convert ldapsearch result into external datagroup format

Give me a couple of hours. I have some perl scripts that do stuff like this. I will dig them out.

 

 

H

 

Hi Guys,

I come cross a requirement to convert a txt file into external datagroup format in background.

The source file is the result from ldapsearch, something like this:

dn: cn=client1+givenName=client1+m...erve

r1+sn=test,ou=people,dc=my-domain,dc=com

cn: client1

mailHost: mailserver1

dn: cn=client2+givenName=client2+m...erve

r2+sn=test,ou=people,dc=my-domain,dc=com

cn: client2

mailHost: mailserver2

I need a scripts to convert it into external datagroup format like this:

"client1" := "mailserver1",

"client2" := "mailserver2",

Then use datagroup as a reference to distribution traffic to mailservers.

If anyone has similar scripts on hand, please share me.

Thanks a lot!!

Here’s a quick and dirty one liner:

cat ldif.txt |egrep "^(cn:|mailHost:)" |cut -d: -f 2- |sed -e 's/^ //' -e 'N;s/\n /|/' |awk -F\| '{print "\""$1"\" := \""$2"\","}'

Sample ldif file

$ cat ldif.txt

dn: cn=client1+givenName=client1+mail=c...=mailserve

r1+sn=test,ou=people,dc=my-domain,dc=com

cn: client1

mailHost: mailserver1

dn: cn=client2+givenName=client2+mail=c...=mailserve

r2+sn=test,ou=people,dc=my-domain,dc=com

cn: client2

mailHost: mailserver2

dn: cn=client3+givenName=client3+mail=c...=mailserve

r3+sn=test,ou=people,dc=my-domain,dc=com

cn: client3:test:with:colons

mailHost: mailserver3

Output for the sample file

$ cat ldif.txt |egrep "^(cn:|mailHost:)" |cut -d: -f 2- |sed -e 's/^ //' -e 'N;s/\n /|/' |awk -F\| '{print "\""$1"\" := \""$2"\","}'

"client1" := "mailserver1",

"client2" := "mailserver2",

"client3:test:with:colons" := "mailserver3",

Breakdown of the commands:

cat ldif.txt - write file to standard output

egrep "^(cn:|mailHost:)" - get lines starting with cn: or mailHost:

cut -d: -f 2- - split the lines on a delimiter of : and return the second through last fields

sed -e 's/^ //' -e 'N;s/\n /|/' - remove leading spaces and replace every other new line with a pipe

awk -F\| '{print "\""$1"\" := \""$2"\","}' - split on the pipe and format the fields with “name” := “value” format

And here's a cleaner option from John Gruber using Perl:

!/usr/bin/perl
if($ARGV < 1) {
    print $ARGV."\n";
    print "$0 [ldifffile] [datagroupfile]\n";
    exit(0);
}
open(LDIFFFILE, "<".$ARGV[0]) or die "Can't open ".$ARGV[0]."\n";
open(DATAGROUP, ">".$ARGV[1]) or die "Can't open ".$ARGV[1]."\n";
while()
{
  my($line) = $_;
  chomp($line);  
  if ($line =~ /^cn:/) {
      print DATAGROUP "\"".substr($line,index($line,': ')+2)."\" := \"";
  }
  if ($line =~ /^mailHost:/) {
      print DATAGROUP substr($line, index($line,':')+2)."\"\n";
  }
}
close(LDIFFFILE);
close(DATAGROUP);

Aaron

And an even simpler awk example from James Quinby:

 

 

 

awk 'BEGIN{RS=" \n";FS="\n"};{print $3,$4}' input.txt | awk '{print "\042"$2"\042 := \042"$4"\042,"}'

 

 

First awk breaks the records by blank-lines (RS) and the fields (FS) by newlines – this prints the last 2 lines of each record side-by-side.

 

Second awk splits by spaces and prints the correct fields, delimiter and double-quotes.

 

 

N.B.: The sample input was a space on a blank line between records – you’ll want to adjust the RS in that first awk accordingly if the actual record separator is a true bare linefeed. This could probably be combined into a single awk script.

 

 

 

Aaron