bug#52018: 28.0.60; Improve documentation for compilation-finish-functions

[Thibault Polge]

> Maybe so, but I'm looking at the calls of compilation-handle-exit, and
> the callers do all kinds of transformations with the original status,
> so I think documenting what happens to be more-or-less correct now is
> fragile.

I assumed this interface was reasonably stable, I may be wrong.

> Why do you think this addition is of such great importance to have it
> spelled out in the doc string?

All the docstring currently states about that argument is that it's “a
string describing how the process finished”, which requires users to
determine its possible values through trial and error, or by digging
into the code.  I assumed it would be nicer (but not of “great
importance”) to just document it a bit further.

Best regards,