Georg Baum <[EMAIL PROTECTED]> writes:
...
> > Gregor> Thinking a bit further, this could be solved by another option
> > Gregor> for tex2lyx to specify which literate programing file format
> > Gregor> would be chosen. I suggest noweb for default name instead of
> > Gregor> literate as literate is far more general than noweb. For
> > Gregor> Sweave file format name sweave could be chosen.
>
> Why a different format? What is different in sweave that requires tex2lyx to
> act differently? I suspect that you simply found a bug in the noweb import
> that would also accor with noweb files, because the tex2lyx noweb support
> was really a quick hack.
I did not found any bug in tex2lyx noweb!
There is no difference in syntax between noweb and Sweave. There is only
difference in what layout should provide - for example Sweave does not need
\usepackage{noweb} - it needs \usepackage{Sweave}, but this is inserted by
Sweave intself. And there is a need to be able to align scrap environment -
current literate layouts do not allow this. I have all the needed modifications,
but that is a hack. Partly because noweb support is a hack.
> > Is there any reason to have a sweave-article and a noweb-article
> > class? What would the difference be between the two? It seems to me
> > that it would be better to pick a tool in Document>Settings to choose
> > a literate programming program.
>
> But that is already possible if you simply redefine the noweb converter. As
> Gregor pointed out this is not nice if you use both. The alternative to a
> new document class would be yet another set of dvi, ps and pdf formats for
> output via sweave, with an appropriate set of converters. That becomes
> unmanageable if it is extended even more. I already do not like the
> different pdf formats, but I have no better idea.
> I always wanted to reply to his original message but did not find the time.
Only literate format can be used as it has specific meaning.
> IMO, there are four problems:
>
> 1) output to sweave format. This is currently only possible with hacks,
> either via redefining the literate format, or by defining a literate->seave
> converter (would be a dummy converter that just copies the file), and new
Hmm. This is interesting. How can I define dummy converter that just copies
literate to sweave layout?
> dvi,ps and pdf formats with an appropriate converter from sweave. The long
This is again a problem, as I have to use literate in output definition in
layout files. Then I would need sweave --> literate converter, but then I again
have problem with noweb specific things.
> term solution would be to implement a sweave backend in LyX (it could just
> be an alias for the existing noweb backend), so that LyX knows what
> converters to call for a sweave doc.
Seems to be the way I proposed. But involves some coding. I can at least
help in testing.
> 2) layout files (or maybe not - depending on the solution to 1)
I have it.
> 3) import: should not be difficult to fix once 1) is decided
Current tex2lyx works just fine we would just need to specify sweave and not
noweb (current literate) format.
> 4) problems with some encironments. I do nut understand that fully yet.
Will be fixed with proper layout and newline fix in align environments.
Thanks!
Gregor
