Re: html, css, and js modes working together

[Tom Tromey <tom@tromey.com>]

>>>>> "Dmitry" == Dmitry Gutov <dgutov@yandex.ru> writes:

Dmitry> I do see the oddities, and it seems the wonkiness is more related to
Dmitry> setting up submode regions, because when the font-lock acts up,
Dmitry> indentation can fail on that line as well.

I spent some time debugging this and I fixed up the fontification here.
I also fixed the bug Clément found, where sub-mode fontification was
being applied even in HTML comments.  I've played with it quite a bit
and used it on real code here, and it's more robust than before.

Dmitry> Here's a broken example, for you reference.

Thanks.  This one works now.


At this point I think I'm nearly ready to start checking things in.
The current state is:

* I have a bunch of small patches to change various things to use
  derived-mode-p rather than eq.  This is needed to deal with having the
  new mode not be html-mode.  I don't really want to test most of these
  (like changes in viper or cedet, about which I know very little), so I
  thought I'd send them for review.  The changes are all very small.

* Most basic functionality works, like fontification, indentation,
  flyspell, and comment-*.

* Imenu and some other things don't work, but I think it's better to
  move forward with what we have than to fix every feature beforehand.
  I wonder if you agree.

* I need to write some tests.  I think this plus documentation is the
  last bit I need to do.

* One open question is whether the new mode should run prog-mode-hook.
  It is sort of both a prog mode and a text mode.

Tom