def test_boosts(self): """Check if boosts are doing their job""" if not self.manager.connected_to_es: return # 1. Create topics (with identical titles), posts (with identical texts), an article and a tuto text = 'test' topic_1_solved_sticky = TopicFactory(forum=self.forum, author=self.user) topic_1_solved_sticky.title = text topic_1_solved_sticky.subtitle = '' topic_1_solved_sticky.solved_by = self.user topic_1_solved_sticky.is_sticky = True topic_1_solved_sticky.save() post_1 = PostFactory(topic=topic_1_solved_sticky, author=self.user, position=1) post_1.text = post_1.text_html = text post_1.save() post_2_useful = PostFactory(topic=topic_1_solved_sticky, author=self.user, position=2) post_2_useful.text = post_2_useful.text_html = text post_2_useful.is_useful = True post_2_useful.like = 5 post_2_useful.dislike = 2 # l/d ratio above 1 post_2_useful.save() topic_2_locked = TopicFactory(forum=self.forum, author=self.user, title=text) topic_2_locked.title = text topic_2_locked.subtitle = '' topic_2_locked.is_locked = True topic_2_locked.save() post_3_ld_below_1 = PostFactory(topic=topic_2_locked, author=self.user, position=1) post_3_ld_below_1.text = post_3_ld_below_1.text_html = text post_3_ld_below_1.like = 2 post_3_ld_below_1.dislike = 5 # l/d ratio below 1 post_3_ld_below_1.save() tuto = PublishableContentFactory(type='TUTORIAL') tuto_draft = tuto.load_version() tuto.title = text tuto.authors.add(self.user) tuto.save() tuto_draft.repo_update_top_container(text, tuto.slug, text, text) chapter1 = ContainerFactory(parent=tuto_draft, db_object=tuto) chapter1.repo_update(text, 'Who cares ?', 'Same here') ExtractFactory(container=chapter1, db_object=tuto) published_tuto = publish_content(tuto, tuto_draft, is_major_update=True) tuto.sha_public = tuto_draft.current_version tuto.sha_draft = tuto_draft.current_version tuto.public_version = published_tuto tuto.save() article = PublishedContentFactory(type='ARTICLE', title=text) published_article = PublishedContent.objects.get(content_pk=article.pk) opinion_not_picked = PublishedContentFactory(type='OPINION', title=text) published_opinion_not_picked = PublishedContent.objects.get( content_pk=opinion_not_picked.pk) opinion_picked = PublishedContentFactory(type='OPINION', title=text) opinion_picked.sha_picked = opinion_picked.sha_draft opinion_picked.date_picked = datetime.datetime.now() opinion_picked.save() published_opinion_picked = PublishedContent.objects.get( content_pk=opinion_picked.pk) for model in self.indexable: if model is FakeChapter: continue self.manager.es_bulk_indexing_of_model(model) self.manager.refresh_index() self.assertEqual( len( self.manager.setup_search(Search().query( MatchAll())).execute()), 10) # 2. Reset all boosts to 1 for doc_type in settings.ZDS_APP['search']['boosts']: for key in settings.ZDS_APP['search']['boosts'][doc_type]: settings.ZDS_APP['search']['boosts'][doc_type][key] = 1.0 # 3. Test posts result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Post.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 3) # score are equals without boost: self.assertTrue(response[0].meta.score == response[1].meta.score == response[2].meta.score) settings.ZDS_APP['search']['boosts']['post']['if_first'] = 2.0 result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Post.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 3) self.assertTrue(response[0].meta.score == response[1].meta.score > response[2].meta.score) self.assertEqual(response[2].meta.id, str( post_2_useful.pk)) # post 2 is the only one not first settings.ZDS_APP['search']['boosts']['post']['if_first'] = 1.0 settings.ZDS_APP['search']['boosts']['post']['if_useful'] = 2.0 result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Post.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 3) self.assertTrue(response[0].meta.score > response[1].meta.score == response[2].meta.score) self.assertEqual(response[0].meta.id, str(post_2_useful.pk)) # post 2 is useful settings.ZDS_APP['search']['boosts']['post']['if_useful'] = 1.0 settings.ZDS_APP['search']['boosts']['post']['ld_ratio_above_1'] = 2.0 result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Post.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 3) self.assertTrue(response[0].meta.score > response[1].meta.score == response[2].meta.score) self.assertEqual(response[0].meta.id, str( post_2_useful.pk)) # post 2 have a l/d ratio of 5/2 settings.ZDS_APP['search']['boosts']['post']['ld_ratio_above_1'] = 1.0 settings.ZDS_APP['search']['boosts']['post'][ 'ld_ratio_below_1'] = 2.0 # no one would do that in real life result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Post.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 3) self.assertTrue(response[0].meta.score > response[1].meta.score == response[2].meta.score) self.assertEqual(response[0].meta.id, str( post_3_ld_below_1.pk)) # post 3 have a l/d ratio of 2/5 settings.ZDS_APP['search']['boosts']['post']['ld_ratio_below_1'] = 1.0 # 4. Test topics result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Topic.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 2) # score are equals without boost: self.assertTrue(response[0].meta.score == response[1].meta.score) settings.ZDS_APP['search']['boosts']['topic']['if_sticky'] = 2.0 result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Topic.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 2) self.assertTrue(response[0].meta.score > response[1].meta.score) self.assertEqual(response[0].meta.id, str(topic_1_solved_sticky.pk)) # topic 1 is sticky settings.ZDS_APP['search']['boosts']['topic']['if_sticky'] = 1.0 settings.ZDS_APP['search']['boosts']['topic']['if_solved'] = 2.0 result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Topic.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 2) self.assertTrue(response[0].meta.score > response[1].meta.score) self.assertEqual(response[0].meta.id, str(topic_1_solved_sticky.pk)) # topic 1 is solved settings.ZDS_APP['search']['boosts']['topic']['if_solved'] = 1.0 settings.ZDS_APP['search']['boosts']['topic'][ 'if_locked'] = 2.0 # no one would do that in real life result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Topic.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 2) self.assertTrue(response[0].meta.score > response[1].meta.score) self.assertEqual(response[0].meta.id, str(topic_2_locked.pk)) # topic 2 is locked settings.ZDS_APP['search']['boosts']['topic'][ 'if_locked'] = 1.0 # no one would do that in real life # 5. Test published contents result = self.client.get(reverse('search:query') + '?q=' + text + '&models=content', follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 5) # score are equals without boost: self.assertTrue( response[0].meta.score == response[1].meta.score == response[2]. meta.score == response[3].meta.score == response[4].meta.score) settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_article'] = 2.0 result = self.client.get(reverse('search:query') + '?q=' + text + '&models=content', follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 5) self.assertTrue(response[0].meta.score > response[1].meta.score) self.assertEqual(response[0].meta.id, str(published_article.pk)) # obvious settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_article'] = 1.0 settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_tutorial'] = 2.0 result = self.client.get(reverse('search:query') + '?q=' + text + '&models=content', follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 5) self.assertTrue(response[0].meta.score > response[1].meta.score) self.assertEqual(response[0].meta.id, str(published_tuto.pk)) # obvious settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_tutorial'] = 1.0 settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_opinion'] = 2.0 settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_opinion_not_picked'] = 4.0 # Note: in "real life", unpicked opinion would get a boost < 1. result = self.client.get(reverse('search:query') + '?q=' + text + '&models=content', follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 5) self.assertTrue(response[0].meta.score > response[1].meta.score > response[2].meta.score) self.assertEqual( response[0].meta.id, str(published_opinion_not_picked.pk)) # unpicked opinion got first self.assertEqual(response[1].meta.id, str(published_opinion_picked.pk)) settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_opinion'] = 1.0 settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_opinion_not_picked'] = 1.0 settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_medium_or_big_tutorial'] = 2.0 result = self.client.get(reverse('search:query') + '?q=' + text + '&models=content', follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 5) self.assertTrue(response[0].meta.score > response[1].meta.score) self.assertEqual(response[0].meta.id, str(published_tuto.pk)) # obvious settings.ZDS_APP['search']['boosts']['publishedcontent'][ 'if_medium_or_big_tutorial'] = 1.0 # 6. Test global boosts # NOTE: score are NOT the same for all documents, no matter how hard it tries to, small differences exists for model in self.indexable: # set a huge number to overcome the small differences: settings.ZDS_APP['search']['boosts'][ model.get_es_document_type()]['global'] = 10.0 result = self.client.get(reverse('search:query') + '?q=' + text, follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 10) self.assertEqual(response[0].meta.doc_type, model.get_es_document_type()) # obvious settings.ZDS_APP['search']['boosts'][ model.get_es_document_type()]['global'] = 1.0
def test_category_and_subcategory_impact_search(self): """If two contents do not belong to the same (sub)category""" if not self.manager.connected_to_es: return text = 'Did you ever hear the tragedy of Darth Plagueis The Wise?' # 1. Create two contents with different subcategories category_1 = 'category 1' subcategory_1 = SubCategoryFactory(title=category_1) category_2 = 'category 2' subcategory_2 = SubCategoryFactory(title=category_2) tuto_1 = PublishableContentFactory(type='TUTORIAL') tuto_1_draft = tuto_1.load_version() tuto_1.title = text tuto_1.authors.add(self.user) tuto_1.subcategory.add(subcategory_1) tuto_1.save() tuto_1_draft.description = text tuto_1_draft.repo_update_top_container(text, tuto_1.slug, text, text) chapter_1 = ContainerFactory(parent=tuto_1_draft, db_object=tuto_1) extract_1 = ExtractFactory(container=chapter_1, db_object=tuto_1) extract_1.repo_update(text, text) published_1 = publish_content(tuto_1, tuto_1_draft, is_major_update=True) tuto_1.sha_public = tuto_1_draft.current_version tuto_1.sha_draft = tuto_1_draft.current_version tuto_1.public_version = published_1 tuto_1.save() tuto_2 = PublishableContentFactory(type='TUTORIAL') tuto_2_draft = tuto_2.load_version() tuto_2.title = text tuto_2.authors.add(self.user) tuto_2.subcategory.add(subcategory_2) tuto_2.save() tuto_2_draft.description = text tuto_2_draft.repo_update_top_container(text, tuto_2.slug, text, text) chapter_2 = ContainerFactory(parent=tuto_2_draft, db_object=tuto_2) extract_2 = ExtractFactory(container=chapter_2, db_object=tuto_2) extract_2.repo_update(text, text) published_2 = publish_content(tuto_2, tuto_2_draft, is_major_update=True) tuto_2.sha_public = tuto_2_draft.current_version tuto_2.sha_draft = tuto_2_draft.current_version tuto_2.public_version = published_2 tuto_2.save() # 2. Index: self.assertEqual( len( self.manager.setup_search(Search().query( MatchAll())).execute()), 0) # index for model in self.indexable: if model is FakeChapter: continue self.manager.es_bulk_indexing_of_model(model) self.manager.refresh_index() result = self.client.get(reverse('search:query') + '?q=' + text, follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 4) # Ok # 3. Test result = self.client.get(reverse('search:query') + '?q=' + text + '&model=content&subcategory=' + subcategory_1.slug, follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 2) self.assertEqual([ int(r.meta.id) for r in response if r.meta.doc_type == 'publishedcontent' ][0], published_1.pk) self.assertEqual([ r.meta.id for r in response if r.meta.doc_type == 'chapter' ][0], tuto_1.slug + '__' + chapter_1.slug) result = self.client.get(reverse('search:query') + '?q=' + text + '&model=content&subcategory=' + subcategory_2.slug, follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 2) self.assertEqual([ int(r.meta.id) for r in response if r.meta.doc_type == 'publishedcontent' ][0], published_2.pk) self.assertEqual([ r.meta.id for r in response if r.meta.doc_type == 'chapter' ][0], tuto_2.slug + '__' + chapter_2.slug)
def test_hidden_forums_give_no_results_if_user_not_allowed(self): """Long name, isn't ?""" if not self.manager.connected_to_es: return # 1. Create a hidden forum belonging to a hidden staff group. text = 'test' group = Group.objects.create(name='Les illuminatis anonymes de ZdS') _, hidden_forum = create_category(group) self.staff.groups.add(group) self.staff.save() topic_1 = TopicFactory(forum=hidden_forum, author=self.staff, title=text) post_1 = PostFactory(topic=topic_1, author=self.user, position=1) post_1.text = post_1.text_html = text post_1.save() self.manager.es_bulk_indexing_of_model(Topic) self.manager.es_bulk_indexing_of_model(Post) self.manager.refresh_index() self.assertEqual( len( self.manager.setup_search(Search().query( MatchAll())).execute()), 2) # indexing ok # 2. search without connection and get not result result = self.client.get(reverse('search:query') + '?q=' + text, follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 0) # 3. Connect with user (not a member of the group), search, and get no result self.assertTrue( self.client.login(username=self.user.username, password='******')) result = self.client.get(reverse('search:query') + '?q=' + text, follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 0) # 4. Connect with staff, search, and get the topic and the post self.client.logout() self.assertTrue( self.client.login(username=self.staff.username, password='******')) result = self.client.get(reverse('search:query') + '?q=' + text, follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 2) # ok !
def test_indexation(self): """test the indexation and deletion of the different documents""" if not self.manager.connected_to_es: return # create a topic with a post topic = TopicFactory(forum=self.forum, author=self.user) post = PostFactory(topic=topic, author=self.user, position=1) topic = Topic.objects.get(pk=topic.pk) post = Post.objects.get(pk=post.pk) self.assertFalse(topic.es_already_indexed) self.assertTrue(topic.es_flagged) self.assertFalse(post.es_already_indexed) self.assertTrue(post.es_flagged) # create a middle-tutorial and publish it tuto = PublishableContentFactory(type='TUTORIAL') tuto.authors.add(self.user) tuto.save() tuto_draft = tuto.load_version() chapter1 = ContainerFactory(parent=tuto_draft, db_object=tuto) ExtractFactory(container=chapter1, db_object=tuto) published = publish_content(tuto, tuto_draft, is_major_update=True) tuto.sha_public = tuto_draft.current_version tuto.sha_draft = tuto_draft.current_version tuto.public_version = published tuto.save() published = PublishedContent.objects.get(content_pk=tuto.pk) self.assertFalse(published.es_already_indexed) self.assertTrue(published.es_flagged) # 1. index all for model in self.indexable: if model is FakeChapter: continue self.manager.es_bulk_indexing_of_model(model, force_reindexing=False) self.manager.refresh_index() topic = Topic.objects.get(pk=topic.pk) post = Post.objects.get(pk=post.pk) self.assertTrue(topic.es_already_indexed) self.assertFalse(topic.es_flagged) self.assertTrue(post.es_already_indexed) self.assertFalse(post.es_flagged) published = PublishedContent.objects.get(content_pk=tuto.pk) self.assertTrue(published.es_already_indexed) self.assertFalse(published.es_flagged) s = Search() s.query(MatchAll()) results = self.manager.setup_search(s).execute() self.assertEqual(len(results), 4) # get 4 results, one of each type must_contain = { 'post': False, 'topic': False, 'publishedcontent': False, 'chapter': False } id_must_be = { 'post': str(post.pk), 'topic': str(topic.pk), 'publishedcontent': str(published.pk), 'chapter': tuto.slug + '__' + chapter1.slug } for hit in results: doc_type = hit.meta.doc_type must_contain[doc_type] = True self.assertEqual(hit.meta.id, id_must_be[doc_type]) self.assertTrue(all(must_contain)) # 2. Test what reindexation will do: new_topic = TopicFactory(forum=self.forum, author=self.user) new_post = PostFactory(topic=new_topic, author=self.user, position=1) pk_of_topics_to_reindex = [] for item in Topic.get_es_indexable(force_reindexing=False): pk_of_topics_to_reindex.append(item.pk) pk_of_posts_to_reindex = [] for item in Post.get_es_indexable(force_reindexing=False): pk_of_posts_to_reindex.append(item.pk) self.assertTrue(topic.pk not in pk_of_topics_to_reindex) self.assertTrue(new_topic.pk in pk_of_topics_to_reindex) self.assertTrue(post.pk not in pk_of_posts_to_reindex) self.assertTrue(new_post.pk in pk_of_posts_to_reindex) for model in self.indexable: # ok, so let's index that if model is FakeChapter: continue self.manager.es_bulk_indexing_of_model(model, force_reindexing=False) self.manager.refresh_index() s = Search() s.query(MatchAll()) results = self.manager.setup_search(s).execute() self.assertEqual(len(results), 6) # good! # 3. Test single deletion: new_post = Post.objects.get(pk=new_post.pk) self.manager.delete_document(new_post) self.manager.refresh_index() s = Search() s.query(MatchAll()) results = self.manager.setup_search(s).execute() self.assertEqual(len(results), 5) # one is missing for hit in results: self.assertTrue(hit.meta.doc_type != Post.get_es_document_type() or hit.meta.id != new_post.es_id) # 4. Test "delete_by_query_deletion": topic = Topic.objects.get(pk=topic.pk) new_topic = Topic.objects.get(pk=new_topic.pk) self.manager.delete_by_query(Topic.get_es_document_type(), MatchAll()) # the two topic are deleted self.manager.refresh_index() s = Search() s.query(MatchAll()) results = self.manager.setup_search(s).execute() self.assertEqual(len(results), 3) for hit in results: self.assertTrue(hit.meta.doc_type != Topic.get_es_document_type() or hit.meta.id != new_topic.es_id) self.assertTrue(hit.meta.doc_type != Topic.get_es_document_type() or hit.meta.id != topic.es_id) # 5. Test that the deletion of an object also triggers its deletion in ES post = Post.objects.get(pk=post.pk) post.delete() self.manager.refresh_index() s = Search() s.query(MatchAll()) results = self.manager.setup_search(s).execute() self.assertEqual(len(results), 2) for hit in results: self.assertTrue(hit.meta.doc_type != Post.get_es_document_type() or hit.meta.id != post.es_id) # 6. Test full desindexation: for model in self.indexable: if model is FakeChapter: continue self.manager.clear_indexing_of_model(model) # note "topic" is gone since "post" is gone, due to relationships at the Django level new_topic = Topic.objects.get(pk=new_topic.pk) new_post = Post.objects.get(pk=new_post.pk) self.assertFalse(new_topic.es_already_indexed) self.assertTrue(new_topic.es_flagged) self.assertFalse(new_post.es_already_indexed) self.assertTrue(new_post.es_flagged) published = PublishedContent.objects.get(content_pk=tuto.pk) self.assertFalse(published.es_already_indexed) self.assertTrue(published.es_flagged)
def test_special_case_of_contents(self): """test that the old publishedcontent does not stay when a new one is created""" if not self.manager.connected_to_es: return # 1. Create a middle-tutorial, publish it, then index it tuto = PublishableContentFactory(type='TUTORIAL') tuto.authors.add(self.user) tuto.save() tuto_draft = tuto.load_version() chapter1 = ContainerFactory(parent=tuto_draft, db_object=tuto) ExtractFactory(container=chapter1, db_object=tuto) published = publish_content(tuto, tuto_draft, is_major_update=True) tuto.sha_public = tuto_draft.current_version tuto.sha_draft = tuto_draft.current_version tuto.public_version = published tuto.save() self.manager.es_bulk_indexing_of_model(PublishedContent, force_reindexing=True) # index self.manager.refresh_index() first_publication = PublishedContent.objects.get(content_pk=tuto.pk) self.assertTrue(first_publication.es_already_indexed) self.assertFalse(first_publication.es_flagged) s = Search() s.query(MatchAll()) results = self.manager.setup_search(s).execute() self.assertEqual( len(results), 2) # get 2 results, one for the content and one for the chapter self.assertEqual(PublishedContent.objects.count(), 1) # 2. Change thet title, which will trigger a change in the slug tuto = PublishableContent.objects.get(pk=tuto.pk) versioned = tuto.load_version(sha=tuto.sha_draft) tuto.title = u'un titre complètement différent!' tuto.save() versioned.repo_update_top_container(tuto.title, tuto.slug, u'osef', u'osef') second_publication = publish_content(tuto, versioned, True) tuto.sha_public = versioned.current_version tuto.sha_draft = versioned.current_version tuto.public_version = second_publication tuto.save() self.assertEqual(PublishedContent.objects.count(), 2) # now there is two objects ... first_publication = PublishedContent.objects.get( pk=first_publication.pk) self.assertTrue(first_publication.must_redirect ) # .. including the first one, for redirection self.manager.refresh_index() s = Search() s.query(MatchAll()) results = self.manager.setup_search(s).execute() self.assertEqual( len(results), 0 ) # the old one is gone (and we need to reindex to get the new one) # 3. Check if indexation brings the new one, and not the old one self.manager.es_bulk_indexing_of_model(PublishedContent, force_reindexing=True) # index self.manager.refresh_index() first_publication = PublishedContent.objects.get( pk=first_publication.pk) second_publication = PublishedContent.objects.get( pk=second_publication.pk) s = Search() s.query(MatchAll()) results = self.manager.setup_search(s).execute() self.assertEqual(len(results), 2) # Still 2, not 4 ! found_old = False found_new = False for hit in results: if hit.meta.doc_type == PublishedContent.get_es_document_type(): if hit.meta.id == first_publication.es_id: found_old = True if hit.meta.id == second_publication.es_id: found_new = True self.assertTrue(found_new) self.assertFalse(found_old)
def test_change_publishedcontents_impacts_chapter(self): if not self.manager.connected_to_es: return # 1. Create middle-size content and index it text = 'test' tuto = PublishableContentFactory(type='TUTORIAL') tuto_draft = tuto.load_version() tuto.title = text tuto.authors.add(self.user) tuto.save() tuto_draft.repo_update_top_container( text, tuto.slug, text, text) # change title to be sure it will match chapter1 = ContainerFactory(parent=tuto_draft, db_object=tuto) chapter1.repo_update(text, text, text) extract = ExtractFactory(container=chapter1, db_object=tuto) extract.repo_update(text, text) published = publish_content(tuto, tuto_draft, is_major_update=True) tuto.sha_public = tuto_draft.current_version tuto.sha_draft = tuto_draft.current_version tuto.public_version = published tuto.save() self.manager.es_bulk_indexing_of_model(PublishedContent) self.manager.refresh_index() self.assertEqual( len( self.manager.setup_search(Search().query( MatchAll())).execute()), 2) # indexing ok result = self.client.get(reverse('search:query') + '?q=' + text + '&models=content', follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 2) chapters = [r for r in response if r.meta.doc_type == 'chapter'] self.assertEqual(chapters[0].meta.doc_type, FakeChapter.get_es_document_type()) self.assertEqual(chapters[0].meta.id, published.content_public_slug + '__' + chapter1.slug) # 2. Change tuto: delete chapter and insert new one ! tuto = PublishableContent.objects.get(pk=tuto.pk) tuto_draft = tuto.load_version() tuto_draft.children[0].repo_delete() # chapter 1 is gone ! another_text = 'another thing' self.assertTrue( text not in another_text ) # to prevent a future modification from breaking this test chapter2 = ContainerFactory(parent=tuto_draft, db_object=tuto) chapter2.repo_update(another_text, another_text, another_text) extract2 = ExtractFactory(container=chapter2, db_object=tuto) extract2.repo_update(another_text, another_text) published = publish_content(tuto, tuto_draft, is_major_update=False) tuto.sha_public = tuto_draft.current_version tuto.sha_draft = tuto_draft.current_version tuto.public_version = published tuto.save() self.manager.es_bulk_indexing_of_model(PublishedContent) self.manager.refresh_index() self.assertEqual( len( self.manager.setup_search(Search().query( MatchAll())).execute()), 2) # 2 objects, not 3 ! result = self.client.get(reverse('search:query') + '?q=' + text + '&models=content', follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() contents = [r for r in response if r.meta.doc_type != 'chapter'] self.assertEqual(response.hits.total, len(contents)) # no chapter found anymore result = self.client.get(reverse('search:query') + '?q=' + another_text + '&models=content', follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() chapters = [r for r in response if r.meta.doc_type == 'chapter'] self.assertEqual(response.hits.total, 1) self.assertEqual(chapters[0].meta.doc_type, FakeChapter.get_es_document_type()) self.assertEqual(chapters[0].meta.id, published.content_public_slug + '__' + chapter2.slug) # got new chapter
def search(self, search): sort = None if search.sort is None else search.sort.translate() if not search.words: query = MatchAll() else: query = MultiMatch(query=search.words, fields=['title^10', 'content^5', 'description^5', 'location^3', 'category^2', 'organization', 'website']) elastic_search = Search(index=self.index_name)\ .doc_type(StarDocument)\ .query(query)\ .highlight('content', fragment_size=50) elastic_search = elastic_search[search.start:search.start + search.size] # Filter results for type and ages if search.types: elastic_search = elastic_search.filter('terms', **{"type": search.types}) if search.ages: elastic_search = elastic_search.filter('terms', **{"ages": search.ages}) # Filter results by date if search.date: elastic_search = elastic_search.filter('range', **{"date": {"gte": search.date}}) else: elastic_search = elastic_search.filter('bool', **{"should": [ {"range": {"date": {"gte": datetime.datetime.now()}}}, # Future events OR {"bool": {"must_not": {"exists": {"field": "date"}}}} # Date field is empty ]}) if sort is not None: elastic_search = elastic_search.sort(sort) if search.category and search.category.id: elastic_search = elastic_search.filter('terms', category=[str(search.category.search_path())]) if search.category.calculate_level() == 0: exclude = ".*\\,.*\\,.*"; include = str(search.category.id) + "\\,.*" aggregation = A("terms", field='category', exclude=exclude, include=include) elif search.category.calculate_level() == 1: include = ".*\\,.*\\,.*"; aggregation = A("terms", field='category', include=include) else: aggregation = A("terms", field='category') else: aggregation = A("terms", field='category', exclude=".*\\,.*") elastic_search.aggs.bucket('terms', aggregation) elastic_search.aggs.bucket('type', A("terms", field='type')) elastic_search.aggs.bucket('ages', A("terms", field='ages')) # KEEPING FOR NOW - THESE WERE THE ORIGINAL FACETS WE HAD SET UP. WILL NEED TO CONVERT TO AGGREGATIONS # IF WE WANT TO KEEP ANY OF THESE. # 'Location': elasticsearch_dsl.TermsFacet(field='location'), # 'Type': elasticsearch_dsl.TermsFacet(field='label'), # 'Age Range': elasticsearch_dsl.TermsFacet(field='age_range'), # 'Category': elasticsearch_dsl.TermsFacet(field='category'), # 'Organization': elasticsearch_dsl.TermsFacet(field='organization'), # 'Status': elasticsearch_dsl.TermsFacet(field='status'), # 'Topic': elasticsearch_dsl.TermsFacet(field='topic'), return elastic_search.execute()
def test_es_manager(self): """Test the behavior of the ``es_manager`` command""" if not self.index_manager.connected_to_es: return # in the beginning: the void self.assertTrue(self.index_manager.index not in self.index_manager.es.cat.indices()) text = 'Ceci est un texte de test' # create a topic with a post topic = TopicFactory(forum=self.forum, author=self.user, title=text) post = PostFactory(topic=topic, author=self.user, position=1) post.text = post.text_html = text post.save() topic = Topic.objects.get(pk=topic.pk) post = Post.objects.get(pk=post.pk) self.assertFalse(topic.es_already_indexed) self.assertTrue(topic.es_flagged) self.assertFalse(post.es_already_indexed) self.assertTrue(post.es_flagged) # create a middle-tutorial and publish it tuto = PublishableContentFactory(type='TUTORIAL') tuto.authors.add(self.user) tuto.save() tuto_draft = tuto.load_version() chapter1 = ContainerFactory(parent=tuto_draft, db_object=tuto) chapter1.repo_update(text, text, text) extract1 = ExtractFactory(container=chapter1, db_object=tuto) version = extract1.repo_update(text, text) published = publish_content(tuto, tuto_draft, is_major_update=True) tuto.sha_public = version tuto.sha_draft = version tuto.public_version = published tuto.save() published = PublishedContent.objects.get(content_pk=tuto.pk) self.assertFalse(published.es_already_indexed) self.assertTrue(published.es_flagged) # 1. test "index-all" call_command('es_manager', 'index_all') self.assertTrue(self.index_manager.es.indices.exists(self.index_manager.index)) self.index_manager.index_exists = True topic = Topic.objects.get(pk=topic.pk) post = Post.objects.get(pk=post.pk) self.assertTrue(topic.es_already_indexed) self.assertFalse(topic.es_flagged) self.assertTrue(post.es_already_indexed) self.assertFalse(post.es_flagged) published = PublishedContent.objects.get(content_pk=tuto.pk) self.assertTrue(published.es_already_indexed) self.assertFalse(published.es_flagged) s = Search() s.query(MatchAll()) results = self.index_manager.setup_search(s).execute() self.assertEqual(len(results), 4) # get 4 results, one of each type must_contain = {'post': False, 'topic': False, 'publishedcontent': False, 'chapter': False} id_must_be = { 'post': str(post.pk), 'topic': str(topic.pk), 'publishedcontent': str(published.pk), 'chapter': tuto.slug + '__' + chapter1.slug } for hit in results: doc_type = hit.meta.doc_type must_contain[doc_type] = True self.assertEqual(hit.meta.id, id_must_be[doc_type]) self.assertTrue(all(must_contain)) # 2. test "clear" self.assertTrue(self.index_manager.index in self.index_manager.es.cat.indices()) # index in call_command('es_manager', 'clear') self.assertFalse(self.index_manager.es.indices.exists(self.index_manager.index)) self.index_manager.index_exists = False # must reset every object topic = Topic.objects.get(pk=topic.pk) post = Post.objects.get(pk=post.pk) self.assertFalse(topic.es_already_indexed) self.assertTrue(topic.es_flagged) self.assertFalse(post.es_already_indexed) self.assertTrue(post.es_flagged) published = PublishedContent.objects.get(content_pk=tuto.pk) self.assertFalse(published.es_already_indexed) self.assertTrue(published.es_flagged) self.assertTrue(self.index_manager.index not in self.index_manager.es.cat.indices()) # index wiped out ! # 3. test "setup" call_command('es_manager', 'setup') self.assertTrue(self.index_manager.es.indices.exists(self.index_manager.index)) self.index_manager.index_exists = True self.assertTrue(self.index_manager.index in self.index_manager.es.cat.indices()) # index back in ... s = Search() s.query(MatchAll()) results = self.index_manager.setup_search(s).execute() self.assertEqual(len(results), 0) # ... but with nothing in it result = self.index_manager.es.indices.get_settings(index=self.index_manager.index) settings_index = result[self.index_manager.index]['settings']['index'] self.assertTrue('analysis' in settings_index) # custom analyzer was setup # 4. test "index-flagged" once ... call_command('es_manager', 'index_flagged') topic = Topic.objects.get(pk=topic.pk) post = Post.objects.get(pk=post.pk) self.assertTrue(topic.es_already_indexed) self.assertFalse(topic.es_flagged) self.assertTrue(post.es_already_indexed) self.assertFalse(post.es_flagged) published = PublishedContent.objects.get(content_pk=tuto.pk) self.assertTrue(published.es_already_indexed) self.assertFalse(published.es_flagged) s = Search() s.query(MatchAll()) results = self.index_manager.setup_search(s).execute() self.assertEqual(len(results), 4) # get the 4 results back
def get_targets(self, es): for target in Search().using(es).index(self.es_index_gene).extra( track_total_hits=True).query(MatchAll()).scan(): yield target.to_dict()
def get_targets(self, es): for target in Search().using(es).index(self.es_index_gene).query( MatchAll()).params(scroll='4h').scan(): yield str(target.meta.id)
def evaluate_score(student, client, num_resp: int = 25): """Takes a student, represented as a dictionary and an elasticsearch-py client and returns an elastic response See above student class for schema """ # Adjust weights here: base_score = 1.0 company_score = 1.0 rural_score = 2.0 tags_score = 3.0 underrep_score = 1.0 # Timezone weights are found in the timezone script query s = elasticsearch_dsl.Search(using=client, index="mentors_index").extra(explain=True) # Start by filtering the search by track s = s.filter("term", track=student["track"]) # And also by requireExtended if student["requireExtended"]: s = s.filter("term", okExtended="true") if not student["underrepresented"]: s = s.exclude("term", preferStudentUnderRep=2) # Adds one to all remaining entries in order to be sure that, in the worst case, # there are enough responses, even if they aren't a good fit base_value = Q("constant_score", filter=MatchAll(), boost=base_score) # Uses a fuzzy query to determine if a student is interested in the mentor's company, # then if so adds `weight` to the score company_q = None for company in student["interestCompanies"]: if company_q is None: company_q = Q( "function_score", query=Q("fuzzy", company=company), weight=company_score, boost_mode="replace", ) else: company_q = company_q | Q( "function_score", query=Q("fuzzy", company=company), weight=company_score, boost_mode="replace", ) if student["rural"]: # If background_rural matches on mentor and student, then add one to the score background_rural = Q("constant_score", filter=Q("term", backgroundRural=student["rural"]), boost=rural_score) else: background_rural = Q("constant_score", filter=MatchNone()) # Adds `weight` * the number of matching tags to score tags_matching = None num_interests = len(student["interestTags"]) for interest in student["interestTags"]: if tags_matching is None: tags_matching = Q( "function_score", query=Q("term", proj_tags=interest), weight=tags_score / num_interests, boost_mode="replace", ) else: tags_matching = tags_matching | Q( "function_score", query=Q("term", proj_tags=interest), weight=tags_score / num_interests, boost_mode="replace", ) combined_query = ( base_value | tags_matching | company_q | background_rural # | prefer_student_underrep ) # Decay the combined score based on the number of students who already voted for that combined_query = Q("function_score", query=combined_query, functions=SF("gauss", numStudentsSelected={ "origin": 0, "scale": 3, "offset": 3, "decay": 0.50 })) # Timezone - this one's a bit more complex. See comments in script for more details. # Multiplies it's value by the previous scores, allowing it to reduce, set to zero, and increase scores. # See below string for python implementation """ if mentor['okTimezoneDifference']: if 16 < student['timezone'] < 22: return True return false else: if abs(student['timezone'] - mentor['timezone']) < 3: return True return False """ s = s.query(combined_query)[0:num_resp] resp = s.execute() return resp
def get_diseases(self, es): for disease in Search().using(es).index(self.es_index_efo).query( MatchAll()).scan(): yield disease.to_dict()
def search(self, search): sort = None if search.sort is None else search.sort.translate() if not search.words: query = MatchAll() else: query = MultiMatch(query=search.words, fields=['content']) elastic_search = Search(index=self.index_name)\ .doc_type(StarDocument)\ .query(query)\ .highlight('content', type='unified', fragment_size=150) elastic_search = elastic_search[search.start:search.start + search.size] # Filter results for type, ages, and languages if search.types: elastic_search = elastic_search.filter('terms', **{"type": search.types}) if search.ages: elastic_search = elastic_search.filter('terms', **{"ages": search.ages}) if search.languages: elastic_search = elastic_search.filter( 'terms', **{"languages": search.languages}) # Filter results by date if search.date: elastic_search = elastic_search.filter( 'range', **{"date": { "gte": search.date }}) else: elastic_search = elastic_search.filter( 'bool', **{ "should": [ { "range": { "date": { "gte": datetime.datetime.utcnow() } } }, # Future events OR { "bool": { "must_not": { "exists": { "field": "date" } } } } # Date field is empty ] }) if sort is not None: elastic_search = elastic_search.sort(sort) if 'user' in g and g.user: if Permission.edit_resource not in g.user.role.permissions(): elastic_search = elastic_search.filter( Q('bool', must_not=[Q('match', is_draft=True)])) else: elastic_search = elastic_search.filter( Q('bool', must_not=[Q('match', is_draft=True)])) if search.category and search.category.id: elastic_search = elastic_search.filter( 'terms', category=[str(search.category.search_path())]) if search.category.calculate_level() == 0: exclude = ".*\\,.*\\,.*" include = str(search.category.id) + "\\,.*" aggregation = A("terms", field='category', exclude=exclude, include=include, size=25) elif search.category.calculate_level() == 1: include = ".*\\,.*\\,.*" aggregation = A("terms", field='category', include=include, size=25) else: aggregation = A("terms", field='category', size=25) else: aggregation = A("terms", field='category', exclude=".*\\,.*", size=25) elastic_search.aggs.bucket('terms', aggregation) elastic_search.aggs.bucket('type', A("terms", field='type')) elastic_search.aggs.bucket('ages', A("terms", field='ages')) elastic_search.aggs.bucket('languages', A("terms", field='languages')) # KEEPING FOR NOW - THESE WERE THE ORIGINAL FACETS WE HAD SET UP. WILL NEED TO CONVERT TO AGGREGATIONS # IF WE WANT TO KEEP ANY OF THESE. # 'Location': elasticsearch_dsl.TermsFacet(field='location'), # 'Type': elasticsearch_dsl.TermsFacet(field='label'), # 'Age Range': elasticsearch_dsl.TermsFacet(field='age_range'), # 'Category': elasticsearch_dsl.TermsFacet(field='category'), # 'Organization': elasticsearch_dsl.TermsFacet(field='organization'), # 'Status': elasticsearch_dsl.TermsFacet(field='status'), # 'Topic': elasticsearch_dsl.TermsFacet(field='topic'), return elastic_search.execute()
from elasticsearch import Elasticsearch from elasticsearch_dsl import Search from elasticsearch_dsl.query import Range, MatchAll, Term, Terms, Match, MatchPhrase, MultiMatch, QueryString, Bool, \ FunctionScore from elasticsearch_dsl.function import FieldValueFactor, Exp, RandomScore import json from pprint import pprint client = Elasticsearch(['127.0.0.1:9300']) search = (Search(using=client, index='resource').query(MatchAll()).source( ['title', 'available_from_date']).sort('-available_from_date')) (search.aggs.bucket('organizations', 'terms', field='resource_organization').metric( 'views_per_month_stats', 'stats', field='views_per_month')) print('Search Body:') pprint(search.to_dict()) print(json.dumps(search.to_dict())) response = search.execute() print(response) for hit in response: print(hit.meta.score, hit.title) for org in response.aggregations.organizations.buckets: print(org.key, org.doc_count, org.views_per_month_stats.to_dict())
def test_basic_search(self): """Basic search and filtering""" if not self.manager.connected_to_es: return # 1. Index and test search: text = 'test' topic_1 = TopicFactory(forum=self.forum, author=self.user, title=text) post_1 = PostFactory(topic=topic_1, author=self.user, position=1) post_1.text = post_1.text_html = text post_1.save() # create a middle-size content and publish it tuto = PublishableContentFactory(type='TUTORIAL') tuto_draft = tuto.load_version() tuto.title = text tuto.authors.add(self.user) tuto.save() tuto_draft.repo_update_top_container( text, tuto.slug, text, text) # change title to be sure it will match chapter1 = ContainerFactory(parent=tuto_draft, db_object=tuto) extract = ExtractFactory(container=chapter1, db_object=tuto) extract.repo_update(text, text) published = publish_content(tuto, tuto_draft, is_major_update=True) tuto.sha_public = tuto_draft.current_version tuto.sha_draft = tuto_draft.current_version tuto.public_version = published tuto.save() # nothing has been indexed yet: self.assertEqual( len( self.manager.setup_search(Search().query( MatchAll())).execute()), 0) # index for model in self.indexable: if model is FakeChapter: continue self.manager.es_bulk_indexing_of_model(model) self.manager.refresh_index() result = self.client.get(reverse('search:query') + '?q=' + text, follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 4) # get 4 results # 2. Test filtering: topic_1 = Topic.objects.get(pk=topic_1.pk) post_1 = Post.objects.get(pk=post_1.pk) published = PublishedContent.objects.get(pk=published.pk) ids = { 'topic': [topic_1.es_id], 'post': [post_1.es_id], 'content': [ published.es_id, published.content_public_slug + '__' + chapter1.slug ], } search_groups = [ k for k, v in settings.ZDS_APP['search']['search_groups'].items() ] group_to_model = { k: v[1] for k, v in settings.ZDS_APP['search']['search_groups'].items() } for doc_type in search_groups: result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + doc_type, follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, len(ids[doc_type])) # get 1 result of each … for i, r in enumerate(response): self.assertIn( r.meta.doc_type, group_to_model[doc_type]) # … and only of the right type … self.assertEqual(r.meta.id, ids[doc_type][i]) # … with the right id !
def get_diseases(self, es): for disease in Search().using(es).index(self.es_index_efo).extra( track_total_hits=True).query(MatchAll()).scan(): yield disease.to_dict()
def test_change_topic_impacts_posts(self): if not self.manager.connected_to_es: return # 1. Create a hidden forum belonging to a hidden group and add staff in it. text = 'test' group = Group.objects.create(name='Les illuminatis anonymes de ZdS') _, hidden_forum = create_category(group) self.staff.groups.add(group) self.staff.save() # 2. Create a normal topic and index it topic_1 = TopicFactory(forum=self.forum, author=self.user, title=text) post_1 = PostFactory(topic=topic_1, author=self.user, position=1) post_1.text = post_1.text_html = text post_1.save() self.manager.es_bulk_indexing_of_model(Topic) self.manager.es_bulk_indexing_of_model(Post) self.manager.refresh_index() self.assertEqual( len( self.manager.setup_search(Search().query( MatchAll())).execute()), 2) # indexing ok result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Post.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 1) # ok self.assertEqual(response[0].meta.doc_type, Post.get_es_document_type()) self.assertEqual(response[0].forum_pk, self.forum.pk) self.assertEqual(response[0].topic_pk, topic_1.pk) self.assertEqual(response[0].topic_title, topic_1.title) # 3. Change topic title and reindex topic_1.title = 'new title' topic_1.save() self.manager.es_bulk_indexing_of_model(Topic) self.manager.es_bulk_indexing_of_model(Post) self.manager.refresh_index() result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Post.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 1) # ok self.assertEqual(response[0].topic_title, topic_1.title) # title was changed # 4. connect with staff and move topic self.assertTrue( self.client.login(username=self.staff.username, password='******')) data = {'move': '', 'forum': hidden_forum.pk, 'topic': topic_1.pk} response = self.client.post(reverse('topic-edit'), data, follow=False) self.assertEqual(302, response.status_code) self.manager.es_bulk_indexing_of_model(Topic) self.manager.es_bulk_indexing_of_model(Post) self.manager.refresh_index() result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Post.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual( response.hits.total, 1) # Note: without staff, would not get any results (see below) self.assertEqual(response[0].forum_pk, hidden_forum.pk) # post was updated with new forum # 5. Topic is now hidden self.client.logout() result = self.client.get(reverse('search:query') + '?q=' + text + '&models=' + Post.get_es_document_type(), follow=False) self.assertEqual(result.status_code, 200) response = result.context['object_list'].execute() self.assertEqual(response.hits.total, 0) # ok
def all(cls): p = cls.search() p.query = MatchAll() return p.execute()
def test_upercase_and_lowercase_search_give_same_results(self): """Pretty self-explanatory function name, isn't it ?""" if not self.manager.connected_to_es: return # 1. Index lowercase stuffs text_lc = 'test' topic_1_lc = TopicFactory(forum=self.forum, author=self.user, title=text_lc) tag_lc = TagFactory(title=text_lc) topic_1_lc.tags.add(tag_lc) topic_1_lc.subtitle = text_lc topic_1_lc.save() post_1_lc = PostFactory(topic=topic_1_lc, author=self.user, position=1) post_1_lc.text = post_1_lc.text_html = text_lc post_1_lc.save() tuto_lc = PublishableContentFactory(type='TUTORIAL') tuto_draft_lc = tuto_lc.load_version() tuto_lc.title = text_lc tuto_lc.authors.add(self.user) subcategory_lc = SubCategoryFactory(title=text_lc) tuto_lc.subcategory.add(subcategory_lc) tuto_lc.tags.add(tag_lc) tuto_lc.save() tuto_draft_lc.description = text_lc tuto_draft_lc.repo_update_top_container(text_lc, tuto_lc.slug, text_lc, text_lc) chapter1_lc = ContainerFactory(parent=tuto_draft_lc, db_object=tuto_lc) extract_lc = ExtractFactory(container=chapter1_lc, db_object=tuto_lc) extract_lc.repo_update(text_lc, text_lc) published_lc = publish_content(tuto_lc, tuto_draft_lc, is_major_update=True) tuto_lc.sha_public = tuto_draft_lc.current_version tuto_lc.sha_draft = tuto_draft_lc.current_version tuto_lc.public_version = published_lc tuto_lc.save() # 2. Index uppercase stuffs text_uc = 'TEST' topic_1_uc = TopicFactory(forum=self.forum, author=self.user, title=text_uc) topic_1_uc.tags.add( tag_lc) # Note: a constraint forces tags title to be unique topic_1_uc.subtitle = text_uc topic_1_uc.save() post_1_uc = PostFactory(topic=topic_1_uc, author=self.user, position=1) post_1_uc.text = post_1_uc.text_html = text_uc post_1_uc.save() tuto_uc = PublishableContentFactory(type='TUTORIAL') tuto_draft_uc = tuto_uc.load_version() tuto_uc.title = text_uc tuto_uc.authors.add(self.user) tuto_uc.subcategory.add(subcategory_lc) tuto_uc.tags.add(tag_lc) tuto_uc.save() tuto_draft_uc.description = text_uc tuto_draft_uc.repo_update_top_container(text_uc, tuto_uc.slug, text_uc, text_uc) chapter1_uc = ContainerFactory(parent=tuto_draft_uc, db_object=tuto_uc) extract_uc = ExtractFactory(container=chapter1_uc, db_object=tuto_uc) extract_uc.repo_update(text_uc, text_uc) published_uc = publish_content(tuto_uc, tuto_draft_uc, is_major_update=True) tuto_uc.sha_public = tuto_draft_uc.current_version tuto_uc.sha_draft = tuto_draft_uc.current_version tuto_uc.public_version = published_uc tuto_uc.save() # 3. Index and search: self.assertEqual( len( self.manager.setup_search(Search().query( MatchAll())).execute()), 0) # index for model in self.indexable: if model is FakeChapter: continue self.manager.es_bulk_indexing_of_model(model) self.manager.refresh_index() result = self.client.get(reverse('search:query') + '?q=' + text_lc, follow=False) self.assertEqual(result.status_code, 200) response_lc = result.context['object_list'].execute() self.assertEqual(response_lc.hits.total, 8) result = self.client.get(reverse('search:query') + '?q=' + text_uc, follow=False) self.assertEqual(result.status_code, 200) response_uc = result.context['object_list'].execute() self.assertEqual(response_uc.hits.total, 8) for responses in zip( response_lc, response_uc): # we should get results in the same order! self.assertEqual(responses[0].meta.id, responses[1].meta.id)
from elasticsearch import Elasticsearch from elasticsearch_dsl import Search, Q from elasticsearch_dsl.query import Type, HasChild, MatchAll from config import * es = Elasticsearch(host=host, timeout=120) q_parent = Type(value='auditd-parent') q_has_child = HasChild(type='auditd-child', query=MatchAll(), inner_hits={"sort": {"@timestamp": "asc"}, 'size': 100}) search = Search(using=es, index=index).sort({"@timestamp": {"order": "asc"}}) search.query = Q((q_has_child | (q_parent & ~q_has_child))) def scan(n=10, verbose=False): i = 0 aggs = dict() for hit in search.scan(): aggs.setdefault(hit.ses, []) aggs[hit.ses].append((hit, hit.meta.inner_hits['auditd-child'].hits.hits)) if i % 1000 == 0 and verbose: print str(round(100*i/float(n), 2)) + "%" if i > n: break i += 1 return aggs