New PDF Engine does not seem to work

[groogiam]

Describe the bug
I am attempting to follow the documentation for the new PDF engine here gttps://dotnet.github.io/docfx/docs/pdf.html. The documentation references a "Download PDF" but after following the instructions and updating my docfx.json file one does not appear after rebuilding.

To Reproduce
Steps to reproduce the behavior:

Added to my docfx.json

{
  "build": {
    "globalMetadata": {
      "pdf": true
    }
  }
}

Then ran

docfx build docfx.json -s

The site builds with no errors and I can open it in my browser but there is no Download PDF button.

Expected behavior
The "Download PDF" button should show as per the documentation.

Context (please complete the following information):

• OS: Windows 10 Pro
• Docfx version: 2.73.1+1d7b96c23465e14eafbcb0dd765029a87b5312d4

Additional context
Add any other context about the problem here.

[groogiam]

After some playing around it does seem to be generating multiple pdf files in the background but still no links. I still have several other issues.

1. Is there a way to generate a single pdf document with all the content instead of a bunch of one off pdf's. The old engine had this support but it appears this was not accounted for in the new engine.
2. pdfCoverPage metadata does not seem to work. The documentation is a bit unclear. The old engine referenced a markdown file in the repository. This seems to reference something from the output but the any pdf title page I generate is not part of the output and I would not want it to show up there.
3. pdfTocPage seems to be adding numbers at the end of the text.

Thanks in advance for your help.

[yufeih]

The "Download PDF link" and TOC page styling is currently only applied to the modern template. Is it feasible to use that template instead?