2. Wrappers for Interface Construction

2.1. Overview

In the workflow system, interfaces are built using wrappers that encapsulate tools and expose them in a standardized way. These wrappers serve as the building blocks of workflows, enabling seamless integration and execution of diverse tools.

There are two primary types of wrappers:

  1. Command-Line Wrappers

  2. Python Wrappers

2.2. Wrapper Implementation

All wrappers are implemented as Python decorators, which provide a clean and expressive way to define interfaces.

Benefits of Using Decorators:

  • Simplifies interface definition

  • Promotes modular and reusable code

  • Enables automatic metadata extraction

  • Supports consistent execution patterns

2.3. Command-Line Wrappers

Command-line wrappers encapsulate tools that are available on the system and can be executed via the shell. These wrappers provide a structured interface to interact with external binaries, scripts, or utilities.

Features:

  • Execute system-level tools using subprocesses.

  • Capture standard output and error streams.

  • Support for input/output file management.

  • Parameter mapping to command-line arguments.

Return Value:

A command-line wrapper returns:

  • A command or list of commands to be executed.

  • A tuple of generated output paths.

2.4. Python Wrappers

Python wrappers encapsulate tools implemented directly in Python. These wrappers allow for tighter integration with the workflow engine and enable more complex logic and data manipulation.

Features:

  • Direct invocation of Python functions or classes.

  • Access to in-memory data structures.

  • Easier debugging and testing.

  • Richer error handling and logging.

Return Value:

A Python wrapper returns:

  • A tuple of generated output paths.

2.5. Usage in Workflows

Both types of wrappers are used to define interfaces (modular units of computation that can be chained together in workflows). By abstracting tool-specific details, wrappers promote reusability, portability, and clarity in workflow design.

Note

Wrappers should be designed with clear input/output specifications and robust error handling to ensure reliability across diverse execution environments.