def __next__(self):
"""Load the next molecule from the SDF file
This method is part of the iterator protocol.
"""
while True:
title = next(self.f)
if len(title) == 0:
raise StopIteration
else:
title = title.strip()
next(self.f) # skip line
next(self.f) # skip empty line
words = next(self.f).split()
if len(words) < 2:
raise FileFormatError(
"Expecting at least two numbers at fourth line.")
try:
num_atoms = int(words[0])
num_bonds = int(words[1])
except ValueError:
raise FileFormatError(
"Expecting at least two numbers at fourth line.")
numbers = np.zeros(num_atoms, int)
coordinates = np.zeros((num_atoms, 3), float)
for i in range(num_atoms):
words = next(self.f).split()
if len(words) < 4:
raise FileFormatError(
"Expecting at least four words on an atom line.")
try:
coordinates[i, 0] = float(words[0])
coordinates[i, 1] = float(words[1])
coordinates[i, 2] = float(words[2])
except ValueError:
raise FileFormatError(
"Coordinates must be floating point numbers.")
atom = periodic[words[3]]
if atom is None:
raise FileFormatError("Unrecognized atom symbol: %s" %
words[3])
numbers[i] = atom.number
coordinates *= angstrom
edges = []
orders = np.zeros(num_bonds, int)
for i in range(num_bonds):
words = next(self.f).split()
if len(words) < 3:
raise FileFormatError(
"Expecting at least three numbers on a bond line.")
try:
edges.append((int(words[0]) - 1, int(words[1]) - 1))
orders[i] = int(words[2])
except ValueError:
raise FileFormatError(
"Expecting at least three numbers on a bond line.")
formal_charges = np.zeros(len(numbers), int)
line = next(self.f)
while line != "M END\n":
if line.startswith("M CHG"):
words = line[6:].split(
)[1:] # drop the first number which is the number of charges
i = 0
while i < len(words) - 1:
try:
formal_charges[int(words[i]) - 1] = int(words[i +
1])
except ValueError:
raise FileFormatError(
"Expecting only integer formal charges.")
i += 2
line = next(self.f)
# Read on to the next molecule
for line in self.f:
if line == "$$$$\n":
break
molecule = Molecule(numbers, coordinates, title)
molecule.formal_charges = formal_charges
molecule.formal_charges.setflags(write=False)
molecule.graph = MolecularGraph(edges, numbers, orders)
return molecule
def next(self):
"""Load the next molecule from the SDF file
This method is part of the iterator protocol.
"""
while True:
title = self.f.next()
if len(title) == 0:
raise StopIteration
else:
title = title.strip()
self.f.next() # skip line
self.f.next() # skip empty line
words = self.f.next().split()
if len(words) < 2:
raise FileFormatError("Expecting at least two numbers at fourth line.")
try:
num_atoms = int(words[0])
num_bonds = int(words[1])
except ValueError:
raise FileFormatError("Expecting at least two numbers at fourth line.")
numbers = numpy.zeros(num_atoms, int)
coordinates = numpy.zeros((num_atoms, 3), float)
for i in xrange(num_atoms):
words = self.f.next().split()
if len(words) < 4:
raise FileFormatError("Expecting at least four words on an atom line.")
try:
coordinates[i, 0] = float(words[0])
coordinates[i, 1] = float(words[1])
coordinates[i, 2] = float(words[2])
except ValueError:
raise FileFormatError("Coordinates must be floating point numbers.")
atom = periodic[words[3]]
if atom is None:
raise FileFormatError("Unrecognized atom symbol: %s" % words[3])
numbers[i] = atom.number
coordinates *= angstrom
edges = []
orders = numpy.zeros(num_bonds, int)
for i in xrange(num_bonds):
words = self.f.next().split()
if len(words) < 3:
raise FileFormatError("Expecting at least three numbers on a bond line.")
try:
edges.append((int(words[0])-1, int(words[1])-1))
orders[i] = int(words[2])
except ValueError:
raise FileFormatError("Expecting at least three numbers on a bond line.")
formal_charges = numpy.zeros(len(numbers), int)
line = self.f.next()
while line != "M END\n":
if line.startswith("M CHG"):
words = line[6:].split()[1:] # drop the first number which is the number of charges
i = 0
while i < len(words)-1:
try:
formal_charges[int(words[i])-1] = int(words[i+1])
except ValueError:
raise FileFormatError("Expecting only integer formal charges.")
i += 2
line = self.f.next()
# Read on to the next molecule
for line in self.f:
if line == "$$$$\n":
break
molecule = Molecule(numbers, coordinates, title)
molecule.formal_charges = formal_charges
molecule.formal_charges.setflags(write=False)
molecule.graph = MolecularGraph(edges, numbers, orders)
return molecule
