REOPENED. PDF + Excel exporters read non-existent dict keys, so demand/supply/ scenario columns silently rendered "—". Tests passed only because the fixtures were stale (hand-typed the same wrong keys → fixture agreed with buggy exporter). - future_market: demand/supply → projected_demand_units/projected_supply_units - scenarios: drop non-existent per-scenario "overall"; show primary-horizon deficit_index from ScenarioForecast.forecasts (scoring.overall was NOT broken) - Excel #991: add missing future_supply (index + breakdown) + confidence.factors sections; add future_supply to PDF for parity - tests: rebuild forecast/scenario fixtures from real DemandSupplyForecast / ScenarioForecast as_dict(); contract-key regression guards fail on key-drift (verified: reintroducing old keys fails the new tests). 28 passed. Refs #989 #991 |
||
|---|---|---|
| .. | ||
| api | ||
| core | ||
| models | ||
| observability | ||
| schemas | ||
| scrapers | ||
| services | ||
| templates | ||
| workers | ||
| __init__.py | ||
| main.py | ||