Output a word in a file

[lexarrow]

Hi all,

Here's my problem:

I made a script that downloads, installs and configures LDAP from source
I created a user input variable with the base dn ( eg. dc=example,dc=com) . Now I want to automatically create and add an init.ldif file and I need to output `example` from `dc=example,dc=com` to a file

I looked at cat and grep help but I couldn't find the answer.

Any suggestions?

THX!

[ghostdog74]

to output something to a file from a script, use the ">" or ">>" (append) operator. Otherwise, describe your problem more clearly, providing example inputs and output.

[sisco311]

sed -i 's/^dc=example*/&,dc=com/' filenamethis command will append the lines, in the file filename, starting with dc=example with ,dc=com

dc=example -> dc=example,dc=com

[lexarrow]

I rephrase a little. I have set a variable like this:


dialog --title "LDAP Server Configuration" --backtitle "Server Install" --inputbox "Enter the desired BASE of the LDAP server" 8 60 2>/tmp/basedn
basedn=`cat /tmp/basedn`

I need a way to output the first word after `dc=` in /tmp/basedn

[risby]

There are many many ways to do this. The most obvious use cut or awk or perl.

To extract the first occurrance of "example" from the string "dc=example -> dc=example,dc=com" held in file /tmp/basedn and store it in variable basedn use any of:


basedn=$( cut -d'=' -f2 /tmp/basedn | cut -d' ' -f1 )

basedn=$( awk -F'=' {'print $2'} /tmp/basedn | awk -F' ' '{print $1}' )

basedn=$( perl -ple 's/dc=(.*?) .*/$1/' /tmp/basedn )

[ghostdog74]

no need to call awk that many times.


# more tmp
dc=example,dc=com
# awk -F"[,=]" '{print $2}' tmp
example

[WW]

Or sed...


basedn=$( sed 's/^dc=\(.*\),.*/\1/' /tmp/basedn )

This assumes that /tmp/basedn has just one line of text.

[risby]

no need to call awk that many times.


# more tmp
dc=example,dc=com
# awk -F"[,=]" '{print $2}' tmp
example


Interesting but that doesn't work in the same way on either my HP-UX or Linux systems:


$ cat /tmp/basedn
dc=example -> dc=example,dc=com
$ awk -F"[,=]" '{print $2}' /tmp/basedn
example -> dc
$
and that is what I would expect. It uses the first two field separators it finds, i.e. in this case both '=' symbols and returns all the text between them.

Your awk appears to be broken ! EDIT: Ah I see you are using a different input example from the OP.

So


awk -F"[ =]" '{print $2}' /tmp/basedn
would work nicely.

RE-EDIT: Oh dammit, I've re-read the spec, it was my input example that was wrong.