Fix seed boolean null handling (#5870)

Signed-off-by: nkwork9999 <143652584+nkwork9999@users.noreply.github.com>

Author: nkwork9999

sqlmesh/core/model/definition.py

@@ -1712,6 +1712,7 @@ def render(
 
     def render_seed(self) -> t.Iterator[QueryOrDF]:
         import numpy as np
+        import pandas as pd
 
         self._ensure_hydrated()
 
@@ -1752,8 +1753,6 @@ def render_seed(self) -> t.Iterator[QueryOrDF]:
 
             # convert all date/time types to native pandas timestamp
             for column in [*date_columns, *datetime_columns]:
-                import pandas as pd
-
                 df[column] = pd.to_datetime(df[column], infer_datetime_format=True, errors="ignore")  # type: ignore
 
             # extract datetime.date from pandas timestamp for DATE columns
@@ -1769,7 +1768,7 @@ def render_seed(self) -> t.Iterator[QueryOrDF]:
                     )
 
             for column in bool_columns:
-                df[column] = df[column].apply(lambda i: str_to_bool(str(i)))
+                df[column] = df[column].apply(lambda i: None if pd.isna(i) else str_to_bool(str(i)))
 
             df.loc[:, string_columns] = df[string_columns].mask(
                 cond=lambda x: x.notna(),  # type: ignore

tests/core/test_model.py

@@ -1541,6 +1541,44 @@ def test_seed_model_custom_types(tmp_path):
     assert df["empty_date"].iloc[0] is None
 
 
+def test_seed_model_boolean_nulls_are_preserved(tmp_path):
+    model_csv_path = (tmp_path / "model.csv").absolute()
+
+    with open(model_csv_path, "w", encoding="utf-8") as fd:
+        fd.write("id,test_ind\n")
+        fd.write("1,null\n")
+        fd.write("2,false\n")
+        fd.write("3,true\n")
+        fd.write("4,null\n")
+
+    model = create_seed_model(
+        "test_db.test_model",
+        SeedKind(path=str(model_csv_path)),
+        columns={
+            "id": "int",
+            "test_ind": "boolean",
+        },
+        dialect="databricks",
+    )
+
+    df = next(model.render_seed())
+
+    assert df["test_ind"].to_list() == [None, False, True, None]
+
+    context = Context(
+        config=Config(
+            default_connection=DuckDBConnectionConfig(),
+            model_defaults=ModelDefaultsConfig(dialect="databricks"),
+        )
+    )
+    context.upsert_model(model)
+
+    rendered_sql = context.render(model).sql("databricks")
+
+    assert "CAST(NULL AS BOOLEAN)" in rendered_sql
+    assert "(4, NULL)" in rendered_sql
+
+
 def test_seed_with_special_characters_in_column(tmp_path, assert_exp_eq):
     config = Config(model_defaults=ModelDefaultsConfig(dialect="duckdb"))
     context = Context(config=config)