From c37f56506759ca19451e76b3202f2d5fc3a84ad9 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Boris=20Cl=C3=A9net?= <boris.clenet@inria.fr>
Date: Mon, 10 Jun 2024 17:29:50 +0200
Subject: [PATCH] simplify_urns method

---
 bids_prov/utils.py | 47 +++++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 44 insertions(+), 3 deletions(-)

diff --git a/bids_prov/utils.py b/bids_prov/utils.py
index dcd51ba03..d5472c26f 100644
--- a/bids_prov/utils.py
+++ b/bids_prov/utils.py
@@ -5,6 +5,7 @@
 import shutil
 import uuid
 from typing import Mapping, Union, Tuple
+import re
 
 CONTEXT_URL = "https://raw.githubusercontent.com/bids-standard/BEP028_BIDSprov/master/context.json"
 
@@ -120,6 +121,44 @@ def compute_sha_256_entity(entities: dict):
         shutil.rmtree(directory)
 
 
+def simplify_urns(graph: str) -> str:
+    """
+    Replace URNs of a json-ld graph with simpler (not random) values defined by their order of appearance in the graph.
+
+    Parameters
+    ----------
+    graph : str
+        The json-ld graph dumped as a string
+
+    Returns
+    -------
+    str
+        The input string with simplified urns.
+    """
+
+    # Get all occurences of ("urn:" + 36 following chars)
+    urn_occurences = [i for i in re.finditer(r'urn:[a-zA-Z0-9\-]{36}', content)]
+
+    # Prepare a dict in which :
+    #     - key is an urn as present in the original graph
+    #     - value is an id corresponding to the order of appearance of the urn in the graph.
+    known_urns = dict()
+    counter = 0
+
+    # Assign one simpler id to each urn
+    for urn in urn_occurences:
+        if urn.group() not in known_urns:
+            known_urns[urn.group()] = f"urn:{str(counter).zfill(36)}"
+            counter += 1
+
+    # Replace all urns in the input graph with their assigned id
+    output_graph = graph
+    for key, value in known_urns.items():
+        output_graph = output_graph.replace(key, value)
+
+    return output_graph
+
+
 def writing_jsonld(graph, indent, output_file):
     """
     Write a json-ld in memory unless it already exists and contains the same content
@@ -140,10 +179,12 @@ def writing_jsonld(graph, indent, output_file):
     """
     if os.path.isfile(output_file):
         with open(output_file, "r") as f:
-            existing_content = f.read()
+            existing_content = simplify_urns(f.read())
+
+        new_content = simplify_urns(json.dumps(graph, indent=indent))
 
-            if existing_content == json.dumps(graph, indent=indent):
-                return True
+        if existing_content == new_content:
+            return True
 
     with open(output_file, "w") as fd:
         json.dump(graph, fd, indent=indent)