Archive parser
We should allow users to provide archives as files. This would give us manual support for "abitrary" data. Data from this parsers needs to be processed differently, because it might vary what normalizer (if any) need to be run on it. The parser could match application/json
files with path *.metainfo.json
.