Merge pull request #252 from dandi/enh-ngff

Add support for detection and addition of ome/ngff "standard" into assets summary
dandi · Nov 4, 2024 · 4b17876 · 4b17876
2 parents d3cc4b4 + f82c121
commit 4b17876
Show file tree

Hide file tree

Showing 3 changed files with 22 additions and 8 deletions.
diff --git a/dandischema/metadata.py b/dandischema/metadata.py
@@ -380,13 +380,18 @@ def _add_asset_to_stats(assetmeta: Dict[str, Any], stats: _stats_type) -> None:
                 stats["tissuesample"].append(sample)
 
     stats["dataStandard"] = stats.get("dataStandard", [])
+
+    def add_if_missing(standard: dict) -> None:
+        if standard not in stats["dataStandard"]:
+            stats["dataStandard"].append(standard)
+
     if "nwb" in assetmeta["encodingFormat"]:
-        if models.nwb_standard not in stats["dataStandard"]:
-            stats["dataStandard"].append(models.nwb_standard)
+        add_if_missing(models.nwb_standard)
     # TODO: RF assumption that any .json implies BIDS
     if set(Path(assetmeta["path"]).suffixes).intersection((".json", ".nii")):
-        if models.bids_standard not in stats["dataStandard"]:
-            stats["dataStandard"].append(models.bids_standard)
+        add_if_missing(models.bids_standard)
+    if Path(assetmeta["path"]).suffixes == [".ome", ".zarr"]:
+        add_if_missing(models.ome_ngff_standard)
 
 
 # TODO?: move/bind such helpers as .from_metadata or alike within

diff --git a/dandischema/models.py b/dandischema/models.py
@@ -815,6 +815,11 @@ class StandardsType(BaseType):
     identifier="RRID:SCR_016124",
 ).model_dump(mode="json", exclude_none=True)
 
+ome_ngff_standard = StandardsType(
+    name="OME/NGFF Standard",
+    identifier="DOI:10.25504/FAIRsharing.9af712",
+).model_dump(mode="json", exclude_none=True)
+
 
 class ContactPoint(DandiBaseModel):
     email: Optional[EmailStr] = Field(

diff --git a/dandischema/tests/test_metadata.py b/dandischema/tests/test_metadata.py
@@ -542,7 +542,7 @@ def test_aggregation_bids() -> None:
         {
             "id": "dandiasset:6668d37f-e842-4b73-8c20-082a1dd0d31a",
             "path": "sub-MITU01/ses-20210703h01m05s04/microscopy/sub-MITU01_"
-            "run-1_sample-163_stain-YO_chunk-5_spim.h5",
+            "run-1_sample-163_stain-YO_chunk-5_spim.ome.zarr",
             "access": [
                 {"status": "dandi:OpenAccess", "schemaKey": "AccessRequirements"}
             ],
@@ -565,7 +565,7 @@ def test_aggregation_bids() -> None:
             "contentSize": 38474544973,
             "dateModified": "2021-07-22T23:59:16.060551-04:00",
             "schemaVersion": "0.4.4",
-            "encodingFormat": "application/x-hdf5",
+            "encodingFormat": "application/x-zarr",
             "wasGeneratedBy": [
                 {
                     "id": "urn:uuid:aef77d59-7a7f-4320-9d4b-9b03f3e25e54",
@@ -588,7 +588,7 @@ def test_aggregation_bids() -> None:
         {
             "id": "dandiasset:84dd580f-8d4a-43f8-bda3-6fb53fb5d3a2",
             "path": "sub-MITU01/ses-20210703h16m32s10/microscopy/sub-MITU01_"
-            "ses-20210703h16m32s10_run-1_sample-162_stain-LEC_chunk-5_spim.h5",
+            "ses-20210703h16m32s10_run-1_sample-162_stain-LEC_chunk-5_spim.ome.zarr",
             "access": [
                 {"status": "dandi:OpenAccess", "schemaKey": "AccessRequirements"}
             ],
@@ -611,7 +611,7 @@ def test_aggregation_bids() -> None:
             "contentSize": 61774316916,
             "dateModified": "2021-10-01T18:28:16.038990-04:00",
             "schemaVersion": "0.6.0",
-            "encodingFormat": "application/x-hdf5",
+            "encodingFormat": "application/x-zarr",
             "wasGeneratedBy": [
                 {
                     "id": "urn:uuid:8f69a248-0e6a-4fa1-8369-ae1cc63d59d8",
@@ -634,3 +634,7 @@ def test_aggregation_bids() -> None:
     ]
     summary = aggregate_assets_summary(data)
     assert summary["numberOfSamples"] == 2
+    assert (
+        sum(_.get("name", "").startswith("OME/NGFF") for _ in summary["dataStandard"])
+        == 1
+    )  # only a single entry so we do not duplicate them