code description handling design

I am building an ETL (Extract, Transform, Load) feed to transfer an old Iseries app's data to a web interface system.

The old system uses many codes. Each code has a description in its own file.

Examples:


* state code TX is described in the STATE file as "TEXAS".

* Status code "A" is described in the activity file as "ACTIVE".

There are 30 or so files.

My choices are:


1. Each pgm goes out to read each file to grab the code description


2. I load all the codes and descriptions into ONE file, and use a common routine to grab the code descriptions.


#1:
Pros: each code file is explicitly named in each pgm
Cons: dozens of files will be added to each pgm, messy code


#2:
Pros: No extra files in each pgm, only the main input & output files
Each ETL pgm is very similar
Coding seems simpler
One common routine handles all code descriptions

Cons: Design may be confusing because app files are not named explicity




Is it worth it to load all these unrelated codes and descriptions into one file??

Any suggestions?

______________________________________________________________________
This outbound email has been scanned for all viruses by the MessageLabs Skyscan service.
For more information please visit http://www.symanteccloud.com
______________________________________________________________________