refactor: read table with system time for sqlglot compiler (#2252)

Fixes internal issue 459579873🦕

Author: chelsea-lin

bigframes/core/compile/sqlglot/compiler.py

@@ -172,6 +172,7 @@ def compile_readtable(node: nodes.ReadTableNode, child: ir.SQLGlotIR):
         col_names=[col.source_id for col in node.scan_list.items],
         alias_names=[col.id.sql for col in node.scan_list.items],
         uid_gen=child.uid_gen,
+        system_time=node.source.at_time,
     )
 
 

bigframes/core/compile/sqlglot/sqlglot_ir.py

@@ -15,6 +15,7 @@
 from __future__ import annotations
 
 import dataclasses
+import datetime
 import functools
 import typing
 
@@ -118,6 +119,7 @@ def from_table(
         col_names: typing.Sequence[str],
         alias_names: typing.Sequence[str],
         uid_gen: guid.SequentialUIDGenerator,
+        system_time: typing.Optional[datetime.datetime] = None,
     ) -> SQLGlotIR:
         """Builds a SQLGlotIR expression from a BigQuery table.
 
@@ -128,6 +130,7 @@ def from_table(
             col_names (typing.Sequence[str]): The names of the columns to select.
             alias_names (typing.Sequence[str]): The aliases for the selected columns.
             uid_gen (guid.SequentialUIDGenerator): A generator for unique identifiers.
+            system_time (typing.Optional[str]): An optional system time for time-travel queries.
         """
         selections = [
             sge.Alias(
@@ -138,10 +141,20 @@ def from_table(
             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",
+                expression=sge.Literal(this=system_time.isoformat(), is_string=True),
+                kind="AS OF",
+            )
+            if system_time
+            else None
+        )
         table_expr = sge.Table(
             this=sg.to_identifier(table_id, quoted=cls.quoted),
             db=sg.to_identifier(dataset_id, quoted=cls.quoted),
             catalog=sg.to_identifier(project_id, quoted=cls.quoted),
+            version=version,
         )
         select_expr = sge.Select().select(*selections).from_(table_expr)
         return cls(expr=select_expr, uid_gen=uid_gen)

tests/unit/core/compile/sqlglot/conftest.py

@@ -97,7 +97,10 @@ def scalar_types_table_schema() -> typing.Sequence[bigquery.SchemaField]:
 def scalar_types_df(compiler_session) -> bpd.DataFrame:
     """Returns a BigFrames DataFrame containing all scalar types and using the `rowindex`
     column as the index."""
-    bf_df = compiler_session.read_gbq_table("bigframes-dev.sqlglot_test.scalar_types")
+    bf_df = compiler_session._loader.read_gbq_table(
+        "bigframes-dev.sqlglot_test.scalar_types",
+        enable_snapshot=False,
+    )
     bf_df = bf_df.set_index("rowindex", drop=False)
     return bf_df
 
@@ -154,8 +157,9 @@ def nested_structs_types_table_schema() -> typing.Sequence[bigquery.SchemaField]
 def nested_structs_types_df(compiler_session_w_nested_structs_types) -> bpd.DataFrame:
     """Returns a BigFrames DataFrame containing all scalar types and using the `rowindex`
     column as the index."""
-    bf_df = compiler_session_w_nested_structs_types.read_gbq_table(
-        "bigframes-dev.sqlglot_test.nested_structs_types"
+    bf_df = compiler_session_w_nested_structs_types._loader.read_gbq_table(
+        "bigframes-dev.sqlglot_test.nested_structs_types",
+        enable_snapshot=False,
     )
     bf_df = bf_df.set_index("id", drop=False)
     return bf_df
@@ -204,8 +208,9 @@ def repeated_types_table_schema() -> typing.Sequence[bigquery.SchemaField]:
 def repeated_types_df(compiler_session_w_repeated_types) -> bpd.DataFrame:
     """Returns a BigFrames DataFrame containing all scalar types and using the `rowindex`
     column as the index."""
-    bf_df = compiler_session_w_repeated_types.read_gbq_table(
-        "bigframes-dev.sqlglot_test.repeated_types"
+    bf_df = compiler_session_w_repeated_types._loader.read_gbq_table(
+        "bigframes-dev.sqlglot_test.repeated_types",
+        enable_snapshot=False,
     )
     bf_df = bf_df.set_index("rowindex", drop=False)
     return bf_df
@@ -237,8 +242,9 @@ def json_types_table_schema() -> typing.Sequence[bigquery.SchemaField]:
 def json_types_df(compiler_session_w_json_types) -> bpd.DataFrame:
     """Returns a BigFrames DataFrame containing JSON types and using the `rowindex`
     column as the index."""
-    bf_df = compiler_session_w_json_types.read_gbq_table(
-        "bigframes-dev.sqlglot_test.json_types"
+    bf_df = compiler_session_w_json_types._loader.read_gbq_table(
+        "bigframes-dev.sqlglot_test.json_types",
+        enable_snapshot=False,
     )
     # TODO(b/427305807): Why `drop=False` will produce two "rowindex" columns?
     bf_df = bf_df.set_index("rowindex", drop=True)

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

@@ -0,0 +1,36 @@
+WITH `bfcte_0` AS (
+  SELECT
+    `bool_col`,
+    `bytes_col`,
+    `date_col`,
+    `datetime_col`,
+    `duration_col`,
+    `float64_col`,
+    `geography_col`,
+    `int64_col`,
+    `int64_too`,
+    `numeric_col`,
+    `rowindex`,
+    `rowindex_2`,
+    `string_col`,
+    `time_col`,
+    `timestamp_col`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types` FOR SYSTEM_TIME AS OF '2025-11-09T03:04:05.678901+00:00'
+)
+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 `bfcte_0`

tests/unit/core/compile/sqlglot/test_compile_readtable.py

@@ -12,6 +12,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+import datetime
+
+import google.cloud.bigquery as bigquery
 import pytest
 
 import bigframes.pandas as bpd
@@ -47,3 +50,20 @@ def test_compile_readtable_w_limit(scalar_types_df: bpd.DataFrame, snapshot):
     bf_df = scalar_types_df[["int64_col"]]
     bf_df = bf_df.sort_index().head(10)
     snapshot.assert_match(bf_df.sql, "out.sql")
+
+
+def test_compile_readtable_w_system_time(
+    compiler_session, scalar_types_table_schema, snapshot
+):
+    table_ref = bigquery.TableReference(
+        bigquery.DatasetReference("bigframes-dev", "sqlglot_test"),
+        "scalar_types",
+    )
+    table = bigquery.Table(table_ref, tuple(scalar_types_table_schema))
+    table._properties["location"] = compiler_session._location
+    compiler_session._loader._df_snapshot[str(table_ref)] = (
+        datetime.datetime(2025, 11, 9, 3, 4, 5, 678901, tzinfo=datetime.timezone.utc),
+        table,
+    )
+    bf_df = compiler_session.read_gbq_table(str(table_ref))
+    snapshot.assert_match(bf_df.sql, "out.sql")

tests/unit/session/test_session.py

@@ -240,7 +240,6 @@ def test_read_gbq_cached_table():
     )
     table._properties["location"] = session._location
     table._properties["numRows"] = "1000000000"
-    table._properties["location"] = session._location
     table._properties["type"] = "TABLE"
     session._loader._df_snapshot[str(table_ref)] = (
         datetime.datetime(1999, 1, 2, 3, 4, 5, 678901, tzinfo=datetime.timezone.utc),