Don't render index on PDF. + delete unused files

[artoonie]

@yezr found by editing generator.py to print out the filename it was using for the match. Why is it ".html" when the documentation says it shouldn't be? Who knows.

Note a potential downside: the numbering no longer matches on the PDF versus the website. The PDF generator adds its own header numbering to the ToC

and in the document itself for H1 and H2

[yezr]

These little formatting conventions were how we ended up with the print-page. All the third party PDF generators were death by a thousand cuts with little random things that didn't match up with the conventions we had in the TDP. This is a move in the right direction.

Can we leave the print-page as is? As is, it isn't navigable except by knowing the slug. It can be a good backup option with a clean PDF until we can address all the formatting assumptions in mkdocs-with-pdf. Tradeoffs with the print-page are that it has some unnecessary stuff in the header and footer, like the time it was printed and the url

[artoonie]

Can do! I removed it to allow for compiling with strict mode (where warnings are failures), but there were a few other warnings that I couldn't get rid of, so using strict mode isn't an option yet. Restored.