🚸(backend) use unaccented full name for user search

CHANGELOG.md

@@ -23,6 +23,7 @@ and this project adheres to
 - ♻️(frontend) preserve @ character when esc is pressed after typing it #1512
 - ♻️(frontend) make summary button fixed to remain visible during scroll #1581
 - ♻️(frontend) pdf embed use full width #1526
+- 🚸(backend) use unaccented full name for user search #1637
 
 ### Fixed
 

src/backend/core/api/viewsets.py

@@ -1,4 +1,5 @@
 """API endpoints"""
+
 # pylint: disable=too-many-lines
 
 import base64
@@ -18,7 +19,7 @@
 from django.db import connection, transaction
 from django.db import models as db
 from django.db.models.expressions import RawSQL
-from django.db.models.functions import Left, Length
+from django.db.models.functions import Greatest, Left, Length
 from django.http import Http404, StreamingHttpResponse
 from django.urls import reverse
 from django.utils import timezone
@@ -37,6 +38,7 @@
 from rest_framework.permissions import AllowAny
 
 from core import authentication, choices, enums, models
+from core.api.filters import remove_accents
 from core.services.ai_services import AIService
 from core.services.collaboration_services import CollaborationService
 from core.services.converter_services import (
@@ -188,13 +190,15 @@ def get_queryset(self):
             queryset = queryset.exclude(documentaccess__document_id=document_id)
 
         filter_data = filterset.form.cleaned_data
-        query = filter_data["q"]
+        query = remove_accents(filter_data["q"])
 
         # For emails, match emails by Levenstein distance to prevent typing errors
         if "@" in query:
             return (
                 queryset.annotate(
-                    distance=RawSQL("levenshtein(email::text, %s::text)", (query,))
+                    distance=RawSQL(
+                        "levenshtein(unaccent(email::text), %s::text)", (query,)
+                    )
                 )
                 .filter(distance__lte=3)
                 .order_by("distance", "email")[: settings.API_USERS_LIST_LIMIT]
@@ -203,11 +207,15 @@ def get_queryset(self):
         # Use trigram similarity for non-email-like queries
         # For performance reasons we filter first by similarity, which relies on an
         # index, then only calculate precise similarity scores for sorting purposes
+
         return (
-            queryset.filter(email__trigram_word_similar=query)
-            .annotate(similarity=TrigramSimilarity("email", query))
+            queryset.annotate(
+                sim_email=TrigramSimilarity("email", query),
+                sim_name=TrigramSimilarity("full_name", query),
+            )
+            .annotate(similarity=Greatest("sim_email", "sim_name"))
             .filter(similarity__gt=0.2)
-            .order_by("-similarity", "email")[: settings.API_USERS_LIST_LIMIT]
+            .order_by("-similarity")[: settings.API_USERS_LIST_LIMIT]
         )
 
     @drf.decorators.action(

src/backend/core/migrations/0027_auto_20251120_0956.py

@@ -0,0 +1,37 @@
+# Generated by Django 5.2.8 on 2025-11-20 09:56
+
+from django.db import migrations
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("core", "0026_comments"),
+    ]
+
+    operations = [
+        migrations.RunSQL(
+            sql="""
+                CREATE OR REPLACE FUNCTION public.immutable_unaccent(regdictionary, text)
+                RETURNS text
+                LANGUAGE c IMMUTABLE PARALLEL SAFE STRICT AS
+                '$libdir/unaccent', 'unaccent_dict';
+
+                CREATE OR REPLACE FUNCTION public.f_unaccent(text)
+                RETURNS text
+                LANGUAGE sql IMMUTABLE PARALLEL SAFE STRICT
+                RETURN public.immutable_unaccent(regdictionary 'public.unaccent', $1);
+
+                CREATE INDEX IF NOT EXISTS user_email_unaccent_trgm_idx
+                ON impress_user
+                USING gin (f_unaccent(email) gin_trgm_ops);
+
+                CREATE INDEX IF NOT EXISTS user_full_name_unaccent_trgm_idx
+                ON impress_user
+                USING gin (f_unaccent(full_name) gin_trgm_ops);
+            """,
+            reverse_sql="""
+                DROP INDEX IF EXISTS user_email_unaccent_trgm_idx;
+                DROP INDEX IF EXISTS user_full_name_unaccent_trgm_idx;
+            """,
+        ),
+    ]

src/backend/core/tests/test_api_users.py

@@ -76,6 +76,131 @@ def test_api_users_list_query_email():
     assert user_ids == []
 
 
+def test_api_users_list_query_email_with_internationalized_domain_names():
+    """
+    Authenticated users should be able to list users and filter by email.
+    It should work even if the email address contains an internationalized domain name.
+    """
+    user = factories.UserFactory()
+
+    client = APIClient()
+    client.force_login(user)
+
+    jean = factories.UserFactory(email="jean.martin@éducation.fr")
+    marie = factories.UserFactory(email="marie.durand@education.fr")
+    kurokawa = factories.UserFactory(email="contact@黒川.日本")
+
+    response = client.get("/api/v1.0/users/?q=jean.martin@education.fr")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(jean.id)]
+
+    response = client.get("/api/v1.0/users/?q=jean.martin@éducation.fr")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(jean.id)]
+
+    response = client.get("/api/v1.0/users/?q=marie.durand@education.fr")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(marie.id)]
+
+    response = client.get("/api/v1.0/users/?q=marie.durand@éducation.fr")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(marie.id)]
+
+    response = client.get("/api/v1.0/users/?q=contact@黒川.日本")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(kurokawa.id)]
+
+
+def test_api_users_list_query_full_name():
+    """
+    Authenticated users should be able to list users and filter by full name.
+    Only results with a Trigram similarity greater than 0.2 with the query should be returned.
+    """
+    user = factories.UserFactory()
+
+    client = APIClient()
+    client.force_login(user)
+
+    dave = factories.UserFactory(email="contact@work.com", full_name="David Bowman")
+
+    response = client.get(
+        "/api/v1.0/users/?q=David",
+    )
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(dave.id)]
+
+    response = client.get("/api/v1.0/users/?q=Bowman")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(dave.id)]
+
+    response = client.get("/api/v1.0/users/?q=bowman")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(dave.id)]
+
+    response = client.get("/api/v1.0/users/?q=BOWMAN")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(dave.id)]
+
+    response = client.get("/api/v1.0/users/?q=BoWmAn")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(dave.id)]
+
+    response = client.get("/api/v1.0/users/?q=Bovin")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == []
+
+
+def test_api_users_list_query_accented_full_name():
+    """
+    Authenticated users should be able to list users and filter by full name with accents.
+    Only results with a Trigram similarity greater than 0.2 with the query should be returned.
+    """
+    user = factories.UserFactory()
+
+    client = APIClient()
+    client.force_login(user)
+
+    fred = factories.UserFactory(
+        email="contact@work.com", full_name="Frédérique Lefèvre"
+    )
+
+    response = client.get("/api/v1.0/users/?q=Frédérique")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(fred.id)]
+
+    response = client.get("/api/v1.0/users/?q=Frederique")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(fred.id)]
+
+    response = client.get("/api/v1.0/users/?q=Lefèvre")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(fred.id)]
+
+    response = client.get("/api/v1.0/users/?q=Lefevre")
+    assert response.status_code == 200
+    user_ids = [user["id"] for user in response.json()]
+    assert user_ids == [str(fred.id)]
+
+    response = client.get("/api/v1.0/users/?q=François Lorfebvre")
+    assert response.status_code == 200
+    users = [user["full_name"] for user in response.json()]
+    assert users == []
+
+
 def test_api_users_list_limit(settings):
     """
     Authenticated users should be able to list users and the number of results