Add Flask context automatically to APScheduler executors

[mxdev88]

This PR adds the Flask context to added or modified jobs automaticall. Most of the time when using Flask-APScheduler, you need to access something from your Flask app which requires to add the context yourself.

This should solve issues like mentioned here #176

[viniciuschiele]

I guess this code only works when the jobs are kept in memory, if you use a storage like sqlite/sqlalchemy it won't work.

[mxdev88]

Another option may be to act on the part that runs the job rather than on the job itself. Any thoughts on this?

[viniciuschiele]

It is APScheduler that runs the jobs and it doesn't allow me to add some sort of "middleware" to initialize a Flask context before calling the actual method. If you find a workaround for that, I will be happy to merge it.

[mxdev88]

Hi @viniciuschiele - made another attempt, this time by decorating executors rather than the jobs. Let me know what you think. thanks!

[savchenko]

This does seem to work as intended. Tested as following:

SQLAlchemy jobstore

    SCHEDULER_JOBSTORES = {
        "default": SQLAlchemyJobStore(
            url=DatabaseConfig.SQLALCHEMY_DATABASE_URI
        )
    }

Decorated function

@scheduler.task("foo", id="foo", seconds=30)
def foo():
    do_something()

After starting Flask, the DB is populated with the corresponding data:

SELECT * FROM apscheduler_jobs
-- "id"	        "next_run_time" 	"job_state"
-- "foo"	1721815605.395674	"binary data"

do_something() is executed every 30 seconds with corresponding logging records created.

@viniciuschiele do you reckon it can be merged prior to the upcoming APScheduler v4.0 release?