dyff-schema 0.32.0__py3-none-any.whl → 0.33.0__py3-none-any.whl

dyff/schema/_version.py

@@ -1,2 +1,2 @@
-__version__ = version = "0.32.0"
-__version_tuple__ = version_tuple = (0, 32, 0)
+__version__ = version = "0.33.0"
+__version_tuple__ = version_tuple = (0, 33, 0)

dyff/schema/v0/r1/oci.py

@@ -0,0 +1,183 @@
+# SPDX-FileCopyrightText: 2024 UL Research Institutes
+# SPDX-License-Identifier: Apache-2.0
+
+from __future__ import annotations
+
+import enum
+import hashlib
+from datetime import datetime, timezone
+from typing import Any, Literal, Optional
+
+import canonicaljson
+import pydantic
+
+from .base import int64
+
+
+class KnownMediaTypes(str, enum.Enum):
+    oci_image_manifest = "application/vnd.oci.image.manifest.v1+json"
+    oci_image_config = "application/vnd.oci.image.config.v1+json"
+    oci_image_layer_tar = "application/vnd.oci.image.layer.v1.tar"
+    oci_image_layer_tar_gzip = "application/vnd.oci.image.layer.v1.tar+gzip"
+
+
+# mypy gets confused because 'dict' is the name of a method in DyffBaseModel
+_ModelAsDict = dict[str, Any]
+
+
+class _OCISchemaBaseModel(pydantic.BaseModel):
+    def model_dump(  # type: ignore [override]
+        self, *, by_alias: bool = True, exclude_none: bool = True, **kwargs
+    ) -> _ModelAsDict:
+        return super().model_dump(
+            by_alias=by_alias, exclude_none=exclude_none, **kwargs
+        )
+
+    def model_dump_json(  # type: ignore [override]
+        self, *, by_alias: bool = True, exclude_none: bool = True, **kwargs
+    ) -> str:
+        return super().model_dump_json(
+            by_alias=by_alias, exclude_none=exclude_none, **kwargs
+        )
+
+    def dict(
+        self, *, by_alias: bool = True, exclude_none: bool = True, **kwargs
+    ) -> _ModelAsDict:
+        return self.model_dump(by_alias=by_alias, exclude_none=exclude_none, **kwargs)
+
+    def json(
+        self, *, by_alias: bool = True, exclude_none: bool = True, **kwargs
+    ) -> str:
+        return self.model_dump_json(
+            by_alias=by_alias, exclude_none=exclude_none, **kwargs
+        )
+
+    @pydantic.model_validator(mode="after")
+    def _ensure_datetime_timezone_utc(cls, values):
+        for field_name, field_value in values.__dict__.items():
+            if isinstance(field_value, datetime):
+                if field_value.tzinfo is None:
+                    # Set UTC timezone for naive datetime
+                    setattr(
+                        values, field_name, field_value.replace(tzinfo=timezone.utc)
+                    )
+                elif field_value.tzinfo != timezone.utc:
+                    # Convert to UTC timezone
+                    setattr(values, field_name, field_value.astimezone(timezone.utc))
+        return values
+
+
+class _DigestMixin(pydantic.BaseModel):
+    def digest(self) -> str:
+        obj = self.model_dump(mode="json")
+        h = hashlib.sha256(usedforsecurity=False)
+        for chunk in canonicaljson.iterencode_canonical_json(obj):
+            h.update(chunk)
+        return "sha256:{}".format(h.hexdigest())
+
+
+class Platform(_OCISchemaBaseModel):
+    architecture: str
+    os: str
+    os_version: Optional[str] = pydantic.Field(alias="os.version", default=None)
+    os_features: Optional[list[str]] = pydantic.Field(alias="os.features", default=None)
+    variant: Optional[str] = pydantic.Field(default=None)
+
+
+class Descriptor(_OCISchemaBaseModel):
+    """
+    https://github.com/opencontainers/image-spec/blob/v1.1.1/descriptor.md
+    """
+
+    mediaType: str = pydantic.Field()
+    digest: str = pydantic.Field()
+    size: int64() = pydantic.Field(  # type: ignore
+        description="This REQUIRED property specifies the size, in bytes,"
+        " of the raw content. This property exists so that a client"
+        " will have an expected size for the content before processing."
+        " If the length of the retrieved content does not match"
+        " the specified length, the content SHOULD NOT be trusted."
+    )
+    urls: Optional[list[str]] = pydantic.Field(default=None)
+    annotations: Optional[dict[str, str]] = pydantic.Field(default=None)
+    data: Optional[str] = pydantic.Field(default=None)
+    artifactType: Optional[str] = pydantic.Field(default=None)
+
+    @staticmethod
+    def empty() -> Descriptor:
+        """
+        https://github.com/opencontainers/image-spec/blob/v1.1.1/manifest.md#guidance-for-an-empty-descriptor
+        """
+        return Descriptor.model_validate(
+            {
+                "mediaType": "application/vnd.oci.empty.v1+json",
+                "digest": "sha256:44136fa355b3678a1146ad16f7e8649e94fb4fc21fe77e8310c060f61caaff8a",
+                "size": 2,
+                "data": "e30=",
+            }
+        )
+
+
+class ImageDescriptor(Descriptor):
+    platform: Platform
+
+
+class ExecutionParameters(_OCISchemaBaseModel):
+    # TODO
+    pass
+
+
+class RootFS(_OCISchemaBaseModel):
+    type_: Literal["layers"] = pydantic.Field(alias="type", default="layers")
+    diff_ids: list[str] = pydantic.Field()
+
+
+class HistoryEntry(_OCISchemaBaseModel):
+    created: Optional[str] = pydantic.Field(default=None)
+    author: Optional[str] = pydantic.Field(default=None)
+    created_by: Optional[str] = pydantic.Field(default=None)
+    comment: Optional[str] = pydantic.Field(default=None)
+    empty_layer: Optional[bool] = pydantic.Field(default=None)
+
+
+class ImageConfig(Platform, _DigestMixin):
+    """
+    https://github.com/opencontainers/image-spec/blob/v1.1.1/config.md
+    """
+
+    created: Optional[str] = pydantic.Field(default=None)
+    author: Optional[str] = pydantic.Field(default=None)
+    config: Optional[ExecutionParameters] = pydantic.Field(default=None)
+    rootfs: RootFS
+    history: Optional[list[HistoryEntry]] = pydantic.Field(default=None)
+
+
+class ImageManifest(_OCISchemaBaseModel):
+    schemaVersion: Literal["2"] = "2"
+    mediaType: Literal["application/vnd.oci.image.manifest.v1+json"] = (
+        "application/vnd.oci.image.manifest.v1+json"
+    )
+    artifactType: Optional[str] = pydantic.Field(default=None)
+    config: Descriptor = pydantic.Field(default_factory=Descriptor.empty)
+    layers: list[Descriptor] = pydantic.Field(
+        default_factory=lambda: [Descriptor.empty()]
+    )
+    subject: Optional[Descriptor] = pydantic.Field(default=None)
+    annotations: Optional[dict[str, str]] = pydantic.Field(default=None)
+
+    def is_image_manifest(self) -> bool:
+        return self.artifactType is None
+
+    def is_artifact_manifest(self) -> bool:
+        return self.artifactType is not None
+
+    @pydantic.model_validator(mode="after")
+    def validate_types(self):
+        if self.artifactType is None:
+            if self.config.mediaType != KnownMediaTypes.oci_image_config.value:
+                raise ValueError(
+                    ".config.mediaType must be"
+                    f"{KnownMediaTypes.oci_image_config.value}"
+                    " if .artifactType is unspecified"
+                )
+        return self

dyff/schema/v0/r1/platform.py

@@ -33,6 +33,7 @@ from typing_extensions import Annotated, TypeAlias
 
 from ... import named_data_schema, product_schema
 from ...version import SomeSchemaVersion
+from . import oci
 from .base import DyffSchemaBaseModel
 from .dataset import arrow, make_item_type, make_response_type
 from .version import SCHEMA_VERSION, SchemaVersion
@@ -193,6 +194,7 @@ class Entities(str, enum.Enum):
 
     Account = "Account"
     Analysis = "Analysis"
+    Artifact = "Artifact"
     Audit = "Audit"
     AuditProcedure = "AuditProcedure"
     Concern = "Concern"
@@ -220,6 +222,7 @@ class Resources(str, enum.Enum):
     """The resource names corresponding to entities that have API endpoints."""
 
     Analysis = "analyses"
+    Artifact = "artifacts"
     Audit = "audits"
     AuditProcedure = "auditprocedures"
     Concern = "concerns"
@@ -266,6 +269,7 @@ class Resources(str, enum.Enum):
 
 EntityKindLiteral = Literal[
     "Analysis",
+    "Artifact",
     "Audit",
     "AuditProcedure",
     "DataSource",
@@ -488,6 +492,7 @@ class DyffEntityMetadata(DyffSchemaBaseModel):
 class DyffEntity(Status, Labeled, SchemaVersion, DyffModelWithID):
     kind: Literal[
         "Analysis",
+        "Artifact",
         "Audit",
         "AuditProcedure",
         "DataSource",
@@ -829,8 +834,10 @@ class Digest(DyffSchemaBaseModel):
     )
 
 
+# TODO: (schema-v1) Rename this to "File" or something -- reserve Artifact
+# for OCI artifacts.
 class Artifact(DyffSchemaBaseModel):
-    # TODO: In v1, rename this to 'contentType' or something and commit to making it the MIME type
+    # TODO: (schema-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"
     )
@@ -1122,6 +1129,10 @@ class ModelSource(DyffSchemaBaseModel):
     )
 
 
+class ContainerImage(DyffSchemaBaseModel):
+    pass
+
+
 class AcceleratorGPU(DyffSchemaBaseModel):
     hardwareTypes: list[str] = pydantic.Field(
         min_length=1,
@@ -2130,6 +2141,25 @@ class Score(ScoreData):
     id: str = pydantic.Field(description="Unique identifier of the entity")
 
 
+# ---------------------------------------------------------------------------
+# OCI artifacts
+
+
+# TODO: (schema-v1) Rename this to Artifact
+class OCIArtifact(Documented, DyffEntity):
+    kind: Literal["Artifact"] = Entities.Artifact.value
+
+    manifest: oci.ImageManifest = pydantic.Field(
+        description="The OCI image manifest of the artifact"
+    )
+
+    def dependencies(self) -> list[str]:
+        return []
+
+    def resource_allocation(self) -> Optional[ResourceAllocation]:
+        return None
+
+
 # ---------------------------------------------------------------------------
 # Status enumerations
 
@@ -2372,6 +2402,7 @@ def is_status_success(status: str) -> bool:
 
 _ENTITY_CLASS = {
     Entities.Analysis: Analysis,
+    Entities.Artifact: OCIArtifact,
     Entities.Audit: Audit,
     Entities.AuditProcedure: AuditProcedure,
     Entities.Dataset: Dataset,
@@ -2406,6 +2437,7 @@ _DyffEntityTypeRevisable = Union[
     Method,
     Model,
     Module,
+    OCIArtifact,
     Report,
     SafetyCase,
     UseCase,
@@ -2579,6 +2611,7 @@ __all__ = [
     "ModelStorage",
     "Module",
     "ModuleBase",
+    "OCIArtifact",
     "QueryableDyffEntity",
     "Report",
     "ReportBase",

dyff/schema/v0/r1/requests.py

@@ -21,7 +21,7 @@ from typing import Any, Literal, Optional, Union
 import pydantic
 
 from ... import upcast
-from . import commands
+from . import commands, oci
 from .base import DyffBaseModel, JsonMergePatchSemantics
 from .platform import (
     AnalysisBase,
@@ -72,6 +72,20 @@ class DyffRequestBase(SchemaVersion, DyffRequestDefaultValidators):
     # TODO: (DYFF-223) I think that exclude_unset=True should be the default
     # for all schema objects, but I'm unsure of the consequences of making
     # this change and we'll defer it until v1.
+    def model_dump(  # type: ignore [override]
+        self, *, by_alias: bool = True, exclude_unset=True, **kwargs
+    ) -> _ModelAsDict:
+        return super().model_dump(
+            by_alias=by_alias, exclude_unset=exclude_unset, **kwargs
+        )
+
+    def model_dump_json(  # type: ignore [override]
+        self, *, by_alias: bool = True, exclude_unset=True, **kwargs
+    ) -> str:
+        return super().model_dump_json(
+            by_alias=by_alias, exclude_unset=exclude_unset, **kwargs
+        )
+
     def dict(
         self, *, by_alias: bool = True, exclude_unset=True, **kwargs
     ) -> _ModelAsDict:
@@ -118,6 +132,10 @@ class AnalysisCreateRequest(DyffEntityCreateRequest, AnalysisBase):
         return scope
 
 
+class ArtifactCreateRequest(DyffEntityCreateRequest):
+    manifest: oci.ImageManifest
+
+
 class ConcernCreateRequest(DyffEntityCreateRequest, ConcernBase):
     @pydantic.field_validator("documentation", check_fields=False)
     def _validate_documentation(
@@ -357,6 +375,10 @@ class _AnalysisProductQueryRequest(DyffEntityQueryRequest):
     inputs: Optional[str] = pydantic.Field(default=None)
 
 
+class ArtifactQueryRequest(DyffEntityQueryRequest):
+    name: Optional[str] = pydantic.Field(default=None)
+
+
 class AuditQueryRequest(DyffEntityQueryRequest):
     name: Optional[str] = pydantic.Field(default=None)
 
@@ -446,6 +468,8 @@ class UseCaseQueryRequest(DyffEntityQueryRequest):
 
 __all__ = [
     "AnalysisCreateRequest",
+    "ArtifactCreateRequest",
+    "ArtifactQueryRequest",
     "AuditQueryRequest",
     "ConcernCreateRequest",
     "DyffEntityCreateRequest",

{dyff_schema-0.32.0.dist-info → dyff_schema-0.33.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dyff-schema
-Version: 0.32.0
+Version: 0.33.0
 Summary: Data models for the Dyff AI auditing platform.
 Author-email: Digital Safety Research Institute <contact@dsri.org>
 License: Apache-2.0
@@ -20,6 +20,7 @@ Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
 License-File: NOTICE
+Requires-Dist: canonicaljson==2.0.0
 Requires-Dist: hypothesis
 Requires-Dist: hypothesis-jsonschema
 Requires-Dist: jsonpath-ng

{dyff_schema-0.32.0.dist-info → dyff_schema-0.33.0.dist-info}/RECORD

@@ -1,5 +1,5 @@
 dyff/schema/__init__.py,sha256=w7OWDFuyGKd6xt_yllNtKzHahPgywrfU4Ue02psYaMA,2244
-dyff/schema/_version.py,sha256=cSUDf-klGYAh3VKPYjjxhNT8qv04wlp3-jjPmvgZjS0,80
+dyff/schema/_version.py,sha256=uTbdsxpcOU7mMXqX3G50NVmPGutL-k0_3StqoCGV-b0,80
 dyff/schema/adapters.py,sha256=YMTHv_2VlLGFp-Kqwa6H51hjffHmk8gXjZilHysIF5Q,123
 dyff/schema/annotations.py,sha256=nE6Jk1PLqlShj8uqjE_EzZC9zYnTDW5AVtQcjysiK8M,10018
 dyff/schema/base.py,sha256=jvaNtsSZyFfsdUZTcY_U-yfLY5_GyrMxSXhON2R9XR0,119
@@ -28,8 +28,9 @@ dyff/schema/v0/r1/__init__.py,sha256=L5y8UhRnojerPYHumsxQJRcHCNz8Hj9NM8b47mewMNs
 dyff/schema/v0/r1/adapters.py,sha256=hpwCSW8lkMkUKCLe0zaMUDu-VS_caSxJvPsECEi_XRA,33069
 dyff/schema/v0/r1/base.py,sha256=fdhAa4hpbSn7m3U0qha4rG7gJiYUvPR8SaM-mwszoy0,20289
 dyff/schema/v0/r1/commands.py,sha256=VoThKeIVz4ZQAqBXhb6TVOG6aG4m_Oa0_6Sc4oxyFhs,9801
-dyff/schema/v0/r1/platform.py,sha256=7ErqSqPLQD2bIEqnJByktBb7RaQYRkZbqQsofb0dFIM,82736
-dyff/schema/v0/r1/requests.py,sha256=C-LX-NAIDgh88xCl3NW7wqLIpurK8SXqmhRnGvy4BM0,17197
+dyff/schema/v0/r1/oci.py,sha256=mvh1BaiRahZbjIbMh7RqZxZjOKZT9GsS1yBpSdPw01Y,6587
+dyff/schema/v0/r1/platform.py,sha256=s1O_mAVfi9owekqYxcjwPI10s3F7RNo8B8WvMunMi4U,83577
+dyff/schema/v0/r1/requests.py,sha256=3XcFcKChubBrxoHCUHkV5GrNl6Ijdbxn4D4UZAAo09g,17965
 dyff/schema/v0/r1/responses.py,sha256=nxy7FPtfw2B_bljz5UGGuSE79HTkDQxKH56AJVmd4Qo,1287
 dyff/schema/v0/r1/test.py,sha256=X6dUyVd5svcPCI-PBMOAqEfK9jv3bRDvkQTJzwS96c0,10720
 dyff/schema/v0/r1/version.py,sha256=NONebgcv5Thsw_ymud6PacZdGjV6ndBrmLnap-obcpo,428
@@ -42,9 +43,9 @@ dyff/schema/v0/r1/dataset/text.py,sha256=MYG5seGODDryRSCy-g0Unh5dD0HCytmZ3FeElC-
 dyff/schema/v0/r1/dataset/vision.py,sha256=aIe0fbfM_g3DsrDTdg2K803YKLjZBpurM_VJcJFuZLc,369
 dyff/schema/v0/r1/io/__init__.py,sha256=L5y8UhRnojerPYHumsxQJRcHCNz8Hj9NM8b47mewMNs,92
 dyff/schema/v0/r1/io/vllm.py,sha256=vWyLg-susbg0JDfv6VExBpgFdU2GHP2a14ChOdbckvs,5321
-dyff_schema-0.32.0.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-dyff_schema-0.32.0.dist-info/licenses/NOTICE,sha256=YONACu0s_Ui6jNi-wtEsVQbTU1JIkh8wvLH6d1-Ni_w,43
-dyff_schema-0.32.0.dist-info/METADATA,sha256=_DgORzGNTURV5pqYZS014MpFJOFufqQ--rpSCfIupKA,3632
-dyff_schema-0.32.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-dyff_schema-0.32.0.dist-info/top_level.txt,sha256=9e3VVdeX73t_sUJOPQPCcGtYO1JhoErhHIi3WoWGcFI,5
-dyff_schema-0.32.0.dist-info/RECORD,,
+dyff_schema-0.33.0.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+dyff_schema-0.33.0.dist-info/licenses/NOTICE,sha256=YONACu0s_Ui6jNi-wtEsVQbTU1JIkh8wvLH6d1-Ni_w,43
+dyff_schema-0.33.0.dist-info/METADATA,sha256=XUEtF3Dgk_LSI4YKUPZp5RJzKvzx7nv7_Jxp30ic3W0,3668
+dyff_schema-0.33.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+dyff_schema-0.33.0.dist-info/top_level.txt,sha256=9e3VVdeX73t_sUJOPQPCcGtYO1JhoErhHIi3WoWGcFI,5
+dyff_schema-0.33.0.dist-info/RECORD,,