Support setting the app name via SparkConnection

hlink/spark/session.py

@@ -21,12 +21,15 @@
 class SparkConnection:
     """Handles initialization of spark session and connection to local cluster."""
 
-    def __init__(self, derby_dir, warehouse_dir, tmp_dir, python, db_name):
+    def __init__(
+        self, derby_dir, warehouse_dir, tmp_dir, python, db_name, app_name="linking"
+    ):
         self.derby_dir = derby_dir
         self.warehouse_dir = warehouse_dir
         self.db_name = db_name
         self.tmp_dir = tmp_dir
         self.python = python
+        self.app_name = app_name
 
     def spark_conf(self, executor_cores, executor_memory, driver_memory, cores):
         spark_package_path = os.path.dirname(hlink.spark.__file__)
@@ -44,7 +47,7 @@ def spark_conf(self, executor_cores, executor_memory, driver_memory, cores):
             )
             .set("spark.executorEnv.SPARK_LOCAL_DIRS", self.tmp_dir)
             .set("spark.sql.legacy.allowUntypedScalaUDF", True)
-            .setAppName("linking")
+            .setAppName(self.app_name)
             # .set("spark.executor.cores", executor_cores) \
         )
         if executor_memory:

hlink/tests/spark_connection_test.py

@@ -0,0 +1,33 @@
+from pathlib import Path
+import sys
+
+from hlink.spark.session import SparkConnection
+
+
+def test_app_name_defaults_to_linking(tmp_path: Path) -> None:
+    derby_dir = tmp_path / "derby"
+    warehouse_dir = tmp_path / "warehouse"
+    tmp_dir = tmp_path / "tmp"
+    connection = SparkConnection(
+        derby_dir, warehouse_dir, tmp_dir, sys.executable, "test"
+    )
+    spark = connection.local(cores=1, executor_memory="1G")
+    app_name = spark.conf.get("spark.app.name")
+    assert app_name == "linking"
+
+
+def test_app_name_argument(tmp_path: Path) -> None:
+    derby_dir = tmp_path / "derby"
+    warehouse_dir = tmp_path / "warehouse"
+    tmp_dir = tmp_path / "tmp"
+    connection = SparkConnection(
+        derby_dir,
+        warehouse_dir,
+        tmp_dir,
+        sys.executable,
+        "test",
+        app_name="test_app_name",
+    )
+    spark = connection.local(cores=1, executor_memory="1G")
+    app_name = spark.conf.get("spark.app.name")
+    assert app_name == "test_app_name"