mapper: Elmhurst path populates roof_construction (int) for cross-mapper parity

The gov-EPC API mapper sets BOTH roof_construction (int) and
roof_construction_type (str, derived via _API_ROOF_CONSTRUCTION_TO_STR),
but the Elmhurst mapper set only the string — leaving roof_construction
None on every site-notes cert. The SAP cascade reads the STRING (so SAP
cross-mapper parity always held), but consumers of the int (e.g.
domain/sap10_ml/transform.py ML aggregates `main_dwelling_roof_
construction`) silently saw None on the Elmhurst path.

New `_elmhurst_roof_construction_int` maps the Elmhurst roof-type code to
the same SAP10 int the API lodges (F→1, PN→3, PA→4, PS→8, S/A→7),
harvested from the committed Summary fixtures. Unlike the wall map it
returns None (not a strict-raise) for unmapped codes: the int is not
cascade-load-bearing, so an unknown roof must not block the cert (vaulted
5 / thatched 6 / NR omitted until a fixture surfaces them).

The 6 hand-built U985 reference fixtures gain the matching
roof_construction int (4/4/3 etc.) so test_from_elmhurst_site_notes_
matches_hand_built_* still asserts structural parity. SAP output is
unchanged (cascade reads the string). §4 suite green (2407 passed); the
two pre-existing stone-§5.6 sap10_ml failures are unrelated/out of scope.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

backend/documents_parser/tests/test_summary_pdf_mapper_chain.py

@@ -4361,6 +4361,32 @@ def test_elmhurst_foam_cylinder_insulation_still_maps_to_factory_code_1() -> Non
     assert code == 1
 
 
+def test_elmhurst_roof_construction_int_matches_api_codes() -> None:
+    # Arrange — cross-mapper structural parity: the gov-EPC API mapper
+    # populates BOTH roof_construction (int) and roof_construction_type
+    # (str derived via `_API_ROOF_CONSTRUCTION_TO_STR`), but the Elmhurst
+    # mapper set only the string, leaving the int None. The SAP cascade
+    # reads the string (so SAP parity held), but consumers of the int
+    # (e.g. domain/sap10_ml ML aggregates) saw None on every site-notes
+    # cert. `_elmhurst_roof_construction_int` closes the gap, mapping the
+    # Elmhurst roof code to the same SAP10 int the API lodges. Unmapped
+    # codes return None (not a raise) — the int is not cascade-load-
+    # bearing, so an unknown roof must not block the cert.
+    from datatypes.epc.domain.mapper import _elmhurst_roof_construction_int  # pyright: ignore[reportPrivateUsage]
+
+    # Act / Assert — each Elmhurst roof code → the gov-EPC API int.
+    assert _elmhurst_roof_construction_int("F Flat") == 1
+    assert _elmhurst_roof_construction_int("PN Pitched (slates/tiles), no access") == 3
+    assert _elmhurst_roof_construction_int("PA Pitched (slates/tiles), access to loft") == 4
+    assert _elmhurst_roof_construction_int("PS Pitched, sloping ceiling") == 8
+    assert _elmhurst_roof_construction_int("S Same dwelling above") == 7
+    assert _elmhurst_roof_construction_int("A Another dwelling above") == 7
+    # Absent / unmapped → None (no raise; not cascade-load-bearing).
+    assert _elmhurst_roof_construction_int(None) is None
+    assert _elmhurst_roof_construction_int("") is None
+    assert _elmhurst_roof_construction_int("NR Non-residential space above") is None
+
+
 def test_elmhurst_wall_is_basement_disambiguates_system_built_from_basement() -> None:
     # Arrange — "SY System build" and "B Basement wall" both map to SAP10
     # wall_construction=6 (canonical WALL_SYSTEM_BUILT). The explicit

datatypes/epc/domain/mapper.py

@@ -2314,6 +2314,40 @@ def _elmhurst_dwelling_type(
     return f"{position} flat"
 
 
+# Elmhurst roof-type codes → SAP10 roof_construction integer, matching the
+# gov-EPC API codes in `_API_ROOF_CONSTRUCTION_TO_STR` so the two
+# front-ends populate the same field. Harvested from the committed
+# Elmhurst Summary fixtures (corpus + cohort): F/PN/PA/PS/S/A. Vaulted (5)
+# and thatched (6) are omitted until a fixture surfaces their Elmhurst
+# codes. NR ("Non-residential space above") is intentionally left
+# unmapped — the gov enum's code 7 is specifically "dwelling above".
+_ELMHURST_ROOF_CODE_TO_SAP10: Dict[str, int] = {
+    "F": 1,    # Flat
+    "PN": 3,   # Pitched (slates/tiles), no access (to loft)
+    "PA": 4,   # Pitched (slates/tiles), access to loft
+    "PS": 8,   # Pitched, sloping ceiling
+    "S": 7,    # Same dwelling above
+    "A": 7,    # Another dwelling above
+}
+
+
+def _elmhurst_roof_construction_int(coded: Optional[str]) -> Optional[int]:
+    """Map an Elmhurst roof_type string ('PA Pitched (slates/tiles),
+    access to loft') to the SAP10 `roof_construction` integer the gov-EPC
+    API lodges (4), so the site-notes and API front-ends populate the
+    same field (cross-mapper structural parity).
+
+    Returns None for an absent or unmapped code — and, unlike
+    `_elmhurst_wall_construction_int`, does NOT raise. `roof_construction`
+    (int) is not read by the SAP cascade (which reads the string
+    `roof_construction_type`, populated on both paths), so an unmapped
+    roof code stays None — the pre-existing Elmhurst behaviour — rather
+    than blocking the cert."""
+    if not coded:
+        return None
+    return _ELMHURST_ROOF_CODE_TO_SAP10.get(_leading_code(coded))
+
+
 def _elmhurst_wall_construction_int(coded: str) -> Optional[int]:
     """Map an Elmhurst wall_type string ('CA Cavity') to the SAP10
     integer code (4). Returns None when the lodging is absent (empty
@@ -3494,6 +3528,7 @@ def _map_elmhurst_building_part(
             if walls.insulation_thickness_mm is not None
             else None
         ),
+        roof_construction=_elmhurst_roof_construction_int(roof.roof_type),
         roof_construction_type=_strip_code(roof.roof_type),
         roof_insulation_location=_strip_code(roof.insulation),
         roof_insulation_thickness=_resolve_sloping_ceiling_thickness(

tests/domain/sap10_calculator/worksheet/_elmhurst_worksheet_000474.py

@@ -65,6 +65,8 @@ def build_epc() -> EpcPropertyData:
     """
     main = SapBuildingPart(
         identifier=BuildingPartIdentifier.MAIN,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=4,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,
@@ -98,6 +100,8 @@ def build_epc() -> EpcPropertyData:
     )
     extension_1 = SapBuildingPart(
         identifier=BuildingPartIdentifier.EXTENSION_1,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=4,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,
@@ -130,6 +134,8 @@ def build_epc() -> EpcPropertyData:
     )
     extension_2 = SapBuildingPart(
         identifier=BuildingPartIdentifier.EXTENSION_2,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=3,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,

tests/domain/sap10_calculator/worksheet/_elmhurst_worksheet_000477.py

@@ -63,6 +63,8 @@ def build_epc() -> EpcPropertyData:
     """
     main = SapBuildingPart(
         identifier=BuildingPartIdentifier.MAIN,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=4,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,

tests/domain/sap10_calculator/worksheet/_elmhurst_worksheet_000480.py

@@ -64,6 +64,8 @@ def build_epc() -> EpcPropertyData:
     """
     main = SapBuildingPart(
         identifier=BuildingPartIdentifier.MAIN,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=4,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,
@@ -133,6 +135,8 @@ def build_epc() -> EpcPropertyData:
     )
     extension = SapBuildingPart(
         identifier=BuildingPartIdentifier.EXTENSION_1,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=4,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,

tests/domain/sap10_calculator/worksheet/_elmhurst_worksheet_000487.py

@@ -60,6 +60,8 @@ def build_epc() -> EpcPropertyData:
     """
     main = SapBuildingPart(
         identifier=BuildingPartIdentifier.MAIN,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=4,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,  # "A As Built"
@@ -130,6 +132,8 @@ def build_epc() -> EpcPropertyData:
     )
     extension = SapBuildingPart(
         identifier=BuildingPartIdentifier.EXTENSION_1,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=4,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,

tests/domain/sap10_calculator/worksheet/_elmhurst_worksheet_000490.py

@@ -65,6 +65,8 @@ def build_epc() -> EpcPropertyData:
     """
     main = SapBuildingPart(
         identifier=BuildingPartIdentifier.MAIN,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=4,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,
@@ -97,6 +99,8 @@ def build_epc() -> EpcPropertyData:
     )
     extension = SapBuildingPart(
         identifier=BuildingPartIdentifier.EXTENSION_1,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=4,
         construction_age_band="B",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,

tests/domain/sap10_calculator/worksheet/_elmhurst_worksheet_000516.py

@@ -69,6 +69,8 @@ def build_epc() -> EpcPropertyData:
     """
     main = SapBuildingPart(
         identifier=BuildingPartIdentifier.MAIN,
+        # API parity: roof_construction int mirrors the gov-EPC mapper
+        roof_construction=3,
         construction_age_band="A",
         wall_construction=_WC_CAVITY,
         wall_insulation_type=4,