Loop command, piping previous command's output - linux

Question

I understand that you can use loops in bash to repeat a command a certain amount of times, although it is more conveniently used within bash scripts.
For example, say I have a file that has been compressed several times, and I wish to fully decompress it.
cat file.txt | gzip -d | gzip -d | gzip -d
This is practical enough for a file that has been compressed 3 times, but it would become unwieldy if the file was compressed, for example, 18 times. How could this be simplified? I want to run the gzip -d command on the previous command's output n times. Is there a way to execute this from the command line?

Answer 1

You could do it with something along those lines (pardon any syntax error, consider this pseudo-code close to bash syntax) :
#!/bin/bash
# $1 = iterations left
# $2 = final output file
recursive_gzip()
{
if
[[ "$1" -gt 0 ]]
then
gzip -d | recursive_gzip $(($1 - 1)) "$2"
else
gzip -d > "$2"
fi
}
recursive_gzip 18 "file.txt" <"file.txt.gz"
Please note I replaced your cat with a redirection.
You could generalize the idea to share the same function for compress / decompress, and actually make it work for an arbitrary command by using it as positional arguments after 2.
#!/bin/bash
# $1 = iterations left
# $2 = final output file
recursive_pipe()
{
if
[[ "$1" -gt 0 ]]
then
"${#:3}" | recursive_pipe $(($1 - 1)) "$2" "${#:3}"
else
"${#:3}" > "$2"
fi
}
# Create gzipped file
recursive_pipe 18 "file.txt.gz" gzip <"file.txt"
# Uncompress gzipped file
recursive_pipe 18 "file.txt" gzip -d <"file.txt.gz"

Answer 2

What if you can't remember or wasn't told how often the file was compressed?
You would like to gunzip until it is a normal file.
When you do not want to overwrite your original file, making a copy is the easiest thing:
tmpfile=/tmp/mightbegz
cp file.txt ${tmpfile};
while [ $? -eq 0 ]; do
echo "Gunzipping Again"
mv ${tmpfile} ${tmpfile}.gz
gunzip -d ${tmpfile}.gz
done
The while loop will stop when gzip can not gunzip the file.