def get_source_worker(self):
"""
Returns a worker who has sent this sample.
"""
return Sample.get_worker(
source_type=self.source_type,
source_val=self.source_val,
)
def create_classify_sample(result, source_type, create_classified=True,
label='', source_val='', *args, **kwargs):
"""
Creates classified sample from existing sample, therefore we don't need
web extraction.
"""
# We are given a tuple (extraction result, sample id)
extraction_result = result[0]
# If extraction failed - return
if not extraction_result:
return False
sample_id = result[1]
# Don't classify already classified samples
if label:
return sample_id
if create_classified:
try:
sample = Sample.objects.get(id=sample_id)
if not label:
label = ''
# Proper sample entry
class_sample = ClassifiedSample.objects.create(
job=sample.job,
url=sample.url,
sample=sample,
label=label,
source_type=source_type,
source_val=source_val,
)
worker = Sample.get_worker(source_type=source_type,
source_val=source_val)
if worker:
# Update cache
worker.get_urls_collected_count_for_job(sample.job, cache=False)
# Sample created sucesfully - pushing event.
send_event(
"EventNewClassifySample",
sample_id=class_sample.id,
)
except DatabaseError, e:
# Retry process on db error, such as 'Database is locked'
create_classify_sample.retry(exc=e,
countdown=min(60 * 2 ** current.request.retries, 60 * 60 * 24))
def forwards(self, orm):
"Write your forwards methods here."
# Note: Remember to use orm['appname.ModelName'] rather than "from appname.models..."
for sample in orm["main.Sample"].objects.filter(goldsample__isnull=True):
worker = Sample.get_worker(source_type=sample.source_type, source_val=sample.source_val)
if not worker:
continue
try:
orm["crowdsourcing.WorkerQualityVote"].objects.new_vote(worker=worker, sample=sample, label=LABEL_YES)
except:
# Such vote already exists - skip.
pass
def forwards(self, orm):
"Write your forwards methods here."
# Note: Remember to use orm['appname.ModelName'] rather than "from appname.models..."
for sample in orm['main.Sample'].objects.filter(goldsample__isnull=True):
worker = Sample.get_worker(
source_type=sample.source_type,
source_val=sample.source_val,
)
if not worker:
continue
try:
orm['crowdsourcing.WorkerQualityVote'].objects.new_vote(
worker=worker,
sample=sample,
label=LABEL_YES,
)
except:
# Such vote already exists - skip.
pass
def get_source_worker(self):
"""
Returns a worker who has sent this sample.
"""
return Sample.get_worker(source_type=self.source_type, source_val=self.source_val)