Its not recommended to store possibly large values into attributes since attributes have size limit and are stored in JVM Memory (heap) which might result in out of memory exception in case you are dealing with large amount of zip files that might have large amount of file contents. May I ask why are you trying to store the file list in an attribute?
Each zip contains a json file along with a bunch of various file types. After unzipping, I do some processing on the other files and then I need to match up json with each of the other files contained in from the original zip file to create an elasticsearch document. I tried using a Content Merge based on the fragment attribute, but that is not working 100%. I am out of ideas on how to get this to work.