New deasrc

This commit is contained in:
2024-06-13 09:25:36 +02:00
parent 430b4de995
commit 14df489b38
4 changed files with 262 additions and 21 deletions
+39 -11
View File
@@ -21,7 +21,7 @@ Global Variables:
# TODO: Wpiemdolić to wszystko w klasę z loggerem przysłanym z góry
from queue import Empty, Queue
from threading import Thread
import time
q = Queue()
# DATA FOR TEST ONLY
@@ -49,9 +49,17 @@ def producer(out_q, control_q, filename, _logger):
_logger: Logger object for logging.
"""
with open(DATABASE_PATH + filename, "r+", encoding=ENCODING) as operated_file:
_logger.info(f"Worker {filename} ")
print(f"Worker {filename} ")
line_no = 0
while True:
line = operated_file.readline()
line_no += 1
print(f"\t \t \t \t \t \t W{filename}{line_no}\r", end="")
if not line:
print(f"EOF {filename}")
break
if not line:
break
out_q.put(line)
@@ -59,11 +67,12 @@ def producer(out_q, control_q, filename, _logger):
check = control_q.get(block=False)
except Empty:
check = False
if check is _sentinel:
_logger.info("TERM signal received")
print("TERM signal received")
control_q.put(check)
break
_logger.info(f"Worker finished: {filename}")
print(f"Worker finished: {filename}")
out_q.put(_sentinel)
@@ -79,19 +88,25 @@ def consumer(in_q, control_q, doi, live_results, result_list, control_dict, no,
live_results (list): List to store the search results.
_logger: Logger object for logging.
"""
_logger.info("Consumer thread started:%s", no)
for item in doi:
result_list.append({"DOI": item[0], "exists": False, "data": item[1]})
print(f"Consumer thread started: {no} no")
empty_counter = 0
alive_no = 0
while True:
done_check = True
try:
data = in_q.get(block=False)
data = in_q.get(block=True, timeout = 1)
if data is _sentinel:
_logger.info("Worker finished")
control_dict["sentinels"] += 1
print(f"Workers finished: {control_dict['sentinels']} reported by consumer {no}")
else:
empty_counter = 0
alive_no += 1
print(f"C{no}__{alive_no}\r", end="")
for item in result_list:
if item["DOI"] in data and not item["exists"]:
print(f"HIT in {no} content {data[0]} line {data[1]} file {data[2]} {item["exists"]}")
_logger.info(data)
_logger.info("HIT")
item["exists"] = True
@@ -100,9 +115,18 @@ def consumer(in_q, control_q, doi, live_results, result_list, control_dict, no,
if done_check:
control_q.put(_sentinel)
except Empty:
pass
empty_counter += 1
time.sleep(1)
print(f"Consumer {no} empty")
if empty_counter > 5:
print("Consumer %s empty lvl 2", no)
time.sleep(2)
elif empty_counter > 10:
print(f"Consumer thread finished {no}")
break
if control_dict["sentinels"] >= MAXTHREADS:
_logger.info("All workers finished")
_logger.info(f"All workers finished {no}")
break
@@ -121,6 +145,10 @@ def search_for_doi(doi, live_results, _logger):
threads = []
work_q = Queue(maxsize=WORK_Q_SIZE)
control_q = Queue()
for item in doi:
result_list.append({"DOI": item[0], "exists": False, "data": item[1]})
for i in range (0, (len(doi)//1000)+2):
t_cons = Thread(
target=consumer, args=(work_q, control_q, doi, live_results, result_list, control_dict, i, _logger)