Coverage for /builds/hweiske/ase/ase/io/nwchem/nwreader.py: 65.57%
273 statements
« prev ^ index » next coverage.py v7.2.7, created at 2024-04-22 11:22 +0000
« prev ^ index » next coverage.py v7.2.7, created at 2024-04-22 11:22 +0000
1import re
2from collections import OrderedDict
4import numpy as np
6from ase import Atoms
7from ase.calculators.singlepoint import (SinglePointDFTCalculator,
8 SinglePointKPoint)
9from ase.units import Bohr, Hartree
11from .parser import _define_pattern
13# Note to the reader of this code: Here and below we use the function
14# _define_pattern from parser.py in this same directory to compile
15# regular expressions. These compiled expressions are stored along with
16# an example string that the expression should match in a list that
17# is used during tests (test/nwchem/nwchem_parser.py) to ensure that
18# the regular expressions are still working correctly.
20# Matches the beginning of a GTO calculation
21_gauss_block = _define_pattern(
22 r'^[\s]+NWChem (?:SCF|DFT) Module\n$',
23 " NWChem SCF Module\n",
24)
27# Matches the beginning of a plane wave calculation
28_pw_block = _define_pattern(
29 r'^[\s]+\*[\s]+NWPW (?:PSPW|BAND|PAW|Band Structure) Calculation'
30 r'[\s]+\*[\s]*\n$',
31 " * NWPW PSPW Calculation *\n",
32)
35# Top-level parser
36def read_nwchem_out(fobj, index=-1):
37 """Splits an NWChem output file into chunks corresponding to
38 individual single point calculations."""
39 lines = fobj.readlines()
41 if index == slice(-1, None, None):
42 for line in lines:
43 if _gauss_block.match(line):
44 return [parse_gto_chunk(''.join(lines))]
45 if _pw_block.match(line):
46 return [parse_pw_chunk(''.join(lines))]
47 raise ValueError('This does not appear to be a valid NWChem '
48 'output file.')
50 # First, find each SCF block
51 group = []
52 atomslist = []
53 header = True
54 lastgroup = []
55 lastparser = None
56 parser = None
57 for line in lines:
58 group.append(line)
59 if _gauss_block.match(line):
60 next_parser = parse_gto_chunk
61 elif _pw_block.match(line):
62 next_parser = parse_pw_chunk
63 else:
64 continue
66 if header:
67 header = False
68 else:
69 atoms = parser(''.join(group))
70 if atoms is None and parser is lastparser:
71 atoms = parser(''.join(lastgroup + group))
72 if atoms is not None:
73 atomslist[-1] = atoms
74 lastgroup += group
75 else:
76 atomslist.append(atoms)
77 lastgroup = group
78 lastparser = parser
79 group = []
80 parser = next_parser
81 if not header:
82 atoms = parser(''.join(group))
83 if atoms is not None:
84 atomslist.append(atoms)
86 return atomslist[index]
89# Matches a geometry block and returns the geometry specification lines
90_geom = _define_pattern(
91 r'\n[ \t]+Geometry \"[ \t\S]+\" -> \"[ \t\S]*\"[ \t]*\n'
92 r'^[ \t-]+\n'
93 r'(?:^[ \t\S]*\n){3}'
94 r'^[ \t]+No\.[ \t]+Tag[ \t]+Charge[ \t]+X[ \t]+Y[ \t]+Z\n'
95 r'^[ \t-]+\n'
96 r'((?:^(?:[ \t]+[\S]+){6}[ \t]*\n)+)',
97 """\
99 Geometry "geometry" -> ""
100 -------------------------
102 Output coordinates in angstroms (scale by 1.889725989 to convert to a.u.)
104 No. Tag Charge X Y Z
105 ---- ---------------- ---------- -------------- -------------- --------------
106 1 C 6.0000 0.00000000 0.00000000 0.00000000
107 2 H 1.0000 0.62911800 0.62911800 0.62911800
108 3 H 1.0000 -0.62911800 -0.62911800 0.62911800
109 4 H 1.0000 0.62911800 -0.62911800 -0.62911800
110""", re.M)
112# Unit cell parser
113_cell_block = _define_pattern(r'^[ \t]+Lattice Parameters[ \t]*\n'
114 r'^(?:[ \t\S]*\n){4}'
115 r'((?:^(?:[ \t]+[\S]+){5}\n){3})',
116 """\
117 Lattice Parameters
118 ------------------
120 lattice vectors in angstroms (scale by 1.889725989 to convert to a.u.)
122 a1=< 4.000 0.000 0.000 >
123 a2=< 0.000 5.526 0.000 >
124 a3=< 0.000 0.000 4.596 >
125 a= 4.000 b= 5.526 c= 4.596
126 alpha= 90.000 beta= 90.000 gamma= 90.000
127 omega= 101.6
128""", re.M)
131# Parses the geometry and returns the corresponding Atoms object
132def _parse_geomblock(chunk):
133 geomblocks = _geom.findall(chunk)
134 if not geomblocks:
135 return None
136 geomblock = geomblocks[-1].strip().split('\n')
137 natoms = len(geomblock)
138 symbols = []
139 pos = np.zeros((natoms, 3))
140 for i, line in enumerate(geomblock):
141 line = line.strip().split()
142 symbols.append(line[1])
143 pos[i] = [float(x) for x in line[3:6]]
145 cellblocks = _cell_block.findall(chunk)
146 if cellblocks:
147 cellblock = cellblocks[-1].strip().split('\n')
148 cell = np.zeros((3, 3))
149 for i, line in enumerate(cellblock):
150 line = line.strip().split()
151 cell[i] = [float(x) for x in line[1:4]]
152 else:
153 cell = None
154 return Atoms(symbols, positions=pos, cell=cell)
157# GTO-specific parser stuff
159# Matches gradient block from a GTO calculation
160_gto_grad = _define_pattern(
161 r'^[ \t]+[\S]+[ \t]+ENERGY GRADIENTS[ \t]*[\n]+'
162 r'^[ \t]+atom[ \t]+coordinates[ \t]+gradient[ \t]*\n'
163 r'^(?:[ \t]+x[ \t]+y[ \t]+z){2}[ \t]*\n'
164 r'((?:^(?:[ \t]+[\S]+){8}\n)+)[ \t]*\n',
165 """\
166 UHF ENERGY GRADIENTS
168 atom coordinates gradient
169 x y z x y z
170 1 C 0.293457 -0.293457 0.293457 -0.000083 0.000083 -0.000083
171 2 H 1.125380 1.355351 1.125380 0.000086 0.000089 0.000086
172 3 H -1.355351 -1.125380 1.125380 -0.000089 -0.000086 0.000086
173 4 H 1.125380 -1.125380 -1.355351 0.000086 -0.000086 -0.000089
175""", re.M)
177# Energy parsers for a variety of different GTO calculations
178_e_gto = OrderedDict()
179_e_gto['tce'] = _define_pattern(
180 r'^[\s]+[\S]+[\s]+total energy \/ hartree[\s]+'
181 r'=[\s]+([\S]+)[\s]*\n',
182 " CCD total energy / hartree "
183 "= -75.715332545665888\n", re.M,
184)
185_e_gto['ccsd'] = _define_pattern(
186 r'^[\s]+Total CCSD energy:[\s]+([\S]+)[\s]*\n',
187 " Total CCSD energy: -75.716168566598569\n",
188 re.M,
189)
190_e_gto['tddft'] = _define_pattern(
191 r'^[\s]+Excited state energy =[\s]+([\S]+)[\s]*\n',
192 " Excited state energy = -75.130134499965\n",
193 re.M,
194)
195_e_gto['mp2'] = _define_pattern(
196 r'^[\s]+Total MP2 energy[\s]+([\S]+)[\s]*\n',
197 " Total MP2 energy -75.708800087578\n",
198 re.M,
199)
200_e_gto['mf'] = _define_pattern(
201 r'^[\s]+Total (?:DFT|SCF) energy =[\s]+([\S]+)[\s]*\n',
202 " Total SCF energy = -75.585555997789\n",
203 re.M,
204)
207# GTO parser
208def parse_gto_chunk(chunk):
209 atoms = None
210 forces = None
211 energy = None
212 dipole = None
213 quadrupole = None
214 for theory, pattern in _e_gto.items():
215 matches = pattern.findall(chunk)
216 if matches:
217 energy = float(matches[-1].replace('D', 'E')) * Hartree
218 break
220 gradblocks = _gto_grad.findall(chunk)
221 if gradblocks:
222 gradblock = gradblocks[-1].strip().split('\n')
223 natoms = len(gradblock)
224 symbols = []
225 pos = np.zeros((natoms, 3))
226 forces = np.zeros((natoms, 3))
227 for i, line in enumerate(gradblock):
228 line = line.strip().split()
229 symbols.append(line[1])
230 pos[i] = [float(x) for x in line[2:5]]
231 forces[i] = [-float(x) for x in line[5:8]]
232 pos *= Bohr
233 forces *= Hartree / Bohr
234 atoms = Atoms(symbols, positions=pos)
236 dipole, quadrupole = _get_multipole(chunk)
238 kpts = _get_gto_kpts(chunk)
240 if atoms is None:
241 atoms = _parse_geomblock(chunk)
243 if atoms is None:
244 return None
246 # SinglePointDFTCalculator doesn't support quadrupole moment currently
247 calc = SinglePointDFTCalculator(atoms=atoms,
248 energy=energy,
249 free_energy=energy, # XXX Is this right?
250 forces=forces,
251 dipole=dipole,
252 # quadrupole=quadrupole,
253 )
254 calc.kpts = kpts
255 atoms.calc = calc
256 return atoms
259# Extracts dipole and quadrupole moment for a GTO calculation
260# Note on the regex: Some, but not all, versions of NWChem
261# insert extra spaces in the blank lines. Do not remove the \s*
262# in between \n and \n
263_multipole = _define_pattern(
264 r'^[ \t]+Multipole analysis of the density[ \t\S]*\n'
265 r'^[ \t-]+\n\s*\n^[ \t\S]+\n^[ \t-]+\n'
266 r'((?:(?:(?:[ \t]+[\S]+){7,8}\n)|[ \t]*\n){12})',
267 """\
268 Multipole analysis of the density
269 ---------------------------------
271 L x y z total alpha beta nuclear
272 - - - - ----- ----- ---- -------
273 0 0 0 0 -0.000000 -5.000000 -5.000000 10.000000
275 1 1 0 0 0.000000 0.000000 0.000000 0.000000
276 1 0 1 0 -0.000001 -0.000017 -0.000017 0.000034
277 1 0 0 1 -0.902084 -0.559881 -0.559881 0.217679
279 2 2 0 0 -5.142958 -2.571479 -2.571479 0.000000
280 2 1 1 0 -0.000000 -0.000000 -0.000000 0.000000
281 2 1 0 1 0.000000 0.000000 0.000000 0.000000
282 2 0 2 0 -3.153324 -3.807308 -3.807308 4.461291
283 2 0 1 1 0.000001 -0.000009 -0.000009 0.000020
284 2 0 0 2 -4.384288 -3.296205 -3.296205 2.208122
285""", re.M)
288# Parses the dipole and quadrupole moment from a GTO calculation
289def _get_multipole(chunk):
290 matches = _multipole.findall(chunk)
291 if not matches:
292 return None, None
293 # This pulls the 5th column out of the multipole moments block;
294 # this column contains the actual moments.
295 moments = [float(x.split()[4]) for x in matches[-1].split('\n')
296 if x and not x.isspace()]
297 dipole = np.array(moments[1:4]) * Bohr
298 quadrupole = np.zeros(9)
299 quadrupole[[0, 1, 2, 4, 5, 8]] = [moments[4:]]
300 quadrupole[[3, 6, 7]] = quadrupole[[1, 2, 5]]
301 return dipole, quadrupole.reshape((3, 3)) * Bohr**2
304# MO eigenvalue and occupancy parser for GTO calculations
305_eval_block = _define_pattern(
306 r'^[ \t]+[\S]+ Final (?:Alpha |Beta )?Molecular Orbital Analysis[ \t]*'
307 r'\n^[ \t-]+\n\n'
308 r'(?:^[ \t]+Vector [ \t\S]+\n(?:^[ \t\S]+\n){3}'
309 r'(?:^(?:(?:[ \t]+[\S]+){5}){1,2}[ \t]*\n)+\n)+',
310 """\
311 ROHF Final Molecular Orbital Analysis
312 -------------------------------------
314 Vector 1 Occ=2.000000D+00 E=-2.043101D+01
315 MO Center= 1.1D-20, 1.5D-18, 1.2D-01, r^2= 1.5D-02
316 Bfn. Coefficient Atom+Function Bfn. Coefficient Atom+Function
317 ----- ------------ --------------- ----- ------------ ---------------
318 1 0.983233 1 O s
320 Vector 2 Occ=2.000000D+00 E=-1.324439D+00
321 MO Center= -2.1D-18, -8.6D-17, -7.1D-02, r^2= 5.1D-01
322 Bfn. Coefficient Atom+Function Bfn. Coefficient Atom+Function
323 ----- ------------ --------------- ----- ------------ ---------------
324 6 0.708998 1 O s 1 -0.229426 1 O s
325 2 0.217752 1 O s
326 """, re.M) # noqa: W291
329# Parses the eigenvalues and occupations from a GTO calculation
330def _get_gto_kpts(chunk):
331 eval_blocks = _eval_block.findall(chunk)
332 if not eval_blocks:
333 return []
334 kpts = []
335 kpt = _get_gto_evals(eval_blocks[-1])
336 if kpt.s == 1:
337 kpts.append(_get_gto_evals(eval_blocks[-2]))
338 kpts.append(kpt)
339 return kpts
342# Extracts MO eigenvalue and occupancy for a GTO calculation
343_extract_vector = _define_pattern(
344 r'^[ \t]+Vector[ \t]+([\S])+[ \t]+Occ=([\S]+)[ \t]+E=[ \t]*([\S]+)[ \t]*\n',
345 " Vector 1 Occ=2.000000D+00 E=-2.043101D+01\n", re.M,
346)
349# Extracts the eigenvalues and occupations from a GTO calculation
350def _get_gto_evals(chunk):
351 spin = 1 if re.match(r'[ \t\S]+Beta', chunk) else 0
352 data = []
353 for vector in _extract_vector.finditer(chunk):
354 data.append([float(x.replace('D', 'E')) for x in vector.groups()[1:]])
355 data = np.array(data)
356 occ = data[:, 0]
357 energies = data[:, 1] * Hartree
359 return SinglePointKPoint(1., spin, 0, energies, occ)
362# Plane wave specific parsing stuff
364# Matches the gradient block from a plane wave calculation
365_nwpw_grad = _define_pattern(
366 r'^[ \t]+[=]+[ \t]+Ion Gradients[ \t]+[=]+[ \t]*\n'
367 r'^[ \t]+Ion Forces:[ \t]*\n'
368 r'((?:^(?:[ \t]+[\S]+){7}\n)+)',
369 """\
370 ============= Ion Gradients =================
371 Ion Forces:
372 1 O ( -0.000012 0.000027 -0.005199 )
373 2 H ( 0.000047 -0.013082 0.020790 )
374 3 H ( 0.000047 0.012863 0.020786 )
375 C.O.M. ( -0.000000 -0.000000 -0.000000 )
376 ===============================================
377""", re.M)
379# Matches the gradient block from a PAW calculation
380_paw_grad = _define_pattern(
381 r'^[ \t]+[=]+[ \t]+Ion Gradients[ \t]+[=]+[ \t]*\n'
382 r'^[ \t]+Ion Positions:[ \t]*\n'
383 r'((?:^(?:[ \t]+[\S]+){7}\n)+)'
384 r'^[ \t]+Ion Forces:[ \t]*\n'
385 r'((?:^(?:[ \t]+[\S]+){7}\n)+)',
386 """\
387 ============= Ion Gradients =================
388 Ion Positions:
389 1 O ( -3.77945 -5.22176 -3.77945 )
390 2 H ( -3.77945 -3.77945 3.77945 )
391 3 H ( -3.77945 3.77945 3.77945 )
392 Ion Forces:
393 1 O ( -0.00001 -0.00000 0.00081 )
394 2 H ( 0.00005 -0.00026 -0.00322 )
395 3 H ( 0.00005 0.00030 -0.00322 )
396 C.O.M. ( -0.00000 -0.00000 -0.00000 )
397 ===============================================
398""", re.M)
400# Energy parser for plane wave calculations
401_nwpw_energy = _define_pattern(
402 r'^[\s]+Total (?:PSPW|BAND|PAW) energy'
403 r'[\s]+:[\s]+([\S]+)[\s]*\n',
404 " Total PSPW energy : -0.1709317826E+02\n",
405 re.M,
406)
408# Parser for the fermi energy in a plane wave calculation
409_fermi_energy = _define_pattern(
410 r'^[ \t]+Fermi energy =[ \t]+([\S]+) \([ \t]+[\S]+[ \t]*\n',
411 " Fermi energy = -0.5585062E-01 ( -1.520eV)\n", re.M,
412)
415# Plane wave parser
416def parse_pw_chunk(chunk):
417 atoms = _parse_geomblock(chunk)
418 if atoms is None:
419 return None
421 energy = None
422 efermi = None
423 forces = None
424 stress = None
426 matches = _nwpw_energy.findall(chunk)
427 if matches:
428 energy = float(matches[-1].replace('D', 'E')) * Hartree
430 matches = _fermi_energy.findall(chunk)
431 if matches:
432 efermi = float(matches[-1].replace('D', 'E')) * Hartree
434 gradblocks = _nwpw_grad.findall(chunk)
435 if not gradblocks:
436 gradblocks = _paw_grad.findall(chunk)
437 if gradblocks:
438 gradblock = gradblocks[-1].strip().split('\n')
439 natoms = len(gradblock)
440 symbols = []
441 forces = np.zeros((natoms, 3))
442 for i, line in enumerate(gradblock):
443 line = line.strip().split()
444 symbols.append(line[1])
445 forces[i] = [float(x) for x in line[3:6]]
446 forces *= Hartree / Bohr
448 if atoms.cell:
449 stress = _get_stress(chunk, atoms.cell)
451 ibz_kpts, kpts = _get_pw_kpts(chunk)
453 # NWChem does not calculate an energy extrapolated to the 0K limit,
454 # so right now, energy and free_energy will be the same.
455 calc = SinglePointDFTCalculator(atoms=atoms,
456 energy=energy,
457 efermi=efermi,
458 free_energy=energy,
459 forces=forces,
460 stress=stress,
461 ibzkpts=ibz_kpts)
462 calc.kpts = kpts
463 atoms.calc = calc
464 return atoms
467# Extracts stress tensor from a plane wave calculation
468_stress = _define_pattern(
469 r'[ \t]+[=]+[ \t]+(?:total gradient|E all FD)[ \t]+[=]+[ \t]*\n'
470 r'^[ \t]+S =((?:(?:[ \t]+[\S]+){5}\n){3})[ \t=]+\n',
471 """\
472 ============= total gradient ==============
473 S = ( -0.22668 0.27174 0.19134 )
474 ( 0.23150 -0.26760 0.23226 )
475 ( 0.19090 0.27206 -0.22700 )
476 ===================================================
477""", re.M)
480# Extract stress tensor from a plane wave calculation
481def _get_stress(chunk, cell):
482 stress_blocks = _stress.findall(chunk)
483 if not stress_blocks:
484 return None
485 stress_block = stress_blocks[-1]
486 stress = np.zeros((3, 3))
487 for i, row in enumerate(stress_block.strip().split('\n')):
488 stress[i] = [float(x) for x in row.split()[1:4]]
489 stress = (stress @ cell) * Hartree / Bohr / cell.volume
490 stress = 0.5 * (stress + stress.T)
491 # convert from 3x3 array to Voigt form
492 return stress.ravel()[[0, 4, 8, 5, 2, 1]]
495# MO/band eigenvalue and occupancy parser for plane wave calculations
496_nwpw_eval_block = _define_pattern(
497 r'(?:(?:^[ \t]+Brillouin zone point:[ \t]+[\S]+[ \t]*\n'
498 r'(?:[ \t\S]*\n){3,4})?'
499 r'^[ \t]+(?:virtual )?orbital energies:\n'
500 r'(?:^(?:(?:[ \t]+[\S]+){3,4}){1,2}[ \t]*\n)+\n{,3})+',
501 """\
502 Brillouin zone point: 1
503 weight= 0.074074
504 k =< 0.333 0.333 0.333> . <b1,b2,b3>
505 =< 0.307 0.307 0.307>
507 orbital energies:
508 0.3919370E+00 ( 10.665eV) occ=1.000
509 0.3908827E+00 ( 10.637eV) occ=1.000 0.4155535E+00 ( 11.308eV) occ=1.000
510 0.3607689E+00 ( 9.817eV) occ=1.000 0.3827820E+00 ( 10.416eV) occ=1.000
511 0.3544000E+00 ( 9.644eV) occ=1.000 0.3782641E+00 ( 10.293eV) occ=1.000
512 0.3531137E+00 ( 9.609eV) occ=1.000 0.3778819E+00 ( 10.283eV) occ=1.000
513 0.2596367E+00 ( 7.065eV) occ=1.000 0.2820723E+00 ( 7.676eV) occ=1.000
515 Brillouin zone point: 2
516 weight= 0.074074
517 k =< -0.000 0.333 0.333> . <b1,b2,b3>
518 =< 0.614 0.000 0.000>
520 orbital energies:
521 0.3967132E+00 ( 10.795eV) occ=1.000
522 0.3920006E+00 ( 10.667eV) occ=1.000 0.4197952E+00 ( 11.423eV) occ=1.000
523 0.3912442E+00 ( 10.646eV) occ=1.000 0.4125086E+00 ( 11.225eV) occ=1.000
524 0.3910472E+00 ( 10.641eV) occ=1.000 0.4124238E+00 ( 11.223eV) occ=1.000
525 0.3153977E+00 ( 8.582eV) occ=1.000 0.3379797E+00 ( 9.197eV) occ=1.000
526 0.2801606E+00 ( 7.624eV) occ=1.000 0.3052478E+00 ( 8.306eV) occ=1.000
527""", re.M) # noqa: E501, W291
529# Parser for kpoint weights for a plane wave calculation
530_kpt_weight = _define_pattern(
531 r'^[ \t]+Brillouin zone point:[ \t]+([\S]+)[ \t]*\n'
532 r'^[ \t]+weight=[ \t]+([\S]+)[ \t]*\n',
533 """\
534 Brillouin zone point: 1
535 weight= 0.074074
536""", re.M) # noqa: W291
539# Parse eigenvalues and occupancies from a plane wave calculation
540def _get_pw_kpts(chunk):
541 eval_blocks = []
542 for block in _nwpw_eval_block.findall(chunk):
543 if 'pathlength' not in block:
544 eval_blocks.append(block)
545 if not eval_blocks:
546 return []
547 if 'virtual' in eval_blocks[-1]:
548 occ_block = eval_blocks[-2]
549 virt_block = eval_blocks[-1]
550 else:
551 occ_block = eval_blocks[-1]
552 virt_block = ''
553 kpts = NWChemKpts()
554 _extract_pw_kpts(occ_block, kpts, 1.)
555 _extract_pw_kpts(virt_block, kpts, 0.)
556 for match in _kpt_weight.finditer(occ_block):
557 index, weight = match.groups()
558 kpts.set_weight(index, float(weight))
559 return kpts.to_ibz_kpts(), kpts.to_singlepointkpts()
562# Helper class for keeping track of kpoints and converting to
563# SinglePointKPoint objects.
564class NWChemKpts:
565 def __init__(self):
566 self.data = {}
567 self.ibz_kpts = {}
568 self.weights = {}
570 def add_ibz_kpt(self, index, raw_kpt):
571 kpt = np.array([float(x.strip('>')) for x in raw_kpt.split()[1:4]])
572 self.ibz_kpts[index] = kpt
574 def add_eval(self, index, spin, energy, occ):
575 if index not in self.data:
576 self.data[index] = {}
577 if spin not in self.data[index]:
578 self.data[index][spin] = []
579 self.data[index][spin].append((energy, occ))
581 def set_weight(self, index, weight):
582 self.weights[index] = weight
584 def to_ibz_kpts(self):
585 if not self.ibz_kpts:
586 return np.array([[0., 0., 0.]])
587 sorted_kpts = sorted(list(self.ibz_kpts.items()), key=lambda x: x[0])
588 return np.array(list(zip(*sorted_kpts))[1])
590 def to_singlepointkpts(self):
591 kpts = []
592 for i, (index, spins) in enumerate(self.data.items()):
593 weight = self.weights[index]
594 for spin, (_, data) in enumerate(spins.items()):
595 energies, occs = np.array(sorted(data, key=lambda x: x[0])).T
596 kpts.append(SinglePointKPoint(weight, spin, i, energies, occs))
597 return kpts
600# Extracts MO/band data from a pattern matched by _nwpw_eval_block above
601_kpt = _define_pattern(
602 r'^[ \t]+Brillouin zone point:[ \t]+([\S]+)[ \t]*\n'
603 r'^[ \t]+weight=[ \t]+([\S])+[ \t]*\n'
604 r'^[ \t]+k[ \t]+([ \t\S]+)\n'
605 r'(?:^[ \t\S]*\n){1,2}'
606 r'^[ \t]+(?:virtual )?orbital energies:\n'
607 r'((?:^(?:(?:[ \t]+[\S]+){3,4}){1,2}[ \t]*\n)+)',
608 """\
609 Brillouin zone point: 1
610 weight= 0.074074
611 k =< 0.333 0.333 0.333> . <b1,b2,b3>
612 =< 0.307 0.307 0.307>
614 orbital energies:
615 0.3919370E+00 ( 10.665eV) occ=1.000
616 0.3908827E+00 ( 10.637eV) occ=1.000 0.4155535E+00 ( 11.308eV) occ=1.000
617 0.3607689E+00 ( 9.817eV) occ=1.000 0.3827820E+00 ( 10.416eV) occ=1.000
618 0.3544000E+00 ( 9.644eV) occ=1.000 0.3782641E+00 ( 10.293eV) occ=1.000
619 0.3531137E+00 ( 9.609eV) occ=1.000 0.3778819E+00 ( 10.283eV) occ=1.000
620 0.2596367E+00 ( 7.065eV) occ=1.000 0.2820723E+00 ( 7.676eV) occ=1.000
621""", re.M) # noqa: E501, W291
624# Extracts kpoints from a plane wave calculation
625def _extract_pw_kpts(chunk, kpts, default_occ):
626 for match in _kpt.finditer(chunk):
627 point, weight, raw_kpt, orbitals = match.groups()
628 index = int(point) - 1
629 for line in orbitals.split('\n'):
630 tokens = line.strip().split()
631 if not tokens:
632 continue
633 ntokens = len(tokens)
634 a_e = float(tokens[0]) * Hartree
635 if ntokens % 3 == 0:
636 a_o = default_occ
637 else:
638 a_o = float(tokens[3].split('=')[1])
639 kpts.add_eval(index, 0, a_e, a_o)
641 if ntokens <= 4:
642 continue
643 if ntokens == 6:
644 b_e = float(tokens[3]) * Hartree
645 b_o = default_occ
646 elif ntokens == 8:
647 b_e = float(tokens[4]) * Hartree
648 b_o = float(tokens[7].split('=')[1])
649 kpts.add_eval(index, 1, b_e, b_o)
650 kpts.set_weight(index, float(weight))
651 kpts.add_ibz_kpt(index, raw_kpt)