def from_json(cls, value):
"""init this request from json back again
:param value: the json (not string) representation
"""
process = Process.from_json(value['process'])
wps_request = WPSRequest()
wps_request.json = json.loads(value['wps_request'])
wps_response = ExecuteResponse(wps_request=wps_request,
uuid=process.uuid,
process=process)
wps_response.store_status_file = True
new_job = Job(process=Process.from_json(value['process']),
wps_request=wps_request,
wps_response=wps_response)
return new_job
def test_insert_request(self):
requests = self.session.query(RequestInstance)
assert requests.count() == 0
obj = {
'operation': 'execute',
'version': '1.0.0',
'language': 'eng',
'identifier': 'multiple-outputs',
'store_execute': True,
'status': True,
'lineage': True,
'inputs': {
'count': [{
'identifier': 'count',
'type': 'literal',
'data': 3
}]
},
'outputs': {},
'raw': False
}
request = WPSRequest()
request.json = obj
from .processes.metalinkprocess import MultipleOutputs
process = MultipleOutputs()
process.uuid = uuid.uuid4()
store_request(process.uuid, request, process)
requests = self.session.query(RequestInstance)
assert requests.count() == 1
stored_request = pop_first_stored()
job = Job.from_json({
"process": json.loads(stored_request.process),
"wps_request": json.loads(stored_request.request)
})
assert job.process.identifier == "multiple-outputs"
assert str(job.process.uuid) == str(process.uuid)
requests = self.session.query(RequestInstance)
assert requests.count() == 0
def setUp(self):
self.request = WPSRequest()
self.tempfile = tempfile.mktemp()
x = open(self.tempfile, 'w')
x.write("ahoj")
x.close()
def execute_job(self, process_id, wps_inputs, wps_outputs, mode, job_uuid):
"""
Real execution of the process by active Celery Worker.
"""
execution = WPSExecution(version="2.0", url="localhost")
xml_request = execution.buildRequest(process_id,
wps_inputs,
wps_outputs,
mode=mode,
lineage=True)
wps_request = WPSRequest()
wps_request.identifier = process_id
wps_request.set_version("2.0.0")
request_parser = wps_request._post_request_parser(
wps_request.WPS.Execute().tag) # noqa: W0212
request_parser(xml_request)
# NOTE:
# Setting 'status = false' will disable async execution of 'pywps.app.Process.Process'
# but this is needed since this job is running within Celery worker already async
# (daemon process can't have children processes)
# Because if how the code in PyWPS is made, we have to re-enable creation of status file
wps_request.status = "false"
wps_response = super(WorkerService,
self).execute(process_id, wps_request, job_uuid)
wps_response.store_status_file = True
# update execution status with actual status file and apply required references
execution = check_wps_status(
location=wps_response.process.status_location,
settings=self.settings)
execution.request = xml_request
return execution
def _handler(self, request: WPSRequest, response: ExecuteResponse):
self.write_log("Processing started", response, 5)
variable = self.get_input_or_none(request.inputs, "variable")
rcp = self.get_input_or_none(request.inputs, "rcp")
lat0 = self.get_input_or_none(request.inputs, "lat0")
lon0 = self.get_input_or_none(request.inputs, "lon0")
output_format = request.inputs["output_format"][0].data
output_filename = f"BCCAQv2_subset_{lat0}_{lon0}"
self.write_log("Fetching BCCAQv2 datasets", response, 6)
request.inputs = get_bccaqv2_inputs(request.inputs, variable, rcp)
self.write_log("Running subset", response, 7)
threads = int(configuration.get_config_value("finch",
"subset_threads"))
metalink = self.subset(
request.inputs,
response,
start_percentage=7,
end_percentage=90,
threads=threads,
)
if not metalink.files:
message = "No data was produced when subsetting using the provided bounds."
raise ProcessError(message)
self.write_log("Subset done, creating zip file", response)
output_files = [mf.file for mf in metalink.files]
if output_format == "csv":
csv_files, metadata_folder = netcdf_to_csv(
output_files,
output_folder=Path(self.workdir),
filename_prefix=output_filename,
)
output_files = csv_files + [metadata_folder]
output_zip = Path(self.workdir) / (output_filename + ".zip")
def log(message_, percentage_):
self.write_log(message_, response, percentage_)
zip_files(output_zip,
output_files,
log_function=log,
start_percentage=90)
response.outputs["output"].file = output_zip
self.write_log("Processing finished successfully", response, 99)
return response
def setUp(self):
self.uuid = uuid.uuid1()
self.dummy_process = InOut()
self.dummy_process._set_uuid(self.uuid)
self.dummy_process.set_workdir('/tmp')
self.wps_request = WPSRequest()
self.wps_response = ExecuteResponse(self.wps_request, self.uuid,
process=self.dummy_process)
self.job = Job(
process=self.dummy_process,
wps_request=self.wps_request,
wps_response=self.wps_response)
def test_parse_wps_input():
obj = {
"operation": "execute",
"version": "1.0.0",
"language": "eng",
"identifier": "subset",
"identifiers": "subset", # TODO: why identifierS?
"store_execute": True,
"status": True,
"lineage": True,
"inputs": {
"time": [{
"identifier": "time",
"type": "literal",
"data_type": "string",
"allowed_values": [{
"type": "anyvalue"
}],
"data": "1970/1980",
}],
"time_components": [{
"identifier": "time",
"type": "literal",
"data_type": "string",
"allowed_values": [{
"type": "anyvalue"
}],
"data": "year:1970,1980|month=01,02,03",
}],
},
"outputs": {},
"raw": False,
}
request = WPSRequest()
request.json = obj
assert parse_wps_input(request.inputs, "time", default=None) == "1970/1980"
assert (parse_wps_input(request.inputs, "time_components",
default=None) == "year:1970,1980|month=01,02,03")
def setUp(self):
def handler(request, response):
response.outputs['output'].data = '42'
return response
self.uuid = 1234
self.dummy_process = Process(
handler=handler,
identifier='dummy',
title='Dummy Process',
outputs=[LiteralOutput('output', 'Output', data_type='string')])
self.wps_request = WPSRequest()
self.wps_response = ExecuteResponse(self.wps_request, self.uuid,
process=self.dummy_process)
def test_json_in(self):
obj = {
'operation': 'getcapabilities',
'version': '1.0.0',
'language': 'eng',
'identifier': 'ahoj',
'identifiers': 'ahoj', # TODO: why identifierS?
'store_execute': True,
'status': True,
'lineage': True,
'inputs': {
'myin': [{
'identifier': 'myin',
'type': 'complex',
'supported_formats': [{
'mime_type': 'tralala'
}],
'file': self.tempfile,
'data_format': {
'mime_type': 'tralala'
}
}],
'myliteral': [{
'identifier': 'myliteral',
'type': 'literal',
'data_type': 'integer',
'allowed_values': [{
'type': 'anyvalue'
}],
'data': 1
}]
},
'outputs': {},
'raw': False
}
self.request = WPSRequest()
self.request.json = obj
self.assertEqual(self.request.inputs['myliteral'][0].data, 1,
'Data are in the file')
self.assertEqual(self.request.inputs['myin'][0].data, 'ahoj',
'Data are in the file')
self.assertListEqual(
self.request.inputs['myliteral'][0].allowed_values, [AnyValue()],
'Any value not set')
self.assertTrue(self.request.inputs['myliteral'][0].any_value,
'Any value set')
def test_wps3_process_step_io_data_or_href():
"""
Validates that 'data' literal values and 'href' file references are both handled as input for workflow steps
corresponding to a WPS-3 process.
"""
test_process = "test-wps3-process-step-io-data-or-href"
test_reached_parse_inputs = False # toggle at operation just before what is being tested here
test_cwl_inputs = {
"single-data-value":
1,
"multi-data-values": [2, 3],
"single-href-file": {
"location": "https://random-place"
},
"multi-href-files": [{
"location": "https://random-place"
}, {
"location": "https://another-place"
}]
}
expected_wps_inputs = [
{
"id": "single-data-value",
"data": 1
},
{
"id": "multi-data-values",
"data": 2
},
{
"id": "multi-data-values",
"data": 3
},
{
"id": "single-href-file",
"href": "https://random-place"
},
{
"id": "multi-href-files",
"href": "https://random-place"
},
{
"id": "multi-href-files",
"href": "https://another-place"
},
]
class TestDoneEarlyExit(Exception):
"""Dummy exception to raise to skip further processing steps after the portion to evaluate was reached."""
def mock_wps_request(method, url, *_, **kwargs):
nonlocal test_reached_parse_inputs
method = method.upper()
if url.endswith("/visibility"):
resp = Response()
resp.status_code = 200
resp._content = json.dumps({
"value": VISIBILITY_PUBLIC
},
ensure_ascii=False).encode()
resp.headers = {"Content-Type": CONTENT_TYPE_APP_NETCDF}
resp.encoding = None
if method == "PUT":
test_reached_parse_inputs = True # last operation before parsing I/O is setting visibility
return resp
if method == "POST" and url.endswith(test_process + "/jobs"):
# actual evaluation of intended handling of CWL inputs conversion to WPS-3 execute request
assert kwargs.get("json", {}).get("inputs") == expected_wps_inputs
raise TestDoneEarlyExit(
"Expected exception raised to skip executed job status monitoring"
)
raise AssertionError("unhandled mocked 'make_request' call")
def mock_update_status(*_, **__):
return None
mock_data_sources = {
"localhost": {
"netloc": "localhost",
"ades": "https://localhost:4001",
"default": True
}
}
with mock.patch(
"weaver.processes.wps_process_base.WpsProcessInterface.make_request",
side_effect=mock_wps_request):
with mock.patch("weaver.processes.sources.fetch_data_sources",
return_value=mock_data_sources):
wps_params = {
"service": "wps",
"request": "execute",
"identifier": test_process,
"version": "1.0.0"
}
req = Request(method="GET", params=wps_params)
setattr(req, "args", wps_params)
req = WPSRequest(req)
wps = Wps3Process({}, {}, test_process, req, mock_update_status)
try:
wps.execute(test_cwl_inputs, "", {})
except TestDoneEarlyExit:
pass # successful test / expected handling
except Exception as exc: # noqa
if not test_reached_parse_inputs:
msg = "Prior error was raised [{}], could not evaluate intended handling of inputs".format(
exc)
pytest.fail(msg)
msg = "Other error was raised [{}], inputs where not correctly handled somewhere".format(
exc)
pytest.fail(msg)
def _handler(self, request: WPSRequest, response: ExecuteResponse):
convert_to_csv = request.inputs["output_format"][0].data == "csv"
if not convert_to_csv:
del self.status_percentage_steps["convert_to_csv"]
write_log(self, "Processing started", process_step="start")
# Temporary backward-compatibility adjustment.
# Remove me when lon0 and lat0 are removed
lon, lat, lon0, lat0 = [
single_input_or_none(request.inputs, var)
for var in "lon lat lon0 lat0".split()
]
if not (lon and lat or lon0 and lat0):
raise ProcessError(
"Provide both lat and lon or both lon0 and lat0.")
request.inputs.setdefault("lon", request.inputs.get("lon0"))
request.inputs.setdefault("lat", request.inputs.get("lat0"))
# End of 'remove me'
output_filename = make_output_filename(self, request.inputs)
write_log(self, "Fetching datasets")
variable = request.inputs["variable"][0].data
variables = None if variable is None else [variable]
rcp = single_input_or_none(request.inputs, "rcp")
dataset_name = single_input_or_none(request.inputs, "dataset_name")
request.inputs["resource"] = get_datasets(dataset_name,
workdir=self.workdir,
variables=variables,
rcp=rcp)
write_log(self, "Running subset", process_step="subset")
output_files = finch_subset_gridpoint(
self,
netcdf_inputs=request.inputs["resource"],
request_inputs=request.inputs,
)
if not output_files:
message = "No data was produced when subsetting using the provided bounds."
raise ProcessError(message)
if convert_to_csv:
write_log(self,
"Converting outputs to csv",
process_step="convert_to_csv")
csv_files, metadata_folder = netcdf_file_list_to_csv(
output_files,
output_folder=Path(self.workdir),
filename_prefix=output_filename,
csv_precision=single_input_or_none(request.inputs,
"csv_precision"))
output_files = csv_files + [metadata_folder]
write_log(self, "Zipping outputs", process_step="zip_outputs")
output_zip = Path(self.workdir) / (output_filename + ".zip")
def _log(message, percentage):
write_log(self, message, subtask_percentage=percentage)
zip_files(output_zip, output_files, log_function=_log)
response.outputs["output"].file = output_zip
write_log(self,
"Processing finished successfully",
process_step="done")
return response
def _handler(self, request: WPSRequest, response: ExecuteResponse):
self.write_log("Processing started", response, 5)
lat = request.inputs["lat"][0].data
lon = request.inputs["lon"][0].data
output_format = request.inputs["output_format"][0].data
output_filename = f"BCCAQv2_subset_heat_wave_frequency_{lat}_{lon}"
self.write_log("Fetching BCCAQv2 datasets", response, 6)
tasmin_inputs = get_bccaqv2_inputs(request.inputs, "tasmin")["resource"]
tasmax_inputs = get_bccaqv2_inputs(request.inputs, "tasmax")["resource"]
request.inputs["resource"] = tasmin_inputs + tasmax_inputs
self.write_log("Running subset", response, 7)
threads = int(configuration.get_config_value("finch", "subset_threads"))
metalink = self.subset(
request.inputs,
response,
start_percentage=7,
end_percentage=50,
threads=threads,
)
if not metalink.files:
message = "No data was produced when subsetting using the provided bounds."
raise ProcessError(message)
self.write_log("Subset done, calculating indices", response)
all_files = [Path(f.file) for f in metalink.files]
start_percentage = 50
end_percentage = 95
pairs = list(self._make_tasmin_tasmax_pairs(all_files))
n_pairs = len(pairs)
output_files = []
for n, (tasmin, tasmax) in enumerate(pairs):
percentage = start_percentage + int(
n / n_pairs * (end_percentage - start_percentage)
)
self.write_log(
f"Processing file {n + 1} of {n_pairs}", response, percentage
)
compute_inputs = [i.identifier for i in self.indices_process.inputs]
inputs = {k: v for k, v in request.inputs.items() if k in compute_inputs}
inputs["tasmin"] = deque([make_nc_input("tasmin")], maxlen=1)
inputs["tasmin"][0].file = str(tasmin)
inputs["tasmax"] = deque([make_nc_input("tasmax")], maxlen=1)
inputs["tasmax"][0].file = str(tasmax)
out = self.compute_indices(self.indices_process.xci, inputs)
out_fn = Path(self.workdir) / tasmin.name.replace(
"tasmin", "heat_wave_frequency"
)
out.to_netcdf(out_fn)
output_files.append(out_fn)
if output_format == "csv":
csv_files, metadata_folder = netcdf_to_csv(
output_files,
output_folder=Path(self.workdir),
filename_prefix=output_filename,
)
output_files = csv_files + [metadata_folder]
output_zip = Path(self.workdir) / (output_filename + ".zip")
def log(message_, percentage_):
self.write_log(message_, response, percentage_)
zip_files(output_zip, output_files, log_function=log, start_percentage=90)
response.outputs["output"].file = output_zip
self.write_log("Processing finished successfully", response, 99)
return response
def test_json_inout_datetime(self):
obj = {
'operation': 'getcapabilities',
'version': '1.0.0',
'language': 'eng',
'identifier': 'moinmoin',
'identifiers': 'moinmoin', # TODO: why identifierS?
'store_execute': True,
'status': True,
'lineage': True,
'inputs': {
'datetime': [{
'identifier': 'datetime',
'type': 'literal',
'data_type': 'dateTime',
'data': '2017-04-20T12:00:00',
'allowed_values': [{
'type': 'anyvalue'
}],
}],
'date': [{
'identifier': 'date',
'type': 'literal',
'data_type': 'date',
'data': '2017-04-20',
'allowed_values': [{
'type': 'anyvalue'
}],
}],
'time': [{
'identifier': 'time',
'type': 'literal',
'data_type': 'time',
'data': '09:00:00',
'allowed_values': [{
'type': 'anyvalue'
}],
}],
},
'outputs': {},
'raw': False
}
self.request = WPSRequest()
self.request.json = obj
self.assertEqual(self.request.inputs['datetime'][0].data,
datetime.datetime(2017, 4, 20, 12), 'Datatime set')
self.assertEqual(self.request.inputs['date'][0].data,
datetime.date(2017, 4, 20), 'Data set')
self.assertEqual(self.request.inputs['time'][0].data,
datetime.time(9, 0, 0), 'Time set')
# dump to json and reload
dump = self.request.json
self.request.json = json.loads(dump)
self.assertEqual(self.request.inputs['datetime'][0].data,
datetime.datetime(2017, 4, 20, 12), 'Datatime set')
self.assertEqual(self.request.inputs['date'][0].data,
datetime.date(2017, 4, 20), 'Data set')
self.assertEqual(self.request.inputs['time'][0].data,
datetime.time(9, 0, 0), 'Time set')
def test_stdout_stderr_logging_for_commandline_tool_failure():
"""
Execute a process and assert that stderr is correctly logged to log file.
"""
process = Process({
"title": "test-stdout-stderr",
"id": "test-stdout-stderr",
"package": {
"cwlVersion": "v1.0",
"class": "CommandLineTool",
"baseCommand": "not_existing_command",
"inputs": {
"message": {
"type": "string",
"inputBinding": {
"position": 1
}
}
},
"outputs": {}
}
})
payload = process
package = process["package"]
title = process["title"]
identifier = process["id"]
# WPSPackage._handle()
log_file = tempfile.NamedTemporaryFile()
status_location = log_file.name
workdir = tempfile.TemporaryDirectory()
class TestWpsPackage(WpsPackage):
@property
def status_location(self):
return status_location
wps_package_instance = TestWpsPackage(identifier=identifier,
title=title,
payload=payload,
package=package)
wps_package_instance.set_workdir(workdir.name)
# WPSRequest mock
wps_request = WPSRequest()
wps_request.json = {
"identifier": "test-stdout-stderr",
"operation": "execute",
"version": "1.0.0",
"language": "null",
"identifiers": "null",
"store_execute": "true",
"status": "true",
"lineage": "true",
"raw": "false",
"inputs": {
"message": [{
"identifier": "message",
"title": "A dummy message",
"type": "literal",
"data_type": "string",
"data": "Dummy message",
"allowed_values": [],
}]
},
"outputs": {}
}
# ExecuteResponse mock
wps_response = type("", (object, ), {
"_update_status": lambda *_, **__: 1
})()
from weaver.exceptions import PackageExecutionError
try:
wps_package_instance._handler(wps_request, wps_response)
except PackageExecutionError as exception:
assert "Completed permanentFail" in exception.args[0]
else:
fail(
"\"wps_package._handler()\" was expected to throw \"PackageExecutionError\" exception"
)
def test_stdout_stderr_logging_for_commandline_tool_success():
"""
Execute a process and assert that stdout is correctly logged to log file.
"""
process = Process({
"title": "test-stdout-stderr",
"id": "test-stdout-stderr",
"package": {
"cwlVersion": "v1.0",
"class": "CommandLineTool",
"baseCommand": "echo",
"inputs": {
"message": {
"type": "string",
"inputBinding": {
"position": 1
}
}
},
"outputs": {}
}
})
payload = process
package = process["package"]
title = process["title"]
identifier = process["id"]
# WPSPackage._handle()
log_file = tempfile.NamedTemporaryFile()
status_location = log_file.name
workdir = tempfile.TemporaryDirectory()
class TestWpsPackage(WpsPackage):
@property
def status_location(self):
return status_location
wps_package_instance = TestWpsPackage(identifier=identifier,
title=title,
payload=payload,
package=package)
wps_package_instance.set_workdir(workdir.name)
# WPSRequest mock
wps_request = WPSRequest()
wps_request.json = {
"identifier": "test-stdout-stderr",
"operation": "execute",
"version": "1.0.0",
"language": "null",
"identifiers": "null",
"store_execute": "true",
"status": "true",
"lineage": "true",
"raw": "false",
"inputs": {
"message": [{
"identifier": "message",
"title": "A dummy message",
"type": "literal",
"data_type": "string",
"data": "Dummy message",
"allowed_values": [],
}]
},
"outputs": {}
}
# ExecuteResponse mock
wps_response = type("", (object, ), {
"_update_status": lambda *_, **__: 1
})()
wps_package_instance._handler(wps_request, wps_response)
# log assertions
with open(status_location + ".log", "r") as file:
log_data = file.read()
assert "Dummy message" in log_data
def _handler(self, request: WPSRequest, response: ExecuteResponse):
convert_to_csv = request.inputs["output_format"][0].data == "csv"
if not convert_to_csv:
del self.status_percentage_steps["convert_to_csv"]
write_log(self, "Processing started", process_step="start")
output_filename = make_output_filename(self, request.inputs)
write_log(self, "Fetching datasets")
variable = request.inputs["variable"][0].data
variables = None if variable is None else [variable]
rcp = single_input_or_none(request.inputs, "rcp")
dataset_name = single_input_or_none(request.inputs, "dataset_name")
request.inputs["resource"] = get_datasets(dataset_name,
workdir=self.workdir,
variables=variables,
rcp=rcp)
write_log(self, "Running subset", process_step="subset")
output_files = finch_subset_bbox(
self,
netcdf_inputs=request.inputs["resource"],
request_inputs=request.inputs,
)
if not output_files:
message = "No data was produced when subsetting using the provided bounds."
raise ProcessError(message)
if convert_to_csv:
write_log(self,
"Converting outputs to csv",
process_step="convert_to_csv")
csv_files, metadata_folder = netcdf_file_list_to_csv(
output_files,
output_folder=Path(self.workdir),
filename_prefix=output_filename,
)
output_files = csv_files + [metadata_folder]
write_log(self, "Zipping outputs", process_step="zip_outputs")
output_zip = Path(self.workdir) / (output_filename + ".zip")
def _log(message, percentage):
write_log(self, message, subtask_percentage=percentage)
zip_files(output_zip, output_files, log_function=_log)
response.outputs["output"].file = output_zip
write_log(self,
"Processing finished successfully",
process_step="done")
return response
