Filter ldapsearch with awk/bash

Question

By writing an ldapsearch command such as:

ldapsearch -h ipaddress -p 389 -D "cn=func_01_acc,ou=admins,dc=akademia,dc=int" \
-w akademia.01 -b "ou=stud01,dc=akademia,dc=int" "(l=Torun)" sn cn telephonenumber -LLL |
grep sn: | awk '{print $2 "|" $1}' | sort | uniq -u 

I got output:

 dn: uid=s21705,ou=stud01,dc=akademia,dc=int
 telephoneNumber: [phone]
 cn: Benny Flowers

 dn: uid=s21711,ou=stud01,dc=akademia,dc=int
 telephoneNumber: [phone]
 cn: Jacqueline Newton

The question is: how to format it to

Benny Flowers|[phone]
JAcqueline Newton|[phone]

The farest I went was:

ldapsearch -h ipaddress -p 389 -D "cn=func_01_acc,ou=admins,dc=akademia,dc=int" \
-w akademia.01 -b "ou=stud01,dc=akademia,dc=int" "(l=Torun)" sn cn telephonenumber -LLL |
grep sn: | grep  dn: | sed 's/telephoneNumber: //' | sed 's/cm: //'

And I have:

 [phone]
 Name Surname

 [phone]
 Name Surname

I tried to invert number and names with:

 for ((i=0;i<$elements;i++)); do
      echo ${array[${i}]}
 done

It printed output, but I cannot find out how to pass value to temporary array for names and for another array handling numbers.

Any suggestions would be helpful.

Liquefy

Post the ldap output before you do the grep sn: | awk '{print $2 "|" $1}' | sort | uniq -u and we can help you. I don't understand how you can get the output you say you do though after a pipe that ends in sort | uniq -u. – Alcohol

Answer 1

If you pipe your first output to:

sed -rn '/ telephoneNumber: /{N;s/[^:]*: (.*)\n[^:]*: (.*)/\2|\1/p;}'

You will get:

Benny Flowers|[phone]
Jacqueline Newton|[phone]

Use -En instead on OSX.

sed will search each telephoneNumber entries, N will place that and the following line together, () will match the name and the phone number and finally \2|\1 formats the resulting string.

Guiscard

Answer 2

awk

$ awk -v OFS='|' '{split($0,a,": ")} /^telephoneNumber:/{tel=a[2]} /^cn/{cn=a[2]; print cn, tel}' ldapoutput.txt 
Benny Flowers|[phone]
Jacqueline Newton|[phone]

shell

$ cat foo.sh 
#!/bin/bash

while IFS=':' read -r key value; do
        case ${key} in
                cn|telephoneNumber)
                        read -r "${key}" <<<"${value## }" ;;
                *)
                        continue
                        ;;
        esac
        [ "${key}" = cn ] && printf "%s|%s\n" "${cn}" "${telephoneNumber}"
done

.

$ ./foo.sh < ldapoutput.txt
Benny Flowers|[phone]
Jacqueline Newton|[phone]

Assignor

Answer 3

grep -E '^(telephoneNumber|cn):' | cut -d':' -f2 | sed 's#\s##' | sed -r '$!N;s/(.*)\n(.*)/\2|\1/'

Heindrick

Maybe a little comment as to how and why this works? – Prink