IRC logs as CSV

First, download or clone the francejs git repository. Then, concatenate all logs files in the logs folder into a single francejs.csv file. Just for your information, this file should contain more than 36,000 lines.

You'll write a program parsing this file (using the csv Python module), to find out and print which pair of nicknames mention each other the most, like:

$ python3 solution.py
mdk, sizeof

Beware, I may test with another dataset (of the exact same format).

Data format of this csv file

There's 4 columns: - Message type (1 means it's a message, other values means it's a system info, we ignore them) - Timestamp of the message (number of seconds since 1st jan 1970 UTC) - Username involved - Actual message

For example, this line has no interest, it simply logs the arrival of a user, notice the 2 in the fist column, meaning it's a system info, not a text message:

2,1382426750452,"_mlb","_mlb join the chan."

Here, user _kud mentions user nfroidure on the third line, that's what we're searching for:

1,1382442768956,"nfroidure","pour effroi, je précise"
1,1382442902196,"fbentz","o/"
1,1382443055876,"_kud","On devrait sponsoriser effroi par morsay, nfroidure"