Nowadays, I’m dealing with a NLP project, there is a step which is about python regular expression and I have a problem.
When program meets texts with some formats that my regular expression cannot match well, it will take really long time to finish or even get endless loop. So I want to set a run time upper limit onto each loop. Python decorator could solve this well.
First method:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
import signal
def set_timeout(num, callback):
def wrap(func):
def handle(signum, frame):
"""
callback function,
:signum: number of signal，
:frame: the interrupted stack frame.
"""
raise RuntimeError
def to_do(*args, **kwargs):
try:
signal.signal(signal.SIGALRM, handle) # set signal and callback func
signal.alarm(num) # set alarm for num second
print('start alarm signal.')
r = func(*args, **kwargs)
print('close alarm signal.')
signal.alarm(0) # close alarm
return r
except RuntimeError as e:
callback()
return to_do
return wrap
@set_timeout(num_sec, callback_func)
def target_func():
pass
Code’s from：https://blog.csdn.net/DylanYuan/article/details/83583981
Second method:
The second way is not as powerful as the first one, but it’s much simpler.
It’s needed to use pip to add package timeout_decorator to out python environment first.
1
2
3
4
5
from timeout_decorator import timeout
@timeout(num_sec)
def target_func():
pass
