Add prov

fairscape_models/activity.py

@@ -1,5 +1,5 @@
-from pydantic import BaseModel, Field, ConfigDict
-from typing import Optional, List
+from pydantic import BaseModel, Field, ConfigDict, model_validator
+from typing import Optional, List, Union
 
 from fairscape_models.fairscape_base import IdentifierValue
 
@@ -13,4 +13,8 @@ class Activity(BaseModel):
     generated: Optional[List[IdentifierValue]] = Field(default=[])
     isPartOf: Optional[List[IdentifierValue]] = Field(default=[])
 
-    model_config = ConfigDict(extra="allow")
+    # PROV-O fields (auto-populated)
+    used: Optional[List[Union[str, IdentifierValue]]] = Field(default=[], alias="prov:used")
+    wasAssociatedWith: Optional[List[Union[str, IdentifierValue]]] = Field(default=[], alias="prov:wasAssociatedWith")
+
+    model_config = ConfigDict(extra="allow", populate_by_name=True)

fairscape_models/annotation.py

@@ -1,12 +1,27 @@
-from pydantic import Field, ConfigDict
-from typing import Optional, List
+from pydantic import Field, ConfigDict, model_validator
+from typing import Optional, List, Union
 
 from fairscape_models.fairscape_base import IdentifierValue, ANNOTATION_TYPE
 from fairscape_models.activity import Activity
 
 class Annotation(Activity):
     metadataType: Optional[str] = Field(default="https://w3id.org/EVI#Annotation", alias="@type")
     additionalType: Optional[str] = Field(default=ANNOTATION_TYPE)
-    createdBy: str
+    createdBy: Union[str, IdentifierValue]
     dateCreated: str
     usedDataset: Optional[List[IdentifierValue]] = Field(default=[])
+
+    @model_validator(mode='after')
+    def populate_prov_fields(self):
+        """Auto-populate PROV-O fields from EVI fields"""
+        # Map usedDataset to prov:used (preserving their types)
+        if self.usedDataset:
+            self.used = self.usedDataset
+        else:
+            self.used = []
+
+        # Map createdBy to prov:wasAssociatedWith (preserve type: str or IdentifierValue)
+        if self.createdBy:
+            self.wasAssociatedWith = [self.createdBy]
+
+        return self

fairscape_models/computation.py

@@ -1,4 +1,4 @@
-from pydantic import Field, ConfigDict
+from pydantic import Field, ConfigDict, model_validator
 from typing import Optional, List, Union
 
 from fairscape_models.fairscape_base import IdentifierValue, COMPUTATION_TYPE
@@ -7,10 +7,28 @@
 class Computation(Activity):
     metadataType: Optional[str] = Field(default="https://w3id.org/EVI#Computation", alias="@type")
     additionalType: Optional[str] = Field(default=COMPUTATION_TYPE)
-    runBy: str
+    runBy: Union[str, IdentifierValue]
     dateCreated: str
     additionalDocumentation: Optional[str] = Field(default=None)
     command: Optional[Union[List[str], str]] = Field(default=None)
     usedSoftware: Optional[List[IdentifierValue]] = Field(default=[])
     usedMLModel: Optional[List[IdentifierValue]] = Field(default=[])
     usedDataset: Optional[List[IdentifierValue]] = Field(default=[])
+
+    @model_validator(mode='after')
+    def populate_prov_fields(self):
+        """Auto-populate PROV-O fields from EVI fields"""
+        # Aggregate all inputs into prov:used
+        used_items = []
+        if self.usedSoftware:
+            used_items.extend(self.usedSoftware)
+        if self.usedMLModel:
+            used_items.extend(self.usedMLModel)
+        if self.usedDataset:
+            used_items.extend(self.usedDataset)
+        self.used = used_items
+
+        if self.runBy:
+            self.wasAssociatedWith = [self.runBy]
+
+        return self

fairscape_models/dataset.py

@@ -1,4 +1,4 @@
-from pydantic import Field, ConfigDict, AliasChoices
+from pydantic import Field, ConfigDict, AliasChoices, model_validator
 from typing import Optional, List, Union
 
 from fairscape_models.fairscape_base import IdentifierValue, DATASET_TYPE
@@ -16,4 +16,30 @@ class Dataset(DigitalObject):
         default=None
     )
     generatedBy: Optional[Union[IdentifierValue, List[IdentifierValue]]] = Field(default=[])
-    derivedFrom: Optional[List[IdentifierValue]] = Field(default=[])
+    derivedFrom: Optional[List[IdentifierValue]] = Field(default=[])
+
+    @model_validator(mode='after')
+    def populate_prov_fields(self):
+        """Auto-populate PROV-O fields from EVI fields"""
+        # Map generatedBy → prov:wasGeneratedBy
+        if self.generatedBy:
+            if isinstance(self.generatedBy, list):
+                self.wasGeneratedBy = self.generatedBy
+            else:
+                self.wasGeneratedBy = [self.generatedBy]
+        else:
+            self.wasGeneratedBy = []
+
+        # Map derivedFrom → prov:wasDerivedFrom
+        self.wasDerivedFrom = self.derivedFrom or []
+
+        # Map author
+        if self.author:
+            if isinstance(self.author, str):
+                self.wasAttributedTo = [IdentifierValue(**{"@id": self.author})]
+            elif isinstance(self.author, list):
+                self.wasAttributedTo = [IdentifierValue(**{"@id": a}) for a in self.author]
+        else:
+            self.wasAttributedTo = []
+
+        return self

fairscape_models/digital_object.py

@@ -1,4 +1,4 @@
-from pydantic import BaseModel, Field, ConfigDict
+from pydantic import BaseModel, Field, ConfigDict, model_validator
 from typing import Optional, List, Union
 
 from fairscape_models.fairscape_base import IdentifierValue
@@ -8,7 +8,7 @@ class DigitalObject(BaseModel):
     guid: str = Field(alias="@id")
     name: str
     metadataType: Optional[str] = Field(default=None, alias="@type")
-    author: Union[str, List[str]]
+    author: Union[str, IdentifierValue, List[Union[str, IdentifierValue]]]
     description: str = Field(min_length=10)
     version: str = Field(default="0.1.0")
     associatedPublication: Optional[Union[str, List[str]]] = Field(default=None)
@@ -17,4 +17,9 @@ class DigitalObject(BaseModel):
     isPartOf: Optional[List[IdentifierValue]] = Field(default=[])
     usedByComputation: Optional[List[IdentifierValue]] = Field(default=[])
 
-    model_config = ConfigDict(extra="allow")
+    # PROV-O fields (auto-populated)
+    wasGeneratedBy: Optional[List[Union[str, IdentifierValue]]] = Field(default=[], alias="prov:wasGeneratedBy")
+    wasDerivedFrom: Optional[List[Union[str, IdentifierValue]]] = Field(default=[], alias="prov:wasDerivedFrom")
+    wasAttributedTo: Optional[List[Union[str, IdentifierValue]]] = Field(default=[], alias="prov:wasAttributedTo")
+
+    model_config = ConfigDict(extra="allow", populate_by_name=True)

fairscape_models/experiment.py

@@ -1,15 +1,35 @@
-from pydantic import Field, ConfigDict
-from typing import Optional, List
+from pydantic import Field, ConfigDict, model_validator
+from typing import Optional, List, Union
 from fairscape_models.fairscape_base import IdentifierValue
 from fairscape_models.activity import Activity
 
 class Experiment(Activity):
     metadataType: Optional[str] = Field(default="https://w3id.org/EVI#Experiment", alias="@type")
     experimentType: str
-    runBy: str
+    runBy: Union[str, IdentifierValue]
     datePerformed: str
     protocol: Optional[str] = Field(default=None)
     usedInstrument: Optional[List[IdentifierValue]] = Field(default=[])
     usedSample: Optional[List[IdentifierValue]] = Field(default=[])
     usedTreatment: Optional[List[IdentifierValue]] = Field(default=[])
-    usedStain: Optional[List[IdentifierValue]] = Field(default=[])
+    usedStain: Optional[List[IdentifierValue]] = Field(default=[])
+
+    @model_validator(mode='after')
+    def populate_prov_fields(self):
+        """Auto-populate PROV-O fields from EVI fields"""
+        # Aggregate all inputs into prov:used
+        used_items = []
+        if self.usedInstrument:
+            used_items.extend(self.usedInstrument)
+        if self.usedSample:
+            used_items.extend(self.usedSample)
+        if self.usedTreatment:
+            used_items.extend(self.usedTreatment)
+        if self.usedStain:
+            used_items.extend(self.usedStain)
+        self.used = used_items
+
+        if self.runBy:
+            self.wasAssociatedWith = [self.runBy]
+
+        return self

fairscape_models/fairscape_base.py

@@ -32,6 +32,7 @@
     "@vocab": "https://schema.org/",
     "evi": "https://w3id.org/EVI#",
     "rai": "http://mlcommons.org/croissant/RAI/",
+    "prov": "http://www.w3.org/ns/prov#",
 
     # TODO fully specify default context
     "usedSoftware": {

fairscape_models/mlmodel.py

@@ -1,4 +1,4 @@
-from pydantic import Field, ConfigDict
+from pydantic import Field, ConfigDict, model_validator
 from typing import Optional, List, Union
 
 from fairscape_models.fairscape_base import IdentifierValue, MLMODEL_TYPE
@@ -12,3 +12,32 @@ class MLModel(DigitalObject):
     modelTask: Optional[str] = Field(default=None)
     modelArchitecture: Optional[str] = Field(default=None)
     trainedOn: Optional[List[IdentifierValue]] = Field(default=[])
+    generatedBy: Optional[Union[IdentifierValue, List[IdentifierValue]]] = Field(default=[])
+    derivedFrom: Optional[List[IdentifierValue]] = Field(default=[])
+
+    @model_validator(mode='after')
+    def populate_prov_fields(self):
+        """Auto-populate PROV-O fields from EVI fields"""
+
+        # Map generatedBy → prov:wasGeneratedBy
+        if self.generatedBy:
+            if isinstance(self.generatedBy, list):
+                self.wasGeneratedBy = self.generatedBy
+            else:
+                self.wasGeneratedBy = [self.generatedBy]
+        else:
+            self.wasGeneratedBy = []
+
+        # Map derivedFrom → prov:wasDerivedFrom
+        self.wasDerivedFrom = self.derivedFrom or []
+
+        # Map author → prov:wasAttributedTo
+        if self.author:
+            if isinstance(self.author, str):
+                self.wasAttributedTo = [IdentifierValue(**{"@id": self.author})]
+            elif isinstance(self.author, list):
+                self.wasAttributedTo = [IdentifierValue(**{"@id": a}) for a in self.author]
+        else:
+            self.wasAttributedTo = []
+
+        return self

fairscape_models/rocrate.py

@@ -245,14 +245,25 @@ def cleanIdentifierUnion(identifier_union):
 
                 cleanIdentifierUnion(elem.generatedBy)
 
+                # Clean PROV fields
+                cleanIdentifierList(elem.wasGeneratedBy)
+                cleanIdentifierList(elem.wasDerivedFrom)
+                cleanIdentifierList(elem.wasAttributedTo)
+
             if isinstance(elem, Software):
                 cleanIdentifierList(elem.usedByComputation)
 
+                # Clean PROV fields
+                cleanIdentifierList(elem.wasAttributedTo)
+
             if isinstance(elem, MLModel):
                 cleanIdentifierList(elem.usedByComputation)
 
                 cleanIdentifierList(elem.trainedOn)
 
+                # Clean PROV fields
+                cleanIdentifierList(elem.wasAttributedTo)
+
             if isinstance(elem, Computation):
 
                 cleanIdentifierList(elem.usedDataset)
@@ -263,12 +274,20 @@ def cleanIdentifierUnion(identifier_union):
 
                 cleanIdentifierList(elem.usedMLModel)
 
+                # Clean PROV fields
+                cleanIdentifierList(elem.used)
+                cleanIdentifierList(elem.wasAssociatedWith)
+
             if isinstance(elem, Annotation):
 
                 cleanIdentifierList(elem.usedDataset)
 
                 cleanIdentifierList(elem.generated)
 
+                # Clean PROV fields
+                cleanIdentifierList(elem.used)
+                cleanIdentifierList(elem.wasAssociatedWith)
+
             if isinstance(elem, Experiment):
 
                 cleanIdentifierList(elem.usedInstrument)
@@ -281,6 +300,10 @@ def cleanIdentifierUnion(identifier_union):
 
                 cleanIdentifierList(elem.generated)
 
+                # Clean PROV fields
+                cleanIdentifierList(elem.used)
+                cleanIdentifierList(elem.wasAssociatedWith)
+
     def getCrateMetadata(self)-> ROCrateMetadataElem:
         """ Filter the Metadata Graph for the Metadata Element Describing the Toplevel ROCrate
 

fairscape_models/software.py

@@ -1,4 +1,4 @@
-from pydantic import Field, ConfigDict
+from pydantic import Field, ConfigDict, model_validator
 from typing import Optional, List
 
 from fairscape_models.fairscape_base import IdentifierValue, SOFTWARE_TYPE
@@ -7,5 +7,20 @@
 class Software(DigitalObject):
     metadataType: Optional[str] = Field(default="https://w3id.org/EVI#Software", alias="@type")
     additionalType: Optional[str] = Field(default=SOFTWARE_TYPE)
-    dateModified: Optional[str]
+    dateModified: Optional[str] = None
     fileFormat: str = Field(title="fileFormat", alias="format")
+
+    @model_validator(mode='after')
+    def populate_prov_fields(self):
+        """Auto-populate PROV-O fields from EVI fields"""
+
+        # Map author → prov:wasAttributedTo
+        if self.author:
+            if isinstance(self.author, list):
+                self.wasAttributedTo = self.author
+            else:
+                self.wasAttributedTo = [self.author]
+        else:
+            self.wasAttributedTo = []
+
+        return self

tests/test_annotation.py

@@ -1,15 +1,37 @@
 import pytest
 from pydantic import ValidationError
 from fairscape_models.annotation import Annotation
+from fairscape_models.fairscape_base import IdentifierValue
 
 def test_annotation_instantiation(annotation_minimal_data):
     """Test successful instantiation of an Annotation model."""
     annotation = Annotation.model_validate(annotation_minimal_data)
     assert annotation.guid == annotation_minimal_data["@id"]
     assert annotation.description == annotation_minimal_data["description"]
 
+    # Test PROV field auto-population
+    assert annotation.used == []  # No usedDataset provided
+    assert len(annotation.wasAssociatedWith) == 1
+    assert annotation.wasAssociatedWith[0] == annotation_minimal_data["createdBy"]
+
 def test_annotation_short_description(annotation_minimal_data):
     """Test that a short description raises a ValidationError."""
     annotation_minimal_data["description"] = "too short"
     with pytest.raises(ValidationError):
         Annotation.model_validate(annotation_minimal_data)
+
+def test_annotation_with_datasets(annotation_minimal_data):
+    """Test PROV field population with usedDataset."""
+    annotation_minimal_data["usedDataset"] = [
+        {"@id": "ark:59852/dataset-1"},
+        {"@id": "ark:59852/dataset-2"}
+    ]
+
+    annotation = Annotation.model_validate(annotation_minimal_data)
+
+    # Test PROV:used is populated from usedDataset
+    assert len(annotation.used) == 2
+    assert all(isinstance(item, IdentifierValue) for item in annotation.used)
+    used_ids = [item.guid for item in annotation.used]
+    assert "ark:59852/dataset-1" in used_ids
+    assert "ark:59852/dataset-2" in used_ids