I wish to include some non-ascii characters in the documentation for an R package, specifically the diacritic ă.
I am using roxygen2, so must not include any non-ASCII characters in the .R
file from which the documentation will be generated. Moreover, the non-ASCII characters may not be available to the software that compiles a PDF manual, so I wish to specify a safe fallback for the PDF documentation.
The obvious approach seems to be to conditionally display HTML entities, i.e. \ifelse{html}{ă}{a}
, but the entities are evidently escaped at some point: the .rd
file reads ă
where I expect an ă
.
Bonus points for answers that can be applied to diacritics included from REFERENCES.bib
via the \insertRef
macro...