Merge pull request #82 from OpenSemanticLab/feat-add-schema-install-policy-param

feat: add schema install policy param

Author: simontaurus

src/osw/express.py

@@ -193,15 +193,30 @@ def shut_down(self):
         # connection can't be reopened except when initializing a new instance)
 
     def install_dependencies(
-        self, dependencies: Dict[str, str] = None, mode: str = "append"
+        self,
+        dependencies: Dict[str, str] = None,
+        mode: str = "append",
+        policy: str = "force",
     ):
         """Expects a dictionary with the keys being the names of the dependencies and
         the values being the full page name of the dependencies.
         To keep existing dependencies, use mode='append'.
+        Default policy is 'force', which will always load dependencies.
+        If policy is 'if-missing', dependencies will only be loaded if they are not already installed.
+        This may lead to outdated dependencies, if the dependencies have been updated on the server.
+        If policy is 'if-outdated', dependencies will only be loaded if they were updated on the server.
+        (not implemented yet)
         """
         if dependencies is None:
             dependencies = DEPENDENCIES
-        schema_fpts = list(dependencies.values())
+        schema_fpts = []
+        for k, v in dependencies.items():
+            if policy != "if-missing" or not hasattr(model, k):
+                schema_fpts.append(v)
+            if policy == "if-outdated":
+                raise NotImplementedError(
+                    "The policy 'if-outdated' is not implemented yet."
+                )
         schema_fpts = list(set(schema_fpts))
         for schema_fpt in schema_fpts:
             if not schema_fpt.count(":") == 1:

tests/integration/performance_test.py

@@ -0,0 +1,58 @@
+import time
+
+import osw.model.entity as model
+from osw.auth import CredentialManager
+
+
+def test_fetch_and_load(wiki_domain, wiki_username, wiki_password, mocker):
+    # create a credential file on the default path for osw express
+    cm = CredentialManager()  # cred_filepath = Path.cwd() / "accounts.pwd.yaml")
+    cm.add_credential(
+        CredentialManager.UserPwdCredential(
+            iri=wiki_domain, username=wiki_username, password=wiki_password
+        )
+    )
+    # cm.save_credentials_to_file()
+    # wtsite = WtSite(WtSite.WtSiteConfig(iri=wiki_domain, cred_mngr=cm))
+    # osw_obj = OSW(site=wtsite)
+
+    # Here the initial connection to the wiki is mocked (passing domain, username and
+    # password)
+    mocked_input = mocker.patch("builtins.input")
+    mocked_getpass = mocker.patch("getpass.getpass")
+    mocked_input.side_effect = [wiki_domain, wiki_username]
+    mocked_getpass.return_value = wiki_password
+    # This import will trigger the install_dependencies method call on the first run
+    import osw.express
+
+    osw_express = osw.express.OswExpress(domain=wiki_domain, cred_mngr=cm)
+
+    # Measure the time taken to store and load an entity
+    start_time = time.time()
+
+    # Load Tutorial Schema on demand
+    # if not hasattr(model, "Tutorial"):
+    DEPENDENCIES = {"Tutorial": "Category:OSW494f660e6a714a1a9681c517bbb975da"}
+    osw_express.install_dependencies(DEPENDENCIES, policy="force")
+
+    end_time = time.time()
+    print(f"Time taken to load Tutorial Schema: {end_time - start_time}")
+    assert hasattr(model, "Tutorial")
+    assert end_time - start_time < 15  # typically takes 8 seconds
+
+    start_time = time.time()
+    osw_express.install_dependencies(DEPENDENCIES, policy="force")
+    end_time = time.time()
+    print(
+        f"Time taken to reload Tutorial Schema with policy 'force': {end_time - start_time}"
+    )
+    assert end_time - start_time < 2  # typically takes 1 seconds using memory cache
+
+    start_time = time.time()
+    osw_express.install_dependencies(DEPENDENCIES, policy="if-missing")
+    end_time = time.time()
+    print(
+        "Time taken to reload Tutorial Schema with policy ",
+        "'if-missing': {end_time - start_time}",
+    )
+    assert end_time - start_time < 0.1  # typically takes 0 seconds