Flask-Apscheduler及其应用上下文

Question

Brief: I am trying to use flask-apscheduler to regularly create and upload csv backups of certain data from a MySQL database using my models (and sqlalchemy as my ORM) to S3. In the past, I wrote a seperate script and ran through a cron job. However, I wanted to experiment with flask-apscheduler and now I need to figure out why it won't work!

Stats:

Flask==0.10.1
Flask-APScheduler==1.6.0
Flask-SQLAlchemy==2.0
</pre>

Structure of App:
<pre>
/app
  /main
  /auth
  __init__.py
  models.py
  utilities.py

config.py

init .py

from flask import Flask
from flask.ext.sqlalchemy import SQLAlchemy
from config import config
from flask_apscheduler import APScheduler
import logging

login_manager                       = LoginManager()
login_manager.session_protection    = 'strong'
login_manager.login_view            = 'auth.login'

moment      = Moment()
db          = SQLAlchemy()
sched       = APScheduler()

def create_app(config_name):
    app = Flask(__name__)
    app.config.from_object(config[config_name])

    config[config_name].init_app(app)
    db.init_app(app)
    login_manager.init_app(app)
    sched.init_app(app)

    sched.start()

    # Set up logging
    if not app.debug:
        app.logger.addHandler(logging.StreamHandler())
        app.logger.setLevel(logging.DEBUG)

        logger = logging.getLogger('flask_logger')
    else:
        logging.basicConfig()

    from .main import main as main_blueprint
    from .auth import auth as auth_blueprint

    app.register_blueprint(main_blueprint)
    app.register_blueprint(auth_blueprint, url_prefix = '/auth')

    return app

if __name__ == "__main__":
    app = create_app('development')

config.py

...
JOBS = [
        {
            'id': 'job1',
            'func': 'app.utilities:backup_data_to_s3',
            'kwargs': {'bucket': AWS_BUCKET},
            'trigger': {'type': 'cron', 'hour': '0'}
        }
    ]
...

utilities.py

from datetime import date, datetime, timedelta
import sendgrid
from flask import current_app, g
import boto3
from app.models import *

def backup_data_to_s3(**kwargs):
    """
    length_of_time: in number of days
    start: datetime.date object
    end: datetime.date object
    """
    t0 = date.today()

    bucket_name = kwargs.pop('bucket')
    start       = kwargs.pop('start', t0 - timedelta(days = 1))
    end         = kwargs.pop('end', t0)

    # Get the number of days
    dt      = (end - start).days

    s3          = boto3.resource('s3')
    bucket      = s3.Bucket(bucket_name)
    updated     = 0

    ...
    return

The traceback for the primary error that I receive is:

Traceback (most recent call last):
  File "/Users/dh/Documents/GitHub/web-tata/web-tata/flask/lib/python2.7/site-packages/apscheduler/executors/base.py", line 125, in run_job
    retval = job.func(*job.args, **job.kwargs)
  File "/Users/dh/Documents/GitHub/web-tata/web-tata/app/utilities.py", line 65, in backup_data_to_s3
    print (g)
  File "/Users/dh/Documents/GitHub/web-tata/web-tata/flask/lib/python2.7/site-packages/werkzeug/local.py", line 357, in <lambda>
    __str__ = lambda x: str(x._get_current_object())

  File "/Users/dh/Documents/GitHub/web-tata/web-tata/flask/lib/python2.7/site-packages/werkzeug/local.py", line 297, in _get_current_object
    return self.__local()
  File "/Users/dh/Documents/GitHub/web-tata/web-tata/flask/lib/python2.7/site-packages/flask/globals.py", line 27, in _lookup_app_object
    raise RuntimeError('working outside of application context')
RuntimeError: working outside of application context

It's pretty obvious that there is just something going on with how flask handles the application context, but I'm not quite aware of exactly how it works...Any help is greatly appreciated.

Answer 1

Traceback says line 65, in backup_data_to_s3 print (g) it means you are using g global object while your app context doesn't push. utilities.py could ask for the current app while app doesn't created yet, it means actually you can't use current_app unless you create one. So in your utilities.py you can do something like this:

from . import create_app

app = create_app("config_name")
app_ctx = app.app_context()
app_ctx.push()

#now you can use g, current_app
print(g)

Note

Even if you run your app with init.py at the same time you would not be able to get current_app in your utilities.py.

When flask get a request it activate application context and after request delete app context.