spring-ai icon indicating copy to clipboard operation
spring-ai copied to clipboard

tika: support using remote server

Open davidkarlsen opened this issue 5 months ago • 3 comments

Expected Behavior

Apache Tika can be run as a convenience server, this gives more wiggle room to install language support and share and instance, have less weight in the calling app etc.

It would be nice with a spring-AI implementation which do remote calls instead of using embedded-mode.

Current Behavior

Runs embedded tika.

Context

  • https://cwiki.apache.org/confluence/display/TIKA/TikaServer
  • https://github.com/apache/tika-docker
  • https://hub.docker.com/r/apache/tika/tags

davidkarlsen avatar Feb 24 '24 12:02 davidkarlsen

I'm not against the idea, but would welcome a PR as I'm not familiar with the Tika Server.

markpollack avatar Feb 29 '24 17:02 markpollack

@markpollack Do you want it in a separate maven module, or should we add it to the existing one?

davidkarlsen avatar Apr 13 '24 14:04 davidkarlsen

@markpollack ping?

davidkarlsen avatar Apr 23 '24 22:04 davidkarlsen