awk replace in multiple columns specific character with one command

Question

Input (a.txt)

aa <tab> c-c-c<tab>k-k-k<tab>
ll <tab> j-j-j <tab>v-v-v<tab>

Needed output (b.txt)

aa <tab> c.c.c<tab> k.k.k<tab>
ll <tab> j.j.j <tab>v.v.v<tab>

Please correct me: (this does not work, works only if I ask to replace in $11 column)

awk -F "\t" -v OFS="\t" '{gsub("-",".",$11,&14,&17); print;}' /home/a.txt > /home/b.txt

@AvinashRaj OP just want to do replacement on those 3 columns. — Kent
– Kent, Commented Dec 2, 2014 at 15:42
So is it important to retain the specific <space><tab><space>, <tab><space> and <space><tab> white space between fields as you show or can all field-separating white space be replaced with, say, a single tab character? The former is difficult, the latter is trivial. — Ed Morton
– Ed Morton, Commented Dec 3, 2014 at 1:08

Avinash Raj · Accepted Answer · 2014-12-02 15:28:39Z

2

You don't need to specify the field number in gsub function if you want to do a global substitution on all the fields.

awk '{gsub(/-/,".")}1' /home/a.txt > /home/b.txt

answered Dec 2, 2014 at 15:21

Avinash Raj

175k32 gold badges247 silver badges289 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Kent · Accepted Answer · 2014-12-02 15:41:46Z

0

it seems that you don't want to do the replacement on whole line, just 3 columns, then just do this:

awk -F "\t" -v OFS="\t" 'BEGIN{a[11]=a[14]=a[17]=7}
                      {for(x in a)gsub("-",".",$x)}7' /home/a.txt > /home/b.txt

answered Dec 2, 2014 at 15:41

Kent

197k36 gold badges248 silver badges317 bronze badges

Don't need the =7 i don't think, BEGIN{a[11];a[14];a[17]} would suffice.