Skip to main content

Sphinxsearch database backend for django>=2.0

Project description


SphinxSearch database backend for Django.

Build Status codecov PyPI version

Installation and usage

  1. Install django-sphinxsearch package

    pip install django_sphinxsearch
  2. Configure Django settings

        'ENGINE': 'sphinxsearch.backend.sphinx',
        'HOST': '',
        'PORT': 9306,
        'OPTIONS': {
            'use_unicode': False # if non-ascii letters looks bad in results
    DATABASE_ROUTERS = ['sphinxsearch.routers.SphinxRouter']
  3. Create index definitions in sphinx.conf

    index testapp_testmodel
        type              = rt
        path              = /data/sphinx/testapp/testmodel/
        rt_field          = sphinx_field
        rt_attr_uint      = attr_uint
        rt_attr_bool      = attr_bool
        rt_attr_bigint    = attr_bigint
        rt_attr_float     = attr_float
        rt_attr_multi     = attr_multi
        rt_attr_multi_64  = attr_multi_64
        rt_attr_timestamp = attr_timestamp
        rt_attr_string    = attr_string
        rt_attr_json      = attr_json
  4. Define Django model for index

    from datetime import datetime
    from django.db import models
    from jsonfield.fields import JSONField
    from sphinxsearch import models as spx_models
    class FieldMixin(spx_models.SphinxModel):
        # Note that NULL values are not allowed for sphinx rt-index.
        # Indexed text field. If no attribute with same name defined, can't be
        # retrieved from index.
        class Meta:
            abstract = True
        # Indexed text field. If no attribute with same name defined, can't be
        # retrieved from index.
        sphinx_field = spx_models.SphinxField(default='')
        other_field = spx_models.SphinxField(default='')
        # Numeric attributes
        attr_uint = spx_models.SphinxIntegerField(default=0, db_column='attr_uint_')
        attr_bigint = spx_models.SphinxBigIntegerField(default=0)
        attr_float = models.FloatField(default=0.0)
        attr_timestamp = spx_models.SphinxDateTimeField(
        attr_bool = models.BooleanField(default=False)
        # String attributes
        attr_string = models.CharField(max_length=32, default='')
        attr_json = JSONField(default={})
        # Multi-value fields (sets of integer values)
        attr_multi = spx_models.SphinxMultiField(default=[])
        attr_multi_64 = spx_models.SphinxMulti64Field(default=[])
    class TestModel(FieldMixin, spx_models.SphinxModel):
  5. Query index from your app

    # Numeric attributes filtering
    TestModel.objects.filter(attr_uint=0, attr_float__gte=10, attr_multi__in=[1, 2])
    # For sphinxsearch>=2.2.7, string attr filtering enabled
    TestModel.objects.filter(attr_string='some test')
    # Use mysql-fulltext-search filtering:
    TestModel.objects.filter(sphinx_field__search='find me')
    # Run match queries
        'find in all fields',
        sphinx_field='only in this field')
    # Insert and update documents to index
    obj = TestModel.objects.create(**values)
    obj.attr_uint = 1

Notes for production usage

  • Sphinxsearch engine has some issues with SQL-syntax support, and they vary from one version to another. I.e. float attributes are not comparable, string attributes were not comparable till v2.2.7.
  • Without limits sphinxsearch returns only 20 matched documents.
  • uint attributes accept -1 but return it as unsigned 32bit integer.
  • bigint accept 2**63 + 1 but return it as signed 64bit integer.
  • Use SphinxIntegerField and SphinxBigIntegerField instead of IntegerField and BigIntegerField from django.db.models, because IN is an expression in SQL (value IN column), but a function (IN(value, column)) in sphinxsearch.
  • Since 3.0.1 multi64 field incorrectly parses values greater than 2**31 and is completely unusable for bigint values.
  • In 3.1.1 you can create rt index with same string field and attr from config but you cant clone this index, so you attr-based filtering does not work.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for django-sphinxsearch, version 3.3.0
Filename, size & hash File type Python version Upload date
django_sphinxsearch-3.3.0.tar.gz (16.2 kB) View hashes Source None

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page