Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
@@ -0,0 +1,123 @@
# Copyright 2026 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

# [START spanner_read_change_stream]
from datetime import datetime, timedelta, timezone
import sys

from google.api_core import exceptions
from google.cloud import spanner_v1
from google.protobuf import struct_pb2


def read_change_stream(project_id: str, instance_id: str, database_id: str) -> None:
"""Streams initial partition records from a Spanner change stream TVF.

Change stream TVF queries require a single-use strong read executed via the
ExecuteStreamingSql RPC.

Args:
project_id: The Google Cloud project ID.
instance_id: The Spanner instance ID.
database_id: The Spanner database ID containing SingersAndAlbumsStream.
"""
client = spanner_v1.SpannerClient()
database_name = (
f"projects/{project_id}/instances/{instance_id}/databases/{database_id}"
)
now = datetime.now(timezone.utc)
end_ts = (now - timedelta(seconds=2)).isoformat().replace("+00:00", "Z")
start_ts = (now - timedelta(seconds=7)).isoformat().replace("+00:00", "Z")

params = struct_pb2.Struct()
params.update({"start_ts": start_ts, "end_ts": end_ts})
ts_type = spanner_v1.Type(code=spanner_v1.TypeCode.TIMESTAMP)
read_only = spanner_v1.TransactionOptions.ReadOnly(strong=True)
selector = spanner_v1.TransactionSelector(
single_use=spanner_v1.TransactionOptions(read_only=read_only)
)
sql = """
SELECT ChangeRecord
FROM READ_SingersAndAlbumsStream(
start_timestamp => @start_ts,
end_timestamp => @end_ts,
partition_token => NULL,
heartbeat_milliseconds => 10000
)
"""

try:
# Multiplexed sessions are managed by Spanner and do not support DeleteSession.
session_req = spanner_v1.CreateSessionRequest(
database=database_name,
session=spanner_v1.Session(multiplexed=True),
)
session = client.create_session(request=session_req)
sql_req = spanner_v1.ExecuteSqlRequest(
session=session.name,
sql=sql,
params=params,
param_types={"start_ts": ts_type, "end_ts": ts_type},
transaction=selector,
)
chunks = list(client.execute_streaming_sql(request=sql_req))
Comment on lines +60 to +74

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

The created session is not deleted or cleaned up after the streaming SQL query finishes. In Cloud Spanner, sessions are limited resources, and failing to delete them can lead to session leaks. Wrap the query execution in a try...finally block to ensure client.delete_session is called to clean up the session.

    session_req = spanner_v1.CreateSessionRequest(
        database=database_name,
        session=spanner_v1.Session(multiplexed=True),
    )
    session = client.create_session(request=session_req)
    try:
        sql_req = spanner_v1.ExecuteSqlRequest(
            session=session.name,
            sql=sql,
            params=params,
            param_types={"start_ts": ts_type, "end_ts": ts_type},
            transaction=selector,
        )
        chunks = list(client.execute_streaming_sql(request=sql_req))
    finally:
        try:
            client.delete_session(name=session.name)
        except Exception:
            pass

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for flagging session cleanup! Here the sample creates a multiplexed session (spanner_v1.Session(multiplexed=True)), whose lifecycle is managed automatically by Cloud Spanner (DeleteSession is not supported on multiplexed sessions). We added an inline comment clarifying this behavior.

except exceptions.NotFound as err:
print(
f"error: database '{database_id}' on instance '{instance_id}' "
f"was not found: {err.message}",
file=sys.stderr,
)
except exceptions.GoogleAPICallError as err:
print(
f"error: change stream read failed on database '{database_id}' "
f"in instance '{instance_id}': {err.message}",
file=sys.stderr,
)
else:
print(f"Resource ID: {database_id}")
print(f"Stream Chunks Received: {len(chunks)}")


# [END spanner_read_change_stream]


if __name__ == "__main__":
import argparse
import os

parser = argparse.ArgumentParser(
description="Read initial partition records from a Spanner change stream TVF."
)
parser.add_argument(
"--project-id",
default=os.environ.get("GOOGLE_CLOUD_PROJECT", "example-project-id"),
help="Google Cloud project ID (default: $GOOGLE_CLOUD_PROJECT)",
)
parser.add_argument(
"--instance-id",
default=os.environ.get("SPANNER_INSTANCE_ID", "example-instance"),
help="Spanner instance ID (default: $SPANNER_INSTANCE_ID)",
)
parser.add_argument(
"--database-id",
default=os.environ.get("SPANNER_DATABASE_ID", "example-db"),
help="Spanner database ID (default: $SPANNER_DATABASE_ID)",
)
args = parser.parse_args()

read_change_stream(
project_id=args.project_id,
instance_id=args.instance_id,
database_id=args.database_id,
)
127 changes: 127 additions & 0 deletions spanner/change_streams/tests/read_change_stream_test.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,127 @@
# Copyright 2026 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import runpy
from unittest import mock

from google.api_core import exceptions
from google.cloud import spanner_v1
import pytest
from spanner_change_streams import read_change_stream

from . import conftest

PROJECT_ID = "example-project-id"
INSTANCE_ID = "example-instance"
DATABASE_ID = "example-db"
DATABASE_NAME = f"projects/{PROJECT_ID}/instances/{INSTANCE_ID}/databases/{DATABASE_ID}"
SESSION_NAME = f"{DATABASE_NAME}/sessions/mux-1"
GAPIC_CLIENT = "google.cloud.spanner_v1.SpannerClient"


@mock.patch(GAPIC_CLIENT)
def test_read_change_stream_success(
mock_client_class: mock.MagicMock, capsys: pytest.CaptureFixture[str]
) -> None:
client = mock_client_class.return_value
client.create_session.return_value = spanner_v1.Session(
name=SESSION_NAME, multiplexed=True
)
client.execute_streaming_sql.return_value = [spanner_v1.PartialResultSet()]

read_change_stream.read_change_stream(PROJECT_ID, INSTANCE_ID, DATABASE_ID)

req = client.execute_streaming_sql.call_args.kwargs["request"]
assert "READ_SingersAndAlbumsStream" in req.sql
assert req.transaction.single_use.read_only.strong is True
out, err = capsys.readouterr()
assert f"Resource ID: {DATABASE_ID}" in out
assert PROJECT_ID not in out
assert "Stream Chunks Received: 1" in out
assert err == ""
Comment on lines +48 to +52

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

Add an assertion to verify that the created session is properly deleted after the streaming SQL query completes.

Suggested change
out, err = capsys.readouterr()
assert f"Resource ID: {DATABASE_NAME}" in out
assert "Change Stream: SingersAndAlbumsStream" in out
assert "Stream Chunks Received: 1" in out
assert err == ""
out, err = capsys.readouterr()
assert f"Resource ID: {DATABASE_NAME}" in out
assert "Change Stream: SingersAndAlbumsStream" in out
assert "Stream Chunks Received: 1" in out
assert err == ""
client.delete_session.assert_called_once_with(name=SESSION_NAME)

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks! Since multiplexed sessions are managed automatically by Cloud Spanner and are not deleted via delete_session, we are skipping this mock assertion.



@pytest.mark.parametrize(
"error, expected_prefix",
[
(
exceptions.NotFound("db missing"),
"error: database 'example-db' on instance 'example-instance'",
),
(
exceptions.InvalidArgument("bad stream"),
"error: change stream read failed",
),
],
)
@mock.patch(GAPIC_CLIENT)
def test_read_change_stream_errors(
mock_client_class: mock.MagicMock,
capsys: pytest.CaptureFixture[str],
error: Exception,
expected_prefix: str,
) -> None:
client = mock_client_class.return_value
client.create_session.side_effect = error

read_change_stream.read_change_stream(PROJECT_ID, INSTANCE_ID, DATABASE_ID)

out, err = capsys.readouterr()
assert out == ""
assert err.startswith(expected_prefix)
assert PROJECT_ID not in err
assert PROJECT_ID not in err


@mock.patch(GAPIC_CLIENT)
def test_read_change_stream_cli_entrypoint(
mock_client_class: mock.MagicMock,
monkeypatch: pytest.MonkeyPatch,
capsys: pytest.CaptureFixture[str],
) -> None:
client = mock_client_class.return_value
client.create_session.return_value = spanner_v1.Session(
name=SESSION_NAME, multiplexed=True
)
client.execute_streaming_sql.return_value = [spanner_v1.PartialResultSet()]
monkeypatch.setattr(
"sys.argv",
[
"read_change_stream.py",
"--project-id",
PROJECT_ID,
"--instance-id",
INSTANCE_ID,
"--database-id",
DATABASE_ID,
],
)

runpy.run_module("spanner_change_streams.read_change_stream", run_name="__main__")

assert "Stream Chunks Received: 1" in capsys.readouterr().out


@conftest.live_only
def test_read_change_stream_live(capsys: pytest.CaptureFixture[str]) -> None:
read_change_stream.read_change_stream(
conftest.LIVE_PROJECT_ID,
conftest.LIVE_INSTANCE_ID,
conftest.LIVE_DATABASE_ID,
)

out, err = capsys.readouterr()
assert err == ""
assert conftest.LIVE_PROJECT_ID not in out
assert "Stream Chunks Received:" in out