SELECT col1, col2 ( i.e. projections) can be implemented with several variants of Unix utilities: cut and awk are the two most obvious. 2. JOIN can be implemented with the… wait for it join utility. You'll need to sort its input first, though. Many GROUP BY operations can be performed with combinations of grep -c , sort with or without the -urnk options (look at the man page — you can apply options to individual sort keys), and uniq with or without the -c option. Many more can be done with 20 or 30 characters of awk. Output formatting is easy with column , especially with the -t option.
via NoSQL databases