pdfr - HTML to PDF over HTTP (powered by htmltopdf)

POST /render
  Option A (JSON):
    {"html": "<h1>Hello</h1>",
     "landscape": false,
     "margin": 36,
     "font": "Georgia",
     "filename": "document.pdf",
     "flatten_css": false,
     "base_url": "https://example.com",
     "selector": "#main-content"}
  Option B (raw body):
    request body = the HTML document; use query params for options:
    ?landscape=true&margin=36&font=Georgia

  Response: application/pdf

Options (JSON fields, or query params for raw bodies; JSON fields win):
  landscape=true|1        force A4 landscape
  margin=<points>         page margin in PDF points
  font=<path|family>      embed a TrueType font by file path or family name
  filename=<name>         download filename (default document.pdf)
  flatten_css=true (?flatten=true)
                          inline <link rel="stylesheet"> targets into <style>
                          blocks before rendering (the engine never fetches
                          CSS itself; relative hrefs need base_url;
                          content-fingerprinted asset URLs are cached on
                          disk, see PDFR_CACHE_DIR)
  base=<url>              base URL for resolving stylesheet hrefs
  selector=<css-sel>      render only the first matching element (plus the
                          document's <style> blocks), e.g. #main-content

Examples:
  curl -X POST http://127.0.0.1:8080/render \
    -H 'Content-Type: application/json' \
    -d '{"html": "<h1>Hello</h1><p>World</p>"}' -o out.pdf

  curl -X POST http://127.0.0.1:8080/render \
    -H 'Content-Type: text/html' \
    --data-binary @invoice.html -o invoice.pdf

  # saved page: inline its linked CSS, keep only the main content
  curl -X POST http://127.0.0.1:8080/render \
    -H 'Content-Type: application/json' \
    -d "$(jq -Rs '{html: ., flatten_css: true, base_url: \"https://example.com\", selector: \"#main-content\"}' page.html)" \
    -o clean.pdf

GET /health   -> 200 ok