lint.yml

name: lint

on:
  push:
    branches:
      - main
  pull_request:
    types: [opened, synchronize]

jobs:
  markdown:
    name: "episode/markdown"
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v4
    - name: Dates are valid
      run: |
        pip install pytz ciso8601
        for episode in _episodes/*/*.md; do
          date=$(grep 'date:' "$episode" | head -n1 | sed 's/^date: //')
          if ! python -c "import ciso8601; ciso8601.parse_rfc3339('$date');"; then
            echo "$episode: bad date '$date'"
            exit 1
          fi
        done
    - name: No smart symbols
      run: |
        for episode in _episodes/*/*.md; do
          # this isn't a normal dash, and doesn't get turned into a list
          if grep -qF '⁃' "$episode"; then
            echo "$(basename "$episode"): abnormal dash won't make a list"
            exit 1
          fi
          if grep -P '[“‘’”]' "$episode"; then
            echo "$(basename "$episode"): found smart quotes"
            exit 1
          fi
        done
    - name: Timecode lists are correctly formatted
      run: |
        for episode in _episodes/*/*.md; do
          # timecodes should never start a line (should be in header or list)
          if grep -qP '^\[@' "$episode"; then
            echo "$(basename "$episode"): timecode not in list or header"
            exit 1
          fi
          # timecode listings need to not have empty lines, or we'll get
          #
          #   <li><p>[@HH:MM:SS]
          #
          # which doesn't render right. it happens to work for timecode
          # listings that have sub-listings, but easiest to check that there
          # just aren't any gaps.
          if ! awk '/^\s*$/ { empty = 1; next; } /^\s*-\s*\[@[0-9]/ { if (in_list == 1 && empty == 1) { exit 1; } else { in_list = 1; empty = 0; next; } } /^\s*-/ { empty = 0; next; } { in_list = 0; empty = 0; }' "$episode"; then
            echo "$(basename "$episode"): empty lines between list items"
            exit 1
          fi
        done
    - name: No duplicate URLs
      run: |
        for episode in _episodes/*/*.md; do
          file=$(grep 'file:' "$episode" | head -n1 | sed -e 's/^file: //' -e 's/"//g')
          n=$(grep -F "$file" -l _episodes/*/*.md | wc -l)
          if [[ $n -gt 1 ]]; then
            echo "$episode: shares file with other episodes:"
            grep -F "$file" -l _episodes/*/*.md | grep -vF "$episode"
            exit 1
          fi
        done
    - name: No duplicate slugs
      # For collections, jekyll _only_ uses the basename (without date) of each
      # post for the slug, and doesn't error on duplicates. So we must check.
      run: |
        for episode in _episodes/*/*.md; do
          slug="$(basename "$episode" | sed 's/^[0-9][0-9][0-9][0-9]-[0-9][0-9]-[0-9][0-9]-//')"
          files=(_episodes/*/????-??-??-$slug)
          if [[ "${#files[@]}" -gt 1 ]]; then
            echo "Duplicate slugs found: ${files[*]}"
            exit 1
          fi
        done
  audio:
    name: "episode/audio"
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v4
    - name: Get all changed episode files
      id: changed-markdown-files
      uses: tj-actions/changed-files@v44
      with:
        # Avoid using single or double quotes for multiline patterns
        files: |
           _episodes/*/*.md
    - name: File size specifications are correct
      # note: we only guard this one since it's slow to download all the
      # episode files; the others are fine to run for all files since they run
      # locally.
      if: steps.changed-markdown-files.outputs.any_changed == 'true'
      env:
        ALL_CHANGED_FILES: ${{ steps.changed-markdown-files.outputs.all_changed_files }}
      run: |
        for episode in ${ALL_CHANGED_FILES}; do
          # skip deleted episodes (if any)
          [ -e "$episode" ] || continue;
          length=$(grep 'length:' "$episode" | head -n1 | awk '{print $2}' | sed 's/"//g')
          file=$(grep 'file:' "$episode" | head -n1 | sed -e 's/^file: //' -e 's/"//g')
          size=$(curl --head -f -s "$file" | grep -i content-length | awk '{print $2}' | sed 's/\r//')
          if [ -z $size ]; then
            echo "couldn't get content length of \"$file\""
            exit 1
          fi
          if [ $size -ne $length ]; then
            echo "$(basename "$episode"): ${length}b (reported) != ${size}b (actual)"
            exit 1
          fi
        done
  transcript:
    name: "episode/transcript"
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v4
    - name: Transcripts match episodes
      run: |
        for transcript in _transcripts/*/*.md; do
          episode="_episodes/$(basename "$(dirname "$transcript")")/$(basename "$transcript")"
          if [[ ! -e $episode ]]; then
            echo "$transcript: no matching episode"
            exit 1
          fi
        done
  feed:
    name: "feed/validate"
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v4
    - uses: ruby/setup-ruby@v1
      with:
        ruby-version: '2.7'
        bundler-cache: true
    - name: Install dependencies
      run: bundle install
    - name: Build site and feed
      run: bundle exec jekyll build
    - name: Grab feed validator
      run: |
        git clone https://github.com/w3c/feedvalidator.git
    - uses: actions/setup-python@v4
      with:
        python-version: '3.x'
        cache: 'pip'
    - name: Validate feed
      run: |
        # Fix Ubuntu MIME type for RSS
        sudo sed -i 's@application/x-rss+xml@application/rss+xml@' /etc/mime.types
        cp _site/podcast.rss validate.rss
        # https://github.com/rubys/feedvalidator/issues/16
        sed -i -e 's/https:/http:/g' \
               -e '/rel="self"/ s@href="[^"]*"@href="file://'"$(pwd)"'/validate.rss"@' \
               -e '/xmlns:content/a \
                    xml:base="https://rustacean-station.org"' \
               validate.rss
        cd feedvalidator
        pip install -r requirements.txt
        if ! python src/demo.py ../validate.rss; then
          cat ../validate.rss
          exit 1
        fi