cache-cleanup.yml

name: Cache Cleanup

on:
  schedule:
    # Run every Sunday at 2 AM UTC
    - cron: "0 2 * * 0"
  workflow_dispatch:
    inputs:
      max_age_days:
        description: "Maximum age of caches to keep (in days)"
        required: false
        default: "30"
        type: string

jobs:
  cleanup:
    name: Clean up old caches
    runs-on: ubuntu-latest
    permissions:
      actions: write
      contents: read

    steps:
      - name: Checkout code
        uses: actions/checkout@v4

      - name: Get cache cleanup parameters
        id: params
        run: |
          MAX_AGE_DAYS="${{ github.event.inputs.max_age_days || '30' }}"
          echo "max_age_days=${MAX_AGE_DAYS}" >> $GITHUB_OUTPUT
          echo "cleanup_date=$(date -d "${MAX_AGE_DAYS} days ago" '+%Y-%m-%dT%H:%M:%SZ')" >> $GITHUB_OUTPUT

      - name: List current caches
        run: |
          echo "Current caches in repository:"
          gh api \
            -H "Accept: application/vnd.github+json" \
            -H "X-GitHub-Api-Version: 2022-11-28" \
            "/repos/${{ github.repository }}/actions/caches" \
            --jq '.actions_caches[] | "\(.key) - \(.size_in_bytes) bytes - \(.created_at)"'
        env:
          GH_TOKEN: ${{ github.token }}

      - name: Clean up old caches
        run: |
          echo "Cleaning up caches older than ${{ steps.params.outputs.cleanup_date }}"

          # Get list of cache IDs older than specified date
          CLEANUP_DATE="${{ steps.params.outputs.cleanup_date }}"
          CACHE_IDS=$(gh api \
            -H "Accept: application/vnd.github+json" \
            -H "X-GitHub-Api-Version: 2022-11-28" \
            "/repos/${{ github.repository }}/actions/caches" \
            --jq --arg cleanup_date "$CLEANUP_DATE" '.actions_caches[] | select(.created_at < $cleanup_date) | .id')

          if [ -z "$CACHE_IDS" ]; then
            echo "No old caches found to clean up"
            exit 0
          fi

          echo "Found caches to delete:"
          echo "$CACHE_IDS"

          # Delete each cache
          for cache_id in $CACHE_IDS; do
            echo "Deleting cache ID: $cache_id"
            gh api \
              -X DELETE \
              -H "Accept: application/vnd.github+json" \
              -H "X-GitHub-Api-Version: 2022-11-28" \
              "/repos/${{ github.repository }}/actions/caches/$cache_id" || true
          done
        env:
          GH_TOKEN: ${{ github.token }}

      - name: Clean up caches from deleted branches
        run: |
          echo "Cleaning up caches from non-existent branches"

          # Get all remote branches
          git fetch --all
          EXISTING_BRANCHES=$(git branch -r | sed 's/origin\///' | grep -v HEAD | tr -d ' ')

          # Get all cache keys
          CACHE_KEYS=$(gh api \
            -H "Accept: application/vnd.github+json" \
            -H "X-GitHub-Api-Version: 2022-11-28" \
            "/repos/${{ github.repository }}/actions/caches" \
            --jq '.actions_caches[] | .key')

          # Find caches that might be from deleted branches
          for cache_key in $CACHE_KEYS; do
            # Extract potential branch name from cache key (this is heuristic)
            # Cache keys often contain branch names or refs
            if [[ "$cache_key" == *"refs/heads/"* ]]; then
              BRANCH_IN_KEY=$(echo "$cache_key" | sed -n 's/.*refs\/heads\/\([^-]*\).*/\1/p')
              if [ -n "$BRANCH_IN_KEY" ] && ! echo "$EXISTING_BRANCHES" | grep -q "^$BRANCH_IN_KEY$"; then
                echo "Found cache from potentially deleted branch: $cache_key (branch: $BRANCH_IN_KEY)"

                # Get cache ID and delete
                CACHE_ID=$(gh api \
                  -H "Accept: application/vnd.github+json" \
                  -H "X-GitHub-Api-Version: 2022-11-28" \
                  "/repos/${{ github.repository }}/actions/caches" \
                  --jq --arg key "$cache_key" \
                  '.actions_caches[] | select(.key == $key) | .id')

                if [ -n "$CACHE_ID" ]; then
                  echo "Deleting cache ID: $CACHE_ID"
                  gh api \
                    -X DELETE \
                    -H "Accept: application/vnd.github+json" \
                    -H "X-GitHub-Api-Version: 2022-11-28" \
                    "/repos/${{ github.repository }}/actions/caches/$CACHE_ID" || true
                fi
              fi
            fi
          done
        env:
          GH_TOKEN: ${{ github.token }}

      - name: Show cache statistics after cleanup
        run: |
          echo "Cache statistics after cleanup:"
          CACHE_INFO=$(gh api \
            -H "Accept: application/vnd.github+json" \
            -H "X-GitHub-Api-Version: 2022-11-28" \
            "/repos/${{ github.repository }}/actions/caches")

          CACHE_COUNT=$(echo "$CACHE_INFO" | jq '.actions_caches | length')
          TOTAL_SIZE=$(echo "$CACHE_INFO" | jq '.actions_caches | map(.size_in_bytes) | add // 0')
          TOTAL_SIZE_MB=$((TOTAL_SIZE / 1024 / 1024))

          echo "Total caches: $CACHE_COUNT"
          echo "Total size: ${TOTAL_SIZE_MB} MB"
          echo "Remaining caches:"
          echo "$CACHE_INFO" | jq '.actions_caches[] | "\(.key) - \(.size_in_bytes) bytes - \(.created_at)"'
        env:
          GH_TOKEN: ${{ github.token }}