Wsrtjtyk

I'm trying to remove a lot of files at once but need to be specific as to not remove any of the files I actually need.

I have a ton of corrupt files that start master- but there are valid files that start with master-2018

So, I want to do something like

rm -rf master-* --exclude master-2018*

Is that I need possible?

Yes you can use more than one pattern with find:

$ find -name 'master-*' ! -name 'master-2018*' -print0 -prune |

     xargs -0 echo rm -fr

(remove the echo if you're satisfied with the dry run)

You should add a -maxdepth 1 predicate just after find if you only want ro remove files from the current directory, ie master-1991 but no subdir/master-1991.

In bash:

shopt -s extglob

echo rm master-!(2018*)

Remove the echo if it looks correct.

The above uses bash's extended globbing facility to match files that start with master- but who do not have 2018 immediately following, then followed by anything (*).

If all files you want to delete has pattern like master-YYYY*, you can use those patterns:

rm -rf master-???[0-79]*

rm -rf master-??[02-9]*

rm -rf master-?[1-9]*

rm -rf master-[013-9]*

The goal is to ommit digits from a year number, so on first place after master- we need to ommit (don't remove) digit 2, on 2nd place digit 0, on 3rd place digit 1, and on 4th place - digit 8
I tried it just a minute before, and is enough to run only first two of them.

Second method:
You can move master-2018 to another dir, e.g. /tmp,
then remove everything with master-*, and move back your master-2018 from tmp.

mkdir /tmp/backup

mv -r master-2018* /tmp/backup

rm -rf master-*

mv -r /tmp/backup/* .