How to split files in Linux from the command line

[I originally published this post on Blogspot.]

It can be useful to split large files, or even smaller files and ensure all the resulting volumes have the same size. For this we will be using split.

Splitting files

We can split a larger file into smaller ones like this:

1

split -b 100M file

The previous command splits file into several 100 Mb volumes, called by default xaa, xab, xac and so on. These default names may be prefixed by a pattern:

1

split -b 100k file pattern_

The previous command splits file into several 100 Kb volumes, called pattern_aa, pattern_ab, pattern_ac and so on. If we want digits instead of letters, we can use the -d flag:

1

split -db 1G file pattern.

The previous command splits file into several 1 Gb volumes, called pattern.00, pattern.01, pattern.02 and so on.

Joining files

To join the volumes, we can cat the sorted file names and redirect them to an output file:

1

cat `echo pattern.* | sort` > new_file

Both the original file and new_file have the same MD5 sum; they are identical.