Differences between revisions 2 and 15 (spanning 13 versions)
Revision 2 as of 2007-05-03 00:51:48
Size: 2057
Editor: GreyCat
Comment: last example doesn't need the directory printed twice
Revision 15 as of 2008-10-11 21:57:53
Size: 2231
Editor: gw
Comment: backticks to $()
Deletions are marked like this. Additions are marked like this.
Line 2: Line 2:
== How can I check whether a directory is empty or not? ==
 ''I just deleted three completely '''wrong''' answers from this question. Please, people, make sure that when you add to the FAQ, your answers''
  * answer the question that was asked, and
  * actually '''work'''
 ''Thanks. -- GreyCat''
== How can I check whether a directory is empty or not? How do I check for any *.mpg files? ==
Line 8: Line 4:
Most modern systems have an "ls -A" which explicitly omits "." and ".." from the directory listing: In Bash, you can do this safely and easily with the nullglob and dotglob options (which change the behaviour of [:glob:globbing]), and arrays:
Line 11: Line 7:
    if [ -n "$(ls -A somedir)" ]
    then
        echo directory is non-empty
    fi
    # Bash
    shopt -s nullglob dotglob
    files=(*)
    (( ${#files[*]} )) || echo directory is empty
    shopt -u nullglob dotglob
Line 17: Line 14:
This can be shortened to: As you can see we unset the nullglob after using it, to prevent it affecting other globs in the script in unexpected ways. `nullglob` also simplifies various other operations:
Line 20: Line 17:
    if [ "$(ls -A somedir)" ]
    then
        echo directory is non-empty
    fi
}}}

Another way, using Bash features, involves setting the special shell option which changes the behavior of globbing. Some people prefer to avoid this approach, because it's so drastically different and could severely alter the behavior of scripts.

Nevertheless, if you're willing to use this approach, it does greatly simplify this particular task:

{{{
    shopt -s nullglob
    if [[ -z $(echo *) ]]; then
        echo directory is empty
    fi
}}}

It also simplifies various other operations:

{{{
    # Bash
Line 44: Line 22:
    shopt -u nullglob
Line 46: Line 25:
Without the {{{shopt}}}, that would have to be: Without the {{{nullglob}}}, that would have to be:
Line 49: Line 28:
    # Bash
Line 57: Line 37:
Finally, you may wish to avoid the ''direct'' question altogether. Usually people want to know whether a directory is empty... ''because'' they want to do something involving the files therein, etc. Look to the larger question. For example, something like this may be an appropriate solution: In fact, you may wish to avoid the ''direct'' question altogether. Usually people want to know whether a directory is empty... ''because'' they want to do something involving the files therein, etc. Look to the larger question. For example, one of these [:UsingFind:find-based examples] may be an appropriate solution:
Line 60: Line 40:
   # Bourne
Line 61: Line 42:
   find "$somedir" -maxdepth 0 -empty -exec echo {} is empty. \; # GNU/BSD
   find "$somedir" -type d -empty -exec cp /my/configfile {} \; # GNU/BSD
Line 63: Line 46:
It's all a matter of addressing the program's actual requirements. If your script needs to run with various shell implementations, you can try using an external program like python, perl, or find as indicated above, or you can try something like:

{{{
    # Bourne
    # (Of course, the system must have printf(1).)
    cd foo || exit 1
    if test "$(printf '%s %s %s' .* *)" = '. .. *' && ! test -f '*'
    then
        echo "directory is empty"
    fi
}}}

Yes, it's extremely ugly, but it should be more portable than anything depending on `ls` output. Even `ls -A` solutions can break (HPUX for one, if you are root).

Anchor(faq4)

How can I check whether a directory is empty or not? How do I check for any *.mpg files?

In Bash, you can do this safely and easily with the nullglob and dotglob options (which change the behaviour of [:glob:globbing]), and arrays:

    # Bash
    shopt -s nullglob dotglob
    files=(*)
    (( ${#files[*]} )) || echo directory is empty
    shopt -u nullglob dotglob

As you can see we unset the nullglob after using it, to prevent it affecting other globs in the script in unexpected ways. nullglob also simplifies various other operations:

    # Bash
    shopt -s nullglob
    for i in *.zip; do
        blah blah "$i"  # No need to check $i is a file.
    done
    shopt -u nullglob

Without the nullglob, that would have to be:

    # Bash
    for i in *.zip; do
        [[ -f $i ]] || continue  # If no .zip files, i becomes *.zip
        blah blah "$i"
    done

(You may want to use the latter anyway, if there's a possibility that the glob may match directories in addition to files.)

In fact, you may wish to avoid the direct question altogether. Usually people want to know whether a directory is empty... because they want to do something involving the files therein, etc. Look to the larger question. For example, one of these [:UsingFind:find-based examples] may be an appropriate solution:

   # Bourne
   find "$somedir" -type f -exec echo Found unexpected file {} \;
   find "$somedir" -maxdepth 0 -empty -exec echo {} is empty. \;  # GNU/BSD
   find "$somedir" -type d -empty -exec cp /my/configfile {} \;   # GNU/BSD

If your script needs to run with various shell implementations, you can try using an external program like python, perl, or find as indicated above, or you can try something like:

    # Bourne
    # (Of course, the system must have printf(1).)
    cd foo || exit 1
    if test "$(printf '%s %s %s' .* *)" = '. .. *' && ! test -f '*'
    then
        echo "directory is empty"
    fi

Yes, it's extremely ugly, but it should be more portable than anything depending on ls output. Even ls -A solutions can break (HPUX for one, if you are root).

BashFAQ/004 (last edited 2023-03-28 07:52:15 by emanuele6)