Question

Given a txt log file, which is in the form:

USER_A timestamp1 otherstuff
USER_B timestamp2 otherstuff
USER_C timestamp3 otherstuff
USER_A timestamp4 otherstuff
USER_A timestamp5 otherstuff
USER_C timestamp6 otherstuff
USER_B timestamp7 otherstuff

How would you count the number of different unique users in erlang? I was thinking about reading the file line by line and using proplists module. Each user will be a key, with a value that will be the number of occurrences. Once the file is read, I call:

length(proplists:get_keys(List)).

Is this the correct way to achieve my result?

Answer 1

I would also use the sets module for this since it is both fast and as a set contains no duplicates.

The following code should do the job:

{ok,Bin} = file:read_file("test"),
List = binary_to_list(Bin),
Usernames = [hd(string:tokens(X," ")) || X <- string:tokens(List,[$\n])],
sets:size(sets:from_list(Usernames)).

Edit: I deleted the one-liner since it didn’t add any value