Fixing #12372: leverage preview to have images in DocBook instead of LaTeX code

[Pavel Sanda]

On Wed, Oct 13, 2021 at 03:12:02PM +0200, Thibaut Cuvelier wrote:
> On Wed, 13 Oct 2021 at 15:07, Pavel Sanda <sanda at lyx.org> wrote:
> 
> > On Wed, Oct 13, 2021 at 02:33:47PM +0200, Thibaut Cuvelier wrote:
> > > The only solution I could find is to use the existing mechanism to
> > generate
> > > preview images. This way, the LaTeX code is transformed into an
> > > intelligible form (i.e. an image), while still keeping the original LaTeX
> > > code if someone wants to modify the image without access to the original
> > > LyX document (this property is quite important to me).
> >
> > Just to be sure - do I understand correctly that what you aim for is to
> > take
> > any piece of ERT, run preview-inset-like machinery, get it's output as an
> > image
> > and include it in docbook output?
> >
> 
> No, that's not for any ERT, it's only for insets that opt into this
> mechanism. In particular, for the linguistic trees, there are no ERT
> involved. It's not automatic at all.

I see, that makes better sense. I briefly went through the code and
it looks sane to me, so go ahead unless someone else has more comments.

Pavel