select star if possible when compiling readtable

Author: chelsea-lin

bigframes/core/compile/sqlglot/compiler.py

@@ -176,12 +176,14 @@ def compile_readlocal(node: nodes.ReadLocalNode, child: ir.SQLGlotIR) -> ir.SQLG
 @_compile_node.register
 def compile_readtable(node: nodes.ReadTableNode, child: ir.SQLGlotIR):
     table = node.source.table
+    is_star_selection = node.is_star_selection
     return ir.SQLGlotIR.from_table(
         table.project_id,
         table.dataset_id,
         table.table_id,
         col_names=[col.source_id for col in node.scan_list.items],
         alias_names=[col.id.sql for col in node.scan_list.items],
+        is_star_selection=is_star_selection,
         uid_gen=child.uid_gen,
         sql_predicate=node.source.sql_predicate,
         system_time=node.source.at_time,

bigframes/core/compile/sqlglot/sqlglot_ir.py

@@ -118,6 +118,7 @@ def from_table(
         table_id: str,
         col_names: typing.Sequence[str],
         alias_names: typing.Sequence[str],
+        is_star_selection: bool,
         uid_gen: guid.SequentialUIDGenerator,
         sql_predicate: typing.Optional[str] = None,
         system_time: typing.Optional[datetime.datetime] = None,
@@ -134,15 +135,7 @@ def from_table(
             sql_predicate (typing.Optional[str]): An optional SQL predicate for filtering.
             system_time (typing.Optional[str]): An optional system time for time-travel queries.
         """
-        selections = [
-            sge.Alias(
-                this=sge.to_identifier(col_name, quoted=cls.quoted),
-                alias=sge.to_identifier(alias_name, quoted=cls.quoted),
-            )
-            if col_name != alias_name
-            else sge.to_identifier(col_name, quoted=cls.quoted)
-            for col_name, alias_name in zip(col_names, alias_names)
-        ]
+
         version = (
             sge.Version(
                 this="TIMESTAMP",
@@ -158,7 +151,19 @@ def from_table(
             catalog=sg.to_identifier(project_id, quoted=cls.quoted),
             version=version,
         )
-        select_expr = sge.Select().select(*selections).from_(table_expr)
+        if is_star_selection:
+            select_expr = sge.Select().select(sge.Star()).from_(table_expr)
+        else:
+            selections = [
+                sge.Alias(
+                    this=sge.to_identifier(col_name, quoted=cls.quoted),
+                    alias=sge.to_identifier(alias_name, quoted=cls.quoted),
+                )
+                if col_name != alias_name
+                else sge.to_identifier(col_name, quoted=cls.quoted)
+                for col_name, alias_name in zip(col_names, alias_names)
+            ]
+            select_expr = sge.Select().select(*selections).from_(table_expr)
         if sql_predicate:
             select_expr = select_expr.where(
                 sg.parse_one(sql_predicate, dialect="bigquery"), append=False

bigframes/core/nodes.py

@@ -819,6 +819,12 @@ def order_ambiguous(self) -> bool:
     def explicitly_ordered(self) -> bool:
         return self.source.ordering is not None
 
+    @property
+    def is_star_selection(self) -> bool:
+        physical_names = tuple(item.name for item in self.source.table.physical_schema)
+        scan_names = tuple(item.source_id for item in self.scan_list.items)
+        return physical_names == scan_names
+
     @functools.cached_property
     def variables_introduced(self) -> int:
         return len(self.scan_list.items) + 1

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_nullary_compiler/test_row_number/out.sql

@@ -1,20 +1,6 @@
 WITH `bfcte_0` AS (
   SELECT
-    `bool_col`,
-    `bytes_col`,
-    `date_col`,
-    `datetime_col`,
-    `geography_col`,
-    `int64_col`,
-    `int64_too`,
-    `numeric_col`,
-    `float64_col`,
-    `rowindex`,
-    `rowindex_2`,
-    `string_col`,
-    `time_col`,
-    `timestamp_col`,
-    `duration_col`
+    *
   FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
 ), `bfcte_1` AS (
   SELECT

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_nullary_compiler/test_size/out.sql

@@ -1,20 +1,6 @@
 WITH `bfcte_0` AS (
   SELECT
-    `bool_col`,
-    `bytes_col`,
-    `date_col`,
-    `datetime_col`,
-    `geography_col`,
-    `int64_col`,
-    `int64_too`,
-    `numeric_col`,
-    `float64_col`,
-    `rowindex`,
-    `rowindex_2`,
-    `string_col`,
-    `time_col`,
-    `timestamp_col`,
-    `duration_col`
+    *
   FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
 ), `bfcte_1` AS (
   SELECT

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

@@ -1,20 +1,6 @@
 WITH `bfcte_0` AS (
   SELECT
-    `bool_col`,
-    `bytes_col`,
-    `date_col`,
-    `datetime_col`,
-    `geography_col`,
-    `int64_col`,
-    `int64_too`,
-    `numeric_col`,
-    `float64_col`,
-    `rowindex`,
-    `rowindex_2`,
-    `string_col`,
-    `time_col`,
-    `timestamp_col`,
-    `duration_col`
+    *
   FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
 ), `bfcte_1` AS (
   SELECT

tests/unit/core/compile/sqlglot/snapshots/test_compile_readtable/test_compile_readtable/out.sql

@@ -1,20 +1,6 @@
 WITH `bfcte_0` AS (
   SELECT
-    `bool_col`,
-    `bytes_col`,
-    `date_col`,
-    `datetime_col`,
-    `geography_col`,
-    `int64_col`,
-    `int64_too`,
-    `numeric_col`,
-    `float64_col`,
-    `rowindex`,
-    `rowindex_2`,
-    `string_col`,
-    `time_col`,
-    `timestamp_col`,
-    `duration_col`
+    *
   FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
 )
 SELECT

tests/unit/core/compile/sqlglot/snapshots/test_compile_readtable/test_compile_readtable_w_json_types/out.sql

@@ -1,7 +1,6 @@
 WITH `bfcte_0` AS (
   SELECT
-    `rowindex`,
-    `json_col`
+    *
   FROM `bigframes-dev`.`sqlglot_test`.`json_types`
 )
 SELECT

tests/unit/core/compile/sqlglot/snapshots/test_compile_readtable/test_compile_readtable_w_nested_structs_types/out.sql

@@ -1,7 +1,6 @@
 WITH `bfcte_0` AS (
   SELECT
-    `id`,
-    `people`
+    *
   FROM `bigframes-dev`.`sqlglot_test`.`nested_structs_types`
 )
 SELECT

tests/unit/core/compile/sqlglot/snapshots/test_compile_readtable/test_compile_readtable_w_repeated_types/out.sql

@@ -1,13 +1,6 @@
 WITH `bfcte_0` AS (
   SELECT
-    `rowindex`,
-    `int_list_col`,
-    `bool_list_col`,
-    `float_list_col`,
-    `date_list_col`,
-    `date_time_list_col`,
-    `numeric_list_col`,
-    `string_list_col`
+    *
   FROM `bigframes-dev`.`sqlglot_test`.`repeated_types`
 )
 SELECT