DAS-2326 Return NoDataException warning when the request returns no data (#58)

* DAS-2326 Return NoDataFound warning instead of a failure when the request retuens no data

* DAS-2326 Improve the code with PR comments

* DAS-2326 update check_range_exception to get_failed_variables

* DAS-2325 Update to throw exception if any requested variable is not valid

* DAS-2326 Add comment to the check valid variables

* DAS-2326 Update change log and some variable naming

* DAS-2326 Add missing unit tests

* DAS-2326 Change set update to add

* DAS-2326 Correct check for exception message in tests

* DAS-2326 Update geographic-grid function to capture all failed dimensions

* DAS-2326 Update comments

* DAS-2326 Update earthdata-varinfo version

* DAS-2326 Fix unit test failure

* DAS-2326 Add unit tests to test_dimension_utilities

* DAS-2326 Add subtest for get_requested_index_ranges

* DAS-2326 Add substest for get_spatial_index_ranges for projected

* DAS-2326 Update incorrect subtest description

* DAS-2326 Update test verification for index ranges for projected grid

* DAS-2326 Add subtest for testing multiple temporal dimensions

* DAS-2326 Fix trailing space

* DAS-2326 Fix bug in get_dimension_indices_from_values

* DAS-2326 Add end to end test for NoDataException

* DAS-2326 Add a subtest with a descending file

* DAS-2326 Update harmony-service-lib version

* DAS-2326 Update harmony-service-lib to 2.11

Author: sudha-murthy

CHANGELOG.md

@@ -1,4 +1,13 @@
-## [v1.1.16] - 2025-10-17
+## [v1.1.17] - 2025-11-25
+
+### Changed
+
+- Change HOSS behavior to return NoDataException warning instead of a failed
+  exception when the variable, spatial, temporal request does not return any
+  data. Also fix a bug where the out of range subset requests were reported
+  with a default output.
+
+## [v1.1.16] - 2025-11-17
 
 ### Fixed
 
@@ -207,6 +216,7 @@ Repository structure changes include:
 
 For more information on internal releases prior to NASA open-source approval,
 see legacy-CHANGELOG.md.
+[v1.1.17]: https://github.com/nasa/harmony-opendap-subsetter/releases/tag/1.1.17
 [v1.1.16]: https://github.com/nasa/harmony-opendap-subsetter/releases/tag/1.1.16
 [v1.1.15]: https://github.com/nasa/harmony-opendap-subsetter/releases/tag/1.1.15
 [v1.1.14]: https://github.com/nasa/harmony-opendap-subsetter/releases/tag/1.1.14

docker/service_version.txt

@@ -1 +1 @@
-1.1.16
+1.1.17

hoss/adapter.py

@@ -29,8 +29,9 @@
 from tempfile import mkdtemp
 
 from harmony_service_lib import BaseHarmonyAdapter
+from harmony_service_lib.exceptions import HarmonyException, NoDataException
 from harmony_service_lib.message import Source
-from harmony_service_lib.util import HarmonyException, generate_output_filename, stage
+from harmony_service_lib.util import generate_output_filename, stage
 from pystac import Asset, Item
 
 from hoss.dimension_utilities import is_index_subset
@@ -132,6 +133,9 @@ def process_item(self, item: Item, source: Source):
                 url, title=staged_filename, media_type=mime, roles=['data']
             )
 
+        except NoDataException as no_data_exception:
+            self.logger.exception(no_data_exception)
+            raise
         except Exception as exception:
             self.logger.exception(exception)
             raise_from_hoss_exception(exception)

hoss/dimension_utilities.py

@@ -14,6 +14,7 @@
 from typing import Dict, Set, Tuple
 
 import numpy as np
+from harmony_service_lib.exceptions import NoDataException
 from harmony_service_lib.message import Message
 from harmony_service_lib.message_utility import rgetattr
 from harmony_service_lib.util import Config
@@ -346,6 +347,12 @@ def get_dimension_indices_from_values(
         dimension_values = np.flip(dimension)
         dimension_indices = np.flip(dimension_indices)
 
+    # Check if the minimum and maximum extents are out of range of the dimension values.
+    if maximum_extent < dimension_values[0] or minimum_extent > dimension_values[-1]:
+        raise InvalidRequestedRange()
+
+    # np.interp does not throw an exception if the extents are outside dimension range
+    # It just returns the lowest index or highest index.
     raw_indices = np.interp(dimension_range, dimension_values, dimension_indices)
 
     if (raw_indices[0] == raw_indices[1]) and (raw_indices[0] % 1 == 0.5):
@@ -546,30 +553,43 @@ def get_requested_index_ranges(
     required_dimensions = varinfo.get_required_dimensions(required_variables)
 
     dim_index_ranges = {}
-
+    out_of_range_dim_variables = set()
     with Dataset(dimensions_path, 'r') as dimensions_file:
         for dim in harmony_message.subset.dimensions:
-            if dim.name in required_dimensions:
-                dim_is_valid = True
-            elif dim.name[0] != '/' and f'/{dim.name}' in required_dimensions:
-                dim.name = f'/{dim.name}'
-                dim_is_valid = True
-            else:
-                dim_is_valid = False
-
-            if dim_is_valid:
-                # Try to extract bounds metadata:
-                bounds_array = get_dimension_bounds(dim.name, varinfo, dimensions_file)
-                # Retrieve index ranges for the specifically named dimension:
-                dim_index_ranges[dim.name] = get_dimension_index_range(
-                    dimensions_file[dim.name][:],
-                    dim.min,
-                    dim.max,
-                    bounds_values=bounds_array,
-                )
-            else:
-                # This requested dimension is not in the required dimension set
-                raise InvalidNamedDimension(dim.name)
+            try:
+                if dim.name in required_dimensions:
+                    dim_is_valid = True
+                elif dim.name[0] != '/' and f'/{dim.name}' in required_dimensions:
+                    dim.name = f'/{dim.name}'
+                    dim_is_valid = True
+                else:
+                    dim_is_valid = False
+
+                if dim_is_valid:
+                    # Try to extract bounds metadata:
+                    bounds_array = get_dimension_bounds(
+                        dim.name, varinfo, dimensions_file
+                    )
+                    # Retrieve index ranges for the specifically named dimension:
+                    dim_index_ranges[dim.name] = get_dimension_index_range(
+                        dimensions_file[dim.name][:],
+                        dim.min,
+                        dim.max,
+                        bounds_values=bounds_array,
+                    )
+                else:
+                    # This requested dimension is not in the required dimension set
+                    raise InvalidNamedDimension(dim.name)
+            # In case subset constraint is out of range for a dimension. Continue
+            # processing the other dimensions and raise exception for all the
+            # dimensions that are invalid.
+            except InvalidRequestedRange:
+                out_of_range_dim_variables.add(dim.name)
+
+        if out_of_range_dim_variables:
+            raise NoDataException(
+                f'Input request outside supported dimension range for {out_of_range_dim_variables}'
+            )
 
     return dim_index_ranges
 

hoss/spatial.py

@@ -24,6 +24,7 @@
 
 from typing import List, Set
 
+from harmony_service_lib.exceptions import NoDataException
 from harmony_service_lib.message import Message
 from netCDF4 import Dataset
 from numpy.ma.core import MaskedArray
@@ -49,6 +50,7 @@
     get_dimension_extents,
     get_dimension_index_range,
 )
+from hoss.exceptions import InvalidRequestedRange
 from hoss.projection_utilities import (
     get_master_geotransform,
     get_projected_x_y_extents,
@@ -101,7 +103,7 @@ def get_spatial_index_ranges(
     non_spatial_variables = required_variables.difference(
         varinfo.get_spatial_dimensions(required_variables)
     )
-
+    out_of_range_variables = set()
     with Dataset(dimensions_path, 'r') as dimensions_file:
         if geographic_dimensions:
             # If there is no bounding box, but there is a shape file, calculate
@@ -111,43 +113,60 @@ def get_spatial_index_ranges(
                 bounding_box = get_geographic_bbox(geojson_content)
 
             for dimension in geographic_dimensions:
-                index_ranges[dimension] = get_geographic_index_range(
-                    dimension, varinfo, dimensions_file, bounding_box
-                )
+                try:
+                    index_ranges[dimension] = get_geographic_index_range(
+                        dimension, varinfo, dimensions_file, bounding_box
+                    )
+                except InvalidRequestedRange:
+                    out_of_range_variables.add(dimension)
 
         if projected_dimensions:
             for non_spatial_variable in non_spatial_variables:
-                index_ranges.update(
-                    get_projected_x_y_index_ranges(
-                        non_spatial_variable,
-                        varinfo,
-                        dimensions_file,
-                        index_ranges,
-                        bounding_box=bounding_box,
-                        shape_file_path=shape_file_path,
+                try:
+                    index_ranges.update(
+                        get_projected_x_y_index_ranges(
+                            non_spatial_variable,
+                            varinfo,
+                            dimensions_file,
+                            index_ranges,
+                            bounding_box=bounding_box,
+                            shape_file_path=shape_file_path,
+                        )
                     )
-                )
+                except InvalidRequestedRange:
+                    out_of_range_variables.add(non_spatial_variable)
+
         variables_with_anonymous_dims = get_variables_with_anonymous_dims(
             varinfo, required_variables
         )
+
         for variable_with_anonymous_dims in variables_with_anonymous_dims:
-            latitude_coordinates, longitude_coordinates = get_coordinate_variables(
-                varinfo, [variable_with_anonymous_dims]
-            )
-            if latitude_coordinates and longitude_coordinates:
-                index_ranges.update(
-                    get_x_y_index_ranges_from_coordinates(
-                        variable_with_anonymous_dims,
-                        varinfo,
-                        dimensions_file,
-                        varinfo.get_variable(latitude_coordinates[0]),
-                        varinfo.get_variable(longitude_coordinates[0]),
-                        index_ranges,
-                        bounding_box=bounding_box,
-                        shape_file_path=shape_file_path,
-                    )
+            try:
+                latitude_coordinates, longitude_coordinates = get_coordinate_variables(
+                    varinfo, [variable_with_anonymous_dims]
                 )
 
+                if latitude_coordinates and longitude_coordinates:
+                    index_ranges.update(
+                        get_x_y_index_ranges_from_coordinates(
+                            variable_with_anonymous_dims,
+                            varinfo,
+                            dimensions_file,
+                            varinfo.get_variable(latitude_coordinates[0]),
+                            varinfo.get_variable(longitude_coordinates[0]),
+                            index_ranges,
+                            bounding_box=bounding_box,
+                            shape_file_path=shape_file_path,
+                        )
+                    )
+            except InvalidRequestedRange:
+                out_of_range_variables.add(variable_with_anonymous_dims)
+
+    if out_of_range_variables:
+        raise NoDataException(
+            f'Spatial subset request outside supported dimension range for {out_of_range_variables}'
+        )
+
     return index_ranges
 
 

hoss/subset.py

@@ -7,6 +7,7 @@
 
 from typing import List, Set
 
+from harmony_service_lib.exceptions import NoDataException
 from harmony_service_lib.message import Message, Source
 from harmony_service_lib.message import Variable as HarmonyVariable
 from harmony_service_lib.message_utility import rgetattr
@@ -217,8 +218,14 @@ def get_required_variables(
     requested_variables = {
         f'/{variable.fullPath.lstrip("/")}' for variable in variables
     }
-
-    if request_is_index_subset and len(requested_variables) == 0:
+    # If request includes variable subsetting, check that all requested
+    # variables exist in the granule.
+    if requested_variables:
+        check_requested_variables_in_granule(varinfo, requested_variables)
+
+    # Otherwise, if request is an index subset and no variables are requested,
+    # include all variables.
+    elif request_is_index_subset:
         requested_variables = varinfo.get_science_variables().union(
             varinfo.get_metadata_variables()
         )
@@ -230,6 +237,25 @@ def get_required_variables(
     return varinfo.get_required_variables(requested_variables)
 
 
+def check_requested_variables_in_granule(
+    varinfo: VarInfoFromDmr, requested_variables: List[str]
+) -> bool:
+    """Return True if all variables are in the granule. Raise NoDataException
+    if any of the requested variables are not in the granule.
+
+    """
+    invalid_requested_variables = {
+        variable_name
+        for variable_name in requested_variables
+        if varinfo.get_variable(variable_name) is None
+    }
+    if invalid_requested_variables:
+        raise NoDataException(
+            f'Requested variables:{invalid_requested_variables} not found in granule'
+        )
+    return True
+
+
 def fill_variables(
     output_path: str,
     varinfo: VarInfoFromDmr,

hoss/temporal.py

@@ -13,6 +13,7 @@
 
 from dateutil import tz
 from dateutil.parser import parse as parse_datetime
+from harmony_service_lib.exceptions import NoDataException
 from harmony_service_lib.message import Message
 from netCDF4 import Dataset
 from varinfo import VarInfoFromDmr
@@ -22,7 +23,7 @@
     get_dimension_bounds,
     get_dimension_index_range,
 )
-from hoss.exceptions import UnsupportedTemporalUnits
+from hoss.exceptions import InvalidRequestedRange, UnsupportedTemporalUnits
 
 units_day = {'day', 'days', 'd'}
 units_hour = {'hour', 'hours', 'hr', 'h'}
@@ -54,6 +55,7 @@ def get_temporal_index_ranges(
 
     """
     index_ranges = {}
+    out_of_range_dim_variables = set()
     temporal_dimensions = varinfo.get_temporal_dimensions(required_variables)
 
     time_start = get_datetime_with_timezone(
@@ -65,22 +67,34 @@ def get_temporal_index_ranges(
 
     with Dataset(dimensions_path, 'r') as dimensions_file:
         for dimension in temporal_dimensions:
-            time_variable = varinfo.get_variable(dimension)
-            time_ref, time_delta = get_time_ref(
-                time_variable.get_attribute_value('units')
-            )
-
-            # Convert the Harmony message start and end datetime values into
-            # integer or floating point values (e.g., a number of seconds since
-            # 1970-01-01) using the variable epoch and unit.
-            minimum_extent = (time_start - time_ref) / time_delta
-            maximum_extent = (time_end - time_ref) / time_delta
-
-            index_ranges[dimension] = get_dimension_index_range(
-                dimensions_file[dimension][:],
-                minimum_extent,
-                maximum_extent,
-                bounds_values=get_dimension_bounds(dimension, varinfo, dimensions_file),
+            try:
+                time_variable = varinfo.get_variable(dimension)
+                time_ref, time_delta = get_time_ref(
+                    time_variable.get_attribute_value('units')
+                )
+
+                # Convert the Harmony message start and end datetime values into
+                # integer or floating point values (e.g., a number of seconds since
+                # 1970-01-01) using the variable epoch and unit.
+                minimum_extent = (time_start - time_ref) / time_delta
+                maximum_extent = (time_end - time_ref) / time_delta
+
+                index_ranges[dimension] = get_dimension_index_range(
+                    dimensions_file[dimension][:],
+                    minimum_extent,
+                    maximum_extent,
+                    bounds_values=get_dimension_bounds(
+                        dimension, varinfo, dimensions_file
+                    ),
+                )
+
+            except InvalidRequestedRange:
+                out_of_range_dim_variables.add(dimension)
+
+        if out_of_range_dim_variables:
+            raise NoDataException(
+                f'Temporal range request outside supported dimension range for '
+                f'{out_of_range_dim_variables}'
             )
 
     return index_ranges

pip_requirements.txt

@@ -1,7 +1,7 @@
 # This file should contain requirements to be installed via Pip.
 # Open source packages available from PyPI
-earthdata-varinfo ~= 3.1.0
-harmony-service-lib ~= 2.9.0
+earthdata-varinfo ~= 3.3.1
+harmony-service-lib ~= 2.11.0
 netCDF4 ~= 1.7.2
 numpy ~= 2.2.6
 pyproj ~= 3.7.1