🚸(backend) make document search on title accent-insensitive

This should work in both cases:
- search for "vélo" when the document title contains "velo"
- search for "velo" when the document title contains "vélo"

src/backend/core/api/filters.py

@@ -1,5 +1,8 @@
 """API filters for Impress' core application."""
 
+import unicodedata
+
+from django.db.models import CharField, Func
 from django.utils.translation import gettext_lazy as _
 
 import django_filters
@@ -7,12 +10,64 @@ import django_filters
 from core import models
 
 
-class DocumentFilter(django_filters.FilterSet):
+def remove_accents(value):
+    """Remove accents from a string (vélo -> velo)."""
+    return "".join(
+        c
+        for c in unicodedata.normalize("NFD", value)
+        if unicodedata.category(c) != "Mn"
+    )
+
+
+# pylint: disable=abstract-method
+class Unaccent(Func):
     """
-    Custom filter for filtering documents.
+    PostgreSQL unaccent function wrapper for use in Django ORM queries.
+
+    This allows you to annotate a field using the unaccented version of a
+    text column, enabling accent-insensitive filtering.
     """
 
-    title = django_filters.CharFilter(
+    function = "unaccent"
+    template = "unaccent(%(expressions)s::text)"
+    output_field = CharField()
+
+
+class AccentInsensitiveCharFilter(django_filters.CharFilter):
+    """
+    A custom CharFilter that performs case-insensitive and accent-insensitive filtering.
+
+    This filter uses PostgreSQL's extension `unaccent` function to remove diacritics (accents)
+    from characters before applying the lookup expression (e.g., `icontains`).
+    """
+
+    def filter(self, qs, value):
+        """
+        Apply the filter to the queryset using the unaccented version of the field.
+
+        Args:
+            qs: The queryset to filter.
+            value: The value to search for in the unaccented field.
+
+        Returns:
+            A filtered queryset.
+        """
+        if value:
+            value = remove_accents(value)
+            field_name = self.field_name
+            annotated_field = f"unaccented_{field_name}"
+            return qs.annotate(**{annotated_field: Unaccent(field_name)}).filter(
+                **{f"{annotated_field}__{self.lookup_expr}": value}
+            )
+        return qs
+
+
+class DocumentFilter(django_filters.FilterSet):
+    """
+    Custom filter for filtering documents on title (accent and case insensitive).
+    """
+
+    title = AccentInsensitiveCharFilter(
         field_name="title", lookup_expr="icontains", label=_("Title")
     )
 

src/backend/core/migrations/0021_activate_unaccent_extension.py

@@ -0,0 +1,14 @@
+from django.db import migrations
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("core", "0020_remove_is_public_add_field_attachments_and_duplicated_from"),
+    ]
+
+    operations = [
+        migrations.RunSQL(
+            "CREATE EXTENSION IF NOT EXISTS unaccent;",
+            reverse_sql="DROP EXTENSION IF EXISTS unaccent;",
+        ),
+    ]

src/backend/core/tests/documents/test_api_documents_descendants_filters.py

@@ -7,6 +7,7 @@ from faker import Faker
 from rest_framework.test import APIClient
 
 from core import factories
+from core.api.filters import remove_accents
 
 fake = Faker()
 pytestmark = pytest.mark.django_db
@@ -49,14 +50,16 @@ def test_api_documents_descendants_filter_unknown_field():
     [
         ("Project Alpha", 1),  # Exact match
         ("project", 2),  # Partial match (case-insensitive)
-        ("Guide", 1),  # Word match within a title
+        ("Guide", 2),  # Word match within a title
         ("Special", 0),  # No match (nonexistent keyword)
         ("2024", 2),  # Match by numeric keyword
-        ("", 5),  # Empty string
+        ("", 6),  # Empty string
+        ("velo", 1),  # Accent-insensitive match (velo vs vélo)
+        ("bêta", 1),  # Accent-insensitive match (bêta vs beta)
     ],
 )
 def test_api_documents_descendants_filter_title(query, nb_results):
-    """Authenticated users should be able to search documents by their title."""
+    """Authenticated users should be able to search documents by their unaccented title."""
     user = factories.UserFactory()
     client = APIClient()
     client.force_login(user)
@@ -70,6 +73,7 @@ def test_api_documents_descendants_filter_title(query, nb_results):
         "User Guide",
         "Financial Report 2024",
         "Annual Review 2024",
+        "Guide du vélo urbain",  # <-- Title with accent for accent-insensitive test
     ]
     for title in titles:
         factories.DocumentFactory(title=title, parent=document)
@@ -85,4 +89,7 @@ def test_api_documents_descendants_filter_title(query, nb_results):
 
     # Ensure all results contain the query in their title
     for result in results:
-        assert query.lower().strip() in result["title"].lower()
+        assert (
+            remove_accents(query).lower().strip()
+            in remove_accents(result["title"]).lower()
+        )