def get_string(self, df_head_only=False):
from pymatgen import Structure
lines, scope = [], []
for key, value in self.document.iterate():
if isinstance(value, Table):
lines[-1] = lines[-1].replace("{", "[+").replace("}", "]")
header = any([isinstance(col, str) for col in value])
if isinstance(value.index, MultiIndex):
value.reset_index(inplace=True)
if df_head_only:
value = value.head()
csv_string = value.to_csv(index=False,
header=header,
float_format="%g",
encoding="utf-8")[:-1]
lines += csv_string.split("\n")
if df_head_only:
lines.append("...")
elif isinstance(value, Structure):
from pymatgen.io.cif import CifWriter
cif = CifWriter(value, symprec=1e-10).__str__()
lines.append(
make_pair("".join([replacements.get(c, c) for c in key]),
cif + ":end"))
elif Quantity is not None and isinstance(value, Quantity):
lines.append(
make_pair(value.display_symbols[0], value.pretty_string()))
else:
level, key = key
# truncate scope
level_reduction = bool(level < len(scope))
if level_reduction:
del scope[level:]
# append scope
if value is None:
scope.append("".join([replacements.get(c, c)
for c in key]))
# correct scope to omit internal 'general' section
scope_corr = scope
if scope[0] == mp_level01_titles[0]:
scope_corr = scope[1:]
# insert scope line
if (value is None and scope_corr) or (value is not None
and level_reduction):
lines.append("\n{" + ".".join(scope_corr) + "}")
# insert key-value line
if value is not None:
val = str(value)
value_lines = ([val] if val.startswith("http") else
textwrap.wrap(val))
if len(value_lines) > 1:
value_lines = [""] + value_lines + [":end"]
lines.append(
make_pair(
"".join([replacements.get(c, c) for c in key]),
"\n".join(value_lines),
))
return "\n".join(lines) + "\n"
def get_string(self, df_head_only=False):
from pymatgen import Structure
lines, scope = [], []
for key,value in self.document.iterate():
if isinstance(value, Table):
lines[-1] = lines[-1].replace('{', '[+').replace('}', ']')
header = any([isinstance(col, str) for col in value])
if isinstance(value.index, MultiIndex):
value.reset_index(inplace=True)
if df_head_only:
value = value.head()
csv_string = value.to_csv(
index=False, header=header, float_format='%g', encoding='utf-8'
)[:-1]
lines += csv_string.split('\n')
if df_head_only:
lines.append('...')
elif isinstance(value, Structure):
from pymatgen.io.cif import CifWriter
cif = CifWriter(value, symprec=symprec).__str__()
lines.append(make_pair(
''.join([replacements.get(c, c) for c in key]), cif+':end'
))
elif Quantity is not None and isinstance(value, Quantity):
lines.append(make_pair(
value.display_symbols[0], value.pretty_string()
))
else:
level, key = key
# truncate scope
level_reduction = bool(level < len(scope))
if level_reduction:
del scope[level:]
# append scope
if value is None:
scope.append(''.join([
replacements.get(c, c) for c in key
]))
# correct scope to omit internal 'general' section
scope_corr = scope
if scope[0] == mp_level01_titles[0]:
scope_corr = scope[1:]
# insert scope line
if (value is None and scope_corr) or \
(value is not None and level_reduction):
lines.append('\n{' + '.'.join(scope_corr) + '}')
# insert key-value line
if value is not None:
val = str(value)
value_lines = [val] if val.startswith('http') \
else textwrap.wrap(val)
if len(value_lines) > 1:
value_lines = [''] + value_lines + [':end']
lines.append(make_pair(
''.join([replacements.get(c, c) for c in key]),
'\n'.join(value_lines)
))
return '\n'.join(lines) + '\n'
def get_string(self, with_comments=False):
lines = []
min_indentor = get_indentor()
table_start = mp_level01_titles[1]+' '
for key,value in self.document.iterate():
if key is None and isinstance(value, dict):
pd_obj = pandas.DataFrame.from_dict(value)
csv_string = pd_obj.to_csv(index=False, float_format='%g')[:-1]
lines += csv_string.split('\n')
else:
level, key = key
key = get_indentor(n=key) if value is None else key
sep = '' if min_indentor in key else ':'
if lines and key == min_indentor:
lines.append('')
if isinstance(value, six.string_types):
if value.startswith(table_start):
value = value[len(table_start):]
if ':' in value: # quote to ignore delimiter
value = '"{}"'.format(value)
lines.append(make_pair(key, value, sep=sep))
if with_comments:
for idx_str, comment in self.comments.iteritems():
idx, ast = self.get_comment_index(idx_str)
if ast: lines.insert(idx, '#'+comment)
else: lines[idx] = ' #'.join([lines[idx], comment])
return '\n'.join(lines) + '\n'
def get_key_value(self):
"""print random key-value pair
- type(key) = str, type(value) = anything
- append comment now and then
"""
while 1:
provider_name = self.fake.random_element(elements=self.providers)
if provider_name != 'python' and \
provider_name != 'profile' and \
provider_name != 'credit_card':
break
provider = self.fake.provider(provider_name)
methods = [
k for k, v in inspect.getmembers(provider,
predicate=inspect.ismethod)
if k != '__init__'
]
while 1:
method_name = self.fake.random_element(elements=methods)
method = getattr(provider, method_name)
argspec = inspect.getargspec(method)
nargs = len(argspec.args)
key = '_'.join([provider_name, method_name])
if (argspec.defaults is None
and nargs == 1) or (argspec.defaults is not None and
nargs - 1 == len(argspec.defaults)):
value = method()
if not isinstance(value, list) and \
not isinstance(value, dict) and \
not ':' in str(value):
break
if isinstance(value, str) and '\n' in value:
value = repr(value)
return make_pair(key, value)
def get_nested_key_values_from_dict(self, d, n=1):
"""convert a dict into nested level-n mp-csv representation"""
for k0,v0 in d.iteritems():
print >>self.section, ' '.join([
get_indentor(n+1), k0, self.get_comment()
])
for k1,v1 in v0.iteritems():
print >>self.section, make_pair(k1, v1)
def get_nested_key_values_from_dict(self, d, n=1):
"""convert a dict into nested level-n mp-csv representation"""
for k0, v0 in d.iteritems():
print >> self.section, ' '.join(
[get_indentor(n + 1), k0,
self.get_comment()])
for k1, v1 in v0.iteritems():
print >> self.section, make_pair(k1, v1)
def _make_level_n_section(self, sec, n):
"""recursively generate nested level-n section
- level-1 mp_level01_titles allowed once per level-0 section
- mp_level01_titles[0] can be arbitrarily deep nested ('general')
- mp_level01_titles[1:] has no subsections ('data')
- mp_level01_titles[2] only has level-2 subsections ('plots')
- mp_level01_titles[1] has csv format, all others key:value
- randomly throw in comment lines
"""
comments = self.get_comments()
if comments != '': print >>self.section, comments
print >>self.section, self._get_level_n_section_line(sec, n)
if self.usable and self.main_general and sec == 0 and n == 0:
self.get_player_general_section(n)
comment = self.get_comment()
if comment != '': print >>self.section, comment
num_subsec = 0 if n == self.max_level or \
self.section_titles[-1] == mp_level01_titles[1] or \
( n == 2 and self.section_titles[-2] == mp_level01_titles[2] ) \
else self.fake.random_int(
min=int(n == 1 and \
self.section_titles[-1] == mp_level01_titles[2]),
max=self.max_num_subsec
)
for i in range(num_subsec):
self._make_level_n_section(sec, n+1)
self.section_structure.append('.'.join(self.section_titles))
self.section_titles.pop()
# all subsections processed
if num_subsec == 0:
if self.section_titles[-1] == mp_level01_titles[1] or (
n == 0 and \
self.section_titles[-1] != mp_level01_titles[0].upper()
):
if self.usable:
self.data_gen.player_data.to_csv(self.section, index=False)
else:
print >>self.section, ' ==> data insert here if --usable'
elif n == 2 and self.section_titles[-2] == mp_level01_titles[2]:
column = self.fake.random_element(
elements=self.data_gen.player_data.columns
)
print >>self.section, make_pair('x', column)
elif self.usable and not self.main_general and \
self.section_titles[-1] == mp_level01_titles[0]:
self.get_player_general_section(n)
elif n != 0 or (
n == 0 and \
self.section_titles[-1] != mp_level01_titles[0].upper()
):
for r in range(self.max_data_rows): self._print_key_value()
def get_key_value(self):
"""print random key-value pair
- type(key) = str, type(value) = anything
- append comment now and then
"""
while 1:
provider_name = self.fake.random_element(elements=self.providers)
if provider_name != 'python' and \
provider_name != 'profile' and \
provider_name != 'credit_card':
break
provider = self.fake.provider(provider_name)
methods = [
k for k,v in inspect.getmembers(
provider, predicate=inspect.ismethod
) if k != '__init__'
]
while 1:
method_name = self.fake.random_element(elements=methods)
method = getattr(provider, method_name)
argspec = inspect.getargspec(method)
nargs = len(argspec.args)
key = '_'.join([provider_name, method_name])
if ( argspec.defaults is None and nargs == 1 ) or (
argspec.defaults is not None and
nargs-1 == len(argspec.defaults)
):
value = method()
if not isinstance(value, list) and \
not isinstance(value, dict) and \
not ':' in str(value):
break
if isinstance(value, str) and '\n' in value:
value = repr(value)
return make_pair(key, value)
def get_string(self):
from pymatgen import Structure
lines, scope = [], []
table_start = mp_level01_titles[1] + '_'
for key, value in self.document.iterate():
if isinstance(value, Table):
header = any([
bool(isinstance(col, unicode) or isinstance(col, str))
for col in value
])
if isinstance(value.index, MultiIndex):
value.reset_index(inplace=True)
csv_string = value.to_csv(index=False,
header=header,
float_format='%g',
encoding='utf-8')[:-1]
lines += csv_string.decode('utf-8').split('\n')
elif isinstance(value, Structure):
from pymatgen.io.cif import CifWriter
cif = CifWriter(value, symprec=symprec).__str__()
lines.append(
make_pair(''.join([replacements.get(c, c) for c in key]),
cif + ':end'))
else:
level, key = key
key = key if isinstance(key, unicode) else key.decode('utf-8')
# truncate scope
level_reduction = bool(level < len(scope))
if level_reduction: del scope[level:]
# append scope and set delimiters
if value is None:
is_table = key.startswith(table_start)
if is_table:
# account for 'data_' prefix
key = key[len(table_start):]
start, end = '\n[+', ']'
else:
start, end = '\n{', '}'
scope.append(''.join([replacements.get(c, c)
for c in key]))
# correct scope to omit internal 'general' section
scope_corr = scope
if scope[0] == mp_level01_titles[0]:
scope_corr = scope[1:]
# insert scope line
if (value is None and scope_corr)or \
(value is not None and level_reduction):
lines.append(start + '.'.join(scope_corr) + end)
# insert key-value line
if value is not None:
val = unicode(value) if not isinstance(value,
str) else value
value_lines = [val] if val.startswith('http') \
else textwrap.wrap(val)
if len(value_lines) > 1:
value_lines = [''] + value_lines + [':end']
lines.append(
make_pair(
''.join([replacements.get(c, c) for c in key]),
'\n'.join(value_lines)))
return '\n'.join(lines) + '\n'
