Use stable ID for contributor, affiliation, and source blank nodes

code/vocab_funcs.py

@@ -1,6 +1,4 @@
-from rdflib import Graph, Namespace
-from rdflib.compare import graph_diff
-from rdflib.term import Literal, URIRef, BNode
+from rdflib.term import BNode, Literal
 from vocab_management import *
 
 
@@ -239,7 +237,7 @@ def construct_source(item, data, namespace, header):
                 label = label.replace('(', '', 1)
             if url.endswith(')'):
                 url = url[::-1].replace(')', '', 1)[::-1] # reverse string
-            node = BNode()
+            node = BNode(hash_id(f'{label}-{url}', 'web-', 16, 16))
             triples.append((node, RDF.type, SCHEMA.WebPage))
             triples.append((node, SCHEMA.name, Literal(label)))
             triples.append((node, SCHEMA.url, Literal(url)))

code/vocab_management.py

@@ -4,9 +4,13 @@
 '''Data and configurations for vocabulary management'''
 
 import csv
-from rdflib import Namespace, BNode, Literal
-
+import hashlib
 import logging
+import re
+import unicodedata
+
+from rdflib import BNode, Literal, Namespace
+
 logging.basicConfig(
     level=logging.DEBUG, format='%(levelname)s - %(funcName)s :: %(lineno)d - %(message)s')
 DEBUG = logging.debug
@@ -109,7 +113,7 @@
 
 # Root folder to import RDF files from
 IMPORT_PATH = f'../{DPV_VERSION}'
-# Root folder to export HTML filese to
+# Root folder to export HTML files to
 EXPORT_PATH = f'../{DPV_VERSION}'
 # Root folder where Jinja2 templates are stored
 TEMPLATE_PATH = './jinja2_resources'
@@ -2679,6 +2683,20 @@ def generate_authors_affiliations(authors):
     return authors
 
 
+def hash_id(input_text: str, output_prefix: str, output_slug_len:int, output_hash_len: int) -> str:
+    """Takes an input text and generates a hash-based identifier, with a prefix if provided.
+    Example:
+        >>> generate_hash_id("Unabhängige Landeszentrum", "org-", 5, 8)
+        'org-Unabhäng-41451ff1'
+    """
+    input_text = unicodedata.normalize("NFKC", input_text.strip())
+    slug = re.sub(r"[^a-zA-Z0-9\u00C0-\u00FF-]", "", input_text, flags=re.UNICODE)[:output_slug_len]
+    # add version to make ID stable only wihtin the version
+    input_text = f"{DPV_VERSION}{input_text}"
+    output_hash = hashlib.sha256(input_text.encode()).hexdigest()[:output_hash_len]
+    return f"{output_prefix}{slug}-{output_hash}"
+
+
 def _person_slugify():
     '''person is a string, slugify means make it IRI compatible'
     - e.g. 'Harshvardhan J. Pandit' should be 'HarshvardhanJPandit'
@@ -2692,11 +2710,16 @@ def _helper(person_name):
         nonlocal people
         person_name = person_name.strip()
         person = person_name.replace(',','').replace('.','').replace(' ','')
-        # if person_name.startswith('n')
         if person in people:
             return people[person]
-        bnode_person = BNode()
-        bnode_org = BNode()
+
+        org_name = generate_author_affiliation(person_name)
+        if not org_name:
+            raise Exception(f"{person_name} org is empty!")
+
+        bnode_person = BNode(hash_id(f"{person_name}{org_name}", "person-", 8, 8))
+        bnode_org = BNode( hash_id(org_name, "org-", 8, 8))
+
         triples = []
         triples.append((bnode_person, RDF.type, FOAF.Person))
         triples.append((bnode_person, RDF.type, DCT.Agent))
@@ -2706,9 +2729,7 @@ def _helper(person_name):
         if generate_author_website(person_name):
             triples.append((bnode_person, FOAF.homepage, Literal(generate_author_website(person_name))))
         triples.append((bnode_org, RDF.type, FOAF.Organization))
-        triples.append((bnode_org, FOAF.name, Literal(generate_author_affiliation(person_name))))
-        if not generate_author_affiliation(person_name):
-            raise Exception(f"{person_name} org is empty!")
+        triples.append((bnode_org, FOAF.name, Literal(org_name)))
         triples.append((bnode_person, ORG.memberOf, bnode_org))
         people[person] = {
             'person': bnode_person,