Hi,
I wanted to close the loop here. I don’t think this is fixed yet on the kotlin scanner.
We eventually ended up implementing a diff based sonar approach inspired from the answer at Slow kotlin sensor - #5 by le.flow and we’re seeing good improvements at the 90th percentile for running sonar (1h10m → 35m).
Hope we get a fix from the kotlin analyzer on this. But for now we’re going ahead with this approach.
Please do let me know when this is finally fixed on the kotlin sensor.