PHYSLITE and PRW auto-configuration

This MR does several things to do with pileup reweighting:

-) It fixes a small bug in the analysis PRW algorithm, which was causing the random number to not be correctly decorated onto the original EventInfo. That resulted in downstream CP tools getting an incorrect run number (and therefore period) for scale factors and calibration.

-) It adds a module for auto-configuration of pileup reweighting files, based on some code in SUSYTools, to get the appropriate ilumicalc and mu files for the sample. It's based on the RecExCommon input file peeking mechanism, so at best it will work in Athena (ie not in AnalysisBase).

-) It adds an option in the analysis PRW algorithm to use that auto configuration.

-) It changes PHYSLITE to use the simple apply PRW algorithm available in the PRW package, rather than the analysis algorithms version. This avoids creating an extra EventInfo object (which we don't really want in this case) and also solves the problem with the random run number for free.

Note that because auto-configuration is used, this means we will have an error in all PHYSLITE jobs (because they are trying to get the PRW file from the "dev" area, where these are stored). In my opinion it is up to our new AMG (@lheinric and @alister ) to solve that one.