BUG: fix overwriting a corrupt fileGDB directory

CHANGES.md

@@ -18,6 +18,7 @@
 -   Fix decode error reading an sqlite file on windows (#568).
 -   Fix wrong layername when creating .gpkg.zip file (#570).
 -   Fix segfault on providing an invalid value for `layer` in `read_info` (#564).
+-   Fix overwriting a corrupt fileGDB directory (#600).
 -   Fix error when reading data with ``use_arrow=True`` after having used the
     Parquet driver with GDAL>=3.12 (#601).
 

pyogrio/_io.pyx

@@ -9,6 +9,7 @@ import locale
 import logging
 import math
 import os
+import shutil
 import sys
 import warnings
 from pathlib import Path
@@ -2391,7 +2392,15 @@ cdef create_ogr_dataset_layer(
                 raise exc
 
             # otherwise create from scratch
-            os.unlink(path)
+            if (
+                driver == "OpenFileGDB"
+                and os.path.isdir(path)
+                and os.path.splitext(path)[1] == ".gdb"
+            ):
+                # An FileGDB "file" is a directory instead of a file, so use rmtree
+                shutil.rmtree(path)
+            else:
+                os.unlink(path)
 
             ogr_dataset = NULL
 

pyogrio/tests/test_geopandas_io.py

@@ -2748,6 +2748,33 @@ def test_write_geojson_rfc7946_coordinates(tmp_path, use_arrow):
     assert np.array_equal(gdf_in_appended.geometry.values, points + points_append)
 
 
+@pytest.mark.requires_arrow_write_api
+def test_write_openfilegdb_overwrite_corrupt(tmp_path, use_arrow):
+    """Test to overwriting an existing corrupt OpenFileGDB.
+
+    Test added in context of https://github.com/geopandas/pyogrio/issues/598
+    """
+    # Create a corrupt OpenFileGDB file. An empty directory suffices.
+    test_path = tmp_path / "test.gdb"
+    test_path.mkdir()
+
+    # Overwrite the corrupt OpenFileGDB
+    gdf = gp.GeoDataFrame(
+        {"id": [1.0, 2.0, 3.0]},
+        geometry=[shapely.Point(x, x) for x in range(3)],
+        crs="EPSG:4326",
+    )
+    write_dataframe(
+        gdf, test_path, layer="test_layer", driver="OpenFileGDB", use_arrow=use_arrow
+    )
+
+    # Read back and verify it was (over)written correctly
+    assert test_path.exists()
+    assert test_path.is_dir()
+    read_gdf = read_dataframe(test_path, use_arrow=use_arrow)
+    assert_geodataframe_equal(gdf, read_gdf)
+
+
 @pytest.mark.requires_arrow_api
 @pytest.mark.skipif(
     not GDAL_HAS_PARQUET_DRIVER, reason="Parquet driver is not available"