We are using a wikibase.cloud instance for Kurdish lexicographical data (final goal after curation process on that Wikibase: transfer to Wikidata, where the same issue applies), and we are having some issues with the following language codes which are currently nonexistent, but are used in our source data:
- hac (Gorani (https://www.wikidata.org/wiki/Q33733), uses arabic script only)
- kmr-arab & kmr-latn (Northern Kurdish aka Kurmanji (https://www.wikidata.org/wiki/Q36163))
- ckb-arab & ckb-latn (Central Kurdish aka Sorani (https://www.wikidata.org/wiki/Q36811))
- sdh-arab & sdh-latn (Southern Kurdish (https://www.wikidata.org/wiki/Q1496597))
Although the ku language code is defined for Kurdish, and ku-latn and ku-arab also exist, the representation of its varieties is not possible (hac and kmr don't exist at all). In other cases, the script-defining versions are missing.
Thanks.