Support CSV output file per table

[ndwhelan]

Adding a feature to allow specifying an output directory, allowing for each table to have its rows in its own file.

Adds optional first parameter --out-dir=. Example usage, which has been added to the README, would be like this:

zcat dumbfile.sql.gz | python mysqldump_to_csv.py --out-dir=./output

OR

python mysqldump_to_csv.py --out-dir=./output ./sql/mysql_dump1.sql ./sql/mysql_dump2.sql

So, if the SQL dumps insert data into the tables users and groups, and --out-dir=./output is specified, two files will contain the data, ./output/users.csv and ./output/groups.csv, respectively.