Monday, 25 May 2015

Cut Command in Linux - Extract Fields and Columns from a file

    In some of my recent articles on text processing, I have explained the use of sed command in Linux/Unix. In case of sed command, we provide an input file to the command, it reads the file line-by-line, processes each line and then prints it on the STDOUT. So, in brief, its a row-wise operation. Similar is the case with cut command - there is an input file, there is processing part and the processed output can be displayed on STDOUT or saved in a file. A minor difference between sed and cut is that, cut command processes the file in vertical manner. So, the outcome of the cut command is a single or multiple columns.

    As of now, just remember that, cut command is just a filter, that processes the file and extracts columns from it. Basically, using cut command, we can process a file in order to extract - either a column of characters or some fields. Thus, to achieve more clarity about cut command, we would study it in two parts.

Here we go!

A. Extracting Column of Characters

To begin with, consider a file cuttest.txt with contents as below:

$ cat cuttest.txt
This is line #1
It is line #2
That is line #3
While, this is line #4
It's line #5
I am line #6
Myself line #7
It's me, line #8
Hello, I am line #9
Last line, line #10
Now, just have a look at the basic syntax of the cut command, to extract column(s) of characters from a file:

To explain this briefly, we are instructing cut command to select on the specific characters specified by RANGE from the file FILENAME.

1. Display a Column of Characters

To begin with, lets display the fourth character from each line of the file cuttest.txt.


$ cat -c 4 cuttest.txt
This does make sense!

2. Display a Group of Columns of Characters

In order to extract a group of columns, we need to specify a range - Start and End, to the cut command. To try with, lets display first five characters of each line of the file.


$ cut -c 1-5 cuttest.txt
It is 
I am 
Conclusion is - a whitespace is also considered as a character.

Another variant of this case is, when you want to start from a particular column and display till the last one. As an example, we will start displaying from the 6th column will the end. So, in this case, we would mention start of the range as '6' and we do not mention any end. Thus, it will print everything after the 6th column.


$ cut -c 6- cuttest.txt
is line #1
 line #2
is line #3
, this is line #4
line #5
line #6
f line #7
me, line #8
, I am line #9
line, line #10
Similarly, to get first 6 characters from the beginning of each line, we would have an example as follows:

$ cut -c -6 cuttest.txt
This i
It is 
That i
It's l
I am l
It's m
Last l
Now, there might be a curiosity that, what if I don't mention the start and the end of the range. Let's see what happens-


$ cut -c - cuttest.txt
cut: invalid range with no endpoint: -
Those who thought that entire columns will be printed, are proved to be wrong. Conclusion is - There has to be a valid range.

B. Extracting Field from a File

In order to understand this usage of cut command, lets consider a csv file as follows:

$ cat employees.txt
Employee ID, Employee Name, Age, Gender, Department, Salary
101, John Davies, 35, M, Finance, $4000
102, Mary Fernandes, 29, F, Human Resources, $3000
103, Jacob Williams, 40, M, Sales, $4700
104, Sean Anderson, 25, M, Production, $2700
105, Nick Jones, 42, M, Finance, $7500
106, Diana Richardson, 29, F, Finance, $3200
Remember, in order to extract a field from a file, we would need a delimiter (i.e. a column separator), based on which the file will be divided into columns and we can extract any of them. In this case, the syntax would be-

Here, we are instructing cut command to use a particular delimiter with option -d and then extract certain fields using option -f.

1. Display a specific field from a file

In case of a csv file, it is crystal clear that our delimiter will be a comma (,). Now, we need to enlist the names of the employees working in our organization, i.e. field number 2.


$ cut -d ',' -f 2 employees.txt
Employee Name
John Davies
Mary Fernandes
Jacob Williams
Sean Anderson
Nick Jones
Diana Richardson
Looks good.

2. Displaying Multiple Fields from a File

Moving forward now, lets display more than one field now. Suppose, we need to include 'Age' and 'Gender' fields also. For this, we must specify the range - again, a start and an end.

$ cut -d ',' -f 2-4 employees.txt
Employee Name, Age, Gender
John Davies, 35, M
Mary Fernandes, 29, F
Jacob Williams, 40, M
Sean Anderson, 25, M
Nick Jones, 42, M
Diana Richardson, 29, F
Conclusion, in this case, is that, Input Delimiter = Output Delimiter.

Lets have a look at a variant in this case. Suppose, we need to extract 'Employee ID', 'Employee Name', 'Department' and 'Salary'. In that case, we need to specify two ranges as below:


$ cut -d ',' -f 1-2,5-6 employees.txt
Employee ID, Employee Name, Department, Salary
101, John Davies, Finance, $4000
102, Mary Fernandes, Human Resources, $3000
103, Jacob Williams, Sales, $4700
104, Sean Anderson, Production, $2700
105, Nick Jones, Finance, $7500
106, Diana Richardson, Finance, $3200
This is just awesome!

3. Change the Delimiter in the Output

As we just saw in one of the examples above, by default, Input Delimiter = Output Delimiter. What if I wish to change the output delimiter? Just have a look at the example below:


$ cut -d ',' -f 2-4 --output-delimiter='|' employees.txt
Employee Name| Age| Gender
John Davies| 35| M
Mary Fernandes| 29| F
Jacob Williams| 40| M
Sean Anderson| 25| M
Nick Jones| 42| M
Diana Richardson| 29| F

4. Do not Display Certain Columns

Just like above example, if we use --complement as an option, cut command will display all the fields, but the specified field.


$ cut -d ',' --complement -f 6 employees.txt
Employee ID, Employee Name, Age, Gender, Department
101, John Davies, 35, M, Finance
102, Mary Fernandes, 29, F, Human Resources
103, Jacob Williams, 40, M, Sales
104, Sean Anderson, 25, M, Production
105, Nick Jones, 42, M, Finance
106, Diana Richardson, 29, F, Finance
That's all for this tutorial. Please do let me know about your views about this article in the comment section below, and stay tuned for more awesome articles.


  1. absolutely awesome
    thanks for sharing

  2. Great stuff! Love the step by step exploration.

  3. This is cool. May be you should also add, cut columns that has specific column name or contains a specific string (regex) as well

  4. I want this to save these extracted fields in same file how could i do so.

    1. $ cut -c -6 cuttest.txt > filename.txt
      This will work for you.

  5. Hi thanks for this tute..
    is there a way in cut to only cut lines starting with a particular character?