Changing design of the GA workflows

[djarecka]

We have been adding more workflows that depend on each other and we started having issues with proper triggering etc. Based on the discussion with @puja-trivedi we decided to redesign the workflows structure, the main points of new design:

• for the models that start from google schema, we will have a workflow generate yaml file that will be triggered automatically and create a PR to the models repo
• for all models we will have generate_other_formats that will be triggered by changes in linkml-schema (push and PR to main so will run for open PRs). It will generate all other formats and push to the branch (so it will be visible as a part of PR). This will be likely one workflow that will run generate function for all the models that the yaml file has changed
• for all the models, we will have workflows push_to_bkbit that will be triggered manually that will create PR in bkbit (and run the bkbit tests)
• we will have a separate workflow for testing new linkml generators for all the yaml files (not sure how we will trigger)

Tasks to complete the change:

• Phase 1:

  • change generate_yaml_model.yaml so it creates PR and is triggered manually
  • generalize generate_yaml_model.yaml and create reusable workflow or model name as an input
  • connecting generate_yaml_model.yaml with generate_yaml_model.yaml
  • update generate_other_formats so it is clear for which model it runs (perhaps we can add some steps from test_other_formats, or just completely remove it)
  • changing push_to_bkbit and adding manual trigger with an input for different model names
  • adjust generate_docs
  • include the description in some md file

• Phase 2:

  • adding worflow that checks new version of linkml generators (and perhaps create a PR with updates?)
  • adding triggering bkbit tests after generate_other_formats is run (to be able to see effect on bkbit before merging)

[djarecka]

@puja-trivedi - for some reason, the generation of other format doesn't work when the PR is open with generate_yaml_model.yaml.... Have no idea why... will be checking this

EDIT:
I don't know why it didn't work last time, but it seems to be working now, both for the assertion and library_generation