Francesco Pretto

> 1. **Division-by-zero exception handling**. Ok. > 3. **Support for `BitsPerComponent` values less than 8**. I understand. I'm too busy to add support for this. If you take the challenge,...

The document is actually corrupted. Adobe Acrobat doesn't crash with this document but it will not open Document Properties at first, and when you ask for properties again it will...

I just thought about not trying to not eagerly loading `/Info`, which it would be not that bad. But the fact is also that the problem would be delayed until...

PoDoFo parser is not terrible but it's not handling some edge cases described in this thread [1], and resolution for those is planned but pending. If `offset1` fails to parse...

> While [p3.pdf](https://github.com/user-attachments/files/19503374/p3.pdf) can be opened with standard PDF readers, it only displays the content of [p2.pdf](https://github.com/user-attachments/files/19503392/p2.pdf). Which is the correct behavior, according to the standard. What happens with PoDoFo?...

> When PoDoFo parses p3.pdf, it locates the last startxref at position 12650. However, attempting to use this offset to find the xref table in p3.pdf will inevitably fail because...

Ok, I just verified: PDF.js goes into [recovery mode](https://github.com/mozilla/pdf.js/blob/3da8901f2661a41f4b43688830f73a8a361b4924/src/core/xref.js#L411) to parse [p3.file](https://github.com/user-attachments/files/19503374/p3.pdf). I don't believe the specification say anything about this situation, but maybe Peter can say more. I'm sorry,...

I'm working on this right now. To not waste time, I'm looking at the [implementation](https://pdfium.googlesource.com/pdfium/+/1953ba96515b3f9b21703afb3f201b3521012aa8/core/fpdfapi/parser/cpdf_parser.cpp#756) of PDFium, which it looks way easier to implement than the one in pdf.js. Stay...

It does work with a couple of tests of mine, but doesn't appear to work correctly with [p3.pdf](https://github.com/user-attachments/files/19503374/p3.pdf) as attached [here](https://github.com/podofo/podofo/issues/242#issuecomment-2760810274). I will check it later.

[p3.pdf](https://github.com/user-attachments/files/19503374/p3.pdf) is now parsed correctly and unit tested. Serious testing is still pending but I'm satisfied enough, for now.