This file is indexed.

/usr/share/pyshared/ase/io/abinit.py is in python-ase 3.6.0.2515-1.1.

This file is owned by root:root, with mode 0o644.

The actual contents of the file can be viewed below.

  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
"""
This module contains functionality for reading an ASE
Atoms object in ABINIT input format.

"""

import os

def read_abinit(filename='abinit.in'):
    """Import ABINIT input file.

    Reads cell, atom positions, etc. from abinit input file
    """

    from ase import Atoms, units

    if isinstance(filename, str):
        f = open(filename)
    else: # Assume it's a file-like object
        f = filename

    lines = f.readlines()
    if type(filename) == str:
        f.close()

    full_file = ''
    for line in lines:
        if '#' in line:
            meat, comment = line.split('#')
        else:
            meat = line
        full_file = full_file + meat + ' '

    full_file.strip()
    tokens = full_file.lower().split()

    # note that the file can not be scanned sequentially

    index = tokens.index("acell")
    unit = 1.0
    if(tokens[index+4].lower()[:3] != 'ang'):
        unit = units.Bohr
    acell = [unit*float(tokens[index+1]),
             unit*float(tokens[index+2]),
             unit*float(tokens[index+3])]

    index = tokens.index("natom")
    natom = int(tokens[index+1])

    index = tokens.index("ntypat")
    ntypat = int(tokens[index+1])

    index = tokens.index("typat")
    typat = []
    for i in range(natom):
        typat.append(int(tokens[index+1+i]))

    index = tokens.index("znucl")
    znucl = []
    for i in range(ntypat):
        znucl.append(int(tokens[index+1+i]))

    index = tokens.index("rprim")
    rprim = []
    for i in range(3):
        rprim.append([acell[i]*float(tokens[index+3*i+1]),
                      acell[i]*float(tokens[index+3*i+2]),
                      acell[i]*float(tokens[index+3*i+3])])

    # create a list with the atomic numbers
    numbers = []
    for i in range(natom):
        ii = typat[i] - 1
        numbers.append(znucl[ii])

    # now the positions of the atoms
    if "xred" in tokens:
        index = tokens.index("xred")
        xred = []
        for i in range(natom):
            xred.append([float(tokens[index+3*i+1]),
                         float(tokens[index+3*i+2]),
                         float(tokens[index+3*i+3])])
        atoms = Atoms(cell=rprim, scaled_positions=xred, numbers=numbers)
        return atoms

    index = None
    if "xcart" in tokens:
        index = tokens.index("xcart")
        unit = units.Bohr
    elif "xangs" in tokens:
        unit = 1.0
        index = tokens.index("xangs")

    if(index != None):
        xangs = []
        for i in range(natom):
            xangs.append([unit*float(tokens[index+3*i+1]),
                          unit*float(tokens[index+3*i+2]),
                          unit*float(tokens[index+3*i+3])])
        atoms = Atoms(cell=rprim, positions=xangs, numbers=numbers)
        return atoms

    raise IOError("No xred, xcart, or xangs keyword in abinit input file")


def write_abinit(filename, atoms, cartesian=False, long_format=True):
    """Method to write abinit input files."""

    import numpy as np
    from ase import Atoms, data

    if isinstance(filename, str):
        f = open(filename, 'w')
    else: # Assume it's a 'file-like object'
        f = filename

    if isinstance(atoms, (list, tuple)):
        if len(atoms) > 1:
            raise RuntimeError("Don't know how to save more than "+
                            "one image to input")
        else:
            atoms = atoms[0]

    # Write atom positions in scaled or cartesian coordinates
    if cartesian:
        coord = atoms.get_positions()
    else:
        coord = atoms.get_scaled_positions()

    # let us order the atoms according to chemical symbol
    ind = np.argsort(atoms.get_chemical_symbols())
    symbols = np.array(atoms.get_chemical_symbols())[ind]
    coord = coord[ind]

    # and now we count how many atoms of which type we have
    sc = []
    psym = symbols[0]
    count = 0
    for sym in symbols:
        if sym != psym:
            sc.append((psym, count))
            psym = sym
            count = 1
        else:
            count += 1
    sc.append((psym, count))

    f.write('\n# Definition of the atom types\n')
    f.write("ntypat  " + str(len(sc)) + "\n")
    f.write("znucl  ")
    for specie in sc:
        f.write(str(data.atomic_numbers[specie[0]]) + " ")
    f.write('\n')

    f.write('\n# Definition of the atoms\n')
    f.write('natom  ' + str(len(symbols)) + '\n')
    f.write('typat  ')
    typat = 1
    for specie in sc:
        for natom in range(specie[1]):
            f.write(str(typat) + ' ')
        typat = typat + 1
    f.write('\n')

    f.write('\n# Definition of the unit cell\n')
    f.write('acell\n')
    f.write('%.14f %.14f %.14f Angstrom\n' %  (1.0, 1.0, 1.0))
    f.write('\n')
    f.write('rprim\n')
    if long_format:
        latt_form = ' %21.16f'
    else:
        latt_form = ' %11.6f'

    for vec in atoms.get_cell():
        f.write(' ')
        for el in vec:
            f.write(latt_form % el)
        f.write('\n')
    f.write('\n')

    # Write atom positions in scaled or cartesian coordinates
    if cartesian:
        f.write('xangst\n')
    else:
        f.write('xred\n')

    if long_format:
        cform = ' %19.16f'
    else:
        cform = ' %9.6f'

    for iatom, atom in enumerate(coord):
        f.write(' ')
        for dcoord in atom:
            f.write(cform % dcoord)
        f.write('\n')

    if type(filename) == str:
        f.close()