- Rename indexes to group by year and month instead of university and add prefix
  to indicate data source (MEC or MS);
- Config files (config.sh) now use associative arrays instead of 2 arrays filter
  and university, which needed to be correlated, having the same length and using
  the same number index to relate data;
- Add flag -w in grep, which filters by matching a whole word instead of
  substring. For example: "PARA" does not match with "PARANA" anymore (which stops
  scripts from getting duplicated data for "UNIVERSIDADE FEDERAL DO PARANA");

Signed-off-by: Cristian Weiland <cw14@inf.ufpr.br>