Cl is a tool for filtering data by columns on the command line.

What it does

Usage: cl [options...] <column_indexes...>

Examples:

# Filter a simple table of data for the second column:
$ echo "1 2 3
> 4 5 6
> 7 8 9" | cl 2
2
5
8

# Grab a list of process IDs from ps, ignoring the header row (-i):
$ ps | cl 1 -i
7958
29855

# Grab the third column of output when there may be spaces, in values, and tabs are the separator (-t):
$ netstat | cl 3 -t

# Or the first 2 columns:
$ netstat | cl 1 2 -t

# Or the first 4 columns (in bash):
$ netstat | cl {1..4} -t

Options:

  -i    ignore the header row (first row)
  -s string
        a character or regex to split lines (default: whitespace)
  -t    use tabs as separator (alias of -s \t)

Install

go get github.com/dankinder/cl

Why

Existing commonly-available bash commands like awk and tail can accomplish what cl does. But they become quite verbose.

To select a column with awk, you use:

my_command | awk '{ print $1 }'

But then, you want to delete the header row too (another very common thing). For that you need:

my_command | awk '{ print $1 }' | tail -n +2

I used to have the following bash function in my .bashrc, and I'm sure many others have equivalents:

# Grab column N from stdin, ex. `ps aux | cl 2` => Just Process IDs
function cl() {
    awk "{print \$$1}"
}

# Ignore the header row of the input
function ih() {
	tail -n +2
}

Further, some tools worked around this specific problem by building another tool (pgrep) giving us yet another command to learn.

It felt to me like we need a small, simple, generalized tool for this task.

So I build cl.