linux - Seperate IP from port in CSV file using sed/awk -

i using following create csv output shown below. now, output csv want split ip address , port names in different columns , display process id in last column.

script is:-

netstat -anputw |  awk '{if ($1 == "tcp") print $1,",",$4,",",$5,",",$6,",",$7}' > $home/mylog/connections_$hostname.csv netstat -anputw |  awk '{if ($1 == "udp") print $1,",",$4,",",$5,",",",",$6}' >> $home/mylog/connections_$hostname.csv 

output in csv like:-

tcp      127.0.0.1:25    0.0.0.0:*   listen      1112/sendmail tcp      192.168.0.38:22     192.168.10.143:62998    established     3084987/sshd 

now, wish split ip address , ports (comma seperated) , last column, trim text , display process id. lastly, first column csv should ip address of host name of script run(displayed each of rows). so, output below:-

192.168.0.22  tcp  127.0.0.1     25  0.0.0.0         *       listen        1112 192.168.0.22  tcp  192.68.0.38   22  192.168.10.143  62998   established   3084987 

any pointers how can achieve in csv output?

like previous question, you're approaching wrong. don't need bunch of separate commands , pipes. don't need multiple awk commands. can want in 1 simple awk command.

you didn't provide output netsat see awk input looks rough guess need like:

netstat -anputw | awk '    begin {fs="[ :/]"; ofs=","}    $1 == "tcp" {x=$8; y=$9}    $1 == "udp" {x=""; y=$8}    {print $4,$1,$5,$6,$7,x,y} ' > $home/mylog/connections_$hostname.csv 

replace "x" , "y" more meaningful names if like, don't know values represent.

if updated question show sample input (i.e. output netstat) plus expected output given input can improve/verify script , more.