[Announcement] PyLD v3.2.0

[Miel Vander Sande]

Hi all, 

Just wanted to let you know that PyLD released v3.2.0 with JSON-LD 1.1 API test-suite compliance for 'fromRDF()' and 'expand()' and extensions to the document loaders. It's available at https://pypi.org/project/PyLD/.

For details, consult the github release or the changelog.

Originally, we planned full JSON-LD 1.1 API test-suite compliance, but as many 'toRdf()' cases involve nquads serialization, it makes more sense to bundle these with the introduction of rdflib in 4.0.0.

What's next?

- v3.3.0 with a local file document loader and a CLI.

- v4.0.0 with a switch of the internal datamodel to rdflib, an rdflib compatible API (with < 4.0.0 compatibility mode), full JSON-LD 1.1 API test-suite compliance and RDF Dataset canon test-suite compliance (including RDFC1.0)

Best

--

Miel Vander Sande Data Architect meemoo vzw | Ham 175, 9000 Gent | https://www.meemoo.be/ t +32 9 298 05 01 | m +32 492 83 21 29 Op de hoogte blijven?  Schrijf je in voor onze nieuwsbrief!

[Nicholas Car]

Thanks Miel and hi everyone,

We have an RDFLib 8.0.0 release coming soon which contains a bunch of things not relevant to JSONLD but it does also include complete deprecation of public-facing ConjunctiveGraph and use of Dataset. This will be relevant here.

There are still PRs coming in like #3515 that are fixing aspects of Dataset so please consider if there are any Dataset-related things you need that you might be able to get in before this 8.0.0 release which I hope to have done by the end of August.

Cheers, Nick

--
http://github.com/RDFLib
---
You received this message because you are subscribed to the Google Groups "rdflib-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rdflib-dev+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/rdflib-dev/CAHeRLWtRBjSpuCXGX_PQpy5nG-6hKBeF7nwHNnPh939%2BbWH0Uw%40mail.gmail.com.

[Miel Vander Sande]

Hi Nick,

There are a couple wrt nquads/ntriples parsing/serialization that I'd like to see included because we currently mitigate them in PyLD:

- RDFC1.0-compliant _quote_encode()https://github.com/RDFLib/rdflib/pull/3431

- omitting the xsd:string datatype by default when serializing strings: https://github.com/RDFLib/rdflib/pull/3433

- serializing without normalizing literals and without having to use the NORMALIZE_LITERALS global. (no PR yet)

These are not Dataset related, but they all minor improvements with code ready to go in PyLD

Best,

Miel

Op di 18 aug 2026 om 03:27 schreef Nicholas Car <ni...@kurrawong.ai>:

To view this discussion visit https://groups.google.com/d/msgid/rdflib-dev/yL-l9YDO3elEPMpoKR0m7_dobOMaxscedLPvKV1TdlOCQzDspiCNq671hwTNMHViAcBmHkybNpxgNDdVL__DV_4zlrNyxKT3tH-hWCJ37J4%3D%40kurrawong.ai.

[Nicholas Car]

Hi Miel,

Ok, I see your PR there but please move it out of draft soon and either add PRs or Issues for the other things you want. I’ll tag then as for v8 release.

Cheers, Nick

-------- Original Message --------

To view this discussion visit https://groups.google.com/d/msgid/rdflib-dev/CAHeRLWthT3BOXq%2B7XYz2jxr76ivjVJ4ba5ddBWn1a%2BZD1PAKhA%40mail.gmail.com.

[Miel Vander Sande]

Hi Nick,

(I forgot to send this 😵‍💫)

I had a look and it seems normalization and datatypes of literals and serializers as a whole need a wider discussion/cleanup before adding my last two points. There are currently too many tests failing because of comparison issues.

Moving literal normalization to NormalizedLiteral, restricting it to Literal.normalize() or simply removing it completely would be a possible starting point. In what version would something like that be discussed?

Best,

Miel

Op di 18 aug 2026 om 12:10 schreef Nicholas Car <ni...@kurrawong.ai>:

To view this discussion visit https://groups.google.com/d/msgid/rdflib-dev/adu9S71XuWvtArIunOMvgN7Cx9L_9pAibz3_LZUwr66CYFh3-jmy9IYriKYk3p5LZGFXiUad-CpoA5Nb4npmZT3CELJZoqmCE56tuuMG32Y%3D%40kurrawong.ai.

[Miel Vander Sande]

Hi all,

It's me again :) Just wanted to let you know that PyLD released v3.3.0 with a CLI and the ability to pass custom identifier issuers. It's available at https://pypi.org/project/PyLD/.

For details, consult the github release, the changelog or the docs.

What's next? v4.0.0 with a switch of the internal datamodel to rdflib, an rdflib compatible API (with < 4.0.0 compatibility mode), full JSON-LD 1.1 API test-suite compliance and RDF Dataset canon test-suite compliance (including RDFC1.0). After that, probably going in maintenance mode for a while😴

Cheers,

Miel

[Nicholas Car]

Hi Miel,

I've done a whole bunch of tidy-up tasks and Edmond's soon (next week?) going to merge in many PRs that are good to go, including all the dependabot usual noise, so we should be able to make a v8 release shortly. 

I've merged your PR below in, so if you're happy enough for now, you can build the PyLD 4.x on RDFLib 8.

Any changes you want regarding improved literal normalisation can be an 8.1 easily enough.

Cheers, Nick

To view this discussion visit https://groups.google.com/d/msgid/rdflib-dev/CAHeRLWsj0sGhnJ9px_TF_P_BhmZXZ6TuNBAqpUBwvGU6QYZ_nw%40mail.gmail.com.

[Miel Vander Sande]

Hi Nick,

Ah neat! Still plenty of time to move PyLD to rdflib 8, so let's do that.  

Probably github discussion are a good place to talk about the future of literal normalization; I'll open a thread there.

Best,

Miel

Op wo 2 sep 2026 om 14:52 schreef Nicholas Car <ni...@kurrawong.net>:

To view this discussion visit https://groups.google.com/d/msgid/rdflib-dev/DeclrPaC0S1nMalJDjDan-Xgik3pwYY0YJ1XiDZ_k__cmFOHUB8vASH6g5uWbna5I96C2_52nCACnXP74dIJTEOMdSZKsNkzb9gS7tDiFH4%3D%40kurrawong.net.