Without a .* except for the end could be done with multiples:

sed "s/^[^/]*//" /tmp/input | sed "s/^///" | sed 's/[_@].*//'

On 2016-11-29 10:11 AM, Paul Sierks wrote:

.* Is greedy,

sed "s/^.*/([^_@]*).*$/\1/"

Kevin McGregor kevin.a.mcgregor@gmail.com wrote:

I'm using sed to massage some input. Specifically, I have input lines like

aaaaaaaaaa/BBBBB_ccc@00000 or aaaaaaaaaa/BBBBB@00000

and I want the output to always be

BBBBB

I've got most of it, but I can't figure out how to get rid of anything at the end of the line after EITHER the underscore OR the '@' (including either of those two characters).

Is this possible in one expression in sed? I can do it with piping the input through sed twice but I was wondering if one pass would do it. Currently I'm using

sed "s/^.*/(.*)[_@].*$/\1/"

Which doesn't get rid of the "_ccc" when it appears, just the "@00000".

Suggestions?

---

Roundtable mailing list Roundtable@muug.ca https://muug.ca/mailman/listinfo/roundtable