def parseLines(self, lines):
"""Parse list of lines in PDB format.
Return Structure object or raise StructureFormatError.
"""
xcfg_Number_of_particles = None
xcfg_A = None
xcfg_H0 = numpy.zeros((3, 3), dtype=float)
xcfg_H0_set = numpy.zeros((3, 3), dtype=bool)
xcfg_NO_VELOCITY = False
xcfg_entry_count = None
xcfg_auxiliary = []
p_nl = 0
p_auxiliary_re = re.compile(r"^auxiliary\[(\d+)\] =")
p_auxiliary = {}
try:
stru = Structure()
# ignore trailing blank lines
stop = len(lines)
while stop > 0 and lines[stop - 1].strip() == "":
stop -= 1
ilines = iter(lines[:stop])
# read XCFG header
for line in ilines:
p_nl += 1
stripped_line = line.strip()
# blank lines and lines starting with # are ignored
if stripped_line == "" or line[0] == '#':
continue
elif xcfg_Number_of_particles is None:
if line.find("Number of particles =") != 0:
emsg = ("%d: first line must " +
"contain 'Number of particles ='") % p_nl
raise StructureFormatError(emsg)
xcfg_Number_of_particles = int(line[21:].split(None, 1)[0])
p_natoms = xcfg_Number_of_particles
elif line.find("A =") == 0:
xcfg_A = float(line[3:].split(None, 1)[0])
elif line.find("H0(") == 0:
i, j = (int(line[3]) - 1, int(line[5]) - 1)
xcfg_H0[i, j] = float(line[10:].split(None, 1)[0])
xcfg_H0_set[i, j] = True
elif line.find(".NO_VELOCITY.") == 0:
xcfg_NO_VELOCITY = True
elif line.find("entry_count =") == 0:
xcfg_entry_count = int(line[13:].split(None, 1)[0])
elif p_auxiliary_re.match(line):
m = p_auxiliary_re.match(line)
idx = int(m.group(1))
p_auxiliary[idx] = line[m.end():].split(None, 1)[0]
else:
break
# check header for consistency
if numpy.any(xcfg_H0_set == False):
emsg = "H0 tensor is not properly defined"
raise StructureFormatError(emsg)
p_auxnum = len(p_auxiliary) and max(p_auxiliary.keys()) + 1
for i in range(p_auxnum):
if not i in p_auxiliary:
p_auxiliary[i] = "aux%d" % i
sorted_aux_keys = p_auxiliary.keys()
sorted_aux_keys.sort()
if p_auxnum != 0:
stru.xcfg = {
'auxiliaries': [p_auxiliary[k] for k in sorted_aux_keys]
}
if 6 - 3 * xcfg_NO_VELOCITY + len(p_auxiliary) != xcfg_entry_count:
emsg = ("%d: auxiliary fields " +
"not consistent with entry_count") % p_nl
raise StructureFormatError(emsg)
# define proper lattice
stru.lattice.setLatBase(xcfg_H0)
# build p_assign_atom function to assign entries to proper fields
p_exprs = [
"a.xyz[0]=fields[0]", "a.xyz[1]=fields[1]",
"a.xyz[2]=fields[2]"
]
if not xcfg_NO_VELOCITY:
p_exprs += [
"a.v=numpy.zeros(3, dtype=float)", "a.v[0]=fields[3]",
"a.v[1]=fields[4]", "a.v[2]=fields[5]"
]
for idx in sorted_aux_keys:
prop = p_auxiliary[idx]
col = idx + 6 - 3 * xcfg_NO_VELOCITY
if prop == "Uiso":
p_exprs.append("a.U[0,0]=a.U[1,1]=a.U[2,2]=" +
"fields[%d]" % col)
elif re.match(r"^U\d\d$", prop) \
and 1<=int(prop[1])<=3 and 1<=int(prop[2])<=3 :
i, j = int(prop[1]) - 1, int(prop[2]) - 1
if i == j:
p_exprs.append("a.U[%i,%i]=fields[%d]" % (i, j, col))
else:
p_exprs.append("a.U[%i,%i]=a.U[%i,%i]=fields[%d]" % \
(i, j, j, i, col) )
else:
p_exprs.append( "a.__dict__[%r]=fields[%d]" % \
(prop, col) )
p_assign_expr = "pass; " + "; ".join(p_exprs[3:])
exec "def p_assign_atom(a, fields) : %s" % p_assign_expr
# here we are inside data
p_element = None
p_nl -= 1
for line in lines[p_nl:stop]:
p_nl += 1
words = line.split()
# ignore atom mass
if len(words) == 1 and isfloat(words[0]):
continue
# parse element allowing empty symbol
elif len(words) <= 1:
w = line.strip()
p_element = w[:1].upper() + w[1:].lower()
elif len(words) == xcfg_entry_count and p_element is not None:
fields = [float(w) for w in words]
stru.addNewAtom(p_element, fields[:3])
a = stru.getLastAtom()
a.xyz *= xcfg_A
p_assign_atom(a, fields)
else:
emsg = "%d: invalid record" % p_nl
raise StructureFormatError(emsg)
if len(stru) != p_natoms:
emsg = "expected %d atoms, read %d" % (p_natoms, len(stru))
raise StructureFormatError(emsg)
except (ValueError, IndexError):
emsg = "%d: file is not in XCFG format" % p_nl
exc_type, exc_value, exc_traceback = sys.exc_info()
raise StructureFormatError, emsg, exc_traceback
return stru
def parseLines(self, lines):
"""Parse list of lines in PDB format.
Return Structure instance or raise StructureFormatError.
"""
try:
stru = Structure()
scale = numpy.identity(3, dtype=float)
scaleU = numpy.zeros(3, dtype=float)
p_nl = 0
for line in lines:
p_nl += 1
# skip blank lines
if not line.strip(): continue
# make sure line has 80 characters
if len(line) < 80:
line = "%-80s" % line
words = line.split()
record = words[0]
if record == "TITLE":
continuation = line[8:10]
if continuation.strip():
stru.description += line[10:].rstrip()
else:
stru.description = line[10:].rstrip()
elif record == "CRYST1":
a = float(line[7:15])
b = float(line[15:24])
c = float(line[24:33])
alpha = float(line[33:40])
beta = float(line[40:47])
gamma = float(line[47:54])
stru.lattice.setLatPar(a, b, c, alpha, beta, gamma)
scale = numpy.transpose(stru.lattice.recbase)
elif record == "SCALE1":
sc = numpy.zeros((3,3), dtype=float)
sc[0,:] = [float(x) for x in line[10:40].split()]
scaleU[0] = float(line[45:55])
elif record == "SCALE2":
sc[1,:] = [float(x) for x in line[10:40].split()]
scaleU[1] = float(line[45:55])
elif record == "SCALE3":
sc[2,:] = [float(x) for x in line[10:40].split()]
scaleU[2] = float(line[45:55])
base = numpy.transpose(numpy.linalg.inv(sc))
abcABGcryst = numpy.array(stru.lattice.abcABG())
stru.lattice.setLatBase(base)
abcABGscale = numpy.array(stru.lattice.abcABG())
reldiff = numpy.fabs(1.0 - abcABGscale/abcABGcryst)
if not numpy.all(reldiff < self.epsilon):
emsg = "%d: " % p_nl + \
"SCALE and CRYST1 are not consistent."
raise StructureFormatError(emsg)
if numpy.any(scaleU != 0.0):
emsg = "Origin offset not yet implemented."
raise NotImplementedError(emsg)
elif record in ("ATOM", "HETATM"):
name = line[12:16].strip()
rc = [float(x) for x in line[30:54].split()]
xyz = numpy.dot(scale, rc) + scaleU
try:
occupancy = float(line[54:60])
except ValueError:
occupancy = 1.0
try:
B = float(line[60:66])
U = numpy.identity(3)*B/(8*pi**2)
except ValueError:
U = numpy.zeros((3,3), dtype=float)
symbol = line[76:78].strip()
if symbol == "":
# get symbol from the first 2 characters of name
symbol = line[12:14].strip()
symbol = symbol[0].upper() + symbol[1:].lower()
#stru.addNewAtom(symbol, occupancy=occupancy, name=name, U=U)
stru.addNewAtom(symbol, occupancy=occupancy, label=name, U=U)
last_atom = stru.getLastAtom()
last_atom.xyz_cartn = rc
elif record == "SIGATM":
sigrc = [float(x) for x in line[30:54].split()]
sigxyz = numpy.dot(scale, sigrc)
try:
sigo = float(line[54:60])
except ValueError:
sigo = 0.0
try:
sigB = float(line[60:66])
sigU = numpy.identity(3)*sigB/(8*pi**2)
except ValueError:
sigU = numpy.zeros((3,3), dtype=float)
last_atom.sigxyz = sigxyz
last_atom.sigo = sigo
last_atom.sigU = sigU
elif record == "ANISOU":
Uij = [ float(x)*1.0e-4 for x in line[28:70].split() ]
for i in range(3):
last_atom.U[i,i] = Uij[i]
last_atom.U[0,1] = last_atom.U[1,0] = Uij[3]
last_atom.U[0,2] = last_atom.U[2,0] = Uij[4]
last_atom.U[1,2] = last_atom.U[2,1] = Uij[5]
elif record == "SIGUIJ":
sigUij = [ float(x)*1.0e-4 for x in line[28:70].split() ]
for i in range(3):
last_atom.sigU[i,i] = sigUij[i]
last_atom.sigU[0,1] = last_atom.sigU[1,0] = sigUij[3]
last_atom.sigU[0,2] = last_atom.sigU[2,0] = sigUij[4]
last_atom.sigU[1,2] = last_atom.sigU[2,1] = sigUij[5]
elif record in P_pdb.validRecords:
pass
else:
emsg = "%d: invalid record name '%r'" % (p_nl, record)
raise StructureFormatError(emsg)
except (ValueError, IndexError):
emsg = "%d: invalid PDB record" % p_nl
exc_type, exc_value, exc_traceback = sys.exc_info()
raise StructureFormatError, emsg, exc_traceback
return stru
def parseLines(self, lines):
"""Parse list of lines in PDB format.
Return Structure object or raise StructureFormatError.
"""
xcfg_Number_of_particles = None
xcfg_A = None
xcfg_H0 = numpy.zeros((3,3), dtype=float)
xcfg_H0_set = numpy.zeros((3,3), dtype=bool)
xcfg_NO_VELOCITY = False
xcfg_entry_count = None
xcfg_auxiliary = []
p_nl = 0
p_auxiliary_re = re.compile(r"^auxiliary\[(\d+)\] =")
p_auxiliary = {}
try:
stru = Structure()
# ignore trailing blank lines
stop = len(lines)
while stop>0 and lines[stop-1].strip() == "":
stop -= 1
ilines = iter(lines[:stop])
# read XCFG header
for line in ilines:
p_nl += 1
stripped_line = line.strip()
# blank lines and lines starting with # are ignored
if stripped_line == "" or line[0] == '#':
continue
elif xcfg_Number_of_particles is None:
if line.find("Number of particles =") != 0:
emsg = ("%d: first line must " +
"contain 'Number of particles ='") % p_nl
raise StructureFormatError(emsg)
xcfg_Number_of_particles = int(line[21:].split(None, 1)[0])
p_natoms = xcfg_Number_of_particles
elif line.find("A =") == 0:
xcfg_A = float(line[3:].split(None, 1)[0])
elif line.find("H0(") == 0:
i, j = ( int(line[3])-1 , int(line[5])-1 )
xcfg_H0[i,j] = float(line[10:].split(None, 1)[0])
xcfg_H0_set[i,j] = True
elif line.find(".NO_VELOCITY.") == 0:
xcfg_NO_VELOCITY = True
elif line.find("entry_count =") == 0:
xcfg_entry_count = int(line[13:].split(None, 1)[0])
elif p_auxiliary_re.match(line):
m = p_auxiliary_re.match(line)
idx = int(m.group(1))
p_auxiliary[idx] = line[m.end():].split(None, 1)[0]
else:
break
# check header for consistency
if numpy.any(xcfg_H0_set == False):
emsg = "H0 tensor is not properly defined"
raise StructureFormatError(emsg)
p_auxnum = len(p_auxiliary) and max(p_auxiliary.keys())+1
for i in range(p_auxnum):
if not i in p_auxiliary:
p_auxiliary[i] = "aux%d" % i
sorted_aux_keys = p_auxiliary.keys()
sorted_aux_keys.sort()
if p_auxnum != 0:
stru.xcfg = {
'auxiliaries' : [ p_auxiliary[k]
for k in sorted_aux_keys ]
}
if 6-3*xcfg_NO_VELOCITY+len(p_auxiliary) != xcfg_entry_count:
emsg = ("%d: auxiliary fields " +
"not consistent with entry_count") % p_nl
raise StructureFormatError(emsg)
# define proper lattice
stru.lattice.setLatBase(xcfg_H0)
# build p_assign_atom function to assign entries to proper fields
p_exprs = [ "a.xyz[0]=fields[0]",
"a.xyz[1]=fields[1]",
"a.xyz[2]=fields[2]" ]
if not xcfg_NO_VELOCITY:
p_exprs += [ "a.v=numpy.zeros(3, dtype=float)",
"a.v[0]=fields[3]",
"a.v[1]=fields[4]",
"a.v[2]=fields[5]" ]
for idx in sorted_aux_keys:
prop = p_auxiliary[idx]
col = idx + 6 - 3*xcfg_NO_VELOCITY
if prop == "Uiso":
p_exprs.append("a.U[0,0]=a.U[1,1]=a.U[2,2]=" +
"fields[%d]" % col)
elif re.match(r"^U\d\d$", prop) \
and 1<=int(prop[1])<=3 and 1<=int(prop[2])<=3 :
i, j = int(prop[1])-1, int(prop[2])-1
if i==j:
p_exprs.append("a.U[%i,%i]=fields[%d]" % (i, j, col) )
else:
p_exprs.append("a.U[%i,%i]=a.U[%i,%i]=fields[%d]" % \
(i, j, j, i, col) )
else:
p_exprs.append( "a.__dict__[%r]=fields[%d]" % \
(prop, col) )
p_assign_expr = "pass; " + "; ".join(p_exprs[3:])
exec "def p_assign_atom(a, fields) : %s" % p_assign_expr
# here we are inside data
p_element = None
p_nl -= 1
for line in lines[p_nl:stop]:
p_nl += 1
words = line.split()
# ignore atom mass
if len(words) == 1 and isfloat(words[0]):
continue
# parse element allowing empty symbol
elif len(words) <= 1:
w = line.strip()
p_element = w[:1].upper() + w[1:].lower()
elif len(words) == xcfg_entry_count and p_element is not None:
fields = [ float(w) for w in words ]
stru.addNewAtom(p_element, fields[:3])
a = stru.getLastAtom()
a.xyz *= xcfg_A
p_assign_atom(a, fields)
else:
emsg = "%d: invalid record" % p_nl
raise StructureFormatError(emsg)
if len(stru) != p_natoms:
emsg = "expected %d atoms, read %d" % (p_natoms, len(stru))
raise StructureFormatError(emsg)
except (ValueError, IndexError):
emsg = "%d: file is not in XCFG format" % p_nl
exc_type, exc_value, exc_traceback = sys.exc_info()
raise StructureFormatError, emsg, exc_traceback
return stru
