def submit(self, sequence):
sequence_id = get_sequence_id(sequence)
output_path = self._get_result_path(sequence_id)
if not os.path.isfile(
output_path) and not self._worker.working_on_sequence_id(
sequence_id):
self._worker.submit(sequence)
return sequence_id
def test_result(mock_run):
sequence = "TRY"
sequence_id = get_sequence_id(sequence)
mock_run.return_value = {sequence_id: 'OK'}
worker = Worker()
worker.submit(sequence)
ok_(worker.has_sequence_id(sequence_id))
worker.start()
sleep(1.0)
ok_(worker.result_for_sequence_id(sequence_id) is not None)
def test_job(mock_interproscan):
sequence = "TRY"
sequence_id = get_sequence_id(sequence)
mock_interproscan.return_value = {sequence_id: "OK"}
job_id = job_manager.submit(sequence)
eq_(job_id, sequence_id)
while True:
status = job_manager.get_status(job_id)
if status == 'SUCCESS':
result = job_manager.get_result(job_id)
ok_(result is not None)
return
ok_(status in ['PENDING', 'STARTED'])
def split_proteins(path):
ns_uri = "http://www.ebi.ac.uk/interpro/resources/schemas/interproscan5"
ET.register_namespace("", ns_uri)
ns_map = {'p': ns_uri}
tree = ET.parse(path)
output = {}
for protein in tree.getroot().findall('p:protein', namespaces=ns_map):
sequence = protein.find('p:sequence', namespaces=ns_map).text
matches = ET.Element('protein-matches')
matches.append(protein)
indent_xml(matches)
sequence_id = get_sequence_id(sequence)
output[sequence_id] = ET.tostring(matches).decode('ascii')
return output
def run(self):
_log.info("starting interproscan worker")
while True:
with self._lock:
self._working_sequences = self._queued_sequences
self._queued_sequences = set()
if len(self._working_sequences) <= 0:
continue
from interproscan_web.controllers.job import job_manager
try:
results = interproscan.run(self._working_sequences)
for sequence_id in results:
self._results[sequence_id] = results[sequence_id]
job_manager.store(sequence_id, self._results[sequence_id])
except Exception as e:
with self._lock:
for sequence in self._working_sequences:
self._exceptions[get_sequence_id(sequence)] = e
def run(self, sequences):
fasta_path = tempfile.mktemp()
xml_path = tempfile.mktemp()
job_name = "interproscan_%s" % str(uuid.uuid4())
write_fasta(
fasta_path,
{get_sequence_id(sequence): sequence
for sequence in sequences})
try:
self._execute([
self.interproscan_path, '--goterms', '--formats', 'xml',
'--disable-precalc', '--input', fasta_path, '--outfile',
xml_path, '--seqtype', 'p'
])
return split_proteins(xml_path)
finally:
for p in [fasta_path, xml_path]:
if os.path.isfile(p):
os.remove(p)
def working_on_sequence_id(self, sequence_id):
with self._lock:
return any([get_sequence_id(sequence) == sequence_id for sequence in self._working_sequences])
def has_sequence_id(self, sequence_id):
with self._lock:
return any([get_sequence_id(sequence) == sequence_id for sequence in (self._working_sequences | self._queued_sequences)])