I'm writing a Bash script and need to match all files and folders inside a directory which contain non-word-characters, uppercase-characters and German Umlauts. Inside my test directory I tried different combinations, but either I match any test file/folder or none despite the fact that it is working at Regex101 (example 1, example 2). I can't figure out if the problem are tokens/character classes i use or something Bash-specific:
$ ls
kukukae.txt KuKuKä.txt 'Replüce-?!% Test.txt.bak' repluece_test.txt.bak 'Sä=?)(&%$§"!°^`´+*~#' somo_oemo.gif 'Somo Ömo.gif' super_ 'why?that' # some example files and one folder
$ find . -maxdepth 1 -regex '[^\w\.\/]' # nothing matched
$ find . -maxdepth 1 -regex '.*[^\w\.\/].*' # matches also files without non-word-characters
./super_
./kukukae.txt
./Sä=?)(&%$§!°^`´+*~#
./somo_oemo.gif
./repluece_test.txt.bak
./why?that
./KuKuKä.txt
./Replüce-?!% Test.txt.bak
./Somo Ömo.gif
$ find . -maxdepth 1 -regex '[[:upper:]]|[[:space:]]|[\?\&\ä\ü\ö\!\"\§\$\\\(\)\[\]\}\{\=\`\´\+\*\~\#\:\;\,\^\°\<\>\|]' # again nothing
$ find . -maxdepth 1 -regex '.*[[:upper:]]|[[:space:]]|[\?\&\ä\ü\ö\!\"\§\$\\\(\)\[\]\}\{\=\`\´\+\*\~\#\:\;\,\^\°\<\>\|].*' # nothing one more time
It should look like the following, where only the files with non-word-characters and Umlauts are matched:
$ find [working command]
./Sä=?)(&%$§!°^`´+*~#
./why?that
./KuKuKä.txt
./Replüce-?!% Test.txt.bak
./Somo Ömo.gif
In both versions I don't want to match / or . intentionally ([^\w\.\/]), because it would include path and dots of file extensions.
Does any Regex-pro or Bash-veteran has an easy solution?
Thanks and best
I don't think you need a regex for this
$ LC_ALL=C find . -name '*[!0-9a-z_.]*'
./Somo Ömo.gif
./Sä=?)(&%$§"!°^`´+*~#
./KuKuKä.txt
./Replüce-?!% Test.txt.bak
./why?that
$