https://bugzilla.novell.com/show_bug.cgi?id=726939 https://bugzilla.novell.com/show_bug.cgi?id=726939#c3 --- Comment #3 from Harald Koenig <koenig@linux.de> 2011-10-31 16:41:28 UTC --- it's a bash completion issue (more preceise the definition of the completions). I just played around a bit with the zypper completions (systemctl later, have to reboot to run systemd again;). I noticed that with complete -o default -o bashdefault -F _zypper zypper or complete -o bashdefault -o default -F _zypper zypper things are *much* better for my typical usage pattern, but your example with a space character after ">" still shows zypper lr opts. without space after ">" filenames get completed, or if you type some letter after "> " file/dir name completion gets triggered too (likely only because all further possible completions for "zypper lr" shall start with "--", so there is no zypper-match for something like a<TAB> but with both -o default and -o bashdefault I get back all the expected behavior for file/dir name completions! order of those options does not seem to be important?! right now there are no "-o comp-option"s set for systemctl: $ complete -p systemctl complete -F _systemctl systemctl I guess for systemd complete -o default -o bashdefault -F _systemctl systemctl might be a much better start (at least for me?!) -- will try later... FYI: my test pattern are like this (in a directory with multiple files starting with au* and exactly one directory starting with au*)": command > au<TAB><TAB> command > au*<TAB><TAB> (note the * !) both shall show all files and the directory. command > au*/<TAB> shall directly expands to the single directory name. "shall" == that's the normal behavior in bash if there are no special completions defined for "command" -- and some additional completion hints shall not obscure or break those well known defaults (which still work fine for all other commands), I very much vote for the principle of least surprise! -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.