Gary Hawkins wrote:
2. Option to inject a clearcut delimiter between tests
Distinct from, say, outputting a diagnostic between groups of tests?
By 'tests' I'm thinking 'file' with its subtests, so-to-speak, so
yes,
anything that is clearly delineatable (LOL) programmatically where
one
'file' output stops and another starts is fine.
Different in what way from the current behaviour of Test::Harness?
Each test script gets its own line in the output currently.
Not if verbose output.
Surely you jest. It is not reliably parsable, there can be garbage text
on the end of those summary lines at times.
There's already a proposal to handle per-test diagnostics.
http://perl-qa.yi.org/index.php/TAP_diagnostic_syntax
