refactor: add agg_ops.ProductOp to the sqlglot compiler

Author: chelsea-lin

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

@@ -411,6 +411,58 @@ def _(
     return apply_window_if_present(expr, window)
 
 
+@UNARY_OP_REGISTRATION.register(agg_ops.ProductOp)
+def _(
+    op: agg_ops.ProductOp,
+    column: typed_expr.TypedExpr,
+    window: typing.Optional[window_spec.WindowSpec] = None,
+) -> sge.Expression:
+    # Need to short-circuit as log with zeroes is illegal sql
+    is_zero = sge.EQ(this=column.expr, expression=sge.convert(0))
+
+    # There is no product sql aggregate function, so must implement as a sum of logs, and then
+    # apply power after. Note, log and power base must be equal! This impl uses natural log.
+    logs = (
+        sge.Case()
+        .when(is_zero, sge.convert(0))
+        .else_(sge.func("LN", sge.func("ABS", column.expr)))
+    )
+    logs_sum = apply_window_if_present(sge.func("SUM", logs), window)
+    magnitude = sge.func("EXP", logs_sum)
+
+    # Can't determine sign from logs, so have to determine parity of count of negative inputs
+    is_negative = (
+        sge.Case()
+        .when(
+            sge.LT(this=sge.func("SIGN", column.expr), expression=sge.convert(0)),
+            sge.convert(1),
+        )
+        .else_(sge.convert(0))
+    )
+    negative_count = apply_window_if_present(sge.func("SUM", is_negative), window)
+    negative_count_parity = sge.Mod(
+        this=negative_count, expression=sge.convert(2)
+    )  # 1 if result should be negative, otherwise 0
+
+    any_zeroes = apply_window_if_present(sge.func("LOGICAL_OR", is_zero), window)
+
+    float_result = (
+        sge.Case()
+        .when(any_zeroes, sge.convert(0))
+        .else_(
+            sge.Mul(
+                this=magnitude,
+                expression=sge.If(
+                    this=sge.EQ(this=negative_count_parity, expression=sge.convert(1)),
+                    true=sge.convert(-1),
+                    false=sge.convert(1),
+                ),
+            )
+        )
+    )
+    return float_result
+
+
 @UNARY_OP_REGISTRATION.register(agg_ops.QcutOp)
 def _(
     op: agg_ops.QcutOp,

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

@@ -0,0 +1,16 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `int64_col`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+), `bfcte_1` AS (
+  SELECT
+    CASE
+      WHEN LOGICAL_OR(`int64_col` = 0)
+      THEN 0
+      ELSE EXP(SUM(CASE WHEN `int64_col` = 0 THEN 0 ELSE LN(ABS(`int64_col`)) END)) * IF(MOD(SUM(CASE WHEN SIGN(`int64_col`) < 0 THEN 1 ELSE 0 END), 2) = 1, -1, 1)
+    END AS `bfcol_1`
+  FROM `bfcte_0`
+)
+SELECT
+  `bfcol_1` AS `int64_col`
+FROM `bfcte_1`

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

@@ -0,0 +1,27 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `int64_col`,
+    `string_col`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+), `bfcte_1` AS (
+  SELECT
+    *,
+    CASE
+      WHEN LOGICAL_OR(`int64_col` = 0) OVER (PARTITION BY `string_col`)
+      THEN 0
+      ELSE EXP(
+        SUM(CASE WHEN `int64_col` = 0 THEN 0 ELSE LN(ABS(`int64_col`)) END) OVER (PARTITION BY `string_col`)
+      ) * IF(
+        MOD(
+          SUM(CASE WHEN SIGN(`int64_col`) < 0 THEN 1 ELSE 0 END) OVER (PARTITION BY `string_col`),
+          2
+        ) = 1,
+        -1,
+        1
+      )
+    END AS `bfcol_2`
+  FROM `bfcte_0`
+)
+SELECT
+  `bfcol_2` AS `agg_int64`
+FROM `bfcte_1`

tests/unit/core/compile/sqlglot/aggregations/test_unary_compiler.py

@@ -443,6 +443,25 @@ def test_pop_var(scalar_types_df: bpd.DataFrame, snapshot):
     snapshot.assert_match(sql_window, "window_out.sql")
 
 
+def test_product(scalar_types_df: bpd.DataFrame, snapshot):
+    col_name = "int64_col"
+    bf_df = scalar_types_df[[col_name]]
+    agg_expr = agg_ops.ProductOp().as_expr(col_name)
+    sql = _apply_unary_agg_ops(bf_df, [agg_expr], [col_name])
+
+    snapshot.assert_match(sql, "out.sql")
+
+    bf_df_str = scalar_types_df[[col_name, "string_col"]]
+    window_partition = window_spec.WindowSpec(
+        grouping_keys=(expression.deref("string_col"),),
+    )
+    sql_window_partition = _apply_unary_window_op(
+        bf_df_str, agg_expr, window_partition, "agg_int64"
+    )
+
+    snapshot.assert_match(sql_window_partition, "window_partition_out.sql")
+
+
 def test_qcut(scalar_types_df: bpd.DataFrame, snapshot):
     if sys.version_info < (3, 12):
         pytest.skip(