Use pdf file name as attribute

Hello everyone

I want to do some simple Text Mining using pdf files in RM but I'm a little stuck right now. 

I created a process using the loop files and process document operator for reading in several pdf files.
As I have a lot of files to analyze, which I also want to compare, I would like to create an attribute which includes the file name to keep track of everything. 

I enabled macros and tried to include the file name by generating a new attribute.
The problem is that the generated attribute only consists of the file name of the last file I uploaded and not the name of the corresponding document. How can I ensure that the attribute value is the respective file name of the document?  

Or is there a way to just include the metadata_file as an attribute? 

I included my process and the first 5 files I want to read.
I would really appreciate every help, thank you already in advance!