Merge pull request #180 from nasa/feature/issue-153-return-empty-file

propagate first empty file if all input files are empty

CHANGELOG.md

@@ -9,6 +9,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Added
   - [Issue #170](https://github.com/nasa/stitchee/issues/170): Add PyPI badges to readme
 ### Changed
+  - [Issue #153](https://github.com/nasa/stitchee/issues/153): propagate first empty granule if all input files are empty
   - [Issue #168](https://github.com/nasa/stitchee/issues/168): remove compression for string array of small size
 ### Deprecated
 ### Removed

concatenator/dataset_and_group_handling.py

@@ -322,10 +322,10 @@ def _get_dimension_size(dataset: nc.Dataset, dim_name: str) -> int:
     return dim_size
 
 
-def validate_workable_files(files_to_concat, logger) -> tuple[list[str], int]:
+def validate_workable_files(files, logger) -> tuple[list[str], int]:
     """Remove files from list that are not open-able as netCDF or that are empty."""
     workable_files = []
-    for file in files_to_concat:
+    for file in files:
         try:
             with nc.Dataset(file, "r") as dataset:
                 is_empty = _is_file_empty(dataset)
@@ -334,6 +334,10 @@ def validate_workable_files(files_to_concat, logger) -> tuple[list[str], int]:
         except OSError:
             logger.debug("Error opening <%s> as a netCDF dataset. Skipping.", file)
 
+    # addressing the issue 153: propagate first empty file if all input files are empty
+    if (len(workable_files)) == 0 and (len(files) > 0):
+        workable_files.append(files[0])
+
     number_of_workable_files = len(workable_files)
 
     return workable_files, number_of_workable_files

tests/unit/test_dataset_and_group_handling.py

@@ -8,11 +8,25 @@
     _flatten_coordinate_attribute,
     regroup_coordinate_attribute,
 )
-from concatenator.dataset_and_group_handling import _is_file_empty
+from concatenator.dataset_and_group_handling import (
+    _is_file_empty,
+    validate_workable_files
+)
 
 from .. import data_for_tests_dir
 
 
+def test_dataset_with_single_empty_input_file():
+    """Ensure that a dataset with a single empty input file is propagating empty granule to the output"""
+    files_to_concat = [
+        data_for_tests_dir
+        / "unit-test-data"
+        / "TEMPO_NO2_L2_V03_20240328T154353Z_S008G01.nc4"
+    ]
+    workable_files, number_of_workable_files = validate_workable_files(files_to_concat, None)
+    assert number_of_workable_files == 1
+
+
 def test_dataset_with_singleton_null_values_is_identified_as_empty():
     """Ensure that a dataset with only null arrays with 1-length dimensions is identified as empty."""
     singleton_null_values_file = (