Feature/add unit tests

.github/workflows/ci.yml

@@ -35,3 +35,7 @@ jobs:
       - name: Run flake8 (linting)
         run: |
           poetry run flake8 .
+
+      - name: Run pytest (testing)
+        run: |
+          poetry run pytest

README.md

@@ -25,6 +25,10 @@
 
 `poetry run flake8 .`
 
+## Test
+
+`poetry run pytest`
+
 <img width="495" alt="image" src="https://github.com/user-attachments/assets/4a5fa7a5-0c20-480e-8f15-00dfd9c72b2e" />
 <img width="802" alt="image" src="https://github.com/user-attachments/assets/5d3dc009-9dce-41bd-b4c9-c660a7c594ad" />
 

pyproject.toml

@@ -27,4 +27,5 @@ build-backend = "poetry.core.masonry.api"
 black = "^25.1.0"
 isort = "^6.0.1"
 flake8 = "^7.3.0"
+pytest = "^8.4.1"
 

tests/conftest.py

@@ -0,0 +1,5 @@
+import os
+import sys
+
+# Ensure the app module is importable for all tests
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..")))

tests/services/test_search.py

@@ -0,0 +1,62 @@
+import contextlib
+from unittest.mock import patch
+
+import pandas as pd
+import pytest
+
+from app.services.search import search_companies
+
+
+@pytest.fixture
+def mock_data():
+    return pd.DataFrame(
+        {
+            "Organisation Name": ["Foo Company", "Xyz Motion", "Test Company"],
+            "Town/City": ["London", "London", "Manchester"],
+            "County": ["Greater London", "Greater London", "Greater Manchester"],
+        }
+    )
+
+
+# Generalized fixture factory for patching skilled_worker_data_current
+@pytest.fixture
+def patch_skilled_worker_data_current():
+    @contextlib.contextmanager
+    def _patch(data):
+        with patch(
+            "app.services.search.skilled_worker_data_current"
+        ) as mock_skilled_worker_data_current:
+            mock_skilled_worker_data_current.__getitem__.side_effect = data.__getitem__
+            mock_skilled_worker_data_current.loc.__getitem__.side_effect = (
+                data.loc.__getitem__
+            )
+            yield mock_skilled_worker_data_current
+
+    return _patch
+
+
+def test_exact_match(patch_skilled_worker_data_current, mock_data):
+    with patch_skilled_worker_data_current(mock_data):
+        results = search_companies("Foo Company", threshold=90)
+        assert any("Foo Company" in r.Organisation_Name for r in results)
+
+
+def test_fuzzy_match(patch_skilled_worker_data_current, mock_data):
+    with patch_skilled_worker_data_current(mock_data):
+        results = search_companies("Foo", threshold=70)
+        assert any("Foo Company" in r.Organisation_Name for r in results)
+
+
+def test_no_match(patch_skilled_worker_data_current, mock_data):
+    with patch_skilled_worker_data_current(mock_data):
+        results = search_companies("Nonexistent Company", threshold=90)
+        assert not results
+
+
+def test_nan_handling(patch_skilled_worker_data_current):
+    nan_data = pd.DataFrame(
+        {"Organisation Name": [None], "Town/City": [None], "County": [None]}
+    )
+    with patch_skilled_worker_data_current(nan_data):
+        results = search_companies("anything", threshold=0)
+        assert results[0].Organisation_Name == ""