Master Elasticsearch with Django: From Basics to Real‑World Integration

What is Elasticsearch?

Elasticsearch is a search engine built on the Lucene library. It provides a distributed, multi‑tenant full‑text search engine with an HTTP web interface and schema‑free JSON documents, and it is developed in Java.

What are the uses of Elasticsearch?

Elasticsearch enables fast, near‑real‑time storage, search, and analysis of large volumes of data, returning results in milliseconds by searching indexes directly rather than raw text.

Basic concepts of Elasticsearch

Index – a collection of documents and their fields, e.g., data from a social‑network app.

Type/Mapping – a set of common fields shared by documents within the same index, such as separate types for user profiles, messages, and comments.

Document – a JSON‑formatted set of fields; each document belongs to a type and resides in an index, identified by a unique UID.

Field – an Elasticsearch field can hold multiple values of the same type (essentially a list), unlike a SQL column which holds a single value of its type.

Using Elasticsearch in Django

Install and configure the Django Elasticsearch DSL package: $ pip install django-elasticsearch-dsl Add django_elasticsearch_dsl to INSTALLED_APPS and define ELASTICSEARCH_DSL in Django settings, for example:

ELASTICSEARCH_DSL = {
    'default': {
        'hosts': 'localhost:9200'
    },
}

Declare the data to be indexed and create a model:

# models.py

class Category(models.Model):
    name = models.CharField(max_length=30)
    desc = models.CharField(max_length=100, blank=True)

    def __str__(self):
        return '%s' % (self.name)

To use this model with Elasticsearch, create a subclass of django_elasticsearch_dsl.Document, define an Index class for the index name and settings, and register the document with the registry.register_document decorator in documents.py:

from django_elasticsearch_dsl import Document
from django_elasticsearch_dsl.registries import registry
from .models import Category

@registry.register_document
class CategoryDocument(Document):
    class Index:
        name = 'category'
    settings = {
        'number_of_shards': 1,
        'number_of_replicas': 0
    }
    class Django:
        model = Category
        fields = [
            'name',
            'desc',
        ]

Create and populate the Elasticsearch index and mapping: $ python manage.py search_index --rebuild Search examples:

# Term filter
s = CategoryDocument.search().filter("term", name="computer")

# Match query
s = CategoryDocument.search().query("match", description="abc")

for hit in s:
    print("Category name : {}, description {}".format(hit.name, hit.desc))

Convert Elasticsearch results to a real Django queryset (incurring one additional SQL query to fetch model instances):

s = CategoryDocument.search().filter("term", name="computer")[:30]
qs = s.to_queryset()
for cat in qs:
    print(cat.name)

Feel free to leave comments if you have any questions.