Developer Productivity 2

Watch on Frontend Masters

Unix Tools

The unix set is super useful to know and we can do quite a bit of useful transformation with them.


















lets transform out! (the json file)

sed is a fun tool

  • lets replace foo with bar
  • lets replace foo or bar with foo,foo or bar,bar
  • show how its like vim find / replace
➜  dev-prod-2 git:(main) cat out | sed 's/: "\(.*\)",/: "\1,\1",/'

Expected

{"type": "foo,foo", "values": [1, 2, 3, 4, 5]}
{"type": "foo,foo", "values": [69, 420, 42, 69420]}
{"type": "bar,bar", "values": {"a": 42, "b": 69}}
{"type": "bar,bar", "values": {"a": 1337, "b": 420}}
{"type": "bar,bar", "values": {"a": 111, "b": 222}}

















find

list out files and do stuff

➜  dev-prod-2 git:(main) ✗ find . -maxdepth 1 -mindepth 1 -type f -exec grep -Hn "foo" {} \;
./out:1:{"type": "foo", "values": [1, 2, 3, 4, 5]}
./out:2:{"type": "foo", "values": [69, 420, 42, 69420]}

















xargs

a great tool

➜  dev-prod-2 git:(main) ✗ echo "1\n2\n3" | xargs -I {} curl https://{}.com
curl: (6) Could not resolve host: 1.com
curl: (6) Could not resolve host: 2.com
curl: (6) Could not resolve host: 3.com

















parallel

such a great tool

lets write a file with the entries of 1 through 9, per line

1
2
3
4
5
6
7
8
9

now lets use parallel to control how fast it requests

➜  dev-prod-2 git:(main) ✗ cat count | parallel -j 5 "curl https://{}.com"

















Just a touch of awk

I don't know a lot of awk and you can do quite a bit, but the little i do know is really good

➜  dev-prod-2 git:(main) ✗ ps aux | grep vim | awk '{ sum += $2 } END { print sum }'