`lime_types`¶

Scope: session | Returns: list[LimeType]

The lime_types fixture defines the set of LimeType objects that describe your test database schema. Other fixtures such as database and lime_app depend on it.

Default implementation¶

By default, lime_types returns limetypes that closely match the structure of the lime-core database. Replace it in your conftest.py to test against your own schema — which is what most projects do.

Basic usage¶

Most tests don't use lime_types directly. Define it in your conftest.py to establish the schema for your test suite:

import pytest
import lime_type

@pytest.fixture(scope="session")
def lime_types():
    return lime_type.create_limetypes_from_dsl({
        "company": {
            "name": {"type": "string", "required": True, "length": 200},
            "status": {
                "type": "option",
                "options": ["prospect", "active", "excustomer"],
                "defaultvalue": "prospect",
            },
        },
        "person": {
            "firstname": {"type": "string", "required": True, "length": 32},
            "lastname": {"type": "string", "length": 50},
            "company": {
                "type": "belongsto",
                "related": "company",
                "backref": "person",
            },
        },
    })

Keep the session scope

Always use scope="session" on your lime_types fixture. Changing it to a narrower scope is possible but requires careful consideration of how dependent fixtures behave.

Multiple definitions in different modules¶

You can define lime_types in multiple conftest.py files at different levels of your test tree. pytest resolves the nearest definition, so sub-directories can use a different schema from the one at the top level:

tests/
    conftest.py            # defines lime_types with company and person
    billing/
        conftest.py        # overrides lime_types with subscription and invoice types
        test_invoicing.py  # uses the billing schema
    crm/
        test_contacts.py   # uses the top-level schema

This is useful when different parts of your test suite exercise different database structures and don't need to carry the full schema everywhere.

Defining the schema in a YAML file¶

create_limetypes_from_dsl accepts both a dict and a YAML string:

tests/dsl.yaml

company:
    name:
        type: string
        required: True
        length: 200
person:
    firstname:
        type: string
        required: True
        length: 32
    company:
        type: belongsto
        related: company
        backref: person

import pytest
import lime_type

@pytest.fixture(scope="session")
def lime_types():
    with open("tests/dsl.yaml") as f:
        return lime_type.create_limetypes_from_dsl(f.read())

lime_types¶