Hello, Rasmus <ras...@gmx.us> writes:
> Thanks for all time you've put into the comments. I appreciate it, > and I will try to revise the patches over the weekend. Thanks. Again, there's no rush. > I worry about this approach based on some observations Alan sent > off-list. When you export the quoted document with num:nil all labels > will be of the form "unnumbered-N", loosing all structure in labels. I don't think how that would be a problem. A label is expected to refer uniquely to a headline. Reflecting structure of the document is not a requisite. "sec-NUM" is but a bonus. As far as labels go, we could also use "headline-X" for all headlines, numbered or not. > Also, some labels are still unassigned in html for unnumbered > headlines, e.g. the "text-" (which is a function of parents' section > numbers) and outline-container-sec-. Couldn't they be turned into "text-unnumbered-X" and "outline-container-unnumbered-X" instead? > Do you think it's better to solve the remaining issues, and accept > that when num:nil exported documents will be quite altered compared to > previously, or should I try to introduce a more informative ID for > numbered an unnumbered headlines alike? The former. > If following the latter path, the most obvious approach (to me) would > be to have a separate :headline-id > and :headline-numbering. :headline-id could be collected using > something like `org-export--collect-headline-numbering', but labels > would not necessarily reflect the printed section numbers, > though :headline-numbering would still be "correct". I think we don't need this added complexity. Regards, -- Nicolas Goaziou
