Extract data from xml 🟥

backend/ecmk_fetcher/tests/test_xml_processor.py

@@ -0,0 +1,74 @@
+from backend.ecmk_fetcher.xml_processor import parse_rdsap
+
+
+SAMPLE_XML = """<RdSAP-Report xmlns="https://epbr.digital.communities.gov.uk/xsd/rdsap">
+<Report-Header>
+    <Property>
+        <Address>
+            <Address-Line-1>1</Address-Line-1>
+            <Address-Line-2>Fake Avenue</Address-Line-2>
+            <Post-Town>Random</Post-Town>
+            <Postcode>AB24 5CD</Postcode>
+        </Address>
+    </Property>
+</Report-Header>
+
+<SAP-Data>
+    <SAP-Property-Details>
+        <Property-Type>0</Property-Type>
+
+        <SAP-Building-Parts>
+            <SAP-Building-Part>
+                <Identifier>Main Dwelling</Identifier>
+
+                <SAP-Floor-Dimensions>
+                    <SAP-Floor-Dimension>
+                        <Heat-Loss-Perimeter>25.31</Heat-Loss-Perimeter>
+                        <Room-Height>2.46</Room-Height>
+                        <Total-Floor-Area>43.61</Total-Floor-Area>
+                        <Party-Wall-Length>0</Party-Wall-Length>
+                    </SAP-Floor-Dimension>
+
+                    <SAP-Floor-Dimension>
+                        <Heat-Loss-Perimeter>26.16</Heat-Loss-Perimeter>
+                        <Room-Height>2.44</Room-Height>
+                        <Total-Floor-Area>42.33</Total-Floor-Area>
+                        <Party-Wall-Length>0</Party-Wall-Length>
+                    </SAP-Floor-Dimension>
+                </SAP-Floor-Dimensions>
+
+                <Roof-Construction>4</Roof-Construction>
+                <Roof-Insulation-Location>2</Roof-Insulation-Location>
+                <Roof-Insulation-Thickness>100mm</Roof-Insulation-Thickness>
+            </SAP-Building-Part>
+        </SAP-Building-Parts>
+
+    </SAP-Property-Details>
+</SAP-Data>
+</RdSAP-Report>
+"""
+
+
+def test_parse_rdsap_wide_flat_contract():
+    # arrange + act
+    result = parse_rdsap(SAMPLE_XML)
+
+    # assert
+    assert result == {
+        "address": "1, Fake Avenue, Random, AB24 5CD",
+        "property_type": "House",
+        # Main Dwelling - floor 0
+        "main_dwelling_floor_index_0_area_m2": 43.61,
+        "main_dwelling_floor_index_0_height_m": 2.46,
+        "main_dwelling_floor_index_0_heat_loss_perimeter_m": 25.31,
+        "main_dwelling_floor_index_0_party_wall_length_m": 0.0,
+        # Main Dwelling - floor 1
+        "main_dwelling_floor_index_1_area_m2": 42.33,
+        "main_dwelling_floor_index_1_height_m": 2.44,
+        "main_dwelling_floor_index_1_heat_loss_perimeter_m": 26.16,
+        "main_dwelling_floor_index_1_party_wall_length_m": 0.0,
+        # Roof (building-level, repeated across floors or stored once)
+        "main_dwelling_roof_construction": "4",
+        "main_dwelling_roof_insulation_location": "2",
+        "main_dwelling_roof_insulation_thickness": "100mm",
+    }

backend/ecmk_fetcher/xml_processor.py

@@ -0,0 +1,8 @@
+from typing import Any
+
+from etl.xml_survey_extraction.XmlParser import PROPERTY_TYPE_LOOKUP
+
+
+# This file should ultimately live somewhere different, probably
+def parse_rdsap(xml_string: str) -> Any:  # TODO: define shape of return object
+    raise NotImplementedError

datatypes/epc/domain/field_mappings.py

@@ -0,0 +1 @@
+PROPERTY_TYPE_LOOKUP = {0: "House", 1: "Bungalow", 2: "Flat", 3: "Maisonette"}

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
+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/
 markers =
     integration: mark a test as an integration test