refactor: fix test_list_apply_callable on agg_ops.AllOp and agg_ops.AnyOp

Author: chelsea-lin

bigframes/core/compile/sqlglot/aggregations/unary_compiler.py

@@ -45,9 +45,13 @@ def _(
     column: typed_expr.TypedExpr,
     window: typing.Optional[window_spec.WindowSpec] = None,
 ) -> sge.Expression:
-    # BQ will return null for empty column, result would be false in pandas.
-    result = apply_window_if_present(sge.func("LOGICAL_AND", column.expr), window)
-    return sge.func("IFNULL", result, sge.true())
+    expr = column.expr
+    if column.dtype != dtypes.BOOL_DTYPE:
+        expr = sge.NEQ(this=expr, expression=sge.convert(0))
+    expr = apply_window_if_present(sge.func("LOGICAL_AND", expr), window)
+
+    # BQ will return null for empty column, result would be true in pandas.
+    return sge.func("COALESCE", expr, sge.convert(True))
 
 
 @UNARY_OP_REGISTRATION.register(agg_ops.AnyOp)
@@ -57,6 +61,8 @@ def _(
     window: typing.Optional[window_spec.WindowSpec] = None,
 ) -> sge.Expression:
     expr = column.expr
+    if column.dtype != dtypes.BOOL_DTYPE:
+        expr = sge.NEQ(this=expr, expression=sge.convert(0))
     expr = apply_window_if_present(sge.func("LOGICAL_OR", expr), window)
 
     # BQ will return null for empty column, result would be false in pandas.

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_all/out.sql

@@ -1,12 +1,15 @@
 WITH `bfcte_0` AS (
   SELECT
-    `bool_col`
+    `bool_col`,
+    `int64_col`
   FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
 ), `bfcte_1` AS (
   SELECT
-    COALESCE(LOGICAL_AND(`bool_col`), TRUE) AS `bfcol_1`
+    COALESCE(LOGICAL_AND(`bool_col`), TRUE) AS `bfcol_2`,
+    COALESCE(LOGICAL_AND(`int64_col` <> 0), TRUE) AS `bfcol_3`
   FROM `bfcte_0`
 )
 SELECT
-  `bfcol_1` AS `bool_col`
+  `bfcol_2` AS `bool_col`,
+  `bfcol_3` AS `int64_col`
 FROM `bfcte_1`

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_all/window_partition_out.sql

@@ -1,12 +1,15 @@
 WITH `bfcte_0` AS (
   SELECT
-    `bool_col`
+    `bool_col`,
+    `int64_col`
   FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
 ), `bfcte_1` AS (
   SELECT
-    COALESCE(LOGICAL_OR(`bool_col`), FALSE) AS `bfcol_1`
+    COALESCE(LOGICAL_OR(`bool_col`), FALSE) AS `bfcol_2`,
+    COALESCE(LOGICAL_OR(`int64_col` <> 0), FALSE) AS `bfcol_3`
   FROM `bfcte_0`
 )
 SELECT
-  `bfcol_1` AS `bool_col`
+  `bfcol_2` AS `bool_col`,
+  `bfcol_3` AS `int64_col`
 FROM `bfcte_1`

…t_unary_compiler/test_all/window_out.sql …unary_compiler/test_all_w_window/out.sqltests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_all/window_out.sql renamed to tests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_all_w_window/out.sql tests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_all/window_out.sql renamed to tests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_all_w_window/out.sql

@@ -63,41 +63,47 @@ def _apply_unary_window_op(
 
 
 def test_all(scalar_types_df: bpd.DataFrame, snapshot):
+    bf_df = scalar_types_df[["bool_col", "int64_col"]]
+    ops_map = {
+        "bool_col": agg_ops.AllOp().as_expr("bool_col"),
+        "int64_col": agg_ops.AllOp().as_expr("int64_col"),
+    }
+    sql = _apply_unary_agg_ops(bf_df, list(ops_map.values()), list(ops_map.keys()))
+
+    snapshot.assert_match(sql, "out.sql")
+
+
+def test_all_w_window(scalar_types_df: bpd.DataFrame, snapshot):
     col_name = "bool_col"
     bf_df = scalar_types_df[[col_name]]
     agg_expr = agg_ops.AllOp().as_expr(col_name)
-    sql = _apply_unary_agg_ops(bf_df, [agg_expr], [col_name])
-
-    snapshot.assert_match(sql, "out.sql")
 
     # Window tests
     window = window_spec.WindowSpec(ordering=(ordering.ascending_over(col_name),))
     sql_window = _apply_unary_window_op(bf_df, agg_expr, window, "agg_bool")
-    snapshot.assert_match(sql_window, "window_out.sql")
-
-    bf_df_str = scalar_types_df[[col_name, "string_col"]]
-    window_partition = window_spec.WindowSpec(
-        grouping_keys=(expression.deref("string_col"),),
-        ordering=(ordering.descending_over(col_name),),
-    )
-    sql_window_partition = _apply_unary_window_op(
-        bf_df_str, agg_expr, window_partition, "agg_bool"
-    )
-    snapshot.assert_match(sql_window_partition, "window_partition_out.sql")
+    snapshot.assert_match(sql_window, "out.sql")
 
 
 def test_any(scalar_types_df: bpd.DataFrame, snapshot):
+    bf_df = scalar_types_df[["bool_col", "int64_col"]]
+    ops_map = {
+        "bool_col": agg_ops.AnyOp().as_expr("bool_col"),
+        "int64_col": agg_ops.AnyOp().as_expr("int64_col"),
+    }
+    sql = _apply_unary_agg_ops(bf_df, list(ops_map.values()), list(ops_map.keys()))
+
+    snapshot.assert_match(sql, "out.sql")
+
+
+def test_any_w_window(scalar_types_df: bpd.DataFrame, snapshot):
     col_name = "bool_col"
     bf_df = scalar_types_df[[col_name]]
     agg_expr = agg_ops.AnyOp().as_expr(col_name)
-    sql = _apply_unary_agg_ops(bf_df, [agg_expr], [col_name])
-
-    snapshot.assert_match(sql, "out.sql")
 
     # Window tests
     window = window_spec.WindowSpec(ordering=(ordering.ascending_over(col_name),))
     sql_window = _apply_unary_window_op(bf_df, agg_expr, window, "agg_bool")
-    snapshot.assert_match(sql_window, "window_out.sql")
+    snapshot.assert_match(sql_window, "out.sql")
 
 
 def test_approx_quartiles(scalar_types_df: bpd.DataFrame, snapshot):