Find package files stored in yarn offline mirror that are no longer referenced/resolved in yarn.lock file

clean_yarn_mirror.sh

# There is no way right now to find packages included in an offline mirror folder that are
# not actually referenced by the current yarn.lock file. This script will compare the files
# in your offline mirror folder and find any that are not in the current set resolved by yarn.
# The closest is https://yarnpkg.com/en/docs/prune-offline-mirror but this only effects
# specific future activities, like calling 'yarn upgrade'.

# example usage: cd some_repo; clean_yarn_mirror.sh ./npm_offline_cache

echo "Scanning files in $1..."
tgz=($(grep -ho -e 'resolved .*\.tgz' yarn.lock | cut -d" " -f2))
git=($(grep -ho -e 'resolved .*\.git-.*#' yarn.lock | cut -d" " -f2 | sed 's/#//'))
files=$(ls $1)

containsElement () {
  local e
  for e in "${@:2}"; do [[ "$e" == "$1" ]] && return 0; done
  return 1
}

for f in $files; do
    containsElement "$f" "${tgz[@]}"
    istgz=$?
    containsElement "$f" "${git[@]}"
    isgit=$?
    if [ $istgz -eq 0 ]
    then
        : # echo "$f is listed as a tgz"
    elif [ $isgit -eq 0 ]
    then
        : # echo "$f is listed as a git"
    else
        echo "MATCH: $f is not mentioned in yarn.lock file"
    fi
done