Re: [Podofo-users] PoDoFo cannot load some PDFs with XMP metadata

[Uli Z. <ul...@ri...>]

Am 18.04.2014 um 04:32 schrieb Uli Zappe <ul...@ri...>:

> Or maybe I should say: It definitely happens because of adding/editing XMP metadata to a PDF file. Maybe this has unintended consequences somewhere else – I don't know anything about PoDoFo's internals yet.

Oooops - rethinking what I just wrote, it occurred to me that maybe PoDoFo for some reason doesn't like my removing a character from the PDF with a hex editor, and this seems indeed to be the case.

Adding one random character somewhere in the WebStatement text within the XMP data, thereby keeping the length of the complete XMP packet constant, does indeed remove the PoDoFo error. And if I add an XMP packet with begin='' (without a character within the quotes) to a PDF file with podofoxmp, podofoxmp can also read this XMP data just fine afterwards.

So contrary to what I assumed, PoDoFo does handle begin='' (without a character within the quotes) correctly, it just doesn't like my tinkering with the PDF with a hex editor - which should not be a "real world" situation ever.

Sorry for the confusion!!

Bye

Uli