def test_retrieve_input_data_files_success_multiple_input_file(
self, retrieve_files, join):
job_data = JobData({'files': {'TEST_FILE_INPUT': [1, 2]}})
retrieve_files.return_value = {
1: '/scale/input/TEST_FILE_INPUT1',
2: '/scale/input/TEST_FILE_INPUT2'
}
data_files = [
SeedInputFiles({
'name': 'TEST_FILE_INPUT',
'multiple': True,
'required': True,
'mediaTypes': [],
'partial': False
})
]
result = job_data.retrieve_input_data_files(data_files)
self.assertEqual(
result, {
'TEST_FILE_INPUT': [
'/scale/input/TEST_FILE_INPUT1',
'/scale/input/TEST_FILE_INPUT2'
]
})
def test_retrieve_input_data_files_missing_plurality_mismatch(self, retrieve_files, join):
job_data = JobData({'files': {'TEST_FILE_INPUT': [1]}})
retrieve_files.return_value = {}
data_files = [SeedInputFiles(
{'name': 'TEST_FILE_INPUT', 'multiple': True, 'required': True, 'mediaTypes': [], 'partial': False})]
with self.assertRaises(Exception):
job_data.retrieve_input_data_files(data_files)
def test_retrieve_input_data_files_failure_multiple_for_single_input_file(self, retrieve_files, join):
job_data = JobData({'files': {'TEST_FILE_INPUT': [1, 2]}})
retrieve_files.return_value = {1: '/scale/input/TEST_FILE_INPUT1', 2: '/scale/input/TEST_FILE_INPUT2'}
data_files = [SeedInputFiles(
{'name': 'TEST_FILE_INPUT', 'multiple': False, 'required': True, 'mediaTypes': [], 'partial': False})]
with self.assertRaises(Exception):
job_data.retrieve_input_data_files(data_files)
def test_retrieve_input_data_files_missing_file_not_required(self, retrieve_files, join):
job_data = JobData({'files': {}})
retrieve_files.return_value = {}
data_files = [SeedInputFiles(
{'name': 'TEST_FILE_INPUT', 'multiple': False, 'required': False, 'mediaTypes': [],
'partial': False})]
result = job_data.retrieve_input_data_files(data_files)
self.assertEqual(result, {})
def handle(self, *args, **options):
"""See :meth:`django.core.management.base.BaseCommand.handle`.
This method starts the command.
"""
job_id = int(os.environ.get('SCALE_JOB_ID'))
exe_num = int(os.environ.get('SCALE_EXE_NUM'))
logger.info(
'Command starting: scale_pre_steps - Job ID: %d, Execution Number: %d',
job_id, exe_num)
try:
job_exe = self._get_job_exe(job_id, exe_num)
job_interface = job_exe.job_type.get_job_interface()
exe_config = job_exe.get_execution_configuration()
logger.info('Validating mounts...')
job_interface.validate_populated_mounts(exe_config)
logger.info('Validating settings...')
job_interface.validate_populated_settings(exe_config)
logger.info('Validating outputs and workspaces...')
job_interface.validate_workspace_for_outputs(exe_config)
self._generate_input_metadata(job_exe)
job_data = job_exe.job.get_job_data()
job_data = JobData(job_data.get_dict())
logger.info('Setting up input files...')
job_interface.perform_pre_steps(job_data)
logger.info('Ready to execute job: %s',
exe_config.get_args('main'))
except ScaleError as err:
err.log()
sys.exit(err.exit_code)
except Exception as ex:
exit_code = GENERAL_FAIL_EXIT_CODE
err = get_error_by_exception(ex.__class__.__name__)
if err:
err.log()
exit_code = err.exit_code
else:
logger.exception('Error performing pre-job steps')
sys.exit(exit_code)
logger.info('Command completed: scale_pre_steps')
def queue_new_job_for_user(self, job_type, data):
"""Creates a new job for the given type and data at the request of a user. The new job is immediately placed on
the queue. The given job_type model must have already been saved in the database (it must have an ID). The new
job, event, job_exe, and queue models are saved in the database in an atomic transaction. If the data is
invalid, a :class:`job.configuration.data.exceptions.InvalidData` will be thrown.
:param job_type: The type of the new job to create and queue
:type job_type: :class:`job.models.JobType`
:param data: JSON description defining the job data to run on
:type data: dict
:returns: The ID of the new job
:rtype: int
"""
description = {'user': '******'}
event = TriggerEvent.objects.create_trigger_event(
'USER', None, description, timezone.now())
# TODO: Remove old JobData in v6 when we transition to only Seed job types
if 'version' in data and '6' == data['version']:
job_data = JobData(data)
else:
job_data = JobData_1_0(data)
job_id = self.queue_new_job(job_type, job_data, event).id
return job_id
def create(interface, data=None):
"""Instantiate an appropriately typed Job data based on version"""
if JobInterfaceSunset.is_seed(interface):
return JobData(data)
else:
return JobData_1_0(data)
def test_retrieve_files_with_id(self, filter, download):
job_data = JobData({'files': {}})
filter.return_value = [
Mock(id=1, file_name='input_thing_file'),
Mock(id=2, file_name='input_other_file')
]
data_files = {
1: ("/scale/input/INPUT_THING", False),
2: ("/scale/input/INPUT_OTHER", True)
}
result = job_data._retrieve_files(data_files)
self.assertEqual(
result, {
1: "/scale/input/INPUT_THING/input_thing_file",
2: "/scale/input/INPUT_OTHER/input_other_file"
})
def test_store_output_files(self, dummy_store, isfile):
workspace = storage_test_utils.create_workspace()
files = {'OUTPUT_TIFFS': [ProductFileMetadata('OUTPUT_TIFFS', 'outfile0.tif', media_type='image/tiff')]}
job_data = JobData({})
job_config = JobConfiguration()
job_config.add_output_workspace('OUTPUT_TIFFS', workspace.name)
job_exe = Mock()
job_exe.job_type.get_job_configuration.return_value = job_config
results = JobResults()._store_output_data_files(files, job_data, job_exe)
self.assertEqual({'OUTPUT_TIFFS': [1]}, results.files)
