Hi Alan,
I agree with you on points mentioned by you within your preceding
mail (the mail below) within this mail thread.
I've actually now, uploaded this JEP's draft document at location
https://xalan.apache.org/xalan-j/xsl3/xalan_java_xslt3.0_jep_request_draft.md
which should be easily accessible instead of as an attachment to the
mailing list (One of mailing list members wrote me off list, that
attachments to this list are banned).
I think, that being able to use javax.xml APIs to do XSLT 3.0
transformations may be a win-win situation for Xalan users. The
existing javax.xml TrAX API can only run XSLT 1.0 transformations. The
TrAX transformer factory value to use Xalan XSLT 3.0 processor has
changed to org.apache.xalan.processor.XSL3TransformerFactoryImpl with
the value of its key remaining the same
(javax.xml.transform.TransformerFactory).
An example of complete use of Xalan XSLT 3.0 TrAX API is available
within the class
https://github.com/apache/xalan-java/blob/xalan-j_xslt3.0_mvn/src/test/java/org/apache/xalan/tests/util/XslTransformTestsUtil.java
I'm curious, whether OpenJDK's javax.xml TrAX API can be enhanced to
support both Xalan XSLT 1.0 and XSLT 3.0. May be a new JEP could be
right for this requirement.
Few of my replies to your questions, are mentioned below with prefix [mukul]
On Fri, Apr 10, 2026 at 4:23 PM Alan Bateman <[email protected]> wrote:
> Do you have a sense as to whether new applications are being developed today
> that use XSLT?
[muku] I think yes
> What is so compelling about XSLT 3.0 and Xpath 3.1 that would would be
> exciting for developers?
[mukul] XSL 3 provides lots of new language features that are not
available within XSL 1. XSL 3 could be thought of as an enhancement
like C++ over C.
--
Regards,
Mukul Gandhi
