Hi Michael, As the author of some MW client library, I cannot find the doc, either. Neither have I looked into lexicographical data, but I had some trials-and-fails on Wikibase entity/properties, and perhaps they can help ^_^ For Wikibase items, I think the essential part of the problem is how to compose "data=" parameter of the request. Basically you need to send a partial of serialized Wikibase Entity (see https://www.mediawiki.org/wiki/Wikibase/DataModel/JSON) as data parameter. Just put in the names/aliases/claims that you want to append/update, send the request, and it's done. The tricky part is, when you want to remove a claim/qualifier or an alias, you need to specify "remove" attribute, along with some identifiable attributes (language, id, etc.) like this { "descriptions": { "zh": { "language": "zh", "remove": "" } }, "aliases": { "en": [ { "language": "en", "value": "alias 1", "remove": "" } ] }, "sitelinks": { "testwiki": { "site": "testwiki", "remove": "" } } } When you are removing a claim/qualifier, you need to identify them by snak id, like this. { "claims": { "P95433": [ { "id": "Q212134$49E9F6B8-9D73-43A8-B45C-0D74DBA92B16", "remove": "" } ] } } You can actually combine all these components above in a single request, to change multiple traits of a Wikibase item. However, you may prefer wbset* actions (e.g. wbsetclaim) sometimes, because they generates user-friendly edit summary on what has been changed, at the expense of only being able to edit 1 aspect of the Wikibase item in 1 request. If it's possible, perhaps you can refer to my C# code for composing data parameter for wbeditentity request: https://github.com/CXuesong/WikiClientLibrary/blob/master/WikiClientLibrary… Hope that helps. Thanks, Xinyan -----Original Message----- From: Wikidata-tech <wikidata-tech-bounces(a)lists.wikimedia.org> On Behalf Of Michael F. Sch?nitzer Sent: Wednesday, April 1, 2020 8:15 AM To: wikidata-tech(a)lists.wikimedia.org Subject: [Wikidata-tech] Wikidata API documentation Hello, Is there any documentation to "wbeditentity", except from the few examples on the API-document-page? Any documentation about the format/grammar of the data? And is there any API-documentation for lexicographical data at al? The few bits on the source-code generated page aren't sufficient. Cheers, M -- Michael F. Schönitzer Prochintalstr. 2 80993 München Mail: michael(a)schoenitzer.de Jabber: schoenitzer(a)jabber.ccc.de Tel: 089/37918949 - Mobil: 017657895702