=====================
Model index reference
=====================

.. module:: django_mongodb_backend.indexes
   :synopsis: Database indexes for MongoDB.

Some MongoDB-specific :doc:`indexes <django:ref/models/indexes>`, for use on a
model's :attr:`Meta.indexes <django.db.models.Options.indexes>` option, are
available in ``django_mongodb_backend.indexes``.

Embedded field indexes
======================

``EmbeddedFieldIndex``
----------------------

.. class:: EmbeddedFieldIndex(**kwargs)

    .. versionadded:: 6.0.2

    Subclass of :class:`~django.db.models.Index` for use on a top-level model
    in order to add an index on subfields of embedded model fields.

    The ``fields`` argument uses dotted paths to reference embedded fields. For
    examples, see :ref:`embedded-model-field-indexes`,
    :ref:`embedded-model-array-field-indexes`,
    :ref:`polymorphic-embedded-model-field-indexes`, and
    :ref:`polymorphic-embedded-model-array-field-indexes`.

    .. versionchanged:: 6.0.3

        Support for subfields of polymorphic model fields was added. (In 6.0.2,
        only subfields of :class:`~.fields.EmbeddedModelField` and
        :class:`~.fields.EmbeddedModelArrayField` are supported.)

Search indexes
==============

MongoDB creates these indexes asynchronously, however, Django's
:class:`~django.db.migrations.operations.AddIndex` and
:class:`~django.db.migrations.operations.RemoveIndex` operations will wait
until the index is created or deleted so that the database state is
consistent in the operations that follow. Adding indexes may take seconds or
minutes, depending on the size of the collection.

.. versionchanged:: 5.2.1

    The aforementioned waiting was added.

``SearchIndex``
---------------

.. class:: SearchIndex(fields=(), field_mappings=None, name=None, analyzer=None, search_analyzer=None)

    Creates a basic :doc:`search index <search:index/index-definitions>` on the
    given field(s).

    Some fields such as :class:`~django.db.models.DecimalField` aren't
    supported. See the :ref:`Search documentation <search:bson-data-chart>` for
    a complete list of unsupported data types.

    Use ``field_mappings`` (instead of ``fields``) to create a complex search
    index. ``field_mappings`` is a dictionary that maps field names to index
    options. It corresponds to ``definition["mappings"]["fields"]`` in the
    :ref:`search:fts-static-mapping-examples`.

    If ``name`` isn't provided, one will be generated automatically. If you
    need to reference the name in your search query and don't provide your own
    name, you can lookup the generated one using ``Model._meta.indexes[0].name``
    (substituting the name of your model as well as a different list index if
    your model has multiple indexes).

    Use ``analyzer`` and ``search_analyzer`` to configure the indexing and
    searching :doc:`analyzer <search:index/analyzers/overview>`. If these
    options aren't provided, the server defaults to ``lucene.standard``. It
    corresponds to ``definition["analyzer"]`` and
    ``definition["searchAnalyzer"]`` in the
    :ref:`search:fts-static-mapping-examples`.

    .. versionchanged:: 5.2.2

        The ``field_mappings``, ``analyzer``, and ``search_analyzer`` arguments
        were added.

``VectorSearchIndex``
---------------------

.. class:: VectorSearchIndex(*, fields=(), name=None, similarities, indexing_methods=None)

    A subclass of :class:`SearchIndex` that creates a :doc:`vector search index
    <vector-search:index/vector-search-type>` on the given field(s).

    The index must reference at least one vector field: an :class:`.ArrayField`
    with a :attr:`~.ArrayField.base_field` of
    :class:`~django.db.models.FloatField` or
    :class:`~django.db.models.IntegerField` and a :attr:`~.ArrayField.size`. It
    cannot reference an :class:`.ArrayField` of any other type.

    It may also have other fields to filter on, provided the field stores
    ``boolean``, ``date``, ``objectId``, ``numeric``, ``string``, or ``uuid``.

    Available values for the required ``similarities`` keyword argument are
    ``"cosine"``, ``"dotProduct"``, and ``"euclidean"`` (see
    :ref:`vector-search:avs-similarity-functions` for how to choose). You can
    provide this value either as a string, in which case that value will be
    applied to all vector fields, or a list or tuple of values with a
    similarity corresponding to each vector field.

    The optional ``indexing_methods`` keyword argument selects the
    :ref:`indexing method <vector-search:avs-vector-index-method>` for each
    vector field. Available values are ``"hnsw"`` (the server's default if
    unspecified) and ``"flat"``. You can provide this value either as a string,
    in which case that value will be applied to all vector fields, or as a list
    or tuple of values with an indexing method corresponding to each vector
    field.

    .. versionchanged:: 6.0.4

        The ``indexing_methods`` argument was added.