How to group "Only in" lines in diff at the beginning of the output

Question

I did:

diff -r directory1/ directory2/

Some files are different, and some files were only in one tree or the other, creating several Only in ... lines.

How do I group these lines at the beginning of the file? I am 99.9% sure that neither directory contains any line that begins with Only in. I was considering something like:

diff -r directory1/ directory2/ | grep -v `^Only in`

But that removes them, rather than groups them. And sort doesn't do what I want either, because I want to keep the actual comparisons in their order.

Answer 1

One approach might be to redirect the diff output to a file and then use grep to segregate the two parts:

diff -r directory1/ directory2/ > temp
{ grep '^Only in' temp; grep -v '^Only in' temp; } > diff.output