dyff-schema 0.3.9__tar.gz → 0.5.0__tar.gz

{dyff_schema-0.3.9 → dyff_schema-0.5.0}/.gitlab-ci.yml

@@ -12,7 +12,7 @@ include:
     file:
       - gitlab-release.yml
   - project: buildgarden/pipelines/detect-secrets
-    ref: 0.1.0
+    ref: 0.2.1
     file:
       - detect-secrets.yml
   - project: buildgarden/pipelines/prettier

{dyff_schema-0.3.9/dyff_schema.egg-info → dyff_schema-0.5.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: dyff-schema
-Version: 0.3.9
+Version: 0.5.0
 Summary: Data models for the Dyff AI auditing platform.
 Author-email: Digital Safety Research Institute <contact@dsri.org>
 License: Apache-2.0

{dyff_schema-0.3.9 → dyff_schema-0.5.0}/dyff/schema/v0/r1/platform.py

@@ -30,7 +30,7 @@ from ... import named_data_schema, product_schema
 from ...version import SomeSchemaVersion
 from .base import DyffSchemaBaseModel
 from .dataset import arrow, make_item_type, make_response_type
-from .version import VERSION, Versioned
+from .version import SCHEMA_VERSION, SchemaVersion
 
 SYSTEM_ATTRIBUTES = frozenset(["creationTime", "status", "reason"])
 
@@ -139,15 +139,6 @@ def _k8s_label_value_maxlen():
     return _k8s_label_maxlen()
 
 
-class StorageSignedURL(DyffSchemaBaseModel):
-    url: str = pydantic.Field(description="The signed URL")
-    method: str = pydantic.Field(description="The HTTP method applicable to the URL")
-    headers: dict[str, str] = pydantic.Field(
-        default_factory=dict,
-        description="Mandatory headers that must be passed with the request",
-    )
-
-
 class Entities(str, enum.Enum):
     """The kinds of entities in the dyff system."""
 
@@ -158,6 +149,7 @@ class Entities(str, enum.Enum):
     DataSource = "DataSource"
     Dataset = "Dataset"
     Evaluation = "Evaluation"
+    Family = "Family"
     InferenceService = "InferenceService"
     InferenceSession = "InferenceSession"
     Measurement = "Measurement"
@@ -166,6 +158,7 @@ class Entities(str, enum.Enum):
     Module = "Module"
     Report = "Report"
     SafetyCase = "SafetyCase"
+    Tag = "Tag"
 
 
 class Resources(str, enum.Enum):
@@ -177,6 +170,7 @@ class Resources(str, enum.Enum):
     Dataset = "datasets"
     DataSource = "datasources"
     Evaluation = "evaluations"
+    Family = "families"
     InferenceService = "inferenceservices"
     InferenceSession = "inferencesessions"
     Measurement = "measurements"
@@ -185,6 +179,7 @@ class Resources(str, enum.Enum):
     Module = "modules"
     Report = "reports"
     SafetyCase = "safetycases"
+    Tag = "tags"
 
     Task = "tasks"
     """
@@ -303,7 +298,96 @@ class Status(DyffSchemaBaseModel):
     )
 
 
-class DyffEntity(Status, Labeled, Versioned, DyffModelWithID):
+class Documentation(DyffSchemaBaseModel):
+    title: Optional[str] = pydantic.Field(
+        default=None,
+        description='A short plain string suitable as a title or "headline".',
+    )
+
+    summary: Optional[str] = pydantic.Field(
+        default=None,
+        description="A brief summary, suitable for display in"
+        " small UI elements. Interpreted as Markdown. Excessively long"
+        " summaries may be truncated in the UI, especially on small displays.",
+    )
+
+    fullPage: Optional[str] = pydantic.Field(
+        default=None,
+        description="Long-form documentation. Interpreted as"
+        " Markdown. There are no length constraints, but be reasonable.",
+    )
+
+
+class Documented(DyffSchemaBaseModel):
+    documentation: Documentation = pydantic.Field(
+        default_factory=Documentation,
+        description="Documentation of the resource. The content is used to"
+        " populate various views in the web UI.",
+    )
+
+
+class FamilyMemberKind(str, enum.Enum):
+    """The kinds of entities that can be members of a Family.
+
+    These are resources for which it makes sense to have different versions or variants
+    that evolve over time.
+    """
+
+    Dataset = "Dataset"
+    InferenceService = "InferenceService"
+    Method = "Method"
+    Model = "Model"
+    Module = "Module"
+
+
+class FamilyMember(DyffSchemaBaseModel):
+    family: Optional[str] = pydantic.Field(
+        default=None,
+        description="ID of the Family to which the resource belongs.",
+    )
+
+
+class TagBase(DyffSchemaBaseModel):
+    tag: str = pydantic.Field(
+        description="An interpretable identifier for the tag that is unique in"
+        " the context of the corresponding Family."
+    )
+
+    resource: str = pydantic.Field(
+        description="ID of the resource this tag references.",
+    )
+
+    description: str = pydantic.Field(
+        description="A short description of the tag. Interpreted as Markdown."
+        " This should include information about how the tagged version is"
+        " different from other versions."
+    )
+
+
+class Tag(TagBase):
+    created: datetime = pydantic.Field(description="Tag creation time.")
+
+
+class Family(Labeled, SchemaVersion, DyffModelWithID):
+    kind: Literal["Family"] = "Family"
+
+    resourceKind: FamilyMemberKind = pydantic.Field(
+        description="The kind of resource that comprises the family.",
+    )
+
+    tags: list[Tag] = pydantic.Field(
+        default_factory=list,
+        description="Tags mapping interpretable names to resource IDs.",
+    )
+
+    documentation: Documentation = pydantic.Field(
+        default_factory=Documentation,
+        description="Documentation of the resource family. The content is used"
+        " to populate various views in the web UI.",
+    )
+
+
+class DyffEntity(Status, Labeled, SchemaVersion, DyffModelWithID):
     kind: Literal[
         "Analysis",
         "Audit",
@@ -382,6 +466,9 @@ class APIFunctions(str, enum.Enum):
     query = "query"
     """Query the resource collection."""
 
+    edit = "edit"
+    """Edit properties of existing resources."""
+
     download = "download"
     """
     .. deprecated:: 0.5.0
@@ -471,6 +558,43 @@ class APIKey(DyffSchemaBaseModel):
     )
 
 
+# ----------------------------------------------------------------------------
+
+
+class Digest(DyffSchemaBaseModel):
+    md5: Optional[str] = pydantic.Field(
+        default=None, description="md5 digest of artifact data"
+    )
+
+
+class Artifact(DyffSchemaBaseModel):
+    # TODO: In v1, rename this to 'contentType' or something and commit to making it the MIME type
+    kind: Optional[str] = pydantic.Field(
+        default=None, description="The kind of artifact"
+    )
+    path: str = pydantic.Field(
+        description="The relative path of the artifact within the tree"
+    )
+    digest: Digest = pydantic.Field(
+        default_factory=Digest,
+        description="One or more message digests (hashes) of the artifact data",
+    )
+
+
+class StorageSignedURL(DyffSchemaBaseModel):
+    url: str = pydantic.Field(description="The signed URL")
+    method: str = pydantic.Field(description="The HTTP method applicable to the URL")
+    headers: dict[str, str] = pydantic.Field(
+        default_factory=dict,
+        description="Mandatory headers that must be passed with the request",
+    )
+
+
+class ArtifactURL(DyffSchemaBaseModel):
+    artifact: Artifact
+    signedURL: StorageSignedURL
+
+
 class AuditRequirement(DyffSchemaBaseModel):
     """An evaluation report that must exist in order to apply an AuditProcedure."""
 
@@ -543,21 +667,6 @@ class ExtractorStep(DyffSchemaBaseModel):
     type: Optional[str] = None
 
 
-class Digest(DyffSchemaBaseModel):
-    md5: Optional[str] = pydantic.Field(None)
-
-
-class Artifact(DyffSchemaBaseModel):
-    # TODO: Enumerate the available kinds
-    kind: str = pydantic.Field(description="The kind of artifact")
-    path: str = pydantic.Field(
-        description="The relative path of the artifact within the dataset"
-    )
-    digest: Digest = pydantic.Field(
-        description="One or more message digests (hashes) of the artifact data"
-    )
-
-
 class DyffDataSchema(DyffSchemaBaseModel):
     components: list[str] = pydantic.Field(
         min_items=1,
@@ -565,7 +674,7 @@ class DyffDataSchema(DyffSchemaBaseModel):
         " the composition of these component schemas.",
     )
     schemaVersion: SomeSchemaVersion = pydantic.Field(
-        default=VERSION, description="The dyff schema version"
+        default=SCHEMA_VERSION, description="The dyff schema version"
     )
 
     def model_type(self) -> Type[DyffSchemaBaseModel]:
@@ -588,6 +697,12 @@ class DataSchema(DyffSchemaBaseModel):
         default=None, description="The schema in JSON Schema format"
     )
 
+    @staticmethod
+    def from_model(model: Type[DyffSchemaBaseModel]) -> "DataSchema":
+        arrowSchema = arrow.encode_schema(arrow.arrow_schema(model))
+        jsonSchema = model.schema()
+        return DataSchema(arrowSchema=arrowSchema, jsonSchema=jsonSchema)
+
     @staticmethod
     def make_input_schema(
         schema: Union[pyarrow.Schema, Type[DyffSchemaBaseModel], DyffDataSchema],
@@ -1713,6 +1828,7 @@ __all__ = [
     "APIKey",
     "ArchiveFormat",
     "Artifact",
+    "ArtifactURL",
     "Audit",
     "AuditProcedure",
     "AuditRequirement",
@@ -1732,6 +1848,9 @@ __all__ = [
     "Evaluation",
     "EvaluationBase",
     "ExtractorStep",
+    "Family",
+    "FamilyMember",
+    "FamilyMemberKind",
     "ForeignInferenceService",
     "ForeignMethod",
     "ForeignModel",
@@ -1793,6 +1912,8 @@ __all__ = [
     "SchemaAdapter",
     "Status",
     "StorageSignedURL",
+    "Tag",
+    "TagBase",
     "TaskSchema",
     "entity_class",
     "JobStatus",

{dyff_schema-0.3.9 → dyff_schema-0.5.0}/dyff/schema/v0/r1/requests.py

@@ -30,11 +30,12 @@ from .platform import (
     ModelSpec,
     ModuleBase,
     ReportBase,
+    TagBase,
 )
-from .version import Versioned
+from .version import SchemaVersion
 
 
-class DyffEntityCreateRequest(Versioned, DyffSchemaBaseModel):
+class DyffEntityCreateRequest(SchemaVersion, DyffSchemaBaseModel):
     account: str = pydantic.Field(description="Account that owns the entity")
 
 
@@ -105,6 +106,10 @@ class ReportCreateRequest(DyffEntityCreateRequest, ReportBase):
     )
 
 
+class TagCreateRequest(TagBase):
+    pass
+
+
 class LabelUpdateRequest(Labeled):
     pass
 
@@ -229,4 +234,5 @@ __all__ = [
     "ReportCreateRequest",
     "ReportQueryRequest",
     "SafetyCaseQueryRequest",
+    "TagCreateRequest",
 ]

{dyff_schema-0.3.9 → dyff_schema-0.5.0}/dyff/schema/v0/r1/version.py

@@ -5,16 +5,16 @@ from typing import Literal
 
 import pydantic
 
-VERSION: str = "0.1"
+SCHEMA_VERSION: str = "0.1"
 
 
-class Versioned(pydantic.BaseModel):
+class SchemaVersion(pydantic.BaseModel):
     schemaVersion: Literal["0.1"] = pydantic.Field(
-        default=VERSION, description="The schema version."
+        default=SCHEMA_VERSION, description="The schema version."
     )
 
 
 __all__ = [
-    "VERSION",
-    "Versioned",
+    "SCHEMA_VERSION",
+    "SchemaVersion",
 ]

{dyff_schema-0.3.9 → dyff_schema-0.5.0/dyff_schema.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: dyff-schema
-Version: 0.3.9
+Version: 0.5.0
 Summary: Data models for the Dyff AI auditing platform.
 Author-email: Digital Safety Research Institute <contact@dsri.org>
 License: Apache-2.0