casework / case-mapping-template-stubs Goto Github PK

The stubs are built for UCO, but do not include CASE classes. As a project on https://github.com/casework/, this repository should include CASE classes as well as UCO classes.

This repository should include a process to generate its delivery files.

The build directions currently in the README refer to a directory outside of this repository's source tree. This is helpful for users who wish to build non-released versions of UCO, but could use some adjustment to work with CI because I'm not sure Github's CI process can reliably write outside of the cloned repository root directory.

There are a couple options for how to retrieve an ontology build:

1. case-utils includes a build of the released monolithic CASE ontology, which incorporates the monolithic UCO build. So, instead of rdflib, case-utils could be in the requirements.txt.
2. The ontology repository could be cloned into a Git-ignored directory within this source tree (e.g. ${top_srcdir}/var/), or tracked as a Git submodule, so we know the edition of the ontology the stubs were built against.

Option 2 would enable a user to reset the ontology Git pointer to the state they want.

Would you be able to run a Github action to guarantee the files under templates/ and templates2/ are built against a certain version? The last step of the CI process could be running a recursive diff after running the generating commands.

If we have a CI process set up, it would reduce technical debt of needing to rely on a person to check.