Plan: Remove emit_per_factor — Always Emit One Emission per Factor¶

Context¶

The prepare_create() method in data_entry_emission_service.py has two code paths:

• emit_per_factor=True (lines 154-198): creates one DataEntryEmission per factor, using factor.emission_type_id
• emit_per_factor=False (lines 200-255): aggregates multiple factors into one DataEntryEmission row, summing kg_co2eq values

Only headcount (member/student) uses emit_per_factor=True. All other handlers use the default False and always resolve exactly 1 factor per computation — so the aggregation loop effectively produces 1 row anyway.

Goal: Remove the branching. Always create one DataEntryEmission per factor. No intermediate aggregation in Python.

Impact Analysis¶

Changes¶

1. backend/app/models/data_entry_emission.py¶

• Remove emit_per_factor field from EmissionComputation (lines 332-335)

2. backend/app/modules/headcount/schemas.py¶

• Remove emit_per_factor=True from both MemberHandler.resolve_computations() (line 193) and StudentHandler.resolve_computations() (line 249)

3. backend/app/services/data_entry_emission_service.py — prepare_create()¶

Replace the if comp.emit_per_factor: ... else: ... block (lines 154-255) with a single unified loop:

for factor in factors:
    per_factor_kg = self._apply_formula(ctx, factor.values or {}, comp)
    if per_factor_kg is None:
        # log warning (keep existing missing-key diagnostics)
        continue

    # Compute quantity for meta (used by chart breakdown)
    quantity: float | None = None
    if comp.quantity_key and ctx.get(comp.quantity_key) is not None:
        base_qty = float(ctx[comp.quantity_key])
        multiplier = float(
            (factor.values or {}).get(comp.multiplier_key, comp.multiplier_default)
            if comp.multiplier_key
            else comp.multiplier_default
        )
        quantity = base_qty * multiplier
    quantity_unit: str | None = (factor.values or {}).get("unit")

    results.append(
        DataEntryEmission(
            data_entry_id=data_entry.id,
            emission_type_id=factor.emission_type_id,
            primary_factor_id=factor.id,
            kg_co2eq=per_factor_kg,
            meta={
                "factors_used": [{"id": factor.id, "values": factor.values}],
                "quantity": quantity,
                "quantity_unit": quantity_unit,
                **ctx,
            },
        )
    )

The CSV kg_co2eq override (lines 143-152) stays but applies per-factor instead of once.

4. backend/app/utils/emission_category.py¶

• Remove the comment referencing "pre-emit_per_factor" (line 731). The HEADCOUNT_PER_FTE_KG fallback itself can stay — it only triggers when no real emission rows exist, which handles truly legacy data.

Files touched¶

1. backend/app/models/data_entry_emission.py — remove field
2. backend/app/modules/headcount/schemas.py — remove kwarg
3. backend/app/services/data_entry_emission_service.py — simplify prepare_create()
4. backend/app/utils/emission_category.py — update comment

Verification¶

1. pytest backend/ — run full backend test suite
2. Specifically check headcount emission tests produce the same per-sub-type rows
3. Check that non-headcount handlers still produce correct single-factor emissions
4. Verify chart breakdown endpoint returns correct values (emission_category tests)