LinuxQuestions.org - [SOLVED] Unwanted blank after word deletion

- Programming (https://www.linuxquestions.org/questions/programming-9/)

- - Unwanted blank after word deletion (https://www.linuxquestions.org/questions/programming-9/unwanted-blank-after-word-deletion-4175432699/)

Unwanted blank after word deletion

Have: a file of lines containing blank-delimited words, such as ...

Code:

one two three four five

Want: a file with the second word in each line removed, such as:

Code:

one three four five

cut does this nicely:

Code:

cut -d" " -f1,3-

However, this awk ...

Code:

awk '{$2=""; print}'

... produces this ...

Code:

one three four five

There are two blanks between "one" and "three". Is there a way to code an awk to generate output with only one blank?

Daniel B. Martin

Code:

{ for (i=1; i<NF; i++)

  if (i!=2)

    printf("%s ", $i)

  printf("%s\n", $i)

}

Changed it a bit. More compact, and removed trailing space.

How about:

Code:

awk 'sub(" "$2,"")' file

@grail

Nice trick ! Must read about awk functions.

Quote:

Originally Posted by grail (Post 4808325)

How about:

Code:

awk 'sub(" "$2,"")' file

Sweet! Thank you! SOLVED!

Daniel B. Martin

Hi.

grail's approach have this little problem:

Code:

$ echo 'one two three four three five' |  awk '{sub(" "$5, "")} 1'

one two four three five

When there are repeated fields, it actually removes first occurrence of the field, not necessarily a given one. In this example third field is removed instead of fifth (which contains same text as 3rd).

I myself yet to discover an elegant way to do this though.

EDIT:

Code:

$ echo 'one two three four five' |  awk '{$2=""; $0=$0; $1=$1} 1'

one three four five

which is equivalent to

Code:

$ echo 'one two three four five' |  awk '{$2=""; gsub(" +", " ")} 1'

one three four five

This way all extra spaces are lost, which may be undesirable.

BTW, just for fun, this also prints on a terminal what you want :)

Code:

$ echo 'one two three four five' |  awk '{$2="\b"} 1'

one three four five