It's possible to make it work in debug mode too

[salomonbrys]

It's possible to make it work in debug mode too

[SalomonBrys]

ANR-Watchdog does not makes any check at startup. If you call start() on it, it will start.

That comment refers to the proposed usage in README:

if (!BuildConfig.DEBUG) {
    new ANRWatchDog().start();
}

If you still want the watchdog enabled in debug when the debugger is not connected, simply changes this for:

if (!Debug,isDebuggerConnected()) {
    new ANRWatchDog().start();
}

Be aware, however, that should you attach a debugger after startup, the watchdog will interpret debug pauses as ANRs.
To prevent this, you could set an ANRListener:

new ANRWatchDog().setANRListener(new ANRWatchDog.ANRListener() {
    @Override
    public void onAppNotResponding(ANRError error) {
        if (!Debug,isDebuggerConnected())
            throw error;
    }
}).start();

[TWiStErRob]

Yep, I understand all that. I'm proposing to use the last code block (if-throw) in the core of the library so it's handled automatically for everyone by default without the need for configuration.

It would also simplify basic library usage to one code line and no caveats.

[SalomonBrys]

That's a good idea ;)

There would have to be an option to revert to previous behaviour.

I'll work on that ;)

[HolenZhou]

new ANRWatchDog().setANRListener(new ANRWatchDog.ANRListener() {
@OverRide
public void onAppNotResponding(ANRError error) {
if (!Debug,isDebuggerConnected())
throw error;
}
}).start();

I think this code should be the basic use, it meets all requirements.

[SalomonBrys]

Done in 1.2.0.
Thanks for the report :)