vcs-scraper-ci.yaml

name: VCS Scraper CI

on:
  push:
    branches: [ "main" ]
    paths:
      - .github/workflows/vcs-scraper-ci.yaml
      - components/resc-vcs-scraper/**
      - '!**.md'
      - '!**.png'
      - '!**.gif'
      - .gitignore
  pull_request:
    branches: [ "main" ]
    paths:
      - .github/workflows/vcs-scraper-ci.yaml
      - components/resc-vcs-scraper/**
      - '!**.md'
      - '!**.png'
      - '!**.gif'
      - .gitignore
  workflow_dispatch:

env:
  RESC_VCS_SCRAPER_DIR: components/resc-vcs-scraper
  CONTAINER_REGISTRY: rescabnamro
  IMAGE_NAME: resc-vcs-scraper

permissions: read-all

jobs:
  python-basic-validation:
    name: Python Basic Validation
    runs-on: ubuntu-latest

    strategy:
      matrix:
        python-version: ["3.9"]

    steps:
    - name: Checkout repository
      uses: actions/checkout@v3

    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v4
      with:
        python-version: ${{ matrix.python-version }}

    - name: Install test dependencies
      run: |
        cd ${{ env.RESC_VCS_SCRAPER_DIR }}
        pip install -r test-requirements.txt

    - name: Test with pytest
      run: |
        cd ${{ env.RESC_VCS_SCRAPER_DIR }}
        tox -e pytest

    - name: Import sorting with isort
      run: |
        cd ${{ env.RESC_VCS_SCRAPER_DIR }}
        tox -e sort
    
    - name: Lint with flake8
      run: |
        cd ${{ env.RESC_VCS_SCRAPER_DIR }}
        tox -e lint

    # This step will only execute if PR is created internally.
    - name: SonarCloud Scan
      if: ${{ (github.event.pull_request.head.repo.full_name == 'abnamro/repository-scanner') || (github.ref == 'refs/heads/main') }}
      uses: SonarSource/sonarcloud-github-action@master
      with:
        projectBaseDir: ${{ env.RESC_VCS_SCRAPER_DIR }}
        args: >
          -Dsonar.organization=abnamro-resc
          -Dsonar.projectKey=abnamro-resc_resc-vcs-scraper
          -Dsonar.python.coverage.reportPaths=coverage.xml
          -Dsonar.sourceEncoding=UTF-8
          -Dsonar.projectName=resc-vcs-scraper
          -Dsonar.groupid=resc
          -Dsonar.sources=src/
          -Dsonar.inclusions=**/*.py
          -Dsonar.exclusions=**/*yml,**/*.xml,**/*.txt,**/*.html,**/*.js
          -Dsonar.tests=tests/
          -Dsonar.cpd.exclusions=**/*yml,**/*.xml,**/*.txt,**/*.html,**/*.js
          -Dsonar.coverage.exclusions=**/*yml,**/*.xml,**/*.txt,**/*.html,**/*.js
          -Dsonar.pdf.skip=true
      env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}  # Needed to get PR information, if any
          SONAR_TOKEN: ${{ secrets.__SONAR_TOKEN_BACKEND__ }}

    - name: Get Branch Name
      id: extract_branch
      run: |
        if [[ ${GITHUB_EVENT_NAME} == 'pull_request' ]]; then
            export BRANCH_NAME="${GITHUB_HEAD_REF}"
            echo "::set-output name=branch_name::${BRANCH_NAME}"
        elif [[ ${GITHUB_EVENT_NAME} == 'push' ]]; then
            export BRANCH_NAME="${GITHUB_REF_NAME}"
            echo "::set-output name=branch_name::${BRANCH_NAME}"
        else
            echo "Event is neither pull_request nor push"
        fi
    outputs:
      branch_name: ${{ steps.extract_branch.outputs.branch_name }}
        
    
  python-build-and-publish:
    name: Python Build and Publish
    needs: python-basic-validation
    runs-on: ubuntu-latest

    strategy:
      matrix:
        python-version: ["3.9"]
    outputs:
      scraper_version: ${{ steps.getversion.outputs.scraper_version }}
    steps:
      - name: Checkout repository
        uses: actions/checkout@v3

      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}

      - name: Building python package
        run: |
          cd ${{ env.RESC_VCS_SCRAPER_DIR }}
          python -m pip install --upgrade pip build
          python -m build

      - name: Publish python package to PyPI
        if: ${{ needs.python-basic-validation.outputs.branch_name == 'main' }}
        uses: pypa/gh-action-pypi-publish@release/v1
        with:
          password: ${{ secrets.__PYPI_TOKEN__ }}
          packages_dir: ${{ env.RESC_VCS_SCRAPER_DIR }}/dist
          skip_existing: true

      - id: getversion
        name: Get package version
        run: | 
          cd ${{ env.RESC_VCS_SCRAPER_DIR }}
          scraper_version=$(python ./setup.py --version)
          echo "scraper_version=$scraper_version" >> $GITHUB_OUTPUT

  dockerize:
    name: Build and Push Docker image
    needs: [python-basic-validation, python-build-and-publish]
    runs-on: ubuntu-latest

    steps:
      - name: Checkout Repository
        uses: actions/checkout@v3

      - name: Docker Lint
        uses: hadolint/hadolint-action@v2.0.0
        with:
          dockerfile: ${{ env.RESC_VCS_SCRAPER_DIR }}/Dockerfile
          failure-threshold: error

      - name: Build an image from Dockerfile
        run: |
          cd ${{ env.RESC_VCS_SCRAPER_DIR }}
          docker build -t ${{ env.CONTAINER_REGISTRY }}/${{ env.IMAGE_NAME }}:${{needs.python-build-and-publish.outputs.scraper_version}} .

      - name: Run Trivy vulnerability scanner
        uses: aquasecurity/trivy-action@master
        with:
          image-ref: ${{ env.CONTAINER_REGISTRY }}/${{ env.IMAGE_NAME }}:${{needs.python-build-and-publish.outputs.scraper_version}}
          format: 'table'
          exit-code: '0'
          ignore-unfixed: true
          vuln-type: 'os,library'
          severity: 'CRITICAL,HIGH'

      - name: Get Branch Name
        run: |
          if [[ ${GITHUB_EVENT_NAME} == 'pull_request' ]]; then
              echo "BRANCH_NAME=$(echo ${GITHUB_HEAD_REF})" >> "$GITHUB_ENV"
          elif [[ ${GITHUB_EVENT_NAME} == 'push' ]]; then
              echo "BRANCH_NAME=$(echo ${GITHUB_REF_NAME})" >> "$GITHUB_ENV"
          else
              echo "Event is neither pull_request nor push"
          fi

      - name: Determine if image needs to be published
        run: |
          if [[ ${{ needs.python-basic-validation.outputs.branch_name }} == 'main' ]]; then
              echo "PUBLISH_IMAGE=true" >> "$GITHUB_ENV"
              echo "Going to publish image to registry"
          else
              echo "PUBLISH_IMAGE=false" >> "$GITHUB_ENV"
              echo "Skipping publishing of image to registry"
          fi

      - name: Log in to Container Registry
        if: ${{ env.PUBLISH_IMAGE == 'true' }}
        uses: docker/login-action@f054a8b539a109f9f41c372932f1ae047eff08c9
        with:
          username: ${{ secrets.__DOCKER_HUB_USER__ }}
          password: ${{ secrets.__DOCKER_HUB_PASS__ }}
      
      - name: Build and push Docker image
        uses: docker/build-push-action@ad44023a93711e3deb337508980b4b5e9bcdc5dc
        with:
          context: ${{ env.RESC_VCS_SCRAPER_DIR }}/
          push: ${{ env.PUBLISH_IMAGE }}
          tags: ${{ env.CONTAINER_REGISTRY }}/${{ env.IMAGE_NAME }}:latest, ${{ env.CONTAINER_REGISTRY }}/${{ env.IMAGE_NAME }}:${{needs.python-build-and-publish.outputs.scraper_version}}