Merge branch 'main' into feature/pashub-to-ara

.devcontainer/backend/requirements.txt

@@ -7,6 +7,7 @@ python-jose==3.3.0
 cryptography==43.0.3
 mangum==0.19.0
 playwright==1.58.0
+pymupdf
 # AWS
 boto3==1.35.44
 # Data

.dockerignore

@@ -18,5 +18,6 @@ utils/tests/*
 etl/epc/tests/*
 etl/epc_clean/tests/*
 etl/spatial/tests/*
+backend/documents_parser/tests/*
 
 

.gitignore

@@ -280,7 +280,7 @@ cache/
 *.pptx
 *.csv
 *.xlsx
-*.pdf
+# *.pdf
 **/Chunks/
 *.ipynb
 

backend/app/db/models/epc_property.py

@@ -0,0 +1,658 @@
+from __future__ import annotations
+
+from typing import Optional
+from sqlmodel import SQLModel, Field
+
+from datatypes.epc.domain.epc_property_data import (
+    EpcPropertyData,
+    EnergyElement,
+    MainHeatingDetail,
+    SapBuildingPart,
+    SapFloorDimension,
+    SapFlatDetails,
+    SapWindow,
+)
+
+
+class EpcPropertyModel(SQLModel, table=True):
+    __tablename__ = "epc_property"
+
+    id: Optional[int] = Field(default=None, primary_key=True)
+    property_id: int = Field(foreign_key="property.id", nullable=False)
+    portfolio_id: int = Field(foreign_key="portfolio.id", nullable=False)
+
+    # Identity / admin
+    uprn: Optional[int] = Field(default=None)
+    uprn_source: Optional[str] = Field(default=None)
+    report_reference: Optional[str] = Field(default=None)
+    report_type: Optional[str] = Field(default=None)
+    assessment_type: Optional[str] = Field(default=None)
+    sap_version: Optional[float] = Field(default=None)
+    schema_type: Optional[str] = Field(default=None)
+    schema_versions_original: Optional[str] = Field(default=None)
+    status: Optional[str] = Field(default=None)
+    calculation_software_version: Optional[str] = Field(default=None)
+
+    # Address
+    address_line_1: Optional[str] = Field(default=None)
+    address_line_2: Optional[str] = Field(default=None)
+    post_town: Optional[str] = Field(default=None)
+    postcode: Optional[str] = Field(default=None)
+    region_code: Optional[str] = Field(default=None)
+    country_code: Optional[str] = Field(default=None)
+    language_code: Optional[str] = Field(default=None)
+
+    # Property description
+    dwelling_type: str
+    property_type: Optional[str] = Field(default=None)
+    built_form: Optional[str] = Field(default=None)
+    tenure: str
+    transaction_type: str
+    inspection_date: str  # store as ISO string; cast on read if needed
+    completion_date: Optional[str] = Field(default=None)
+    registration_date: Optional[str] = Field(default=None)
+    total_floor_area_m2: float
+    measurement_type: Optional[int] = Field(default=None)
+
+    # Flags
+    solar_water_heating: bool
+    has_hot_water_cylinder: bool
+    has_fixed_air_conditioning: bool
+    has_conservatory: Optional[bool] = Field(default=None)
+    has_heated_separate_conservatory: Optional[bool] = Field(default=None)
+    conservatory_type: Optional[int] = Field(default=None)
+
+    # Counts
+    door_count: int
+    wet_rooms_count: int
+    extensions_count: int
+    heated_rooms_count: int
+    open_chimneys_count: int
+    habitable_rooms_count: int
+    insulated_door_count: int
+    cfl_fixed_lighting_bulbs_count: int
+    led_fixed_lighting_bulbs_count: int
+    incandescent_fixed_lighting_bulbs_count: int
+    blocked_chimneys_count: Optional[int] = Field(default=None)
+    draughtproofed_door_count: Optional[int] = Field(default=None)
+    energy_rating_average: Optional[int] = Field(default=None)
+    low_energy_fixed_lighting_bulbs_count: Optional[int] = Field(default=None)
+    fixed_lighting_outlets_count: Optional[int] = Field(default=None)
+    low_energy_fixed_lighting_outlets_count: Optional[int] = Field(default=None)
+    number_of_storeys: Optional[int] = Field(default=None)
+    any_unheated_rooms: Optional[bool] = Field(default=None)
+
+    # Misc
+    hydro: Optional[bool] = Field(default=None)
+    photovoltaic_array: Optional[bool] = Field(default=None)
+    waste_water_heat_recovery: Optional[str] = Field(default=None)
+    pressure_test: Optional[int] = Field(default=None)
+    pressure_test_certificate_number: Optional[int] = Field(default=None)
+    percent_draughtproofed: Optional[int] = Field(default=None)
+    insulated_door_u_value: Optional[float] = Field(default=None)
+    multiple_glazed_proportion: Optional[int] = Field(default=None)
+    windows_transmission_u_value: Optional[float] = Field(default=None)
+    windows_transmission_data_source: Optional[int] = Field(default=None)
+    windows_transmission_solar_transmittance: Optional[float] = Field(default=None)
+
+    # Energy source
+    energy_mains_gas: bool
+    energy_meter_type: str
+    energy_pv_battery_count: int
+    energy_wind_turbines_count: int
+    energy_gas_smart_meter_present: bool
+    energy_is_dwelling_export_capable: bool
+    energy_wind_turbines_terrain_type: str
+    energy_electricity_smart_meter_present: bool
+    energy_pv_connection: Optional[str] = Field(default=None)
+    energy_pv_percent_roof_area: Optional[int] = Field(default=None)
+    energy_pv_battery_capacity: Optional[float] = Field(default=None)
+    energy_wind_turbine_hub_height: Optional[float] = Field(default=None)
+    energy_wind_turbine_rotor_diameter: Optional[float] = Field(default=None)
+
+    # Heating config
+    heating_cylinder_size: Optional[str] = Field(default=None)
+    heating_water_heating_code: Optional[int] = Field(default=None)
+    heating_water_heating_fuel: Optional[int] = Field(default=None)
+    heating_immersion_heating_type: Optional[str] = Field(default=None)
+    heating_cylinder_insulation_type: Optional[str] = Field(default=None)
+    heating_cylinder_thermostat: Optional[str] = Field(default=None)
+    heating_secondary_fuel_type: Optional[int] = Field(default=None)
+    heating_secondary_heating_type: Optional[str] = Field(default=None)
+    heating_cylinder_insulation_thickness_mm: Optional[int] = Field(default=None)
+    heating_wwhrs_index_number_1: Optional[int] = Field(default=None)
+    heating_wwhrs_index_number_2: Optional[int] = Field(default=None)
+    heating_shower_outlet_type: Optional[str] = Field(default=None)
+    heating_shower_wwhrs: Optional[int] = Field(default=None)
+
+    # Ventilation
+    ventilation_type: Optional[str] = Field(default=None)
+    ventilation_draught_lobby: Optional[bool] = Field(default=None)
+    ventilation_pressure_test: Optional[str] = Field(default=None)
+    ventilation_open_flues_count: Optional[int] = Field(default=None)
+    ventilation_closed_flues_count: Optional[int] = Field(default=None)
+    ventilation_boiler_flues_count: Optional[int] = Field(default=None)
+    ventilation_other_flues_count: Optional[int] = Field(default=None)
+    ventilation_extract_fans_count: Optional[int] = Field(default=None)
+    ventilation_passive_vents_count: Optional[int] = Field(default=None)
+    ventilation_flueless_gas_fires_count: Optional[int] = Field(default=None)
+    ventilation_in_pcdf_database: Optional[bool] = Field(default=None)
+    mechanical_ventilation: Optional[int] = Field(default=None)
+    mechanical_vent_duct_type: Optional[int] = Field(default=None)
+    mechanical_vent_duct_placement: Optional[int] = Field(default=None)
+    mechanical_vent_duct_insulation: Optional[int] = Field(default=None)
+    mechanical_ventilation_index_number: Optional[int] = Field(default=None)
+    mechanical_vent_measured_installation: Optional[str] = Field(default=None)
+
+    @classmethod
+    def from_epc_property_data(
+        cls,
+        data: EpcPropertyData,
+        property_id: int,
+        portfolio_id: int,
+    ) -> EpcPropertyModel:
+        es = data.sap_energy_source
+        h = data.sap_heating
+        v = data.sap_ventilation
+        shower = h.shower_outlets.shower_outlet if h.shower_outlets else None
+        pv = es.photovoltaic_supply
+        wt = es.wind_turbine_details
+        pvb = es.pv_batteries
+
+        return cls(
+            property_id=property_id,
+            portfolio_id=portfolio_id,
+            uprn=data.uprn,
+            uprn_source=data.uprn_source,
+            report_reference=data.report_reference,
+            report_type=data.report_type,
+            assessment_type=data.assessment_type,
+            sap_version=data.sap_version,
+            schema_type=data.schema_type,
+            schema_versions_original=data.schema_versions_original,
+            status=data.status,
+            calculation_software_version=data.calculation_software_version,
+            address_line_1=data.address_line_1,
+            address_line_2=data.address_line_2,
+            post_town=data.post_town,
+            postcode=data.postcode,
+            region_code=data.region_code,
+            country_code=data.country_code,
+            language_code=data.language_code,
+            dwelling_type=data.dwelling_type,
+            property_type=data.property_type,
+            built_form=data.built_form,
+            tenure=data.tenure,
+            transaction_type=data.transaction_type,
+            inspection_date=data.inspection_date.isoformat(),
+            completion_date=(
+                data.completion_date.isoformat() if data.completion_date else None
+            ),
+            registration_date=(
+                data.registration_date.isoformat() if data.registration_date else None
+            ),
+            total_floor_area_m2=data.total_floor_area_m2,
+            measurement_type=data.measurement_type,
+            solar_water_heating=data.solar_water_heating,
+            has_hot_water_cylinder=data.has_hot_water_cylinder,
+            has_fixed_air_conditioning=data.has_fixed_air_conditioning,
+            has_conservatory=data.has_conservatory,
+            has_heated_separate_conservatory=data.has_heated_separate_conservatory,
+            conservatory_type=data.conservatory_type,
+            door_count=data.door_count,
+            wet_rooms_count=data.wet_rooms_count,
+            extensions_count=data.extensions_count,
+            heated_rooms_count=data.heated_rooms_count,
+            open_chimneys_count=data.open_chimneys_count,
+            habitable_rooms_count=data.habitable_rooms_count,
+            insulated_door_count=data.insulated_door_count,
+            cfl_fixed_lighting_bulbs_count=data.cfl_fixed_lighting_bulbs_count,
+            led_fixed_lighting_bulbs_count=data.led_fixed_lighting_bulbs_count,
+            incandescent_fixed_lighting_bulbs_count=data.incandescent_fixed_lighting_bulbs_count,
+            blocked_chimneys_count=data.blocked_chimneys_count,
+            draughtproofed_door_count=data.draughtproofed_door_count,
+            energy_rating_average=data.energy_rating_average,
+            low_energy_fixed_lighting_bulbs_count=data.low_energy_fixed_lighting_bulbs_count,
+            fixed_lighting_outlets_count=data.fixed_lighting_outlets_count,
+            low_energy_fixed_lighting_outlets_count=data.low_energy_fixed_lighting_outlets_count,
+            number_of_storeys=data.number_of_storeys,
+            any_unheated_rooms=data.any_unheated_rooms,
+            hydro=data.hydro,
+            photovoltaic_array=data.photovoltaic_array,
+            waste_water_heat_recovery=data.waste_water_heat_recovery,
+            pressure_test=data.pressure_test,
+            pressure_test_certificate_number=data.pressure_test_certificate_number,
+            percent_draughtproofed=data.percent_draughtproofed,
+            insulated_door_u_value=data.insulated_door_u_value,
+            multiple_glazed_proportion=data.multiple_glazed_propertion,
+            windows_transmission_u_value=(
+                data.windows_transmission_details.u_value
+                if data.windows_transmission_details
+                else None
+            ),
+            windows_transmission_data_source=(
+                data.windows_transmission_details.data_source
+                if data.windows_transmission_details
+                else None
+            ),
+            windows_transmission_solar_transmittance=(
+                data.windows_transmission_details.solar_transmittance
+                if data.windows_transmission_details
+                else None
+            ),
+            energy_mains_gas=es.mains_gas,
+            energy_meter_type=str(es.meter_type),
+            energy_pv_battery_count=es.pv_battery_count,
+            energy_wind_turbines_count=es.wind_turbines_count,
+            energy_gas_smart_meter_present=es.gas_smart_meter_present,
+            energy_is_dwelling_export_capable=es.is_dwelling_export_capable,
+            energy_wind_turbines_terrain_type=str(es.wind_turbines_terrain_type),
+            energy_electricity_smart_meter_present=es.electricity_smart_meter_present,
+            energy_pv_connection=(
+                str(es.pv_connection) if es.pv_connection is not None else None
+            ),
+            energy_pv_percent_roof_area=(
+                pv.none_or_no_details.percent_roof_area if pv else None
+            ),
+            energy_pv_battery_capacity=pvb.pv_battery.battery_capacity if pvb else None,
+            energy_wind_turbine_hub_height=wt.hub_height if wt else None,
+            energy_wind_turbine_rotor_diameter=wt.rotor_diameter if wt else None,
+            heating_cylinder_size=(
+                str(h.cylinder_size) if h.cylinder_size is not None else None
+            ),
+            heating_water_heating_code=h.water_heating_code,
+            heating_water_heating_fuel=h.water_heating_fuel,
+            heating_immersion_heating_type=(
+                str(h.immersion_heating_type)
+                if h.immersion_heating_type is not None
+                else None
+            ),
+            heating_cylinder_insulation_type=(
+                str(h.cylinder_insulation_type)
+                if h.cylinder_insulation_type is not None
+                else None
+            ),
+            heating_cylinder_thermostat=h.cylinder_thermostat,
+            heating_secondary_fuel_type=h.secondary_fuel_type,
+            heating_secondary_heating_type=(
+                str(h.secondary_heating_type)
+                if h.secondary_heating_type is not None
+                else None
+            ),
+            heating_cylinder_insulation_thickness_mm=h.cylinder_insulation_thickness_mm,
+            heating_wwhrs_index_number_1=h.instantaneous_wwhrs.wwhrs_index_number1,
+            heating_wwhrs_index_number_2=h.instantaneous_wwhrs.wwhrs_index_number2,
+            heating_shower_outlet_type=(
+                str(shower.shower_outlet_type) if shower else None
+            ),
+            heating_shower_wwhrs=shower.shower_wwhrs if shower else None,
+            ventilation_type=v.ventilation_type if v else None,
+            ventilation_draught_lobby=v.draught_lobby if v else None,
+            ventilation_pressure_test=v.pressure_test if v else None,
+            ventilation_open_flues_count=v.open_flues_count if v else None,
+            ventilation_closed_flues_count=v.closed_flues_count if v else None,
+            ventilation_boiler_flues_count=v.boiler_flues_count if v else None,
+            ventilation_other_flues_count=v.other_flues_count if v else None,
+            ventilation_extract_fans_count=v.extract_fans_count if v else None,
+            ventilation_passive_vents_count=v.passive_vents_count if v else None,
+            ventilation_flueless_gas_fires_count=(
+                v.flueless_gas_fires_count if v else None
+            ),
+            ventilation_in_pcdf_database=v.ventilation_in_pcdf_database if v else None,
+            mechanical_ventilation=data.mechanical_ventilation,
+            mechanical_vent_duct_type=data.mechanical_vent_duct_type,
+            mechanical_vent_duct_placement=data.mechanical_vent_duct_placement,
+            mechanical_vent_duct_insulation=data.mechanical_vent_duct_insulation,
+            mechanical_ventilation_index_number=data.mechanical_ventilation_index_number,
+            mechanical_vent_measured_installation=data.mechanical_vent_measured_installation,
+        )
+
+
+class EpcPropertyEnergyPerformanceModel(SQLModel, table=True):
+    __tablename__ = "epc_property_energy_performance"
+
+    id: Optional[int] = Field(default=None, primary_key=True)
+    epc_property_id: int = Field(
+        foreign_key="epc_property.id", nullable=False, unique=True
+    )
+
+    energy_rating_current: Optional[int] = Field(default=None)
+    energy_consumption_current: Optional[int] = Field(default=None)
+    environmental_impact_current: Optional[int] = Field(default=None)
+    heating_cost_current: Optional[float] = Field(default=None)
+    lighting_cost_current: Optional[float] = Field(default=None)
+    hot_water_cost_current: Optional[float] = Field(default=None)
+    co2_emissions_current: Optional[float] = Field(default=None)
+    co2_emissions_current_per_floor_area: Optional[int] = Field(default=None)
+    current_energy_efficiency_band: Optional[str] = Field(default=None)
+    energy_rating_potential: Optional[float] = Field(default=None)
+    energy_consumption_potential: Optional[int] = Field(default=None)
+    environmental_impact_potential: Optional[int] = Field(default=None)
+    heating_cost_potential: Optional[float] = Field(default=None)
+    lighting_cost_potential: Optional[float] = Field(default=None)
+    hot_water_cost_potential: Optional[float] = Field(default=None)
+    co2_emissions_potential: Optional[float] = Field(default=None)
+    potential_energy_efficiency_band: Optional[str] = Field(default=None)
+
+    @classmethod
+    def from_epc_property_data(
+        cls, data: EpcPropertyData, epc_property_id: int
+    ) -> EpcPropertyEnergyPerformanceModel:
+        return cls(
+            epc_property_id=epc_property_id,
+            energy_rating_current=data.energy_rating_current,
+            energy_consumption_current=data.energy_consumption_current,
+            environmental_impact_current=data.environmental_impact_current,
+            heating_cost_current=data.heating_cost_current,
+            lighting_cost_current=data.lighting_cost_current,
+            hot_water_cost_current=data.hot_water_cost_current,
+            co2_emissions_current=data.co2_emissions_current,
+            co2_emissions_current_per_floor_area=data.co2_emissions_current_per_floor_area,
+            current_energy_efficiency_band=(
+                data.current_energy_efficiency_band.value
+                if data.current_energy_efficiency_band
+                else None
+            ),
+            energy_rating_potential=data.energy_rating_potential,
+            energy_consumption_potential=data.energy_consumption_potential,
+            environmental_impact_potential=data.environmental_impact_potential,
+            heating_cost_potential=data.heating_cost_potential,
+            lighting_cost_potential=data.lighting_cost_potential,
+            hot_water_cost_potential=data.hot_water_cost_potential,
+            co2_emissions_potential=data.co2_emissions_potential,
+            potential_energy_efficiency_band=(
+                data.potential_energy_efficiency_band.value
+                if data.potential_energy_efficiency_band
+                else None
+            ),
+        )
+
+
+class EpcFlatDetailsModel(SQLModel, table=True):
+    __tablename__ = "epc_flat_details"
+
+    id: Optional[int] = Field(default=None, primary_key=True)
+    epc_property_id: int = Field(
+        foreign_key="epc_property.id", nullable=False, unique=True
+    )
+
+    level: int
+    top_storey: str
+    flat_location: int
+    heat_loss_corridor: int
+    storey_count: Optional[int] = Field(default=None)
+    unheated_corridor_length_m: Optional[int] = Field(default=None)
+
+    @classmethod
+    def from_domain(
+        cls, flat: SapFlatDetails, epc_property_id: int
+    ) -> EpcFlatDetailsModel:
+        return cls(
+            epc_property_id=epc_property_id,
+            level=flat.level,
+            top_storey=flat.top_storey,
+            flat_location=flat.flat_location,
+            heat_loss_corridor=flat.heat_loss_corridor,
+            storey_count=flat.storey_count,
+            unheated_corridor_length_m=flat.unheated_corridor_length_m,
+        )
+
+
+class EpcMainHeatingDetailModel(SQLModel, table=True):
+    __tablename__ = "epc_main_heating_detail"
+
+    id: Optional[int] = Field(default=None, primary_key=True)
+    epc_property_id: int = Field(foreign_key="epc_property.id", nullable=False)
+
+    has_fghrs: bool
+    main_fuel_type: str
+    heat_emitter_type: str
+    emitter_temperature: str
+    main_heating_control: str
+    fan_flue_present: Optional[bool] = Field(default=None)
+    boiler_flue_type: Optional[int] = Field(default=None)
+    boiler_ignition_type: Optional[int] = Field(default=None)
+    central_heating_pump_age: Optional[int] = Field(default=None)
+    central_heating_pump_age_str: Optional[str] = Field(default=None)
+    main_heating_index_number: Optional[int] = Field(default=None)
+    sap_main_heating_code: Optional[int] = Field(default=None)
+    main_heating_number: Optional[int] = Field(default=None)
+    main_heating_category: Optional[int] = Field(default=None)
+    main_heating_fraction: Optional[int] = Field(default=None)
+    main_heating_data_source: Optional[int] = Field(default=None)
+    condensing: Optional[bool] = Field(default=None)
+    weather_compensator: Optional[bool] = Field(default=None)
+
+    @classmethod
+    def from_domain(
+        cls, detail: MainHeatingDetail, epc_property_id: int
+    ) -> EpcMainHeatingDetailModel:
+        return cls(
+            epc_property_id=epc_property_id,
+            has_fghrs=detail.has_fghrs,
+            main_fuel_type=str(detail.main_fuel_type),
+            heat_emitter_type=str(detail.heat_emitter_type),
+            emitter_temperature=str(detail.emitter_temperature),
+            main_heating_control=str(detail.main_heating_control),
+            fan_flue_present=detail.fan_flue_present,
+            boiler_flue_type=detail.boiler_flue_type,
+            boiler_ignition_type=detail.boiler_ignition_type,
+            central_heating_pump_age=detail.central_heating_pump_age,
+            central_heating_pump_age_str=detail.central_heating_pump_age_str,
+            main_heating_index_number=detail.main_heating_index_number,
+            sap_main_heating_code=detail.sap_main_heating_code,
+            main_heating_number=detail.main_heating_number,
+            main_heating_category=detail.main_heating_category,
+            main_heating_fraction=detail.main_heating_fraction,
+            main_heating_data_source=detail.main_heating_data_source,
+            condensing=detail.condensing,
+            weather_compensator=detail.weather_compensator,
+        )
+
+
+class EpcBuildingPartModel(SQLModel, table=True):
+    __tablename__ = "epc_building_part"
+
+    id: Optional[int] = Field(default=None, primary_key=True)
+    epc_property_id: int = Field(foreign_key="epc_property.id", nullable=False)
+
+    identifier: str
+    construction_age_band: str
+    wall_construction: str
+    wall_insulation_type: str
+    wall_thickness_measured: bool
+    party_wall_construction: str
+    building_part_number: Optional[int] = Field(default=None)
+    wall_dry_lined: Optional[bool] = Field(default=None)
+    wall_thickness_mm: Optional[int] = Field(default=None)
+    wall_insulation_thickness: Optional[str] = Field(default=None)
+    floor_heat_loss: Optional[int] = Field(default=None)
+    floor_insulation_thickness: Optional[str] = Field(default=None)
+    flat_roof_insulation_thickness: Optional[str] = Field(default=None)
+    floor_type: Optional[str] = Field(default=None)
+    floor_construction_type: Optional[str] = Field(default=None)
+    floor_insulation_type_str: Optional[str] = Field(default=None)
+    floor_u_value_known: Optional[bool] = Field(default=None)
+    roof_construction: Optional[int] = Field(default=None)
+    roof_insulation_location: Optional[str] = Field(default=None)
+    roof_insulation_thickness: Optional[str] = Field(default=None)
+    room_in_roof_floor_area: Optional[float] = Field(default=None)
+    room_in_roof_construction_age_band: Optional[str] = Field(default=None)
+    alt_wall_1_area: Optional[float] = Field(default=None)
+    alt_wall_1_dry_lined: Optional[str] = Field(default=None)
+    alt_wall_1_construction: Optional[int] = Field(default=None)
+    alt_wall_1_insulation_type: Optional[int] = Field(default=None)
+    alt_wall_1_thickness_measured: Optional[str] = Field(default=None)
+    alt_wall_1_insulation_thickness: Optional[str] = Field(default=None)
+    alt_wall_2_area: Optional[float] = Field(default=None)
+    alt_wall_2_dry_lined: Optional[str] = Field(default=None)
+    alt_wall_2_construction: Optional[int] = Field(default=None)
+    alt_wall_2_insulation_type: Optional[int] = Field(default=None)
+    alt_wall_2_thickness_measured: Optional[str] = Field(default=None)
+    alt_wall_2_insulation_thickness: Optional[str] = Field(default=None)
+
+    @classmethod
+    def from_domain(
+        cls, part: SapBuildingPart, epc_property_id: int
+    ) -> EpcBuildingPartModel:
+        rir = part.sap_room_in_roof
+        aw1 = part.sap_alternative_wall_1
+        aw2 = part.sap_alternative_wall_2
+        return cls(
+            epc_property_id=epc_property_id,
+            identifier=part.identifier,
+            construction_age_band=part.construction_age_band,
+            wall_construction=str(part.wall_construction),
+            wall_insulation_type=str(part.wall_insulation_type),
+            wall_thickness_measured=part.wall_thickness_measured,
+            party_wall_construction=str(part.party_wall_construction),
+            building_part_number=part.building_part_number,
+            wall_dry_lined=part.wall_dry_lined,
+            wall_thickness_mm=part.wall_thickness_mm,
+            wall_insulation_thickness=part.wall_insulation_thickness,
+            floor_heat_loss=part.floor_heat_loss,
+            floor_insulation_thickness=part.floor_insulation_thickness,
+            flat_roof_insulation_thickness=(
+                str(part.flat_roof_insulation_thickness)
+                if part.flat_roof_insulation_thickness is not None
+                else None
+            ),
+            floor_type=part.floor_type,
+            floor_construction_type=part.floor_construction_type,
+            floor_insulation_type_str=part.floor_insulation_type_str,
+            floor_u_value_known=part.floor_u_value_known,
+            roof_construction=part.roof_construction,
+            roof_insulation_location=(
+                str(part.roof_insulation_location)
+                if part.roof_insulation_location is not None
+                else None
+            ),
+            roof_insulation_thickness=(
+                str(part.roof_insulation_thickness)
+                if part.roof_insulation_thickness is not None
+                else None
+            ),
+            room_in_roof_floor_area=float(rir.floor_area) if rir else None,
+            room_in_roof_construction_age_band=(
+                rir.construction_age_band if rir else None
+            ),
+            alt_wall_1_area=aw1.wall_area if aw1 else None,
+            alt_wall_1_dry_lined=aw1.wall_dry_lined if aw1 else None,
+            alt_wall_1_construction=aw1.wall_construction if aw1 else None,
+            alt_wall_1_insulation_type=aw1.wall_insulation_type if aw1 else None,
+            alt_wall_1_thickness_measured=aw1.wall_thickness_measured if aw1 else None,
+            alt_wall_1_insulation_thickness=(
+                aw1.wall_insulation_thickness if aw1 else None
+            ),
+            alt_wall_2_area=aw2.wall_area if aw2 else None,
+            alt_wall_2_dry_lined=aw2.wall_dry_lined if aw2 else None,
+            alt_wall_2_construction=aw2.wall_construction if aw2 else None,
+            alt_wall_2_insulation_type=aw2.wall_insulation_type if aw2 else None,
+            alt_wall_2_thickness_measured=aw2.wall_thickness_measured if aw2 else None,
+            alt_wall_2_insulation_thickness=(
+                aw2.wall_insulation_thickness if aw2 else None
+            ),
+        )
+
+
+class EpcFloorDimensionModel(SQLModel, table=True):
+    __tablename__ = "epc_floor_dimension"
+
+    id: Optional[int] = Field(default=None, primary_key=True)
+    epc_building_part_id: int = Field(
+        foreign_key="epc_building_part.id", nullable=False
+    )
+
+    floor: Optional[int] = Field(default=None)
+    room_height_m: float
+    total_floor_area_m2: float
+    party_wall_length_m: float
+    heat_loss_perimeter_m: float
+    floor_insulation: Optional[int] = Field(default=None)
+    floor_construction: Optional[int] = Field(default=None)
+
+    @classmethod
+    def from_domain(
+        cls, dim: SapFloorDimension, epc_building_part_id: int
+    ) -> EpcFloorDimensionModel:
+        return cls(
+            epc_building_part_id=epc_building_part_id,
+            floor=dim.floor,
+            room_height_m=dim.room_height_m,
+            total_floor_area_m2=dim.total_floor_area_m2,
+            party_wall_length_m=dim.party_wall_length_m,
+            heat_loss_perimeter_m=dim.heat_loss_perimeter_m,
+            floor_insulation=dim.floor_insulation,
+            floor_construction=dim.floor_construction,
+        )
+
+
+class EpcWindowModel(SQLModel, table=True):
+    __tablename__ = "epc_window"
+
+    id: Optional[int] = Field(default=None, primary_key=True)
+    epc_property_id: int = Field(foreign_key="epc_property.id", nullable=False)
+
+    pvc_frame: str
+    glazing_gap: str
+    orientation: str
+    window_type: str
+    glazing_type: str
+    window_width: float
+    window_height: float
+    draught_proofed: bool
+    window_location: str
+    window_wall_type: str
+    permanent_shutters_present: bool
+    frame_factor: Optional[float] = Field(default=None)
+    permanent_shutters_insulated: Optional[str] = Field(default=None)
+    transmission_u_value: Optional[float] = Field(default=None)
+    transmission_data_source: Optional[int] = Field(default=None)
+    transmission_solar_transmittance: Optional[float] = Field(default=None)
+
+    @classmethod
+    def from_domain(cls, window: SapWindow, epc_property_id: int) -> EpcWindowModel:
+        td = window.window_transmission_details
+        return cls(
+            epc_property_id=epc_property_id,
+            pvc_frame=str(window.pvc_frame),
+            glazing_gap=str(window.glazing_gap),
+            orientation=str(window.orientation),
+            window_type=str(window.window_type),
+            glazing_type=str(window.glazing_type),
+            window_width=window.window_width,
+            window_height=window.window_height,
+            draught_proofed=bool(window.draught_proofed),
+            window_location=str(window.window_location),
+            window_wall_type=str(window.window_wall_type),
+            permanent_shutters_present=bool(window.permanent_shutters_present),
+            frame_factor=window.frame_factor,
+            permanent_shutters_insulated=window.permanent_shutters_insulated,
+            transmission_u_value=td.u_value if td else None,
+            transmission_data_source=td.data_source if td else None,
+            transmission_solar_transmittance=td.solar_transmittance if td else None,
+        )
+
+
+class EpcEnergyElementModel(SQLModel, table=True):
+    __tablename__ = "epc_energy_element"
+
+    id: Optional[int] = Field(default=None, primary_key=True)
+    epc_property_id: int = Field(foreign_key="epc_property.id", nullable=False)
+
+    element_type: str  # roof | wall | floor | main_heating | window | lighting | hot_water | secondary_heating | main_heating_controls
+    description: str
+    energy_efficiency_rating: int
+    environmental_efficiency_rating: int
+
+    @classmethod
+    def from_domain(
+        cls, element: EnergyElement, element_type: str, epc_property_id: int
+    ) -> EpcEnergyElementModel:
+        return cls(
+            epc_property_id=epc_property_id,
+            element_type=element_type,
+            description=element.description,
+            energy_efficiency_rating=element.energy_efficiency_rating,
+            environmental_efficiency_rating=element.environmental_efficiency_rating,
+        )

backend/app/db/models/portfolio.py

@@ -113,8 +113,8 @@ class PropertyModel(Base):
         Enum(PortfolioStatus, values_callable=lambda x: [e.value for e in x]),
         nullable=False,
     )
-    address = Column(Text)
-    postcode = Column(Text)
+    address = Column(Text, nullable=False)
+    postcode = Column(Text, nullable=False)
     has_pre_condition_report = Column(Boolean)
     has_recommendations = Column(Boolean)
     created_at = Column(

backend/documents_parser/extractor.py

@@ -0,0 +1,692 @@
+from datetime import datetime
+from typing import List, Optional
+
+from datatypes.epc.surveys.pashub_rdsap_site_notes import (
+    BuildingConstruction,
+    InspectionMetadata,
+    BuildingMeasurements,
+    Conservatories,
+    CustomerResponse,
+    ExtensionConstruction,
+    ExtensionMeasurements,
+    ExtensionRoofSpace,
+    FloorConstruction,
+    FloorMeasurement,
+    General,
+    HeatingAndHotWater,
+    MainBuildingConstruction,
+    MainBuildingMeasurements,
+    MainHeating,
+    PasHubRdSapSiteNotes,
+    Renewables,
+    RoomCountElements,
+    RoofSpace,
+    RoofSpaceDetail,
+    SecondaryHeating,
+    Shower,
+    SurveyAddendum,
+    Ventilation,
+    WaterHeating,
+    WaterUse,
+    Window,
+)
+
+
+class PasHubRdSapSiteNotesExtractor:
+    def __init__(self, text_list: list[str]) -> None:
+        self.text_list = text_list
+
+    # --- generic helpers ---
+
+    def _get_in_doc(self, key: str, offset: int = 1) -> Optional[str]:
+        return self._get_in(self.text_list, key, offset)
+
+    def _bool(self, key: str, offset: int = 1) -> bool:
+        val = self._get_in_doc(key, offset)
+        return val is not None and val.lower() == "yes"
+
+    def _get_in(
+        self, list_to_process: List[str], key: str, offset: int = 1
+    ) -> Optional[str]:
+        try:
+            idx = list_to_process.index(key)
+            return list_to_process[idx + offset].strip() or None
+        except (ValueError, IndexError):
+            return None
+
+    def _bool_in(self, list_to_process: List[str], key: str, offset: int = 1) -> bool:
+        val = self._get_in(list_to_process, key, offset)
+        return val is not None and val.lower() == "yes"
+
+    def _optional_bool_in(self, list_to_process: List[str], key: str) -> Optional[bool]:
+        val = self._get_in(list_to_process, key)
+        return None if val is None else val.lower() == "yes"
+
+    def _is_known_in(self, list_to_process: List[str], key: str) -> bool:
+        val = self._get_in(list_to_process, key)
+        return val is not None and val.lower() != "not known"
+
+    def _wall_thickness_in(self, list_to_process: List[str]) -> int:
+        val = self._get_in(list_to_process, "Wall thickness:")
+        return int(val.split()[0]) if val else 0
+
+    def _section(self, start: str, end: str) -> List[str]:
+        try:
+            start_idx = self.text_list.index(start)
+            end_idx = self.text_list.index(end, start_idx)
+            return self.text_list[start_idx:end_idx]
+        except ValueError:
+            return []
+
+    # --- public extract methods ---
+
+    def extract_inspection_metadata(self) -> InspectionMetadata:
+        try:
+            addr_start = self.text_list.index("Property Address:") + 1
+            addr_end = self.text_list.index("Property Photo", addr_start)
+            property_address = ", ".join(
+                t.rstrip(",") for t in self.text_list[addr_start:addr_end]
+            )
+        except ValueError:
+            property_address = ""
+
+        created_on_raw = self._get_in_doc("Created On:")
+        created_on = (
+            datetime.strptime(created_on_raw, "%d %B %Y").strftime("%Y-%m-%d")
+            if created_on_raw
+            else ""
+        )
+
+        date_of_inspection_raw = self._get_in_doc("Date of Inspection:")
+        if not date_of_inspection_raw:
+            raise ValueError("Date of Inspection not found in document")
+        date_of_inspection = datetime.strptime(
+            date_of_inspection_raw, "%d %B %Y"
+        ).date()
+
+        return InspectionMetadata(
+            inspection_surveyor=self._get_in_doc("Inspection Surveyor:") or "",
+            email_address=self._get_in_doc("E-Mail Address:") or "",
+            report_reference=self._get_in_doc("Report Reference:") or "",
+            created_on=created_on,
+            date_of_inspection=date_of_inspection,
+            property_address=property_address,
+            property_photo="Property Photo" in self.text_list,
+        )
+
+    def extract(self) -> PasHubRdSapSiteNotes:
+        return PasHubRdSapSiteNotes(
+            inspection_metadata=self.extract_inspection_metadata(),
+            general=self.extract_general(),
+            building_construction=self.extract_building_construction(),
+            building_measurements=self.extract_building_measurements(),
+            roof_space=self.extract_roof_space(),
+            windows=self.extract_windows(),
+            heating_and_hot_water=self.extract_heating_and_hot_water(),
+            ventilation=self.extract_ventilation(),
+            conservatories=self.extract_conservatories(),
+            renewables=self.extract_renewables(),
+            room_count_elements=self.extract_room_count_elements(),
+            water_use=self.extract_water_use(),
+            customer_response=self.extract_customer_response(),
+            addendum=self.extract_addendum(),
+        )
+
+    def extract_general(self) -> General:
+        inspection_date_raw = self._get_in_doc("Inspection Date:")
+        if not inspection_date_raw:
+            raise ValueError("Inspection Date not found in document")
+        inspection_date = datetime.strptime(inspection_date_raw, "%d/%m/%Y").date()
+
+        storeys_raw = self._get_in_doc("Number of storeys:") or "0"
+        extensions_raw = self._get_in_doc("Number of Extensions:") or "0"
+        _extensions_first = extensions_raw.split()[0]
+        extensions_count = int(_extensions_first) if _extensions_first.isdigit() else 0
+
+        return General(
+            epc_checked_before_assessment=self._bool(
+                "Confirm you have checked for the existence of an", offset=2
+            ),
+            epc_exists_at_point_of_assessment=self._bool(
+                "Does an EPC exist at the point of carrying out this", offset=2
+            ),
+            inspection_date=inspection_date,
+            transaction_type=self._get_in_doc("Transaction Type:") or "",
+            tenure=self._get_in_doc("Tenure:") or "",
+            property_type=self._get_in_doc("Type of Property:") or "",
+            detachment_type=self._get_in_doc("Detachment Type:") or "",
+            number_of_storeys=int(storeys_raw.split()[0]),
+            terrain_type=self._get_in_doc("Terrain Type:") or "",
+            number_of_extensions=extensions_count,
+            electricity_smart_meter=self._bool(
+                "Is an electricity smart meter present?"
+            ),
+            electric_meter_type=self._get_in_doc("Electric meter type:") or "",
+            dwelling_export_capable=self._bool("Is the dwelling export-capable?"),
+            mains_gas_available=self._bool("Is mains gas available?"),
+            gas_smart_meter=self._bool("Is there a gas smart meter?"),
+            gas_meter_accessible=self._bool("Is the gas meter accessible?"),
+            measurements_location=self._get_in_doc("Select Measurements Location:")
+            or "",
+        )
+
+    def extract_building_construction(self) -> BuildingConstruction:
+        bc_section = self._section("Building Construction", "Building Measurements")
+
+        # Find extension markers within this section
+        extension_markers = []
+        i = 1
+        while f"Extension {i}" in bc_section:
+            extension_markers.append(f"Extension {i}")
+            i += 1
+
+        # Slice main building data: from "Main Building" to first extension or end
+        main_start = bc_section.index("Main Building")
+        main_end = (
+            bc_section.index(extension_markers[0])
+            if extension_markers
+            else len(bc_section)
+        )
+        main_data = bc_section[main_start:main_end]
+
+        # Slice each extension's data
+        extensions = []
+        for n, marker in enumerate(extension_markers):
+            ext_start = bc_section.index(marker)
+            ext_end = (
+                bc_section.index(extension_markers[n + 1])
+                if n + 1 < len(extension_markers)
+                else len(bc_section)
+            )
+            ext_data = bc_section[ext_start:ext_end]
+            extensions.append(self._parse_extension_construction(n + 1, ext_data))
+
+        return BuildingConstruction(
+            main_building=self._parse_main_building_construction(main_data),
+            floor=self._parse_floor_construction(main_data),
+            extensions=extensions if extensions else None,
+        )
+
+    # --- private parsing helpers ---
+
+    def _parse_main_building_construction(
+        self, data: List[str]
+    ) -> MainBuildingConstruction:
+        return MainBuildingConstruction(
+            age_range=self._get_in(data, "Age Range:") or "",
+            age_indicators=self._get_in(data, "Record indicators of property age:")
+            or "",
+            walls_construction_type=self._get_in(data, "Walls - Construction Type:")
+            or "",
+            cavity_construction_indicators=self._get_in(
+                data, "Record external indicators of Cavity Construction:"
+            )
+            or "",
+            walls_insulation_type=self._get_in(data, "Walls - Insulation Type:") or "",
+            filled_cavity_indicators=self._get_in(
+                data, "Record indicators of filled cavity:"
+            ),
+            thermal_conductivity_of_wall_insulation=self._get_in(
+                data, "Thermal conductivity of wall insulation:"
+            )
+            or "",
+            wall_u_value_known=self._is_known_in(data, "Wall U-Value known?"),
+            wall_thickness_mm=self._wall_thickness_in(data),
+            party_wall_construction_type=self._get_in(
+                data, "Party wall construction type:"
+            )
+            or "",
+        )
+
+    def _parse_extension_construction(
+        self, ext_id: int, data: List[str]
+    ) -> ExtensionConstruction:
+        return ExtensionConstruction(
+            id=ext_id,
+            age_range=self._get_in(data, "Age Range:") or "",
+            age_indicators=self._get_in(data, "Record indicators of property age:")
+            or "",
+            walls_construction_type=self._get_in(data, "Walls - Construction Type:")
+            or "",
+            cavity_construction_indicators=self._get_in(
+                data, "Record external indicators of Cavity Construction:"
+            )
+            or "",
+            walls_insulation_type=self._get_in(data, "Walls - Insulation Type:") or "",
+            filled_cavity_indicators=self._get_in(
+                data, "Record indicators of filled cavity:"
+            ),
+            thermal_conductivity_of_wall_insulation=self._get_in(
+                data, "Thermal conductivity of wall insulation:"
+            )
+            or "",
+            wall_u_value_known=self._is_known_in(data, "Wall U-Value known?"),
+            wall_thickness_mm=self._wall_thickness_in(data),
+            party_wall_construction_type=self._get_in(
+                data, "Party wall construction type:"
+            )
+            or "",
+        )
+
+    def extract_building_measurements(self) -> BuildingMeasurements:
+        bm_section = self._section("Building Measurements", "Roof Space")
+
+        extension_markers = []
+        i = 1
+        while f"Extension {i}" in bm_section:
+            extension_markers.append(f"Extension {i}")
+            i += 1
+
+        main_start = bm_section.index("Main Building")
+        main_end = (
+            bm_section.index(extension_markers[0])
+            if extension_markers
+            else len(bm_section)
+        )
+        main_floors = self._parse_floor_measurements(bm_section[main_start:main_end])
+
+        extensions = []
+        for n, marker in enumerate(extension_markers):
+            ext_start = bm_section.index(marker)
+            ext_end = (
+                bm_section.index(extension_markers[n + 1])
+                if n + 1 < len(extension_markers)
+                else len(bm_section)
+            )
+            extensions.append(
+                ExtensionMeasurements(
+                    id=n + 1,
+                    floors=self._parse_floor_measurements(
+                        bm_section[ext_start:ext_end]
+                    ),
+                )
+            )
+
+        return BuildingMeasurements(
+            main_building=MainBuildingMeasurements(floors=main_floors),
+            extensions=extensions if extensions else None,
+        )
+
+    def extract_roof_space(self) -> RoofSpace:
+        rs_section = self._section("Roof Space", "Windows")
+
+        extension_markers = []
+        i = 1
+        while f"Extension {i}" in rs_section:
+            extension_markers.append(f"Extension {i}")
+            i += 1
+
+        main_start = rs_section.index("Main Building")
+        main_end = (
+            rs_section.index(extension_markers[0])
+            if extension_markers
+            else len(rs_section)
+        )
+        main_data = rs_section[main_start:main_end]
+
+        extensions = []
+        for n, marker in enumerate(extension_markers):
+            ext_start = rs_section.index(marker)
+            ext_end = (
+                rs_section.index(extension_markers[n + 1])
+                if n + 1 < len(extension_markers)
+                else len(rs_section)
+            )
+            ext_data = rs_section[ext_start:ext_end]
+            extensions.append(self._parse_extension_roof_space(n + 1, ext_data))
+
+        return RoofSpace(
+            main_building=self._parse_roof_space_detail(main_data),
+            extensions=extensions if extensions else None,
+        )
+
+    def extract_windows(self) -> List[Window]:
+        w_section = self._section("Windows", "Heating & Hot Water")
+
+        windows = []
+        n = 1
+        while f"Window {n}" in w_section:
+            start = w_section.index(f"Window {n}")
+            end = (
+                w_section.index(f"Window {n + 1}")
+                if f"Window {n + 1}" in w_section
+                else len(w_section)
+            )
+            windows.append(self._parse_window(n, w_section[start:end]))
+            n += 1
+
+        return windows
+
+    def extract_heating_and_hot_water(self) -> HeatingAndHotWater:
+        hhw_section = self._section("Heating & Hot Water", "Ventilation")
+        return HeatingAndHotWater(
+            main_heating=self._parse_main_heating(hhw_section),
+            secondary_heating=self._parse_secondary_heating(hhw_section),
+            water_heating=self._parse_water_heating(hhw_section),
+        )
+
+    def extract_ventilation(self) -> Ventilation:
+        v_section = self._section("Ventilation", "Conservatories")
+        return Ventilation(
+            ventilation_type=self._get_in(v_section, "Ventilation type:") or "",
+            has_fixed_air_conditioning=self._bool_in(
+                v_section, "Has fixed air conditioning?"
+            ),
+            number_of_open_flues=int(
+                self._get_in(v_section, "Number of open flues:") or 0
+            ),
+            number_of_closed_flues=int(
+                self._get_in(v_section, "Number of closed flues:") or 0
+            ),
+            number_of_boiler_flues=int(
+                self._get_in(v_section, "Number of boiler flues:") or 0
+            ),
+            number_of_other_flues=int(
+                self._get_in(v_section, "Number of other flues:") or 0
+            ),
+            number_of_extract_fans=int(
+                self._get_in(v_section, "Number of extract fans:") or 0
+            ),
+            number_of_passive_vents=int(
+                self._get_in(v_section, "Number of passive vents:") or 0
+            ),
+            number_of_flueless_gas_fires=int(
+                self._get_in(v_section, "Number of flueless gas fires:") or 0
+            ),
+            pressure_test=self._get_in(v_section, "Pressure test:") or "",
+            draught_lobby=self._bool_in(v_section, "Is there a draught lobby?"),
+            ventilation_in_pcdf_database=self._optional_bool_in(
+                v_section, "Is the ventilation in the PCDF database?"
+            ),
+        )
+
+    def extract_conservatories(self) -> Conservatories:
+        c_section = self._section("Conservatories", "Renewables")
+        val = self._get_in(c_section, "Is there conservatory?")
+        return Conservatories(
+            has_conservatory=val is not None and val.lower() != "no conservatory"
+        )
+
+    def extract_renewables(self) -> Renewables:
+        r_section = self._section("Renewables", "Room Count Elements")
+        batteries_raw = self._get_in(r_section, "Number of PV batteries:")
+        batteries = (
+            0
+            if batteries_raw is None or batteries_raw.lower() == "none"
+            else int(batteries_raw)
+        )
+        pv_connection = self._get_in(r_section, "PV Connection:")
+        percent_raw = self._get_in(r_section, "Percentage of roof covered with photovoltaic array?")
+        percent_roof = int(percent_raw.split()[0]) if percent_raw else None
+        return Renewables(
+            wind_turbines=self._bool_in(r_section, "Has wind turbines?"),
+            solar_hot_water=self._bool_in(r_section, "Has solar hot water?"),
+            photovoltaic_array=self._bool_in(r_section, "Has photovoltaic array?"),
+            number_of_pv_batteries=batteries,
+            hydro=self._bool_in(r_section, "Is the dwelling connected to Hydro?"),
+            pv_connection=pv_connection,
+            percent_roof_covered_pv=percent_roof,
+        )
+
+    def extract_room_count_elements(self) -> RoomCountElements:
+        rce_section = self._section("Room Count Elements", "Customer Response")
+        heated_rooms_raw = self._get_in(rce_section, "Number of heated rooms?")
+        return RoomCountElements(
+            number_of_habitable_rooms=int(
+                self._get_in(rce_section, "Number of habitable rooms?") or 0
+            ),
+            any_unheated_rooms=self._bool_in(
+                rce_section, "Are any of these rooms unheated?"
+            ),
+            number_of_heated_rooms=int(heated_rooms_raw) if heated_rooms_raw else None,
+            number_of_external_doors=int(
+                self._get_in(rce_section, "Number of external doors?") or 0
+            ),
+            number_of_insulated_external_doors=int(
+                self._get_in(rce_section, "Number of insulated external doors?") or 0
+            ),
+            number_of_draughtproofed_external_doors=int(
+                self._get_in(rce_section, "Number of draughtproofed external doors?")
+                or 0
+            ),
+            number_of_open_chimneys=int(
+                self._get_in(rce_section, "Number of open chimneys?") or 0
+            ),
+            number_of_blocked_chimneys=int(
+                self._get_in(rce_section, "Number of blocked chimneys?") or 0
+            ),
+            number_of_fixed_incandescent_bulbs=int(
+                self._get_in(rce_section, "Number of fixed incandescent bulbs:") or 0
+            ),
+            exact_led_cfl_count_known=self._bool_in(
+                rce_section, "Is the exact number of LED and CFL bulbs known?"
+            ),
+            number_of_fixed_led_bulbs=int(
+                self._get_in(rce_section, "Number of fixed LED bulbs:") or 0
+            ),
+            number_of_fixed_cfl_bulbs=int(
+                self._get_in(rce_section, "Number of fixed CFL bulbs:") or 0
+            ),
+            waste_water_heat_recovery=self._get_in(
+                rce_section, "Are there any waste water heat recovery systems?"
+            )
+            or "",
+        )
+
+    def extract_water_use(self) -> WaterUse:
+        wu_section = self._section("Room Count Elements", "Customer Response")
+        baths_raw = self._get_in(wu_section, "Number of baths:") or "0"
+        special_raw = (
+            self._get_in(
+                wu_section, "How many special features are there at the", offset=2
+            )
+            or "0"
+        )
+
+        showers = []
+        n = 1
+        while f"Shower {n}" in wu_section:
+            start = wu_section.index(f"Shower {n}")
+            end = (
+                wu_section.index(f"Shower {n + 1}")
+                if f"Shower {n + 1}" in wu_section
+                else len(wu_section)
+            )
+            shower_data = wu_section[start:end]
+            showers.append(
+                Shower(
+                    id=n,
+                    outlet_type=self._get_in(shower_data, "Shower outlet type:") or "",
+                )
+            )
+            n += 1
+
+        return WaterUse(
+            number_of_baths=int(baths_raw),
+            number_of_special_features=int(special_raw),
+            showers=showers,
+        )
+
+    def extract_customer_response(self) -> CustomerResponse:
+        cr_section = self._section(
+            "Customer Response", "Addendum + Related Party Disclosure"
+        )
+        return CustomerResponse(
+            customer_present=self._bool_in(cr_section, "Customer present?"),
+            willing_to_answer_satisfaction_survey=self._bool_in(
+                cr_section, "Customer willing to answer satisfaction survey?"
+            ),
+        )
+
+    def extract_addendum(self) -> SurveyAddendum:
+        a_section = self._section(
+            "Addendum + Related Party Disclosure", "Photographs Required"
+        )
+        return SurveyAddendum(
+            addendum=self._get_in(a_section, "Addendum") or "",
+            related_party_disclosure=self._get_in(a_section, "Related party disclosure")
+            or "",
+            hard_to_treat_cavity_access_issues=self._bool_in(
+                a_section,
+                "Hard to treat cavity walls: Property has access",
+                offset=2,
+            ),
+            hard_to_treat_cavity_high_exposure=self._bool_in(
+                a_section,
+                "Hard to treat cavity walls: Property has high",
+                offset=2,
+            ),
+            hard_to_treat_cavity_narrow_cavities=self._bool_in(
+                a_section,
+                "Hard to treat cavity walls: Property has narrow",
+                offset=2,
+            ),
+        )
+
+    def _parse_main_heating(self, data: List[str]) -> MainHeating:
+        return MainHeating(
+            selection_method=self._get_in(
+                data, "How would you like to select the Heating System?"
+            )
+            or "",
+            system_type=self._get_in(data, "System type:") or "",
+            product_id=int(self._get_in(data, "Product Id") or 0),
+            manufacturer=self._get_in(data, "Manufacturer") or "",
+            model=self._get_in(data, "Model") or "",
+            orig_manufacturer=self._get_in(data, "Orig Manuf") or "",
+            fuel=self._get_in(data, "Fuel") or "",
+            summer_efficiency=float(self._get_in(data, "S. Efficiency") or 0),
+            type=self._get_in(data, "Type") or "",
+            condensing=self._bool_in(data, "Condensing"),
+            year=self._get_in(data, "Year") or "",
+            mount=self._get_in(data, "Mount") or "",
+            open_flue=self._get_in(data, "Open Flue") or "",
+            fan_assist=self._bool_in(data, "Fan Assist"),
+            status=self._get_in(data, "Status") or "",
+            central_heating_pump_age=self._get_in(data, "Central heating pump age:")
+            or "",
+            controls=self._get_in(data, "Controls:") or "",
+            flue_gas_heat_recovery_system=self._bool_in(
+                data, "Does the boiler have a Flue Gas Heat Recover", offset=2
+            ),
+            weather_compensator=self._bool_in(data, "Is there a weather compensator?"),
+            emitter=self._get_in(data, "Emitter:") or "",
+            emitter_temperature=self._get_in(data, "Emitter Temperature:") or "",
+        )
+
+    def _parse_secondary_heating(self, data: List[str]) -> SecondaryHeating:
+        system_raw = self._get_in(data, "Secondary System:")
+        return SecondaryHeating(
+            secondary_fuel=self._get_in(data, "Secondary Fuel") or "",
+            secondary_system=system_raw if system_raw else None,
+        )
+
+    def _parse_water_heating(self, data: List[str]) -> WaterHeating:
+        thickness_raw = self._get_in(data, "Insulation Thickness (mm):") or self._get_in(data, "Thickness:")
+        thickness_mm = int(thickness_raw.split()[0]) if thickness_raw else None
+        return WaterHeating(
+            type=self._get_in(data, "Water Heating Type:") or "",
+            system=self._get_in(data, "Water Heating System:") or "",
+            cylinder_size=self._get_in(data, "Cylinder Size:") or "",
+            cylinder_measured_heat_loss=self._get_in(
+                data, "Cylinder Measured Heat Loss:"
+            ),
+            insulation_type=self._get_in(data, "Insulation Type:"),
+            insulation_thickness_mm=thickness_mm,
+            has_thermostat=self._optional_bool_in(data, "Cylinder Thermostat:") or self._optional_bool_in(data, "Has thermostat?"),
+            immersion_type=self._get_in(data, "Immersion:"),
+        )
+
+    def _parse_window(self, window_id: int, data: List[str]) -> Window:
+        height_raw = self._get_in(data, "Window height:")
+        width_raw = self._get_in(data, "Window width:")
+        return Window(
+            id=window_id,
+            location=self._get_in(data, "Window location:") or "",
+            wall_type=self._get_in(data, "Window wall type:") or "",
+            glazing_type=self._get_in(data, "Glazing Type:") or "",
+            window_type=self._get_in(data, "Window type:") or "",
+            frame_type=self._get_in(data, "Window frame type:") or "",
+            glazing_gap=self._get_in(data, "What size is the glazing gap?") or "",
+            draught_proofed=self._bool_in(data, "Is the window draught proofed?"),
+            permanent_shutters=self._bool_in(
+                data, "Are there permanent shutters present?"
+            ),
+            height_m=float(height_raw.split()[0]) if height_raw else 0.0,
+            width_m=float(width_raw.split()[0]) if width_raw else 0.0,
+            orientation=self._get_in(data, "Orientation:") or "",
+        )
+
+    def _parse_insulation_thickness(
+        self, val: Optional[str]
+    ) -> tuple[Optional[int], Optional[str]]:
+        if val is None:
+            return None, None
+        try:
+            return int(val.split()[0]), None
+        except (ValueError, IndexError):
+            return None, val
+
+    def _parse_roof_space_detail(self, data: List[str]) -> RoofSpaceDetail:
+        thickness_mm, thickness_str = self._parse_insulation_thickness(
+            self._get_in(data, "Roofs - Insulation Thickness:")
+        )
+        return RoofSpaceDetail(
+            construction_type=self._get_in(data, "Roofs - Construction Type:") or "",
+            insulation_at=self._get_in(data, "Roofs - Insulation At:") or "",
+            roof_u_value_known=self._is_known_in(data, "Roof U-Value:"),
+            cavity_wall_construction_indicators=self._get_in(
+                data, "Record indicators of Cavity Wall Construction in roof", offset=2
+            )
+            or "",
+            rooms_in_roof=self._bool_in(data, "Are there rooms in the roof?"),
+            insulation_thickness_mm=thickness_mm,
+            insulation_thickness=thickness_str,
+        )
+
+    def _parse_extension_roof_space(
+        self, ext_id: int, data: List[str]
+    ) -> ExtensionRoofSpace:
+        thickness_mm, thickness_str = self._parse_insulation_thickness(
+            self._get_in(data, "Roofs - Insulation Thickness:")
+        )
+        return ExtensionRoofSpace(
+            id=ext_id,
+            construction_type=self._get_in(data, "Roofs - Construction Type:") or "",
+            insulation_at=self._get_in(data, "Roofs - Insulation At:") or "",
+            roof_u_value_known=self._is_known_in(data, "Roof U-Value:"),
+            cavity_wall_construction_indicators=self._get_in(
+                data, "Record indicators of Cavity Wall Construction in roof", offset=2
+            )
+            or "",
+            rooms_in_roof=self._bool_in(data, "Are there rooms in the roof?"),
+            insulation_thickness_mm=thickness_mm,
+            insulation_thickness=thickness_str,
+        )
+
+    def _parse_floor_measurements(self, data: List[str]) -> List[FloorMeasurement]:
+        floors = []
+        i = 0
+        while i < len(data):
+            if data[i].startswith("Floor") and i + 4 < len(data):
+                floors.append(
+                    FloorMeasurement(
+                        name=data[i],
+                        area_m2=float(data[i + 1]),
+                        height_m=float(data[i + 2]),
+                        heat_loss_perimeter_m=float(data[i + 3]),
+                        pwl_m=float(data[i + 4]),
+                    )
+                )
+                i += 5
+            else:
+                i += 1
+        return floors
+
+    def _parse_floor_construction(self, data: List[str]) -> FloorConstruction:
+        return FloorConstruction(
+            floor_type=self._get_in(data, "Floor type:") or "",
+            floor_construction=self._get_in(data, "Floor Construction:") or "",
+            floor_insulation_type=self._get_in(data, "Floor Insulation Type:") or "",
+            floor_u_value_known=self._is_known_in(data, "Floor U-Value known?"),
+        )

backend/documents_parser/handler/handler.py

@@ -0,0 +1,39 @@
+import os
+from typing import Any, Mapping
+
+
+from utils.logger import setup_logger
+from utils.s3 import upload_file_to_s3
+
+logger = setup_logger()
+
+BUCKET = "retrofit-energy-assessments-dev"
+PDF_S3_KEY = "example/SiteNotesExample.pdf"
+PDF_LOCAL_PATH = os.path.join(
+    os.path.dirname(__file__),
+    "..",
+    "..",
+    "tests",
+    "test_data",
+    "SiteNotesExample.pdf",
+)
+
+
+def upload_pdf(local_path: str, bucket: str, key: str) -> None:
+    logger.info(f"Uploading {local_path} to s3://{bucket}/{key}")
+    upload_file_to_s3(local_path, bucket, key)
+    logger.info("Upload complete")
+
+
+def handler(event: Mapping[str, Any], context: Any) -> None:
+    logger.info("Entered handler")
+
+    output_path = os.path.join(os.path.dirname(__file__), "..", "textract_blocks.json")
+
+    upload_pdf(PDF_LOCAL_PATH, BUCKET, PDF_S3_KEY)
+
+    logger.info(f"Uploaded file to {output_path}")
+
+
+if __name__ == "__main__":
+    handler({}, None)

backend/documents_parser/pdf.py

@@ -0,0 +1,12 @@
+from typing import List
+
+import pymupdf
+
+
+def pdf_to_text_list(pdf_bytes: bytes) -> List[str]:
+    tokens: List[str] = []
+    with pymupdf.open(stream=pdf_bytes, filetype="pdf") as doc:
+        for page in doc:
+            for line in page.get_text().split("\n"):
+                tokens.append(line)
+    return tokens

backend/documents_parser/tests/fixtures/site_notes_example_2_text.json

@@ -0,0 +1,737 @@
+[
+  "SMART EPC: Record of",
+  "Inspection & Site Notes",
+  "Inspection Surveyor:",
+  "Rebecca Mcwilliam",
+  "E-Mail Address:",
+  "rebeccamcdea@gmail.com",
+  "Report Reference:",
+  "ADD992DF-BE9B-47EB-8121-4C6143C6A7C9",
+  "Created On:",
+  "15 October 2025",
+  "Date of Inspection:",
+  "13 October 2025",
+  "Property Address:",
+  "13,",
+  "Sandfield Avenue ,",
+  "Wrenbury,",
+  "CW5 8EU",
+  "Property Photo",
+  "Page 1",
+  "",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "RdSAP Assessment",
+  "General",
+  "Confirm you have checked for the existence of an",
+  "EPC before carrying out another energy assessment.",
+  "Yes",
+  "Does an EPC exist at the point of carrying out this",
+  "energy assessment?",
+  "Yes",
+  "Please select why another energy assessment needs",
+  "to be undertaken:",
+  "Assessor instructed to produce a new EPC upon request from building",
+  "owner/tenant/landlord after confirming to the requestor that a valid EPC",
+  "already exists",
+  "Inspection Date:",
+  "13/10/2025",
+  "Transaction Type:",
+  "None of the Above",
+  "Tenure:",
+  "Rented Social",
+  "Type of Property:",
+  "House",
+  "Detachment Type:",
+  "End-terrace",
+  "Number of storeys:",
+  "2 Storeys",
+  "Terrain Type:",
+  "Suburban",
+  "Number of Extensions:",
+  "No Extensions",
+  "Is an electricity smart meter present?",
+  "Yes",
+  "Electric meter type:",
+  "Single",
+  "Page 2",
+  "",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "External indicators of Cavity Wall Construction:",
+  "Is the dwelling export-capable?",
+  "Yes",
+  "Is mains gas available?",
+  "No",
+  "Select Measurements Location:",
+  "Internal",
+  "Building Construction",
+  "Main Building",
+  "Age Range:",
+  "1950-1966",
+  "Record indicators of property age:",
+  "local knowledge, enquiries of owner, period building features",
+  "Walls - Construction Type:",
+  "Cavity",
+  "Record external indicators of Cavity Construction:",
+  "stretcher bond, wall thickness over 270 mm",
+  "Walls - Insulation Type:",
+  "Filled Cavity",
+  "Record indicators of filled cavity:",
+  "evidence of cavity fill drill holes, Boroscope",
+  "Page 3",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 4",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 5",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 6",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 7",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 8",
+  "",
+  "Photo wall thickness:",
+  "Photo wall thickness:",
+  "Thermal conductivity of wall insulation:",
+  "Unknown",
+  "Wall U-Value known?",
+  "Not Known",
+  "Wall thickness:",
+  "300 mm",
+  "Party wall construction type:",
+  "Unable to determine",
+  "Floor type:",
+  "Ground Floor",
+  "Floor Construction:",
+  "Solid",
+  "Floor Insulation Type:",
+  "As Built",
+  "Floor U-Value known?",
+  "Not Known",
+  "Building Measurements",
+  "Area (m2)",
+  "Height (m)",
+  "Heat Loss Perimeter (m)",
+  "PWL (m)",
+  "Main Building",
+  "Floor 1",
+  "43.29",
+  "2.45",
+  "19.74",
+  "6.58",
+  "Floor 0",
+  "43.29",
+  "2.46",
+  "19.74",
+  "6.58",
+  "Roof Space",
+  "Main Building",
+  "Roofs - Construction Type:",
+  "Pitched roof (Slates or tiles), Access to loft",
+  "Roofs - Insulation At:",
+  "Joists",
+  "Roof U-Value:",
+  "Not Known",
+  "Roofs - Insulation Thickness:",
+  "150 mm",
+  "Page 9",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 10",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 11",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 12",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 13",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 14",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 15",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Record indicators of Cavity Wall Construction in roof",
+  "space:",
+  "No indicator of construction visible",
+  "Page 16",
+  "",
+  "Record indicators of party wall construction in roof space:",
+  "Are there rooms in the roof?",
+  "No",
+  "Windows",
+  "Window 1",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Manufacturing date stamp",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "1.27 m",
+  "Window width:",
+  "1.55 m",
+  "Orientation:",
+  "North West",
+  "Page 17",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window 2",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Manufacturing date stamp",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.97 m",
+  "Window width:",
+  "1.04 m",
+  "Orientation:",
+  "North West",
+  "Page 18",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window 3",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Manufacturing date stamp",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.97 m",
+  "Window width:",
+  "1.53 m",
+  "Orientation:",
+  "North West",
+  "Window 4",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Page 19",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "What size is the glazing gap?",
+  "12 mm",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.98 m",
+  "Window width:",
+  "0.57 m",
+  "Orientation:",
+  "North West",
+  "Window 5",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "1.29 m",
+  "Window width:",
+  "0.57 m",
+  "Orientation:",
+  "South West",
+  "Page 20",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window 6",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Manufacturing date stamp",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.98 m",
+  "Window width:",
+  "0.53 m",
+  "Orientation:",
+  "South West",
+  "Window 7",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.98 m",
+  "Window width:",
+  "1.09 m",
+  "Orientation:",
+  "South East",
+  "Page 21",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window 8",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Manufacturing date stamp",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.96 m",
+  "Window width:",
+  "1.54 m",
+  "Orientation:",
+  "South East",
+  "Page 22",
+  "",
+  "Photo of glazing type:",
+  "Window 9",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Manufacturing date stamp",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.97 m",
+  "Window width:",
+  "1.53 m",
+  "Orientation:",
+  "South East",
+  "Window 10",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "12 mm",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.95 m",
+  "Window width:",
+  "1.04 m",
+  "Orientation:",
+  "South East",
+  "Page 23",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window 11",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "12 mm",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.97 m",
+  "Window width:",
+  "1.53 m",
+  "Orientation:",
+  "South East",
+  "Page 24",
+  "",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Heating & Hot Water",
+  "Main Heating Systems",
+  "Main Heating 1",
+  "How would you like to select the Heating System?",
+  "Manual Entry",
+  "System type:",
+  "Room heaters",
+  "Fuel:",
+  "Dual fuel appliance (mineral and wood)",
+  "Heating System (Other):",
+  "Open fire with back boiler (No radiators)",
+  "Controls:",
+  "No thermostatic control of room temperature",
+  "Secondary Heating System",
+  "Secondary Fuel",
+  "Dual Fuel Appliance (Mineral and Wood)",
+  "Secondary System:",
+  "Open fire in grate",
+  "Page 25",
+  "",
+  "Photo of secondary heating system",
+  "Photo of water heating system:",
+  "Photo of water heating system:",
+  "Water Heating & Cylinder",
+  "Water Heating Type:",
+  "Regular",
+  "Water Heating System:",
+  "From main heating 1",
+  "Page 26",
+  "",
+  "Photo of water heating system:",
+  "Photo of water heating system:",
+  "Photo of water heating system:",
+  "Photo of water heating system:",
+  "Page 27",
+  "",
+  "Photo of water heating system:",
+  "Photo of cylinder and thermostat if present:",
+  "Photo of cylinder and thermostat if present:",
+  "Photo of ventilation type:",
+  "Cylinder Size:",
+  "Normal (90-130 litres)",
+  "What is the cylinder measured heat loss:",
+  "Not known",
+  "Insulation Type:",
+  "Factory fitted",
+  "Thickness:",
+  "38 mm",
+  "Has thermostat?",
+  "No",
+  "Ventilation",
+  "Ventilation type:",
+  "Mechanical Extract - Decentralised",
+  "Has fixed air conditioning?",
+  "No",
+  "Is the ventilation in the PCDF database?",
+  "No",
+  "Page 28",
+  "",
+  "Photo of boiler flues:",
+  "Photo of extract fans:",
+  "Number of open flues:",
+  "0",
+  "Number of closed flues:",
+  "0",
+  "Number of boiler flues:",
+  "1",
+  "Number of other flues:",
+  "0",
+  "Number of extract fans:",
+  "1",
+  "Number of passive vents:",
+  "0",
+  "Number of flueless gas fires:",
+  "0",
+  "Pressure test:",
+  "No test",
+  "Is there a draught lobby?",
+  "No",
+  "Conservatories",
+  "Is there conservatory?",
+  "No conservatory",
+  "Page 29",
+  "",
+  "Photo of open chimneys:",
+  "Renewables",
+  "Wind Turbines",
+  "Has wind turbines?",
+  "No",
+  "Solar hot water",
+  "Has solar hot water?",
+  "No",
+  "Photovoltaics",
+  "Has photovoltaic array?",
+  "No",
+  "Number of PV batteries:",
+  "None",
+  "Hydro",
+  "Is the dwelling connected to Hydro?",
+  "No",
+  "Room Count Elements",
+  "Number of habitable rooms?",
+  "5",
+  "Are any of these rooms unheated?",
+  "Yes",
+  "Please enter the number of HEATED rooms:",
+  "0",
+  "Number of external doors?",
+  "2",
+  "Number of insulated external doors?",
+  "0",
+  "Number of draughtproofed external doors?",
+  "2",
+  "Number of open chimneys?",
+  "1",
+  "Number of blocked chimneys?",
+  "1",
+  "Page 30",
+  "",
+  "Photo of blocked chimneys:",
+  "Photo of blocked chimneys:",
+  "Photo of blocked chimneys:",
+  "Photo of incandescent bulbs:",
+  "Number of fixed incandescent bulbs:",
+  "1",
+  "Is the exact number of LED and CFL bulbs known?",
+  "Yes",
+  "Number of fixed LED bulbs:",
+  "4",
+  "Page 31",
+  "",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Photo of CFL bulbs:",
+  "Photo of CFL bulbs:",
+  "Number of fixed CFL bulbs:",
+  "4",
+  "Page 32",
+  "",
+  "Photo of CFL bulbs:",
+  "Photo of CFL bulbs:",
+  "Photo of shower:",
+  "Photo of shower:",
+  "Are there any waste water heat recovery systems?",
+  "None",
+  "Number of baths:",
+  "1",
+  "How many special features are there at the",
+  "property?",
+  "0",
+  "Showers",
+  "Shower 1",
+  "Shower outlet type:",
+  "Electric Shower",
+  "Page 33",
+  "",
+  "Photo of shower:",
+  "Customer Response",
+  "Customer present?",
+  "Yes",
+  "Customer willing to answer satisfaction survey?",
+  "No",
+  "Addendum + Related Party Disclosure",
+  "Addendum",
+  "None",
+  "Related party disclosure",
+  "No related party",
+  "Hard to treat cavity walls: Property has access",
+  "issues?",
+  "No",
+  "Hard to treat cavity walls: Property has high",
+  "exposure?",
+  "No",
+  "Hard to treat cavity walls: Property has narrow",
+  "cavities?",
+  "No",
+  "Photographs Required",
+  "Page 34",
+  "",
+  "General Photos:",
+  "General Photos:",
+  "General Photos:",
+  "General Photos:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 35",
+  "",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 36",
+  "",
+  "External Elevations:",
+  "Page 37",
+  "",
+  "Page 38",
+  "",
+  "Page 39",
+  "",
+  "Page 40",
+  "",
+  "Additional Notes",
+  "Additional Notes",
+  "Over the years Tenant has refused any sort of heating system upgrade so the",
+  "property only has an open fire with back boiler in the living room. There are no",
+  "radiators in habitable rooms. The only radiator is in the bathroom. The hot",
+  "water cylinder is connected to the back boiler and there is also a backup",
+  "emersion.",
+  "Page 41",
+  ""
+]

backend/documents_parser/tests/fixtures/site_notes_example_3_text.json

@@ -0,0 +1,799 @@
+[
+  "SMART EPC: Record of",
+  "Inspection & Site Notes",
+  "Inspection Surveyor:",
+  "Rebecca Mcwilliam",
+  "E-Mail Address:",
+  "rebeccamcdea@gmail.com",
+  "Report Reference:",
+  "Not Applicable",
+  "Created On:",
+  "14 October 2025",
+  "Date of Inspection:",
+  "14 October 2025",
+  "Property Address:",
+  "19,",
+  "Sandfield ,",
+  "Wrenbury,",
+  "CW5 8EU",
+  "Property Photo",
+  "Page 1",
+  "",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "RdSAP Assessment",
+  "General",
+  "Confirm you have checked for the existence of an",
+  "EPC before carrying out another energy assessment.",
+  "Yes",
+  "Does an EPC exist at the point of carrying out this",
+  "energy assessment?",
+  "Yes",
+  "Please select why another energy assessment needs",
+  "to be undertaken:",
+  "Assessor instructed to produce a new EPC upon request from building",
+  "owner/tenant/landlord after confirming to the requestor that a valid EPC",
+  "already exists",
+  "Inspection Date:",
+  "14/10/2025",
+  "Transaction Type:",
+  "None of the Above",
+  "Tenure:",
+  "Rented Social",
+  "Type of Property:",
+  "House",
+  "Detachment Type:",
+  "End-terrace",
+  "Number of storeys:",
+  "2 Storeys",
+  "Terrain Type:",
+  "Suburban",
+  "Number of Extensions:",
+  "No Extensions",
+  "Is an electricity smart meter present?",
+  "Yes",
+  "Electric meter type:",
+  "Dual",
+  "Is the dwelling export-capable?",
+  "Yes",
+  "Is mains gas available?",
+  "No",
+  "Page 2",
+  "",
+  "External indicators of Cavity Wall Construction:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Select Measurements Location:",
+  "Internal",
+  "Building Construction",
+  "Main Building",
+  "Age Range:",
+  "1950-1966",
+  "Record indicators of property age:",
+  "local knowledge, enquiries of owner, period building features",
+  "Walls - Construction Type:",
+  "Cavity",
+  "Record external indicators of Cavity Construction:",
+  "stretcher bond",
+  "Walls - Insulation Type:",
+  "Filled Cavity",
+  "Record indicators of filled cavity:",
+  "evidence of cavity fill drill holes",
+  "Page 3",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 4",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 5",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 6",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 7",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 8",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 9",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo wall thickness:",
+  "Thermal conductivity of wall insulation:",
+  "Unknown",
+  "Wall U-Value known?",
+  "Not Known",
+  "Wall thickness:",
+  "300 mm",
+  "Party wall construction type:",
+  "Unable to determine",
+  "Floor type:",
+  "Ground Floor",
+  "Floor Construction:",
+  "Solid",
+  "Floor Insulation Type:",
+  "As Built",
+  "Floor U-Value known?",
+  "Not Known",
+  "Page 10",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Building Measurements",
+  "Area (m2)",
+  "Height (m)",
+  "Heat Loss Perimeter (m)",
+  "PWL (m)",
+  "Main Building",
+  "Floor 1",
+  "42.7",
+  "2.45",
+  "19.61",
+  "6.53",
+  "Floor 0",
+  "42.7",
+  "2.45",
+  "19.61",
+  "6.53",
+  "Roof Space",
+  "Main Building",
+  "Roofs - Construction Type:",
+  "Pitched roof (Slates or tiles), Access to loft",
+  "Roofs - Insulation At:",
+  "Joists",
+  "Roof U-Value:",
+  "Not Known",
+  "Roofs - Insulation Thickness:",
+  "200 mm",
+  "Page 11",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 12",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 13",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 14",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 15",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 16",
+  "",
+  "Loft insulation:",
+  "Record indicators of party wall construction in roof space:",
+  "Record indicators of Cavity Wall Construction in roof",
+  "space:",
+  "No indicator of construction visible",
+  "Are there rooms in the roof?",
+  "No",
+  "Windows",
+  "Window 1",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Page 17",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "1.26 m",
+  "Window width:",
+  "1.54 m",
+  "Orientation:",
+  "North East",
+  "Window 2",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.97 m",
+  "Window width:",
+  "1.04 m",
+  "Orientation:",
+  "North East",
+  "Page 18",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window 3",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.97 m",
+  "Window width:",
+  "1.53 m",
+  "Orientation:",
+  "North East",
+  "Window 4",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Page 19",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "1.27 m",
+  "Window width:",
+  "0.58 m",
+  "Orientation:",
+  "North East",
+  "Window 5",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.98 m",
+  "Window width:",
+  "0.54 m",
+  "Orientation:",
+  "South East",
+  "Page 20",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window 6",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "1.27 m",
+  "Window width:",
+  "0.58 m",
+  "Orientation:",
+  "South East",
+  "Window 7",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.98 m",
+  "Window width:",
+  "1.06 m",
+  "Page 21",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Orientation:",
+  "South West",
+  "Window 8",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.97 m",
+  "Window width:",
+  "1.56 m",
+  "Orientation:",
+  "South East",
+  "Window 9",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Page 22",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.96 m",
+  "Window width:",
+  "1.53 m",
+  "Orientation:",
+  "South East",
+  "Window 10",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing installed between 2002 - 2021",
+  "Identify basis of window dating:",
+  "Thermal spacer bar",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.95 m",
+  "Window width:",
+  "1.04 m",
+  "Orientation:",
+  "South East",
+  "Page 23",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Page 24",
+  "",
+  "Heating & Hot Water",
+  "Main Heating Systems",
+  "Main Heating 1",
+  "How would you like to select the Heating System?",
+  "PCDF Search",
+  "System type:",
+  "Electric storage heaters",
+  "Number of storage heater types:",
+  "3",
+  "Product Id",
+  "230026",
+  "Manufacturer",
+  "Dimplex",
+  "Model",
+  "Quantum",
+  "Orig Manuf",
+  "Dimplex",
+  "S. Efficiency",
+  "0",
+  "Year",
+  "2019 - current",
+  "Open Flue",
+  "No",
+  "Status",
+  "Normal status for an actual product",
+  "Number of heaters of this type:",
+  "1",
+  "Product Id",
+  "230024",
+  "Manufacturer",
+  "Dimplex",
+  "Model",
+  "Quantum",
+  "Orig Manuf",
+  "Dimplex",
+  "S. Efficiency",
+  "0",
+  "Year",
+  "2019 - current",
+  "Open Flue",
+  "No",
+  "Status",
+  "Normal status for an actual product",
+  "Number of heaters of this type:",
+  "2",
+  "Product Id",
+  "230023",
+  "Manufacturer",
+  "Dimplex",
+  "Model",
+  "Quantum",
+  "Orig Manuf",
+  "Dimplex",
+  "S. Efficiency",
+  "0",
+  "Year",
+  "2019 - current",
+  "Open Flue",
+  "No",
+  "Status",
+  "Normal status for an actual product",
+  "Number of heaters of this type:",
+  "3",
+  "Controls:",
+  "Controls for high heat retention storage heaters",
+  "Page 25",
+  "",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Page 26",
+  "",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating controls:",
+  "Photo of heating controls:",
+  "Photo of heating controls:",
+  "Page 27",
+  "",
+  "Photo of secondary heating system",
+  "Photo of cylinder and thermostat if present:",
+  "Photo of cylinder and thermostat if present:",
+  "Secondary Heating System",
+  "Secondary Fuel",
+  "Electricity",
+  "Secondary System:",
+  "Panel, convector or radiant heaters",
+  "Water Heating & Cylinder",
+  "Water Heating Type:",
+  "Regular",
+  "Water Heating System:",
+  "Electric immersion",
+  "Immersion:",
+  "Dual",
+  "Cylinder Size:",
+  "Medium (131-170 litres)",
+  "What is the cylinder measured heat loss:",
+  "Not known",
+  "Insulation Type:",
+  "Factory fitted",
+  "Thickness:",
+  "50 mm",
+  "Page 28",
+  "",
+  "Photo of cylinder and thermostat if present:",
+  "Photo of cylinder and thermostat if present:",
+  "Photo of cylinder and thermostat if present:",
+  "Photo of ventilation type:",
+  "Has thermostat?",
+  "Yes",
+  "Ventilation",
+  "Ventilation type:",
+  "Mechanical Extract - Decentralised",
+  "Has fixed air conditioning?",
+  "No",
+  "Page 29",
+  "",
+  "Photo of extract fans:",
+  "Is the ventilation in the PCDF database?",
+  "No",
+  "Number of open flues:",
+  "0",
+  "Number of closed flues:",
+  "0",
+  "Number of boiler flues:",
+  "0",
+  "Number of other flues:",
+  "0",
+  "Number of extract fans:",
+  "1",
+  "Number of passive vents:",
+  "0",
+  "Number of flueless gas fires:",
+  "0",
+  "Pressure test:",
+  "No test",
+  "Is there a draught lobby?",
+  "No",
+  "Conservatories",
+  "Is there conservatory?",
+  "No conservatory",
+  "Page 30",
+  "",
+  "Photo of photovoltaic array:",
+  "Photo of photovoltaic array:",
+  "Photo of photovoltaic array:",
+  "Photo of photovoltaic array:",
+  "Renewables",
+  "Wind Turbines",
+  "Has wind turbines?",
+  "No",
+  "Solar hot water",
+  "Has solar hot water?",
+  "No",
+  "Photovoltaics",
+  "Has photovoltaic array?",
+  "Yes",
+  "Is there a PV diverter?",
+  "No",
+  "PV Connection:",
+  "Connected to dwellings electricity meter",
+  "Photovoltaic array kWp Known?",
+  "No",
+  "Percentage of roof covered with photovoltaic array?",
+  "45 %",
+  "Number of PV batteries:",
+  "None",
+  "Hydro",
+  "Is the dwelling connected to Hydro?",
+  "No",
+  "Room Count Elements",
+  "Number of habitable rooms?",
+  "5",
+  "Are any of these rooms unheated?",
+  "No",
+  "Page 31",
+  "",
+  "Photo of open chimneys:",
+  "Photo of incandescent bulbs:",
+  "Photo of incandescent bulbs:",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Number of external doors?",
+  "2",
+  "Number of insulated external doors?",
+  "0",
+  "Number of draughtproofed external doors?",
+  "2",
+  "Number of open chimneys?",
+  "1",
+  "Number of blocked chimneys?",
+  "0",
+  "Number of fixed incandescent bulbs:",
+  "2",
+  "Is the exact number of LED and CFL bulbs known?",
+  "Yes",
+  "Number of fixed LED bulbs:",
+  "9",
+  "Page 32",
+  "",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Number of fixed CFL bulbs:",
+  "0",
+  "Are there any waste water heat recovery systems?",
+  "None",
+  "Number of baths:",
+  "1",
+  "How many special features are there at the",
+  "property?",
+  "0",
+  "Showers",
+  "Shower 1",
+  "Shower outlet type:",
+  "Electric Shower",
+  "Page 33",
+  "",
+  "Photo of shower:",
+  "Photo of shower:",
+  "Customer Response",
+  "Customer present?",
+  "Yes",
+  "Customer willing to answer satisfaction survey?",
+  "No",
+  "Addendum + Related Party Disclosure",
+  "Addendum",
+  "None",
+  "Related party disclosure",
+  "No related party",
+  "Hard to treat cavity walls: Property has access",
+  "issues?",
+  "No",
+  "Hard to treat cavity walls: Property has high",
+  "exposure?",
+  "No",
+  "Hard to treat cavity walls: Property has narrow",
+  "cavities?",
+  "No",
+  "Photographs Required",
+  "Page 34",
+  "",
+  "General Photos:",
+  "General Photos:",
+  "General Photos:",
+  "General Photos:",
+  "General Photos:",
+  "General Photos:",
+  "Page 35",
+  "",
+  "General Photos:",
+  "General Photos:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 36",
+  "",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 37",
+  "",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 38",
+  "",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 39",
+  "",
+  "Page 40",
+  "",
+  "Page 41",
+  "",
+  "Additional Notes",
+  "Additional Notes",
+  "dMEV in kitchen, IEV in bathroom. Vent on chimney more than 300x100mm so",
+  "recorded as open chimney.",
+  "Page 42",
+  ""
+]

backend/documents_parser/tests/fixtures/site_notes_example_4_text.json

@@ -0,0 +1,480 @@
+[
+  "SMART EPC: Record of",
+  "Inspection & Site Notes",
+  "Inspection Surveyor:",
+  "Rebecca Mcwilliam",
+  "E-Mail Address:",
+  "rebeccamcdea@gmail.com",
+  "Report Reference:",
+  "DA1A93D9-354C-4B4B-A299-3A681231D4B5",
+  "Created On:",
+  "15 October 2025",
+  "Date of Inspection:",
+  "13 October 2025",
+  "Property Address:",
+  "18,",
+  "Oakfield Close,",
+  "Wrenbury,",
+  "CW5 8ET",
+  "Property Photo",
+  "Page 1",
+  "",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "RdSAP Assessment",
+  "General",
+  "Confirm you have checked for the existence of an",
+  "EPC before carrying out another energy assessment.",
+  "Yes",
+  "Does an EPC exist at the point of carrying out this",
+  "energy assessment?",
+  "Yes",
+  "Please select why another energy assessment needs",
+  "to be undertaken:",
+  "Assessor instructed to produce a new EPC upon request from building",
+  "owner/tenant/landlord after confirming to the requestor that a valid EPC",
+  "already exists",
+  "Inspection Date:",
+  "13/10/2025",
+  "Transaction Type:",
+  "None of the Above",
+  "Tenure:",
+  "Rented Social",
+  "Type of Property:",
+  "Bungalow",
+  "Detachment Type:",
+  "End-terrace",
+  "Number of storeys:",
+  "1 Storey",
+  "Terrain Type:",
+  "Suburban",
+  "Number of Extensions:",
+  "No Extensions",
+  "Is an electricity smart meter present?",
+  "Yes",
+  "Electric meter type:",
+  "Single",
+  "Page 2",
+  "",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "External indicators of Cavity Wall Construction:",
+  "External indicators of Cavity Wall Construction:",
+  "Is the dwelling export-capable?",
+  "Yes",
+  "Is mains gas available?",
+  "No",
+  "Select Measurements Location:",
+  "Internal",
+  "Building Construction",
+  "Main Building",
+  "Age Range:",
+  "1950-1966",
+  "Record indicators of property age:",
+  "local knowledge, enquiries of owner, period building features",
+  "Walls - Construction Type:",
+  "Cavity",
+  "Record external indicators of Cavity Construction:",
+  "stretcher bond, wall thickness over 270 mm",
+  "Walls - Insulation Type:",
+  "Filled Cavity",
+  "Record indicators of filled cavity:",
+  "evidence of cavity fill drill holes, Boroscope",
+  "Page 3",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 4",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo wall thickness:",
+  "Thermal conductivity of wall insulation:",
+  "Unknown",
+  "Wall U-Value known?",
+  "Not Known",
+  "Wall thickness:",
+  "300 mm",
+  "Party wall construction type:",
+  "Unable to determine",
+  "Floor type:",
+  "Ground Floor",
+  "Floor Construction:",
+  "Solid",
+  "Floor Insulation Type:",
+  "As Built",
+  "Floor U-Value known?",
+  "Not Known",
+  "Page 5",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Building Measurements",
+  "Area (m2)",
+  "Height (m)",
+  "Heat Loss Perimeter (m)",
+  "PWL (m)",
+  "Main Building",
+  "Floor 0",
+  "46.08",
+  "2.44",
+  "20.49",
+  "6.67",
+  "Roof Space",
+  "Main Building",
+  "Roofs - Construction Type:",
+  "Pitched roof (Slates or tiles), No access",
+  "Identify the reason for restricted access:",
+  "access hatch blocked",
+  "Roofs - Insulation At:",
+  "Unknown",
+  "Page 6",
+  "",
+  "Loft insulation:",
+  "Record indicators of Cavity Wall Construction in roof",
+  "space:",
+  "No indicator of construction visible",
+  "Are there rooms in the roof?",
+  "No",
+  "Windows",
+  "Window 1",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "12 mm",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.97 m",
+  "Window width:",
+  "1.54 m",
+  "Orientation:",
+  "North East",
+  "Page 7",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window 2",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "12 mm",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "1.01 m",
+  "Window width:",
+  "1.04 m",
+  "Orientation:",
+  "North East",
+  "Window 3",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Page 8",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "What size is the glazing gap?",
+  "12 mm",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.99 m",
+  "Window width:",
+  "1.07 m",
+  "Orientation:",
+  "South West",
+  "Window 4",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "12 mm",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "1.28 m",
+  "Window width:",
+  "2.07 m",
+  "Orientation:",
+  "South",
+  "Page 9",
+  "",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Heating & Hot Water",
+  "Main Heating Systems",
+  "Main Heating 1",
+  "How would you like to select the Heating System?",
+  "PCDF Search",
+  "System type:",
+  "Heat pump with radiators or underfloor heating",
+  "Product Id",
+  "102421",
+  "Manufacturer",
+  "Daikin Altherma",
+  "Model",
+  "EDLQ05CAV3",
+  "Year",
+  "2014 - current",
+  "Fuel",
+  "Electricity, any tariff",
+  "Status",
+  "Normal status for an actual product",
+  "Central heating pump age:",
+  "Unknown",
+  "MCS installed heat pump:",
+  "Yes",
+  "Controls:",
+  "Programmer, room thermostat and TRVs",
+  "Emitter:",
+  "Radiators",
+  "Emitter Temperature:",
+  "Unknown",
+  "Page 10",
+  "",
+  "Photo of heating controls:",
+  "Photo of heating controls:",
+  "Photo of heating controls:",
+  "Photo of heating controls:",
+  "Photo of heating controls:",
+  "Photo of heating controls:",
+  "Page 11",
+  "",
+  "Photo of heating controls:",
+  "Photo of secondary heating system",
+  "Secondary Heating System",
+  "Secondary Fuel",
+  "Electricity",
+  "Secondary System:",
+  "Panel, convector or radiant heaters",
+  "Water Heating & Cylinder",
+  "Water Heating Type:",
+  "Regular",
+  "Water Heating System:",
+  "From main heating 1",
+  "Page 12",
+  "",
+  "Photo of water heating system:",
+  "Photo of water heating system:",
+  "Photo of water heating system:",
+  "Cylinder Size:",
+  "Medium (131-170 litres)",
+  "What is the cylinder measured heat loss:",
+  "Not known",
+  "Insulation Type:",
+  "Factory fitted",
+  "Thickness:",
+  "50 mm",
+  "Page 13",
+  "",
+  "Photo of cylinder and thermostat if present:",
+  "Photo of cylinder and thermostat if present:",
+  "Photo of ventilation type:",
+  "Photo of ventilation type:",
+  "Has thermostat?",
+  "Yes",
+  "Ventilation",
+  "Ventilation type:",
+  "Mechanical Extract - Decentralised",
+  "Has fixed air conditioning?",
+  "No",
+  "Is the ventilation in the PCDF database?",
+  "No",
+  "Number of open flues:",
+  "0",
+  "Number of closed flues:",
+  "0",
+  "Number of boiler flues:",
+  "0",
+  "Page 14",
+  "",
+  "Photo of extract fans:",
+  "Number of other flues:",
+  "0",
+  "Number of extract fans:",
+  "1",
+  "Number of passive vents:",
+  "0",
+  "Number of flueless gas fires:",
+  "0",
+  "Pressure test:",
+  "No test",
+  "Is there a draught lobby?",
+  "No",
+  "Conservatories",
+  "Is there conservatory?",
+  "No conservatory",
+  "Renewables",
+  "Wind Turbines",
+  "Has wind turbines?",
+  "No",
+  "Solar hot water",
+  "Has solar hot water?",
+  "No",
+  "Photovoltaics",
+  "Has photovoltaic array?",
+  "No",
+  "Number of PV batteries:",
+  "None",
+  "Hydro",
+  "Is the dwelling connected to Hydro?",
+  "No",
+  "Room Count Elements",
+  "Number of habitable rooms?",
+  "2",
+  "Are any of these rooms unheated?",
+  "No",
+  "Page 15",
+  "",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Photo of LED bulbs:",
+  "Number of external doors?",
+  "2",
+  "Number of insulated external doors?",
+  "0",
+  "Number of draughtproofed external doors?",
+  "2",
+  "Number of open chimneys?",
+  "0",
+  "Number of blocked chimneys?",
+  "0",
+  "Number of fixed incandescent bulbs:",
+  "0",
+  "Is the exact number of LED and CFL bulbs known?",
+  "Yes",
+  "Number of fixed LED bulbs:",
+  "7",
+  "Page 16",
+  "",
+  "Photo of LED bulbs:",
+  "Photo of shower:",
+  "Photo of shower:",
+  "Number of fixed CFL bulbs:",
+  "0",
+  "Are there any waste water heat recovery systems?",
+  "None",
+  "Number of baths:",
+  "1",
+  "How many special features are there at the",
+  "property?",
+  "0",
+  "Showers",
+  "Shower 1",
+  "Shower outlet type:",
+  "Electric Shower",
+  "Customer Response",
+  "Customer present?",
+  "Yes",
+  "Customer willing to answer satisfaction survey?",
+  "No",
+  "Addendum + Related Party Disclosure",
+  "Addendum",
+  "None",
+  "Page 17",
+  "",
+  "General Photos:",
+  "General Photos:",
+  "External Elevations:",
+  "External Elevations:",
+  "Related party disclosure",
+  "No related party",
+  "Hard to treat cavity walls: Property has access",
+  "issues?",
+  "No",
+  "Hard to treat cavity walls: Property has high",
+  "exposure?",
+  "No",
+  "Hard to treat cavity walls: Property has narrow",
+  "cavities?",
+  "No",
+  "Photographs Required",
+  "Page 18",
+  "",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 19",
+  "",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 20",
+  "",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 21",
+  "",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 22",
+  "",
+  "Additional Notes",
+  "Additional Notes",
+  "The loft hatch is a drop-down hatch with a ladder above. The hatch is broken",
+  "and it could not be opened. She is about to have a new door made so loft will",
+  "be able to be accessed at some point soon but on the day I could not.",
+  "Page 23",
+  ""
+]

backend/documents_parser/tests/fixtures/site_notes_example_text.json

@@ -0,0 +1,643 @@
+[
+  "SMART EPC: Record of",
+  "Inspection & Site Notes",
+  "Inspection Surveyor:",
+  "Benjamin Burke",
+  "E-Mail Address:",
+  "ben@mbsolutionsgroup.co.uk",
+  "Report Reference:",
+  "6EA2A86D-94CE-4792-8D49-AB495C744EDD",
+  "Created On:",
+  "10 November 2025",
+  "Date of Inspection:",
+  "25 September 2025",
+  "Property Address:",
+  "40,",
+  "Abbey Place,",
+  "Crewe,",
+  "Cheshire,",
+  "CW1 4JR",
+  "Property Photo",
+  "Page 1",
+  "",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "Photo of electricity meter:",
+  "RdSAP Assessment",
+  "General",
+  "Confirm you have checked for the existence of an",
+  "EPC before carrying out another energy assessment.",
+  "Yes",
+  "Does an EPC exist at the point of carrying out this",
+  "energy assessment?",
+  "No",
+  "Inspection Date:",
+  "25/09/2025",
+  "Transaction Type:",
+  "Grant-Scheme (ECO, RHI, etc.)",
+  "Tenure:",
+  "Rented Social",
+  "Type of Property:",
+  "House",
+  "Detachment Type:",
+  "Mid-terrace",
+  "Number of storeys:",
+  "2 Storeys",
+  "Terrain Type:",
+  "Suburban",
+  "Number of Extensions:",
+  "1 Extension",
+  "Is an electricity smart meter present?",
+  "Yes",
+  "Electric meter type:",
+  "Single",
+  "Is the dwelling export-capable?",
+  "Yes",
+  "Is mains gas available?",
+  "Yes",
+  "Is there a gas smart meter?",
+  "Yes",
+  "Is the gas meter accessible?",
+  "Yes",
+  "Page 2",
+  "",
+  "Photo of Gas Meter:",
+  "External indicators of Cavity Wall Construction:",
+  "External indicators of Cavity Wall Construction:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Select Measurements Location:",
+  "Internal",
+  "Building Construction",
+  "Main Building",
+  "Age Range:",
+  "1950-1966",
+  "Record indicators of property age:",
+  "local knowledge, enquiries of owner",
+  "Walls - Construction Type:",
+  "Cavity",
+  "Record external indicators of Cavity Construction:",
+  "wall thickness over 270 mm",
+  "Walls - Insulation Type:",
+  "Filled Cavity",
+  "Record indicators of filled cavity:",
+  "evidence of cavity fill drill holes",
+  "Page 3",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Page 4",
+  "",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Photo indicators of filled cavity insulation:",
+  "Thermal conductivity of wall insulation:",
+  "Unknown",
+  "Wall U-Value known?",
+  "Not Known",
+  "Wall thickness:",
+  "310 mm",
+  "Page 5",
+  "",
+  "Photo wall thickness:",
+  "External indicators of Cavity Wall Construction:",
+  "Party wall construction type:",
+  "Cavity Masonry, Filled",
+  "Floor type:",
+  "Ground Floor",
+  "Floor Construction:",
+  "Solid",
+  "Floor Insulation Type:",
+  "As Built",
+  "Floor U-Value known?",
+  "Not Known",
+  "Extension 1",
+  "Age Range:",
+  "2003-2006",
+  "Record indicators of property age:",
+  "local knowledge, enquiries of owner",
+  "Walls - Construction Type:",
+  "Cavity",
+  "Record external indicators of Cavity Construction:",
+  "wall thickness over 270 mm",
+  "Walls - Insulation Type:",
+  "As built",
+  "Thermal conductivity of wall insulation:",
+  "Unknown",
+  "Wall U-Value known?",
+  "Not Known",
+  "Wall thickness:",
+  "310 mm",
+  "Page 6",
+  "",
+  "Photo wall thickness:",
+  "Party wall construction type:",
+  "Cavity Masonry, Filled",
+  "Floor type:",
+  "Ground Floor",
+  "Floor Construction:",
+  "Solid",
+  "Floor Insulation Type:",
+  "As Built",
+  "Floor U-Value known?",
+  "Not Known",
+  "Building Measurements",
+  "Area (m2)",
+  "Height (m)",
+  "Heat Loss Perimeter (m)",
+  "PWL (m)",
+  "Main Building",
+  "Floor 1",
+  "35.68",
+  "2.19",
+  "13.44",
+  "10.62",
+  "Floor 0",
+  "35.68",
+  "2.17",
+  "11",
+  "10.62",
+  "Extension 1",
+  "Floor 0",
+  "3.8",
+  "2",
+  "5.7",
+  "0",
+  "Roof Space",
+  "Main Building",
+  "Roofs - Construction Type:",
+  "Pitched roof (Slates or tiles), Access to loft",
+  "Roofs - Insulation At:",
+  "Joists",
+  "Roof U-Value:",
+  "Not Known",
+  "Roofs - Insulation Thickness:",
+  "100 mm",
+  "Page 7",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 8",
+  "",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Loft insulation:",
+  "Page 9",
+  "",
+  "Loft insulation:",
+  "Indicators of Cavity Wall Construction in roof space:",
+  "Indicators of Cavity Wall Construction in roof space:",
+  "Record indicators of party wall construction in roof space:",
+  "Record indicators of party wall construction in roof space:",
+  "Record indicators of Cavity Wall Construction in roof",
+  "space:",
+  "cavity visible in roof space",
+  "Are there rooms in the roof?",
+  "No",
+  "Extension 1",
+  "Roofs - Construction Type:",
+  "Pitched roof, Sloping ceiling",
+  "Page 10",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Roofs - Insulation At:",
+  "Sloping ceiling insulation",
+  "Roof U-Value:",
+  "Not Known",
+  "Roofs - Insulation Thickness:",
+  "As built",
+  "Record indicators of Cavity Wall Construction in roof",
+  "space:",
+  "No indicator of construction visible",
+  "Are there rooms in the roof?",
+  "No",
+  "Windows",
+  "Window 1",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "16 mm or more",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "1.2 m",
+  "Window width:",
+  "2.3 m",
+  "Orientation:",
+  "North West",
+  "Window 2",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "16 mm or more",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Page 11",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window height:",
+  "1.2 m",
+  "Window width:",
+  "1 m",
+  "Orientation:",
+  "North West",
+  "Window 3",
+  "Window location:",
+  "Main Building",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "16 mm or more",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.9 m",
+  "Window width:",
+  "1 m",
+  "Orientation:",
+  "North East",
+  "Window 4",
+  "Window location:",
+  "Extension 1",
+  "Window wall type:",
+  "External wall",
+  "Page 12",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "16 mm or more",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.9 m",
+  "Window width:",
+  "1 m",
+  "Orientation:",
+  "North",
+  "Window 5",
+  "Window location:",
+  "Extension 1",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "16 mm or more",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.9 m",
+  "Window width:",
+  "1.7 m",
+  "Orientation:",
+  "North East",
+  "Page 13",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Window 6",
+  "Window location:",
+  "Extension 1",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "16 mm or more",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.9 m",
+  "Window width:",
+  "2.3 m",
+  "Orientation:",
+  "North West",
+  "Window 7",
+  "Window location:",
+  "Extension 1",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "Page 14",
+  "",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "Photo of glazing type:",
+  "What size is the glazing gap?",
+  "16 mm or more",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "1 m",
+  "Window width:",
+  "1.2 m",
+  "Orientation:",
+  "North West",
+  "Window 8",
+  "Window location:",
+  "Extension 1",
+  "Window wall type:",
+  "External wall",
+  "Glazing Type:",
+  "Double glazing, Unknown install date",
+  "Window type:",
+  "Window",
+  "Window frame type:",
+  "Wooden or PVC",
+  "What size is the glazing gap?",
+  "16 mm or more",
+  "Is the window draught proofed?",
+  "Yes",
+  "Are there permanent shutters present?",
+  "No",
+  "Window height:",
+  "0.9 m",
+  "Window width:",
+  "1 m",
+  "Orientation:",
+  "North East",
+  "Page 15",
+  "",
+  "Photo of heating system:",
+  "Heating & Hot Water",
+  "Main Heating Systems",
+  "Main Heating 1",
+  "How would you like to select the Heating System?",
+  "PCDF Search",
+  "System type:",
+  "Boiler with radiators or underfloor heating",
+  "Product Id",
+  "16839",
+  "Manufacturer",
+  "Vaillant",
+  "Model",
+  "ecoTEC pro 28",
+  "Orig Manuf",
+  "Vaillant",
+  "Fuel",
+  "Mains gas",
+  "S. Efficiency",
+  "0",
+  "Type",
+  "Combi",
+  "Condensing",
+  "Yes",
+  "Year",
+  "2005 - 2015",
+  "Mount",
+  "Wall",
+  "Open Flue",
+  "Room-sealed",
+  "Fan Assist",
+  "Yes",
+  "Status",
+  "Normal status for an actual product",
+  "Central heating pump age:",
+  "Unknown",
+  "Controls:",
+  "Programmer, room thermostat and TRVs",
+  "Does the boiler have a Flue Gas Heat Recover",
+  "System (FGHRS)?",
+  "No",
+  "Is there a weather compensator?",
+  "No",
+  "Emitter:",
+  "Radiators",
+  "Emitter Temperature:",
+  "Unknown",
+  "Page 16",
+  "",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Page 17",
+  "",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating system:",
+  "Photo of heating controls:",
+  "Photo of heating controls:",
+  "Secondary Heating System",
+  "Secondary Fuel",
+  "No Secondary Heating",
+  "Water Heating & Cylinder",
+  "Water Heating Type:",
+  "Regular",
+  "Water Heating System:",
+  "From main heating 1",
+  "Cylinder Size:",
+  "No Cylinder",
+  "Ventilation",
+  "Ventilation type:",
+  "Mechanical Extract - Decentralised",
+  "Page 18",
+  "",
+  "Photo of ventilation type:",
+  "Has fixed air conditioning?",
+  "No",
+  "Is the ventilation in the PCDF database?",
+  "No",
+  "Number of open flues:",
+  "0",
+  "Number of closed flues:",
+  "0",
+  "Number of boiler flues:",
+  "0",
+  "Number of other flues:",
+  "0",
+  "Number of extract fans:",
+  "0",
+  "Number of passive vents:",
+  "0",
+  "Number of flueless gas fires:",
+  "0",
+  "Pressure test:",
+  "No test",
+  "Is there a draught lobby?",
+  "No",
+  "Conservatories",
+  "Is there conservatory?",
+  "No conservatory",
+  "Page 19",
+  "",
+  "Photo of incandescent bulbs:",
+  "Photo of incandescent bulbs:",
+  "Renewables",
+  "Wind Turbines",
+  "Has wind turbines?",
+  "No",
+  "Solar hot water",
+  "Has solar hot water?",
+  "No",
+  "Photovoltaics",
+  "Has photovoltaic array?",
+  "No",
+  "Number of PV batteries:",
+  "None",
+  "Hydro",
+  "Is the dwelling connected to Hydro?",
+  "No",
+  "Room Count Elements",
+  "Number of habitable rooms?",
+  "3",
+  "Are any of these rooms unheated?",
+  "No",
+  "Number of external doors?",
+  "2",
+  "Number of insulated external doors?",
+  "0",
+  "Number of draughtproofed external doors?",
+  "2",
+  "Number of open chimneys?",
+  "0",
+  "Number of blocked chimneys?",
+  "0",
+  "Number of fixed incandescent bulbs:",
+  "4",
+  "Page 20",
+  "",
+  "Photo of incandescent bulbs:",
+  "Photo of incandescent bulbs:",
+  "Photo of CFL bulbs:",
+  "Is the exact number of LED and CFL bulbs known?",
+  "Yes",
+  "Number of fixed LED bulbs:",
+  "0",
+  "Number of fixed CFL bulbs:",
+  "1",
+  "Are there any waste water heat recovery systems?",
+  "None",
+  "Number of baths:",
+  "1",
+  "How many special features are there at the",
+  "property?",
+  "0",
+  "Showers",
+  "Shower 1",
+  "Shower outlet type:",
+  "Non-Electric Shower",
+  "Page 21",
+  "",
+  "Photo of shower:",
+  "Photo of shower:",
+  "General Photos:",
+  "Customer Response",
+  "Customer present?",
+  "Yes",
+  "Customer willing to answer satisfaction survey?",
+  "No",
+  "Addendum + Related Party Disclosure",
+  "Addendum",
+  "None",
+  "Related party disclosure",
+  "No related party",
+  "Hard to treat cavity walls: Property has access",
+  "issues?",
+  "No",
+  "Hard to treat cavity walls: Property has high",
+  "exposure?",
+  "No",
+  "Hard to treat cavity walls: Property has narrow",
+  "cavities?",
+  "No",
+  "Photographs Required",
+  "Page 22",
+  "",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "External Elevations:",
+  "Page 23",
+  "",
+  "Page 24",
+  "",
+  "Page 25",
+  "",
+  "Page 26",
+  "",
+  "Page 27",
+  ""
+]

backend/documents_parser/tests/test_end_to_end.py

@@ -0,0 +1,422 @@
+import os
+from datetime import date
+
+import pytest
+
+from backend.documents_parser.extractor import PasHubRdSapSiteNotesExtractor
+from backend.documents_parser.pdf import pdf_to_text_list
+from datatypes.epc.domain.epc_property_data import (
+    EpcPropertyData,
+    InstantaneousWwhrs,
+    MainHeatingDetail,
+    SapBuildingPart,
+    SapEnergySource,
+    SapFloorDimension,
+    SapHeating,
+    SapVentilation,
+    SapWindow,
+    ShowerOutlet,
+    ShowerOutlets,
+)
+from datatypes.epc.domain.mapper import EpcPropertyDataMapper
+
+PDF_PATH = os.path.join(os.path.dirname(__file__), "fixtures", "ExampleSiteNotes.pdf")
+PDF_PATH_2 = os.path.join(
+    os.path.dirname(__file__), "fixtures", "ExampleSiteNotes_2.pdf"
+)
+
+
+class TestPdfToEpcPropertyData:
+    @pytest.fixture
+    def result(self) -> EpcPropertyData:
+        with open(PDF_PATH, "rb") as f:
+            pdf_bytes = f.read()
+        site_notes = PasHubRdSapSiteNotesExtractor(
+            pdf_to_text_list(pdf_bytes)
+        ).extract()
+        return EpcPropertyDataMapper.from_site_notes(site_notes)
+
+    def test_full_epc_property_data(self, result: EpcPropertyData) -> None:
+        assert result == EpcPropertyData(
+            dwelling_type="Mid-terrace house",
+            inspection_date=date(2025, 9, 25),
+            tenure="Rented Social",
+            transaction_type="Grant-Scheme (ECO, RHI, etc.)",
+            roofs=[],
+            walls=[],
+            floors=[],
+            main_heating=[],
+            door_count=2,
+            sap_heating=SapHeating(
+                instantaneous_wwhrs=InstantaneousWwhrs(),
+                main_heating_details=[
+                    MainHeatingDetail(
+                        has_fghrs=False,
+                        main_fuel_type="Mains gas",
+                        heat_emitter_type="Radiators",
+                        emitter_temperature="Unknown",
+                        main_heating_control="Programmer, room thermostat and TRVs",
+                        fan_flue_present=True,
+                        condensing=True,
+                        weather_compensator=False,
+                        central_heating_pump_age_str="Unknown",
+                    )
+                ],
+                has_fixed_air_conditioning=False,
+                shower_outlets=ShowerOutlets(
+                    shower_outlet=ShowerOutlet(
+                        shower_outlet_type="Non-Electric Shower"
+                    ),
+                ),
+            ),
+            sap_windows=[
+                SapWindow(
+                    pvc_frame="Wooden or PVC",
+                    glazing_gap="16 mm or more",
+                    orientation="North West",
+                    window_type="Window",
+                    glazing_type="Double glazing, Unknown install date",
+                    window_width=2.3,
+                    window_height=1.2,
+                    draught_proofed=True,
+                    window_location="Main Building",
+                    window_wall_type="External wall",
+                    permanent_shutters_present=False,
+                ),
+                SapWindow(
+                    pvc_frame="Wooden or PVC",
+                    glazing_gap="16 mm or more",
+                    orientation="North West",
+                    window_type="Window",
+                    glazing_type="Double glazing, Unknown install date",
+                    window_width=1.0,
+                    window_height=1.2,
+                    draught_proofed=True,
+                    window_location="Main Building",
+                    window_wall_type="External wall",
+                    permanent_shutters_present=False,
+                ),
+                SapWindow(
+                    pvc_frame="Wooden or PVC",
+                    glazing_gap="16 mm or more",
+                    orientation="North East",
+                    window_type="Window",
+                    glazing_type="Double glazing, Unknown install date",
+                    window_width=1.0,
+                    window_height=0.9,
+                    draught_proofed=True,
+                    window_location="Main Building",
+                    window_wall_type="External wall",
+                    permanent_shutters_present=False,
+                ),
+                SapWindow(
+                    pvc_frame="Wooden or PVC",
+                    glazing_gap="16 mm or more",
+                    orientation="North",
+                    window_type="Window",
+                    glazing_type="Double glazing, Unknown install date",
+                    window_width=1.0,
+                    window_height=0.9,
+                    draught_proofed=True,
+                    window_location="Extension 1",
+                    window_wall_type="External wall",
+                    permanent_shutters_present=False,
+                ),
+                SapWindow(
+                    pvc_frame="Wooden or PVC",
+                    glazing_gap="16 mm or more",
+                    orientation="North East",
+                    window_type="Window",
+                    glazing_type="Double glazing, Unknown install date",
+                    window_width=1.7,
+                    window_height=0.9,
+                    draught_proofed=True,
+                    window_location="Extension 1",
+                    window_wall_type="External wall",
+                    permanent_shutters_present=False,
+                ),
+                SapWindow(
+                    pvc_frame="Wooden or PVC",
+                    glazing_gap="16 mm or more",
+                    orientation="North West",
+                    window_type="Window",
+                    glazing_type="Double glazing, Unknown install date",
+                    window_width=2.3,
+                    window_height=0.9,
+                    draught_proofed=True,
+                    window_location="Extension 1",
+                    window_wall_type="External wall",
+                    permanent_shutters_present=False,
+                ),
+                SapWindow(
+                    pvc_frame="Wooden or PVC",
+                    glazing_gap="16 mm or more",
+                    orientation="North West",
+                    window_type="Window",
+                    glazing_type="Double glazing, Unknown install date",
+                    window_width=1.2,
+                    window_height=1.0,
+                    draught_proofed=True,
+                    window_location="Extension 1",
+                    window_wall_type="External wall",
+                    permanent_shutters_present=False,
+                ),
+                SapWindow(
+                    pvc_frame="Wooden or PVC",
+                    glazing_gap="16 mm or more",
+                    orientation="North East",
+                    window_type="Window",
+                    glazing_type="Double glazing, Unknown install date",
+                    window_width=1.0,
+                    window_height=0.9,
+                    draught_proofed=True,
+                    window_location="Extension 1",
+                    window_wall_type="External wall",
+                    permanent_shutters_present=False,
+                ),
+            ],
+            sap_energy_source=SapEnergySource(
+                mains_gas=True,
+                meter_type="Single",
+                pv_battery_count=0,
+                wind_turbines_count=0,
+                gas_smart_meter_present=True,
+                is_dwelling_export_capable=True,
+                wind_turbines_terrain_type="Suburban",
+                electricity_smart_meter_present=True,
+            ),
+            sap_building_parts=[
+                SapBuildingPart(
+                    identifier="main",
+                    construction_age_band="1950-1966",
+                    wall_construction="Cavity",
+                    wall_insulation_type="Filled Cavity",
+                    wall_thickness_measured=True,
+                    party_wall_construction="Cavity Masonry, Filled",
+                    sap_floor_dimensions=[
+                        SapFloorDimension(
+                            room_height_m=2.19,
+                            total_floor_area_m2=35.68,
+                            party_wall_length_m=10.62,
+                            heat_loss_perimeter_m=13.44,
+                            floor=1,
+                        ),
+                        SapFloorDimension(
+                            room_height_m=2.17,
+                            total_floor_area_m2=35.68,
+                            party_wall_length_m=10.62,
+                            heat_loss_perimeter_m=11.0,
+                            floor=0,
+                        ),
+                    ],
+                    wall_thickness_mm=310,
+                    roof_insulation_location="Joists",
+                    roof_insulation_thickness=100,
+                    floor_type="Ground Floor",
+                    floor_construction_type="Solid",
+                    floor_insulation_type_str="As Built",
+                    floor_u_value_known=False,
+                ),
+                SapBuildingPart(
+                    identifier="extension_1",
+                    construction_age_band="2003-2006",
+                    wall_construction="Cavity",
+                    wall_insulation_type="As built",
+                    wall_thickness_measured=True,
+                    party_wall_construction="Cavity Masonry, Filled",
+                    sap_floor_dimensions=[
+                        SapFloorDimension(
+                            room_height_m=2.0,
+                            total_floor_area_m2=3.8,
+                            party_wall_length_m=0.0,
+                            heat_loss_perimeter_m=5.7,
+                            floor=0,
+                        ),
+                    ],
+                    wall_thickness_mm=310,
+                    roof_insulation_location="Sloping ceiling insulation",
+                    roof_insulation_thickness="As built",
+                ),
+            ],
+            solar_water_heating=False,
+            has_hot_water_cylinder=False,
+            has_fixed_air_conditioning=False,
+            wet_rooms_count=0,
+            extensions_count=1,
+            heated_rooms_count=0,
+            open_chimneys_count=0,
+            habitable_rooms_count=3,
+            insulated_door_count=0,
+            cfl_fixed_lighting_bulbs_count=1,
+            led_fixed_lighting_bulbs_count=0,
+            incandescent_fixed_lighting_bulbs_count=4,
+            total_floor_area_m2=75.16,
+            built_form="Mid-terrace",
+            property_type="House",
+            has_conservatory=False,
+            blocked_chimneys_count=0,
+            draughtproofed_door_count=2,
+            address_line_1="40, Abbey Place",
+            post_town="Crewe",
+            postcode="CW1 4JR",
+            report_reference="6EA2A86D-94CE-4792-8D49-AB495C744EDD",
+            number_of_storeys=2,
+            any_unheated_rooms=False,
+            waste_water_heat_recovery="None",
+            hydro=False,
+            photovoltaic_array=False,
+            sap_ventilation=SapVentilation(
+                ventilation_type="Mechanical Extract - Decentralised",
+                draught_lobby=False,
+                pressure_test="No test",
+                open_flues_count=0,
+                closed_flues_count=0,
+                boiler_flues_count=0,
+                other_flues_count=0,
+                extract_fans_count=0,
+                passive_vents_count=0,
+                flueless_gas_fires_count=0,
+                ventilation_in_pcdf_database=False,
+            ),
+        )
+
+
+class TestPdfToEpcPropertyDataFixture2:
+    @pytest.fixture
+    def result(self) -> EpcPropertyData:
+        with open(PDF_PATH_2, "rb") as f:
+            pdf_bytes = f.read()
+        site_notes = PasHubRdSapSiteNotesExtractor(
+            pdf_to_text_list(pdf_bytes)
+        ).extract()
+        return EpcPropertyDataMapper.from_site_notes(site_notes)
+
+    def test_cylinder_insulation_thickness(self, result: EpcPropertyData) -> None:
+        assert result.sap_heating.cylinder_insulation_thickness_mm == 38
+
+    def test_cylinder_size(self, result: EpcPropertyData) -> None:
+        assert result.sap_heating.cylinder_size == "Normal (90-130 litres)"
+
+    def test_secondary_heating_type(self, result: EpcPropertyData) -> None:
+        assert result.sap_heating.secondary_heating_type == "Open fire in grate"
+
+
+PDF_PATH_3 = os.path.join(
+    os.path.dirname(__file__), "fixtures", "ExampleSiteNotes_3.pdf"
+)
+
+
+class TestPdfToEpcPropertyDataFixture3:
+    @pytest.fixture
+    def result(self) -> EpcPropertyData:
+        with open(PDF_PATH_3, "rb") as f:
+            pdf_bytes = f.read()
+        site_notes = PasHubRdSapSiteNotesExtractor(
+            pdf_to_text_list(pdf_bytes)
+        ).extract()
+        return EpcPropertyDataMapper.from_site_notes(site_notes)
+
+    def test_immersion_heating_type(self, result: EpcPropertyData) -> None:
+        assert result.sap_heating.immersion_heating_type == "Dual"
+
+    def test_pv_connection(self, result: EpcPropertyData) -> None:
+        assert (
+            result.sap_energy_source.pv_connection
+            == "Connected to dwellings electricity meter"
+        )
+
+    def test_photovoltaic_supply_percent_roof(self, result: EpcPropertyData) -> None:
+        assert result.sap_energy_source.photovoltaic_supply is not None
+        assert (
+            result.sap_energy_source.photovoltaic_supply.none_or_no_details.percent_roof_area
+            == 45
+        )
+
+    def test_electric_storage_heater_fuel_type(self, result: EpcPropertyData) -> None:
+        assert (
+            result.sap_heating.main_heating_details[0].main_fuel_type == "Electricity"
+        )
+
+
+PDF_PATH_4 = os.path.join(
+    os.path.dirname(__file__), "fixtures", "ExampleSiteNotes_4.pdf"
+)
+
+
+class TestPdfToEpcPropertyDataFixture4:
+    @pytest.fixture
+    def result(self) -> EpcPropertyData:
+        with open(PDF_PATH_4, "rb") as f:
+            pdf_bytes = f.read()
+        site_notes = PasHubRdSapSiteNotesExtractor(
+            pdf_to_text_list(pdf_bytes)
+        ).extract()
+        return EpcPropertyDataMapper.from_site_notes(site_notes)
+
+    def test_cylinder_insulation_type(self, result: EpcPropertyData) -> None:
+        assert result.sap_heating.cylinder_insulation_type == "Factory fitted"
+
+    def test_heat_pump_fuel_type(self, result: EpcPropertyData) -> None:
+        assert (
+            result.sap_heating.main_heating_details[0].main_fuel_type == "Electricity"
+        )
+
+    def test_roof_insulation_location_unknown(self, result: EpcPropertyData) -> None:
+        assert result.sap_building_parts[0].roof_insulation_location == "Unknown"
+
+    def test_roof_insulation_thickness_none(self, result: EpcPropertyData) -> None:
+        assert result.sap_building_parts[0].roof_insulation_thickness is None
+
+
+PDF_PATH_5 = os.path.join(
+    os.path.dirname(__file__), "fixtures", "ExampleSiteNotes_5.pdf"
+)
+
+
+class TestPdfToEpcPropertyDataFixture5:
+    @pytest.fixture
+    def result(self) -> EpcPropertyData:
+        with open(PDF_PATH_5, "rb") as f:
+            pdf_bytes = f.read()
+        site_notes = PasHubRdSapSiteNotesExtractor(
+            pdf_to_text_list(pdf_bytes)
+        ).extract()
+        return EpcPropertyDataMapper.from_site_notes(site_notes)
+
+    def test_cfl_bulb_count(self, result: EpcPropertyData) -> None:
+        assert result.cfl_fixed_lighting_bulbs_count == 2
+
+    def test_secondary_heating_type(self, result: EpcPropertyData) -> None:
+        assert (
+            result.sap_heating.secondary_heating_type
+            == "Panel, convector or radiant heaters"
+        )
+
+    def test_electric_shower_outlet_type(self, result: EpcPropertyData) -> None:
+        assert result.sap_heating.shower_outlets is not None
+        assert (
+            result.sap_heating.shower_outlets.shower_outlet.shower_outlet_type
+            == "Electric Shower"
+        )
+
+
+PDF_PATH_6 = os.path.join(
+    os.path.dirname(__file__), "fixtures", "ExampleSiteNotes_6.pdf"
+)
+
+
+class TestPdfToEpcPropertyDataFixture6:
+    @pytest.fixture
+    def result(self) -> EpcPropertyData:
+        with open(PDF_PATH_6, "rb") as f:
+            pdf_bytes = f.read()
+        site_notes = PasHubRdSapSiteNotesExtractor(
+            pdf_to_text_list(pdf_bytes)
+        ).extract()
+        return EpcPropertyDataMapper.from_site_notes(site_notes)
+
+    def test_party_wall_construction(self, result: EpcPropertyData) -> None:
+        assert (
+            result.sap_building_parts[0].party_wall_construction
+            == "Solid Masonry, Timber Frame, or System Built"
+        )

backend/documents_parser/tests/test_extractor.py

@@ -0,0 +1,799 @@
+import json
+import os
+from datetime import date
+
+import pytest
+
+from backend.documents_parser.extractor import PasHubRdSapSiteNotesExtractor
+from datatypes.epc.surveys.pashub_rdsap_site_notes import (
+    BuildingConstruction,
+    BuildingMeasurements,
+    Conservatories,
+    CustomerResponse,
+    ExtensionConstruction,
+    ExtensionMeasurements,
+    ExtensionRoofSpace,
+    FloorConstruction,
+    FloorMeasurement,
+    General,
+    HeatingAndHotWater,
+    InspectionMetadata,
+    MainBuildingConstruction,
+    MainBuildingMeasurements,
+    MainHeating,
+    Renewables,
+    RoomCountElements,
+    RoofSpace,
+    RoofSpaceDetail,
+    SecondaryHeating,
+    Shower,
+    SurveyAddendum,
+    Ventilation,
+    WaterHeating,
+    WaterUse,
+)
+
+FIXTURES = os.path.join(os.path.dirname(__file__), "fixtures")
+
+
+def load_text_fixture() -> list[str]:
+    with open(os.path.join(FIXTURES, "site_notes_example_text.json")) as f:
+        return json.load(f)
+
+
+def load_text_fixture_2() -> list[str]:
+    with open(os.path.join(FIXTURES, "site_notes_example_2_text.json")) as f:
+        return json.load(f)
+
+
+def load_text_fixture_3() -> list[str]:
+    with open(os.path.join(FIXTURES, "site_notes_example_3_text.json")) as f:
+        return json.load(f)
+
+
+def load_text_fixture_4() -> list[str]:
+    with open(os.path.join(FIXTURES, "site_notes_example_4_text.json")) as f:
+        return json.load(f)
+
+
+def load_text_fixture_5() -> list[str]:
+    with open(os.path.join(FIXTURES, "site_notes_example_5_text.json")) as f:
+        return json.load(f)
+
+
+def load_text_fixture_6() -> list[str]:
+    with open(os.path.join(FIXTURES, "site_notes_example_6_text.json")) as f:
+        return json.load(f)
+
+
+class TestInspectionMetadata:
+    def test_full_inspection_metadata(self) -> None:
+        result = PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract_inspection_metadata()
+        assert result == InspectionMetadata(
+            inspection_surveyor="Benjamin Burke",
+            email_address="ben@mbsolutionsgroup.co.uk",
+            report_reference="6EA2A86D-94CE-4792-8D49-AB495C744EDD",
+            created_on="2025-11-10",
+            date_of_inspection=date(2025, 9, 25),
+            property_address="40, Abbey Place, Crewe, Cheshire, CW1 4JR",
+            property_photo=True,
+        )
+
+
+class TestGeneral:
+    @pytest.fixture
+    def general(self) -> General:
+        return PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract_general()
+
+    def test_epc_checked_before_assessment(self, general: General) -> None:
+        assert general.epc_checked_before_assessment is True
+
+    def test_epc_exists_at_point_of_assessment(self, general: General) -> None:
+        assert general.epc_exists_at_point_of_assessment is False
+
+    def test_inspection_date(self, general: General) -> None:
+        assert general.inspection_date == date(2025, 9, 25)
+
+    def test_transaction_type(self, general: General) -> None:
+        assert general.transaction_type == "Grant-Scheme (ECO, RHI, etc.)"
+
+    def test_tenure(self, general: General) -> None:
+        assert general.tenure == "Rented Social"
+
+    def test_property_type(self, general: General) -> None:
+        assert general.property_type == "House"
+
+    def test_detachment_type(self, general: General) -> None:
+        assert general.detachment_type == "Mid-terrace"
+
+    def test_number_of_storeys(self, general: General) -> None:
+        assert general.number_of_storeys == 2
+
+    def test_number_of_extensions(self, general: General) -> None:
+        assert general.number_of_extensions == 1
+
+    def test_electricity_smart_meter(self, general: General) -> None:
+        assert general.electricity_smart_meter is True
+
+    def test_mains_gas_available(self, general: General) -> None:
+        assert general.mains_gas_available is True
+
+    def test_measurements_location(self, general: General) -> None:
+        assert general.measurements_location == "Internal"
+
+    def test_full_general(self, general: General) -> None:
+        assert general == General(
+            epc_checked_before_assessment=True,
+            epc_exists_at_point_of_assessment=False,
+            inspection_date=date(2025, 9, 25),
+            transaction_type="Grant-Scheme (ECO, RHI, etc.)",
+            tenure="Rented Social",
+            property_type="House",
+            detachment_type="Mid-terrace",
+            number_of_storeys=2,
+            terrain_type="Suburban",
+            number_of_extensions=1,
+            electricity_smart_meter=True,
+            electric_meter_type="Single",
+            dwelling_export_capable=True,
+            mains_gas_available=True,
+            gas_smart_meter=True,
+            gas_meter_accessible=True,
+            measurements_location="Internal",
+        )
+
+
+class TestGeneralNoExtensions:
+    @pytest.fixture
+    def general(self) -> General:
+        return PasHubRdSapSiteNotesExtractor(load_text_fixture_2()).extract_general()
+
+    def test_number_of_extensions_when_no_extensions(self, general: General) -> None:
+        assert general.number_of_extensions == 0
+
+
+class TestBuildingConstruction:
+    @pytest.fixture
+    def construction(self) -> BuildingConstruction:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture()
+        ).extract_building_construction()
+
+    def test_main_building_wall_u_value_known_is_false(
+        self, construction: BuildingConstruction
+    ) -> None:
+        assert construction.main_building.wall_u_value_known is False
+
+    def test_main_building_wall_thickness_mm(
+        self, construction: BuildingConstruction
+    ) -> None:
+        assert construction.main_building.wall_thickness_mm == 310
+
+    def test_main_building_filled_cavity_indicators_present(
+        self, construction: BuildingConstruction
+    ) -> None:
+        assert (
+            construction.main_building.filled_cavity_indicators
+            == "evidence of cavity fill drill holes"
+        )
+
+    def test_extension_filled_cavity_indicators_absent(
+        self, construction: BuildingConstruction
+    ) -> None:
+        assert construction.extensions is not None
+        assert construction.extensions[0].filled_cavity_indicators is None
+
+    def test_one_extension(self, construction: BuildingConstruction) -> None:
+        assert construction.extensions is not None
+        assert len(construction.extensions) == 1
+
+    def test_extension_id(self, construction: BuildingConstruction) -> None:
+        assert construction.extensions is not None
+        assert construction.extensions[0].id == 1
+
+    def test_full_building_construction(
+        self, construction: BuildingConstruction
+    ) -> None:
+        assert construction == BuildingConstruction(
+            main_building=MainBuildingConstruction(
+                age_range="1950-1966",
+                age_indicators="local knowledge, enquiries of owner",
+                walls_construction_type="Cavity",
+                cavity_construction_indicators="wall thickness over 270 mm",
+                walls_insulation_type="Filled Cavity",
+                filled_cavity_indicators="evidence of cavity fill drill holes",
+                thermal_conductivity_of_wall_insulation="Unknown",
+                wall_u_value_known=False,
+                wall_thickness_mm=310,
+                party_wall_construction_type="Cavity Masonry, Filled",
+            ),
+            floor=FloorConstruction(
+                floor_type="Ground Floor",
+                floor_construction="Solid",
+                floor_insulation_type="As Built",
+                floor_u_value_known=False,
+            ),
+            extensions=[
+                ExtensionConstruction(
+                    id=1,
+                    age_range="2003-2006",
+                    age_indicators="local knowledge, enquiries of owner",
+                    walls_construction_type="Cavity",
+                    cavity_construction_indicators="wall thickness over 270 mm",
+                    walls_insulation_type="As built",
+                    thermal_conductivity_of_wall_insulation="Unknown",
+                    wall_u_value_known=False,
+                    wall_thickness_mm=310,
+                    party_wall_construction_type="Cavity Masonry, Filled",
+                    filled_cavity_indicators=None,
+                )
+            ],
+        )
+
+
+class TestBuildingMeasurements:
+    @pytest.fixture
+    def measurements(self) -> BuildingMeasurements:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture()
+        ).extract_building_measurements()
+
+    def test_main_building_has_two_floors(
+        self, measurements: BuildingMeasurements
+    ) -> None:
+        assert len(measurements.main_building.floors) == 2
+
+    def test_main_building_floor_area(
+        self, measurements: BuildingMeasurements
+    ) -> None:
+        assert measurements.main_building.floors[0].area_m2 == 35.68
+
+    def test_integer_token_parses_to_float(
+        self, measurements: BuildingMeasurements
+    ) -> None:
+        # "11" in the PDF (no decimal) should parse to 11.0
+        assert measurements.main_building.floors[1].heat_loss_perimeter_m == 11.0
+
+    def test_extension_measurements_present(
+        self, measurements: BuildingMeasurements
+    ) -> None:
+        assert measurements.extensions is not None
+        assert len(measurements.extensions) == 1
+
+    def test_extension_id(self, measurements: BuildingMeasurements) -> None:
+        assert measurements.extensions is not None
+        assert measurements.extensions[0].id == 1
+
+    def test_full_building_measurements(
+        self, measurements: BuildingMeasurements
+    ) -> None:
+        assert measurements == BuildingMeasurements(
+            main_building=MainBuildingMeasurements(
+                floors=[
+                    FloorMeasurement(
+                        name="Floor 1",
+                        area_m2=35.68,
+                        height_m=2.19,
+                        heat_loss_perimeter_m=13.44,
+                        pwl_m=10.62,
+                    ),
+                    FloorMeasurement(
+                        name="Floor 0",
+                        area_m2=35.68,
+                        height_m=2.17,
+                        heat_loss_perimeter_m=11.0,
+                        pwl_m=10.62,
+                    ),
+                ]
+            ),
+            extensions=[
+                ExtensionMeasurements(
+                    id=1,
+                    floors=[
+                        FloorMeasurement(
+                            name="Floor 0",
+                            area_m2=3.8,
+                            height_m=2.0,
+                            heat_loss_perimeter_m=5.7,
+                            pwl_m=0.0,
+                        )
+                    ],
+                )
+            ],
+        )
+
+
+class TestRoofSpace:
+    @pytest.fixture
+    def roof_space(self) -> RoofSpace:
+        return PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract_roof_space()
+
+    def test_main_building_insulation_thickness_mm(
+        self, roof_space: RoofSpace
+    ) -> None:
+        assert roof_space.main_building.insulation_thickness_mm == 100
+
+    def test_main_building_insulation_thickness_string_absent(
+        self, roof_space: RoofSpace
+    ) -> None:
+        assert roof_space.main_building.insulation_thickness is None
+
+    def test_main_building_rooms_in_roof(self, roof_space: RoofSpace) -> None:
+        assert roof_space.main_building.rooms_in_roof is False
+
+    def test_main_building_roof_u_value_known(self, roof_space: RoofSpace) -> None:
+        assert roof_space.main_building.roof_u_value_known is False
+
+    def test_extension_uses_string_thickness(self, roof_space: RoofSpace) -> None:
+        assert roof_space.extensions is not None
+        assert roof_space.extensions[0].insulation_thickness == "As built"
+        assert roof_space.extensions[0].insulation_thickness_mm is None
+
+    def test_full_roof_space(self, roof_space: RoofSpace) -> None:
+        assert roof_space == RoofSpace(
+            main_building=RoofSpaceDetail(
+                construction_type="Pitched roof (Slates or tiles), Access to loft",
+                insulation_at="Joists",
+                roof_u_value_known=False,
+                cavity_wall_construction_indicators="cavity visible in roof space",
+                rooms_in_roof=False,
+                insulation_thickness_mm=100,
+                insulation_thickness=None,
+            ),
+            extensions=[
+                ExtensionRoofSpace(
+                    id=1,
+                    construction_type="Pitched roof, Sloping ceiling",
+                    insulation_at="Sloping ceiling insulation",
+                    roof_u_value_known=False,
+                    cavity_wall_construction_indicators="No indicator of construction visible",
+                    rooms_in_roof=False,
+                    insulation_thickness_mm=None,
+                    insulation_thickness="As built",
+                )
+            ],
+        )
+
+
+class TestWindows:
+    @pytest.fixture
+    def windows(self) -> list:
+        return PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract_windows()
+
+    def test_window_count(self, windows: list) -> None:
+        assert len(windows) == 8
+
+    def test_ids_are_sequential(self, windows: list) -> None:
+        assert [w.id for w in windows] == list(range(1, 9))
+
+    def test_first_window_location(self, windows: list) -> None:
+        assert windows[0].location == "Main Building"
+
+    def test_extension_window_location(self, windows: list) -> None:
+        assert windows[3].location == "Extension 1"
+
+    def test_height_parses_to_float(self, windows: list) -> None:
+        assert windows[0].height_m == 1.2
+
+    def test_draught_proofed_true(self, windows: list) -> None:
+        assert windows[0].draught_proofed is True
+
+    def test_permanent_shutters_false(self, windows: list) -> None:
+        assert windows[0].permanent_shutters is False
+
+    def test_first_window_full(self, windows: list) -> None:
+        from datatypes.epc.surveys.pashub_rdsap_site_notes import Window
+        assert windows[0] == Window(
+            id=1,
+            location="Main Building",
+            wall_type="External wall",
+            glazing_type="Double glazing, Unknown install date",
+            window_type="Window",
+            frame_type="Wooden or PVC",
+            glazing_gap="16 mm or more",
+            draught_proofed=True,
+            permanent_shutters=False,
+            height_m=1.2,
+            width_m=2.3,
+            orientation="North West",
+        )
+
+
+class TestWaterHeatingCylinderThickness:
+    @pytest.fixture
+    def hhw(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_2()
+        ).extract_heating_and_hot_water()
+
+    @pytest.fixture
+    def hhw_no_cylinder(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture()
+        ).extract_heating_and_hot_water()
+
+    def test_cylinder_insulation_thickness_mm(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.water_heating.insulation_thickness_mm == 38
+
+    def test_cylinder_insulation_thickness_mm_absent(self, hhw_no_cylinder: HeatingAndHotWater) -> None:
+        assert hhw_no_cylinder.water_heating.insulation_thickness_mm is None
+
+    def test_cylinder_size(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.water_heating.cylinder_size == "Normal (90-130 litres)"
+
+
+class TestImmersionType:
+    @pytest.fixture
+    def hhw(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_3()
+        ).extract_heating_and_hot_water()
+
+    def test_immersion_type(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.water_heating.immersion_type == "Dual"
+
+
+class TestCylinderThermostat:
+    @pytest.fixture
+    def hhw(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_3()
+        ).extract_heating_and_hot_water()
+
+    def test_has_thermostat_true(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.water_heating.has_thermostat is True
+
+
+class TestSecondaryHeating:
+    @pytest.fixture
+    def hhw(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_2()
+        ).extract_heating_and_hot_water()
+
+    @pytest.fixture
+    def hhw_no_secondary(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture()
+        ).extract_heating_and_hot_water()
+
+    def test_secondary_system(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.secondary_heating.secondary_system == "Open fire in grate"
+
+    def test_secondary_system_absent(self, hhw_no_secondary: HeatingAndHotWater) -> None:
+        assert hhw_no_secondary.secondary_heating.secondary_system is None
+
+
+class TestHeatingAndHotWater:
+    @pytest.fixture
+    def hhw(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture()
+        ).extract_heating_and_hot_water()
+
+    def test_product_id_parses_to_int(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.main_heating.product_id == 16839
+
+    def test_summer_efficiency_parses_to_float(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.main_heating.summer_efficiency == 0.0
+
+    def test_condensing_true(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.main_heating.condensing is True
+
+    def test_fghrs_false(self, hhw: HeatingAndHotWater) -> None:
+        # multi-line label
+        assert hhw.main_heating.flue_gas_heat_recovery_system is False
+
+    def test_secondary_fuel(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.secondary_heating.secondary_fuel == "No Secondary Heating"
+
+    def test_water_heating_no_cylinder(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.water_heating.cylinder_size == "No Cylinder"
+        assert hhw.water_heating.insulation_type is None
+        assert hhw.water_heating.has_thermostat is None
+
+    def test_full_heating_and_hot_water(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw == HeatingAndHotWater(
+            main_heating=MainHeating(
+                selection_method="PCDF Search",
+                system_type="Boiler with radiators or underfloor heating",
+                product_id=16839,
+                manufacturer="Vaillant",
+                model="ecoTEC pro 28",
+                orig_manufacturer="Vaillant",
+                fuel="Mains gas",
+                summer_efficiency=0.0,
+                type="Combi",
+                condensing=True,
+                year="2005 - 2015",
+                mount="Wall",
+                open_flue="Room-sealed",
+                fan_assist=True,
+                status="Normal status for an actual product",
+                central_heating_pump_age="Unknown",
+                controls="Programmer, room thermostat and TRVs",
+                flue_gas_heat_recovery_system=False,
+                weather_compensator=False,
+                emitter="Radiators",
+                emitter_temperature="Unknown",
+            ),
+            secondary_heating=SecondaryHeating(
+                secondary_fuel="No Secondary Heating",
+            ),
+            water_heating=WaterHeating(
+                type="Regular",
+                system="From main heating 1",
+                cylinder_size="No Cylinder",
+                cylinder_measured_heat_loss=None,
+                insulation_type=None,
+                insulation_thickness_mm=None,
+                has_thermostat=None,
+            ),
+        )
+
+
+class TestVentilation:
+    @pytest.fixture
+    def ventilation(self) -> Ventilation:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture()
+        ).extract_ventilation()
+
+    def test_ventilation_type(self, ventilation: Ventilation) -> None:
+        assert ventilation.ventilation_type == "Mechanical Extract - Decentralised"
+
+    def test_number_of_open_flues(self, ventilation: Ventilation) -> None:
+        assert ventilation.number_of_open_flues == 0
+
+    def test_ventilation_in_pcdf_database(self, ventilation: Ventilation) -> None:
+        assert ventilation.ventilation_in_pcdf_database is False
+
+    def test_full_ventilation(self, ventilation: Ventilation) -> None:
+        assert ventilation == Ventilation(
+            ventilation_type="Mechanical Extract - Decentralised",
+            has_fixed_air_conditioning=False,
+            number_of_open_flues=0,
+            number_of_closed_flues=0,
+            number_of_boiler_flues=0,
+            number_of_other_flues=0,
+            number_of_extract_fans=0,
+            number_of_passive_vents=0,
+            number_of_flueless_gas_fires=0,
+            pressure_test="No test",
+            draught_lobby=False,
+            ventilation_in_pcdf_database=False,
+        )
+
+
+class TestConservatories:
+    def test_full_conservatories(self) -> None:
+        result = PasHubRdSapSiteNotesExtractor(
+            load_text_fixture()
+        ).extract_conservatories()
+        assert result == Conservatories(has_conservatory=False)
+
+
+class TestRenewables:
+    def test_number_of_pv_batteries_none_string_becomes_zero(self) -> None:
+        result = PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract_renewables()
+        assert result.number_of_pv_batteries == 0
+
+    def test_full_renewables(self) -> None:
+        result = PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract_renewables()
+        assert result == Renewables(
+            wind_turbines=False,
+            solar_hot_water=False,
+            photovoltaic_array=False,
+            number_of_pv_batteries=0,
+            hydro=False,
+        )
+
+
+class TestRenewablesPvConnection:
+    @pytest.fixture
+    def renewables(self) -> Renewables:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_3()
+        ).extract_renewables()
+
+    def test_pv_connection(self, renewables: Renewables) -> None:
+        assert renewables.pv_connection == "Connected to dwellings electricity meter"
+
+    def test_percent_roof_covered_pv(self, renewables: Renewables) -> None:
+        assert renewables.percent_roof_covered_pv == 45
+
+
+class TestRoomCountElements:
+    @pytest.fixture
+    def rce(self) -> RoomCountElements:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture()
+        ).extract_room_count_elements()
+
+    def test_habitable_rooms(self, rce: RoomCountElements) -> None:
+        assert rce.number_of_habitable_rooms == 3
+
+    def test_heated_rooms_null(self, rce: RoomCountElements) -> None:
+        assert rce.number_of_heated_rooms is None
+
+    def test_full_room_count_elements(self, rce: RoomCountElements) -> None:
+        assert rce == RoomCountElements(
+            number_of_habitable_rooms=3,
+            any_unheated_rooms=False,
+            number_of_heated_rooms=None,
+            number_of_external_doors=2,
+            number_of_insulated_external_doors=0,
+            number_of_draughtproofed_external_doors=2,
+            number_of_open_chimneys=0,
+            number_of_blocked_chimneys=0,
+            number_of_fixed_incandescent_bulbs=4,
+            exact_led_cfl_count_known=True,
+            number_of_fixed_led_bulbs=0,
+            number_of_fixed_cfl_bulbs=1,
+            waste_water_heat_recovery="None",
+        )
+
+
+class TestWaterUse:
+    def test_full_water_use(self) -> None:
+        result = PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract_water_use()
+        assert result == WaterUse(
+            number_of_baths=1,
+            number_of_special_features=0,
+            showers=[Shower(id=1, outlet_type="Non-Electric Shower")],
+        )
+
+
+class TestCustomerResponse:
+    def test_full_customer_response(self) -> None:
+        result = PasHubRdSapSiteNotesExtractor(
+            load_text_fixture()
+        ).extract_customer_response()
+        assert result == CustomerResponse(
+            customer_present=True,
+            willing_to_answer_satisfaction_survey=False,
+        )
+
+
+class TestExtract:
+    def test_full_extract(self) -> None:
+        result = PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract()
+        assert result.inspection_metadata.inspection_surveyor == "Benjamin Burke"
+        assert result.general.inspection_date == date(2025, 9, 25)
+        assert result.building_construction.main_building.wall_thickness_mm == 310
+        assert result.building_measurements.main_building.floors[0].area_m2 == 35.68
+        assert result.roof_space.main_building.insulation_thickness_mm == 100
+        assert len(result.windows) == 8
+        assert result.heating_and_hot_water.main_heating.product_id == 16839
+        assert result.ventilation.ventilation_type == "Mechanical Extract - Decentralised"
+        assert result.conservatories.has_conservatory is False
+        assert result.renewables.number_of_pv_batteries == 0
+        assert result.room_count_elements.number_of_habitable_rooms == 3
+        assert result.water_use.number_of_baths == 1
+        assert result.customer_response.customer_present is True
+        assert result.addendum.addendum == "None"
+
+
+class TestSurveyAddendum:
+    def test_hard_to_treat_flags(self) -> None:
+        result = PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract_addendum()
+        assert result.hard_to_treat_cavity_access_issues is False
+        assert result.hard_to_treat_cavity_high_exposure is False
+        assert result.hard_to_treat_cavity_narrow_cavities is False
+
+    def test_full_addendum(self) -> None:
+        result = PasHubRdSapSiteNotesExtractor(load_text_fixture()).extract_addendum()
+        assert result == SurveyAddendum(
+            addendum="None",
+            related_party_disclosure="No related party",
+            hard_to_treat_cavity_access_issues=False,
+            hard_to_treat_cavity_high_exposure=False,
+            hard_to_treat_cavity_narrow_cavities=False,
+        )
+
+
+# --- fixture 4: heat pump, factory-fitted cylinder, blocked loft ---
+
+
+class TestCylinderInsulationType:
+    @pytest.fixture
+    def hhw(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_4()
+        ).extract_heating_and_hot_water()
+
+    def test_insulation_type_extracted(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.water_heating.insulation_type == "Factory fitted"
+
+    def test_insulation_thickness_mm(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.water_heating.insulation_thickness_mm == 50
+
+    def test_cylinder_size(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.water_heating.cylinder_size == "Medium (131-170 litres)"
+
+
+class TestHeatPumpFuelExtraction:
+    @pytest.fixture
+    def hhw(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_4()
+        ).extract_heating_and_hot_water()
+
+    def test_fuel_raw_value(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.main_heating.fuel == "Electricity, any tariff"
+
+    def test_system_type(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.main_heating.system_type == "Heat pump with radiators or underfloor heating"
+
+
+class TestRoofSpaceUnknownInsulation:
+    @pytest.fixture
+    def roof_space(self) -> RoofSpace:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_4()
+        ).extract_roof_space()
+
+    def test_insulation_at_unknown(self, roof_space: RoofSpace) -> None:
+        assert roof_space.main_building.insulation_at == "Unknown"
+
+    def test_insulation_thickness_mm_none(self, roof_space: RoofSpace) -> None:
+        assert roof_space.main_building.insulation_thickness_mm is None
+
+    def test_insulation_thickness_str_none(self, roof_space: RoofSpace) -> None:
+        assert roof_space.main_building.insulation_thickness is None
+
+
+class TestCflBulbCount:
+    @pytest.fixture
+    def rce(self) -> RoomCountElements:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_5()
+        ).extract_room_count_elements()
+
+    def test_cfl_count(self, rce: RoomCountElements) -> None:
+        assert rce.number_of_fixed_cfl_bulbs == 2
+
+    def test_led_count(self, rce: RoomCountElements) -> None:
+        assert rce.number_of_fixed_led_bulbs == 7
+
+    def test_incandescent_count(self, rce: RoomCountElements) -> None:
+        assert rce.number_of_fixed_incandescent_bulbs == 1
+
+
+class TestSecondaryHeatingPanel:
+    @pytest.fixture
+    def hhw(self) -> HeatingAndHotWater:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_5()
+        ).extract_heating_and_hot_water()
+
+    def test_secondary_system(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.secondary_heating.secondary_system == "Panel, convector or radiant heaters"
+
+    def test_secondary_fuel(self, hhw: HeatingAndHotWater) -> None:
+        assert hhw.secondary_heating.secondary_fuel == "Electricity"
+
+
+class TestElectricShowerExtraction:
+    @pytest.fixture
+    def wu(self) -> WaterUse:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_5()
+        ).extract_water_use()
+
+    def test_shower_outlet_type(self, wu: WaterUse) -> None:
+        assert wu.showers[0].outlet_type == "Electric Shower"
+
+
+class TestSolidMasonryPartyWall:
+    @pytest.fixture
+    def bc(self) -> BuildingConstruction:
+        return PasHubRdSapSiteNotesExtractor(
+            load_text_fixture_6()
+        ).extract_building_construction()
+
+    def test_party_wall_construction_type(self, bc: BuildingConstruction) -> None:
+        assert (
+            bc.main_building.party_wall_construction_type
+            == "Solid Masonry, Timber Frame, or System Built"
+        )

backend/documents_parser/tests/test_pdf.py

@@ -0,0 +1,31 @@
+import json
+import os
+
+import pytest
+
+from backend.documents_parser.pdf import pdf_to_text_list
+
+PDF_PATH = os.path.join(os.path.dirname(__file__), "fixtures", "ExampleSiteNotes.pdf")
+FIXTURE_PATH = os.path.join(os.path.dirname(__file__), "fixtures", "site_notes_example_text.json")
+
+
+@pytest.fixture
+def pdf_bytes() -> bytes:
+    with open(PDF_PATH, "rb") as f:
+        return f.read()
+
+
+class TestPdfToTextList:
+    def test_returns_list(self, pdf_bytes: bytes) -> None:
+        result = pdf_to_text_list(pdf_bytes)
+        assert isinstance(result, list)
+
+    def test_all_elements_are_strings(self, pdf_bytes: bytes) -> None:
+        result = pdf_to_text_list(pdf_bytes)
+        assert all(isinstance(t, str) for t in result)
+
+    def test_matches_fixture(self, pdf_bytes: bytes) -> None:
+        with open(FIXTURE_PATH) as f:
+            expected = json.load(f)
+        result = pdf_to_text_list(pdf_bytes)
+        assert result == expected

datatypes/epc/domain/epc_property_data.py

@@ -29,18 +29,21 @@ class MainHeatingDetail:
     boiler_flue_type: Optional[int] = None  # TODO: make enum?
     boiler_ignition_type: Optional[int] = None  # TODO: make enum?
     central_heating_pump_age: Optional[int] = None
+    central_heating_pump_age_str: Optional[str] = None  # str from site notes e.g. "Unknown", "Pre 2013"
     main_heating_index_number: Optional[int] = None
     sap_main_heating_code: Optional[int] = None  # TODO: make enum?
     main_heating_number: Optional[int] = None
     main_heating_category: Optional[int] = None
     main_heating_fraction: Optional[int] = None
     main_heating_data_source: Optional[int] = None
+    condensing: Optional[bool] = None
+    weather_compensator: Optional[bool] = None
 
 
 @dataclass
 class ShowerOutlet:
-    shower_wwhrs: int
-    shower_outlet_type: int
+    shower_outlet_type: Union[int, str]
+    shower_wwhrs: Optional[int] = None
 
 
 @dataclass
@@ -54,18 +57,33 @@ class SapHeating:
     instantaneous_wwhrs: InstantaneousWwhrs
     main_heating_details: List[MainHeatingDetail]
     has_fixed_air_conditioning: bool
-    cylinder_size: Optional[int] = (
-        None  # int code from API; not directly available from site notes
+    cylinder_size: Optional[Union[int, str]] = (
+        None  # int code from API; str (e.g. "Normal (90-130 litres)") from site notes
     )
     water_heating_code: Optional[int] = None  # TODO: make enum?
     water_heating_fuel: Optional[int] = None  # TODO: make enum?
     immersion_heating_type: Optional[Union[int, str]] = None  # TODO: make enum?
     shower_outlets: Optional[ShowerOutlets] = None
-    cylinder_insulation_type: Optional[int] = None
+    cylinder_insulation_type: Optional[Union[int, str]] = None
     cylinder_thermostat: Optional[str] = None
     secondary_fuel_type: Optional[int] = None
-    secondary_heating_type: Optional[int] = None
-    cylinder_insulation_thickness: Optional[int] = None
+    secondary_heating_type: Optional[Union[int, str]] = None  # int from API; str from site notes
+    cylinder_insulation_thickness_mm: Optional[int] = None
+
+
+@dataclass
+class SapVentilation:
+    ventilation_type: Optional[str] = None
+    draught_lobby: Optional[bool] = None
+    pressure_test: Optional[str] = None  # str from site notes e.g. "No test"; int in API via mechanical_ventilation
+    open_flues_count: Optional[int] = None
+    closed_flues_count: Optional[int] = None
+    boiler_flues_count: Optional[int] = None
+    other_flues_count: Optional[int] = None
+    extract_fans_count: Optional[int] = None
+    passive_vents_count: Optional[int] = None
+    flueless_gas_fires_count: Optional[int] = None
+    ventilation_in_pcdf_database: Optional[bool] = None
 
 
 @dataclass
@@ -130,7 +148,7 @@ class SapEnergySource:
     wind_turbines_terrain_type: str  # int in API, str (e.g. "Suburban") in site notes
     electricity_smart_meter_present: bool
 
-    pv_connection: Optional[int] = None
+    pv_connection: Optional[Union[int, str]] = None  # int from API; str from site notes
     photovoltaic_supply: Optional[PhotovoltaicSupply] = None
     wind_turbine_details: Optional[WindTurbineDetails] = None
     pv_batteries: Optional[PvBatteries] = None
@@ -200,6 +218,10 @@ class SapBuildingPart:
     flat_roof_insulation_thickness: Optional[Union[str, int]] = (
         None  # TODO: make enum/mapping?
     )
+    floor_type: Optional[str] = None  # str from site notes e.g. "Ground Floor"
+    floor_construction_type: Optional[str] = None  # str from site notes; distinct from floor_construction: int in SapFloorDimension
+    floor_insulation_type_str: Optional[str] = None  # str from site notes e.g. "As Built"
+    floor_u_value_known: Optional[bool] = None
 
     roof_construction: Optional[int] = None
     roof_insulation_location: Optional[Union[int, str]] = (
@@ -235,6 +257,9 @@ class EpcPropertyData:
     inspection_date: date
     tenure: str  # str in site notes; stringified int (e.g. "1") from API
     transaction_type: str  # str in site notes; stringified int from API
+    address_line_1: str
+    postcode: str
+    post_town: str
 
     # Elements
     roofs: List[EnergyElement]
@@ -270,9 +295,6 @@ class EpcPropertyData:
     assessment_type: Optional[str] = None  # not available from site notes
     sap_version: Optional[float] = None  # not available from site notes
     uprn: Optional[int] = None  # not available from site notes
-    address_line_1: Optional[str] = None  # not available from site notes
-    postcode: Optional[str] = None  # not available from site notes
-    post_town: Optional[str] = None  # not available from site notes
     status: Optional[str] = None  # not available from site notes
     window: Optional[EnergyElement] = None  # not available from site notes
     lighting: Optional[EnergyElement] = None  # not available from site notes
@@ -280,6 +302,7 @@ class EpcPropertyData:
     schema_type: Optional[str] = None
     schema_versions_original: Optional[str] = None
     report_type: Optional[str] = None  # TODO: make enum?
+    report_reference: Optional[str] = None
     uprn_source: Optional[str] = None
     address_line_2: Optional[str] = None
     region_code: Optional[str] = None  # TODO: make enum?
@@ -346,3 +369,10 @@ class EpcPropertyData:
     # survey_addendum: Optional[Any] = None # not sure how to handle, skip for now
     fixed_lighting_outlets_count: Optional[int] = None
     low_energy_fixed_lighting_outlets_count: Optional[int] = None
+    # Site-notes-only fields
+    sap_ventilation: Optional[SapVentilation] = None
+    number_of_storeys: Optional[int] = None
+    any_unheated_rooms: Optional[bool] = None
+    waste_water_heat_recovery: Optional[str] = None
+    hydro: Optional[bool] = None
+    photovoltaic_array: Optional[bool] = None

datatypes/epc/domain/mapper.py

@@ -1,5 +1,5 @@
 from datetime import date
-from typing import List, Sequence, Union
+from typing import List, Optional, Sequence, Union
 
 from datatypes.epc.domain.epc_property_data import (
     EnergyElement,
@@ -16,6 +16,7 @@ from datatypes.epc.domain.epc_property_data import (
     SapFloorDimension,
     SapHeating,
     SapRoomInRoof,
+    SapVentilation,
     SapWindow,
     ShowerOutlet,
     ShowerOutlets,
@@ -55,10 +56,14 @@ from datatypes.epc.surveys.pashub_rdsap_site_notes import (
     BuildingMeasurements,
     ExtensionConstruction,
     ExtensionMeasurements,
+    ExtensionRoofSpace,
+    FloorConstruction,
     FloorMeasurement,
     HeatingAndHotWater,
     PasHubRdSapSiteNotes,
+    RoofSpaceDetail,
     Ventilation,
+    WaterUse,
     Window,
 )
 
@@ -78,6 +83,24 @@ class EpcPropertyDataMapper:
     @staticmethod
     def from_site_notes(survey: PasHubRdSapSiteNotes) -> EpcPropertyData:
         general = survey.general
+        metadata = survey.inspection_metadata
+        address_parts = [p.strip() for p in metadata.property_address.split(", ")]
+        postcode = address_parts[-1] if address_parts else ""
+        post_town = (
+            address_parts[-3]
+            if len(address_parts) >= 4
+            else (address_parts[-2] if len(address_parts) >= 3 else "")
+        )
+        address_line_1 = (
+            ", ".join(address_parts[:-3])
+            if len(address_parts) >= 4
+            else (
+                ", ".join(address_parts[:-2])
+                if len(address_parts) >= 3
+                else address_parts[0] if address_parts else ""
+            )
+        )
+
         construction = survey.building_construction
         measurements = survey.building_measurements
         heating = survey.heating_and_hot_water
@@ -85,13 +108,23 @@ class EpcPropertyDataMapper:
         renewables = survey.renewables
         room_counts = survey.room_count_elements
 
-        sap_building_parts = [_map_main_building_part(construction, measurements)]
+        roof_space = survey.roof_space
+        sap_building_parts = [
+            _map_main_building_part(
+                construction, measurements, roof_space.main_building
+            )
+        ]
         if construction.extensions and measurements.extensions:
             for ext_c in construction.extensions:
-                matching = [m for m in measurements.extensions if m.id == ext_c.id]
-                if matching:
+                matching_m = [m for m in measurements.extensions if m.id == ext_c.id]
+                matching_r = [
+                    r for r in (roof_space.extensions or []) if r.id == ext_c.id
+                ]
+                if matching_m:
                     sap_building_parts.append(
-                        _map_extension_building_part(ext_c, matching[0])
+                        _map_extension_building_part(
+                            ext_c, matching_m[0], matching_r[0] if matching_r else None
+                        )
                     )
 
         total_floor_area = round(
@@ -101,11 +134,11 @@ class EpcPropertyDataMapper:
                 for floor in part.sap_floor_dimensions
             ),
             2,
-        )
+        )  # TODO: verify that is the correct approach
 
         return EpcPropertyData(
             dwelling_type=f"{general.detachment_type} {general.property_type.lower()}",
-            inspection_date=date.fromisoformat(general.inspection_date),
+            inspection_date=general.inspection_date,
             tenure=general.tenure,
             transaction_type=general.transaction_type,
             roofs=[],
@@ -113,7 +146,7 @@ class EpcPropertyDataMapper:
             floors=[],
             main_heating=[],
             door_count=room_counts.number_of_external_doors,
-            sap_heating=_map_sap_heating(heating, ventilation),
+            sap_heating=_map_sap_heating(heating, ventilation, survey.water_use),
             sap_windows=[_map_sap_window(w) for w in survey.windows],
             sap_energy_source=SapEnergySource(
                 mains_gas=general.mains_gas_available,
@@ -124,6 +157,16 @@ class EpcPropertyDataMapper:
                 is_dwelling_export_capable=general.dwelling_export_capable,
                 wind_turbines_terrain_type=general.terrain_type,
                 electricity_smart_meter_present=general.electricity_smart_meter,
+                pv_connection=renewables.pv_connection,
+                photovoltaic_supply=(
+                    PhotovoltaicSupply(
+                        none_or_no_details=PhotovoltaicSupplyNoneOrNoDetails(
+                            percent_roof_area=renewables.percent_roof_covered_pv,
+                        )
+                    )
+                    if renewables.percent_roof_covered_pv is not None
+                    else None
+                ),
             ),
             sap_building_parts=sap_building_parts,
             solar_water_heating=renewables.solar_hot_water,
@@ -145,6 +188,16 @@ class EpcPropertyDataMapper:
             has_conservatory=survey.conservatories.has_conservatory,
             blocked_chimneys_count=room_counts.number_of_blocked_chimneys,
             draughtproofed_door_count=room_counts.number_of_draughtproofed_external_doors,
+            address_line_1=address_line_1,
+            post_town=post_town,
+            postcode=postcode,
+            report_reference=metadata.report_reference,
+            number_of_storeys=general.number_of_storeys,
+            any_unheated_rooms=room_counts.any_unheated_rooms,
+            waste_water_heat_recovery=room_counts.waste_water_heat_recovery,
+            hydro=renewables.hydro,
+            photovoltaic_array=renewables.photovoltaic_array,
+            sap_ventilation=_map_sap_ventilation(ventilation),
         )
 
     @staticmethod
@@ -227,7 +280,7 @@ class EpcPropertyDataMapper:
                 cylinder_thermostat=None,
                 secondary_fuel_type=None,
                 secondary_heating_type=None,
-                cylinder_insulation_thickness=None,
+                cylinder_insulation_thickness_mm=None,
             ),
             sap_windows=[],
             sap_energy_source=SapEnergySource(
@@ -368,7 +421,7 @@ class EpcPropertyDataMapper:
                 cylinder_thermostat=schema.sap_heating.cylinder_thermostat,
                 secondary_fuel_type=schema.sap_heating.secondary_fuel_type,
                 secondary_heating_type=schema.sap_heating.secondary_heating_type,
-                cylinder_insulation_thickness=schema.sap_heating.cylinder_insulation_thickness,
+                cylinder_insulation_thickness_mm=schema.sap_heating.cylinder_insulation_thickness,
             ),
             sap_windows=[],
             sap_energy_source=SapEnergySource(
@@ -509,7 +562,7 @@ class EpcPropertyDataMapper:
                 cylinder_thermostat=schema.sap_heating.cylinder_thermostat,
                 secondary_fuel_type=schema.sap_heating.secondary_fuel_type,
                 secondary_heating_type=schema.sap_heating.secondary_heating_type,
-                cylinder_insulation_thickness=schema.sap_heating.cylinder_insulation_thickness,
+                cylinder_insulation_thickness_mm=schema.sap_heating.cylinder_insulation_thickness,
             ),
             sap_windows=[],
             sap_energy_source=SapEnergySource(
@@ -657,7 +710,7 @@ class EpcPropertyDataMapper:
                 cylinder_thermostat=schema.sap_heating.cylinder_thermostat,
                 secondary_fuel_type=schema.sap_heating.secondary_fuel_type,
                 secondary_heating_type=schema.sap_heating.secondary_heating_type,
-                cylinder_insulation_thickness=schema.sap_heating.cylinder_insulation_thickness,
+                cylinder_insulation_thickness_mm=schema.sap_heating.cylinder_insulation_thickness,
             ),
             # 19.0 has no per-window list; individual window fields are at schema root
             sap_windows=[],
@@ -808,7 +861,7 @@ class EpcPropertyDataMapper:
                 cylinder_thermostat=schema.sap_heating.cylinder_thermostat,
                 secondary_fuel_type=schema.sap_heating.secondary_fuel_type,
                 secondary_heating_type=schema.sap_heating.secondary_heating_type,
-                cylinder_insulation_thickness=schema.sap_heating.cylinder_insulation_thickness,
+                cylinder_insulation_thickness_mm=schema.sap_heating.cylinder_insulation_thickness,
             ),
             # 20.0.0 SapWindow lacks frame/gap/draught fields present in later schemas
             sap_windows=[
@@ -986,7 +1039,7 @@ class EpcPropertyDataMapper:
                 cylinder_thermostat=schema.sap_heating.cylinder_thermostat,
                 secondary_fuel_type=schema.sap_heating.secondary_fuel_type,
                 secondary_heating_type=schema.sap_heating.secondary_heating_type,
-                cylinder_insulation_thickness=schema.sap_heating.cylinder_insulation_thickness,
+                cylinder_insulation_thickness_mm=schema.sap_heating.cylinder_insulation_thickness,
             ),
             sap_windows=[
                 SapWindow(
@@ -1219,7 +1272,7 @@ class EpcPropertyDataMapper:
                 cylinder_thermostat=schema.sap_heating.cylinder_thermostat,
                 secondary_fuel_type=schema.sap_heating.secondary_fuel_type,
                 secondary_heating_type=schema.sap_heating.secondary_heating_type,
-                cylinder_insulation_thickness=schema.sap_heating.cylinder_insulation_thickness,
+                cylinder_insulation_thickness_mm=schema.sap_heating.cylinder_insulation_thickness,
             ),
             # SAP windows
             sap_windows=[
@@ -1412,16 +1465,33 @@ def _map_floor_dimensions(floors: List[FloorMeasurement]) -> List[SapFloorDimens
             total_floor_area_m2=floor.area_m2,
             party_wall_length_m=floor.pwl_m,
             heat_loss_perimeter_m=floor.heat_loss_perimeter_m,
+            floor=int(floor.name.split()[-1]),
         )
         for floor in floors
     ]
 
 
+def _map_roof(
+    roof: Optional[Union[RoofSpaceDetail, ExtensionRoofSpace]],
+) -> tuple[Optional[str], Optional[Union[str, int]]]:
+    if roof is None:
+        return None, None
+    thickness: Optional[Union[str, int]] = (
+        roof.insulation_thickness_mm
+        if roof.insulation_thickness_mm is not None
+        else roof.insulation_thickness
+    )
+    return roof.insulation_at or None, thickness
+
+
 def _map_main_building_part(
     construction: BuildingConstruction,
     measurements: BuildingMeasurements,
+    roof: RoofSpaceDetail,
 ) -> SapBuildingPart:
     main = construction.main_building
+    floor = construction.floor
+    roof_location, roof_thickness = _map_roof(roof)
     return SapBuildingPart(
         identifier="main",
         construction_age_band=_extract_age_band(main.age_range),
@@ -1431,13 +1501,21 @@ def _map_main_building_part(
         party_wall_construction=main.party_wall_construction_type,
         sap_floor_dimensions=_map_floor_dimensions(measurements.main_building.floors),
         wall_thickness_mm=main.wall_thickness_mm,
+        roof_insulation_location=roof_location,
+        roof_insulation_thickness=roof_thickness,
+        floor_type=floor.floor_type,
+        floor_construction_type=floor.floor_construction,
+        floor_insulation_type_str=floor.floor_insulation_type,
+        floor_u_value_known=floor.floor_u_value_known,
     )
 
 
 def _map_extension_building_part(
     ext_c: ExtensionConstruction,
     ext_m: ExtensionMeasurements,
+    roof: Optional[ExtensionRoofSpace],
 ) -> SapBuildingPart:
+    roof_location, roof_thickness = _map_roof(roof)
     return SapBuildingPart(
         identifier=f"extension_{ext_c.id}",
         construction_age_band=_extract_age_band(ext_c.age_range),
@@ -1447,6 +1525,8 @@ def _map_extension_building_part(
         party_wall_construction=ext_c.party_wall_construction_type,
         sap_floor_dimensions=_map_floor_dimensions(ext_m.floors),
         wall_thickness_mm=ext_c.wall_thickness_mm,
+        roof_insulation_location=roof_location,
+        roof_insulation_thickness=roof_thickness,
     )
 
 
@@ -1467,7 +1547,7 @@ def _map_sap_window(window: Window) -> SapWindow:
 
 
 def _map_sap_heating(
-    heating: HeatingAndHotWater, ventilation: Ventilation
+    heating: HeatingAndHotWater, ventilation: Ventilation, water_use: WaterUse
 ) -> SapHeating:
     main = heating.main_heating
     secondary = heating.secondary_heating
@@ -1479,18 +1559,61 @@ def _map_sap_heating(
         None if secondary.secondary_fuel == "No Secondary Heating" else None
     )
 
+    shower_outlets = (
+        ShowerOutlets(
+            shower_outlet=ShowerOutlet(
+                shower_outlet_type=water_use.showers[0].outlet_type,
+            )
+        )
+        if water_use.showers
+        else None
+    )
+
+    _ELECTRIC_SYSTEM_TYPES = {"electric storage heaters", "electric underfloor heating"}
+    _raw_fuel = main.fuel.split(", ")[0] if main.fuel else ""
+    fuel_type = (
+        _raw_fuel
+        if _raw_fuel
+        else ("Electricity" if main.system_type.lower() in _ELECTRIC_SYSTEM_TYPES else _raw_fuel)
+    )
+
     return SapHeating(
         instantaneous_wwhrs=InstantaneousWwhrs(),
         main_heating_details=[
             MainHeatingDetail(
                 has_fghrs=main.flue_gas_heat_recovery_system,
-                main_fuel_type=main.fuel,
+                main_fuel_type=fuel_type,
                 heat_emitter_type=main.emitter,
                 emitter_temperature=main.emitter_temperature,
                 fan_flue_present=main.fan_assist,
                 main_heating_control=main.controls,
+                condensing=main.condensing,
+                weather_compensator=main.weather_compensator,
+                central_heating_pump_age_str=main.central_heating_pump_age,
             )
         ],
         has_fixed_air_conditioning=ventilation.has_fixed_air_conditioning,
         secondary_fuel_type=secondary_fuel_type,
+        secondary_heating_type=heating.secondary_heating.secondary_system,
+        shower_outlets=shower_outlets,
+        cylinder_size=heating.water_heating.cylinder_size if heating.water_heating.cylinder_size != "No Cylinder" else None,
+        cylinder_insulation_type=heating.water_heating.insulation_type,
+        cylinder_insulation_thickness_mm=heating.water_heating.insulation_thickness_mm,
+        immersion_heating_type=heating.water_heating.immersion_type,
+    )
+
+
+def _map_sap_ventilation(ventilation: Ventilation) -> SapVentilation:
+    return SapVentilation(
+        ventilation_type=ventilation.ventilation_type,
+        draught_lobby=ventilation.draught_lobby,
+        pressure_test=ventilation.pressure_test,
+        open_flues_count=ventilation.number_of_open_flues,
+        closed_flues_count=ventilation.number_of_closed_flues,
+        boiler_flues_count=ventilation.number_of_boiler_flues,
+        other_flues_count=ventilation.number_of_other_flues,
+        extract_fans_count=ventilation.number_of_extract_fans,
+        passive_vents_count=ventilation.number_of_passive_vents,
+        flueless_gas_fires_count=ventilation.number_of_flueless_gas_fires,
+        ventilation_in_pcdf_database=ventilation.ventilation_in_pcdf_database,
     )

datatypes/epc/domain/tests/test_from_site_notes.py

@@ -13,7 +13,10 @@ from datatypes.epc.domain.epc_property_data import (
     SapEnergySource,
     SapFloorDimension,
     SapHeating,
+    SapVentilation,
     SapWindow,
+    ShowerOutlet,
+    ShowerOutlets,
 )
 from datatypes.epc.domain.mapper import EpcPropertyDataMapper
 from datatypes.epc.schema.tests.helpers import from_dict
@@ -330,14 +333,14 @@ class TestFromSiteNotesExample1:
     def test_uprn_absent(self, result: EpcPropertyData) -> None:
         assert result.uprn is None
 
-    def test_address_absent(self, result: EpcPropertyData) -> None:
-        assert result.address_line_1 is None
+    def test_address_line_1(self, result: EpcPropertyData) -> None:
+        assert result.address_line_1 == "1, Test Street"
 
-    def test_postcode_absent(self, result: EpcPropertyData) -> None:
-        assert result.postcode is None
+    def test_postcode(self, result: EpcPropertyData) -> None:
+        assert result.postcode == "TE1 1ST"
 
-    def test_post_town_absent(self, result: EpcPropertyData) -> None:
-        assert result.post_town is None
+    def test_post_town(self, result: EpcPropertyData) -> None:
+        assert result.post_town == "Test Town"
 
     def test_status_absent(self, result: EpcPropertyData) -> None:
         assert result.status is None
@@ -352,9 +355,9 @@ class TestFromSiteNotesExample1:
             sap_version=None,
             dwelling_type="Mid-terrace house",
             uprn=None,
-            address_line_1=None,
-            postcode=None,
-            post_town=None,
+            address_line_1="1, Test Street",
+            postcode="TE1 1ST",
+            post_town="Test Town",
             inspection_date=date(2026, 3, 31),
             status=None,
             tenure="Rented Social",
@@ -379,9 +382,18 @@ class TestFromSiteNotesExample1:
                         emitter_temperature="Unknown",
                         fan_flue_present=True,
                         main_heating_control="Programmer, room thermostat and TRVs",
+                        condensing=True,
+                        weather_compensator=False,
+                        central_heating_pump_age_str="Unknown",
                     )
                 ],
                 has_fixed_air_conditioning=False,
+                cylinder_size="Normal (90-130 litres)",
+                cylinder_insulation_type="Factory fitted",
+                cylinder_insulation_thickness_mm=12,
+                shower_outlets=ShowerOutlets(
+                    shower_outlet=ShowerOutlet(shower_outlet_type="Non-Electric Shower"),
+                ),
             ),
             # Windows
             sap_windows=[
@@ -464,15 +476,23 @@ class TestFromSiteNotesExample1:
                             total_floor_area_m2=24.78,
                             party_wall_length_m=6.15,
                             heat_loss_perimeter_m=14.21,
+                            floor=1,
                         ),
                         SapFloorDimension(
                             room_height_m=2.35,
                             total_floor_area_m2=24.78,
                             party_wall_length_m=6.15,
                             heat_loss_perimeter_m=14.21,
+                            floor=0,
                         ),
                     ],
                     wall_thickness_mm=280,
+                    roof_insulation_location="Joists",
+                    roof_insulation_thickness=100,
+                    floor_type="Ground Floor",
+                    floor_construction_type="Suspended, not timber",
+                    floor_insulation_type_str="As Built",
+                    floor_u_value_known=False,
                 )
             ],
             solar_water_heating=False,
@@ -495,5 +515,182 @@ class TestFromSiteNotesExample1:
             has_conservatory=False,
             blocked_chimneys_count=0,
             draughtproofed_door_count=2,
+            report_reference="49D422A9-0779-44DD-9665-464D35DFF1A8",
+            number_of_storeys=2,
+            any_unheated_rooms=True,
+            waste_water_heat_recovery="None",
+            hydro=False,
+            photovoltaic_array=False,
+            sap_ventilation=SapVentilation(
+                ventilation_type="Natural",
+                open_flues_count=0,
+                closed_flues_count=0,
+                boiler_flues_count=0,
+                other_flues_count=0,
+                extract_fans_count=2,
+                passive_vents_count=0,
+                flueless_gas_fires_count=0,
+                pressure_test="No test",
+                draught_lobby=False,
+            ),
         )
         assert result == expected
+
+
+class TestFromSiteNotesVentilation:
+    """
+    Fixture: pashub_rdsap_site_notes_example1.json
+    Ventilation: Natural, 2 extract fans, no flues, no test, no draught lobby.
+    """
+
+    @pytest.fixture
+    def survey(self) -> PasHubRdSapSiteNotes:
+        return from_dict(
+            PasHubRdSapSiteNotes, load("pashub_rdsap_site_notes_example1.json")
+        )
+
+    @pytest.fixture
+    def result(self, survey: PasHubRdSapSiteNotes) -> EpcPropertyData:
+        return EpcPropertyDataMapper.from_site_notes(survey)
+
+    def test_sap_ventilation_present(self, result: EpcPropertyData) -> None:
+        assert result.sap_ventilation is not None
+
+    def test_ventilation_type(self, result: EpcPropertyData) -> None:
+        # ventilation.ventilation_type: "Natural"
+        assert result.sap_ventilation.ventilation_type == "Natural"
+
+    def test_open_flues_count(self, result: EpcPropertyData) -> None:
+        # ventilation.number_of_open_flues: 0
+        assert result.sap_ventilation.open_flues_count == 0
+
+    def test_closed_flues_count(self, result: EpcPropertyData) -> None:
+        # ventilation.number_of_closed_flues: 0
+        assert result.sap_ventilation.closed_flues_count == 0
+
+    def test_boiler_flues_count(self, result: EpcPropertyData) -> None:
+        # ventilation.number_of_boiler_flues: 0
+        assert result.sap_ventilation.boiler_flues_count == 0
+
+    def test_other_flues_count(self, result: EpcPropertyData) -> None:
+        # ventilation.number_of_other_flues: 0
+        assert result.sap_ventilation.other_flues_count == 0
+
+    def test_extract_fans_count(self, result: EpcPropertyData) -> None:
+        # ventilation.number_of_extract_fans: 2
+        assert result.sap_ventilation.extract_fans_count == 2
+
+    def test_passive_vents_count(self, result: EpcPropertyData) -> None:
+        # ventilation.number_of_passive_vents: 0
+        assert result.sap_ventilation.passive_vents_count == 0
+
+    def test_flueless_gas_fires_count(self, result: EpcPropertyData) -> None:
+        # ventilation.number_of_flueless_gas_fires: 0
+        assert result.sap_ventilation.flueless_gas_fires_count == 0
+
+    def test_pressure_test(self, result: EpcPropertyData) -> None:
+        # ventilation.pressure_test: "No test"
+        assert result.sap_ventilation.pressure_test == "No test"
+
+    def test_draught_lobby(self, result: EpcPropertyData) -> None:
+        # ventilation.draught_lobby: false
+        assert result.sap_ventilation.draught_lobby is False
+
+
+class TestFromSiteNotesFloorConstruction:
+    """
+    Fixture: pashub_rdsap_site_notes_example1.json
+    Floor: Suspended not timber, As Built insulation, Ground Floor type.
+    """
+
+    @pytest.fixture
+    def survey(self) -> PasHubRdSapSiteNotes:
+        return from_dict(
+            PasHubRdSapSiteNotes, load("pashub_rdsap_site_notes_example1.json")
+        )
+
+    @pytest.fixture
+    def result(self, survey: PasHubRdSapSiteNotes) -> EpcPropertyData:
+        return EpcPropertyDataMapper.from_site_notes(survey)
+
+    def test_floor_type(self, result: EpcPropertyData) -> None:
+        # building_construction.floor.floor_type: "Ground Floor"
+        assert result.sap_building_parts[0].floor_type == "Ground Floor"
+
+    def test_floor_construction_type(self, result: EpcPropertyData) -> None:
+        # building_construction.floor.floor_construction: "Suspended, not timber"
+        assert result.sap_building_parts[0].floor_construction_type == "Suspended, not timber"
+
+    def test_floor_insulation_type_str(self, result: EpcPropertyData) -> None:
+        # building_construction.floor.floor_insulation_type: "As Built"
+        assert result.sap_building_parts[0].floor_insulation_type_str == "As Built"
+
+    def test_floor_u_value_known(self, result: EpcPropertyData) -> None:
+        # building_construction.floor.floor_u_value_known: false
+        assert result.sap_building_parts[0].floor_u_value_known is False
+
+
+class TestFromSiteNotesHeatingBoiler:
+    """
+    Fixture: pashub_rdsap_site_notes_example1.json
+    Boiler: Vaillant ecoFIT sustain, condensing, no weather compensator, pump age unknown.
+    """
+
+    @pytest.fixture
+    def survey(self) -> PasHubRdSapSiteNotes:
+        return from_dict(
+            PasHubRdSapSiteNotes, load("pashub_rdsap_site_notes_example1.json")
+        )
+
+    @pytest.fixture
+    def result(self, survey: PasHubRdSapSiteNotes) -> EpcPropertyData:
+        return EpcPropertyDataMapper.from_site_notes(survey)
+
+    def test_condensing(self, result: EpcPropertyData) -> None:
+        # heating_and_hot_water.main_heating.condensing: true
+        assert result.sap_heating.main_heating_details[0].condensing is True
+
+    def test_weather_compensator(self, result: EpcPropertyData) -> None:
+        # heating_and_hot_water.main_heating.weather_compensator: false
+        assert result.sap_heating.main_heating_details[0].weather_compensator is False
+
+    def test_central_heating_pump_age_str(self, result: EpcPropertyData) -> None:
+        # heating_and_hot_water.main_heating.central_heating_pump_age: "Unknown"
+        assert result.sap_heating.main_heating_details[0].central_heating_pump_age_str == "Unknown"
+
+
+class TestFromSiteNotesMiscTopLevel:
+    """
+    Fixture: pashub_rdsap_site_notes_example1.json
+    Misc fields: 2 storeys, unheated rooms present, no hydro, no PV array, no WWHR.
+    """
+
+    @pytest.fixture
+    def survey(self) -> PasHubRdSapSiteNotes:
+        return from_dict(
+            PasHubRdSapSiteNotes, load("pashub_rdsap_site_notes_example1.json")
+        )
+
+    @pytest.fixture
+    def result(self, survey: PasHubRdSapSiteNotes) -> EpcPropertyData:
+        return EpcPropertyDataMapper.from_site_notes(survey)
+
+    def test_number_of_storeys(self, result: EpcPropertyData) -> None:
+        # general.number_of_storeys: 2
+        assert result.number_of_storeys == 2
+
+    def test_any_unheated_rooms(self, result: EpcPropertyData) -> None:
+        # room_count_elements.any_unheated_rooms: true
+        assert result.any_unheated_rooms is True
+
+    def test_waste_water_heat_recovery(self, result: EpcPropertyData) -> None:
+        # room_count_elements.waste_water_heat_recovery: "None"
+        assert result.waste_water_heat_recovery == "None"
+
+    def test_hydro(self, result: EpcPropertyData) -> None:
+        # renewables.hydro: false
+        assert result.hydro is False
+
+    def test_photovoltaic_array(self, result: EpcPropertyData) -> None:
+        # renewables.photovoltaic_array: false
+        assert result.photovoltaic_array is False

datatypes/epc/schema/tests/helpers.py

@@ -1,5 +1,6 @@
 import dataclasses
 import typing
+from datetime import date
 from typing import Any, Dict, Type, TypeVar
 
 T = TypeVar("T")
@@ -70,4 +71,7 @@ def _coerce(value: Any, hint: Any) -> Any:
     if dataclasses.is_dataclass(hint) and isinstance(value, dict):
         return _from_dict_impl(hint, value)
 
+    if hint is date and isinstance(value, str):
+        return date.fromisoformat(value)
+
     return value

datatypes/epc/surveys/pashub_rdsap_site_notes.py

@@ -1,4 +1,5 @@
 from dataclasses import dataclass
+from datetime import date
 from typing import List, Optional
 
 
@@ -8,7 +9,7 @@ class InspectionMetadata:
     email_address: str
     report_reference: str
     created_on: str
-    date_of_inspection: str
+    date_of_inspection: date
     property_address: str
     property_photo: Optional[bool] = None
 
@@ -17,7 +18,7 @@ class InspectionMetadata:
 class General:
     epc_checked_before_assessment: bool
     epc_exists_at_point_of_assessment: bool
-    inspection_date: str
+    inspection_date: date
     transaction_type: str
     tenure: str
     property_type: str
@@ -178,6 +179,7 @@ class MainHeating:
 @dataclass
 class SecondaryHeating:
     secondary_fuel: str
+    secondary_system: Optional[str] = None
 
 
 @dataclass
@@ -189,6 +191,7 @@ class WaterHeating:
     insulation_type: Optional[str] = None
     insulation_thickness_mm: Optional[int] = None
     has_thermostat: Optional[bool] = None
+    immersion_type: Optional[str] = None
 
 
 @dataclass
@@ -226,6 +229,8 @@ class Renewables:
     photovoltaic_array: bool
     number_of_pv_batteries: int
     hydro: bool
+    pv_connection: Optional[str] = None
+    percent_roof_covered_pv: Optional[int] = None
 
 
 @dataclass

datatypes/epc/surveys/tests/fixtures/pashub_rdsap_site_notes_example1.json

@@ -5,7 +5,7 @@
     "report_reference": "49D422A9-0779-44DD-9665-464D35DFF1A8",
     "created_on": "2026-03-31",
     "date_of_inspection": "2026-03-31",
-    "property_address": "test"
+    "property_address": "1, Test Street, Test Town, Test County, TE1 1ST"
   },
   "general": {
     "epc_checked_before_assessment": true,
@@ -229,4 +229,4 @@
     "hard_to_treat_cavity_high_exposure": false,
     "hard_to_treat_cavity_narrow_cavities": false
   }
-}
+}

datatypes/epc/surveys/tests/test_pashub_rdsap_site_notes_loading.py

@@ -1,5 +1,6 @@
 import json
 import os
+from datetime import date
 from typing import Any, Dict
 
 import pytest
@@ -224,7 +225,7 @@ class TestExample2:
         self, survey: PasHubRdSapSiteNotes
     ) -> None:
         assert survey.inspection_metadata.created_on == "2025-11-10"
-        assert survey.inspection_metadata.date_of_inspection == "2025-09-25"
+        assert survey.inspection_metadata.date_of_inspection == date(2025, 9, 25)
 
     # --- general ---
 

pytest.ini

@@ -3,6 +3,6 @@ pythonpath = .
 log_cli = true
 log_cli_level = INFO
 addopts = --cov-report term-missing --cov=etl/epc --cov=recommendations --cov=backend --cov=etl/epc_clean --cov=etl/spatial
-testpaths = recommendations/tests backend/tests etl/epc/tests etl/epc_clean/tests etl/spatial/tests backend/condition/tests backend/address2UPRN/tests backend/onboarders/tests backend/categorisation/tests backend/export/tests etl/hubspot/tests backend/hubspot_trigger_orchestrator/tests datatypes/epc/schema/tests datatypes/epc/surveys/tests datatypes/epc/domain/tests backend/ecmk_fetcher/tests/
+testpaths = recommendations/tests backend/tests etl/epc/tests etl/epc_clean/tests etl/spatial/tests backend/condition/tests backend/address2UPRN/tests backend/onboarders/tests backend/categorisation/tests backend/export/tests etl/hubspot/tests backend/hubspot_trigger_orchestrator/tests datatypes/epc/schema/tests datatypes/epc/surveys/tests datatypes/epc/domain/tests backend/ecmk_fetcher/tests/ backend/documents_parser/tests
 markers =
     integration: mark a test as an integration test

test.requirements.txt

@@ -6,4 +6,5 @@ dotenv
 psycopg[binary]
 pytest-postgresql
 hubspot-api-client
-fuzzywuzzy
+fuzzywuzzy
+pymupdf