Shell script for files matching a pattern and count occurence per directory

Question

I have the following problem: I want to count migration files per directory.

My structure always looks like this:

/some/path/app1/migrations/0001_hello_world.py /some/path/app2/migrations/0001_foo.py /some/path/app2/migrations/0002_bar.py

What I need is a shell script, in the best case a one-liner to get the following result:

app1: 1 app2: 2

So I get the numbers of files per app directory.

I got as far as getting my list of files:

git diff --color --name-only --diff-filter=A origin/develop my-branch | grep "\/migrations\/[0-9]"

I can count all of those matches with wc -l but I cannot count per directory.

Unfortunately even with google and stack, I'm stuck.

Any ideas on that topic?

pLumo · Accepted Answer · 2019-06-13 15:10:45Z

Try,

git diff --color --name-only --diff-filter=A origin/develop my-branch \ | grep '/migrations/[0-9]' \ | cut -d/ -f4 \ | sort \ | uniq -c

adapt the field number of cut (here: 4) as necessary.

Output:

 2 app1 1 app2

slm · Accepted Answer · 2019-06-13 14:53:19Z

-1

This will work:

$ for folder in *; do echo "$folder: "; \ find ./$folder -type f | wc -l; done | grep -A 1 "some/path/"`

slm♦

380k127 gold badges793 silver badges897 bronze badges

answered Jun 13, 2019 at 14:34

4371 gold badge8 silver badges22 bronze badges

2 Answers 2