giohappy
commented
2 weeks ago @t-book if you agree I would close this PR, since it has conflicts now and the proposed solution doesn't cover the other parts of the same module with the same problem.
Your solution might work but it should be put in a method to be reused where needed.
The current code splits the URL based on the string geoserver. In case the domain contains geoserver somewhere else, like https://geoserver.example.com the loading of resources fails. I'd like to suggest to explitcitly search for /geoserver/ which would target geoserver.example.com /geoserver/ ows...