[FEAT] consolidate Spark session fixture into conftest.py

daft/daft/__init__.pyi

@@ -1208,7 +1208,7 @@ def sql_expr(sql: str) -> PyExpr: ...
 def list_sql_functions() -> list[SQLFunctionStub]: ...
 def utf8_count_matches(expr: PyExpr, patterns: PyExpr, whole_words: bool, case_sensitive: bool) -> PyExpr: ...
 def to_struct(inputs: list[PyExpr]) -> PyExpr: ...
-def connect_start(addr: str) -> ConnectionHandle: ...
+def connect_start(addr: str) -> tuple[ConnectionHandle, int]: ...
 
 class ConnectionHandle:
     def shutdown(self) -> None: ...

src/daft-connect/Cargo.toml

@@ -1,5 +1,6 @@
 [dependencies]
 arrow2 = {workspace = true}
+async-stream = "0.3.6"
 common-daft-config = {workspace = true}
 daft-local-execution = {workspace = true}
 daft-local-plan = {workspace = true}

src/daft-connect/src/lib.rs

@@ -49,10 +49,13 @@ impl ConnectionHandle {
     }
 }
 
-pub fn start(addr: &str) -> eyre::Result<ConnectionHandle> {
+pub fn start(addr: &str) -> eyre::Result<(ConnectionHandle, u16)> {
     info!("Daft-Connect server listening on {addr}");
     let addr = util::parse_spark_connect_address(addr)?;
 
+    let listener = std::net::TcpListener::bind(addr)?;
+    let port = listener.local_addr()?.port();
+
     let service = DaftSparkConnectService::default();
 
     info!("Daft-Connect server listening on {addr}");
@@ -65,21 +68,35 @@ pub fn start(addr: &str) -> eyre::Result<ConnectionHandle> {
 
     std::thread::spawn(move || {
         let runtime = tokio::runtime::Runtime::new().unwrap();
-        let result = runtime
-            .block_on(async {
-                tokio::select! {
-                    result = Server::builder()
-                        .add_service(SparkConnectServiceServer::new(service))
-                        .serve(addr) => {
-                        result
-                    }
-                    _ = shutdown_receiver => {
-                        info!("Received shutdown signal");
-                        Ok(())
+        let result = runtime.block_on(async {
+            let incoming = {
+                let listener = tokio::net::TcpListener::from_std(listener)
+                    .wrap_err("Failed to create TcpListener from std::net::TcpListener")?;
+
+                async_stream::stream! {
+                    loop {
+                        match listener.accept().await {
+                            Ok((stream, _)) => yield Ok(stream),
+                            Err(e) => yield Err(e),
+                        }
                     }
                 }
-            })
-            .wrap_err_with(|| format!("Failed to start server on {addr}"));
+            };
+
+            let result = tokio::select! {
+                result = Server::builder()
+                    .add_service(SparkConnectServiceServer::new(service))
+                    .serve_with_incoming(incoming)=> {
+                    result
+                }
+                _ = shutdown_receiver => {
+                    info!("Received shutdown signal");
+                    Ok(())
+                }
+            };
+
+            result.wrap_err_with(|| format!("Failed to start server on {addr}"))
+        });
 
         if let Err(e) = result {
             eprintln!("Daft-Connect server error: {e:?}");
@@ -88,7 +105,7 @@ pub fn start(addr: &str) -> eyre::Result<ConnectionHandle> {
         eyre::Result::<_>::Ok(())
     });
 
-    Ok(handle)
+    Ok((handle, port))
 }
 
 #[derive(Default)]
@@ -364,7 +381,7 @@ impl SparkConnectService for DaftSparkConnectService {
 #[cfg(feature = "python")]
 #[pyo3::pyfunction]
 #[pyo3(name = "connect_start")]
-pub fn py_connect_start(addr: &str) -> pyo3::PyResult<ConnectionHandle> {
+pub fn py_connect_start(addr: &str) -> pyo3::PyResult<(ConnectionHandle, u16)> {
     start(addr).map_err(|e| pyo3::exceptions::PyRuntimeError::new_err(format!("{e:?}")))
 }
 

tests/connect/conftest.py

@@ -0,0 +1,29 @@
+from __future__ import annotations
+
+import pytest
+from pyspark.sql import SparkSession
+
+
+@pytest.fixture(scope="session")
+def spark_session():
+    """
+    Fixture to create and clean up a Spark session.
+
+    This fixture is available to all test files and creates a single
+    Spark session for the entire test suite run.
+    """
+    from daft.daft import connect_start
+
+    # Start Daft Connect server
+    (server, port) = connect_start("sc://localhost:0")
+
+    url = f"sc://localhost:{port}"
+
+    # Initialize Spark Connect session
+    session = SparkSession.builder.appName("DaftConfigTest").remote(url).getOrCreate()
+
+    yield session
+
+    # Cleanup
+    server.shutdown()
+    session.stop()

tests/connect/test_config_simple.py

@@ -1,29 +1,5 @@
 from __future__ import annotations
 
-import time
-
-import pytest
-from pyspark.sql import SparkSession
-
-
-@pytest.fixture
-def spark_session():
-    """Fixture to create and clean up a Spark session."""
-    from daft.daft import connect_start
-
-    # Start Daft Connect server
-    server = connect_start("sc://localhost:50051")
-
-    # Initialize Spark Connect session
-    session = SparkSession.builder.appName("DaftConfigTest").remote("sc://localhost:50051").getOrCreate()
-
-    yield session
-
-    # Cleanup
-    server.shutdown()
-    session.stop()
-    time.sleep(2)  # Allow time for session cleanup
-
 
 def test_set_operation(spark_session):
     """Test the Set operation with various data types and edge cases."""

tests/connect/test_range_simple.py

@@ -1,26 +1,5 @@
 from __future__ import annotations
 
-import pytest
-from pyspark.sql import SparkSession
-
-
-@pytest.fixture
-def spark_session():
-    """Fixture to create and clean up a Spark session."""
-    from daft.daft import connect_start
-
-    # Start Daft Connect server
-    server = connect_start("sc://localhost:50051")
-
-    # Initialize Spark Connect session
-    session = SparkSession.builder.appName("DaftConfigTest").remote("sc://localhost:50051").getOrCreate()
-
-    yield session
-
-    # Cleanup
-    server.shutdown()
-    session.stop()
-
 
 def test_range_operation(spark_session):
     # Create a range using Spark