On 02/11/2014 07:20 PM, stefano franchi wrote:



On Tue, Feb 11, 2014 at 4:27 PM, Steve Litt <sl...@troubleshooters.com <mailto:sl...@troubleshooters.com>> wrote:

    On Tue, 11 Feb 2014 15:57:13 -0600
    stefano franchi <stefano.fran...@gmail.com
    <mailto:stefano.fran...@gmail.com>> wrote:

    > My next struggle with word conversions came much sooner than I
    > thought.
    >
    > Suggestions are welcome on how to tackle the conversion a  document
    > with the following characteristics:
    >
    > ~ 16,000 words
    > class: article
    > engine: LuaTex
    > Bib: biblatex + biber
    > no math
    > no images
    > no X-references, branches, etc.
    > lots of footnotes


    If I'm understanding you correctly, you're trying to import a LaTeX
    file (?) into (LyX? XHTML? MSWord? ODT?)


I am "just" trying to go from LyX to Word because the publisher wants Word only. An unfortunately very common situation in my field.


    If the existing LaTeX file works, why not just keep working with it in
    Vim? Unless you're trying to bang out voluminous content, LaTeX in an
    editor is reasonable to deal with.


Not a problem for me, working in Lyx is fine.

    For LaTeX conversions, I often found it handy to compile the LaTeX all
    the way to at least .dvi, and then use all the created intermediate
    files as the basis of a home-grown conversion program. If one of the
    intermediate files has all your biblatex info, then you can parse that
    with a Python file and do what you want with it in the destination
    format.

    Personally, if it were me, on the way to (eeeuuuuu) doc/odt, I'd
    make a
    side trip to either xhtml or HTML5 formatted as well-formed xml, with
    styles completely intact. From there you could probably find a
    converter to ODT, and from ODT you could export MSWord .doc, but with
    the xml, you'd have an easy gateway to eBooks or anything else you
    wanted to do with it later. I might even be so crazy as to suggest you
    use the HTML5 well formed xml as your source, editing it in Bluefish,
    and convert *that* to other formats.


That's was my idea too: LyX-->(X)HTML-->Odt-->Doc

But it fails on several counts, with the tools I tried, for the reasons listed in my message. To which I should add

5. htlatex
Does not work with Luatex. And switching from LuaTeX to pdflatex is not really an option, as the document and the bibliography have a mix of English, French, and Italian, with lots of diacritics. I would have to manually convert everything to Latex's codes, I guess. Or find a tool that does it for me.

I really cannot see the way out of this problem, which is bad, as we are currently proposing a LyX-->doc/odt (possibly roundtrip) conversion as a GSOC 2014 project.

As far as I can tell, the stumbling block, in this particular case, is biblatex/biber. The conversion must happen after they have successfully run and produced references and bibliography. BUT: neither LyX, nor eLyxer, nor pandoc have any awareness of biblatex. Lyx because it does not currently support it, and, consequently, eLyxer, because it uses Lyx's data, not LaTeX's. I do not know about pandoc, and I am not sure how to find out.

Is it possible to mix and match here? The footnote issue is annying, no doubt. I don't know how to get any sort of XHTML intermediary to be recognized as footnotes in LibreOffice.

How hard is it to convert your bibliography to BibTeX? I've had good success exporting that via LyXHTML.

rh

Reply via email to