* Overhauled archiva-discoverer to use a walker / publish / consumer type interface instead of in-memory lists of files. This has had a reduction in memory over large repositories. Expect an average of memory reduction of (22MB * Managed Repository Count).
* Overhauled archiva-reports-standard to use jpox as opposed to xml store to reduce memory usage. Expect an average memory reduction of (45MB * Managed Repository Count).
git-svn-id: https://svn.apache.org/repos/asf/maven/archiva/branches@507550 13f79535-47bb-0310-9956-ffa450edef68
Yes, I could have just fixed the problem in trunk, but the fix has quite an impact on the codebase.
I wanted to flesh out all of the problems on a branch, and allow collaboration, before these fixes hit trunk.
git-svn-id: https://svn.apache.org/repos/asf/maven/archiva/branches@506385 13f79535-47bb-0310-9956-ffa450edef68
