Indexes

Indexes setup

There are more than one way to set up indexes using Beanie

Indexed function

To set up an index over a single field, the Indexed function can be used to wrap the type and does not require a Settings class:

from bunnet import Document
from bunnet import Indexed


class Sample(Document):
    num: Indexed(int)
    description: str

The Indexed function takes an optional index_type argument, which may be set to a pymongo index type:

import pymongo

from bunnet import Document
from bunnet import Indexed


class Sample(Document):
    description: Indexed(str, index_type=pymongo.TEXT)

The Indexed function also supports PyMongo's IndexModel kwargs arguments (see the PyMongo Documentation for details).

For example, to create a unique index:

from bunnet import Document
from bunnet import Indexed


class Sample(Document):
    name: Indexed(str, unique=True)

Multi-field indexes

The indexes field of the inner Settings class is responsible for more complex indexes. It is a list where items can be:

• Single key. Name of the document's field (this is equivalent to using the Indexed function described above without any additional arguments)
• List of (key, direction) pairs. Key - string, name of the document's field. Direction - pymongo direction ( example: pymongo.ASCENDING)
• pymongo.IndexModel instance - the most flexible option. PyMongo Documentation

import pymongo
from pymongo import IndexModel

from bunnet import Document


class Sample(Document):
    test_int: int
    test_str: str

    class Settings:
        indexes = [
            "test_int",
            [
                ("test_int", pymongo.ASCENDING),
                ("test_str", pymongo.DESCENDING),
            ],
            IndexModel(
                [("test_str", pymongo.DESCENDING)],
                name="test_string_index_DESCENDING",
            ),
        ]