mjhjmtu

Input:

1,012018,111
2,1-2018,111
3,10-2018,111

Output:

1,01/2018,111
2,01/2018,111
3,10/2018,111

How can i achieve that which meaning edit of month and year to be mm/yyyy

With GNU Awk, using patsplit to split the second comma delimited field into subfields of at most two decimal digits:

$ gawk 'BEGINOFS=FS="," patsplit($2,a,/[0-9][0-9]?/); $2 = sprintf("%02d/%d%d", a[1], a[2], a[3]) 1' file
1,01/2018,111
2,01/2018,111
3,10/2018,111

If Perl is an option

perl -F, -lpe '
 $F[1] =~ s(dd?)-?(d+)sprintf "%02d/%d", $1, $2e; 
 $_ = join ",", @F
' file

sed -e 's/^(.),(.),(.)/2/g' -e 's/([0-9]1,2)([0-9]4)/1-2/g' file*

s/^(.),(.),(.*)/2/g - splits based on commas and gets 2nd value
s/([0-9]1,2)([0-9]4)/1-2/g - splits and inserts '-' between year and month

sed -e 's/^(.),([0-9]1,2)([0-9]4),(.)/1,2-3,4/g' test.txt

• ^(.*), = capture until first ',' and stored in 1


• ([0-9]1,2) = capture 1 or 2 digit number and store in 2


• ([0-9]4) = capture 4 didit no and store in 3


• ,(.*) 4 = capture the rest from ',' and store it in 4


• And replace accordingly