RFR: 8345431: Improve jar --validate to detect duplicate or invalid entries [v9]

[Lance Andersen]

On Tue, 20 May 2025 18:14:15 GMT, Henry Jen <henryjen at openjdk.org> wrote:

>>> I am wondering if we should not explicitly said status code of 1 to be more flexible. s/status code of 1/non-zero status code
>> 
>> Well, now would be a good time to consider say a value of 2 for duplicates, 3 for invalid names, etc...
>
> We report all issues, not just one, so that would also involve what do we prioritized for the exit code...
> In general, a non-zero code for invalid jar needs attention. We still 0 with unmatched order.
> Also the process could fail with other issues like I/O or non-zip file.

Yes I understand and currently we have no documentation as to what the return values might be.  An inconsistency in the order should also be 1 based we only return a 0 if there are no issues.  Any warning should be treated the same

My point is that it is worth considering with these changes

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

PR Review Comment: https://git.openjdk.org/jdk/pull/24430#discussion_r2098625246