Add US long-run bundle manifest

changelog.d/crfb-longrun-bundle.added.md

@@ -0,0 +1 @@
+Added the US long-run managed dataset bundle.

pyproject.toml

@@ -46,8 +46,8 @@ uk = [
     "policyengine-uk==2.88.14",
 ]
 us = [
-    "policyengine_core>=3.26.1",
-    "policyengine-us==1.691.10",
+    "policyengine_core==3.26.1",
+    "policyengine-us==1.691.12",
 ]
 dev = [
     "pytest",
@@ -60,9 +60,9 @@ dev = [
     "plotly>=5.0.0",
     "pytest-asyncio>=0.26.0",
     "ruff>=0.9.0",
-    "policyengine_core>=3.26.1",
+    "policyengine_core==3.26.1",
     "policyengine-uk==2.88.14",
-    "policyengine-us==1.691.10",
+    "policyengine-us==1.691.12",
     "towncrier>=24.8.0",
     "mypy>=1.11.0",
     "pytest-cov>=5.0.0",

src/policyengine/data/release_manifests/us.trace.tro.jsonld

@@ -17,7 +17,6 @@
         "schema:name": "PolicyEngine",
         "schema:url": "https://policyengine.org"
       },
-      "schema:dateCreated": "2026-05-16T04:38:05.331368Z",
       "schema:description": "TRACE TRO for certified runtime bundle us-4.5.1 covering the bundle manifest, the certified dataset artifact, the country model wheel, and the country data release manifest when it is available.",
       "schema:name": "policyengine us certified bundle TRO",
       "trov:createdWith": {
@@ -39,14 +38,6 @@
               },
               "trov:hasLocation": "data/release_manifests/us.json"
             },
-            {
-              "@id": "arrangement/1/location/data_release_manifest",
-              "@type": "trov:ArtifactLocation",
-              "trov:hasArtifact": {
-                "@id": "composition/1/artifact/data_release_manifest"
-              },
-              "trov:hasLocation": "https://huggingface.co/policyengine/policyengine-us-data/resolve/69fc39a7fece4c49ba87291e598e76b40568cc5d/releases/1.115.3/release_manifest.json"
-            },
             {
               "@id": "arrangement/1/location/dataset",
               "@type": "trov:ArtifactLocation",
@@ -61,7 +52,7 @@
               "trov:hasArtifact": {
                 "@id": "composition/1/artifact/model_wheel"
               },
-              "trov:hasLocation": "https://files.pythonhosted.org/packages/a7/a3/53630e30285cb441cc96e5264621b1b9e5af3b648ccd176b376f0e03a458/policyengine_us-1.691.10-py3-none-any.whl"
+              "trov:hasLocation": "https://files.pythonhosted.org/packages/83/6f/b605fc1d8e06e377ae50870dc44a28cfe6562f0032e36dd53a5ef49472db/policyengine_us-1.691.12-py3-none-any.whl"
             }
           ]
         }
@@ -75,54 +66,43 @@
             "@type": "trov:ResearchArtifact",
             "schema:name": "policyengine.py bundle manifest for us",
             "trov:mimeType": "application/json",
-            "trov:sha256": "55cd4433702834bacd1057d6c728ff5730f35e508f78821040be6c48dd304045"
-          },
-          {
-            "@id": "composition/1/artifact/data_release_manifest",
-            "@type": "trov:ResearchArtifact",
-            "schema:name": "policyengine-us-data release manifest 1.115.3",
-            "trov:mimeType": "application/json",
-            "trov:sha256": "50fec0baf199cf7f5bc5a16d9956d7c8d0fca3e2097b69dbb7f26cbbcf9114a9"
+            "trov:sha256": "3229be76641f8c8d79b44604da04453cf8aa1b28169a4816a3ef818f6f93bfee"
           },
           {
             "@id": "composition/1/artifact/dataset",
             "@type": "trov:ResearchArtifact",
             "schema:name": "enhanced_cps_2024",
-            "trov:mimeType": "application/x-hdf5",
             "trov:sha256": "96965b0d9931d36beb29486b25fffee8b630e3b62c9376e7265976e02f7ab3ca"
           },
           {
             "@id": "composition/1/artifact/model_wheel",
             "@type": "trov:ResearchArtifact",
-            "schema:name": "policyengine-us==1.691.10 wheel",
+            "schema:name": "policyengine-us==1.691.12 wheel",
             "trov:mimeType": "application/zip",
-            "trov:sha256": "a2c8365f152c2ec3332e996c4003d560b433b6c204bdfcc73bf094a0ae49f10e"
+            "trov:sha256": "ef43482bd8c6cc16f8f1d4050423f5dc1d045af15931f5d1b089715a31c839d2"
           }
         ],
         "trov:hasFingerprint": {
           "@id": "composition/1/fingerprint",
           "@type": "trov:CompositionFingerprint",
-          "trov:sha256": "8fc5a40f13183c7c041b3dc6e76e92aa14ee959f300f39f434cdaca0f95b1617"
+          "trov:sha256": "c14a4009d824044cd386f08b93b31f6af4cc608619ee22d83314f481fddfd47b"
         }
       },
       "trov:hasPerformance": {
         "@id": "trp/1",
         "@type": "trov:TransparentResearchPerformance",
-        "pe:builtWithModelVersion": "1.691.10",
-        "pe:certifiedBy": "policyengine.py bundled manifest",
-        "pe:certifiedForModelVersion": "1.691.10",
-        "pe:ciGitRef": "refs/heads/main",
-        "pe:ciGitSha": "4c052fa345868f1eb92757ee779a653a82c72b71",
-        "pe:ciRunUrl": "https://github.com/PolicyEngine/policyengine.py/actions/runs/25953758405",
-        "pe:compatibilityBasis": "exact_build_model_version",
-        "pe:dataBuildFingerprint": "sha256:b80625974285a9cff23bc2436fc46df98477997d50aea61e73b7f137b5121584",
-        "pe:dataBuildId": "policyengine-us-data-1.115.3",
-        "pe:emittedIn": "github-actions",
-        "rdfs:comment": "Certification of build policyengine-us-data-1.115.3 for policyengine-us 1.691.10.",
+        "pe:builtWithModelVersion": "1.691.12",
+        "pe:certifiedBy": "policyengine.py candidate long-term manifest",
+        "pe:certifiedForModelVersion": "1.691.12",
+        "pe:compatibilityBasis": "candidate_long_term_bundle",
+        "pe:dataBuildFingerprint": "sha256:ae675b873404ec1fdc461e056a02958816fd0f8ab08ba5883fb05c4cf40f5b49",
+        "pe:dataBuildId": "policyengine-us-data-crfb-longrun-20260517",
+        "pe:dataReleaseManifestStatus": "unavailable",
+        "pe:emittedIn": "local",
+        "rdfs:comment": "Certification of build policyengine-us-data-crfb-longrun-20260517 for policyengine-us 1.691.12.",
         "trov:accessedArrangement": {
           "@id": "arrangement/1"
         },
-        "trov:startedAtTime": "2026-05-16T04:38:05.331368Z",
         "trov:wasConductedBy": {
           "@id": "trs"
         }

src/policyengine/provenance/manifest.py

@@ -57,6 +57,8 @@ class DataBuildInfo(BaseModel):
 
 class ArtifactPathReference(BaseModel):
     path: str
+    sha256: Optional[str] = None
+    metadata_sha256: Optional[str] = None
 
 
 class ArtifactPathTemplate(BaseModel):
@@ -502,6 +504,23 @@ def resolve_local_managed_dataset_source(
     _, _, path_in_repo = parts
 
     model_module_name, data_repo_name, data_package_name = local_hint
+    explicit_repo_roots = []
+    country_env = f"POLICYENGINE_{country_id.upper()}_DATA_REPO"
+    for env_name in (country_env, "POLICYENGINE_LOCAL_DATA_REPO_ROOT"):
+        env_value = os.environ.get(env_name)
+        if env_value:
+            explicit_repo_roots.extend(
+                [
+                    Path(env_value).expanduser(),
+                    Path(env_value).expanduser() / data_repo_name,
+                ]
+            )
+
+    for candidate_repo_root in explicit_repo_roots:
+        local_path = candidate_repo_root / data_package_name / "storage" / path_in_repo
+        if local_path.exists():
+            return str(local_path)
+
     try:
         model_module = import_module(model_module_name)
     except ImportError:

src/policyengine/tax_benefit_models/us/__init__.py

@@ -38,6 +38,7 @@
         ensure_datasets,
         load_datasets,
         load_long_term_datasets,
+        load_managed_long_term_datasets,
         validate_long_term_dataset_metadata,
     )
     from .household import calculate_household
@@ -63,6 +64,7 @@
         "create_datasets",
         "load_datasets",
         "load_long_term_datasets",
+        "load_managed_long_term_datasets",
         "ensure_datasets",
         "validate_long_term_dataset_metadata",
         "PolicyEngineUS",