Instead of adapting URI generation logic in the PBW script, a viable option for URI consoldiation between old UUID-based PBW URIs and hash-based URIs according to the URI Minting Policy would be a SPARQL Update request that generates the respective URIs.
The following is a Perplexity-generated draft of such an Update request:
l SPARQL UPDATE: Replace URIs for crm:E21_Person entities
# whose identifier was assigned by the Prosopography of the Byzantine World.
#
# URI minting policy for E21_Person:
# Hash input: LCASE(STRIP(Identifier)) + " / " + LCASE(STRIP(Service))
# where Service = "https://pbw2016.kdl.kcl.ac.uk/"
# New URI = <https://r11.eu/rdf/resource/{first 36 chars of SHA256 hex digest}>
#
# IMPORTANT: This query assumes your triple store supports SHA256() returning
# a hex string (SPARQL 1.1 standard). Some stores (e.g. Jena) return the hash
# with a "^^xsd:hexBinary" datatype and in uppercase — adjust STR()/LCASE()
# wrapping as needed for your store.
#
# Strategy:
# 1. Find each E21_Person that has an E15_Identifier_Assignment
# carried out by the PBW agent, and retrieve the symbolic content
# (the identifier string) from the associated E42_Identifier.
# 2. Compute the new URI per the minting policy.
# 3. For every triple where the old person URI appears as subject or object,
# delete the old triple and insert a replacement with the new URI.
#
# NOTE: Run the SELECT query at the bottom first to preview the mappings
# before executing the DELETE/INSERT.
PREFIX crm: <http://www.cidoc-crm.org/cidoc-crm/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
PREFIX r11res: <https://r11.eu/rdf/resource/>
### Step 1: Replace triples where the person URI is the SUBJECT
DELETE {
?person ?p ?o .
}
INSERT {
?newURI ?p ?o .
}
WHERE {
# Identify persons with a PBW identifier assignment
?person a crm:E21_Person .
# The E15_Identifier_Assignment that targets this person
?assignment a crm:E15_Identifier_Assignment ;
crm:P140_assigned_attribute_to ?person ;
crm:P14_carried_out_by ?agent ;
crm:P37_assigned ?identifier .
# The agent must be the Prosopography of the Byzantine World
?agent rdfs:label "Prosopography of the Byzantine World"@en .
# The E42_Identifier and its symbolic content (the PBW identifier string)
?identifier a crm:E42_Identifier ;
crm:P190_has_symbolic_content ?idValue .
# All triples with the person as subject
?person ?p ?o .
# Compute new URI per minting policy:
# hash_input = lcase(strip(idValue)) + " / " + "https://pbw2016.kdl.kcl.ac.uk/"
BIND(CONCAT(LCASE(xsd:string(?idValue)), " / ", "https://pbw2016.kdl.kcl.ac.uk/") AS ?hashInput)
BIND(SHA256(?hashInput) AS ?fullHash)
BIND(SUBSTR(STR(?fullHash), 1, 36) AS ?segment)
BIND(IRI(CONCAT("https://r11.eu/rdf/resource/", ?segment)) AS ?newURI)
# Only process persons whose URI actually needs changing
FILTER(?person != ?newURI)
};
### Step 2: Replace triples where the person URI is the OBJECT
DELETE {
?s ?p ?person .
}
INSERT {
?s ?p ?newURI .
}
WHERE {
?person a crm:E21_Person .
?assignment a crm:E15_Identifier_Assignment ;
crm:P140_assigned_attribute_to ?person ;
crm:P14_carried_out_by ?agent ;
crm:P37_assigned ?identifier .
?agent rdfs:label "Prosopography of the Byzantine World"@en .
?identifier a crm:E42_Identifier ;
crm:P190_has_symbolic_content ?idValue .
# All triples with the person as object
?s ?p ?person .
BIND(CONCAT(LCASE(xsd:string(?idValue)), " / ", "https://pbw2016.kdl.kcl.ac.uk/") AS ?hashInput)
BIND(SHA256(?hashInput) AS ?fullHash)
BIND(SUBSTR(STR(?fullHash), 1, 36) AS ?segment)
BIND(IRI(CONCAT("https://r11.eu/rdf/resource/", ?segment)) AS ?newURI)
FILTER(?person != ?newURI)
};
### ---------------------------------------------------------------
### PREVIEW QUERY: Run this SELECT first to inspect the mappings
### before executing the UPDATE above.
### ---------------------------------------------------------------
# PREFIX crm: <http://www.cidoc-crm.org/cidoc-crm/>
# PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
# PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
#
SELECT ?person ?idValue ?hashInput ?newURI
WHERE {
?person a crm:E21_Person .
?assignment a crm:E15_Identifier_Assignment ;
crm:P140_assigned_attribute_to ?person ;
crm:P14_carried_out_by ?agent ;
crm:P37_assigned ?identifier .
?agent rdfs:label "Prosopography of the Byzantine World"@en .
?identifier a crm:E42_Identifier ;
crm:P190_has_symbolic_content ?idValue .
BIND(CONCAT(LCASE(xsd:string(?idValue)), " / ", "https://pbw2016.kdl.kcl.ac.uk/") AS ?hashInput)
BIND(SHA256(?hashInput) AS ?fullHash)
BIND(SUBSTR(STR(?fullHash), 1, 36) AS ?segment)
BIND(IRI(CONCAT("https://r11.eu/rdf/resource/", ?segment)) AS ?newURI)
FILTER(?person != ?newURI)
}
ORDER BY ?idValueI reviewed the query and find the URI generation to be sound. Also spot-checking the generated URIs against URIs produced by the Master Spreadsheet Conversion logic confirms that URI consolidation works as intended.
However, I would suggest INSERTing owl:sameAs insertions into a <https://r11.eu/rdf/resource/consolidation> named graph; this would avoid destructive modification of the main PBW graph and keep things separated.
Instead of adapting URI generation logic in the PBW script, a viable option for URI consoldiation between old UUID-based PBW URIs and hash-based URIs according to the URI Minting Policy would be a SPARQL Update request that generates the respective URIs.
The following is a Perplexity-generated draft of such an Update request:
I reviewed the query and find the URI generation to be sound. Also spot-checking the generated URIs against URIs produced by the Master Spreadsheet Conversion logic confirms that URI consolidation works as intended.
However, I would suggest INSERTing
owl:sameAsinsertions into a<https://r11.eu/rdf/resource/consolidation>named graph; this would avoid destructive modification of the main PBW graph and keep things separated.