Merge branch 'dev'

.github/workflows/python-app.yml

@@ -4,29 +4,27 @@ name: Python application
 
 on:
   push:
-    branches: [ "main", "dev" ]
+    branches: ["main", "dev"]
   pull_request:
-    branches: [ "main", "dev" ]
+    branches: ["main", "dev"]
 
 permissions:
   contents: read
 
 jobs:
   build:
-
     runs-on: ubuntu-latest
 
     steps:
-    - uses: actions/checkout@v3
-    - name: Set up Python 3.10
-      uses: actions/setup-python@v3
-      with:
-        python-version: "3.10"
-    - name: Install dependencies
-      run: |
-        python -m pip install --upgrade pip
-        pip install -r requirements.txt
-    - name: Test with pytest
-      run: |
-        pytest
-
+      - uses: actions/checkout@v3
+      - name: Set up Python 3.10
+        uses: actions/setup-python@v3
+        with:
+          python-version: "3.10"
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install -r requirements.txt
+      - name: Test with pytest
+        run: |
+          pytest

CHANGELOG.md

@@ -9,6 +9,14 @@ Changes to data are ignored for the purpose of this changelog.
 
 ## [Unreleased]
 
+## [2.1.0] (2023-09-13)
+
+### Added
+
+- Support for deprecated terms:
+  - Any term having `owl:deprecated  "true"^^xsd:boolean` will be present in the RDF API but omitted from the JSON API.
+  - `dcterms:replaces` is mirrored with `dcterms:isReplacedBy`
+
 ## [2.0.2] (2023-09-06)
 
 ### Added
@@ -85,7 +93,8 @@ Features:
 - A build script that compiles multiple RDF/SKOS files (terms) to one (thesaurus).
 - An HTTP server that exposes the data as RDF as well as more web-app-friendly JSON
 
-[unreleased]: https://github.com/gu-gridh/queerlit-terms/compare/v2.0.2...HEAD
+[unreleased]: https://github.com/gu-gridh/queerlit-terms/compare/v2.1.0...HEAD
+[2.1.0]: https://github.com/gu-gridh/queerlit-terms/compare/v2.0.2...v2.1.0
 [2.0.2]: https://github.com/gu-gridh/queerlit-terms/compare/v2.0.1...v2.0.2
 [2.0.1]: https://github.com/gu-gridh/queerlit-terms/compare/v2.0.0...v2.0.1
 [2.0.0]: https://github.com/gu-gridh/queerlit-terms/compare/v1.3.1...v2.0.0

build.py

@@ -56,7 +56,7 @@ def replace_identifier(thesaurus: Thesaurus, old_id: str, new_id: str):
 def check_changes(thesaurus: Thesaurus, thesaurus_prev: Thesaurus):
     uris = thesaurus.refs()
     uris_prev = thesaurus_prev.refs()
-    
+
     count_new = 0
     count_changed = 0
     count_removed = len(list(uri for uri in uris_prev if uri not in uris))
@@ -97,7 +97,10 @@ def is_infile(fn):
     thesaurus = Thesaurus()
 
     # Prepare parsing.
-    fns = [join(indir, fn) for indir in INDIR.split(':') for fn in os.listdir(indir) if is_infile(fn)]
+    indirs = INDIR.split(":")
+    for indir in indirs:
+        print(f'Reading files from {indir}')
+    fns = [join(indir, fn) for indir in indirs for fn in os.listdir(indir) if is_infile(fn)]
     print(f'Parsing {len(fns)} files...')
     skipped = []
 

qlit.nt

@@ -6780,7 +6780,7 @@
 <https://queerlit.dh.gu.se/qlit/v1/nj44aa03> <http://www.w3.org/2004/02/skos/core#scopeNote> "Används för skildringar av homosexuella, bisexuella, transpersoner, queera och/eller intersexpersoner som beskrivs som icke-vita. Icke-vit förstås här som en rasifierande kategori som markerar personer med melaninrik hud eller andra karakteristika som rasifieras som icke-vita. Ämnesordet används för skildringar där identiteten som icke-vit är i fokus." .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://purl.org/dc/terms/identifier> "nj45ko84" .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://purl.org/dc/terms/issued> "2022-05-19T10:12:00"^^<http://www.w3.org/2001/XMLSchema#dateTime> .
-<https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://purl.org/dc/terms/modified> "2023-02-07T17:15:49"^^<http://www.w3.org/2001/XMLSchema#dateTime> .
+<https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://purl.org/dc/terms/modified> "2023-09-13T11:44:16"^^<http://www.w3.org/2001/XMLSchema#dateTime> .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.w3.org/2004/02/skos/core#Concept> .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/2004/02/skos/core#altLabel> "Bikvinnor" .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/2004/02/skos/core#altLabel> "Kvinnliga bisexuella" .
@@ -6793,7 +6793,6 @@
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/2004/02/skos/core#narrower> <https://queerlit.dh.gu.se/qlit/v1/cq47se04> .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/2004/02/skos/core#narrower> <https://queerlit.dh.gu.se/qlit/v1/os42ys07> .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/2004/02/skos/core#narrower> <https://queerlit.dh.gu.se/qlit/v1/pc36se94> .
-<https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/2004/02/skos/core#narrower> <https://queerlit.dh.gu.se/qlit/v1/px69nf52> .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/2004/02/skos/core#prefLabel> "Bisexuella kvinnor" .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/2004/02/skos/core#related> <https://queerlit.dh.gu.se/qlit/v1/dq66qs98> .
 <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> <http://www.w3.org/2004/02/skos/core#related> <https://queerlit.dh.gu.se/qlit/v1/px69nf52> .
@@ -8213,10 +8212,9 @@
 <https://queerlit.dh.gu.se/qlit/v1/px55pp67> <http://www.w3.org/2004/02/skos/core#scopeNote> "Används för skildringar av HBTQI-personer som är anhängare av romersk-katolska kyrkan. Används även för skildringar av andra katolikers förhållningssätt till HBTQI." .
 <https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://purl.org/dc/terms/identifier> "px69nf52" .
 <https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://purl.org/dc/terms/issued> "2022-05-19T10:12:00"^^<http://www.w3.org/2001/XMLSchema#dateTime> .
-<https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://purl.org/dc/terms/modified> "2023-02-07T17:15:49"^^<http://www.w3.org/2001/XMLSchema#dateTime> .
+<https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://purl.org/dc/terms/modified> "2023-09-13T11:44:16"^^<http://www.w3.org/2001/XMLSchema#dateTime> .
 <https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.w3.org/2004/02/skos/core#Concept> .
 <https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://www.w3.org/2004/02/skos/core#broader> <https://queerlit.dh.gu.se/qlit/v1/cb49gf75> .
-<https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://www.w3.org/2004/02/skos/core#broader> <https://queerlit.dh.gu.se/qlit/v1/nj45ko84> .
 <https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://www.w3.org/2004/02/skos/core#inScheme> <https://queerlit.dh.gu.se/qlit/v1> .
 <https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://www.w3.org/2004/02/skos/core#narrower> <https://queerlit.dh.gu.se/qlit/v1/qo25ba41> .
 <https://queerlit.dh.gu.se/qlit/v1/px69nf52> <http://www.w3.org/2004/02/skos/core#prefLabel> "Kärlek mellan kvinnor" .

qlit/simple.py

@@ -5,7 +5,7 @@
 from os.path import basename
 import re
 from dotenv import load_dotenv
-from rdflib import SKOS, URIRef, Literal
+from rdflib import OWL, SKOS, URIRef, Literal
 from .thesaurus import BASE, Termset, Thesaurus
 from collections.abc import Generator
 
@@ -177,6 +177,7 @@ def add_hit(ref, score):
 
         scored_hits = []
         for ref, score in hits.items():
+            if self.th.value(ref, OWL.deprecated): continue
             term = SimpleTerm.from_subject(self.t, ref)
             term['score'] = score
             scored_hits.append(term)

qlit/skos.py

@@ -43,3 +43,9 @@ def skos_complete_graph(g: Thesaurus) -> None:
     if not g.value(term, SKOS.broader):
       g.set((term, SKOS.topConceptOf, g.scheme))
       g.add((g.scheme, SKOS.hasTopConcept, term))
+
+    # replaces <-> isReplacedBy
+    for replaced in g[term:DCTERMS.replaces]:
+      g.add((replaced, DCTERMS.isReplacedBy, term))
+    for replacer in g[term:DCTERMS.isReplacedBy]:
+      g.add((replacer, DCTERMS.replaces, term))

qlit/test_simple.py

@@ -1,7 +1,6 @@
 from rdflib import URIRef, RDF, SKOS
 from .thesaurus import Thesaurus, Termset
 from .simple import SimpleThesaurus, SimpleTerm, ref_to_name, name_to_ref, Tokenizer
-from json import dumps
 
 def test_tokenizer():
     assert list(Tokenizer.split("foo bar")) == ["foo", "bar"]
@@ -46,12 +45,28 @@ def test_simple_term_from_termset():
 
     terms = SimpleTerm.from_termset(termset)
     assert len(terms) == 2
-    # Convert to json to compare objects
-    terms_json = dumps(terms, sort_keys=True)
-    term1 = SimpleTerm.from_subject(T, URIRef("https://queerlit.dh.gu.se/qlit/v1/ez04as46"))
-    term2 = SimpleTerm.from_subject(T, URIRef("https://queerlit.dh.gu.se/qlit/v1/lx88hn91"))
-    assert dumps(term1, sort_keys=True) in terms_json
-    assert dumps(term2, sort_keys=True) in terms_json
+    term1 = next(term for term in terms if term['name'] == 'ez04as46')
+    term2 = next(term for term in terms if term['name'] == 'lx88hn91')
+    assert term1['prefLabel'] == "Syskon"
+    assert term2['prefLabel'] == "Intersexseparatism"
+
+    sterm1 = SimpleTerm.from_subject(T, URIRef("https://queerlit.dh.gu.se/qlit/v1/ez04as46"))
+    sterm2 = SimpleTerm.from_subject(T, URIRef("https://queerlit.dh.gu.se/qlit/v1/lx88hn91"))
+    for term, sterm in ((term1, sterm1), (term2, sterm2)):
+
+        # Compare string values
+        for prop in ['name', 'uri', 'prefLabel', 'scopeNote']:
+            assert term[prop] == sterm[prop]
+
+        # Compare lists of strings
+        for prop in ['altLabels', 'hiddenLabels', 'broader', 'narrower', 'related']:
+            assert sorted(term[prop]) == sorted(sterm[prop])
+
+        # Compare lists of terms
+        for prop in ['exactMatch', 'closeMatch']:
+            def sort_by_uri(dicts):
+                return sorted(dicts, key=lambda t: t['uri'])
+            assert sort_by_uri(term[prop]) == sort_by_uri(sterm[prop])
 
 def test_simple_term_get_labels():
     term = SimpleTerm.from_subject(T, URIRef("https://queerlit.dh.gu.se/qlit/v1/xy93px60"))

qlit/test_skos.py

@@ -1,4 +1,4 @@
-from rdflib import Graph, URIRef, Literal, RDF, SKOS
+from rdflib import Graph, URIRef, Literal, DCTERMS, RDF, SKOS
 from .skos import skos_validate_partial, skos_validate_graph, skos_complete_graph
 from .thesaurus import Thesaurus
 
@@ -32,13 +32,16 @@ def test_skos_complete_graph():
     t = Thesaurus()
     food = URIRef("https://queerlit.dh.gu.se/qlit/v1/food")
     fruit = URIRef("https://queerlit.dh.gu.se/qlit/v1/fruit")
+    friut = URIRef("https://queerlit.dh.gu.se/qlit/v1/friut")
     vegetable = URIRef("https://queerlit.dh.gu.se/qlit/v1/vegetable")
     t.add((food, RDF.type, SKOS.Concept))
     t.add((fruit, RDF.type, SKOS.Concept))
+    t.add((friut, RDF.type, SKOS.Concept))
     t.add((vegetable, RDF.type, SKOS.Concept))
     t.add((food, SKOS.narrower, fruit))
     t.add((vegetable, SKOS.broader, food))
     t.add((vegetable, SKOS.related, fruit))
+    t.add((friut, DCTERMS.isReplacedBy, fruit))
 
     skos_complete_graph(t)
     assert (fruit, SKOS.inScheme, t.scheme) in t
@@ -47,4 +50,5 @@ def test_skos_complete_graph():
     assert (fruit, SKOS.related, vegetable) in t
     assert (food, SKOS.topConceptOf, t.scheme) in t
     assert (fruit, SKOS.topConceptOf, t.scheme) not in t
-    assert (t.scheme, SKOS.hasTopConcept, food) in t
+    assert (t.scheme, SKOS.hasTopConcept, food) in t
+    assert (fruit, DCTERMS.replaces, friut) in t

qlit/thesaurus.py

@@ -1,4 +1,4 @@
-from rdflib import RDF, SKOS, Graph, Literal, URIRef
+from rdflib import RDF, OWL, SKOS, Graph, Literal, URIRef
 
 BASE = 'https://queerlit.dh.gu.se/qlit/v1/'
 
@@ -39,6 +39,9 @@ def terms_if(self, f) -> Termset:
         g = Termset(base=self.base)
         for term in self.refs():
             if f(term):
+                # Skip any deprecated term.
+                if self.value(term, OWL.deprecated):
+                    continue;
                 g += self.triples((term, None, None))
         return g