/*--------------------------------------------------------------------------*/
int sci_fft_gen(void* _pvCtx, char *fname, int ndimsA, int *dimsA, double *Ar, double *Ai, int isn, int iopt, guru_dim_struct gdim)
{
/* API variables */
SciErr sciErr;
/* Input array variables */
int isrealA = (Ai == NULL), issymA = 1, lA = 1;
/*for MKL*/
int isrealA_save = isrealA ;
/*FFTW specific library variable */
enum Scaling scale = None;
enum Plan_Type type;
fftw_plan p;
/* input/output address for transform variables */
double *ri = NULL, *ii = NULL, *ro = NULL, *io = NULL;
/* for MKL special cases */
int * dims1 = NULL;
int * incr1 = NULL;
/* local variable */
int one = 1;
int i = 0;
int errflag = 0;
for (i = 0; i < ndimsA; i++)
{
lA *= dimsA[i];
}
if (iopt == 0)
{
/* automatically selected algorithm*/
issymA = check_array_symmetry(Ar, Ai, gdim);
if (issymA < 0 )
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
}
else if (iopt == 1)
{
issymA = 1; /* user forces symmetry */
}
else
{
issymA = 0;
}
AssignOutputVariable(_pvCtx, 1) = 1;/* assume inplace transform*/
if (WITHMKL)
{
double dzero = 0.0;
if (isrealA)
{
/*MKL does not implement the r2c nor r2r guru split methods, make A complex */
if (issymA)
{
/* result will be real, the imaginary part of A can be allocated alone */
sciErr = allocMatrixOfDouble(pvApiCtx, *getNbInputArgument(_pvCtx) + 1, 1, lA, &Ai);
if (sciErr.iErr)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
C2F(dset)(&lA, &dzero, Ai, &one);
}
else
{
/* result will be complex, realloc A for inplace computation */
sciErr = allocComplexArrayOfDouble(pvApiCtx, *getNbInputArgument(_pvCtx) + 1, ndimsA, dimsA, &ri, &Ai);
if (sciErr.iErr)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
C2F(dcopy)(&lA, Ar, &one, ri, &one);
Ar = ri;
C2F(dset)(&lA, &dzero, Ai, &one);
AssignOutputVariable(_pvCtx, 1) = nbInputArgument(_pvCtx) + 1;
isrealA = 0;
}
}
}
if (!isrealA && issymA) /* A is complex but result is real */
{
/* result will be complex, realloc real part of A for real part inplace computation */
sciErr = allocArrayOfDouble(pvApiCtx, *getNbInputArgument(_pvCtx) + 1, ndimsA, dimsA, &ri);
if (sciErr.iErr)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
C2F(dcopy)(&lA, Ar, &one, ri, &one);
Ar = ri;
AssignOutputVariable(pvApiCtx, 1) = nbInputArgument(_pvCtx) + 1;
}
/* Set pointers on real and imaginary part of the input */
ri = Ar;
ii = Ai;
scale = None; /*no scaling needed */
if (isn == FFTW_BACKWARD) scale = Divide;
if (isrealA & !WITHMKL) /* To have type = C2C_PLAN*/
{
/*A is real */
if (issymA)
{
/*r2r = isrealA && issymA*/
/* there is no general plan able to compute r2r transform so it is tranformed into
a R2c plan. The computed imaginary part will be zero*/
sciErr = allocMatrixOfDouble(pvApiCtx, *getNbInputArgument(_pvCtx) + 1, 1, lA, &io);
if (sciErr.iErr)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
type = R2C_PLAN;
ro = Ar;
}
else
{
/*r2c = isrealA && ~issymA;*/
/* transform cannot be done in place */
sciErr = allocComplexArrayOfDouble(pvApiCtx, *getNbInputArgument(_pvCtx) + 1, ndimsA, dimsA, &ro, &io);
if (sciErr.iErr)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
AssignOutputVariable(pvApiCtx, 1) = nbInputArgument(_pvCtx) + 1;
type = R2C_PLAN; /* fftw_plan_guru_split_dft_r2c plans for an FFTW_FORWARD transform*/
if (isn == FFTW_BACKWARD)
{
/*transform problem into a FORWARD fft*/
/*ifft(A)=conj(fft(A/N)) cas vect*/
/* pre traitement A must be divided by N cas vect*/
/* post treatment result must conjugated */
}
}
}
else
{
/* A is complex */
if (!WITHMKL && issymA) /*result is real*/
{
/*c2r = ~isrealA && issymA*/
ro = ri;
io = NULL;
type = C2R_PLAN; /*fftw_plan_guru_split_dft_c2r plans for an FFTW_BACKWARD transform*/
if (isn == FFTW_FORWARD)
{
/*transform problem into a BACKWARD fft : fft(A)=ifft(conj(A))*/
double minusone = -1.0;
C2F(dscal)(&lA, &minusone, ii, &one);
}
}
else
{
/*c2c = ~isrealA && ~issymA;*/
/* use inplace transform*/
isrealA = 0;
type = C2C_PLAN; /* fftw_plan_guru_split_dft plans for an FFTW_FORWARD transform*/
if (isn == FFTW_BACKWARD)
{
/*transform problem into a FORWARD fft*/
/* ifft(A) = %i*conj(fft(%i*conj(A)/N) */
/* reverse input */
ri = Ai;
ii = Ar;
/* reverse output */
ro = Ai;
io = Ar;
}
else
{
ro = ri;
io = ii;
}
}
}
/* pre-treatment */
if (scale != None)
{
double ak = 1.0;
for (i = 0; i < gdim.rank; i++) ak = ak * ((double)(gdim.dims[i].n));
if (scale == Divide) ak = 1.0 / ak;
C2F(dscal)(&lA, &ak, ri, &one);
if (isrealA == 0) C2F(dscal)(&lA, &ak, ii, &one);
}
if (!WITHMKL || gdim.howmany_rank <= 1)
{
/* Set Plan */
p = GetFFTWPlan(type, &gdim, ri, ii, ro, io, getCurrentFftwFlags(), isn , (fftw_r2r_kind *)NULL,&errflag);
if (errflag == 1)
{
Scierror(999, _("%s: No more memory.\n"), fname);
return 0;
}
else if (errflag == 2)
{
Scierror(999, _("%s: Creation of requested fftw plan failed.\n"), fname);
return 0;
}
/* execute FFTW plan */
ExecuteFFTWPlan(type, p, ri, ii, ro, io);
}
else
{
/*FFTW MKL does not implement yet guru plan with howmany_rank>1 */
/* associated loops described in gdim.howmany_rank and gdim.howmany_dims */
/* are implemented here by a set of call with howmany_rank==1 */
fftw_iodim *howmany_dims = gdim.howmany_dims;
int howmany_rank = gdim.howmany_rank;
int i1 = 0, i2 = 0;
int nloop = 0;
int t = 0;
gdim.howmany_rank = 0;
gdim.howmany_dims = NULL;
p = GetFFTWPlan(type, &gdim, ri, ii, ro, io, getCurrentFftwFlags(), isn , (fftw_r2r_kind *)NULL,&errflag);
if (errflag == 1)
{
Scierror(999, _("%s: No more memory.\n"), fname);
FREE(dims1);
FREE(incr1);
return 0;
}
else if (errflag == 2)
{
Scierror(999, _("%s: Creation of requested fftw plan failed.\n"), fname);
FREE(dims1);
FREE(incr1);
return 0;
}
/* flatten nested loops: replace howmany_rank nested loops by a single one*/
/* Build temporary arrays used by flatened loop */
if ((dims1 = (int *)MALLOC(sizeof(int) * howmany_rank)) == NULL)
{
Scierror(999, _("%s: No more memory.\n"), fname);
FREE(dims1);
FREE(incr1);
return 0;
}
dims1[0] = howmany_dims[0].n;
for (i = 1; i < howmany_rank; i++) dims1[i] = dims1[i - 1] * howmany_dims[i].n;
nloop = dims1[howmany_rank - 1];
if ((incr1 = (int *)MALLOC(sizeof(int) * howmany_rank)) == NULL)
{
Scierror(999, _("%s: No more memory.\n"), fname);
FREE(dims1);
FREE(incr1);
return 0;
}
t = 1;
for (i = 0; i < howmany_rank; i++)
{
t += (howmany_dims[i].n - 1) * howmany_dims[i].is;
incr1[i] = t;
}
/*loop on each "plan" */
i = 0; /*index on the first plan entry */
for (i1 = 1; i1 <= nloop; i1++)
{
/* the input and output are assumed to be complex because
within MKL real cases are transformed to complex ones in
previous steps of sci_fft_gen*/
ExecuteFFTWPlan(type, p, &ri[i], &ii[i], &ro[i], &io[i]);
i += howmany_dims[0].is;
/* check if a loop ends*/
for (i2 = howmany_rank - 2; i2 >= 0; i2--)
{
if ((i1 % dims1[i2]) == 0)
{
/*loop on dimension i2 ends, compute jump on the first plan entry index*/
i += howmany_dims[i2 + 1].is - incr1[i2];
break;
}
}
}
/* free temporary arrays */
FREE(dims1);
FREE(incr1);
/* reset initial value of gdim for post treatment*/
gdim.howmany_rank = howmany_rank;
gdim.howmany_dims = howmany_dims;
}
/* Post treatment */
switch (type)
{
case R2R_PLAN:
if (complete_array(ro, NULL, gdim) == -1)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
break;
case C2R_PLAN:
break;
case R2C_PLAN:
if (issymA)
{
/*R2C has been used to solve an r2r problem*/
if (complete_array(ro, NULL, gdim) == -1)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
}
else
{
if (complete_array(ro, io, gdim) == -1)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
if (isn == FFTW_BACKWARD)
{
/*conjugate result */
double ak = -1.0;
C2F(dscal)(&lA, &ak, io, &one);
}
}
break;
case C2C_PLAN:
if (WITHMKL && isrealA_save)
{
if (isn == FFTW_FORWARD)
{
if (complete_array(ro, io, gdim) == -1)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
}
else
{
if (complete_array(io, ro, gdim) == -1)
{
Scierror(999, _("%s: Cannot allocate more memory.\n"), fname);
return 0;
}
}
}
break;
}
return 1;
}
/*--------------------------------------------------------------------------*/
int sci_fftw_flags(char *fname, unsigned long fname_len)
{
/* declaration of variables to store scilab parameters address */
static int m1 = 0, n1 = 0;
char **Str1 = NULL;
char **Str3 = NULL;
unsigned int uiVar1 = 0;
int* piDataOut = NULL;
int* piAddr1 = NULL;
int* piLen = NULL;
int iType = 0;
/* please update me ! */
static int nb_flag = 22;
static char *Str[] =
{
/* documented flags */
"FFTW_MEASURE",
"FFTW_DESTROY_INPUT",
"FFTW_UNALIGNED",
"FFTW_CONSERVE_MEMORY",
"FFTW_EXHAUSTIVE",
"FFTW_PRESERVE_INPUT",
"FFTW_PATIENT",
"FFTW_ESTIMATE",
/* undocumented beyond-guru flags */
"FFTW_ESTIMATE_PATIENT",
"FFTW_BELIEVE_PCOST",
"FFTW_NO_DFT_R2HC",
"FFTW_NO_NONTHREADED",
"FFTW_NO_BUFFERING",
"FFTW_NO_INDIRECT_OP",
"FFTW_ALLOW_LARGE_GENERIC",
"FFTW_NO_RANK_SPLITS",
"FFTW_NO_VRANK_SPLITS",
"FFTW_NO_VRECURSE",
"FFTW_NO_SIMD",
"FFTW_NO_SLOW",
"FFTW_NO_FIXED_RADIX_LARGE_N",
"FFTW_ALLOW_PRUNING"
};
static unsigned flagt[] =
{
/* documented flags */
FFTW_MEASURE,
FFTW_DESTROY_INPUT,
FFTW_UNALIGNED,
FFTW_CONSERVE_MEMORY,
FFTW_EXHAUSTIVE,
FFTW_PRESERVE_INPUT,
FFTW_PATIENT,
FFTW_ESTIMATE,
/* undocumented beyond-guru flags */
FFTW_ESTIMATE_PATIENT,
FFTW_BELIEVE_PCOST,
FFTW_NO_DFT_R2HC,
FFTW_NO_NONTHREADED,
FFTW_NO_BUFFERING,
FFTW_NO_INDIRECT_OP,
FFTW_ALLOW_LARGE_GENERIC,
FFTW_NO_RANK_SPLITS,
FFTW_NO_VRANK_SPLITS,
FFTW_NO_VRECURSE,
FFTW_NO_SIMD,
FFTW_NO_SLOW,
FFTW_NO_FIXED_RADIX_LARGE_N,
FFTW_ALLOW_PRUNING
};
unsigned flagv = 0;
int i = 0, j = 0;
SciErr sciErr;
CheckInputArgument(pvApiCtx, 0, 1);
if (nbInputArgument(pvApiCtx) == 0)
{
// nothing
}
else
{
//get variable address of the input argument
sciErr = getVarAddressFromPosition(pvApiCtx, 1, &piAddr1);
if (sciErr.iErr)
{
printError(&sciErr, 0);
return 1;
}
getVarType(pvApiCtx, piAddr1, &iType);
switch (iType)
{
case sci_ints:
{
/* int */
int iPrecision = 0;
int* pi32Data = NULL;
unsigned int* pui32Data = NULL;
getMatrixOfIntegerPrecision(pvApiCtx, piAddr1, &iPrecision);
if (iPrecision != SCI_INT32 && iPrecision != SCI_UINT32)
{
Scierror(999, _("%s: Wrong type for input argument #%d: A int32 expected.\n"), fname, 1);
return 1;
}
if (iPrecision == SCI_INT32)
{
sciErr = getMatrixOfInteger32(pvApiCtx, piAddr1, &m1, &n1, pi32Data);
uiVar1 = (unsigned int)pi32Data[0];
}
else
{
sciErr = getMatrixOfUnsignedInteger32(pvApiCtx, piAddr1, &m1, &n1, pui32Data);
uiVar1 = pui32Data[0];
}
if (sciErr.iErr)
{
Scierror(999, _("%s: Can not read input argument #%d.\n"), fname, 1);
printError(&sciErr, 0);
return 1;
}
break;
}
case sci_matrix:
{
/* double */
double* pdblData = NULL;
sciErr = getMatrixOfDouble(pvApiCtx, piAddr1, &m1, &n1, &pdblData);
if (sciErr.iErr)
{
Scierror(999, _("%s: Can not read input argument #%d.\n"), fname, 1);
printError(&sciErr, 0);
return 1;
}
uiVar1 = (unsigned int)pdblData[0];
break;
}
case sci_strings:
{
/* string */
//fisrt call to retrieve dimensions
sciErr = getMatrixOfString(pvApiCtx, piAddr1, &m1, &n1, NULL, NULL);
if (sciErr.iErr)
{
printError(&sciErr, 0);
return 1;
}
piLen = (int*)malloc(sizeof(int) * m1 * n1);
//second call to retrieve length of each string
sciErr = getMatrixOfString(pvApiCtx, piAddr1, &m1, &n1, piLen, NULL);
if (sciErr.iErr)
{
printError(&sciErr, 0);
return 1;
}
Str1 = (char**)malloc(sizeof(char*) * m1 * n1);
for (i = 0 ; i < m1 * n1 ; i++)
{
Str1[i] = (char*)malloc(sizeof(char) * (piLen[i] + 1));//+ 1 for null termination
}
//third call to retrieve data
sciErr = getMatrixOfString(pvApiCtx, piAddr1, &m1, &n1, piLen, Str1);
if (sciErr.iErr)
{
printError(&sciErr, 0);
return 1;
}
for (j = 0; j < m1 * n1; j++)
{
for (i = 0; i < nb_flag; i++)
{
if (strcmp(Str1[j], Str[i]) == 0)
{
break;
}
}
if (i == nb_flag)
{
freeArrayOfString(Str1, m1 * n1);
Scierror(999, _("%s: Wrong values for input argument #%d: FFTW flag expected.\n"), fname, 1);
return 0;
}
else
{
if (i > 0)
{
flagv = ( flagv | (1U << (i - 1)) );
}
}
}
uiVar1 = (unsigned int)flagv;
freeArrayOfString(Str1, m1 * n1);
m1 = 1;
n1 = 1;
break;
}
default:
Scierror(53, _("%s: Wrong type for input argument #%d.\n"), fname, 1);
return 1;
}
CheckDims(1, m1, n1, 1, 1);
setCurrentFftwFlags(uiVar1);
}
/* return value of Sci_Plan.flags in position 2 */
sciErr = allocMatrixOfInteger32(pvApiCtx, nbInputArgument(pvApiCtx) + 2, 1, 1, &piDataOut);
if (sciErr.iErr)
{
printError(&sciErr, 0);
Scierror(999, _("%s: No more memory.\n"), fname);
return 1;
}
piDataOut[0] = (int) getCurrentFftwFlags();
/*Test for only FFTW_MEASURE*/
if (getCurrentFftwFlags() == 0)
{
j = 1;
if ((Str3 = (char **)MALLOC(sizeof(char *))) == NULL)
{
Scierror(999, _("%s: No more memory.\n"), fname);
return 1;
}
Str3[0] = strdup(Str[0]);
if (Str3[0] == NULL)
{
Scierror(999, _("%s: No more memory.\n"), fname);
return 1;
}
}
else
{
j = 0;
for (i = 1; i < nb_flag; i++)
{
if ((getCurrentFftwFlags() & flagt[i]) == flagt[i])
{
j++;
if (Str3)
{
Str3 = (char **)REALLOC(Str3, sizeof(char *) * j);
}
else
{
Str3 = (char **)MALLOC(sizeof(char *) * j);
}
if ( Str3 == NULL)
{
Scierror(999, _("%s: No more memory.\n"), fname);
return 1;
}
Str3[j - 1] = strdup(Str[i]);
if (Str3[j - 1] == NULL)
{
freeArrayOfString(Str3, j);
Scierror(999, _("%s: No more memory.\n"), fname);
return 1;
}
}
}
}
/* Create the string matrix as return of the function */
sciErr = createMatrixOfString(pvApiCtx, nbInputArgument(pvApiCtx) + 3, j, 1, Str3);
freeArrayOfString(Str3, j); // Data have been copied into Scilab memory
if (sciErr.iErr)
{
freeArrayOfString(Str3, j); // Make sure everything is cleanup in case of error
printError(&sciErr, 0);
return 1;
}
AssignOutputVariable(pvApiCtx, 1) = nbInputArgument(pvApiCtx) + 2;
AssignOutputVariable(pvApiCtx, 2) = nbInputArgument(pvApiCtx) + 3;
ReturnArguments(pvApiCtx);
return 0;
}
/*--------------------------------------------------------------------------*/
int sci_fftw_flags(char *fname, unsigned long fname_len)
{
/* declaration of variables to store scilab parameters address */
static int l1 = 0, m1 = 0, n1 = 0;
SciIntMat M1;
char **Str1 = NULL;
static int l2 = 0, m2 = 0, n2 = 0;
char **Str3 = NULL;
/* please update me ! */
static int nb_flag = 22;
static char *Str[] =
{
/* documented flags */
"FFTW_MEASURE",
"FFTW_DESTROY_INPUT",
"FFTW_UNALIGNED",
"FFTW_CONSERVE_MEMORY",
"FFTW_EXHAUSTIVE",
"FFTW_PRESERVE_INPUT",
"FFTW_PATIENT",
"FFTW_ESTIMATE",
/* undocumented beyond-guru flags */
"FFTW_ESTIMATE_PATIENT",
"FFTW_BELIEVE_PCOST",
"FFTW_NO_DFT_R2HC",
"FFTW_NO_NONTHREADED",
"FFTW_NO_BUFFERING",
"FFTW_NO_INDIRECT_OP",
"FFTW_ALLOW_LARGE_GENERIC",
"FFTW_NO_RANK_SPLITS",
"FFTW_NO_VRANK_SPLITS",
"FFTW_NO_VRECURSE",
"FFTW_NO_SIMD",
"FFTW_NO_SLOW",
"FFTW_NO_FIXED_RADIX_LARGE_N",
"FFTW_ALLOW_PRUNING"
};
static unsigned flagt[] =
{
/* documented flags */
FFTW_MEASURE,
FFTW_DESTROY_INPUT,
FFTW_UNALIGNED,
FFTW_CONSERVE_MEMORY,
FFTW_EXHAUSTIVE,
FFTW_PRESERVE_INPUT,
FFTW_PATIENT,
FFTW_ESTIMATE,
/* undocumented beyond-guru flags */
FFTW_ESTIMATE_PATIENT,
FFTW_BELIEVE_PCOST,
FFTW_NO_DFT_R2HC,
FFTW_NO_NONTHREADED,
FFTW_NO_BUFFERING,
FFTW_NO_INDIRECT_OP,
FFTW_ALLOW_LARGE_GENERIC,
FFTW_NO_RANK_SPLITS,
FFTW_NO_VRANK_SPLITS,
FFTW_NO_VRECURSE,
FFTW_NO_SIMD,
FFTW_NO_SLOW,
FFTW_NO_FIXED_RADIX_LARGE_N,
FFTW_ALLOW_PRUNING
};
unsigned flagv = 0;
int i = 0, j = 0;
CheckRhs(0, 1);
if (Rhs == 0)
{
// nothing
}
else
{
switch(VarType(1))
{
case sci_ints:
/* int */
GetRhsVar(1, MATRIX_OF_VARIABLE_SIZE_INTEGER_DATATYPE, &m1, &n1, &M1);
CheckDims(1, m1, n1, 1, 1);
setCurrentFftwFlags(((int *)M1.D)[0]);
break;
case sci_matrix:
/* double */
GetRhsVar(1, MATRIX_OF_DOUBLE_DATATYPE, &m1, &n1, &l1);
CheckDims(1, m1, n1, 1, 1);
setCurrentFftwFlags((int)*stk(l1));
break;
case sci_strings:
/* string */
GetRhsVar(1, MATRIX_OF_STRING_DATATYPE, &m1, &n1, &Str1);
for (j = 0; j < m1 * n1; j++)
{
for (i = 0; i < nb_flag; i++)
{
if (strcmp(Str1[j], Str[i]) == 0) break;
}
if (i == nb_flag)
{
freeArrayOfString(Str1, m1 * n1);
Scierror(999, _("%s: Wrong values for input argument #%d: FFTW flag expected.\n"), fname, 1);
return 0;
}
else
{
if (i > 0)
{
flagv = ( flagv | (1U << (i - 1)) );
}
}
}
setCurrentFftwFlags(flagv);
freeArrayOfString(Str1, m1 * n1);
break;
default:
Scierror(53, _("%s: Wrong type for input argument #%d.\n"), fname, 1);
return 0;
}
}
/* return value of Sci_Plan.flags in position 2 */
m2 = 1;
n2 = m2;
l2 = I_INT32;
CreateVar(Rhs + 2, MATRIX_OF_VARIABLE_SIZE_INTEGER_DATATYPE, &m2, &n2, &l2);
*istk(l2) = (int) getCurrentFftwFlags();
/*Test for only FFTW_MEASURE*/
if (getCurrentFftwFlags() == 0)
{
j = 1;
if ((Str3 = (char **)MALLOC(sizeof(char *))) == NULL)
{
Scierror(999, _("%s: No more memory.\n"), fname);
return 0;
}
Str3[0] = strdup(Str[0]);
if (Str3[0] == NULL)
{
Scierror(999, _("%s: No more memory.\n"), fname);
return 0;
}
}
else
{
j = 0;
for (i = 1;i < nb_flag; i++)
{
if((getCurrentFftwFlags() & flagt[i]) == flagt[i])
{
j++;
if (Str3)
{
Str3 = (char **)REALLOC(Str3,sizeof(char *) * j);
}
else
{
Str3 = (char **)MALLOC(sizeof(char *) * j);
}
if ( Str3 == NULL)
{
Scierror(999, _("%s: No more memory.\n"), fname);
return 0;
}
Str3[j - 1] = strdup(Str[i]);
if (Str3[j - 1] == NULL)
{
freeArrayOfString(Str3, j);
Scierror(999, _("%s: No more memory.\n"), fname);
return 0;
}
}
}
}
n1 = 1;
CreateVarFromPtr(Rhs + 3, MATRIX_OF_STRING_DATATYPE, &j, &n1, Str3);
freeArrayOfString(Str3, j);
LhsVar(1) = Rhs + 2;
LhsVar(2) = Rhs + 3;
PutLhsVar();
return 0;
}