Lightweight scheduler and worker for Django using a database as backend.
The task worker and scheduler use the database as a queue to store and find tasks to process
and you can forget about using transaction.on_commit.
If you want a billion messages per second, get out of here.
from django.http import HttpRequest, HttpResponse
from django.db import transaction
from barn.decorators import task
@transaction.atomic
def index(request: HttpRequest) -> HttpResponse:
html = "<html><body>Hello</body></html>"
# do something with database
dummy.delay(view="index", code="12")
return HttpResponse(html)
@task
def dummy(view: str, code: str) -> str:
# task called inside transaction
return "ok"You can specify your own model for schedule and task and this approach is preferred.
from django.db import models
from django.core.mail import send_mail
from barn.models import AbstractTask
class EndOfTrialPeriodTask(AbstractTask):
user = models.ForeignKey('auth.User', models.CASCADE)
def process(self) -> None:
# go somewhere and do something...
self.user.is_active = False
self.user.save()
NotifyEndOfTrialPeriodTask.objects.create(user_id=self.user_id)
class NotifyEndOfTrialPeriodTask(AbstractTask):
user = models.ForeignKey('auth.User', models.CASCADE)
def process(self) -> None:
if self.user.email:
send_mail(
"End of trial period",
f"Hello {self.user.get_full_name()}. Your trial period has ended.",
"[email protected]",
[self.user.email],
fail_silently=False,
)
class EndOfSubscriptionTask(AbstractTask):
user = models.ForeignKey('auth.User', models.CASCADE)
def process(self) -> None:
# go somewhere and do something...