You are viewing a plain text version of this content. The canonical link for it is here.
Posted to j-dev@xerces.apache.org by "Radu Coravu (JIRA)" <xe...@xml.apache.org> on 2018/05/08 06:22:00 UTC
[jira] [Commented] (XERCESJ-1694) Avoid stack overflow when using
xpointer between two resources
[ https://issues.apache.org/jira/browse/XERCESJ-1694?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16466946#comment-16466946 ]
Radu Coravu commented on XERCESJ-1694:
--------------------------------------
The problem is that the "XPointerParserConfiguration" has both an xiinclude and a xpointer component added to the pipeline.
So when parsing "a.xml", the "XPointerParserConfiguration" will be used to parse "b.xml". The XPointerHandler will identify the part which is pointed to and issue SAX events for it but the XIncludeHandler will continue to be used and receive events for the entire contents of the b.xml document, including the "xi:include" back to "a.xml", thus creating a stack overflow.
As a possible solution, the XIncludeHandler added to the "XPointerParserConfiguration" needs to know about the "XPointerHandler" and if the XPointerHandler has found its target, the XIncludeHandler needs to start ignoring xi:include elements because it no longer needs to expand them.
> Avoid stack overflow when using xpointer between two resources
> --------------------------------------------------------------
>
> Key: XERCESJ-1694
> URL: https://issues.apache.org/jira/browse/XERCESJ-1694
> Project: Xerces2-J
> Issue Type: Bug
> Components: SAX
> Reporter: Radu Coravu
> Priority: Major
>
> Let' say I have two documents:
> 1) "a.xml":
> {code}<book xmlns="http://docbook.org/ns/docbook"
> xmlns:xlink="http://www.w3.org/1999/xlink"
> xmlns:xi="http://www.w3.org/2001/XInclude"
> xml:id="a-book"
> version="5.0">
> <title>A</title>
> <chapter xml:id="a-chapter">
> <title>a chapter</title>
> <para>foo</para>
> </chapter>
> <xi:include href="b.xml" xpointer="b-chapter"/>
> </book>{code}
> 2) "b.xml":
> {code}<book xmlns="http://docbook.org/ns/docbook"
> xmlns:xlink="http://www.w3.org/1999/xlink"
> xmlns:xi="http://www.w3.org/2001/XInclude"
> xml:id="b-book"
> version="5.0">
> <title>B</title>
> <chapter xml:id="b-chapter">
> <title>b chapter</title>
> <para>foo</para>
> </chapter>
> <xi:include href="a.xml" xpointer="a-chapter"/>
> </book>{code}
> When using a SAX parser to parse "a.xml" a stack overflow occurs although according to the XInclude specs this is a legal situation, the xpointer parts to a part of b.xml.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: j-dev-unsubscribe@xerces.apache.org
For additional commands, e-mail: j-dev-help@xerces.apache.org