Get documents by ids

[steffenkleinle]

Motivation

For several purposes like favorites we need to save documents in the app. While saving the whole documents is of course possible, there is no way of updating these if e.g. images, alternatives or the words of a document is changed. Fur us the better alternative would be to just save document ids and request the actual documents using a cms endpoint.

Proposed Solution

Implement an endpoint accepting one or multiple document ids and returns an array of the corresponding documents. If an id does not point to a valid document (anymore), the request should not fail but instead just not return a document for that id.

Alternatives

Additional Context

Corresponding app issue: https://issues.tuerantuer.org/browse/LUN-307

[timobrembeck]

Just to clarify, the new words endpoint does not satisfy your needs?

• Get list of all words: https://lunes-test.tuerantuer.org/api/words/
• Get single word by id: https://lunes-test.tuerantuer.org/api/words/2/

If it's really necessary to filter the documents by multiple ids, what is the preferred way for you to pass them? As GET-parameters, e.g. words/?ids[]=1&ids[]=2, words/?ids=1,2 or in the POST data as JSON list?

[steffenkleinle]

Ah yes, it does, sorry and thanks! Didn't know about that. I think it is necessary to filter by multiple ids. Query params would be my preferred way to do so. Is there documentation for the endpoints somewhere already?

[timobrembeck]

Alright!
Yes, we do have documentations, in my opinion even a bit too many 🙈:

• GitHub README
• Django Rest Framework
• Swagger
• ReDoc

@lukasboehm97 I think we should decide on a single source of truth, keep that one up to date and remove all other documentations.

[batcapricorn]

Ideally we would get our sphinx documentation up-to-date and keep it as single point of truth

[batcapricorn]

Summary of To-Do's regardings this issue:

• currently, we have two similar endpoints discipline_by_id (which is only used internally within the cms GUI) and words. It makes sense to merge these endpoints into one
• handle access rights (API-Token, Authentication for replacing discipline_by_id etc.)
• set up filtering and url-creation