On Sat, Aug 12, 2017 at 01:42:42PM +0200, Kornel Benko wrote: > There are 2 issues here. > > 1.) The reference with non-ascii characters > "sec:Insertar-Código-TeX"
I personally think the change you propose here is fine (but the opinion of a non-English speaker would be more valuable here). Could this issue explain a lot of our systemF failures (that we've probably inverted long ago and forgotten about)? I imagine it is quite common. Do we know if there are any plans to allow non-ascii labels by whichever part of the LaTeX machinery is responsible for this part of the code? If not, I wonder if it would be a reasonable feature request to convert such labels to ascii on export with system fonts. It seems we would want to allow non-ascii labels for export to e.g. pdflatex, so we should not e.g. only allow ascii characters in our label dialog. But I wonder if converting them would make sense. I'm not sure if we would want to convert the labels as accurately as possible (e.g. código to codigo), or since an accurate conversion would be impossible for non-latin languages just use e.g. "label1", "label2", etc. I'm guessing this might not be a good idea, though, because aren't labels used in output in some cases? For example, I think I remember labels of math equations being printed if I set a certain LaTeX option (maybe draft?). > 2.) Language specification in connection with ERT in argument to labelitem > > Annoyingly, it is not possible to remove the language setting unless one > first removes the ERT. > (See the section "3.5.3 Cómo se emplea" with the item '\(star\)' After this change, does the document appear any different? That is, do we sacrifice anything just to make the tests pass? If not, I would say go ahead. Scott
signature.asc
Description: PGP signature
