Generate CSV from NS Zone File

Question

I have a large file that contains records in multilines. ex:

domain1 NS ns1
domain1 NS ns2
domain1 NS ns3
domain2 NS dnsx

What might be the fastest way to generate a CSV of the form

domain1,ns1,ns2,ns3
domain2, dnsx

I have tried php and groovy scripts but the processor load time is too high for a file (read) of a 1Gb file (and a consequent write to a csv file).

I am thinking there should be a better programatic approach than what I am doing. Basically I am creating a list/array to contain records and check against the last element if the same domain is found in the current line being checked.

P.S Mentioned groovy/php, but response may have no relation to these specific scripting languages.

Answer 1

Assuming all the records with the same domain are grouped together, this awk program will have a very small memory footprint (can't say what the CPU load will be)

awk '
    $1 != domain {
        if (domain) print ""
        printf "%s", $1
        domain = $1
    } 
    {printf ",%s", $3} 
    END {print ""}
' file