[spanner and sqlalchemy-spanner] Assess approaches to avoiding naming collisions. #18670

Determine this is the right repository

• I determined this is the correct repository in which to report this feature request.

Summary of the feature request

Recently two PRs were put forward to assist with management of Spanner Table creation and teardown in the spanner and the sqlalchemy-spanner packages and to ensure that tests do not fail due to duplicate names.

There is some discussion about whether the two are complementary OR not.
TASK: Assess the two sets of changes and identify whether the changes to sqlalchemy-spanner might mask issues with how the spanner library is expected to behave.

• sqlalchemy-spanner PR #18623: focused on ensuring that all tables during testing are created with unique names, DDL statements are committed, and then during teardown, the teardown steps are executed, the DDL statements are committed and extra metadata is flushed to ensure that no lingering content causes any other test to fail.
• spanner PR #18625: appears to be focused on ensuring that in unexpected situations DDL statements are not retried in production.