[Factors] in-process runtime and conformance tests #2732

rylev · 2024-08-20T18:30:59Z

This allows us to run runtime and conformance tests using the "in-process runtime" (as opposed to the Spin CLI spawned process runtime).

This requires a few refactorings that I think push us more towards other goals as well (namely spin-trigger being generic over the RuntimeFactors type).

The actual code for instantiating and triggering an HTTP trigger without running an actual HTTP server looks like this:

    let app = spin_app::App::new("my-app", locked_app);
    let trigger = HttpTrigger::new(&app, "127.0.0.1:80".parse().unwrap(), None)?;
    let mut builder = TriggerAppBuilder::new(trigger, PathBuf::from("."));
    let trigger_app = builder.build(app, TriggerAppOptions::default()).await?;
    let server = builder.trigger.into_server(trigger_app)?;
    server
        .handle(
              req,
              http::uri::Scheme::HTTP,
              (std::net::Ipv4Addr::LOCALHOST, 7000).into(),
         )
        .await?;

Signed-off-by: Ryan Levick <[email protected]>

lann · 2024-08-20T18:47:51Z

crates/trigger-http2/src/server.rs

@@ -333,6 +288,111 @@ impl HttpServer {
    }
 }

+/// Handles a routed HTTP trigger request.
+pub struct RequestHandler {


Could you say more about this change? It isn't clear how it relates to the conformance tests.

To be fair this was from an evolutionary dead end. I do think it helps make things a bit easier to follow (at the expense of being slightly more verbose) by only giving the OutboundHttpInterceptor what it needs and nothing more. It encapsulates just the component handling and leaves out all the other HTTP sever stuff to the encapsulating HttpServer types.

I can revert it if you want me to though.

Its mostly just really hard to review the diff 😅

Sorry! If you aren't able to review, I'll revert that change tomorrow.

I've reverted that change so hopefully the diff is a bit easier to read. Unfortunately, there's one spot where git's diffing still gets a bit confused.

Signed-off-by: Ryan Levick <[email protected]>

rylev · 2024-08-21T10:58:47Z

crates/trigger2/src/cli.rs

+        let (abortable, abort_handle) = futures::future::abortable(run_fut);
+        ctrlc::set_handler(move || abort_handle.abort())?;
+        match abortable.await {
+            Ok(Ok(())) => {
+                tracing::info!("Trigger executor shut down: exiting");
+                Ok(())
+            }
+            Ok(Err(err)) => {
+                tracing::error!("Trigger executor failed");
+                Err(err)
+            }
+            Err(_aborted) => {
+                tracing::info!("User requested shutdown: exiting");
+                Ok(())
            }
+        }
+    }
+
+    fn follow_components(&self) -> FollowComponents {
+        if self.silence_component_logs {
+            FollowComponents::None
+        } else if self.follow_components.is_empty() {
+            FollowComponents::All
+        } else {
+            let followed = self.follow_components.clone().into_iter().collect();
+            FollowComponents::Named(followed)
+        }
+    }
+}
+
+const SLOTH_WARNING_DELAY_MILLIS: u64 = 1250;
+
+fn warn_if_wasm_build_slothful() -> sloth::SlothGuard {
+    #[cfg(debug_assertions)]
+    let message = "\
+        This is a debug build - preparing Wasm modules might take a few seconds\n\
+        If you're experiencing long startup times please switch to the release build";
+
+    #[cfg(not(debug_assertions))]
+    let message = "Preparing Wasm modules is taking a few seconds...";
+
+    sloth::warn_if_slothful(SLOTH_WARNING_DELAY_MILLIS, format!("{message}\n"))
+}
+
+fn help_heading<T: Trigger>() -> Option<&'static str> {
+    if T::TYPE == help::HelpArgsOnlyTrigger::TYPE {
+        Some("TRIGGER OPTIONS")
+    } else {
+        let heading = format!("{} TRIGGER OPTIONS", T::TYPE.to_uppercase());
+        let as_str = Box::new(heading).leak();
+        Some(as_str)
+    }
+}


This is all just moved from down below.

rylev · 2024-08-21T11:05:12Z

crates/trigger-http2/src/server.rs

@@ -333,6 +288,111 @@ impl HttpServer {
    }
 }

+/// Handles a routed HTTP trigger request.
+pub struct RequestHandler {


I've reverted that change so hopefully the diff is a bit easier to read. Unfortunately, there's one spot where git's diffing still gets a bit confused.

rylev added 2 commits August 20, 2024 17:58

Refactor http-trigger to be more reusable outside of main flow

c578fdd

Signed-off-by: Ryan Levick <[email protected]>

Get things compiling

6a67cef

Signed-off-by: Ryan Levick <[email protected]>

rylev requested a review from lann August 20, 2024 18:30

rylev changed the title ~~[Factors] in-process conformance tests~~ [Factors] in-process runtime and conformance tests Aug 20, 2024

lann reviewed Aug 20, 2024

View reviewed changes

Revert RequestHandler type refactoring

32f6da4

Signed-off-by: Ryan Levick <[email protected]>

rylev force-pushed the factors-conformance-tests branch from 363e400 to 32f6da4 Compare August 21, 2024 11:03

rylev requested a review from lann August 21, 2024 11:04

rylev commented Aug 21, 2024

View reviewed changes

lann approved these changes Aug 21, 2024

View reviewed changes

rylev merged commit 66f357a into factors Aug 21, 2024
2 checks passed

rylev deleted the factors-conformance-tests branch August 21, 2024 14:17

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Factors] in-process runtime and conformance tests #2732

[Factors] in-process runtime and conformance tests #2732

rylev commented Aug 20, 2024 •

edited

Loading

lann Aug 20, 2024

rylev Aug 20, 2024

lann Aug 20, 2024

rylev Aug 20, 2024

rylev Aug 21, 2024

lann Aug 21, 2024

rylev Aug 21, 2024

rylev Aug 21, 2024

[Factors] in-process runtime and conformance tests #2732

[Factors] in-process runtime and conformance tests #2732

Conversation

rylev commented Aug 20, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rylev commented Aug 20, 2024 •

edited

Loading