Rasmus <ras...@gmx.us> writes: > 'Cause usually org-export-data gives you the right value automatically. > Anyway, I don't mind using org-element-parse-secondary-string.
`org-element-parse-secondary-string' is meant for cases like this one. >> DESCRIPTION could be moved to `org-element-document-properties'. But >> then, we need to introduce `org-export-with-description', which implies >> export back-ends have to deal with it somehow (ATM "ox-ascii" ignores >> it). > >> Another option is to simply remove DESCRIPTION from "ox.el" and let >> export back-ends in need of it to add it to their definition. > > I don't oppose that. > >> However, I don't see why KEYWORDS should accept markup. I also think it >> should be moved to back-end definitions instead of "ox.el". > > Keywords should also be printed in the document as well IMO, e.g. after > the abstract. An example of where you need markup: σ-algebra. Good point. Since KEYWORDS and DESCRIPTION are really back-end dependant, I vote for moving them from `org-export-options-alist' to back-end definitions. Using `org-element-parse-secondary-string' will be required in this case. WDYT? Also, supposing you agree, do you want to do a patch (caveat: there's "org.texi" messing involved)? Regards,