Adds quoted triples and asserted triples from RDF-star CG report #32

gkellogg · 2023-04-10T20:52:26Z

Add quoted and asserted triple definitions.
Add "classic" and "full" conformance levels.

Some license taken to make terminology more native to RDF, rather than a modification of RDF.

Fixes #23.
References #34 for potentially adding additional terminology.
Relates to w3c/rdf-star-wg#33 and w3c/rdf-star-wg#23

Preview | Diff

spec/index.html

Note, this updates the raw EBNF to remove production numbers, and auto-generate the production numbers in the HTML serialization. It is dependent on w3c/rdf-concepts#32.

afs · 2023-04-13T12:10:07Z

This is prejudging and splitting the discussion of w3c/rdf-star-wg#33. There are changes out for other specs.

"triple" is the defined term. Concepts has "generalized triple" which is an extension, not a restriction so the terminology "triple" in the semantics works. But if there are triples that are restrictions, the reader may reasonably ask "what happens for X ro Y triples?". "complex triple" means nothing to me. The use of special case terminology on RDF concepts detracts from the foundational nature of a triple.

As a triple (an item with three parts, etc etc) this is not any more "complex" that any other variation.

We don't have special names for triple-with-literal or triple-with-subject-blankNode.

rdf-canon is already explaining it is URDNA2015 which is RDF 1.1 timeframe.

Regardless of this PR, it will note the situation which makes it the better place to explain that it, quite reasonably, targets RDF 1.1 while taking the canonicalized N-Quads syntax.

spec/index.html

afs · 2023-04-13T15:15:31Z

spec/index.html

+    <li><dfn class="no-export lint-ignore">Full</dfn> conformance
+      supports <a>complex graphs</a> or <a>complex datasets</a>,
+      and allows the use of <a>quoted triples</a>.
+      Such input syntaxes include
+      [[RDF12-N-TRIPLES]], [[RDF12-N-QUADS]], [[RDF12-TURTLE]], and [[RDF12-TRIG]].</li>


Not needed - it is "conformance". "complex" is a judgement.

I'm gong to add a note in SOTD that explains that these terms are tentative placeholders, some of which may not continue going forward.

Changed to "containing quoted triples".

spec/index.html

gkellogg · 2023-04-13T22:42:36Z

I added something to the SOTD; at this point I'd rather not remove terminology until there is more discussion. We can create some new sections, possibly informative, to move around definitions as a next step. It also needs to be rebased on recent commits to main, which may make some existing comments obsolete requiring them to be re-considered, but we can wait on that until we're closer to concensus.

spec/index.html

afs · 2023-04-16T10:04:46Z

at this point I'd rather not remove terminology until there is more discussion.

(referring to plain/complex)

The introduced of this terminology, which is not in the CG report, is of significant size, does not reflect the discussion and concerns of the telecon (2023-04-13). The terminology is described as "adapted from the CG report" - that's misleading.

I'd rather not put terminology into the drafts that is highlighting one position over another. In this case, another possibility is not having top-level terminology or having terminology to describe one situation.

(referring to w3c/rdf-star-wg#19, w3c/rdf-star-wg#23)

Put in issue markers without opinion or direction.

gkellogg · 2023-04-16T16:46:15Z

I think an issue marker for everything other than “quoted triple” is a reasonable interim solution.

Note, this updates the raw EBNF to remove production numbers, and auto-generate the production numbers in the HTML serialization. It is dependent on w3c/rdf-concepts#32.

spec/index.html

afs · 2023-04-18T07:49:08Z

spec/index.html

+      as the subject.</p>
+
+    <figure id="fig-quoted-triple">
+      <a href="quoted-triple.svg">


(There is an access problem seeing this diagram in the preview.)

Yes, a shortcoming of PR Preview. But, you can either see it in the file view, or using the Google drawing referenced in the comment: https://docs.google.com/drawings/d/1I_QxbUgnQXumXzb8c0WNJHIQ-mtRs2S80dDG6i9aOD8.

spec/index.html

afs · 2023-04-18T08:00:36Z

spec/index.html

@@ -427,14 +498,26 @@ <h2>RDF Documents and Syntaxes</h2>
    <em>RDF 1.2 Concepts and Abstract Syntax</em>,
    but can conform to such other specifications that normatively
    reference terms defined here.</p>
+
+  <p>This specification establishes two conformance levels:</p>


The WG has not agreed that nor the words used. It's WIP.

"Profiles" has also been mentioned.

Also - "This specification" would be the semantics.

At a minimum this text needs a note that it is tentative.
As concepts does not need this language, and the dfn's are unstable, removing it for now would be better.

See note added below. My opinion is that if it relates to the data model/abstract syntax, it should be described in Concepts rather than Semantics. I suggested elsewhere that the text could just be removed from this PR and taken up another time. But, the notion that there may be some variation allowed or supported is probably worth signaling sooner rather than later.

spec/index.html

afs · 2023-04-18T08:09:00Z

spec/quoted-triple.svg

I'm note sure this diagram works. The two lines are different things.

spec/index.html

Note, this updates the raw EBNF to remove production numbers, and auto-generate the production numbers in the HTML serialization. It is dependent on w3c/rdf-concepts#32.

* Add issue marker for #14 related to potential additional terminology. * Add "classic" and "full" conformance levels. * Add quoted and asserted triple definitions.

gkellogg · 2023-04-20T21:44:40Z

Squashed and rebased to address conflicts and simplify history.

Co-authored-by: Olaf Hartig <[email protected]> Co-authored-by: Ted Thibodeau Jr <[email protected]>

pfps · 2023-04-21T00:09:55Z

Here are the relevant definitions for RDF graphs and triples from the PR.

An RDF graph is a set of RDF triples.
An RDF triple consists of three components:

the subject, which is an IRI, a blank node, or a quoted triple

the predicate, which is an IRI

the object, which is an IRI, a literal, a blank node, or a quoted triple

An asserted triple is an RDF triple that is an element of an RDF graph.
A quoted triple is an RDF triple that is used as the subject or object of another triple.
In an RDF graph, a triple may occur as either a quoted triple, an asserted triple, or both.

The above definitions do not forbid triples that contain themselves.

Neither cycles of quoted triples nor creation is defined so the statement below is meaningless.

However, by this definition, cycles of quoted triples cannot be created.

afs · 2023-04-21T07:57:35Z

The above definitions do not forbid triples that contain themselves.

We ought to clarify to make sure this isn't a possible reading.

@pfps - do you have a suggestion?

A temporary measure, for FPWG, is to say (informally) "Cycles are not permitted".
Or (temporary) note that all RDF syntaxes can not describe RDF data with cycles.

pchampin · 2023-04-21T14:04:00Z

I agree that the current definition of 'RDF triple' leaves the door open for "self-containing" triples, and that the notion of "cycles of quoted triples" is confusing...

It seems to me that the definition in the CG report is a better way to avoid "self-containing" triples (@pfps do you agree?), but it relies on the distinction between "RDF triple" and "RDF-star triple"... so in order to reuse this phrasing, we would need a distinction between "simple/complex" triples... :-/

pfps · 2023-04-21T14:08:13Z

@pchampin Yes the CG definition rules out triples that contain themselves, so it is different than the definition in this PR. Whether this is better depends on whether one wants to allow such triples.

spec/index.html

hartig · 2023-04-21T20:52:51Z

It is possible to adapt the definition of the CG report as follows.

An RDF triple is a 3-tuple defined recursively as follows:

if s is an IRI or a blank node, p is an IRI, o is an IRI, a blank node or a literal, then (s, p, o) is an RDF triple;
if t and t' are RDF triples, s is an IRI or a blank node, p is an IRI, o is an IRI, a blank node or a literal, then (t, p, o), (s, p, t) and (t, p, t') are RDF triples.

(Since I am just writing this on my phone, I am posting this just as text rather than as a directly commit-able suggestion. However, I can create such a suggestion comment later (tomorrow) if you agree with this direction for the definition.)

hartig · 2023-04-22T16:38:53Z

@gkellogg I have my proposed change properly written in my local copy of the branch now. However, I cannot put it here as a directly commit-able suggestion because it also changes line 440 ("<p>An <dfn data-local-lt="triple">RDF t...") which is too many lines away from the changes in this PR and, thus, I cannot include it when selecting lines in the "Files changed" tab of this PR. As an alternative, I could simply push a commit with the change to the branch of this PR. Would that be okay with you? Another option is to merge this PR now (in which case I would approve it), and then I create another PR with the change. Which one do you prefer?

gkellogg · 2023-04-22T17:10:59Z

Yes, I would just commit to the branch. You can always clarify with a comment after the fact, if it seems necessary.

…t contain themselves

hartig · 2023-04-22T19:42:32Z

Thanks @gkellogg! I have pushed the commit (f24e320).

gkellogg added needs discussion Proposed for discussion in an upcoming meeting spec:substantive Change in the spec affecting its normative content (class 3) –see also spec:bug, spec:new-feature labels Apr 10, 2023

gkellogg requested review from afs, pchampin and hartig April 10, 2023 20:52

TallTed suggested changes Apr 10, 2023

View reviewed changes

TallTed reviewed Apr 11, 2023

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

TallTed reviewed Apr 11, 2023

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

gkellogg mentioned this pull request Apr 11, 2023

Support for quoted triples. w3c/rdf-n-triples#21

Merged

gkellogg mentioned this pull request Apr 13, 2023

Use Case: Concept of Graph/Dataset without quoted triples w3c/rdf-star-wg#33

Closed

afs mentioned this pull request Apr 13, 2023

work with antoine and others to come up with a proposal for weak and strong compliance w3c/rdf-star-wg#19

Closed

afs requested changes Apr 13, 2023

View reviewed changes

afs reviewed Apr 13, 2023

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

hartig reviewed Apr 14, 2023

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

gkellogg mentioned this pull request Apr 15, 2023

work on conformance proposal w3c/rdf-star-wg#23

Closed

afs reviewed Apr 16, 2023

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

gkellogg changed the title ~~Adds quoted triples and related terminology from RDF-star CG report~~ Adds quoted triples and asserted triples from RDF-star CG report Apr 16, 2023

gkellogg requested a review from afs April 16, 2023 21:14

afs requested changes Apr 18, 2023

View reviewed changes

gkellogg commented Apr 18, 2023

View reviewed changes

spec/index.html Show resolved Hide resolved

TallTed suggested changes Apr 18, 2023

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

spec/index.html Show resolved Hide resolved

spec/index.html Outdated Show resolved Hide resolved

spec/index.html Outdated Show resolved Hide resolved

spec/index.html Outdated Show resolved Hide resolved

afs approved these changes Apr 20, 2023

View reviewed changes

hartig requested changes Apr 20, 2023

View reviewed changes

gkellogg removed the needs discussion Proposed for discussion in an upcoming meeting label Apr 20, 2023

* Minor grammar, spelling and simplifications.

43ca048

* Add issue marker for #14 related to potential additional terminology. * Add "classic" and "full" conformance levels. * Add quoted and asserted triple definitions.

gkellogg force-pushed the quoted-triples branch from 7dc51c4 to 43ca048 Compare April 20, 2023 21:43

Updates from Olaf and Ted

06ddd30

Co-authored-by: Olaf Hartig <[email protected]> Co-authored-by: Ted Thibodeau Jr <[email protected]>

gkellogg commented Apr 21, 2023

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

Revert expanded definition of nodes in the graph.

97ca502

gkellogg requested a review from hartig April 21, 2023 20:41

new (recursive) definition of 'RDF triple' that rules out triples tha…

f24e320

…t contain themselves

hartig approved these changes Apr 22, 2023

View reviewed changes

gkellogg merged commit 1d02fff into main Apr 27, 2023

gkellogg deleted the quoted-triples branch April 27, 2023 22:12

hartig mentioned this pull request Apr 29, 2023

Reordering of paragraphs and notes in Sec.3.1 (Triples) #35

Merged

pchampin mentioned this pull request Dec 16, 2024

add section about 'unstar' mapping #115

Merged

afs mentioned this pull request Jan 16, 2025

Adds selected issue markers and removes those already closed. #130

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Adds quoted triples and asserted triples from RDF-star CG report #32

Adds quoted triples and asserted triples from RDF-star CG report #32

gkellogg commented Apr 10, 2023 •

edited by pr-preview bot

Loading

afs commented Apr 13, 2023 •

edited

Loading

afs Apr 13, 2023

gkellogg Apr 13, 2023

gkellogg Apr 16, 2023

gkellogg commented Apr 13, 2023

afs commented Apr 16, 2023

gkellogg commented Apr 16, 2023

afs Apr 18, 2023

gkellogg Apr 18, 2023

afs Apr 18, 2023

gkellogg Apr 18, 2023

afs Apr 18, 2023

gkellogg commented Apr 20, 2023

pfps commented Apr 21, 2023 •

edited

Loading

afs commented Apr 21, 2023

pchampin commented Apr 21, 2023

pfps commented Apr 21, 2023

hartig commented Apr 21, 2023

hartig commented Apr 22, 2023

gkellogg commented Apr 22, 2023

hartig commented Apr 22, 2023

Adds quoted triples and asserted triples from RDF-star CG report #32

Adds quoted triples and asserted triples from RDF-star CG report #32

Conversation

gkellogg commented Apr 10, 2023 • edited by pr-preview bot Loading

afs commented Apr 13, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gkellogg commented Apr 13, 2023

afs commented Apr 16, 2023

gkellogg commented Apr 16, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gkellogg commented Apr 20, 2023

pfps commented Apr 21, 2023 • edited Loading

afs commented Apr 21, 2023

pchampin commented Apr 21, 2023

pfps commented Apr 21, 2023

hartig commented Apr 21, 2023

hartig commented Apr 22, 2023

gkellogg commented Apr 22, 2023

hartig commented Apr 22, 2023

gkellogg commented Apr 10, 2023 •

edited by pr-preview bot

Loading

afs commented Apr 13, 2023 •

edited

Loading

pfps commented Apr 21, 2023 •

edited

Loading