Files
2026-07-13 13:22:34 +08:00

161 lines
5.7 KiB
Python

from importlib import reload
from unittest import mock
import pytest
import mlflow.tracking.context.registry
from mlflow.tracking.context.databricks_job_context import DatabricksJobRunContext
from mlflow.tracking.context.databricks_notebook_context import DatabricksNotebookRunContext
from mlflow.tracking.context.databricks_repo_context import DatabricksRepoRunContext
from mlflow.tracking.context.default_context import DefaultRunContext
from mlflow.tracking.context.git_context import GitRunContext
from mlflow.tracking.context.jupyter_notebook_context import JupyterNotebookRunContext
from mlflow.tracking.context.registry import RunContextProviderRegistry, resolve_tags
def test_run_context_provider_registry_register():
provider_class = mock.Mock()
registry = RunContextProviderRegistry()
registry.register(provider_class)
assert set(registry) == {provider_class.return_value}
def test_run_context_provider_registry_register_entrypoints():
provider_class = mock.Mock()
mock_entrypoint = mock.Mock()
mock_entrypoint.load.return_value = provider_class
with mock.patch(
"mlflow.utils.plugins._get_entry_points", return_value=[mock_entrypoint]
) as mock_get_group_all:
registry = RunContextProviderRegistry()
registry.register_entrypoints()
assert set(registry) == {provider_class.return_value}
mock_entrypoint.load.assert_called_once_with()
mock_get_group_all.assert_called_once_with("mlflow.run_context_provider")
@pytest.mark.parametrize(
"exception", [AttributeError("test exception"), ImportError("test exception")]
)
def test_run_context_provider_registry_register_entrypoints_handles_exception(exception):
mock_entrypoint = mock.Mock()
mock_entrypoint.load.side_effect = exception
with mock.patch(
"mlflow.utils.plugins._get_entry_points", return_value=[mock_entrypoint]
) as mock_get_group_all:
registry = RunContextProviderRegistry()
# Check that the raised warning contains the message from the original exception
with pytest.warns(UserWarning, match="test exception"):
registry.register_entrypoints()
mock_entrypoint.load.assert_called_once_with()
mock_get_group_all.assert_called_once_with("mlflow.run_context_provider")
def _currently_registered_run_context_provider_classes():
return {
provider.__class__
for provider in mlflow.tracking.context.registry._run_context_provider_registry
}
def test_registry_instance_defaults():
expected_classes = {
DefaultRunContext,
GitRunContext,
JupyterNotebookRunContext,
DatabricksNotebookRunContext,
DatabricksJobRunContext,
DatabricksRepoRunContext,
}
assert expected_classes.issubset(_currently_registered_run_context_provider_classes())
def test_registry_instance_loads_entrypoints():
class MockRunContext:
pass
mock_entrypoint = mock.Mock()
mock_entrypoint.load.return_value = MockRunContext
with mock.patch(
"mlflow.utils.plugins._get_entry_points", return_value=[mock_entrypoint]
) as mock_get_group_all:
# Entrypoints are registered at import time, so we need to reload the module to register the
# entrypoint given by the mocked entrypoints.get_group_all
reload(mlflow.tracking.context.registry)
assert MockRunContext in _currently_registered_run_context_provider_classes()
mock_get_group_all.assert_called_once_with("mlflow.run_context_provider")
def test_run_context_provider_registry_with_installed_plugin(tmp_path, monkeypatch):
monkeypatch.chdir(tmp_path)
reload(mlflow.tracking.context.registry)
from mlflow_test_plugin.run_context_provider import PluginRunContextProvider
assert PluginRunContextProvider in _currently_registered_run_context_provider_classes()
# The test plugin's context provider always returns False from in_context
# to avoid polluting tags in developers' environments. The following mock overrides this to
# perform the integration test.
with mock.patch.object(PluginRunContextProvider, "in_context", return_value=True):
assert resolve_tags()["test"] == "tag"
@pytest.fixture
def mock_run_context_providers():
base_provider = mock.Mock()
base_provider.in_context.return_value = True
base_provider.tags.return_value = {"one": "one-val", "two": "two-val", "three": "three-val"}
skipped_provider = mock.Mock()
skipped_provider.in_context.return_value = False
exception_provider = mock.Mock()
exception_provider.in_context.return_value = True
exception_provider.tags.return_value = {
"random-key": "This val will never make it to tag resolution"
}
exception_provider.tags.side_effect = Exception(
"This should be caught by logic in resolve_tags()"
)
override_provider = mock.Mock()
override_provider.in_context.return_value = True
override_provider.tags.return_value = {"one": "override", "new": "new-val"}
providers = [base_provider, skipped_provider, exception_provider, override_provider]
with mock.patch("mlflow.tracking.context.registry._run_context_provider_registry", providers):
yield
skipped_provider.tags.assert_not_called()
def test_resolve_tags(mock_run_context_providers):
tags_arg = {"two": "arg-override", "arg": "arg-val"}
assert resolve_tags(tags_arg) == {
"one": "override",
"two": "arg-override",
"three": "three-val",
"new": "new-val",
"arg": "arg-val",
}
def test_resolve_tags_no_arg(mock_run_context_providers):
assert resolve_tags() == {
"one": "override",
"two": "two-val",
"three": "three-val",
"new": "new-val",
}