[
https://issues.apache.org/jira/browse/PDFBOX-3592?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15691186#comment-15691186
]
Pat Hickey commented on PDFBOX-3592:
------------------------------------
* Diff NonSequentialPDFParser.java (1.8.9):
{noformat}569c569,571
< parseXrefObjStream(prev, false);
---
> if ( streamoffset != prev ) {
> parseXrefObjStream( prev, false );
> }
{noformat}
* DIff COSParser.java (2.0.0):
{noformat}258c258,260
< parseXrefObjStream(prev, false);
---
> if ( streamOffset != prev ) {
> parseXrefObjStream(prev, false);
> }
{noformat}
> Missing stream in /XrefStm causes parse to break
> ------------------------------------------------
>
> Key: PDFBOX-3592
> URL: https://issues.apache.org/jira/browse/PDFBOX-3592
> Project: PDFBox
> Issue Type: Bug
> Components: Parsing
> Affects Versions: 1.8.9, 2.0.0
> Reporter: Pat Hickey
> Priority: Trivial
>
> When the pdf document contains a startref and xref table (a la PDF1.4) as
> well as an /XRefStm, but the offset given by /XRefStm does not describe a
> xref (stream or otherwise), PDFBox does its level best to find the xref info
> using a brute force search. In the case where the document only contains the
> xref table given by startxref, and does not contain an object of /Type /XRef,
> the brute force search results in the xref table.
> See diffs for COSParser.java(2.0.0) and NonSequentialParser.java (1.8.9),
> below.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
