- implement opt-in unpacking/caching of the jar collection archives,
such as spring boot fat jars and WARs, to the temp dir, and creating
a valid classpath from extracted archive
- turning it on for the new default jsr223 engine
- fallback to extracting classpath from resource URLs if the classloader
is not known url-based one
- refactor and optimize extraction
- cache the last retrieved classpath in the default JSR-223 script engine
factory