[
https://issues.apache.org/jira/browse/PDFBOX-5190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17341272#comment-17341272
]
Andreas Lehmkühler commented on PDFBOX-5190:
--------------------------------------------
The given pdf is just a middle part of a pdf. The exception occurs when
rebuilding the header using the brute force parser.
Object 20 0 is a compressed object stream which was modified by adding a lot
0-bytes. That leads to an interesting stream when decompressing it using a
FlateFilter. In the beginning the first part of the stream is repeated several
times, which isn't a real issue at all, as those object aren't added to the
corresponding COSDictionary as they are already there. But after some time the
following content occurs:
{code}
/Parent 0 R/Resources 13 0 R/MediaBox[0 0 595.2756 841.8898]/TrimBox [ox [ox
[ox [ox [ox [ox [ox [ox [ox
{code}
"/Parent 0 R" leads to a parser error but PDFBox is more or less able to
recover from it. But after "/TrimBox" the parser tries to read a COSArray. As
there are a lot of consecutive "[ox" strings PDFBox runs into a recursive loop
and ended up in a stack overflow.
> COSParser: stack overflow when reading a corrupt pdf
> ----------------------------------------------------
>
> Key: PDFBOX-5190
> URL: https://issues.apache.org/jira/browse/PDFBOX-5190
> Project: PDFBox
> Issue Type: Bug
> Components: Parsing
> Affects Versions: 2.0.23, 3.0.0 PDFBox
> Reporter: Andreas Lehmkühler
> Assignee: Andreas Lehmkühler
> Priority: Major
> Attachments: crash-23adb315cc495bfd4761b16f877639d73d124d5b.pdf
>
>
> The attached file leads to a StackOverflowError while loading it.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
