Home/ Questions/Q 8518229
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-06-11T05:55:26+00:00

I have this kind of table: classA, s1, ss1 classA, s1, ss2 classA, s2,

  • 0

I have this kind of table:

classA, s1, ss1
classA, s1, ss2
classA, s2, ss1
classB, s1, ss1
classB, s1, ss2
classC, s1, ss1
classC, s2, ss1
classC, s2, ss2

Which I would like to split in 3 files called “classA”, “classB”, and “classC”, according to the consecutive values in 1st column. I tried first to find a way to do that using csplit, but it seems to work only on a specified pattern, not when the pattern changes. Is there any awk way or any other command-line tool to do that?

[UPDATE]
I have also some slashes in my 1st column that lead to that error, example:

classA, s1, ss1
classA, s1, ss2
classA, s2, ss1
classB, s1, ss1
classB, s1, ss2
classC, s1, ss1
classC, s2, ss1
classC, s2, ss2
classA/classA.1/classA.2, s1, ss1
classA/classA.1/classA.2, s1, ss2

And when I run the command:

awk -F, '{ print $0 > $1}' infile

It works with until “classC” but then I have that error because I guess it interprets the “/” as a path:

fatal: can't redirect to `classA/classA.1/classA.2' (No such file or directory)
Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team
    awk -F, '{ print $0 > $1}' infile

    output

    -rw-r--r-- 1 shellter root  48 Sep 13 14:01 classC
-rw-r--r-- 1 shellter root  32 Sep 13 14:01 classB
-rw-r--r-- 1 shellter root  48 Sep 13 14:01 classA

    where $1 has “/…/path/info, try

    awk -F, '{ outFile="$1"; gsub(/\//, "", outFile); print $0 > outFile }' infile

    If you don’t do the gsub(), any ‘/’ chars will be intrepreted to follow a path to create the output file. Of course the path specfied will need to exist, or you’ll get error messages to that affect.

    Note that some awks prefer to have the ‘/’ char like gsub(/[\/]/,"", fileName), or you might even need [\\/] as the search target.

    IHTH

    • 0