CODEWITHSUNDEEP
bashshell
Jon Abraham
Jon Abraham

Reputation: 955

How to iterate over directory and sub directories and get the name of parent directory?

I am iterating through folders and finding the folders that has specific file extension. Based on the find result, I am getting the directory path as -

e.g.

./Config/SysMapping.txt

./Config/exclusion_list_old.txt

I want to simply get the name:

Config

and remove

./

before

Config

and everything after Config, i.e.

/SysMapping.txt

Upvotes: 0

Views: 121

Answers (3)

mickp
mickp

Reputation: 1809

You can use the globstar setting:

shopt -s globstar nullglob
files=(**/*.txt)

# if there cannot be newlines in filenames
printf '%s\n' "${files[@]%/*}" | sort -u

# if there can be newlines in filenames
printf '%s\0' "${files[@]%/*}" | sort -zu | tr '\0' '\n'

The tr is there just to print the dirnames nicely at the end. You can choose to process the null delimited dirnames differently, with a while loop for example.

Upvotes: 0

jhnc
jhnc

Reputation: 16692

This will extract the first path element after ./ and delete duplicates:

echo ./Config/SysMapping.txt | awk -F/ '{print $2}' | sort -u

In bash, you can do:

P=./Config/SysMapping.txt
P=${P%/*}  # --> ./Config
P=${P#./}  # --> Config
echo "$P"

Upvotes: 3

klog
klog

Reputation: 486

dirname <path> would remove the file at the end of your path. In your example you'd end up with ./Config. You can then pipe this to awk '{ gsub(/^\.\//, ""); print }' to remove the ./.

Upvotes: 1

Related Questions