Test dependencies

Testing often requires replacing real dependencies with mocks, stubs, or test-specific implementations. The injection.testing module provides dedicated registration decorators that work alongside your regular dependencies without interfering with production code.

Activating test dependencies

Before using test dependencies, you must activate the test profile using load_test_profile:

from injection.testing import load_test_profile

load_test_profile()

If you're using a ProfileLoader, pass it to load_test_profile:

from injection.loaders import ProfileLoader
from injection.testing import load_test_profile

profile_loader = ProfileLoader(...)

load_test_profile(profile_loader)

load_test_profile can also be used as a context manager to automatically unload the test profile after test complete:

from injection.testing import load_test_profile

with load_test_profile():
    # Test dependencies are active here
    run_test()

# Test profile is unloaded, cache is cleared

Test-specific decorators

All the registration decorators you've seen earlier have test equivalents in injection.testing:

from injection.testing import (
    set_test_constant,
    test_constant,
    test_injectable,
    test_scoped,
    test_singleton,
)

These decorators work exactly like their production counterparts but register dependencies in a separate test registry.

Using test decorators

test_injectable

from injection.testing import test_injectable

@test_injectable(on=AbstractDependency)
class MockDependency(AbstractDependency):
    ...

test_singleton

from injection.testing import test_singleton

@test_singleton(on=AbstractDependency)
class MockDependency(AbstractDependency):
    ...

set_test_constant

from injection.testing import set_test_constant

type APIKey = str

test_api_key = set_test_constant("test_key_123", APIKey, alias=True)

test_constant

from dataclasses import dataclass
from injection import constant
from injection.testing import test_constant
from os import environ

@dataclass(frozen=True)
class Settings:
    api_key: str
    debug: bool

@constant
def _settings_factory() -> Settings:
    return Settings(environ["API_KEY"], debug=False)

@test_constant
def _settings_test_factory() -> Settings:
    return Settings("test_key_123", debug=True)

test_scoped

from injection.testing import test_scoped

@test_scoped("<scope_name>", on=AbstractDependency)
class MockDependency(AbstractDependency):
    ...

How test dependencies work

Test dependencies are registered in a separate registry and take precedence over production dependencies when tests are running. This means:

• You can override production implementations with test versions
• Production code remains unchanged
• Test dependencies are isolated from production dependencies
• No need to modify your production registration code

Best practice

Register test dependencies in your test files or in a dedicated test configuration module. This keeps test setup close to your tests and makes it clear which dependencies are being mocked.