def parse_cif_pycodcif(fileobj):
"""Parse a CIF file using pycodcif CIF parser"""
blocks = []
if not isinstance(fileobj, basestring):
fileobj = fileobj.name
try:
from pycodcif import parse
except ImportError:
raise ImportError(
'parse_cif_pycodcif requires pycodcif ' +
'(http://wiki.crystallography.net/cod-tools/pycodcif/)')
data, _, _ = parse(fileobj)
for datablock in data:
tags = datablock['values']
for tag in tags.keys():
values = [convert_value(x) for x in tags[tag]]
if len(values) == 1:
tags[tag] = values[0]
else:
tags[tag] = values
blocks.append((datablock['name'], tags))
return blocks
def validate():
if flask.request.method == 'POST':
file = flask.request.files['cif']
filename = ("testfile_" + str(random.randint(1, 1000000)) + ".cif")
file.save(filename)
try:
conf = {}
for option in flask.request.form.items():
conf[option[0]] = 1
data, err_count, err_msg = pycodcif.parse(filename, conf)
data[0]['err_count'] = err_count
data[0]['err_msg'] = err_msg
except Exception as e:
e = str(e).replace("\n", " ")
error = 'Failed to parse the cif file: ' + e
data = [{
'err_msg': error,
}]
try:
os.remove(filename)
except Exception as e:
e = str(e).replace("\n", " ")
error = 'Error: ' + e
data = [{
'err_msg': error,
}]
return json.dumps(data)
def validate_example():
"""
API Route for some example files for validation.
Parameters: Accepts the name of structure.
Function: Validates the file using PyCodCIF.
Response: A JSON Object containing the key 'status' which may hold the values 'valid' or 'error', and the key 'message'.
"""
filename = 'code/webservice/static/' + flask.request.form.get("name")
response = {'status': '', 'message': []}
try:
conf = {}
for option in flask.request.form.items():
conf[option[0]] = 1
data, err_count, err_msg = pycodcif.parse(filename, conf)
response['status'] = 'valid'
response['message'].append(err_msg)
except Exception as e:
# Pycodcif could not parse the file
e = str(e).replace("\n", " ")
error = 'Failed to parse the cif file: ' + e
response['status'] = 'error'
response['message'].append(error)
return json.dumps(response) # Return JSON
def validate():
"""
API Route for validation
Parameters: Accepts a file with key 'cif' via POST
Function: Validates the file using PyCodCIF
Response: A JSON Object containing the key 'status' which may hold the values 'valid' or 'error', and the key 'message'.
"""
file = flask.request.files['cif'] # The file must have the key 'cif'
filename = ("testfile_" + str(random.randint(1, 1000000)) + ".cif"
) # Generate Filename
file.save(filename) # Save the file, PyCodCIF reads only saved files
response = {'status': '', 'message': []}
try:
conf = {}
for option in flask.request.form.items():
conf[option[0]] = 1
data, err_count, err_msg = pycodcif.parse(filename, conf)
response['status'] = 'valid'
response['message'].append(err_msg)
except Exception as e:
# Pycodcif could not parse the file
e = str(e).replace("\n", " ")
error = 'Failed to parse the cif file: ' + e
response['status'] = 'error'
response['message'].append(error)
try:
# Remove saved file
os.remove(filename)
except Exception as e:
e = str(e).replace("\n", " ")
response['message'].append(error)
return json.dumps(response) # Return JSON
def parse_cif_pycodcif(fileobj) -> Iterator[CIFBlock]:
"""Parse a CIF file using pycodcif CIF parser."""
if not isinstance(fileobj, str):
fileobj = fileobj.name
try:
from pycodcif import parse
except ImportError:
raise ImportError(
'parse_cif_pycodcif requires pycodcif ' +
'(http://wiki.crystallography.net/cod-tools/pycodcif/)')
data, _, _ = parse(fileobj)
for datablock in data:
tags = datablock['values']
for tag in tags.keys():
values = [convert_value(x) for x in tags[tag]]
if len(values) == 1:
tags[tag] = values[0]
else:
tags[tag] = values
yield CIFBlock(datablock['name'], tags)
def get_content(file):
datablocks, error_count, error_messages = parse(file)
return datablocks
def cif_to_ase(cif_string):
"""
Naive pycodcif usage
FIXME:
as soon as pycodcif supports CIFs as strings,
the tempfile below should be removed
Args:
cif_string: (str) WYSIWYG
Returns:
ASE atoms (object) *or* None
None *or* error (str)
"""
with tempfile.NamedTemporaryFile(suffix='.cif') as tmp:
tmp.write(cif_string)
tmp.flush()
try:
parsed_cif = parse(tmp.name)[0][0]['values']
except:
return None, 'Invalid or non-standard CIF'
if '_symmetry_int_tables_number' in parsed_cif:
try:
spacegroup = int(parsed_cif['_symmetry_int_tables_number'][0])
except ValueError:
return None, 'Invalid space group info in CIF'
elif '_symmetry_space_group_name_h-m' in parsed_cif:
spacegroup = parsed_cif['_symmetry_space_group_name_h-m'][0].strip(
) # NB ase is very strict to whitespaces in HM symbols, so this is the most frequent error source
if not spacegroup:
return None, 'Empty space group info in CIF'
else:
return None, 'Absent space group info in CIF'
try:
cellpar = (float(parsed_cif['_cell_length_a'][0].split('(')[0]),
float(parsed_cif['_cell_length_b'][0].split('(')[0]),
float(parsed_cif['_cell_length_c'][0].split('(')[0]),
float(parsed_cif['_cell_angle_alpha'][0].split('(')[0]),
float(parsed_cif['_cell_angle_beta'][0].split('(')[0]),
float(parsed_cif['_cell_angle_gamma'][0].split('(')[0]))
basis = np.transpose(
np.array([[
char.split('(')[0]
for char in parsed_cif['_atom_site_fract_x']
],
[
char.split('(')[0]
for char in parsed_cif['_atom_site_fract_y']
],
[
char.split('(')[0]
for char in parsed_cif['_atom_site_fract_z']
]]).astype(np.float))
occupancies = [
float(occ.split('(')[0])
for occ in parsed_cif.get('_atom_site_occupancy', [])
]
except:
return None, 'Unexpected non-numerical values occured in CIF'
symbols = parsed_cif.get('_atom_site_type_symbol')
if not symbols:
symbols = parsed_cif.get('_atom_site_label')
if not symbols:
return None, 'Cannot find atomic positions in CIF'
symbols = [
char.encode('ascii').translate(None, ".0123456789")
for char in symbols
]
occ_data = None
if occupancies and any([occ != 1 for occ in occupancies]):
basis = basis.tolist()
partial_pos, occ_data = {}, {}
for n in range(len(occupancies) - 1, -1, -1):
if occupancies[n] != 1:
disordered_pos = basis.pop(n)
disordered_el = symbols.pop(n)
partial_pos.setdefault(tuple(disordered_pos),
{})[disordered_el] = occupancies[n]
for xyz, occs in partial_pos.items():
index = len(symbols)
symbols.append(sorted(occs.keys())[0])
basis.append(xyz)
occ_data[index] = occs
atom_data = []
for n, xyz in enumerate(basis):
atom_data.append(Atom(symbols[n], tuple(xyz), tag=n))
try:
return crystal(
atom_data,
spacegroup=spacegroup,
cellpar=cellpar,
primitive_cell=True,
onduplicates='error',
info=dict(disordered=occ_data) if occ_data else {}), None
except:
return None, 'Unrecognized sites or invalid site symmetry in CIF'
