Re: improve latex-lyx roundtripping (GSoC)

[Daniel Vainsencher]

On 04/18/2013 07:58 AM, Liviu Andronic wrote: I would strongly agree with this. Round-tripping LyX <--> LaTeX documents is one of the long-term goals of LyX. And in my understanding this is something that lies in the realm of achievable. What we need is, mostly, for a devel or student to show more love to the tex2lyx conversion routines. I think roundtripping in particular will require working on the converters in both directions simultaneously, in particular for storing lyx information in latex comments and then parsing it (as in my math-macros suggestion). - Automatically generate styles for environments defined in imported documents. For example if the imported latex has a keywords environment, that should not suddenly become ERT in LyX. This should allow me to start from the conference sample file, import it into LyX, and obtain a first class lyx environment. ... although I'm wondering how achievable is the 2nd one. I'd love to hear what established devels think of this. Me too: I am not technical enough about LyX and latex to know. Just to be clear, I am *not* proposing that LyX automatically customize how such environment look in LyX: just that \begin{theenv} ... \end{theenv} be converted into something like \begin_layout theenvLyxLayout ... \end_layout, with a minimal definition for the layout inserted into an appropriate definition file. I think the main requirements are that the content be editable as standard lyx (no ERT), and that the resulting latex on re-export is equivalent (turns back into a \begin{theenv} etc). This might also serve some of the same purposes as the layout file editor project, though neither seems to make the other redundant. Either way, I think this is a project worthy for GSoC and if you can come up with a suitably worded project proposal feel free to place it here: http://wiki.lyx.org/Devel/SummerOfCode2013Ideas (and let us know on lyx-devel). This would greatly increase the chances of a student picking it up. And if we have no student for such a project this year, we should certainly consider it for subsequent GSoC summers. I have written an initial project description there. I would appreciate help fleshing out the features and knowledge requirements. Of course a precondition would be a technical mentor that knows LyX well enough to guide students (I would be glad to provide input from the requirements side). Thanks, Daniel