html-export-workflow.yml

name: Exportera till HTML-format och ladda upp till Cloudflare R2

on:
  workflow_dispatch:  # Tillåter manuell körning
    inputs:
      source_ref:
        description: 'Git ref att bygga från'
        required: false
        default: 'main'
      filter:
        description: 'Filtrera filer efter år (YYYY) eller specifik beteckning (YYYY:NNN). Kommaseparerad lista.'
        required: false
        type: string
  workflow_call:  # Tillåter anrop från andra workflows
    inputs:
      source_ref:
        required: false
        type: string
        default: 'main'
      filter:
        required: false
        type: string

permissions:
  contents: read

jobs:
  html-export:
    runs-on: ubuntu-latest
    environment: Test
    
    steps:
    - name: Checkout repository
      uses: actions/checkout@v4
      with:
        ref: ${{ inputs.source_ref || 'main' }}

    - name: Set up Python
      uses: actions/setup-python@v4
      with:
        python-version: '3.11'

    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        pip install -r requirements.txt

    - name: Get JSON source files (from git or R2)
      run: |
        # Try to use JSON files from git first
        if [ -d "data/sfs_json" ] && [ -n "$(ls -A data/sfs_json 2>/dev/null)" ]; then
          echo "✅ Found $(find data/sfs_json -name '*.json' | wc -l) JSON files in git"
          echo "Using JSON files from git checkout"
        else
          echo "⚠️  No JSON files in git, downloading from Cloudflare R2..."

          # Configure AWS CLI for R2
          aws configure set aws_access_key_id ${{ secrets.CLOUDFLARE_R2_ACCESS_KEY_ID }}
          aws configure set aws_secret_access_key ${{ secrets.CLOUDFLARE_R2_SECRET_ACCESS_KEY }}
          aws configure set region us-east-1
          aws configure set output json

          # Download all JSON files from R2
          mkdir -p data/sfs_json
          aws s3 sync s3://${{ secrets.CLOUDFLARE_R2_BUCKET_NAME }}/sfs_json/ data/sfs_json/ \
            --endpoint-url https://${{ secrets.CLOUDFLARE_R2_ACCOUNT_ID }}.r2.cloudflarestorage.com \
            --exclude "*" \
            --include "*.json"

          # Verify download
          if [ ! -d "data/sfs_json" ] || [ -z "$(ls -A data/sfs_json)" ]; then
            echo "::error::Failed to download JSON files from R2"
            exit 1
          fi
          echo "✅ Downloaded $(find data/sfs_json -name '*.json' | wc -l) JSON files from R2"
        fi
      env:
        AWS_DEFAULT_REGION: us-east-1

    - name: Generate HTML export
      run: |
        if [ -n "${{ inputs.filter }}" ]; then
          python sfs_processor.py --input data/sfs_json --output output/html --formats html --filter "${{ inputs.filter }}"
        else
          python sfs_processor.py --input data/sfs_json --output output/html --formats html
        fi
      env:
        PYTHONPATH: ${{ github.workspace }}

    - name: Regenerate index pages for HTML export
      run: |
        python exporters/html/populate_index_pages.py --input data/sfs_json --output index.html --limit 30
        python exporters/html/populate_index_pages.py --input data/sfs_json --output latest.html --limit 10
      env:
        PYTHONPATH: ${{ github.workspace }}

    - name: Check required secrets
      run: |
        if [ -z "${{ secrets.CLOUDFLARE_R2_ACCESS_KEY_ID }}" ]; then
          echo "Error: CLOUDFLARE_R2_ACCESS_KEY_ID secret is not set"
          exit 1
        fi
        if [ -z "${{ secrets.CLOUDFLARE_R2_SECRET_ACCESS_KEY }}" ]; then
          echo "Error: CLOUDFLARE_R2_SECRET_ACCESS_KEY secret is not set"
          exit 1
        fi
        if [ -z "${{ secrets.CLOUDFLARE_R2_BUCKET_NAME }}" ]; then
          echo "Error: CLOUDFLARE_R2_BUCKET_NAME secret is not set"
          exit 1
        fi
        if [ -z "${{ secrets.CLOUDFLARE_R2_ACCOUNT_ID }}" ]; then
          echo "Error: CLOUDFLARE_R2_ACCOUNT_ID secret is not set"
          exit 1
        fi
        echo "All required secrets are configured"

    - name: Configure AWS CLI for Cloudflare R2
      run: |
        aws configure set aws_access_key_id ${{ secrets.CLOUDFLARE_R2_ACCESS_KEY_ID }}
        aws configure set aws_secret_access_key ${{ secrets.CLOUDFLARE_R2_SECRET_ACCESS_KEY }}
        aws configure set region us-east-1
        aws configure set output json

    - name: Upload HTML folder to Cloudflare R2
      run: |
        aws s3 sync output/html/ s3://${{ secrets.CLOUDFLARE_R2_BUCKET_NAME }}/sfs/ \
          --endpoint-url https://${{ secrets.CLOUDFLARE_R2_ACCOUNT_ID }}.r2.cloudflarestorage.com \
          --delete \
          --cache-control "public, max-age=3600" \
          --content-type "text/html" \
          --exclude "*.md" \
          --include "*.html"
      env:
        AWS_DEFAULT_REGION: us-east-1

    - name: Upload index pages to Cloudflare R2
      run: |
        aws s3 cp index.html s3://${{ secrets.CLOUDFLARE_R2_BUCKET_NAME }}/index.html \
          --endpoint-url https://${{ secrets.CLOUDFLARE_R2_ACCOUNT_ID }}.r2.cloudflarestorage.com \
          --cache-control "public, max-age=1800" \
          --content-type "text/html"
        aws s3 cp latest.html s3://${{ secrets.CLOUDFLARE_R2_BUCKET_NAME }}/latest.html \
          --endpoint-url https://${{ secrets.CLOUDFLARE_R2_ACCOUNT_ID }}.r2.cloudflarestorage.com \
          --cache-control "public, max-age=1800" \
          --content-type "text/html"
      env:
        AWS_DEFAULT_REGION: us-east-1

    - name: Upload summary
      run: |
        echo "HTML export completed at $(date)" > upload-summary.txt
        echo "Files uploaded to Cloudflare R2 bucket: ${{ secrets.CLOUDFLARE_R2_BUCKET_NAME }}/sfs/" >> upload-summary.txt
        echo "Index pages uploaded: index.html (30 senaste), latest.html (10 senaste)" >> upload-summary.txt
        aws s3 cp upload-summary.txt s3://${{ secrets.CLOUDFLARE_R2_BUCKET_NAME }}/sfs/last-update.txt \
          --endpoint-url https://${{ secrets.CLOUDFLARE_R2_ACCOUNT_ID }}.r2.cloudflarestorage.com