[SOLVED] change desired field in passwd or shadow

vikas027 · 06-09-2008, 10:03 AM

Hi,

I need to change field as desired in /etc/passwd or /etc/shadow manually separated by delimiter ":"

For eg. I have an entry of /etc/shadow file

vikas:$1$WYD1mVMs$0ntLUvT4UgZ2FDa4V9spw:13991:0:99999:7:::

I need to change 2nd field of it manually which is between 2nd and 3rd colon , ":"

I did some awk and sed work but could suceed.

Pls help.

Thanks in adv.

VIKAS

colucix · 06-09-2008, 10:28 AM

In awk you can try something like this

Code:

awk -F: 'BEGIN{OFS=":"}{gsub(/.*/,"newstring",$3)}1' /etc/passwd

or if you want to selectively change the third field - e.g. based on the user name:

Code:

awk -F: 'BEGIN{OFS=":"}/vikas/{gsub(/.*/,"newstring",$3)}1' /etc/passwd

vikas027 · 06-09-2008, 10:32 AM

Quote:

Originally Posted by colucix

In awk you can try something like this

Code:

awk -F: 'BEGIN{OFS=":"}{gsub(/.*/,"newstring",$3)}1' /etc/passwd

or if you want to selectively change the third field - e.g. based on the user name:

Code:

awk -F: 'BEGIN{OFS=":"}/vikas/{gsub(/.*/,"newstring",$3)}1' /etc/passwd

Thank you so much, you people rock.

vikas027 · 06-09-2008, 11:35 AM

Hi Colucix,

Your code is working perfect BUT when I am passing variable in it, the desired position is not changing.

Like in this example, I am looking for user "vikas" in /etc/shadow

Code:

[root@RHEL vikas]# awk -F: 'BEGIN{OFS=":"}/vikas/{gsub(/.*/,"police",$2)}1' /etc/shadow
vikas:police:13991:0:99999:7:::
[root@RHEL vikas]# 

[root@RHEL vikas]# a=$(cat file.txt | head -1 | awk '{print$1}')
[root@RHEL vikas]# echo $a
ZENITH

[root@RHEL vikas]# echo $b
vikas

I have tried passing the value of this variable a in this, but its not giving desired output.
I am looking for user vikas which is in variable b.

Code:

[root@RHEL vikas]# awk -F: 'BEGIN{OFS=":"}/$b/{gsub(/.*/,$a,$2)}1' /etc/shadow | grep vikas

Please help, guess I am making some small mistake.

Thanks in advance.
VIKAS

colucix · 06-09-2008, 11:59 AM

You cannot pass the value of a shell variable to an awk script in this way. Note that the awk code is enclosed in single quotes, hence $a is passed literally to awk (it is not expanded/evaluated by the shell). To pass an external variable to awk you have to use the -v option like this

Code:

awk -F: -v string=$a 'BEGIN{OFS=":"}/vikas/{gsub(/.*/,string,$2)}1' /etc/shadow

here you assign the value $a to the variable "string" and use this inside the awk code.

vikas027 · 06-09-2008, 12:03 PM

Hi,

I also meanwhile found other option

Quote:

[root@RHEL vikas]# awk -F: 'BEGIN{OFS=":"}/^'$b'/{gsub(/.*/,'$a',$2)}1' /etc/shadow | grep ^$b: | sed 's/'^$b':/&'$a'/1'
vikas:ZENITH:13991:0:99999:7:::

trying your option too.

Thanks mate, thanks a ton.