optimize code for proxy handling of `urllib.request`

Feature or enhancement

Proposal:

1. the second loop of line starts from 1845 is duplicated from the first loop.
   https://github.com/python/cpython/blob/2041a95e68ebf6d13f867e214ada28affa830669/Lib/urllib/request.py#L1817-L1853
2. the registry access should be separated to a standalone function for cache, to be used by code like https://github.com/aio-libs/aiohttp/blob/e79b2d5df70a2644e81925cc49558962af91848d/aiohttp/client.py#L609-L617
   https://github.com/python/cpython/blob/2041a95e68ebf6d13f867e214ada28affa830669/Lib/urllib/request.py#L2071-L2104
3. the proxy setting from registry should be ignored when if proxies from environment exists? like the proxy_bypass()

Should open a PR?

Has this already been discussed elsewhere?

This is a minor feature, which does not need previous discussion elsewhere

Links to previous discussion of this feature:

No response

Linked PRs

• gh-127767

Lib/urllib/request.py の 1817-1853 行付近と 2071-2104 行付近から始め、リンク先の aiohttp のプロキシ処理と比較してください。目標にループの重複排除、キャッシュされたレジストリへのアクセスの切り出し、環境プロキシの優先が含まれるかを明確にしてください。合意したプロキシ動作が実装され、適切な urllib.request テストでカバーされていれば完了です。