[ENG-7879] Ability to resync preprints that have missing doi with Crossref

admin/management/views.py

@@ -150,10 +150,12 @@ def post(self, request):
 class BulkResync(ManagementCommandPermissionView):
 
     def post(self, request):
+        missing_dois_only = request.POST.get('missing_preprint_dois_only', False)
         sync_doi_metadata.apply_async(kwargs={
             'modified_date': timezone.now(),
             'batch_size': None,
-            'dry_run': False
+            'dry_run': False,
+            'missing_preprint_dois_only': missing_dois_only
         })
         messages.success(request, 'Resyncing with CrossRef and DataCite! It will take some time.')
         return redirect(reverse('management:commands'))

admin/templates/management/commands.html

@@ -74,7 +74,7 @@ <h4> <u>Ban spam users by regular expression</u>
                     <label>Nodes:</label>   <input type="checkbox" name="node_ban" checked /><br>
                     <label>Registrations:</label>   <input type="checkbox" name="registration_ban" checked /><br>
                     <label>Preprints:</label>  <input type="checkbox" name="preprint_ban" checked /><br>
-                    <input class="btn btn-danger" type="submit" value="Run" style="color: red" />
+                    <input class="btn btn-danger" type="submit" value="Run" style="color: white" />
                 </form>
             </ul>
             <section>
@@ -133,6 +133,7 @@ <h4><u>Resync with CrossRef and DataCite</u></h4>
                 <form method="post"
                       action="{% url 'management:bulk-resync'%}">
                     {% csrf_token %}
+                    <label>Only preprints missing DOI:</label>  <input type="checkbox" name="missing_preprint_dois_only"/><br>
                     <nav>
                         <input class="btn btn-success" type="submit" value="Run" />
                     </nav>

osf/management/commands/sync_doi_metadata.py

@@ -5,7 +5,7 @@
 
 from django.contrib.contenttypes.models import ContentType
 from django.core.management.base import BaseCommand
-from osf.models import GuidMetadataRecord, Identifier, Registration
+from osf.models import GuidMetadataRecord, Identifier, Registration, Preprint
 from framework.celery_tasks import app
 
 logger = logging.getLogger(__name__)
@@ -14,26 +14,30 @@
 RATE_LIMIT_RETRY_DELAY = 60 * 5
 
 
-@app.task(name='osf.management.commands.sync_doi_metadata', max_retries=5, default_retry_delay=RATE_LIMIT_RETRY_DELAY)
-def sync_identifier_doi(identifier_id):
+@app.task(name='osf.management.commands.sync_doi_metadata', bind=True, acks_late=True, max_retries=5, default_retry_delay=RATE_LIMIT_RETRY_DELAY)
+def sync_identifier_doi(self, identifier_id):
     try:
         identifier = Identifier.objects.get(id=identifier_id)
         identifier.referent.request_identifier_update('doi')
         identifier.save()
         logger.info(f'Doi update for {identifier.value} complete')
     except Exception as err:
         logger.warning(f'[{err.__class__.__name__}] Doi update for {identifier.value} failed because of error: {err}')
-        sync_identifier_doi.retry(exc=err, countdown=RATE_LIMIT_RETRY_DELAY)
+        self.retry()
 
 
 @app.task(name='osf.management.commands.sync_doi_metadata_command', max_retries=5, default_retry_delay=RATE_LIMIT_RETRY_DELAY)
-def sync_doi_metadata(modified_date, batch_size=100, dry_run=True, sync_private=False, rate_limit=100):
+def sync_doi_metadata(modified_date, batch_size=100, dry_run=True, sync_private=False, rate_limit=100, missing_preprint_dois_only=False):
     identifiers = Identifier.objects.filter(
         category='doi',
         deleted__isnull=True,
         modified__lte=modified_date,
         object_id__isnull=False,
     )
+    if missing_preprint_dois_only:
+        sync_preprint_missing_dois.apply_async(kwargs={'rate_limit': rate_limit})
+        identifiers = identifiers.exclude(content_type=ContentType.objects.get_for_model(Preprint))
+
     if batch_size:
         identifiers = identifiers[:batch_size]
         rate_limit = batch_size if batch_size > rate_limit else rate_limit
@@ -55,6 +59,27 @@ def sync_doi_metadata(modified_date, batch_size=100, dry_run=True, sync_private=
             sync_identifier_doi.apply_async(kwargs={'identifier_id': identifier.id})
 
 
+@app.task(name='osf.management.commands.sync_preprint_missing_dois', max_retries=5, default_retry_delay=RATE_LIMIT_RETRY_DELAY)
+def sync_preprint_missing_dois(rate_limit):
+    preprints = Preprint.objects.filter(preprint_doi_created=None)
+    for record_number, preprint in enumerate(preprints, 1):
+        # in order to not reach rate limit that CrossRef has, we make delay
+        if not record_number % rate_limit:
+            time.sleep(RATE_LIMIT_RETRY_DELAY)
+
+        async_request_identifier_update.apply_async(kwargs={'preprint_id': preprint._id})
+
+
+@app.task(name='osf.management.commands.async_request_identifier_update', bind=True, acks_late=True, max_retries=5, default_retry_delay=RATE_LIMIT_RETRY_DELAY)
+def async_request_identifier_update(self, preprint_id):
+    preprint = Preprint.load(preprint_id)
+    try:
+        preprint.request_identifier_update('doi', create=True)
+    except Exception as err:
+        logger.warning(f'[{err.__class__.__name__}] Doi creation failed for the preprint with id {preprint._id} because of error: {err}')
+        self.retry()
+
+
 @app.task(name='osf.management.commands.sync_doi_empty_metadata_dataarchive_registrations_command', max_retries=5, default_retry_delay=RATE_LIMIT_RETRY_DELAY)
 def sync_doi_empty_metadata_dataarchive_registrations(modified_date, batch_size=100, dry_run=True, sync_private=False, rate_limit=100):
     registrations_ids = list(