r", dataRef ); if( !file ) goto finish;

	// read file then immediately close file to minimize file-open-lock time
	size = npFileRead( buffer, 1, kNPfileBlockSize, file, dataRef );
	
	// close file only if all contents have been read
	if (size < kNPfileBlockSize)
		npFileClose( file, dataRef );
			
	// exit if file is empty, otherwise process it in chunks
	if( !size ) goto finish;
		
//	printf( "Bytes Read: %d", size );

	//find the first field by skipping over any white space at beginning of file
	curs = npSkipWhiteSpace (buffer, size);
	if (!curs)
	{
		npPostMsg("err 8311 - file buffer is blank", kNPmsgErr, data);
		goto finish;
	}

	//id the table format and data type, native tables start with np_...._id
	mapFormat = npGetFormat( filePath, curs, size, data );
	mapType = npGetMapType( mapFormat, curs, size, data );
		
// npLoadMapFile (buffer, wordSize, count, dataRef);
//------------------------------------------------------------------------------
//zz tokenize fields into ptr structure and convert quotes " to cstring '/0'
/*
//	mapItemFuncPtr = npGetMapItemFuncPtr( mapFormat, mapType, data );

	void* mapItemFuncPtr = NULL;

	switch( mapFormat )
	{	case kNPformatCSV : 
			switch( mapType )
			{	case kNPmapGlobals : mapItemFuncPtr = npMapItemFromCSV(); break;
				case kNPmapNode : mapItemFuncPtr = npMapNodeFromCSV(); break;
				case kNPmapTag : mapItemFuncPtr = npMapTagFromCSV(); break;
				case kNPmapOSC : mapItemFuncPtr = npMapOscFromCSV(); break;
				default : mapItemFuncPtr = npMapItemFromCSV(); break;
			}
		case kNPformatDB : break;		//MySQL formatted
		case kNPformatOSC : break;	//this is an actual OSC binary stream...
		case kNPformatJSON : break;
		case kNPformatSNMP : break;
		case kNPformatXML : break;
		case kNPformatKML : break;
	}
*/
	//now ready to parse data
	//start processing items (CSV/table rows)

	//find the chunk endpoint of the last record	//zz debug, add handling for record size > chunk size
	endPoint = npLastEOL( buffer, size );
	if (!endPoint){ printf("err 8312 - no EOL in file chunk\n"); goto finish;}

/* //zz
		// set read ptr to beginning of first data row
	
		//zz debug, replace below with comma and slash aware getMapItemPtr() 
		// directly STREAM parse csv/json/osc... different separators ','',''',""
		//
		// tokenizing converts the comma to a NULL-terminated cstr
		// stream parsing has to detect various separators such as ""
		// both methods avoid the need to copy the string
		i=0; while( curs[i] != ',' ) i++;			//find the next comma

		if (i>0) curs[i-1] = '\0'; curs++;			// i=0 could cause bad ptr 
													// i-1 to remove ending quote ...",

		//then search the globals mapType and get the element data ptr
		intPtr = (int*)npMapItemPtr (NULL, NULL, curs, NULL, data);
*/	
	//goto start of first record, typically starts with its id
	count += i = npNextLineLimit(curs, endPoint - count); curs = &curs[i];

	while( endPoint > (int)count )
	{
		id = curs;															//id

		count += i = npNextField(curs); mapPathA	= curs = &curs[i];	//array or element ptr
        //printf("curs:%s\n", curs); // lv mac port
		count += i = npNextField(curs); itemA		= curs = &curs[i];	//array index int, 0 if element such as np_globals

		count += i = npNextField(curs); elementA	= curs = &curs[i];	//element ptr
		count += i = npNextField(curs); typeTagA	= curs = &curs[i];	//type_tag_a descriptor
		curs -= 2; *curs = '\0'; curs+=2; 	//zz debug, i-2 to convert ending quote ...", to cstr for element ptr

		count += i = npNextField(curs); permissions = curs = &curs[i];	//direction

		count += i = npNextField(curs); name		= curs = &curs[i];	//UI name
		count += i = npNextField(curs); desc		= curs = &curs[i];	//UI description

		count += i = npNextField(curs); valueStr	= curs = &curs[i];	//value

		//seek next record
		count += i = npNextLineLimit( curs, endPoint - count ); curs = &curs[i];

		//set the element value ptr, (at this time) type tags must match
		//zz add custom type transcoding and a generic defualt handler
		if(0)//strcmp( typeTagA, typeTagB ) )
		{
			printf("warn 8323 - type tags do not match, id: %d\n", id);
			// npAutoTranslate();
		}
		else
		{
		//switch (typeTagA)
		//     case 'ii'
		//			switch (typeTagB)
			elementA++; typeTagA++; valueStr++;
            
			printf("valueStr : %.32s", elementA, valueStr);
			elementRef = npMapItemPtr (NULL, NULL, elementA, NULL, data);
			if( !elementRef )
			{							//zz todo add a default handler
				if( (int)count < endPoint )
					printf("warn 8324 - unknown npMapItemPtr element\n");	 
			}
			else if( strncmp( typeTagA, "i", 1 ) == 0 ) 
			{
				while( *typeTagA++ == 'i' ) 
				{
                    printf("Integer\n"); // lv mac port O key debug
					(*(int*)elementRef) = npatoi(valueStr);
					//	printf ("subElement value: %d\n", *(int*)elementRef );
					printf (" %d", *(int*)elementRef );

					if (*typeTagA == 'i')	//goto next field if more then one element
					{
                        elementRef += sizeof(int*); // lv mac port window bug fix
						i = npNextField(valueStr); valueStr = &valueStr[i];
					}
				}
			}
			else if( strncmp( typeTagA, "f", 1 ) == 0 ) 
			{
                printf("Float\n"); // lv mac port O key debug
				while( *typeTagA++ == 'f' ) //loop processes each type tag
				{	
					*(float*)elementRef = npatof(valueStr);
					printf (" %f", *(float*)elementRef );
					
					if (*typeTagA == 'f')	//goto next field if more then one element
					{
					//	(float*)elementRef += 1; // lv mac port
                        elementRef += 1; // lv mac port
						i = npNextField( valueStr ); valueStr = &valueStr[i];	//zz debug, what if no more fields... mem violation?
					}
				}
			}
			else if( strncmp( typeTagA, "s", 1 ) == 0 ) 
			{
                printf("String\n"); // lv mac port O key debug
				while( *typeTagA++ == 's' ) //loop processes each type tag
				{	
					//zz debug, how do we know if a char* or allocated char[array] ??? currently  arrays ONLY!!!!
					//need to (re)allocate when assigning to ptr
					
					//npCSVtoCstr (valueStr);
					i=0; while ( *(valueStr+i) != '\"' ) i++; *(valueStr+i) = '\0';
					strcpy( (char*)elementRef, valueStr );
                    printf("elementRef:%s\n", (char*)elementRef); // lv mac port window size loading problem
				//	(char*)elementRef = npNewStrcpy( valueStr, data );		//zz debug, need to handle free() and NULL etc...
					
					printf (" %.20s", (char*)elementRef );
					
					if (*typeTagA == 's')	//goto next field if more then one element
					{
					//	(char*)elementRef += 1; // lv mac port
                        elementRef += 1; // lv mac port
                      //  elementRef = (char*)elementRef + 1;
                      //  *(char*)elementRef += 1; // lv mac port
						i = npNextField( valueStr ); valueStr = &valueStr[i];	//zz debug, what if no more fields... mem violation?
					}
				}
//				printf( " strings not yet supports: %3s...\n", *(char*)elementRef );
			}
			else
				printf( "err 8333 - unknown type tag: %.20s\n", typeTagA );
		}
		printf("\n");

		result = 1;

	} //end loop

	if( mapType == kNPmapGlobals )
	{	
		printf("\n7382 kNPmapGlobals");
		npUpdateGlobals(dataRef);
		//printf("\nafter 7382");
	}

//		return result;


	//loop and write out all globals in the mapType[] list
	//populates the rows using the globals map descriptor array 
/*	for (id = 1; id <= itemCount; id++)
	{
		mapItem = &data->map.typeMapGlobals[id];

		n += sprintf ((curs + n), "%d,%s,\"%s\",\"%s\"\n",	//zz optimize with npitoa...
					id,
	//				type?
					mapItem->name,
					mapItem->typeTagB,
					npGetValueStr(str, mapItem, data)	//format value str			
				 );
	}
*/

	// 
	//similar to npCSVtoTag()
	//npLoadMapFile()

	//get the first row
	while( !eof )
	{
		//goto start of next row

		//id = npstrtoi(&cursor);
		//typeA = kNPgFullscreen
		//name =
		//typeTagB =	
		//valueStr =

		eof = true;
	}
//------------------------------------------------------------------------------

finish:

	npFree( buffer, data );

	if( msg )
	{
		npPostMsg( msg, kNPmsgFile, data );
	}

	return result;
}

/*
//-----------------------------------------------------------------------------
int npSaveGlobalsCSV (char* fileName, void* dataRef)
{
	char* path = NULL;
	FILE* fileRef = NULL;

	pData data = (pData) dataRef;

	csvStr = npMalloc (kNPfileBlock, kNPfileBlockSize, data);

	//if no fileName then assign default time-stamped name
	if(!fileName)
	{
		path = npMalloc(0, kNPmaxPath, data);
		strcpy (path, "antzglobals.csv");			//add time stamp and CWD	//zz debug
	}
	else
		path = fileName;

	//open the file for writing
	fileRef = npFileOpen ((const char*)path, "rw+", data);


	//get the first row header, includes CRLF line ending
	size = npGetGlobalsCSV (csvStr, data);

	npFileWrite ((const void*)csvStr, 1, size, fileRef, data);
	
	npFree(path);
	npFree(csvStr);

	return size;
}
*/


//-----------------------------------------------------------------------------
int npOpenGlobalsCSV (char* filePath, int wordSize, int size, void* dataRef)
{
	return npOpenMapCSV(filePath, kNPmapGlobals, dataRef);
}


// appends map table name to filePath and adds '.csv' extension
//-----------------------------------------------------------------------------
int npSaveMapToCSV( char* datasetName, int mapTypeID, void* dataRef )
{
	pData data = (pData) dataRef;

	int i = 0;
	int err = 0;
	int size = 0;
	int total = 0;
	int index = 0;		//same as rootIndex...
	int rootIndex = 0;
	size_t count = 0;
	FILE* file = NULL;

	char* path = data->io.file.csvPath;
	char* name = NULL;
	char* mapTypeName = NULL;
		
	char* buffer = NULL;
	char msg[kNPurlMax];
	char filePath[kNPurlMax];
	pNPnode node = NULL;


	//convert lookup table
	mapTypeName = npMapTypeName( mapTypeID, data );

	//construct the file path name for this map type using passed in dataset name
	sprintf( filePath, "%s%s%s%s", path, datasetName, mapTypeName, ".csv" );
	sprintf( msg, "Save: %s",  filePath);
	npPostMsg (msg, kNPmsgCtrl, data );

	if( mapTypeID == kNPmapNode )
		return npFileSaveMap( filePath, 1, strlen( filePath ), data );

	buffer = (char*) npMalloc( 0, kNPfileBlockSize, data);
	if( !buffer ) return 0;

	// open the file, "w+" overwrites existing files
//	printf( "\nSave File: %s\n", filePath);
	file = npFileOpen (filePath, "w+", dataRef);

	if (file == NULL)
	{
		printf("err 4215 - file open failed, cannot write file\n");							//debug, add err reporting, zz
		free (buffer);
		return 0;
	}
	
	// copies current state to the write buffer, formats as CSV
	// file buffer needs to be larger then kNPfileBlockSize
	printf("Writing...");
	count = npMapToCSV( buffer, mapTypeID, kNPmapFileBufferMax, &index, data );

	while (count > 0)
	{
		printf(".");
		total += count = npFileWrite( buffer, 1, count, file, data );
	
		if (count >= kNPfileBlockSize) //0)//zz debug file block methods //
		{
			printf("-");
			count = npMapToCSV( buffer, mapTypeID, kNPmapFileBufferMax, &index, data );
		}
		else
			count = 0;
	}
	printf("\n");

	// print partial file contents
	if (total > 0)
	{
		printf("Bytes Written: %d\n\n", total);
//		npFileRewind(file);
//		size = npFileRead (buffer, 1, 79, file, dataRef);
//		printf("File Contents:\n");
//		for( i = 0; i < size; i++ )
//			printf("%c", buffer[i]);
//		printf("\n");
	}
	else
		printf("err 4326 - file write failure, zero bytes written\n");

	err = npFileClose(file, dataRef);
	if(err)
		npPostMsg("err 4327 - file close failure", kNPmsgErr, data);

	free (buffer);

//	npPostMsg("Done writing CSV", kNPmsgCtrl, data);	//zz debug, ...writing Globals to...

	return total;
}

 
// appends map table name to filePath and adds '.csv' extension
//-----------------------------------------------------------------------------
int npSaveMapToCSV2( char* datasetName, int mapTypeID, void* dataRef )
{
	pData data = (pData) dataRef;

	int i = 0;
	int err = 0;
	int size = 0;
	int total = 0;
	int index = 0;		//same as rootIndex...
	int rootIndex = 0;
	size_t count = 0;
	FILE* file = NULL;

	char* path = data->io.file.csvPath;
	char* name = NULL;
	char* mapTypeName = NULL;
		
	char* buffer = NULL;
	char msg[kNPurlMax];
	char filePath[kNPurlMax];
	pNPnode node = NULL;


	//convert lookup table
	mapTypeName = npMapTypeName( mapTypeID, data );

	//construct the file path name for this map type using passed in dataset name
	sprintf( filePath, "%s/antz-%s%s", datasetName, mapTypeName, ".csv" );
	sprintf( msg, "Save: %s",  filePath);
	npPostMsg (msg, kNPmsgCtrl, data );

	if( mapTypeID == kNPmapNode )
		return npFileSaveMap( filePath, 1, strlen( filePath ), data );

	buffer = (char*) npMalloc( 0, kNPfileBlockSize, data);
	if( !buffer ) return 0;

	// open the file, "w+" overwrites existing files
//	printf( "\nSave File: %s\n", filePath);
	file = npFileOpen (filePath, "w+", dataRef);

	if (file == NULL)
	{
		printf("err 4215 - file open failed, cannot write file\n");							//debug, add err reporting, zz
		free (buffer);
		return 0;
	}
	
	// copies current state to the write buffer, formats as CSV
	// file buffer needs to be larger then kNPfileBlockSize
	printf("Writing...");
	count = npMapToCSV( buffer, mapTypeID, kNPmapFileBufferMax, &index, data );

	while (count > 0)
	{
		printf(".");
		total += count = npFileWrite( buffer, 1, count, file, data );
	
		if (count >= kNPfileBlockSize) //0)//zz debug file block methods //
		{
			printf("-");
			count = npMapToCSV( buffer, mapTypeID, kNPmapFileBufferMax, &index, data );
		}
		else
			count = 0;
	}
	printf("\n");

	// print partial file contents
	if (total > 0)
	{
		printf("Bytes Written: %d\n\n", total);
//		npFileRewind(file);
//		size = npFileRead (buffer, 1, 79, file, dataRef);
//		printf("File Contents:\n");
//		for( i = 0; i < size; i++ )
//			printf("%c", buffer[i]);
//		printf("\n");
	}
	else
		printf("err 4326 - file write failure, zero bytes written\n");

	err = npFileClose(file, dataRef);
	if(err)
		npPostMsg("err 4327 - file close failure", kNPmsgErr, data);

	free (buffer);

//	npPostMsg("Done writing CSV", kNPmsgCtrl, data);	//zz debug, ...writing Globals to...

	return total;
}