Open turbomam opened 2 days ago
Starting by generating Markdown documentation for all of the scripts in src/scripts
. Using ChatGPT 4.
find . -name "*.py" | sort
./src/mixs/datamodel/mixs.py ./src/mixs/_version.py ./src/scripts/camel_case_enums.py ./src/scripts/combinations_list_generator.py ./src/scripts/enumerations_list_generator.py ./src/scripts/extension_distances.py ./src/scripts/extension_slot_diffrences.py ./src/scripts/__init__.py ./src/scripts/isolate_slots.py ./src/scripts/mixs_slots_report.py ./src/scripts/organize_excel_files.py ./src/scripts/term_list_generator.py ./tests/__init__.py ./tests/test_data.py
I moved some of those scripts to an inactive
directory
src/scripts/isolate_slots.py
is derrived from output of linkml2schemasheets-template
and comes closest to generating a tabular representation of the schema, but it creates one global slot report, and we want per-class slot reports, showing the induced slot usage.
(I don't think we have determined what takes precedent yet, when a class has a is_a
parent and uses mixins
.)
I am inclined to read the schema into a SchemaView
, iterate over the in-scope classes, and report selected properties of the induced slots
/attributes
.
annotations
@cmungall would prefer that this kind of thing is developed in a LinkML repo, but the turnaround is slower. So develop with in a generalizable way with good coding practices so that it can eventually be moved into LinkML
I don't think the stock linkml2sheets
will work here either.
form @pbuttigieg: generate the reports in a dev path and then move them to a root path upon release, like ODK does
This could be backported to older releases. but the reports for previous releases wouldn't be bundled cumulatively with each new release
I suspect that this overlaps with a lot of existing issues, along the lines of "make it easy for people to find release artifacts"
380