I'll work on editing that tomorrow. It's surprisingly simple.
If you look at the dbstr_us.txt file, you will see a ^ delimited text file.
The first field is the "item number", but of what, is answered in field 2. Then field 3 is the description
Code 11 is the code for races.
So the line:
1^11^Human^(some other garbage probably on this line, but we don't care about that).
So we find all the fields with "11" as field 2, then race # is field 1 and race name is field 3.
I somehow figured that out many years ago and had a shell script to generate it based on the input file, but as I said before, that's lost.
I'll edit my program to generate races.h in addition to Races.txt and post the result when it's done.