Prototyping Five Safes RO-Crate 1.0 tooling

[douglowe]

This is inspired by the SACRO RO-Crate example request from Tripti.

It is also inspired by Alex's recent updates to the Five Safes RO-Crate profile, in which we are moving away from workflow based profile to versioned profiles.

Thirdly, this is inspired by the provenance (pff) scoping.

Deliverables:

1. Example ro-crate for Tripti - for us to build on at the hackathon - Milestone: SACRO RO-Crate integration demonstrator #232
2. Python library for 5-Safe RO-Crate creation
3. Python library for gathering provenance information, and preparing this for inclusion in RO-Crate creation

All these should be done to some level by the hackathon (1st week in September).

[douglowe]

I am prototyping here: https://github.com/douglowe/sacro_provenance_crate

This will be split into 2 or 3 eScience repositories (and GH issues) soon - one for each python library, and another for the 'sacro' prototype tool.

[douglowe]

New eScience repositories:

• https://github.com/eScienceLab/provff-py
• https://github.com/eScienceLab/5s-crate-py
• https://github.com/eScienceLab/sacro-5s-rocrate

[douglowe]

The provenance library is probably sufficiently developed for this prototype.

The main work now is to build the 5-Safes RO-Crate library, and the command line tool interface.

[douglowe]

5-Safes RO-Crate defined here: https://github.com/eScienceLab/Five-Safes-RO-Crate

see markdown file: https://github.com/eScienceLab/Five-Safes-RO-Crate/blob/1.0/docs/index.md

[douglowe]

Items to add (non-definitive list):

• Process (CreateAction & prov:Activity) describing the running of (s)acro tool
• Output assets - note that we identify assets, then any copies of these assets included within the RO-Crate are linked to these identified assets (see https://github.com/eScienceLab/Five-Safes-RO-Crate/blob/1.0/docs/index.md#packaged-copies-of-identified-assets for a better description of this)
• FAIR assets - for each fair metadata record - make sure to link to the output asset object
• File assets - for each file output - again link to the relevant output asset object

[douglowe]

See this function as an example of how to reuse metadata from one crate within a new crate: https://github.com/workflowhub-eu/workflowhub-graph/pull/83/changes#diff-5b8128fc98506223730ddfc7c4ce9309b774cd3d646add070d03fe44dd2fef35R281