substitue part of a string using sed

Question

Assume the regular expression of a email is [a-zA-Z0-9]+@[a-zA-Z0-9]+.[a-zA-Z0-9]+, I would like to substitute all the id part into "customer", for example

mike@yahoo.com => customer@yahoo.com john@gmail.com => customer@gmail.com

I can write something like

$ echo mike@yahoo.com | sed 's/[a-zA-Z0-9][a-zA-Z0-9]*@[a-zA-Z0-9][a-zA-Z0-9]*\.[a-zA-Z0-9][a-zA-Z0-9]*/customer/g' 

But how can I get the domain part not changed? Basically, the question is find a pattern in a string and substitute just part of it and the remaining part not changed.

Answer 1

You can capture parts of the matched pattern with \(..\) and reuse it in your replacement string using \1:

echo mike@yahoo.com | sed 's/[a-zA-Z0-9][a-zA-Z0-9]*@\([a-zA-Z0-9][a-zA-Z0-9]*\.[a-zA-Z0-9][a-zA-Z0-9]*\)/customer@\1/g'

Answer 2

I think you're making it too complicated:

echo mike@yahoo.com | sed -e 's/.*@/customer@/'