Hi
Can some tell to use AWK or SED in the following situtation

file A has 10 columns, whenever column 10 has valuues AorBorCorDorE replace the column value with X else if it has GorH replace the column value with Y.

I'm not too user which one to use, whether mix of both of anone wil do ?

Thanks in advance

BrainDrain

Hi,
I was just trying the following, however the value i'm trying to replace($7) with 'Z' for the occurences of 'A' retains it's orginal value. Am I doing something stupid ?

awk -F"~" '{if ($7=='A') $7='Z' ; print $1"~"$7; }' cif_in.tsv

I guess what you need to use is $7 ~ /A/
here the ~ is used to match the regular expression /A/

awk -F"~" '{
if($7 ~ /A/) {
print $1"~"$2"~"$3"~"$4"~Z~"$6"~"$7
}
else {
print $0
}
}' cif_in.tsv

Hi,
Thanks for the reply, Well i'm able to evaluvate
if ($7=='A') in AWK, however assignment was not happening. May be i should try printing directly the value as you mentioned, instead of assignment. I thought there must be much neat way of doing this in awk.

Try this

$ cat adata
aa bb cc dd ee ff gg hh ii jj
qq ww ee rr tt yy uu ii oo pp
aa ss dd ff gg hh jj kk ll lo
$ awk ' { if ($7 == "uu") $7="zz"; print $0 } ' adata
aa bb cc dd ee ff gg hh ii jj
qq ww ee rr tt yy zz ii oo pp
aa ss dd ff gg hh jj kk ll lo
$

A slight refinement to Perderabo's solution. This is a bit more awk-like.

awk '$10 == "uu" { $10 = "@@" } { print }' adata

awk '$10 == "uu" || $10 == "99" || $10 == "**" { $10 = "zz" }
$10 == "aa" || $10 == "bb" { $10 = "yy" }
{ print }' adata