Use cached namespaces during validation (#531)

* update pynwb validation to use cached namespaces * add zarr check when searching for nwbfiles * add dir test, fix parallel test indent * update zarr dir detection * update CHANGELOG.md --------- Co-authored-by: Ryan Ly <[email protected]>
NeurodataWithoutBorders · Nov 20, 2024 · 3aa3dab · 3aa3dab
1 parent c26924b
commit 3aa3dab
Show file tree

Hide file tree

Showing 3 changed files with 84 additions and 68 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -18,6 +18,7 @@
 
 ### Fixes
 * Fixed incorrect error message for OptogeneticStimulusSite. [#524](https://github.com/NeurodataWithoutBorders/nwbinspector/pull/524)
+* Fixed detection of Zarr directories for inspection. [#531](https://github.com/NeurodataWithoutBorders/nwbinspector/pull/531)
 
 
 # v0.5.2

diff --git a/src/nwbinspector/_nwb_inspection.py b/src/nwbinspector/_nwb_inspection.py
@@ -9,12 +9,13 @@
 from warnings import filterwarnings, warn
 
 import pynwb
+from hdmf_zarr import ZarrIO
 from natsort import natsorted
 from tqdm import tqdm
 
 from ._configuration import configure_checks
 from ._registration import Importance, InspectorMessage, available_checks
-from .tools._read_nwbfile import read_nwbfile, read_nwbfile_and_io
+from .tools._read_nwbfile import read_nwbfile
 from .utils import (
     OptionalListOfStrings,
     PathType,
@@ -126,7 +127,9 @@ def inspect_all(
     if progress_bar_options is None:
         progress_bar_options = dict(position=0, leave=False)
 
-    if in_path.is_dir():
+    if in_path.is_dir() and (in_path.match("*.nwb*")) and ZarrIO.can_read(in_path):
+        nwbfiles = [in_path]  # if it is a zarr directory
+    elif in_path.is_dir():
         nwbfiles = list(in_path.rglob("*.nwb*"))
 
         # Remove any macOS sidecar files
@@ -271,10 +274,10 @@ def inspect_nwbfile(
     filterwarnings(action="ignore", message="Ignoring cached namespace .*")
 
     try:
-        in_memory_nwbfile, io = read_nwbfile_and_io(nwbfile_path=nwbfile_path)
+        in_memory_nwbfile = read_nwbfile(nwbfile_path=nwbfile_path)
 
         if not skip_validate:
-            validation_errors = pynwb.validate(io=io)
+            validation_errors, _ = pynwb.validate(paths=[nwbfile_path])
             for validation_error in validation_errors:
                 yield InspectorMessage(
                     message=validation_error.reason,

diff --git a/tests/test_inspector.py b/tests/test_inspector.py
@@ -268,74 +268,86 @@ def test_inspect_all(self):
         ]
         self.assertCountEqual(first=test_results, second=true_results)
 
-        def test_inspect_all_parallel(self):
-            test_results = list(
-                inspect_all(path=Path(self.nwbfile_paths[0]).parent, select=[x.__name__ for x in self.checks], n_jobs=2)
+    def test_inspect_all_parallel(self):
+        test_results = list(
+            inspect_all(
+                path=Path(self.nwbfile_paths[0]).parent,
+                select=[x.__name__ for x in self.checks],
+                n_jobs=2,
+                skip_validate=self.skip_validate,
             )
-            true_results = [
-                InspectorMessage(
-                    message="data is not compressed. Consider enabling compression when writing a dataset.",
-                    importance=Importance.BEST_PRACTICE_SUGGESTION,
-                    severity=Severity.LOW,
-                    check_function_name="check_small_dataset_compression",
-                    object_type="TimeSeries",
-                    object_name="test_time_series_1",
-                    location="/acquisition/test_time_series_1",
-                    file_path=self.nwbfile_paths[0],
-                ),
-                InspectorMessage(
-                    message=(
-                        "TimeSeries appears to have a constant sampling rate. Consider specifying starting_time=1.2 "
-                        "and rate=0.5 instead of timestamps."
-                    ),
-                    importance=Importance.BEST_PRACTICE_VIOLATION,
-                    severity=Severity.LOW,
-                    check_function_name="check_regular_timestamps",
-                    object_type="TimeSeries",
-                    object_name="test_time_series_2",
-                    location="/acquisition/test_time_series_2",
-                    file_path=self.nwbfile_paths[0],
-                ),
-                InspectorMessage(
-                    message=(
-                        "Data may be in the wrong orientation. Time should be in the first dimension, and is usually "
-                        "the longest dimension. Here, another dimension is longer."
-                    ),
-                    importance=Importance.CRITICAL,
-                    severity=Severity.LOW,
-                    check_function_name="check_data_orientation",
-                    object_type="SpatialSeries",
-                    object_name="my_spatial_series",
-                    location="/processing/behavior/Position/my_spatial_series",
-                    file_path=self.nwbfile_paths[0],
+        )
+        true_results = [
+            InspectorMessage(
+                message="data is not compressed. Consider enabling compression when writing a dataset.",
+                importance=Importance.BEST_PRACTICE_SUGGESTION,
+                severity=Severity.LOW,
+                check_function_name="check_small_dataset_compression",
+                object_type="TimeSeries",
+                object_name="test_time_series_1",
+                location="/acquisition/test_time_series_1",
+                file_path=self.nwbfile_paths[0],
+            ),
+            InspectorMessage(
+                message=(
+                    "TimeSeries appears to have a constant sampling rate. Consider specifying starting_time=1.2 "
+                    "and rate=0.5 instead of timestamps."
                 ),
-                InspectorMessage(
-                    message=(
-                        "The length of the first dimension of data (4) does not match the length of timestamps (3)."
-                    ),
-                    importance=Importance.CRITICAL,
-                    severity=Severity.LOW,
-                    check_function_name="check_timestamps_match_first_dimension",
-                    object_type="TimeSeries",
-                    object_name="test_time_series_3",
-                    location="/acquisition/test_time_series_3",
-                    file_path=self.nwbfile_paths[0],
+                importance=Importance.BEST_PRACTICE_VIOLATION,
+                severity=Severity.LOW,
+                check_function_name="check_regular_timestamps",
+                object_type="TimeSeries",
+                object_name="test_time_series_2",
+                location="/acquisition/test_time_series_2",
+                file_path=self.nwbfile_paths[0],
+            ),
+            InspectorMessage(
+                message=(
+                    "Data may be in the wrong orientation. Time should be in the first dimension, and is usually "
+                    "the longest dimension. Here, another dimension is longer."
                 ),
-                InspectorMessage(
-                    message=(
-                        "TimeSeries appears to have a constant sampling rate. Consider specifying starting_time=1.2 "
-                        "and rate=0.5 instead of timestamps."
-                    ),
-                    importance=Importance.BEST_PRACTICE_VIOLATION,
-                    severity=Severity.LOW,
-                    check_function_name="check_regular_timestamps",
-                    object_type="TimeSeries",
-                    object_name="test_time_series_2",
-                    location="/acquisition/test_time_series_2",
-                    file_path=self.nwbfile_paths[1],
+                importance=Importance.CRITICAL,
+                severity=Severity.LOW,
+                check_function_name="check_data_orientation",
+                object_type="SpatialSeries",
+                object_name="my_spatial_series",
+                location="/processing/behavior/Position/my_spatial_series",
+                file_path=self.nwbfile_paths[0],
+            ),
+            InspectorMessage(
+                message=("The length of the first dimension of data (4) does not match the length of timestamps (3)."),
+                importance=Importance.CRITICAL,
+                severity=Severity.LOW,
+                check_function_name="check_timestamps_match_first_dimension",
+                object_type="TimeSeries",
+                object_name="test_time_series_3",
+                location="/acquisition/test_time_series_3",
+                file_path=self.nwbfile_paths[0],
+            ),
+            InspectorMessage(
+                message=(
+                    "TimeSeries appears to have a constant sampling rate. Consider specifying starting_time=1.2 "
+                    "and rate=0.5 instead of timestamps."
                 ),
-            ]
-            self.assertCountEqual(first=test_results, second=true_results)
+                importance=Importance.BEST_PRACTICE_VIOLATION,
+                severity=Severity.LOW,
+                check_function_name="check_regular_timestamps",
+                object_type="TimeSeries",
+                object_name="test_time_series_2",
+                location="/acquisition/test_time_series_2",
+                file_path=self.nwbfile_paths[1],
+            ),
+        ]
+        self.assertCountEqual(first=test_results, second=true_results)
+
+    def test_inspect_all_directory(self):
+        """Test that inspect_all will find the file when given a valid path (in the case of Zarr, this path may be a directory)."""
+        test_results = list(
+            inspect_all(
+                path=self.nwbfile_paths[0], select=[x.__name__ for x in self.checks], skip_validate=self.skip_validate
+            )
+        )
+        self.assertGreater(len(test_results), 0)
 
     def test_inspect_nwbfile(self):
         test_results = list(