Fix search to ignore accents and case in trigram similarity

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)

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
 

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
 

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
 

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