Ignore only the outline whose named destination is not defined - #129

When an outline refers to a named destination that the document doesn't define, doc.get_dest() raises PDFDestinationNotFound inside the loop in _get_outlines(). process_file() catches it and logs

WARNING: Failed to retrieve outlines: b'_TableParagraph71_MS.F0'

but by then the generator has stopped, so every outline after that one is lost too, and annotations in those sections get no heading.

I found it with 2017.pdf from pypdf's test files (132 pages): it has 530 outlines, 10 of which name a destination that isn't defined, and pdfannots keeps only the first 70.

_get_outlines() now catches PDFDestinationNotFound for that one outline, logs a warning naming it, and goes on with the next one. On 2017.pdf it keeps 520 outlines.

tests/missing-named-destination.pdf is a small hand-written file with three outlines, the second of which names a destination that isn't in /Dests. The MissingNamedDestination tests check that Chapter 1 and Chapter 3 are kept and that one warning names Chapter 2; both fail on main. pytest tests.py, mypy ., flake8 and autopep8 --diff pass.

Used Claude (Opus) to help find and patch this; I reviewed and tested the change