[NTLK] Converting NewtonBooks to ePub

[J Caffiney]

I've been looking at the NewtonBook binary format lately, trying to figure
out if there's a reasonable way to extract book content from .pkg files and
turn it into something readable on today's devices.

I've managed to get some basic conversions working — NewtonBook .pkg to
ePub. Results are rough but readable. So far I've run it against a few
titles from UNNA: The Art of Newton, InvesTerms, and the NS BASIC Tech
Notes. Chapter structure and text come through. Styles are hit or miss. I'm
happy to share the converted ePubs if anyone wants to see how they turned
out.

The interesting challenge is Newton's internal format — UTF-16 big endian
text, the ref encoding for objects, and the way book content is nested
inside NOS frames. Apple documented some of this in the Formats spec but
there are gaps, especially around how style runs map to the actual text
boundaries. Robert Sundling's PKGDUMP (
https://github.com/RobertSundling/Newton-PKGDUMP) was helpful for
sanity-checking the package structure, but the book-specific content layer
is mostly undocumented territory.

Does anyone have deeper knowledge of the format — especially the
relationship between style runs and text content in book parts? I've been
filling in the gaps by trial and error, but I'd rather not reinvent
the wheel.

-J