geonames-create-graph.yaml

name: Update Knowledge Graph with data from dataset 'GeoNames', based on IRIs in the Knowledge Graph

# Every 3 hours. Don't run it more often; otherwise you might hit
# GeoNames' hourly and daily request limits - http://www.geonames.org/export/
on:
  schedule:
    - cron: "0 */3 * * *"

# Run a single workflow at a time
concurrency:
  group: knowledge-graph

jobs:
  run:
    runs-on: ubuntu-latest
    timeout-minutes: 120
    steps:
      - name: Checkout data repository
        uses: actions/checkout@v4
        with:
          path: ./data
          sparse-checkout: geonames
      - name: Checkout code repository
        uses: actions/checkout@v4
        with:
          path: ./code
          repository: colonial-heritage/integration-layer
      - name: Install Node
        uses: actions/setup-node@v4
        with:
          node-version: "20"
      - name: Install app
        run: |
          cd ./code
          npm install --no-progress
          npx turbo run build --filter=@colonial-collections/graph-create-geonames
      - name: Create graph
        run: |
          cd ./data
          mkdir -p "$RUNNER_TEMP"
          test -f ./geonames/data/data.tar.zst && tar --zstd -xf ./geonames/data/data.tar.zst -C "$RUNNER_TEMP"
          ../code/apps/graph-create-geonames/dist/cli.js create \
            --resource-dir ./geonames/resources \
            --data-file "$RUNNER_TEMP/data.sqlite" \
            --iterate-endpoint-url "${{ vars.SPARQL_ENDPOINT_URL_KG }}" \
            --iterate-locations-query-file ./geonames/queries/iterate-locations.rq \
            --iterate-countries-query-file ./geonames/queries/iterate-countries.rq \
            --dereference-batch-size 500 \
            --triplydb-instance-url "${{ vars.TRIPLYDB_INSTANCE_URL }}" \
            --triplydb-api-token "${{ secrets.TRIPLYDB_API_TOKEN }}" \
            --triplydb-account "${{ vars.TRIPLYDB_ACCOUNT }}" \
            --triplydb-dataset "${{ vars.TRIPLYDB_DATASET_KG }}" \
            --triplydb-service "kg" \
            --graph-name "https://data.colonialcollections.nl/geonames" \
            --temp-dir "$RUNNER_TEMP"
      - name: Save changes
        run: |
          cd ./data
          mkdir -p ./geonames/data
          test -f "$RUNNER_TEMP/data.sqlite" && tar --zstd -cf ./geonames/data/data.tar.zst -C "$RUNNER_TEMP" data.sqlite
          git config --global user.name "github-actions[bot]"
          git config --global user.email "github-actions[bot]@users.noreply.github.com"
          git add .
          git commit --quiet -a -m "Save changes" || true
          git push --force -u origin