How could I copy digits from one end of a string to another end of a string? So example,
Input -
Example123:Hello
Exp12:Hey1
Exp:heylo
expected output -
Example123:Hello123
Exp12:Hey112
Exp:heylo
I'm open to using sed or awk, seperator must be accounted for, so row1 is the row to extract digits from and row 2 is the row to place digits
Assuming there's one and only one occurrence of : in each line of the input, you could do something like:
sed 's/\([[:digit:]]*\):.*/&\1/' < input
If there can be more than one : (and you want to append the digits to the end of the line, not the second field), that becomes more complicated, like:
sed 's/^\([^:]*[^:[:digit:]]\)\{0,1\}\([[:digit:]]*\):.*/&\2/' < input
You can use awk here
awk -F':' '{ j=$1; gsub(/[^0-9]+/,"",j); print $0 j }'
Explanation
-F':': tells awk to use colon as the field separator
j=$1; gsub(/[^0-9]+/,"",j) : assigns the first column to to a temporary variable j, and then removes anything that is not a digit from j
print $0 j : finally prints the original string appended with j that carries our number
sub(/.*[^0-9]/, "", j) in case $1 may also contain digits other than at the end.
– Stéphane Chazelas
Jun 10 '19 at 07:04
With the match() function of GNU awk to store the matching group in an array, you could do
awk -F: -v OFS=: 'match($1, /([0-9]+)$/ , arr) { $2 = $2 arr[1] } 1' file
On any POSIX compliant awk you could do
awk -F: -v OFS=: 'match($1, /[[:digit:]]+$/) { $2 = $2 substr($0, RSTART, RLENGTH) }; 1' file
See How to permanently change a file using awk? ("in-place" edits, as with "sed -i") to make the file change persistent.
Tested with below script and worked fine
Count_line=`awk '{print NR}' filename| sort -nr| sed -n '1p'`
for ((i=1;i<=$count_line;i++)); do awk -v i="$i" -F ":" 'NR==i{print $1}' filename| grep -o "[0-9]$">/dev/null; if [[ $? == 0 ]]; then h=`awk -v i="$i" -F ":" 'NR==i{print $1}' filename|grep -o "[0-9]*$"`; awk -v i="$i" -v h="$h" 'NR==i{print $0h}' filename; else awk -v i="$i" 'NR==i{print $0}' filename; fi; done
output
Example123:Hello123
Exp12:Hey112
Exp:heylo
As short as possible in bash... I think...
sed 's/\([0-9]\+:\).*/&\1/' < input
(With a bit a help from @Stèphane's earlier answer.)
