From 0bbc73a16d7f02448e6709f1a3f68120263ed797 Mon Sep 17 00:00:00 2001
From: "Sofian A. Thibaut" <sofian.ab@hotmail.fr>
Date: Mon, 30 Sep 2024 11:43:51 +0200
Subject: [PATCH] Fix search to ignore accents and case in trigram similarity

---
 public_data/models/administration/Commune.py     | 3 ++-
 public_data/models/administration/Departement.py | 3 ++-
 public_data/models/administration/Epci.py        | 3 ++-
 public_data/models/administration/Region.py      | 3 ++-
 public_data/models/administration/Scot.py        | 3 ++-
 5 files changed, 10 insertions(+), 5 deletions(-)

diff --git a/public_data/models/administration/Commune.py b/public_data/models/administration/Commune.py
index 8dbca5ce0..afd2fe5c5 100644
--- a/public_data/models/administration/Commune.py
+++ b/public_data/models/administration/Commune.py
@@ -1,6 +1,7 @@
 from django.contrib.gis.db import models
 from django.contrib.postgres.search import TrigramSimilarity
 from django.core.validators import MaxValueValidator, MinValueValidator
+from django.db.models.functions import Lower
 
 from public_data.models.cerema import Cerema
 from public_data.models.enums import SRID
@@ -87,7 +88,7 @@ def get_official_id(self) -> str:
 
     @classmethod
     def search(cls, needle, region=None, departement=None, epci=None):
-        qs = cls.objects.annotate(similarity=TrigramSimilarity("name", needle))
+        qs = cls.objects.annotate(similarity=TrigramSimilarity(Lower("name__unaccent"), needle.lower()))
 
         if needle.isdigit():
             qs = cls.objects.filter(insee__icontains=needle)
diff --git a/public_data/models/administration/Departement.py b/public_data/models/administration/Departement.py
index 218e5e373..bff61fd03 100644
--- a/public_data/models/administration/Departement.py
+++ b/public_data/models/administration/Departement.py
@@ -1,6 +1,7 @@
 from django.contrib.gis.db import models
 from django.contrib.postgres.fields import ArrayField
 from django.contrib.postgres.search import TrigramSimilarity
+from django.db.models.functions import Lower
 
 from public_data.models.cerema import Cerema
 from public_data.models.enums import SRID
@@ -48,7 +49,7 @@ def __str__(self):
 
     @classmethod
     def search(cls, needle, region=None, departement=None, epci=None):
-        qs = cls.objects.annotate(similarity=TrigramSimilarity("name", needle))
+        qs = cls.objects.annotate(similarity=TrigramSimilarity(Lower("name__unaccent"), needle.lower()))
         qs = qs.filter(similarity__gt=0.15)  # Filtrer par un score minimum de similarité
         qs = qs.order_by("-similarity")  # Trier par score décroissant
 
diff --git a/public_data/models/administration/Epci.py b/public_data/models/administration/Epci.py
index 74f260de9..e741f5127 100644
--- a/public_data/models/administration/Epci.py
+++ b/public_data/models/administration/Epci.py
@@ -1,6 +1,7 @@
 from django.apps import apps
 from django.contrib.gis.db import models
 from django.contrib.postgres.search import TrigramSimilarity
+from django.db.models.functions import Lower
 
 from public_data.models.enums import SRID
 from utils.db import IntersectManager
@@ -58,7 +59,7 @@ def __str__(self):
 
     @classmethod
     def search(cls, needle, region=None, departement=None, epci=None):
-        qs = cls.objects.annotate(similarity=TrigramSimilarity("name", needle))
+        qs = cls.objects.annotate(similarity=TrigramSimilarity(Lower("name__unaccent"), needle.lower()))
         qs = qs.filter(similarity__gt=0.15)  # Filtrer par un score minimum de similarité
         qs = qs.order_by("-similarity")  # Trier par score décroissant
 
diff --git a/public_data/models/administration/Region.py b/public_data/models/administration/Region.py
index 436b0200c..9748c83b6 100644
--- a/public_data/models/administration/Region.py
+++ b/public_data/models/administration/Region.py
@@ -1,6 +1,7 @@
 from django.apps import apps
 from django.contrib.gis.db import models
 from django.contrib.postgres.search import TrigramSimilarity
+from django.db.models.functions import Lower
 
 from public_data.models.cerema import Cerema
 from public_data.models.enums import SRID
@@ -42,7 +43,7 @@ def get_ocsge_millesimes(self) -> set:
 
     @classmethod
     def search(cls, needle, region=None, departement=None, epci=None):
-        qs = cls.objects.annotate(similarity=TrigramSimilarity("name", needle))
+        qs = cls.objects.annotate(similarity=TrigramSimilarity(Lower("name__unaccent"), needle.lower()))
         qs = qs.filter(similarity__gt=0.15)  # Filtrer par un score minimum de similarité
         qs = qs.order_by("-similarity")  # Trier par score décroissant
 
diff --git a/public_data/models/administration/Scot.py b/public_data/models/administration/Scot.py
index 7ba74fca9..4b5557f90 100644
--- a/public_data/models/administration/Scot.py
+++ b/public_data/models/administration/Scot.py
@@ -1,5 +1,6 @@
 from django.contrib.gis.db import models
 from django.contrib.postgres.search import TrigramSimilarity
+from django.db.models.functions import Lower
 
 from public_data.models.cerema import Cerema
 from public_data.models.enums import SRID
@@ -49,7 +50,7 @@ def get_official_id(self) -> str:
 
     @classmethod
     def search(cls, needle, region=None, departement=None, epci=None):
-        qs = cls.objects.annotate(similarity=TrigramSimilarity("name", needle))
+        qs = cls.objects.annotate(similarity=TrigramSimilarity(Lower("name__unaccent"), needle.lower()))
         qs = qs.filter(similarity__gt=0.15)  # Filtrer par un score minimum de similarité
         qs = qs.order_by("-similarity")  # Trier par score décroissant