Sometimes we need to skip tests because they are interactive, they are broken... usually we show that reasons in a comment just above "RESTRICT=test". Would be possible to make portage show that commented message when available instead of always show the canonical message of tests being skipped due ebuild restriction? That way we could inform users a bit better about real reasons to skip the test and, how knows... maybe some reader could help on "unbreaking" that tests ;)
Why not just define the src_test function to display your message? I think it would work well, possibly even better since you'll have the expressive power of bash to do conditional messages if you want.
But it would be like duplicating the ways of restricting tests, instead of simply using "RESTRICT=test" and a comment above it, we would let tests go for skipping them later :/
Hmm, maybe a pkg_notest phase would be a good compromise.
Adding pms-bugs to CC, since a pkg_notest extension should be documented there. I imagine it would not be much different from pkg_nofetch. The only difference is that it's called when src_test is skipped due to RESTRICT=test being set.
Can be solved using src_test(). Seen no reply in 2.5yr, so assuming interest was lost. WONTFIX.