|Problems? Is your data what you think it is?|
Ha! No, I don't want a single * and, yes, I could have done a better job of articulating the constraints.
Here's the deal. Each of the files in that original list of 5,000 will have a prefix of 3 characters, a middle pattern of [\w+\d]*[l|n]n, and a suffix of n digits. Now that I think about it, you can ignore the suffix altogether and just replace the n digits with *. So all we need to focus on is the prefix and middle pattern. Now I need to do this:-
1. Fracture up the start list into n new lists, where n is the number of unique prefixes. So if n = 3, the sum total of the file names in the 3 lists will equal 5,000.
2. Now take each list from #1 and reduce each one down to the fewest number of unique patterns that end in *. The prefix must be preserved, as well as, at a minimum, the first character of the middle string.
3. There can only be one *, and it has to be at the end of each pattern. The original example should make this clear. (Again, it closely resembles the actual file names and the outcome I want.)
Hopefully this helps!