refactor: add ops.where_op to the sqlglot compiler

Author: chelsea-lin

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

@@ -23,6 +23,7 @@
 import bigframes.core.compile.sqlglot.scalar_compiler as scalar_compiler
 
 register_unary_op = scalar_compiler.scalar_op_compiler.register_unary_op
+register_ternary_op = scalar_compiler.scalar_op_compiler.register_ternary_op
 
 
 @register_unary_op(ops.AsTypeOp, pass_op=True)
@@ -94,6 +95,13 @@ def _(expr: TypedExpr) -> sge.Expression:
     return sge.Not(this=sge.Is(this=expr.expr, expression=sge.Null()))
 
 
+@register_ternary_op(ops.where_op)
+def _(
+    original: TypedExpr, condition: TypedExpr, replacement: TypedExpr
+) -> sge.Expression:
+    return sge.If(this=condition.expr, true=original.expr, false=replacement.expr)
+
+
 # Helper functions
 def _cast_to_json(expr: TypedExpr, op: ops.AsTypeOp) -> sge.Expression:
     from_type = expr.dtype

tests/system/small/engines/test_generic_ops.py

@@ -314,7 +314,7 @@ def test_engines_astype_timedelta(scalars_array_value: array_value.ArrayValue, e
     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_where_op(scalars_array_value: array_value.ArrayValue, engine):
     arr, _ = scalars_array_value.compute_values(
         [

tests/unit/core/compile/sqlglot/expressions/snapshots/test_generic_ops/test_where/out.sql

@@ -0,0 +1,15 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `bool_col` AS `bfcol_0`,
+    `int64_col` AS `bfcol_1`,
+    `float64_col` AS `bfcol_2`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+), `bfcte_1` AS (
+  SELECT
+    *,
+    IF(`bfcol_0`, `bfcol_1`, `bfcol_2`) AS `bfcol_3`
+  FROM `bfcte_0`
+)
+SELECT
+  `bfcol_3` AS `result_col`
+FROM `bfcte_1`

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

@@ -202,3 +202,19 @@ def test_map(scalar_types_df: bpd.DataFrame, snapshot):
     )
 
     snapshot.assert_match(sql, "out.sql")
+
+
+def test_where(scalar_types_df: bpd.DataFrame, snapshot):
+    op_expr = ops.where_op.as_expr("int64_col", "bool_col", "float64_col")
+
+    array_value = scalar_types_df._block.expr
+    result, col_ids = array_value.compute_values([op_expr])
+
+    # Rename columns for deterministic golden SQL results.
+    assert len(col_ids) == 1
+    result = result.rename_columns({col_ids[0]: "result_col"}).select_columns(
+        ["result_col"]
+    )
+
+    sql = result.session._executor.to_sql(result, enable_cache=False)
+    snapshot.assert_match(sql, "out.sql")