Decapitalize as part of normalization; add tests

app/routes/thematic/geoencoder.py

@@ -38,9 +38,9 @@ async def geoencode(
         None,
         description="Name of the subregion to match.",
     ),
-    search_unaccented: bool = Query(
+    normalize_search: bool = Query(
         True,
-        description="Whether or not to unaccent names in request.",
+        description="Whether or not to perform a case- and accent-insensitive search.",
     ),
 ):
     """Look up administrative boundary IDs matching a specified country name
@@ -64,13 +64,13 @@ async def geoencode(
     await version_is_valid(dataset, version_str)
 
     names: List[str | None] = sanitize_names(
-        search_unaccented, country, region, subregion
+        normalize_search, country, region, subregion
     )
 
     adm_level: str = determine_admin_level(*names)
 
     sql: str = _admin_boundary_lookup_sql(
-        adm_level, search_unaccented, admin_source, *names
+        adm_level, normalize_search, admin_source, *names
     )
 
     json_data: List[Dict[str, Any]] = await _query_dataset_json(
@@ -111,13 +111,13 @@ async def geoencode(
 
 
 def sanitize_names(
-    search_unaccented: bool,
+    normalize_search: bool,
     country: str | None,
     region: str | None,
     subregion: str | None,
 ) -> List[str | None]:
-    """Turn any empty strings into Nones, enforce the admin level hierarchy,
-    and optionally unaccent names.
+    """Turn any empty strings into Nones, enforces the admin level hierarchy,
+    and optionally unaccents and decapitalizes names.
     """
     names = []
 
@@ -128,8 +128,8 @@ def sanitize_names(
         )
 
     for name in (country, region, subregion):
-        if name and search_unaccented:
-            names.append(unidecode(name))
+        if name and normalize_search:
+            names.append(unidecode(name).lower())
         elif name:
             names.append(name)
         else:
@@ -155,7 +155,7 @@ def determine_admin_level(
 
 def _admin_boundary_lookup_sql(
     adm_level: str,
-    search_unaccented: bool,
+    normalize_search: bool,
     dataset: str,
     country_name: str,
     region_name: str | None,
@@ -165,8 +165,8 @@ def _admin_boundary_lookup_sql(
     IDs by name.
     """
     name_fields: List[str] = ["country", "name_1", "name_2"]
-    if search_unaccented:
-        match_name_fields = [name_field + "_unaccented" for name_field in name_fields]
+    if normalize_search:
+        match_name_fields = [name_field + "_normalized" for name_field in name_fields]
     else:
         match_name_fields = name_fields
 

tests_v2/unit/app/routes/thematic/geoencoder/test_geoencoder.py

@@ -1,12 +1,63 @@
 from typing import Any, Dict, List, Optional
 
 import pytest
+from fastapi import HTTPException
 from httpx import AsyncClient
 
 from app.models.pydantic.geostore import GeostoreCommon
-from app.routes.datasets.versions import get_version
 from app.routes.thematic import geoencoder
-from app.routes.thematic.geoencoder import _admin_boundary_lookup_sql
+from app.routes.thematic.geoencoder import _admin_boundary_lookup_sql, sanitize_names
+
+
+@pytest.mark.asyncio
+async def test_sanitize_names_pass_through() -> None:
+    country = "A Country"
+    region = "Some region"
+    subregion = "SUBREGION"
+    normalize = False
+
+    names = sanitize_names(normalize, country, region, subregion)
+
+    assert names == [country, region, subregion]
+
+
+@pytest.mark.asyncio
+async def test_sanitize_names_normalize() -> None:
+    country = "Fictîcious de San México"
+    region = "Söme Reğion"
+    subregion = "SÜBREGION"
+    normalize = True
+
+    names = sanitize_names(normalize, country, region, subregion)
+
+    assert names == ["ficticious de san mexico", "some region", "subregion"]
+
+
+@pytest.mark.asyncio
+async def test_sanitize_names_tolerate_empty() -> None:
+    country = "México"
+    region = "Tijuana"
+    subregion = ""
+    normalize = False
+
+    names = sanitize_names(normalize, country, region, subregion)
+
+    assert names == [country, region, None]
+
+
+@pytest.mark.asyncio
+async def test_sanitize_names_tolerate_enforce_hierarchy() -> None:
+    country = "México"
+    region = None
+    subregion = "some subregion"
+    normalize = False
+
+    try:
+        _ = sanitize_names(normalize, country, region, subregion)
+    except HTTPException as e:
+        assert (
+            e.detail == "If subregion is specified, region must be specified as well."
+        )
 
 
 @pytest.mark.asyncio
@@ -48,15 +99,15 @@ async def test__admin_boundary_lookup_sql_all() -> None:
 
 
 @pytest.mark.asyncio
-async def test__admin_boundary_lookup_sql_all_unaccented() -> None:
+async def test__admin_boundary_lookup_sql_all_normalized() -> None:
     sql = _admin_boundary_lookup_sql(
         "2", True, "some_dataset", "some_country", "some_region", "some_subregion"
     )
     assert sql == (
         "SELECT gid_0, gid_1, gid_2, country, name_1, name_2 FROM some_dataset"
-        " WHERE country_unaccented='some_country'"
-        " AND name_1_unaccented='some_region'"
-        " AND name_2_unaccented='some_subregion'"
+        " WHERE country_normalized='some_country'"
+        " AND name_1_normalized='some_region'"
+        " AND name_2_normalized='some_subregion'"
         " AND adm_level='2'"
     )