django-el 0.2.2

Quickstart

Configure your models to be indexable:

from django.db import models
from el.models import Indexed

class Article(models.Model, Indexed):
    title = models.CharField(max_length=78)

@classmethod
def get_indexable(cls):
    return cls.objects.all()

@classmethod
def configure_mapping(cls, mapping):
    # mapping is an elasticsearch_dsl Mapping object
    mapping.field('title', 'string')
    return mapping

From this moment, the Article model will be autodiscovered and indexed.

Update search indexes:

./manage.py update_index

Use elasticsearch_dsl to query:

# articles is a list of an Article instances
articles = Article.search().query('match', title="Bob's article").execute()

# articles is a list of elasticsearch_dsl hits
articles = Article.search().query('match', title="Bob's article").execute(cast=False)

In contrast with elasticsearch_dsl, django-el provides modified Search object which return django model instances instead of raw elasticsearch results by default. You can control this feature using the cast argument.

Installation

Install django-el as usual python package using pip:

pip install django-el

Configuration

Django-el is build on top of elasticsearch_dsl library and provides django-way connections configuration through settings.py:

ELASTICSEARCH_CONNECTIONS = {
    'default': {
        'hosts': ['127.0.0.1:9200'],
        'serializer': 'project.serializers.MySerializer',
    }
}

You can define project connections using ELASTICSEARCH_CONNECTIONS setting. It is just a hight-level interface over low-level elasticsearch_dsl.connections.connections.create_connection function.

The keys are (default, in this example) are connection aliases, and it’s values are create_connection arguments.