yaml-validation

README.md

@@ -3,3 +3,21 @@
 Converts data from SpikeGadgets to the NWB Data Format
 
 This will supersede [rec_to_nwb](https://github.com/LorenFrankLab/rec_to_nwb) once it is operational.
+
+## Notes
+
+### Unit Test
+
+Unit test are quick-running test to ensure that no code is broken as new code is added. It is recommended you run them frequently and before and after adding new code. To run the unit test, run -
+
+```bash
+pytest src/spikegadgets_to_nwb/tests/unit_tests/
+```
+
+### Integration Test
+
+Integration test are to ensure the system works. Functionality and third-party libraries are tested together and noting is mocked. They are best ran prior to pushing a commit as they can be slow. To run the test, run -
+
+```bash
+pytest src/spikegadgets_to_nwb/tests/integration_tests/
+```

environment.yml

@@ -1,18 +1,19 @@
 name: spikegadgets_to_nwb
 channels:
-- defaults
-- conda-forge
-- franklab
+  - defaults
+  - conda-forge
+  - franklab
 dependencies:
-- numpy
-- scipy
-- pandas
-- pynwb
-- ipython
-- black
-- pytest
-- pytest-cov
-- pip
-- pyyaml
-- pip:
-    - ndx-franklab-novela
+  - numpy
+  - scipy
+  - pandas
+  - pynwb
+  - ipython
+  - black
+  - pytest
+  - pytest-cov
+  - pip
+  - pyyaml
+  - jsonschema
+  - pip:
+      - ndx-franklab-novela

pyproject.toml

@@ -50,3 +50,22 @@ exclude = [".git_archival.txt"]
 [tool.hatch.build.targets.wheel]
 packages = ["src/spikegadgets_to_nwb"]
 exclude = ["src/spikegadgets_to_nwb/data/test_data"]
+
+[tool.black]
+exclude = '''
+
+(
+  /(
+      \.eggs         # exclude a few common directories in the
+    | \.git          # root of the project
+    | \.hg
+    | \.mypy_cache
+    | \.tox
+    | \.venv
+    | _build
+    | buck-out
+    | build
+    | dist
+  )/
+)
+'''

src/spikegadgets_to_nwb/convert_yaml.py

@@ -1,5 +1,6 @@
 import yaml
 from xml.etree import ElementTree
+from spikegadgets_to_nwb import metadata_validation
 
 from pynwb import NWBFile, TimeSeries
 from pynwb.file import Subject, ProcessingModule
@@ -39,6 +40,11 @@ def load_metadata(
     """
     with open(metadata_path, "r") as stream:
         metadata = yaml.safe_load(stream)
+
+    is_metadata_valid, metadata_errors = metadata_validation.validate(metadata)
+    if not is_metadata_valid:
+        raise ValueError("".join(metadata_errors))
+
     probe_metadata = []
     for path in probe_metadata_paths:
         with open(path, "r") as stream:

src/spikegadgets_to_nwb/metadata_validation.py

@@ -0,0 +1,74 @@
+import jsonschema
+import os
+import yaml
+import copy
+import datetime
+
+
+def _get_nwb_json_schema_path() -> str:
+    """Get the NWB JSON Schema file path
+
+    Returns
+    -------
+    str
+        NWB Schema file Path
+    """
+    current_path = os.path.dirname(os.path.abspath(__file__))
+    json_schema_file = "/./nwb_schema.json"
+    return f"{current_path}{os.path.normpath(json_schema_file)}"
+
+
+def _get_json_schema() -> str:
+    """Get JSON Schema
+
+    Returns
+    -------
+    str
+        JSON Schema content
+    """
+    json_schema = None
+    json_schema_path = _get_nwb_json_schema_path()
+    with open(json_schema_path, "r") as stream:
+        json_schema = yaml.safe_load(stream)
+    return json_schema
+
+
+def validate(metadata: dict) -> tuple:
+    """Validates metadata
+
+    Parameters
+    ----------
+    metadata : dict
+        metadata documenting the particulars of a session
+
+    Returns
+    -------
+    tuple
+        information of the validity of the metadata data and any errors
+    """
+    assert metadata is not None  # metadata cannot be null
+    assert isinstance(metadata, dict)  # cannot proceed if metadata is not a dictionary
+
+    # date_of_birth is set to a datetime by the YAML-to-dict converter.
+    # This code converts date_of_birth  to string
+    metadata_content = copy.deepcopy(metadata) or {}
+    if (
+        metadata_content["subject"]
+        and metadata_content["subject"]["date_of_birth"]
+        and type(metadata_content["subject"]["date_of_birth"]) is datetime.datetime
+    ):
+        metadata_content["subject"]["date_of_birth"] = (
+            metadata_content["subject"]["date_of_birth"].utcnow().isoformat()
+        )
+
+    schema = _get_json_schema()
+    validator = jsonschema.Draft202012Validator(schema)
+    metadata_validation_errors = validator.iter_errors(metadata_content)
+    errors = []
+
+    for metadata_validation_error in metadata_validation_errors:
+        errors.append(metadata_validation_error.message)
+
+    is_valid = len(errors) == 0
+
+    return is_valid, errors