1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
|
"""Helper functions for read_fdf."""
from os import fstat
from re import compile
_label_strip_re = compile(r'[\s._-]')
def _labelize(raw_label):
# Labels are case insensitive and -_. should be ignored, lower and strip it
return _label_strip_re.sub('', raw_label).lower()
def _is_block(val):
# Tell whether value is a block-value or an ordinary value.
# A block is represented as a list of lists of strings,
# and a ordinary value is represented as a list of strings
if isinstance(val, list) and \
len(val) > 0 and \
isinstance(val[0], list):
return True
return False
def _get_stripped_lines(fd):
# Remove comments, leading blanks, and empty lines
return [_f for _f in [L.split('#')[0].strip() for L in fd] if _f]
def _read_fdf_lines(file, inodes=[]):
# Read lines and resolve includes
if isinstance(file, str):
file = open(file, 'r')
fst = fstat(file.fileno())
inode = (fst.st_dev, fst.st_ino)
if inode in inodes:
raise IOError('Cyclic include in fdf file')
inodes = inodes + [inode]
lbz = _labelize
lines = []
for L in _get_stripped_lines(file):
w0 = lbz(L.split(None, 1)[0])
if w0 == '%include':
# Include the contents of fname
fname = L.split(None, 1)[1].strip()
lines += _read_fdf_lines(fname, inodes)
elif '<' in L:
L, fname = L.split('<', 1)
w = L.split()
fname = fname.strip()
if w0 == '%block':
# "%block label < filename" means that the block contents
# should be read from filename
if len(w) != 2:
raise IOError('Bad %%block-statement "%s < %s"' %
(L, fname))
label = lbz(w[1])
lines.append('%%block %s' % label)
lines += _get_stripped_lines(open(fname))
lines.append('%%endblock %s' % label)
else:
# "label < filename.fdf" means that the label
# (_only_ that label) is to be resolved from filename.fdf
label = lbz(w[0])
fdf = _read_fdf(fname, inodes)
if label in fdf:
if _is_block(fdf[label]):
lines.append('%%block %s' % label)
lines += [' '.join(x) for x in fdf[label]]
lines.append('%%endblock %s' % label)
else:
lines.append('%s %s' % (label, ' '.join(fdf[label])))
# else:
# label unresolved!
# One should possibly issue a warning about this!
else:
# Simple include line L
lines.append(L)
return lines
# The reason for creating a separate _read_fdf is simply to hide the
# inodes-argument
def _read_fdf(fname, inodes=[]):
# inodes is used to detect cyclic includes
fdf = {}
lbz = _labelize
lines = _read_fdf_lines(fname, inodes)
while lines:
w = lines.pop(0).split(None, 1)
if lbz(w[0]) == '%block':
# Block value
if len(w) == 2:
label = lbz(w[1])
content = []
while True:
if len(lines) == 0:
raise IOError('Unexpected EOF reached in %s, '
'un-ended block %s' % (fname, label))
w = lines.pop(0).split()
if lbz(w[0]) == '%endblock' and lbz(w[1]) == label:
break
content.append(w)
if label not in fdf:
# Only first appearance of label is to be used
fdf[label] = content
else:
raise IOError('%%block statement without label')
else:
# Ordinary value
label = lbz(w[0])
if len(w) == 1:
# Siesta interpret blanks as True for logical variables
fdf[label] = []
else:
fdf[label] = w[1].split()
return fdf
def read_fdf(fname):
"""Read a siesta style fdf-file.
The data is returned as a dictionary
( label:value ).
All labels are converted to lower case characters and
are stripped of any '-', '_', or '.'.
Ordinary values are stored as a list of strings (splitted on WS),
and block values are stored as list of lists of strings
(splitted per line, and on WS).
If a label occurres more than once, the first occurrence
takes precedence.
The implementation applies no intelligence, and does not
"understand" the data or the concept of units etc.
Values are never parsed in any way, just stored as
split strings.
The implementation tries to comply with the fdf-format
specification as presented in the siesta 2.0.2 manual.
An fdf-dictionary could e.g. look like this::
{'atomiccoordinatesandatomicspecies': [
['4.9999998', '5.7632392', '5.6095972', '1'],
['5.0000000', '6.5518100', '4.9929091', '2'],
['5.0000000', '4.9746683', '4.9929095', '2']],
'atomiccoordinatesformat': ['Ang'],
'chemicalspecieslabel': [['1', '8', 'O'],
['2', '1', 'H']],
'dmmixingweight': ['0.1'],
'dmnumberpulay': ['5'],
'dmusesavedm': ['True'],
'latticeconstant': ['1.000000', 'Ang'],
'latticevectors': [
['10.00000000', '0.00000000', '0.00000000'],
['0.00000000', '11.52647800', '0.00000000'],
['0.00000000', '0.00000000', '10.59630900']],
'maxscfiterations': ['120'],
'meshcutoff': ['2721.139566', 'eV'],
'numberofatoms': ['3'],
'numberofspecies': ['2'],
'paobasissize': ['dz'],
'solutionmethod': ['diagon'],
'systemlabel': ['H2O'],
'wavefunckpoints': [['0.0', '0.0', '0.0']],
'writedenchar': ['T'],
'xcauthors': ['PBE'],
'xcfunctional': ['GGA']}
"""
return _read_fdf(fname)
def read_struct_out(fname):
"""Read a siesta struct file"""
from ase.atoms import Atoms, Atom
f = fname
cell = []
for i in range(3):
cell.append([float(x) for x in f.readline().split()])
natoms = int(f.readline())
atoms = Atoms()
for atom in f:
Z, pos_x, pos_y, pos_z = atom.split()[1:]
atoms.append(Atom(int(Z),
position=(float(pos_x), float(pos_y), float(pos_z))))
if len(atoms) != natoms:
raise IOError('Badly structured input file')
atoms.set_cell(cell, scale_atoms=True)
return atoms
|
