def do_closure(self, text):
num = normalize_closure(text)
if self.closures.has_key(num):
prev_atom, bond = self.closures[num]
del self.closures[num]
assert self._pending_bond is not None, "Can't happen"
if self._pending_bond is not implicit_bond and \
bond is not implicit_bond and \
self._pending_bond.symbol != "-": # according to toolkit
# need to verify they are compatible
prev_symbol = bond.symbol
symbol = self._pending_bond.symbol
if (prev_symbol == symbol) or \
(prev_symbol == "/" and symbol == "\\") or \
(prev_symbol == "\\" and symbol == "/"):
pass
else:
raise AssertionError("bond types don't match")
elif bond is implicit_bond and self._pending_bond is not implicit_bond:
# see if one of the bonds is not implicit and keep it
bond = self._pending_bond
elif bond is implicit_bond:
# both are implicit so make a new one
bond = Bond()
bond._closure = 1
atom = self._prev_atoms[-1]
if prev_atom is atom:
raise AssertionError("cannot close a ring with itself")
bond.atoms[:] = [prev_atom, atom]
prev_atom._closure = 1
atom._closure = 1
##self.mol.add_bond(bond, prev_atom, atom)
bond.atoms = [prev_atom, atom]
atom.bonds.append(bond)
prev_atom.bonds.append(bond)
atom.oatoms.append(prev_atom)
prev_atom.oatoms.append(atom)
self.bonds.append(bond)
if self.vfgraph:
index1, index2 = prev_atom.index, atom.index
insert_edge = self.insert_edge
insert_edge(index1, index2, bond)
insert_edge(index2, index1, bond)
else:
self.closures[num] = (self._prev_atoms[-1], self._pending_bond)
self._pending_bond = implicit_bond
def do_closure(self, text):
num = normalize_closure(text)
if self.closures.has_key(num):
prev_atom, bond = self.closures[num]
del self.closures[num]
assert self._pending_bond is not None, "Can't happen"
if self._pending_bond is not implicit_bond and \
bond is not implicit_bond and \
self._pending_bond.symbol != "-": # according to toolkit
# need to verify they are compatible
prev_symbol = bond.symbol
symbol = self._pending_bond.symbol
if (prev_symbol == symbol) or \
(prev_symbol == "/" and symbol == "\\") or \
(prev_symbol == "\\" and symbol == "/"):
pass
else:
raise AssertionError("bond types don't match")
elif bond is implicit_bond and self._pending_bond is not implicit_bond:
# see if one of the bonds is not implicit and keep it
bond = self._pending_bond
elif bond is implicit_bond:
# both are implicit so make a new one
bond = Bond()
bond._closure = 1
atom = self._prev_atoms[-1]
if prev_atom is atom:
raise AssertionError("cannot close a ring with itself")
bond.atoms[:] = [prev_atom, atom]
prev_atom._closure = 1
atom._closure = 1
##self.mol.add_bond(bond, prev_atom, atom)
bond.atoms = [prev_atom, atom]
atom.bonds.append(bond)
prev_atom.bonds.append(bond)
atom.oatoms.append(prev_atom)
prev_atom.oatoms.append(atom)
self.bonds.append(bond)
if self.vfgraph:
index1, index2 = prev_atom.index, atom.index
insert_edge = self.insert_edge
insert_edge(index1, index2, bond)
insert_edge(index2, index1, bond)
else:
self.closures[num] = (self._prev_atoms[-1], self._pending_bond)
self._pending_bond = implicit_bond
def do_bond(self, text):
assert self._pending_bond in (implicit_bond, None)
symbol, bondorder, bondtype, equiv_class, stereo = BONDLOOKUP[text]
# if the bond came in as aromatic (which it
# CAN'T!))
if bondtype == 4:
assert 0, "Bond's shouldn't come in as ':'"
fixed = 0
else:
fixed = 1
bond = Bond(text, bondorder, bondtype, fixed, stereo)
bond.equiv_class = equiv_class
self._pending_bond = bond
def do_bond(self, text):
assert self._pending_bond in (implicit_bond, None)
symbol, bondorder, bondtype, equiv_class, stereo = BONDLOOKUP[text]
# if the bond came in as aromatic (which it
# CAN'T!))
if bondtype == 4:
assert 0, "Bond's shouldn't come in as ':'"
fixed = 0
else:
fixed = 1
bond = Bond(text, bondorder, bondtype, fixed, stereo)
bond.equiv_class = equiv_class
self._pending_bond = bond
def addHydrogens(mol):
if not mol.explicitHydrogens:
for atom in mol.atoms:
for i in range(atom.hcount):
hatom = Atom()
hatom.symbol = 'H'
#is the coordinates of hydrogen stored somewhere in the
#atom the hydrogen is attached to?
## hatom.x =
## hatom.y =
## hatom.z =
#does the hcount need to be changed?
## atom.hcount -= 1
mol.add_atom(hatom)
bond = Bond()
mol.add_bond(bond, atom, hatom)
#reset atom indices
index = 0
for atom in mol.atoms:
atom.index = index
index += 1
mol.explicitHydrogens = 1 #change flag to let know that hydrogens are explicit in this mol
return mol
def add_atom(self, atom):
atoms = self.atoms
atom.index = len(atoms)
atoms.append(atom)
if self.vfgraph:
index = self.insert_node(atom)
## assert index == atom.index, "%s <--> %s"%(index, atom.index)
## self.mol.add_atom(atom)
if self._pending_bond == implicit_bond:
# Implicit single or aromatic bond
self._pending_bond = Bond()
if self._pending_bond is not None:
bond = self._pending_bond
prev_atom = self._prev_atoms[-1]
bond.atoms[:] = [prev_atom, atom]
##self.mol.add_bond(bond, prev_atom, atom)
bond.atoms = [prev_atom, atom]
atom.bonds.append(bond)
prev_atom.bonds.append(bond)
atom.oatoms.append(prev_atom)
prev_atom.oatoms.append(atom)
self.bonds.append(bond)
if self.vfgraph:
index1, index2 = prev_atom.index, atom.index
insert_edge = self.insert_edge
insert_edge(index1, index2, bond)
insert_edge(index2, index1, bond)
self._pending_bond = implicit_bond
if not self._prev_atoms:
self._prev_atoms.append(atom)
else:
self._prev_atoms[-1] = atom
def reader(file, stripHydrogens=1):
lines = collector(file)
while 1:
try:
fields = {}
name = lines.next().strip()
userLine = lines.next().strip()
comment = lines.next().strip()
molinfo = lines.next()
numAtoms, numBonds = int(molinfo[0:3]), int(molinfo[3:6])
atoms = [] # this is the full list of atoms
_atoms = [] # this is the (potentially stripped list
# of atoms. I.e. no hydrogens.)
i = 0
for index in range(numAtoms):
line = lines.next()
x,y,z,symbol,mass,charge,stereo,hcount,hcount_fixed = parse_atom(line)
if symbol == "H" and stripHydrogens:
atoms.append(None)
else:
atom = Atom()
atoms.append(atom)
_atoms.append(atom)
atom.set_symbol(symbol)# = symbol
atom.explicit_hcount = hcount
atom.charge = charge
atom._line = line
atom.x = x
atom.y = y
atom.z = z
if hcount_fixed:
print "hcount fixed"
atom.fixed_hcount = 1 # oops, we shouldn't use this
atom.has_explicit_hcount = True
if mass:
atom.weight = atom.mass + mass
atom.index = i
i = i + 1
bonds = []
for index in range(numBonds):
line = lines.next()
a1, a2, bondtype, stereo, remainder = parse_bond(line)
symbol, bondorder, bondtype, fixed = BOND_SYMBOL[bondtype]
atom1, atom2 = atoms[a1], atoms[a2]
if atom1 is not None and atom2 is not None:
h1, h2 = atom1.handle, atom2.handle
bond = Bond(symbol, bondorder, bondtype, fixed)
bonds.append(bond)
bond._line = remainder
bond.index = index
bond.atoms = [atom1, atom2]
try:
bond.stereo = BOND_LOOKUP_STEREO[bondtype-1][stereo]
except KeyError:
raise MolReaderError("An SD record cannot have a bondtype of %s and a stereo value of %s"%(bondtype, stereo))
except IndexError:
print "*"*44
print line
print "bondtype, stereo", bondtype, stero
raise
atom1.bonds.append(bond)
atom2.bonds.append(bond)
atom1.oatoms.append(atom2)
atom2.oatoms.append(atom1)
if atom1.symbol == "H": atom2.explicit_hcount += 1
if atom2.symbol == "H": atom1.explicit_hcount += 1
else:
if atom1 is None and atom2 is not None:
atom2.explicit_hcount += 1
elif atom2 is None and atom1 is not None:
atom1.explicit_hcount += 1
##############################################################
# read the mprops if necessary
line = lines.next().strip()
while 1:
if line and line[0:6] == "M END":
line = lines.next().strip()
break
elif line == "M CHG":
groups = line[6:].split()[1:]
index = 0
while index < len(groups):
atomIndex = int(groups[index]) - 1
atom = self.atoms[atomIndex]
charge = int(groups[index+1])
self.atoms[atomIndex].charge = charge
index += 2
line = lines.next().strip()
elif line and line[0] == ">":
break
elif line[0:4] == "$$$$":
break
line = lines.next().strip()
# What about end of mol?
#############################################################
# read the fields if necessary
while line != "$$$$":
if line and line[0] == ">":
res = FIELDPATTERN.match(line)
if res: field, potentialID = res.groups()
else:
res = ALTFIELDPATTERN.match(line)
if res:
field = res.groups()[0]
potentialID = None
else:
field, potentialID = None, None
if name is None: name = potentialID
if field:
data = []
line = lines.next().strip()
while line and line != "$$$$":
data.append(line)
line = lines.next().strip()
fields[field] = os.linesep.join(data)
line = lines.next().strip()
mol = Molecule(_atoms, bonds)
mol.name = name
mol.fields = fields
mol.name = name
yield mol, lines.dump(), None
except StopIteration:
break
except Exception:
line = lines.current.strip()
while line[0:4] != "$$$$":
line = lines.next().strip()
stdout, stderr = sys.stdout, sys.stderr
sys.stdout = sys.stderr = io = StringIO()
traceback.print_exc()
sys.stdout = stdout
sys.stderr = stderr
yield None, lines.dump(), io.getvalue()
class MolReader:
def __init__(self, file, stripHydrogens=1):
self.file = file
self.iterator = iter(file)
self._lastlines = [] # lastlines stores the original lines that made up the
# last molecule read
self.stripHydrogens = stripHydrogens
self._lastline = None
def _readline(self, endOk=0):
"""internal readline function, if endOk is 0 then upon an end of
line a MolReaderError is generated"""
if self._lastline:
res = self._lastline
self._lastline = None
return res
try:
line = self.iterator.next()
self._lastlines.append(line)
except StopIteration:
line = None
if not line and not endOk:
raise MolReaderError, "Unexpected end of file"
return line
def _pushback(self, line):
self._lastline = line
def _clear(self):
"""Clear the _lastlines buffer"""
self._lastlines = []
def get_text(self):
"""->text that formed the last molecule read"""
return "".join(self._lastlines)
def get_lines(self):
"""->the lines of text that formed the last molecule read"""
return self._lastlines
def _read_to_next(self):
readline = self._readline
endOfMol = self._endOfMol
while 1:
line = readline(endOk=1)
if not line:
break
if endOfMol(line):
break
def _endOfMol(self, line):
"""(line)-> return 1 if the line signifies the end of molecule
0 otherwise"""
if line[0:4] == "$$$$":
return 1
return 0
def _readFields(self,
pattern=re.compile(">\s+<([^>]+)>\s+\(*([^)]*)")
):
"""Read the database field component at the end of a molecule
record. Sets a dictionary of key->values"""
readline = self._readline
endOfMol = self._endOfMol
fields = {}
name = None
while 1:
# by setting endOk = 1 we can read mol files as
# well as sdfiles
line = readline(endOk=1)
if not line:
break
if endOfMol(line):
break
elif line[0] == ">":
# we have a data line so get the field
# and potentialID values
if not endOfMol(line):
res = pattern.match(line)
if res:
field, potentialID = res.groups()
else:
field, potentialID = None, None
if name is None:
name = potentialID
elif name != potentialID:
name = "UNKNOWN (id clash)"
# read the data from the next line
if field:
line = readline().strip()
data = []
while line:
data.append(line)
line = readline().strip()
if not endOfMol(line):
fields[field] = os.linesep.join(data)
else:
break
if endOfMol(line):
break
if not endOfMol(line):
# by setting endok = 1 here we can read
# mol files as well as sd files
line = readline(endOk=1)
return fields, name
def readMProps(self):
readline = self._readline
while 1:
line = readline()
if line[0] == ">":
# need to push back the last line
self._pushback(line)
return
if line[0:6] == "M END":
break
if line[0:6] == "M CHG":
# parse the charge line and add charges
# to the correct atoms
groups = line[6:].split()[1:]
index = 0
while index < len(groups):
atomIndex = int(groups[index]) - 1
atom = self.atoms[atomIndex]
charge = int(groups[index+1])
self.atoms[atomIndex].charge = charge
index += 2
def read_one(self):
"""Read one molecule from the sd file"""
self._clear()
readline = self._readline
endOfMol = self._endOfMol
try:
name = readline().strip()
userLine = readline()
comment = readline()
line = readline()
except MolReaderError, msg:
if str(msg) == "Unexpected end of file":
return None
raise
try:
numAtoms, numBonds = map(int, (line[0:3], line[3:6]))
except ValueError: # XXX FIX ME - trap exceptions and stuff
print "cannot parse atom, bond line"
self._read_to_next()
return None
atoms = self.atoms = []
for index in range(numAtoms):
line = readline()
try:
x,y,z,symbol,mass,charge,stereo,hcount,hcount_fixed = parse_atom(line)
except: # XXX FIX ME - trap exceptions and stuff
self._read_to_next()
return None
atom = Atom()
atom._line = line
atom.symbol = symbol
atom.explicit_hcount = hcount
atom.charge = charge
#if hcount_fixed:
#symbol, hcount, charge, weight=0, aromatic=0)
# XXX FIX ME
# a really bad hack here.
# ignore please!
atom._line = line
atom.x = x
atom.y = y
atom.z = z
if hcount_fixed: atom.fixed_hcount = 1
if mass:
atom.weight = atom.mass + mass
atom.index = len(atoms)
atoms.append(atom)
if vfgraph:
insert_node(atom)
bonds = []
mappings = []
bondCount = [0] * len(atoms)
closures = {}
for index in range(numBonds):
line = readline()
try:
a1, a2, bondtype, remainder = parse_bond(line)
except:
self._read_to_next()
return None
a1 -= 1
a2 -= 1
symbol, bondorder, bondtype, fixed = BOND_SYMBOL[bondtype]
atom1 = atoms[a1]
atom2 = atoms[a2]
if stripHydrogen:
if atom1.symbol == "H":
atom2.hcount += 1
atom2.hcount_fixed = 1
if atom2.symbol == "H":
atom1.hcount += 1
atom1.hcount_fixed = 1
else:
bond = Bond(symbol, bondorder, bondtype, fixed)
bond._line = line
# XXX FIX ME
# a really bad hack here
# ignore please!
bond._line = remainder
bond.atoms = [a1, a2]
a1.bonds.append(bond)
a2.bonds.append(bond)
a1.oatoms.append(a1)
a2.oatoms.append(a2)
bonds.append(bond)
## mappings.append((bond, a1, a2))
## bondCount[a1] += 1
## bondCount[a2] += 1
self.readMProps()
fields, potentialName = self._readFields()
if not name:
name = potentialName
elif name != potentialName:
# XXX FIX ME, what do I do here?
pass
# we've tokenized the molecule, now we need to build one
# XXX FIX ME - Split this up into a builder and a tokenizer ?
mol = Molecule()
mol.name = name
mol.fields = fields
# for atom in atoms:
# mol.add_atom(atom)
# for bond, a1, a2 in mappings:
# atom1, atom2 = atoms[a1], atoms[a2]
# XXX FIX ME
# does this format mean the atom's hcount can't
# change?
# stripHydrogens = self.stripHydrogens
# if not hasattr(atom1, "number"):
# print atom
# print atom.__dict__
## if stripHydrogens and atom1.symbol == "H" and bondCount[a1] == 1:
## atom2.hcount += 1
## atom2.hcount_fixed = 1
## bondCount[a2] -=1
## mol.remove_atom(atom1)
## elif stripHydrogens and atom2.symbol == "H" and bondCount[a2] == 1:
## atom1.hcount += 1
## atom1.hcount_fixed = 1
## bondCount[a1] -= 1
## mol.remove_atom(atom2)
## else:
## mol.add_bond(bond, atom1, atom2)
## if bond.bondtype == 4:
## atom1.aromatic = 1
## atom2.aromatic = 1
## # get rid of any non-bonded hydrogens
## atomsToDelete = []
## for atom in mol.atoms:
## if atom.symbol == "H":
## assert len(atom.bonds) == 0
## atomsToDelete.append(atom)
## for atom in atomsToDelete:
## mol.remove_atom(atom)
index = 0
for atom in mol.atoms:
assert atom.symbol != "H"
if len(atom.bonds) > 1:
atom._closure = 1
atom.index = index
index += 1
return mol