Merge pull request #3004 from catalyst-cooperative/datapackage-update

Update sources, DOI and copyright link in PUDL
catalyst-cooperative · Nov 2, 2023 · e7aedda · e7aedda
2 parents e848d64 + 6e6bf5e
commit e7aedda
Show file tree

Hide file tree

Showing 4 changed files with 43 additions and 26 deletions.
diff --git a/src/pudl/metadata/classes.py b/src/pudl/metadata/classes.py
@@ -42,7 +42,7 @@
 )
 from pudl.metadata.resources import FOREIGN_KEYS, RESOURCE_METADATA, eia861
 from pudl.metadata.sources import SOURCES
-from pudl.workspace.datastore import Datastore
+from pudl.workspace.datastore import Datastore, ZenodoDoi
 from pudl.workspace.setup import PudlPaths
 
 logger = pudl.logging_helpers.get_logger(__name__)
@@ -869,6 +869,27 @@ class Contributor(Base):
     role: Literal[
         "author", "contributor", "maintainer", "publisher", "wrangler"
     ] = "contributor"
+    zenodo_role: Literal[
+        "contact person",
+        "data collector",
+        "data curator",
+        "data manager",
+        "distributor",
+        "editor",
+        "hosting institution",
+        "other",
+        "producer",
+        "project leader",
+        "project member",
+        "registration agency",
+        "registration authority",
+        "related person",
+        "researcher",
+        "rights holder",
+        "sponsor",
+        "supervisor",
+        "work package leader",
+    ] = "project member"
     organization: String = None
     orcid: String = None
 
@@ -911,10 +932,10 @@ class DataSource(Base):
     field_namespace: String = None
     keywords: list[str] = []
     path: HttpUrl = None
-    contributors: list[Contributor] = []  # Or should this be compiled from Resources?
+    contributors: list[Contributor] = []
     license_raw: License
     license_pudl: License
-    # concept_doi: Doi = None  # Need to define a Doi type?
+    concept_doi: ZenodoDoi = None
     working_partitions: dict[SnakeCase, Any] = {}
     source_file_dict: dict[SnakeCase, Any] = {}
     # agency: Agency  # needs to be defined

diff --git a/src/pudl/metadata/constants.py b/src/pudl/metadata/constants.py
@@ -62,7 +62,7 @@
     "us-govt": {
         "name": "other-pd",
         "title": "U.S. Government Works",
-        "path": "https://www.usa.gov/government-works",
+        "path": "https://www.usa.gov/publicdomain/label/1.0/",
     },
 }
 """License attributes."""
@@ -83,91 +83,105 @@
         "email": "[email protected]",
         "path": "https://catalyst.coop",
         "role": "publisher",
+        "zenodo_role": "distributor",
         "organization": "Catalyst Cooperative",
     },
     "zane-selvans": {
         "title": "Zane Selvans",
         "email": "[email protected]",
         "path": "https://amateurearthling.org",
         "role": "wrangler",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
         "orcid": "0000-0002-9961-7208",
     },
     "christina-gosnell": {
         "title": "Christina Gosnell",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
         "orcid": "0009-0004-2979-6142",
     },
     "steven-winter": {
         "title": "Steven Winter",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
     },
     "alana-wilson": {
         "title": "Alana Wilson",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
     },
     "karl-dunkle-werner": {
         "title": "Karl Dunkle Werner",
         "email": "[email protected]",
         "path": "https://karldw.org",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "UC Berkeley",
         "orcid": "0000-0003-0523-7309",
     },
     "greg-schivley": {
         "title": "Greg Schivley",
         "path": "https://gschivley.github.io",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Carbon Impact Consulting",
         "orcid": "0000-0002-8947-694X",
     },
     "austen-sharpe": {
         "title": "Austen Sharpe",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
     },
     "katherine-lamb": {
         "title": "Katherine Lamb",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
     },
     "bennett-norman": {
         "title": "Bennett Norman",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
     },
     "trenton-bush": {
         "title": "Trenton Bush",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
     },
     "ethan-welty": {
         "title": "Ethan Welty",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
         "orcid": "0000-0001-8046-2210",
     },
     "dazhong-xia": {
         "title": "Dazhong Xia",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
     },
     "ella-belfer": {
         "title": "Ella Belfer",
         "email": "[email protected]",
         "role": "contributor",
+        "zenodo_role": "project member",
         "organization": "Catalyst Cooperative",
         "orcid": "0000-0001-9784-8531",
     },

diff --git a/src/pudl/metadata/sources.py b/src/pudl/metadata/sources.py
@@ -20,6 +20,7 @@
         ),
         "license_raw": LICENSES["us-govt"],
         "license_pudl": LICENSES["cc-by-4.0"],
+        "contributors": [CONTRIBUTORS["catalyst-cooperative"]],
     },
     "eia176": {
         "title": "EIA Form 176 -- Annual Report of Natural and Supplemental Gas Supply and Disposition",
@@ -75,10 +76,6 @@
         "field_namespace": "eia",
         "contributors": [
             CONTRIBUTORS["catalyst-cooperative"],
-            CONTRIBUTORS["zane-selvans"],
-            CONTRIBUTORS["christina-gosnell"],
-            CONTRIBUTORS["steven-winter"],
-            CONTRIBUTORS["alana-wilson"],
         ],
         "working_partitions": {
             "years": sorted(set(range(2001, 2023))),
@@ -195,10 +192,6 @@
         },
         "contributors": [
             CONTRIBUTORS["catalyst-cooperative"],
-            CONTRIBUTORS["zane-selvans"],
-            CONTRIBUTORS["christina-gosnell"],
-            CONTRIBUTORS["steven-winter"],
-            CONTRIBUTORS["katherine-lamb"],
         ],
         "keywords": sorted(
             set(
@@ -240,8 +233,6 @@
         "working_partitions": {},
         "contributors": [
             CONTRIBUTORS["catalyst-cooperative"],
-            CONTRIBUTORS["zane-selvans"],
-            CONTRIBUTORS["trenton-bush"],
         ],
         "keywords": sorted(
             set(
@@ -286,8 +277,6 @@
         },
         "contributors": [
             CONTRIBUTORS["catalyst-cooperative"],
-            CONTRIBUTORS["karl-dunkle-werner"],
-            CONTRIBUTORS["zane-selvans"],
         ],
         "keywords": sorted(
             set(
@@ -325,7 +314,6 @@
         "working_partitions": {},
         "contributors": [
             CONTRIBUTORS["catalyst-cooperative"],
-            CONTRIBUTORS["austen-sharpe"],
         ],
         "keywords": sorted(
             set(
@@ -369,11 +357,6 @@
         },
         "contributors": [
             CONTRIBUTORS["catalyst-cooperative"],
-            CONTRIBUTORS["zane-selvans"],
-            CONTRIBUTORS["christina-gosnell"],
-            CONTRIBUTORS["steven-winter"],
-            CONTRIBUTORS["alana-wilson"],
-            CONTRIBUTORS["austen-sharpe"],
         ],
         "keywords": sorted(
             set(
@@ -580,9 +563,6 @@
         ),
         "contributors": [
             CONTRIBUTORS["catalyst-cooperative"],
-            CONTRIBUTORS["zane-selvans"],
-            CONTRIBUTORS["christina-gosnell"],
-            CONTRIBUTORS["steven-winter"],
         ],
         "field_namespace": "pudl",
         "keywords": ["us", "electricity", "open data", "open source"],

diff --git a/src/pudl/workspace/datastore.py b/src/pudl/workspace/datastore.py
@@ -27,7 +27,9 @@
 logger = pudl.logging_helpers.get_logger(__name__)
 
 PUDL_YML = Path.home() / ".pudl.yml"
-ZenodoDoi = constr(regex=r"(10\.5072|10\.5281)/zenodo.([\d]+)")
+ZenodoDoi = constr(
+    strict=True, min_length=16, regex=r"(10\.5072|10\.5281)/zenodo.([\d]+)"
+)
 
 
 class ChecksumMismatchError(ValueError):