Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
1 changed file
with
141 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,141 @@ | ||
| ## Telecon 2018-02-26 Minutes | ||
|
|
||
| * Participants:Kat, LarsG, Gregg, Tom, Dimitris, EricP, Andra, Lucas, Harold, Jeremie | ||
| * Regrets: | ||
|
|
||
| Call in Jitsi # https://meet.jit.si/ShEx | ||
| => https://zoom.us/j/959347434 | ||
|
|
||
| * Chair: Dimitris | ||
| * Scribe: Tom, Dimitris | ||
|
|
||
| last minutes:https://github.com/shexSpec/shex/blob/master/meetings/20180213-minutes.md | ||
|
|
||
| Agenda: https://github.com/shexSpec/shex/blob/master/meetings/20180226-agenda.md | ||
|
|
||
|
|
||
| Everyone switch to Zoom : https://zoom.us/j/959347434 | ||
|
|
||
| * Dimitris: Enyone has any objection to switch to Thursdays from now on? Eric cannot make it on Mondays | ||
| (no objection) | ||
| * Dimitris: Next meeting on March 15th | ||
|
|
||
| ## Community Updates | ||
|
|
||
| ### ShEx talk at https://wiki.dnb.de/display/DINIAGKIM/KIM+Workshop+2018 | ||
| * Tom: Library related workshop on json-ld/Wikidata, and will give an introduction to ShEx, will add some Wikidata examples | ||
| * Andra: Also the wikidata community is just learning about ShEx. https://www.wikidata.org/wiki/Wikidata:WikiProject_ShEx | ||
| * Eric: can you swap the order of the talks | ||
| * Tom: So it makes sense as is | ||
|
|
||
| ### Jupyter notebooks | ||
|
|
||
| * Tom: Anaconda is developing Jupyter notebook with IDE functionality and support for cells that use kernels in multiple languages. | ||
|
|
||
| http://www.zdnet.com/article/can-data-science-notebooks-get-real-jupyter-lab-releases-to-users/ | ||
|
|
||
| Notebook: https://colab.research.google.com/drive/1CB4EPj_JaXZNWkv2PHQ3Ryn5JahcstXU | ||
|
|
||
| Visualizations: https://etherpad.wikimedia.org/p/Diagrams_to_ShEx_OWL_RML_FHIR | ||
|
|
||
| Andra shows https://colab.research.google.com/drive/1_pGP63DjvvqEHSI17HzACP9en-S9iu6d | ||
|
|
||
| * Andra: (showing notebook) Two things, 1) I made a shex to python with GraphML that is actually XML. | ||
| (explining doc) using Harold's implementation | ||
| 2) using Python impl by Greg in Jypiter notebook | ||
|
|
||
| * Harold: should be connectable to the manifest. Keep me posted. I am in idle mode. | ||
| * Andra: Might ping you after this meeting. Also want to mention: maintaining GeneWiki shapes files - currently implementing | ||
| using Git submodules. Re-organizing. Example here: https://github.com/SuLab/Genewiki-ShEx/tree/e6ab1947c775bdf9a48ab2bba2114bc468571820 | ||
|
|
||
| ### Continue discussion of multiple raised issues on the spec by Harold | ||
| 79 - nodeIn appears incorrect for blank nodes:https://github.com/shexSpec/shex/issues/79 | ||
| PROPOSAL: remove blank node text from nodeIn | ||
|
|
||
| #### Issue-79 nodeIn appears incorrect for blank nodes: | ||
| * Eric: Does anyone have a use case for testing bnodes and value sets? | ||
|
|
||
| * Dimitris: Bnodes not referencable. | ||
|
|
||
| * Harold: Parser has to make additional [distinctions]. If we are going to do bnodes, will have to do work. | ||
| * Gregg: Have been opposed from the beginning. Bnodes only valid within document. | ||
| * Harold: Whole series of tests I cannot do because rdflib is religious about keeping them blank, so I | ||
| do not implement anything that. | ||
| * Dimitris: Two options: Forbid, or ignore (and treat in their own systems). | ||
| * Gregg: Should be in grammar. | ||
| * Harold: My recollection is that it is not in the grammar. If I'm not mistaken, this is the area where we have | ||
| language nodes and stems but not bnodes and stems. So I got down to parser and asked: how do I even | ||
| recognize these things? | ||
| * Gregg: Legit to use [] in a shape expression itself. | ||
| * Harold: I couldn't even parse to it. Problem: IRI/Literal/Language stem range, but nothing in here to | ||
| deal with BNode stem range which, to be honest, seems really insane. BNodes cannot just be blank, | ||
| but can also have patterns...?!? | ||
| * Dimitris: Seems like not valid grammar. | ||
| * Gregg: I see no "blank node range". | ||
| * Harold: In semantics, talks as if you can get down to blank nodes, but I couldn't find.... | ||
| * Dimitris: If in theory one could have a ShEx schema in RDF, could work: but is this not the case? | ||
| * Harold: Tried to document this. "Blank node can match any stem". Four conditions: "matches if | ||
| N is a blank node, and there is nothing about S. So the way it is written, if I have an IRI stem. We | ||
| would need "blank node stem" but crazy: do not want to pattern-match on blank nodes! | ||
| * Gregg: Agree. Need Eric to defend it. He says: some preserve it. | ||
| * Dimitris: | ||
|
|
||
| http://shex.io/shex-semantics/#values | ||
|
|
||
| * Harold: Would like to remove [] and _:xxx. Don't think it is even worth our time to get Blank Node | ||
| Stem into the grammar. No way to implement the little bit that is in here now. I see it as an error | ||
| in the 2.0 Semantic spec. Could be an issue for 2.2. But I do not want to implement that line. | ||
| Would not want it to match if it happened to start with "en" (like language). | ||
| * Gregg: +1 | ||
|
|
||
| Dimitris: everynone vote here with Thumbs up/down: https://github.com/shexSpec/shex/issues/79#issuecomment-357547513 | ||
| Dimitris: Resolved | ||
|
|
||
| ### Issue 81 (second raised by Harold) | ||
| https://github.com/shexSpec/shex/issues/81 | ||
|
|
||
|
|
||
| * Harold: In SPARQL, xsd:date not listed as a datatype but used as example in document. | ||
| Eric said: support them, but not do type inference based on them. | ||
|
|
||
| * Gregg: SPARQL grammar does use xsd:date in examples. I think there are tests for it. | ||
| * Harold: Eric concluded that: change rules to say: SPARQL 11 plus xsd:date and time. | ||
| * Gregg: Many tests uses xsd:date. | ||
| * Eric: SPARQL has a set of supported datatypes. Behavior of SPARQL is that you are | ||
| responsible for. My intention for ShEx: same behavior. My proposal: clarify proposal, | ||
| because everyone wondering why they are not in SPARQL. | ||
| * Dimitris: Eric has a proposal we can vote for: | ||
| * Dimitris: Everyone please vote on https://github.com/shexSpec/shex/issues/81#issuecomment-368446352 | ||
| * Dimitris: Resolved | ||
|
|
||
| ### Inheritance discussion | ||
|
|
||
| * Eric: No progress to report (was working on query maps). Was implementing examples | ||
| for Primer and ran into bugs. Implementation still works - tests a partition instead of an | ||
| entire neighborhood. Implementation of extension and restriction. A bit of new text in the | ||
| Primer. Wondering if Harold could work through, if I created tests "on shape expression", | ||
| could you implement to see if it makes sense to you? | ||
| * Harold: Happy to try next weekend. | ||
| * Gregg: Longer to get to it, but yes. | ||
|
|
||
| ### Other business | ||
|
|
||
| * Harold: Work for a common CLI - for invoking the ShEx interpreter? WOuld like to conform to | ||
| this as well as possible. | ||
| * Gregg: There was discussion of an API to use with test suite. | ||
| * Eric: Jose and I implement the same. | ||
| * Harold: Please point me to what you are doing. | ||
| * Eric: Will ping Jose about having a Gitter chat to nail down manifest and CLI. Not the same | ||
| as the SPARQL manifest, which has stuff that we wouldn't want into the CLI. If the manifest | ||
| format is more useful than... would we want to deviate from the SPARQL test suite if the | ||
| manifest is terser, better? If useful, I will make a proposal. | ||
|
|
||
| * Harold: What is the status of Shape Map document? | ||
| * Eric: Been awhile since eyes were on it. When I was working on Query Map (fatter | ||
| version of Shape Map). | ||
| * Harold: Ended up making wrapper for Jupyter. | ||
| * Eric: How many people in critical path? Harold, Andra, | ||
| * Gregg: me too, but gone next week | ||
|
|
||
|
|
||
|
|