cli: Warn when metadata-only deposit without metadata provenance

Related to T3677

swh/deposit/cli/client.py

@@ -604,16 +604,24 @@ def metadata_only(ctx, url, username, password, metadata_path, output_format):
     from xml.etree import ElementTree
 
     from swh.deposit.client import PublicApiDepositClient
-    from swh.deposit.utils import parse_swh_reference
+    from swh.deposit.utils import parse_swh_metadata_provenance, parse_swh_reference
 
     # Parse to check for a swhid presence within the metadata file
     with open(metadata_path, "r") as f:
         metadata_raw = f.read()
-    actual_swhid = parse_swh_reference(ElementTree.fromstring(metadata_raw))
+    metadata_tree = ElementTree.fromstring(metadata_raw)
+    actual_swhid = parse_swh_reference(metadata_tree)
 
     if not actual_swhid:
         raise InputError("A SWHID must be provided for a metadata-only deposit")
 
+    meta_prov_url = parse_swh_metadata_provenance(metadata_tree)
+    if not meta_prov_url:
+        logger.warning(
+            "A '<swh:metadata-provenance>' should be provided for a metadata-only "
+            "deposit"
+        )
+
     with trap_and_report_exceptions():
         client = PublicApiDepositClient(url=_url(url), auth=(username, password))
         collection = _collection(client)

swh/deposit/tests/cli/test_client.py

@@ -8,6 +8,7 @@ import contextlib
 import json
 import logging
 import os
+from typing import Optional
 from unittest.mock import MagicMock
 from xml.etree import ElementTree
 
@@ -860,8 +861,17 @@ def test_cli_update_metadata_with_swhid_on_other_status_deposit(
     }
 
 
+@pytest.mark.parametrize(
+    "metadata_entry_key", ["entry-data-with-swhid", "entry-data-with-swhid-no-prov"]
+)
 def test_cli_metadata_only_deposit_full_metadata_file(
-    datadir, requests_mock_datadir, cli_runner, atom_dataset, tmp_path,
+    datadir,
+    requests_mock_datadir,
+    cli_runner,
+    atom_dataset,
+    tmp_path,
+    metadata_entry_key,
+    caplog,
 ):
     """Post metadata-only deposit through cli
 
@@ -870,7 +880,7 @@ def test_cli_metadata_only_deposit_full_metadata_file(
     """
     api_url_basename = "deposit.test.metadataonly"
     swhid = "swh:1:dir:ef04a768181417fbc5eef4243e2507915f24deea"
-    metadata = atom_dataset["entry-data-with-swhid"].format(swhid=swhid)
+    metadata = atom_dataset[metadata_entry_key].format(swhid=swhid)
     metadata_path = os.path.join(tmp_path, "entry-data-with-swhid.xml")
     with open(metadata_path, "w") as m:
         m.write(metadata)
@@ -901,6 +911,19 @@ def test_cli_metadata_only_deposit_full_metadata_file(
     assert "error" not in actual_deposit_status
     assert actual_deposit_status == expected_deposit_status
 
+    count_warnings = 0
+    warning_record: Optional[str] = None
+    for (_, log_level, msg) in caplog.record_tuples:
+        if log_level == logging.WARNING:
+            count_warnings += 1
+            warning_record = msg
+
+    if "no-prov" in metadata_entry_key:
+        assert count_warnings == 1
+        assert "metadata-provenance>' should be provided" in warning_record
+    else:
+        assert count_warnings == 0
+
 
 def test_cli_metadata_only_deposit_invalid_swhid(
     datadir, requests_mock_datadir, cli_runner, atom_dataset, tmp_path,

swh/deposit/tests/data/atom/entry-data-with-swhid-no-prov.xml

+<?xml version="1.0"?>
+<entry xmlns="http://www.w3.org/2005/Atom"
+         xmlns:codemeta="https://doi.org/10.5063/SCHEMA/CODEMETA-2.0"
+         xmlns:swh="https://www.softwareheritage.org/schema/2018/deposit">
+    <title>Awesome Compiler</title>
+    <id>urn:uuid:1225c695-cfb8-4ebb-aaaa-80da344efa6a</id>
+    <author>dudess</author>
+    <swh:deposit>
+      <swh:reference>
+        <swh:object swhid="{swhid}" />
+      </swh:reference>
+    </swh:deposit>
+</entry>

swh/deposit/tests/data/atom/entry-data-with-swhid.xml

 <?xml version="1.0"?>
 <entry xmlns="http://www.w3.org/2005/Atom"
          xmlns:codemeta="https://doi.org/10.5063/SCHEMA/CODEMETA-2.0"
+         xmlns:schema="http://schema.org/"
          xmlns:swh="https://www.softwareheritage.org/schema/2018/deposit">
     <title>Awesome Compiler</title>
     <id>urn:uuid:1225c695-cfb8-4ebb-aaaa-80da344efa6a</id>
@@ -9,5 +10,8 @@
       <swh:reference>
         <swh:object swhid="{swhid}" />
       </swh:reference>
+      <swh:metadata-provenance>
+        <schema:url>https://inria.halpreprod.archives-ouvertes.fr/hal-abcdefgh</schema:url>
+      </swh:metadata-provenance>
     </swh:deposit>
 </entry>