Count the total lines of code for a user or organization in GitHub (excludes forks)

cloc-gh

#!/usr/bin/env bash

# Author: C.J. May @lawndoc
# Usage: cloc-gh <username>
# Prereqs: cloc gh

cloc_repo () {
  gh repo clone "$1" temp-linecount-repo -- --depth 1 > /dev/null 2>&1 &&
  cloc temp-linecount-repo | grep SUM | awk '{ print $5 }' >> line_count.txt &&
  rm -rf temp-linecount-repo
}

progress_bar () {
  # Process data
  let _progress=(${1}*100/${2}*100)/100
  let _done=(${_progress}*4)/10
  let _left=40-$_done
  # Build progressbar string lengths
  _fill=$(printf "%${_done}s")
  _empty=$(printf "%${_left}s")
  printf "\r\033[KProgress : [${_fill// /\#}${_empty// /-}] ${_progress}%% -- "
}

echo "Enumerating $1 repos..."
# ignores forked repos
gh repo list $1 --limit 10000 --json nameWithOwner,isFork --jq '.[] | select( .isFork | not ) | .nameWithOwner' > repos.txt
let _total=$(awk 'END{print NR}' repos.txt)
let _current=1

echo "Counting lines of code..."
while read repo; do
  progress_bar $_current $_total
  printf "$repo"
  cloc_repo $repo
  let "_current+=1"
done < repos.txt
printf "\nDone.\n"

printf "Total lines of code in $1: "
awk '{ sum += $1 } END { print sum }' line_count.txt

rm repos.txt line_count.txt

To use this script, you'll need to install cloc and gh, and you'll also need to sign into gh with your account. It should then have access to all private repos and organizations you have access to.

Update the --json fields and --jq query to filter on different sets of repos.