On Thu, 4 Jul 2024 23:08:37 GMT, David Holmes <dhol...@openjdk.org> wrote:
> My concern is that the logic was wrong and so you fixed it, but this then > screams out for a test that would have detected the error, but you can't > write a test because the line can never be empty, so that becomes an > invariant rather than a runtime check. Do I understand this right that the suggestion is to add an explicit runtime check for non-emptiness and report an error if an empty line has been discovered? If so, then the `is_empty()`check can be removed from the `CmdLine::is_executable()`. ------------- PR Comment: https://git.openjdk.org/jdk/pull/20006#issuecomment-2224336709
