How to use logger in flask app for outside app contexts (flask-apscheduler)

[Shivappa]

Facing a context issue for the scheduled jobs.
Issue is explained here.
https://stackoverflow.com/questions/68320718/how-to-use-logger-in-flask-app-for-outside-app-contexts-flask-apscheduler

[viniciuschiele]

You need to create a new Flask context within your task like this: https://github.com/viniciuschiele/flask-apscheduler/blob/master/examples/flask_context.py#L22

You can also the the Flask app from the APScheduler object:

with apscheduler.app.app_context();
    pass

[mxdev88]

I ran into the same issue and fixed it by adding what you suggest in the func. However I was just wondering whether this could be improved within the library to make it automatic? I was thinking decorating func within APScheduler.add_job should work?

e.g. l.166 https://github.com/viniciuschiele/flask-apscheduler/blob/master/flask_apscheduler/scheduler.py, when self.app is known at time of adding a job, could something like below be done?

...
        fix_job_def(job_def)
        job = self._scheduler.add_job(**job_def)
        if self.app:
            job.func = with_app_context(job.func)
        return job
...

This should ensure the func is always run with an app context. Any thoughts on this?

[viniciuschiele]

This doesn't work because jobs can be stored in database and when you shutdown your app and start it again, the jobs would not come attached to with_app_context.
I couldn't find a nice solution for this problem yet.

[mxdev88]

@viniciuschiele I see your point. How about decorating BaseExecutor.submit_job instead then? After initializing the scheduler, could loop through the scheduler executors to make sure they will submit job with the app context.
Do you think that would work then?