chore: implement mul_op and div_op compilers

bigframes/core/compile/sqlglot/expressions/binary_compiler.py

@@ -73,6 +73,51 @@ def _(op, left: TypedExpr, right: TypedExpr) -> sge.Expression:
     )
 
 
+@BINARY_OP_REGISTRATION.register(ops.div_op)
+def _(op, left: TypedExpr, right: TypedExpr) -> sge.Expression:
+    left_expr = left.expr
+    if left.dtype == dtypes.BOOL_DTYPE:
+        left_expr = sge.Cast(this=left_expr, to="INT64")
+    right_expr = right.expr
+    if right.dtype == dtypes.BOOL_DTYPE:
+        right_expr = sge.Cast(this=right_expr, to="INT64")
+
+    result = sge.func("IEEE_DIVIDE", left_expr, right_expr)
+    if left.dtype == dtypes.TIMEDELTA_DTYPE and dtypes.is_numeric(right.dtype):
+        return sge.Cast(this=sge.Floor(this=result), to="INT64")
+    else:
+        return result
+
+
+@BINARY_OP_REGISTRATION.register(ops.ge_op)
+def _(op, left: TypedExpr, right: TypedExpr) -> sge.Expression:
+    return sge.GTE(this=left.expr, expression=right.expr)
+
+
+@BINARY_OP_REGISTRATION.register(ops.JSONSet)
+def _(op, left: TypedExpr, right: TypedExpr) -> sge.Expression:
+    return sge.func("JSON_SET", left.expr, sge.convert(op.json_path), right.expr)
+
+
+@BINARY_OP_REGISTRATION.register(ops.mul_op)
+def _(op, left: TypedExpr, right: TypedExpr) -> sge.Expression:
+    left_expr = left.expr
+    if left.dtype == dtypes.BOOL_DTYPE:
+        left_expr = sge.Cast(this=left_expr, to="INT64")
+    right_expr = right.expr
+    if right.dtype == dtypes.BOOL_DTYPE:
+        right_expr = sge.Cast(this=right_expr, to="INT64")
+
+    result = sge.Mul(this=left_expr, expression=right_expr)
+
+    if (dtypes.is_numeric(left.dtype) and right.dtype == dtypes.TIMEDELTA_DTYPE) or (
+        left.dtype == dtypes.TIMEDELTA_DTYPE and dtypes.is_numeric(right.dtype)
+    ):
+        return sge.Cast(this=sge.Floor(this=result), to="INT64")
+    else:
+        return result
+
+
 @BINARY_OP_REGISTRATION.register(ops.sub_op)
 def _(op, left: TypedExpr, right: TypedExpr) -> sge.Expression:
     if dtypes.is_numeric(left.dtype) and dtypes.is_numeric(right.dtype):
@@ -113,13 +158,3 @@ def _(op, left: TypedExpr, right: TypedExpr) -> sge.Expression:
     raise TypeError(
         f"Cannot subtract type {left.dtype} and {right.dtype}. {constants.FEEDBACK_LINK}"
     )
-
-
-@BINARY_OP_REGISTRATION.register(ops.ge_op)
-def _(op, left: TypedExpr, right: TypedExpr) -> sge.Expression:
-    return sge.GTE(this=left.expr, expression=right.expr)
-
-
-@BINARY_OP_REGISTRATION.register(ops.JSONSet)
-def _(op, left: TypedExpr, right: TypedExpr) -> sge.Expression:
-    return sge.func("JSON_SET", left.expr, sge.convert(op.json_path), right.expr)

tests/system/small/engines/test_numeric_ops.py

@@ -71,7 +71,7 @@ def test_engines_project_sub(
     assert_equivalence_execution(arr.node, REFERENCE_ENGINE, engine)
 
 
-@pytest.mark.parametrize("engine", ["polars", "bq"], indirect=True)
+@pytest.mark.parametrize("engine", ["polars", "bq", "bq-sqlglot"], indirect=True)
 def test_engines_project_mul(
     scalars_array_value: array_value.ArrayValue,
     engine,
@@ -80,7 +80,7 @@ def test_engines_project_mul(
     assert_equivalence_execution(arr.node, REFERENCE_ENGINE, engine)
 
 
-@pytest.mark.parametrize("engine", ["polars", "bq"], indirect=True)
+@pytest.mark.parametrize("engine", ["polars", "bq", "bq-sqlglot"], indirect=True)
 def test_engines_project_div(scalars_array_value: array_value.ArrayValue, engine):
     # TODO: Duration div is sensitive to zeroes
     # TODO: Numeric col is sensitive to scale shifts
@@ -90,7 +90,7 @@ def test_engines_project_div(scalars_array_value: array_value.ArrayValue, engine
     assert_equivalence_execution(arr.node, REFERENCE_ENGINE, engine)
 
 
-@pytest.mark.parametrize("engine", ["polars", "bq"], indirect=True)
+@pytest.mark.parametrize("engine", ["polars", "bq", "bq-sqlglot"], indirect=True)
 def test_engines_project_div_durations(
     scalars_array_value: array_value.ArrayValue, engine
 ):

tests/unit/core/compile/sqlglot/expressions/snapshots/test_binary_compiler/test_div_numeric/out.sql

@@ -0,0 +1,54 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `bool_col` AS `bfcol_0`,
+    `int64_col` AS `bfcol_1`,
+    `rowindex` AS `bfcol_2`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+), `bfcte_1` AS (
+  SELECT
+    *,
+    `bfcol_2` AS `bfcol_6`,
+    `bfcol_1` AS `bfcol_7`,
+    `bfcol_0` AS `bfcol_8`,
+    IEEE_DIVIDE(`bfcol_1`, `bfcol_1`) AS `bfcol_9`
+  FROM `bfcte_0`
+), `bfcte_2` AS (
+  SELECT
+    *,
+    `bfcol_6` AS `bfcol_14`,
+    `bfcol_7` AS `bfcol_15`,
+    `bfcol_8` AS `bfcol_16`,
+    `bfcol_9` AS `bfcol_17`,
+    IEEE_DIVIDE(`bfcol_7`, 1) AS `bfcol_18`
+  FROM `bfcte_1`
+), `bfcte_3` AS (
+  SELECT
+    *,
+    `bfcol_14` AS `bfcol_24`,
+    `bfcol_15` AS `bfcol_25`,
+    `bfcol_16` AS `bfcol_26`,
+    `bfcol_17` AS `bfcol_27`,
+    `bfcol_18` AS `bfcol_28`,
+    IEEE_DIVIDE(`bfcol_15`, CAST(`bfcol_16` AS INT64)) AS `bfcol_29`
+  FROM `bfcte_2`
+), `bfcte_4` AS (
+  SELECT
+    *,
+    `bfcol_24` AS `bfcol_36`,
+    `bfcol_25` AS `bfcol_37`,
+    `bfcol_26` AS `bfcol_38`,
+    `bfcol_27` AS `bfcol_39`,
+    `bfcol_28` AS `bfcol_40`,
+    `bfcol_29` AS `bfcol_41`,
+    IEEE_DIVIDE(CAST(`bfcol_26` AS INT64), `bfcol_25`) AS `bfcol_42`
+  FROM `bfcte_3`
+)
+SELECT
+  `bfcol_36` AS `rowindex`,
+  `bfcol_37` AS `int64_col`,
+  `bfcol_38` AS `bool_col`,
+  `bfcol_39` AS `int_div_int`,
+  `bfcol_40` AS `int_div_1`,
+  `bfcol_41` AS `int_div_bool`,
+  `bfcol_42` AS `bool_div_int`
+FROM `bfcte_4`

tests/unit/core/compile/sqlglot/expressions/snapshots/test_binary_compiler/test_div_timedelta/out.sql

@@ -0,0 +1,21 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `int64_col` AS `bfcol_0`,
+    `rowindex` AS `bfcol_1`,
+    `timestamp_col` AS `bfcol_2`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+), `bfcte_1` AS (
+  SELECT
+    *,
+    `bfcol_1` AS `bfcol_6`,
+    `bfcol_2` AS `bfcol_7`,
+    `bfcol_0` AS `bfcol_8`,
+    CAST(FLOOR(IEEE_DIVIDE(86400000000, `bfcol_0`)) AS INT64) AS `bfcol_9`
+  FROM `bfcte_0`
+)
+SELECT
+  `bfcol_6` AS `rowindex`,
+  `bfcol_7` AS `timestamp_col`,
+  `bfcol_8` AS `int64_col`,
+  `bfcol_9` AS `timedelta_div_numeric`
+FROM `bfcte_1`

tests/unit/core/compile/sqlglot/expressions/snapshots/test_binary_compiler/test_mul_numeric/out.sql

@@ -0,0 +1,54 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `bool_col` AS `bfcol_0`,
+    `int64_col` AS `bfcol_1`,
+    `rowindex` AS `bfcol_2`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+), `bfcte_1` AS (
+  SELECT
+    *,
+    `bfcol_2` AS `bfcol_6`,
+    `bfcol_1` AS `bfcol_7`,
+    `bfcol_0` AS `bfcol_8`,
+    `bfcol_1` * `bfcol_1` AS `bfcol_9`
+  FROM `bfcte_0`
+), `bfcte_2` AS (
+  SELECT
+    *,
+    `bfcol_6` AS `bfcol_14`,
+    `bfcol_7` AS `bfcol_15`,
+    `bfcol_8` AS `bfcol_16`,
+    `bfcol_9` AS `bfcol_17`,
+    `bfcol_7` * 1 AS `bfcol_18`
+  FROM `bfcte_1`
+), `bfcte_3` AS (
+  SELECT
+    *,
+    `bfcol_14` AS `bfcol_24`,
+    `bfcol_15` AS `bfcol_25`,
+    `bfcol_16` AS `bfcol_26`,
+    `bfcol_17` AS `bfcol_27`,
+    `bfcol_18` AS `bfcol_28`,
+    `bfcol_15` * CAST(`bfcol_16` AS INT64) AS `bfcol_29`
+  FROM `bfcte_2`
+), `bfcte_4` AS (
+  SELECT
+    *,
+    `bfcol_24` AS `bfcol_36`,
+    `bfcol_25` AS `bfcol_37`,
+    `bfcol_26` AS `bfcol_38`,
+    `bfcol_27` AS `bfcol_39`,
+    `bfcol_28` AS `bfcol_40`,
+    `bfcol_29` AS `bfcol_41`,
+    CAST(`bfcol_26` AS INT64) * `bfcol_25` AS `bfcol_42`
+  FROM `bfcte_3`
+)
+SELECT
+  `bfcol_36` AS `rowindex`,
+  `bfcol_37` AS `int64_col`,
+  `bfcol_38` AS `bool_col`,
+  `bfcol_39` AS `int_mul_int`,
+  `bfcol_40` AS `int_mul_1`,
+  `bfcol_41` AS `int_mul_bool`,
+  `bfcol_42` AS `bool_mul_int`
+FROM `bfcte_4`

tests/unit/core/compile/sqlglot/expressions/snapshots/test_binary_compiler/test_mul_timedelta/out.sql

@@ -0,0 +1,31 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `int64_col` AS `bfcol_0`,
+    `rowindex` AS `bfcol_1`,
+    `timestamp_col` AS `bfcol_2`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+), `bfcte_1` AS (
+  SELECT
+    *,
+    `bfcol_1` AS `bfcol_6`,
+    `bfcol_2` AS `bfcol_7`,
+    `bfcol_0` AS `bfcol_8`,
+    CAST(FLOOR(86400000000 * `bfcol_0`) AS INT64) AS `bfcol_9`
+  FROM `bfcte_0`
+), `bfcte_2` AS (
+  SELECT
+    *,
+    `bfcol_6` AS `bfcol_14`,
+    `bfcol_7` AS `bfcol_15`,
+    `bfcol_8` AS `bfcol_16`,
+    `bfcol_9` AS `bfcol_17`,
+    CAST(FLOOR(`bfcol_8` * 86400000000) AS INT64) AS `bfcol_18`
+  FROM `bfcte_1`
+)
+SELECT
+  `bfcol_14` AS `rowindex`,
+  `bfcol_15` AS `timestamp_col`,
+  `bfcol_16` AS `int64_col`,
+  `bfcol_17` AS `timedelta_mul_numeric`,
+  `bfcol_18` AS `numeric_mul_timedelta`
+FROM `bfcte_2`

tests/unit/core/compile/sqlglot/expressions/test_binary_compiler.py

@@ -82,6 +82,26 @@ def test_add_unsupported_raises(scalar_types_df: bpd.DataFrame):
         _apply_binary_op(scalar_types_df, ops.add_op, "int64_col", "string_col")
 
 
+def test_div_numeric(scalar_types_df: bpd.DataFrame, snapshot):
+    bf_df = scalar_types_df[["int64_col", "bool_col"]]
+
+    bf_df["int_div_int"] = bf_df["int64_col"] / bf_df["int64_col"]
+    bf_df["int_div_1"] = bf_df["int64_col"] / 1
+
+    bf_df["int_div_bool"] = bf_df["int64_col"] / bf_df["bool_col"]
+    bf_df["bool_div_int"] = bf_df["bool_col"] / bf_df["int64_col"]
+
+    snapshot.assert_match(bf_df.sql, "out.sql")
+
+
+def test_div_timedelta(scalar_types_df: bpd.DataFrame, snapshot):
+    bf_df = scalar_types_df[["timestamp_col", "int64_col"]]
+    timedelta = pd.Timedelta(1, unit="d")
+    bf_df["timedelta_div_numeric"] = timedelta / bf_df["int64_col"]
+
+    snapshot.assert_match(bf_df.sql, "out.sql")
+
+
 def test_json_set(json_types_df: bpd.DataFrame, snapshot):
     bf_df = json_types_df[["json_col"]]
     sql = _apply_binary_op(
@@ -122,3 +142,25 @@ def test_sub_unsupported_raises(scalar_types_df: bpd.DataFrame):
 
     with pytest.raises(TypeError):
         _apply_binary_op(scalar_types_df, ops.sub_op, "int64_col", "string_col")
+
+
+def test_mul_numeric(scalar_types_df: bpd.DataFrame, snapshot):
+    bf_df = scalar_types_df[["int64_col", "bool_col"]]
+
+    bf_df["int_mul_int"] = bf_df["int64_col"] * bf_df["int64_col"]
+    bf_df["int_mul_1"] = bf_df["int64_col"] * 1
+
+    bf_df["int_mul_bool"] = bf_df["int64_col"] * bf_df["bool_col"]
+    bf_df["bool_mul_int"] = bf_df["bool_col"] * bf_df["int64_col"]
+
+    snapshot.assert_match(bf_df.sql, "out.sql")
+
+
+def test_mul_timedelta(scalar_types_df: bpd.DataFrame, snapshot):
+    bf_df = scalar_types_df[["timestamp_col", "int64_col"]]
+    timedelta = pd.Timedelta(1, unit="d")
+
+    bf_df["timedelta_mul_numeric"] = timedelta * bf_df["int64_col"]
+    bf_df["numeric_mul_timedelta"] = bf_df["int64_col"] * timedelta
+
+    snapshot.assert_match(bf_df.sql, "out.sql")