Don't update search on HTMLFile save

readthedocs/search/signals.py

@@ -76,40 +76,3 @@ def remove_project_delete(instance, *args, **kwargs):
     # Do not index if autosync is disabled globally
     if DEDConfig.autosync_enabled():
         delete_objects_in_es.delay(**kwargs)
-
-
-@receiver(post_save, sender=HTMLFile)
-def index_html_file_save(instance, *args, **kwargs):
-    """
-    Save a HTMLFile instance based on the post_save signal.post_save.
-
-    This uses Celery to do it async, replacing how django-elasticsearch-dsl does
-    it.
-    """
-    from readthedocs.search.documents import PageDocument
-    kwargs = {
-        'app_label': HTMLFile._meta.app_label,
-        'model_name': HTMLFile.__name__,
-        'document_class': str(PageDocument),
-        'objects_id': [instance.id],
-    }
-
-    # Do not index if autosync is disabled globally
-    if DEDConfig.autosync_enabled():
-        index_objects_to_es.delay(**kwargs)
-
-
-@receiver(pre_delete, sender=HTMLFile)
-def remove_html_file_delete(instance, *args, **kwargs):
-    from readthedocs.search.documents import PageDocument
-
-    kwargs = {
-        'app_label': HTMLFile._meta.app_label,
-        'model_name': HTMLFile.__name__,
-        'document_class': str(PageDocument),
-        'objects_id': [instance.id],
-    }
-
-    # Do not index if autosync is disabled globally
-    if DEDConfig.autosync_enabled():
-        delete_objects_in_es.delay(**kwargs)

readthedocs/search/tests/conftest.py

@@ -7,6 +7,7 @@
 from django_dynamic_fixture import G
 
 from readthedocs.projects.models import Project, HTMLFile
+from readthedocs.search.documents import PageDocument
 from .dummy_data import ALL_PROJECTS, PROJECT_DATA_FILES
 
 
@@ -30,7 +31,8 @@ def all_projects(es_index, mock_processed_json, db, settings):
             # file_basename in config are without extension so add html extension
             file_name = file_basename + '.html'
             version = project.versions.all()[0]
-            G(HTMLFile, project=project, version=version, name=file_name)
+            html_file = G(HTMLFile, project=project, version=version, name=file_name)
+            PageDocument().update(html_file)
 
         projects_list.append(project)
 

readthedocs/search/tests/test_api.py

@@ -6,6 +6,7 @@
 from readthedocs.builds.models import Version
 from readthedocs.projects.models import HTMLFile
 from readthedocs.search.tests.utils import get_search_query_from_project_file
+from readthedocs.search.documents import PageDocument
 
 
 @pytest.mark.django_db
@@ -67,6 +68,7 @@ def test_doc_search_filter_by_version(self, api_client, project):
             # Make primary key to None, so django will create new object
             f.pk = None
             f.save()
+            PageDocument().update(f)
 
         search_params = {'q': query, 'project': project.slug, 'version': dummy_version.slug}
         resp = api_client.get(self.url, search_params)
@@ -88,6 +90,7 @@ def test_doc_search_pagination(self, api_client, project):
             # Make primary key to None, so django will create new object
             html_file.pk = None
             html_file.save()
+            PageDocument().update(html_file)
 
         search_params = {'q': query, 'project': project.slug, 'version': latest_version.slug}
         resp = api_client.get(self.url, search_params)

readthedocs/search/tests/test_views.py

@@ -130,21 +130,6 @@ def test_file_search_exact_match(self, client, project):
 
         assert len(result) == 1
 
-    def test_page_search_not_return_removed_page(self, client, project):
-        """Check removed page are not in the search index."""
-        query = get_search_query_from_project_file(project_slug=project.slug)
-        # Make a query to check it returns result
-        result, _ = self._get_search_result(url=self.url, client=client,
-                                            search_params={'q': query, 'type': 'file'})
-        assert len(result) == 1
-
-        # Delete all the HTML files of the project
-        HTMLFile.objects.filter(project=project).delete()
-        # Run the query again and this time there should not be any result
-        result, _ = self._get_search_result(url=self.url, client=client,
-                                            search_params={'q': query, 'type': 'file'})
-        assert len(result) == 0
-
     def test_file_search_show_projects(self, client, all_projects):
         """Test that search result page shows list of projects while searching
         for files."""