def ControlVocab(self, ncfile, variable=None, print_all=True):
"""
Check CMIP6 global attributes against Control Vocabulary file.
1. Validate required attribute if presents and some values.
2. Validate registered institution and institution_id
3. Validate registered source and source_id
4. Validate experiment, experiment_id and all attributes associated with this experiment.
Make sure that all attributes associate with the experiment_id found in CMIP6_CV.json
are set to the appropriate values.
5. Validate grid_label and grid_resolution
6. Validate creation time in ISO format (YYYY-MM-DDTHH:MM:SS)
7. Validate furtherinfourl from CV internal template
8. Validate variable attributes with CMOR JSON table.
9. Validate parent_* attribute
10. Validate sub_experiment_* attributes.
11. Validate that all *_index are integers.
"""
filename = os.path.basename(ncfile)
# -------------------------------------------------------------------
# Initialize arrays
# -------------------------------------------------------------------
# If table_path is the table directory
# Deduce corresponding JSON from filename
if os.path.isdir(self.cmip6_table_path):
cmip6_table = '{}/CMIP6_{}.json'.format(
self.cmip6_table_path, self._get_table_from_filename(filename))
else:
cmip6_table = self.cmip6_table_path
table_id = os.path.basename(
os.path.splitext(cmip6_table)[0]).split('_')[1]
# Check and get JSON table
cmor_table = self._check_json_table(cmip6_table)
# -------------------------------------------------------------------
# Load CMIP6 table into memory
# -------------------------------------------------------------------
table = cmip6_cv.load_table(cmip6_table)
# -------------------------------------------------------------------
# Deduce variable
# -------------------------------------------------------------------
# If variable can be deduced from the filename (Default)
# If not variable submitted on command line with --variable is considered
variable_id = self._get_variable_from_filename(filename)
if not variable:
variable = variable_id
# -------------------------------------------------------------------
# Distinguish similar CMOR entries with the same out_name if exist
# -------------------------------------------------------------------
# Apply test on variable only if a particular treatment if required
prepare_path = os.path.dirname(os.path.realpath(__file__))
out_names_tests = json.loads(
open(os.path.join(prepare_path, 'out_names_tests.json')).read())
# -------------------------------------------------------------------
# Open file in processing
# The file needs to be open before the calling the test.
# -------------------------------------------------------------------
infile = Cdunif.CdunifFile(ncfile, "r")
key = '{}_{}'.format(table_id, variable_id)
variable_cmor_entry = None
if key in list(out_names_tests.keys()):
for test, cmor_entry in list(out_names_tests[key].items()):
if getattr(self, test)(**{
'infile': infile,
'variable': variable,
'filename': filename
}):
# If test successfull, the CMOR entry to consider is given by the test
variable_cmor_entry = cmor_entry
else:
# If not, CMOR entry to consider is the variable from filename or from input command-line
variable_cmor_entry = variable
else:
# By default, CMOR entry to consider is the variable from filename or from input command-line
variable_cmor_entry = variable
# -------------------------------------------------------------------
# Get variable out name in netCDF record
# -------------------------------------------------------------------
# Variable record name should follow CMOR table out names
if variable_cmor_entry not in list(
cmor_table['variable_entry'].keys()):
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("The entry " + variable_cmor_entry +
" could not be found in CMOR table")
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
raise KeyboardInterrupt
variable_record_name = cmor_table['variable_entry'][
variable_cmor_entry]['out_name']
# Variable id attribute should be the same as variable record name
# in any case to be CF- and CMIP6-compliant
variable_id = variable_record_name
# -------------------------------------------------------------------
# Create a dictionary of all global attributes
# -------------------------------------------------------------------
self.dictGbl = infile.__dict__
for key, value in list(self.dictGbl.items()):
cmip6_cv.set_cur_dataset_attribute(key, value)
# Set member_id attribute depending on sub_experiment_id and variant_label
member_id = ""
if "sub_experiment_id" in list(self.dictGbl.keys()):
if self.dictGbl["sub_experiment_id"] not in ['none']:
member_id = '{}-{}'.format(self.dictGbl['sub_experiment_id'],
self.dictGbl['variant_label'])
else:
member_id = self.dictGbl['variant_label']
cmip6_cv.set_cur_dataset_attribute(cmip6_cv.GLOBAL_ATT_MEMBER_ID,
member_id)
# -------------------------------------------------------------------
# Create a dictionary of attributes for the variable
# -------------------------------------------------------------------
try:
self.dictVar = infile.variables[variable_record_name].__dict__
except BaseException:
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("The variable " + variable_record_name +
" could not be found in file")
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
raise KeyboardInterrupt
# -------------------------------------------------------------------
# Check global attributes
# -------------------------------------------------------------------
self.errors += cmip6_cv.check_requiredattributes(table)
self.errors += cmip6_cv.check_institution(table)
self.errors += cmip6_cv.check_sourceID(table)
self.errors += cmip6_cv.check_experiment(table)
self.errors += cmip6_cv.check_grids(table)
self.errors += cmip6_cv.check_ISOTime()
self.errors += cmip6_cv.check_furtherinfourl(table)
self.errors += cmip6_cv.check_subExpID(table)
for attr in ['branch_time_in_child', 'branch_time_in_parent']:
if attr in list(self.dictGbl.keys()):
self.set_double_value(attr)
if not isinstance(self.dictGbl[attr], numpy.float64):
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("{} is not a double: ".format(attr),
type(self.dictGbl[attr]))
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
self.errors += 1
for attr in [
'realization_index', 'initialization_index', 'physics_index',
'forcing_index'
]:
if not isinstance(self.dictGbl[attr], numpy.ndarray):
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("{} is not an integer: ".format(attr),
type(self.dictGbl[attr]))
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
self.errors += 1
self.errors += cmip6_cv.check_parentExpID(table)
for attr in ['table_id', 'variable_id']:
try:
if locals()[attr] != self.dictGbl[attr]:
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("{} attribute is not consistent: ".format(attr),
self.dictGbl[attr])
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
self.errors += 1
except KeyError:
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("{} attribute is missing in global attributes".format(
attr))
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
self.errors += 1
# -------------------------------------------------------------------
# Get time axis properties
# -------------------------------------------------------------------
# Get calendar and time units
try:
calendar = infile.variables['time'].calendar
timeunits = infile.variables['time'].units
except BaseException:
calendar = "gregorian"
timeunits = "days since ?"
# Get first and last time bounds
climatology = self.is_climatology(filename)
if climatology:
if cmip6_table.find('Amon') != -1:
variable = '{}Clim'.format(variable)
clim_idx = variable.find('Clim')
if climatology and clim_idx != -1:
var = [variable[:clim_idx]]
try:
if 'bounds' in list(infile.variables['time'].__dict__.keys()):
bndsvar = infile.variables['time'].__dict__['bounds']
elif 'climatology' in list(
infile.variables['time'].__dict__.keys()):
bndsvar = infile.variables['time'].__dict__['climatology']
else:
bndsvar = 'time_bnds'
startimebnds = infile.variables[bndsvar][0][0]
endtimebnds = infile.variables[bndsvar][-1][1]
except BaseException:
startimebnds = 0
endtimebnds = 0
try:
startime = infile.variables['time'][0]
endtime = infile.variables['time'][-1]
except BaseException:
startime = 0
endtime = 0
# -------------------------------------------------------------------
# Setup variable
# -------------------------------------------------------------------
varid = cmip6_cv.setup_variable(variable_cmor_entry,
self.dictVar['units'],
self.dictVar['_FillValue'][0],
startime, endtime, startimebnds,
endtimebnds)
if varid == -1:
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("Could not find variable {} in table {} ".format(
variable_cmor_entry, cmip6_table))
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
raise KeyboardInterrupt
# -------------------------------------------------------------------
# Check filename
# -------------------------------------------------------------------
self.errors += cmip6_cv.check_filename(table, varid, calendar,
timeunits, filename)
# -------------------------------------------------------------------
# Check variable attributes
# -------------------------------------------------------------------
cv_attrs = cmip6_cv.list_variable_attributes(varid)
for key in cv_attrs:
if key == "long_name":
continue
if key == "comment":
continue
if key == "cell_measures":
if cv_attrs[key].find("OPT") != -1 or cv_attrs[key].find(
"MODEL") != -1:
continue
# Is this attribute in file?
if key in list(self.dictVar.keys()):
# Verify that attribute value is equal to file attribute
table_value = cv_attrs[key]
file_value = self.dictVar[key]
# PrePARE accept units of 1 or 1.0 so adjust the table_value
if key == "units":
if (table_value == "1") and (file_value == "1.0"):
table_value = "1.0"
if (table_value == "1.0") and (file_value == "1"):
table_value = "1"
if isinstance(table_value, str) and isinstance(
file_value, numpy.ndarray):
if numpy.array(
[int(value) for value in table_value.split()
] == file_value).all():
file_value = True
table_value = True
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(file_value, numpy.ndarray):
file_value = file_value[0]
if isinstance(table_value, float):
if abs(table_value -
file_value) <= 0.00001 * abs(table_value):
table_value = file_value
if key == "cell_methods":
idx = file_value.find(" (")
if idx != -1:
file_value = file_value[:idx]
table_value = table_value[:idx]
if key == "cell_measures":
pattern = re.compile(
'(?P<param>[\w.-]+): (?P<val1>[\w.-]+) OR (?P<val2>[\w.-]+)'
)
values = re.findall(pattern, table_value)
table_values = [
""
] # Empty string is allowed in case of useless attribute
if values:
tmp = dict()
for param, val1, val2 in values:
tmp[param] = [
str('{}: {}'.format(param, val1)),
str('{}: {}'.format(param, val2))
]
table_values.extend([
' '.join(i) for i in list(
itertools.product(*list(tmp.values())))
])
if str(file_value) not in list(map(str, table_values)):
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("Your file contains \"" + key + "\":\"" +
str(file_value) + "\" and")
print("CMIP6 tables requires \"" + key + "\":\"" +
str(table_value) + "\".")
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
self.errors += 1
continue
if str(table_value) != str(file_value):
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("Your file contains \"" + key + "\":\"" +
str(file_value) + "\" and")
print("CMIP6 tables requires \"" + key + "\":\"" +
str(table_value) + "\".")
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
self.errors += 1
else:
# That attribute is not in the file
table_value = cv_attrs[key]
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(table_value, float):
table_value = "{0:.2g}".format(table_value)
print(BCOLORS.FAIL)
print(
"====================================================================================="
)
print("CMIP6 variable " + variable + " requires \"" + key +
"\":\"" + str(table_value) + "\".")
print(
"====================================================================================="
)
print(BCOLORS.ENDC)
self.errors += 1
# Print final message
if self.errors != 0:
print(BCOLORS.FAIL + "└──> :: CV FAIL :: {}".format(ncfile) +
BCOLORS.ENDC)
raise KeyboardInterrupt
elif print_all:
print(BCOLORS.OKGREEN + " :: CV SUCCESS :: {}".format(ncfile) +
BCOLORS.ENDC)
def ControlVocab(self, args):
'''
Check CMIP6 global attributes against Control Vocabulary file.
1. Validate required attribute if presents and some values.
2. Validate registered institution and institution_id
3. Validate registered source and source_id
4. Validate experiment, experiment_id and all attributes associated with this experiment.
Make sure that all attributes associate with the experiment_id found in CMIP6_CV.json
are set to the appropriate values.
5. Validate grid_label and grid_resolution
6. Validate creation time in ISO format (YYYY-MM-DDTHH:MM:SS)
7. Validate furtherinfourl from CV internal template
8. Validate variable attributes with CMOR JSON table.
9. Validate parent_* attribute
10. Validate sub_experiment_* atributes.
11. Validate that all *_index are integers.
'''
self.variable = args.variable
self.infile = args.infile
# -------------------------------------
# Create alist of all Global Attributes
# -------------------------------------
self.dictGbl = {
key: self.infile.__dict__[key]
for key in self.infile.__dict__.keys()
}
self.attributes = self.infile.__dict__.keys()
self.variables = self.infile.variables.keys()
ierr = [
cmip6_cv.set_cur_dataset_attribute(key, value)
for key, value in self.dictGbl.iteritems()
]
member_id = ""
if ("sub_experiment_id" in self.dictGbl.keys()):
if (self.dictGbl["sub_experiment_id"] not in ["none"]):
member_id = self.dictGbl["sub_experiment_id"] + \
'-' + self.dictGbl["variant_label"]
else:
member_id = self.dictGbl["variant_label"]
cmip6_cv.set_cur_dataset_attribute(cmip6_cv.GLOBAL_ATT_MEMBER_ID,
member_id)
self.setDoubleValue('branch_time_in_parent')
self.setDoubleValue('branch_time_in_child')
if self.variable is not None:
self.var = [self.variable]
else:
# -------------------------------------------------------------------
# find variable that contains a "history" (should only be one)
# -------------------------------------------------------------------
self.var = [self.infile.variable_id]
if ((self.var == []) or (len(self.var) > 1)):
print bcolors.FAIL
print "!!!!!!!!!!!!!!!!!!!!!!!!!"
print "! Error: The input file does not have an history attribute and the CMIP6 variable could not be found"
print "! Please use the --variable option to specify your CMIP6 variable"
print "! Check your file or use CMOR 3.x to achieve compliance for ESGF publication."
print "!!!!!!!!!!!!!!!!!!!!!!!!!"
print bcolors.ENDC
raise KeyboardInterrupt
try:
self.keys = self.infile.variables[self.var[0]].__dict__.keys()
except BaseException:
print bcolors.FAIL
print "!!!!!!!!!!!!!!!!!!!!!!!!!"
print "! Error: The variable " + self.var[
0] + " could not be found"
print "! Check your file variables "
print "!!!!!!!!!!!!!!!!!!!!!!!!!"
print bcolors.ENDC
raise
# -------------------------------------------------------------------
# Create a dictionnary of attributes for var
# -------------------------------------------------------------------
self.dictVars = dict(
(y, x) for y, x in
[(key, value) for key in self.keys
if self.infile.variables[self.var[0]].__dict__[key] is not None
for value in [self.infile.variables[self.var[0]].__dict__[key]]])
try:
self.calendar = self.infile.variables['time'].calendar
self.timeunits = self.infile.variables['time'].units
except BaseException:
self.calendar = "gregorian"
self.timeunits = "days since ?"
cmip6_cv.check_requiredattributes(self.table_id)
cmip6_cv.check_institution(self.table_id)
cmip6_cv.check_sourceID(self.table_id)
cmip6_cv.check_experiment(self.table_id)
cmip6_cv.check_grids(self.table_id)
cmip6_cv.check_ISOTime()
cmip6_cv.check_furtherinfourl(self.table_id)
cmip6_cv.check_parentExpID(self.table_id)
cmip6_cv.check_subExpID(self.table_id)
try:
startimebnds = self.infile.variables['time_bnds'][0][0]
endtimebnds = self.infile.variables['time_bnds'][-1][1]
except BaseException:
startimebnds = 0
endtimebnds = 0
try:
startime = self.infile.variables['time'][0]
endtime = self.infile.variables['time'][-1]
except BaseException:
startime = 0
endtime = 0
varunits = self.infile.variables[self.var[0]].units
varmissing = self.infile.variables[self.var[0]]._FillValue[0]
varid = cmip6_cv.setup_variable(self.var[0], varunits, varmissing,
startime, endtime, startimebnds,
endtimebnds)
if (varid == -1):
print bcolors.FAIL
print "====================================================================================="
print " Could not find variable '%s' in table '%s' " % (
self.var[0], self.cmip6_table)
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
return
# fn = os.path.basename(self.infile.id)
fn = os.path.basename(str(self.infile).split('\'')[1])
cmip6_cv.check_filename(self.table_id, varid, self.calendar,
self.timeunits, fn)
if not isinstance(self.dictGbl['realization_index'], numpy.ndarray):
print bcolors.FAIL
print "====================================================================================="
print "realization_index is not an integer: ", type(
self.dictGbl['realization_index'])
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
if not isinstance(self.dictGbl['initialization_index'], numpy.ndarray):
print bcolors.FAIL
print "====================================================================================="
print "initialization_index is not an integer: ", type(
self.dictGbl['initialization_index'])
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
if not isinstance(self.dictGbl['physics_index'], numpy.ndarray):
print bcolors.FAIL
print "====================================================================================="
print "physics_index is not an integer: ", type(
self.dictGbl['physics_index'])
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
if not isinstance(self.dictGbl['forcing_index'], numpy.ndarray):
print bcolors.FAIL
print "====================================================================================="
print "forcing_index is not an integer: ", type(
self.dictGbl['forcing_index'])
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
prepLIST = cmip6_cv.list_variable_attributes(varid)
for key in prepLIST:
if (key == "long_name"):
continue
if (key == "comment"):
continue
# Is this attritue in file?
if (key in self.dictVars.keys()):
# Verify that attribute value is equal to file attribute
table_value = prepLIST[key]
file_value = self.dictVars[key]
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(file_value, numpy.ndarray):
file_value = file_value[0]
if isinstance(table_value, float):
if (file_value == 0):
if (table_value != file_value):
file_value = False
else:
if (1 - (table_value / file_value) < 0.00001):
table_value = file_value
if key == "cell_methods":
idx = file_value.find(" (interval:")
file_value = file_value[:idx]
table_value = table_value[:idx]
file_value = str(file_value)
table_value = str(table_value)
if table_value != file_value:
print bcolors.FAIL
print "====================================================================================="
print "You file contains \"" + key + "\":\"" + str(
file_value) + "\" and"
print "CMIP6 tables requires \"" + key + "\":\"" + str(
table_value) + "\"."
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
else:
# That attribute is not in the file
table_value = prepLIST[key]
if key == "cell_measures":
if ((table_value.find("OPT") != -1)
or (table_value.find("MODEL") != -1)):
continue
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(table_value, float):
table_value = "{0:.2g}".format(table_value)
print bcolors.FAIL
print "====================================================================================="
print "CMIP6 variable " + self.var[
0] + " requires \"" + key + "\":\"" + str(
table_value) + "\"."
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
if (cmip6_cv.get_CV_Error()):
raise KeyboardInterrupt
pass
print bcolors.OKGREEN
print "*************************************************************************************"
print "* This file is compliant with the CMIP6 specification and can be published in ESGF. *"
print "*************************************************************************************"
print bcolors.ENDC
def ControlVocab(self):
'''
Check CMIP6 global attributes against Control Vocabulary file.
1. Validate required attribute if presents and some values.
2. Validate registered institution and institution_id
3. Validate registered source and source_id
4. Validate experiment, experiment_id and attribute associated with the experiment.
5. Validate grid_label and grid_resolution
6. Validate creation time in ISO format (YYYY-MM-DDTHH:MM:SS)
7. Validate furtherinfourl from CV internal template
8. Validate variable attributes with CMOR JSON table.
'''
cmip6_cv.check_requiredattributes(self.table_id)
cmip6_cv.check_institution(self.table_id)
cmip6_cv.check_sourceID(self.table_id)
cmip6_cv.check_experiment(self.table_id)
cmip6_cv.check_grids(self.table_id)
cmip6_cv.check_ISOTime()
cmip6_cv.check_furtherinfourl(self.table_id)
varid = cmip6_cv.setup_variable(self.var[0], 'm', 1e20)
prepLIST = cmip6_cv.list_variable_attributes(varid)
for key in prepLIST:
if (key == "comment"):
continue
# Is this attritue in file?
if (key in self.dictVars.keys()):
# Verify that attribute value is equal to file attribute
table_value = prepLIST[key]
file_value = self.dictVars[key]
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(file_value, numpy.ndarray):
file_value = file_value[0]
if isinstance(table_value, float):
if (table_value / file_value < 1.1):
table_value = file_value
if key == "cell_methods":
idx = file_value.find(" (interval:")
file_value = file_value[:idx]
table_value = table_value[:idx]
file_value = str(file_value)
table_value = str(table_value)
if table_value != file_value:
print bcolors.FAIL
print "====================================================================================="
print "You file contains \"" + key + "\":\"" + str(
file_value) + "\" and"
print "CMIP6 tables requires \"" + key + "\":\"" + str(
table_value) + "\"."
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
else:
# That attribute is not in the file
table_value = prepLIST[key]
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(table_value, float):
table_value = "{0:.2g}".format(table_value)
print bcolors.FAIL
print "====================================================================================="
print "CMIP6 variable " + self.var[
0] + " requires \"" + key + "\":\"" + str(
table_value) + "\"."
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
if (cmip6_cv.get_CV_Error()):
raise KeyboardInterrupt
print bcolors.OKGREEN
print "*************************************************************************************"
print "* This file is compliant with the CMIP6 specification and can be published in ESGF. *"
print "*************************************************************************************"
print bcolors.ENDC
def ControlVocab(self, ncfile, variable=None):
"""
Check CMIP6 global attributes against Control Vocabulary file.
1. Validate required attribute if presents and some values.
2. Validate registered institution and institution_id
3. Validate registered source and source_id
4. Validate experiment, experiment_id and all attributes associated with this experiment.
Make sure that all attributes associate with the experiment_id found in CMIP6_CV.json
are set to the appropriate values.
5. Validate grid_label and grid_resolution
6. Validate creation time in ISO format (YYYY-MM-DDTHH:MM:SS)
7. Validate furtherinfourl from CV internal template
8. Validate variable attributes with CMOR JSON table.
9. Validate parent_* attribute
10. Validate sub_experiment_* attributes.
11. Validate that all *_index are integers.
"""
err = 0
cmip6_cv.reset_CV_Error()
filename = os.path.basename(ncfile)
# -------------------------------------------------------------------
# Initialize arrays
# -------------------------------------------------------------------
# If table_path is the table directory
# Deduce corresponding JSON from filename
if os.path.isdir(self.cmip6_table_path):
cmip6_table = '{}/CMIP6_{}.json'.format(
self.cmip6_table_path, self._get_table_from_filename(filename))
else:
cmip6_table = self.cmip6_table_path
table_id = os.path.basename(os.path.splitext(cmip6_table)[0]).split('_')[1]
# Check JSON file
self._check_json_table(cmip6_table)
# -------------------------------------------------------------------
# Load CMIP6 table into memory
# -------------------------------------------------------------------
table = cmip6_cv.load_table(cmip6_table)
# -------------------------------------------------------------------
# Deduce variable
# -------------------------------------------------------------------
# If variable can be deduced from the filename (Default)
# If not variable submitted on command line with --variable is considered
variable_id = self._get_variable_from_filename(filename)
if not variable:
variable = variable_id
# -------------------------------------------------------------------
# Open file in processing
# -------------------------------------------------------------------
infile = Cdunif.CdunifFile(ncfile, "r")
# -------------------------------------------------------------------
# Create a dictionary of all global attributes
# -------------------------------------------------------------------
self.dictGbl = infile.__dict__
for key, value in self.dictGbl.iteritems():
cmip6_cv.set_cur_dataset_attribute(key, value)
# Set member_id attribute depending on sub_experiment_id and variant_label
member_id = ""
if "sub_experiment_id" in self.dictGbl.keys():
if self.dictGbl["sub_experiment_id"] not in ['none']:
member_id = '{}-{}'.format(self.dictGbl['sub_experiment_id'],
self.dictGbl['variant_label'])
else:
member_id = self.dictGbl['variant_label']
cmip6_cv.set_cur_dataset_attribute(cmip6_cv.GLOBAL_ATT_MEMBER_ID, member_id)
self.set_double_value('branch_time_in_parent')
self.set_double_value('branch_time_in_child')
# -------------------------------------------------------------------
# Create a dictionary of attributes for the variable
# -------------------------------------------------------------------
try:
self.dictVar = infile.variables[variable].__dict__
except BaseException:
print BCOLORS.FAIL
print "====================================================================================="
print "The variable " + variable + " could not be found in file"
print "====================================================================================="
print BCOLORS.ENDC
raise KeyboardInterrupt
# -------------------------------------------------------------------
# Check global attributes
# -------------------------------------------------------------------
self.errors += cmip6_cv.check_requiredattributes(table)
self.errors += cmip6_cv.check_institution(table)
self.errors += cmip6_cv.check_sourceID(table)
self.errors += cmip6_cv.check_experiment(table)
self.errors += cmip6_cv.check_grids(table)
self.errors += cmip6_cv.check_ISOTime()
self.errors += cmip6_cv.check_furtherinfourl(table)
self.errors += cmip6_cv.check_parentExpID(table)
self.errors += cmip6_cv.check_subExpID(table)
for attr in ['branch_time_in_child', 'branch_time_in_parent']:
if attr in self.dictGbl.keys():
if not isinstance(self.dictGbl[attr], numpy.float64):
print BCOLORS.FAIL
print "====================================================================================="
print "{} is not a double: ".format(attr), type(self.dictGbl[attr])
print "====================================================================================="
print BCOLORS.ENDC
self.errors += 1
for attr in ['realization_index', 'initialization_index', 'physics_index', 'forcing_index']:
if not isinstance(self.dictGbl[attr], numpy.ndarray):
print BCOLORS.FAIL
print "====================================================================================="
print "{} is not an integer: ".format(attr), type(self.dictGbl[attr])
print "====================================================================================="
print BCOLORS.ENDC
self.errors += 1
for attr in ['table_id', 'variable_id']:
try:
if locals()[attr] != self.dictGbl[attr]:
print BCOLORS.FAIL
print "====================================================================================="
print "{} attribute is not consistent: ".format(attr), self.dictGbl[attr]
print "====================================================================================="
print BCOLORS.ENDC
self.errors += 1
except KeyError:
print BCOLORS.FAIL
print "====================================================================================="
print "{} attribute is missing in global attributes".format(attr)
print "====================================================================================="
print BCOLORS.ENDC
self.errors += 1
# -------------------------------------------------------------------
# Get time axis properties
# -------------------------------------------------------------------
# Get calendar and time units
try:
calendar = infile.variables['time'].calendar
timeunits = infile.variables['time'].units
except BaseException:
calendar = "gregorian"
timeunits = "days since ?"
# Get first and last time bounds
try:
if 'bounds' in infile.variables['time'].__dict__.keys():
bndsvar = infile.variables['time'].__dict__['bounds']
startimebnds = infile.variables[bndsvar][0][0]
endtimebnds = infile.variables[bndsvar][-1][1]
else:
startimebnds = infile.variables['time_bnds'][0][0]
endtimebnds = infile.variables['time_bnds'][-1][1]
except BaseException:
startimebnds = 0
endtimebnds = 0
# Get first and last time steps
try:
startime = infile.variables['time'][0]
endtime = infile.variables['time'][-1]
except BaseException:
startime = 0
endtime = 0
# -------------------------------------------------------------------
# Distinguish similar CMOR entries with the same out_name if exist
# -------------------------------------------------------------------
# Apply test on variable only if a particular treatment if required
prepare_path = os.path.dirname(os.path.realpath(__file__))
out_names_tests = json.loads(open(os.path.join(prepare_path, 'out_names_tests.json')).read())
key = '{}_{}'.format(table_id, variable_id)
if key in out_names_tests.keys():
for test, cmor_entry in out_names_tests[key].iteritems():
if getattr(self, test)(**{'infile': infile,
'variable': variable,
'filename': filename}):
variable = cmor_entry
# -------------------------------------------------------------------
# Setup variable
# -------------------------------------------------------------------
varid = cmip6_cv.setup_variable(variable,
self.dictVar['units'],
self.dictVar['_FillValue'][0],
startime,
endtime,
startimebnds,
endtimebnds)
if varid == -1:
print BCOLORS.FAIL
print "====================================================================================="
print "Could not find variable {} in table {} ".format(variable, cmip6_table)
print "====================================================================================="
print BCOLORS.ENDC
raise KeyboardInterrupt
# -------------------------------------------------------------------
# Check filename
# -------------------------------------------------------------------
self.errors += cmip6_cv.check_filename(table,
varid,
calendar,
timeunits,
filename)
# -------------------------------------------------------------------
# Check variable attributes
# -------------------------------------------------------------------
fn = os.path.basename(str(infile).split('\'')[1])
err += cmip6_cv.check_filename(
table,
varid,
calendar,
timeunits,
fn)
if (err != 0) or (cmip6_cv.get_CV_Error() == 1):
self.cv_error = True
if 'branch_time_in_child' in self.dictGbl.keys():
if not isinstance(self.dictGbl['branch_time_in_child'], numpy.float64):
print BCOLORS.FAIL
print "====================================================================================="
print "branch_time_in_child is not a double: ", type(self.dictGbl['branch_time_in_child'])
print "====================================================================================="
print BCOLORS.ENDC
self.cv_error = True
if 'branch_time_in_parent' in self.dictGbl.keys():
if not isinstance(self.dictGbl['branch_time_in_parent'], numpy.float64):
print BCOLORS.FAIL
print "====================================================================================="
print "branch_time_in_parent is not an double: ", type(self.dictGbl['branch_time_in_parent'])
print "====================================================================================="
print BCOLORS.ENDC
self.cv_error = True
if not isinstance(self.dictGbl['branch_time_in_child'], numpy.float64):
print bcolors.FAIL
print "====================================================================================="
print "realization_index is not a double: ", type(self.dictGbl['branch_time_in_child'])
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
if not isinstance(self.dictGbl['branch_time_in_parent'], numpy.float64):
print bcolors.FAIL
print "====================================================================================="
print "initialization_index is not an double: ", type(self.dictGbl['branch_time_in_parent'])
print "====================================================================================="
print bcolors.ENDC
cmip6_cv.set_CV_Error()
if not isinstance(self.dictGbl['realization_index'], numpy.ndarray):
print BCOLORS.FAIL
print "====================================================================================="
print "realization_index is not an integer: ", type(self.dictGbl['realization_index'])
print "====================================================================================="
print BCOLORS.ENDC
self.cv_error = True
if not isinstance(self.dictGbl['initialization_index'], numpy.ndarray):
print BCOLORS.FAIL
print "====================================================================================="
print "initialization_index is not an integer: ", type(self.dictGbl['initialization_index'])
print "====================================================================================="
print BCOLORS.ENDC
self.cv_error = True
if not isinstance(self.dictGbl['physics_index'], numpy.ndarray):
print BCOLORS.FAIL
print "====================================================================================="
print "physics_index is not an integer: ", type(self.dictGbl['physics_index'])
print "====================================================================================="
print BCOLORS.ENDC
self.cv_error = True
if not isinstance(self.dictGbl['forcing_index'], numpy.ndarray):
print BCOLORS.FAIL
print "====================================================================================="
print "forcing_index is not an integer: ", type(self.dictGbl['forcing_index'])
print "====================================================================================="
print BCOLORS.ENDC
self.cv_error = True
# -----------------------------
# variable attribute comparison
# -----------------------------
cv_attrs = cmip6_cv.list_variable_attributes(varid)
for key in cv_attrs:
if key == "long_name":
continue
if key == "comment":
continue
if key == "cell_measures":
if cv_attrs[key].find("OPT") != -1 or cv_attrs[key].find("MODEL") != -1:
continue
# Is this attribute in file?
if key in self.dictVar.keys():
# Verify that attribute value is equal to file attribute
table_value = cv_attrs[key]
file_value = self.dictVar[key]
# PrePARE accept units of 1 or 1.0 so adjust the table_value
if key == "units":
if (table_value == "1") and (file_value == "1.0"):
table_value = "1.0"
if (table_value == "1.0") and (file_value == "1"):
table_value = "1"
if isinstance(table_value, str) and isinstance(file_value, numpy.ndarray):
if numpy.array([int(value) for value in table_value.split()] == file_value).all():
file_value = True
table_value = True
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(file_value, numpy.ndarray):
file_value = file_value[0]
if isinstance(table_value, float):
if file_value == 0:
if table_value != file_value:
file_value = False
else:
if abs(1 - (table_value / file_value)) < 0.00001:
table_value = file_value
if key == "cell_methods":
idx = file_value.find(" (")
if idx != -1:
file_value = file_value[:idx]
table_value = table_value[:idx]
if key == "cell_measures":
pattern = re.compile('(?P<param>[\w.-]+): (?P<val1>[\w.-]+) OR (?P<val2>[\w.-]+)')
values = re.findall(pattern, table_value)
table_values = [""] # Empty string is allowed in case of useless attribute
if values:
tmp = dict()
for param, val1, val2 in values:
tmp[param] = [str('{}: {}'.format(param, val1)), str('{}: {}'.format(param, val2))]
table_values.extend([' '.join(i) for i in list(itertools.product(*tmp.values()))])
if str(file_value) not in map(str, table_values):
print BCOLORS.FAIL
print "====================================================================================="
print "Your file contains \"" + key + "\":\"" + str(file_value) + "\" and"
print "CMIP6 tables requires \"" + key + "\":\"" + str(table_value) + "\"."
print "====================================================================================="
print BCOLORS.ENDC
self.errors += 1
continue
if str(table_value) != str(file_value):
print BCOLORS.FAIL
print "====================================================================================="
print "Your file contains \"" + key + "\":\"" + str(file_value) + "\" and"
print "CMIP6 tables requires \"" + key + "\":\"" + str(table_value) + "\"."
print "====================================================================================="
print BCOLORS.ENDC
self.errors += 1
else:
# That attribute is not in the file
table_value = cv_attrs[key]
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(table_value, float):
table_value = "{0:.2g}".format(table_value)
print BCOLORS.FAIL
print "====================================================================================="
print "CMIP6 variable " + variable + " requires \"" + key + "\":\"" + str(table_value) + "\"."
print "====================================================================================="
print BCOLORS.ENDC
self.errors += 1
if self.errors != 0:
raise KeyboardInterrupt
else:
print BCOLORS.OKGREEN
print "*************************************************************************************"
print "* This file is compliant with the CMIP6 specification and can be published in ESGF *"
print "*************************************************************************************"
print BCOLORS.ENDC
def ControlVocab(self, ncfile, variable=None, print_all=True, no_text_color=False):
"""
Check CMIP6 global attributes against Control Vocabulary file.
1. Validate required attribute if presents and some values.
2. Validate registered institution and institution_id
3. Validate registered source and source_id
4. Validate experiment, experiment_id and all attributes associated with this experiment.
Make sure that all attributes associate with the experiment_id found in CMIP6_CV.json
are set to the appropriate values.
5. Validate grid_label and grid_resolution
6. Validate creation time in ISO format (YYYY-MM-DDTHH:MM:SS)
7. Validate furtherinfourl from CV internal template
8. Validate variable attributes with CMOR JSON table.
9. Validate parent_* attribute
10. Validate sub_experiment_* attributes.
11. Validate that all *_index are integers.
"""
filename = os.path.basename(ncfile)
# -------------------------------------------------------------------
# Initialize arrays
# -------------------------------------------------------------------
# If table_path is the table directory
# Deduce corresponding JSON from filename
if os.path.isdir(self.cmip6_table_path):
cmip6_table = '{}/CMIP6_{}.json'.format(
self.cmip6_table_path, self._get_table_from_filename(filename))
else:
cmip6_table = self.cmip6_table_path
table_id = os.path.basename(os.path.splitext(cmip6_table)[0]).split('_')[1]
# Check and get JSON table
cmor_table = self._check_json_table(cmip6_table)
# -------------------------------------------------------------------
# Load CMIP6 table into memory
# -------------------------------------------------------------------
table = cmip6_cv.load_table(cmip6_table)
# -------------------------------------------------------------------
# Deduce variable
# -------------------------------------------------------------------
# If variable can be deduced from the filename (Default)
# If not variable submitted on command line with --variable is considered
variable_id = self._get_variable_from_filename(filename)
if not variable:
variable = variable_id
# -------------------------------------------------------------------
# Distinguish similar CMOR entries with the same out_name if exist
# -------------------------------------------------------------------
# Apply test on variable only if a particular treatment if required
prepare_path = os.path.dirname(os.path.realpath(__file__))
out_names_tests = json.loads(open(os.path.join(prepare_path, 'out_names_tests.json')).read())
# -------------------------------------------------------------------
# Open file in processing
# The file needs to be open before the calling the test.
# -------------------------------------------------------------------
infile = netCDF4.Dataset(ncfile, "r")
key = '{}_{}'.format(table_id, variable_id)
variable_cmor_entry = None
if key in list(out_names_tests.keys()):
for test, cmor_entry in list(out_names_tests[key].items()):
if getattr(self, test)(**{'infile': infile,
'variable': variable,
'filename': filename}):
# If test successfull, the CMOR entry to consider is given by the test
variable_cmor_entry = cmor_entry
break
else:
# If not, CMOR entry to consider is the variable from filename or from input command-line
variable_cmor_entry = variable
else:
# By default, CMOR entry to consider is the variable from filename or from input command-line
variable_cmor_entry = variable
# -------------------------------------------------------------------
# Get variable out name in netCDF record
# -------------------------------------------------------------------
# Variable record name should follow CMOR table out names
if variable_cmor_entry not in list(cmor_table['variable_entry'].keys()):
msg = "The entry " + variable_cmor_entry + " could not be found in CMOR table"
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.prepare_print("└──> :: CV FAIL :: {}".format(ncfile), 'FAIL', no_text_color)
raise KeyboardInterrupt
variable_record_name = cmor_table['variable_entry'][variable_cmor_entry]['out_name']
# Variable id attribute should be the same as variable record name
# in any case to be CF- and CMIP6-compliant
variable_id = variable_record_name
# -------------------------------------------------------------------
# Create a dictionary of all global attributes
# -------------------------------------------------------------------
self.dictGbl = infile.__dict__
for key, value in list(self.dictGbl.items()):
cmip6_cv.set_cur_dataset_attribute(key, value)
# Set member_id attribute depending on sub_experiment_id and variant_label
member_id = ""
if "sub_experiment_id" in list(self.dictGbl.keys()):
if self.dictGbl["sub_experiment_id"] not in ['none']:
member_id = '{}-{}'.format(self.dictGbl['sub_experiment_id'],
self.dictGbl['variant_label'])
else:
member_id = self.dictGbl['variant_label']
cmip6_cv.set_cur_dataset_attribute(cmip6_cv.GLOBAL_ATT_MEMBER_ID, member_id)
# -------------------------------------------------------------------
# Create a dictionary of attributes for the variable
# -------------------------------------------------------------------
try:
self.dictVar = infile.variables[variable_record_name].__dict__
except BaseException:
msg = "The variable " + variable_record_name + " could not be found in file"
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.prepare_print("└──> :: CV FAIL :: {}".format(ncfile), 'FAIL', no_text_color)
raise KeyboardInterrupt
# -------------------------------------------------------------------
# Check global attributes
# -------------------------------------------------------------------
if cmip6_cv.check_requiredattributes(table) != 0:
self.errors += 1
if cmip6_cv.check_institution(table) != 0:
self.errors += 1
if cmip6_cv.check_sourceID(table) != 0:
self.errors += 1
if cmip6_cv.check_experiment(table) != 0:
self.errors += 1
if cmip6_cv.check_grids(table) != 0:
self.errors += 1
if cmip6_cv.check_ISOTime() != 0:
self.errors += 1
if cmip6_cv.check_furtherinfourl(table) != 0:
self.errors += 1
if cmip6_cv.check_subExpID(table) != 0:
self.errors += 1
for attr in ['branch_time_in_child', 'branch_time_in_parent']:
if attr in list(self.dictGbl.keys()):
self.set_double_value(attr)
if not numpy.issubdtype(type(self.dictGbl[attr]), numpy.float64):
msg = "{} is not a double: {}".format(attr, type(self.dictGbl[attr]))
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
for attr in ['realization_index', 'initialization_index', 'physics_index', 'forcing_index']:
try:
if not numpy.issubdtype(type(self.dictGbl[attr]), numpy.integer):
msg = "{} is not an integer: {}".format(attr, type(self.dictGbl[attr]))
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
except KeyError:
msg = "{} attribute is missing in global attributes".format(attr)
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
if cmip6_cv.check_parentExpID(table) != 0:
self.errors += 1
for attr in ['table_id', 'variable_id']:
try:
if locals()[attr] != self.dictGbl[attr]:
msg = "{} attribute is not consistent: {}".format(attr, self.dictGbl[attr])
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
except KeyError:
msg = "{} attribute is missing in global attributes".format(attr)
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
# -------------------------------------------------------------------
# Get time axis properties
# -------------------------------------------------------------------
# Get calendar and time units
try:
calendar = infile.variables['time'].calendar
timeunits = infile.variables['time'].units
except BaseException:
calendar = "gregorian"
timeunits = "days since ?"
# Get first and last time bounds
climatology = self.is_climatology(filename)
if climatology:
if cmip6_table.find('Amon') != -1:
variable = '{}Clim'.format(variable)
clim_idx = variable.find('Clim')
if climatology and clim_idx != -1:
var = [variable[:clim_idx]]
try:
if 'bounds' in list(infile.variables['time'].__dict__.keys()):
bndsvar = infile.variables['time'].__dict__['bounds']
elif 'climatology' in list(infile.variables['time'].__dict__.keys()):
bndsvar = infile.variables['time'].__dict__['climatology']
else:
bndsvar = 'time_bnds'
startimebnds = infile.variables[bndsvar][0][0]
endtimebnds = infile.variables[bndsvar][-1][1]
except BaseException:
startimebnds = 0
endtimebnds = 0
try:
startime = infile.variables['time'][0]
endtime = infile.variables['time'][-1]
except BaseException:
startime = 0
endtime = 0
# -------------------------------------------------------------------
# Setup variable
# -------------------------------------------------------------------
varid = cmip6_cv.setup_variable(variable_cmor_entry,
cmor_table['variable_entry'][variable_cmor_entry]['units'],
float(cmor_table['Header']['missing_value']),
int(cmor_table['Header']['int_missing_value']),
startime,
endtime,
startimebnds,
endtimebnds)
if varid == -1:
msg = "Could not find variable {} in table {} ".format(variable_cmor_entry, cmip6_table)
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.prepare_print("└──> :: CV FAIL :: {}".format(ncfile), 'FAIL', no_text_color)
raise KeyboardInterrupt
# -------------------------------------------------------------------
# Check filename
# -------------------------------------------------------------------
if cmip6_cv.check_filename(table, varid, calendar, timeunits, filename) != 0:
self.errors += 1
# -------------------------------------------------------------------
# Check variable attributes
# -------------------------------------------------------------------
cv_attrs = cmip6_cv.list_variable_attributes(varid)
for key in cv_attrs:
if key == "long_name":
continue
if key == "comment":
continue
if key == "cell_measures":
if " OR " in cv_attrs[key] or "OPT" in cv_attrs[key] or "MODEL" in cv_attrs[key] or "UGRID" in cv_attrs[key]:
continue
# Is this attribute in file?
if key in list(self.dictVar.keys()):
# Verify that attribute value is equal to file attribute
table_value = cv_attrs[key]
file_value = self.dictVar[key]
# PrePARE accept units of 1 or 1.0 so adjust the table_value
if key == "units":
if (table_value == "1") and (file_value == "1.0"):
table_value = "1.0"
if (table_value == "1.0") and (file_value == "1"):
table_value = "1"
if isinstance(table_value, str) and isinstance(file_value, numpy.ndarray):
if numpy.array([int(value) for value in table_value.split()] == file_value).all():
file_value = True
table_value = True
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(file_value, numpy.ndarray):
file_value = file_value[0]
if isinstance(table_value, float):
if abs(table_value - file_value) <= 0.00001 * abs(table_value):
table_value = file_value
if key == "cell_methods":
# Remove text that is inside parentheses i.e. comments
file_value = re.sub(r"\(.*\)", "", file_value)
table_value = re.sub(r"\(.*\)", "", table_value)
# Remove extra whitespace
file_value = " ".join(file_value.split())
table_value = " ".join(table_value.split())
if key == "cell_measures":
# Check if area and volume values from the table's cell_measures are found in the file's external_variables
pattern = re.compile('(?:area|volume): (\w+)')
values = re.findall(pattern, table_value)
for v in values:
if 'external_variables' in list(self.dictGbl.keys()):
if not re.search(r"\b{}\b".format(v), self.dictGbl['external_variables']):
msg = "Your file contains external_variables = \"" + self.dictGbl['external_variables'] + "\", and"
if len(values) == 2:
msg += "\nCMIP6 tables requires \"" + values[0] + "\" and \"" + values[1] + "\" in external_variables."
else:
msg += "\nCMIP6 tables requires \"" + values[0] + "\" in external_variables."
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
else:
if len(values) == 2:
msg = "Your file contains \"" + values[0] + "\" and \"" + values[1] + "\" in cell_measures and"
else:
msg = "Your file contains \"" + values[0] + "\" in cell_measures and"
msg += "\nCMIP6 tables require attribute \"external_variables\" in global attributes."
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
continue
if str(table_value) != str(file_value):
msg = "Your file contains \"" + key + "\":\"" + str(file_value) + "\" and"
msg += "\nCMIP6 tables requires \"" + key + "\":\"" + str(table_value) + "\"."
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
else:
# That attribute is not in the file
table_value = cv_attrs[key]
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(table_value, float):
table_value = "{0:.2g}".format(table_value)
msg = "CMIP6 variable " + variable + " requires \"" + key + "\":\"" + str(table_value) + "\"."
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
# Check if cell_measures is defined in the file but not in the table
if "cell_measures" in list(self.dictVar.keys()) and "cell_measures" not in cv_attrs:
msg = "Your file contains \"cell_measures\":\"" + str(self.dictVar["cell_measures"]) + "\" but"
msg += "\nCMIP6 tables do not define \"cell_measures\"."
self.prepare_print(msg, 'FAIL', no_text_color, lines=True)
self.errors += 1
# Print final message
if self.errors != 0:
self.prepare_print("└──> :: CV FAIL :: {}".format(ncfile), 'FAIL', no_text_color)
raise KeyboardInterrupt
elif print_all:
self.prepare_print(" :: CV SUCCESS :: {}".format(ncfile), 'OKGREEN', no_text_color)
def ControlVocab(self, ncfile, variable=None, print_all=True):
"""
Check CMIP6 global attributes against Control Vocabulary file.
1. Validate required attribute if presents and some values.
2. Validate registered institution and institution_id
3. Validate registered source and source_id
4. Validate experiment, experiment_id and all attributes associated with this experiment.
Make sure that all attributes associate with the experiment_id found in CMIP6_CV.json
are set to the appropriate values.
5. Validate grid_label and grid_resolution
6. Validate creation time in ISO format (YYYY-MM-DDTHH:MM:SS)
7. Validate furtherinfourl from CV internal template
8. Validate variable attributes with CMOR JSON table.
9. Validate parent_* attribute
10. Validate sub_experiment_* attributes.
11. Validate that all *_index are integers.
"""
filename = os.path.basename(ncfile)
# -------------------------------------------------------------------
# Initialize arrays
# -------------------------------------------------------------------
# If table_path is the table directory
# Deduce corresponding JSON from filename
if os.path.isdir(self.cmip6_table_path):
cmip6_table = '{}/CMIP6_{}.json'.format(
self.cmip6_table_path, self._get_table_from_filename(filename))
else:
cmip6_table = self.cmip6_table_path
table_id = os.path.basename(os.path.splitext(cmip6_table)[0]).split('_')[1]
# Check and get JSON table
cmor_table = self._check_json_table(cmip6_table)
# -------------------------------------------------------------------
# Load CMIP6 table into memory
# -------------------------------------------------------------------
table = cmip6_cv.load_table(cmip6_table)
# -------------------------------------------------------------------
# Deduce variable
# -------------------------------------------------------------------
# If variable can be deduced from the filename (Default)
# If not variable submitted on command line with --variable is considered
variable_id = self._get_variable_from_filename(filename)
if not variable:
variable = variable_id
# -------------------------------------------------------------------
# Distinguish similar CMOR entries with the same out_name if exist
# -------------------------------------------------------------------
# Apply test on variable only if a particular treatment if required
prepare_path = os.path.dirname(os.path.realpath(__file__))
out_names_tests = json.loads(open(os.path.join(prepare_path, 'out_names_tests.json')).read())
# -------------------------------------------------------------------
# Open file in processing
# The file needs to be open before the calling the test.
# -------------------------------------------------------------------
infile = Cdunif.CdunifFile(ncfile, "r")
key = '{}_{}'.format(table_id, variable_id)
variable_cmor_entry = None
if key in list(out_names_tests.keys()):
for test, cmor_entry in list(out_names_tests[key].items()):
if getattr(self, test)(**{'infile': infile,
'variable': variable,
'filename': filename}):
# If test successfull, the CMOR entry to consider is given by the test
variable_cmor_entry = cmor_entry
else:
# If not, CMOR entry to consider is the variable from filename or from input command-line
variable_cmor_entry = variable
else:
# By default, CMOR entry to consider is the variable from filename or from input command-line
variable_cmor_entry = variable
# -------------------------------------------------------------------
# Get variable out name in netCDF record
# -------------------------------------------------------------------
# Variable record name should follow CMOR table out names
if variable_cmor_entry not in list(cmor_table['variable_entry'].keys()):
print(BCOLORS.FAIL)
print("=====================================================================================")
print("The entry " + variable_cmor_entry + " could not be found in CMOR table")
print("=====================================================================================")
print(BCOLORS.ENDC)
raise KeyboardInterrupt
variable_record_name = cmor_table['variable_entry'][variable_cmor_entry]['out_name']
# Variable id attribute should be the same as variable record name
# in any case to be CF- and CMIP6-compliant
variable_id = variable_record_name
# -------------------------------------------------------------------
# Create a dictionary of all global attributes
# -------------------------------------------------------------------
self.dictGbl = infile.__dict__
for key, value in list(self.dictGbl.items()):
cmip6_cv.set_cur_dataset_attribute(key, value)
# Set member_id attribute depending on sub_experiment_id and variant_label
member_id = ""
if "sub_experiment_id" in list(self.dictGbl.keys()):
if self.dictGbl["sub_experiment_id"] not in ['none']:
member_id = '{}-{}'.format(self.dictGbl['sub_experiment_id'],
self.dictGbl['variant_label'])
else:
member_id = self.dictGbl['variant_label']
cmip6_cv.set_cur_dataset_attribute(cmip6_cv.GLOBAL_ATT_MEMBER_ID, member_id)
# -------------------------------------------------------------------
# Create a dictionary of attributes for the variable
# -------------------------------------------------------------------
try:
self.dictVar = infile.variables[variable_record_name].__dict__
except BaseException:
print(BCOLORS.FAIL)
print("=====================================================================================")
print("The variable " + variable_record_name + " could not be found in file")
print("=====================================================================================")
print(BCOLORS.ENDC)
raise KeyboardInterrupt
# -------------------------------------------------------------------
# Check global attributes
# -------------------------------------------------------------------
self.errors += cmip6_cv.check_requiredattributes(table)
self.errors += cmip6_cv.check_institution(table)
self.errors += cmip6_cv.check_sourceID(table)
self.errors += cmip6_cv.check_experiment(table)
self.errors += cmip6_cv.check_grids(table)
self.errors += cmip6_cv.check_ISOTime()
self.errors += cmip6_cv.check_furtherinfourl(table)
self.errors += cmip6_cv.check_subExpID(table)
for attr in ['branch_time_in_child', 'branch_time_in_parent']:
if attr in list(self.dictGbl.keys()):
self.set_double_value(attr)
if not isinstance(self.dictGbl[attr], numpy.float64):
print(BCOLORS.FAIL)
print("=====================================================================================")
print("{} is not a double: ".format(attr), type(self.dictGbl[attr]))
print("=====================================================================================")
print(BCOLORS.ENDC)
self.errors += 1
for attr in ['realization_index', 'initialization_index', 'physics_index', 'forcing_index']:
if not isinstance(self.dictGbl[attr], numpy.ndarray):
print(BCOLORS.FAIL)
print("=====================================================================================")
print("{} is not an integer: ".format(attr), type(self.dictGbl[attr]))
print("=====================================================================================")
print(BCOLORS.ENDC)
self.errors += 1
self.errors += cmip6_cv.check_parentExpID(table)
for attr in ['table_id', 'variable_id']:
try:
if locals()[attr] != self.dictGbl[attr]:
print(BCOLORS.FAIL)
print("=====================================================================================")
print("{} attribute is not consistent: ".format(attr), self.dictGbl[attr])
print("=====================================================================================")
print(BCOLORS.ENDC)
self.errors += 1
except KeyError:
print(BCOLORS.FAIL)
print("=====================================================================================")
print("{} attribute is missing in global attributes".format(attr))
print("=====================================================================================")
print(BCOLORS.ENDC)
self.errors += 1
# -------------------------------------------------------------------
# Get time axis properties
# -------------------------------------------------------------------
# Get calendar and time units
try:
calendar = infile.variables['time'].calendar
timeunits = infile.variables['time'].units
except BaseException:
calendar = "gregorian"
timeunits = "days since ?"
# Get first and last time bounds
climatology = self.is_climatology(filename)
if climatology:
if cmip6_table.find('Amon') != -1:
variable = '{}Clim'.format(variable)
clim_idx = variable.find('Clim')
if climatology and clim_idx != -1:
var = [variable[:clim_idx]]
try:
if 'bounds' in list(infile.variables['time'].__dict__.keys()):
bndsvar = infile.variables['time'].__dict__['bounds']
elif 'climatology' in list(infile.variables['time'].__dict__.keys()):
bndsvar = infile.variables['time'].__dict__['climatology']
else:
bndsvar = 'time_bnds'
startimebnds = infile.variables[bndsvar][0][0]
endtimebnds = infile.variables[bndsvar][-1][1]
except BaseException:
startimebnds = 0
endtimebnds = 0
try:
startime = infile.variables['time'][0]
endtime = infile.variables['time'][-1]
except BaseException:
startime = 0
endtime = 0
# -------------------------------------------------------------------
# Setup variable
# -------------------------------------------------------------------
varid = cmip6_cv.setup_variable(variable_cmor_entry,
self.dictVar['units'],
self.dictVar['_FillValue'][0],
startime,
endtime,
startimebnds,
endtimebnds)
if varid == -1:
print(BCOLORS.FAIL)
print("=====================================================================================")
print("Could not find variable {} in table {} ".format(variable_cmor_entry, cmip6_table))
print("=====================================================================================")
print(BCOLORS.ENDC)
raise KeyboardInterrupt
# -------------------------------------------------------------------
# Check filename
# -------------------------------------------------------------------
self.errors += cmip6_cv.check_filename(table,
varid,
calendar,
timeunits,
filename)
# -------------------------------------------------------------------
# Check variable attributes
# -------------------------------------------------------------------
cv_attrs = cmip6_cv.list_variable_attributes(varid)
for key in cv_attrs:
if key == "long_name":
continue
if key == "comment":
continue
if key == "cell_measures":
if cv_attrs[key].find("OPT") != -1 or cv_attrs[key].find("MODEL") != -1:
continue
# Is this attribute in file?
if key in list(self.dictVar.keys()):
# Verify that attribute value is equal to file attribute
table_value = cv_attrs[key]
file_value = self.dictVar[key]
# PrePARE accept units of 1 or 1.0 so adjust the table_value
if key == "units":
if (table_value == "1") and (file_value == "1.0"):
table_value = "1.0"
if (table_value == "1.0") and (file_value == "1"):
table_value = "1"
if isinstance(table_value, str) and isinstance(file_value, numpy.ndarray):
if numpy.array([int(value) for value in table_value.split()] == file_value).all():
file_value = True
table_value = True
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(file_value, numpy.ndarray):
file_value = file_value[0]
if isinstance(table_value, float):
if abs(table_value - file_value) <= 0.00001 * abs(table_value):
table_value = file_value
if key == "cell_methods":
idx = file_value.find(" (")
if idx != -1:
file_value = file_value[:idx]
table_value = table_value[:idx]
if key == "cell_measures":
pattern = re.compile('(?P<param>[\w.-]+): (?P<val1>[\w.-]+) OR (?P<val2>[\w.-]+)')
values = re.findall(pattern, table_value)
table_values = [""] # Empty string is allowed in case of useless attribute
if values:
tmp = dict()
for param, val1, val2 in values:
tmp[param] = [str('{}: {}'.format(param, val1)), str('{}: {}'.format(param, val2))]
table_values.extend([' '.join(i) for i in list(itertools.product(*list(tmp.values())))])
if str(file_value) not in list(map(str, table_values)):
print(BCOLORS.FAIL)
print("=====================================================================================")
print("Your file contains \"" + key + "\":\"" + str(file_value) + "\" and")
print("CMIP6 tables requires \"" + key + "\":\"" + str(table_value) + "\".")
print("=====================================================================================")
print(BCOLORS.ENDC)
self.errors += 1
continue
if str(table_value) != str(file_value):
print(BCOLORS.FAIL)
print("=====================================================================================")
print("Your file contains \"" + key + "\":\"" + str(file_value) + "\" and")
print("CMIP6 tables requires \"" + key + "\":\"" + str(table_value) + "\".")
print("=====================================================================================")
print(BCOLORS.ENDC)
self.errors += 1
else:
# That attribute is not in the file
table_value = cv_attrs[key]
if isinstance(table_value, numpy.ndarray):
table_value = table_value[0]
if isinstance(table_value, float):
table_value = "{0:.2g}".format(table_value)
print(BCOLORS.FAIL)
print("=====================================================================================")
print("CMIP6 variable " + variable + " requires \"" + key + "\":\"" + str(table_value) + "\".")
print("=====================================================================================")
print(BCOLORS.ENDC)
self.errors += 1
# Print final message
if self.errors != 0:
print(BCOLORS.FAIL + "└──> :: CV FAIL :: {}".format(ncfile) + BCOLORS.ENDC)
raise KeyboardInterrupt
elif print_all:
print(BCOLORS.OKGREEN + " :: CV SUCCESS :: {}".format(ncfile) + BCOLORS.ENDC)