Docbook 5.0 Relax Schema and Xpointer Resolution
Attempting to utilize the Docbook 5.01 Relax schema with xsl-1.69.1 and oXygen-6.2. Problem with xpointers. When utilized in conjunction with xincludes, xpointers resolve via absolute positioning but not via id. Could this be an issue with xinclude.mod and the xml:id designation in v5.0 or my misconfiguration? Ray
Hi Ray, As you probably know oXygen uses Xerces for parsing XML documents and for providing XInclude support. The XPointer support is Xerces is limited, IDs are supported only when they are defined in DTDs. See more information at: Does the XInclude processor support XPointer? http://xerces.apache.org/xerces2-j/faq-xinclude.html#faq-7 What types of IDs are currently supported for XPointers? http://xerces.apache.org/xerces2-j/faq-xinclude.html#faq-8 Best Regards, George --------------------------------------------------------------------- George Cristian Bina <oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger http://www.oxygenxml.com Ray Miller wrote:
Attempting to utilize the Docbook 5.01 Relax schema with xsl-1.69.1 and oXygen-6.2.
Problem with xpointers.
When utilized in conjunction with xincludes, xpointers resolve via absolute positioning but not via id. Could this be an issue with xinclude.mod and the xml:id designation in v5.0 or my misconfiguration?
Ray
------------------------------------------------------------------------
_______________________________________________ oXygen-user mailing list oXygen-user@oxygenxml.com http://www.oxygenxml.com/mailman/listinfo/oxygen-user
Hi Ray, We will look into this, however it is not an easy task. To understand what happens here it is how the processing is performed. Let's assume you have an XML document master.xml and a document from which you want to include a fragment, fragment.xml. The Relax NG validator acts as a SAX content handler placed after the XML parser that parses the document. The XML parser does not know about the Relax NG validator. The XInclude processing is done inside the XML parser So, when the XML parser parses master.xml it will already resolve the XInclude element and will replace it with the actual included content. This reach the Relax NG validator. In the parsing phase there is no schema processing involved as the XML parser does not even know about a schema processor. In order to have a fragment from fragment.xml included you need to know what is an ID in fragment.xml and for that you need to parse it against a schema (but that document may not be valid against some schema or there may be no schema associated with it). In any case, finding the IDs in the fragment.xml requires parsing it against a schema from the XML parser which does not know about any schema. Now, the possible solution here will be to forget about schema and rely on the xml:id for determining what values represent IDs in the included document. This I think is doable but again it is not a trivial task. I will add this on our bugzilla and we will look into it. As a workaround for the current situation you can place the content you want to include in a separate file and include it both from the master document and from the fragment document. For instance if you have a para in the fragment.xml then you can place that in a new file, xinclude that file in fragment.xml in the place the para used to be and xinclude it also in the master xml where you included it using xpointer. Best Regards, George --------------------------------------------------------------------- George Cristian Bina <oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger http://www.oxygenxml.com Ray Miller wrote:
IDs are supported only when they are defined in DTDs. <<
So how will the oXygen community cope with Docbook-rng schemas; surely oXygen has a plan other than to state nonsupport.
Xinclude / XPointer is essential for many of us docbook users.
Advice please.
Ray
_______________________________________________ oXygen-user mailing list oXygen-user@oxygenxml.com http://www.oxygenxml.com/mailman/listinfo/oxygen-user
Just to update also the oxygen-user list with the latest developments on support for xml:id IDs for XInclude/XPointer. I logged an issue and a patch with this on Xerces Jira: http://issues.apache.org/jira/browse/XERCESJ-1113 We made a build starting from Xerces 2.7.1 and applying that patch: http://www.oxygenxml.com/update/xmlid/LICENSE http://www.oxygenxml.com/update/xmlid/xercesImpl.jar http://www.oxygenxml.com/update/xmlid/xml-apis.jar To get this running in oXygen 6.2 copy the above jars over the ones from [oXygen]/lib folder. Best Regards, George --------------------------------------------------------------------- George Cristian Bina <oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger http://www.oxygenxml.com George Cristian Bina wrote:
Hi Ray,
We will look into this, however it is not an easy task. To understand what happens here it is how the processing is performed. Let's assume you have an XML document master.xml and a document from which you want to include a fragment, fragment.xml.
The Relax NG validator acts as a SAX content handler placed after the XML parser that parses the document.
The XML parser does not know about the Relax NG validator.
The XInclude processing is done inside the XML parser
So, when the XML parser parses master.xml it will already resolve the XInclude element and will replace it with the actual included content. This reach the Relax NG validator. In the parsing phase there is no schema processing involved as the XML parser does not even know about a schema processor. In order to have a fragment from fragment.xml included you need to know what is an ID in fragment.xml and for that you need to parse it against a schema (but that document may not be valid against some schema or there may be no schema associated with it). In any case, finding the IDs in the fragment.xml requires parsing it against a schema from the XML parser which does not know about any schema.
Now, the possible solution here will be to forget about schema and rely on the xml:id for determining what values represent IDs in the included document. This I think is doable but again it is not a trivial task. I will add this on our bugzilla and we will look into it.
As a workaround for the current situation you can place the content you want to include in a separate file and include it both from the master document and from the fragment document. For instance if you have a para in the fragment.xml then you can place that in a new file, xinclude that file in fragment.xml in the place the para used to be and xinclude it also in the master xml where you included it using xpointer.
Best Regards, George --------------------------------------------------------------------- George Cristian Bina <oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger http://www.oxygenxml.com
Ray Miller wrote:
IDs are supported only when they are defined in DTDs. <<
So how will the oXygen community cope with Docbook-rng schemas; surely oXygen has a plan other than to state nonsupport.
Xinclude / XPointer is essential for many of us docbook users.
Advice please.
Ray
_______________________________________________ oXygen-user mailing list oXygen-user@oxygenxml.com http://www.oxygenxml.com/mailman/listinfo/oxygen-user
oXygen-user mailing list oXygen-user@oxygenxml.com http://www.oxygenxml.com/mailman/listinfo/oxygen-user
participants (2)
-
George Cristian Bina -
Ray Miller