RFR: 8343791: Socket.connect API should document whether the socket will be closed when hostname resolution fails or another error occurs [v23]

[Volkan Yazıcı]

On Fri, 29 Nov 2024 13:39:08 GMT, Alan Bateman <alanb at openjdk.org> wrote:

>> I deviated from the first-letter-in-upper-case convention deliberately to match the message in `Socket`. Are you sure you want me to replace `already connected` with `Already connected` in `Net`?
>
> There are 40+ places in Socket that throw a SocketException with a fixed string. It looks like all but two start with an uppercase letter. So maybe we fix Socket.connect instead. I realise this is adding more to this PR but touching older code is always likely to find little oddities like this.

In 840fe7aca9e0b7b35d464fc7706fe39de879143a, capitalized only the `Already connected` message in `Socket` – left the other not-capitalized `SocketException` message (i.e., `factory already defined`) as is. (I have the suspicion that this has triggered some test failures... I am checking them in the meantime.)

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/22160#discussion_r1863596170