Re-order ClinGen Variant Submissions

[bencap]

This pull request refactors the job submission and error handling logic for mapping variant data and submitting it to external services (ClinGen Allele Registry and LDH). These changes necessitated a rework of how publication testing is handled to reduce duplication.

Error handling and exception cleanup:

• Removed the unused LinkingEnqueueError exception class from src/mavedb/lib/exceptions.py and its imports, simplifying the exception hierarchy. [1] [2]

Job submission and queue logic improvements:

• Refactored the job enqueuing logic in publish_score_set to use more descriptive variable names (refresh_published_variants_job, submit_to_ldh_job) and improved logging context keys for better traceability.
• Removed unnecessary redis job enqueuing for LDH submission from the CAR submission job, streamlining the workflow and eliminating unnecessary job chaining.
• Invoked the LDH submission job on score set publication, rather than after CAR submission.
• Construct hgvs_assay_level after mapping completes, so it is available directly for the CAR and LDH submission jobs.
• Removed the unnecessary link_clingen_variants job, as this linkage is handled directly by CAR submission.

HGVS string handling and logging enhancements:

• Changed the mapped variant selection to use the new hgvs_assay_level field instead of post_mapped, and added logging for variants missing this field, improving error visibility and data integrity. [1] [2]
• Updated error messages and logging to consistently refer to CAR (ClinGen Allele Registry) instead of LDH where appropriate, clarifying the target service and improving log accuracy. [1] [2]

Imports and code cleanup:

• Cleaned up imports in src/mavedb/worker/jobs.py, removing unused functions and exception imports, and adding func for SQL operations. [1] [2]