/* Read full (with hash chains, for future modifying) *
* lzw dictionary function */
void LzwDict::load( Scanner &_scanner )
{
int i, j, HashCode, n, k;
_modified = false;
_alphabetSize = _scanner.readBinaryInt();
_nextCode = _scanner.readBinaryInt();
n = _scanner.readBinaryInt();
_bitcodeSize = _scanner.readBinaryInt();
_maxCode = (1 << _bitcodeSize) - 1;
_storage.clear();
_storage.resize(n);
for (i = 0; i < n; i++)
{
k = _scanner.readBinaryDword();
_storage[i].Prefix = k;
_storage[i].AppendChar = _scanner.readByte();
}
_freePtr = _scanner.readBinaryInt();
_hashKeys.clear_resize(SIZE);
_nextPointers.clear_resize(SIZE);
for (i = 0; i < SIZE; i++)
{
_nextPointers[i] = -1;
_hashKeys[i] = -1;
}
while (!_scanner.isEOF())
{
HashCode = _scanner.readBinaryInt();
j = _scanner.readBinaryInt();
_hashKeys[HashCode] = j;
i = j;
j = _scanner.readBinaryInt();
while (j != -1)
{
_nextPointers[i] = j;
i = j;
j = _scanner.readBinaryInt();
}
}
}
void CmfLoader::loadXyz (Scanner &scanner)
{
if (_mol == 0)
throw Error("loadMolecule() must be called prior to loadXyz()");
int i;
CmfSaver::VecRange range;
range.xyz_min.x = scanner.readBinaryFloat();
range.xyz_min.y = scanner.readBinaryFloat();
range.xyz_min.z = scanner.readBinaryFloat();
range.xyz_range.x = scanner.readBinaryFloat();
range.xyz_range.y = scanner.readBinaryFloat();
range.xyz_range.z = scanner.readBinaryFloat();
range.have_z = (scanner.readByte() != 0);
for (i = 0; i < _atoms.size(); i++)
{
Vec3f pos;
_readVec3f(scanner, pos, range);
int idx = i;
if (has_mapping)
idx = inv_atom_mapping_to_restore[i];
_mol->setAtomXyz(idx, pos.x, pos.y, pos.z);
}
// Read sgroup coordinates data
int idx[5] = {0};
for (int i = 0; i < _sgroup_order.size(); i++)
_readSGroupXYZ(scanner, _sgroup_order[i], idx, *_mol, range);
_mol->have_xyz = true;
}
bool MoleculeAutoLoader::tryMDLCT (Scanner &scanner, Array<char> &outbuf)
{
int pos = scanner.tell();
bool endmark = false;
QS_DEF(Array<char>, curline);
outbuf.clear();
while (!scanner.isEOF())
{
int len = scanner.readByte();
if (len > 90) // Molfiles and Rxnfiles actually have 80 characters limit
{
scanner.seek(pos, SEEK_SET);
// Garbage after endmark means end of data.
// (See the note below about endmarks)
if (endmark)
return true;
return false;
}
curline.clear();
while (len-- > 0)
{
if (scanner.isEOF())
{
scanner.seek(pos, SEEK_SET);
return false;
}
int c = scanner.readChar();
curline.push(c);
}
curline.push(0);
if (endmark)
{
// We can not properly read the data to the end as there
// is often garbage after the actual MDLCT data.
// Instead, we are doing this lousy check:
// "M END" or "$END MOL" can be followed only
// by "$END CTAB" (in Markush queries), or
// by "$MOL" (in Rxnfiles). Otherwise, this
// is actually the end of data.
if (strcmp(curline.ptr(), "$END CTAB") != 0 &&
strcmp(curline.ptr(), "$MOL") != 0)
{
scanner.seek(pos, SEEK_SET);
return true;
}
}
if (strcmp(curline.ptr(), "M END") == 0)
endmark = true;
else if (strcmp(curline.ptr(), "$END MOL") == 0)
endmark = true;
else
endmark = false;
outbuf.appendString(curline.ptr(), false);
outbuf.push('\n');
}
scanner.seek(pos, SEEK_SET);
// It happened once that a valid Molfile had successfully
// made its way through the above while() cycle, and thus
// falsely recognized as MDLCT. To fight this case, we include
// here a check that the last line was actually an endmark
return endmark;
}