Add GraphDB fixture for tests

haystack/knowledge_graph/graphdb.py

@@ -44,7 +44,6 @@ def delete_index(self):
         if response.status_code > 299:
             raise Exception(response.text)
 
-
     def import_from_ttl_file(self, index: str, path: Path):
         url = f"{self.url}/repositories/{index}/statements"
         headers = {"Content-type": "application/x-turtle"}

test/conftest.py

@@ -6,6 +6,7 @@
 import pytest
 import requests
 from elasticsearch import Elasticsearch
+from haystack.knowledge_graph.graphdb import GraphDBKnowledgeGraph
 from milvus import Milvus
 
 from haystack.document_store.milvus import MilvusDocumentStore
@@ -54,6 +55,8 @@ def pytest_collection_modifyitems(items):
             item.add_marker(pytest.mark.tika)
         elif "elasticsearch" in item.nodeid:
             item.add_marker(pytest.mark.elasticsearch)
+        elif "graphdb" in item.nodeid:
+            item.add_marker(pytest.mark.graphdb)
         elif "pipeline" in item.nodeid:
             item.add_marker(pytest.mark.pipeline)
         elif "slow" in item.nodeid:
@@ -96,6 +99,29 @@ def milvus_fixture():
         time.sleep(40)
 
 
+@pytest.fixture(scope="session")
+def graphdb_fixture():
+    # test if a GraphDB instance is already running. If not, download and start a GraphDB instance locally.
+    try:
+        kg = GraphDBKnowledgeGraph()
+        # TODO fail if not running GraphDB
+        raise RuntimeError
+    except:
+        print("Starting GraphDB ...")
+        status = subprocess.run(
+            ['docker rm haystack_test_graphdb'],
+            shell=True
+        )
+        status = subprocess.run(
+            ['docker run -d -p 7200:7200 --name haystack_test_graphdb docker-registry.ontotext.com/graphdb-free:9.4.1-adoptopenjdk11'],
+            shell=True
+        )
+        if status.returncode:
+            raise Exception(
+                "Failed to launch GraphDB. Please check docker container logs.")
+        time.sleep(30)
+
+
 @pytest.fixture(scope="session")
 def tika_fixture():
     try:

test/pytest.ini

@@ -4,6 +4,7 @@ markers =
     slow: marks tests as slow (deselect with '-m "not slow"')
     tika: marks tests which require tika container (deselect with '-m "not tika"')
     elasticsearch: marks tests which require elasticsearch container (deselect with '-m "not elasticsearch"')
+    graphdb: marks tests which require graphdb container (deselect with '-m "not graphdb"')
     generator: marks generator tests (deselect with '-m "not generator"')
     pipeline: marks tests with pipeline
     summarizer: marks summarizer tests

test/test_knowledge_graph.py

@@ -2,24 +2,21 @@
 
 import pytest
 
-from haystack.graph_retriever.text_to_sparql import Text2SparqlRetriever
+from haystack.graph_retriever import Text2SparqlRetriever
 from haystack.knowledge_graph.graphdb import GraphDBKnowledgeGraph
+from haystack.preprocessor.utils import fetch_archive_from_http
 
-
-@pytest.mark.graph
+@pytest.mark.graphdb
 def test_graph_retrieval(retriever_with_docs, document_store_with_docs):
-    # doc_dir = "../data/tutorial10_knowledge_graph/"
-    # s3_url = "https://fandom-qa.s3-eu-west-1.amazonaws.com/triples_and_config.zip"
-    # fetch_archive_from_http(url=s3_url, output_dir=doc_dir)
-    #
-    # # Fetch a pre-trained BART model that translates natural language questions to SPARQL queries
-    # doc_dir = "../saved_models/tutorial10_knowledge_graph/"
-    # s3_url = "https://fandom-qa.s3-eu-west-1.amazonaws.com/saved_models/hp_v3.4.zip"
-    # fetch_archive_from_http(url=s3_url, output_dir=doc_dir)
-
-    # status = subprocess.run(
-    #             ['docker run -d -p 7200:7200 --name graphdb-instance-tutorial docker-registry.ontotext.com/graphdb-free:9.4.1-adoptopenjdk11'], shell=True
-    #         )
+    # TODO rename doc_dir
+    doc_dir = "../data/tutorial10_knowledge_graph/"
+    s3_url = "https://fandom-qa.s3-eu-west-1.amazonaws.com/triples_and_config.zip"
+    fetch_archive_from_http(url=s3_url, output_dir=doc_dir)
+
+    # Fetch a pre-trained BART model that translates natural language questions to SPARQL queries
+    doc_dir = "../saved_models/tutorial10_knowledge_graph/"
+    s3_url = "https://fandom-qa.s3-eu-west-1.amazonaws.com/saved_models/hp_v3.4.zip"
+    fetch_archive_from_http(url=s3_url, output_dir=doc_dir)
 
     kg = GraphDBKnowledgeGraph(index="tutorial_10_index")
     kg.create_index(config_path=Path("../data/tutorial10_knowledge_graph/repo-config.ttl"))