gene_name.py filter of fragments, no gene name, and duplicates

* Fragments are taged as  'no gene name' if `fragment` is set a `true` in config file. Shouldn't it be something else ? Like 'Fragment' or so.

* Proteins with no gene name are taged as 'no gene name' which is fine. BUT if `noname` is set to `false` and `ducplicate` to `true` in the config file, all the 'no gene name' proteins are consider as duplicates and removed when they should be kept.