Refactor Docker setup and migration scripts

- Updated Dockerfile to set permissions for the entrypoint script and defined the entrypoint for the container.
- Consolidated Alembic migration history into a single initial migration file and removed obsolete revision files.
- Added a new script to run Alembic migrations before starting the application.
- Updated changelog to reflect changes in migration handling and Docker setup.
- Enhanced pytest configuration for coverage reporting and excluded specific files from coverage calculations.
This commit is contained in:
2025-11-11 18:30:15 +01:00
parent 795a9f99f4
commit 4364927965
9 changed files with 794 additions and 435 deletions

View File

@@ -46,3 +46,6 @@
- Authored `calminer-docs/specifications/financial_metrics.md` capturing DCF assumptions, solver behaviours, and worked examples, and cross-linked the architecture concepts to the new reference for consistent navigation.
- Implemented `services/simulation.py` Monte Carlo engine with configurable distributions, summary aggregation, and reproducible RNG seeding, introduced regression tests in `tests/test_simulation.py`, and documented configuration/usage in `calminer-docs/specifications/monte_carlo_simulation.md` with architecture cross-links.
- Polished reporting HTML contexts by cleaning stray fragments in `routes/reports.py`, adding download action metadata for project and scenario pages, and generating scenario comparison download URLs with correctly serialised repeated `scenario_ids` parameters.
- Consolidated Alembic history into a single initial migration (`20251111_00_initial_schema.py`), removed superseded revision files, and ensured Alembic metadata still references the project metadata for clean bootstrap.
- Added `scripts/run_migrations.py` and a Docker entrypoint wrapper to run Alembic migrations before `uvicorn` starts, removed the fallback `Base.metadata.create_all` call, and updated `calminer-docs/admin/installation.md` so developers know how to apply migrations locally or via Docker.
- Configured pytest defaults to collect coverage (`--cov`) with an 80% fail-under gate, excluded entrypoint/reporting scaffolds from the calculation, updated contributor docs with the standard `pytest` command, and verified the suite now reports 83% coverage.