Update Zenodo entry upon release

.github/scripts/publish_to_zenodo.py

@@ -0,0 +1,197 @@
+# Based on examples on https://developers.zenodo.org/
+
+import json
+import os
+import requests
+import yaml
+
+ACCESS_TOKEN = os.environ["ZENODO_SANDBOX_TOKEN"]
+CONCEPT_ID = os.environ["ZENODO_SANDBOX_CONCEPT_ID"]
+
+print(CONCEPT_ID)
+print(ACCESS_TOKEN)
+
+BASE_URL = "https://sandbox.zenodo.org/api"
+
+
+json_headers = {
+    "Authorization": f"Bearer {ACCESS_TOKEN}",
+    "Content-Type": "application/json"
+}
+
+upload_headers = {
+    "Authorization": f"Bearer {ACCESS_TOKEN}"
+}
+
+def load_citation():
+    with open("CITATION.cff", encoding="utf-8") as f:
+        return yaml.safe_load(f)
+
+
+def cff_to_creators(authors):
+    creators = []
+
+    for author in authors:
+        if "family-names" in author:
+            creators.append(
+                {
+                    "name": (
+                        f"{author['family-names']}, "
+                        f"{author['given-names']}"
+                    )
+                }
+            )
+        else:
+            creators.append({"name": author["name"]})
+
+    return creators
+
+
+
+# Load metadata from CITATION.cff
+
+cff = load_citation()
+
+metadata = {
+    "title": cff["title"],
+    "upload_type": "dataset",
+    "description": cff.get("abstract", ""),
+    "creators": cff_to_creators(cff["authors"]),
+    "keywords": cff.get("keywords", []),
+    "version": str(cff.get("version")) if cff.get("version") is not None else None,
+    "license": cff.get("license")
+}
+
+
+# Discard any unpublished draft left over from a previous failed run.
+# Zenodo only allows one unpublished new-version draft per concept at
+# a time, so a stray one would make the newversion call below fail.
+
+r = requests.get(
+    f"{BASE_URL}/deposit/depositions",
+    params={"q": f"conceptrecid:{CONCEPT_ID}", "all_versions": "true"},
+    headers=json_headers,
+)
+
+r.raise_for_status()
+
+for dep in r.json():
+    if not dep.get("submitted"):
+        print(f"Discarding leftover draft {dep['id']}")
+        requests.delete(dep["links"]["self"], headers=json_headers).raise_for_status()
+
+
+# Find the current latest version of this concept.
+# newversion only works when called on the latest version's own
+# deposit ID, and that ID changes with every new release.
+
+r = requests.get(
+    f"{BASE_URL}/records",
+    params={"q": f"conceptrecid:{CONCEPT_ID}", "all_versions": "true"},
+    headers=json_headers,
+)
+
+r.raise_for_status()
+
+hits = r.json()["hits"]["hits"]
+
+latest_id = hits[0]["id"]
+for hit in hits[1:]:
+    if hit["id"] > latest_id:
+        latest_id = hit["id"]
+
+print(f"Latest existing version: {latest_id}")
+
+
+# Create new version of existing concept DOI
+
+r = requests.post(
+    f"{BASE_URL}/deposit/depositions/{latest_id}/actions/newversion",
+    headers=json_headers)
+
+r.raise_for_status()
+
+latest_draft = r.json()["links"]["latest_draft"]
+
+
+# Get draft deposition information
+
+r = requests.get(
+    latest_draft,
+    headers=json_headers,
+)
+
+r.raise_for_status()
+
+deposition = r.json()
+
+deposition_id = deposition["id"]
+bucket_url = deposition["links"]["bucket"]
+
+print(f"Draft deposition: {deposition_id}")
+
+
+# Upload release archive
+
+
+tag = os.environ["GITHUB_REF_NAME"]
+repo_name = os.environ['GITHUB_REPOSITORY']
+
+
+archive_url = (
+    f"https://github.com/"
+    f"{repo_name}"
+    f"/archive/refs/tags/{tag}.zip"
+)
+
+archive_name = f"{tag}.zip"
+
+download = requests.get(archive_url)
+download.raise_for_status()
+
+with open(archive_name, "wb") as fp:
+    fp.write(download.content)
+
+with open(archive_name, "rb") as fp:
+    r = requests.put(
+        f"{bucket_url}/{archive_name}",
+        data=fp,
+        headers=upload_headers
+    )
+
+r.raise_for_status()
+
+print(f"Uploaded {archive_name}")
+
+
+# Update metadata
+
+data = {
+    "metadata": metadata
+}
+
+r = requests.put(
+    f"{BASE_URL}/deposit/depositions/{deposition_id}",
+    data=json.dumps(data),
+    headers=json_headers
+)
+
+r.raise_for_status()
+
+print("Metadata updated")
+
+
+# Publish
+
+
+r = requests.post(
+    f"{BASE_URL}/deposit/depositions/{deposition_id}/actions/publish",
+    headers=json_headers,
+)
+
+r.raise_for_status()
+
+record = r.json()
+
+print(f"Published DOI: {record.get('doi')}")
+print(f"Record URL: {record['links']['html']}")

.github/workflows/zenodo.yml

@@ -0,0 +1,27 @@
+name: Publish Release to Zenodo
+
+on:
+  release:
+    types: [published]
+
+jobs:
+  zenodo:
+    runs-on: ubuntu-latest
+    environment: Zenodo_sandbox
+
+    permissions:
+      contents: read
+
+    env:
+      ZENODO_TOKEN: ${{ secrets.ZENODO_TOKEN }}
+      ZENODO_CONCEPT_ID: ${{ vars.ZENODO_CONCEPT_ID }}
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install dependencies
+        run: |
+          pip install requests pyyaml
+
+      - name: Create Zenodo deposition
+        run: python .github/scripts/publish_to_zenodo.py

get_schema_info_from_cff.py

@@ -0,0 +1,65 @@
+"""Fill bioschemas.yml with metadata sourced from CITATION.cff."""
+
+import json
+from datetime import date, datetime
+
+import yaml
+
+CITATION_FILE = "CITATION.cff"
+BIOSCHEMAS_FILE = "bioschemas.yml"
+
+SPDX_LICENSE_URLS = {
+    "CC-BY-4.0": "https://creativecommons.org/licenses/by/4.0/"
+}
+
+
+def load_yaml(path):
+    with open(path, encoding="utf-8") as f:
+        return yaml.safe_load(f)
+
+
+def cff_to_authors(authors):
+    result = []
+    for author in authors:
+        if "family-names" in author:
+            result.append(
+                {
+                    "@type": "Person",
+                    "name": f"{author['given-names']} {author['family-names']}",
+                }
+            )
+        else:
+            result.append({"@type": "Organization", "name": author["name"]})
+    return result
+
+
+def license_url(spdx_id):
+    return SPDX_LICENSE_URLS.get(spdx_id, spdx_id)
+
+
+def stringify(value):
+    if isinstance(value, (date, datetime)):
+        return value.isoformat()
+    return value
+
+
+def main():
+    cff = load_yaml(CITATION_FILE)
+    bioschemas = load_yaml(BIOSCHEMAS_FILE)
+
+    bioschemas["name"] = cff["title"]
+    bioschemas["description"] = cff.get("abstract", "")
+    bioschemas["author"] = cff_to_authors(cff["authors"])
+    bioschemas["version"] = stringify(cff.get("version", ""))
+    bioschemas["url"] = cff.get("url", "")
+    bioschemas["license"] = license_url(cff.get("license", ""))
+    if cff.get("doi"):
+        bioschemas["identifier"] = f"https://doi.org/{cff['doi']}"
+
+    with open(BIOSCHEMAS_FILE, "w", encoding="utf-8") as f:
+        json.dump(bioschemas, f, indent=4)
+        f.write("\n")
+
+
+if __name__ == "__main__":
+    main()