This PR makes connection-pool exhaustion observable through polling error callbacks without changing the handling of regular request timeouts.

HTTPXRequest currently maps httpx.PoolTimeout to TimedOut. The polling retry loop intentionally handles every TimedOut as an expected request timeout and retries it silently. As a result, Updater.start_polling(error_callback=...) cannot distinguish or react to a request that was never sent because no connection was available from the pool.

The change introduces telegram.error.PoolTimeout as a subclass of TimedOut, translates httpx.PoolTimeout to that backend-independent exception, and forwards only PoolTimeout to the polling error callback. Existing except TimedOut and isinstance(error, TimedOut) checks remain compatible, while regular long-poll timeouts retain their current behavior.

Check-list for PRs

• Added .. versionadded:: NEXT.VERSION, .. versionchanged:: NEXT.VERSION, .. deprecated:: NEXT.VERSION or .. versionremoved:: NEXT.VERSION to the docstrings for user facing changes (for methods/class descriptions, arguments and attributes)

• Created new or adapted existing unit tests

• Documented code changes according to the CSI standard

• Added myself alphabetically to AUTHORS.rst (optional)

• Added new classes & modules to the docs and all suitable __all__ s

• Checked the Stability Policy in case of deprecations or changes to documented behavior

• If Relevant

  • Added or updated documentation for the changed class(es) and/or method(s)