[DUX-1897]: more clear output for interpreting modules for eval

src/ghci/mod.rs

@@ -564,7 +564,6 @@ impl Ghci {
         // restore the old value when the function returns.
         for (path, commands) in self.eval_commands.clone() {
             for command in commands {
-                tracing::info!("{path}:{command}");
                 // If the `module` was already compiled, `ghci` may have loaded the interface file instead
                 // of the interpreted bytecode, giving us this error message when we attempt to
                 // load the top-level scope with `:module + *{module}`:
@@ -574,8 +573,10 @@ impl Ghci {
                 // We use `:add *{module}` to force interpreting the module. We do this here instead of in
                 // `add_module` to save time if eval commands aren't used (or aren't needed for a
                 // particular module).
+                tracing::info!("Loading {path} in interpreted mode for eval commands");
                 self.interpret_module(&path, log).await?;
                 let module = self.search_paths.path_to_module(&path)?;
+                tracing::info!("Eval {path}:{command}");
                 self.stdin
                     .eval(&mut self.stdout, &module, &command.command, log)
                     .await?;

src/ghci/parse/eval.rs

@@ -39,7 +39,11 @@ pub struct EvalCommand {
 
 impl Display for EvalCommand {
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        write!(f, "{}:{}: {}", self.line, self.column, self.display_command)
+        write!(
+            f,
+            "{}:{}: -- $> {}",
+            self.line, self.column, self.display_command
+        )
     }
 }
 

tests/eval.rs

@@ -33,7 +33,7 @@ async fn can_eval_commands() {
     let defined_in_multiple_files =
         BaseMatcher::message("Read stderr line").with_field("line", "defined in multiple files");
 
-    let eval_message = BaseMatcher::message(r"MyModule.hs:\d+:\d+: example \+\+ example")
+    let eval_message = BaseMatcher::message(r"MyModule.hs:\d+:\d+: -- \$> example \+\+ example")
         .but_not(defined_in_multiple_files.clone());
     session
         .assert_logged_or_wait(eval_message.clone())
@@ -95,8 +95,10 @@ async fn can_load_new_eval_commands_multiline() {
         .await
         .unwrap();
 
-    let eval_message =
-        BaseMatcher::message(&format!(r"MyModule.hs:\d+:\d+: {}", regex::escape(cmd)));
+    let eval_message = BaseMatcher::message(&format!(
+        r"MyModule.hs:\d+:\d+: -- \$> {}",
+        regex::escape(cmd)
+    ));
     session
         .wait_for_log(&eval_message)
         .await
@@ -177,7 +179,7 @@ async fn can_eval_commands_in_non_interpreted_modules() {
             BaseMatcher::message("All good!")
                 // Evals the command.
                 .and(BaseMatcher::message(
-                    r"MyModule.hs:\d+:\d+: example \+\+ example",
+                    r"MyModule.hs:\d+:\d+: -- \$> example \+\+ example",
                 ))
                 // Reads eval output.
                 .and(BaseMatcher::message("Read line").with_field("line", "exampleexample"))
@@ -220,7 +222,7 @@ async fn can_eval_commands_twice() {
     let ok_reload = BaseMatcher::message("All good!")
         // Evals the command.
         .and(BaseMatcher::message(
-            r"MyModule.hs:\d+:\d+: example \+\+ example",
+            r"MyModule.hs:\d+:\d+: -- \$> example \+\+ example",
         ))
         // Reads eval output.
         .and(BaseMatcher::message("Read line").with_field("line", "exampleexample"))