I have a list of files (see below) I need to merge into separate files (merge1, merge2). merge1 would contain file1a_1 and file1a_2; merge2 would contain file2a_1, file2a_2.
I tried
find . -name "file*_*" -exec cat {} \; >> mergefile*
i.e. Files:
file1a_1
X1a_2
file1a_3
file1b_1
file1b_2
file1b_3
The shell script puts all the files into merge file and doesn't separate them out individually.
Any assistance would be appreciated
If the files are all in the current directory, you could hard-code a loop through the highest number that you might merge (100 in the example below):
shopt -s nullglob
for((index=1;index<100;index++))
do
inputs=( "file${index}"*_* )
if [ "${#inputs[@]}" -gt 0 ]
then
cat -- "${inputs[@]}" > "merge${index}"
fi
done
Maybe something like:
find . -name 'file[0-9]*_*' -type f -exec awk '
FNR == 1 {
output = FILENAME
sub(/.*\/file/, "", output)
sub(/[^0-9].*/, "", output)
output = "merge" output
}
{print > output}' {} +
Note that find finds files in an unspecified order, so files will be merged in an unspecified order.
Note that if all the files for a given merge file are empty, the merge file won't be created. If you'd rather get an empty merge file in those cases, and if your awk is the GNU ones, you can change the above to:
find . -name 'file[0-9]*_*' -type f -exec gawk '
BEGINFILE {
output = FILENAME
sub(/.*\/file/, "", output)
sub(/[^0-9].*/, "", output)
output = "merge" output
printf "" > output
}
{print > output}' {} +
debian 9.3
find . -name 'file[0-9]*_*' -type f -exec awk '{print FILENAME}' {} + why?!
find . -name 'file[0-9]*_*' -type f work and show file name ..
fish or rc, you'd need to quote {}.
bash in default! bayanbox.ir/view/4393624869319520231/…
file*_*files in subdirectories, or all in the current directory? I'm assuming theX1a_2is meant as an example of a file to omit?