Programming Forums > Script Programming > Sed and Awk

keenlearner · Sep 7th, 2004, 10:49 AM

hi!

how can i store the output of awk command in a script..
for eg:

in TRY.sh

awk '{ printf "%d", $1 }' filename
var=$1

i want something of this sort...
but this particular thing does'nt work..
i mean if the $1 value is 50 than i am not able to store it in a different variable
for further calculations..

so pls help me out...

TAKE CARE
Regards
Keen Learner

Infinite Recursion · Sep 7th, 2004, 11:34 AM

your_awk_command > yourfile

awk -F":" '/Grendel/{print $1 " - " $2 " - " $3}' data.txt > results.txt

The results of the parsing of data.txt will be stored in results.txt.

let us know if you have problems with this.

keenlearner · Sep 7th, 2004, 1:11 PM

It works definetely.. but..
i dont want to store the output in another file..
i want the output in a variable so that i can further use that value in my script.

actually i am picking a particular value from a file..
and that is the first field of dat file..
i am able to get it but cannot able to store it..

i want it in a variable...dat value is a numeric value.

hope u understand my problem...

Infinite Recursion · Sep 7th, 2004, 6:45 PM

Oh... sorry, I mis-read. The easiest way would be to set an environment variable and read in that variable into the other program.

erebus · Sep 7th, 2004, 9:20 PM

Actually, most newer implementations of awk (POSIX, Nawk, or Gawk) has the -v switch so you can interface awk and shell variables. Something like this should suffice:

(Toggle Plain Text)

awk -v name="$USER" '{ print name }'

Or you can use a built-in associative array that accepts a current environment variable name as its index, which is quite handy for working with the shell:

(Toggle Plain Text)

awk '{ print ENVIRON["USER"] }'

Another less efficient way would be to use the export command, though I won't give an example because I like the above methods, but it can be done.

For your problem, it's a little bit tricky. Maybe something like:

(Toggle Plain Text)

var=`awk '{ printf "%d", $1 }' filename`

Not sure if that's what you were looking for, but working with the above information will get you where you need to go. This will substitue anything in back-quotes with the output of the command sequence within the two, thus, saving what should have gone to stdout to a variable. Sorry I couldn't be of much help.

keenlearner · Sep 8th, 2004, 1:33 PM

definitely i want something like..

var=`awk '{printf "%d".$1}' filename`

but this particular command will not work...
this is the problem i am facing that i am not able to write that particular syntax of awk....

could u pls suggest something more...

Thanx
With Regards

erebus · Sep 8th, 2004, 8:46 PM

What shell are you using that doesn't let you use this to put command output in a variable. If you're using csh or tcsh, use set to do variables(man set). If not, wish ya luck...

EDIT: I noticed that the extension on your script is *.sh, so you probably want to use that. If sh is not your default shell(which i'm not making the assumption it's not, but it's a possibilty), use the shebang line at the beggining of the script(#!/bin/sh). If you don't know what your shell is, 'echo $SHELL' and get back to us.

sarumont · Sep 9th, 2004, 8:51 PM

Dupe, as called by erebus. B)

erebus · Sep 9th, 2004, 10:42 PM

I wish I got the joke, dude :-P.

Vorlin · Feb 27th, 2005, 4:34 PM

In trying to understand your script and what you're doing with it, here's what I did.

I created a 'foo.dat' file in my home directory ($HOME in env var). In this file, there's a simple line that has '10:20:30:40' in it. I didn't know the format of your dat file but you said it was the first variable so here we go.

Then I created a global variable that was exported. Something to remember about exporting a variable from the command line is that it will only be available in that session unless you add it as an alias executed during your login or add it to the skeleton profile, etc...

(Toggle Plain Text)

$ /home/booyaka > export FOO=`cat $HOME/foo.dat | head -1 | awk -F: '{ print $1 }'`
$ /home/booyaka > echo $FOO
10

Or you could do something like this, just as an alternative:

(Toggle Plain Text)

$ /home/booyaka > export FOO=`cat $HOME/foo.dat | head -1 | cut -d: -f1`
$ /home/booyaka > echo $FOO
10

If you're trying to do something else, and I misinterpreted, my bad...

Sep 7th, 2004, 11:34 AM	#2
Infinite Recursion Programming Guru Join Date: Jul 2004 Location: United States Posts: 3,467 Rep Power: 8	your_awk_command > yourfile awk -F":" '/Grendel/{print $1 " - " $2 " - " $3}' data.txt > results.txt The results of the parsing of data.txt will be stored in results.txt. let us know if you have problems with this. __________________ http://jasonpowers.net "There are a thousand hacking at the branches of evil to one who is striking at the root."

Sep 7th, 2004, 6:45 PM	#4
Infinite Recursion Programming Guru Join Date: Jul 2004 Location: United States Posts: 3,467 Rep Power: 8	Oh... sorry, I mis-read. The easiest way would be to set an environment variable and read in that variable into the other program. __________________ http://jasonpowers.net "There are a thousand hacking at the branches of evil to one who is striking at the root."

Sep 7th, 2004, 9:20 PM	#5
erebus Programmer Join Date: Aug 2004 Location: /dev/null Posts: 65 Rep Power: 5	Actually, most newer implementations of awk (POSIX, Nawk, or Gawk) has the -v switch so you can interface awk and shell variables. Something like this should suffice: (Toggle Plain Text) awk -v name="$USER" '{ print name }' awk -v name="$USER" '{ print name }' Or you can use a built-in associative array that accepts a current environment variable name as its index, which is quite handy for working with the shell: (Toggle Plain Text) awk '{ print ENVIRON["USER"] }' awk '{ print ENVIRON["USER"] }' Another less efficient way would be to use the export command, though I won't give an example because I like the above methods, but it can be done. For your problem, it's a little bit tricky. Maybe something like: (Toggle Plain Text) var=`awk '{ printf "%d", $1 }' filename` var=`awk '{ printf "%d", $1 }' filename` Not sure if that's what you were looking for, but working with the above information will get you where you need to go. This will substitue anything in back-quotes with the output of the command sequence within the two, thus, saving what should have gone to stdout to a variable. Sorry I couldn't be of much help. __________________ <span style='font-size:14pt;line-height:100%'><span style='color:red'>"Political power grows out of the barrel of a gun" - Mao Tse-Tung</span></span>

Sep 8th, 2004, 8:46 PM	#7
erebus Programmer Join Date: Aug 2004 Location: /dev/null Posts: 65 Rep Power: 5	What shell are you using that doesn't let you use this to put command output in a variable. If you're using csh or tcsh, use set to do variables(man set). If not, wish ya luck... EDIT: I noticed that the extension on your script is .sh, so you probably want to use that. If sh is not your default shell(which i'm not making the assumption it's not, but it's a possibilty), use the shebang line at the beggining of the script(#!/bin/sh). If you don't know what your shell is, 'echo $SHELL' and get back to us. __________________ <span style='font-size:14pt;line-height:100%'><span style='color:red'>"Political power grows out of the barrel of a gun" - Mao Tse-Tung</span>*</span>

Sep 9th, 2004, 8:51 PM	#8
sarumont Hobbyist Programmer Join Date: Apr 2004 Location: /dev/urandom Posts: 154 Rep Power: 5	Dupe, as called by erebus. B) __________________ "Time is an illusion. Lunchtime doubly so." -the late, great Douglas Adams

Sep 7th, 2004, 10:49 AM	#1
keenlearner Newbie Join Date: Sep 2004 Posts: 13 Rep Power: 0	hi! how can i store the output of awk command in a script.. for eg: in TRY.sh awk '{ printf "%d", $1 }' filename var=$1 i want something of this sort... but this particular thing does'nt work.. i mean if the $1 value is 50 than i am not able to store it in a different variable for further calculations.. so pls help me out... TAKE CARE Regards Keen Learner

Sep 7th, 2004, 1:11 PM	#3
keenlearner Newbie Join Date: Sep 2004 Posts: 13 Rep Power: 0	It works definetely.. but.. i dont want to store the output in another file.. i want the output in a variable so that i can further use that value in my script. actually i am picking a particular value from a file.. and that is the first field of dat file.. i am able to get it but cannot able to store it.. i want it in a variable...dat value is a numeric value. hope u understand my problem...

Sep 8th, 2004, 1:33 PM	#6
keenlearner Newbie Join Date: Sep 2004 Posts: 13 Rep Power: 0	definitely i want something like.. var=`awk '{printf "%d".$1}' filename` but this particular command will not work... this is the problem i am facing that i am not able to write that particular syntax of awk.... could u pls suggest something more... Thanx With Regards

Sep 9th, 2004, 10:42 PM	#9
erebus Programmer Join Date: Aug 2004 Location: /dev/null Posts: 65 Rep Power: 5	I wish I got the joke, dude :-P. __________________ <span style='font-size:14pt;line-height:100%'><span style='color:red'>"Political power grows out of the barrel of a gun" - Mao Tse-Tung</span></span>

Feb 27th, 2005, 4:34 PM	#10
Vorlin Newbie Join Date: Feb 2005 Posts: 12 Rep Power: 0	In trying to understand your script and what you're doing with it, here's what I did. I created a 'foo.dat' file in my home directory ($HOME in env var). In this file, there's a simple line that has '10:20:30:40' in it. I didn't know the format of your dat file but you said it was the first variable so here we go. Then I created a global variable that was exported. Something to remember about exporting a variable from the command line is that it will only be available in that session unless you add it as an alias executed during your login or add it to the skeleton profile, etc... (Toggle Plain Text) $ /home/booyaka > export FOO=`cat $HOME/foo.dat \| head -1 \| awk -F: '{ print $1 }'` $ /home/booyaka > echo $FOO 10 $ /home/booyaka > export FOO=`cat $HOME/foo.dat \| head -1 \| awk -F: '{ print $1 }'` $ /home/booyaka > echo $FOO 10 Or you could do something like this, just as an alternative: (Toggle Plain Text) $ /home/booyaka > export FOO=`cat $HOME/foo.dat \| head -1 \| cut -d: -f1` $ /home/booyaka > echo $FOO 10 $ /home/booyaka > export FOO=`cat $HOME/foo.dat \| head -1 \| cut -d: -f1` $ /home/booyaka > echo $FOO 10 If you're trying to do something else, and I misinterpreted, my bad... Last edited by Vorlin; Feb 27th, 2005 at 4:35 PM. Reason: Forgot to add tail part...

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Programming Forums > Script Programming > Sed and Awk

Awk Help Needed...urgent!