In bash, I make use of the autocomplete with TAB feature a lot, and sometimes I have filenames starting with special characters that I can't type with my default keyboard layout.

My question is, how can I find a filename starting with a character which is almost impossible to type with an English keyboard? For example, if the filename starts with a Japanese character, like ファイル.txt that might be difficult to type with an English keyboard. Sure, I can find the filename by its extension but what if it does not contain any Latin character at all?

This becomes even more difficult without a graphical environment in which I can easily select the filename and copy from it.

If isn't possible to get a solution similar to the TAB autocomplete, how could I find this kind of filenames easily?

I already read this question but has no accepted answer, also I don't know Japanese to input kanjis or similar.

1 Answer 1


You might be able to use the builtin bindings for the readline completion functions seen with:

$ bind -p | grep glob
"\eg":   glob-complete-word
"\C-x*": glob-expand-word
"\C-xg": glob-list-expansions

You can then type some globbing characters (* and ?) and have them expanded into the line with Ctrl-x*, or just listed with the other commands.

  • How would this help? Wouldn't the globs match all files in the directory? "*.txt" would match too much, and there are no Latin characters to be more specific
    – oligofren
    Apr 25, 2018 at 5:23
  • You end up with the filenames in the input line and you can delete the ones you dont want.
    – meuh
    Apr 25, 2018 at 7:23
  • Thank you, this is the closest thing to what I was looking for, although for directories with a lot of files, this might be not as useful. For that matter, I could resort to a script. May 11, 2018 at 9:50

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .