fix: resolve merge conflicts

Author: nlebovits

data/src/data_utils/city_owned_properties.py

@@ -4,7 +4,10 @@
 import geopandas as gpd
 
 from src.validation.base import ValidationResult, validate_output
-from src.validation.city_owned_properties import CityOwnedPropertiesOutputValidator
+from src.validation.city_owned_properties import (
+    CityOwnedPropertiesOutputValidator,
+    CityOwnedPropertiesInputValidator,
+)
 
 from ..classes.loaders import EsriLoader
 from ..constants.services import CITY_OWNED_PROPERTIES_TO_LOAD
@@ -50,6 +53,7 @@ def city_owned_properties(
         esri_urls=CITY_OWNED_PROPERTIES_TO_LOAD,
         cols=["OPABRT", "AGENCY", "SIDEYARDELIGIBLE"],
         opa_col="opabrt",
+        validator=CityOwnedPropertiesInputValidator(),
     )
 
     city_owned_properties, input_validation = loader.load_or_fetch()

data/src/data_utils/council_dists.py

@@ -4,7 +4,10 @@
 import pandas as pd
 
 from src.validation.base import ValidationResult, validate_output
-from src.validation.council_dists import CouncilDistrictsOutputValidator
+from src.validation.council_dists import (
+    CouncilDistrictsInputValidator,
+    CouncilDistrictsOutputValidator,
+)
 
 from ..classes.loaders import EsriLoader
 from ..constants.services import COUNCIL_DISTRICTS_TO_LOAD
@@ -42,6 +45,7 @@ def council_dists(
         name="Council Districts",
         esri_urls=COUNCIL_DISTRICTS_TO_LOAD,
         cols=["district"],
+        validator=CouncilDistrictsInputValidator(),
         input_crs="EPSG:4326",  # Load in geographic coordinates since the data appears to be lat/lon
     )
 

data/src/test/data_utils/test_data_utils.py

@@ -112,11 +112,26 @@ def test_park_priority_basic_functionality(
         self.assertIn("park_priority", second_arg.columns)
         self.assertNotIn("parkneed", second_arg.columns)
 
-        # Check return values - focus on object identity and type, not exact equality
-        self.assertIs(result_gdf, mock_spatial_join.return_value)
+        # Check return values - use data equality instead of object identity
         self.assertIsInstance(result_gdf, gpd.GeoDataFrame)
         self.assertIsInstance(validation_result, ValidationResult)
-        # Use the mock return value instead of comparing objects
+
+        # Check data equality for the GeoDataFrame
+        pd.testing.assert_frame_equal(
+            result_gdf.drop(columns=["geometry"]),
+            expected_result_gdf.drop(columns=["geometry"]),
+            check_dtype=False,
+        )
+
+        # Check geometry column separately
+        for i, (expected_geom, actual_geom) in enumerate(
+            zip(expected_result_gdf.geometry, result_gdf.geometry)
+        ):
+            self.assertTrue(
+                expected_geom.equals(actual_geom), f"Geometry mismatch at index {i}"
+            )
+
+        # Check validation result
         self.assertIs(validation_result, mock_validation_result)
 
     @patch("src.data_utils.park_priority.EsriLoader")
@@ -176,8 +191,25 @@ def test_park_priority_column_renaming(
         self.assertNotIn("parkneed", second_arg.columns)
         self.assertEqual(second_arg["park_priority"].iloc[0], 5.0)
 
-        # Check return value - focus on object identity
-        self.assertIs(result_gdf, mock_spatial_join.return_value)
+        # Check return value - use data equality instead of object identity
+        self.assertIsInstance(result_gdf, gpd.GeoDataFrame)
+
+        # Check data equality for the GeoDataFrame
+        pd.testing.assert_frame_equal(
+            result_gdf.drop(columns=["geometry"]),
+            expected_result_gdf.drop(columns=["geometry"]),
+            check_dtype=False,
+        )
+
+        # Check geometry column separately
+        for i, (expected_geom, actual_geom) in enumerate(
+            zip(expected_result_gdf.geometry, result_gdf.geometry)
+        ):
+            self.assertTrue(
+                expected_geom.equals(actual_geom), f"Geometry mismatch at index {i}"
+            )
+
+        # Check validation result
         self.assertIs(validation_result, mock_validation_result)
 
     @patch("src.data_utils.park_priority.EsriLoader")
@@ -261,9 +293,25 @@ def test_park_priority_empty_data(self, mock_spatial_join, mock_esri_loader_clas
         result_gdf, validation_result = _park_priority_logic(input_gdf)
 
         # Check that the function handles empty data gracefully
-        self.assertIs(result_gdf, mock_spatial_join.return_value)
         self.assertIsInstance(result_gdf, gpd.GeoDataFrame)
         self.assertIsInstance(validation_result, ValidationResult)
+
+        # Check data equality for the GeoDataFrame
+        pd.testing.assert_frame_equal(
+            result_gdf.drop(columns=["geometry"]),
+            input_gdf.drop(columns=["geometry"]),
+            check_dtype=False,
+        )
+
+        # Check geometry column separately
+        for i, (expected_geom, actual_geom) in enumerate(
+            zip(input_gdf.geometry, result_gdf.geometry)
+        ):
+            self.assertTrue(
+                expected_geom.equals(actual_geom), f"Geometry mismatch at index {i}"
+            )
+
+        # Check validation result
         self.assertIs(validation_result, mock_validation_result)
 
         # Check that spatial_join was still called (even with empty data)
@@ -324,8 +372,23 @@ def test_park_priority_return_format(
         # Check second element is ValidationResult
         self.assertIsInstance(validation_result, ValidationResult)
 
-        # Check the actual values - focus on object identity
-        self.assertIs(result_gdf, mock_spatial_join.return_value)
+        # Check the actual values - use data equality instead of object identity
+        # Check data equality for the GeoDataFrame
+        pd.testing.assert_frame_equal(
+            result_gdf.drop(columns=["geometry"]),
+            expected_result_gdf.drop(columns=["geometry"]),
+            check_dtype=False,
+        )
+
+        # Check geometry column separately
+        for i, (expected_geom, actual_geom) in enumerate(
+            zip(expected_result_gdf.geometry, result_gdf.geometry)
+        ):
+            self.assertTrue(
+                expected_geom.equals(actual_geom), f"Geometry mismatch at index {i}"
+            )
+
+        # Check validation result
         self.assertIs(validation_result, mock_validation_result)
 
     @pytest.mark.skip

data/src/validation/city_owned_properties.py

@@ -29,18 +29,43 @@
     coerce=True,
 )
 
+# Expecting ~7,796 records returned (within ±20% tolerance).
+# This is checked in CityOwnedPropertiesInputSchema
+expected = 7796
+lower = int(expected * 0.8)
+upper = int(expected * 1.2)
+
+CityOwnedPropertiesInputSchema = pa.DataFrameSchema(
+    columns={
+        "opa_id": pa.Column(pa.Int, checks=pa.Check(lambda s: s.dropna() != "")),
+        "agency": pa.Column(pa.String, nullable=True),
+        "sideyardeligible": pa.Column(
+            pa.Category, nullable=True, checks=pa.Check.isin(["Yes", "No"])
+        ),
+        "geometry": pa.Column("geometry"),
+    },
+    checks=pa.Check(lambda df: lower <= df.shape[0] <= upper),
+    strict=True,
+)
+
 
 class CityOwnedPropertiesInputValidator(BaseValidator):
-    """Validator for city owned properties service input."""
+    """
+    Validator for the city-owned properties dataset input.
+    schema and _custom_validation() are used by validate() in the parent class.
+    """
 
-    schema = None  # No schema validation for input
+    schema = CityOwnedPropertiesInputSchema
 
     def _custom_validation(self, gdf: gpd.GeoDataFrame):
         pass
 
 
 class CityOwnedPropertiesOutputValidator(BaseValidator):
-    """Validator for city owned properties service output."""
+    """
+    Validator for the city-owned properties dataset output.
+    schema and _custom_validation() are used by validate() in the parent class.
+    """
 
     schema = CityOwnedPropertiesSchema
 

data/src/validation/council_dists.py

@@ -26,11 +26,27 @@
     coerce=True,
 )
 
+CouncilDistrictsInputSchema = pa.DataFrameSchema(
+    columns={
+        "district": pa.Column(
+            str,
+            nullable=True,
+        ),
+        "geometry": pa.Column("geometry"),
+    },
+    # district should contain 10 records of strings 1-10
+    checks=pa.Check(
+        lambda df: set(df["district"].dropna().unique())
+        == {str(i) for i in range(1, 11)}
+    ),
+    strict=True,
+)
+
 
 class CouncilDistrictsInputValidator(BaseValidator):
     """Validator for council districts service input."""
 
-    schema = None  # No schema validation for input
+    schema = CouncilDistrictsInputSchema
 
     def _custom_validation(self, gdf: gpd.GeoDataFrame, check_stats: bool = True):
         pass