Replace wkhtmltopdf with a Modern PDF API

# wkhtmltopdf: installation headaches, old rendering engine
# Ubuntu: apt install wkhtmltopdf (outdated version)
# macOS: brew install wkhtmltopdf (ARM issues)

wkhtmltopdf \
  --page-size A4 \
  --margin-top 20mm \
  --margin-bottom 20mm \
  input.html output.pdf

# Problems:
# - Flexbox/Grid may not render correctly
# - JavaScript execution is limited
# - No longer actively maintained

# HTML2DocHub: REST API, current Chromium, works everywhere
curl -X POST https://api.html2dochub.com/v1/render \
  -H "X-API-Key: sk_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "type": "pdf",
    "html": "'"$(cat input.html)"'",
    "options": {
      "format": "A4",
      "margin_top": "20mm",
      "margin_bottom": "20mm",
      "print_background": true
    }
  }' | python3 -c "import json,sys; print(json.load(sys.stdin)['''download_url'''])"

# Before (pdfkit / wkhtmltopdf)
import pdfkit
pdfkit.from_file("input.html", "output.pdf")
# Requires wkhtmltopdf binary, PATH config, etc.

# After (HTML2DocHub) — no binary dependencies
import requests

with open("input.html") as f:
    html = f.read()

resp = requests.post(
    "https://api.html2dochub.com/v1/render",
    headers={"X-API-Key": "sk_live_YOUR_KEY"},
    json={"type": "pdf", "html": html, "options": {"format": "A4"}},
)
pdf_url = resp.json()["download_url"]

# Download
import urllib.request
urllib.request.urlretrieve(pdf_url, "output.pdf")