From 97ee4376c162fa0acf804de158c72d1583cb4713 Mon Sep 17 00:00:00 2001
From: Moritz Schott <moritz.schott@heigit.org>
Date: Mon, 8 Jul 2024 11:26:48 +0200
Subject: [PATCH] fix(response): assert tags are exploded even if not present
 in the response (#159)

by None-filling them

closes #149
---
 ohsome/response.py           |  2 +-
 ohsome/test/conftest.py      | 22 ++++++++++
 ohsome/test/test_response.py | 80 ++++++++++++++++++++++++++++++++++++
 3 files changed, 103 insertions(+), 1 deletion(-)

diff --git a/ohsome/response.py b/ohsome/response.py
index ebc02ea..c6880b1 100644
--- a/ohsome/response.py
+++ b/ohsome/response.py
@@ -94,7 +94,7 @@ def _as_geodataframe(
                 for feature in self.data["features"]:
                     properties = feature["properties"]
                     tags = {}
-                    new_properties = {}
+                    new_properties = {k: None for k in explode_tags}
                     for k in properties.keys():
                         if (
                             (k.startswith("@"))
diff --git a/ohsome/test/conftest.py b/ohsome/test/conftest.py
index 9b460f5..2b7ee16 100644
--- a/ohsome/test/conftest.py
+++ b/ohsome/test/conftest.py
@@ -4,10 +4,14 @@
 import logging
 from unittest.mock import patch, PropertyMock
 
+import geopandas as gpd
 import pytest
+from requests import Response
+from shapely import Point
 from urllib3 import Retry
 
 import ohsome
+from ohsome import OhsomeResponse
 
 logger = logging.getLogger(__name__)
 
@@ -95,3 +99,21 @@ def vcr_config():
             "headers",
         ]
     }
+
+
+@pytest.fixture()
+def dummy_ohsome_response() -> OhsomeResponse:
+    """Mocked ohsome response with a single point geometry."""
+    test_gdf = gpd.GeoDataFrame(
+        data={
+            "highway": ["primary"],
+            "width": ["10"],
+            "@snapshotTimestamp": ["2024-01-01"],
+            "@osmId": ["node/1234"],
+        },
+        geometry=[Point(0, 0)],
+        crs="EPSG:4326",
+    )
+    response = Response()
+    response._content = test_gdf.to_json().encode()
+    return OhsomeResponse(response=response)
diff --git a/ohsome/test/test_response.py b/ohsome/test/test_response.py
index 69a9692..e1514d4 100644
--- a/ohsome/test/test_response.py
+++ b/ohsome/test/test_response.py
@@ -2,10 +2,13 @@
 # -*- coding: utf-8 -*-
 """Tests for ohsome API response"""
 import warnings
+from datetime import datetime
 
 import geopandas as gpd
 import pandas as pd
 import pytest
+from geopandas.testing import assert_geodataframe_equal
+from shapely import Point
 
 
 @pytest.mark.vcr
@@ -463,3 +466,80 @@ def test_all_columns_with_timestamps_to_be_without_timezone(base_client):
     assert at_timestamp.tz is None
     assert timestamp.tz is None
     assert at_snapshotTimestamp.tz is None
+
+
+def test_explode_tags(dummy_ohsome_response):
+    """Test if the explode_tags parameter explodes tags."""
+    expected_df = gpd.GeoDataFrame(
+        data={
+            "highway": ["primary"],
+            "@other_tags": [{"width": "10"}],
+            "@snapshotTimestamp": [datetime(2024, 1, 1)],
+            "@osmId": ["node/1234"],
+        },
+        geometry=[Point(0, 0)],
+        crs="EPSG:4326",
+    )
+
+    computed_df = dummy_ohsome_response.as_dataframe(
+        explode_tags=("highway",), multi_index=False
+    )
+
+    assert_geodataframe_equal(computed_df, expected_df, check_like=True)
+
+
+def test_explode_tags_none(dummy_ohsome_response):
+    """Test if the explode_tags parameter can be set to explode all (to get previous behaviour)."""
+    expected_df = gpd.GeoDataFrame(
+        data={
+            "highway": ["primary"],
+            "width": ["10"],
+            "@snapshotTimestamp": [datetime(2024, 1, 1)],
+            "@osmId": ["node/1234"],
+        },
+        geometry=[Point(0, 0)],
+        crs="EPSG:4326",
+    )
+
+    computed_df = dummy_ohsome_response.as_dataframe(
+        explode_tags=None, multi_index=False
+    )
+
+    assert_geodataframe_equal(computed_df, expected_df, check_like=True)
+
+
+def test_explode_tags_empy(dummy_ohsome_response):
+    """Test if explode_tags parameter can be disabled."""
+    expected_df = gpd.GeoDataFrame(
+        data={
+            "@other_tags": [{"width": "10", "highway": "primary"}],
+            "@snapshotTimestamp": [datetime(2024, 1, 1)],
+            "@osmId": ["node/1234"],
+        },
+        geometry=[Point(0, 0)],
+        crs="EPSG:4326",
+    )
+
+    computed_df = dummy_ohsome_response.as_dataframe(explode_tags=(), multi_index=False)
+
+    assert_geodataframe_equal(computed_df, expected_df, check_like=True)
+
+
+def test_explode_tags_missing_in_response(dummy_ohsome_response):
+    """Test if the explode_tags keys are always present in the result, even if they are not part of the response."""
+    expected_df = gpd.GeoDataFrame(
+        data={
+            "this_key_does_not_exist": [None],
+            "@other_tags": [{"width": "10", "highway": "primary"}],
+            "@snapshotTimestamp": [datetime(2024, 1, 1)],
+            "@osmId": ["node/1234"],
+        },
+        geometry=[Point(0, 0)],
+        crs="EPSG:4326",
+    )
+
+    computed_df = dummy_ohsome_response.as_dataframe(
+        explode_tags=("this_key_does_not_exist",), multi_index=False
+    )
+
+    assert_geodataframe_equal(computed_df, expected_df, check_like=True)