maintainers

[gromgull]

I believe rdflib-json is essentially maintaner free at the moment.

I have no great interest in json-ld and no time - would anyone who has submitted PRs recently be interested in being made maintainers?

@ksahlmann, @RinkeHoekstra , @hsolbrig ?

[hsolbrig]

I have a strong interest in json-ld, rdflib and python. A problem, however, is that there are two potentially overlapping libraries -- this and pyld. Will be happy to pitch in, but I'd like to understand the relationship between these two before we proceed. (Will cross-post an issue on that side)

[shafaypro]

Recently, my team has been working on the realtime conversions for the Linkage based mapping, therefore, I might be interested in improving and understanding the missing functionalities regarding JSON, jsonld, and RDF XML conversion.

[RinkeHoekstra]

Unfortunately I do not have bandwidth myself to act as maintainer, but I will ask around. @albertmeronyo, @pyvandenbussche ?

The link with pyld is indeed something that needs to be considered. PyLD is for parsing and manipulating JSON-LD documents, where RDFLib is for interacting with RDF graphs. Ideally we'd have PyLD be a parser plugin for the RDFLib library.

This would bring JSON-LD 1.1 parsing and compacting/expansion and framing algorithms to RDFLib.

Currently PyLD can (only) produce an NQuads serialization of JSON-LD documents: parsing that into RDFLib drops any namespace prefix bindings defined in the context of the JSON-LD document. These will have to be preserved somehow. I don't know enough about PyLD whether that's easy to do.

[niklasl]

I'm terribly sorry for basically abandoning this project for quite some time! I'm working to allot time to it again, and would love to give access (including to PyPI) to anyone who is prepared to assist. Alas I can make no promises (family and work taking precedence), but work intersects enough here to give me some hope of picking up where I left off.

I've been pondering if or how to leverage pyld (at least for compaction, as that is the most complex part to update to 1.1), and will attempt to delineate some options as I go further.

[niklasl]

This library came about as I reworked an earlier effort on RDF-in-JSON (Gluon) and funneled that into working on JSON-LD 1.0. This then became one of the required independent implementations, alongside pyld, for "ratifying" making JSON-LD 1.0 become a W3C REC.

As pyld is isolated to the JSON-handling part, this library could be reworked as an adapter of that. It could make sense to keep this as an independent implementation for consuming JSON-LD, i.e. expanding it and directing the results directly into RDFLib. Or reduce it to just consume already expanded JSON-LD and delegate to pyld for all the context-driven JSON wrangling back and forth. Although I'd be inclined to create an rdflib-pyld adapter for such a reduced implementation.

I would like to have a go at implementing independent JSON-LD 1.1 support for at least expansion in this library, to handle RDFLib ingestion (initial support for scoped contexts and @nest has been simple enough so far). I consider including that in a 0.5.x release series, to attempt support of at least one known use case (described in #54).

Given my on-and-off time I need to decide which of these somewhat diverging directions (reduce and delegate to pyld, or level up this code) I should invest in.

(It should be noted that if you can cope with the overhead, you can use this library and pyld together today, by letting pyld handle the JSON-LD I/O, and sending serializations of its expanded form into rdflib.Graph(data=jsonld_string, format='application/ld+json') and conversely jsonld_string = graph.serialize(format='application/ld+json') and let pyld compact from that jsonld_string.)

[jpmccu]

What would it take for us to get a new release? It's been 5 years, and I've been on the development release for 3 of them. Of course, RDFlib itself is similarly delayed. These are pretty key projects (in my mind) to have stagnate.

[niklasl]

I pondered the value of a 0.4.1, but as it stands, I'd say there's a bit more than a patch in master, so I'd rather cut a 0.5.0 (since it signals a bit more than patching).

However, since 1.1 is around the corner, and I have just added support for some features (se #74 ), we might consider cutting a 0.5.0 from that state instead.

I'd love for any help in assessing the quality of both master and #74 to determine which is more viable to base a release on. (And if anyone wants to be a co-maintainer of rdflib-jsonld @PyPI, drop me a line.)

[ashleysommer]

Hi @niklasl @RinkeHoekstra @hsolbrig

I'm the maintainer and release-manager for RDFLib/pySHACL and RDFLib/OWL-RL.

rdflib-jsonld is used in both of those projects, and is an integral part of the RDF-in-Python ecosystem.

Outside of RDF, I am a core-developer on Sanic, and project owner for Sanic-CORS, Sanic-Plugins-Framework, Sanic-Restplus, and others.

I'd like to put up my hand to help with maintaining and releasing rdflib-jsonld, as well as helping with project administration, issue triaging, etc.

[niklasl]

@ashleysommer @hsolbrig Thank you for your willingness to help! I've added you both to the RDFLib organization, which should give you access to these repositories.

I also added ashleysommer to rdflib-jsonld @PyPI. (@hsolbrig let me know if you want to be added there too.)

@jimmccusker are you also interested in any or both of these roles?

My current aim is to get a 0.5.0 ready with #74 merged, to let rdflib-jsonld consume the main new JSON-LD 1.1 features. Not necessarily feature-complete nor bug-free, but usable enough (for some definition of "enough"). Please assess the state of that as it progresses.

If anyone is up for a 0.4.1 maintenance release from current master until that happens, let me/us know if it is in a capable state for that as it stands.

[hsolbrig]

Probably wouldn't hurt to add me as well...

[jpmccu]

If you have a healthy set of maintainers (and it seems that you're getting plenty), I'm happy to leave it at that. But if things fall down again, let me know. I would appreciate a maintenance release, as we have been deploying production code against the github master for quite a while.

Thanks, Jim

On Wed, Feb 19, 2020 at 12:14 PM Harold Solbrig notifications@github.com wrote:

Probably wouldn't hurt to add me as well...

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/RDFLib/rdflib-jsonld/issues/68?email_source=notifications&email_token=AAETCEJYO3YB57BP6GY5IXTRDVSGRA5CNFSM4IT3AZMKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEMIUONQ#issuecomment-588334902, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAETCEK3CNQRZC2LOBGC35TRDVSGRANCNFSM4IT3AZMA .

-- Jim McCusker

Director, Data Operations Tetherless World Constellation Rensselaer Polytechnic Institute mccusj2@rpi.edu mccusj@cs.rpi.edu http://tw.rpi.edu

[niklasl]

I just released rdflib-jsonld 0.5.0 from current master. I hope this works out for everyone. If not, @nicholascar may be able to assist, as I have very limited capacity (due to an illness in my family).

[nicholascar]

@jimmccusker @hsolbrig @RinkeHoekstra @shafaypro: are any of you able to help with adding support for JSON-LD 1.1 that is started here so we- RDFlib core maintainers - can then merge this package into the main RDFlib package? If so, please contact me or the rdflib-dev mailing list. We have a couple of volunteers but likely need more.

[shafaypro]

HI @nicholascar Sorry for the late response, Sure I am up for it.

[nicholascar]

@shafaypro we’ve already merged rdflib-jsonld functionality into rdflib, so this package is now redundant. However, we’ve still not implemented all the features of JSON-LD 1.1.

If you are keen to help, please could you review the Issues and Pull Requests in rdflib to do with JSON-LD (they aren’t all tagged well so please search!) and then see if you can work on getting RDFlib to pass all the W3C’s JSON-LD 1.1 tests?

There is growing dependence on JSON-LD for much RDF work so I wouldn’t be surprised if helpers appear if they see positive action here! If you were to take the lead and summarise the state of implementation, that would be the most valuable service!