April 2026 - BaseX-Talk - mailman.uni-konstanz.de

BaseX 12.3: JNodes, Tweaks, Fixes
by Christian Grün 20 May '26

20 May '26

Dear all, We are happy to announce a new major minor (minor major ?) release of BaseX, with support for XQuery 4.0 jnodes! XPath syntax can now be used to navigate in maps and arrays: https://docs.basex.org/13/XQuery_4.0#jnodes As the feature is still fresh, all your feedback is welcome to improve it further. Here is a change log of the other additions: XQUERY 4.0 - JNodes: XPath traversal in map/array structures - Improved string representation of function items - XML, JSON: canonical serialization - Namespace declarations: support for 'fixed' and '##any' XQUERY - HTML Parsing without parser: wrap HTML contents as string value - file:delete: raise no error if target does not exist - file:copy: Preserve capitalization of target path GUI - Support for FlatLaf Look and Feels - Editor: Improved closing of tags FIXES, PERFORMANCE - xquery:fork-join was fixed - info/debugging strings: increased output length - numerous tweaks to take advantage of recent JVMs - fn:path: caching of repeatedly accessed steps Visit the following page to get a full list of new features that will officially be supported with BaseX 13: https://docs.basex.org/13/Changelog Visit our homepage to download the latest version: https://basex.org Have fun, yours, BaseX Team Christian

3 3

Error handling according to the server protocol
by Ben Engbers 18 May '26

18 May '26

Hi, I am currently busy writing C wrapper functions around the public functions from my libBasexCpp library (see https://github.com/BenEngbers/libBasexCpp) In doing so, I discovered an imperfection in that library and an ambiguity in the server protocol. The description of the command protocol states that in the event of an error, the server response consists of {partial result} {error} \01. However, at the bottom of the description of the Query protocol, it is stated that the server response ends with a \01, followed by an error string. In the current implementation, I assumed that all successfully executed requests end with a \00 and that a \01 at the end means an error has occurred. My question is whether this assumption is correct or whether I need to adjust my implementation so that the \01 is followed by the error message? Ben

2 4

TSV inputs and OOM errors
by Bridger Dyson-Smith 01 May '26

01 May '26

Hi all - I'm trying to work with some files that are TSV (tab separated values) that are fairly small (510K) but are causing OOM errors, even when launching the BaseX GUI with 8GB of memory. I've attached an example. Parsing options are separator = tab, format = direct, parse first line as table header = true; no indexes or full-text are selected. I recall this was an issue with v12.1, but I think I found an alternate representation of the data and didn't report the problem. Is this a bug, or just user/operator error? I'm using v12.3. Thanks in advance for any suggestions! Best, Bridger

2 8

update depth?
by Graydon Saunders 28 Apr '26

28 Apr '26

Hello! If I try to get rid of nested inline markup with no attributes let $test as element() := <p>First <span><span><span>word</span></span></span> was excited.</p> return $test update {(.//span[empty(@*)]) ! (replace node . with ./node()) } I get <p>First <span><span>word</span></span> was excited.</p> Which tells me (or at least I think it tells me) that the outermost replace happens last and the result is replacing the outer span with its node children. I'm not getting free recursion by matching all the span elements. What I want is <p>First word was excited.</p> where all the span elements with no associated attributes unwrap. Is there a plausibly elegant way to do this? (Stacking the same update several times works, and a fully recursive typeswitch approach works, but I can't avoid thinking there's a better way.) Thanks! Graydon -- Graydon Saunders | graydonish(a)fastmail.com Þæs oferéode, ðisses swá mæg. -- Deor ("That passed, so may this.")

3 4

BaseX 12.3: xml:id namespace lost when loading XInclude document
by Jonathan Robie 28 Apr '26

28 Apr '26

I am hitting a namespace bug when loading XInclude files with -i. I am using BaseX 12.3. Consider the following two files: main.xml: <?xml version="1.0"?> <root xmlns:xi="http://www.w3.org/2001/XInclude"> <xi:include href="included.xml"/> </root> included.xml: <?xml version="1.0"?> <item xml:id="test123"/> If I load main.xml with -i, the xml namespace is lost on xml:id attributes: basex -i main.xml '//item/@xml:id' This returns empty. But if I load it using doc(), it works fine: basex 'doc("main.xml")//item/@xml:id' This returns test123. My workaround is to use doc() with an external variable instead of -i. This may be related to issue #2642. Jonathan

2 1

BaseX UI main window class name - SunAwtFrame
by ykhabins＠bellsouth.net 23 Apr '26

23 Apr '26

Hello, I have quite an unusual request. BaseX UI has a main window class on a source code level. Would it be possible to rename its current main windows class name, i.e. SunAwtFrame, to something else? For example, BaseXUI or BaseXFrame, to make it unique. Here is a context of my request. I am using "Dark Mode Any App" to apply a dark mode for the BaseX UI: https://darkmodeanyapp.com/ "Dark Mode Any App is a simple utility which allows you to apply a dark mode filter to any Windows application, even when it does not natively support dark mode." Unfortunately, both BaseX and Oxygen XML IDE are using the same main windows class name: SunAwtFrame. It is causing a conflict between BaseX and Oxygen XML IDE for the "Dark Mode Any App" utility. It cannot distinguish between them. The dark mode is being turned on for both apps at the same time. Which is not an intended result. Thank you in advance. Regards, Yitzhak Khabinsky

2 1

Inconsistent behavior when trying to use @*:id instead of @xml:id ?
by Orderic-Vital Pain 21 Apr '26

21 Apr '26

Hello, So we came across a strange behaviour while building a webapp with BaseX and aren’t sure if it’s a bug or an edge case we can’t wrap our heads around. I’ll try to explain it the best I can, with basic examples attached. First some context: - our app needs to fetch xml fragments stored in a database (one db per app instance) which can contains anywhere from 1 to X document(s) - these xml documents can be encoded in xml-TEI (which use attribute @xml:id) or xml-ead (which use @id) ; we plan to add support for more xml vocabularies in the future and thus can’t presume whichether attribute is used - so far we’ve used a function which checks for both attributes by writing it as `@*:id`, similar to the attached query `query2` Now here’s the problem : this doesn’t seem to work to fetch element with an @xml:id attribute if and only if: - the document containing the target node only has 1 namespace declared in an @xmlns attribute (see `doc1a` and `doc1b`) - AND the document is the only one in the database we’re trying to fetch from Here are the tests we conducted before reaching out to you: - `doc1a` and `doc1b` are identicals, and contain only 1 namespace encoded in an @xmlns attribute - `doc2` is the same as them except it has 2 (identical) namespaces, one declared in an @xmlns attribute and the other in @xmlns:ns - `query1` search for specific(s) <div> using `@xml:id='div2'` - `query2` is the same but writes it as `@*:id='div2'` - If we search in doc1a (OR doc1b) ALONE, query1 find the searched <div> [test1] but query2 fails [test2] - If we search in doc2 ALONE, both query1 and query2 find the searched <div> [test3 and test4] - If we search in any combination of 2 or more of these documents, both query1 and query2 find the searched <div> [test5 and test6] This as been tested with BaseX versions 10.7, 11.7 and 12.2, both in our webapp and on the GUI – to the same result each time. Attached is a basex script which reproduce these tests. So as I said, we’re not sure why it would do that – failing with 1 doc but working with 2 or more ? In any case we can build around that (by writing eg. `[@*:id or @xml:id]`), we’re just curious to know if it’s working as intended and if yes, if there is more to this that we should know of. Thank you for your time. Best regards, O.-V. Pain

2 2

changed "function-look"-behavior in BaseX version 12.3
by Rob Stapper 21 Apr '26

21 Apr '26

Hi, Version 12.3 now requiers the module where the function is fetched from, to be imported in the same module as where the "function-lookup"-function is called from. Before this wasn't the case, the module only needed to be imported in the module where the qName for the function is created. So I could have a module for generic function retrieval. Those days seem to be over. Is this behaviour to stay or can it be reversed. (app.xqm) import module namespace c = "_function-lookup-test.c" ; import module namespace m = "_function-lookup-test.m" ; c:method( xs:QName( "m:test") , 0 )() (m.xqm) module namespace m = "_function-lookup-test.m" ; declare %public function m:test( ) { "test"} ; (c.xqm) module namespace c = "_function-lookup-test.c" ; (: Basex 12.3 requiers the lookup-module to be imported in the module that executes the 'function-lookup'-function in previous versions this was not necessary, the-lookup-module only needed to be imported in the calling module, here: "m.xqm" import module namespace m = "_function-lookup-test.m" ; (: not working anymore since 123 when not active :) :) declare %public function c:method ( $qName as xs:QName , $arity as xs:integer ) as function(*)? { function-lookup( $qName , $arity ) otherwise message( "WARNING!: Method not found: " || xs:string( $qName) || "#" || $arity ) } ; Best. Rob Stapper

3 2

Can not set HTTP status message in response anymore comparing 10.6 to 12.3
by Omar Siam 20 Apr '26

20 Apr '26

Hi list, I just started to update my error reporting library [1] to 12 and XQuery 4.0. This used HTTP messages in previous versions to report the exact kind of problem the server encountert. It relies on this snippet working: declare %rest:path("response_message") function api:response_message() { (web:response-header( {}, { 'Cache-Control': 'max-age=3600,public' }, { 'status': 200, 'message': '*Something else*' } ), "Something else") }; To my surprise this does not show "Something else" as the HTTP status message. curl -i http://localhost:8984/response_message HTTP/1.1 200 *OK* Server: Jetty(12.1.8) Date: Fri, 17 Apr 2026 15:45:09 GMT Cache-Control: max-age=3600,public Content-Type: application/xml; charset=UTF-8 Server-Timing: parse;dur=5.49 ms,compile;dur=0.68 ms,optimize;dur=0.01 ms,evaluate;dur=0.00 ms,serialize;dur=0.00 ms Content-Length: 14 Something else As setting the HTTP status message to anything is a rather obscure feature, I would say updating the docs and stating, that this no longer works would "solve" this error. The last version I tested this with and it worked is 10.6 (probably any version running in Jetty 9.4). BaseX 12.3 [HTTP Server] [main] INFO org.eclipse.jetty.server.Server - jetty-12.1.8; built: 2026-04-01T01:09:18.991Z; git: c9cdc9aaa434a3665b8a53b4d1cc3684992da649; jvm 17.0.18+8-LTS [main] INFO org.eclipse.jetty.ee9.webapp.StandardDescriptorProcessor - NO JSP Support for /, did not find org.eclipse.jetty.ee9.jsp.JettyJspServlet [main] INFO org.eclipse.jetty.session.DefaultSessionIdManager - Session workerName=node0 Server was started (port: 1984). [main] INFO org.eclipse.jetty.server.handler.ContextHandler - Started oeje9n.ContextHandler$CoreContextHandler@3543df7d{BaseX: The XML Database and XQuery Processor,/,b=file:///V:/basexapiproblem/webapp/,a=AVAILABLE,h=oeje9n.ContextHandler$CoreContextHandler$CoreToNestedHandler@7c541c15{STARTED}} [main] INFO org.eclipse.jetty.server.AbstractConnector - Started oejs.ServerConnector(a)7af30760{HTTP/1.1, (http/1.1)}{0.0.0.0:8984} [main] INFO org.eclipse.jetty.server.Server - Started oejs.Server(a)758c83d8{STARTING}[12.1.8,sto=0] @435ms HTTP STOP Server was started (port: 8081). HTTP Server was started (port: 8984). Best regards [1] https://github.com/acdh-oeaw/api-problem4restxq -- Mag. Ing. Omar Siam Austrian Center for Digital Humanities Österreichische Akademie der Wissenschaften | Austrian Academy of Sciences Stellvertretende Behindertenvertrauensperson | Deputy representative for disabled persons Bäckerstraße 13, 1010 Wien, Österreich | Vienna, Austria T: +43 1 51581-7295 omar.siam(a)oeaw.ac.at | www.oeaw.ac.at/acdh

2 1

previous mail
by Rob Stapper 17 Apr '26

17 Apr '26

I allready adjusted my software. So no need there. Just wondering if function-lookup behaviour is changed by design. Cheers. Rob Stapper

1 0