good

pipeline/download/test_places.py

@@ -4,6 +4,8 @@ from pipeline.download.places import (
     _is_dlr_station,
     _is_tram_station,
     _naptan_dlr_stations,
+    _ofs_universities,
+    _select_university_name,
     _station_display_name,
 )
 
@@ -79,3 +81,68 @@ def test_naptan_dlr_stations_are_deduplicated_by_atco_code(tmp_path):
     assert shadwell["lat"] == (51.51156 + 51.511693) / 2
     assert shadwell["place_type"] == "station"
     assert shadwell["travel_destination"] is True
+
+
+def test_select_university_name_prefers_public_trading_name_for_noisy_legal_name():
+    assert (
+        _select_university_name(
+            "The Chancellor, Masters and Scholars of the University of Oxford",
+            "Oxford University\nThe University of Oxford",
+        )
+        == "Oxford University"
+    )
+    assert (
+        _select_university_name(
+            "Bournemouth University Higher Education Corporation",
+            "Bournemouth University",
+        )
+        == "Bournemouth University"
+    )
+    assert (
+        _select_university_name("The University of Surrey", "Not applicable")
+        == "University of Surrey"
+    )
+
+
+def test_ofs_universities_extracts_university_title_rows_with_postcode_coords():
+    raw_register = pl.DataFrame(
+        [
+            ["OfS Register", None, None, None],
+            ["Note row", None, None, None],
+            [
+                "Provider's legal name",
+                "Provider's trading name(s)",
+                "Provider's contact address",
+                "Does the provider have the right to use university in its title?",
+            ],
+            [
+                "The Chancellor, Masters and Scholars of the University of Oxford",
+                "Oxford University\nThe University of Oxford",
+                "University Offices\nWellington Square\nOxford\nOX1 2JD\nUnited Kingdom",
+                "Yes",
+            ],
+            [
+                "Example College",
+                "Not applicable",
+                "Example Street\nLondon\nSW1A 1AA\nUnited Kingdom",
+                "No",
+            ],
+        ],
+        orient="row",
+    )
+
+    universities, skipped = _ofs_universities(
+        raw_register, {"OX12JD": (51.7585, -1.2643)}
+    )
+
+    assert skipped == 0
+    assert universities == [
+        {
+            "name": "Oxford University",
+            "place_type": "university",
+            "lat": 51.7585,
+            "lon": -1.2643,
+            "population": 0,
+            "travel_destination": True,
+        }
+    ]