fix(bigquery): pass timeout to the lookup after an exists_ok conflict

[abhishekacharya200]

Fixes #18766

create_dataset, create_table and create_routine with exists_ok=True fall back to get_dataset, get_table or get_routine when the insert returns 409. That fallback passed retry but not timeout. BigQuery's DEFAULT_TIMEOUT is None, so the lookup ran with no timeout at all, even when the caller set one, and a slow or stuck GET could block the call far past the timeout it asked for.

This passes the caller's timeout to the three lookups as well. When no timeout is given, the lookups get None exactly as before, so default behavior does not change. Each request still gets its own timeout and retry keeps working the same way, which matches the docstring: the number of seconds to wait for the HTTP transport before using retry.

Added test_create_dataset_alreadyexists_w_exists_ok_true_w_timeout, test_create_table_alreadyexists_w_exists_ok_true_w_timeout and test_create_routine_w_conflict_exists_ok_w_timeout, which check that both the POST and the GET get the caller's timeout. All three fail on main (the GET goes out with timeout=None) and pass with this change. The full tests/unit suite passes locally (2408 passed), and black==23.7.0 and flake8 report nothing.

I also ran the real client against a small local server that answers 409 to every insert and waits 10 seconds before answering every GET, with timeout=2. A minimal version of the script and server is in #18766. I used retry=None so each run shows a single request, because DEFAULT_RETRY treats a read timeout as retryable and keeps trying until its deadline. The last case uses the default retry with a 5 second deadline instead.

Call	main (57366ee)	This PR (55dc9a2)
create_dataset(exists_ok=True, timeout=2, retry=None)	returned after 10.1s	ReadTimeout after 2.0s
create_table(exists_ok=True, timeout=2, retry=None)	returned after 10.0s	ReadTimeout after 2.0s
create_routine(exists_ok=True, timeout=2, retry=None)	returned after 10.0s	ReadTimeout after 2.0s
create_table(exists_ok=True, timeout=2, retry=DEFAULT_RETRY.with_deadline(5))	returned after 10.0s	RetryError after 4.3s

• Issue opened with a reproduction (BigQuery: create_dataset, create_table and create_routine with exists_ok=True ignore timeout on the lookup after a 409 #18766)
• Tests and linter pass
• Code coverage does not decrease
• No docs changes needed

[gemini-code-assist]

Code Review

This pull request ensures that the caller's timeout parameter is preserved and passed to the subsequent get_* calls when handling resource conflicts (where exists_ok=True and a Conflict exception is raised) in the create_dataset, create_routine, and create_table methods of the BigQuery client. Additionally, corresponding unit tests have been added to verify this behavior for datasets, routines, and tables. There are no review comments, so I have no feedback to provide.