#include <stdlib.h>
#include <stdio.h>
#include "wv.h"

void wvReleaseCLX(CLX *clx)
	{
	U16 i;
	for(i=0;i<clx->grpprl_count;i++)
		wvFree(clx->grpprl[i]);
	wvFree(clx->grpprl);
	wvFree(clx->cbGrpprl);
	wvReleasePCD_PLCF(clx->pcd,clx->pos);
	}

void wvBuildCLXForSimple6(CLX *clx,FIB *fib)
	{
	wvInitCLX(clx);
    clx->nopcd=1;;

	clx->pcd = (PCD *) malloc(clx->nopcd * sizeof(PCD));
	clx->pos = (U32 *) malloc( (clx->nopcd+1) * sizeof(U32));
	
    clx->pos[0] = 0;
    clx->pos[1] = fib->ccpText;

	wvInitPCD(&(clx->pcd[0]));			
	clx->pcd[0].fc = fib->fcMin;

	/* reverse the special encoding thing they do for word97 
	if we are using the usual 8 bit chars*/
	
	if (fib->fExtChar == 0)
		{
		clx->pcd[0].fc *= 2;
		clx->pcd[0].fc |= 0x40000000UL;
		}

	clx->pcd[0].prm.fComplex = 0;
	clx->pcd[0].prm.para.var1.isprm = 0;
	/*
	these set the ones that *I* use correctly, but may break for other wv
	users, though i doubt it, im just marking a possible firepoint for the
	future
	*/
	}
/*
The complex part of a file (CLX) is composed of a number of variable-sized
blocks of data. Recorded first are any grpprls that may be referenced by the
plcfpcd (if the plcfpcd has no grpprl references, no grpprls will be
recorded) followed by the plcfpcd. Each block in the complex part is
prefaced by a clxt (clx type), which is a 1-byte code, either 1 (meaning the
block contains a grpprl) or 2 (meaning this is the plcfpcd). A clxtGrpprl
(1) is followed by a 2-byte cb which is the count of bytes of the grpprl. A
clxtPlcfpcd (2) is followed by a 4-byte lcb which is the count of bytes of
the piece table. A full saved file will have no clxtGrpprl's.
*/
void wvGetCLX(version ver,CLX *clx,U32 offset,U32 len,U8 fExtChar,FILE *fd)
	{
	U8 clxt;
	U16 cb;
	U32 lcb,i,j=0;

	wvTrace(("offset %x len %d\n",offset,len));
	fseek(fd,offset,SEEK_SET);

	wvInitCLX(clx);

	while (j<len)
		{
		clxt = getc(fd);
		j++;
		if (clxt == 1)
			{
			cb = read_16ubit(fd);
			j+=2;
			clx->grpprl_count++;
			clx->cbGrpprl = (U16 *)realloc(clx->cbGrpprl,sizeof(U16)*clx->grpprl_count);
			clx->cbGrpprl[clx->grpprl_count-1] = cb;
			clx->grpprl = (U8 **)realloc(clx->grpprl,sizeof(U8 *)*(clx->grpprl_count));
			clx->grpprl[clx->grpprl_count-1] = (U8 *)malloc(cb);
			for(i=0;i<cb;i++)
				clx->grpprl[clx->grpprl_count-1][i] = getc(fd);
			j+=i;
			}
		else if (clxt == 2)
			{
			if (ver == WORD8)
				{
				lcb = read_32ubit(fd);
				j+=4;
				}
			else
				{
				wvTrace(("Here so far\n"));
#if 0
				lcb = read_16ubit(fd);		/* word 6 only has two bytes here */
				j+=2;
#endif

				lcb = read_32ubit(fd);		/* word 6 specs appeared to have lied ! */
				j+=4;
				}
			wvGetPCD_PLCF(&clx->pcd,&clx->pos,&clx->nopcd,ftell(fd),lcb,fd);
			j+=lcb;

			if (ver == WORD7)
				{
#if 0
				/* DANGER !!, this is a completely mad attempt to differenciate 
				between word 95 files that use 16 and 8 bit characters. It may
				not work, it attempt to err on the side of 8 bit characters.
				*/
				if (!(wvGuess16bit(clx->pcd,clx->pos,clx->nopcd)))
#else
				/* I think that this is the correct reason for this behaviour */
				if (fExtChar == 0)
#endif
					for (i=0;i<clx->nopcd;i++)
						{
						clx->pcd[i].fc *= 2;
						clx->pcd[i].fc |= 0x40000000UL;
						}
				}
			}
		else
			{
			wvError(("clxt is not 1 or 2, it is %d\n",clxt));
			return;
			}
		}
	}


void wvInitCLX(CLX *item)
	{
    item->pcd=NULL;
	item->pos=NULL;
	item->nopcd=0;

    item->grpprl_count=0;
    item->cbGrpprl=NULL;
    item->grpprl=NULL;
    }


int wvGetPieceBoundsFC(U32 *begin,U32 *end,CLX *clx,U32 piececount)
	{
    int type;
    if ( (piececount+1) > clx->nopcd)
		{
		wvTrace(("piececount is > nopcd, i.e.%d > %d\n", piececount+1,clx->nopcd));
        return(-1);
		}
    *begin = wvNormFC(clx->pcd[piececount].fc,&type);

	if (type)
	    *end = *begin+(clx->pos[piececount+1]-clx->pos[piececount]);
	else
	    *end = *begin+((clx->pos[piececount+1]-clx->pos[piececount])*2);

    return(type);
    }

int wvGetPieceBoundsCP(U32 *begin,U32 *end,CLX *clx,U32 piececount)
	{
    if ( (piececount+1) > clx->nopcd)
        return(-1);
    *begin = clx->pos[piececount];
    *end = clx->pos[piececount+1];
    return(0);
    }


char *wvAutoCharset(wvParseStruct *ps)
	{
	U16 i=0;
	int flag;
	char *ret;
	ret="iso-8859-15";

	/* 
	If any of the pieces use unicode then we have to assume the
	worst and use utf-8
	*/
	while(i<ps->clx.nopcd)
        {
		wvNormFC(ps->clx.pcd[i].fc,&flag);
		if (flag==0)
			{
			ret="UTF-8";
			break;
			}
        i++;
        }

	/* 
	Also if the document fib is not codepage 1252 we also have to 
	assume the worst 
	*/
	if (strcmp(ret,"UTF-8"))
		{
		if (
			(ps->fib.lid != 0x407) &&
			(ps->fib.lid != 0x807) &&
			(ps->fib.lid != 0x409) &&
			(ps->fib.lid != 0x807) &&
			(ps->fib.lid != 0xC09)
			)
		ret="UTF-8";
		}
	return(ret);
	}



int wvQuerySamePiece(U32 fcTest,CLX *clx,U32 piece)
	{
	/*
	wvTrace(("Same Piece, %x %x %x\n",fcTest,wvNormFC(clx->pcd[piece].fc,NULL),wvNormFC(clx->pcd[piece+1].fc,NULL)));
    if ( (fcTest >= wvNormFC(clx->pcd[piece].fc,NULL)) && (fcTest < wvNormFC(clx->pcd[piece+1].fc,NULL)) )
	*/
	wvTrace(("Same Piece, %x %x %x\n",fcTest,clx->pcd[piece].fc,wvGetEndFCPiece(piece,clx)));
    if ( (fcTest >= wvNormFC(clx->pcd[piece].fc,NULL)) && (fcTest < wvGetEndFCPiece(piece,clx)) )
		return(1);
    return(0);
	}


U32 wvGetPieceFromCP(U32 currentcp,CLX *clx)
	{
	U32 i=0;
	while(i<clx->nopcd)
		{
		wvTrace(("i %d: currentcp is %d, clx->pos[i] is %d, clx->pos[i+1] is %d\n",i,currentcp,clx->pos[i],clx->pos[i+1]));
		if ( (currentcp >= clx->pos[i]) && (currentcp < clx->pos[i+1]) )
			return(i);
		i++;
		}
	wvTrace(("cp was not in any piece ! \n",currentcp));
	return(0xffffffffL);
	}

U32 wvGetEndFCPiece(U32 piece,CLX *clx)
	{
	int flag;
	U32 fc;
	U32 offset = clx->pos[piece+1] - clx->pos[piece];

	wvTrace(("offset is %x, befc is %x\n",offset,clx->pcd[piece].fc));
	fc = wvNormFC(clx->pcd[piece].fc,&flag);
	wvTrace(("fc is %x, flag %d\n",fc,flag));
	if (flag) fc+=offset;
	else fc+=offset*2;
	wvTrace(("fc is finally %x\n",fc));
	return(fc);
	}

/*
1) search for the piece containing the character in the piece table.

2) Then calculate the FC in the file that stores the character from the piece
    table information.
*/
U32 wvConvertCPToFC(U32 currentcp,CLX *clx)
	{
	U32 currentfc=0xffffffffL;
	U32 i=0;
	int flag;

	while(i<clx->nopcd)
        {
        if ( (currentcp >= clx->pos[i]) && (currentcp < clx->pos[i+1]) )
            {
            currentfc = wvNormFC(clx->pcd[i].fc,&flag);
			if (flag) currentfc+=(currentcp-clx->pos[i]);
			else currentfc+=((currentcp-clx->pos[i])*2);
            break;
            }
        i++;
        }

	if (currentfc == 0xffffffffL)
		{
		i--;
		currentfc = wvNormFC(clx->pcd[i].fc,&flag);
		if (flag) currentfc+=(currentcp-clx->pos[i]);
		else currentfc+=((currentcp-clx->pos[i])*2);
		wvTrace(("flaky cp to fc conversion underway\n"));
		}

	return(currentfc);
	}

struct test
	{
	U32 fc;
	U32 offset;
	};

int compar(const void *a, const void *b)
	{
	struct test *one,*two;
	one = (struct test *)a;
	two = (struct test *)b;
	
	if (one->fc < two->fc)
		return(-1);
	else if (one->fc == two->fc)
		return(0);
	return(1);
	}

/* 
In word 95 files there is no flag attached to each
offset as there is in word 97 to tell you that we are
talking about 16 bit chars, so I attempt here to make
an educated guess based on overlapping offsets to
figure it out, If I had some actual information as
the how word 95 actually stores it it would help.
*/

int wvGuess16bit(PCD *pcd,U32 *pos,U32 nopcd)
	{
	struct test *fcs;
	U32 i;
	int ret=1;
	fcs = (struct test *)malloc(sizeof(struct test) * nopcd);
	for(i=0;i<nopcd;i++)
		{
		fcs[i].fc = pcd[i].fc;
		fcs[i].offset = (pos[i+1]-pos[i])*2;
		}

	qsort(fcs, nopcd, sizeof(struct test), compar);

	for(i=0;i<nopcd-1;i++)
		{
		if (fcs[i].fc+fcs[i].offset > fcs[i+1].fc)
			{
			wvTrace(("overlap, my guess is 8 bit\n"));
			ret = 0;
			break;
			}
		}

	wvFree(fcs);
	return(ret);
	}