@qt4cg statuses

This page displays recent status updates about the QT4CG project.

The are also captured in an RSS feed.

By year: 2025, 2024, 2023, 2022, 2021, 2020

QT4 CG meeting 140 draft minutes #minutes-10-28

28 Oct at 17:15:00 GMT

Link: https://qt4cg.org/meeting/minutes/2025/10-28

Draft minutes published.

Issue #2205 closed #closed-2205

28 Oct at 17:10:19 GMT

Link: https://github.com/qt4cg/qtspecs/pull/2205

2190 Drop binary input for parse-csv and parse-json

Issue #2217 closed #closed-2217

28 Oct at 17:08:46 GMT

Link: https://github.com/qt4cg/qtspecs/issues/2217

`bin:decode-string`: Interpretation of `$encoding` argument

Issue #2222 closed #closed-2222

28 Oct at 17:08:45 GMT

Link: https://github.com/qt4cg/qtspecs/pull/2222

2217 bin:decode-string: Input encoding

Issue #2214 closed #closed-2214

28 Oct at 17:01:59 GMT

Link: https://github.com/qt4cg/qtspecs/issues/2214

Split XSLT streaming into a separate specification

Issue #2264 closed #closed-2264

28 Oct at 17:01:58 GMT

Link: https://github.com/qt4cg/qtspecs/pull/2264

2214 Separate streaming into its own spec

Issue #2258 closed #closed-2258

28 Oct at 17:00:15 GMT

Link: https://github.com/qt4cg/qtspecs/issues/2258

array:split() isn't exactly the inverse of array:join()?

Issue #2262 closed #closed-2262

28 Oct at 17:00:14 GMT

Link: https://github.com/qt4cg/qtspecs/pull/2262

2258 Correct the statement that split is the inverse of join

Pull request #2266 created #created-2266

28 Oct at 13:06:45 GMT

Link: https://github.com/qt4cg/qtspecs/pull/2266

540 system-property equivalent for XQuery

Meets the requirement by adding magic environment variable names to fn:environment-variables.

QT4 CG meeting 140 draft agenda #agenda-10-28

27 Oct at 12:00:00 GMT

Link: https://qt4cg.org/meeting/agenda/2025/10-28

Draft agenda published.

Issue #2265 created #created-2265

27 Oct at 07:28:50 GMT

Link: https://github.com/qt4cg/qtspecs/issues/2265

XDM 4.0: a question about mutability/immutability of sequence types

As the maintainer of an implementation of XPath for Python (on ElementTree/lxml data), I'm examining the changes in XPath/XDM 4.0 and in particular the formalization of a Sequence Type and the substitution of the dynamic context item with a context value that possibly include a sequence instance within possible values.

In my implementation the sequences are generated only when necessary (e.g. the comma operator) or at the end for exposing results.

Currently sequences are expanded as Python lists (mutable sequences). But reading the sequence type definition in XDM 4.0, my interpretation is that the formalized sequence type is intended to be an immutable sequence. Is it correct or mutability VS immutability of sequence objects is considered an implementation choice?

Thank you

Pull request #2264 created #created-2264

23 Oct at 16:45:04 GMT

Link: https://github.com/qt4cg/qtspecs/pull/2264

2214 Separate streaming into its own spec

This is work in progress on splitting off streaming into a separate specification.

I've been mainly concerned with (a) a fairly crude extraction of text from the XSLT spec into the new streaming spec (b) fixing up the mechanics of making it build (c) dealing with cross-spec references. The current state is that it builds (for me) but with quite a few cross-referencing errors. Sharing it so that (a) we can see what the dashboard does with it, and (b) to give some visibility of progress, for feedback. There's a lot of detail to be sorted, and of course when we see how the new spec has landed, there will be scope for reorganising the material into a more logical order.

Fix #2214

Issue #2148 closed #closed-2148

23 Oct at 08:21:33 GMT

Link: https://github.com/qt4cg/qtspecs/issues/2148

fn:base-uri: Raise errors?

Issue #2263 created #created-2263

22 Oct at 14:57:45 GMT

Link: https://github.com/qt4cg/qtspecs/issues/2263

Named record types: TODOs

Resulting from the conversation in #2101:

The following records have not been named yet:

(: return types :)
fn:divide-decimals
fn:xsd-validator
array:of-members
array:members

(: for parameter types as well? :)
fn:sort-by

We should add at least one example that demonstrates the usage of a built-in-record constructor.
@line-o suggested we could have records for all options.