Welcome to EasyRabbit’s documentation!¶

Actually easy RabbitMQ utilities for common tasks. Hides the complexities of complete control packages like pika by wrapping common use cases in minimalistic wrappers.

Installation¶

You can install directly from Pypi:

pip install easyrabbit

or from git using pip:

pip install git+https://github.com/scnerd/easyrabbit

Usage¶

Simple routing¶

Reader¶

RoutingReader provides a fully asynchronous way to read from an exchange via a queue bound with a routing key. It launches a subprocess to free the calling program from being responsible for its computational overhead; within its own process, it uses pika’s asynchronous connection, enabling the highest possible performance client. The API exposed mimics a simple queue, if, after all, what you really want is to use a RabbitMQ queue as if it were a Python queue.

with RoutingReader(url, exchange, queue_name, routing_key) as reader:
    for msg in reader:
        print("Received the following message: {}".format(msg))

If the reader is needed persistently, you can also launch and terminate it yourself:

reader = RoutingReader(url, exchange, queue_name, routing_key)
reader.start()
# Do things with the reader
reader.close()

Note that even though the reader is asynchronous, you don’t need to wait for it to be ready before using it. All calls hang on the process pipe that sends data from the client process to your parent code. If you want to make sure that the connection is fully established before using it, however, you can use wait_till_ready:

reader = RoutingReader(url, exchange, queue_name, routing_key)
reader.start()
try:
   reader.wait_till_ready(timeout=5)
except TimeoutError:
   raise RuntimeError("RabbitMQ reader took more than 5 seconds to launch")

While reader.get is blocking, a non-blocking equivalent is reader.get_nowait, or your code can explicitly check that a value is available first using not reader.empty(). Iterating over reader just repeatedly calls get, and hence is blocking and will only end when the reader is closed. Use reader.getall_nowait() to obtain all values currently in the queue.

All result objects are byte arrays.

Writing¶

The analogous utility for writing is RoutingWriter, which provides a nearly identical interface for writing messages to RabbitMQ

with RoutingWriter(url, exchange, routing_key) as writer:
    for msg in msgs:
        writer.put(msg)

RoutingWriter exposes much the same API as RoutingReader, except of course exposing put instead of get.