Remove a character from second occurence

Question

How can I remove , after the first column of every line

$ cat tmp.txt
Name, Charles, James, Criss 
Age, 21, 25, 23

There may be n number of columns in the file.

And I need the output like

$ cat tmp.txt
Name, Charles James Criss
Age, 21 25 23

the , should not be deleted from first column.

I changed the title to something that it is more likely to be searched for. Any hints how to improve it? It still sounds a bit strange to me. — fedorqui, Nov 27 '14 at 12:39

score 6 · Accepted Answer · answered Nov 27 '14 at 11:10

6

You can tell sed to do it globally from the 2nd match:

$ sed 's/,//g2' file
Name, Charles James Criss 
Age, 21 25 23

answered Nov 27 '14 at 11:10

fedorqui

275,237
103
548
598

score 2 · Answer 2 · answered Nov 27 '14 at 11:59

2

sed ':still
s/;//2
t still' tmp.txt

For non GNU sed where /g2 only change occurance n° 2

answered Nov 27 '14 at 11:59

NeronLeVelu

9,908
1
23
43

But `g2` worked fine on my `sed (GNU sed) 4.2.2.`. Could you expand when isn't it working? – fedorqui Nov 27 '14 at 12:03
yes, you are right, it work **on** GNU sed, **not on** posix one (my aix/sun does not make an error but only change second occurance) – NeronLeVelu Nov 27 '14 at 14:24
Aaah ok, I understood it the other way round. +1 for portability then! – fedorqui Nov 27 '14 at 14:35

score 1 · Answer 3 · edited May 23 '17 at 12:30

In Perl, i would do like the below through PCRE verb (*SKIP)(*F).

$ perl -pe 's/^\S+(*SKIP)(*F)|,//g' file
Name, Charles James Criss 
Age, 21 25 23

^\S+ matches one or more non-space characters which are present at the start of a line. (*SKIP)(*F) causes the match to fail. So now it matches all the commas except the one at the first column.

$ perl -pe 's/^[^,]*,(*SKIP)(*F)|,//g' file
Name, Charles James Criss 
Age, 21 25 23

^[^,]*, Match upto the first , then (*SKIP)(*F) make the match to fail, | from the remaining substring , match all the commas.

score 0 · Answer 4 · answered Nov 29 '14 at 02:29

With awk:

'BEGIN{FS=","}{for(i=1;i<NF+1;i++){if(i==1){printf $i", "}else{printf $i" "}};printf "\n"}'

Not the most elegant way to do it, but works nicely. The key here is setting the field separator to the character you wanted to remove anyway "," and then using printf (the formatting print function of awk) to add the "," back in between your field of interest (in this case if(i==1), adds the comma back in after only field 1.

score 0 · Answer 5 · answered Dec 05 '15 at 01:43

0

awk '{gsub(/,/,"")}{$1=$1","}1' file
Name, Charles James Chriss
Age, 21 25 23

Remove every comma and put it back in first column.

answered Dec 05 '15 at 01:43

Claes Wikner

1,457
1
9
8

Remove a character from second occurence

5 Answers5