monthly-update.yml

name: n8n nodes catalog monthly update

on:
  schedule:
    # 1st of each month, 09:00 UTC
    - cron: "0 9 1 * *"
  workflow_dispatch:
    inputs:
      tag:
        description: "n8n release tag (leave blank for latest)"
        required: false
        default: ""

jobs:
  update:
    runs-on: ubuntu-latest
    permissions:
      contents: write

    steps:
      - name: Checkout repo
        uses: actions/checkout@v6

      - name: Set up Python
        uses: actions/setup-python@v6
        with:
          python-version: "3.11"

      - name: Install dependencies
        run: pip install requests pyarrow huggingface_hub

      - name: Run extraction
        id: extract
        env:
          N8N_TAG: ${{ github.event.inputs.tag }}
        run: |
          if [ -n "$N8N_TAG" ]; then
            python extract.py --tag "$N8N_TAG"
          else
            python extract.py
          fi

      - name: Check for data changes
        id: diff
        run: |
          NEW_HASH=$(sha256sum nodes.json | cut -d' ' -f1)
          echo "new_hash=$NEW_HASH" >> "$GITHUB_OUTPUT"

          if python - <<'PY'
          import sys
          try:
              from huggingface_hub import hf_hub_download
              import os, hashlib
              path = hf_hub_download(
                  repo_id="automatelab/n8n-nodes-catalog",
                  filename="nodes.json",
                  repo_type="dataset",
                  token=os.environ.get("HF_TOKEN"),
                  local_dir="/tmp/hf_prev",
              )
              h = hashlib.sha256(open(path, "rb").read()).hexdigest()
              print(h)
              sys.exit(0)
          except Exception as e:
              print(f"download failed: {e}", file=sys.stderr)
              sys.exit(1)
          PY
          then
            PREV_HASH=$(python - <<'PY'
          import sys
          try:
              from huggingface_hub import hf_hub_download
              import os, hashlib
              path = hf_hub_download(
                  repo_id="automatelab/n8n-nodes-catalog",
                  filename="nodes.json",
                  repo_type="dataset",
                  token=os.environ.get("HF_TOKEN"),
                  local_dir="/tmp/hf_prev2",
              )
              print(hashlib.sha256(open(path,"rb").read()).hexdigest())
          except:
              print("none")
          PY
            )
            echo "prev_hash=$PREV_HASH" >> "$GITHUB_OUTPUT"
            if [ "$NEW_HASH" = "$PREV_HASH" ]; then
              echo "changed=false" >> "$GITHUB_OUTPUT"
            else
              echo "changed=true" >> "$GITHUB_OUTPUT"
            fi
          else
            echo "changed=true" >> "$GITHUB_OUTPUT"
            echo "prev_hash=none" >> "$GITHUB_OUTPUT"
          fi
        env:
          HF_TOKEN: ${{ secrets.HF_TOKEN }}

      - name: Update dataset card Last Updated timestamp
        if: steps.diff.outputs.changed == 'true'
        run: |
          MONTH=$(date -u +%Y-%m)
          sed -i "s/Last updated: [0-9][0-9][0-9][0-9]-[0-9][0-9]\./Last updated: $MONTH./" dataset-card.md
          echo "Updated dataset-card.md timestamp to $MONTH"

      - name: Upload to HuggingFace
        if: steps.diff.outputs.changed == 'true'
        env:
          HF_TOKEN: ${{ secrets.HF_TOKEN }}
        run: |
          python - <<'PY'
          import os
          from huggingface_hub import HfApi

          api = HfApi(token=os.environ["HF_TOKEN"])
          repo_id = "automatelab/n8n-nodes-catalog"

          uploads = [
              ("nodes.json", "nodes.json"),
              ("nodes.parquet", "nodes.parquet"),
              ("dataset-card.md", "README.md"),
          ]
          for src, dst in uploads:
              api.upload_file(
                  path_or_fileobj=src,
                  path_in_repo=dst,
                  repo_id=repo_id,
                  repo_type="dataset",
                  commit_message=f"monthly update: {dst}",
              )
              print(f"uploaded: {src} -> {dst}")
          print("done")
          PY

      - name: No changes — skip upload
        if: steps.diff.outputs.changed == 'false'
        run: |
          echo "nodes.json hash unchanged vs HuggingFace — skipping upload."
          echo "prev=${{ steps.diff.outputs.prev_hash }}"
          echo "new=${{ steps.diff.outputs.new_hash }}"

      - name: Send failure email
        if: failure()
        uses: dawidd6/action-send-mail@v3
        with:
          server_address: ${{ secrets.SMTP_HOST }}
          server_port: ${{ secrets.SMTP_PORT }}
          username: ${{ secrets.SMTP_USER }}
          password: ${{ secrets.SMTP_PASS }}
          subject: "[n8n-nodes-catalog] monthly update FAILED — ${{ github.run_id }}"
          to: ${{ secrets.SMTP_USER }}
          from: ${{ secrets.SMTP_USER }}
          body: |
            The monthly n8n nodes catalog update workflow failed.

            Workflow run: https://github.com/${{ github.repository }}/actions/runs/${{ github.run_id }}
            Triggered by: ${{ github.event_name }}
            Ref: ${{ github.ref }}

            Check the run log above for the exact error. Common causes:
            - HF_TOKEN secret missing or expired
            - extraction script error (upstream n8n repo changed layout)
            - HuggingFace API error

            Fix and re-trigger manually via workflow_dispatch or wait for next month.