Clever Geek Handbook

How to extract two different columns separated by different delimiters into one row

I have below dataset in unix txt file:

/this/is/a/directory_name/which/i/want:listen= tcp://someurl
/this/is/a/another_directory_name/which/i/want:listen= tcp://someotherurl

output, which is mainly intended:

directory_name <whitespace> tcp://someurl
another_directory_name <whitespace> tcp://someotherurl

Below is the command I'm trying, but I get either url or directory name

cat the_file.txt|awk -F '/' '{print $5}'
cat the_file.txt|awk -F '=' '{print $2}'

could there be a way to run both of the above command at the same time and get the output on the same line? Significantly apciciated

+3


source to share


3 answers


Just save the first command and make a setting for the second: split the full text based on =

and print the second resulting field:

$ awk -F'/' '{split($0, a,"="); print $5, a[2]}' file
directory_name  tcp://someurl
another_directory_name  tcp://someotherurl


Or the last one if there are many =

s:

$ awk -F'/' '{n=split($0, a,"="); print $5, a[n]}' file
directory_name  tcp://someurl
another_directory_name  tcp://someotherurl
+3


source


FROM sed 

$ sed -E 's|^([^/]*/){4}([^/]+).*=|\2|' ip.txt 
directory_name tcp://someurl
another_directory_name tcp://someotherurl


  • -E

    use ERE, some versions will require -r

    instead of
  • ^([^/]*/){4}

    matches upto /

    before a required field
  • ([^/]+)

    required field
  • .*=

    to the last =

    in the line
  • \2

    replace with required field, given patterns have a space before tcp

    , if not guaranteed, uses|^([^/]*/){4}([^/]+).*= *|\2 |

+1


source 


$ awk '{split($1,d,"/"); print d[5], $2}' file
directory_name tcp://someurl
another_directory_name tcp://someotherurl
0


source






More articles:


All Articles