PR-checks.yml

name: on PR
# This workflow is triggered on pull requests to the repository.
# It runs a job to check for broken links in the files changed in the PR against the online built webpage, and against the files on the PR branch.
# It checks for spelling errors in the files changed in the PR.
# It builds the static site and checks the size of the built site.

on:
  pull_request

jobs:
  build:
    name: 📦 Build
    runs-on: ubuntu-latest
    outputs:
      html_lychee_exited: ${{ steps.stringify.outputs.lychee_exited }}
      html_lychee_file: ${{ steps.stringify.outputs.lychee_file }}
    steps:
      - uses: actions/checkout@v3

      - uses: actions/setup-python@v4
        with:
          python-version: '3.x'
      - run: pip install nbconvert packaging
      - name: Convert Jupyter notebooks
        run: |
          shopt -s globstar
          jupyter-nbconvert --to markdown **/*.ipynb --ExtractOutputPreprocessor.enabled=False
          shopt -u globstar
        shell: bash

      - name: Use Node.js
        uses: actions/setup-node@v3
        with:
          node-version: 16

      - name: Install vuepress
        run: yarn global add vuepress
      - name: Install package.json
        run: yarn
      - name: Build static site with vuepress
        # alias for vuepress build src defined in package.json
        run: yarn build

      - name: Get size of files in dist
        shell: bash
        run: |
          du -ah src/.vuepress/dist

      - name: 'Get built html files'
        id: html-files
        run: |
          html_files=$(find src/.vuepress/dist -name '*.html')
          html_files="${html_files//$'\n'/ }"
          html_files="${html_files//$'\r'/ }"
          echo $html_files
          echo "all_html_files=${html_files}" >> $GITHUB_OUTPUT
      
      - name: 'Check Links'
        id: lychee
        uses: lycheeverse/lychee-action@v1.5.4
        env:
          GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}
        with:
          args: -v -n -a 403,503,429,200 -i -b 'https://docs.pupil-labs.com/' --exclude-mail --include-verbatim -- ${{ steps.html-files.outputs.all_html_files }}
          output: ./lychee/htmlfiles.md
          jobSummary: true
      
      - name: 'Outputs as strings'
        id: stringify
        run: |
          EXIT_CODE_STRING="$( printf '%d' "${{ env.lychee_exit_code }}" )"
          echo "lychee_exited=${EXIT_CODE_STRING}" >> $GITHUB_OUTPUT
          MD_OUT="$( cat ./lychee/htmlfiles.md )"
          MD_OUT="${MD_OUT//'%'/'%25'}"
          MD_OUT="${MD_OUT//$'\n'/'%0A'}"
          MD_OUT="${MD_OUT//$'\r'/'%0D'}" 
          echo "text<<EOF" >> $GITHUB_OUTPUT
          echo "$MD_OUT" >> $GITHUB_OUTPUT
          echo "EOF" >> $GITHUB_OUTPUT
          echo "lychee_file=${MD_OUT}" >> $GITHUB_OUTPUT
      
  check_links_on_pr_files_online:
    runs-on: ubuntu-latest
    name: 🌍 Test (.md/ipynb) changed-files for broken links
    outputs:
      lychee_exited: ${{ steps.stringify.outputs.lychee_exited }}
      lychee_file: ${{ steps.stringify.outputs.lychee_file }}
    steps:
      # Compares the default generated "test merge commit" by Github of the PR branch into the base branch
      - uses: actions/checkout@v3
        with:
          fetch-depth: 2

      # Gets a list of files from the PR with a git status of Added (A) or Modified (M) | (via outputs.all_changed_files)
      - name: 'Get a list of changed files to process(Only html/md/ipynb files)'
        id: get-changed-files
        uses: tj-actions/changed-files@v34
        with:
          files: |
            **/*.md
            **/*.ipynb
          separator: ','

      # lychee-action requires pre-processing for file paths with spaces (boundary quotes added in lychee-action input):
      - name: 'Quote wrap paths to support spaces'
        id: changed-files
        run: |
          QUOTE_WRAPPED="$( sed "s/,/' '/g" <<< '${{ steps.get-changed-files.outputs.all_changed_files }}' )"
          echo "all_changed_files=${QUOTE_WRAPPED}" >> $GITHUB_OUTPUT

      - name: 'Check Links'
        id: lychee
        uses: lycheeverse/lychee-action@v1.5.4
        env:
          GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}
        with:
          args: -v -n -a 403,503,429,200 -i -b 'https://docs.pupil-labs.com/' --exclude-mail --include-verbatim -- '${{ steps.changed-files.outputs.all_changed_files }}'
          output: ./lychee/results.md
          jobSummary: true
      
      - name: 'Outputs as strings'
        id: stringify
        run: |
          EXIT_CODE_STRING="$( printf '%d' "${{ env.lychee_exit_code }}" )"
          echo "lychee_exited=${EXIT_CODE_STRING}" >> $GITHUB_OUTPUT
          MD_OUT="$( cat ./lychee/results.md )"
          MD_OUT="${MD_OUT//'%'/'%25'}"
          MD_OUT="${MD_OUT//$'\n'/'%0A'}"
          MD_OUT="${MD_OUT//$'\r'/'%0D'}" 
          echo "text<<EOF" >> $GITHUB_OUTPUT
          echo "$MD_OUT" >> $GITHUB_OUTPUT
          echo "EOF" >> $GITHUB_OUTPUT
          echo "lychee_file=${MD_OUT}" >> $GITHUB_OUTPUT
        
  check_failed_links_using_new_branch_url:
    if: contains(['2'],${{ needs.check_links_on_pr_files_online.outputs.lychee_exited }})
    runs-on: ubuntu-latest
    name: 🐙 Test failed links using branch url
    needs: [check_links_on_pr_files_online, build]
    steps:
      - uses: actions/checkout@v3
        with:
          fetch-depth: 0

      - name: Get branch name
        id: branch-name
        uses: tj-actions/branch-names@v6

      - name: 'Get branch url'
        id: git-branch-name
        run: |
          url="https://github.com/"
          url+=${{github.repository}}
          url+="/tree/"
          url+=${{ steps.branch-name.outputs.current_branch }}
          url+="/src/"
          echo "branch_name=${url}" >> $GITHUB_OUTPUT
      
      - name: 'Get broken URLs from results.md'
        id: get-broken-urls
        run: |
          broken_urls=""
          if [ ${{ needs.check_links_on_pr_files_online.outputs.lychee_exited }} == 2 ]
          then
          broken_urls+="${{ needs.check_links_on_pr_files_online.outputs.lychee_file }}"
          fi
          if [ ${{ needs.build.outputs.html_lychee_exited }} == 2 ]
          then
          broken_urls+="${{ needs.build.outputs.html_lychee_file }}"
          fi
          broken_urls=$(echo "$broken_urls" | sed "s/%25/%/g")
          body=$(echo "$broken_urls" | sed "s/%0A/$'\n'/g")
          broken_urls=$(echo "$broken_urls" | sed "s/%0D/$'\r'/g")
          broken_urls=$(echo "$broken_urls" | sed "s/https:\/\/docs.pupil-labs.com\///g")
          echo ${broken_urls}
          mkdir -p temp
          echo ${broken_urls} > temp/broken_urls.md

      - name: 'Test again these broken links with the new branch url'
        id: lychee
        uses: lycheeverse/lychee-action@v1.5.4
        env:
          GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}
        with:
          args: -v -n -a 403,503,429,200 -i -b '${{ steps.git-branch-name.outputs.branch_name }}' --exclude-mail --include-verbatim temp/broken_urls.md
          output: ./lychee/results_final.md
          jobSummary: true
          fail: true

  check_spelling:
    name: ✍️ Check spelling
    runs-on: ubuntu-latest
    outputs:
      fail_spelling: ${{ steps.fail.outputs.fail_spelling }}
    steps:
      - uses: actions/checkout@v3
        with:
          fetch-depth: 2
      - name: 'Check Grammar'
        id: cspell
        uses: streetsidesoftware/cspell-action@v2
        with:
          files: |
            **/*.{md,html,ipynb}
          strict: false
          verbose: true
        
      - name: 'Fail if spelling errors are found'
        id: fail
        if: steps.cspell.outputs.number_of_issues > 0
        run: |
          echo "fail_spelling="Spelling errors found" >> $GITHUB_OUTPUT
          exit 1