I'm interested in dealing with archive contents in a similar way to dealing with Images through the awt.imageio api: Just get them as a file and see if you know how to decode them.
Obviously, there's the jar apis but I believe they only work with zip formats.
End use is Clojure code.
I think that some of the types you can handle through Apache Compress library. I think, that you can also make a close look to Apache Tika library, that extracts text and metadata from different file types, and as I remember, that they want to extend Apache Compress with more archive types (look onto patches in the JIRA)