search all files and only empty directories using find command

Firerat · 06-13-2013, 02:37 PM

Quote:

Originally Posted by cli

I understood it now.

Code:

# ls -l this?is?an?example
-rw-r--r-- 1 root root 0 Jun 14 00:53 this is an example
-rw-r--r-- 1 root root 0 Jun 14 00:54 this_is_an_example
-rw-r--r-- 1 root root 0 Jun 14 00:54 thismismanmexample

Thanks a lot. And waiting for your kind reply form BASH Guru & Firerat for tar.bz2 format.

oh sorry, forgot

Code:

tar --help

Code:

 Compression options:

  -a, --auto-compress        use archive suffix to determine the compression
                             program
  -I, --use-compress-program=PROG
                             filter through PROG (must accept -d)
  -j, --bzip2                filter the archive through bzip2
  -J, --xz                   filter the archive through xz
      --lzip                 filter the archive through lzip
      --lzma                 filter the archive through xz
      --lzop
      --no-auto-compress     do not use archive suffix to determine the
                             compression program
  -z, --gzip, --gunzip, --ungzip   filter the archive through gzip
  -Z, --compress, --uncompress   filter the archive through compress

so

Code:

tar jc . -f Archive.tar.bz2
tar Jc . -f Archive.tar.xz
...

Edit:
never seen this before

Code:

tar ac . -f Archive.tar.bz2

don't know if it works

David the H. · 06-14-2013, 01:13 PM

I really dislike the idea of using tools like sed for processing filenames in bulk. It's a sloppy and unsafe practice.

If you want to use the output of find in another program, it's better to load the filenames into an array first. This will allow you to safely work with both the list as a whole and the individual entries in it.

Loading the array should be done with null separators for maximum safety (-print0 instead of -print).

How can I use array variables?
http://mywiki.wooledge.org/BashFAQ/005/

How can I find and deal with file names containing newlines, spaces or both?
http://mywiki.wooledge.org/BashFAQ/020

Code:

unset array                        #make sure the array starts empty, if necessary

while IFS='' read -r -d '' fname; do
    array+=( "$fname" )
done < <( find . -type f -print0 )   #simplified find example

Using them afterwards generally just requires using the "@" index expansion, which produces the entire list as a series of separate items. Be sure to quote it! That will ensure that any spaces and reserved characters are escaped.

It's also possible to modify the contents of every entry with any of the standard parameter substitution forms at the same time, although I don't think it's needed in this case.

Code:

tar -cpjvf "${array[@]}"

The main possible drawback with arrays is that there's a possibility for the expanded command to be longer than your system's ARG_MAX, the maximum length a command can be. In which case you'd have to break it up into multiple commands with a loop or xargs.

I'm getting "Argument list too long". How can I process a large list in chunks?
http://mywiki.wooledge.org/BashFAQ/095

David the H. · 06-14-2013, 01:24 PM

Also:

Code:

sed s/\ /?/g

The pattern on the left hand side is considered a regular expression, so characters like * and ? are special there. But the right hand side treats everything as simple text, except for a few special backslash-escape patterns. So the above command will simply replace spaces with question marks.

However, you should always* enclose your grep/sed/awk/whatever expressions in single quotes, otherwise they are subject to parsing by the shell before execution.

Code:

sed 's/ /?/g'

*Unless you know what you are doing and have a special need for it, such as including variables in the expression.

Firerat · 06-14-2013, 01:58 PM

Quote:

Originally Posted by David the H.

Also:

Code:

sed s/\ /?/g

<snip>

However, you should always* enclose your grep/sed/awk/whatever expressions in single quotes, otherwise they are subject to parsing by the shell before execution.

Code:

sed 's/ /?/g'

*Unless you know what you are doing and have a special need for it, such as including variables in the expression.

Very true, I get lazy with the short stuff sometimes
'' can be used if you want to include variables
example

Code:

Foo=Foo
Bar=Bar
echo $Foo | sed 's/$Foo/$Bar/'
echo $Foo | sed 's/'$Foo'/'$Bar'/'

David the H. · 06-14-2013, 02:59 PM

Code:

echo $Foo | sed 's/'$Foo'/'$Bar'/'

Except that this leaves the variables completely unprotected, so word-splitting and glob expansion will still be done on their contents.

The usual rule of thumb is to simply enclose the longest string possible with the necessary quotes. And remember that you can concatenate multiple quoted strings together too. Strings that don't have shell reserved characters in them don't have to be quoted at all.

So any of these would be better:

Code:

echo "$Foo" | sed "s/$Foo/$Bar/"
echo "$Foo" | sed 's/'"$Foo"'/'"$Bar"'/'
echo "$Foo" | sed s/"$Foo"/"$Bar"/

The first one is the cleanest and easiest, obviously.

Actually, though, once a string is stored in a variable, you can usually just use parameter substitution, which is much faster.

Code:

echo "${Foo/$Bar}"