Add test for 321/support all on schema change options

tests/functional/iceberg/test_incremental_models.py

@@ -3,7 +3,7 @@
 
 from pathlib import Path
 
-from dbt.tests.util import run_dbt, run_dbt_and_capture
+from dbt.tests.util import run_dbt, run_dbt_and_capture, write_file
 
 
 _SEED_INCREMENTAL_STRATEGIES = """
@@ -32,6 +32,7 @@
     incremental_strategy='{strategy}',
     unique_key="world_id",
     external_volume = "s3_iceberg_snow",
+    on_schema_change = "sync_all_columns"
   )
 }}}}
 select * from {{{{ ref('upstream_table') }}}}
@@ -123,3 +124,46 @@ def test_incremental_strategies_with_update(self, project, setup_class):
         self.__check_correct_operations(self.append, rows_affected=2)
         self.__check_correct_operations("merge", rows_affected=1)
         self.__check_correct_operations("delete_insert", rows_affected=1)
+
+
+class TestIcebergIncrementalOnSchemaChangeMutatesRelations:
+    @pytest.fixture(scope="class")
+    def project_config_update(self):
+        return {"flags": {"enable_iceberg_materializations": True}}
+
+    @pytest.fixture(scope="class")
+    def seeds(self):
+        return {
+            "seed.csv": _SEED_INCREMENTAL_STRATEGIES,
+        }
+
+    @pytest.fixture(scope="function", autouse=True)
+    def setup_class(self, project):
+        run_dbt(["seed"])
+        run_dbt(["run"])
+        yield
+
+    @pytest.fixture(scope="class")
+    def models(self):
+        return {
+            "upstream_table.sql": _MODEL_BASIC_TABLE_MODEL,
+            "merge.sql": _MODEL_INCREMENTAL_ICEBERG_MERGE,
+        }
+
+    def test_sync_and_append_semantics(self, project, setup_class):
+        model_file = project.project_root / Path("models") / Path("merge.sql")
+        sql = f"show columns in {project.database}.{project.test_schema}.merge;"
+        column_names = [column[2] for column in project.run_sql(sql, fetch="all")]
+        assert len(column_names) == 3
+
+        write_file(_MODEL_INCREMENTAL_ICEBERG_MERGE.replace("*", "*, 1 as new_column"), model_file)
+        run_dbt()
+        column_names = [column[2].lower() for column in project.run_sql(sql, fetch="all")]
+        assert len(column_names) == 4
+        assert "new_column" in column_names
+
+        write_file(_MODEL_INCREMENTAL_ICEBERG_MERGE, model_file)
+        run_dbt()
+        column_names = [column[2].lower() for column in project.run_sql(sql, fetch="all")]
+        assert len(column_names) == 3
+        assert "new_column" not in column_names