def conc_register(self, user_id, corpus_id, subc_name, subchash, query,
samplesize, time_limit):
"""
Register concordance calculation and initiate the calculation.
arguments:
user_id -- an identifier of the user who entered the query (used to specify subc. directory if needed)
corpus_id -- a corpus identifier
subc_name -- a sub-corpus identifier (None if not used)
subchash -- a MD5 checksum of the sub-corpus data file
query -- a query tuple
samplesize -- a row number limit (if 0 then unlimited - see Manatee API)
time_limit -- a time limit (in seconds) for the main conc. task
returns:
a dict(cachefile=..., pidfile=..., stored_pidfile=...)
"""
reg_fn = concworker.TaskRegistration(task_id=self.request.id)
subc_path = os.path.join(settings.get('corpora', 'users_subcpath'),
str(user_id))
pub_path = os.path.join(settings.get('corpora', 'users_subcpath'),
'published')
initial_args = reg_fn(corpus_id, subc_name, subchash,
(subc_path, pub_path), query, samplesize)
if not initial_args['already_running']: # we are first trying to calc this
app.send_task('worker.conc_calculate',
args=(initial_args, user_id, corpus_id, subc_name,
subchash, query, samplesize),
soft_time_limit=time_limit)
return initial_args
def create_task(user_id, corp, subchash, q, samplesize):
task_id = str(uuid.uuid1())
reg_fn = concworker.TaskRegistration(task_id=task_id)
corpus_id = corp.corpname
subcname = getattr(corp, 'subcname', None)
initial_args = reg_fn(corpus_id, subcname, subchash, q, samplesize)
if not initial_args['already_running']: # we are first trying to calc this
def run():
with plugins.runtime.CONC_CACHE as cc:
task = concworker.ConcCalculation(task_id=task_id, cache_factory=cc.fork())
subc_path = '%s/%s' % (settings.get('corpora', 'users_subcpath'), user_id)
return task(initial_args, subc_path, corpus_id, subcname, subchash, q, samplesize)
proc = Process(target=run)
else:
proc = EmptyTask()
return proc
def conc_register(self, user_id, corpus_id, subc_name, subchash, query, samplesize):
"""
Register concordance calculation and initiate the calculation.
arguments:
user_id -- an identifier of the user who entered the query (used to specify subc. directory if needed)
corpus_id -- a corpus identifier
subc_name -- a sub-corpus identifier (None if not used)
subchash -- a MD5 checksum of the sub-corpus data file
query -- a query tuple
samplesize -- a row number limit (if 0 then unlimited - see Manatee API)
returns:
a dict(cachefile=..., pidfile=..., stored_pidfile=...)
"""
reg_fn = concworker.TaskRegistration(task_id=self.request.id)
subc_path = '%s/%s' % (settings.get('corpora', 'users_subcpath'), user_id)
initial_args = reg_fn(corpus_id, subc_name, subchash, subc_path, query, samplesize)
if not initial_args['already_running']: # we are first trying to calc this
conc_calculate.delay(initial_args, user_id, corpus_id,
subc_name, subchash, query, samplesize)
return initial_args