Markdown: hotfix for failing test (#60591)
Another unfortunate interaction between PRs. A proper fix for this
regression will come, but for now let's make the tests pass again so
that nobody else has to suffer from this mistake.
---
To explain: locally I work on an integration branches (were all my
Markdown work comes together), and then split out work for individual
commits. I tested both the integration branch and the individual PRs,
and all was fine. But then in master suddenly the failure. Why did my
tests on my integration branch not catch this? Because there's another
in-flight change on there, which *hides* the issue (but does not fix
it), namely, I've switched the spec tests to not use the "Julia flavor"
but rather the "common flavor". I'm sorry for any inconvenience I may
have caused. On the upside, I am now aware of this "hides the issue"
problem and will think about ways to deal with it. Most likely I'll just
let it test both with the Julia and the Common(Mark) flavor, so that
regressions in both are caught.