w3c / rdf-star

RDF-star specification
https://w3c.github.io/rdf-star/
Other
120 stars 23 forks source link

Squashed commit of the following: #205

Closed gatemezing closed 2 years ago

gatemezing commented 3 years ago

commit dd8b43515e0c9bbe55723543967303f441dde22c Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Mon Sep 13 22:47:14 2021 +0200

solving parsing errors manifest

commit 08aee86852e8af4788fadfc515b217f0bf07f9bd Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Fri Sep 10 09:32:23 2021 +0200

solving conflict

commit 691278ca402efe5c45e7f398f3a03d214912ae24 Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Thu Sep 9 14:28:41 2021 +0200

fixing ttl conflict

commit bdfa01e5fa430cbbeb41511479d01d83c3f134c0 Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Thu Sep 9 14:21:38 2021 +0200

next file fix conflict

commit 22f1f734ea75d34599a8e1649a7df08db7eec178 Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Thu Sep 9 14:11:08 2021 +0200

new fix conflict file

commit ec645e090bfb4843822ae41a96b9f9914f15ceef Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Thu Sep 9 14:04:21 2021 +0200

solving conflict files

commit 94ce29d5826026ed641880309e11953140c863ce Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Thu Sep 9 10:14:35 2021 +0200

alignment works wrt tp @andy suggestions

commit 2899371e0ac60f43e0eb7d1963a8dce3a37c6d19 Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Sat Sep 4 00:28:55 2021 +0200

updating tests manifest jsonld files

commit 9235e8de09e7486d2add9b2802a63a8f61235859 Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Fri Sep 3 15:52:19 2021 +0200

reviewing missing prefixes

commit f22befcd709fbf2db515ef55af6a64d79715015a Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Fri Sep 3 15:48:55 2021 +0200

alignment continues

commit 26827298f259a0b00ef7cbc0bddfe7de99dc43a3 Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Fri Sep 3 15:01:55 2021 +0200

aligning test manifest

commit 65d580c5e93d120a51918b916a197ab97e27478c Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Fri Sep 3 14:43:46 2021 +0200

removed individual names + trying to align stuff

commit bd17f5f1dbb41291382dfbf6dfcf607c999f076d Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Thu Jul 22 19:17:23 2021 +0200

added more metada in other manifest files

commit 6ec36cf6e7f13f0f0658a1121893ce70c145fc62 Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Thu Jul 22 19:06:17 2021 +0200

removed  testsuite license and used PREFIX.

commit 6a7677521e9a8e911bc78bac9fde6eb81698826b Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Sun Jul 18 15:35:28 2021 +0200

more metadata

commit e2fff0e26dec75fd12d09da06261514b8b75d8b2 Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Sun Jul 18 15:27:38 2021 +0200

added metadata turtle manifest

commit ff814c79424e9b0af53197fd85afc4b6d4c36ba5 Author: Ghislain Atemezing ghislain.atemezing@mondeca.com Date: Sun Jul 18 14:27:54 2021 +0200

suggest metadata in manifest turtle file
afs commented 3 years ago

Your personal repo gatemezing/rdf-star has both "master" and "main".

The squash is against "master" which itself is 25 commits behind w3c:main.

RDF-Star repo is "main". Grabbing the PR branch, I see:

[893e6d6] ghislain.atemezing@mondeca.com (27 hours ago) - Squashed commit of the following:
[5484872] pierre-antoine@w3.org (9 weeks ago) - today's minutes
[4c5bd52] pierre-antoine@w3.org (3 months ago) - today's minutes

The squash itself has problems. We now have a commit with log message "Squashed commit of the following:" which isn't very helpful. The body contains the day-to-day activity messages.

A PR should have a first line gives a clear indication of what it is - e.g. "Add metadata to manifest files". The rest of the body of the log message can add details if you want but those details should refer to the change not e.g. "alignment continues".

It's still a bit messy - alignment is wrong in a couple of places and when I checked by taking the diff #205 of moving changes around) and got this from "git apply" - the ones at the end look serious even if the PR can merge.

The test suite needs manual checking to make sure nothing has been lost but better would be a clean PR.

/home/afs/Desktop/rs-205.diff:20: trailing whitespace.
            "@id": "rdfs:label", 
/home/afs/Desktop/rs-205.diff:121: trailing whitespace.
      "@id": "rdfs:label", 
/home/afs/Desktop/rs-205.diff:212: trailing whitespace.
      "@id": "rdfs:label", 
/home/afs/Desktop/rs-205.diff:668: trailing whitespace.
PREFIX rdf:    <http://www.w3.org/1999/02/22-rdf-syntax-ns#> 
/home/afs/Desktop/rs-205.diff:669: trailing whitespace.
PREFIX rdfs:   <http://www.w3.org/2000/01/rdf-schema#> 
error: patch failed: tests/nt/syntax/manifest.ttl:79
error: tests/nt/syntax/manifest.ttl: patch does not apply
error: patch failed: tests/semantics/manifest.ttl:1
error: tests/semantics/manifest.ttl: patch does not apply
error: patch failed: tests/trig/eval/manifest.ttl:23
error: tests/trig/eval/manifest.ttl: patch does not apply
error: patch failed: tests/turtle/eval/manifest.ttl:23
error: tests/turtle/eval/manifest.ttl: patch does not apply
gatemezing commented 3 years ago

Your personal repo gatemezing/rdf-star has both "master" and "main".

The squash is against "master" which itself is 25 commits behind w3c:main.

RDF-Star repo is "main". Grabbing the PR branch, I see:

[893e6d6] ghislain.atemezing@mondeca.com (27 hours ago) - Squashed commit of the following:
[5484872] pierre-antoine@w3.org (9 weeks ago) - today's minutes
[4c5bd52] pierre-antoine@w3.org (3 months ago) - today's minutes

The squash itself has problems. We now have a commit with log message "Squashed commit of the following:" which isn't very helpful. The body contains the day-to-day activity messages.

A PR should have a first line gives a clear indication of what it is - e.g. "Add metadata to manifest files". The rest of the body of the log message can add details if you want but those details should refer to the change not e.g. "alignment continues".

It's still a bit messy - alignment is wrong in a couple of places and when I checked by taking the diff #205 of moving changes around) and got this from "git apply" - the ones at the end look serious even if the PR can merge.

The test suite needs manual checking to make sure nothing has been lost but better would be a clean PR.

/home/afs/Desktop/rs-205.diff:20: trailing whitespace.
            "@id": "rdfs:label", 
/home/afs/Desktop/rs-205.diff:121: trailing whitespace.
      "@id": "rdfs:label", 
/home/afs/Desktop/rs-205.diff:212: trailing whitespace.
      "@id": "rdfs:label", 
/home/afs/Desktop/rs-205.diff:668: trailing whitespace.
PREFIX rdf:    <http://www.w3.org/1999/02/22-rdf-syntax-ns#> 
/home/afs/Desktop/rs-205.diff:669: trailing whitespace.
PREFIX rdfs:   <http://www.w3.org/2000/01/rdf-schema#> 
error: patch failed: tests/nt/syntax/manifest.ttl:79
error: tests/nt/syntax/manifest.ttl: patch does not apply
error: patch failed: tests/semantics/manifest.ttl:1
error: tests/semantics/manifest.ttl: patch does not apply
error: patch failed: tests/trig/eval/manifest.ttl:23
error: tests/trig/eval/manifest.ttl: patch does not apply
error: patch failed: tests/turtle/eval/manifest.ttl:23
error: tests/turtle/eval/manifest.ttl: patch does not apply

I did not see this comment before my last push 10h ago generating manifest.html. I'll see what I can do based on your suggestions

afs commented 2 years ago

This https://github.com/w3c/rdf-star/pull/198#issuecomment-921212395 seems to have got lost:

Running the manifest.ttl files, I get suite names in a mix of English, French and Spanish. The code looks for one rdfs:label.

Suggestion: Use rdfs:label with non-lang-tag text. Use skos for the language tagged literals.

gkellogg commented 2 years ago

This #198 (comment) seems to have got lost:

Running the manifest.ttl files, I get suite names in a mix of English, French and Spanish. The code looks for one rdfs:label. Suggestion: Use rdfs:label with non-lang-tag text. Use skos for the language tagged literals.

It doesn't look like any of the JSON-LD manifests actually use a language map for either "label", which they should to be consistent with the Turtle. If they did, then the code to generate the HTML could potentially make use of this to select the English version of the labels. But, separating to rdfs:label and skos:altLabel or similar might help.

pchampin commented 2 years ago

This was discussed during today's call: https://w3c.github.io/rdf-star/Minutes/2021-10-15.html#a01

afs commented 2 years ago

I get quite a few conflicts when I tried to apply the state of this PR as a diff file. The easiest way I can think of to fix these is to start with the current repo main branch. I don't think there is anything major bu a start point of the current repo reduces the risk of undoing material that now current. It because there have been changes to the current version since this was forked.

  1. There was a renaming "embed-annotation" has become "quoted-annotation" inline with the group decision on terminology "quoted triple".
  2. JSON-LD: comment about language maps
  3. Use rdfs:label for labels and skos:prefLabel for the language tagged literals.
  4. HTML files need regenerating to be inline with state of the current repo main branch

3 may address 2.

We can sort the mechanics out after if we have a clean PR from the current state covering 1 and 3.

gatemezing commented 2 years ago

Thanks @afs. I've done yet another PR here Metadata #217. HTH

TallTed commented 2 years ago

@gatemezing

  • Use rdfs:label for labels and skos:prefLabel for the language tagged literals.

As I read @afs' comment, rdfs:label values should have no lang-tag, en or otherwise.

As I read #217, you've put en-tagged literals in rdfs:label and all other-lang-tagged literals into skos:prefLabel.

I would recommend echoing the en-tagged literals in skos:prefLabel -- because skos:prefLabel should hold the preferred label for each language.

afs commented 2 years ago

@TallTed - yes.

afs commented 2 years ago

@gatemezing - If this PR is no longer active, could you close it please?

gatemezing commented 2 years ago

@afs - sure! Makes sense!