Datum
nseq_concat (PG_FUNCTION_ARGS)
{
NSEQ *a = PG_GETARG_NSEQ_P (0);
NSEQ *b = PG_GETARG_NSEQ_P (1);
NSEQ *retval = NULL;
int32 totalsize, totalblocks, overflow;
char *newval = NULL;
int i;
if(a->rna != b->rna)
{
elog(ERROR,"Cannot concatenate DNA and RNA");
}
totalsize = a->size+b->size;
//elog(INFO, "%i", totalsize);
overflow = totalsize % BLOCKSIZE;
//elog(INFO, "%i", overflow);
totalblocks = totalsize / BLOCKSIZE;
//elog(INFO, "%i", totalblocks);
if(overflow > 0)
{
newval = nseq_concat_slow(a,b);
PG_RETURN_NSEQ_P (make_nseq(newval,strlen(newval),a->rna));
}
else
{
newval = (char *) palloc0 (totalblocks);
memcpy(newval, a->data, a->compressed_size);
memcpy(newval+(a->size / BLOCKSIZE), b->data, b->compressed_size);
}
retval = palloc(CALCDATASZ(a->compressed_size+b->compressed_size));
retval->rna = a->rna;
retval->size = totalsize;
retval->compressed_size = a->compressed_size+b->compressed_size;
for(i=0; i<HISTSZ; i++)
{
retval->histogram[i] = a->histogram[i]+b->histogram[i];
}
memcpy(DATAPTR(retval), newval, retval->compressed_size);
SET_VARSIZE (retval,CALCDATASZ(retval->compressed_size));
PG_RETURN_NSEQ_P (retval);
}
static NSEQ *make_nseq(const char* sequence, const size_t seqlen, const bool isRNA)
{
NSEQ *retval = NULL;
char *buffer = NULL;
char tmp;
uint32 i, offset = 0;
uint32 bufsize = seqlen / BLOCKSIZE;
//COMPRESSED_DATA *compressed_data;
int32 histogram[HISTSZ] = {0,0,0,0};
bool run = true;
if((seqlen % BLOCKSIZE) != 0)
{
bufsize += 1;
}
buffer = (char*) palloc0(bufsize*sizeof(char));
while (run && (offset < bufsize))
{
for(i = 0; i<BLOCKSIZE; i++)
{
tmp = toupper(sequence[(offset*BLOCKSIZE)+i]);
switch(tmp)
{
case 'A':
buffer[offset] |= (0x0 << (i*2));
histogram[0]++;
break;
case 'C':
buffer[offset] |= (0x1 << (i*2));
histogram[1]++;
break;
case 'G':
buffer[offset] |= (0x2 << (i*2));
histogram[2]++;
break;
case 'U':
if(isRNA)
{
buffer[offset] |= (0x3 << (i*2));
histogram[3]++;
}
else
{
elog(ERROR, "Unknown nucleotide for DNA: %c\n", tmp);
run = false;
}
break;
case 'T':
if(isRNA)
{
elog(ERROR, "Unknown nucleotide for RNA: %c\n", tmp);
run = false;
}
else
{
buffer[offset] |= (0x3 << (i*2));
histogram[3]++;
}
break;
case '\0':
run = false;
break;
}
}
offset++;
}
//compressed_data = compress_data(buffer, bufsize);
retval = palloc(CALCDATASZ(bufsize));
retval->rna = isRNA;
retval->size = seqlen;
retval->compressed_size = bufsize;
memcpy(retval->histogram, &histogram, sizeof(histogram));
memcpy(DATAPTR(retval), buffer, bufsize);
SET_VARSIZE (retval,CALCDATASZ(bufsize));
//elog(INFO,"make %d %d", seqlen, bufsize);
return retval;
}
MOLECULE *
new_molecule (char *smiles, char *molfile)
{
unsigned int sizemf;
unsigned int sizesmi;
size_t totalsize;
MOLECULE *result;
char *inchikey = NULL;
char *ancillarydata = NULL;
char *aidata = NULL;
uint32 ancsize = 0;
ancillarydata = ob_lyophilize_molecule(smiles);
if(ancillarydata == NULL) {
elog (ERROR, "Molecule lyophilization failed! SMILES:\n %s", smiles);
}
ancsize = *(unsigned int*) ancillarydata;
sizemf = strlen (molfile)+1;
sizesmi = strlen (smiles)+1;
totalsize = CALCDATASZ (sizemf, sizesmi, ancsize);
result = (MOLECULE *) palloc (totalsize);
memset (result, 0x0, totalsize);
if (strchr (smiles, '.') != NULL)
result->disconnected = true;
result->sizemf = sizemf;
result->sizesmi = sizesmi;
strncpy (SMIPTR(result), smiles, sizesmi);
strncpy (MFPTR(result), molfile, sizemf);
aidata = (char*) &((unsigned int*)ancillarydata)[1];
memcpy(ANCPTR(result), aidata, ancsize);
inchikey = ob_smiles_to_inchikey (smiles);
if(inchikey == NULL || strlen(inchikey) != INCHIKEYSZ) {
elog (ERROR, "InChI key generation failled ! SMILES:\n %s", smiles);
} else {
memcpy(result->inchikey, inchikey, INCHIKEYSZ);
}
if (inchikey != NULL) {
free (inchikey);
inchikey=NULL;
}
ob_fp_bin(aidata, result->fp);
if(ancillarydata != NULL) {
free(ancillarydata);
ancillarydata=NULL;
}
SET_VARSIZE (result,totalsize);
return result;
}
