You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@pdfbox.apache.org by "Dennis Noordsij (Jira)" <ji...@apache.org> on 2020/09/24 18:37:00 UTC
[jira] [Comment Edited] (PDFBOX-4965) Stack overflow in
PDFCloneUtility
[ https://issues.apache.org/jira/browse/PDFBOX-4965?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17201584#comment-17201584 ]
Dennis Noordsij edited comment on PDFBOX-4965 at 9/24/20, 6:36 PM:
-------------------------------------------------------------------
I've found the culprit, there is a chain of dictionary entries like:
{{%% Original object ID: 615 0 }}
{{627 0 obj }}
{{<< }}
{{/Dest [ }}
{{ 15 0 R }}
{{ /XYZ }}
{{ 428 }}
{{ 392 }}
{{ 0 }}
{{ ] }}
{{ /Next 625 0 R }}
{{ /Parent 5 0 R }}
{{ /Prev 629 0 R }}
{{ /Title (position_1516796_2160) }}
{{>> }}
{{endobj }}
that is over 5000 entries in length.
The document is just a couple of pages of text, but it was created by Microsoft Word .....
With a larger stack size it works. However it is still possible to craft a PDF file that will cause a stackoverflow when used in the merger utility. Perhaps the cloneForNewDocument can be implemented without recursion.
was (Author: dennisn):
java.lang.StackOverflowError
at java.util.HashMap.putVal(HashMap.java:629)
at java.util.HashMap.put(HashMap.java:612)
at java.util.HashSet.add(HashSet.java:220)
at org.apache.pdfbox.util.SmallMap.entrySet(SmallMap.java:383)
at org.apache.pdfbox.cos.COSDictionary.entrySet(COSDictionary.java:1398)
at org.apache.pdfbox.multipdf.PDFCloneUtility.cloneForNewDocument(PDFCloneUtility.java:142)
at org.apache.pdfbox.multipdf.PDFCloneUtility.cloneForNewDocument(PDFCloneUtility.java:109)
> Stack overflow in PDFCloneUtility
> ---------------------------------
>
> Key: PDFBOX-4965
> URL: https://issues.apache.org/jira/browse/PDFBOX-4965
> Project: PDFBox
> Issue Type: Bug
> Components: Utilities
> Affects Versions: 2.0.21
> Environment: FreeBSD
> Reporter: Dennis Noordsij
> Priority: Major
>
> My production system crashed with the following trace:
> java.lang.StackOverflowError
> at java.util.HashMap.hash(HashMap.java:339)
> at java.util.HashMap.get(HashMap.java:557)
> at org.apache.pdfbox.multipdf.PDFCloneUtility.cloneForNewDocument(PDFCloneUtility.java:81)
> at org.apache.pdfbox.multipdf.PDFCloneUtility.cloneForNewDocument(PDFCloneUtility.java:146)
> at org.apache.pdfbox.multipdf.PDFCloneUtility.cloneForNewDocument(PDFCloneUtility.java:109)
> at org.apache.pdfbox.multipdf.PDFCloneUtility.cloneForNewDocument(PDFCloneUtility.java:146)
> at org.apache.pdfbox.multipdf.PDFCloneUtility.cloneForNewDocument(PDFCloneUtility.java:109)
> at org.apache.pdfbox.multipdf.PDFCloneUtility.cloneForNewDocument(PDFCloneUtility.java:146)
> (repeated)
>
> We do a lot with PDF files and I don't have the actual PDF or action which caused this at the moment, but perhaps this already rings a bell with someone of what kind of bug this could be or wether I am doing something strange?
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@pdfbox.apache.org
For additional commands, e-mail: dev-help@pdfbox.apache.org