Please provide description for each of the URL within the tornado-celery/handlers.py example.
@mher I don't entirely understand how tornado-celery consumes tasks.
I was reading this blog post to try to understand the various methods of integrating celery and tornado.
It seems to me that tornado-celery provides an HTTP API for interacting with celery tasks but there's no easy way to notify the main tornado app when a task completes with polling -- is this correct?
As I'm reading further, I see that tornado-celery uses pika for a non-blocking AMQP client and supports using send_task or apply_async with a non-blocking producer.
Is there any way to set up a non-blocking consumer? I'd like to be able to utilize my tornado eventloop (or perhaps another tornado eventloop in a separate process) to perform non-blocking tasks that simply scrape web pages. What's the easiest way to make this happen?
Okay so I see that tcelery maintains its own ioloop, https://github.com/mher/tornado-celery/blob/master/tcelery/__main__.py#L41.
In my tasks.py module, can I create an @celery.task that uses @gen.coroutine
http_client = AsyncHTTPClient()
response = yield http_client.fetch("http://example.com")
I've setup documentation for tornado-celery at http://tornado-celery.readthedocs.org. Feel free to send pull requests.