Using the jedi autocompletion library, I'm finding that each call to Script.completions() makes successive calls slower.
In the following code, I repeat my run function three times. It's not clear to me why, but each one takes longer than the previous.
import jedi
import time
def measure(source):
start = time.time()
script = jedi.Script(source, line=1, column=0)
script.completions()
print(' %-20s%f' % (source, time.time() - start))
def run():
start = time.time()
measure('min(1,2)')
measure('max(1,2)')
measure('print("Hello")')
measure('abs(1)')
measure('set()')
measure('dict()')
print('Total: %f' % (time.time() - start))
run()
run()
run()
Results:
min(1,2) 0.016168
max(1,2) 0.014470
print("Hello") 0.016843
abs(1) 0.019889
set() 0.023725
dict() 0.025874
Total: 0.117067
min(1,2) 0.029772
max(1,2) 0.034207
print("Hello") 0.034982
abs(1) 0.038538
set() 0.041346
dict() 0.054610
Total: 0.233565
min(1,2) 0.047249
max(1,2) 0.050380
print("Hello") 0.053113
abs(1) 0.056774
set() 0.059072
dict() 0.062129
Total: 0.328825
Thank you for any advice on why might this be happening and what I can do to prevent it.
