def test_no_tags(self):
'''Tests calling get_data_type_tags() with no tags'''
ingest = Ingest()
tags = ingest.get_data_type_tags()
self.assertSetEqual(tags, set())
def test_no_tags(self):
'''Tests calling get_data_type_tags() with no tags'''
ingest = Ingest()
tags = ingest.get_data_type_tags()
self.assertSetEqual(tags, set())
def test_deduplicate_ingest_list_no_existing(self, ingests_by_scan):
"""Tests calling S3Scanner._deduplicate_ingest_list() without existing"""
ingests_by_scan.return_value = []
ingests = [Ingest(file_name='test1'), Ingest(file_name='test2')]
final_ingests = S3Scanner._deduplicate_ingest_list(None, ingests)
self.assertItemsEqual(ingests, final_ingests)
def test_deduplicate_ingest_list_with_duplicate_file_names(
self, ingests_by_scan):
"""Tests calling S3Scanner._deduplicate_ingest_list() with duplicates"""
ingests_by_scan.return_value = []
ingests = [Ingest(file_name='test1'), Ingest(file_name='test1')]
final_ingests = S3Scanner._deduplicate_ingest_list(None, ingests)
self.assertEquals(len(final_ingests), 1)
self.assertEquals(final_ingests[0].file_name, 'test1')
def test_deduplicate_ingest_list_with_existing_no_other_dups(
self, ingests_by_scan):
"""Tests calling S3Scanner._deduplicate_ingest_list() with existing and no other dups"""
ingests_by_scan.return_value = [Ingest(file_name='test1')]
ingests = [Ingest(file_name='test1'), Ingest(file_name='test2')]
final_ingests = S3Scanner._deduplicate_ingest_list(None, ingests)
self.assertEquals(len(final_ingests), 1)
self.assertEquals(final_ingests[0].file_name, 'test2')
def test_tags(self):
'''Tests calling get_data_type_tags() with tags'''
ingest = Ingest(data_type='A,B,c')
tags = ingest.get_data_type_tags()
correct_set = set()
correct_set.add('A')
correct_set.add('B')
correct_set.add('c')
self.assertSetEqual(tags, correct_set)
def test_tags(self):
"""Tests calling get_data_type_tags() with tags"""
ingest = Ingest(data_type_tags=['A', 'B', 'c'])
tags = ingest.get_data_type_tags()
correct_set = set()
correct_set.add('A')
correct_set.add('B')
correct_set.add('c')
self.assertSetEqual(tags, correct_set)
def test_same_tag(self):
"""Tests calling add_data_type_tag() with the same tag twice"""
ingest = Ingest()
ingest.add_data_type_tag('Hello1')
ingest.add_data_type_tag('Hello1')
tags = ingest.get_data_type_tags()
correct_set = set()
correct_set.add('Hello1')
self.assertSetEqual(tags, correct_set)
def test_tags(self):
'''Tests calling get_data_type_tags() with tags'''
ingest = Ingest(data_type='A,B,c')
tags = ingest.get_data_type_tags()
correct_set = set()
correct_set.add('A')
correct_set.add('B')
correct_set.add('c')
self.assertSetEqual(tags, correct_set)
def test_invalid(self):
"""Tests calling add_data_type_tag() with invalid tags"""
ingest = Ingest()
self.assertRaises(InvalidDataTypeTag, ingest.add_data_type_tag, 'my.invalid.tag')
self.assertRaises(InvalidDataTypeTag, ingest.add_data_type_tag, 'my\invalid\tag!')
def _create_ingest(self, file_name):
"""Creates a new ingest for the given file name. The database save is the caller's responsibility.
:param file_name: The name of the file being ingested
:type file_name: string
:returns: The new ingest model
:rtype: :class:`ingest.models.Ingest`
"""
ingest = Ingest()
ingest.file_name = file_name
ingest.strike_id = self.strike_id
ingest.media_type = get_media_type(file_name)
ingest.workspace = self._monitored_workspace
logger.info('New file on %s: %s', ingest.workspace.name, file_name)
return ingest
def test_valid(self):
'''Tests calling add_data_type_tag() with valid tags'''
ingest = Ingest()
ingest.add_data_type_tag('Hello1')
ingest.add_data_type_tag('foo_BAR')
tags = ingest.get_data_type_tags()
correct_set = set()
correct_set.add('Hello1')
correct_set.add('foo_BAR')
self.assertSetEqual(tags, correct_set)
def test_valid(self):
'''Tests calling add_data_type_tag() with valid tags'''
ingest = Ingest()
ingest.add_data_type_tag('Hello1')
ingest.add_data_type_tag('foo_BAR')
tags = ingest.get_data_type_tags()
correct_set = set()
correct_set.add('Hello1')
correct_set.add('foo_BAR')
self.assertSetEqual(tags, correct_set)
def handle(self, *args, **options):
"""See :meth:`django.core.management.base.BaseCommand.handle`.
This method migrates existing data files into scale.
"""
logger.info(u'Command starting: migratedata')
workspace, workspace_path, local_path, data_types = None, None, None, []
if options['workspace'] is not None and options[
'workspace_path'] is not None:
workspace, workspace_path = options['workspace'], options[
'workspace_path']
tmp = Workspace.objects.filter(name=workspace)
if tmp.count() > 0:
workspace = tmp.first()
else:
workspace = Workspace.objects.get(id=int(workspace))
else:
logger.error('Must specify workspace and workspace-path.')
return False
if options['data_type'] is not None:
data_types.extend(options['data_type'])
mnt_dirs = None
if options['local_path'] is not None:
local_path = options['local_path']
else: # mount
mnt_dirs = "/tmp", tempfile.mkdtemp()
workspace.setup_download_dir(*mnt_dirs)
local_path = os.path.join(mnt_dirs[1], workspace_path)
logger.info("Ingesting files from %s/%s", workspace.name,
workspace_path)
filenames = self.generate_file_list(local_path, options['include'],
options['exclude'])
logger.info("Found %d files", len(filenames))
# prepare for ingest ala strike
ingest_records = {}
for filename in filenames:
logger.info("Generating ingest record for %s" % filename)
ingest = Ingest()
ingest.file_name = os.path.basename(filename)
ingest.file_path = os.path.join(
workspace_path, os.path.relpath(filename, local_path))
ingest.transfer_started = datetime.utcfromtimestamp(
os.path.getatime(filename))
ingest.file_size = ingest.bytes_transferred = os.path.getsize(
filename)
ingest.transfer_ended = timezone.now()
ingest.media_type = get_media_type(filename)
ingest.workspace = workspace
for data_type in data_types:
ingest.add_data_type_tag(data_type)
ingest.status = 'TRANSFERRED'
if options['no_commit']:
s = IngestDetailsSerializer()
logger.info(s.to_representation(ingest))
else:
ingest.save()
ingest_records[filename] = ingest.id
logging.info("Ingests records created")
# start ingest tasks for all the files
if not options['no_commit']:
logging.info("Starting ingest tasks")
for filename in filenames:
ingest = Ingest.objects.get(id=ingest_records[filename])
logging.info("Processing ingest %s" % ingest.file_name)
with transaction.atomic():
ingest.ingest_started = timezone.now()
sf = ingest.source_file = SourceFile.create()
sf.update_uuid(ingest.file_name)
for tag in ingest.get_data_type_tags():
sf.add_data_type_tag(tag)
sf.media_type = ingest.media_type
sf.file_name = ingest.file_name
sf.file_size = ingest.file_size
sf.file_path = ingest.file_path
sf.workspace = workspace
sf.is_deleted = False
sf.deleted = None
sf.save()
sf.set_countries()
sf.save()
ingest.status = 'INGESTED'
ingest.ingest_ended = timezone.now()
ingest.source_file = sf
ingest.save()
IngestTriggerHandler().process_ingested_source_file(
ingest.source_file, ingest.ingest_ended)
logging.info(
"Ingests processed, monitor the queue for triggered jobs.")
if mnt_dirs is not None:
workspace.cleanup_download_dir(*mnt_dirs)
logger.info(u'Command completed: migratedata')
def handle(self, *args, **options):
"""See :meth:`django.core.management.base.BaseCommand.handle`.
This method migrates existing data files into scale.
"""
logger.info(u'Command starting: migratedata')
workspace, workspace_path, local_path, data_types = None, None, None, []
if options['workspace'] is not None and options['workspace_path'] is not None:
workspace, workspace_path = options['workspace'], options['workspace_path']
tmp = Workspace.objects.filter(name=workspace)
if tmp.count() > 0:
workspace = tmp.first()
else:
workspace = Workspace.objects.get(id=int(workspace))
else:
logger.error('Must specify workspace and workspace-path.')
return False
if options['data_type'] is not None:
data_types.extend(options['data_type'])
mnt_dirs = None
if options['local_path'] is not None:
local_path = options['local_path']
else: # mount
mnt_dirs = "/tmp", tempfile.mkdtemp()
workspace.setup_download_dir(*mnt_dirs)
local_path = os.path.join(mnt_dirs[1], workspace_path)
logger.info("Ingesting files from %s/%s", workspace.name, workspace_path)
filenames = self.generate_file_list(local_path, options['include'], options['exclude'])
logger.info("Found %d files", len(filenames))
# prepare for ingest ala strike
ingest_records = {}
for filename in filenames:
logger.info("Generating ingest record for %s" % filename)
ingest = Ingest()
ingest.file_name = os.path.basename(filename)
ingest.transfer_path = filename
ingest.file_path = os.path.join(workspace_path, os.path.relpath(filename, local_path))
ingest.transfer_started = datetime.utcfromtimestamp(os.path.getatime(filename))
ingest.file_size = ingest.bytes_transferred = os.path.getsize(filename)
ingest.transfer_ended = datetime.utcnow()
ingest.media_type = get_media_type(filename)
ingest.workspace = workspace
for data_type in data_types:
ingest.add_data_type_tag(data_type)
ingest.status = 'TRANSFERRED'
if options['no_commit']:
s = IngestDetailsSerializer()
logger.info(s.to_representation(ingest))
else:
ingest.save()
ingest_records[filename] = ingest.id
logging.info("Ingests records created")
# start ingest tasks for all the files
if not options['no_commit']:
logging.info("Starting ingest tasks")
for filename in filenames:
ingest = Ingest.objects.get(id=ingest_records[filename])
logging.info("Processing ingest %s" % ingest.file_name)
with transaction.atomic():
ingest.ingest_started = datetime.utcnow()
sf = ingest.source_file = SourceFile()
sf.update_uuid(ingest.file_name)
for tag in ingest.get_data_type_tags():
sf.add_data_type_tag(tag)
sf.media_type = ingest.media_type
sf.file_name = ingest.file_name
sf.file_size = ingest.file_size
sf.file_path = ingest.file_path
sf.workspace = workspace
sf.is_deleted = False
sf.deleted = None
sf.save()
sf.set_countries()
sf.save()
ingest.status = 'INGESTED'
ingest.ingest_ended = datetime.utcnow()
ingest.source_file = sf
ingest.save()
IngestTriggerHandler().process_ingested_source_file(ingest.source_file, ingest.ingest_ended)
logging.info("Ingests processed, monitor the queue for triggered jobs.")
if mnt_dirs is not None:
workspace.cleanup_download_dir(*mnt_dirs)
logger.info(u'Command completed: migratedata')
def _process_file(self, file_name, ingest):
'''Processes the given file in the Strike directory. The file_name
argument represents a file in the Strike directory to process. If
file_name is None, then the ingest argument represents an ongoing
transfer where the file is unexpectedly not in the Strike directory.
If file_name is not None and ingest is None, then this is a
new transfer without an ingest record yet. If both arguments are None
an exception is thrown.
:param file_name: The name of the file to process (possibly None)
:type file_name: str
:param ingest: The ingest model for the file (possibly None)
:type ingest: :class:`ingest.models.Ingest`
'''
if file_name is None and ingest is None:
raise Exception('Nothing for Strike to process')
if file_name is None:
file_name = ingest.file_name
file_path = os.path.join(self.strike_dir, file_name)
final_name = self._final_filename(file_name)
# Create ingest model for new transfer
if ingest is None:
msg = 'New file %s has arrived, creating ingest for %s'
logger.info(msg, file_path, final_name)
ingest = Ingest()
# Ingest model should record the actual name of the file (no
# temporary suffix)
ingest.file_name = final_name
ingest.strike_id = self.strike_id
# TODO: investigate better way to get start time of transfer
last_access = os.path.getatime(file_path)
ingest.transfer_path = os.path.join(self.strike_dir, final_name)
ingest.transfer_started = datetime.utcfromtimestamp(last_access)
if ingest.status == 'TRANSFERRING':
# Update bytes transferred
size = os.path.getsize(file_path)
ingest.bytes_transferred = size
# Ensure that file is still in Strike dir as expected
if not os.path.exists(file_path):
msg = '%s was being transferred, but the file is now lost'
logger.error(msg, file_path)
ingest.status = 'ERRORED'
ingest.save()
logger.info('Ingest for %s marked as ERRORED', final_name)
return
if self._is_still_transferring(file_name):
# Update with current progress of the transfer
ingest.save()
logger.info('%s is still transferring, progress updated', file_path)
else:
# Transfer is complete, will move on to next section
self._complete_transfer(ingest, size)
if ingest.status == 'TRANSFERRED':
if ingest.ingest_path:
self._prepare_file_for_ingest(ingest)
else:
self._defer_file(ingest)
elif not ingest.status == 'TRANSFERRING':
msg = 'Strike not expecting to process file with status %s'
raise Exception(msg, ingest.status)
def _process_file(self, file_name, ingest):
'''Processes the given file in the Strike directory. The file_name
argument represents a file in the Strike directory to process. If
file_name is None, then the ingest argument represents an ongoing
transfer where the file is unexpectedly not in the Strike directory.
If file_name is not None and ingest is None, then this is a
new transfer without an ingest record yet. If both arguments are None
an exception is thrown.
:param file_name: The name of the file to process (possibly None)
:type file_name: str
:param ingest: The ingest model for the file (possibly None)
:type ingest: :class:`ingest.models.Ingest`
'''
if file_name is None and ingest is None:
raise Exception('Nothing for Strike to process')
if file_name is None:
file_name = ingest.file_name
file_path = os.path.join(self.strike_dir, file_name)
final_name = self._final_filename(file_name)
# Create ingest model for new transfer
if ingest is None:
msg = 'New file %s has arrived, creating ingest for %s'
logger.info(msg, file_path, final_name)
ingest = Ingest()
# Ingest model should record the actual name of the file (no
# temporary suffix)
ingest.file_name = final_name
ingest.strike_id = self.strike_id
# TODO: investigate better way to get start time of transfer
last_access = os.path.getatime(file_path)
ingest.transfer_path = os.path.join(self.strike_dir, final_name)
ingest.transfer_started = datetime.utcfromtimestamp(last_access)
if ingest.status == 'TRANSFERRING':
# Update bytes transferred
size = os.path.getsize(file_path)
ingest.bytes_transferred = size
# Ensure that file is still in Strike dir as expected
if not os.path.exists(file_path):
msg = '%s was being transferred, but the file is now lost'
logger.error(msg, file_path)
ingest.status = 'ERRORED'
ingest.save()
logger.info('Ingest for %s marked as ERRORED', final_name)
return
if self._is_still_transferring(file_name):
# Update with current progress of the transfer
ingest.save()
logger.info('%s is still transferring, progress updated', file_path)
else:
# Transfer is complete, will move on to next section
self._complete_transfer(ingest, size)
if ingest.status == 'TRANSFERRED':
if ingest.ingest_path:
self._prepare_file_for_ingest(ingest)
else:
self._defer_file(ingest)
elif not ingest.status == 'TRANSFERRING':
msg = 'Strike not expecting to process file with status %s'
raise Exception(msg, ingest.status)
