BUG: infinite loop on damaged pdf file

[pubpub-zz]

          (for issue creation/tracking)

Note : With the latest dev build

an infinite loop occurs when getting number of pages
https://github.com/user-attachments/files/17162634/7f40cb209fb97d1782bffcefc5e7be40.pdf

Originally posted by @pubpub-zz in #2876 (comment)

[pubpub-zz]

the pdf contains some dictionnaries where the << and >> are not balanced. this creates an infinite loop when parsing file

[stefan6419846]

How is this different from #2876?

[pubpub-zz]

from my understanding, #2876 deals with memory. Here, the infinite has been generated because of #2872.
maybe at the end the fix will cover both.