Re: doc for multi-line font-lock

[Dave Love]

Stefan Monnier <address@hidden> writes:

> Setting font-lock-multiline to t should allow font-lock-keywords that
> typically span a single line but sometimes a bit more (things like
> <title>blabla</title>) to work correctly most of the time.

Then the doc needs fixing, doesn't it?

>> I find the doc rather hard to follow, though I'm not sure how to
>> improve it.
>
> I think many people agree on the first part, and sadly they also agreed on
> the second until now.

It's difficult if you haven't hacked on the stuff and don't know how
it is supposed to work.  Perhaps even adding examples would help.

> Documenting its behavior is not something I'm looking forward to.

I don't understand, but no matter.

> I'm responsible for introducing `font-lock-syntactic-face-function' as well
> as for all of the code that uses it in Emacs (AFAIK), but I think
> it's just a hack that's screaming for a replacement based on a significant
> improvement to font-lock rather than a quick hack.

I don't understand that either.  However, there's a definite need to
maintain syntax properties orthogonally to font-lock for other
purposes not tied to redisplay (as discussed long ago, at least with
gerd).

The specific case I was concerned with was treating

\end{code}
...
\begin{code}

as a comment.  If the syntax part of that was taken care of separately
I'd want to be able to fontify arbitrarily long blocks in either
font-lock-comment-face or in the default face.  Would
font-lock-multiline actually work for that?

[I have another solution to literate comments -- this was fixing
existing stuff.]