Bash: Grab part of string from a command line output

Question

I am running a command in CentOS that gives me an output of a string and I want to grab a certain part of that output and set it to a variable.

I run the command ebi-describe-env.

My output as follows:

ApplicationName | CNAME | DATECreated | DateUpdated | Description | EndpointURL |   
EnvironmentID | EnvironmentName | Health | Stack | Status | TemplateName | 
Version Label --------------------------
Web App | domain.com | 2012-02-23 | 2012-08-31 | |
anotherdomain.com | e-8sgkf3eqbj | Web-App-Name | Status | 
Linux | Ready | N/A | 20120831 - daily

I want to grab the '20120831 - daily' part of the string (this string will always change but stays in the same place) and set it to a variable.

Originally I thought I could use grep or sed and print a line after each '|' and set the 13th line to a variable.

I'm very new to bash scripting, so any help would be great. Thank you.

Is that all supposed to be on one line?

Neil
– Neil

2012-08-31 23:44:02 +00:00
Commented Aug 31, 2012 at 23:44 — Neil
– Neil, Commented Aug 31, 2012 at 23:44
Yes, it is supposed to be on one line.

Michael Le
– Michael Le

2012-09-04 20:48:03 +00:00
Commented Sep 4, 2012 at 20:48 — Michael Le
– Michael Le, Commented Sep 4, 2012 at 20:48

Levon · Accepted Answer · 2012-09-04 21:41:55Z

9

Using awk:

awk -F"|" '{print $NF}'

this will work:

echo " Web App | domain.com | 2012-02-23 | 2012-08-31 | | anotherdomain.com | 
       e-8sgkf3eqbj | Web-App-Name | Status | Linux | Ready | N/A | 
       20120831 - daily" | awk -F"|" '{print $NF}'

and yield:

20120831 - daily

To assign to a variable (data.txt contains your string just for simplicity, it also works the echo above):

$ myvar=$(awk -F"|" '{print $NF}' data.txt)
$ echo $myvar
20120831 - daily

Explanation

the -F sets the input field separator, in this case to |. NF is a built-in awk variable that denotes the number of input fields, and the $ in front of the variable accesses that element, i.e., in this case the last field in the line ($NF).

Alternatively: You could grab each of the last three fields separated by white space (the awk default) with this:

awk '{print $(NF-2), $(NF-1), $NF}'

edited Sep 4, 2012 at 21:41

answered Aug 31, 2012 at 23:31

Levon

144k35 gold badges205 silver badges194 bronze badges

Sign up to request clarification or add additional context in comments.

6 Comments

Levon Over a year ago

@MichaelLe Updated the answer to show the assignment to a shell variable

Levon Over a year ago

@MichaelLe Happy to have been able to help.

Michael Le Over a year ago

I realized that I had a larger output and formatted my original question. I tried using $(NF-1) and such to grab the next fields, but it turns out that it is on the incorrect line. How would I go to the next line in the output? Thank you.

Levon Over a year ago

@MichaelLe That's a very different problem and really merits its own new question. It's one thing to grab part of a line (which is what you asked for originally) and then to have a different problem that consists of a multi-line (always? sometimes?) input where parts you are interested in are on potentially different lines (if I understood your comment). I would suggest you accept an answer that solved the problem you posted, and then carefully formulate another separate question to post new (including accurate input data and desired output). AWK, like other solutions here, is line-oriented.

Michael Le Over a year ago

Thank you, I didn't realize that there was more effort because it was multiline.

|

Jeremy J Starcher · Accepted Answer · 2012-08-31 23:37:38Z

3

Levon's answer works great, but I just had to show there are always other ways with shell scripting.

This one uses the basic tool called cut

echo "Web App | domain.com | 2012-02-23 | 2012-08-31 | | anotherdomain.com |  e-8sgkf3eqbj | Web-App-Name | Status | Linux | Ready | N/A |   20120831 - daily" | cut -d"|" -f13

answered Aug 31, 2012 at 23:37

Jeremy J Starcher

24k6 gold badges56 silver badges75 bronze badges

2 Comments

Michael Le Over a year ago

Can you explain what the -f13 does?

Jeremy J Starcher Over a year ago

If you have access to a Unix/Linux/*nix box, pull up a terminal and type man cut. In short, the -d"|" indicates that the | symbol separates fields, and the -f13 part asks for the 13th field. cut -d"|" -f13

Neil · Accepted Answer · 2012-09-04 23:59:27Z

0

I know that this has been accepted already, but here's how to do it in pure bash:

string="Web App | domain.com | 2012-02-23 | 2012-08-31 | | anotherdomain.com | e-8sgkf3eqbj | Web-App-Name | Status | Linux | Ready | N/A | 20120831 - daily"
myvar="${string##*| }"
echo "$myvar"

answered Sep 4, 2012 at 23:59

Neil

55.5k8 gold badges65 silver badges75 bronze badges

3 Comments

user2065875 Over a year ago

Is there a reason to avoid Awk in this case? Awk is specified by POSIX and provided by Busybox. Smaller embedded systems don't have it, maybe? This is not a rhetorical question.

Neil Over a year ago

@ngks No, I just felt like pointing out that bash's internal string constructs are up to this particular job.

user2065875 Over a year ago

Cool, as @Levon said there are always other ways.

Collectives™ on Stack Overflow

Bash: Grab part of string from a command line output

3 Answers 3

6 Comments

2 Comments

3 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

6 Comments

2 Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related