Description

This is WIP. Do not merge!
This is a new feature to support preparation for preservation checks.

Closes #235

ToDo List

@astrochun:
As of 07/26/2021, here is the current status of this PR:
While most methods have been tested at a minimum level for the preservation check workflow, it is still not fully tested/explored. The methods of Preserve that have been tested include: get_metadata, save_metadata, check_files, make_symbolic_links, delete_old_readme_files, delete_hidden_files, delete_files (used by delete_hidden_files).

I have not tested:

• The Preserve.update_files method
• I have not done full end-to-end testing with instances where there is (1) no issue (i.e., curation version matches published version), (2) cases where the are differences between curation version and published version (we were not immediately aware of such a version for testing).
• I have not also tested this against dataset that does not have data (e.g., NPN datasets). Presumeably the --metadata_only option should work for this purpose.

In addition, we should have the script run as a "dry run" by default and to have a --update option. This will ensure that files are not replaced and to not download metadata. This would be a keyword input option that would need to be available to all method, probably through a (self) instance variable in Preserve.

Test plan

Update Changelog

• CHANGELOG.md

Resources

Screenshots or additional context