Question

I’m using the following command with sed : sed s/1[^\.]/2/g test

I want to replace every 1 with a 2 except the one followed by a point. It works fine.

But in my file I’ve got a {{1e}} that is replaced by {{2}}.

Why does sed delete the e?

PS: I’m a beginner with regex so it might be an obvious mistake.

Answer 1

Capture the character after 1 and insert it again after 2:

sed 's/1\([^.]\)/2\1/g' test

The only problem is 1 at the end of line. You can replace it by 2 with

s/1$/2/