Styjun

fairly new to using linux on shell.

I want to reduce the amount of pipes I used to extract the following data.

V 190917135635Z 1005 unknown /C=DE/ST=City/L=City/O=something/OU=Somewhat/CN=someserver.com/emailAddress=test@toast.com

My goal is to put the following values into a separate file

190917135635 someserver.com

The command I use right now is fairly long, piped and looks like this

grep -v '^R' $file | awk 'print $2, $6' | awk -F'[=|/]' 'print $1, $3' | awk 'print $1, $3' | awk -F 'Z ' 'print $1, $2' > sdata.txt

(The file contains other lines starting with 'R' so I exclude those in my grep)

Is this a legit way of doing it?

Is there a way to get this in a shorter command?

Thanks a lot!

Another awk. Using match to find CN entry and substr to extract it for print to print, if it exists.

$ awk '!/^R/
 print $2,
 (match($0,/CN=[^/]+/)?substr($0,RSTART+3,RLENGTH-3):"") # 3==length("CN=")
' file

Output:

190917135635Z someserver.com

Looks some of your data fields are used as creating SSL certificates, thus many fields might contain SPACES, i.e. City, Organization Name etc. That's why you need many awk lines(???). Here is one way which might help you overcome these issues. So instead of transforming your existing code logic, the target is to find the domain name by searching the substring CN= and fetching its corresponding value.

awk '
 !/^R/
 start = index($0, "CN=")+3
 end = index(substr($0, start), "/")
 domain = end ? substr($0, start, end-1) : substr($0, start)
 print $2, domain
 
' file.txt

Where:

• we use index() to find the start-position of the substring CN=, +3 will be the starting point of the domain name


• then we search the next / to get the end-position of this domain. if it's at the end of the line, there will be no / and thus end will be '0'


• then we get the domain name between the substring CN= and the next '/' by using substr($0, start, end-1) or the end of line by using substr($0, start).

A short version:

awk '!/^R/s=index($0, "CN=")+3; e=index(substr($0, s), "/"); print $2, substr($0, s, e ? e-1 : 253)' file.txt

where 253 is the longest possible domain name which might be enough to fit your needs.

Update:

Actually, it's much easier just use match(), but the point is the same:

awk '!/^R/if(match($0, "/CN=([^/]*)")) print $2, substr($0, RSTART+4, RLENGTH-4)' file.txt

If this:

$ awk -F'[[:space:]/=]+' '!/^R/print $2+0, $16' file
190917135635 someserver.com

isn't all you need then updated your question to clarify your requirements and provide more truly representative sample input/output.

Using GNU sed:

sed -E -n '/^R/d; s/^[A-Za-z]s+([0-9]+)s+[0-9]+s+.*/CN=(.*)/.*/1 2/p' input_file > new_file

EDIT: Strictly considering that OP's Input_file is same as shown samples only. After seeing OP's samples one could try following.

awk -F"[ =/Z]" '!/^R/print $8,$37' Input_file

For FUN :) in case one want to try in OP's approach then we could try following.

awk '
!/^R/
 val=$2 OFS $5
 split(val,array,"[ /Z]")
 val1=array[1] OFS array[9] OFS array[10]
 split(val1,array1,"[ =]")
 print array1[1],array1[3]

' Input_file

You are using $6 in the second awk command, that means your 5th column has potentially spaces inside unlike the sample data you showed, also it is extracting CN= part (CNAME?).

So here's a more compatible and more exact sed way which does not require GNU sed:

sed -n -e '/^R/!p;'

If you just want digits in the second column and it begins with digit, then you can change to use this:

sed -n -e '/^R/!p;'