Improve callResourceUri caching

[tomasg2012]

Current caching relies on python's built in lru_cache function. When URIs contain fragments (i.e. /redfish/v1/.../Thermal#/Fans/1), it will call the resource /redfish/v1/.../Thermal for each link in a Collection, when once would be satisfactory. Fix could use a wrapper that strips the fragment, then cache that URI.

Consider adding an option for cache size and how much memory it consumes while running the tool on big services.