Add user-controlled-only flag to metadata retrieval command

changelog/13.feature.md

@@ -0,0 +1,2 @@
+Added the `--user-controlled-only` flag to `openscm-zenodo retrieve-metadata`.
+This is the flag to use if you want to use the retrieved metadata as the starting point for the next version of a deposit.

src/openscm_zenodo/cli/app.py

@@ -150,6 +150,18 @@ def cli(
 def retrieve_metadata_command(
     deposition_id: DEPOSITION_ID_TYPE,
     token: TOKEN_TYPE = None,
+    user_controlled_only: Annotated[
+        bool,
+        typer.Option(
+            help="""Only return metadata keys that the user can control.
+
+If this is `True`, the metadata keys controlled by Zenodo (e.g. the DOI)
+are removed from the returned metadata.
+This flag is important to use
+if you want to use the retrieved metadata
+as the starting point for the next version of a deposit."""
+        ),
+    ] = False,
     zenodo_domain: ZENODO_DOMAIN_TYPE = ZenodoDomain.production,
 ) -> None:
     """
@@ -160,7 +172,9 @@ def retrieve_metadata_command(
         zenodo_domain=zenodo_domain,
     )
 
-    metadata = zenoodo_interactor.get_metadata(deposition_id)
+    metadata = zenoodo_interactor.get_metadata(
+        deposition_id, user_controlled_only=user_controlled_only
+    )
 
     print(json.dumps(metadata, indent=2, sort_keys=True))
 

src/openscm_zenodo/zenodo.py

@@ -218,6 +218,7 @@ def get_deposition(
     def get_metadata(
         self,
         deposition_id: str,
+        user_controlled_only: bool = False,
     ) -> MetadataType:
         """
         Get the metadata for a given deposition ID
@@ -227,6 +228,16 @@ def get_metadata(
         deposition_id
             The ID of the deposition
 
+        user_controlled_only
+            Only return metadata keys that the user can control.
+
+            If this is `True`, the metadata keys controlled by Zenodo
+            (e.g. the DOI)
+            are removed from the returned metadata.
+            This flag is important to use
+            if you want to use the retrieved metadata
+            as the starting point for the next version of a deposit.
+
         Returns
         -------
         :
@@ -244,6 +255,17 @@ def get_metadata(
 
         metadata = {"metadata": deposition.json()["metadata"]}
 
+        if user_controlled_only:
+            for k in [
+                "doi",
+                "imprint_publisher",
+                "prereserve_doi",
+                "publication_date",
+                "relations",
+            ]:
+                if k in metadata["metadata"]:
+                    metadata["metadata"].pop(k)
+
         return metadata
 
     def get_bibtex_entry(

tests/integration/test_cli.py

@@ -123,6 +123,36 @@ def test_retrieve_metadata(test_data_dir):
     assert res.stdout == f"{json.dumps(exp, indent=2, sort_keys=True)}\n"
 
 
+@pytest.mark.zenodo_token
+def test_retrieve_metadata_user_controlled_only(test_data_dir):
+    """
+    Test we can retrieve only user-controlled metadata from Zenodo
+    """
+    deposition_id = "101709"
+
+    res = runner.invoke(
+        app,
+        [
+            "retrieve-metadata",
+            deposition_id,
+            "--zenodo-domain",
+            "https://sandbox.zenodo.org",
+            "--user-controlled-only",
+        ],
+    )
+    assert res.exit_code == 0, res.stderr
+    exp = {
+        "metadata": {
+            "access_right": "open",
+            "creators": [{"affiliation": None, "name": "Nicholls, Zebedee"}],
+            "license": "cc-by-4.0",
+            "title": "OpenSCM-Zenodo testing 0",
+            "upload_type": "dataset",
+        }
+    }
+    assert res.stdout == f"{json.dumps(exp, indent=2, sort_keys=True)}\n"
+
+
 def test_retrieve_bibtex(test_data_dir):
     """
     Test we can retrieve a bibtex entry from Zenodo

tests/integration/test_flow.py

@@ -4,8 +4,11 @@
 
 from __future__ import annotations
 
+import copy
+import datetime as dt
 import json
 import os
+from pathlib import Path
 
 import pytest
 import requests
@@ -14,14 +17,14 @@
 
 
 @pytest.mark.zenodo_token
-def test_default_end_to_end_flow(test_data_dir):
+def test_default_end_to_end_flow(test_data_dir, tmpdir):
     """
     Test we can start with an ID and end up publishing a new version
     """
+    tmpdir = Path(tmpdir)
+
     any_deposition_id = "101709"
-    metadata_file = test_data_dir / "test-deposit-metadata.json"
     sub_dir_file = test_data_dir / "sub-dir" / "file-in-sub-dir.txt"
-    files_to_upload = [metadata_file, sub_dir_file]
 
     zenoodo_interactor = ZenodoInteractor(
         token=os.environ["ZENODO_TOKEN"],
@@ -48,21 +51,37 @@ def test_default_end_to_end_flow(test_data_dir):
         for response in remove_all_files_responses
     )
 
-    with open(metadata_file) as fh:
-        metadata = json.load(fh)
+    # Retrieve metadata from existing version
+    metadata_current = zenoodo_interactor.get_metadata(
+        latest_deposition_id, user_controlled_only=True
+    )
+
+    # Update this metadata for this version
+    timestamp = dt.datetime.now(dt.timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
+    metadata_updated = copy.deepcopy(metadata_current)
+    metadata_updated["metadata"][
+        "title"
+    ] = f"OpenSCM-Zenodo testing test run {timestamp}"
+
+    # Save the metadata to a file
+    metadata_file = tmpdir / "test-deposit-metadata.json"
+    with open(metadata_file, "w") as fh:
+        json.dump(metadata_current, fh)
 
     update_metadata_response = zenoodo_interactor.update_metadata(
         deposition_id=new_deposition_id,
-        metadata=metadata,
+        metadata=metadata_updated,
     )
     assert isinstance(update_metadata_response, requests.models.Response)
 
     # Just a test that this exists really, but handy trick to know
     reserved_doi = get_reserved_doi(update_metadata_response)
     assert "10.5281/zenodo" in reserved_doi
 
+    # Upload files
     bucket_url = zenoodo_interactor.get_bucket_url(deposition_id=new_deposition_id)
 
+    files_to_upload = [metadata_file, sub_dir_file]
     for file in files_to_upload:
         resp = zenoodo_interactor.upload_file_to_bucket_url(
             file,
@@ -79,12 +98,14 @@ def test_default_end_to_end_flow(test_data_dir):
     zenodo_altered_keys = ["prereserve_doi"]
 
     comparable_metadata_from_user = {
-        k: v for k, v in metadata["metadata"].items() if k not in zenodo_altered_keys
+        k: v
+        for k, v in metadata_updated["metadata"].items()
+        if k not in zenodo_altered_keys
     }
     comparable_metadata_from_publish_response = {
         k: v
         for k, v in publish_response_json["metadata"].items()
-        if k in metadata["metadata"] and k not in zenodo_altered_keys
+        if k in metadata_updated["metadata"] and k not in zenodo_altered_keys
     }
     assert comparable_metadata_from_user == comparable_metadata_from_publish_response