Directory Setup
This page outlines the unit test coverage for the directory_setup.py module, which is responsible for initializing the automl_workspace directory structure.
Coverage Overview
The tests validate that the function create_automl_workspace():
Creates a well-defined directory tree for data pipeline, model registry, config, and label studio
Supports custom base paths
Gracefully handles already existing directories
Responds to permission errors with informative messages
Properly constructs nested directories as intended
Function Tests
test_create_automl_workspace_default_path
Verifies that the default workspace is created correctly when no custom path is provided.
test_create_automl_workspace_custom_path
Checks that the workspace is created under a user-specified base directory.
test_create_automl_workspace_already_exists
Confirms the function does not overwrite existing files and directories when re-run.
test_create_automl_workspace_permission_error
Simulates a PermissionError during directory creation and verifies that the error is logged appropriately.
test_create_automl_workspace_nested_structure
Ensures that deeply nested directories like label_studio/pending are created as expected.
test_create_automl_workspace_makedirs_called_correctly
Mocks os.makedirs and asserts it is called for all required subdirectories.
Summary
These tests guarantee that create_automl_workspace() reliably sets up the workspace environment in various conditions, making the AutoML pipeline reproducible and structured across different setups.