refactor: add struct_op for the sqlglot compiler

Author: chelsea-lin

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

@@ -24,6 +24,7 @@
 from bigframes.core.compile.sqlglot.expressions.typed_expr import TypedExpr
 import bigframes.core.compile.sqlglot.scalar_compiler as scalar_compiler
 
+register_nary_op = scalar_compiler.scalar_op_compiler.register_nary_op
 register_unary_op = scalar_compiler.scalar_op_compiler.register_unary_op
 
 
@@ -40,3 +41,13 @@ def _(expr: TypedExpr, op: ops.StructFieldOp) -> sge.Expression:
         this=sge.to_identifier(name, quoted=True),
         catalog=expr.expr,
     )
+
+
+@register_nary_op(ops.StructOp, pass_op=True)
+def _(*exprs: TypedExpr, op: ops.StructOp) -> sge.Struct:
+    return sge.Struct(
+        expressions=[
+            sge.PropertyEQ(this=sge.to_identifier(col), expression=expr.expr)
+            for col, expr in zip(op.column_names, exprs)
+        ]
+    )

tests/unit/core/compile/sqlglot/expressions/snapshots/test_struct_ops/test_struct_op/out.sql

@@ -0,0 +1,21 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `bool_col` AS `bfcol_0`,
+    `int64_col` AS `bfcol_1`,
+    `float64_col` AS `bfcol_2`,
+    `string_col` AS `bfcol_3`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+), `bfcte_1` AS (
+  SELECT
+    *,
+    STRUCT(
+      `bfcol_0` AS bool_col,
+      `bfcol_1` AS int64_col,
+      `bfcol_2` AS float64_col,
+      `bfcol_3` AS string_col
+    ) AS `bfcol_4`
+  FROM `bfcte_0`
+)
+SELECT
+  `bfcol_4` AS `result_col`
+FROM `bfcte_1`

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

@@ -12,15 +12,39 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+import typing
+
 import pytest
 
 from bigframes import operations as ops
+from bigframes.core import expression as ex
 import bigframes.pandas as bpd
 from bigframes.testing import utils
 
 pytest.importorskip("pytest_snapshot")
 
 
+def _apply_nary_op(
+    obj: bpd.DataFrame,
+    op: ops.NaryOp,
+    *args: typing.Union[str, ex.Expression],
+) -> str:
+    """Applies a nary op to the given DataFrame and return the SQL representing
+    the resulting DataFrame."""
+    array_value = obj._block.expr
+    op_expr = op.as_expr(*args)
+    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)
+    return sql
+
+
 def test_struct_field(nested_structs_types_df: bpd.DataFrame, snapshot):
     col_name = "people"
     bf_df = nested_structs_types_df[[col_name]]
@@ -34,3 +58,11 @@ def test_struct_field(nested_structs_types_df: bpd.DataFrame, snapshot):
     sql = utils._apply_unary_ops(bf_df, list(ops_map.values()), list(ops_map.keys()))
 
     snapshot.assert_match(sql, "out.sql")
+
+
+def test_struct_op(scalar_types_df: bpd.DataFrame, snapshot):
+    bf_df = scalar_types_df[["bool_col", "int64_col", "float64_col", "string_col"]]
+    op = ops.StructOp(column_names=tuple(bf_df.columns.tolist()))
+    sql = _apply_nary_op(bf_df, op, *bf_df.columns.tolist())
+
+    snapshot.assert_match(sql, "out.sql")