How to Divide Files with split

What is split?

split is a Unix command-line utility similar to grep or tail. It allows you to divide a larger file into several smaller files.

Example Files

1. Create example.txt in a text editor and add the following content:

   File: example.txt

   1 2 3 4 5 6 7 8 9 10

   example line 1 example line 2 example line 3 example line 4 example line 5 example line 6 example line 7 example line 8 example line 9 example line 10

2. Download the text of Moby Dick to demonstrate working with larger files:

   wget -O moby-dick.txt https://archive.org/stream/mobydickorwhale01melvuoft/mobydickorwhale01melvuoft_djvu.txt
   

Basic Usage

1. Run the split command with default options:

   split moby-dick.txt
   

2. Check your working directory:

   ls
   

   moby-dick.txt  xaa  xab  xac  xad  xae  xaf  xag  ...

   The new files present in the directory (xaa, xab, etc.) each contain a portion of the original file. By default, split divides a file into subfiles of 1000 lines each. The original moby-dick.txt file had 16,000 lines, resulting in 16 subfiles. The original moby-dick.txt file is left unchanged.

Options and Parameters

Prefix

The first argument to split is the name of the file, as demonstrated above. An optional second argument allows you to specify the prefix for the output files. By default, this value is x.

split moby-dick.txt moby-dick

Each of the files will begin with moby-dick.

moby-dick.txt  moby-dickaa  moby-dickab  moby-dickac  ...

Split by Number of Lines

The -l option sets the length in lines of each subfile. This value is 1000 by default. The files output by the following command will each contain two lines of text:

split -l 2 example.txt

$ cat xaa
example line 1
example line 2

Split by Size

The -b (or --size) option divides files by size rather than number of lines. The following command will split the input file into subfiles of 100KB each:

split -b 100k moby-dick.txt

You can specify this value in different formats:

• megabytes - m
• gigabytes - g
• terabytes - t

Split by Number of Files

To split a file into a specific number of subfiles, regardless of size or length, use the -n option. For example, to split a file into 3 parts:

split -n 3 example.txt

Label Files Numerically

Use the -d option to label the output files numerically rather than alphabetically:

split -l 2 -d example.txt

x00  x01  x02  x03  x04

Set Suffix Length

Use the -a option to set the number of digits or letters used when labeling the output files. This option defaults to two (i.e. x00).

split -a 1 -d -l 2 example.txt

x0  x1  x2  x3  x4

Advanced Examples

The following command combines the options above to split example.txt into 4 subfiles, each prefixed with example- and labeled numerically:

split -a 1 -n 4 -d example.txt example-

example-0  example-1  example-2  example-3  example.txt

split can also be used to display portions of files without creating subfiles. The following command will break Moby Dick into 100 pieces (without creating any new files) and display the 10th of those pieces:

split -n 10/100 moby-dick.txt

Like many shell commands, split can also accept input from the output of another command using the pipe operator:

grep whale moby-dick.txt | split -l 100