On Monday 29 November 2004 16:55, Jean-Marc Lasgouttes wrote: > >>>>> "Jose'" == Jose' Matos <[EMAIL PROTECTED]> writes: > > Jose'> Actually I think that this kind of transformation could go > Jose'> inside lyx2lyx, or as a different module, called by lyx2lyx. > > Jose'> I say this since I use a python script to convert ancient > Jose'> documents, that used inset latex accents to latin1. I can > Jose'> expand this easily, if needed, and if relevant to others. > > I should not be only for latin1, but be able to track the encoding.
I understand that. :-) I said that I convert to latin1 since that is what I use, that is why I stated this can be easily expanded to other encodings. There isn't nothing intrinsic to latin1 in the transformation I do. > Jose'> The advantage of doing this in lyx2lyx and not only in > Jose'> tex2lyx is that it can be applied to lyx documents. > > I thought about that (actually doing that when parsing the document in > lyx), but then I wondered whether it is always what people expect. Can > there be cases where people want to exchange latex documents with this > markup? I guess it is less and less useful, but I used to do that a > lot when exchanging documents between msdos and unix. I am not proposing to do it automatically, I remember that the spellchecker didn't support the latex accents, that is why people asked me to fix their documents. :-) > I we decide that using native characters is always a win, then doing > it in lyx2lyx seems reasonable. I was thinking about calling lyx2lyx as: lyx2lyx --mode=accents [--encoding=latin1] file1.lyx > file2.lyx FWIW I base my proposal in valgrind calling scheme. :-) > JMarc -- José Abílio
