Issue #114/#211/#197 tests for load_geojson and its properties arg

openeo_driver/ProcessGraphDeserializer.py

@@ -1581,17 +1581,23 @@ def load_uploaded_files(args: dict, env: EvalEnv) -> Union[DriverVectorCube,Driv
     .returns("vector-cube", schema={"type": "object", "subtype": "vector-cube"})
 )
 def to_vector_cube(args: Dict, env: EvalEnv):
-    # TODO: standardization of something like this? https://github.com/Open-EO/openeo-processes/issues/346
+    _log.warning("Experimental process `to_vector_cube` is deprecated, use `load_geojson` instead")
+    # TODO: remove this experimental/deprecated process
     data = extract_arg(args, "data", process_id="to_vector_cube")
     if isinstance(data, dict) and data.get("type") in {"Polygon", "MultiPolygon", "Feature", "FeatureCollection"}:
         return env.backend_implementation.vector_cube_cls.from_geojson(data)
-    # TODO: support more inputs: string with geojson, string with WKT, list of WKT, string with URL to GeoJSON, ...
     raise FeatureUnsupportedException(f"Converting {type(data)} to vector cube is not supported")
 
 
 @process_registry_100.add_function(spec=read_spec("openeo-processes/2.x/proposals/load_geojson.json"))
 def load_geojson(args: ProcessArgs, env: EvalEnv) -> DriverVectorCube:
-    data = args.get_required("data", validator=ProcessArgs.validator_geojson_dict())
+    data = args.get_required(
+        "data",
+        validator=ProcessArgs.validator_geojson_dict(
+            # TODO: also allow LineString and MultiLineString?
+            allowed_types=["Point", "MultiPoint", "Polygon", "MultiPolygon", "Feature", "FeatureCollection"]
+        ),
+    )
     properties = args.get_optional("properties", default=[], expected_type=(list, tuple))
     vector_cube = env.backend_implementation.vector_cube_cls.from_geojson(data, columns_for_cube=properties)
     return vector_cube

openeo_driver/datacube.py

@@ -8,6 +8,7 @@
 
 import geopandas as gpd
 import numpy
+import pandas
 import pyproj
 import shapely.geometry
 import shapely.geometry.base
@@ -284,14 +285,21 @@ def from_geodataframe(
         elif columns_for_cube == cls.COLUMN_SELECTION_ALL:
             columns_for_cube = available_columns
         elif isinstance(columns_for_cube, list):
-            # TODO #114 limit to subset with available columns (and automatically fill in missing columns with nodata)?
             columns_for_cube = columns_for_cube
         else:
             raise ValueError(columns_for_cube)
         assert isinstance(columns_for_cube, list)
 
         if columns_for_cube:
-            cube_df = data[columns_for_cube]
+            existing = [c for c in columns_for_cube if c in available_columns]
+            to_add = [c for c in columns_for_cube if c not in available_columns]
+            if existing:
+                cube_df = data[existing]
+                if to_add:
+                    cube_df.loc[:, to_add] = numpy.nan
+            else:
+                cube_df = pandas.DataFrame(index=data.index, columns=to_add)
+
             # TODO: remove `columns_for_cube` from geopandas data frame?
             #   Enabling that triggers failure of som existing tests that use `aggregate_spatial`
             #   to "enrich" a vector cube with pre-existing properties
@@ -311,6 +319,7 @@ def from_geodataframe(
             return cls(geometries=geometries_df, cube=cube)
 
         else:
+            # TODO: add a dummy 1D no-data cube?
             return cls(geometries=data)
 
     @classmethod
@@ -429,6 +438,16 @@ def to_wkt(self) -> List[str]:
         wkts = [str(g) for g in self._geometries.geometry]
         return wkts
 
+    def to_internal_json(self) -> dict:
+        """
+        Export to an internal JSON-style representation.
+        Subject to change any time: not intended for public consumption, just for (unit) test purposes.
+        """
+        return {
+            "geometries": shapely.geometry.mapping(self._geometries),
+            "cube": self._cube.to_dict(data="array") if self._cube is not None else None,
+        }
+
     def get_crs(self) -> pyproj.CRS:
         return self._geometries.crs or pyproj.CRS.from_epsg(4326)
 

openeo_driver/testing.py

@@ -6,6 +6,7 @@
 import http.server
 import json
 import logging
+import math
 import multiprocessing
 import re
 import urllib.request
@@ -494,6 +495,11 @@ def approxify(x: Any, rel: Optional = None, abs: Optional[float] = None) -> Any:
         raise ValueError(x)
 
 
+class IsNan:
+    def __eq__(self, other):
+        return isinstance(other, float) and math.isnan(other)
+
+
 class ApproxGeometry:
     """Helper to compactly and approximately compare geometries."""
 

tests/test_testing.py

@@ -1,6 +1,8 @@
 import logging
 import re
 import subprocess
+
+import numpy
 import sys
 import textwrap
 import urllib.error
@@ -22,6 +24,7 @@
     ephemeral_fileserver,
     preprocess_check_and_replace,
     ApproxGeoJSONByBounds,
+    IsNan,
 )
 
 
@@ -261,6 +264,18 @@ def test_approxify_tolerance_rel():
     assert {"a": [10.1, 2.1]} != approxify({"a": [10, 2.3]}, rel=0.01)
 
 
+@pytest.mark.parametrize("other", [float("nan"), numpy.nan])
+def test_is_nan(other):
+    assert other == IsNan()
+    assert IsNan() == other
+
+
+@pytest.mark.parametrize("other", [0, 123, False, True, None, "dfd", [], {}, ()])
+def test_is_not_nan(other):
+    assert other != IsNan()
+    assert IsNan() != other
+
+
 @pytest.mark.parametrize(
     "format",
     [

tests/test_vectorcube.py

@@ -10,7 +10,7 @@
 
 from openeo_driver.errors import OpenEOApiException
 from openeo_driver.datacube import DriverVectorCube
-from openeo_driver.testing import DictSubSet, ApproxGeometry
+from openeo_driver.testing import DictSubSet, ApproxGeometry, IsNan
 from openeo_driver.util.geometry import as_geojson_feature_collection
 from openeo_driver.utils import EvalEnv
 
@@ -83,6 +83,144 @@ def test_to_wkt(self, gdf):
             ['POLYGON ((1 1, 3 1, 2 3, 1 1))', 'POLYGON ((4 2, 5 4, 3 4, 4 2))']
         )
 
+    def test_to_internal_json_defaults(self, gdf):
+        vc = DriverVectorCube(gdf)
+        assert vc.to_internal_json() == {
+            "geometries": DictSubSet(
+                {
+                    "type": "FeatureCollection",
+                    "features": [
+                        DictSubSet(
+                            {
+                                "type": "Feature",
+                                "geometry": {
+                                    "type": "Polygon",
+                                    "coordinates": (((1.0, 1.0), (3.0, 1.0), (2.0, 3.0), (1.0, 1.0)),),
+                                },
+                                "properties": {"id": "first", "pop": 1234},
+                            }
+                        ),
+                        DictSubSet(
+                            {
+                                "type": "Feature",
+                                "geometry": {
+                                    "type": "Polygon",
+                                    "coordinates": (((4.0, 2.0), (5.0, 4.0), (3.0, 4.0), (4.0, 2.0)),),
+                                },
+                                "properties": {"id": "second", "pop": 5678},
+                            }
+                        ),
+                    ],
+                }
+            ),
+            "cube": None,
+        }
+
+    @pytest.mark.parametrize(
+        ["columns_for_cube", "expected_cube"],
+        [
+            (
+                "numerical",
+                {
+                    "name": None,
+                    "dims": ("geometries", "properties"),
+                    "coords": {
+                        "geometries": {"attrs": {}, "data": [0, 1], "dims": ("geometries",)},
+                        "properties": {"attrs": {}, "data": ["pop"], "dims": ("properties",)},
+                    },
+                    "data": [[1234], [5678]],
+                    "attrs": {},
+                },
+            ),
+            (
+                "all",
+                {
+                    "name": None,
+                    "dims": ("geometries", "properties"),
+                    "coords": {
+                        "geometries": {"attrs": {}, "data": [0, 1], "dims": ("geometries",)},
+                        "properties": {"attrs": {}, "data": ["id", "pop"], "dims": ("properties",)},
+                    },
+                    "data": [["first", 1234], ["second", 5678]],
+                    "attrs": {},
+                },
+            ),
+            ([], None),
+            (
+                ["pop", "id"],
+                {
+                    "name": None,
+                    "dims": ("geometries", "properties"),
+                    "coords": {
+                        "geometries": {"attrs": {}, "data": [0, 1], "dims": ("geometries",)},
+                        "properties": {"attrs": {}, "data": ["pop", "id"], "dims": ("properties",)},
+                    },
+                    "data": [[1234, "first"], [5678, "second"]],
+                    "attrs": {},
+                },
+            ),
+            (
+                ["pop", "color"],
+                {
+                    "name": None,
+                    "dims": ("geometries", "properties"),
+                    "coords": {
+                        "geometries": {"attrs": {}, "data": [0, 1], "dims": ("geometries",)},
+                        "properties": {"attrs": {}, "data": ["pop", "color"], "dims": ("properties",)},
+                    },
+                    "data": [[1234.0, IsNan()], [5678.0, IsNan()]],
+                    "attrs": {},
+                },
+            ),
+            (
+                ["color"],
+                {
+                    "name": None,
+                    "dims": ("geometries", "properties"),
+                    "coords": {
+                        "geometries": {"attrs": {}, "data": [0, 1], "dims": ("geometries",)},
+                        "properties": {"attrs": {}, "data": ["color"], "dims": ("properties",)},
+                    },
+                    "data": [[IsNan()], [IsNan()]],
+                    "attrs": {},
+                },
+            ),
+        ],
+    )
+    def test_to_internal_json_columns_for_cube(self, gdf, columns_for_cube, expected_cube):
+        vc = DriverVectorCube.from_geodataframe(gdf, columns_for_cube=columns_for_cube)
+        internal = vc.to_internal_json()
+        assert internal == {
+            "geometries": DictSubSet(
+                {
+                    "type": "FeatureCollection",
+                    "features": [
+                        DictSubSet(
+                            {
+                                "type": "Feature",
+                                "geometry": {
+                                    "type": "Polygon",
+                                    "coordinates": (((1.0, 1.0), (3.0, 1.0), (2.0, 3.0), (1.0, 1.0)),),
+                                },
+                                "properties": {"id": "first", "pop": 1234},
+                            }
+                        ),
+                        DictSubSet(
+                            {
+                                "type": "Feature",
+                                "geometry": {
+                                    "type": "Polygon",
+                                    "coordinates": (((4.0, 2.0), (5.0, 4.0), (3.0, 4.0), (4.0, 2.0)),),
+                                },
+                                "properties": {"id": "second", "pop": 5678},
+                            }
+                        ),
+                    ],
+                }
+            ),
+            "cube": expected_cube,
+        }
+
     def test_get_crs(self, gdf):
         vc = DriverVectorCube(gdf)
         assert vc.get_crs() == pyproj.CRS.from_epsg(4326)