[ENG-6314] Set private title instead of an empty string when sync wit…

…h CrossRef (#10969)

## Purpose

CrossRef fails silently when a status of preprint is not public. In this case we should pass a specific string instead of an empty string

## Notes

1. Matt's comment: "The XML file that we send has an empty <title>, which causes CrossRef to (silently) not accept it. If preprint spam_status is.SPAM, title should be REMOVED DUE TO POLICY VIOLATIONS, otherwise WITHDRAWN."
2. Import is not moved at the top of the file because of circular imports

## Ticket

https://openscience.atlassian.net/browse/ENG-6314

tests/identifiers/test_crossref.py

@@ -196,7 +196,6 @@ def test_metadata_for_deleted_node(self, crossref_client, preprint):
         assert not root.find('.//{%s}contributors' % crossref.CROSSREF_NAMESPACE)
 
         assert root.find('.//{%s}group_title' % crossref.CROSSREF_NAMESPACE).text == preprint.provider.name
-        assert not root.find('.//{%s}title' % crossref.CROSSREF_NAMESPACE).text
         assert not root.find('.//{%s}abstract/' % crossref.JATS_NAMESPACE)
         assert not root.find('.//{%s}license_ref' % crossref.CROSSREF_ACCESS_INDICATORS)
 
@@ -348,4 +347,26 @@ def test_metadata_for_affiliated_institutions(self, crossref_client, preprint):
         root = lxml.etree.fromstring(crossref_xml)
         contributors = root.find('.//{%s}contributors' % crossref.CROSSREF_NAMESPACE)
         assert contributors.find('.//{%s}institution_name' % crossref.CROSSREF_NAMESPACE).text == institution.name
-        assert contributors.find('.//{%s}institution_id' % crossref.CROSSREF_NAMESPACE).text == institution.ror_uri
+        assert contributors.find('.//{%s}institution_id' % crossref.CROSSREF_NAMESPACE).text == institution.ror_uri
+
+    def test_public_preprint_title_is_composed_correctly(self, crossref_client, preprint):
+        preprint.title = 'My Title'
+        preprint.save()
+
+        xml = crossref_client.build_metadata(preprint)
+        assert '<titles><title>My Title</title></titles>' in str(xml)
+
+    def test_private_preprint_title_is_composed_correctly(self, crossref_client, preprint):
+        preprint.title = 'My Title'
+        preprint.is_public = False
+        preprint.save()
+        xml = crossref_client.build_metadata(preprint)
+        assert '<titles><title>WITHDRAWN</title></titles>' in str(xml)
+
+    def test_spam_preprint_title_is_composed_correctly(self, crossref_client, preprint):
+        preprint.title = 'My Title'
+        preprint.save()
+
+        preprint.confirm_spam(save=True)
+        xml = crossref_client.build_metadata(preprint)
+        assert '<titles><title>REMOVED DUE TO POLICY VIOLATIONS</title></titles>' in str(xml)

website/identifiers/clients/crossref.py

@@ -86,6 +86,8 @@ def build_posted_content(self, preprint, element, include_relation):
         preprint - preprint to build posted_content for
         element - namespace element to use when building parts of the XML structure
         """
+        from osf.models import SpamStatus
+
         posted_content = element.posted_content(
             element.group_title(preprint.provider.name),
             type='preprint'
@@ -94,7 +96,8 @@ def build_posted_content(self, preprint, element, include_relation):
         if status == 'public':
             posted_content.append(element.contributors(*self._crossref_format_contributors(element, preprint)))
 
-        title = element.title(remove_control_characters(preprint.title)) if status == 'public' else element.title('')
+        private_title = 'REMOVED DUE TO POLICY VIOLATIONS' if preprint.spam_status == SpamStatus.SPAM else 'WITHDRAWN'
+        title = element.title(remove_control_characters(preprint.title)) if status == 'public' else element.title(private_title)
         posted_content.append(element.titles(title))
 
         posted_content.append(element.posted_date(*self._crossref_format_date(element, preprint.date_published)))