carrot 0.5.0

Introduction

carrot is an AMQP messaging queue framework. AMQP is the Advanced Message Queuing Protocol, an open standard protocol for message orientation, queuing, routing, reliability and security.

The aim of carrot is to make messaging in Python as easy as possible by providing a high-level interface for producing and consuming messages. At the same time it is a goal to re-use what is already available as much as possible.

carrot has pluggable messaging back-ends, so it is possible to support several messaging systems. At the time of release, the py-amqplib based backend is considered suitable for production use.

Several AMQP message broker implementations exists, including RabbitMQ, ZeroMQ and Apache ActiveMQ. You’ll need to have one of these installed, personally we’ve been using RabbitMQ.

Before you start playing with carrot, you should probably read up on AMQP, and you could start with the excellent article about using RabbitMQ under Python, Rabbits and warrens. For more detailed information, you can refer to the Wikipedia article about AMQP.

Documentation

Carrot is using Sphinx, and the latest documentation is available at GitHub:

http://github.com/ask/carrot/

$ pip install carrot

$ easy_install carrot

$ python setup.py build
# python setup.py install # as root

Creating a connection

You can set up a connection by creating an instance of carrot.messaging.AMQPConnection, with the appropriate options for your AMQP server:

>>> from carrot.connection import AMQPConnection
>>> amqpconn = AMQPConnection(hostname="localhost", port=5672,
...                           userid="test", password="test",
...                           vhost="test")

If you’re using Django you can use the carrot.connection.DjangoAMQPConnection class instead, which loads the connection settings from your settings.py:

AMQP_SERVER = "localhost"
AMQP_PORT = 5672
AMQP_USER = "test"
AMQP_PASSWORD = "secret"
AMQP_VHOST = "/test"

Then create a connection by doing:

>>> from carrot.connection import DjangoAMQPConnection
>>> amqpconn = DjangoAMQPConnection()

Receiving messages using a Consumer

First we open up a Python shell and start a message consumer.

This consumer declares a queue named "feed", receiving messages with the routing key "importer" from the "feed" exchange.

The example then uses the consumers wait() method to go into consume mode, where it continuously polls the queue for new messages, and when a message is received it passes the message to all registered callbacks.

>>> from carrot.messaging import Consumer
>>> consumer = Consumer(connection=amqpconn, queue="feed",
...                     exchange="feed", routing_key="importer")
>>> def import_feed_callback(message_data, message)
...     feed_url = message_data["import_feed"]
...     print("Got feed import message for: %s" % feed_url)
...     # something importing this feed url
...     # import_feed(feed_url)
...     message.ack()
>>> consumer.register_callback(import_feed_callback)
>>> consumer.wait() # Go into the consumer loop.

Sending messages using a Publisher

Then we open up another Python shell to send some messages to the consumer defined in the last section.

>>> from carrot.messaging import Publisher
>>> publisher = Publisher(connection=amqpconn,
...                       exchange="feed", routing_key="importer")
>>> publisher.send({"import_feed": "http://cnn.com/rss/edition.rss"})
>>> publisher.close()

Look in the first Python shell again (where consumer.wait() is running), where the following text has been printed to the screen:

Got feed import message for: http://cnn.com/rss/edition.rss

Serialization of Data

By default every message is encoded using JSON, so sending Python data structures like dictionaries and lists works. YAML and Python’s built-in pickle module is also supported, and if needed you can register any custom serialization scheme you want to use.

Each option has its advantages and disadvantages.

json – JSON is supported in many programming languages, is now

a standard part of Python (since 2.6), and is fairly fast to decode using the modern Python libraries such as cjson or ``simplejson.

The primary disadvantage to JSON is that it limits you to the following data types: strings, unicode, floats, boolean, dictionaries, and lists. Decimals and dates are notably missing.

Also, binary data will be transferred using base64 encoding, which will cause the transferred data to be around 34% larger than an encoding which supports native binary types.

However, if your data fits inside the above constraints and you need cross-language support, the default setting of JSON is probably your best choice.

pickle – If you have no desire to support any language other than

Python, then using the pickle encoding will gain you the support of all built-in Python data types (except class instances), smaller messages when sending binary files, and a slight speedup over JSON processing.

yaml – YAML has many of the same characteristics as json,

except that it natively supports more data types (including dates, recursive references, etc.)

However, the Python libraries for YAML are a good bit slower than the libraries for JSON.

If you need a more expressive set of data types and need to maintain cross-language compatability, then YAML may be a better fit than the above.

To instruct carrot to use an alternate serialization method, use one of the following options.

1. Set the serialization option on a per-Publisher basis:

   >>> from carrot.messaging import Publisher
   >>> publisher = Publisher(connection=amqpconn,
   ...                       exchange="feed", routing_key="importer",
   ...                       serializer="yaml")
   

2. Set the serialization option on a per-call basis

   >>> from carrot.messaging import Publisher
   >>> publisher = Publisher(connection=amqpconn,
   ...                       exchange="feed", routing_key="importer")
   >>> publisher.send({"import_feed": "http://cnn.com/rss/edition.rss"},
   ...                serializer="pickle")
   >>> publisher.close()
   

Note that Consumer``s do not need the serialization method specified in their code. They can auto-detect the serialization method since we supply the ``Content-type header as part of the AMQP message.

Sending raw data without Serialization

In some cases, you don’t need your message data to be serialized. If you pass in a plain string or unicode object as your message, then carrot will not waste cycles serializing/deserializing the data.

You can optionally specify a content_type and content_encoding for the raw data:

>>> from carrot.messaging import Publisher
>>> publisher = Publisher(connection=amqpconn,
...                       exchange="feed",
                          routing_key="import_pictures")
>>> publisher.send(open('~/my_picture.jpg','rb').read(),
                   content_type="image/jpeg",
                   content_encoding="binary")
>>> publisher.close()

The message object returned by the Consumer class will have a content_type and content_encoding attribute.

Receiving messages without a callback

You can also poll the queue manually, by using the fetch method. This method returns a Message object, from where you can get the message body, de-serialize the body to get the data, acknowledge, reject or re-queue the message.

>>> consumer = Consumer(connection=amqpconn, queue="feed",
...                     exchange="feed", routing_key="importer")
>>> message = consumer.fetch()
>>> if message:
...    message_data = message.payload
...    message.ack()
... else:
...     # No messages waiting on the queue.
>>> consumer.close()

Sub-classing the messaging classes

The Consumer, and Publisher classes can also be subclassed. Thus you can define the above publisher and consumer like so:

>>> from carrot.messaging import Publisher, Consumer

>>> class FeedPublisher(Publisher):
...     exchange = "feed"
...     routing_key = "importer"
...
...     def import_feed(self, feed_url):
...         return self.send({"action": "import_feed",
...                           "feed_url": feed_url})

>>> class FeedConsumer(Consumer):
...     queue = "feed"
...     exchange = "feed"
...     routing_key = "importer"
...
...     def receive(self, message_data, message):
...         action = message_data["action"]
...         if action == "import_feed":
...             # something importing this feed
...             # import_feed(message_data["feed_url"])
                message.ack()
...         else:
...             raise Exception("Unknown action: %s" % action)

>>> publisher = FeedPublisher(connection=amqpconn)
>>> publisher.import_feed("http://cnn.com/rss/edition.rss")
>>> publisher.close()

>>> consumer = FeedConsumer(connection=amqpconn)
>>> consumer.wait() # Go into the consumer loop.

Mailing list

Join the carrot-users mailing list.