Skip to content


Subversion checkout URL

You can clone with
Download ZIP
An extension to micromodels to fetch/store/delete models in Riak
Branch: master


Micromodels Riak

This is a really simple micromodels loader for Riak


Here's an example use:

from micromodelsext.riak import Loader
import riak
import micromodels

class MyModel(micromodels.Model):
    title = micromodels.CharField()

client = riak.RiakClient()
bucket = client.bucket("mymodel")

# create a loader for the bucket and the model
loader = Loader(bucket, MyModel)

m = MyModel.from_dict({"title": "This is a test"})

# Store model
loader['test'] = m

# Fetch model
m = loader['test']

# Delete model
del loader['test']

Sometimes you want to put scoping information in the bucket's name, but the model will be the same. For instance:

glenn_entries = Loader(client.bucket("glenn_entries"), MyModel)
eric_entries = Loader(client.bucket("eric_entries"), MyModel)

glenn_story = MyModel.from_dict({"title": "I love riding my motorcycle"})
eric_story = MyModel.from_dict({"title": "I love riding my bicycle"})

glenn_entries['i-love-my-bike'] = glenn_story
eric_entries['i-love-my-bike'] = eric_story

The same model is used, but different buckets.

Handling Lists

Riak is a k/v store and this maps nicely to Python dictionaries, however sometimes you have a list of RiakObjects. Here's an example from a search result using the Riak Search Solr interface:

bucket = client.bucket("eric_entries")
loader = Loader(bucket, MyModel)

result ="title:bicycle", rows=20)
obj_gen = (bucket.get(r['id']) for r in result)
model_list = [loader.cast(obj) for obj in obj_gen]

This actually could be shorten by doing the following:

bucket = client.bucket("eric_entries")
loader = Loader(bucket, MyModel)

result ="title:bicycle", rows=20)
model_list = [loader[r['id']] for r in result]
Something went wrong with that request. Please try again.