Home / ubuntu / Questions / 584702
Accepted
Roboman1723
Asked: 2015-02-13 06:54:59 +0800 CST

Awk variable passing question

  • 772

This is the code I wrote to resolve DNSs and output to a file. Is there a way to forgo the while loop entirely and just write in awk? It works perfectly but it seems very cumbersome and inefficient, takes about 10 mins for 25k lines of IPs. Will gladly elaborate on script if clarity is needed.

#!/bin/bash

while read line; do
  echo -en " ${startCount} / ${endCount} IPs resolved\r"
  ip=$(echo ${line} | cut -d "," -f1)
  col2=$(nslookup ${ip} | fgrep "name" | sed -e 's/\t/,/g' -e 's/name = //g' -e 's/.uncc.edu.//g' | cut -d "," -f2)
  if [[ ! -z ${col2} ]]; then
    echo "${line}" | awk -F"," -v var="${col2}" '{print $1","var","$2","$3","$4","$5","$6}' >> ${outFile}
  else
    echo "${line}" | awk -F"," '{print $1",""UNRESOLVED"","$2","$3","$4","$5","$6}' >> ${outFile}
  fi
  ((startCount++))
done < ${1}
bash

1 Answers

  1. Best Answer

    Try this:

    gawk -F, '
    # spawns nslookup as a coprocess, passes the IP into its stdin
    # then reads the output of the coprocess to find the hostname
    function get_name(ip,      line, name) { 
        name = "UNRESOLVED"
        print ip |& "nslookup"
        close("nslookup", "to")
        while (("nslookup" |& getline line) > 0) {
            if (match(/name =(.+)\.uncc\.edu\./, line, m)) 
                name = m[1]
        }
        close("nslookup")
        return name
    }
    { print $1, get_name($1), $2, $3, $4, $5, $6 }
' "$1" > "$outfile"

    ref: https://www.gnu.org/software/gawk/manual/html_node/Two_002dway-I_002fO.html

    • 2