Search code examples
bashsedcygwingnuwin32

Sed - Remove digits from end of line/string

I'm trying to remove digits from the end of each line within a text file, I've tried using:

sed 's/[0-9]\+$//' <input.txt >output.txt

but it doesn't work, however if I a echo input, ie

echo "Bla1Bla123" | sed 's/[0-9]\+$//'

it will work correctly and output as expected

Bla1Bla

removing " 123 " from the end.

however when I try using input.txt it fails to do that for each line..

example of input.txt is

heyexample1:123example123
h123h:i123i123

which should then output

heyexample1:123example
h123h:i123i

instead it makes no alterations

Solution

  • Your files are probably encoded with the \r\n line endings (since you are on Windows). Try 's/[0-9]\+\r$/\r/' replace command (preserving the \r intentionally).

    There might be a way for sed to handle CRLF line endings, but I cannot find it after a brief search.