openeo.UDF: automatically un-indent UDF code

Author: soxofaan

CHANGELOG.md

@@ -11,6 +11,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Changed
 
+- `openeo.UDF()`: automatically un-indent given UDF code ([#782](https://github.com/Open-EO/openeo-python-client/issues/782))
+
 ### Removed
 
 ### Fixed

openeo/rest/_datacube.py

@@ -3,6 +3,7 @@
 import logging
 import pathlib
 import re
+import textwrap
 import typing
 import uuid
 import warnings
@@ -142,6 +143,10 @@ class UDF:
         Specifying the ``data`` argument is not necessary anymore, and actually deprecated.
         Added :py:meth:`from_file` to simplify loading UDF code from a file.
         See :ref:`old_udf_api` for more background about the changes.
+
+    .. versionchanged:: 0.43.0
+        Automatically un-indent given UDF code,
+        to simplify writing valid and properly formatted inline UDF code.
     """
 
     # TODO: eliminate dependency on `openeo.rest.connection` and move to somewhere under `openeo.internal`?
@@ -155,7 +160,9 @@ def __init__(
         data=None,  # TODO #181 remove `data` argument
         version: Optional[str] = None,
         context: Optional[dict] = None,
+        *,
         _source=None,
+        auto_dedent: bool = True,
     ):
         """
         Construct a UDF object from given code string and other argument related to the ``run_udf`` process.
@@ -167,7 +174,8 @@ def __init__(
         :param context: optional additional UDF context data
         :param _source: (for internal use) source identifier
         """
-        # TODO: automatically dedent code (when literal string) ?
+        if auto_dedent:
+            code = textwrap.dedent(code)
         self.code = code
         self._runtime = runtime
         self.version = version

tests/rest/datacube/test_datacube100.py

@@ -3841,6 +3841,46 @@ def test_apply_udf_basic(self, con100):
             },
         }
 
+    def test_apply_udf_auto_dedent(self, con100):
+        udf = openeo.UDF(
+            """
+            import xarray
+
+            def apply_datacube(cube: xarray.DataArray, context: dict) -> xarray.DataArray:
+                cube.values = 0.0001 * cube.values
+                return cube
+            """
+        )
+        cube = con100.load_collection("S2")
+        res = cube.apply(udf)
+
+        assert res.flat_graph()["apply1"]["arguments"] == {
+            "data": {"from_node": "loadcollection1"},
+            "process": {
+                "process_graph": {
+                    "runudf1": {
+                        "process_id": "run_udf",
+                        "arguments": {
+                            "data": {"from_parameter": "x"},
+                            "runtime": "Python",
+                            "udf": "\n".join(
+                                [
+                                    "",
+                                    "import xarray",
+                                    "",
+                                    "def apply_datacube(cube: xarray.DataArray, context: dict) -> xarray.DataArray:",
+                                    "    cube.values = 0.0001 * cube.values",
+                                    "    return cube",
+                                    "",
+                                ]
+                            ),
+                        },
+                        "result": True,
+                    }
+                },
+            },
+        }
+
     def test_apply_udf_runtime_detection(self, con100, requests_mock):
         udf = UDF("def foo(x):\n    return x\n")
         cube = con100.load_collection("S2")