test: additional tests with record index for entity empty check functions

Author: stevenhsd

src/dve/core_engine/backends/implementations/duckdb/duckdb_helpers.py

@@ -116,7 +116,7 @@ def relation_is_empty(relation: DuckDBPyRelation) -> bool:
     cl_all_null_check_exprs = [
         f'bool_and("{cl}" is NULL)' for cl in relation.columns if cl != RECORD_INDEX_COLUMN_NAME
     ]
-    return relation.aggregate(" AND ".join(cl_all_null_check_exprs)).fetchone()[0] # type: ignore
+    return relation.aggregate(" AND ".join(cl_all_null_check_exprs)).fetchone()[0]  # type: ignore
 
 
 def get_duckdb_type_from_annotation(type_annotation: Any) -> DuckDBPyType:

tests/test_core_engine/test_backends/test_implementations/test_duckdb/test_duckdb_helpers.py

@@ -25,6 +25,7 @@
     relation_is_empty
 )
 from dve.core_engine.backends.utilities import datetime_format_to_regex
+from dve.core_engine.constants import RECORD_INDEX_COLUMN_NAME
 from dve.metadata_parser.domain_types import formatteddatetime
 
 @pytest.fixture
@@ -272,6 +273,12 @@ def test_relation_is_empty(temp_ddb_conn: DuckDBPyConnection):
 
     rel3 = con.sql("SELECT 1 as fld1, NULL as fld2 UNION ALL SELECT NULL as fld1, NULL as fld2")
     assert not relation_is_empty(rel3)
+    
+    rel4 = con.sql(f"SELECT 1 as {RECORD_INDEX_COLUMN_NAME}, NULL as fld1, NULL as fld2 UNION ALL SELECT 2, NULL, NULL")
+    assert relation_is_empty(rel4)
+    
+    rel5 = con.sql(f"SELECT 1 as {RECORD_INDEX_COLUMN_NAME}, NULL as fld1, NULL as fld2 UNION ALL SELECT 2, NULL, 'hello'")
+    assert not relation_is_empty(rel5)
 
 @pytest.mark.parametrize("date_format, expected_regex", [("%Y-%m-%d", r"^[0-9]{4}\-[0-9]{2}\-[0-9]{2}$"),
                                                           ("%Y-%m-%dT%H:%M:%S", r"^[0-9]{4}\-[0-9]{2}\-[0-9]{2}T[0-9]{2}:[0-9]{2}:[0-9]{2}$"),

tests/test_core_engine/test_backends/test_implementations/test_spark/test_spark_helpers.py

@@ -32,6 +32,7 @@
     python_to_java_datetime_format,
     df_is_empty
 )
+from dve.core_engine.constants import RECORD_INDEX_COLUMN_NAME
 from dve.metadata_parser.domain_types import conformatteddate
 
 from .....fixtures import spark  # pylint: disable=unused-import
@@ -366,5 +367,20 @@ def test_df_is_empty(spark):
     df3 = spark.createDataFrame([(None, None), (None, None), (None, 3)],
                                schema=StructType([StructField("col1", StringType()), StructField("col2", LongType())]))
     assert not df_is_empty(df3)
+    
+    _schema_w_ri = schema=StructType([
+                                   StructField(RECORD_INDEX_COLUMN_NAME, LongType()),
+                                   StructField("col1", StringType()),
+                                   StructField("col2", LongType())
+                                   ])
+    
+    df4 = spark.createDataFrame([(1, None, None), (2, None, None), (3, None, None)],
+                               schema=_schema_w_ri)
+    assert df_is_empty(df4)
+    
+    df5 = df4.union(spark.createDataFrame([(4, None, 8)], schema=_schema_w_ri))
+    assert not df_is_empty(df5)
+    
+