I have a folder with 28 gz files with the extension .gz and 28 files with the extension .gz.bam.
I would like to unzip all the 28 .gz files and send them to another folder. I was doing one by one as follows:
gunzip -c file1.gz > /mnt/s3/data_transfer/file1
How can I specify I want the .gz and not the .gz.bam?
With bash:
(
ret=0
shopt -s nullglob
for file in *.gz; do
gunzip < "$file" > /mnt/s3/data_transfer/"${file%.gz}" || ret="$?"
done
exit "$ret"
)
(also set the dotglob option if you also want to process the hidden .gz files, those whose name starts with a .).
In zsh:
(
ret=0
for file (*.gz(N.)) gunzip < $file > /mnt/s3/data_transfer/$file:r || ret=$?
exit $ret
)
(also adding the . glob qualifier to restrict the glob expansion to regular files; add the D qualifier to process hidden files).
Just iterate over them:
for file in *gz; do
gunzip -c -- "$file" > /mnt/s3/data_transfer/"${file%.gz}"
done
The "${file%.gz}" is shell syntax that will return the value of $file with the final .gz removed.
Note that this quick-n-dirty approach is less robust than Stéphane's below, and if you run it in a directory with no gz files, it will try to gunzip the *gz pattern itself. This won't be a problem as long as you do actually have file names ending in gz. Also, unlike his, this one will not exit with a fail status if one of the files fails to be decompressed correctly. If these are problems for you, please use his approach instead.
