ENH: Preserve links in added pages also when links come from merged-in pages

[larsga]

This is the last part of the implementation of issue 3290.

It causes the tests to be slow because named_destinations is not cached, but I'm leaving that as a separate task.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.80%. Comparing base (23339e5) to head (a9da8f6).
⚠️ Report is 36 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3400      +/-   ##
==========================================
+ Coverage   97.70%   97.80%   +0.10%     
==========================================
  Files          55       55              
  Lines       10410    10526     +116     
  Branches     1928     1965      +37     
==========================================
+ Hits        10171    10295     +124     
+ Misses        133      128       -5     
+ Partials      106      103       -3     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[stefan6419846]

It causes the tests to be slow because named_destinations is not cached, but I'm leaving that as a separate task.

This is an unexpected side effect of these changes and as we know that it decreases performance, I cannot merge this PR safely without addressing the speed issues - ideally as a separate PR.

[larsga]

Now that PR #3442 has been merged, and I've fixed the issue you found, I guess this one might be ready to merge as well?

[stefan6419846]

Thanks for the updates. I have left some remarks where I see further changes to consider.

[larsga]

This should be ready for review now.

[stefan6419846]

Thanks for the updates.

Besides the inline comments: The modifications to pypdf/generic/__init__.py and NamedReferenceLink.__init__ breaks backwards compatibility. As they have been exposed as public API, we unfortunately need to keep this backwards-compatible with a proper deprecation period and cannot simply remove them. Please indicate if you need help with this, especially with the __init__.py part.

[stefan6419846]

Could we please add an assertion here?

[larsga]

As they have been exposed as public API, we unfortunately need to keep this backwards-compatible with a proper deprecation period and cannot simply remove them. Please indicate if you need help with this, especially with the init.py part.

@stefan6419846 Yes, please. I don't know the deprecation process for PyPDF, so I'm at a loss for how to proceed here.

[stefan6419846]

Yes, please. I don't know the deprecation process for PyPDF, so I'm at a loss for how to proceed here.

I will try to look into this as fast as possible, but for personal reasons, it might take a few weeks.