One of the under appreciated toils of changing jobs (and laptops for that matter) is the setup from scratch of my various projects. Like in DevOps, this repeated exercise every few years results in a certain discipline in and appreciation of both documentation and automation. It becomes basic hygiene like release deployment in DevOps: ‘if it hurts you are not doing it often enough’.

Given the high number of repos in one of my Microservices projects, I quickly got tiered of copying and pasting the URL to the terminal to clone them to my laptop. So I wrote a small bash script for reading the repo names from a file and cloning them in turn.

#!/bin/bash
filename=$1
n=1
while read line; do
# reading each line
git clone git@bitbucket.org:<your_org_here>/`echo $line`.git
n=$((n+1))
done < $filename

The corresponding input file is simple enough

acd-api
acd-api-backend
acd-api-queues
acd-apis-store
acd-api-topics
acd-app-api
acd-app-backend
...

And no, I did not crawl the web for some git voodoo magic command which would could do it for me - sometimes bash is good enough for me. Writing bash scripts is a fun ride down the memory lane to my humble beginnings as sysadmin in the IBM Nordic Labs in Stockholm and makes me appreciate the genius of the now ancient UNIX design principles.

May the (scripting) power be strong in you.