Use addobject-Endpoint in API-Clients supported?

[futurearchitec]

Hello everyone,

I am evaluating using txtai for a new AI application we are going to develop. The software looks very promising, as it is a very cool package for using documents with AI-embedings and vector database. Also it is very good, to see quite a few API-client packages for different platforms. Additionally, there are a lot of Jupyter Notebook examples for using txtai.

But my question is, how can I upload documents and images to a running txtai-instance? All examples I have seen, use local files that are already on the server. That is not very realistic. Then I saw your enhancement off the /addobject-endpoint in the embeddings router, that shall work exactly as needed. But it is not available in any of the client-api. Is this on purpose, as the endpoint should not be used by a client-api. Or is this more of a kind of a TODO, in the sense that the enhanced endpoint is not fully reflected yet, in the api-clients.

Thanks for your reply.

[davidmezzetti]

Hello. Thank you for writing this up.

Managing each of the different APIs has become a cumbersome over time. So I've only added functionality as it's been requested.

Which one of the APIs are you using (Python, JS etc)?

[futurearchitec]

Hello David.

Thank you very much for answering my question so quickly. Actually we are using the Java-API for our new software. An update of it, would be awesome. Thank you for efforts and the great software txtai.

[davidmezzetti]

Ok, I'll see what I can add. It's not going to be in time for the 9.0 release though.

PRs are always welcome if it's something you want to try to tackle. My Java is getting more and more out of date.