validate-content.yml

name: validate-content

on:
  push:
    branches: ['**']
  pull_request:
    branches: ['**']

jobs:
  validate:
    name: Validate blog frontmatter, JSON, and build scripts
    runs-on: ubuntu-latest
    steps:
      - name: Checkout
        uses: actions/checkout@v4

      - name: Set up Python 3.11
        uses: actions/setup-python@v5
        with:
          python-version: '3.11'

      - name: Install Python deps
        run: pip install pyyaml

      - name: Validate blog frontmatter
        run: |
          python3 - <<'PY'
          import pathlib, sys, yaml

          REQUIRED = {"title", "date", "category", "excerpt"}
          blog_dir = pathlib.Path("website/assets/blog")
          if not blog_dir.exists():
              print(f"::warning::{blog_dir} not found, skipping")
              sys.exit(0)

          posts = sorted(blog_dir.glob("*.md"))
          if not posts:
              print(f"::warning::No posts found in {blog_dir}")
              sys.exit(0)

          errors = []
          categories = set()
          for p in posts:
              text = p.read_text(encoding="utf-8")
              if not text.startswith("---"):
                  errors.append(f"{p.name}: missing frontmatter")
                  continue
              try:
                  _, fm, _ = text.split("---", 2)
              except ValueError:
                  errors.append(f"{p.name}: malformed frontmatter delimiters")
                  continue
              try:
                  data = yaml.safe_load(fm) or {}
              except yaml.YAMLError as e:
                  errors.append(f"{p.name}: YAML parse error: {e}")
                  continue
              missing = REQUIRED - set(data.keys())
              if missing:
                  errors.append(f"{p.name}: missing required fields: {sorted(missing)}")
                  continue
              if data.get("category"):
                  categories.add(str(data["category"]))

          if errors:
              for e in errors:
                  print(f"::error::{e}")
              sys.exit(1)

          print(f"OK: {len(posts)} posts, {len(categories)} categories")
          PY

      - name: Validate JSON data files
        run: |
          python3 - <<'PY'
          import json, pathlib, sys

          roots = [
              pathlib.Path("website/assets/data"),
              pathlib.Path("website/assets/data/pages"),
          ]
          errors = []
          count = 0
          for root in roots:
              if not root.exists():
                  continue
              for jf in sorted(root.glob("*.json")):
                  count += 1
                  try:
                      json.loads(jf.read_text(encoding="utf-8"))
                  except json.JSONDecodeError as e:
                      errors.append(f"{jf}: {e}")

          if errors:
              for e in errors:
                  print(f"::error::{e}")
              sys.exit(1)
          print(f"OK: {count} JSON files parsed cleanly")
          PY

      - name: Set up Node 20
        uses: actions/setup-node@v4
        with:
          node-version: '20'

      - name: Dry-run build-blog-index
        run: |
          if [ -f website/scripts/build-blog-index.py ]; then
            python3 website/scripts/build-blog-index.py
          else
            echo "::warning::website/scripts/build-blog-index.py not found"
          fi

      - name: Dry-run build-sitemap
        run: |
          if [ -f website/scripts/build-sitemap.py ]; then
            python3 website/scripts/build-sitemap.py
          else
            echo "::warning::website/scripts/build-sitemap.py not found"
          fi

      - name: Summary
        run: |
          python3 - <<'PY'
          import json, pathlib

          posts = list(pathlib.Path("website/assets/blog").glob("*.md")) \
                  if pathlib.Path("website/assets/blog").exists() else []
          idx = pathlib.Path("website/assets/blog/index.json")
          cats = []
          if idx.exists():
              try:
                  data = json.loads(idx.read_text(encoding="utf-8"))
                  cats = data.get("categories", [])
              except Exception:
                  pass
          pages = list(pathlib.Path("website").glob("*.html"))
          print(f"Posts:      {len(posts)}")
          print(f"Categories: {len(cats)}")
          print(f"Pages:      {len(pages)}")
          PY