is the file sorted? if so, a simple uniq -d foo.txt > dups.txt will do it..

I wrote one coding but I have 'Out of memory Error'

What's the code look like? Someone here may be able to provide suggestions on how to make it more efficient if you post the snippet.