Added Reference Dataset Support

[djriffle]

This pull request adds support for specifying an optional reference dataset in the CLI workflow, improving flexibility for users who need to work with both a primary and reference dataset. The changes include updates to command-line options, user prompts, resource management, and sandbox setup to ensure both datasets are handled correctly.

Support for Reference Dataset:

• Added a new --reference-dataset (-ref) CLI option to allow users to specify an optional reference dataset file, with corresponding updates to user prompts and context handling. [1] [2]
• Introduced a static in-container path (SANDBOX_REF_DATA_PATH) for the reference dataset, ensuring consistent handling within the sandbox environment.

Resource and Sandbox Management:

• Updated resource collection and mounting logic to include the reference dataset if provided, ensuring it is available inside the sandbox container. [1] [2]
• Modified sandbox setup and data copying procedures to bind mount or copy both the primary and reference datasets as needed for both Docker and Singularity backends. [1] [2]

User Experience Improvements:

• Improved dataset prompts to clarify the distinction between primary and reference datasets, enhancing usability.

These changes collectively make it easier for users to work with workflows that require both a primary and a reference dataset, while maintaining backward compatibility and clear user interaction.