From b0f14bd048c06fa7d36bed066a7a5467ab9f6a0b Mon Sep 17 00:00:00 2001
From: Chelsea Lin <chelsealin@google.com>
Date: Wed, 17 Sep 2025 23:59:12 +0000
Subject: [PATCH] refactor: add agg_ops.MeanOp for sqlglot compiler

---
 .../sqlglot/aggregations/unary_compiler.py    | 20 ++++++++++++++
 .../system/small/engines/test_aggregation.py  |  2 +-
 .../test_unary_compiler/test_mean/out.sql     | 27 +++++++++++++++++++
 .../aggregations/test_unary_compiler.py       | 27 +++++++++++++++++++
 4 files changed, 75 insertions(+), 1 deletion(-)
 create mode 100644 tests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_mean/out.sql

diff --git a/bigframes/core/compile/sqlglot/aggregations/unary_compiler.py b/bigframes/core/compile/sqlglot/aggregations/unary_compiler.py
index 4cb0000894..8ed5510ec2 100644
--- a/bigframes/core/compile/sqlglot/aggregations/unary_compiler.py
+++ b/bigframes/core/compile/sqlglot/aggregations/unary_compiler.py
@@ -56,6 +56,26 @@ def _(
     return apply_window_if_present(sge.func("MAX", column.expr), window)
 
 
+@UNARY_OP_REGISTRATION.register(agg_ops.MeanOp)
+def _(
+    op: agg_ops.MeanOp,
+    column: typed_expr.TypedExpr,
+    window: typing.Optional[window_spec.WindowSpec] = None,
+) -> sge.Expression:
+    expr = column.expr
+    if column.dtype == dtypes.BOOL_DTYPE:
+        expr = sge.Cast(this=expr, to="INT64")
+
+    expr = sge.func("AVG", expr)
+
+    should_floor_result = (
+        op.should_floor_result or column.dtype == dtypes.TIMEDELTA_DTYPE
+    )
+    if should_floor_result:
+        expr = sge.Cast(this=sge.func("FLOOR", expr), to="INT64")
+    return apply_window_if_present(expr, window)
+
+
 @UNARY_OP_REGISTRATION.register(agg_ops.MedianOp)
 def _(
     op: agg_ops.MedianOp,
diff --git a/tests/system/small/engines/test_aggregation.py b/tests/system/small/engines/test_aggregation.py
index 98d5cd4ac8..9b4efe8cbe 100644
--- a/tests/system/small/engines/test_aggregation.py
+++ b/tests/system/small/engines/test_aggregation.py
@@ -71,7 +71,7 @@ def test_engines_aggregate_size(
     assert_equivalence_execution(node, REFERENCE_ENGINE, engine)
 
 
-@pytest.mark.parametrize("engine", ["polars", "bq"], indirect=True)
+@pytest.mark.parametrize("engine", ["polars", "bq", "bq-sqlglot"], indirect=True)
 @pytest.mark.parametrize(
     "op",
     [agg_ops.min_op, agg_ops.max_op, agg_ops.mean_op, agg_ops.sum_op, agg_ops.count_op],
diff --git a/tests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_mean/out.sql b/tests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_mean/out.sql
new file mode 100644
index 0000000000..6d4bb6f89a
--- /dev/null
+++ b/tests/unit/core/compile/sqlglot/aggregations/snapshots/test_unary_compiler/test_mean/out.sql
@@ -0,0 +1,27 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `bool_col` AS `bfcol_0`,
+    `int64_col` AS `bfcol_1`,
+    `duration_col` AS `bfcol_2`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+), `bfcte_1` AS (
+  SELECT
+    *,
+    `bfcol_1` AS `bfcol_6`,
+    `bfcol_0` AS `bfcol_7`,
+    `bfcol_2` AS `bfcol_8`
+  FROM `bfcte_0`
+), `bfcte_2` AS (
+  SELECT
+    AVG(`bfcol_6`) AS `bfcol_12`,
+    AVG(CAST(`bfcol_7` AS INT64)) AS `bfcol_13`,
+    CAST(FLOOR(AVG(`bfcol_8`)) AS INT64) AS `bfcol_14`,
+    CAST(FLOOR(AVG(`bfcol_6`)) AS INT64) AS `bfcol_15`
+  FROM `bfcte_1`
+)
+SELECT
+  `bfcol_12` AS `int64_col`,
+  `bfcol_13` AS `bool_col`,
+  `bfcol_14` AS `duration_col`,
+  `bfcol_15` AS `int64_col_w_floor`
+FROM `bfcte_2`
\ No newline at end of file
diff --git a/tests/unit/core/compile/sqlglot/aggregations/test_unary_compiler.py b/tests/unit/core/compile/sqlglot/aggregations/test_unary_compiler.py
index 4f0016a6e7..a5ffda0e65 100644
--- a/tests/unit/core/compile/sqlglot/aggregations/test_unary_compiler.py
+++ b/tests/unit/core/compile/sqlglot/aggregations/test_unary_compiler.py
@@ -56,6 +56,33 @@ def test_max(scalar_types_df: bpd.DataFrame, snapshot):
     snapshot.assert_match(sql, "out.sql")
 
 
+def test_mean(scalar_types_df: bpd.DataFrame, snapshot):
+    col_names = ["int64_col", "bool_col", "duration_col"]
+    bf_df = scalar_types_df[col_names]
+    bf_df["duration_col"] = bpd.to_timedelta(bf_df["duration_col"], unit="us")
+
+    # The `to_timedelta` creates a new mapping for the column id.
+    col_names.insert(0, "rowindex")
+    name2id = {
+        col_name: col_id
+        for col_name, col_id in zip(col_names, bf_df._block.expr.column_ids)
+    }
+
+    agg_ops_map = {
+        "int64_col": agg_ops.MeanOp().as_expr(name2id["int64_col"]),
+        "bool_col": agg_ops.MeanOp().as_expr(name2id["bool_col"]),
+        "duration_col": agg_ops.MeanOp().as_expr(name2id["duration_col"]),
+        "int64_col_w_floor": agg_ops.MeanOp(should_floor_result=True).as_expr(
+            name2id["int64_col"]
+        ),
+    }
+    sql = _apply_unary_agg_ops(
+        bf_df, list(agg_ops_map.values()), list(agg_ops_map.keys())
+    )
+
+    snapshot.assert_match(sql, "out.sql")
+
+
 def test_median(scalar_types_df: bpd.DataFrame, snapshot):
     bf_df = scalar_types_df
     ops_map = {